Developing Inclusive and Sustainable Economic and Financial Systems

Volume 3
Islamic banking and finance – Essays on corporate
finance, efficiency and product development
Editorial Board
Dr. Hatem A. El-Karanshawy
Dr. Azmi Omar

Dr. Tariqullah Khan

Dr. Salman Syed Ali
Dr. Hylmun Izhar
Wijdan Tariq

Karim Ginena
Bahnaz Al Quradaghi

ISBN: 978-9927-118-23-4
Cover design: Natacha Fares

Copyright © 2015 The Authors

 CONTENTS
Foreword v
Acknowledgments vii
Preface ix
Introduction xi

PART 1: ISLAMIC CAPITAL MARKETS AND CORPORATE FINANCE

Chapter 1 Valuation of Islamic debt instruments, the Sukuk: Lessons

for market development
Mohamed Ariff and Meysam Safari 1
Chapter 2 The impact of Islamic debt on company value
Fitriya Fauzi, Stuart Locke, Abdul Basyith and
Muhammad Idris 19
Chapter 3 Islamic financing and bank characteristics in a dual banking
system: Evidence from Malaysia
Muhamed Zulkhibri 37
Chapter 4 Leverage risk, financial crisis, and stock returns:
A comparison among Islamic, conventional, and
socially responsible stocks
Vaishnavi Bhatt and Jahangir Sultan 47
Chapter 5 Is Shariah-compliant investment universally sustainable?
A comparative study
Mehdi Sadeghi 81

PART 2: I SLAMIC BANKING – EFFICIENCY, PROFITABILITY &

INTERNATIONAL PERSPECTIVES

Chapter 6 The nexus between economic freedom and Islamic bank

performance: Empirical evidence from the MENA banking
sectors
Fadzlan Sufian, Muhamed Zulkhibri and Abdul Majid 93
Chapter 7 Efficiency of performance of banks in the Gulf region before,
during and after crises (financial and political)
Abdel Latef Anouze 111
Chapter 8 The relationship between Islamic bank efficiency
and stock market performance: Evidence from GCC
countries
Samir Srairi, Imen Kouki and Nizar Harrathi 125

iii
Chapter 9 Conventional banks versus Islamic banks: What makes the
difference?
Huseyin Aytug and Huseyin Ozturk 137

PART 3: PRODUCT DEVELOPMENT IN ISLAMIC FINANCE

Chapter 10 Estimating expected returns on Mudaraba time deposits

of Islamic banks
Zeynep Topaloglu Calkan 151
Chapter 11 Self-adjusting profit sharing ratios for Musharakah financing
Volker Nienhaus 161
Chapter 12 Indexing government debt to GDP: A risk sharing mechanism
for government financing in Muslim countries
Syed Aun R. Rizvi and Shaista Arshad 173
Chapter 13 Concept and mathematics of Islamic valuation and financial
engineering
Nadi Serhan Aydin, FRM and Martin Rainer 181
Foreword
Hatem A. El-Karanshawy
Founding Dean, Qatar Faculty of Islamic Studies, Hamad bin Khalifa University, Qatar Foundation, Doha

The International Conference on Islamic Economics and system. It is through such a setting that thoughts can
Finance (ICIEF) is the leading academic conference in the be debated with the objective of advancing knowledge
discipline organized by the International Association for creation, facilitating policymaking and promoting genuine
Islamic Economics (IAIE) in collaboration with other key innovation for the industry and the markets. Disseminating
stakeholders, including the Islamic Research and Training research presented at ICIEF to the greatest number of
Institute, Islamic Development Bank. It is the pioneering researchers interested in the topic is important. It not only
international conference on Islamic economics organized advances the discourse, but also grants those who did not
first in Makkah Al Mukaramah, Kingdom of Saudi Arabia, have the privilege of attending the conference to partake in
in 1976 under the auspices of King Abdulaziz University the discussion.
and has since been held in numerous locations around the
world. The conference as such has contributed immensely To this end, this series of five volumes (two in Arabic to
to the promotion of Islamic economics and finance. follow) presents the proceedings of 8th and 9th conferences,
Since 2011, the Qatar Faculty of Islamic Studies (QFIS), of which were held in Doha and Istanbul respectively in 2011
Hamad bin Khalifa University, Qatar Foundation, has also and 2013. Each volume focuses on a particular sub-theme
become a key partner in organizing the conference. within the broader theme of Developing Inclusive and
Sustainable Economic and Financial Systems.
The global economy continues to face the perennial
problems of poverty, persistent youth unemployment, The volumes are as follows:
excessive inequalities of income and wealth, high levels of
inflation, large macroeconomic and budgetary imbalances, Volume 1: Access to Finance – Essays on Zakah, Awqaf and
exorbitant debt-servicing burdens, inadequate and aging Microfinance
public utilities and infrastructure, skyrocketing energy Volume 2: Islamic Economics and Social Justice – Essays
prices, and growing food insecurity. The reoccurring on Theory and Policy
regional and global financial crises further intensify Volume 3: Islamic Banking and Finance – Essays on
and magnify these problems, particularly for the Corporate Finance, Efficiency, and Product
underprivileged segments of the world population. As a Development
result, many countries are at the risk of failing to achieve by Volume 4: Ethics, Governance, and Regulation in Islamic
2015 the Millennium Development Goals (MDGs) set by the Finance
United Nations. Hence the achievement of an inclusive and Volume 5: Financial Stability and Risk Management in
sustainable economic and financial system has remained Islamic Financial Institutions
highly illusive.
We hope that this academic endeavor in partnership with
The ICIEF presents an excellent opportunity for those the Bloomsbury Qatar Foundation Publishing will benefit
interested in Islamic economics and finance to present the Islamic economics and finance community and policy
their research and contribute to the development of an makers and that it will promote further academic study of
inclusive and sustainable global economic and financial the discipline.

Cite this chapter as: El-Karanshawy H A (2015). Foreword. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

Acknowledgements
Tariqullah Khan
President, International Association for Islamic Economics

At the International Association for Islamic Economics Farooq, Mohamad Akram Laldin, Mohamad Aslam Haneef,
(IAIE), we are grateful to acknowledge the unprecedented Mohamed Ariff Syed Mohamed, Mohammed Benbouziane,
success of the 8th and 9th International Conferences on Mohammed El-Komi, Monzer Kahf, Muhammad Syukri
Islamic Economics and Finance, which were respectively Salleh, Murat Çizakça, Nabil Dabour, Nafis Alam, Nasim
organized in the Qatar National Convention Centre, Doha, Shirazi, Nazim Zaman, Necdet Sensoy, Nejatullah Siddiqi,
December 19–21, 2011, and in the WoW Convention Centre Rifki Ismal, Rodney Wilson, Ruhaya Atan, Sabur Mollah,
Istanbul, September 9–10, 2013. We greatly appreciate Salman Syed Ali, Savas Alpay, Sayyid Tahir, Serap Oguz
the financial, academic and logistic support provided by Gonulal, Shamim Siddiqui, Shinsuke Nagaoka, Simon
the Qatar Faculty of Islamic Studies, Hamad bin Khalifa Archer, Tariqullah Khan, Toseef Azid, Turan Erol, Usamah
University at Qatar Foundation; Islamic Research and Ahmed Uthman, Volker Nienhaus, Wafica Ghoul, Wijdan
Training Institute at the Islamic Development Bank; and Tariq, Zamir Iqbal, Zarinah Hamid, Zeynep Topaloglu
the Statistical, Economic and Social Research and Training Calkan, Zubair Hasan, and Zulkifli Hasan. The reviewers
Centre for Islamic Countries. of the Arabic papers and abstracts included Abdelrahman
Elzahi, Abdulazeem Abozaid, Abdullah Turkistani,
We offer our sincere thanks to the sponsors of the 8th Abdulrahim Alsaati, Ahmed Belouafi, Ali Al-Quradaghi,
International Conference on Islamic Economics and Aly Khorshid, Anas Zarqa, Bahnaz Al-Quradaghi, Layachi
Finance in Doha. Without their partnership and generous Feddad, Mabid Al-Jarhi, Mohammed El-Gamal, Nabil
contributions, the conference would not have been Dabour, Ridha Saadallah, Sami Al-Suwailem, Seif El-Din
possible. In addition to the Qatar Foundation and the Taj El-Din, Shehab Marzban and Usamah A. Uthman.
Islamic Development Bank, other sponsors included: Qatar
Central Bank (QCB), Qatar Financial Centre Authority The primary objective of the conferences is to further
(QFCA), Qatar National Research Fund (QNRF), Qatar the frontiers of knowledge in the area of Islamic economics
National Bank, Qatar Islamic Bank, Qatar International and finance. Without the hard work and creativity of the
Islamic Bank, Masraf Al Rayan, and Qatar Airways. researchers who shared their work with us, the pool of
knowledge generated in the form of the conference papers
We owe our deepest gratitude to the highly-esteemed panel and presentations would not have been possible. We thank
of reviewers who volunteered to dedicate their time and all the authors who submitted their abstracts and papers to
energy in reviewing all the thousands of abstracts and papers the two conferences.
that were submitted to the conferences. The reviewers
of the English papers and abstracts included: Abdallah The IAIE has always endeavored to publish most of the
Zouache, Abdel Latef Anouze, Abdelaziz Chazi, Abdul Azim significant research papers contributed to its conferences.
Islahi, Abdullah Turkistani, Abdulrahim AlSaati, Ahmet Currently the selected papers of the 8th and 9th conference
Tabako lu, Anowar Zahid, Asad Zaman, Asyraf Dusuki, are being published in five volumes under the common
Ercument Aksak, Evren Tok, Habib Ahmed, Hafas Furqani, theme of Developing Inclusive and Sustainable Economic
Hafsa Orhan Astrom, Haider Ala Hamoudi, Hossein Askari, and Financial Systems. On behalf of the Editorial Board we
Humayon Dar, Ibrahim Warde, Iraj Toutounchian, Jahangir acknowledge that the partnership with the Bloomsbury
Sultan, John Presley, Kabir Hassan, Karim Ginena, Kazem Qatar Foundation Publishing in this regard will be
Yavari, Kenan Bagci, Mabid Al-Jarhi, Maliah Sulaiman, highly beneficial in disseminating research output and in
Marwan Izzeldin, Masooda Bano, Masudul Alam promoting the academic cause.
Choudhury, Mehdi Sadeghi, Mehmet Asutay, Moazzam

Cite this chapter as: Khan T (2015). Acknowledgements. In H A El-Karanshawy et al. (Eds.), Islamic banking and
finance – Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

Preface
Dr. Moh’d Azmi Omar
Director General, Islamic Research & Training Institute (IRTI)

Research in the area of Islamic banking and finance help to align the practices of Islamic banks to its originally
has mushroomed since the new millennium. As the Islamic intended objectives of offering a just and equitable form of
banking industry continues to grow at impressive rates, financing based on sound ethical principles.
researchers have pondered on fundamental questions
relating to whether Islamic banks are unique from an This volume titled Islamic Banking and Finance – Essays
economic standpoint. Islamic banks have been compared on Corporate Finance, Efficiency, and Product Development
to their conventional counterparts on various measures aims to bring together papers on these themes. The volume
such as efficiency and profitability. Capital market consists of selected papers from the 8th International
products such as Islamic mutual funds have also been the Conference on Islamic Economics and Finance held in
subject of recent research. In order to achieve favorable Doha during 19–21 December 2011 and from the 9th
outcomes for the industry, it is important to empirically International Conference on Islamic Economics and Finance
explore and determine the economic characteristics of held in Istanbul during 9–11 September 2013. The papers
Islamic banking and finance. This may inform how we are presented here in their original form as presented at
may go about determining the usefulness of a company to the conferences, with changes limited to copyediting and
issue Islamic debt, for example, and the impact that such correcting typographical errors. The conferences were
an issuance would have on the stock performance of the organized by the Center for Islamic Economics and Finance,
company. In most parts of the world, Islamic banks operate Qatar Faculty of Islamic Studies (QFIS), Hamad bin Khalifa
in a dual banking system together with conventional University; Islamic Research and Training Institute (IRTI),
banks. Researchers may be curious to explore whether Islamic Development Bank (IDB); International Association
operating in a dual banking system affects the type of for Islamic Economics and Finance (IAIE); and Statistical,
Islamic banks that exist in the economy and the kind Economic, and Social Research and Training Centre for
of products that they offer. In light of the recent LIBOR Islamic Countries (SESRIC).
scandals, debates surrounding benchmarking of Islamic
products will recapture the attention of researchers; hence, We have selected the papers in this volume to reflect the
product development in Islamic banking will also remain diversity of research in the field of Islamic banking and
an important area. With the increasing availability of finance. We hope that the papers that have been selected
data, empirical research in Islamic banking is expected to here will help to inform and motivate researchers to
continue to inform industry stakeholders. Understanding conduct rigorous empirical research to explore questions
the Islamic banking phenomenon from an economic that remained to be answered.
perspective with the support of empirical evidence may

Cite this chapter as: Omar M A (2015). Preface. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays
on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

Islamic banking and finance – Essays on corporate
finance, efficiency and product development:
An introduction to the issues and papers
Wijdan Tariq
Senior Researcher, Center for Islamic Economics and Finance, Qatar Faculty of Islamic Studies, Hamad bin Khalifa University

Academic interest in Islamic banking and finance has made at highlighting policy-relevance of research where
grown over the years. The appeal of Islamic banking to possible. The first volume covers issues related to access
the Muslim world is decades old. However, curiosity from to finance and human development including essays on
the wider academic community has increased particularly zakah, awqaf, and microfinance. The second volume
in the aftermath of the global financial crisis and ongoing is a selection of academic and policy-relevant papers
Eurozone crises. In recent years, prominent academics on economic thought from an Islamic perspective with
from conventional economics have either written papers discussions on topics such as fiscal and monetary policy,
on Islamic finance or commented in conferences on the among others. The third volume (this volume) presents
viability of the Islamic finance proposition (e.g., Beck, a wealth of empirical evidence on various issues related
et  al., 2013; Abedifar, et  al., 2013; Zaheer, et  al., 2013; to Islamic banks such as profitability, efficiency, product
Ongena and Şendeniz-Yüncü, 2011; Rogoff, 2011a; Rogoff, development, and Islamic corporate financing. The
2011b; Roubini, 2013, Baele, et al., 2014) This is a welcome fourth volume touches upon ethical, governance, and
development and may help add more rigor and diversity to regulatory issues in Islamic finance. The fifth volume is
the Islamic finance discourse. about financial stability and risk management in Islamic
financial institutions.
Due to the increasing availability of data, empirical work
in Islamic finance is on the rise. The selection of papers in In the following sections, I briefly introduce the papers
this volume reflects this empirical trend. The majority of in this volume. Part 1  includes a selection of papers on
the existing literature attempts to answer questions of a Islamic capital markets and corporate finance. Part 2
comparative nature. Islamic banks are often compared to contains empirical papers on Islamic banking – topics
conventional banks on various measures. Which banks are such as efficiency and profitability comparisons between
more stable, more profitable, and more efficient? A more Islamic and conventional banks. Part 3 comprises papers
fundamental question that some academics have asked is: on product development in Islamic banks using innovative
Are Islamic banks unique to begin with? approaches.

Efficiency and profitability comparisons are perhaps

the most common type of study in the field.1 Generally 1.  Islamic capital markets and corporate
speaking, the more robust studies seem to show that there finance
were few differences between Islamic and conventional The first part of this volume is a collection of papers on
banks before the crisis. During the crisis, however, evidence Islamic capital markets and corporate finance. Islamic
seems to suggest that Islamic banks were less profitable. capital markets refer to those capital market products that
The size of the bank is an important control variable in such comply with Islamic finance principles. These investment
comparative studies. Studies have shown that small Islamic products can be categorized into two broad asset classes:
banks appear to perform better than small conventional Islamic equity markets and Islamic debt markets.
banks, but large conventional banks appear to perform
better than large Islamic banks. Furthermore, recent The Islamic equity markets are stocks and mutual funds that
anecdotal evidence appears to suggest that Islamic banks pass certain financial and nonfinancial screening criteria
may face challenges in liquidity management, hitherto an that render them to be permissible for investment by Islamic
area of lesser concern (Ali, 2013). Product development institutional and individual investors. Financial screening
will thus remain a key topic for research in Islamic banking of stocks is generally based on Shariah principles that
for reasons related to risk management as well as increasing discourage debt and prohibit interest-based transactions. In
competitive advantage. identifying a stock as being Shariah-compliant, a common
financial screen would be, for example, the debt-to-total asset
While organizing these conferences and selecting the ratio. If this ratio exceeds 33% for a company, its stock may
papers for this volume series, a deliberate attempt was be deemed impermissible for investment. Due to the nature

Cite this chapter as: Tariq W (2015). Islamic banking and finance – Essays on corporate finance, efficiency and product
development: An introduction to the issues and papers. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance
– Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Tariq

of modern enterprise, interest income may be tolerated up explores the relationship between bank financing, financing
to a certain percentage, e.g., as long as interest income is rate and bank-specific characteristics in Malaysia, a
less than 5% of all income. Nonfinancial screens are based country whose financial system operates on a dual-banking
on the underlying business that the company is engaged in. system. Using Bankscope data, the result of his pooled
All industries are considered permissible as long as there is panel estimations is that Islamic banks financing behavior
no contravening of basic Shariah principles regarding trade. is dependent on the characteristics of the banks, such as
If a company is engaged in the trading of goods and services its size, liquidity, and capital. This is consistent with the
deemed impermissible by Shariah, then its stock would also behavior of conventional banks. The lending behaviors of
be impermissible. Examples of such goods and services that both types of banks do not differ significantly with respect
Shariah scholars would prohibit include alcohol, gambling, to interest rates. The author also echoes sentiments of
and profane/violent entertainment. Since the introduction purist Islamic economists that Islamic banks should move
of Dow Jones Islamic Indices, a number of similar indices away from debt-based instruments and towards profit-
have been developed in different countries. Each Islamic loss-sharing in order to differentiate themselves from
stock index would have slightly different financial screening conventional banks.
criteria depending on which Shariah scholars are on the
respective Shariah boards. The fourth paper by Bhatt and Sultan adds a leverage risk
factor to the Fama and French multifactor asset pricing
The second asset class in the Islamic capital markets is the model in order to investigate whether Islamic stocks are less
Islamic debt markets, which largely constitute products sensitive to leverage than other stocks such as conventional
known as sukuk. In theory, sukuk are meant to be securitized stocks and socially responsible stocks. In widely cited
representations of undivided shares in an underlying papers, academics have reported anomalies in their
asset or service. In practice, however, the majority of results on asset pricing models (Fama and French (1992);
sukuk prospectuses are drafted by lawyers with the aim Lakonishok, Shleifer and Vishny (1994); Kothari, Shanken
of replicating bond structures. Effectively, the majority of and Sloan (1995)). Despite the methodological challenges
the over $600  billion sukuk that have been issued can be of studying asset pricing models, such as concerns about
considered as the Islamic equivalent of bonds. data-snooping, sample selection biases, irrational behavior
of market participants, and the empirical difficulties of
Conventional corporate finance research is devoted distinguishing between competing hypotheses, this area
mainly to the study of valuation of securities, portfolio of research remains quite popular. The application of such
management, capital structure of firms, dividend policy, factor models to Shariah-compliant stocks is a nascent area
and market inefficiencies. A significant source of corporate of research. In Bhatt and Sultan’s paper, readers will find
financing is lending by banks; a large theoretical and a valuable reference on this topic. The authors find that
empirical literature exists on this topic as well. In recent Shariah-compliant stocks exhibit lower risk premiums to
years, research in the vein of corporate Islamic financing traditional risk factors, but that they are also sensitive to the
has grown with the increasing availability of reliable data. leverage risk factor. This last result may have implications
for asset management practices and will also be of interest
The first paper in this section is about the valuation of to researchers in this field.
sukuk. Ariff and Safari first explore whether sukuk are
equivalent to bonds. Using a unique dataset from the The final paper in this section by Sadeghi, explores a
Bondstream database, they compare sukuk and bonds number of issues surrounding Shariah-compliant stocks.
issued by the same issuers with the same rating in the same He first compares Shariah-compliant sustainable stocks
market. They find evidence suggesting that bonds and and conventional sustainable stocks, and then, using
sukuk are priced differently and that sukuk yields are not event study methodology, Sadeghi investigates the market
determined by bond yields. Their paper is also among the reaction to the addition and deletion of a stock from a
very few papers that attempts to propose valuation models Shariah-compliant index. It is suggested that readers
for pricing sukuk. should regard the results in this conference paper simply
as preliminary findings. Further research is needed before
The second paper by Fauzi, Locke, Basyith and Idris is drawing any conclusions.
an empirical analysis that explores the impact of Islamic
debt (sukuk) on the value of the issuing company. They
find evidence in support of the tradeoff theory of capital 2.  Islamic banking – Efficiency, profitability
structure, which posits that companies actively decide and international perspectives
between equity or debt issuance by assessing costs and The first paper in this section by Sufian and Zulkhibri
benefits. In their sample, they found evidence that suggests presents evidence on the relationship between economic
that the issuance of sukuk was positively associated with freedom and the performance of Islamic banks. Economic
an improvement in the financial performance of the freedom in this paper broadly refers to the extent of
firm, perhaps due to the associated tax benefits. They private ownership of resources as compared to government
conduct further analysis on whether subsequent sukuk control. Measurements of economic freedom are taken
issuances by the same firm also positively affect financial from the Heritage Foundation and include measures of
performance, with mixed results. This paper is relevant to business freedom, trade freedom, investment freedom, etc.
readers interested in theories on capital structure and their The results of the study are mixed. The authors present
application in Islamic finance evidence of a positive association between financial/
business freedom and the profitability of Islamic banks in
The third paper by Zulkhibri of the Islamic Development the MENA region. However, monetary freedom is negatively
Bank focuses on bank financing as a source of capital. He associated with the performance of Islamic banks, lending

xii Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic banking and finance – Essays on corporate finance, efficiency and product development

support to the benefits of government intervention. The propensity score matching generally confirm the findings
authors also offer some preliminary policy evidence based of the regression results. More research in this area will be
on the results of their panel regressions. required to uncover the distinctiveness of using NIM as a
profitability ratio for Islamic banks as opposed to ROA and
Anouze’s paper is in application of two nonparametric ROE, and whether such differences may be driving the
techniques to measure the efficiency and performance of results in this study.
Islamic banks, namely: Data Envelopment Analysis (DEA)
and the Classification and Regression Tree (CART). The
use of DEA to analyze efficiency of Islamic banks is widely 3.  Product development in Islamic finance
applied. It essentially involves grouping data into inputs The first paper in this section by Calkan is an application
and outputs, where input variables are the resources of simple mathematical techniques to enhance the
to be minimized and output variables are the resources competitiveness of Islamic banking products. The
and products/services to be maximized to ensure a high mudharabah-based deposits of Islamic banks – Profit
relative efficiency score. On the other hand, there are only Sharing  Investment Accounts (PSIAs) – have an inherent
a handful of papers that use the CART technique to analyze competitive disadvantage compared to conventional time
Islamic banks. This data mining technique can be used to deposits because profit-rates are not known in advance,
descriptively represent and explore data in a decision-tree which may deter some customers. Calkan proposes that
type of format that makes it relatively simple to understand. while Shariah scholars prohibit the fixing of profit rates
Combining these two techniques allows this paper to make in mudharabah transactions in advance, there is no
a useful contribution to the literature. Anouze analyzes the prohibition on reliably estimating the expected profit rates
period between 1997–2007 to explore the performance and disclosing this to the customer. Through simulation
of Islamic and conventional banks during the period of techniques, the author provides results suggesting that
geo-political crises as well as the onset of the financial Islamic banks may be able to estimate the PSIA profit rates
crises.2 Overall, Anouze identifies 15 factors that future within a 95% confidence interval. This may potentially
researchers can consider as being important in predicting be beneficial to banks by increasing their competitive
efficient banks. advantage and improving their fund management;
hence, this academic research paper has direct practical
The third paper by Srairi, Kouki, and Harrathi explores implications for the industry.
the relationship between Islamic bank efficiency and
stock market performance for 25 Islamic banks in the GCC Nienhaus’s paper is an innovative proposal to structure a
region using Data Envelopment Analysis (DEA) during sukuk product based on musharakah with a self-adjusting
the period 2003–2009. They find evidence that suggests profit sharing ratio. He promotes the use of profit-loss
that pure technical efficiency is positively associated sharing (PLS) products, which are viewed by some Islamic
with stock returns while changes in scale efficiency have economists as the ideal form of Islamic financing. However,
no association with stock performance. Pure technical the uptake of these products has been met with resistance
efficiency here refers to the ability of managers to efficiently from lenders due to the adverse selection and moral hazard
utilize the resources of the bank. Scale efficiency refers to problems that PLS financing arrangements pose. Nienhaus’s
the proportional reduction in input usage if the bank can proposal of self-adjusting profit sharing ratios is an attempt
operate at optimal scale in terms of outputs produced. So to overcome the information problems  in standard
in other words, the authors find evidence that suggests musharakah financing arrangements. Interestingly, the
that stock returns respond positively to improvements concept is based upon existing Accounting and Auditing
in managerial efficiency. The authors also find evidence Organization for Islamic Financial Institutions (AAOIFI)
suggestive of increasing technical efficiency over the years standards that allow for parties in a musharakah
in the sample of Islamic banks. arrangement to adjust profit-sharing ratios, if mutually
agreeable. Nienhaus puts forward a formula (which
Aytug and Ozturk’s paper is an empirical study investigating can be viewed as a customizable template) that would
the differences between Islamic banks and conventional automatically adjust the profit-sharing ratio as new
banks in Turkey between 2003 and 2011. They specifically information emerges on the profitability of the venture.
investigate whether being Islamic has any unique effect This method, he argues, would be less costly than ex-post
on profitability ratios. One of the distinguishing features renegotiations regarding the profit-sharing ratios, because
of this paper is the application of the propensity score in ex-post renegotiations, each party would strive to protect
matching method to estimate the average treatment effect their gains/losses at the expense of the other party. On
of being an Islamic bank. This statistical matching method the other hand, the ex-ante model of self-adjusting ratios
can overcome the concerns of self-selection bias that may proposed by Nienhaus links the profit-sharing ratio to
occur when using more conventional techniques such the actual performance of the venture, so the adjustment
as the Ordinary Least Squares (OLS) and Generalized is non-discretionary. Prior research has been limited to
Method of Moments (GMM) estimators. For the sake of theoretical treatment of the topic (Ahmed, 2002); hence,
comparison, however, the authors do also use the OLS readers would welcome the numerical example that
method. According to the regression results, there is an Nienhaus provides.3 In short, PLS financing arrangements
insignificant relationship between being Islamic and do have a role to play in the financial system particularly
profitability if profitability is measured using the Return in the financing of Small and Medium Enterprises (SMEs),
on Assets (ROA) and Return on Equity (ROE) ratios. When which have been shown to be disproportionally affected by
using the Net Interest Margin (NIM) ratio as a measure for economic downturns. Researchers would be keen to build
profitability, the results suggest that Islamic banks may upon the model provided here to develop more feasible PLS
have a negative relationship with NIM. The results of the financing products for Islamic banks to use.

Eds. Hatem A. El-Karanshawy et al. xiii

 Tariq

The third paper by Rizvi and Arshad is a proposal ponder in the spirit of academic and intellectual discourse.
to introduce Gross Domestic Product (GDP)-linked It is expected that new large-scale evidence on the effects
government securities. The authors also attempt to test of high levels of debt in the economy (including household
the performance/benefits of their proposal using simple debt) will help to inform this debate. The topic of Shariah
simulations on four countries, as well as using descriptive equity screens will continue to evolve and there is expected
statistics and correlation analysis. Their focus is on the to be a convergence between Shariah-compliant and socially
introduction of a GDP-linked sukuk for government responsible investing. In addition to prevailing negative
financing. The idea is that indebted countries may stock screening approaches, perhaps future research will
face economic shocks that reduce their ability to repay explore the viability of positive-screening approaches and
international debt. If the rate of repayment is linked to the role that ‘impact investing’ has to play in Islamic equity
the country’s GDP as opposed to an external benchmark markets. Questions on how policy-makers can promote
such as the London Interbank Offered Rate (LIBOR), economic freedom and protect shareholders and creditors
this would ensure greater risk-sharing between the rights in a way that makes it attractive for Islamic banks
sovereign borrower and international lenders. In times in particular to do business, will continue to be explored.
of economic downturns, for example, the repayments Increasing the competitive advantage of Islamic banks can
on a GDP-linked sukuk would be lower. The authors be achieved perhaps through product development that
provide an overview of GDP-linked sukuk. They present increases information transparency and caters to the needs
some Shariah issues that may arise in structuring such of customers. Finally, attempts at using alternative pricing
products. They provide simple simulations to show the benchmarks for Islamic securities that tie more closely to
potential benefits for countries that adopt GDP-linked the real economy, such as GDP-linked securities, remains a
sukuk. Finally, the authors realize that investors/lenders fruitful area for researchers.
would be reluctant to partake in greater risk sharing;
hence, they discuss the viability of such products from
the perspective of lenders. More work would be needed Notes
to ensure the viability of these products; in particular, 1. For an excellent overview of the empirical literature, I
the Shariah issues that would arise from a musharakah- recommend readers to refer to the chapter by Ongena
based GDP-linked sukuk merits a lengthier discussion and Zaheer (2013).
than the authors have provided. Nonetheless, the paper 2. It can be argued, however, that limiting the period until
makes a reasonable contribution to the growing literature 2007  may not sufficiently capture the effects of the
on GDP-linked securities. financial crisis.
3. Diaw et  al (2011) also provide hypothetical examples
The final paper in this section is by Aydin and Reyner. for their GDP-linked sukuk based on the ijarah.
The paper makes a contribution to the Islamic financial
engineering literature. Unlike previous studies, this
paper uses a rigorous mathematical framework of risk- References
neutral valuation and arbitrage-free contract prices. The
Abedifar P, Molyneux P, Tarazi A. (2013) Risk in Islamic
authors argue that the reference to risk-free zero bonds
Banking. Review of Finance. forthcoming.
in conventional finance is unrealistic. Instead, the authors
propose the use of a commodity basket-linked currency, Ahmed H. (2002) Incentive-Compatible Profit-Sharing
which they describe in detail in the paper. The authors Contracts: A Theoretical Treatment. In: Munawar I,
compare valuation of a conventional forward contract Llewellyn DT (Eds.). Islamic Banking and Finance – New
with a forward contract from an Islamic perspective and Perspectives on Profit-Sharing and Risk. Edward Elgar.
discuss the challenges of using a commodity basket in Cheltenham. 40–54.
the latter case, particularly because their forward prices
Ali SS. (2013) State of Liquidity Management in Islamic
may not be available. Stochastic processes may be useful
Financial Institutions. Islamic Economic Studies.
in this instance for estimation purposes. In short, readers
21(1):63–98.
interested in a financial mathematics and probability
theory treatment of Islamic financial engineering will find Baele L, Farooq M, Ongena S. (2014) Of Religion and
this paper a valuable starting point. The use of financial Redemption: Evidence from Default on Islamic Loans.
mathematics in Islamic finance literature is scarce indeed; Journal of Banking & Finance. 44:141–159.
this fact alone makes this paper noteworthy.
Beck T, Demirgüç-Kunt A, Merrouche O. (2013) Islamic vs.
Conventional Banking: Business Model, Efficiency and
3. Conclusion Stability. Journal of Banking and Finance. 37:433–447.
The papers presented in this volume represent significant Diaw A, Bacha O, Lahsasna A. (2011) Public Sector Funding
contributions to the body of research in Islamic banking, and Debt Management: A Case for GDP-Linked Sukuk.
and suggest fruitful areas of future research. Looking ahead, Paper presented at the 8th International Conference on
the lack of standardized sukuk contracts and the absence Islamic Economics and Finance, 19–21 December 2011,
of liquidity in these markets calls for further research in Doha, Qatar.
financial engineering and regulatory frameworks. There
Fama E, French K. (1992) The Cross-Section of Expected
are continuing calls for Islamic banks to differentiate
Stock Returns. Journal of Finance. 47(2):427–465.
themselves from conventional banks in order to reduce their
dependence on markup-based debt financing; this has been Kothari S, Shanken J, Sloan R. (1995) Another Look at the
a contentious issue that has been ongoing for decades and Cross-Section of Expected Stock Returns. Journal of
that will remain a topic for researchers and practitioners to Finance. 50(1):185–224.

xiv Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic banking and finance – Essays on corporate finance, efficiency and product development

Lakonishok J, Shleifer A, Vishny R. (1994) Contrarian Regulation in the Face of Global Imbalances. Banque
Investment, Extrapolation, and Risk. Journal of Finance. de France, 4 March 2011. Paris, France. Available  at:
49(5):1541–1578. https://www.banque-france.fr/fileadmin/user_
upload/banque_de_france/Economie-et-Statistiques/
Ongena S, Şendeniz-Yüncü I. (2011) Which Firms Engage
La_recherche/GB/Session1-Rogoff.pdf.
Small, Foreign, or State Banks? And Who Goes Islamic?
Evidence from Turkey. Journal of Banking and Finance. Roubini N. (2013) There is a Lot Islamic Finance Can
35:3213–3224. Teach Us. Blog. Available at: http://www.roubiniblog.
com/2013/12/roubini-there-is-lot-islamic-finance.html.
Ongena S, Zaheer S. (2013) Chapter 4: Empirical evidence.
In: Islamic Finance in Europe, Occasional paper series. Zaheer S, Ongena S, Van Wijnbergen S. (2013) The
No. 146, June. European Central Bank. Transmission of Monetary Policy Through Conventional
and Islamic Banks. International Journal of Central
Rogoff K. (2011a). Global Imbalances Without Tears. Project
Banking. 8(5):175–224.
Syndicate. Available at: http://www.project-syndicate.org
/commentary/global-imbalances-without-tears.
Rogoff K. (2011b) What Imbalances After the Crisis? Speech
at International Symposium of the Banque de France  –

Eds. Hatem A. El-Karanshawy et al. xv

Valuation of Islamic debt instruments, the Sukuk:
Lessons for market development
Mohamed Ariff1, Meysam Safari2
Faculty of Business, Bond University, Gold Coast, Australia, T: (617)-5595–2296, E: [email protected]
1

SEGi University, Selangor, Malaysia, T: (60)-17271–4676, E: [email protected]

2

Abstract - A key issue for the fast growing Islamic debt market is whether sukuk instruments
are equivalent to conventional bonds as is practiced today by the market operators. This major
research question is explored here using a large traded data set on sukuk matched with a sample
of conventional bonds issued by the same issuers with same risk-rating and traded in the same
market. Matched samples of sukuk and conventional bonds traded over seven years are used in
our analyses, using statistical and causality tests. The results suggest that sukuk instruments are
priced significantly differently and that their yields are not Granger-caused by conventional security
yields or vice versa. This empirical finding does not support the market’s current practices based
on the assumption that sukuk are like normal bonds. In addition, we describe the basic contract
specifications, core cash flow structures in order to develop and suggest valuation models for three
selected sukuk types. The major implication of these findings is that there is much more research
needed to first document the independent nature of sukuk market behavior and the need for a
re-examination of current market practices.

Keywords: Sukuk, bond, yield curve, yield to maturity, Islamic finance, fixed income securities,
securitization

1. Introduction tangible asset transaction either in ownership or in a master

lease agreement.
Sukuk is the plural form of sakk, which in Arabic means
legal instrument, deed, or cheque.1 It was used in pre-
Islamic Financial Services Board (IFSB, 2009) definition of
Islamic era as withdrawal certificate (a form of cheque) on
Sukuk is “Sukuk (plural of sakk), frequently referred to as
deposits in financial firm or authorized merchants. Later,
“Islamic bonds,” are certificates with each sakk representing
these certificates became instrument for trading. And then
a proportional undivided ownership right in tangible assets,
it became debt instruments traded among willing holders
or a pool of predominantly tangible assets, or a business
of already issued debt just as is a conventional bond.
venture (such as a Mudarabah). These assets may be in a
specific project or investment activity in accordance with
Being a relatively new product, there is no consensus on the
Shariah rules and principles.”
exact definition of sukuk. Some of international regulatory
bodies such as AAOIF, attempted to define sukuk.
Accounting and Auditing Organization for Islamic Financial
Institutions (AAOIFI, 2004) defines sukuk as “Certificates of 2.  Sukuk securities are not the same
equal value representing, after closing subscription, receipt of as bonds
the value of the certificates and putting it to use as planned, This section provides empirical evidences that sukuk
common title to shares and rights in tangible assets, usufructs securities are different from conventional bonds. The
and services, or equity of a given project or equity of a special empirical evidence is first shown as yield curves; second,
investment activity.” International Islamic Financial Market paired sample mean test between pairs of sukuk and bonds
(IIFM, 2010) defined sukuk as a commercial paper that issued by same issuer and for the same maturity period;
provides an investor with ownership in an underlying asset. and third, Granger causality test between abovementioned
It is asset-backed trust certificate evidencing ownership of security pairs. Tests were conducted using data on
an asset or its usufruct (earnings or fruits). It has a stable aggregated monthly yield to maturity (YTM) data2 from
income and complies with the principle of Shariah. Unlike Malaysian securities over August 2005 to April 2013 (total
conventional bonds, sukuk needs to have an underlying of 93 observations for each security).

Cite this chapter as: Ariff M, Safari M (2015). Valuation of Islamic debt instruments, the Sukuk: Lessons for market
development. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Ariff and Safari

Descriptive statistics than 10 years; it is more for periods beyond 10 years. The
Descriptive statistics for various sukuk securities and maximum difference between the yields of sukuk securities
conventional bonds are presented in Table  1. The and the conventional bonds issued by corporate issuers
statistics suggest that the mean yield of sukuk securities with maturities less than 10 years is for those with 2-year
for all types of issuers and for all forms of maturities is maturity with a -7.28 basis points. However, the maximum
3.92 percent. The yields vary within a minimum of 2.91 amount for securities with maturities longer than 10-
(3  months maturity Treasury sukuk securities) and the year term is +6.38 basis points for securities with 20-year
maximum of 5.66 (sukuk securities issued by AAA rated maturity. Long-dated sukuk securities are perceived by
corporations with 20-year maturity). On the other hand, the market as being more risky, thereby attracting higher
the mean yield of conventional bonds of all types of issuers yields. Long dated sukuk are perhaps more risky given the
and all maturities is 3.91 percent. These vary between a risk of greater uncertainty beyond 10 years. It is puzzle
minimum of 2.90 (3 months maturity Treasury bills) and why the same firms issuing short-dated securities provide
the maximum of 5.60 (conventional bonds issued by AAA a safer investment. Obviously again, the yield differences
rated corporate with 20-year maturity). At the issuer level, are systematic.
AAA rated corporate issue of sukuk securities yielded 4.29
percent while the mean of Treasury bill yields is 3.55 On average, there is a -2.44 basis points difference between
percent. On the other hand, the highest conventional yields of sukuk securities and conventional bonds. The total
mean yield for AAA rated corporate issuers is 4.31 percent outstanding value of sukuk securities issued by Malaysian
while the lowest mean yield for conventional bills and AAA Corporate issuers in April 2013 was RM 64.48 billion
notes issued is by the Government of Malaysia with 3.51 (US$ 22  billion). Multiplying yield difference and the
percent. market size indicates that the AAA corporate issuers would
save RM 1.57  billion (US$ 0.53  billion) per year on their
sukuk securities.
Yield curves
The result data are presented in this section, as yield curves
in two plots. Yield curve is the relation between the cost of Comparison of yields of Sukuk securities and
borrowing and the time to maturity of a debt security for a conventional bonds
given issuer for a given rating quality. Yield curves for sukuk Results of the paired sample t-tests are summarized in
securities and conventional bonds issued by government Table 2 (panels A and B) on the equality of means. Out of
and corporations are plotted as in Figure 1 A and Figure 1B. the 20 tested pairs of mean yields of sukuk and conventional
The plots are presented as YTM of (i) conventional against bonds, 19 cases (95 percent) of all pairs showed significant
(ii) sukuk issues in two plots. The two issuer types are of differences in their yields to maturities. In 15 cases, the null
increasingly higher risk rating with sovereign being the hypotheses are rejected at 0.01  significance levels. Thus,
lowest risk – therefore with the lowest yields – on the one one can conclude that the yields of sukuk securities differ
end, and the AAA corporate issues with higher yield at the from the case of conventional bonds, although the issuer
other end. and the issue tenure are the same. Table 2 A is a summary
of the statistics pertaining to the mean yield of sukuk and
As Figure  1(A) suggests, the yields of Government conventional bonds.
Islamic Issues (GII) are higher than those of conventional
bonds issued by the same issuer (Malaysian Government As the t-statistics suggests, the mean yield of sukuk
Securities, or MGS). The difference between sukuk yield securities and conventional bonds are significantly
and conventional bond yield is larger as maturities increase different for all issues by Government. The difference
from 2 years to 15 years. The maximum difference between between the means
the yields of sukuk securities and those of conventional
bonds for this category is with 3 years maturities. The of these are positive, indicating that sukuk securities
difference is 8.28 basis points. On average, there is a 4.04 tend to yield more than conventional bonds issued by the
basis points difference between yields of sukuk securities Government of Malaysia, ceteris paribus. Thus, the market
and conventional bonds. The total outstanding value of associates higher risks (for reasons, we still do not know) to
sukuk securities issued by Malaysian government as at sukuk structures rather than conventional structures.
April 2013 was RM 155.9 billion (US$ 55 billion).
Table 2B is a summary of the statistics for sukuk securities
Multiplying yield difference and market size indicates and conventional bonds issued by AAA rated corporations.
that the Malaysian government needs to pay an extra RM For these securities, the mean yields of sukuk securities and
6.3  billion per year to investors holding sukuk securities conventional bonds are significantly different in all cases
compared to the amount conventional issues of same term except in the cases of 10-year maturity. The differences are
and quality. This means that the sukuk investors earn RM negative for securities with tenure 7 year or less, while, for
6.3  billion higher return compared to the investors in securities with 10 years maturity or more, the difference is
conventional bond market. Obviously sukuk yields are positive. This means the mean of yield of sukuk securities
systematically higher. issued by AAA rated corporations is lower than the yield of
conventional bonds for issues with 7 years or less maturity.
Figure 1(B) shows a plot of the yields of securities issued by For the securities with long-term maturities (10 years and
AAA rated corporate issuers. Yields of sukuk securities are more) the mean yield of sukuk securities is higher than the
less than yields of conventional bonds for maturities of less conventional bonds.

2 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 1. Descriptive statistics of Sukuk vs. Conventional bonds.

Issuer Tenure Mean Median Std. Dev Range Min Max Mean Median Std. Dev Range Min Max

Conventional Sukuk

Eds. Hatem A. El-Karanshawy et al.

Government 3-Month 2.903 2.89 0.519 1.84 1.82 3.66 2.9183 2.89 0.538 1.96 1.82 3.78
6-Month 2.942 2.92 0.527 1.98 1.85 3.83 2.9544 2.92 0.544 2.04 1.85 3.89
1-Year 3.015 2.99 0.525 2.06 1.92 3.98 3.0280 2.99 0.530 2.11 1.97 4.08
2-Year 3.17 3.1 0.447 2.1 2.2 4.3 3.2039 3.11 0.436 2 2.3 4.3
3-Year 3.315 3.21 0.380 2.13 2.37 4.5 3.3976 3.31 0.340 1.84 2.63 4.47
5-Year 3.564 3.5 0.318 1.8 2.78 4.58 3.6359 3.63 0.310 1.8 2.85 4.65
7-Year 3.759 3.74 0.319 1.84 2.91 4.75 3.8003 3.75 0.312 1.79 3 4.79
10-Year 3.943 3.95 0.391 1.93 3.09 5.02 3.9967 3.99 0.358 1.81 3.17 4.98
15-Year 4.186 4.13 0.401 1.77 3.35 5.12 4.2285 4.18 0.374 1.67 3.45 5.12
20-Year 4.357 4.3 0.393 1.58 3.6 5.18 4.3946 4.32 0.372 1.5 3.68 5.18
Mean 3.515 3.473 0.422 1.90 2.58 4.49 3.556 3.509 0.411 1.85 2.67 4.45
Corporate 3-Month 3.3394 3.33 0.453 2.17 2.28 4.45 3.3133 3.29 0.455 2.17 2.24 4.41
6-Month 3.4333 3.39 0.412 1.9 2.6 4.5 3.4009 3.35 0.405 1.9 2.56 4.46
1-Year 3.6178 3.52 0.375 1.59 3.1 4.69 3.5534 3.47 0.359 1.61 3.04 4.65
2-Year 3.8454 3.7 0.359 1.61 3.32 4.93 3.7726 3.67 0.324 1.56 3.3 4.86
3-Year 4.0665 3.97 0.366 1.61 3.62 5.23 4.0016 3.93 0.322 1.5 3.62 5.12
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

5-Year 4.3576 4.27 0.385 1.57 3.87 5.44 4.3082 4.24 0.352 1.56 3.84 5.4
7-Year 4.6380 4.57 0.420 1.7 4.03 5.73 4.5942 4.53 0.405 1.69 4 5.69
10-Year 4.9529 4.93 0.477 1.85 4.21 6.06 4.9433 4.93 0.483 1.84 4.18 6.02
15-Year 5.3006 5.31 0.494 1.98 4.41 6.39 5.3559 5.3 0.580 2.15 4.38 6.53
20-Year 5.6013 5.6 0.521 2.12 4.61 6.73 5.6651 5.66 0.573 2.11 4.58 6.69
Mean 4.315 4.259 0.429 1.81 3.60 5.41 4.291 4.237 0.425 1.80 3.57 5.38

3
 Ariff and Safari

A Yield curve of government issued securities B Yield curve of corporate issued securities
4.4
5.7000

4.2

4 5.2000

3.8
4.7000

3.6

3.4 4.2000

3.2
3.7000
3 Corporate Bonds
Government Conventional Issues
Corporate Sukuk
Government Sukuk Issues
2.8 3.2000
0 5 10 15 20 0 5 10 15 20

Figure 1.  Yield curve for Sukuk securities vs. Conventional bonds.

Table 2A.  Paired samples T-test results: Government. Table 2B.  Paired samples T-test results: Corporate issues.

Δ (Sukuk Δ (Sukuk
Tenure Sukuk Conv - Conv) t-Stat Tenure Sukuk Conv -Conv) t-Stat

3M 2.9183 2.9031 0.0152 2.789*** 3M 3.3133 3.3394 -0.0260 -6.173***

6M 2.9544 2.9424 0.0120 2.461** 6M 3.4009 3.4333 -0.0325 -7.483***
1Y 3.0280 3.0147 0.0132 2.419** 1Y 3.5534 3.6178 -0.0644 -7.360***
2Y 3.2039 3.1699 0.0340 4.789*** 2Y 3.7726 3.8454 -0.0728 -6.789***
3Y 3.3976 3.3148 0.0828 7.686*** 3Y 4.0016 4.0665 -0.0648 -6.047***
5Y 3.6359 3.5641 0.0718 8.465*** 5Y 4.3082 4.3576 -0.0495 -5.685***
7Y 3.8003 3.7588 0.0415 6.418*** 7Y 4.5942 4.6380 -0.0438 -7.318***
10Y 3.9967 3.9433 0.0533 7.037*** 10Y 4.9433 4.9529 -0.0096 -0.953
15Y 4.2285 4.1865 0.0420 5.887*** 15Y 5.3559 5.3006 0.0553 2.414**
20Y 4.3946 4.3567 0.0380 5.043*** 20Y 5.6651 5.6013 0.0638 2.571**
**, ***: Significant at 0.05, and 0.01 significance levels, **, ***: Significant at 0.05, and 0.01 significance levels,
respectively. respectively.

Granger causality test of yields of Sukuk and change in yield of sukuk to verify if it causes a change in
conventional bonds yield of conventional bonds. Second test is on the change
The previous section showed that the mean yield of sukuk in yield of conventional bonds to verify if it causes a
is statistically different from yield of conventional bonds. change in yield of sukuk. The latter test is on the yields
Since each pair of securities issued by the same issuer for of conventional bonds Granger causing yields of sukuk.
the same period of time and for same rating, it is expected Results of pair-wise Granger causality test on each pair is
that the correlation between yields of these securities presented in Table 3.
may be high. This may be a valid reason for a hypothetical
argument that this difference arises from an unidentified The first null hypothesis tested was “yield of sukuk security
causal relationship. One may wish to test if changes in yield does not Granger cause the yield of conventional bond
of one type of security may cause change in the other series? counterparts.” As the statistics in Table 3 suggest, out of 20
In other words, one may want to test for Granger Causality pairs of securities tested, the null hypothesis are rejected in
(Granger, 1969) between yields of sukuk securities and those only 6 pairs at the 0.10 significance level. Yields of sukuk
of conventional bonds to identify if some causal link exists. securities Granger cause yield of conventional bonds in only
6 out of 20 (or 30 percent) pairs. This indicates that one
In order to test this, two Granger causality tests were may not generally conclude that yield of sukuk securities
conducted on each pair of securities. First test is on the Granger cause the yields of conventional bonds. Results also

4 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

Table 3. Pair-wise granger causality tests with lags = 2.

Sukuk security does not Granger Conventional bond does not

cause conventional bond Granger cause sukuk security

Issuer Maturity F-Statistic Prob F-Statistic Prob

Government 3M 0.190 0.8266 0.8972 0.4115

6M 0.7611 0.4702 1.2937 0.2795
1Y 2.5477** 0.0842 1.5399 0.2202
2Y 3.7064** 0.0286 3.7080** 0.0285
3Y 2.4141* 0.0955 5.1703*** 0.0076
5Y 3.6773** 0.0293 0.8004 0.4525
7Y 0.6627 0.5180 0.0314 0.9690
10Y 0.1797 0.8358 0.2506 0.7788
15Y 0.1263 0.8815 0.8322 0.4325
20Y 0.2575 0.7736 1.1982 0.3067
Corporate 3M 0.0684 0.9339 1.6904 0.1905
6M 0.9525 0.3898 3.3457** 0.0399
1Y 3.1337** 0.486 5.6886*** 0.0048
2Y 3.0097* 0.0545 1.6613 0.1959
3Y 0.8961 0.4119 0.7585 0.4714
5Y 1.0998 0.3376 0.3039 0.7387
7Y 1.7843 0.1724 1.2941 1.2794
10Y 0.0696 0.9327 0.7896 0.4573
15Y 0.6228 0.5389 0.5011 0.6078
20Y 0.9677 0.3840 0.9878 0.3766

Note: *, **, ***: significant at 0.10, 0.05, and 0.01 significance levels, respectively.

show that yields of Government sukuk (1, 2, 3 and 5 years) (1 year). In summary, it is reasonable to conclude that,
and those of AAA rated corporate (1 and 2 years) issues with few exceptions, there is no causal relationship
Granger cause their conventional bond yields. Results do between sukuk and conventional bonds. This is the
not show a consistent pattern in terms of issuer or maturity second statistically supported evidence to affirm an
of the security for having a Granger causal effect. argument that the two types of debt instruments are not
the same. This conclusion has important implication for
The second test conducted was to check for the presence of market operation, valuation practices, risk estimation
Granger causal relation between conventional bonds and and regulatory rule setting. These are challenges to be
sukuk. The null hypothesis tested is “yield of conventional addressed in future research.
bonds does not Granger cause the yield of sukuk security.”
Out of the 20 pairs of securities tested, the null hypothesis
is rejected in 4 pairs at the 0.05  significance levels. This 3. Structuring sukuk contracts
indicates that one may not generally conclude that the This section reviews the structure of three major types
conventional bond yields Granger cause sukuk security of sukuk securities namely Ijarah, Musharakah, and
yields. These results show that yield of conventional bonds Mudarabah.
issued by Government (2 and 3 years) and AAA rated
corporate (6  months and 1 year) Granger cause their
sukuk counterparts. Results do not show consistent pattern Ijarah sukuk
in terms of issuer or maturity of the security for having a Ijarah, which means “to give something on rent” (Lewis and
Granger causal effect. Algaoud, 2001), is the reward or recompense that proceeds
from a rental contract between two parties, where the
Finally, as in Table  3, bi-directional Granger causality lessor (the owner of the asset) leases capital asset to the
(as expressed in: Enders, 1995, Hossain, 2005) between lessee (the user of the asset) (Gait and Worthington,
yield of sukuk and yield of conventional bonds is 2007). There is a tendency toward lease financing (Ijarah)
observable in 3 out of 20 (or 15 percent) pairs. In other in Islamic banking sector, since it promises higher yields
words, in 3 pairs of securities, both null hypotheses are than in trade finance (Murabahah) and it also has longer
significantly rejected, or, yield of sukuk Granger cause financing horizon, which is an important feature for
yield of conventional bonds and the other way around. business investments (Daryanani, 2008). In order to be
This may signal that both variables are Granger caused permissible under the Shariah, the Ijarah contract should
by a third variable yet to be explored. Results show satisfy some conditions. The primary requirement is that
that yield of sukuk and conventional bonds have bi- the lessor must be the real owner and in possession of the
directional Granger causal relation in securities issued asset to be leased under contract. As a result, the lessor
by Government (2 and 3 years) and AAA rated corporate should solely bear all risks and uncertainties associated

Eds. Hatem A. El-Karanshawy et al. 5

 Ariff and Safari

to the asset and be responsible for all damage, repair, Then the SPV leases back the assets to the issuer at a specific
insurance, and depreciation of the asset (Khan and Bhatti, predetermined rental fee and then the SPV securitize the
2008). ownership in the assets by issuing sukuk certificates to the
public investors (Lewis, 2007). These sukuk certificates
It could be inferred that charging rental payment is not represent an undividable share in the ownership of the
allowed until the lessee actually receives the possession assets, which entitle the sukuk holders to distribution of
of the asset and shall pay the rental only as long as it is the rental payments on the underlying assets. However,
in usable condition. Moreover, in case of manufacturing the rental payment could be fixed or floating for the whole
defects, which are beyond the lessee’s control, the lessor period, dependent on the leasing contract between the SPV
is responsible. However, the lessee is responsible for the and originator. Since these sukuk certificates represent
proper upkeep and maintenance of the leased asset. The ownership in real assets, they can be traded in a secondary
intention of posing such restriction in Ijarah contract by market.
Shariah is to protect both parties to the contract by reducing
the uncertainty and ambiguity from the agreement (Wilson, The SPV manages the cash flows of the sukuk contract
2004). In addition to that, both lessor and lessee should be by receiving periodic rentals and installments from the
clear on purpose of Ijarah and the usage of assets, moreover, originator and then disbursing the cash flows to the
the Ijarah purpose must comply with Shariah (Al-Omar sukuk holders (Aseambankers, 2005). SPV also manages
and Abdel-Haq, 1996). disbursement of lump sum maturity payments. At the
maturity of a sukuk contract, the SPV no longer has a role
There are two forms of leasing contracts, or Ijarah, in and consequently will cease to exist. However, the Ijarah
Islamic finance. Ijarah, or direct leasing contract, is the case sukuk is typically issued for periods longer than five years
where the lessee uses the capital asset owned by the lessor, and could be considered as long-term debt certificates.
with his/her permission, for a specific period of time for This may raise the issue of SPV’s default risk, so, the
a monthly or annual rent. The owner assumes ownership investors typically receive a direct guarantee from the
title during the whole contract period, and the owner issuer’s guarantee of the SPV obligations (Wilson, 2008).
should performs the ownership responsibilities such as This guarantee also includes the obligation by the issuer to
insurance (Zaher and Hassan, 2001). In this Ijarah contract repurchase the asset from the SPV at the end of the Ijarah
possession of asset should be transferred back to the owner contract at the original sale price.
after the contract matures. In other words, in pure Ijarah
contracts, there is no option to transfer the ownership of Wilson (2008) suggest that SPV does not have any of the
the asset at maturity. risks associated with banks due to SPV’s nature. In other
words, SPV is bankruptcy remote. If the issuer faces
Ijarah wa Iqtina, or hire purchase, is the case of contract the bankruptcy, the creditors to the issuer cannot claim
where the basic intention is transferring the ownership after the assets held by the SPV or otherwise interfere with the
completing the leasing period. Ijarah wa Iqtina is popularly rights of the sukuk-holders with respect to the underlying
practiced when Islamic bank purchases equipment or some assets (Gurgey and Keki, 2008). As a result, SPV would be
other capital asset based on the request of an individual attractive to both issuers and investors, and this may justify
or institutional customer and then rents it to the customer the relatively high legal establishment costs. Figure 2 is a
for a certain fixed rent. On the other hand, the customer contract specification.
promises to purchase the equipment or asset within a
specified period to transfer the ownership from the Islamic Kamali (2007) claims that the fixed and predetermined
bank to the customer (Al-Jarhi and Iqbal, 2001). However, nature of rental cash flow introduces additional risk
it should be noted that the lease contract is completely because the Ijarah sukuk holders receive steady income,
separate and independent from the contract of purchase
of residuals, which has to be valued on a market-basis and
cannot be fixed in advance.

The purchase contract should be an optional, non-binding

contract because the quality and the market price of the
asset at the end of the lease period are unclear (Chapra,
1998). One other approach is the case where the ownership
is gradually transferred to the customer. In this case, and
in addition to the regular rental payment, the customer
shall pay installments of the value of the asset in order to
reduce the ownership share of the lessor in the asset until
the ownership is fully transferred to the lessee (Metwally,
2006). Ijarah wa Iqtina, having strong agreement among
Shariah scholars, is widely used in the real estate, retail,
industry, and manufacturing sectors (Iqbal, 1998, Warde,
2000).

Ijarah sukuk is based on the Ijarah contract. In order to

issue Ijarah sukuk, the originator, who primarily owns the Figure 2. Ijarah sukuk structure – adapted from Wilson
assets, sells the assets to a Special Purpose Vehicle (SPV), (2008), Abdul Majid (2003), El-Gamal (2007),
which is typically a company in an offshore tax-free site. Bose and McGee (2008).

6 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

which is more risk averse than is the case of common shatter one of the most important pillars of the Islamic value
stocks. However, he mentioned general market conditions, system, but also lead to dissatisfaction and conflict among the
price movements of real assets, ability of the lessee to pay partners and destabilize the partnership. The losses must,
the rental or installments, maintenance and insurance cost however, be shared in proportion to capital contribution and
are sources of risks to the Ijarah sukuk. He concluded that the stipulation of any other proportion would be ultra vires
because of these risk factors, the expected return on some and unenforceable.”
of Ijarah sukuk may not be precisely predetermined and
fixed. Thus, the fixed rental may only represent a maximum Lewis and Algaoud (2001) suggest two ways to structure a
that is subject to some possible deductions. Musharakah contact. However, both types are based on the
same general concept of Musharakah, where parties (capital
The major criticism of Ijarah sukuk is that the return is owner and entrepreneur) are ensured an equitable share in
variable or floating in most cases. Moreover, this variable the profit or loss on pre-agreed terms. The difference lies
rate, sometimes for simplification reasons, is mostly in the pre-agreed sharing ratio. In the first method, this
benchmarked or “pegged” to an interest-based index such pre-agreed ratio is pre-fixed and remains constant for the
as the London Interbank Offered Rate (LIBOR) for US$ whole period of the contract while in the second type, the
based sukuk and in local currencies the local rates. Usmani ratio is declining. The diminishing Musharakah contract
(2002) criticized this practice by associating riba to this is preferred by some financiers since it allows release of
form of Ijarah sukuk practice. Shariah scholars suggest the their capital from the investment by reducing its equity
usage of other non-interest benchmarks for pricing and share each year and receiving periodic profits based on the
evaluation purposes. In order to overcome the riba issue, remaining balance. On the other hand, the equity share of
government sukuk could be assessed by macroeconomic the entrepreneur increases over time to the extent that he/
indicators and corporate sukuk could be assessed based on she becomes the sole owner of the firm.
the company performance indicators.
Sukuk based on diminishing Musharakah are gaining
momentum since they enable Islamic banks or Shariah-
Musharakah sukuk compliant investment companies to provide up-front
Iqbal and Molyneux (2005) defined Musharakah as “an investment funding to the issuer. In this regard, both parties
arrangement where two or more parties establish a joint establish a Special Purpose Vehicle (SPV) to administer the
commercial enterprise and all contribute capital as well sukuk. In order to issue a diminishing Musharakah sukuk,
as labor and management as a general rule.” In contrast to the issuer transfers the ownership of an asset to the SPV to
Mudarabah contract, Musharakah investors have the right enter the partnership agreement. The investors enter the
to participate in management of the business partnership, agreement by paying cash. Therefore, both the investors
however, this right is entrusted to each investor (Shinsuke, and the issuer are equity partners in the SPV. However,
2007). It could be argued that Musharakah contract may the investors share in the SPV diminishes over time as the
require establishment of a partnership or company, where issuer pays installments to investors to repurchase their
Musharakah contract parties are the participants and respective shares in the asset. These installment payments
owners (Wilson, 2004). plus the issuer’s rental payments for use of asset (asset’s
generated income) so the contract becomes a Musharakah
Musharakah sukuk securities could be issued based on sukuk with cash flow stream for sukuk-holders. In fixed-
such financing concept. Musharakah type of equity finance ratio Musharakah sukuk, the cash flow stream for the sukuk-
demands that both a profit-sharing ratio and length of holder is only from the income generated from the asset
the joint venture agreement is decided in advance. Similar and not the installment part. The structure of diminishing
to Mudarabah, loss is shared in proportion to the capital Musharakah sukuk is depicted in the Figure 3.
contribution unless the loss is proven to be due to negligence
of one party (Daryanani, 2008). Therefore, all profits and
losses generated from the Musharakah are shared among
the parties on the basis of the pre-agreed ratio. As a result,
Musharakah is basically suitable for financing private or
public companies as well as projects as also practiced by
Islamic banks, where it is typically performed through joint
ventures between banks and business firm for a certain
operation (Gait and Worthington, 2007).

Musharakah, due to its nature and advantages in providing

equal (but proportionate) benefits for all parties, has the
support of all scholars and is valid under Shariah principles.
However, El-Gamal (2000) suggests that parties to
Musharakah usually need the help of legal expert to ensure
that any potential Riba or Gharar is carefully avoided. On
the discussion about Musharakah contracts, Chapra (1998)
concluded that “The only requirement of the Shariah would
be justice, which would imply that the proportional shares
of partners in profit must reflect the contribution made to the
business by their capital, skill, time, management ability, Figure 3. Diminishing musharakah sukuk structure
goodwill and contacts. Anything otherwise would not only (Wilson, 2008).

Eds. Hatem A. El-Karanshawy et al. 7

 Ariff and Safari

Flexibility in payment schedule and amounts has made

diminishing Musharakah sukuk more convenient to use.
However, it should be highlighted that all arrangements
should be agreed upon ex-ante by all parties to the SPV. The
payments are usually monthly or quarterly, but not necessary
in equal amounts (Wilson, 2008). Smaller installments Figure 4. Cash flows pattern for sukuk with fixed promised
could be made during the initial period of the sukuk, with regular payments and promised maturity
most of the asset value or SPV capital remaining with the payment.
investors, but the amount of the installment payments
could increase in a linear fashion, or according to some
predetermined formula. As the issuer’s share in the asset of predetermined maturity payment, r is the discount rate,
increases through the buy-back process, the periodic rental and N is the number of time periods remaining to maturity.
might be expected to decrease due to a decline in remaining
share. However, this does not necessarily have to be the case, Another variation of cash flows of Ijarah sukuk happens
especially if there is capital appreciation in the value of the when the periodic payments follow a growth pattern. In
asset. In other words, when installment and rental payments Ijarah sukuk, maturity payment (M) should be based on
are aggregated, they might be constant, diminishing or market value (hence, a priori undetermined), while the
increasing over time, provided both parties agree to the Ri, the amount of promised regular payment (rental fees)
formula used and the documentation is transparent. in period i is following a predetermined constant growth
model with growth rate of g. However, in practice, maturity
payment of Ijarah sukuk is predetermined and mentioned
in contract documents. Therefore, its cash flow is similar to
4.  Sukuk cash flows the one in Figure 5.
This section reviews cash flow patterns of two major types
of sukuk securities namely Ijarah sukuk and Musharakah
sukuk.

Ijarah sukuk
Ijarah sukuk can have various types of payback structures.
In the simplest form, Ijarah sukuk payback could be fixed
promised regular payments and not a predetermined
Figure 5. Growing promised regular payments pattern
promised maturity payment. The formal Ijarah contract
with predetermined promised maturity
does not have the option for parties to transfer the
payment.
ownership of the asset at the end of the period. Thus, at
the end of an Ijarah contract, the asset should be returned The cash flow pattern of sukuk consists of a growing
to the owner (capital owner or the SPV). In order to annuity of promised regular payments and a promised
transfer the ownership back to the issuer at the maturity, maturity payment. Thus, using the formula for calculating
one should use Ijarah wa Iqtina (lease and purchase) the present value of an annuity, one can formulate the price
contract. Ijarah wa Iqtina sukuk is form of Ijarah contract of a sukuk security as Equation 2.
where the ownership of the asset will be transferred to
lessee (issuer) at the maturity of the sukuk. However, the N
R M R1  1+ g N
 M
maturity payment is not determined at the issuance time P=∑ + = . 1 –   +
of sukuk. The valuation of the asset in this case should be t =1 (1 + r ) (1 + r )
t N
(r - g )   1 + r   (1 + r )
N

conducted at the maturity time, when the market value

of the asset is revealed and maturity payment is set to be  (2)
equal to that.
In Equation 2, P is the price of sukuk, r is the discount rate,
N is the number of periods to maturity, g is the growth
Ijarah sukuk has fixed and predetermined rental payment
rate of promised regular payments, and R1 is the amount
(rewards) and a market valued maturity payment.
of first promised regular payment. It is assumed that the
However, in practice, the maturity value of the property is,
promised payments are growing at a constant rate of g,
sometimes, fixed and predetermined to both sides of the
thus, R2 = R1(1 + g).
contract. Therefore, the cash flow pattern of Ijarah sukuk
would be similar to that of a coupon-bearing conventional
bond such securities are depicted in Figure 4. The valuation Musharakah Sukuk
of this form of sukuk is similar to a straight bond because
The simplest form of cash flow generated by a Musharakah
of the similarity of cash flows pattern and the tradability.
sukuk security is obtained from those securities that
Thus, a sukuk could be priced by using the Equation 1.
only pay a lump sum amount of cash in a certain and
predetermined point of time in future at maturity.
N
R M R 1  M There is no cash payment to investors prior to maturity
P=∑ + = 1 - + (1) as shown in Figure 6. The amount of maturity payment,
t =1 (1 + r ) (1 + r )
t N
r  (1 + r ) N  (1 + r ) N
however, is not determined and should be based on the
performance of the underlying investment. However, as
In Equation  1, P is the price of sukuk security, R is the in most practice cases, the maturity payment is fixed and
amount of periodical promised payment, M is the amount predetermined.

8 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

Figure 6. Zero-promised regular payment cash flow

Figure 8. Declining promised payments cash flows
pattern of Musharakah Sukuk securities.
pattern.

This form of cash flow is the same as conventional discount Thus, price of a diminishing Musharakah sukuk can be
bond cash flow or a bond with balloon payment. Thus the formulated as Equation 5.
same valuation process is applicable. The valuation of these
forms of sukuk could be performed using the conventional  N

. 1 –  1 + g   
N
R R1
pricing approach as follow. The current price of sukuk is the P=∑ = (5)
maturity payment (face value) discounted to the present t =1 (1 + r )t (r - g )   1 + r  
time, similar to discount bonds.
In Equation  5, P is the price of diminishing Musharakah
M 
P= (3) sukuk, r is the discount rate, N is the number of periods to
(1 + r )T -t
maturity, g is the negative growth rate of promised regular
payments, and R1 is the amount of first promised regular
In Equation  3, P is the price of sukuk, M is maturity payment. It is assumed that the promised regular payments
payment (face value), T is maturity date, t is time, and r are declining at a constant rate of g, thus, R2 = R1 (1 + g) < R1.
is the discount rate. It should be noted that r should not be
based on any interest bearing benchmark.
5.  Sukuk world markets
Diminishing Musharakah sukuk may possess payoff The sukuk market has grown rapidly in recent years.
structure in a manner that it only pays some promised Emergence of more than 250 Takaful (Islamic Insurance)
regular payments at certain periods of time with zero companies, 350 Islamic equity funds, and 370 Islamic banks
maturity payment. Amount of promised regular payments worldwide has made a great demand for sukuk. Khan and
are fixed and predetermined. Cash flow pattern of such Bhatti (2008) highlighted that sukuk constitute about 85
security is depicted in Figure 7. percent of the Middle Eastern capital market, US$13bn of
them have been issued there with an average growth rate
of over 45 percent during 2002–2007. The Middle East and
Asian regions will primarily rely on sukuk to meet their US$1.5
trillion infrastructure needs over the next ten years (John,
2007) is the kind of statements one reads in commentaries.

McKenzie (2008) investigated the sukuk market statistics

Figure 7. Fixed promise regular payment cash flows and reported that, while sovereign sukuk issues by Bahrain
pattern of Sukuk without promised maturity and Malaysia played an initial role in establishing the
payment. market, about 80 percent of issues between 2001 and
2006 have been corporate issues. However as reported in
McKenzie (2010), most sukuk issuers are government or
Cash flow represented in Figure 7 is identical to a constant quasi-government organizations in 2009. In addition to
annuity. Thus, in order to evaluate the price of a diminishing that, he found that the most important market for corporate
Musharakah sukuk, present value of annuity is applicable. sukuk issues totaling US$44 billion over the period of 2001–
This will result into Equation 4. 2006 has been for infrastructure finance, with issuance
of US$17billion, 39 percent of the total. The next largest
N
R R 1  markets were for financial services (18 percent) and energy
P=∑ = 1 -  (4) (6 percent). McKenzie (2009) claims that there is a growing
t =1 (1 + r ) t
r  (1 + r ) N 
appetite and demand for investment in sukuk that goes well
beyond Islamic investors amongst those investors wishing
In Equation  4, P is the price of diminishing Musharakah to gain exposure to diverse but high quality assets. The IFSL
sukuk security, R is the amount of periodical promised Islamic Finance (2012) reported that the sukuk issuance
payment, r is the discount rate, and N is the number of time has increased significantly in past three years to the level of
periods remaining to maturity. US$84.4billion in 2011. The amount of new sukuk issues in
last ten years is shown in plots in Figure 9.
In diminishing Musharakah sukuk, amount of promised
regular payment in period i is following a predetermined According to International Islamic Financial Market
constant negative growth (declining) model with growth (IIFM) 2010 report, the sukuk market size grew to over
rate of g, which is a negative number. In other words, the US$136billion in mid-2009. As at the end of 2010, the total
cash flow stream of regular payments declines to zero at asset value of publicly-listed sukuk was more than US$
maturity (M is also equal to zero). The cash flow diagram is 197 billion in 13 markets (Alvi, et al., 2011). Mass media
presented in Figure 8. reports suggest a much higher asset value by including

Eds. Hatem A. El-Karanshawy et al. 9

 Ariff and Safari

90.0 mostly practiced among Islamic investment companies

80.0 and banks. In the process of issuing sukuk, according to
70.0
Liquidity Management Center of Bahrain (2008), Shariah
60.0
50.0
advisors should study proposed sukuk structures and
40.0 suggest a Shariah structure which fulfills the set economic
30.0 aims sought in the issue. In addition, they should work
20.0 closely with legal counsel of the issuer and the arranger
10.0 (investment banker) to ensure that the legal documents
0.0 are in line with Shariah requirements. Finally, they should
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Issue a fatwa, which is an opinion, as to the compliance of
Figure 9. Amount of worldwide Sukuk issues (Source: the sukuk with Islamic principles—similar to the opinion
IFSL, Zawya Sukuk monitor). tax authorities provide on a tax question—on the whole
sukuk deal before it can be put into circulation.

privately-issued sukuk in a number major financial centers Islamic financial institutions are subject to rules and
such as Zurich, London, Frankfurt, Singapore and others. regulations of the local Shariah authority. Some countries
Hence, a figure of US$ 840  billion is suggested in Ariff such as Malaysia have set up their own Shariah Advisory
et al. (2012). Currently, sukuk are offered in specialized Council (SAC) at the national level, which oversees the
exchanges such as the Labuan Exchange in Malaysia, the consistent application across similar situations in financial
Third market in Vienna, the Dubai International Finance interactions. These councils are part of the Securities
Exchange, and the London Stock Exchange (Asad, 2009b). Commission or also part of the Central Bank of the country
Governments and regulators in a variety of countries have at national level. At the international level (the Organization
recognized the important role that sukuk can play in capital of Islamic Countries) there is a Shariah Council in Saudi
markets and have been giving priority to developing their Arabia.
countries as sukuk centers (Abd Razak and Abdul Karim,
2008). In a global perspective, there are few international
organizations that attempt to regulate and screen the
conduct of sukuk issuance and trade. Among these
international organizations, AAOIFI, IFSB, and IIFM are
the most influential ones (DIFC, 2009). Although these
3% organizations try to base their rulings on Shariah principles,
2%
3% 3%
there are occasions that sukuk based on their guidelines have
Malaysia
variations in formation. Not being national bodies, these
4% rules cannot be imposed, so remain voluntary. Therefore,
Qatar
Siddiqui (2008) highlighted that more communication
UAE between these organizations will bridge the differences
5% existing between sukuk contracts.
Indonesia
The decisions regarding permissibility of each sukuk
11% S. Arabia contract, as mentioned above, is made by expert Muslim
scholars who are appointed to the Shariah Board. Number
Bahrain
69% of these experts are estimated to fall between 100 and
Pakistan 200, worldwide (Asad, 2009a). This indicates the urgent
need for training of expert Muslim scholars to sit on
Others Shariah boards in Islamic financial institutions. In order
to address this shortfall, some Islamic institutes such as
ISRA3 and INCEIF4 are planning to design special programs
Sukuk issuance by country, 2011 (Source:
Figure 10.  for training certified Shariah scholars. However, there are
IFSL, Zawya Sukuk monitor). some other institutions that are currently providing short
time courses on this topic.5
6.  Sukuk regulations
All Islamic financial and banking interactions, similar to AAOIFI
their conventional counterparts, are subject to regulations The Accounting and Auditing Organization for Islamic
of professional authorities. Moreover, in order to be Financial Institutions (AAOIFI) in its website6 introduces
recognized as an “Islamic” transaction, sukuk products itself as “an Islamic international autonomous non-profit
must follow some extra procedures required by Islamic corporate body that prepares accounting, auditing,
authorities (Iqbal, 1999). Shariah regulations governing governance, ethical and Shariah standards for Islamic
sukuk, similar to other Islamic finance and banking financial institutions and the industry.” AAOIFI was
practices in general, are dictated from three sources; established in accordance with the Agreement of
international organizations, local authorities, or in-house Association, which was signed by Islamic financial
Shariah boards. institutions in 1990  in Algiers. Then, it was registered
in 1991  in Bahrain. As an independent international
Some major Islamic financial institutions have their organization, AAOIFI is supported by institutional
own in-house Shariah Supervisory Boards (SSB). This is members (more than 200  members from 45 countries).

10 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

Members include central banks, Islamic financial a) Sukuk issuances have to be backed by real assets, the
institutions, and other participants from the international ownership of which has to be legally transferred to
Islamic banking and finance industry, worldwide. sukuk holders in order to be tradable
Esta­blishment of AAOIFI represented a shift from the b) Sukuk must not represent receivables or debts, except
authority of independent Shariah supervisory boards set in the case of a trading or financial entity selling
up in individual Islamic banking and finance operations, all its assets or a portfolio with a standing financial
to a centralized model for the dissemination of standards, obligation, in which, some debts owing by third
procedures and best practices (Gambling et  al., 1993; parties, incidental to physical assets or usufruct, are
Maurer, 2002, Pomeranz, 1997) unintentionally included
c) The manager of the sukuk is prohibited from extending
In May 2003, AAOIFI issued Shariah standard FAS 17 “loans” to make up for the shortfall in the return on
titled “Investment sukuk.” In this standard, AAOIFI issued the assets, whether acting as a mudarib (investment
standard for 14 different types of sukuk, where some of these manager), or sharik (partner) or wakil (agent)
sukuk are classified as tradable and others are classified d) Guarantees to repurchase the assets at nominal value
as non-tradable based on the type and characteristics upon maturity with the exception of Ijarah sukuk
of the issued sukuk (AlBuolayan, 2006). AAOIFI and structures are also prohibited
its Shariah Board chairman, Shaikh Muhammad Taqi e) Closer scrutiny of documentation and subsequent
Usmani, have special attention to sukuk because it is one execution of the transaction is required by Shariah
of the most favored Islamic financing instruments and Supervisory Boards
there is a possibility of variation in contract formation by
practitioner. Shaikh Usmani has given some comments Maurer (2010) investigated this controversial issue
on the practice of sukuk over time. For instance, he has and concluded that for some, Usmani’s salvo was a long
highlighted that “… since Ijarah sukuk represent the pro overdue and much needed corrective to what they saw as
rata ownership of their holders in the tangible assets of the the excesses of sukuk issuances and structured financing
fund, and not the liquid amounts or debts, they are fully vehicles that came very close to mimicking conventional
negotiable and can be sold and purchased in the secondary bonds. To others, it was an overreaction, born of impatience
market. Anyone who purchases these sukuk replaces the with the pace of development of Islamic financial
sellers in the pro rata ownership of the relevant assets and institutions and markets, and an unrealistic appraisal of
all the rights and obligations of the original subscriber are what Islamic finance can actually accomplish in a globally
passed on to him. The price of these sukuk will be determined interconnected and interdependent world.
on the basis of market forces, and are normally based on their
profitability” (Usmani, 2001).
IFSB
However, his most cited and debated comment regarding The Islamic Financial Services Board (IFSB), which is based
sukuk is his statement in November 2007 which declared in Kuala Lumpur, was established in 2002 and started
that some 85 percent of outstanding sukuk had failed the operations in early 2003. It serves as an international
Shariah-compliance test on the basis that they were ‘asset- standard-setting body of regulatory and supervisory
based’ rather than ‘asset-backed’ with the guaranteed agencies that have vested interest in ensuring the soundness
return of the face value of the sukuk on maturity and in the and stability of the Islamic financial services industry, which
absence of a transfer in asset ownership to sukuk holders is defined broadly to include banking, capital market and
(Usmani, 2007). Since then, the juridical validity of sukuk insurance. In advancing this mission, the IFSB promotes
became suspect (Hasan, 2010, Alsayyed and Malik, 2010). the development of a prudent and transparent Islamic
His reasons for such declaration were in brief as under. financial services industry through introducing new or
adapting existing international standards consistent with
• There have been cases where the assets in the sukuk Shariah principles and recommend them for adoption.
were the shares of companies that do not confer true To this end, the work of the IFSB complements that of the
ownership but which merely offer to sukuk holders a Basel Committee on Banking Supervision, International
right to returns. Organization of Securities Commissions and the
• Most sukuk issued are identical to conventional International Association of Insurance Supervisors.
bonds with regard to the distribution of profits from
their enterprises at fixed percentage bench-marked Drafting of standards in IFSB is done on a task force working
on interest rates. The legal presumption regarding group method. The IFSB council appoints members of
sukuk is that no fixed rate of profit or the refund of technical committee which are responsible for advising the
capital can be guaranteed. council on technical issues within its terms of reference.
• Virtually, all sukuk issues guarantee the return Islamic Development Bank (IDB) Shariah Supervisory
of the principal to holders at maturity (just as Board is responsible for Shariah supervision of IFSB’s
in conventional bonds) through a binding promise standards. IFSB has issued three standards that affect the
from either the issuer or the manager to repurchase issuance, trading, or investing in sukuk:
the assets at the stated price regardless of their true
or market value at maturity. • IFSB-1: Guiding principles of risk management
for institutions (other than insurance institutions)
Later on, in February 2008, AAOIFI issued a guidance offering only Islamic financial services, issued in
statement on accounting for investments and amendment December 2005. This guideline also includes the
in FAS 17 (AAOIFI, 2008). Summary of important issues various risk elements affecting institutions offering
raised in this guideline are: investment certificates such as sukuk and operational

Eds. Hatem A. El-Karanshawy et al. 11

 Ariff and Safari

consideration regarding them. This guideline offers The introduction of sukuk rating in the 90  s represents
a general perspective toward risk sources and risk another critical milestone in the development of sukuk
management and is not specific for sukuk. market. Bond ratings are principally designed to arrive at
• IFSB-2: Capital adequacy standard for institutions a reasoned judgment on credit risk via a careful analysis of
(other than insurance institutions) offering only the critical issues surrounding a specific debt on the issuer
Islamic financial services, issued in December 2005. (Mohd Asri, 2004).
This standard overviews various Islamic contracts
(some of which are underlying contracts of sukuk) From the global capital markets point of view, sukuk can be
and provide capital requirement for each. rated just like any conventional bond, and can be traded as
• IFSB-7: Capital adequacy requirements for sukuk, such, as well (Maurer, 2010). In general, rating agencies
securitizations and real estate investment, issued have the same criteria for corporate bonds rating. The
in January 2009. The first part of this guideline criteria incorporate issue structure (repayment schedule
investigates the sukuk and securitization of it, sukuk and debt types), business risk analysis, financial risk
structures, operational requirements pertaining to analysis, management, ownership and other qualitative
sukuk, treatment for regulatory capital purposes of factors (Mohd Asri, 2004). However, realizing the
sukuk and securitization exposures, and treatment of uniqueness and types of sukuk, the rating methodology
credit risk exposures of sukuk. should be different to that of conventional bonds rating
(Jalil, 2005). Rosly (2007) argued that sukuk structures
falls in 2 categories:
IIFM
The International Islamic Financial Market (IIFM), located • Asset-Backed sukuk, for which ratings are dependent
in Bahrain, is a global standardization body for the on a risk analysis of the asset. However, investors
Islamic capital and money market segment of the Islamic hold rights to underlying assets through SPV and
financial services industry. Its primary focus lies in the not directly; hence, sukuk performance is driven by
standardization of Islamic products, documentation and assets and not linked to the originator.
related processes. IIFM was founded with the collective • Unsecured sukuk, for which ratings are primarily
efforts of Central Bank of Bahrain, Bank Indonesia, Central dependent on the riskiness of the sponsor, originator,
Bank of Sudan, Labuan Financial Services Authority or the borrower.
(Malaysia), Ministry of Finance (Brunei Darussalam) and
Islamic Development Bank (a multilateral institution based Therefore, similar to conventional bonds, risk elements
in Saudi Arabia). Besides the founding members, IIFM is affecting sukuk should be thoroughly investigated by rating
supported by its permanent members, namely State Bank agencies. Among risk elements, Rosly (2007) mentioned
of Pakistan and Dubai International Financial Centre that credit risk is the most critical one. Other factors he
Authority (UAE). IIFM is further supported by a number highlighted are currency risk (for international issues),
of regional and international financial institutions as well tax risk, and reserve funds. In contrast to highly sensitive
as other market participants as its members. IIFM activities conventional bonds, sukuk are less sensitive to interest
are under supervision of its Shariah Advisory Panel, rate. Zurich based investment bank Credit Suisse believes
which currently has ten members. Focus of IIFM’s work investment in Islamic Finance and Banking products
is on Islamic capital and money markets. Presently, IIFM are not expose to interest rates since Islam prohibits
has no specific standard or guideline pertaining to sukuk. charging interest and sukuk securities are unaffected to
However, it has released two reports on sukuk in 2010 the credit crisis in the international finance and banking
and 2011. In these reports, IIFM investigated the current industry (Farook, 2009). Tariq (2004) summarized risk
sukuk market from an international as well as domestic characteristics of each type of sukuk structures which is
perspective. It also investigated various sukuk structures depicted in Table 4.
international issues. Moreover, it has studied some of sukuk
issues as case study.
Security commission
Security Commissions in each country is a statutory body
Sukuk rating that investigates the conduct of financial markets and has
Rating is an evaluation of a corporate or municipal enforcement power. Among all of their responsibilities,
bond’s relative safety from an investment standpoint. some are related to origination and exchange of sukuk
In conventional sense, it scrutinizes the issuer’s ability securities. These responsibilities are supervising
to repay principal and make interest payments. Then, a exchanges, clearing houses, and central depositors;
grade (the most controversial part of rating process) is approving authority for corporate bond issues (including
given to the bond that indicates its credit quality. Private sukuk securities); and regulating all matters relating to
independent international rating companies such as securities and futures contracts. Hence, one may assume
Standard & Poor’s, Moody’s, and Fitch, or domestic rating that the all sukuk securities issued require permission
agencies like RAM and MARC of Malaysia, provide these and approval by Security Commission of that particular
evaluations of issuer’s financial strength, or the ability jurisdiction. Beyond the regulatory obligations, they also
to pay a bond’s principal and interest in a timely fashion. act as a strategic policy making body on financial markets.
As a result, bonds are rated in a range from AAA or Aaa
(the highest), to C or D, which represents a company that Among these incentives,7 some directly affect the issuance
has already defaulted. Each rating company has its own of sukuk securities in Malaysia. Issuers of sukuk securities
definition and methodology for rating and own set of in Malaysia are offered tax deduction on expenses incurred
rating ranges. in due course of issuance of Wakalah, Murabahah, Bai

12 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 4.  Summary of risk characteristics of sukuk structures – source: (Tariq, 2004).

Description of Sukuk Rate of return

Type of Sukuk structure Credit Risk (Interest rate risk) FX risk Price risk Other risks

Zero coupon Sukuk Istisna, Murabahah Unique basis of Very high due to If all other conditions Related to the Liquidity Risk is serious
debt certificates – credit risks exist, fixed rate, are similar, FX risk underlying as far as the non-tradable
non-tradable (Khan and remains for the will be the same for commodities prices Sukuk are concerned.
Ahmed, 2001) entire maturity all cases of Sukuk. and assets in relation Business risk of the
of the issue However, those to the market prices. issuer is a risk underlying
Fixed Rate Ijarah Securitized Ijarah, Default on rent Very high due to Sukuk which are Ijarah Sukuk is most Sukuk as compared to

Eds. Hatem A. El-Karanshawy et al.

Sukuk certificate holder payment, fixed fixed rate, liquid or which are exposed to this as traditional fixed incomes.
owns part of asset rate makes credit remains for the relatively short term the values of the Shariah compliance risk
or usufructs and earns risk more serious entire maturity in nature will be underlying assets is another one unique in
fixed rent – tradable of the issue less exposed. may depreciate case of Sukuk.
Floating Certificate holder Default on rent Exists only within The composition faster as compared Infrastructure rigidities,
Rate Ijarah Sukuk owns part of asset payment, floating the time of the of assets in the pool to market prices. lack of efficient
or usufructs and rate makes default floating period will also contribute Maintenance of the institutional support
earns floating rent risk lesser serious – normally to the FX risk in assets will play an increases the risk of
indexed to market see previous case 6 months different ways. important part in this Sukuk as compared
benchmark such as Hence this can be process. Liquidity to traditional fixed
LIBOR – tradable very useful tool to of the Sukuk will also incomes, (Sundararajan
Fixed rate Hybrid/ Securitized pool of Credit risk of debt Very high due to overcome the FX play an important part and Errico, 2002)
Pooled Sukuk assets; debts must part of pool, fixed rate, risk by diversifying in the risk. Salam is
not be more than default on rents, remains for the the pool in different also exposed to price
49%, floating rate fixed rate makes entire maturity currencies. risks. However,
possibility exists – credit risk serious of the issue through parallel
tradable contracts these risks
can be overcome
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

Musharakah Medium term Musharakah has Similar to floating

Term Finance redeemable high default risk rate, but, unique
Sukuk (MTFS) Musharakah (Khan and Ahmed, because the rate
certificate based 2001), however, is not indexed
on diminishing MTFS could be with a benchmark
Musharakah – based on the like LIBOR,
tradable as well strength of the hence least
as redeemable entire balance sheet exposed to
this risk
Salam Sukuk Securitized salam, Unique credit risk Very high due
fixed-rate and (Khan and Ahmed, to fixed rate
non-tradable 2001)

13
 Ariff and Safari

Bithaman Ajil, Musharakah, Mudarabah, Ijarah, and Istisna Amin (2011) suggests that as there are more attempts to
sukuk securities until 2015. Moreover, the Malaysian develop an internationally accepted accounting standards
Security Commission (SC) has recognized SPV as a by IASB, Islamic accounting should follow their lead
channel to transfer funds and hence, exempted them from and focus more on requiring such standards to adhere to
income taxes. In addition to that, the issuer is permitted to the demands of the Muslim world. He suggests that IFRS
tax deduct the cost of issuance of sukuk incurred by SPV may include Islamic requirements in the standards by
Company. On the other hand, sukuk investors are receiving mandating practitioners to provide such information as
some incentives such as tax exemption on profits earned footnote disclosures or in other formats.
from investing in sukuk securities (except for the profits
due to convertible loan stocks). Former issues indicate that there are potential conflicts
between the outcome of following AAOIFI standards
To add to the sukuk market flavor, SC offers intermediaries and IFRS standards in practice. Hence, firms may face a
income tax exemption for qualified institutions “in respect dilemma in adopting the proper accounting standards.
of statutory income derived from regulated activity of dealing Since issuance of sukuk securities is one of the many
in securities and advising on corporate finance relating to accounting and finance practices of firms, they may
arranging, underwriting and distributing non-ringgit sukuk generally follow the IFRS standards, which are more
originating from Malaysia which are issued or guaranteed comprehensive. In addition to that, audit firms may prefer
by the Government or approved by the SC until the year of conducting audit service for firms who follow a more
assessment 2014.” acceptable standard (IFRS) rather than a less common
one (AAOIFI). This might result in higher fees for auditing
firms that adopt AAOIFI standards and make them less
7.  Institutional developments favored. Moreover, majority of accountant practitioners
Similar to conventional practices of accounting, there is are educated with IAS methods and standards. The
no unique standard for Islamic accounting and finance. number of well-versed accountants in Islamic regulations
However, in a similar fashion to conventional counterparts, is still lower than the demand; hence, many firms ought to
there are attempts to develop one. AAOIFI, which gained hire conventional accountants.
respect of many Islamic countries and institutions, has
developed standards for accounting statements. In the These issues have hindered the acceptance and adoption
section  6.2 of the “Statement of Financial Accounting,” of AAOIFI standards by many firms. Therefore, similar
AAOIFI states that the following is the main focus of to Amin’s (2011) recommendation, we suggest that
financial accounting in Islam: “on the fair reporting of the the AAOIFI should focus more on collaboration with
entity’s financial position and results of its operations, in a international standard setting bodies and negotiate with
manner that would reveal what is halal (permissible) and them to incorporate and embed the demands of Muslim
haram (forbidden). Moreover, they highlighted one of the practitioners as well as investors in the internationally
key objectives of financial reports: Information about the recognized standards such as IFRS.
Islamic bank’s compliance with the Islamic Shariah and its
objectives and to establish such compliance; and information
establishing the separation of prohibited earnings and 8.  Future developments
expenditures, if any, which occurred, and of the manner in Pro-market regulatory framework is a key feature for long-
which these were disposed of.” term growth and sustainability of any financial market
development. Presently, the contracts are arranged and
Such an objective is not merely the same as the objectives organized individually and hence, the cost of issuance
set by the IASB (International Accounting Standards of sukuk securities is higher compared to conventional
Board) in IFRS (International Financial Reporting bonds. Lack of standardized sukuk contracts results in
Standards). In the 2001 “Framework for the Preparation similar costly structuring of contracts. Standardization of
and Presentation of Financial Statements,” IASB contracts strengthens the convenience of securitization
mentioned that the objective of financial statements process, and will boost the cooperation among regulatory
is to “provide information about the financial position, bodies. This will lead to one-stop servicing of primary
performance and changes in financial position of an entity market making, thus lessening the burdens of issuance
that is useful to a wide range of users in making economic process.
decisions.” Hence, one may conclude that these two
objectives are not necessarily the same. Market deepening activities should be undertaken to
improve the lack of liquidity in the market. Majority of
Amin (2011) examined the implication of such difference sukuk markets suffers from illiquidity to the extent that 70
in objectives and showed that there would be an actual percent or more are not traded as at 2011 has never been
difference in outcomes. He claimed that due to the publicly traded since the issuance until their maturity.
difference in standards, particularly with respect to the Introduction of discount houses as the intermediary
presence of SPV in the structuring of sukuk securities, between the exchange and the custormers would improve
there would be a different outcome in financial reports. liquidity.
He argues that the root cause of such difference lies in the
different interpretation of the role of the SPV, especially the Market broadening activities are also required to further
fact that AAOIFI requires the ownership transfer of assets develop sukuk markets. Introducing more securities
to it, while the IFRS considers sukuk transaction purely as a targeting at specific needs of customers would help
financing transaction and does not require the transferring introduce a broader list of securities. For example, the
of title of assets in the balance sheet statement. sukuk market has 6 traded ones although conceptually

14 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

there are 14 different types. These securities, though References

combination of two or more types of existing contracts into
AAOIFI (2004) Shariah Standards for Financial Institutions.
one structure, would result in new a security to address the
Accounting and Auditing Organization for Islamic
unique needs of the issuer. More efforts should be put into
Financial Institutions (AAOIFI).
financial engineering of sukuk securities.
AAOIFI (2008) Guidance Statement on Accounting for
Finally, education is another critical area that needs more Investments and Amendment in FAS 17. Accounting and
attention in order to ensure the continuity and sustainability Auditing Organization for Islamic Financial Institutions
of sukuk markets. Currently, education and training efforts (AAOIFI).
across the world is highly un-coordinated. Standardization
Abd Razak, D. & Abdul Karim, M.A. (2008) Development
of education and training on Islamic finance in general
of Islamic Finance in Malaysia: A Conceptual Paper. 8th
and in particularly sukuk securities is absent. There is no
Global Conference on Business & Economics. Florence, Italy.
accreditation for different programs offered to the public.
Although an international accreditation body for education Abdul Majid, A.R. (2003) Development of Liquidity
programs is ideal, national accrediting bodies may serve Management Instruments: Challenges and
the need for the time being. Existence of accreditation for Opportunities. International Conference on Islamic
Islamic finance programs may not only ensure the quality Banking: Risk Management, Regulation and Supervision.
of the programs, but also is vehicle for standardization of Jakarta –Indonesia.
these programs.
Al-Jarhi, M.A. & Iqbal, M. (2001) Islamic Banking: Answers
to Some Frequently Asked Questions Jeddah, Saudi
Although some limited number institutions (like INCEIF in
Arabia, Islamic Research and Training Institute, Islamic
Malaysia) offer specialized programs, majority of universities
Development Bank.
do not offer Islamic finance programs. Underlying reasons
for shortage should be further investigated. Some potential Al-Omar, F. & Abdel-Haq, M. (1996) Islamic Banking: Theory,
reasons might be lack of enough syllabuses, or even lack Practice & Challenges, Karachi, Oxford University Press.
of demand for such programs. However, universities may
Albuolayan, A. (2006) Rapid Surge in Sukuk Market.
collaborate and share their resources to develop “market
ArabNews, Jeddah, Saudi Arabia. Jeddah, Saudi Arabia,
intelligence” database. Such database can be used to
Saudi Research & Publishing Co.
identify the potential programs, courses and syllabus. For
instance, especial programs may be developed for training Alsayyed, N.A. & Malik, F. (2010) Sukukization: Islamic
of Shariah board members. Economic Risk Factors in Shariah View. Working Paper
of International Shari’ah Research Academy for Islamic
Finance (ISRA). Kuala Lumpur, Malaysia, International
Notes Shari’ah Research Academy for Islamic Finance (ISRA).
1 During the 3rd century AD, financial firms in Persia
Alvi, I.A., Mohammed, A.R., Khan, G.Z., Nasser, U.M., Naseer,
(currently known as Iran) and other territories in
B. & Khan, M.S. (2010) Sukuk Report, A Comprehensive
the Persian Sassanid Dynasty issued letters of credit
Study of The International Sukuk Market. In Alvi, I.A. (Ed.
known as “chak.” KHARAZMI, A. A.-A. M. B. A. (1895)
International Islamic financial market. 1 ed. Manama,
Maf t h al-Ool m. IN VLOTEN, G. V. (Ed. Leiden. In post-
Bahrain, International Islamic financial market.
Islamic Arabic documents this word has transformed
into “sakk” FLOOR, W. (1990) AK (legal document, Alvi, I.A., Mohammed, A.R. & Naseer, U.M. (2011) Sukuk
testament, money draft, check). IN YARSHATER, E. Report, A Comprehensive Study of The International
(Ed. Encyclopedia Iranica Online Edition. New York., Sukuk Market. In Alvi, I. A. (Ed. International Islamic
Columbia University. Financial Market. 2 ed. Manama, Bahrain, IIFM.
2 Data is collected from Bondstream database, a product
Amin, M. (2011) Accounting for sukuk under IFRS and
of Bond Pricing Agency Malaysia. Sample is limited
AAOIFI accounting standards. In Jaffer, S. (Ed.) Global
to Malaysia due to unavailability of parallel markets
Growth, Opportunities and Challenges in the Sukuk
for sukuk and bonds in other countries. Moreover,
Market. Euromoney Trading Inc.
Malaysia accommodate almost two third of world
sukuk issues. Asad, S.I.-U.-D. (2009a) The business of shariah advice The
3 International Shariah Research Academy for Islamic DAWN newspaper, Karachi, Pakistan. Karachi, Pakistan,
Finance, based in Kuala Lumpur, Malaysia. Pakistan Herald Publication (Pvt) Limited.
4 International Centre for Education in Islamic Finance,
Asad, S.I.-U.-D. (2009b) An overview of the sukuk market.
based in Kuala Lumpur, Malaysia.
Jang Newspaper. Karachi, Pakistan, Jang Group of
5 Besides the ISRA and INCEIF which offer long term
Newspapers.
programs, we can mention, among all, Dr. Kahf
Institute, REDmoney Islamic Finance training institute, Aseambankers (2005) Capitalising on Opportunities in the
and UK Islamic Finance Council. Sukuk Industry. Aseambankers Malaysia Berhad. Kuala
6 http://www.aaoifi.com/aaoifi/TheOrganization/Over­ Lumpur, Aseambankers.
view/tabid/62/language/en-US/Default.aspx accessed
Bose, S., SR. & McGee, R.W. (2008) Islamic Investment
on 9/4/2013.
Funds: An Analysis of Risks and Returns. SSRN eLibrary.
7 For complete list of such incentives refer to the website
of Security Commission of Malaysia http://www. Chapra, M.U. (1998) The Major Modes of Islamic Finance. 6th
sc.com.my/main.asp?pageid=593&menuid=322&ne Intensive Orientation Course on “Islamic Economics, Banking
wsid=&linkid=&type=(Accessed on 11/4/2013) and Finance.” the Islamic Foundation, Leicester, U.K.

Eds. Hatem A. El-Karanshawy et al. 15

 Ariff and Safari

Daryanani, N. (2008) A Deeper Understanding on Iqbal, Z. (1999) Financial engineering in Islamic finance.
the Prohibition of Riba. Management. Nottingham, Thunderbird International Business Review, 41, 541–559.
University of Nottingham, UK.
Jalil, A. (2005) Islamic Bonds Issues: The Malaysian
DIFC (2009) Dubai International Financial Centre Sukuk Experience. In T. Ramanayanah, Mohamed Sulaiman,
Guidebook, Dubai, UAE. Hasnah Harun, Ruhaini Ali, Aizzat Mohd. Nasurdin,
Nabsiah Abdul Wahid & Intan Osman (Eds.) The
El-Gamal, M.A. (2000) A Basic Guide to Contemporary
6th Asian Academy of Management Conference, 9–11
Islamic Banking and Finance. Plainfield, Indiana, Islamic
December 2005. Casuarina Ipoh, Perak, Malaysia, Asian
Society of North America.
Academy of Management.
El-Gamal, M.A. (2007) Mutuality as an antidote to rent-
John, I. (2007) Sukuks key to meeting $1.5tr infrastructure
seeking Shariah arbitrage in Islamic finance. Thunderbird
needs of ME, Asia. Khaleej Times. Dubai, Galadari
International Business Review, 49, 187–202.
Printing and Publishing.
Enders, W. (1995) Applied Econometrics Time Series,
Kamali, M.H. (2007) A Shari’ah Analysis of Issues in Islamic
New York, John. Wiley & Sons.
Leasing. J.KAU: Islamic Econ, 20, 3–22.
Farook, R. (2009) Global Financial Crisis Unthinkable
Khan, M.M. & Bhatti, M. I. (2008) Development in Islamic
Under Islamic Banking Principles. Sunday Observer. Sri
banking: a financial risk-allocation approach. The
Lanka, The Associated Newspapers of Ceylon Ltd.
Journal of Risk Finance, 9, 40–51.
Floor, W. (1990) AK (legal document, testament, money
Khan, T. & Ahmed, H. (2001) Risk Management: An Analysis
draft, check). In Yarshater, E. (Ed. Encyclopedia Iranica
of Issues in the Islamic Financial Industry, Jeddah, Saudi
Online Edition. New York., Columbia University.
Arabia, Islamic Development Bank – Islamic Research
Gait, A.H. & Worthington, A.C. (2007) A Primer on Islamic and Training Institute.
Finance: Definitions, Sources, Principles and Methods.
Kharazmi, A.A.-A.M.B.A. (1895) Maf t h al-Ool m. In
Working Papers Series of University of Wollongong.
Vloten, G.V. (Ed.) Leiden.
Wollongong, Australia, University of Wollongong.
Lewis, M.K. (2007) Islamic Banking in Theory and Practice.
Gambling, T., Jones, R. & Karim, R.A.A. (1993) Credible
Monash Business Review, 3, 1–8.
Organizations: Self-Regulation V. External Standard-
Setting in Islamic Banks and British Charities. Financial Lewis, M.K. & Algaoud, L.M. (2001) Islamic banking,
Accountability & Management, 9, 195–207. Cheltenham, UK, Edward Elgar.
Granger, C.W.J. (1969) Investigating Causal Relations Liquidity Management Center (2008) The Guide to Sukuk
by Econometric Models and Cross-spectral Methods. Market. Bahrain, November 2008.
Econometrica, 37, 424–438.
Maurer, B. (2002) Anthropological and Accounting
Gurgey, U. & Keki, E. (2008) Sukuk in Turkey. International Knowledge in Islamic Banking and Finance:
Financial Law Review, 27, 111–111. Rethinking Critical Accounts. The Journal of the Royal
Anthropological Institute, 8, 645–667.
Hasan, Z. (2010) Islamic Finance: What Does It Change,
What It Does Not – The Structure – Objectives Mismatch Maurer, B. (2010) Form versus substance: AAOIFI projects
and Its Consequences. Working Paper. Kuala Lumpur, and Islamic fundamentals in the case of sukuk. Journal
Malaysia, International Centre for Education in Islamic of Islamic Accounting and Business Research, 1, 32–41.
Finance (INCEIF).
McKenzie, D. (2008) Islamic Finance 2008. International
Hossain, A. (2005) Granger-Causality Between Inflation, Financial Services London, Islamic Finance Working
Money Growth, Currency Devaluation and Economic Group, 8.
Growth in Indonesia, 1951–2002. International Journal
McKenzie, D. (2009) Islamic Finance 2009. International
of Applied Econometrics and Quantitative Studies, 2, 23.
Financial Services London (IFSL), 8.
IFSB (2009) Capital Adequacy Requirements for Sukuk,
McKenzie, D. (2010) Islamic Finance 2010. International
Securitizations and Real Estate Investment. In Board,
Financial Services London (IFSL), 8.
I.F.S. (Ed. IFSB. Kuala Lumpur, Malaysia, Islamic
Financial Services Board. Metwally, M.M. (2006) Economic Consequences of
Applying Islamic Principles in Muslim Societies. Journal
IFSL (2012) Islamic Finance. IN Mckenzie, D. (Ed. Finaicial
of Islamic Banking and Finance, 23, 11–33.
Market Series. London, International Financial Services
London. Mohd Asri, N. (2004) The Effect of Islamic Private Debt
Securities Rating Changes on Firm’s Common Stock
Iqbal, M. (1998) Islamic banking. In Kahf, M. (Ed.) Lessons
Returns. The Journal of Muamalat and Islamic Finance
in Islamic Economics. Jeddah, Islamic Research and
Research, 1, 25–38.
Training Institute.
Pomeranz, F. (1997) The Accounting and Auditing
Iqbal, M. & Molyneux, P. (2005) Thirty years of Islamic
Organization for Islamic Financial Institutions: An
banking: History, Performance and Prospects,
Important Regulatory Debut. International Journal of
Houndmills, Basingstoke, Hampshire; New York,
Accounting, Auditing and Taxation, 6, 123–30.
Palgrave Macmillan.

16 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development

Rosly, S.A. (2007) Islamic Capital Market: Shariah Stocks, Usmani, M.T. (2007) Sukuk and their Contemporary
Sukuk, i-Reits, Islamic unit trust funds. LOFSA-INCEIF Applications. Accounting and Auditing Organization for
Workshop on Islamic Finance, 14th-15thNovember 2007, Islamic Financial Institutions (AAOIFI), 14.
Labuan, Malaysia. Labuan, Malaysia, International
Warde, I. (2000) Islamic Finance in the Global Economy,
Center For Education in Islamic Finance (INCEIF).
Edinburgh Edinburgh University Press.
Shinsuke, N. (2007) Beyond the Theoretical Dichotomy in
Wilson, R. (2004) Overview of the sukuk market. In Adam,
Islamic Finance: Analytical Reflections on Mur ba ah
N.J. & Thomas, A. (Eds.) Islamic Bonds: Your Guide to
Contracts and Islamic Debt Securities. Kyoto Bulletin of
Issuing, Structuring and Investing in Sukuk. University
Islamic Area Studies,, 1, 72–91.
of Durham, United Kingdom, Euromoney Institutional
Siddiqui, R. (2008) Contributing to the Development of the Investor PLC.
Islamic Capital Market: The Dow Jones Citigroup® Sukuk
Wilson, R. (2008) Innovation in the structuring of Islamic
Index. Dow Jones Islamic Market Indexes NewsLetter,
sukuk securities. Humanomics, 24, 170–181.
December, 1–2.
Zaher, T.S. & Hassan, M.K. (2001) A Comparative Literature
Sundararajan, V. & Errico, L. (2002) Islamic financial
Survey of Islamic Finance and Banking. Financial
institutions and products in the global financial system:
Markets, Institutions & Instruments, 10, 155–199.
Key issues in risk management and challenges ahead. Risk
Management in an Islamic Financial System, September 1,
2002, Tehran, Iran. Tehran, Iran Iran Banking Institute, Acknowledgment
Central Bank of the Islamic Republic of Iran. An earlier version of this paper was presented at the 23rd
Global Finance Conference, Chicago, USA in May 2012
Tariq, A.A. (2004) Managing Financial Risks of Sukuk
at De Paul University. Authors would like to appreciate
Structures. School of Business and Economics.
the feedbacks received from reviewers and participants of
Leicestershire, UK, Loughborough University.
that conference. This research was funded by Khazanah
Usmani, M.T. (2001) Principles Of Shari’ah Governing Nasional Bhd., Malaysia, and the Maybank Endowed
Islamic Investment Funds. chair professorship at the University Putra Malaysia. The
remaining errors are solely those of the authors.
Usmani, M.T. (2002) An Introduction to Islamic Finance,
The Hague, Kluwer Law International.

Eds. Hatem A. El-Karanshawy et al. 17

The impact of Islamic debt on company value
Fitriya Fauzi1, Stuart Locke2, Abdul Basyith, Muhammad Idris
Faculty of Economics, University of Muhammadiyah, Palembang, Indonesia, Email: [email protected]
1

Waikato Management School, The University Waikato

2

Abstract - This study uses micro-econometric analysis to examine the impact of Islamic debt on
firm value and firm financial performance by observing Malaysian firms. A number of significant
contributions to corporate finance arise from this research in relation to Islamic debt instruments
and firm financial performance. First, it provides evidence of the Islamic debt impact on firm value
and firm financial performance. Second, and very importantly it provides new insights, adding
substantially to the very few studies that have been conducted on these types of instruments.

The choice of model employed is specified according to its diagnostic testing results for non-
normality, heteroskedasticity, multicollinearity, endogeneity and linearity in. A test is conducted
to confirm that there are no outliers in the data set prior to the diagnostic testing. Poolability and
co-integration testing are also included. Based on the diagnostic results, data are analysed using
the dynamic panel generalised method of moment (GMM using a quarterly balanced panel of 80
Malaysian firms issuing Islamic debt which spans from 2000 to 2009. This method is employed to
investigate the impact of Islamic debt issues on firm value and/or firm financial performance.

The result reveals that Islamic debt has a significant positive impact on company value and firm
financial performance. It also confirms that trade-off theory holds well in the Malaysian context for
Islamic debt financing. Furthermore, the coefficient for Islamic debt is higher than the coefficient
for non-Islamic debt, suggesting that the Islamic debt provides a higher contribution to firm value
and to the improvement of firms’ financial performance compared to non-Islamic debt.

Keywords: Islamic debt, firm value

1. Introduction it is widely accepted that the world is not made up of perfect

markets, the issue of optimal D/E ratio remains contentious
Modes of financing are a part of capital structure, and the
and unresolved.
relationship between capital structure and firm value is
important and continues to be debated in the literature.
Moreover, there are considerable studies that have
An unleveraged firm can be seen as an all equity firm,
investigated the impact of conventional debt on firm
whereas a leveraged firm is made up of ownership equity
value, and those studies have generated few theories
and debt. A firm’s debt to equity ratio provides a measure
which are implemented up to now. However, so far there
of the leverage or gearing. The influence of debt to equity
is no investigation of the financial performance and value
on the value of a firm is the subject of multiple articles,
consequences of using Islamic debt as opposed to non-
including the seminal work of Modigliani and Miller
Islamic debt (conventional debt). The lack of prior research
(M&M). Modigliani and Miller achieved notoriety with
into the financial aspect of Islamic debt does not reflect a
their proof that capital structure made no different to firm
lack of concern for such matters, but rather is a reflection
value. Subsequent relaxation of the M&M assumptions
of the newness of the topic. Therefore, the extent to which
suggested that debt to equity choice does impact on firm
prior research is applicable to Islamic debt, particularly in
value. More recent research considering bankruptcy costs
emerging market, requires analysis, and this study is useful
has contributed to empirical estimates of optimal leverage.
to supplement existing studies in this field and serves as a
The M&M argument is that, in a perfect market, how a firm
reference for studies in the future.
is financed is irrelevant to its value. However, even though

Cite this chapter as: Fauzi F, Locke S, Basyith A, Idris M (2015). The impact of Islamic debt on company value. In
H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Fauzi et al.

The literature relating to Islamic debts has predominantly riba, known as adding any interest payments to a loan or
focused on the legal aspects of Islamic law, concept, other financing contract. The second is the prohibition from
basic requirements and the validity of how the debts are gharar and maisir, known as uncertainty and gambling; so
conducted in Islamic finance as general Islamic debt (Cakir, transactions embodying these attributes will be considered
2007; Mirakhor, 1996; Ashhari, 2009; Somolo, 2009; invalid. The third is the prohibition of non-halal business
Tariq, 2007; Wilson, 2008). So far, researchers have been transactions, such as alcohol, gambling and any other
unable to find research that looks at the effects of Islamic things that are prohibited and considered as non-halal. The
debts on the value of the company in international contexts. fourth is the general prohibition of contracts that fail to
Haneef (2009) discusses the history of Sukuk, explaining meet the highest Shariah standards (Ayub, 2007).
how it has evolved from an asset backed structure, where
Sukuk holders have ownership rights over the underlying The development of Sukuk is supported by many factors,
asset, to an asset based structure, where Sukuk holders including the development of Islamic banking (takaful)
rank paripassu with unsecured creditors. Other scholars and an increasing demand for Islamic products in the debt
(Abd.Sukor, 2008; Al-Amine, 2001; Juan, 2008; Kamali, market. The development of Sukuk with its associated
2007; Mohd Yatim, 2009; Mokhtar, 2009; Al-Amine, n.d.; types of structure has given rise to much discussion and
Al Amine, 2008; Usmani, 1999, n.d.; Vishwanath, 2009; debate among scholars of Islamic law. The uniqueness of
Wilson, 2008, n.d.; Yean, n.d.) also discuss the structure Islamic debt compared to non-Islamic debt is that Islamic
and the regulation of the Sukuk market in relation to debt offers a secure investment based on the principle of
Shariah perspective and Shariah compliancy. Therefore, rent and profit sharing without legalised interest system.
this study attempts to examine the impact of Islamic debt It is constituted by pure motive of cooperation based on
on company value. Islamic law. How the market prices this security in term
of the yield curve and how risk pricing is embedded in the
The rest of this paper is organised as follows. Section two value and performance of the firms.
presents the significance emergence of Islamic debt in the
fast growing form of financing in emerging and mature Recent innovations in Islamic finance have changed the
markets. Section three provides literature review followed dynamics of the Islamic finance industry, especially in
by section four, five and six which present the methodology, the debt markets. Sukuk became increasingly popular
analysis and conclusion. as companies sought to raise funds by offering corporate
Sukuk. It has become significant for raising funds in the
international capital markets through Islamic Shariah.
2.  Significant emergence of Islamic debt Increases in this market have been strong all over the
The Islamic financial and economic system has existed since world, especially in Malaysia, UAE and Saudi Arabia. In
the time of the prophet Muhammad SAW. During that time, 1996 total Sukuk issued was USD 0.05b rising to USD
buying and selling, and savings and loans activities were 15.5b by the end of 2008. The most significant increase
not as extensive as they are now. However, the principle occurred in 2007 with more than 130 issues valued at
remains the same; no interest charged and no non halal USD 34.3b. The trend is apparent in Table 1 which shows
products and activities permitted. The interest system is a rapid expansion by value through to the financial
not used at all because it is forbidden by Allah SWT. The crisis in 2008. Malaysia accounts for 43.7% of Sukuk
banning was declared in the Quran and the Hadith. issues followed by UAE with 30.1% and Saudi Arabia
representing 10.4%. The size of offering by country for
Islamic debt, known as Sukuk, has evolved to become a 2009 is shown in Table 2.
significant part of corporate capital trading in the secondary
market. The Accounting and Auditing Organization of In the early years of Sukuk’s emergence as a financial
Islamic Financial Institutions (AAOIFI) has also defined instrument, murabahah and istisna were the most
Islamic debt as certificates of equal value representing significant forms of issuance, accounting for 62.5% and
undivided shares in the ownership of tangible assets, 19.5% respectively. This changed between 2002 and 2007
usufruct and services or (in the ownership of) assets of when musyarakah and ijarah become the largest type of
the particular projects or any specified investment activity. issue, accounting for 36.3% and 28.3% of the total market.
Investment of Sukuk should be distinguished from common In 2008 to 2009 the ranking reversed with ijarah and
shares and bonds. While shares represent the ownership musyarakah accounting for 43.4% and 20.8% respectively
of a company as a whole and are for an indefinite period, as reflected in Table 3.
Sukuk represent specified assets and are for a given period
of time. Sukuk, unlike bonds, carry returns based on cash The increase in the issue size from year to year indicates that
flow originating from the assets on the basis of which they this market was gradually developing. It became lucrative
are issued (Ayub, 2007; p. 392). for both the Sukuk issuer and the Sukuk holder, receiving
increased support in the form of market surveillance and
Islamic debt includes no periodic interest payments and regulation, and from market participants.
provides a different cash flow profile when compared with
non-Islamic debt instruments for borrowing companies The evolution of Sukuk structure is presented in Figure 1.
and lenders. There is a socio-religious dimension relating The evolving of the original structure is due to the
to major principles that underlie all business transactions needs of this market for its product development. At the
under Islamic law. All business transactions must adhere beginning of the emergence of Sukuk is debt based Sukuk.
the teaching of the Islamic foundation, which is the Quran Murabahah Sukuk is one of the debt-based forms. The
and Sunnah. There are at least four major prohibitions in second stage of the evolution is asset-based Sukuk. One of
Islamic business transactions. The first is the prohibition of the forms of this structure is Ijarah Sukuk. The last stage

20 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

Table 1. Global Sukuk issuance by year. 200

180 Value USD billion
Year Value USD billion Number
160 Number
1996 0.05   1
140
1997  0.9   1
1998 – – 120
1999  0.2   4 100
2001  1.6  16
80
2002  2.9  23
2003  4.2  32 60
2004  3.5  50 40
2005  7.8  96
20
2006 19.5  99
2007 34.3 130 0
2008 15.5 174

96
97

98
99
01
02

03
04
05
06
07
08
19
19

19
19
20
20

20
20
20
20
20
20
Source: ZawyaSukuk Monitor, 2009

50
Table 2. Global Sukuk issuance by country in 2009.
45 Value USD Billion
Country Value USD Billion Value in%
40 Value in %
Malaysia 31.5  43.67 35
Bahrain  5.2   7.21
30
Indonesia  0.3   0.42
UAE 21.7  30.08 25
Pakistan  1.5   2.08 20
Brunei Darussalam  0.7   0.97
15
Kuwait  1.8   2.50
Saudi Arabia  7.5  10.40 10
Qatar  1.3   1.80 5
UK  0.2   0.28
0
Sudan  0.13   0.18
USA  0.16   0.22
Ba ia
In ain
sia

ei Pak E
us n
m

a
ar

n
G SA

y
i A it
bi

an
ar ta

da
ud wa
UA
s

U
la

at
ay

ne
hr

D is

ra

U
m
Su
sa

Q
Sa u

Germany  0.14   0.19

al

do

er
M

Total 72.13 100

un
Br

Source: ZawyaSukuk Monitor, 2009

of the evolution is equity-based Sukuk or partnership- regardless of capital structure. On the other hand Jensen
based Sukuk. The forms of this structure are musyarakah and Meckling (1976) states that the amount of leverage in a
and istisna’. firm’s capital structure is associated with its performance.

Furthermore, several researchers have conducted numerous

3.  Literature review studies which aim to examine the relationship between
Capital structure is widely discussed in the finance literature. capital structure and firm performance. However, until now
The mixture of debt to equity in the financial structure the evidence regarding this study is contradictory and mixed.
of companies and whether it will impact upon financial (Ebaid, 2009; Ni & Yu, 2008; Phillips, 2004) find consistent
performance risk and valuation is the subject of theoretical results with Modigliani and Miller (M&M) theorem. On the
and empirical studies. In general, capital structure theories other hand, (Abor, 2007; Bhabra, Liu & Tirtiroglu., 2008)
are classified into three categories; first, the zero impact find inconsistent results with M&M theorem. The focus on
hypotheses or the Modigliani and Miller theory; second, country effect, for example, developed and emerging markets
the positive impact hypotheses or trade-off theory; third, has been done in some of the studies (Bhabra et al., 2008;
the negative impact hypotheses or pecking order theory. Ebaid, 2009). Other studies document a focus on firm size,
such as large, medium or small companies (Abor, 2007).
Modigliani and Miller theory The Modigliani and Miller theory (1958) assumes that
Modigliani and Miller (1958) argue that capital structure a capital market is perfect (no transaction or bankruptcy
is irrelevant, thus the total cash flows a company makes for costs, perfect information), individuals and corporations
all investors (debt holders and shareholders) are the same can borrow at the same rate, and no taxes. It does not

Eds. Hatem A. El-Karanshawy et al. 21

 Fauzi et al.

Table 3. Global Sukuk issuance by structure type. structure almost entirely composed of debt. But in the real
world, firms cannot stand only with debt or a hundred
Type of Value USD Value percent leverage because an increase in debt will increase
Year structures billion in% bankruptcy cost and agency cost. Consequently, it means
that no optimal capital structure exists.
Phase I Murabahah  1.6   62.5
(1996–2001) Al Salaam  0.16    6.3
  Istisna  0.5    19.5 Trade-off theory
  Ijarah  0.25    9.8 After the seminal work of M&M, little research has
  Mudarabah  0.05    2.0 been done to explore capital structure in which some
  Musyarakah – – assumptions were proposed; trade off theory and pecking
  Al Istithmar – – order theory. The trade-off theory derived from the models
  Hybrid – – based on taxes and agency cost. Modigliani and Miller
  Other – – (1963), DeAngelo and Masulis (1980) and Jensen and
Meckling (1976) suggest the firm has an optimal capital
  Total   2.56 100.0 structure by offsetting the advantages of debt and the cost
of debt. Therefore, trade off theory refers to the idea that
Phase II Murabahah  4.9    6.8 a company chooses how much debt finance and how much
(2002–2007) Al Salaam  1.9     2.6 equity finance to use by balancing the costs and benefits. It
Istisnaa  4.1    5.7 states that there is an advantage to financing with debt, the
Ijarah 20.5   28.3 tax benefits of debt, and tax benefits to be had, but there
Mudarabah  8   11.0 is also a cost to financing with debt, the costs of financial
Musyarakah 26.3   36.3 distress including bankruptcy costs, and agency costs.
Al Istithmar  2.9    4.0 This theory suggests that there is a positive relationship
Hybrid  2.8    3.9 between debt level and firm performance. Moreover, the
Other  1    1.4 implication of this trade off theory is that firms have target
leverage and they adjust their leverage toward the target
Total 72.4 100.0 over time. In addition, Harris and Raviv (1990) imply
that higher leverage can be expected to be associated with
Phase III Murabahah  4   12.8 larger firm value, higher debt level relative to expected
(2008–2009) Al Salaam  0.05    0.2 income, and lower probability of reorganization following
Istisnaa  0.08    0.3 default.
Ijarah 13.6   43.4
Mudarabah  2.5    8.0 The empirical relevance of the trade-off theory has often
Musyarakah  6.5   20.8 been questioned. Some research has been conducted
Al Istithmar  3.5   11.2 to investigate this theory and the results from various
Hybrid  0.075    0.2 contexts are mixed and inconclusive. The evidence does
Al Wakalah  1    3.2 indicate there are likely to be differences attributable
to firm size, country and the maturity of the respective
Total 31.305 100.0 capital market.

Source: ZawyaSukuk Monitor, 2009.

Pecking order theory
Pecking order theory was developed by Myers and
matter if the firm’s capital is raised by issuing stock or debt, Majluf (1984). Myers and Majluf (1984) consider firms
or what the firm’s dividend policy is. Therefore, the M&M must issue common stock to raise cash to undertake a
theory concludes that capital structure is irrelevant. valuable investment opportunity. Management is assumed
to know more about the firm’s value than potential
Modigliani and Miller made two propositions under these investors. Investors interpret the firm’s actions rationally.
conditions. Their first proposition was that the value of a
leveraged firm is the same as the value of an unleveraged firm.
Their second proposition was that the expected return on The evolution of Sukuk
equity is positively related to leverage because the risk to equity
holders increases with leverage. These two propositions stand
on the assumption that taxes are ignored, and bankruptcy cost 1. Debt based Sukuk
and other agency costs were not considered.

When taxes were taken into account, they had two 2. Asset based Sukuk
propositions as well. First, they state that the value of the
firm is positively related to leverage. It means that corporate 3. Equity based Sukuk
leverage lowers tax payments because corporations can
deduct interest payments but not dividend payments.
Secondly, they state that the cost of equity rises with 4. Hybrid/mixed Sukuk
leverage because the risk to equity rises with leverage.
These propositions assume that firms have a capital Figure 1.  The evolution of Sukuk structure.

22 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

An equilibrium mode1 of the issue-investment decision is 4. Methodology

developed under these assumptions. The model shows that
firms may refuse to issue stock, and therefore may pass up Data
valuable investment opportunities. The model suggests
The data for this study were obtained from the Islamic
explanations for several aspects of corporate financing
Finance Information Service (IFIS) database. The sampling
behaviour, including the tendency to rely on internal
period is 2000 to 2009, which is ten years and, this study
sources of funds, and to prefer debt to equity if external
used quarterly data. This quarterly data is important since
financing is required.
the issuance of Islamic debt for every firm is in different
quarters. Initially, this study proposed to investigate the
In addition, Frank and Goyal (2003) state that capital
debt choice impact on a company value and firm’s financial
structure is acquired in accordance with the priority of the
performance using Malaysian firms as a sample. Further,
firm in which internal funding is preferable and external
this study notes that 227 companies from Malaysia issued
funding is less preferable. If it is needed, firms could use
Islamic debt from 2000 to 2009. From those 227 Malaysian
external funding from the lowest risk debt. Therefore,
companies, 106 companies are public companies, and 121
pecking order theory refers to the idea that companies prefer
companies are limited companies. From the 106 public
to use their sources of financing from internal financing
companies, 31 companies have been excluded from the
to equity. If external financing is required, firms issue the
data list because of the unavailability of their financial
safest security first. That is, they start with secure debt, then
statement data. In addition, the sample of Islamic debt
perhaps equity as a last choice. In addition, issue costs are
offering must have data availability on the size of the
least for internal funds, low for debt and highest for equity.
offering, the maturity length, the history of the issuance,
There is also the negative signaling to the stock market
and other accounting data information.
associated with issuing equity, positive signaling associated
with debt, and asymmetric information between managers
For panel data analysis, the availability of ten years’ worth
and investors. This theory suggests that there is a negative
of data is required, particularly quarterly data. To mitigate
relationship between debt level and firm performance.
the problem of missing values, this study uses multiple
Therefore, the implication of this pecking order theory is
imputations by including the weighted value to compensate
that firms prefer to depend on internal sources of funds and
the missing value excluded in the model (Raghunathan,
prefer debt to equity if external financing is required. Thus,
2004).
a firm’s leverage is not driven by the trade-off theory, but
rather by results of the firm’s attempts to mitigate signalling
effect and information asymmetry.
Variables
The majority of studies have been conducted in mature Firm value and firm financial performance are dependent
markets with some based on developing markets including variables. Each of the performance indicators measures a
Asia, Africa and the Middle East. Previous studies also different aspect of performance. Tobin’s Q is used as a firm
consider firm size, investigating whether large corporations value indicator,and it is considered as a market reflection
and small and medium sized corporations behave differently. of the firm’s activities and performances. Return on Asset
The results are mixed and inconclusive. In addition, previous (ROA) and Return on Equity (ROE) metrics are used as a
studies examined different institutional structure (Booth, firm financial performance indicator. ROE measures the
Aivazian, Demirguc-Kunt & Maksimovic., 2001); different performance from the perspective of the equity holders;
governance mechanisms (Wiwattanakantang, 1999); meanwhile ROA measures the asset productivity and
different market power and firms investment (Eriotis, operating profit margin. It is important to note that none
Frangouli & Neokosmides., 2002); different regional of these measures truly reflect the complete picture by
risk (Zeitun & Tian, 2007); different firm characteristics, themselves but have to be seen in conjunction with other
ownership structure and industry membership (Bhabra metrics.
et al., 2008). In comparison with the abundance of studies
on the relationship between capital structure and firm The proportion of the Islamic debt to non-Islamic debt,
performance and the determinant factors of capital structure the proportion of Islamic debt, the frequency of Islamic
in conventional debt, only a few studies focus on Islamic debt issuance and the type of Islamic debt issued are used
debt. This study will provide evidence about how Sukuk as independent variables. The proportion of Islamic debt
impacts upon financial performance and corporate value in is calculated as the total Islamic debt divided by the total
markets. assets/or total Islamic debt divided by the total Islamic
debt plus total equity. The proportion of non-Islamic debt
Furthermore, the study of leverage impact on firm value is calculated as the total of non-Islamic debt divided by the
has become an important aspect of capital structure theory. total assets/or total non-Islamic debt divided by the total
Myers (1984) claims that the firm value depends on the debt non-Islamic debt plus total equity.
ratio, similarly, many studies have focused on the impact
of the debt level and the debt type on a firm’s financial For the proportion of Islamic debt, the frequency of Islamic
performance (Ebaid, 2009; Ghosh & Cai, 1999; Hatfield, debt issued and the type of Islamic debt, dummy variables
Cheng & Davidson, 1994; Coleman, 2007; Talberg, Winge, are employed. To avoid too many parameters and to find
Frydenberg & Westgaard, 2008). However, few recent the unique least square estimates for the model, this study
studies investigate the impact of the Islamic debt type on a uses only n-1 dummy; therefore, the baseline is chosen
firm’s value and financial performance. This study attempts for every set of the specifications. In addition, n-1 dummy
to investigate the impact of Islamic debt on company value may mitigate the problem of multicollienarity among the
and a firm’s financial performance. regressors (Baltagi, 2005). The choice of a baseline category

Eds. Hatem A. El-Karanshawy et al. 23

 Fauzi et al.

is essentially arbitrary, for this study fits precisely with all Akhtar, 2005; Zeitun & Tian, 2007; Talberg et al., 2008).
regression models regardless of which category is selected The natural logarithm is applied for the firm size variable
for this role. The value and meaning of the individual owing to the skewness and kurtosis problem. Further,
dummy-variable coefficients δ1, δ2, ζ1, ζ1, η1 and η2 depend, natural logarithm ensures that the actual regressor has less
however, on which category is chosen as the baseline. statistical noise in the regression model, and moderates the
effects of the large size of the firm.
The first group dummy is aimed at examining the effect of
the Islamic debt proportion on each company, and three
categories are set: first, a company having an Islamic Model specification
debt proportion below the average of the Islamic debt This study uses the panel data method which allows
proportion; second, a company having an average Islamic the unobservable heterogeneity for each observation
debt proportion; and third, a company having an Islamic in the sample to be eliminated and multicollinearity
debt proportion higher than the average of the Islamic debt among variables to be alleviated. Unobservable
proportion. The first category will be set as “1” if companies heterogeneity might result in spurious correlations with
have a below average Islamic debt proportion; otherwise the dependent variables, which would bias the coefficient
it is set equal to “0”. The second category will be set as “1” obtained (Baltagi, 2005). Before proceeding to the
if companies have an average Islamic debt proportion; model specification, diagnostic testing of normality,
otherwise it is set equal to “0”. The third category will be set heteroskedasticity, multicollinearity, and autocorrelation,
as “0” if companies have a higher than average Islamic debt was conducted to determine the appropriate method used
proportion. The baseline category is used for this dummy in this study. The specification testing results are provided
if the company has a higher than average proportion in Tables 4 and 5.
Islamic debt. The average of the Islamic debt proportion is
8.06%, which is calculated by the total of the Islamic debt The heteroskedasticity result is 462.99 with p-value 0.0000,
proportion over the total number of companies. suggesting that there is a heteroskedaticity problem.
Therefore, this problem needs to be catered to obtain
The second group dummy is aimed at examining the effect efficient and unbiased results. The skewness and kurtosis
of the Islamic debt issuance frequency on each company, results are 45.72 with p-value 0.000 and 4.25 with p-value
and three categories are set: first, a company issuing an 0.0392, suggesting that non-normal distribution, thus
Islamic debt only once; second, a company issuing an this non-normal distribution has to be treated. Therefore,
Islamic debt for the second time; and third, a company outliers’ checking is conducted prior to data analysing.
issuing an Islamic debt more than twice. The first category
will be set as “1” if companies issue an Islamic debt only The multicollinearity result is 10.9100 with p-value 0.000,
once; otherwise it is set equal to “0”. The second category suggesting no multicollinearity problem among the
will be set as “1” if companies issue an Islamic debt for explanatory variables. Before proceeding to the endogeneity
the second time; otherwise it is set equal to “0”. The third test and linearity test results, a brief conclusion made is that
category will be set as “0” if companies issue an Islamic heteroskedasticity and non-normality problems have to be
debt more than twice. The baseline category is used for this
dummy if the company has more than twice of the Islamic
debt issuance.

The third group dummy is aimed at examining the effect of Table 4. Summary of the specification testing results.
the Islamic debt type on each company, and three categories
are set: first, a company issuing a debt-based type of Islamic Tests p-value
debt; second, a company issuing an asset-based type of
Heteroskedasticity 462.99* 0.0000
Islamic debt; and third, a company issuing an equity-based
type of Islamic debt. The first category will be set as “1” if Skewness  45.72* 0.0000
companies issue a debt-based type; otherwise it is set equal Kurtosis   4.25* 0.0392
to “0”. The second category will be set as “1” if companies Multicollinearity  10.91* 0.0000
issue an asset-based type; otherwise it is set equal to “0”. Linearity   2.9055* 0.0000
The third category will be set as “0” if companies issue an Endogeneity Endogeneity exist
equity-based type. The baseline category is used for this
dummy if the company issues an equity-based type of the *Sig. at 1% significance level.
Islamic debt.

Prior research suggests that the performance of each firm

may differ according to their size, because larger firms Table 5. The DWH test for endogeneity of regressors.
have greater economies of scale in the transaction costs
associated with long term debt, which may influence the Variables Tobin’s Q ROA ROE
results and inferences (Ramaswamy, 2001; Frank & Goyal,
2003; Coleman, 2007; Jermias, 2008; Ebaid, 2009). In Islamic Debt 0.0820 6.9848* 0.0035
addition, larger firms have less potential of bankruptcy Proportion (0.7745) (0.0083) (0.9526)
cost; therefore, firm size should be positively related to the Non-Islamic 37.4723* 0.0036 14.2038*
borrowing capacity (Krishnan & Moyer, 1997). This study Debt Proportion (0.0000) (0.9523) (0.0002)
uses a natural logarithm of the total assets as a proxy for
firm size as the control variable (Naceur & Goaied, 2002; *Sig. at 1% significance level.

24 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

treated. The linearity test result for group 1 is 2.905506 Taking into account the first difference, one can elide the
with p-value 0.0000, which rejects the null hypothesis of unobserved firm-specific effect:
nonlinearity. Similar to the result for group 1, the linearity
test result for group 2 is -4.659 with p-value 0.000, which y it - y it-1 = a ( y it-1 - y it-2 ) + b ′( X it - X it-1 ) + (e it - e it-1 ) (6)
supports the linear model. The endogeneity test result  
reveals that the regressors in the model present endogeneity.
where X includes lag performance for explanatory
Supported by numerous previous studies, by the varia­bles, yit−1 as well as dependent variables. The first-
assumptions above, by the endogeneity tests and by differencing eliminates potential bias that arises from
the linearity tests, the Generalised Method of Moments unobservable heterogeneity. After first-differencing, GMM
is appropriate as it corrects for heteroskedasticity, the estimation uses lagged values as instruments for Xit − Xit−1:
endogeneity problems and reduces multicollinearity, hence
improving the efficiency of the estimates.In conclusion, Supposing that the regressors are predetermined, it is
according to specification testing results, a linear dynamic possible to obtain consistent estimates of coefficients
panel GMM is employed. performing a GMM estimator that exploits the following
orthogonality conditions;
Before constructing the dynamic panel GMM model, the
equation below is a starting point for this study to establish E  y it- s ( e it - e it-1 )  = 0  for s ≥ 2  and t = 3, …, T  (7)
 
if the debt choice has an impact on a firm’s value and
firm’s financial performance. A model for the regression of
Islamic debt, non-Islamic debt, the proportion of Islamic E  X it- s (e it - e it-1 )  = 0 for s ≥ 2 and t = 3, …, T  (8)
 
debt, the frequency of Islamic debt issuance, and the type
of the Islamic debt issued is then: Then, the instrumental variable estimation in the first
difference model is:
y it = a + b i1 X i1 + b i 2 X i 2 + δ 1 K i1 + δ 2 K i 2 + ζ 1 N i1 Δy it = γ 1 Δy it-1 +…+ γ p Δy it- p + ΔX ′it b + Δe it ,
+ζ 2 N i 2 + η1 Z i1 + η 2 Z i 2 + uit  (1) (9)
t = p + 1, …, T

uit = µ i + λ t + v it ps(2) where ∆ is the first difference operator. The first difference
002.e in equations and levels using their past levels/first
differences are used for the instrumented variables.
i = 1, …, N; t = 1, …, T,
Further, a test of overidentifiying restrictions is necessary
where yi is firm’s value and/or firm’s financial performance. to test the validity of overidentifying instruments in an
Xi1 is Islamic debt, Xi2 is non-Islamic debt and, Xi3 is firm overidentified model to identify that the parameters of
size. K is the dummy proportion for Islamic debt, N is the the model are estimated using optimal GMM. This test
dummy frequency for Islamic debt and Z is the dummy is called Hansen’s test, and the null hypothesis is that all
Islamic debt type. mi denotes the unobservable individual instruments are valid. At last, weak instruments testing is
effect, lt denotes the unobservable time effect, and vit is done to identify whether the instrument is weak, and the
the remainder stochastic disturbance term. This model overidentified model is used because the model has only
describes three parallel regression planes, which can differ one endogenous regressor that is overidentified (Cameron
in their intercepts. Hereafter, the X, K, N, Z will be referred & Trivedi, 2010; p. 191–199).
as Xit (set of regressors):
5. Analysis
y it = a i + β ′ x it + uit , i = 1, …, N and t = 1, … T , The sample used consists of 80 listed firms issuing
3.eps(3) Islamic debt for the period of 2000 to 2009. Therefore,
there are approximately 3,200 observations used. Table 6
where xit is a K × 1 vector of regressors, β is a K × 1 vector provides the descriptive statistics used in this study.
of parameters to be estimated, and ai represents time- The table depicts the number of observations, mean,
invariant individual nuisance parameters. Under the standard deviation, minimum and maximum value of
null hypothesis, uit is assumed to be independent and each variable. The dependent variables are Tobin’s Q,
identically distributed (i.i.d.) over periods and across ROA and ROE,and each of these dependent variables is
cross-sectional units. regressed toward its explanatory variables.This study
divides all explanatory variables into four categories. The
The GMM equation model (Blundell & Bond, 1998) is first category is the debt structure used by the firm. The
specified as follows (Cameron & Trivedi, 2010): second category is the frequency of Islamic debt issuance.
The third category is the Islamic debt proportion issued.
y it - y it-1 = (a - 1) y it-1 + b ′ X it + η i + e it  (4) The fourth category is the Islamic debt type issued. Firm
size and year of Islamic debt issued are used as control
where yit is the Tobin’s Q at time t for firm i, Xit is a set of variables.
regressors, hi is an unobserved firm-specific effect and eit is
a stochastic error. Moving to the right yit−1, it is to obtain: The mean value for Tobin’s Q is 0.1679 with a range of
-1.6600 to 1.9938, suggesting that most of the firms
experienced low firm performance based on the market
y it = a y it-1 + b ′ X it + η i + e it . (5) measure. A low Tobin’s Q may indicate that the stock is

Eds. Hatem A. El-Karanshawy et al. 25

 Fauzi et al.

Table 6. Descriptive statistics.

Variables Obs. Mean Std. Dev. Min Max

Dependent variables
Tobin’s Q 80 0.1679 0.2129 -1.6600 1.9938
ROA 80 0.0925 0.0004  0.0100 0.1526
ROE 80 0.0156 0.0352  0.0021 0.2292
Explanatory variables the debt structure of the firm
Islamic Debt Proportion 80 0.0806 0.0847  0.0102 0.4576
Non-Islamic Debt Proportion 80 0.2174 0.1725  0.0598 0.8732
The frequency of Islamic debt issuance
First Issuance 80 0.0000 0.0000  0.0000 0.0000
Second Issuance 80 0.1316 0.3381  0.0000 1.0000
More Than two Issuance 80 0.4211 0.4938  0.0000 1.0000
The proportion of Islamic debt issued
Islamic Debt Below Average 80 0.8813 0.3235  0.0000 1.0000
Islamic Debt Average 80 0.0000 0.0000  0.0000 0.0000
Islamic Debt Above Average 80 0.1164 0.3208  0.0000 1.0000
The type of Islamic debt issued 80  
Debt Type of Islamic Debt 80 0.0000 0.0000  0.0000 0.0000
Asset Type of Islamic Debt 80 0.1053 0.3069  0.0000 1.0000
Equity Type of Islamic Debt 80 0.1316 0.3381  0.0000 1.0000
Control Variables Size effect
Firm Size 80 6.0388 0.7254  4.6032 8.4924
Year effect
Year 2001 80 0.0119 0.1081  0.0000 1.0000
Year 2003 80 0.0625 0.2440  0.0000 1.0000
Year 2004 80 0.1563 0.3660  0.0000 1.0000
Year 2005 80 0.3438 0.4787  0.0000 1.0000
Year 2006 80 0.1719 0.3803  0.0000 1.0000
Year 2007 80 0.1406 0.3504  0.0000 1.0000
Year 2008 80 0.0938 0.2938  0.0000 1.0000
Year 2009  80 0.0313  0.1754  0.0000 1.0000

undervalued. Theoretically, stock being undervalued is indicates asset-light firms (for example, agency firms,
likely to happen in a firm which has a stable earning history, software firms, advertising firms, etc). The ROA is
a historically consistent return on equity and a higher approximately 9% which may indicate that the majority
earnings growth rate compared to the market average. of the firms used in the sample are asset-heavy firms, and
Apparently, this seems to be consistent with the sample represent a variety of sectors. These are a few examples of
used for this group, in which the majority of firms are large the firms used in the sample: Esso Malaysia is one of the
firms (see the mean value of firm size, which suggests that biggest fuel providers in Malaysia, Hubline is one of the
most of the firms are big firms). biggest shipping service providers, Kinsteel is one of the
largest steel millers, Kuala Kepong is the largest rubber
The mean value for ROA is 0.0925 with a range of 0.0100 plantation and manufacturer, and Zecon is a construction,
to 0.1526. Though the mean value of ROA is considerably infrastructure, toll concession and property development
small, this positive value indicates that the firms in the company.
sample create shareholder value over the sampling period.
This positive value also indicates an effective utilisation The mean value for ROE is 0.0156 with a range of 0.0021
of firm assets in generating an operating surplus in the to 0.2292, suggesting that most of the firms experienced
business. This lower value of ROA may indicate that the low firm performance based on accounting measures.
firms are asset-intensive firms. If so, they thus require However, the positive value indicates that the firms in the
more money to be invested into the business to continue sample create shareholder value and operating efficiency
generating earnings. According to a common rule, ROA is positively translated into benefits to the owners.
below 5% indicates asset-heavy firms (for example; Furthermore, the lower value of ROE may indicate that
manufacturing, railroads, telecommunication providers, the majority of the firms require more capital invested as
car manufacturers, etc); meanwhile ROA above 20% discussed in point two, where it is noted that the majority

26 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

of the firms are asset-heavy. Therefore, the lower value of Table  7 provides a pairwise correlation matrix of the
ROE does not mean that they have lower performance. explanatory variables. The highest correlation is between
Moreover, those asset-heavy firms have less competition as the Islamic debt proportion and Tobin’s Q, which counts
the entry barrier is high. This can be said to be one of the for 0.4379 (p-value 0.0000) and this value is significant.
competitive advantages of these firms. The second highest correlation is between the proportion
of Islamic debt and the Islamic debt above average, which
The mean value for Islamic debt proportion is 0.0806 with a counts for 0.5979 (p-value 0.0000). The third highest
range of 0.0102 to 0.4576, indicating that most of the firms correlation is between the proportion of Islamic debt and
issued small amounts of Islamic debt. This may be due to the Islamic debt below average, which counts for −0.5965
the fact that this Islamic debt is traded in the thin trading, (p-value 0.0000). The rest of the correlation coefficient
moreover, some of the Islamic debt type certificates cannot is less than 0.5, and it is considered as a low correlation
be traded in the stock exchange due to its Islamic law issue. between the explanatory variables, thus, giving less cause
The mean value for non-Islamic debt proportion is 0.2174 for concern about the multicollinearity problem.
with a range of 0.0598 to 0.8732, indicating that most
of the firms are not highly leveraged. This also suggests Table  8 presents the dynamic GMM panel regression
that the majority of the firms are less risky since excessive results. There are three regression equations, and there are
debt can lead to greater interest payments and principal four explanatory variable categories.
repayment burden.

First issuance is used as a baseline category for the The debt structure of the firm and Tobin’s
frequency of Islamic debt issuance, and it takes the Q, ROA and ROE
value of zero. The mean value for the second issuance of The coefficient of Islamic debt and non-Islamic debt are a
Islamic debt is 0.1316 with a range of 0.0000 to 1.0000, positive and significant, indicating that these two variables
suggesting that only 13.16% of the firms issued Islamic have a positive effect on a firm’s financial performance.
debt for the second time. The mean value for more than Both variables are statistically significant at a 1% level.
two issuance is 0.4211 with a range of 0.0000 to 1.0000, This finding can be better explained by trade-off theory.
suggesting that most of the firms issued Islamic debt more According to prior literature, a firm has an optimal capital
than twice. structure by offsetting the advantages of debt and the cost
of debt (Modigliani & Miller, 1963; DeAngelo & Masulis,
The average of the Islamic debt proportion is 8.06%, which 1980; Jensen & Meckling, 1976; Haris & Raviv, 1990;
is calculated by the total of the Islamic debt proportion over Frank & Goyal, 2003), and this theory apparently can
the total firms in the sample, and thus, this 8.06% average also be applied to Islamic debt. Trade-off theory refers to
value is used as the average category. The mean value for the idea that a company chooses how much debt finance
Islamic debt below average is 0.8831 with a range of 0.0000 and how much equity finance to use by balancing the
to 1.000, suggesting that most of the firms issued Islamic costs and benefits. It states that there is an advantage
debt no greater than 10% (below the average). Islamic debt to financing with debt and the tax benefits of debt, and
average is used as a baseline category for the proportion fortunately Islamic debt is exempted from the taxes.
of Islamic debt issued, and it takes the value of zero. The Moreover, the use of leverage is one way to improve firm
mean value of Islamic debt above average is 0.1164 with a performance (Champion, 1999), and firms prefer debt
range of 0.0000 to 1.0000, suggesting that only a few firms financing because they anticipate a higher return (Hadlock
issued Islamic debt greater than the average. This may be & James, 2002). Furthermore, this finding is in line with
due to the fact that excessive debt issued might increase Krishnan and Moyer (1997) and Abor (2005) who find a
the probability of default. Therefore, the issuers have to positive relationship between capital structure choice and
assess the trade-off between the Islamic debt and any other firm financial performance in developing countries. In
potential risks arising as a result of this debt. particular, Krishnan and Moyer (1997) include Malaysia as
one of the sample in their study.
Debt type is used as a baseline category for the Islamic
debt type and it takes the value of zero. The mean value for The positive result for Islamic debt coefficient obtained
asset type of Islamic debt is 0.1053 with a range of 0.0000 supports the trade-off theory, which was derived from the
to 1.0000, suggesting that only 10.53% of the firms in the models based on taxes and agency cost. From the point of
sample issued this type of Islamic debt. The mean value for view of internal management, having Islamic debt in their
the equity type of Islamic debt is 0.1316 with a range of debt structure brings more pressure to the management as
0.0000 to 1.0000, suggesting that only 13.16% of the firms Islamic debt is more expensive compared to non-Islamic
in the sample issued this type of Islamic debt. debt, hence, improving the firm’s efficiency is important
to maximise asset utilisation due to the Islamic debt
The mean value for firm size is 6.0388 with a range of obtained. At the end, this action leads to improvement in
4.6032 to 8.4924, suggesting that most of the firms are big the firm’s performance. Moreover, debt may reduce agency
firms (see explanation on point two). During the sampling costs by reducing cash flows available for expropriation
period 2000 to 2009, Islamic debt is only issued during and investments in negative net present value projects
these eight years – 2001, 2003 to 2009. Islamic debt is (Harris & Raviv, 1990; Jensen, 1986), as does Islamic
mostly issued in 2005 which accounted for 34.38%. The debt. Furthermore, compared to equity issues, the issue
mean value for 2001, 2003, 2004, 2005, 2006, 2007, 2008 of debt will not dilute the managers’ equity holdings
and 2009 are 1.19%, 6.25%, 15.63%, 34.38%, 17.19%, as a proportion of total equity, but further enhance the
14.06%, 9.38% and 3.13% respectively from the total alignment of interests (Fleming, Heaney & McCosker,
sample. 2005). In addition, though conventional debt and Islamic

Eds. Hatem A. El-Karanshawy et al. 27

 Table 7. Pairwise correlation matrix for explanatory variables.

28
Islamic Debt Non-Islamic First
Variables Tobin’s Q ROA ROE Proportion Debt Proportion Issuance

Tobin’s Q 1.0000
ROA 0.1525*** 1.0000
(0.0000)
ROE 0.0382** 0.1829*** 1.0000
(0.0353) (0.0000)
Islamic Debt Proportion 0.4379*** 0.0393** 0.1006*** 1.0000
(0.0000) (0.0304) (0.0000)
Non-Islamic Debt Proportion 0.0997*** 0.0528*** 0.0466*** -0.2773*** 1.0000
(0.0000) (0.0036) (0.0101) (0.0000)
First Issuance 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Second Issuance -0.0681*** 0.0285 0.003 -0.0847*** -0.0919*** 0.0000
(0.0002) (0.1162) (0.8666) (0.0000) (0.0000) (0.0000)
More than 2 Issuance -0.0699*** 0.0346** 0.0920*** -0.0334* -0.0780*** 0.0000
(0.0001) (0.0565) (0.0000) (0.0659) (0.0000) (0.0000)
Islamic Debt Below Average -0.3119*** 0.0780*** 0.1048*** -0.5965*** 0.0788*** 0.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Islamic Debt Average 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Islamic Debt Above Average 0.3119*** 0.0813*** 0.1059*** 0.5979*** -0.0817*** 0.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Debt Type of Islamic Debt 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Asset Type of Islamic Debt -0.0556*** 0.0228 0.0087 -0.0400** -0.0587*** 0.0000
(0.0021) (0.2080) (0.6304) (0.0274) (0.0012) (0.0000)
Equity Type of Islamic Debt 0.0045 0.0073 0.1006*** 0.1094*** 0.2026*** 0.0000
(0.8056) (0.6892) (0.0000) (0.0000) (0.0000) (0.0000)
Firm Size -0.2079*** 0.0082 0.1971*** -0.1369*** 0.4492*** 0.0000
(0.0000) (0.6878) (0.0000) (0.0000) (0.0000) (0.0000)
Year 2001 -0.0056 0.0000 -0.0483*** 0.1882*** 0.1362*** 0.0000
(0.7587) (1.0000) (0.0078) (0.0000) (0.0000) (0.0000)
Year 2003 0.0573*** 0.0000 -0.0292 0.0611*** 0.0971*** 0.0000
(0.0016) (1.0000) (0.1074) (0.0007) (0.0000) (0.0000)
Year 2004 0.0348** -0.1001*** -0.0884*** 0.0271 -0.1106*** 0.0000
(0.0547) (0.0000) (0.0000) (0.1349) (0.0000) (0.0000)
Year 2005 0.2012*** 0.0097 -0.0105 0.3000*** -0.0901*** 0.0000
(0.0000) (0.5919) (0.5624) (0.0000) (0.0000) (0.0000)
Year 2006 0.0689*** 0.0000 0.0037 0.0496*** -0.0866*** 0.0000
(0.0001) (1.0000) (0.8368) (0.0062) (0.0000) (0.0000)
Year 2007 0.0078 0.0355** 0.0542*** 0.0237 0.0624*** 0.0000
(0.6681) (0.0501) (0.0028) (0.1917) (0.0006) (0.0000)
Year 2008 -0.0361** 0.0000 0.0732*** 0.0153 0.0330*** 0.0000
(0.0466) (1.0000) (0.0001) (0.398) (0.0690) (0.0000)
Year 2009 -0.0688*** 0.0000 0.0010 -0.0177 -0.0148 0.0000

Islamic banking and finance – Essays on corporate finance, efficiency and product development
Fauzi et al.

(0.0001) (1.0000) (0.9572) (0.3304) (0.4135) (0.0000)

 Table 7. (Continued)

Second More Than Islamic Debt Islamic Debt Debt Type of Asset Type of
Issuance 2 Issuance Below Average Above Average Islamic Debt Islamic Debt

Second Issuance 1.0000

More than 2 Issuance -0.3320*** 1.0000
(0.0000)
Islamic Debt Below Average 0.1188*** -0.0309* 1.0000
(0.0000) (0.0886)

Eds. Hatem A. El-Karanshawy et al.

Islamic Debt Average 0.0000 0.0000 0.0000
(0.0000) (0.0000) (0.0000)
Islamic Debt Above Average -0.1170*** 0.0290 -0.9890*** 1.0000
The impact of Islamic debt on company value

(0.0000) (0.1103) (0.0000)

Debt Type of Islamic Debt 0.0000 0.0000 0.0000 0.0000 1.0000
(0.0000) (0.0000) (0.0000) (0.0000)
Asset Type of Islamic Debt -0.1335*** 0.0548*** 0.0199 -0.0209 0.0000 1.0000
(0.0000) (0.0025) (0.2732) (0.2487) (0.0000)
Equity Type of Islamic Debt -0.0364** -0.0166 0.0075 -0.0109 0.0000 -0.1335***
(0.0450) (0.3603) (0.6785) (0.5496) (0.0000) (0.0000)
Firm Size 0.1010*** -0.0076 0.2083*** -0.2060*** 0.0000 -0.0363*
(0.0000) (0.7086) (0.0000) (0.0000) (0.0000) (0.0740)
Year 2001 -0.0426*** -0.0934*** -0.1008*** 0.1025*** 0.0000 -0.0375**
(0.0188) (0.0000) (0.0000) (0.0000) (0.0000) (0.0384)
Year 2003 0.0088 0.0000 -0.1189*** 0.1218*** 0.0000 -0.0900***
(0.6293) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Year 2004 0.0458*** -0.0926*** -0.0096 0.0119 0.0000 -0.0144
(0.0115) (0.0000) (0.5979) (0.5132) (0.0000) (0.4266)
Year 2005 0.0449*** -0.0038 -0.3430*** 0.3406*** 0.0000 -0.0897***
(0.0132) (0.8321) (0.0000) (0.0000) (0.0000) (0.0000)
Year 2006 -0.1011*** 0.1730*** -0.0761*** 0.0744*** 0.0000 -0.0186
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.3065)
Year 2007 0.0231 0.0733* -0.0665*** 0.0591*** 0.0000 0.1052***
(0.2020) (0.0001) (0.0002) (0.0011) (0.0000) (0.0000)
Year 2008 0.0046 -0.0772*** -0.0102 0.0037 0.0000 0.0168
(0.8011) (0.0000) (0.5737) (0.8405) (0.0000) (0.3550)
Year 2009 -0.0200 0.0082 0.0189 -0.0186 0.0000 0.0661***
(0.2704) (0.6508) (0.2987) (0.3040) (0.0000) (0.0003)

29
30
Table 7. (Continued)

Equity Type of
Islamic Debt Firm Size Year 2001 Year 2004 Year 2005 Year 2006

Equity Type of Islamic Debt 1.0000

Firm Size 0.2151*** 1.0000

(0.0000)
Year 2001 -0.0426*** 0.0085 1.0000
(0.0188) (0.6746)
Year 2003 -0.1022*** 0.0167 -0.0287
(0.0000) (0.4107) (0.1131)
Year 2004 -0.1200*** -0.1020*** -0.0338* 1.0000
(0.0000) (0.0000) (0.0627)
Year 2005 -0.0618*** -0.0172 -0.0474*** -0.1335*** 1.0000
(0.0007) (0.3959) (0.0089) (0.0000)
Year 2006 -0.1011*** -0.0618*** -0.0284 -0.0801*** -0.1124*** 1.0000
(0.0000) (0.0023) (0.1171) (0.0000) (0.0000)
Year 2007 0.0781*** 0.0917*** -0.0244 -0.0688*** -0.0966*** -0.0579***
(0.0000) (0.0000) (0.1784) (0.0001) (0.0000) (0.0014)
Year 2008 0.2361*** 0.1215*** -0.0150 -0.0422*** -0.0593*** -0.0356**
(0.0000) (0.0000) (0.4085) (0.0198) (0.0011) (0.0499)
Year 2009 -0.0200 0.0244 -0.0056 -0.0158 -0.0222 -0.0133
(0.2704) (0.2299) (0.7566) (0.3826) (0.2202) (0.4623)
Year 2007 Year 2008 Year 2009
Year 2007 1.0000
Year 2008 -0.0305* 1.0000
(0.0922)
Year 2009 -0.0115 -0.0070 1.0000
(0.5278) (0.6981)

***Sig. at 1% significance level, **Sig. at 5% significance level and *Sig. at 10% significance level.

Islamic banking and finance – Essays on corporate finance, efficiency and product development
Fauzi et al.
The impact of Islamic debt on company value

Table 8. Summary of the regression result.

Dynamic GMM

Variables Tobin’s Q ROA ROE

L1 0.5742*** 0.1220*** 0.1547***

(0.0174) (0.0041) (0.0015)
Constant 0.3180*** 8.7500*** 0.8463**
(0.0988) (0.0410) (0.4264)
The debt structure of the firm
Islamic Debt Proportion 0.7441*** 0.0159*** 0.0276**
(0.0723) (0.0025) (0.2540)
Non-Islamic Debt Proportion 0.5372*** 0.0177*** 0.3357***
  (0.0159) (0.0015) (0.0544)
The frequency of Islamic debt issuance
D1_First Issuance  (Omitted) (Omitted) (Omitted)
D2_Second Issuance -0.0961* -0.0774*** -0.7018***
  (0.0951) (0.0140) (0.2332)
D3_More Than 2 Issuance 0.2090*** 0.0323*** 1.2480***
(0.0562) (0.0026) (0.1135)
The proportion of Islamic debt issued
D1_ID Below Average 0.0006* 0.0181*** 0.1319*
(0.0129) (0.0010) (0.1146)
D2_ID Average (Omitted)  (Omitted)  (Omitted)
D3_ID Above Average -0.0248** -0.0036*** -0.0135**
  (0.0111) (0.0010) (0.1266)
The type of Islamic debt issued
D1_Debt Type of ID (Omitted) (Omitted) (Omitted)
       
D2_Asset Type of ID 0.0423 0.0250*** 0.3337
  (0.1076) (0.0074) (0.2695)
D3_Equity Type of ID 0.1764* 0.0346*** 1.0458***
(0.1049) (0.0027) (0.0532)
Control Variables Size effect
Firm Size -0.0844*** -0.0040*** -0.2354***
(0.0142) (0.0005) (0.0654)
Year effect
Year 2001 -1.5848* -0.0169* -0.5066***
(0.9510) (0.0097) (0.1019)
Year 2003  0.0458*** 0.0092*** -0.0786*
(0.0098) (0.0024) (0.0433)
Year 2004 0.0130 0.0173*** 0.3969***
(0.0109) (0.0019) (0.0657)
Year 2005 0.0171*** 0.0058*** 0.0981***
(0.0048) (0.0014) (0.0159)
Year 2006 -0.0346*** -0.0018*** -0.0392***
(0.0072) (0.0003) (0.0078)
Year 2007 0.1160*** 0.0080*** -0.0889***
(0.0177) (0.0020) (0.0298)
Year 2008 0.0397* 0.0017*** 0.8061***
(0.0236) (0.0006) (0.1146)
Year 2009 0.0587** 0.0048*** 0.7980**
(0.0294) (0.0019) (2.8289)
J-Statistics 21514.85 78177.74 1.27e+06
Chi2 0.0000 0.0000 0.0000

Eds. Hatem A. El-Karanshawy et al. 31

 Fauzi et al.

debt are fundamentally different, they perform similarly in in turn have had an important influence in shaping the
a competitive market as these two instruments are affected increasing complexity and sophisticated nature of its capital
by the same factors (Kraciska & Nowak, 2012). market along with the implementation of regulations, and
these changes support firms to operate more effectively and
Although both debt types have a positive impact, the efficiently, increasing the confidence of markets. Moreover,
coefficient for Islamic debt is higher than the coefficient a more diversified financial system, in particular, the rapid
for non-Islamic debt (only for Tobins’ Q), suggesting that growth of the Malaysian Islamic Capital Market and the
the Islamic debt provides a higher contribution to the Malaysian debt market, has increased the alternative
improvement of firms’ financial performance compared to sources of financing available to corporations.
non-Islamic debt. Furthermore, it can be concluded that
when Islamic debt is chosen as a tool of firm financing, These key benefits supports the theory that the choice of
(1) the markets react positively to firm performance, thus capital structure may help mitigate agency costs (Jensen
this positive reaction might lead to the stock becoming & Meckling, 1976). According to the agency costs theory,
overvalued;(2) Islamic debt not only improves the high leverage or a low equity/asset ratio reduces the
effectiveness of the firm’s management s in managing their agency costs of outside equity and increases firm value by
assets to generate profits, but it also improves the operating constraining or encouraging managers to act more in the
efficiency of the total business; (3) firms are effective in interests of shareholders. Moreover, corporate debt has a
managing their operation efficiency which in the end disciplining effect on management, since it serves to reduce
contributes to the owners’ wealth because ROE measures the free cash flow and therefore minimises management’s
the performance from the perspective of the equity- discretionary spending.
holders. There are a few reasons for this significant positive
contribution of Islamic debt issuance. First, Islamic debt is Overall, the finding for ROA and ROE are similar to Tobin’s
claimed and advertised as a secure investment due to its Q which also supports the trade-off theory (Modigliani &
structure. Second, Islamic debt is given a special privilege Miller, 1963; DeAngelo & Masulis, 1980; Jensen & Meckling,
such as stamp duty and exempted tax for both issuers and 1976; Haris & Raviv, 1990; Frank & Goyal, 2003), and this
investors. Third, Islamic debt is guaranteed by the special theory apparently can also be applied to Islamic debt.
purpose vehicle (SPV); in case of default the Islamic debt
holders may recourse the assets underlying the Islamic
debt. Fourth, though there were a few cases of default in The frequency of Islamic debt issuance and Tobin’s
Middle East, those cases have no impact on the investors’ Q, ROA and ROE
perspective, as some investors investing in Islamic debt only The coefficient for first issuance of Islamic debt is a positive
do so only to comply with the religious matter. Fifth, the and significant at 1% level of significance (but for Tobins’ Q
majority of investors are non-Muslim, with an increasing which significant at 10% level of significance), suggesting
presence of foreign investors (PricewaterhouseCoopers that the first issuance of Islamic debt affects higher firm
Malaysia, 2008). Sixth, the Islamic debt issuance contributes performance. This also indicates that (1) the markets
to an increase in the issuer’s stock returns (Nagano, nd.). react positively to the issuance of Islamic debt when it is
first introduced to the market; (2) the firm effectively
Moreover, from the issuers’ perspective, there are benefits utilises its assets to generate profits for the shareholders,
issuing Islamic securities, in particular, Islamic debt. The and additional debt, in particular Islamic debt, pushes the
key benefits are tax incentives, value proposition and management to perform better. There are several factors
regulatory process. First, for tax incentives, the issuers are that might contribute to this positive finding. First, the
exempted from stamp duty, tax deductible of issuance cost, managers of the firms are compelled to put more effort into
and the special purpose vehicle (SPV) is exempted from tax, generating more profits. Because some of Islamic debt is in
and tax neutrality. Second, for value proposition, there is a the form of partnership (profit and loss sharing agreement),
wider investors’ base, Islamic debt is attractively priced due Islamic debt tends to place greater pressure on the managers
to the strong demand, there is strong structuring expertise to manage the firms effectively. Second, there is a broad-
in the Islamic finance industry, and Islamic debt enhances based coordination of government policies which resulted
the issuers’ profile. Third, in terms of regulatory process, in a comprehensive public policy that supports growth and
the process facilitates the issuance process, the rating innovation in the Islamic financial market, in particular,
of Islamic debt is automatically approved for AAA-rated Islamic debt. Third, the importance of government
for Islamic debt issued in domestic (Malaysian) currency intervention, such as tax incentives and required ratings
and A-rated Islamic debt issued in foreign currency, any improves issuers’ and investors’ confidence. Fourth, the
amendment to terms of approved Islamic debt need only rapid growth of Islamic finance signifies that Islamic debt
to inform the Securities Commission, and exchangeable has moved from the pioneering stage to being an established
Islamic debt is exempted from rating. From the point of financing instrument that serves as a commercially viable
view of shareholders, the usage of debt increases their and effective tool for mobilising investment assets to finance
wealth, and because of this, markets believe that Islamic productive economic activities. Fifth, in the beginning of
debt positively contributes to the firm performance. the Islamic finance initiation, Islamic debt offered those
Moreover, Islamic debt issuance contributes to an increase competitiveness features, particularly cost effectiveness,
in the issuers’ total factor productivity (Nagano, n.d.). secureness and efficiency. As such, the market had high
expectations of this new instrument, the upshot was
Furthermore, the positive result may be due to the stabilised that Islamic debt brought more pressure on managers
nature of the Malaysian financial system which has evolved to manage their firms effectively in order to meet market
in line with the changing structure of the economy. The expectations. Sixth, apart from being well-regulated by
changes in the economic structure and financial system various standards and guidelines, Malaysia is also the only

32 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

country that makes it compulsory for all tradable corporate stream of cash flow over time. Thus, Islamic debt also
debt securities to be rated to enhance investors’ confidence reduces the agency problem within the company and hence
and to assist in the investment decision-making process. increases firm value. (2) That as the industry grows, it is
Another distinguishing factor for the Malaysian Islamic more apparent that there is more demand by non-Muslim
debt market is the establishment of a centralised, national investors and issuers to play a role in the industry. Here in
level Shariah supervisory board, which ensures that every Malaysia, for instance, there is just as strong a demand for
Islamic debt issued in Malaysia, is in full compliance with Shariah compliant products among non-Muslims as there
the Shariah. All these factors provide sufficient protection is among Muslims (PricewaterhouseCoopers Malaysia,
to investors in the Islamic debt and conventional debt 2008).From the view point of markets, this may indicate
markets. that the markets have learnt through several issuances of
Islamic debt and therefore they have greater confidence in
However, the coefficient for the second issuance of Islamic subsequent issuances compared to the second issuance of
debt is a negative and significant at 5% level of significance, Islamic debt. However, investors are irrational according
suggesting that the issuance of Islamic debt for a second to the behavioural finance theory. Their decision may be
time lowers firm performance. This negative finding is influenced by the magnitude issue, their bias selection and
similar to the study by Godlewski et  al. (2010), which the lucky event issue.
suggest that Islamic debt expansion has a detrimental
effect on firm value. This negative finding may indicate
that (1) either the management of the firms have loosened The proportion of Islamic debt issued
their control because of overconfidence from the first and Tobin’s Q
successful issuance of Islamic debt or that the management The coefficients for the proportion of Islamic debt below
have expropriated the firms’ previous profits; (2) the the average and at the average are a positive and significant
markets have experienced, observed and learnt from the varies at 10% and 1% level of significance. These positive
first Islamic debt issuance, leading underconfidence in and significant results may be caused by internal and
the markets over this second issuance, which in turn may external factors. In terms of internal factors, the proportion
affect the share price of those firms issuing Islamic debt; of Islamic debt issued at a certain level stimulates the
(3) low credit rating of firms issuing Islamic debt as this management to work effectively. For external factors,
is associated with high risk. The gap between the first and there are two views; first from the markets’ view, second
the second Islamic debt issuance ranges between two to six from the view of government support. From the markets’
years. Presumably, in that time period, investors observed view, the proportion of a certain level of Islamic debt may
the firm’s performance, their Islamic debt rating, the be considered as tax exempted stimulation as the profits
market conditions such as the frequency of default cases of derived from Islamic debt are exempted from the taxes.
Islamic debt. In Malaysia, cases of Islamic debt default were Furthermore, the markets have confidence over the assets/
few and it is something that raises concern on the investors’ projects underlying the Islamic debt contract which may
protection because a default occurs due to the breach of bring profits in future; therefore, this market confidence
any binding obligations under the original terms of the affects their stock price. With regards to government
agreement between the issuer and the Sukuk holders. Thus support, the Malaysian government has provided an
this factor may contribute to the negative result. interesting model to promote the co-existence of an ethical
and societal-based finance through issuing a few regulations
Furthermore, debt is also a source of information which that appeal to Muslim and non-Muslim investors; hence
indicates the firm’s current condition that investors can these regulations issued can assure the credibility of this
use to monitor and evaluate major operating decisions of instrument. Furthermore, the regulating body has taken
the firm in two ways. Firstly, the mere ability of the firm vital steps to develop a facilitative regulatory framework, to
to make its contractual payments to debt-holders provides create a large pool of players, to introduce a comprehensive
information. Secondly, in the event that the organisation range of innovative and competitive Islamic financial
fails to make the payments, their ways to resolve the product and services, and to ensure sufficient depth to
matter either through informal negotiation or formal facilitate liquidity management, hence creating market
bankruptcy proceedings will disseminate considerable confidence.
information to the investors (Harris & Raviv, 1990). In
sum, the negative relationship of the second issuance of Though debt reduces the agency costs of free cash flow
Islamic debt and its firm’s performance is probably either by reducing the cash flow available for spending at the
a result of the previous firm performance in meeting their discretion of managers (Jensen, 1986), an increased
obligation of payment or a result of inefficient utilisation leverage also has costs; as leverage increases the risk of
of their firms’s assets. default also increases. This theory supports the result for
Islamic debt above the average which is a negative and
Fortunately, the coefficient for more than two issuance significant at 1% level of significance. This finding suggests
of Islamic debt is a positive and significant at 1% level of that the greater the proportion of Islamic debt issued, the
significance, suggesting the issuance of Islamic debt for lower the firm performance. This result is similar to the
more than two improves a firm’s financial performance. empirical result for non-Islamic debt, in that the proportion
This may indicate that after having a few experiences of debt at a certain level may hamper firm performance as
in issuing Islamic debt, the issuance of Islamic debt later an additional incurrence of debt gives no guarantee that
on impacts positively on firm performance. This may be firm performance will be higher. This is mainly because as
caused by the fact that (1) the debt-holders of Islamic debt the leverage increases, so does the risk of default, which
closely monitor the management of the firm to ensure provides a greater incentive for lenders to monitor the firm.
that the firm can generate profits and distribute a periodic Though it is claimed that Islamic debt is more secure than

Eds. Hatem A. El-Karanshawy et al. 33

 Fauzi et al.

the conventional debt, this result finds no support for that infrastructure, vibrant business environment and quality
claim. On the contrary, this finding supports the notion of life, has always been an attractive market for foreign
that as the leverage increases, the probability of default investors. Therefore, the coefficients for year 2003, 2004
also increases, and Islamic debt is no exception to this rule. and 2005 are supported.
Overall, the result for Islamic debt proportion Tobins’ Q,
ROA and ROE has similarity. Despite the challenging global economy, Malaysia
has continued to pursue liberalisation, enhancing the
entrepreneurial and investment environments. The
The type of Islamic debt and Tobin’s Q economy scores above the world average in many of the
The coefficients for the debt-type and equity-type are a ten economic freedoms (World Bank, 2011). The trade
positive and significant at 1% and 10% level of significance regime is relatively open despite lingering non-tariff
for Tobins’ Q and ROE. While all types of Islamic debt are barriers. However, corruption and a judicial system that
a positive and significant at 1% level of significance for remains vulnerable to political influence pose significant
ROA. Though the finding for ROA is slightly different than challenges to economic freedom. 2001 and 2006 were two
for Tobin’s Q, this result does not impair on the Tobin’s Q years which yielded a negative and significant impact. The
result, as it is common for different methods of calculation first, 2001, may be due to the global economic slowdown
to give different results. The finding suggests that debt-types overall. Significantly, though, a general election was held in
and equity-types affect higher firm performance. The result 2003 and again in 2008, revealing a pattern in which there
supports the notion that certain types of debts have a different is a two year gap between this political event and a year
impact on shareholders’ wealth (Mikkelson & Partch, 1986); yielding a negative and significant impact. This may indicate
hence, this finding can also be applied to Islamic debt. that before the general election, the political situation in
Malaysia heats up, which affects the market players.
Furthermore, the finding can be explained by the different
Islamic debt structure. This is important since the structure The Malaysian economy has been surprisingly resilient in
determines the obligation of the originator/issuers. There is spite of the global slowdown in 2007.Malaysia has only
typically a requirement that on maturity of the Islamic debt felt a minor impact from the slowing US economy, but
or upon an event of default, the originator has a purchase emerging challenges in the form of soaring food prices and
obligation to repurchase the assets which enables the the persistent rise in global oil prices are weighing down
Special Purpose Vehicle (SPV) to redeem the outstanding heavily on economic prospects. Furthermore, to avoid the
certificates and repay the Sukuk holders. In this regard, fiscal deficit, the government announced a revamp in oil
the rights of Sukuk holders in the event of default will vary subsidies, pushing up the price of petrol diesel, which has
depending on whether the Sukuk structure is an asset- adverse implications for inflation and economic growth.
based or an asset-backed structure. The positive result However, in 2008 and 2009, the business confidence index
for debt-based and equity-based Sukuk may be caused by increased as it indicates by the rise of sales and production,
their structure. The assumptions that may be raised is that higher export sales, higher capacity utilisation, higher
debt-based and equity-based are in the structure of asset- domestic demands and higher capital investment. The
backed Sukuk, and asset-based is in the structure of asset- gross domestic product growth was sustained at a certain
based Sukuk. Thus, the rights of the Sukuk-holders depend targeted level. This growth was driven by high commodity
on the structure of Islamic debt. For example, in the case prices, strong private consumption and steady investment,
of Sukuk ijarah, if the Sukuk is asset-backed, this allows and supported by fiscal spending. The business condition
the holders to liquidate the underlying asset in the event index would be a better indicator of current economic
of default to recover most of their investments. On the activity as it relies on firm-level information. Therefore,
other hand, if the Sukuk is asset-based, this only represents the positive and significant coefficients for year 2007, 2008
beneficial ownership on the underlying asset and it restricts and 2009 are supported.
the holders’ rights in the event of a default.

The coefficient for firm size is a negative and significant for 6. Conclusion
all four regression equations, suggesting that bigger firms In sum, the findings for all three categories of explanatory
which having Islamic debt in their debt structure have a variables, along with their control variables for all metrics
lower firm performance. The negative result may be due to (Tobins Q, ROA and ROE), are only slightly different in
the fact that bigger firms are already well-stabilised in terms their coefficient value. Almost all the coefficient signs and
of cash flows and profits because of their well-stabilised significance values reveal the same direction and a similar
capital structure; hence changing its capital structure significance value. The coefficients for Islamic debt is higher
with a new unproven instrument may endanger the firm’s than the coefficient for non-Islamic debt and, overall, the
credibility and ability to maintain their stable cash flows findings for Tobin’s Q, ROA, ROE and EVA support the
and profits. This notion leads to the markets’ perspective trade-off theory (Modigliani & Miller, 1963; DeAngelo &
on the firms’ capability in the future; the markets may have Masulis, 1980; Jensen & Meckling, 1976; Haris & Raviv,
lower confidence and in turn, this affects the stock price of 1990; Frank & Goyal, 2003) and this theory apparently can
the firms. also be applied to Islamic debt.

Apart from year 2004, all the years reveal a significant

result. All the years (2003, 2004, 2005, 2007, 2008, and References
2009) have a positive coefficient except year 2001 and year Abd. Sukor, M.E., Muhamad, R, Gunawa, A.Y. (2008)
2006. Malaysia, with its economic strength, supportive Malaysian Sukuk: Issues in accounting standard.
government policies, educated workforce, developed Shariah Journal, 16(1), 63–74.

34 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value

Abor, J. (2005) The effect of capital structure on Ghosh, A., Cai, F. (1999) Capital structure: New evidence of
profitability: An empirical analysis of listed firms in optimality and pecking order theory. American Business
Ghana. Journal of Risk Finance, 6, 438–447. Review, 17(1): 32–38.
Abor, J. (2007) Debt policy and performance of SMEs: Hadlock, C.J., James, C.M. (2002) Do banks provide
Evidence from Ghanaian and South African firms. The financial slack? Journal of Finance, 57, 1383–1420.
Journal of Risk Finance, 8(4), 364–379.
Haneef, R. (2009) From asset-backed to asset-light
Al Amine, M.A.B. (2008) Sukuk market: Innovations and structures: The intricate history of Sukuk. ISRA
challenges. Islamic Economic Studies, 15(2), 1–22. International Journal of Islamic Finance, 1(1).
Al Amine, M.A.B. (n.d.). The Islamic bonds market: Harris, M., Raviv, A. (1990) Capital structure and the
Possibilities and challenges. International Journal of informational role of debt. The Journal of Finance,
Islamic Financial Services, 3(1). 45(2), 321–349.
Al-Amine, M.A.B.M. (2001) Istisna’ and Its application in Hatfield, G.B., Cheng, L.T.W., Davidson, W.N. (1994) The
Islamic banking. Arab Law Quarterly, 16(1), 22–48. determination of optimal capital structure: The effect of
firm and industry debt ratios on market value. Journal of
Ashhari, M.Z., Chun, L.S., Nassir, A. Md. (2009)
Financial and Strategic Decisions, 7(3):1–14.
Conventional vs Islamic bond announcements: The
effects on shareholders’ wealth. International Journal of Jensen, M.C. (1986) Agency costs of free cash flow,
Business and Management, 4(6). corporate finance and takeovers. American Economic
Review, 76(2), 323–329.
Ayub, M. (2007) Understanding Islamic Finance. England:
John Wiley & Sons, Ltd. Jensen, M.C., Meckling, W.H. (1976) Theory of the firm:
Managerial behavior, agency costs and ownership
Baltagi, B.H. (2005) Econometric Analysis of Panel Data (3 rd
structure. Journal of Financial Economics, 3(4),
ed.). West Sussex, England: John Wiley & Son Ltd.
305–360.
Bhabra, H., Liu, T., Tirtiroglu, D. (2008) Capital Structure
Jermias, J. (2008) The relative influence of competitive
Choice in a Nascent Market: Evidence from Listed Firms
intensity and business strategy on the relationship
in China. Financial Management, 37(2), 341.
between financial leverage and performance. The British
Blundell, R., Bond, S. (1998) Initial conditions and moment Accounting Review, 40(1), 71–88.
restrictions in dynamic panel data models. Journal of
Juan, S. (2008) Prospects and challenges for developing
Econometrics, 87, 115–143.
corporate Sukuk and bond markets: International
Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovic, V. Monetary Fund. International Journal of Islamic and
(2001) Capital Structures in Developing Countries. Journal Middle Eastern Finance and Management, 1(1), 20–30.
of Finance, 56(1), 87–130.
Kamali, M.H. (2007) A shari‘ah analysis of issues in Islamic
Cakir, S., Raei, F. (2007) Sukuk vs. Eurobonds: Is there leasing. Islamic Economic, 20(1), 3–22.
a difference in value-at-risk, IMF Working Paper
Kraciska, O., Nowak, S. (2012) What’s in it for me? A
(Vol. WP/07/237): International Monetary Fund.
primer differences between Islamic and conventional
Cameron, A.C., Trivedi, P.K. (2010) Microeconometrics using finance in Malaysia. IMF Working Paper, 151.
stata (Rev. Ed.).College Station, Texas: Stata Press.
Krishnan, V.S., Moyer, R.C. (1997) Performance, capital
Champion, D. (1999) Finance: The joy of leverage. Harvard structure and home country: An analysis of Asian
Business Review, 77(4), 19–22. corporations. Global Finance Journal, 8(1), 129–143.
Coleman, A.K. (2007) The impact of capital structure Mikkelson, W.H., Partch, M.M. (1986) Valuation effects of
on the performance of microfinance institutions. The security offerings and the issuance process. The Journal
Journal of Risk Finance, 8(1), 56–71. of Financial Economics, 15, 31–60.
DeAngelo, H. & Masulis, R.W. (1980) Optimal capital Mirakhor, A. (1996) Cost of capital and investment in a
structure under corporate and personal taxation. non-interest economy. Islamic Economic Studies, 4(1),
Journal of Financial Economics, 8(1), 3–29. 35–46.
Ebaid, I.E.S. (2009) The impact of capital-structure choice Modigliani, F., Miller, M. (1963) The cost of capital,
on firm performance: empirical evidence from Egypt. corporation finance and the theory of investment.
The Journal of Risk Finance, 10(5), 477–487. American Economic Review, 48, p.261–297.
Eriotis, N.P., Frangouli, Z., Neokosmides, Z.V. (2002) Profit Modigliani, F., Miller, M.H. (1958) The cost of capital,
margin and capital structure: An empirical relationship. corporation finance and the theory of investment.
The Journal of Applied Business Research, 18(2), 85–88. American Economic Review, 48(3), 261.
Fleming, G., Heaney, R., McCosker, R. (2005) Agency Mohd Yatim, M.N., Shah, M. (2009) Sukuk (Islamic bond): A
costs and ownership structure in Australia. Pacific-Basin crucial financial instrument for securitisation of debt for
Finance Journal, 13, 29–52. the debt-holders in shari’ah-compliant capital market.
International Journal of Business and Management.
Frank, M.Z., Goyal, V.K. (2003) Testing the pecking
order theory of capital structure.Journal of Financial Mokhtar, S. (2009) A synthesis of Shariah issues and
Economics, 67(2), 217–248. market challenges in the application of wa’d in equity-

Eds. Hatem A. El-Karanshawy et al. 35

 Fauzi et al.

based Sukuk. ISRA International Journal of Islamic empirical study of the Indian Manufacturing sectors.
Finance, 1(139–145). Strategic Management Journal, 22, 989–998.
Myers, S., C., Majluf, N., S. (1984) Corporate financing and Somolo, E. (2009) Islamic Pricing Benchmark. ISRA Inter­
investment decisions when firms have information that national Journal of Islamic Finance, 1(1).
investors do not have. Journal of Financial Economics,
Talberg, M., Winge, C., Frydenberg, S. & Westgaard, S.
13, 187–221.
(2008) Capital structure across industries. International
Myers, S.C. (1984) The capital structure puzzle. Journal of Journal of the Economics of Business, 15(12), 181–200.
Finance, 39, 575–592.
Tariq, A.A., Dar, H. (2007) Risks of Sukuk structures:
Naceur, S.B., Goaied, M. (2002) The relationship between implications for resource mobilization. Thunderbird
dividend policy, financial structure, profitability and International Business Review, 49(2), 203–233.
firm value. Applied Financial Economics, 12; 843–849
Usmani, M.T. (1999) The Concept of Musharakah and Its
Nagano, M. (n.d.). Islamic finance and the theory of capital Application as an Islamic Method of Financing. Arab
structure. Discussion Papers in Economics, Society of Law Quarterly, 14(3), 203–220.
Economics Nagoya City University, 51.
Vishwanath, S.R., Azmi, S. (2009) An overview of Islamic
Ni, J., Yu, M. (2008) Testing the Pecking-Order Theory: Sukuk bonds. The Journal of Structured Finance, 14(4),
Evidence from Chinese Listed Companies. The Chinese 58–67.
Economy, 41(1), 97–113.
Wilson, R. (2008) Innovation in the structuring of Islamic
Phillips, P.A., Sipahioglub, M.A. (2004) Performance Sukuk securities. Humanomics, 24(3), 170–181.
implications of capital structure: evidence from quoted
Wilson, R. (n.d.). Islamic Bonds: Your Guide to Issuing,
UK organisations with hotel interests. Service Industries
Structuring and investing in Sukuk.
Journal, 24(5), 31–51.
Wiwattanakantang, Y. (1999) An empirical study on the
PricewaterhouseCoopers Malaysia.(2008) Malaysia, Asia’s
determinants of the capital structure of Thai firms.
Islamic Finance hub (Kuala Lumpur). Retrieved from:
Pacific-Basin Finance Journal, 7(3–4), 371–403.
http://www.pwc.com/my/en/publications/islamic-
finance-hub.jhtml World Bank. (2011).
Raghunathan, T.E. (2004) What do we do with missing Yean, T.W. (n.d.). Sukuk: Issues and the way forward.
data? Some options for analysis of incomplete data.
Zeitun, R., Tian, G.G. (2007) Capital structure and
Annual Reviews Public Health, 25, 99–117.
corporate performance: Evidence from Jordan.
Ramaswamy, K. (2001) Organisational ownership, Australasian Accounting Business and Finance Journal,
competitive intensity and firm performance: An 1(4): 40–61.

36 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a
dual banking system: Evidence from Malaysia
Muhamed Zulkhibri
Economic Research and Policy Department, Islamic Development Bank, P.O Box 5925, Jeddah 21432, Saudi Arabia,
E-mail: [email protected]

Abstract - An understanding of financing behaviour explains the performance of Islamic banks

as an alternative to the conventional finance model as suppliers of capital for businesses and
entrepreneurs. In the Islamic banking system, banks are suppliers of capital and not lenders unlike
that stipulated in a traditional banking system. To date, Islamic banks have become the major
source of capital in Malaysia and lending behaviour is an important policy variable. In this context,
the paper examines the relationship between bank financing, bank financing rate and bank-specific
characteristics in a dual banking system. The evidence suggests that the bank-specific characteristics
are important for Islamic banking financing behaviour. The Islamic banks’ financing behaviour is
consistent with behaviour of conventional banks in that the bank lending operates through banks
depending on the size, and level of liquidity and capital. The findings also suggest that that there is
no significant difference between Islamic bank financing and conventional bank lending behaviour
with respect to interest rates.

Keywords: Islamic banks, base financing rate, bank financing, panel regression analysis

1. Introduction understanding of the functioning of banks must lead to the

conclusion that Islamic banks are not viable.
An Islamic bank is a deposit-taking institution, which
includes all functions currently known as banking activities.
The distinguishing features of Islamic banks are
The bank mobilizes funds on the basis of mudaraba (profit-
the prohibition of charging or paying interest, the
sharing) or wakalah (as an agent charging a fixed fee for
impermissibility of demanding collateral and, to a small
managing funds), which form part of its liabilities, while
extent, compulsory charitable spending (Khan and
financing on a profit-and-loss sharing (PLS) basis or through
Mirakhor 1992). Profit has to be generated merely by
the purchase of goods (on cash) and sale (on credit) or
primary and secondary modes of Islamic finance (Chapra
other trading, leasing and manufacturing activities, form
2000). Primary modes include profit-sharing arrangements
part of the assets. Apart from demand deposits, which are
such as mudaraba (partnership) and musharakah (equity
treated as interest-free loans from the clients to the bank
participation), while secondary modes are essentially
and are guaranteed to be repaid in full, it plays the role of
mark-up pricing or leasing arrangements.
an investment manager for the owners of deposits, akin to
a universal bank.
Many scholars also argue that Islamic banking resembles
conventional banking schemes. This includes the claim that
In contrast, conventional banks are understood in a
the majority of Islamic lending has a debt-like character
generic sense as financial intermediaries. Their main task
(Aggarwal and Youssef 1996). El-Gamal (2005) has
is to provide indirect finance, in contrast to direct finance
concluded that Islamic finance is primarily a form of rent-
through financial markets. Banks are understood to
seeking legal arbitrage and simply seeks to replicate the
channel surplus funds from the household sector into the
operations of conventional financial instruments. However,
corporate sector facing a deficit, as they invest more than
some researchers have argued that Islamic financial
internal or direct finance. Banks play a vital role in the
institutions have huge potential over the conventional
economy enabling more productive investment than would
banking model as an alternative finance to absorb macro-
be possible merely on the basis of profits and financial
financial shocks and promote economic growth (Dridi and
market funds. Hence, an asymmetric information and
Hasan 2010; Mills and Presley 1999).

Cite this chapter as: Zulkhibri M (2015). Islamic financing and bank characteristics in a dual banking system: Evidence
from Malaysia. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Zulkhibri

In term of deposit, Islamic banks use mainly the risk- 1. A full-fledged Islamic banking system operating on a
sharing PLS instruments, while in financing, most Islamic parallel basis with a full-fledged conventional system
banks rely on debt-like instruments (mark-up financing (dual banking system).
and a guaranteed profit margin) that are based on deferred 2. A step-by-step approach, in the context of an overall
obligation contracts. Moreover, conventional interest long term strategy.
rates (the London Interbank Offered Rate (LIBOR) or 3. A comprehensive set of Islamic banking legislation
a domestic equivalent) will always be a benchmark for and a common Shariah Supervising Council for all
Islamic banks’ mark-up. As a result, in the case of such debt- Islamic banking institutions.
like instruments, the pricing of Islamic financing is not a 4. A practical and open-minded approach in developing
function of real economic activity but is based on a pre- Islamic financial interests.
determined interest rate plus a credit risk premium.
An important feature in the implementation of Islamic
The objective of the paper is to investigate the determinants banking in Malaysia and creating a viable Islamic banking
of Islamic financing while taking into consideration bank- system is that three basic elements have been adopted:
specific characteristics. Understanding this behaviour
indicates how efficiently Islamic banks perform their roles 1. A large number of instruments and range of different
as suppliers of capital for businesses and entrepreneurs. types of financial instruments must be available to
However, little is known about the determinants of bank meet the different needs of different investors and
financing which operate alongside conventional banks in borrowers.
the dual banking system. Moreover, due to the fact that the 2. A large number of institutions with an adequate
rate of return on retail PLS accounts closely follows interest number of different types of institutions participating
rates offered by conventional banks in Malaysia (Chong in the Islamic banking system to provide depth to the
and Liu 2009; Cervik and Charap 2011), the paper employs Islamic banking system.
a panel-pooled regression methodology by investigating 3. An Islamic interbank market to support an efficient
the cross-sectional differences in the way that Islamic and effective system linking the system to the
banks respond to base financing rates across bank-specific institutions and the instruments.
characteristics.
As it can be observed in Figure 1, the share of Islamic assets
in the overall banking system is growing significantly, from
2.  Overview of Malaysian Islamic financial around 7% in 2006 to 20% in 2012. As at the end of 2012, the
industry country’s Islamic banking system had accumulated a total of
Malaysia’s Islamic finance industry has been in existence RM119 billion in assets, or about 20% of the total assets of
for over 30 years. The enactment of the Islamic Banking the banking sector, which is RM0.6 trillion. To date, Malaysia
Act 1983 enabled the country’s first Islamic Bank to be has 16 Islamic Banks, which comprises nine local Islamic
established. Malaysia’s overall strategy in the development Banks and seven foreign Islamic Banks. Figure 2 shows the
of Islamic banking can be summarized under four pillars: composition of the Islamic financing modes. It shows that

100% 3%
6% 6% 4% 4% 4% 4%
90% 7% 8% 14% 16% 17% 18% 20%
80%

70%

60%

50%
87% 86%
40% 82% 80% 79% 78% 77%
30%

20%

10%

0%
2006 2007 2008 2009 2010 2011 2012

Commercial Bank Islamic Bank Investment Bank

Source: Bank Negara Malaysia (2013)

Figure 1.  Total assets: Islamic and conventional banks.

38 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia

100%

90% 19% 16% 15% 18%

21% 20% 20%

80%

70%

60%
29%
30%
50% 30% 29% 27%
26% 23%
40%

30%

20% 41% 37% 33% 32% 34% 32% 32%

10%

0%
2006 2007 2008 2009 2010 2011 2012

Bai Bithaman Ajil Ijarah Ijarah Thumma Al-Bai Murabahah

Musyarakah Mudharabah Istisna' Others

Source: Bank Negara Malaysia (2013)

Figure 2.  Composition of Islamic financing modes.

the Bai Bithaman Ajil and Ijarah Thumma Al Bai dominate financing and conventional lending rates on loans in
the composition with 32% and 23%, respectively, over the Malaysia. Between 2009 and 2012, the correlation of
period 2006–2012. This dominant trend of using mark-up base lending rate of conventional banks and the Islamic
and debt-like instruments in Islamic financing practices base financing rate was about 76 percent. Accordingly,
support some of the arguments that Islamic banking is akin despite the fact that conventional and Shariah-compliant
to conventional banking in practical terms. Islamic banks operate in different banking environments,
it is surprising that the Islamic base financing rate closely
As shown in Figure  3, the data reveal a high degree tracks interest rates offered by conventional banks in
of correlation between the base financing rate on retail Malaysia.

7.00
Correlation=0.76

6.50

Islamic
6.00

5.50

5.00
Conventional

4.50
2009 2010 2011 2012

Source: Bank Negara Malaysia (2013)

Figure 3.  Islamic rate of return and conventional average lending rate.

Eds. Hatem A. El-Karanshawy et al. 39

 Zulkhibri

3.  Literature review More recent literature on Islamic finance tries to establish the
difference between Islamic rates of return and conventional
Islamic banking is different from conventional banking
banks interest rates based on empirical assessments. Cervik
from a theoretical perspective because interest (riba) is
and Charap (2011) compare the empirical behaviour of
prohibited in Islam (rate of return on deposits cannot be fixed
conventional bank deposit rates and the rate of returns
by the bank and interest cannot be charged on loans). The
on retail Islamic PLS investment accounts in Malaysia and
prohibition of interest is combined with the common belief
Turkey. The findings show that conventional bank deposit
that banks channel funds towards productive investment,
rates and PLS rate of returns exhibit co-integration in the
which makes Islamic banking and Western economic
long-run, and that conventional bank deposit rates cause
theory inconsistent with each other. A unique feature of
returns on PLS accounts. Moreover, the time-varying
Islamic banking is the PLS paradigm, which is largely based
volatility of conventional bank deposit rates and PLS
on the mudaraba (profit-sharing) and musharakah (equity
returns is correlated and is statistically significant.
participation) concepts of Islamic contracting.
Such correlations have been observed in other studies. In
The concepts of Islamic finance in using the rate of returns as
the case of Malaysia, Chong and Liu (2009), for example,
a replacement for interest can be divided into two strands of
find that retail Islamic deposit rates mimic the behaviour
argument. The idealist literature attempts to look at the key
of conventional interest rates. The study shows that only
concepts of Islamic finance such as PLS, money, interest and
a small portion of Islamic bank financing is strictly PLS-
profit from an ideal perspective. A pre-determined return to the
based, and that Islamic deposits are not interest-free,
lender, dependent on the borrowing period and independent
but are very much pegged to conventional deposits. The
of the borrower’s uncertainty, is not permissible under Islamic
findings also suggest that the Islamic resurgence worldwide
banking. This means that the ideal and most ‘Islamic’ form
drives the rapid growth in Islamic banking rather than the
of each concept should be accepted as valid. Much of the
advantages of the PLS paradigm, implying that regulations
literature on Islamic banking and finance in the 1960s and the
similar to those of conventional banks should be applied for
theoretical studies on Islamic banking fall under this category.
the Islamic bank.
Another line of argument based on maslaha-oriented
Similarly, Kasri and Kassim (2009) examined the
literature would be at the extreme end of the continuum.
relationship between investment deposits and rate
According to this view, riba should not be interpreted in a
of return, including interest rate for Islamic banks in
simplistic fashion as modem bank interest. Any interest-
Indonesia over the period 2000 to 2007. Using a vector
based bank could theoretically be an Islamic bank provided
autoregressive model (VAR) model, the study reveals that
that the Islamic ideals of justice, equity, fairness, non-
the mudaraba investment deposit in the Islamic banks are
exploitation were its guiding principles; humane terms of
co-integrated with return of the Islamic deposit, interest
providing finance to those ‘needing’ them were practised;
rate of the conventional banks’ deposit, number of Islamic
and, it provided one way of helping the economically
banks’ branches, and national income in the long-run. The
disadvantaged classes of society to raise their standard of
finding also suggests that rate of return and interest rate
living. Nonetheless, it can be seen from this that there has
move in tandem, indicating that Islamic banks in Indonesia
been a gradual shift from the idealist position to a more
are exposed to benchmark risk and rate of return risk.
pragmatic, mark-up based and less risky version.
In practice, the main explanation of the similarity between
In the conventional literature, the interest rate has long
Islamic bank profit rate and conventional banks can be
been recognized not only by classical and neo-classical
attributed to the differences in perceptions of riskiness
economists, but also by contemporary economists as
(theoretically and practically) at the institutional and
one of the factors that determine the level of savings in
systemic level, particularly on the asset side. In addition,
the economy, and that the interest rate has a positive
Islamic banks lose on the grounds of liquidity, assets
relationship with savings. However, Haron (2001) found
and liabilities concentrations and operational efficiency,
similar positive relationship behaviour for the profit rate
whereas they tend to win in the field of profitability.
declared by Islamic banks. In other words, Islamic bank
Nevertheless, Islamic banking could provide a further
customers are guided by the profit maximization theory
guarantee, albeit still marginal, against systemic risks in
since there is no pre-determined rate of return involved
certain emerging financial markets.
in the Islamic banking system. Since depositors at Islamic
banks possess similar attitudes to those at the conventional
banks, the interest rate will continue to have an influence 4.  Data and estimation methodology
on the operations of Islamic banks. The study employs a panel of annual bank level data of
all Islamic banks operating in the Malaysia covering the
On a similar note, while funding activities are carried out period 2006–2012. The financial statements of Islamic
mainly through the participatory PLS model, it is well- banks operating in Malaysian Islamic banking sectors are
established in the literature that Islamic banks follow their collected from the Bankscope database of Bureau van Dijk’s
conventional counterparts in creating assets through non- company. The macroeconomic variables: consumer price
PLS, debt-like instruments with a pre-determined, fixed index, real gross domestic product, Islamic base financing
rate of return; in line with the findings of Beck et al. (2010), rate, and monetary policy rate are taken from various issues
there are “few significant differences in business orientation, of Quarterly Statistical Bulletin published by Central Bank
efficiency, asset quality or stability” between conventional of Malaysia.
and retail Islamic banks. As a result, given the implicit link
to interest rates on the asset side of the balance sheet, PLS Table  1 reports the basic descriptive statistics for the
rates of return follow conventional bank deposit rates. sample:

40 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia

Table 1.  Descriptive statistics. The choice of bank-specific characteristics is based on

the theoretical assumption that a certain type of bank
# of Std. is expected to be more responsive to financing shocks
Variable Obs. Mean Dev. Min Max since it operates in a dual banking system, and these
characteristics are widely used in the empirical literature.
Financing  69 14.15 1.28  7.41 16.57 Following Gambarcota (2005), the three measures for
Assets  71 14.91 0.95 11.83 16.96 bank characteristics of size (SIZE), liquidity (LIQUIDITY)
Liquidity  70  0.08 0.04 -0.02  0.20 and capitalisation (CAPITAL) are defined as follows:
Capital  66  0.09 0.04 -0.02  0.24
GDP 119 20.32 0.08 20.17 20.42 Size:
Base Financing 119  4.60 0.05  4.51  4.66
Rate
∑
N
ln A it
Overnight Rate 119  2.92 0.46  2.00  3.50 SIZE it = ln A it - i=1

Prices 119  3.29 0.24  2.93  3.70 NT

s
Note: Financing, Assets, GDP and Prices are in logarithmic
forms; Liquidity is defined as a ratio of liquid assets (cash Liquidity:
and short-term funds) to total assets; Capital is defined as
LAit  T ∑ i=1 LAit / Ait 
N
ratio of capital and reserve to total assets.
Source: Author’s own computation from Bankscope. LIQUIDITYit = -∑  /T
Ait  t=1 NT 
ps

Using a panel of information for 17  individual banks, Capitalisation:

we initially estimate the benchmark model for Islamic
bank financing to allow for asymmetric response to bank
K it  T ∑ i=1 K it / Ait 
N

characteristics and monetary policy. This has usually been CAPITALit = -∑  /T
done by introducing interaction terms between Islamic base Ait  t=1 NT 
financing rate and bank discriminatory variables. Beside s
these variables, we control for economic activities and
consumer prices, which allow us to control for demand- Bank size (SIZE) is measured by the logarithm of total
side effects on Islamic bank financing. By combining time assets (A). Relatively, banks with a smaller size may face
series of cross-sectional observation, panel data give more higher constraints in raising external funds, thus forcing
informative data, more variability, less co-linearity among them to reduce their lending (Kasyhap and Stein 1995,
the variables, more degrees of freedom and more efficiency 2000). Liquidity (LIQUIDITY) is measured by the ratio of
(Gujarati and Sangeetha 2007). The panel-data estimation liquid assets (cash and short-term funds) to total assets
method of both pooled-regression and fixed-effect model (LA). More liquid banks can draw down on their liquid
is preferred. Fixed-effects specification is mainly used to assets to shield their financing portfolios and are less likely
account for time-invariant unobservable heterogeneity that to cut back on financing in the face of rising cost or rate
is potentially correlated with the dependent variable. To test of return. Capitalisation (CAPITAL) is measured by the
for estimation robustness of the models, we employ random- ratio of capital and reserve to total assets (K). Since raising
effect estimations and use all diagnostic tests to validate the bank capital is costly, the bank tends to adjust the lending
models. Our baseline model specification is as follows: behaviour to meet the required level of capital. In the face
of a rising rate of return, a bank’s cost of financing rises
l while the remuneration of bank assets remains the same.
ΔFIN it = m i + ∑ β j ΔBFRt- j + γ 0 SIZE it-1 + ω 0 LIQUIDITYit-1 Hence, the financing of highly-leveraged banks is expected
j =1 to be more responsive to changes in the rate of return than
l
the financing of well-capitalised banks (Kishan and Opiela
+ϕ 0CAPITALit-1 + ∑ γ j ΔBFRt- j * SIZE it-1
j =1
2006).
l
+ ∑ ω j ΔBFRt- j * LIQUIDITYit-1 All three criteria are normalised with respect to their average
j =1
l
(NT) across all the banks in the respective sample in order to
+ ∑ ϕ j ΔBFRt- j * CAPITALit-1 get indicators that sum to zero over all observations. For the
j =1 Eqation (1), the average of the interaction term (∆BFR*SIZE,
l l ∆BFR*LIQUIDITY and ∆BFR*CAPITAL) is, therefore, zero
+ ∑κ j ΔGDPt-1 + ∑ l j ΔPRICESt-1 + υ i + e i ,t  and the parameters are directly interpretable as the overall
j =1 j =1
(1) Islamic rate of return effect on Islamic bank financing. To
remove the upward trend in the case of size (reflecting the
where ∆ is the first-difference operator, FIN is the Islamic fact that size is measured in nominal terms), or the overall
banks financing, GDP is the logarithm of real GDP, PRICES mean in the case of liquidity and capitalisation, the bank
is the logarithm of consumer price index, BFR is the Islamic characteristic variables are defined as deviations from their
financing rate, SIZE, LIQUIDITY and CAPITAL are the bank cross-sectional means at each time period.
size, liquidity and capitalisation respectively. The subscript
i denotes banks where i  =  1, …, N; t denotes time where The assumption is that small, less liquid and poorly
t  =  2006–2012; ni denotes individual bank effects and ei,t capitalised banks react more strongly to changes in base
denotes error-term. financing rate. This would correspond to a significant

Eds. Hatem A. El-Karanshawy et al. 41

 Zulkhibri

positive coefficient for the interaction terms ∆BFR*SIZE, the bank lending behaviour. The variable of SIZE is positive
∆ BFR*LIQUIDITY, and ∆BFR*CAPITAL, and means that and highly significant for all models. In the fixed-effect
banks with these characteristics reduce their financing and random-effect model, SIZE is positive and significant,
growth rate more strongly in response to a restrictive shock ranging from 0.51 to 1.39. Larger banks might be more
of base financing rate than do larger, more liquid and well- efficient due to scale economies, while the theoretical and
capitalised banks. empirical literature on the relationship between size and
stability is ambiguous (Beck et  al. 2013). This suggests
Since Islamic banks operate in the dual banking system, that size is an important factor characterising the banks’
conventional interest rates may influence the Islamic financing reaction with large banks being expected to
bank financing behaviour. Equation (1) may represent the minimise cost. This finding is also consistent with Fadzlan
overall effect of Islamic bank financing without monetary and Zulkhibri (2009), who suggest that larger financial
policy. Huang (2003) argues that, under the conventional institutions in Malaysia attain a higher level of technical
system, changes in interest rates have a larger effect on efficiency in their operations and exhibit an inverted
bank loans supplies because banks’ ability to insulate their U-shape behaviour.
financing supplies from changes in monetary policy will be
restricted, in particular, during periods of tight monetary In the case of the liquidity characteristic, the results show
conditions. We try to test this hypothesis for Islamic that the coefficient of LIQUIDITY is positively associated
financing behaviour by including monetary policy rate, and highly significant with bank financing, and is between
where MP is monetary policy shock proxy by overnight 1.16 and 9.47. Only banks that have a larger share of liquid
policy rate in Equation (2) and estimate the following assets, or that are bigger, are able to shield their lending
model: relationships. This evidence points to the fact that Islamic
banks are able to protect their financing portfolios by
ΔFIN it drawing down on their liquid assets and are, therefore, less
l likely to cut on financing, whereas the latter have better
= m i + ∑ β j ΔBFRt- j + γ 0 SIZE it-1 + δ 0 LIQUIDITYit-1 access to external finance due to their size in order to retain
j =1
l their preferred liquidity ratio. This finding also implies that,
+ ϕ 0CAPITALit-1 + ∑ γ j ΔBFRt- j * SIZE it-1 in periods of rising base financing rate, a borrower from a
j =1 less liquid bank, on average, tends to suffer from a sharp
l l
decline in financing more than does a customer of a more
+ ∑ ω j ΔBFRt- j * LIQUIDITYit-1 + ∑ ϕ j ΔBFRt- j * CAPITALit-1
j =1 j =1
liquid bank. The result is in line with the findings by Brooks
l l l (2007) that liquidity is the main determinant explaining
+ ∑ ζ MPt- j + ∑κ j ΔGDPt-1 + ∑ l j ΔPRICESt-1 + υ i + e i ,t credit supply in Turkey.
 j =1 j =1 j =1
(2)
Looking at the coefficient of capitalisation, CAPITAL, it
appears that bank financing is positively associated with
bank capitalisation, or the bank capital structure. The
5.  Empirical results results suggest that market participants may perceive
Table  2 reports the results for our benchmark model of highly capitalised banks as being less risky (Kishan and
Islamic bank financing, while Table  3 to Table  4 report Opiela 2000). Consequently, it should be more expensive
the results from fixed-effect and random-effect. The direct for poorly-capitalised banks to finance externally. Such
impact of changes in the base financing rate on bank poorly-capitalised banks try to avoid the cost of falling
financing is negative and significant. The coefficients for below the regulatory minimum capital requirements or
base financing rate range from 1.78 to 5.47, which means the increased risk of violating the capital requirement by
that an increase of base financing rate by one percentage holding capital buffers and asset buffers (short-term risk-
point leads to a decrease in the bank financing in the range weighted assets rather than customer financing) that
of between 1.7% to 5.5%. The result of our benchmark can be liquidated if the bank runs into problems with the
models in line with the basic theoretical prediction is capital requirement. The more short-term risk-weighted
similar to the bank lending channel of conventional banks assets (other than customer loans) the bank holds on its
(Ehrmann et  al. 2003). Since the Islamic rate of return balance sheet (i.e., the higher the bank’s asset buffer),
implicitly tracks interest rates offered by conventional the lower the risk of violating the capital requirements
banks (Chong and Liu 2009), the results also explain will be. The short-term risk-weighted assets will soon be
that the reduction in Islamic bank deposits may not liquid, thereby reducing the capital requirement in the
be completely substituted by other forms of financing near future. Also, the higher the bank’s capital buffer,
in order to continue to meet financing demand, thus the lower the risk of violating the capital requirement
leading to a reduction in Islamic bank financing. The will be.
results for fixed-effect and random-effect provide similar
observations, albeit with a lower impact of base financing The macroeconomic variables included in the bank
rate on bank financing between -0.21 to -1.76. The financing models control for the demand-side effect, and
estimated regression equations for all models explain only the real GDP growth variable is significant in the
the behaviour of financing in the range of 30% - 97%. All equation, where it has a positive coefficient. The response
diagnostic tests confirm the good fit of the models. of credit to economic activity is consistent with expectation.
The facts that the coefficient of real GDP is significant may
The results from Table  2 to Table  4 also show the imply that the economic activities are taken into account in
importance of bank-specific characteristics with respect to financing decisions in an important way. On the other hand,

42 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia

Table 2.  OLS estimation: Islamic bank financing and characteristics models.

Bank-specific characteristics

ΔFINANCEit SIZE SIZE LIQUIDITY LIQUIDITY CAPITAL CAPITAL

ΔBFRt -1.841*** -1.787*** -3.156*** -3.509*** -4.363*** -5.478***

(0.165) (0.259) 0.272 0.516 (0.155) (0.681)

Bank-specific char.
SIZE 1.110*** 1.113*** – – – –
(0.050) (0.052)
LIQUIDITY 2.282*** 1.811*** – –
– – (1.827) (1.924)
CAPITAL – – – – 3.065*** 2.887***
(0.884) (0.873)
Impact of BFR
ΔBFRxSIZE -0.145*** -0.146*** – – – –
(0.021) (0.021)
ΔBFRxLIQUIDITY – – -4.591*** -4.489*** – –
(0.657) (0.671)
ΔBFRxCAPITAL – – – – -1.806*** -1.032***
(0.550) (0.542)
Impact of MP
ΔONt – 0.053*** – 0.052*** – 1.237***
(0.194) (0.194) (0.737)
Control Variables
ΔGDPt 1.495* 1.142* 6.227* 8.363* 8.636*v 9.778**
(0.635) (0.710) (4.329) (4.969) (2.115) (7.256)
ΔPRICESt -1.311 -1.828 -6.205 -9.292 -2.899 -2.277
(4.105) (4.561) (9.318) 10.106 (16.869) (18.852)
R-square 0.848 0.848 0.332 0.341 0.771 0.664

Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticity-
consistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price index;
*significant at 10%, **significant at 5%, and ***significant at 1%.

the price variable is negatively related to bank financing, specifications. This finding broadly supports the findings
but is insignificant. The rise in inflation may be associated by Chong and Liu (2009), Cervik and Charap (2011),
with the variability of the inflation rate and will generate and Beck et  al. (2013) that there is no significant
uncertainty about the future return on investments. This, difference in bank financing behaviour with respect to
in turn, discourages firms from undertaking investments interest rates. Furthermore, Kasri and Kassim (2009)
which, consequently, reduces their financing demand. confirm that the conventional interest rate is one of
However, the price variable is insignificant for the results the determinants for saving deposits in Indonesia. This
of all regressions. evidence explains why the bulk of Islamic bank financing
is based on the mark-up principle and is very debt-like
Due to the potential interrelations between bank in nature (i.e. murabaha and ijarah), rather than using
financing and conventional interest rate, all bank the principle of PLS. Despite their operations being
financing models are run with overnight policy rate different from those of conventional banks, Islamic
(ON). The coefficients in all regressions are negatively banks seem to face asymmetric information, severe
related to bank financing and vary within a reasonable adverse selection and moral hazard problems similar to
magnitude (0.05 to 0.52), but are broadly lower than those of their counterparts in their attempts to provide
the base financing rate. The results of these regressions funds to entrepreneurs. However, the use of debt-like
suggest that the reaction of banks to changes in interest instruments is a rational response on the part of Islamic
rates remains the same as the change base financing rate banks to informational asymmetries in the environments
and is robust in terms of a different type of econometric in which they operate.

Eds. Hatem A. El-Karanshawy et al. 43

 Zulkhibri

Table 3.  Fixed effect estimation: Islamic bank financing and characteristics models.

Bank-specific characteristics

ΔFINANCEit SIZE SIZE LIQUIDITY LIQUIDITY CAPITAL CAPITAL

ΔBFRt -0.208*** -0.294*** -0.470*** -1.764*** -0.820*** -0.339***

(0.128) (0.194) (0.415) (0.516) (0.171) (0.186)
Bank-specific char.
SIZE 0.595*** 0.508** – – – –
(0.036) (0.554)
LIQUIDITY 1.792*** 1.161** – –
– – (1.894) (0.815)
CAPITAL – – – – 6.067** 8.815**
(2.354) (2.526)
Impact of BFR
ΔBFRxSIZE -0.222** -0.021** – – – –
(0.010) (0.159)
ΔBFRxLIQUIDITY – – -1.605** -1.458* – –
(0.511) (1.108)
ΔBFRxCAPITAL – – – – -3.011** -3.671**
(1.599) (1.335)
Impact of MP
ΔONt – -0.043*** – -0.528* – -0.368**
(0.017) (0.317) (0.169)
Control Variables
ΔGDPt 1.193* 1.538* 5.170* 5.021* 5.386** 8.222**
(0.746) (1.207) (3.116) (3.410) (2.494) (4.319)
ΔPRICESt -0.844 -0.357 -0.361 -0.305 -0.980 -0.229
(2.197) (2.385) (7.153) (7.446) (7.967) (9.035)
Constant 7.588*** 8.784 10.902*** 19.164*** 18.058*** 17.516***
(0.357) (7.917) (0.937) (16.003) (2.494) (2.553)
Firm Dummy Yes Yes Yes Yes Yes Yes
R-square 0.972 0.972 0.738 0.751 0.735 0.743
F-statistic 72.29 66.77 5.989 5.856 5.231 4.995

Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticity-
consistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price
index; *significant at 10%, **significant at 5%, and ***significant at 1%.

To analyze further the impact of banks reducing their significant at the conventional level. These results suggest
bank financing in response to a change in base financing that banks’ lending and their ability to obtain other sources
rate with respect to specific characteristics, we have of funding are factors that are affected indirectly through
interacted the base financing rate variable with bank bank-specific characteristics.
size (∆BFR*SIZE), liquidity (∆BFR*LIQUIDITY) and
capitalization (∆BFR*CAPITAL); this is to investigate the In summary, we find strong evidence of the asymmetric
economic arguments that there is a unique role for Islamic adjustment of bank financing through bank-specific
banks in the dual banking system, and the importance of characteristics, as reported in the literature for conventional
heterogeneity among banks. Tables 2, 3 and 4 also report banks and in line with the arguments of Kashyap and Stein
the results of the base financing rate with respect to bank- (1995, 2000) and Kishan and Opiela (2000). Moreover,
specific characteristics for bank financing using a fixed-effect banks react differently to the base financing rate depending
and random-effect model. The estimates of bank-specific on their own specific characteristics; a bank with higher
characteristics coefficients provide interesting results. The capitalisation, in particular, is expected to increase
estimate coefficients of BFR*LIQUIDITY, BFR*CAPITAL and financing more than a bank with greater size and liquidity.
BFR*SIZE consistently show a positive sign, and are highly Furthermore, since an Islamic bank is operating in a dual

44 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia

Table 4.  Random effect estimation: Islamic bank financing and characteristics models.

Bank-specific characteristics

ΔFINANCEit SIZE SIZE LIQUIDITY LIQUIDITY CAPITAL CAPITAL

ΔBFRt -0.368*** -0.324*** -0.452*** -0.334*** -0.377*** -0.371***

(0.028) (0.025) (0.214) (0.215) (0.142) (0.105)
Bank-specific char.
SIZE 0.618*** 1.399*** – – – –
(0.035) (0.496)
LIQUIDITY 9.471*** 4.639** – –
– – (1.805) (2.858)
CAPITAL – – – – 7.307*** 8.521***
(2.284) (2.331)
Impact of BFR
ΔBFRxSIZE -0.221** -0.244** – – – –
(0.010) (0.143)
ΔBFRxLIQUIDITY – – -1.539*** -0.189*** – –
(0.102) 0.075
ΔBFRxCAPITAL – – – – -3.406*** -3.667**
(1.129) (1.777)
Impact of MP
ΔONt – -0.041** – -0.472** – -0.323***
(0.021) (0.231) (0.036)
Control Variables
ΔGDPt 1.222* 1.423* 5.365* 9.664* 6.131* 8.530*
(0.842) (1.112) (4.104) (4.380) (3.517) (5.198)
ΔPRICESt -0.657 -1.394 -0.698 -0.808 -3.444 -5.773
(2.191) (0.502) (7.112) (7.406) (2.066) (8.947)
Constant 7.101*** 4.052 10.873*** 13.830 17.810*** 17.412
(0.346) (7.067) (0.861) (14.852) (2.297) (2.340)
Firm Dummy Yes Yes Yes Yes Yes Yes
R-square 0.908 0.904 0.394 0.413 0.400 0.409
F-statistic 93.58 72.83 6.256 5.515 5.455 4.968

Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticity-
consistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price
index. *significant at 10%, **significant at 5%, and ***significant at 1%.

banking system, the asymmetric information problems contingent on observable bank-specific characteristics.
faced by Islamic banks are expected to affect the ability to Understanding this mechanism is crucially important, in
protect the financing lines from policy-induced reductions the context that Islamic banks have increasingly played a
in deposits, and result in Islamic bank financing behaviour. dominant role in the Malaysian financial system.

This paper analyses the importance of bank-specific

6. Conclusion characteristics with respect to Islamic bank financing
A significant number of empirical studies have explored the in Malaysia. The results obtained from pooled panel
bank lending behaviour of conventional banks over the past estimation allow us to draw several significant conclusions
decades, while studies on Islamic bank financing behaviour about Islamic bank financing behaviour in Malaysia within
remain scarce due to lack of bank-level data. Under the dual a dual banking system. The evidence gathered in this study
banking system, to the extent that financial constraints vary suggests that bank-specific characteristics are important
with banks’ ability to access other sources of financing, the for Islamic banking financing behaviour. The financing
implication is that Islamic bank financing responses to the behaviour of Islamic banks is consistent with the behaviour
bank financing rate and the conventional interest rate are of conventional banks in that the bank lending operates

Eds. Hatem A. El-Karanshawy et al. 45

 Zulkhibri

according to banks’ size, and level of liquidity and capital El-Gamal M. “Mutuality as an Antidote to Rent-Seeking
(Goldniuk 2006). The results of these regressions also Sharia-Arbitrage in Islamic Finance”. http://www.ruf.
suggest that the reaction of Islamic banks financing to rice.edu/~elgamal/. 2005.
changes in interest rates is the same as for conventional
Ehrmann M, Gambacorta L, Martinez-Pages J, Sevestre P,
banks, and are robust to different types of econometric
Worms A. “Financial Systems and the Role of Bank in
specifications.
Monetary Transmission in the Euro Area”. In I. Angeloni,
A. Kashyap, and B. Mojon, (eds.), Monetary Transmission
Many problems and challenges relating to Islamic
in the Euro Area: A Study by the Eurosystem Monetary
instruments, financial markets, and regulations must
Transmission Network, Cambridge University Press,
be addressed and resolved. A complete Islamic financial
2003. 235–269.
system with its identifiable instruments and markets is
still at a relatively early stage of evolution. The functioning Fadzlan S, Zulkhibri M. “Post-Crisis Productivity Change
of Islamic banks should rapidly differentiate itself from in Non-Bank Financial Institutions: Efficiency Increase
conventional banking. Due to the existence of moral or Technological Progress?”Journal of Transnational
hazards and adverse selection in the industry, an Islamic Management, 14(2), 2009. 124–154.
bank is not able to provide a full-fledged alternative
Haron S. “Islamic banking and finance.” Leading issues in
finance to conventional finance. Moreover, an Islamic bank
Islamic Banking and Finance, 20 (1), 2001. 17–32.
does not develop in the path that was envisioned by the
Islamic scholars (Saeed 1996). One of the drawbacks is Huang Z. “Evidence of Bank Lending Channel in the U.K.”
the low level of participation in PLS arrangements, which Journal of Banking and Finance, 27, 2003. 491–510.
seems contradictory to the essential concept of Islamic
Gambacorta L. “Inside the Bank Lending Channel.”
banking. In practice, it would beneficial for Islamic banks
European Economic Review, 49, 2005. 1737–1759.
to stop replicating the conventional banking models that
concentrate mainly on debt-based instruments and mark- Gujarati DN, Sangeetha S. “Basic Econometric.” 4th Edition,
up models, but instead to move over to the PLS model. McGraw-Hill Education Books Ltd., India, 2007.
Golodniuk I. “Evidence on the Bank-Lending Channel
References in Ukraine.” Research in International Business and
Finance, 20, 2006. 180–99.
Aggarwal R, Yousef T. “Islamic Banks and Investment
Financing.” Journal of Money, Credit, and Banking, Kasri RA, Kassim, S. “Empirical Determinants of Saving of
32(1), 2000. 93–120. the Islamic Banks in Indonesia.” Journal of King Abdul
Aziz University, Islamic Economics, 22(2), 2009.
Bank Negara Malaysia, Quarterly Bulletins, Kuala Lumpur.
Khan M, Mirakhor A. “Islamic Banking”. The New Palgrave
Beck T, Demirguc-Kunt A, Merrouche O. “Islamic vs.
Dictionary of Money and Finance, London: Macmillan
Conventional Banking: Business Model, Efficiency and
Press, 1992.
Stability.” Journal Banking and Finance, 37(2), 2013.
433–447. Kashyap A, Stein J. “The Impact of Monetary Policy on Bank
Balance Sheets.” Carnegie Rochester Conference Series
Brooks PK. “Does the Bank Lending Channel of Monetary
on Public Policy, 1995. 51–195.
Transmission Work in Turkey?” IMF Working Paper
07/272, Washington DC: International Monetary Fund. Kashyap A, Stein J. “What Do a Million Observations Say
2007. About the Transmission Mechanism of Monetary Policy.”
American Economic Review, 90(3), 2000. 407–428.
Cervik S, Charap J. “The Behavior of Conventional and
Islamic Bank Deposit Returns in Malaysia and Turkey.” Kishan P, Opelia T. “Bank Capital and Loan Asymmetry
IMF Working Paper, No. WP/11/156, Washington DC: in the Transmission of Monetary Policy.” Journal of
International Monetary Fund, 2011. Banking and Finance, 30, 2006. 259–285.
Chong B, M-H Liu. “Islamic Banking: Interest-Free or Khan M. “Time Value of Money and Discounting in Islamic
Interest-Based?” Pacific- Basin Finance Journal, 17, Perspective.” Review of Islamic Economics, Vol. 1, No.
2009.124–144. 2, 1991. 35–45.
Chapra MU. “The Future of Economics: An Islamic Mills P, Presley J. “Islamic Finance: Theory and Practice.”
Perspective.” The Islamic Foundation, Leicester, UK, London, UK: Macmillan, 1999.
2000.
Saeed A. “Islamic Banking and Interest: A Study of
Dridi J, Hasan M. “Have Islamic Banks Been Impacted the Prohibition of Riba’ and its Contemporary
Differently than Conventional Banks During the Recent Interpretation.” Leiden: EJ Brill, 1996.
Global Crisis?” IMF Working Paper, No. WP/10/201,
Washington DC: International Monetary Fund, 2010.

46 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Leverage risk, financial crisis, and stock returns:
A comparison among Islamic, conventional, and
socially responsible stocks
Vaishnavi Bhatt1, Jahangir Sultan2
Ramaiah Institute of Management Studies, Banglaore, India, Email: [email protected]
1

The Hughey Center for Financial Services, Bentley University, Email: [email protected]
2

We thank Marcia Cornet, Vishwanathan Iyer, Kartik Raman, and seminar participants at Bentley University for their
useful suggestions. We thank Dow Jones Indexes for providing us with information on proprietary indexes. We remain
responsible for all remaining errors. Please do not quote without permission.

Abstract - According to the financial press, firms with low leverage have lower distress risk due
to their reduced exposure to the credit market, especially during credit crises. Compared to their
conventional and socially responsible (SRI) counterparts, Shariah compliant (SC) stocks are
low-leverage stocks. Our hypothesis is that SC firms would be less sensitive to leverage risk and
thus would be ideal for wealth preservation during declining market environment. We find that
the leverage risk factor performs consistently across various categories of firms and its impact is
more pronounced during the recent financial crisis. However, we also find that compared to the
conventional stocks, SC stocks are also quite sensitive to the leverage factor. In contrast, the SRI
class of stocks has the least sensitivity to leverage risk factor, suggesting they can be attractive for
wealth preservation during credit crises.

Keywords: asset pricing, leverage, returns, financial crisis

JEL Codes: G00, G01, G10, G11, G12, G32

1. Introduction the return on low leverage stocks) is a non-diversifiable risk

premium and therefore should be included in any multifactor
As the divine code of law, the Shariah is a code of conduct
asset pricing model. The evidence that high leverage requires
that guides business transactions for the Muslims and are
higher risk premium can be indicative of the notion that high
based on the Quran and the edicts of Prophet Muhammad
leverage can be value destructive, especially when equity
(pbuh). Hence, the guidelines set forth in the Shariah
prices are falling in a persistent fashion. The fact that Islamic
become imperative to every Muslim and govern all aspects
stocks may have lower credit market exposure is important
of life, whether they may be of personal, social, political,
for wealth preservation during both good and bad times.
economic or financial nature. Shariah compliant (SC)
Milly and Sultan (2009) report that Islamic stocks listed
stocks1 are low-leverage stocks with high asset backing,
globally have outperformed conventional stocks and SRI
compared to their conventional and socially responsible
stocks during the 2007–2009 economic crisis3. It would
(SRI) counterparts. It is a widely held belief that firms
be interesting to examine how these stocks respond to the
with low leverage have lower distress risk due to their
traditional risk factors (such as market risk premium, size,
reduced exposure to the credit market. Naturally, these
and value) as well as the leverage risk factor. If indeed
firms are capable of promoting flight to safety, especially in
Islamic stocks have lower sensitivity to the leverage risk
a declining market environment.
factor, it would be indicative of their attractiveness for wealth
preservation when investors are looking for safer assets.
In this paper, we examine if SC stocks have lower sensitivity
to economy wide leverage risk. To this extent, we create a
In this paper, using a sample of 3704 globally traded stocks
new leverage risk factor (LEV) on the basis of firm-specific
for the period January 2000- April 2009, we construct a risk
financial leverage (total debt over assets)2. The risk factor
factor based on firm-specific leverage find that the inclusion
LEV (defined as the return on high leverage stocks minus

Cite this chapter as: Bhatt V, Sultan J (2015). Leverage risk, financial crisis, and stock returns: A comparison among
Islamic, conventional, and socially responsible stocks. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Bhatt and Sultan

of the leverage risk factor leads to a weakening of the and Vassalou and Xing (2004)). A common finding in the
significance of the traditional FF variables. Furthermore, literature is that value stocks earn a premium over growth
we show that, in comparison to the traditional FF factors, stocks. Similarly there is evidence that small sized stocks
the economic and statistical significance of the leverage earn a premium over big stocks.
risk factor is high, especially during the financial crisis. We
also demonstrate that the leverage risk factor contributes to These so-called empirical anomalies continue to generate
the systematic risk of a firm and represents the underlying controversies in the literature. For instance, are value and
macroeconomic fundamentals. Finally, we show that size premiums caused by the underlying risk factors of
compared to the conventional stocks, SC stocks display firms falling within these categories? Similarly, the notion
substantially lower risk premium to traditional risk factors. of whether value and size premiums reflect incorrect
We also find that similar to the conventional stocks, Islamic extrapolation of past earnings growth by the market and
stocks are also sensitive to the leverage factor, thus leading subsequent correction of the mispricing errors, continues
us to suggest that a leveraged based screening of Islamic to receive attention in the literature (see Eom and Park
stocks may not be ideal for wealth preservation especially (2008) for a recent survey).
during a credit crises. An investor must search for other
redeemable characteristics in Islamic stocks that can help How well do FF risk factors capture financial distress risk?
preserve equity value during falling equity prices. Fama and French (1992) note that the combination of book
to market and size describes the cross-section of average
The remainder of the paper is as follows. In Section II, we stock returns and absorb the apparent roles of other variables
review the link between leverage and stock returns. In like leverage and E/P. The authors note that the SMB and
Section III, we discuss the recent financial crisis to motivate HML factors are correlated with leverage and, therefore,
the empirical model. In Section IV we offer empirical adequately represent financial distress. The ability of the
results, and the final section concludes the paper. traditional FF factors to directly capture leverage risk is
critical for asset management, especially when leverage
risk becomes a source of systemic risk in the economy. The
2.  Review of literature implication for an investor facing such catastrophic shocks
A detailed analysis of the sensitivity of SC stocks to is simple. If size and value based strategies do not perform
the leverage risk is tricky. In the first place, one must consistently well across good and bad times, the rationale
demonstrate that, in the context of a multifactor asset behind such investing strategy is at risk.
pricing model, the previous risk factors are incapable of
capturing economy wide leverage risk. Once a reliable risk However, Fama and French (1992 and 1993) deal with the
factor is constructed, a researcher can proceed to the next market leverage (assets over market value of equity) and
stage to investigate whether such risk factor is significant the book leverage (assets over book value of equity), which
in an asset pricing model. Finally, the analysis can proceed may not directly capture the sensitivity of the firms to
to examine if there are differences in the way different economy wide leverage risk4. In particular, the debt market
categories of firms respond to this newly created risk factor. exposure of a firm is a major determinant of the distress
risk that may not be directly captured by the FF factors.
Consider the following multifactor asset pricing model Furthermore, to the extent that excessive leveraging and
(Fama-French (1992)) major credit events can lead to correlated defaults, we
may find that the debt market exposure is monotonically
rt - r ft = b 0 + b 1(rmt - r ft ) + b 2 Rt ,SMB + b 3 Rt ,HML + e t (1) increasing in financial leverage. In essence, the resulting
  credit crisis produces contagion-like effects with leverage
risk as being the primary catalyst. According to Fama and
shows that excess return on a portfolio (rt – rft) is explained French (1996), if default risk becomes correlated across
by the sensitivity of its return to three factors: the excess firms, market participants, especially workers in distressed
return on a broad market portfolio (rmt – rft); the difference firms, tend to avoid all distressed firms in general. We
between the return on a portfolio of small stocks and the believe that this presents an ideal opportunity for volatility
return on a portfolio of large stocks (SMB, small minus spillover among firms in the economy, with the extent of
big); and the difference between the return on a portfolio spillover monotonically rising in leverage.
of high-book-to-market stocks and the return on a portfolio
of low-book-to-market stocks (HML, high minus low). Surprisingly, very few studies have empirically examined
the role of leverage risk factor in asset pricing. Chan and
Our analysis thus leads us to first address an important Chen (1991) examine the effects of financial leverage
question which has largely been ignored in the literature. (book value of debt and preferred stock over market value
Fama and French (1992) note that SMB (return on a of equity) on stock returns and find a positive relationship.
portfolio of small firms minus the return on a portfolio of Unfortunately, their analysis does not investigate if factor
large firms) and HML (return on a portfolio of high book loadings on the financial leverage can subsume the effects
to market firms minus return on a portfolio of low book of HML and SMB. As Fama and French (1992) write, “It
to market firms) are statistically important in explaining would be interesting to check whether loadings on their
the cross-section of equity returns. Subsequent work by distress factors absorb the size and book-to-market equity
academics and practitioners has sought to verify the effects effects in average returns documented here.” Ferguson and
of these factors (FF factors, from hereafter) on cross-section Shockley (20003) write, “... a three-factor empirical model
of equity returns (for example, see Fama and French (1993, that includes factors based on relative leverage and relative
1995, and 1998), Liew and Vassalou (2000), Davis, Fama distress should outperform the Fama and French (1993)
and French (2000), Sivaprasad and Muradoglu (2009), three-factor model in the cross section”.

48 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

An investigation into this topic is timely given the recent of debt financing (Jensen and Meckling (1976) and Fama
financial crisis when economy-wide leverage played a key and Jensen (1983)); and a reduction of the manager’s ability
role in exacerbating the risk exposure especially for the to waste free cash (Jensen (1986)). Overall, these studies
leveraged5 financial and non-financial firms. As the subprime imply that debt reduces agency costs and managerial waste,
crisis deepened, coupled with escalating liquidity crisis, the improves disclosure, and thus reduces equity risk premium.
credit market virtually dried up, limiting access to funds. As a result, leverage is decreasing in stock returns.
The TED spread (difference between the interest rates on
Eurodollar loans and short-term U.S. T-bill) rose in July 2007, The previous discussion suggests that the leverage risk
then spiked even higher in September 2008, reaching as high factor is important for asset pricing models. Our focus in
as 4.65% on October 10, 2008. While the impact was felt this paper is to examine the extent to which the well-known
mostly by the hedge funds, insurance agencies, banks, and anomalies (size and book to market effects) are resolved
firms directly involved in construction business and mortgage by directly adding leverage as a systematic risk factor.
lending, the effects of the liquidity crisis also had affected the Leverage risk becomes fundamental risk especially when
non-financial firms as well. Thus, the financial crisis in 2007– firms’ exposure to the debt market becomes pervasive and
2008 had a devastating contagion-like effect on credit risk, correlated across the economy. Fama and French (1996)
with leverage risk acting as the centrepiece. An analysis of the recognize that investors avoid financially distressed firms
Islamic stocks and their conventional counterparts is critical because distress risk is correlated across the economy.
from the point of view of academic as well as the practitioner We suggest that when leverage risk becomes correlated
community. If Islamic stocks have lower sensitivity to the across the economy, it has a contagion-like effect on firms
leverage risk factor, then these stocks would be ideal for in general, especially those with high exposure to the
wealth management, especially during financial crises. debt market. To this extent, while size and book to market
factors are correlated with the leverage of the firm, they
There are several studies on the relationship between may not adequately capture the firm’s direct exposure to
leverage and stock returns. See Chou, Ko, and Lin (2010) the economy wide systemic risk due to excessive leverage.
for a recent survey. In one strand of the literature, leverage Finally, to the extent that Islamic stocks tend to have low
is positively related to stock returns, especially for weak leverage and are involved only in permissible economic
firms with poor investment opportunities. Accordingly, as activities under the guidelines of the Quran and Sunnah,
debt increases the risk exposure of such firms, investors may have reduced exposure to interest rate volatility.
demand a premium. Sivaprasad and Muradoglu (2009) This simple and powerful proposition has not been fully
find that leverage has a significant positive relation with addressed in the literature. If Islamic stocks continue to act
stock returns. Gomes and Schmid (2009) show that equity like their conventional counterparts, it only goes to reaffirm
returns are increasing in market leverage. Ho, Strange the harmful effect of riba as firms take on more debt.
and Piesse (2008) conduct a similar study for the Hong
Kong stock exchange and conclude that market leverage Our suggestion is consistent with the anecdotal evidence
(Assets/Market value of equity) exhibit a significant from the recent financial crisis when leverage risk became
conditional relationship with the stock returns. Bhandari one of the primary drivers of the global economic crisis.
(1988) performs cross sectional regressions between There was plenty of evidence of such systemic risk in the
monthly average returns and the leverage ratios for the recent financial crisis: debt markets such as the commercial
period 1948–1979 and finds that the debt equity ratio has paper market, the repo market, and short-term bank
a positive effect on stock returns. Ferguson and Shockley borrowing virtually dried up. Altogether, increased leverage
(2003) include relative leverage (D/E) and relative of firms, especially of hedge funds, insurance agencies,
distress risk, based on Altman’s Z score. They find that banks, and mortgage companies, coupled with a liquidity
their model performs better than the three factor FF model crisis, took a heavy toll on the global economy.
in explaining stock returns. On similar lines, Chou et  al
(2010) propose an augmented five factor model which In the next section, we discuss the link between leverage
incorporates both FF factors as well as Ferguson and risk factor and selected macroeconomic variables such as
Shockley factors and demonstrate that this augmented five the industrial production, unemployment, inflation, credit
factor model explains most of the asset pricing anomalies. spread and term spread. Our intent is to draw inferences on
the effects of the leverage risk factor on stock returns across
In contrast, there are several studies that offer rationales various time periods.
for supporting a negative relationship between financial
leverage and stock returns. The debt-overhang theory
(Meyers, 1977) provides a convenient framework to suggest 3.  Leverage risk and the financial
why leverage reduces equity return. Accordingly, as leverage crisis—contemporary evidence
increases, the distress risk increases, and shareholders pass In 2004, the US Securities and Exchange commission
up positive NPV projects. As a result, the stock price decreases, granted a waiver of the international standards of
reflecting underinvestment in successful projects and a maximum accounting leverage ratio6 (which was about
decline in firm value (Meyers (1977)). Other explanations 12) for five major securities firms – Goldman Sachs, Merrill
include firms substituting debt for equity especially during Lynch, Morgan Stanley, Lehman Brothers and Bear Sterns.7
economic crisis when the cost of equity financing is higher Subsequently, many of the investment banks boosted their
than the cost of debt financing (Dimitrov and Jain (2006)); leverage ratios to as high as 30. Mortgage giants Freddie
managerial preference for equity over debt because high Mac and Fannie Mae had leverage levels close to 60 to 1
debt payments can reduce equity returns, especially when (2008 data), which can be very lucrative if the asset prices
firms do not take advantage of growth opportunities (Lang, rise, but is disastrous when asset prices fall. A recent
et al (1995)); the benefit of external disciplining mechanism report8 cites excessively high leverage ratios prevailing in

Eds. Hatem A. El-Karanshawy et al. 49

 Bhatt and Sultan

the housing market and the underlying mortgage backed of 15% to 25%. Such macroeconomic instability has the
securities as the culprit behind the credit crisis. Towards potential to push investors away from the stock and bond
the end of the year 2009, the global economy was afflicted markets. Furthermore, an increase in the perceived risk in
with excessive indebtedness which adversely affected the the financial markets would prompt investors requiring a
worldwide economy. For example, average household higher risk premium, which directly affects the expected
sector debt increased 141 per cent of disposable income in returns on these stocks. So, deleveraging could have
the United States and 177 percent in the United Kingdom. negative effects and is expected to reduce productivity.
Furthermore, the best known banks in the US and Europe Overall, leverage affects expected returns not as a firm
had their leverage (assets/equity) rising to forty, sixty or specific variable but as a systematic risk factor.
even hundred times the size of their equity capital.9

There is a broad consensus that increased leverage affects 4.  Empirical results
stock returns during the financial crisis. According to the Our initial sample includes weekly data for approximately
popular press10, under normal circumstances where stock 4000 stocks from 55 countries from January 2000 to April
prices deviate from their underlying fundamentals, prices 2009. Our sample includes both financials (banks, S&Ls,
tend to bounce back to their intrinsic values, thereby credit unions, mortgage financing companies, real estate
restoring the efficiency of the equity markets. However, firms, and insurance companies) and non-financial firms.
during a prolonged crisis, price discovery process takes Since financial firms, especially banks and insurance firms,
longer, and stocks move away from their intrinsic values operate with high leverage, we will also separate financials
for a longer period of time. In addition, when investors from the aggregate sample to examine if financials stocks
are pessimistic about the financial markets, they may miss have different sensitivity to the risk factors.
out on profitable arbitrage opportunities as prices move.
In fact, due to the significant mispricing in the market, the We eliminate stocks having negative book to market equity
US subprime crisis caused share prices of various US and from the sample in the construction of the risk factors.15 Also,
European banks to fall and exerted immense pressure on the number of stocks each year used in the construction of
these banks in the form of deteriorating profit margins.11 factors varies depending on the availability of data for
the corresponding year. This eliminates the problem of
From a balance sheet perspective, companies reduce their survivorship bias in the sample. The data for the weekly
leverage ratios either by selling off their assets (thereby stock returns are extracted from Datastream, while the data
restructuring their balance sheets) or by issuing new related to economic fundamentals like size, book to market
shares. Both of these strategies have different implications equity and leverage are extracted from FactSet. Stock returns
on the expected returns from the investor perspective. are in US dollar terms and are based upon log relatives of
According to James Lee, Vice Chairman of JP Morgan,12 in weekly stock prices. The Dow Jones Global Index is used as
spite of the efforts by the financial sector to augment their the market benchmark, and the US risk-free rate is used as
capital levels to as high as $300 billion firms have not been a proxy for global risk free rate16. We use previous year-end
able to bring down the leverage to pre-crisis levels. fundamentals to form portfolios for each successive year;
the rationale behind this is that investors use information
While many financial institutions and asset managers contained in the balance sheets and financial statements to
have been deleveraging since 2008, the process might predict future returns. Investors are assumed to follow a buy
eventually diminish the ability of these institutions to and hold policy with annual portfolio rebalancing.
produce attractive returns, especially when they are unable
to grow their balance sheets. In such circumstances, as
financing gets costlier, firms focus on augmenting their Construction of risk factors
capital level rather than investing it. In this process—“The We sort all stocks in the sample by size, book to market and
big get bigger and the rest get smaller”13—has a direct leverage and categorize them in 3 groups (top 30%, middle
impact on the stock returns of these firms. In other words, 40%, and bottom 30%). Using the independent sorting
higher leverage levels increase the risk exposure of the procedure we construct value weighted portfolios formed
firms and present higher growth opportunities, which by the intersection of three portfolios based on size, three
should lead to higher stock returns. In contrast, lower portfolios based on book to market equity and three portfolios
leverage levels shrink the balance sheet of the firm and also based on leverage (Debt/Assets). In all, we have 3*3*3 = 27
reduce their competitiveness, having a negative impact on portfolios. The returns on these annually rebalanced
the shareholder value and stock returns. portfolios create the dependent variable. In addition to
the XMKT (market risk premium), the FF factors are: SMB
Leverage risk during the financial crisis has macroeconomic (size mimicking portfolio constructed each week by taking
implications. Notwithstanding the de-leveraging efforts of the simple average of the returns on small sized portfolios
banks and other financial institutions, as of November 6, minus returns on big sized portfolios), HML (book to market
2009, banks in particular exhibited 40 to 1 leverage (assets mimicking portfolios constructed each week by taking the
over equity capital). Similarly, the deleveraging efforts simple average of the returns on high book to market portfolios
undertaken by many governments have also led to adopting minus the returns on low book to market portfolios) and
restrictive monetary policy, resulting in higher interest LEV (leverage mimicking portfolios constructed each week
rates. However, analysts argue that increasing interest by taking simple average of the returns on high leveraged
rates and withdrawing funds from the financial system may portfolios minus the returns on low leverage portfolios).
cause the economy to exacerbate the effects of the credit
risk. It has also been forecasted14 that deliberate attempts Table 1 reports the number of stocks used for the construction
by the governments to deleverage will lead to lower wages of factors and portfolios each year which varies depending
in developed countries and a permanent unemployment on the availability of data and meeting specific requirements

50 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 1 A. Number of stocks. Table  3. Descriptive statistics of the returns on Market

factor, SMB, HML, and LEV factors.
Year No. of stocks
XMKT SMB LEV HML
2000 3745
2001 4344 Mean -0.002 0.002 0.000 0.001
2002 4378 Median 0.001 0.002 0.000 0.001
2003 4379 Maximum 0.115 0.054 0.045 0.053
2004 4396 Minimum -0.221 -0.063 -0.035 -0.049
2005 4403 Std. Dev. 0.027 0.017 0.008 0.013
2006 4399 Skewness -1.398 -0.272 -0.077 0.127
2007 4399 Kurtosis 13.623 4.883 6.630 6.155
2008 4406
2009 4391 XMKT is defined as rm-rf where rf is the return on the risk
free asset and rm is the return on the market portfolio. SMB
is the return on the size mimicking portfolio constructed
by taking the simple average of the returns each week
(for e.g. positive book to market equity). The correlation
of all “small” portfolios minus “big” portfolios. HML
matrices for the sample across the three periods are reported
is the return on book to market mimicking portfolio
in Tables 2A-2C. For the aggregate period, there is a positive
constructed by taking the simple average of the returns
correlation of .45 between HML and LEV, which is expected
each week of all “high BE/ME” portfolios minus “low BE/
since both these factors closely represent the distress
ME” portfolios. LEV is the return on leverage mimicking
risk of the firm. In Table  2B, for the non-crisis period, the
portfolios constructed by taking the simple average of the
correlation coefficients are as follows: 0.407 (LEV, HML),
returns each week of all “high leverage” portfolios minus
-.1451 (LEV, XMKT), and .205 (LEV, SMB). In Table 2C, for
“low leverage portfolios”.
the crisis period, there are some interesting changes. For
example, the correlation between LEV and HML increases
further, and the correlation between LEV and XMKT actually
turns positive. Finally, the correlation between LEV and SMB
actually turns negative. Table  3 reports summary statistics Macroeconomic variables and factor loadings
for the risk factors, XMKT, SMB, HML and LEV. In this section we demonstrate that the FF and leverage
risk factors have macroeconomic implications. Several
studies have shown that macroeconomic variables predict
expected returns on stocks and bonds. See for example,
Table  2A. Correlation matrix of the explanatory factors Abel (1999), Fama (1981), Elton, et. al. (2001), Vassolou
for—all stocks (Jan 2000–April 2009). (2003) and Petkova (2006) and references therein. These
studies show a significant positive relationship between the
XMKT SMB LEV HML excess market returns and indicators of economic growth.
We extend this analysis and test for the relationship
XMKT 1.000 -0.087 0.031 -0.057
between selected macroeconomic variables and returns on
SMB -0.087 1.000 -0.030 0.086 SMB, HML, and LEV factors.
LEV 0.031 -0.030 1.000 0.451
HML -0.057 0.086 0.451 1.000 We choose the following variables to represent the
world economic environment for these globally traded
stocks: growth rate in industrial production (world),
Table 2B. Correlation matrix of the explanatory factors— unemployment rate (world), inflation (U.S.), credit
all stocks (Jan 2000–June 2007). spread (U.S.) and term spread (U.S.) during our sample
period. Credit spread is defined as the difference in the
XMKT SMB LEV HML weekly yield on Moody’s AAA corporate bonds and 1
year maturity government Treasury notes. Term spread
XMKT 1.000 -0.151 -0.145 -0.213 is calculated as the difference between the weekly yield
SMB -0.151 1.000 0.205 0.283 on 1 year treasury notes and 3 month treasury bills. The
LEV -0.145 0.205 1.000 0.407 source of the data is the FRED database at the St. Louis
HML -0.213 0.283 0.407 1.000 Federal Reserve. Monthly data for industrial production
for the world and unemployment rates have been obtained
from the database of IHS Global Insights (http://www.
Table 2C. Correlation matrix of the explanatory factors— ihsglobalinsight.com/EconomicFinancialData). Monthly
all stocks (July 2007–April 2009). inflation rates for the U.S. are obtained from the website
www.Inflationdata.com.
XMKT SMB LEV HML
Table 4 represents the results for multivariate regressions
XMKT 1.000 -0.007 0.222  0.143 of each of the macroeconomic variables on lagged excess
SMB -0.007 1.000 -0.448 -0.301 market returns and returns on SMB, HML and LEV. We use
LEV 0.222 -0.448 1.000 0.523 three lags to extract the maximum information content of
HML 0.143 -0.301 0.523 1.000 these factors. The regressions are estimated at following
periodicity: monthly for inflation, industrial production

Eds. Hatem A. El-Karanshawy et al. 51

 Bhatt and Sultan

growth rate, unemployment rate and weekly for credit pool of assets like mortgages or credit card receivables
spread and term spread. The regression model is: (commonly known as collateralized debt obligations).
Investors relied on the major credit rating agencies like
Ykt = b 0 + b 1(rm - r ft )t-i + b 2 Rt-i ,SMB Moody’s and Standard & Poor’s for the acceptance of these
+ b 3 Rt-i ,HML + b 4 Rt-i ,LEV + e t (2) products. The collapse of the subprime lending sector
and the resulting credit crisis in 2007 and 2008 exposed
a colossal failure of the credit rating agencies; which also
where, Ykt represents each of the following macroeconomic paved the way for a near-complete closure of markets for
variables: monthly percentage change in industrial these products. In a nutshell, the credit spreads did not
production growth rate, inflation and unemployment rate, reflect the true economic risk underlying the corporate
weekly credit spread and term spread, and i represents the debt, hence it is difficult to establish a true empirical
number of lagged terms 1 to 3 to reduce serial correlation. relationship between the leverage risk factor and the credit
All macroeconomic variables have been tested for unit root spread variables.
and those with unit root have been differenced once to
induce stationarity. Notwithstanding the previous discussion, we find a positive
relationship between LEV and the credit spread (see
Panel A represents the coefficients and t-statistics for each Table 4, Panel C). This is consistent with the evidence that
of the above macroeconomic variables regressed against the firms hit hard by the credit crisis were those that relied
one lag of the independent variables. The results indicate heavily on debt to finance growth like Home Depot, Toyota
that the leverage risk factor affects the unemployment rate Motor and FedEx.18 Stock prices of these firms, including
and inflation (at 5% level of significance) while remains investment banks like Citigroup and UBS AG, plummeted
insignificant for term spread, industrial production and during the recent market meltdown. Bear Stearns and
credit spread. In Panel B, unemployment exhibits significant American Home Mortgage are notable examples of firms
sensitivity to LEV lagged one period and inflation shows which were coerced to sell their holdings at far below their
significant sensitivity to LEV lagged two periods. Panel C book values. In general, there was a continuous re-pricing
also shows significant factor loadings on LEV for all the of risk in the stock market and stock prices plummeted.
macroeconomic variables at different lag lengths. With In contrast, the US treasury yields were falling due to
respect to the other factors, SMB shows significant factor flight to safety, while the rates on mortgage debts failed
loadings for industrial production and term spread (at the to decline at the same pace. This resulted in higher credit
first and second lags) while the impact of excess market spreads because mortgage debts were most risky and
returns on these macroeconomic variables seems to be demanded a premium over Treasury bonds. Thus, the
weak. Note that HML seems to have limited ability to positive relationship between leverage risk factor and
predict these economic variables at the first and the second the credit spread as seen in Table 4 is plausible since both
lags but tends to exhibit a significant impact on these the variables are representative of the increased exposure
variables at the third lag (significant for unemployment and of the firm to distress risk caused by over leveraging.
inflation). Overall, the results emphasize that the “leverage
risk factor” is a systematic risk factor, though its effects on Finally, when inflation is uncertain, investors demand
macroeconomic variables are not uniform. inflation risk premium. Inflation induces volatility in
the returns on debt and hence there is a leverage risk
These results have powerful implications for the US premium. Whether the relationship between inflation and
economy bouncing back from a severe financial crisis. First, leverage risk premium is positive or negative depends on
researchers argue that “deeper the decline in GDP, peak the interaction between inflation, taxes (corporate tax and
to trough, the more rapid the post recession rebound.” A personal tax), expected return on assets, and the amount of
recent report17 suggests that this is the case only if there debt used in the project. According to Armitrage (2005), as
is a significant increase in the private sector liabilities. inflation increases, the real tax adjusted weighted average
According to the report, a 0.3% drop in employment rate cost of capital decreases because higher inflation alleviates
requires the real GDP growth higher than 3%, which in the corporate taxes on the firms’ real profits and increases
turn requires a 5% rise in the private sector liabilities, the tax advantage on debt. However in the presence of
and subsequently, has a significant impact on the level personal taxes, higher inflation causes an increase in the tax
of industrial production. In fact, we have witnessed slow rates on real returns to debt. This increases the leverage risk
moving recoveries following the 1980, 1991 and 2001 of the firms which are heavily dependent on debt and thus
recessions, with the slowness being attributed to low levels demand a premium over firms which rely less on external
of private liabilities during these periods. This supports debt. For our sample, we find mixed evidence (positive and
the positive relationship between the leverage risk factor negative) of the relationship between inflation and the risk
and industrial production and a consistently negative factors (See Table 4).
relationship between unemployment rate and the leverage
risk factor which has been documented in the earlier
section. Explaining cross-section of returns
In this section, we present our regression results by
Next, the credit spread is a representative of firm’s default including leverage factor as a systematic risk factor. First, we
risk. A high credit spread indicates stringent credit markets test for the significance of the FF risk factors. Next we add
and higher risk levels. However, the last quarter century LEV to the regression model to compare results across three
witnessed some of the major developments in finance, periods: January 2000 –April 2009 (aggregate), January
for e.g. “securitisation” and introduction of “structured 2000 – June 2007 (non-crisis), and July 2007 – April 2009
products” which generate cash flows from underlying (crisis)19. To check on the robustness of these results, we will

52 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 4. Multivariate regressions of macroeconomic variables conditional on factor returns during the aggregate period.
The following regression is estimated to demonstrate the link between Fama-French factors and economic variables:

Ykt = b 0 + b 1 (rm - rft )t -i + b 2 R t -i ,SMB + b 3 R t -i ,HML + b 4 R t -i ,LEV + e t

where Ykt represents each of these macroeconomic variables (monthly Industrial production growth rates, monthly unemployment
rate, monthly data for percentage change in inflation rates and weekly data for credit spread and term spread) for the combined period
(January 2000 to April 2009). i represents the number of lagged terms 1 to 3. rf is the return on the risk free asset and rm is the return
on the market portfolio. rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size
mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big” portfolios.
RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of all “high
BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple
average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”.

Panel A
% change in
Industrial Unemployment rate Credit Spread Term Spread inflation rate
Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats
XMKT t-1 −3.241 −0.307 0.007 0.044 −0.0001 −2.081** 0.0000 −1.660* 1.847 −0.172
SMB t-1 −41.724 −2.249** 0.338 1.110 0.0001 1.326 −0.0001 −2.965** 1.532 1.365
HML t-1 −29.364 −1.090 0.729 1.721* 0.0001 0.681 −0.0001 −1.393 −6.337 0.959
LEV t-1 62.315 1.540 −1.667 −2.302** −0.0003 −1.404 0.0001 1.348 5.998 −2.207**
R-square 0.105 0.114** 0.030** 0.028** 0.152**

Panel B

% change in
Industrial Unemployment rate Credit Spread Term Spread inflation rate
Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats
XMKT t-1 −2.582 −0.205 0.016 0.098 −0.0001 −1.619 0.0000 −1.501 2.308 1.707
SMB t-1 −47.422 −3.096** 0.438 1.786* 0.0001 1.928** −0.0001 −2.992** 1.172 0.792
HML t-1 −20.096 −0.794 0.578 1.588 0.0001 0.507 −0.0001 −1.802* −5.263 −1.897*
LEV t-1 24.896 0.903 −0.910 −1.840* −0.0002 −0.960 0.0001 1.362 3.367 0.814
XMKTt-2 13.845 2.104** −0.350 −2.413** −0.0001 −3.493** 0.0000 −0.388 0.361 0.454
SMB t-2 −23.054 −1.497 0.087 0.305 0.0001 0.835 0.0000 −0.519 0.187 0.123
HML t-2 −21.410 −0.897 0.447 1.018 0.0000 −0.183 −0.0002 −2.547** 0.306 0.165
LEV t-2 61.648 1.553 −1.043 −1.412 −0.0001 −0.687 0.0001 0.951 5.379 2.274**
R-square 0.224 0.219 0.054 0.042 0.186

Panel C

% change in
Industrial Unemployment rate Credit Spread Term Spread inflation rate
Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats Coeff. t-stats
XMKT t-1 −5.246 −0.410 0.090 0.535 −0.0001 −1.651* 0.0000 −1.402 2.403 1.678*
SMB t-1 −48.352 −3.397 0.480 2.004** 0.0001 1.835* −0.0001 −2.961 0.803 0.537
HML t-1 −7.562 −0.297 0.230 0.579 0.0001 0.540 −0.0001 −1.733 −4.378 −1.353
LEV t-1 44.872 1.699 −1.224 −2.597** −0.0002 −1.038 0.0001 1.267 3.673 0.859
XMKTt-2 18.343 2.842 −0.458 −3.220** −0.0001 −4.013** 0.0000 −0.505 0.739 0.991
SMB t-2 −25.884 −2.343 0.196 0.941 0.0000 0.603 0.0000 −0.585 0.067 0.049
HML t-2 −6.985 −0.384 0.270 0.946 0.0000 −0.320 −0.0001 −2.321 0.110 0.053
LEV t-2 1.522 0.045 −0.043 −0.084 −0.0001 −0.787 0.0001 0.812 6.470 2.284**
XMKTt-3 16.418 2.282 −0.476 −3.596** 0.0000 −0.529 0.0000 0.473 0.308 0.338
SMB t-3 −25.738 −1.619 0.087 0.316 0.0000 0.354 0.0000 0.225 0.393 0.214
HML t-3 2.303 0.149 −0.631 −2.107** −0.0001 −0.923 0.0000 0.738 5.296 2.655**
LEV t-3 2.740** −1.109 −2.042** 0.0003 1.953* 0.0000 0.153 −2.493 −0.861
R-square 0.430 0.412 0.066 0.044 0.215
(*indicates significant at 10% level, ** indicates significance at 5% level)

Eds. Hatem A. El-Karanshawy et al. 53

 Bhatt and Sultan

further classify firms into two groups: financial and non- XMKT is consistent with the single factor CAPM model.
financial. Financial firms include all financial institution as The distribution of the SMB is about half positive and half
well as real estate and mortgage firms. The popular adage is negative. The HML is positive in 1,539 and negative in
that leverage is a two-way sword. It magnifies returns in an 216 cases. When LEV is added to the model (Model 2), we
up market and magnifies losses in a down market. Finally, find that, there is a .24% increase in the number of cases
we test our main hypothesis that Islamic stocks would be ((3312/3304)-1) where XMKT is significant. With the
less sensitive to the leverage risk factor than conventional addition of LEV, there is a 2.54% increase in the number
and socially responsible stocks. Our primary rationale is of cases where SMB is significant. Surprisingly, the number
that low leverage of Islamic stocks would lessen the interest of cases HML is positive and significant drops by 16.58%.
rate exposure of these firms. Finally, in 2,208  instances, LEV is positive, though in
125 instances it is negative.
We use the following firm-specific GARCH model:
The results (Panel B) for the non-crisis period (2000-June
rt - r ft = β 0 + β 1(rmt - r ft ) + β 2 Rt ,SMB + β 3 Rt ,HML 2007) are similar. The number of cases where the factors
+ β 4 Rt ,LEV + e t (3) is significant changed as follows: .39% (XMKT), 1.83%
  (SMB), and -2.05% (HML). With regard to positive and
negative impact of the factors on stock returns, there are
  e t|ψ t-1 ~ N(0, σ t2 ), (4) some changes compared to the aggregate period (Model
2). For example, SMB, the number of negative cases is now
q p 680, representing a 40% decline from the previous model.
σ t2 = W + ∑ a ie 2 t-i + ∑ δ iσ t- j (5) In contrast, HML, now has 539  instances for which the
  i=1 j =1 coefficients are negative, indicating a 17.43% increase from
the previous value. Finally, we have 862 instances of positive
where rt – rft in the mean equation is the weekly excess and 149 cases of negative coefficients for LEV. It appears
return on asset i, rft is the weekly risk free rate (US T-bill), that, compared to Model 2 (aggregate period), there is a
rmt – rft is the market risk premium (XMKT), and SMB, HML large number of instances the regression coefficients are
and LEV are Fama-French factors and the leverage risk insignificant. Altogether, the number of significant cases
factor, defined earlier. The variance equation (5) models drops by 56%, suggesting that the LEV factor is able to
the conditional variance as a GARCH(p,q) process where capture systemic risk in the economy across good and bad
p and q denote the lag length. W is the intercept term, a times quite well.
is the ARCH term and d is the GARCH term. a and d terms
are expected to be positive and significant determinants However, the contribution of LEV in capturing leverage risk
of the conditional variance of changes in the excess is evident when we estimate firm-specific regressions for
return. The primary reason for using the GARCH model the crisis period (Panel C). During July 2007-April 2009,
is that preliminary diagnostics suggest that the weekly compared to the non-crisis period, there is a 201.07%
excess returns have time varying variance with volatility increase in the number of the cases where LEV is significant.
clustering and fat tails. The GARCH models are estimated This increase is indicative of several stylized facts during
using the Bollerslev-Wooldridge (1992) corrections to deal the escalating financial crisis afflicting the global economy.
with excess kurtosis. As noted earlier, standard t-statistics It appears that the credit crisis had a contagion-like effect,
based inferences in the presence of excess kurtosis in the impacting firms across all spectrums of leverage. In essence,
residuals are asymptotically invalid because standard errors firms were hard hit especially when access to the debt market
are biased downward, leading to false acceptances. was severely limited because of reluctance among financial
institutions to lend. The results suggest that for 3,038 firms,
Factor loadings at the firm level the sensitivity to LEV is positive and significant. Only in
We test the above model at both the firm and portfolio level 66 cases the variable has negative coefficients. Compared
for all 3,707 financial and non-financial firms. Each week to the non-crisis period, the addition of LEV during the
from January 2000 to April 2009 we run cross sectional credit crisis leads to a change in the number of significant
regressions of weekly excess stock returns on XMKT, cases for the remaining factors: XMKT (-84.91%), SMB
SMB, and HML factors. Next, we add LEV to test for its (-14.97%), and HML (-27.54%). In particular, the number
significance in addition to the market factor and the Fama- of negative coefficients for HML is higher than the positive
French factors. For robustness check, we test for the partial ones, indicating that during the recent credit crisis, a value
F-statistics of LEV to see whether this additional factor based investment strategy would have earned investors
contributes significantly in explaining the cross section of negative risk premium. Again, it supports the notion that
expected returns (in addition to the market factor and the the HML may not have been a good proxy for the distress
traditional FF factors). risk during this period.

Tables 5 exhibit the summary of the impact of XMKT, SMB, Factor loadings at the portfolio level
HML and LEV factors on the returns of firm and portfolios. Table  5 also the highlights portfolio-specific regressions
Model 1 is the traditional FF case and Model 2  includes (Panels D-F) for the three periods. Based on the intersection
the LEV factor in addition to the FF factors. As shown, we of these three factors, we have 27 portfolios with annual
have 3707stocks in the sample. Note that in Table  5 and rebalancing. The results reconfirm our earlier finding
subsequent tables, we only include regression results that that the addition of LEV weakens the significance of the
are significant at least at the 5% level. In Panel A, the traditional FF factors. For the aggregate period (Panel D),
results show that for the aggregate period (2000–2009), we find that LEV is significant and positive for 19 out of 27
in 3,304 instances XMKT is positive. A positive sign for the portfolios. The number drops to 17 when we estimate the

54 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 5. Summary of results showing the number of stocks and portfolios which showed significant sensitivities to XMKT, SMB, HML, and LEV factors.
Model 1 Model 2

rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t

2
e t |ψ t -1 ~ N(0, σ ), t e t |ψ t -1 ~ N(0, σ t2 ),
q p q p

i t- j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑δ σ
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking
the simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple
average of the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average

Eds. Hatem A. El-Karanshawy et al.

of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with (*) are significant at least at the 5% level of significance.

All stocks: 3707 All Portfolios: 27

Panel A: Aggregate Period Panel D: Aggregate Period

XMKT SMB HML LEV XMKT SMB HML LEV

Model 1 Positive 3304 1208 1539 Model 1 Positive 27 13 18

Negative 0 1114 216   Negative 0 9 3
Total 3304 2322 1755   Total 27 22 21
Model 2 Positive 3312 1234 1005 2208 Model 2 Positive 27 12 17 19
Negative 0 1147 459 125 Negative 0 9 6 1
Total 3312 2381 1464 2333 Total 27 21 23 20
A comparison among Islamic, conventional, and socially responsible stocks

% change in 0.24% 2.54% -16.58%   % change in 0.00% -4.55% 9.52%  

significance (by model) significance (by model)
Panel B: Non-crisis Period Panel E: Non-crisis Period

XMKT SMB HML LEV XMKT SMB HML LEV

Model 1 Positive 3570 1461 1244 0 Model 1 Positive 27 15 18

Negative 0 612 417   Negative 0 9 3
Total 3570 2073 1661   Total 27 24 21
Model 2 Positive 3584 1431 1088 882 Model 2 Positive 27 15 18 17
Negative 0 680 539 149 Negative 0 9 5 3
Total 3584 2111 1627 1031 Total 27 24 23 20
%change in 0.39% 1.83% -2.05%   %change in 0.00% 0.00% 9.52%  
significance (by model) significance (by model)
(Continued)

55
 Bhatt and Sultan

model for the non-crisis period (Panel E). In contrast, we

35.00%
find that in all cases, LEV is positive and significant during
LEV the crisis period (Panel F). We also note that, in comparison

27
27
to the aggregate period, there is a 100% reduction in the

 
number of cases XMKT is significant during the crisis
period. For the remaining variables, percentage change

-18.75%

-43.48%
in significance is as follows: SMB (-41.67%) and for HML
(-43.48%), indicating an across the board weakening of
HML

the FF factors during the financial distress. In contrast,

13
16

16

12
0

1
there is a 35% increase in the number of instances where
LEV is positive and significant.
-17.65%

-41.67%
Overall, the FF factors seem to lose their significance when
SMB

LEV as a systemic risk is included in the model. In particular,

14
17
17
0

7
7

during a financial crisis period, sensitivity to LEV at the firm

and portfolio level suggests that the traditional FF factors
-100.00%
may not be adequately capturing the effects of economy-
wide distress arising from excess leverage. Therefore,
XMKT

0.00%

sensitivity to this systemic risk translates into additional

risk premium that is not adequately captured by the FF risk
0
4
0
4
0
0

factors.
significance (by period)
significance (by model)
Panel F: Crisis period

Negative

Negative

Portfolio-specific regression results across aggregate, non-

Positive

Positive

crisis and crisis periods are provided next in Tables 6–8 to

Total
Total

highlight the magnitude of the coefficients and to check

for robustness of adding LEV. In Table 6, there are several
%change in

%change in

stylized facts for the aggregate period. First, there is a

noticeable increase in the adjusted R2 when LEV is added as
Model 1

Model 2

an explanatory variable. Second, as reported earlier, with

the addition of LEV in Model 2, the traditional FF factors
tend to lose their statistical significance. Finally, we find
that in many instances, the coefficient of HML actually
turns negative.
201.07%

In Table  7, we report the results for the non-crisis period

and the results indicate that while the variable LEV is an
3104
3038
LEV

important explanatory power, its addition makes only

66
0

marginal impact on Model 2. There is an increase in the

 
 
 

adjusted R2 but by a small margin. In contrast, during

the crisis period (Table 8), the addition of the LEV makes
-17.32%

-27.54%

a substantial contribution to the overall forecast ability

1179
1330

1426
HML

of Model 2. The adjusted R2 increases by a considerable

439
740
96

margin. In addition, the size of the coefficients across the

27 portfolios is large, ranging from 1.56% (portfolio #7)
to 4% (portfolio #18). The magnitude of the coefficients
-14.97%

indicates the heightened sensitivity of firms to the economic

-8.65%

distress during the period. As indicated earlier, the results

1795
1564
1965
SMB

401

854
941

indicate that our leverage risk factor performs quite well in

representing systemic risk in the global economy. Also note
that the number of negative significant coefficients for HML
increases considerably (from 5 to 12)20 which suggests that
-84.91%

the value based investment strategy may not with falling

-2.35%
XMKT

equity prices. In contrast, LEV has a positive relationship

541
439
115
554
157
384

with stock returns for all 27 portfolios, suggesting that

investors demand a premium for investing in high leverage
portfolios during the credit crisis.
significance (by period)
significance (by model)
Panel C: Crisis period

Negative

Negative
Positive

Positive
Table 5. (Continued)

The negative effects of leverage risk factor on stock

Total
Total

returns
In a number of cases (see Tables  5–8), leverage risk has
a negative effect on stock returns, which is consistent
% change in
%change in

with several existing studies. For example, Penman et  al.

Model 1

Model 2

(2007) decompose the book to price ratio of a firm into two

components. The first component is the enterprise book
to price (which represents the operating risk of the firm),

56 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 6. Factor loadings of all firms for the aggregate period (January 2000 to April 2009).

Model 1 Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t-1 ~ N(0,σ t2 ), e t |ψ t -1 ~ N(0,σ t2 ),
q p q p

σ t2 = W + ∑ a i e t2-i + ∑δ σ i t- j
σ t2 = W + ∑ a i e t2-i + ∑δ σ i t- j

i=1 j =1
i=1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.

Aggregate Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 0.002* 0.794* 0.388* -0.235* -0.015 0.002* 0.794* 0.389* -0.231* -0.023 -0.021
2 0.002* 0.663* 0.344* -0.047 0.000 0.002* 0.642* 0.301* -0.193* 0.722* 0.103
3 0.001 0.742* 0.598* 0.053 -0.004 0.001 0.727* 0.560* -0.254* 1.303* 0.169
4 0.002 0.557* 0.390* 0.261* 0.014 0.001 0.547* 0.364* 0.248* 0.145 0.042
5 0.000 0.661* 0.546* 0.431* -0.001 0.000 0.665* 0.521* 0.353* 0.396* 0.064
6 0.000 0.723* 0.565* 0.565* 0.024 0.000 0.701* 0.510* 0.444* 0.457* 0.103
7 0.000 0.622* 0.640* 0.744* 0.055 -0.001 0.639* 0.671* 0.753* -0.118 0.021
8 0.000 0.765* 0.731* 0.812* 0.022 0.000 0.767* 0.720* 0.740* 0.306* 0.059
9 0.000 0.840* 0.842* 0.923* 0.080 0.000 0.862* 0.890* 0.725* 0.741* 0.141
10 0.000 0.896* 0.053 -0.264* -0.041 0.000 0.900* 0.044 -0.289* 0.141 -0.028
11 0.001 0.690* 0.062 -0.072 0.009 0.001 0.700* 0.023 -0.169* 0.453* 0.079
12 0.000 0.718* 0.009 0.016 -0.021 0.000 0.733* -0.050 -0.155 0.807* 0.101
13 0.000 0.630* 0.083 0.372* -0.014 0.000 0.628* 0.050 0.306* 0.337* 0.034
14 0.001 0.631* 0.052 0.340* 0.042 0.001 0.648* -0.003 0.214* 0.548* 0.129
15 0.001 0.632* 0.138* 0.462* 0.032 0.001 0.634* 0.089 0.306* 0.659* 0.151
16 0.000 0.630* 0.226* 0.892* 0.108 0.000 0.628* 0.222* 0.874* 0.053 0.112
17 0.001 0.749* 0.250* 0.684* 0.051 0.001 0.752* 0.203* 0.590* 0.515* 0.132
18 0.001 0.810* 0.188* 0.872* 0.086 0.001 0.814* 0.130* 0.716* 0.939* 0.215
19 0.000 0.772* -0.470* -0.288* 0.198 0.000 0.775* -0.467* -0.274* -0.100 0.189
20 0.000 0.656* -0.396* 0.035 0.094 0.000 0.651* -0.406* -0.022* 0.409* 0.148
21 0.000 0.613* -0.379* 0.082 0.122 0.000 0.622* -0.358* -0.075 0.650* 0.202
22 0.001 0.698* -0.387* 0.255* 0.064 0.001 0.671* -0.414* 0.180* 0.312 0.104
23 0.001 0.721* -0.283* 0.352* 0.073 0.001 0.718* -0.338* 0.261* 0.599* 0.158
24 0.001 0.671* -0.253* 0.505* 0.103 0.001 0.667* -0.262* 0.357* 0.543* 0.184
25 0.002* 0.804* -0.839* 0.856* 0.297 0.002* 0.806* -0.848* 0.937* -0.430* 0.277
26 0.000 0.773* -0.321* 0.901* 0.179 0.000 0.769* -0.290* 0.670* 1.035* 0.255
27 0.002* 0.846* -0.486* 0.880* 0.255 0.002* 0.857* -0.562 0.688* 1.271* 0.401

measured as the ratio of book value of operating assets to with the expected stock returns. Johnson (2004) documents
their market value. The second component is the financial a negative relationship between leverage and cross section
leverage component (which represents the financing risk of expected returns after controlling for firm specific
of the firm), measured as the ratio of market value of debt characteristics like volatility. See Arditti (1967), Dimitrov
to market value of equity. The authors find that enterprise and Jain (2006) for similar results. In particular, Dimitrov
book to price ratio has a significant positive relationship and Jain (2006) note that during economic distress, raising
with the expected stock returns while the “leverage” equity is costlier than debt (e.g., bank financing or line
component of book to price ratio has negative relationship of credit), so firms would prefer to increase leverage. So,

Eds. Hatem A. El-Karanshawy et al. 57

 Bhatt and Sultan

Table 7. Factor loadings of all firms for non-crisis period (January 2000 to June 2007).

Model 1 Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t -1~N( 0 ,σ t2 ), e t |ψ t -1 ~ N(0, σ t2 ),
q p q p

σ t2 = W + ∑ a i e t2-i + ∑δ σ i t- j
σ t2 = W + ∑ a i e 2t -i + ∑δ σ i t- j

i=1 j =1
i=1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Non-crisis Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept XMKT SMB HML R-square Intercept XMKT SMB HML LEV R-square

 1 0.002* 0.856 0.449* -0.259* 0.473 0.002* 0.851* 0.454* -0.224* -0.195 0.473
 2 0.002* 0.778 0.457* -0.139 0.320 0.002* 0.783* 0.447* -0.204* 0.343 0.328
 3 0.001 0.822 0.678* 0.005 0.378 0.001 0.846* 0.646* -0.253* 1.119* 0.419
 4 0.001 0.673 0.397* 0.278* 0.394 0.001 0.674* 0.396* 0.277* 0.005 0.392
 5 0.000 0.709 0.565* 0.435* 0.440 0.000 0.722* 0.540* 0.378* 0.313* 0.449
 6 0.000 0.781 0.618* 0.562* 0.433 0.000 0.777* 0.572* 0.479* 0.311* 0.450
 7 -0.001 0.785 0.686* 0.712* 0.532 -0.001 0.777* 0.699* 0.772* -0.221* 0.531
 8 0.000 0.807 0.764* 0.817* 0.501 0.000 0.810* 0.760* 0.766* 0.212* 0.502
 9 0.000 0.878 0.882* 0.925* 0.471 0.000 0.899* 0.916* 0.775* 0.637 0.475
10 -0.001 0.973 0.142* -0.367* 0.466 -0.001 0.974* 0.142* -0.369* 0.008 0.464
11 0.000 0.743 0.069 -0.073 0.446 0.000 0.764* 0.045 -0.143 0.322* 0.456
12 0.000 0.791 0.081 -0.070 0.324 0.000 0.812* 0.027 -0.178 0.661* 0.361
13 0.000 0.699 0.127* 0.346* 0.383 0.000 0.703* 0.106* 0.310* 0.213* 0.386
14 0.001 0.687 0.077 0.346* 0.405 0.001 0.715* 0.049 0.256* 0.429* 0.422
15 0.001 0.662 0.176* 0.436* 0.400 0.001 0.682* 0.119* 0.338* 0.559* 0.437
16 0.000 0.701 0.294* 0.846* 0.482 0.000 0.700* 0.300* 0.870* -0.084 0.481
17 0.001 0.781 0.298* 0.668* 0.469 0.001 0.797* 0.271* 0.599* 0.392* 0.484
18 0.001 0.852 0.233* 0.834* 0.446 0.001 0.886* 0.171* 0.729* 0.832* 0.473
19 0.000 0.865 -0.360* -0.354* 0.542 0.000 0.929* -0.349* -0.415* -0.430* 0.552
20 0.000 0.693 -0.367* 0.040 0.502 0.000 0.694* -0.374* -0.003 0.314* 0.511
21 0.000 0.664 -0.308* 0.037 0.437 0.000 0.682* -0.304* -0.085 0.539* 0.461
22 0.001 0.823 -0.335* 0.255* 0.420 0.001 0.814* -0.345* 0.221* 0.170 0.422
23 0.001 0.769 -0.226* 0.324* 0.478 0.001 0.779* -0.280* 0.266* 0.523* 0.504
24 0.001 0.715 -0.205* 0.498* 0.453 0.001 0.721* -0.215* 0.392* 0.447* 0.480
25 0.002* 0.883 -0.747* 0.812* 0.401 0.002* 0.867* -0.753 0.930* -0.699* 0.441
26 0.000 0.951 -0.496* 0.940* 0.363 0.000 1.014* -0.478 0.707* 0.666* 0.363
27 0.002* 0.891 -0.416* 0.846* 0.395 0.002* 0.921* -0.450 0.677* 0.992* 0.456

falling equity returns during economic distress and rising emphasize that the negative relationship between leverage
leverage support the empirical finding that leverage and and growth is more visible for firms with a low Tobin’s q
return on equity may be negatively correlated. since these firms are characterised by negligible growth
opportunities not recognised by the capital markets. The
Managerial preference for debt over equity financing is also study further rationalises that managers of firms with
related to the value of the firm and its future prospects. considerably lucrative growth opportunities generally do
Lang et  al. (1995) find a negative relationship between not opt for a high leverage21 because high interest payments
financial leverage and future growth of a firm. The authors on debt tend to erode the profitability of the firm which

58 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 8. Factor loadings of all firms during crisis period (July 2007 to April 2009).
Model 1 Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1 ~ N(0, σ ), 2
t e t |ψ t -1~N( 0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Crisis period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept XMKT SMB HML R-square Intercept XMKT SMB HML LEV R-square

1 -0.003 0.075 -0.320 0.161 -0.034 -0.001 -0.034 0.264 -0.942* 2.931* 0.410
2 -0.005 0.084 -0.164 0.221 -0.040 -0.002 0.020 0.461* -0.974* 3.193* 0.444
3 -0.003 0.077 -0.326 0.498 -0.005 0.000 -0.028 0.393* -0.831* 3.633* 0.549
4 -0.003 0.087 -0.079 0.608* -0.010 -0.003 0.038 0.238 -0.180 1.784* 0.312
5 -0.003 0.145 -0.151 0.711* 0.012 -0.003 0.029 0.385* -0.364 2.610* 0.450
6 -0.004 0.167* 0.060 0.757* 0.004 -0.004 0.064 0.598* -0.196 2.643* 0.427
7 -0.004 0.078 0.129 0.782* 0.053 -0.003 -0.033 0.436* 0.177 1.563* 0.311
8 -0.002 0.196* -0.081 0.855* 0.053 -0.001 0.019 0.544* -0.087 2.736* 0.473
9 -0.002 0.112 0.077 1.347* 0.136 0.000 0.018 0.748* -0.080 3.851* 0.587
10 -0.003 0.086 -0.520* 0.361 0.026 -0.001 0.063 0.022 -0.927* 3.042* 0.458
11 -0.004 0.037 -0.619* 0.522 0.053 -0.003 -0.007 -0.128 -0.910* 3.149* 0.506
12 -0.004 0.109 -0.791* 0.463 0.105 -0.003 0.063 -0.247 -0.859* 3.078* 0.538
13 -0.003 0.100 -0.370* 0.702* 0.059 -0.003 0.057 0.031 -0.274 2.140* 0.385
14 -0.003 0.127 -0.728* 0.724* 0.132 -0.002 0.039 -0.119 -0.442* 3.061* 0.552
15 -0.004 0.095 -0.586* 0.908* 0.130 -0.003 0.034 -0.016 -0.490* 3.328* 0.561
16 -0.003 0.098 -0.275 0.871* 0.134 -0.003 0.028 0.114 0.096 1.895* 0.373
17 -0.003 0.031 -0.556* 1.278* 0.175 -0.002 0.012 0.020 0.039 3.287* 0.609
18 -0.005 0.051 -0.863* 1.541* 0.264 -0.003 0.045 -0.134 -0.112 4.033* 0.657
19 -0.002 0.173* -1.069* -0.026 0.182 -0.002 0.028 -0.554* -0.877* 2.388* 0.522
20 -0.001 0.119 -0.964* -0.011 0.113 -0.002 0.000 -0.349 -1.069* 2.744* 0.479
21 -0.003 0.194* -1.017* 0.092 0.117 -0.002 0.044 -0.475* -0.853* 2.596* 0.531
22 -0.002 0.146 -0.903* 0.278 0.089 -0.001 -0.042 -0.480* -0.416 2.083* 0.356
23 -0.002 0.160 -1.064* 0.445 0.178 -0.002 0.009 -0.485* -0.600* 2.883* 0.534
24 -0.004 0.157 -0.897* 0.749* 0.199 -0.003 0.024 -0.386 -0.413 2.875* 0.561
25 -0.004 0.196 -1.176* 0.887* 0.374 -0.002 0.129 -0.735* -0.004 2.702* 0.581
26 -0.004 0.043 -1.124* 1.435* 0.386 -0.003 0.048 -0.587* 0.333 2.972* 0.680
27 -0.005 -0.008 -1.782* 1.765* 0.454 -0.002 -0.048 -1.153* 0.486* 3.693* 0.771

prevents the firm from utilizing the benefits of these growth have direct implications on the cash flow of the company by
opportunities. Hence, a negative relationship between enforcing regular interest payments on debt which controls
leverage and growth seems rational, which implies a managerial expropriation. Fama and Jensen (1983)
negative relationship between leverage and stock returns.22 explain that increased debt levels adds to the default risk
of the firm and affects the manager’s reputations adversely
The negative effect of leverage on return on equity is also in case the firm defaults on its interest payments or debt.
consistent from a corporate governance perspective. Jensen This imposes a constraint on manager expropriation and
and Meckling (1976) suggest that increased debt levels leads to better corporate disclosures. In addition, Jensen

Eds. Hatem A. El-Karanshawy et al. 59

 Bhatt and Sultan

(1986) suggests that leverage increasing transactions such stocks25. Such an examination is critical because it removes
as LBOs, new debt issues (bonds), and stock repurchase industry-specific effects of the credit crisis since the effects
reduce the manager’s access to free cash, thus reducing may not have been uniformly distributed among financial
their waste. He further suggests that debt reduces the and non-financial firms. Financial firms included in the
agency cost. This implies that as leveraging increases, sample include banks, S&Ls, credit unions, mortgage
external monitoring increases, and managerial efficiency financing companies, real estate firms, and insurance
is expected to rise. Furthermore, this may be imply that companies. Clearly, these firms bore the brunt of the credit
as firms become efficient, shareholders demand less risk crisis due to over speculation, deregulation, and over
premium for leverage, and as a result, stock prices fall with leveraging. We re-construct FF and LEV factors and estimate
higher leverage. firm26 and portfolio-specific regressions using two separate
samples of firms: the first sample with 645 financial stocks
Consistent with the above discussion, there are also more and the second sample with 2,975 non-financial stocks.
instances of significant negative coefficients for LEV during
the non-crisis period (which was a period with profitable A summary of the regression results are reported in Panels
investment opportunities in the market). For example A-F, Table  9. In Panels A-C, there is evidence that the
as reported in Table  5, there are 149 cases of negative leverage risk factor performs well across the three periods,
coefficients on LEV (Model 2, Panel B) during the non- especially during the crisis period. The results support
crisis period, but the number reduces to 66 during the crisis the hypothesis that the addition of LEV weakens the
period (Panel C). At the portfolio level (Panel F), compared significance of the traditional FF factors. For the aggregate
to the non-crisis period (Panel E), the number of negative period (Panel A), LEV is significant and positive for 17 out
coefficients for LEV reduces from 3 to 0. of 27 portfolios. During the non-crisis period (Panel B), the
significance of LEV drops, we now have 14 positive and 4
Leverage and investment strategy negative instances. During the crisis period (Panel C), in 23
These results have investment implications that suggest out of 27 cases, LEV is positive and significant. Note that,
investing in highly leveraged firms. However, an investor in comparison to the non-crisis period, there is a -85.19%
needs to decide between excessively high leverage level change in the number of cases XMKT is significant during
and the negative effects of leverage on financial distress the crisis period. For the remaining risk factors, the change
(Luoma and Spiller (2002)). See Bris and Koskinen (2002) in significance is as follows: SMB (−5.26%) and HML
for further evidences. A recent report23 elaborates that the (−59%), suggesting a weakening of the FF factors during
regular interest payments on debt for those companies the financial distress. In contrast, there is an increase of
which fund their investments through debt tend to erode the 27.78% in the number of instances where LEV is positive
cash flow levels of the company by adding to the operating and significant during the crisis period.
expenses of the firm. The flip side of the argument is that a
firm with highly profitable growth opportunities and with In Panels D-F, we report a summary of statistically
a strong cash flow position would still earn a higher return significant results for the non-financial firms in the sample.
on equity since they yield high profit margins. The report We confirm our previous findings that the addition of
claims that a period of economic recovery is characterized LEV weakens the significance of the traditional FF factors
by a strong economic momentum which bolsters earnings considerably. For the aggregate period (Panel D), LEV is
potentials of levered firms. The rationale behind this is that positive in 16 out of 27 portfolios. During the non-crisis
debt is cheaper for firms with promising growth prospects, period (Panel E), in 14  instances LEV has positive and
and such they perform at the peak levels when debt is significant coefficients, and in 4  instances the coefficients
easily available24. The economic recovery in 2003 provides are negative and significant. Similar to our earlier findings,
strong evidence to this fact when the federal funds rate was in 27 out of 27 cases, LEV is positive and significant during
approximately 1.25%, which in turn stimulated economic the crisis period (Panel F). In comparison to the non-crisis
growth to jump from 1% to 7%. During this period, levered period, there is a -100% change in the number of cases
companies, high yield bonds and bank loans yielded where XMKT is significant during the crisis period. For the
attractive returns. SMB, the number of significant cases changes by -39.13%
and for HML, the number of significant cases changes by
These results do not suggest that as efficiency increases, -57%, confirming the fact that the power of the FF factors
stock price decreases. Rather, as firms become more weakens during the financial distress. In contrast, there is
efficient, debt becomes cheaper and such companies can a 50% increase in the number of instances where LEV is
afford to have high debt levels in their capital structure positive during the crisis period.
(thereby decreasing the overall cost of capital) without
increasing their credit risk. Due to lower risk levels, Details of these portfolio-specific regressions across
investors do not need additional compensation for excessive aggregate, non-crisis and crisis periods are provided in
leverage as in the case of firms which are not efficient. Also, Tables  10–15 to demonstrate the contribution of the LEV
in efficient markets, due to strong corporate governance on a case by case basis. Tables 10–12 report the results for
principles and better disclosures, the probability of insider the financial stocks while Tables  13–15 report the results
information is reduced and information of the company is for the non-financial stocks. In these tables we also report
quickly reflected in the stock prices. Hence there is no scope the adjusted R2 for Model 1 (without LEV) and Model 2
for mispricing or arbitrage opportunities; so returns fall. (with LEV) and the results confirm our earlier results.
In Table  10, first, there is a noticeable increase in the
Leverage risk of financial and real estate firms adjusted R2 when LEV is added as an explanatory variable,
We perform additional robustness tests by separating indicating increased forecasting power of Model 2 during
the financial stocks in the sample from the non-financial the aggregate period. Second, as reported earlier, with the

60 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 9. Summary of factor loadings for financial and non financial stock portfolios.
Model 1 Model 2

rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t

e t |ψ t -1~N( 0 , σ t2 ), e t |ψ t -1~N( 0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking
the simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple aver-

Eds. Hatem A. El-Karanshawy et al.

age of the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average of
the returns each week of all “high leverage” portfolios minus “low leverage portfolios”.

Financial stock portfolios Non financial stock portfolios

Pane A: Aggregate Period Panel D: Aggregate Period

XMKT SMB HML LEV XMKT SMB HML LEV

Model 1 Positive 27 7 15 Model 1 Positive 27 13 18

Negative 0 15 7 Negative 0 8 3
Total 27 22 22 Total 27 21 21
Model 2 Positive 27 7 15 17 Model 2 Positive 27 13 18 16
Negative 0 16 7 0 Negative 0 9 6 1
Total 27 23 22 17 Total 27 22 24 17
A comparison among Islamic, conventional, and socially responsible stocks

% change in 0.00% 4.55% 0.00% % change in 0.00% 4.76% 14.29%

significance (by model) significance (by model)
Panel B: Non-crisis Period Panel E: Non-crisis Period

XMKT SMB HML LEV XMKT SMB HML LEV

Model 1 Positive 27 8 15 0 Model 1 Positive 27 14 18

Negative 0 12 7 Negative 0 9 3
Total 27 20 22 Total 27 23 21
Model 2 Positive 27 8 15 14 Model 2 Positive 27 14 18 14
Negative 0 11 7 4 Negative 0 9 3 4
Total 27 19 22 18 Total 27 23 21 18

% change in 0.00% -5.00% 0.00% % change in significance 0.00% 0.00% 0.00%

significance (by model) (by model)

61
 Bhatt and Sultan

addition of LEV in Model 2, the statistical significance of

50.00%
the traditional FF factors tend to weaken. Finally, we find

LEV
that in many instances the coefficient of HML actually

27
27
0
turns negative. During the non-crisis period (Table 11), the
addition of LEV to the model makes only marginal impact
on the forecast power of the Model 2. The adjusted the R2

-10.00%

-57.14%
changes by a small margin. In contrast, we find that during
the crisis period (Table 12), the addition of the LEV makes
HML

a substantial contribution to the overall forecast ability of

10

9
9
6
4

0
Model 2. The adjusted R2 increases by a substantial margin.
Furthermore, the size of the coefficient for LEV across
133.33% portfolios is large, similar to the results reported earlier. The

-39.13%
magnitude of the coefficient clearly indicates an increased
SMB

sensitivity of firms to the economic distress. Similar results

14
10

are obtained for non-financial stocks in Tables 13–15.

4
0
6
6

Overall, our analysis shows that financial and non-

financial categories of stocks have similar exposure to the
-100.00%

debt market, despite the fact that the concept of leverage

XMKT

0.00%

and its use varies across these two categories of firms. It

again reinforces the notion that the financial crisis had a
0
0
0
0
0
0

contagion-like effect on all types of firms. It also establishes

the fact that our leverage risk factor is able to capture
% change in significance

% change in significance

economy-wide risk from over leveraging during the

Panel F: Crisis period

Negative
Negative
Positive

Positive

financial crisis period.

Total
Total

Test for robustness

Earlier, we reported that the correlation between LEV and
(by period)
(by model)

HML as follows: .45 (aggregate period), .40 (non-crisis

Model 1

Model 2

period), and .51 (crisis period). These correlations may be

viewed as high, raising a criticism that LEV factor may be
collinear with HML, and as such, rendering the effects of
HML insignificant in majority of the cases. We argue that
while HML and LEV are both balance sheet variables and
are therefore should be correlated, they do not represent
27.78%

similar risk factor in the economy. That is to say that HML

LEV

does not represent LEV and LEV does not represent HML.
23
23
0
0

Both are capturing economy wide risk. However, while

HML is supposed to be capturing distress risk (Fama and
French 1993)), it does not adequately capture systemic risk
-59.09%
28.57%

of firms when their exposure to the debt market rises due

HML

to economy-wide problems with over leveraging. To this

extent, LEV adds unique information to the model and does
9
3
4
7
6
3

a good job in capturing systemic risk related to the over

exposure of firms to the debt market.
-28.00%

-5.26%

While a correlation of .51  may not be indicative of

SMB

multicollinearity, it is important to examine if these results

25
25

18
18

are robust to such statistical artefact. In the present context,

0

we do not find a glaring evidence of multicollinearity

because it would have been reflected in high F-statistics
-73.33%

-85.19%

with insignificant t-statistics for the estimated coefficients.

XMKT

Despite the fact that multicollinearity is not an issue, we

15

15
0

4
0
4

decided to estimate the partial F-statistics to check the

robustness of these results to multicollinearity. The partial
% change in significance

significance (by period)

F-statistic determines the incremental explanatory power

Panel C: Crisis period

Negative

Negative
Positive

Positive
Table 9. (Continued)

of adding additional variables to the basic model. In the

Total
Total

present context,  a significant partial F statistic (critical

value is 3.32 at the 1% significance level) provides
justification for adding LEV to the model containing the
% change in
(by model)

traditional FF factors.
Model 1

Model 2

Table 16 reports the partial F-statistics (across all the three

groups - all stock portfolios, financial stocks only portfolios,

62 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 10. Factor loadings for financial and non-financial firms for the aggregate period (January 2000 to April 2009).
Model 1 Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t -1~N( 0 , σ t2 ), e t |ψ t -1~N( 0 , σ t2 ),
q p q p

σ t2 = W + ∑ a i e t2-i + ∑ δ iσ t - j σ t2 = W + ∑ a i e t2-i + ∑ δ iσ t - j
i=1 j =1
i=1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.

Aggregate Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 0.002* 0.709* -0.026 -0.189* -0.060 0.002* 0.709* -0.026 -0.189* -0.001 -0.063
2 0.001 0.379* 0.514* -0.395* -0.098 0.001 0.263* 0.431* -0.277* 0.857* 0.113
3 0.003* 0.783* 0.138* -0.253* 0.060 0.002 0.716* 0.163* -0.189* 1.112* 0.348
4 0.001 0.485* 0.151* 0.337* 0.023 0.001 0.478* 0.159* 0.333* 0.149 0.043
5 0.000 0.473* 0.069 0.203* -0.015 0.000 0.441* 0.066 0.256* 0.419* 0.149
6 0.003* 0.463* 0.082* 0.188* 0.038 0.003* 0.423* 0.085* 0.192* 0.462* 0.204
7 0.000 0.383* 0.173* 0.822* 0.121 0.000 0.409* 0.175* 0.827* -0.105 0.094
8 0.002 0.796* 0.211* 0.687* -0.052 0.001 0.702* 0.193* 0.753* 0.544* 0.121
9 0.002* 0.756* 0.399* 0.783* -0.002 0.001 0.673* 0.368* 0.797* 0.802* 0.267
10 0.002 0.546* -0.315* -0.245* 0.178 0.002 0.535* -0.317* -0.230* 0.112 0.209
11 0.001 0.552* -0.190* -0.125* 0.088 0.001 0.505* -0.220* -0.126* 0.332* 0.214
12 0.002* 0.520* -0.154* -0.302* 0.109 0.001 0.491* -0.139* -0.247* 0.809* 0.359
13 0.001 0.503* -0.120* 0.312* 0.133 0.001 0.501* -0.122* 0.312* 0.030 0.140
14 0.001 0.683* -0.117* 0.164* -0.022 0.001 0.633* -0.127* 0.202* 0.384* 0.123
15 0.002* 0.369* -0.085 0.085 0.071 0.002* 0.356* -0.092* 0.092 0.489* 0.262
16 0.000 0.413* -0.064 0.718* 0.187 0.000 0.423* -0.068 0.713* -0.071 0.175
17 0.003* 0.713* -0.074 0.627* 0.072 0.002* 0.643* -0.108 0.707* 0.602* 0.241
18 0.002* 0.414* -0.112* 0.482* 0.142 0.002* 0.371* -0.177* 0.584* 0.737* 0.408
19 0.001 0.668* -0.591* -0.067 0.316 0.001 0.653* -0.598* -0.070 0.119 0.340
20 0.001* 0.593* -0.582* -0.171* 0.341 0.001 0.564* -0.605* -0.158* 0.169 0.391
21 0.001 0.743* -0.598* 0.007 0.298 0.001 0.706* -0.609* 0.010 0.374* 0.392
22 0.001 0.670* -0.911* 0.035 0.498 0.001 0.661* -0.905* 0.037 0.108* 0.510
23 0.002* 0.693* -0.458* 0.051 0.270 0.001 0.664* -0.475* 0.118 0.324 0.368
24 0.002* 0.740* -0.676* 0.207* 0.420 0.002 0.682* -0.771* 0.190* 0.577* 0.566
25 0.001 0.609* -0.617* 0.711* 0.373 0.001 0.608* -0.617* 0.711* 0.028* 0.375
26 0.001 0.627* -0.639* 0.611* 0.357 0.001 0.579* -0.646* 0.689* 0.452* 0.439
27 0.001 0.679* -1.120* 0.785* 0.478 0.000 0.505* -1.087* 1.113* 1.562* 0.699

and non-financial stock only portfolios) for the aggregate, of LEV is predominantly high during the crisis period with
non-crisis, and crisis periods, respectively. For the combined significant partial F statistics in 27 cases for combined and
stock portfolios, the partial F statistic is significant in 22 non-financial stock portfolios, and in 26 cases for financial
out of 27 portfolios during the aggregate period. During stocks only portfolios. This supports the evidence presented
the non-crisis period, the number of cases of significant earlier suggesting that compared to HML, LEV incorporates
partial F statistics is reduced to 18. Similar results can be additional and unique information concerning distress
seen for financial and non-financial stock portfolios during risk exposure of the firms. In particular, the effect of LEV is
the aggregate and the non-crisis period. However, the effect particularly dominant during the crisis period.

Eds. Hatem A. El-Karanshawy et al. 63

 Bhatt and Sultan

Table 11. Factor loadings for financial and non-financial firms for non-crisis period (January 2000 to June 2007).
Model 1 Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ), 2
t e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Non-crisis Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 0.002* 0.757* 0.110 -0.172* 0.168 0.002* 0.784* 0.110 -0.223* -0.269* 0.170
2 0.001 0.547* 0.922* -0.542* 0.143 0.001 0.521* 0.897* -0.496* 0.375* 0.155
3 0.003* 0.873* 0.365* -0.325* 0.265 0.002 0.867* 0.411* -0.283* 0.673* 0.317
4 0.002 0.586* 0.311* 0.296* 0.262 0.002 0.589* 0.316* 0.294* -0.035 0.260
5 0.000 0.508* 0.167* 0.171* 0.182 0.000 0.499* 0.160* 0.193* 0.186* 0.185
6 0.003* 0.497* 0.145* 0.208* 0.236 0.003* 0.466* 0.136* 0.208* 0.296* 0.257
7 0.000 0.643* 0.362* 0.800* 0.342 -0.001 0.683* 0.443* 0.826* -0.408* 0.333
8 0.001 0.842* 0.354* 0.678* 0.265 0.001 0.834* 0.312* 0.693* 0.359* 0.274
9 0.002* 0.792* 0.547* 0.759* 0.307 0.002 0.753* 0.513* 0.786* 0.477* 0.347
10 0.002* 0.592* -0.252* -0.282* 0.301 0.002* 0.594* -0.251* -0.286* -0.038 0.300
11 0.001 0.600* -0.104* -0.143* 0.224 0.001 0.589* -0.121* -0.143* 0.126 0.232
12 0.002* 0.590* -0.056 -0.311* 0.168 0.002 0.562* -0.055 -0.276* 0.411* 0.215
13 0.001 0.551* -0.054 0.295* 0.299 0.001 0.557* -0.038 0.293* -0.153 0.301
14 0.002* 0.444* -0.095* 0.137* 0.280 0.001 0.732* -0.045 0.149* 0.277* 0.198
15 0.003* 0.362* -0.033 0.030 0.158 0.002* 0.370* -0.035 0.052 0.335* 0.192
16 0.001 0.524* 0.063 0.702* 0.316 0.001 0.553* 0.065 0.666* -0.480* 0.351
17 0.003* 0.778* 0.043 0.588* 0.233 0.003* 0.768* 0.020 0.624* 0.377* 0.249
18 0.002* 0.417* -0.017 0.419* 0.242 0.002* 0.405* -0.020 0.452* 0.417* 0.283
19 0.001 0.727* -0.493* -0.040 0.452 0.001 0.731* -0.483* -0.044 -0.096 0.449
20 0.001* 0.663* -0.454* -0.224* 0.504 0.001* 0.663* -0.454* -0.224* 0.001 0.503
21 0.001 0.776* -0.512* 0.003 0.522 0.001 0.765* -0.535* 0.017 0.209* 0.537
22 0.001 0.745* -0.795* 0.061 0.518 0.002 0.755* -0.781* 0.056 -0.110 0.516
23 0.002* 0.747* -0.367* -0.008 0.376 0.002* 0.749* -0.366* 0.009 0.088 0.378
24 0.002 0.789* -0.523* 0.235* 0.442 0.002 0.773* -0.544* 0.234* 0.212* 0.461
25 0.001 0.655* -0.380* 0.651* 0.236 0.001 0.668* -0.360* 0.602* -0.436* 0.253
26 0.002 0.728* -0.472* 0.497* 0.272 0.002 0.726* -0.472* 0.525* 0.184 0.268
27 0.002 0.780* -0.850* 0.655* 0.347 0.001 0.712* -0.862* 0.956* 1.241* 0.497

Factor loadings by types of firms exchanges as meeting desired criteria for various style of
Previously, we noted that LEV is a good proxy for distress investing. For instance, the Dow Jones classifies investing
risk across financial and non-financial stocks. We find that in certain stocks (popular household names) under broad
stocks have similar sensitivities to the leverage risk factor. categories such as socially responsible investing because
In this section, we further conduct an additional robustness these firms promote social, environmental, and corporate
check to examine whether there are differences in the responsibility. To this extent, we consider conventional,
way various categories of firms respond to the economy Islamic and Socially Responsible Investing (SRI) stocks,
wide risk factors because they are classified by stock where each group exhibits distinct characteristics27. There

64 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 12. Factor loadings for financial and non-financial firms for the crisis period (July 2007 to April 2009).
Model 1 Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ), 2
t e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Crisis period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 -0.004 -0.126 -0.577* -0.168 0.130 -0.003 -0.154 -0.283 0.051 0.529 0.203
2 -0.003 -0.051 -0.644* -0.708* 0.137 0.001 -0.104 -0.002 -0.187 1.207* 0.357
3 -0.005 0.596* -1.159* -1.536* 0.348 -0.002 0.266* -0.263* -0.952* 1.642* 0.653
4 -0.003 0.101 -0.197 0.265 0.069 -0.003 0.048 -0.065 0.336 0.257 0.102
5 -0.005 0.332* -0.794* -0.222 0.217 -0.003 0.074 -0.140 0.184 1.169* 0.483
6 -0.005 0.390* -0.636* -0.394 0.136 -0.003 0.201 0.227 0.131 1.551* 0.408
7 -0.005 0.092 -0.115 0.554* 0.059 -0.004 0.069 -0.027 0.619* 0.167 0.085
8 -0.006* -0.034 -0.553* 0.103 0.144 -0.004 0.284* -0.184 0.241 0.813* 0.251
9 -0.003 0.252 -0.724* 0.190 0.292 0.001 0.178 0.094 0.978* 1.691* 0.542
10 -0.005 0.344* -0.998* -0.588 0.361 -0.003 0.238* -0.695* -0.404 0.637* 0.449
11 -0.004 0.261* -1.037* -0.531 0.398 -0.003 0.140 -0.664* -0.276 0.733* 0.506
12 -0.006 0.257* -1.394* -1.299* 0.499 -0.004 0.133 -0.611* -0.809* 1.446* 0.696
13 -0.004 0.239* -0.568* -0.145 0.238 -0.004 0.193* -0.434* -0.083 0.228 0.277
14 -0.003 0.266* -1.062* -0.332 0.350 -0.002 0.152 -0.544* 0.050 0.843* 0.466
15 -0.009* 0.326* -1.256* -0.687 0.383 -0.005 0.103 -0.422* -0.132 1.616* 0.603
16 -0.002 0.143 -0.338* 0.393* 0.231 -0.002 0.054 -0.161 0.442* 0.268* 0.265
17 -0.004 0.203* -1.145* -0.110 0.423 -0.002 0.140 -0.600* 0.421 1.005* 0.539
18 -0.007* 0.304* -1.314* 0.109 0.488 -0.006 0.087 -0.596* 0.519* 1.275* 0.671
19 -0.001 0.308* -1.277* -0.542 0.469 0.000 0.227 -0.869* -0.225 0.758* 0.543
20 -0.001 0.207 -1.383* -0.437 0.552 0.000 0.107 -0.913* -0.011 0.917* 0.644
21 -0.006 0.264 -1.458* -1.107* 0.473 -0.003 0.118 -0.837* -0.497* 1.275* 0.626
22 -0.003 0.292* -1.459* -0.528 0.614 -0.002 0.178* -1.127* -0.292 0.615* 0.661
23 -0.004 0.197* -1.537* -0.452 0.625 -0.002 0.107 -0.881* 0.056 1.182* 0.736
24 -0.005 0.154 -1.729* 0.039 0.704 -0.001 0.106 -1.108* 0.268 1.275* 0.809
25 -0.006 0.235 -1.570* 0.070 0.647 -0.005 0.130 -1.146* 0.462 0.725* 0.671
26 -0.001 0.034 -1.683* 0.908* 0.695 0.002 -0.052 -0.970* 1.247* 1.255* 0.770
27 -0.008* 0.393* -2.280* 0.121 0.720 -0.005 0.161 -1.374* 0.854* 1.686* 0.826

are distinct differences among these groups with respect to stocks are withheld though some of the common household
the fundamentals such as size, ROA, ROE, leverage, return names in the US may be classified as Islamic stocks because
on capital, PE ratio, and EPS. (See Milly and Sultan (2009) they meet the requirements set by the Dow Jones Shariah
for further evidences.) Board. On October 29, 2010, the market capitalization of
the Dow Jones Islamic World Index was $20  billion with
We use stocks included in the Dow Jones Islamic Index 2,369  stocks. The weights (%) for some of the major
(DJIM) which is a proprietary index of stocks classified as countries in the index are as follows: US (50.54), UK
Islamic stocks by the Dow Jones Shariah Board. Because of (6.71), Japan (5.42), Canada (5.27), Switzerland (3.45),
proprietary nature of such classifications, the names of the Australia (3.26), France (2.97), India (2.5), Taiwan (2.2),

Eds. Hatem A. El-Karanshawy et al. 65

 Bhatt and Sultan

Table 13. Factor loadings for the non-financial stock portfolios for aggregate period (January 2000 to April 2009).
Model 1 Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ), 2
t e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.

Aggregate Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 0.002* 0.776* 0.453* -0.425* 0.076 0.002 0.774* 0.468* -0.395* -0.196 0.057
2 0.002* 0.646* 0.380* -0.097 0.011 0.002 0.630* 0.391* -0.246* 0.679* 0.095
3 0.000 0.698* 0.614* 0.048 0.041 0.000 0.686* 0.548* -0.185* 1.123* 0.166
4 0.001 0.584* 0.345* 0.215* 0.023 0.001 0.587* 0.342* 0.206* 0.043 0.023
5 0.000 0.668* 0.533* 0.376* 0.048 0.000 0.677* 0.521* 0.336* 0.211* 0.071
6 0.000 0.775* 0.570* 0.586* 0.014 0.000 0.779* 0.554* 0.551* 0.181 0.031
7 0.000 0.621* 0.594* 0.690* 0.109 0.000 0.626* 0.613* 0.708* -0.101 0.094
8 0.000 0.741* 0.729* 0.787* 0.075 0.000 0.748* 0.723* 0.752* 0.136 0.083
9 0.000 0.838* 0.812* 0.893* 0.103 0.000 0.871* 0.818* 0.767* 0.454* 0.130
10 -0.001 0.958* 0.098 -0.309* -0.054 -0.001 0.943* 0.112 -0.276* -0.161 -0.060
11 0.001 0.709* 0.094 -0.144 0.025 0.001 0.727* 0.084 -0.197* 0.238* 0.047
12 0.000 0.704* 0.021 -0.094 0.004 0.000 0.740* -0.007 -0.224* 0.558* 0.064
13 0.000 0.682* 0.092 0.281* -0.016 0.000 0.696* 0.081 0.246* 0.189 -0.007
14 0.001 0.629* 0.026 0.287* 0.036 0.001 0.673* -0.019 0.185* 0.463* 0.088
15 0.000 0.665* 0.168* 0.474* 0.032 0.000 0.692* 0.125* 0.358* 0.531* 0.100
16 0.000 0.686* 0.341* 0.790* 0.056 0.000 0.687* 0.331* 0.757* 0.137 0.065
17 0.000 0.723* 0.270* 0.641* 0.051 0.000 0.753* 0.210* 0.528* 0.498* 0.110
18 0.001 0.848* 0.253* 0.796* 0.033 0.000 0.856* 0.182* 0.699* 0.809* 0.140
19 0.001 0.745* -0.459* -0.372* 0.194 0.001 0.743* -0.454* -0.347* -0.166 0.187
20 0.000 0.638* -0.372* -0.008 0.054 0.000 0.651* -0.392* -0.053 0.369* 0.098
21 0.000 0.589* -0.353* 0.000 0.102 0.000 0.604* -0.364* -0.131 0.628* 0.185
22 0.001 0.685* -0.394* 0.224* 0.043 0.001 0.685* -0.396* 0.210* 0.049 0.043
23 0.001 0.672* -0.243* 0.257* 0.031 0.001 0.682* -0.301* 0.178* 0.493* 0.094
24 0.001 0.647* -0.165* 0.438* -0.003 0.000 0.659* -0.195* 0.312* 0.504* 0.067
25 0.001 0.666* -1.074* 1.002* 0.236 0.001 0.672* -1.089* 1.105* -0.505* 0.236
26 0.001 0.772* -0.187* 0.701* 0.054 0.000 0.767* -0.233* 0.585* 0.814* 0.140
27 0.002 0.791* -0.408* 0.796* 0.125 0.001 0.794* -0.443* 0.589* 1.068* 0.241

Germany (1.73), South Korea (1.56), Brazil (1.5), Russia relative to conventional investment strategies, which lack
(1.47), China (1.39), Hong Kong (1.25), and Sweden such ethical ambition. SC stocks are popular among a
(1.09). Among some of the traditionally Muslim majority new class of investors that, in addition to profit motives,
countries, the weights are: Malaysia (.35), Kuwait (.22), is also driven by their desire to live ethically and invest
Qatar (.08), UAE (.03), and Bahrain (.01). morally. Compared to the conventional Western financial
system, Islamic finance is a newcomer to the global
Our selection of SC stocks is in line with the recent financial world, encompassing somewhere between
interest in the performance of faith based investing, with $750  billion to $1 trillion of investments in firms and
its overarching goal to promote the betterment of society, projects that are classified as SC. Yet, over the past few

66 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 14. Factor loadings for non-financial stocks portfolio for non-crisis period (January 2000 to June 2007)
Model 1 Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ), 2
t e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Non-crisis Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 0.002 0.846* 0.462* -0.390* 0.483 0.002 0.830* 0.482* -0.321* -0.364* 0.496
2 0.002* 0.732* 0.415* -0.098 0.238 0.002* 0.741* 0.407* -0.164 0.360* 0.251
3 0.000 0.782* 0.627* 0.115 0.347 0.000 0.821* 0.599* -0.126 0.923* 0.371
4 0.001 0.673* 0.363* 0.268* 0.406 0.001 0.666* 0.366* 0.278* -0.057 0.408
5 0.000 0.718* 0.518* 0.415* 0.451 0.000 0.726* 0.513* 0.394* 0.109 0.450
6 0.000 0.825* 0.573* 0.613* 0.453 0.000 0.835* 0.561* 0.590* 0.130 0.451
7 -0.001 0.758* 0.638* 0.673* 0.551 -0.001 0.750* 0.647* 0.719* -0.147 0.554
8 0.000 0.789* 0.729* 0.815* 0.527 0.000 0.795* 0.726* 0.791* 0.092 0.525
9 0.000 0.903* 0.807* 0.953* 0.471 0.000 0.939* 0.800* 0.862* 0.385* 0.469
10 -0.001 1.023* 0.121 -0.323* 0.465 -0.001 0.982* 0.125 -0.243* -0.292* 0.480
11 0.001 0.775* 0.057 -0.095 0.458 0.001 0.790* 0.051 -0.127 0.150 0.457
12 0.000 0.764* 0.044 -0.038 0.357 0.000 0.812* 0.026 -0.125 0.401* 0.365
13 0.000 0.756* 0.140* 0.277* 0.401 0.000 0.766* 0.136* 0.255* 0.111 0.396
14 0.001 0.689* 0.026 0.328* 0.411 0.001 0.731* 0.000 0.257* 0.366* 0.419
15 0.000 0.711* 0.160* 0.499* 0.444 0.000 0.751* 0.132* 0.408* 0.440* 0.458
16 0.000 0.759* 0.368* 0.798* 0.475 0.000 0.761* 0.366* 0.791* 0.031 0.473
17 0.000 0.759* 0.280* 0.657* 0.495 0.000 0.794* 0.240* 0.570* 0.394* 0.502
18 0.001 0.898* 0.249* 0.801* 0.447 0.001 0.932* 0.192* 0.726* 0.710* 0.461
19 0.000 0.849* -0.432* -0.375* 0.549 0.000 0.912* -0.380* -0.428* -0.402* 0.565
20 0.000 0.664* -0.367* 0.013 0.475 0.000 0.676* -0.385* -0.026 0.312* 0.486
21 0.000 0.639* -0.323* -0.002 0.414 -0.001 0.769* -0.400* -0.109 0.649* 0.418
22 0.001 0.801* -0.370* 0.278* 0.403 0.001 0.799* -0.369* 0.290* -0.048 0.404
23 0.001 0.727* -0.242* 0.308* 0.460 0.001 0.769* -0.268* 0.242* 0.396* 0.470
24 0.000 0.699* -0.173* 0.472* 0.416 0.000 0.731* -0.208* 0.381* 0.422* 0.435
25 0.001 0.820* -1.110* 1.026* 0.436 0.001 0.764* -1.118* 1.202* -0.918* 0.492
26 0.001 0.927* -0.342* 0.879* 0.358 0.001 0.977* -0.360* 0.763* 0.552* 0.358
27 0.001 0.865* -0.388* 0.849* 0.410 0.001 0.890* -0.412* 0.656* 0.885* 0.444

years Islamic investments have become more competitive market capitalization of $300  billion and numerous
and consequently attractive not only to Muslim but also traditional US financial institutions joining to partake in
non-Muslim investors seeking alternative investments this development.
opportunities, which live up to high ethical as well as
nominal performance standards. As a result, the number Similarly, the SRI class of stocks is a relative newcomer,
of Islamic mutual funds and exchange traded funds which has gained popularity in recent years. In the
world-wide has increased considerably from merely 8 early 2000s, we have seen a dramatic interest in socially
before 1992 to more than 300 in 2008, with an estimated responsible investing that poured billions of dollars

Eds. Hatem A. El-Karanshawy et al. 67

 Bhatt and Sultan

Table 15. Factor loadings for non-financial stock portfolios during crisis period (July 2007 to April 2009).
Model 1 Model 2

rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t

e t |ψ t -1~N(0 , σ t2 ), e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.

Crisis Period

Model 1 Model 2

Adj. Adj.
Portfolio Intercept MKT SMB HML R-square Intercept MKT SMB HML LEV R-square

1 -0.005 0.156 0.586 -1.208* -0.092 -0.004* 0.051 0.718* -1.277* 2.808* 0.391
2 -0.005 0.124 0.500 -0.967* -0.087 -0.003* 0.044 0.832* -1.075* 2.866* 0.442
3 -0.004 0.043 0.392 -0.652 -0.092 -0.002 0.017 0.756* -0.717* 3.192* 0.535
4 -0.004 0.062 -0.008 0.459 -0.073 -0.004 0.047 0.377* -0.081 1.710* 0.292
5 -0.005 0.086 0.047 0.359 -0.043 -0.005 0.049 0.537* -0.216 2.292* 0.439
6 -0.006 0.143 0.427 0.191 -0.030 -0.005 0.053 0.733* -0.074 2.438* 0.417
7 -0.004 0.079 0.233 0.579* 0.033 -0.003 -0.019 0.540* 0.307 1.427* 0.315
8 -0.004 0.110 0.393 0.305 0.001 -0.003 -0.008 0.716* 0.141 2.412* 0.471
9 -0.004 0.151 0.646 0.032 -0.040 -0.002 0.018 0.840* 0.150 3.242* 0.575
10 -0.005 -0.004 -0.268 0.055 -0.066 -0.002 0.078 0.212 -0.604* 2.539* 0.417
11 -0.006* -0.004 -0.549* 0.112 -0.066 -0.004 0.019 0.055 -0.769* 2.855* 0.479
12 -0.005 0.047 -0.535 -0.161 -0.043 -0.003 0.070 -0.087 -0.736* 2.957* 0.504
13 -0.004 0.025 -0.410 0.674* -0.073 -0.003 0.092 0.032 -0.116 1.839* 0.332
14 -0.005 0.041 -0.471 0.241 -0.042 -0.003 0.034 -0.056 -0.287 2.786* 0.487
15 -0.006* 0.064 -0.234 0.221 -0.045 -0.003 0.032 0.147 -0.227 2.963* 0.515
16 -0.004 0.033 -0.027 0.539 -0.010 -0.002 -0.008 0.445* 0.150 1.938* 0.359
17 -0.005 0.015 -0.527 0.772* -0.022 -0.002 -0.014 -0.015 0.161 3.016* 0.533
18 -0.006 -0.004 -0.415 0.834* 0.004 -0.003 0.090 0.104 0.035 4.151* 0.612
19 -0.004 0.177 -0.643* -0.673* -0.020 -0.002 0.054 -0.451* -0.773* 2.140* 0.444
20 -0.004 0.034 -0.667* -0.626* -0.020 -0.003 -0.012 -0.255 -0.933* 2.589* 0.445
21 -0.005* 0.114 -0.818* -0.241 -0.009 -0.003 0.058 -0.427* -0.616* 2.331* 0.487
22 -0.003 0.118 -0.640* -0.076 -0.070 -0.003 -0.021 -0.302 -0.430 1.827* 0.239
23 -0.004 0.098 -0.543 -0.309 -0.040 -0.003 0.000 -0.276 -0.420 2.608* 0.448
24 -0.005 0.101 -0.468 0.036 -0.044 -0.003 0.017 -0.119 -0.201 2.566* 0.453
25 -0.003 -0.033 -0.337 0.756* 0.039 -0.003 -0.013 -0.085 0.351 1.865* 0.331
26 -0.006* 0.171 -1.072* 0.708* 0.010 -0.004 0.141 -0.379* 0.093 2.762* 0.488
27 -0.005 -0.018 -0.622 -0.079 -0.024 -0.003 0.002 -0.546* 0.192 3.637* 0.609

into companies known for their efforts to offer ethical conceived in a religious context as well, socially responsible
investments and projects that promoted environmental investing has expanded to take in consideration “the so-
sustainability. In terms of the portfolio allocation and called ‘triple bottom line’, commonly known as the ‘three
structure, Islamic and socially responsible investing (SRI) P’s rule: people, planet and profit’” (Forte & Miglietta,
stocks exhibit strong similarities, whereas conventional 2007, p. 3). Most recently, assets under SRI management
stocks are not subject to any other qualitative or were estimated to have increased “from $639  billion in
quantitative constraints. Although SRI funds were initially 1995 … to $2.71 trillion in 2007”, while “assets in all types

68 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Table 16. Partial f-statistics testing for the significance of contribution made by the LEV factor.
Restricted Model Unrestricted Model

rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t

e t |ψ t -1~N(0 , σ t2 ), e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week
of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking
the simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. Partial f-statistics and the
p-values test for the significance in the contribution of R-square made by the new model (which includes the LEV factor). The factors
SMB, HML and LEV have been rebalanced for financial stock portfolios and non financial stock portfolios. (*) indicates significance at 5%
level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors. Model 1 excludes
LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are available upon
request.

All stock portfolios Financial stock portfolios Non-financial stock portfolios

Aggregate Non-crisis Crisis Aggregate Non-crisis Crisis Aggregate Non-crisis Crisis

period period period period period period period period period

Partial Partial Partial Partial Partial Partial Partial Partial Partial

Portfolio f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic

1 -1.62 1.04 71.90* -0.17 2.04 9.64* -8.36 10.65* 75.57*

2 56.73* 5.39* 82.82* 115.83* 6.41* 33.22* 45.46* 7.58* 90.27*
3 101.36* 28.68* 116.58* 214.21* 30.94* 83.68* 72.96* 16.11* 127.92*
4 14.96* -0.01 45.12* 11.11* -0.07 4.60* 1.26 1.73 49.54*
5 34.46* 7.13* 75.91* 94.36* 2.16 49.55* 12.75* 0.13 81.80*
6 43.57* 13.12* 70.42* 101.84* 11.67* 44.31* 9.16* -0.18 73.26*
7 -15.57 0.34 36.19* -13.06 -3.84 3.66 -6.98 2.99 39.88*
8 19.58* 2.22 75.98* 95.79* 5.93* 14.53* 5.07* -0.88 84.74*
9 35.62* 4.19* 103.75* 178.07* 24.36* 52.41* 15.69* -0.56 137.36*
10 6.96* -0.11 76.02* 19.93* 0.39 16.06* -1.71 12.03* 78.90*
11 37.50* 8.58* 87.33* 78.43* 4.73* 21.61* 12.12* 0.92 99.36*
12 66.64* 23.42* 89.13* 189.67* 24.14* 62.10* 31.73* 5.81* 104.67*
13 25.01* 2.55 50.96* 4.82* 2.34 6.09* 5.04* -1.87 58.03*
14 49.24* 12.04* 89.40* 80.92* -38.57 21.53* 28.36* 6.12* 97.85*
15 68.41* 26.13* 93.21* 126.42* 17.04* 53.15* 37.70* 10.87* 109.78*
16 3.29 0.71 36.76* -6.22 22.03* 5.42* 5.44* -0.52 55.13*
17 46.24* 12.02* 105.33* 108.33* 9.49* 24.76* 33.00* 6.07* 112.73*
18 80.45* 21.10* 108.63* 217.60* 23.24* 53.18* 61.30* 11.21* 148.07*
19 -4.56* 9.18* 68.14* 18.88* -0.99 16.25* -3.45 15.24* 79.61*
20 31.43* 8.18* 67.16* 41.14* 0.03 25.47* 24.38* 9.57* 79.64*
21 49.63* 18.60* 83.87* 75.18* 13.06* 39.69* 50.31* 3.95* 91.88*
22 22.87* 2.70 39.97* 12.65* -0.83 13.91* 1.05 1.35 39.27*
23 49.86* 21.54* 72.94* 75.48* 2.52 40.51* 34.80* 8.93* 84.35*
24 48.31* 21.56* 78.73* 162.88* 15.23* 52.25* 37.14* 13.95* 86.59*
25 -12.15* 28.19* 47.55* 2.89* 9.85* 7.90* 0.62 43.70* 42.22*
26 50.41* 1.07 87.66* 72.14* -0.97 31.74* 49.04* 1.20 88.80*
27 118.71* 44.78* 131.34* 355.94* 116.64* 58.15* 74.25* 25.09* 153.14*

of socially and environmentally screened funds [… in the Socially responsible investing (SRI) is an investment
US] rose to $201.8  billion.” (2007 Report on Socially process that considers the social and environmental
Responsible Investing Trends in the United States, 2008, consequences of investments, both positive and negative,
p. ii) The premise of the “three P’s rule” is reflected in a within the context of rigorous financial analysis… It is a
definition of socially responsible investing, which can be process of identifying and investing in companies that meet
found in the 2005 Report on SRI Trends in the United States certain standards of Corporate Social Responsibility (CSR)
released by the Social Investment Forum:

Eds. Hatem A. El-Karanshawy et al. 69

 Bhatt and Sultan

(2004 Report on Socially Responsible Investing Trends in Factor loadings of Islamic stocks
the United States. 10 Year Review, 2005, p. 2). Panel B reports the results for the Islamic group of stocks.
Compared to the aggregate period, there is a remarkable
The congruence of Islamic and SRI stocks stems from the change in the number of cases of where XMKT is significant
fact that both do not have profit maximization as their sole (-87.6% at the firm and by -92.59% at the portfolio level).
objective, but rather strive to achieve a paramount, ethical The change in significance for SMB is as follows: 9.89%
obligation and a social-utilitarian function. In the case of at the firm and -44.44% at the portfolio level. The results
Islamic funds, the religious responsibilities and regulations for the HML are again consistent across both the firm and
outlined in the Shariah, take precedence over profit in the portfolio level. The change in statistical significance
order to further the establishment of a just and moral for HML is as follows: -56.44% at the firm and by -60%
Islamic economic system and ultimately society. at the portfolio level. Finally, the number of instances
LEV is significant increases by 98.9% at the firm and by
In contrast, profit maximization is the dominant objective 73.33% at the portfolio level. Again, our results are quite
in traditional fund management. Conventional equity consistent with the previous results reported without
portfolio strategies include neither positive nor negative the index classifications. Islamic stocks behave similar to
screens, whose purpose it is to align the portfolio with certain the conventional stocks when it comes to sensitivities to
ethical, qualitative standards. As such, conventional funds economic risk factors.
are not subject to the qualitative screening procedures that
are so imperative to Islamic and SRI funds. Additionally, Factor loadings of SRI stocks
Islamic funds differ from SRI and conventional ones, since In Panel C, we report the results for 238  stocks classified
their provisions incorporate quantitative screens that are as SRI group of stocks. Compared to the previous groups,
directly based on ethical paradigms found in the Shariah. we have some unusual results. We find that, compared to
Furthermore, Islamic funds have to comply with certain the aggregate period, the number of instances where the
income purification requirements, which are derived from XMKT is significant drops by 89% at the firm and by 100%
the teachings of the Holy Quran and Sunnah. at the portfolio level. For SMB, the changes in the number
of significant cases are: -66% (firm-level) and -80.95%
The hypothesis tested is that high leverage increases (portfolio-level). In contrast to our previous results, the
exposure to the credit market and subsequently translates number of instances where HML is significant at the firm
into shareholders demanding higher risk premium. Recall level increases by 56.43% and by 35% at the portfolio level.
that Islamic stocks have low leverage, they are significantly Finally, the number of instances where LEV is significant
more asset-backed than conventional firms, and are drops by 11.11% the firm and by 56.25% at the portfolio
not involved in the business of speculation, production level.
of weapons, alcohol, pork, and entertainment. More
specifically, Islamic funds typically screen out companies With respect to the effects of LEV risk factor, the results for
with excessive reliance on debt, where the typical the SRI group are quite different from the Conventional
maximum level of total debt to market capitalization is set and Islamic stocks, suggesting that stocks in this category
at 33 percent28. are less sensitive to the economy-wide leverage risk
factor. Certainly, leverage risk for this type of firms is not
The first step towards applying our leverage risk factor to unusually different but perhaps the nature of the business
these index classifications is to recreate the FF and LEV these firms are involved may make it less susceptive to
specific to each category of stocks. This is followed by economy wide leverage risk. It may also be possible that
estimating GARCH regressions at the firm and portfolio during the financial crisis, while socially responsible
level. investing would have earned positive risk premium with
respect the HML, SRI investors would have earned a
Factor loadings of conventional stocks negative risk premium when leverage was employed as
In Table  17, we report a summary of firm and portfolio a stock picking strategy. Whether SRI investing produces
specific regressions by groups. The first panel reports the a lower return because these stocks are generally less
results for the firms belonging to the conventional stock sensitive to the economy wide risk factors suggests that
category. We find that at the firm level, the inclusion of these stocks may offer significant diversification benefits.
LEV produced some interesting results. Compared to the Overall, further research along these lines would offer
aggregate period, the number of instances where the more clues as to why SRI stocks have negative risk
XMKT is significant drops by 79.57% at the firm and by premium for leverage risk.
100% at the portfolio level. The change in significance
for SMB is as follows: 8.89% at the firm and -17.39% at Partial F-test
the portfolio level. The results for the HML are consistent Table  18 reports partial F-statistics (across all the three
across both the firm and the portfolio level. The number of groups - all stock portfolios), for the aggregate, non-crisis,
instances where HML is significant at the firm level drops and crisis periods  in order to test for the significance of
by 56.13% and by 44.44% at the portfolio level. Finally, the the contribution made by LEV. For the combined stock
number of instances LEV is significant increases by 231% portfolios, the partial F statistic is significant in 24 out of
at the firm and by 58.82% at the portfolio level. Overall, 27 portfolios during the aggregate period. During the non-
these results are qualitatively similar to the ones reported crisis period, the number of significant partial F statistics
earlier and confirm our earlier finding that the inclusion of is reduced to 17 cases. However, the effect of LEV is
LEV subsumes the effects of the traditional FF factors to a prominent during the crisis period with significant partial
great extent. F statistics in all 27 portfolios. For the Islamic stocks, the

70 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 17.  Factor loadings by types of firms.
Model 1 Model 2

rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t

e t |ψ t -1~N(0 , σ t2 ), e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking the
simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of
the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average of the returns
each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with (*) are significant at 5% level of significance.

Eds. Hatem A. El-Karanshawy et al.

PANEL A
Conventional stocks: 2308 Conventional portfolios: 27
Aggregate Period Aggregate Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 1757 8 1023 Model 1 Positive 27 4 18
Negative 0 1439 225   Negative 0 17 0
Total 1757 1447 1248   Total 27 21 18
Model 2 Positive 1697 8 797 1208 Model 2 Positive 27 2 12 20
Negative 0 1582 482 84 Negative 0 18 3 0
Total 1697 1590 1279 1292 Total 27 20 15 20
A comparison among Islamic, conventional, and socially responsible stocks

%change in significance −3.41% 9.88% 2.48% %change in significance 0.00% -4.76% -16.67%
(by model) (by model)
Non-crisis Period Non-crisis Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 2076 7 917 0 Model 1 Positive 27 7 17
Negative 0 1108 378   Negative 0 14 0
Total 2076 1115 1295   Total 27 21 17
Model 2 Positive 2056 6 863 459 Model 2 Positive 27 7 15 15
Negative 0 1152 499 80 Negative 0 16 3 2
Total 2056 1158 1362 539 Total 27 23 18 17
%change in significance −0.96% 3.86% 5.17% %change in significance 0.00% 9.52% 5.88%
(by model) (by model)

71
(Continued)
72
Table 17. (Continued)
Crisis period Crisis period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 290 262 1026 0 Model 1 Positive 10 0 27
Negative 76 996 12   Negative 0 16 0
Total 366 1258 1038   Total 10 16 27
Model 2 Positive 92 525 480 1744 Model 2 Positive 0 5 9 27
Negative 328 736 115 41 Negative 0 14 1 0
Total 420 1261 595 1785 Total 0 19 10 27
%change in significance 14.75% 0.24% -42.68% %change in significance 0.00% 18.75% -62.96%
(by model) (by model)

%change in significance -79.57% 8.89% -56.31% 231.17% %change in significance -100.00% -17.39% -44.44% 58.82%
(by period) (by period)

PANEL B
Islamic stocks: 1161 Islamic portfolios: 1161
Aggregate Aggregate Period
Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 1004 402 252 Model 1 Positive 27 12 12
Negative 0 214 196   Negative 0 5 8
Total 1004 616 448   Total 27 17 20
Model 2 Positive 1019 401 248 388 Model 2 Positive 27 12 12 10
Negative 0 244 239 202 Negative 0 7 9 4
Total 1019 645 487 590 Total 27 19 21 14
%change in significance 1.49% 4.71% 8.71% %change in significance 0.00% 11.76% 5.00%
(by model) (by model)

Islamic banking and finance – Essays on corporate finance, efficiency and product development
Bhatt and Sultan
 Non-crisis Period Non-crisis Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 1087 236 259 Model 1 Positive 27 10 13
Negative 0 276 234   Negative 0 7 8
Total 1087 512 493   Total 27 17 21
Model 2 Positive 1089 254 265 178 Model 2 Positive 27 11 12 4
Negative 0 254 263 280 Negative 0 7 8 11
Total 1089 508 528 458 Total 27 18 20 15
%change in significance 0.18% −0.78% 7.10% %change in significance 0.00% 5.88% −4.76%

Eds. Hatem A. El-Karanshawy et al.

(by model) (by model)
Crisis period Crisis period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 83 508 191 Model 1 Positive 0 0 7
Negative 33 48 64   Negative 0 11 3
Total 116 556 255   Total 0 11 10
Model 2 Positive 92 489 133 873 Model 2 Positive 2 10 6 26
Negative 43 69 97 38 Negative 0 0 2 0
Total 135 558 230 911 Total 2 10 8 26
%change in significance 16.38% 0.36% −9.80% %change in significance 0.00% −9.09% −20.00%
A comparison among Islamic, conventional, and socially responsible stocks

(by model) (by model)

%change in significance −87.60% 9.84% −56.44% 98.91% %change in significance -92.59% −44.44% −60.00% 73.33%
(by period) (by period)
(Continued)

73
74
Table 17. (Continued)
PANEL C
SRI stocks: 238 SRI Portfolios: 27
Aggregate Period Aggregate Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 224 80 183 Model 1 Positive 27 9 20
Negative 0 69 0   Negative 0 10 1
Total 224 149 183   Total 27 19 21
Model 2 Positive 224 79 191 12 Model 2 Positive 27 11 21 8
Negative 0 69 0 51 Negative 0 10 2 9
Total 224 148 191 63 Total 27 21 23 17
%change in significance 0.00% -0.67% 4.37% %change in significance 0.00% 10.53% 9.52%
(by model) (by model)
Non-crisis Period Non-crisis Period
XMKT SMB HML LEV XMKT SMB HML LEV
Model 1 Positive 235 78 124 0 Model 1 Positive 27 9 17
Negative 0 82 11   Negative 0 10 3
Total 235 160 135   Total 27 19 20
Model 2 Positive 237 76 133 18 Model 2 Positive 27 11 17 8
Negative 0 83 7 27 Negative 0 10 3 8
Total 237 159 140 45 Total 27 21 20 16
%change in significance 0.85% −0.63% 3.70% %change in significance 0.00% 10.53% 0.00%
(by model) (by model)

Islamic banking and finance – Essays on corporate finance, efficiency and product development
Bhatt and Sultan
A comparison among Islamic, conventional, and socially responsible stocks

results are quite strong. The number of cases F-statistics is

−56.25%
significant is 17 (aggregate period), 15 (non-crisis period),

LEV
and 27 (crisis period). Finally, we find that the SRI stocks

7
0
7
only portfolios are not sensitive to the leverage risk during
the credit crisis.

Overall, the regression results suggest that while the

35.00%
0.00%
HML

sensitivities of the portfolio returns to the FF factors are

27

27
27

27
significant during the aggregate and the non-crisis periods,
0

0
there are important changes in the sign and significance of
these factors during the crisis period. Their significance also
weakens with the introduction of leverage as a risk factor,

−80.95%
0.00%
almost to the tune of being subsumed by the leverage risk
SMB

factor. The effects of the market factor are persistent before

3
1
4
2
2
4

the crisis period but surprisingly became insignificant

during the crisis period. Leverage factor is consistently
significant across all the periods and its effect is more
prominent during the crisis period due to the greater debt
−100.00%
exposure of the firms and higher macroeconomic risk. The
XMKT

0.00%

results further support the conclusions drawn in the earlier

tables.
0
0
0
0
0
0

Leverage risk factor for US stocks

%change in significance

%change in significance
Negative

Negative

A potential shortcoming of the preceding results is

Positive

Positive

due to the fact that our previous samples include

Total

Total

stocks traded globally and may not accurately quantify

the effects of the credit crisis on the US market. We
Crisis period

therefore conduct another experiment using US

(by model)

(by period)

stocks only. This additional exercise is carried out

Model 1

Model 2

by excluding all non-US stocks, creating traditional

Fama-French factors, and adding our newly created
LEV factor to represent financial distress. In addition,
we also estimate GARCH regressions to demonstrate
that our risk factors represent macroeconomic shocks
−11.11%

as well. While the results are not presented to save

space, we can summarize the results as follows. We
LEV

28
12
40

find that for predicting US industrial production, the

 
 

following variables are statistically significant: SMB,

HML, XMKT, HML (t-2), and LEV (t-2). For predicting
US unemployment rate, variables such as SMB, HML,
−1.35%

56.43%

LEV and several interaction variables on LEV (for 2008

HML
222

222
219

219

credit crisis period) are significant at various lag length.

0

We find that several interaction variables using LEV are

significant in predicting the credit spread and term
spread. Finally, dummy variables on LEV are highly
significant in affecting changes in the US inflation
−66.04%
12.50%

rate. Overall, it is safe to conclude that the risk factors

SMB

contain adequate information on the US economy.

36
12
48
39
15
54

Next, GARCH regressions are estimated. For the aggregate

period (Jan2000 – April 2009), out of 27 portfolios, SMB
−89.03%
44.44%

has 18 (6) positive (negative) coefficients. HML has 16

XMKT

(8) positive (negative) coefficients. In contrast, for Model

10
18
12
14
26

2 (where we add LEV), the results are as follows: SMB

8

has 18 positive and 4 negative coefficients, HML has

17 positive and 9 negative coefficients, and LEV has 6
%change in significance

%change in significance
Negative

Negative

positive and 17 negative coefficients. For the non-crisis

Positive

Positive

period (January 2000 – June 2007), SMB has 18 (7)

Total

Total

positive (negative) coefficients. HML has 13 (10) positive

(negative) coefficients. In contrast, for Model 2, SMB
Crisis period

has 18 positive and 7 negative coefficients, HML has

(by model)

(by period)

14 positive and 9 negative coefficients, and LEV has 6

Model 1

Model 2

positive and 18 negative coefficients. In all cases, there

is a significant increase in R2 when we add LEV in the
model.

Eds. Hatem A. El-Karanshawy et al. 75

 Bhatt and Sultan

Table 18. Partial f-statistics testing for the significance of contribution made by the LEV factor.
Old Model New Model
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ), 2
t e t |ψ t -1~N(0 , σ t2 ),
q p q p

σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1 j =1 i =1 j =1

where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week
of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking
the simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients
with (*) are significant at 5% level of significance. Partial f-statistics and the p-values test for the significance in the contribution of
R-square made by the new model (which includes the LEV factor). GARCH models are estimated using the Bollerslev-Wooldridge
corrections to the standard errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not
reported to conserve space. They are available upon request.

Conventional Portfolios Islamic Portfolios SRI Portfolios

Aggregate Non-crisis Crisis Aggregate Non-crisis Crisis Aggregate Non-crisis Crisis

period period period period period period period period period

Portfolio Partial Partial Partial Partial Partial Partial Partial Partial Partial
f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic

1 21.91* -1.27 69.32* -9.88 40.84* 20.16* 43.73* 24.27* 2.56

2 34.89* 5.64* 83.46* 16.99* -0.38 45.00* 0.86 0.74 -2.26
3 59.14* 17.21* 119.34* 92.84* 53.73* 51.58* 1.26 37.32* 1.22
4 4.67* -0.71 25.16* -3.19 6.05* 24.07* 29.93* 18.52* 0.12
5 8.03* -0.21 51.10* 7.80* 0.02 20.48* -1.26 0.02 -1.37
6 28.87* 4.78* 95.20* 16.85* -0.50 43.11* -3.29 20.11* -2.12
7 6.48* 0.03 21.11* -5.00 13.93* 17.95* 39.16* 22.08* 4.96*
8 12.42* 1.32 80.78* -10.27 2.37 36.12* 3.40 0.46 -1.58
9 73.34* 45.25* 110.37* 11.15* 0.00 102.05* -16.20 16.59* 6.18*
10 39.51* 0.63 71.80* -8.71 21.96* 38.52* 17.58* 8.50* 0.08
11 37.85* 7.25* 74.64* -3.90 5.45* 34.93* 2.31 0.13 -0.48
12 107.88* 39.23* 96.55* 12.21* 0.02 47.64* -3.58 13.58* 0.42
13 21.39* 1.59 35.39* -9.71 6.82* 22.37* 16.78* 1.64 2.30
14 44.05* 9.55* 82.74* 7.34* -0.35 35.52* 4.54* 0.87 -2.03
15 57.86* 14.22* 98.65* 36.39* 10.23* 57.74* -8.10 7.48* -0.94
16 5.55* 2.44 29.10* -1.82 14.17* 15.63* 16.09* -0.36 1.87
17 36.48* 4.02* 109.43* 4.21* 0.32 38.63* 4.81* 1.16 -0.16
18 49.86* 19.90* 103.38* 30.52* 2.30 74.60* -6.02 15.96* 4.00*
19 0.84 -2.49 65.33* -15.51 40.79* 45.68* 45.41* 39.12* -0.74
20 27.10* 6.44 72.09* 35.40* 6.89* 44.07* 3.41 -2.41 -1.16
21 108.17* 29.41* 69.69* 26.46* 2.18 48.00* -7.28 16.48* -0.76
22 -8.32 -5.24 48.33* -10.31 2.29 10.15* 22.20* 6.17* -0.59
23 92.18* 24.24* 79.91* 22.92* 5.23* 39.53* 0.15 0.56 -0.66
24 98.98* 29.60* 45.80* 36.11* 4.99* 72.64* -1.92 0.55 1.66
25 -5.79 13.44* 20.00* 54.37* 108.31* 3.51 53.43* 71.05* 2.78
26 25.65* 2.07 77.86* 19.86* 0.54 43.08* -2.53 1.10 -0.97
27 258.76* 81.47* 134.57* 88.06* 47.13* 92.43* 3.03 53.30* -0.37

For the crisis period (July 2007- April 2009), SMB Finally, we estimated the partial F-statistics to measure
has 20 positive coefficients. HML has 17 (8) positive the marginal significance of LEV in the model. Similar to
(negative) coefficients. For Model 2, SMB has 19 positive the results for global stocks, we find that LEV contributes
and 1 negative coefficients, HML has 17 positive and to improving the overall significance of the model.
8 negative coefficients, and LEV has 9 positive and 5 In all three periods (aggregate, non-crisis and turbulent),
negative coefficients. As noted earlier, there is a marked the partial F-statistics is significant in majority of the
improvement in the regression R2 when LEV is added. cases.

76 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

5. Conclusions data from the emerging and Muslim countries are

included in the Dow Jones Islamic Index.
Fama and French (1993) note that the traditional FF factors
 2. The leverage measure which we are using is the
SMB and HML are good proxies for the underlying distress
market value of debt to market value of assets and not
risk of the firms. As of today, a comparison of how well SMB
book value of debt to market value of assets. Both debt
and HML explain stock returns across good times and bad is
to equity and debt to assets are measures of capital
missing from the literature. In particular, an investigation
structure of a company reflecting the amount of fixed
into whether the economy wide leverage factor replicates
liabilities. The only difference being that debt to equity
the underlying economic fundamentals and contributes to
ratio is more specific to the overall capital used in the
systematic risk especially during bad times is still abstruse;
company while debt to assets ratio is a much broader
and we attempt to unravel this puzzle in the present study.
measure.
Our hypothesis is that, compared to conventional stocks
  3. The authors used weekly data to examine the relative
with high leverage, we would expect SC stocks to have lower
performance of investing in three different types of
sensitivity to risk factors, as well as lower risk premium.
stocks –conventional, Islamic, and SRI stocks. Both
This finding would significantly reaffirm the notion that
in sample (Jan 2000-June 2007) and out of sample
excessive leverage and engaging in economic activities that
(July 2007-April 2009) mean-variance optimization
are not consistent with the principles of Islamic transactions
indicated a portfolio with Islamic stocks generated
can destroy economic and social values, especially during
significantly larger Sharp ratios. The authors claim
falling market environment.
that a low credit market exposure of Islamic stocks
was largely responsible for the relative superior
Using weekly data on stock returns for 3704 firms, we test
performance. The results are robust even when
for the significance of the factors constructed on the basis
financial and real estate companies are removed from
of size, book to market equity, and leverage. We find that
the sample.
the significance of the market factor is drastically reduced
  4. A combination of these two leverage factors produces
during the recent crisis while the explanatory power of the
the book to market ratio. See Fama and French (1992).
Fama-French factors, SMB and HML is reduced considerably.
 5. The crisis had a major impact in September and
In contrast, leverage risk factor performs considerably well
October 2008 when there was a huge withdrawal
across all there periods, especially well during the financial
of $144.5  billion from the money market. Major
crisis, in capturing systemic risk in the economy. Its addition
institutions like Lehman Brothers, Bear Stearns,
to the model is directly correlated with the reduction of
Merrill Lynch, Fannie Mae, Freddie Mac and AIG had
the economic and statistical significance of the traditional
to bear the brunt of high debt market exposure.
Fama-French factors.
 6. Accounting leverage is defined as assets/(assets-
liabilities).
The main result of this paper is that the effects of leverage
 7. Source: http://neutralobserver.blogspot.com/2008/
risk are robust to heterogeneity of the firms in the sample.
11/understanding-financial-crisis-leverage.html
To show that, we perform cross-sectional regressions across
 8. This part of the discussion has been adapted from
three distinct categories of stocks i.e. Conventional, Islamic,
“Leverage 101: The Real Cause of Financial Crisis”,
and SRI stocks. First, as indicated in the earlier section, excess
Sept. 25, 2008, extracted from http://seekingalpha.
market returns play a leading role in explaining the cross
com/article/97299-leverage-101-the-real-cause-of-
section of expected returns prior to the crisis period, but the
the-financial-crisis
effects of the market factor consistently phased out across
  9. Source: http://www.huffingtonpost.com/niall-
all the three categories of stocks during the crisis period.
ferguson/beyond-the-age-of-leverag_b_163872.html
The effects of the leverage factor are consistently significant
10. Source: http://sg.biz.yahoo.com/080625/67/4hbq2.
(except in the case of the socially responsible investing stocks)
html
throughout; however leverage factor gains momentum
11. Source: http://sg.biz.yahoo.com/080625/67/4hbq2.
during the crisis period and has a significant effect on the
html
cross-section of expected returns on stocks and portfolios.
12. Source: http://www.financialweek.com/article/
The sensitivities of stock returns to the Fama-French factors
20080624/REG/705337846/-1/FWDAILYALERT01
are lower after the introduction of the leverage factor.
13. As stated in a research note by James Lee, vice
chairman of J.P.Morgan Chase – extracted from
In a nutshell, the contribution of leverage risk to asset
http://www.financialweek.com/article/20080624/
pricing has been quite strong. The results indicate that
REG/705337846/-1/FWDAILYALERT01
leverage based risk factor can explain a substantial portion
14. As stated by Bob Chapman, “Upsurge of Global
of the cross-section of stock returns across financial and
Leveraged Speculation: The Financial Crisis is
non-financial stocks, as well as, various categories of stocks
not over”, Global Research, November 6, 2009 –
including conventional, Islamic, and SRI stocks. These
extracted from http://www.globalresearch.ca/index.
results have powerful implications for asset management
php?context = va&aid = 15959
using various types of stocks and also during periods of
15. This is consistent with the portfolio formation
great uncertainties.
procedure as suggested in Fama and French (1993).
However, for the purpose of firm specific analysis, we
consider all stocks.
Notes 16. To avoid complications, we restrict the 3-month T-bill
 1. Sharia compliant stocks are household names in return to zero for the months of December 2008 and
mostly developed countries. Surprisingly, only few January, 2009 when intraday return on T-bills was
stocks with enough liquidity and strong balance sheet often negative.

Eds. Hatem A. El-Karanshawy et al. 77

 Bhatt and Sultan

17. Source: http://www.lombardodier.com/annexes/ References

23056/23074/Investment_Strategy_Bulletin_06.10.09.
Abel, AB. 1988. Stock prices under time-varying dividend
pdf, “Is de-leveraging an obstacle to recovery?”
risk: An exact solution in an infinite-horizon general
18. http://www.wikinvest.com/concept/2007_Credit_
equilibrium model. Journal of Monetary Economics.
Crunch
22:375–393.
19. We split the periods to specifically test the impact of
LEV factor during the non-crisis and the crisis period. Abel, AB. 1999. Risk premia and term premia in general
Given the fact the overleveraging leads to increased equilibrium. Journal of Monetary Economics. 43:3–33.
risk exposure in the economy, we believe that this
Arditti, FD. 1967. Risk and the required return on equity.
part of the systematic risk was not captured fully by
Journal  of  Finance, 22:19–36.
the traditional FF factors. This leads us to conjecture
that LEV factor has more direct implications for the Armitrage, S. 2005. The cost of capital: Intermediate
performance of the stocks during the recent credit theory. Cambridge University Press.
crisis and hence we expect the LEV factor to exhibit
Bhandari, LC. 1988. Debt/Equity ratio and expected
stronger effects during this period.
common stock returns: Empirical evidence. Journal of
20. See Table 5.
Finance. 43:507–528.
21. According to the study, managers choose leverage on
the basis of the private information about the future Black, F. 1972. Capital market equilibrium with restricted
growth prospects and hence, the financial health of borrowing. Journal of Business. 45:444–455.
the firm.
Black, F. 1990. Mean reversion and consumption smoothing.
22. Furthermore, a firm with low leverage (having low
Review of Financial Studies. 3:107–114.
Tobin’s q and insignificant growth opportunities)
are harder hit during distress periods as compared to Bris, A, Koskinen, Y. 2002. Corporate leverage and currency
firms with higher leverage ratios (with major growth crises. Journal of Financial Economics.  632, 275–310.
prospects and positive NPV projects). This also
Chan, KC, Chen, N. 1991. Structural and return
explains a negative relationship between credit spread
characteristics of small and large firms. Journal of
and firm’s leverage.
Finance. 46, 1467–1484.
23. Source: “How leverage can increase a company’s
return on equity”, Putnam Spectral Funds, extracted Chou, PH, Ko KC, Lin S. 2010. Do relative leverage and
from: http://www.putnam.com/spectrum/return-on- relative distress really explain size and book-to-market
equity.htm anomalies? Journal of Financial Markets. 13, 77–100.
24. However the risk substantially increases with the
Cochrane, J. 2001. Asset Pricing. University Press. New
excessive use of debt since the firm is under a pressure
Jersey.
to service its debt on a regular basis. In addition,
during economic distress, the assumption that debt is Connor, G, Korajczyk, R. 1989. An intertemporal
available at a lower cost may not hold true. The recent equilibrium beta pricing model. Review of Financial
credit crisis of 2007 presents plentiful evidence where Studies. 2:373–392.
debt became costly. In fact, a firm which undertakes
Daniel, K, Titman S. 1997. Evidence on the characteristics
risky projects may not enjoy the low cost of debt
of cross-sectional variation in stock returns. Journal of
because the riskiness of its operations may require the
Finance. 52, 1–33.
debt holders to be paid a higher interest.
25. It is believed that financial firms exhibit different Davis, JL, Fama, E, French K. 2000. Characteristics,
characteristics as compared to non-financial firms covariances, and average returns: 1929 to 1997. Journal
and hence show different sensitivities to the risk of Finance. 55:389–406.
factors. For instance, high leverage for a financial
Dempsey, M. 2009. The Fama and French three-factor
firm has different implication as compared to a non-
model and leverage: compatibility with Modigliani
financial firm with high debt levels. This further
and Miller propositions. Investment Management and
rationalizes the idea of conducting a robustness
Financial Innovations. 6:48–53.
check by separating out financial firms from the
sample. Dimitrov, V, Jain, P. 2006. The value relevance of changes
26. Firm-specific regressions are not reported to conserve in financial leverage. Working Paper. http://ssrn.com/
space. They are available on request. abstract = 708281.
27. We thank Dow Jones for providing us with the
Dow Jones & Company. (2009, February 27). Dow Jones
proprietary list of stocks classified as conventional,
Indexes. Retrieved August 9, 2009, from Dow Jones &
Islamic and SRI stocks.
Company: http://www.djindexes.com/mdsidx/index.
28 Specifically, the debt ratio (short-term plus long-term
cfm?event = showIslamicStats#fund.
debt as a percent of market capitalization) must not
exceed 33%, interest income should not represent Dow Jones & Company. (2009, February 28). Dow Jones
more than 5% of total revenue, the ratio of accounts Sustainability Index World Components. SAM Indexes.
receivables to total assets does not exceed 45%, and Retrieved September 8, 2009, from Dow Jones &
the ratio of cash and interest bearing securities to Company: http://www.sustainability-index.com/djsi_
market capitalization does not exceed 33%. See Dow protected/djsi_world/components/SAM_DJSIWorld_
Jones website for more. Components.pdf.

78 Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks

Dow Jones & Company. (2009, March 2). Guide to the Dow Kandel, S, Stambaugh, RF. 1990. Expectations and
Jones Islamic Market Indexes. Dow Jones & Company. volatility of consumption and asset returns. Review of
Financial Studies. 3:207–232.
Dow Jones Sustainability World Index. SAM Indexes.
(2009, February 1). Dow Jones Sustainability World Lakonishok, J, Shleifer AR, Vishny. 1994. Contrarian
Index. SAM Indexes. Retrieved September 8, 2009, from investment, extrapolation, and risk. Journal of Finance.
Dow Jones & Company: http://www.sustainability- 49:1541–1578.
index.com/djsi_pdf/publications/Factsheets/SAM_
Lally, M. 2004. The Fama and French model, leverage, and
IndexesMonthly_DJSIWorld.pdf.
the Modigliani-Miller propositions. Journal of Financial
Elton, EJ, Gruber, MJ, Agrawal, D, Mann, C. 2001. Research.  27:341–349.
Explaining the rate spread on corporate bonds. Journal
Lang, L, Ofek, E, Stulz, R. 1996. Leverage, Investment and
of Finance. 247–277.
firm growth. Journal of Financial Economics. 40:3–29.
Eom, KS, Park JH. 2008. Evidence on the three-factor and
Liew, J, Vassalou, M. 2000. Can book-to-market, size
characteristics models: Korea, SSRN: http://ssrn.com/
and momentum be risk factors that predict economic
abstract = 1329664.
growth. Journal of Financial Economics. 57:221–245.
Fama, E, French, K. 1993. Common risk factors in the
Luoma, GA, Spiller, EA, Jr. 2002. Financial accounting
returns on stocks and bonds. Journal of Financial
return on investment and financial leverage. Journal of
Economics. 33:3–56.
Accounting Education. 202:131–138.
Fama, E, French, K. 1992. The cross-section of expected
Milly, M, Sultan, J. 2009. Portfolio diversification during
stock returns. Journal of Finance. 47:427–465.
financial crisis: An analysis of Islamic asset  allocation
Fama, E, French, K. 1995. Size and book-to-market factors strategy. Bentley University Working Paper.
in earnings and returns. Journal of Finance. 50:131–155.
Modigliani, F, Miller, M. 1958. The cost of capital,
Fama, E, French K. 1996. Multifactor explanations of asset corporation finance and the theory of investment.
pricing anomalies. Journal of Finance. 51: 55–84. American Economic Review. 48:261–297.
Fama, EF, French K. 1998. Value versus growth: The Penman, SS, Richardson, Irem, T. 2007. The book-to-price
international evidence. Journal of Finance. 53:1975–1999. effect in stock returns: Accounting for Leverage. Journal
of Accounting Research. 45:427–467.
Fama, EF. 1981. Stock returns, real activity, inflation and
money. American Economic Review. 71, 545–565. Petkova, R. 2006. Do the Fama-French factors proxy for
innovations in predictive variables? Journal of Finance.
Fama, EF, Jensen M. 1998. Separation of ownership and
61:581–612.
control. Michael C, Jensen Foundations of Organizational
Strategy, Harvard University Press. Report on Socially Responsible Investing Trends in the
United States. 10 Year Review. Washington D.C.: Social
Fama, E, Schwert, WG. 1977. Asset returns and inflation.
Investment Forum. 2004.
Journal of Financial Economics. 5:115–146.
Report on Socially Responsible Investing Trends in the United
Ferguson, M, Shockley R. 2003. Equilibrium anomalies.
States. Washington D.C.: Social Investment Forum.
The Journal of Finance. 58(6):2549–2580.
2008.
Gomes, JF, Schmid L. 2009. Levered returns. Forthcoming.
Ron, YWH, Strange, R, Jenifer, P. 2008. Corporate financial
Journal of Finance.
leverage and asset pricing in the Hong Kong market.
Jensen, M, Meckling WH. 1976. Theory of the firm: International Business Review. 17(1):1–7.
Managerial behavior, agency costs and ownership
Sivaprasad, S, Muradoglu, YG. 2009. An Empirical test on
structure. Journal  of Financial  Economics. 3:305–360.
leverage and stock returns. University of Westminster
Jensen, M. 1986. The agency costs of free cash flow: Working Paper.
Corporate finance and takeovers. American Economic
Vassalou, M, Yuhang X. 2004. Default risk in equity returns.
Review. 76:323–329.
Journal of Finance. April: 831–868.
Johnson, T, 2004. Forecast dispersion and the cross
Vassalou, M. 2003. News related to future GDP growth
section of expected returns. Journal of Finance.
as a risk factor in equity returns. Journal of Financial
59:1957–1978.
Economics. 68:47–73.

Eds. Hatem A. El-Karanshawy et al. 79

Is Shariah-compliant investment universally
sustainable? A comparative study
Mehdi Sadeghi
Department of Applied Finance and Actuarial Studies, Macquarie University, Sydney-Australia,
T: (61) 2–98508527, E: [email protected]

Abstract - The current paper reports the outcome of investigating the sustainability and efficiency
of Shariah–compliant investment from the global and cross-country perspectives. Our findings,
thus far, suggest that global Shariah compliant sustainable shares performed slightly better than
global sustainable shares in general, during 2006-2011, and Shariah compliant shares performed
substantially better than global market during the same period. The superior performances of
Islamic market indexes suggest that Shariah compliant investment is more resilient and sustainable
compared to their counterparts in the long term. Further evidence from our cross country study
suggests that, Shariah compliant investments perform better than the market on whole in Muslim
countries, and worse than the market in predominately non Muslim countries. These findings have
important implications for investors, regulators, customers, and Islamic financial institutions.

Keywords: Shariah-compliant sustainable investment, Shariah-compliant investment, index addition

and deletion, event study, abnormal returns, liquidity changes

JEL Classification: G14, G15

1. Introduction equity funds (303), mixed asset (101), money markets

(77), and sukuk funds (77).
The market for Islamic financial services is growing at
an impressive rate, reaffirming its position as one of
With the recent troubles in the global economy, finance
the most dynamic sectors in international finance. The
industry has been looking at Islamic contracts as the
Islamic ­
­ finance industry enjoyed a compound annual
possible means of preventing such meltdowns from ever
growth rate for 2006–2009 of 28%1. The current value of
materializing again. Another area that has received more
Shariah compliant assets managed worldwide, according
in-depth media coverage is the field of sustainability.
to the International Monetary Fund (IMF) estimates, now
Recent changes in the world of investment have made asset
tops USD 1 trillion. The value of these assets is forecasted
owners and managers increasingly aware of the potential
to hit US$1.6 trillion by 2013. This growth represents
risk and value impact of environmental, social, and
a major achievement, as well as new challenges for
governance (ESG) factors, on an investment profile. There
investors, regulators, customers, and also Islamic financial
are arguments in financial literature in favour of both areas
institutions ­themselves.
as safer approaches, and less vulnerable to questionable
financial transactions, which may have led to the global
The biggest share of Islamic financial belongs to the
recession beginning in 2008. These arguments have been
Islamic banks. The S&P2 report indicate that the assets of
substantiated by some empirical findings that suggest some
top 500 Islamic banks in 2008 was $639bn, and grew by
Islamic financial institutions and companies focused on
28.6 percent to $822 bn in 2009. There are also Shariah-
sustainability have been more resilient to financial crisis.
compliant investment funds within Islamic financial system
For instance, Hasan and Dridi (2010), report that Islamic
that cover a wide range of sectors including real estate,
banks have been more resilient than conventional banks
equities, infrastructure, and energy. According to Lipper
during recent global financial crises. This view was also
data for 2010, 586 Islamic funds were in operation with
corroborated by external rating agencies’ reassessment
$37bn of assets under management, with a bias towards
of Islamic banks’ risk, which was generally found to be

Cite this chapter as: Sadeghi M (2015). Is Shariah-compliant investment universally sustainable? A comparative study.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Sadeghi

more favourable than—or similar to—that of conventional Islamic Market Index and Dow Jones Islamic Market
banks (with the exception of UAE) (ibid). Some studies Sustainability Indexes and their constituents to see if there
also suggest that companies with a strong commitment to is any significant difference between the performances of
sustainability have outperformed their industry averages these indexes with Dow Jones Global Stock Market Index.
by 17%3. We also investigate whether there is any significant change
in the efficiency and liquidity of market following Islamic
But are Islamic finance and sustainability finance compa­ index addition and deletion events.
tible? What’s really involved in incorporating sustainability
criteria and Islamic principles into investment decisions? This study is important for several reasons. First, although
Can they make a material difference to investment Shariah-compliant investment is similar to SRI, an area
performance? We start answering these questions by that has already attracted a great deal of research interest,
highlighting similarities and differences between these certain differences is evident in the screening procedures
two. Islamic finance and socially responsible investing that make Shariah-compliant investment different. For
(SRI) approaches have a lot in common with respect to instance, some Islamic funds do not exclude weapons
the screening process, and criteria used for stock selection. manufacturers but they do exclude conventional banks,
Sustainability, on the other hand, goes above and beyond while SRI funds normally exclude weapon manufacturing
SRI by considering positive screens, promoting investment firms and do not exclude banks. As another difference,
in companies with best practices. According to World concerns about environmental issues are not as important
Economic Forum Report (2011) “Sustainable investing in screening Shariah-compliant companies as they are for
is an investment approach that integrates long-term SRI funds. Furthermore, Shariah-compliant companies are
environmental, social and governance (ESG) criteria subject to certain financial ratio tests that are not relevant
into investment and ownership decision-making with the to conventional SRI companies6.
objective of generating superior risk-adjusted financial
returns”4. As the financial crisis receded into a period Second, Miller-Modigliani capital structure theory
of uncertainty in the past two years, recognition that contemplates that in an imperfect capital market with
sustainability, corporate governance and transparency are corporate taxes, companies can increase their assets’ value
important factors in portfolio management has emerged. by increasing their leverage. Given that Shariah-compliant
This is a fundamental shift away from the ideological companies are constrained by their level of borrowing, it
and political corner of SRI to the real performance of would be interesting to investigate how this constraint can
sustainability. affect their value.

Some researches assert that Islamic finance holistic and Third, finance theory based on the efficient market
dynamic perception of SRI is more effective in taking into hypothesis (EMH) considers shares with identical risk and
consideration the reality and ever-changing circumstances return as perfect substitutes for each other. This makes
of societies in contrast to Western humanistic theories. market demand for securities elastic and horizontal. Since
They conclude that corporations operation on a piety-based Shariah-compliant equities are not a perfect substitute
business paradigm acknowledge their social responsibility for the conventional equities, their demand may not be
to their workers, managers, other corporations, customers, horizontal. This can bring about a different outcome to the
and society as a whole more significantly (Dusuki and study of a Shariah-compliant index revision.
Abdullah, 2007). However, regardless of their similarities,
and theoretical arguments in support of one or another, Fourth, Islamic screening criteria reduce the number of
sustainability and Shariah-compliant investments are available shares to invest. It is claimed by critics that the
assessed on the basis of long-term trends in yield, profitability, reduction of the investment universe through screening will
and efficiency in use of limited financial resources. reduce the performance. Similar counterarguments have
been raised regarding sustainability criteria (Freidman,
In January 2006, Dow Jones Indexes launched the world’s 1996). It would be interesting to investigate how this
first Dow Jones Islamic Market Sustainability Index. This constraint can affect Shriah-compliant portfolios.
index merges Islamic investing principles and sustaina­
bility criteria by combining the methodology of Dow Jones Finally, academic research on the performance of Shariah-
Islamic Market Indexes5 and Dow Jones Sustainability compliant investments is rare, and to the best of our
Indexes. To be included in the index, companies must knowledge, no similar study on the impacts of the Shariah-
be components of both the Dow Jones Islamic Market compliant index revisions has been conducted before.
Index and the Dow Jones Sustainability World Index.
Linking Shariah compliant investment performance to Our results, thus far suggest that global Shariah compliant
sustainability is, perhaps, the most effective way to highlight sustainable shares perform worse than global Shariah
the importance of ESG governing factors to Islamic finance. compliant shares in the long term. However, they both
The time series data provided by Dow Jones Indexes is an perform better than global stock market as a whole.
invaluable resource to help us investigate whether Islamic Further evidence from individual countries suggests that,
finance is a sustainable practice in the long term. Shariah compliant investments perform better than the
market in Muslim countries, and worse than the market in
Current paper is a progress reports on our ongoing long term predominately non Muslim world. The rest of this study is
research objective of testing the efficiency and sustainability organized as follows: Section II is allocated to a short review
of Shariah compliant investment opportunities around of research background. We outline our methodology,
the world. We have used time series data on Dow Jones data and hypothesis development in Section III. Empirical

82 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study

findings are discussed in Section IV. Section V articulates in Shariah compliant companies in non-Muslim countries
our conclusions, and describes the limits of our study. around the world11. In the case of equities, the differences
between Shariah compliant shares and their conventional
forms are even less significant, only requires screenings.
2.  Research background and literature This screening process is similar to the screening of Socially
review Responsible Investing (SRI) instruments.
Research background In the third stage of our research we decided to investigate
We started our study with an investigation of the market Shariah-compliant index addition and deletion to
performance and liquidity of Shariah-compliant Index predominantly non-Muslim countries, starting with
(SI) portfolio following its introduction by Bursa Malaysia. Australia as the first sample. Australia’s skills in complex
Malaysia has one of the largest Islamic fund markets in the financial engineering and experience in infrastructure,
world. It had 155 unit trusts and mutual funds at the end resources, property and agriculture provide her with
of June 2010 with a total volume of about RM22.69 billion. a unique opportunity to develop Shariah-compliant
Our findings show that, overall, introduction of SI had investments. This country also has easy access to rapidly
a positive impact on the financial performance and the growing Islamic financial markets with over a billion in
liquidity of included shares in this country7. population to accommodate their demand12.

As time series data on Shariah compliant indexes become A through presentation of our findings on all eight countries
more readily available for other parts of Muslim world studied so far is too long to report here. In order to show the
through index providers, such as Dow Jones Islamic contrasting nature of market reaction to Index addition and
Market Index8, we decided to extend our study to the deletion events in predominately Muslim and non-Muslim
MENA (Middle East and North Africa) market in the second countries, we report the report the results on two sample
stage of our study. MENA region is another important countries of Egypt and Australia in section IV.
hub in Islamic finance, with large market and appropriate
financial infrastructure. Constrained by the availability of
times series data, we used event study methodology and
the improved models of liquidity measures, first to index Literature review
addition to equity markets in Qatar, Kuwait, Oman, and From a theoretical perspective, there are two explanations
UAE. Our findings showed an even stronger result than for for the effects of stock additions to an index: demand-based
Malaysia in that market reacts positively to the introduction and information-based. The demand-based explanation
of Shariah compliant shares in these countries. This was sees index changes as information-free events. For example,
reflected in short and long-term market performance Shleifer (1986), by employing the downward-sloping
and the improvement in the liquidity of shares9. One of demand curve hypothesis, showed that the price effects
the limitations of recent study was the small number of following index changes are due to the demand from index
companies in our sample. To test the robustness of findings tracking. These effects can be temporary or permanent.
with larger samples, we extended our investigation to The temporary effect is explained by the price pressure
Jordan and Egypt. Our results overwhelmingly supported hypothesis, predicting a reversal of initial price increases
the robustness of our earlier findings of countries in the in the long run (Harris and Gurel, 1986). The permanent
Gulf region10. effect is explained by the imperfect-substitute hypothesis,
which assumes that there would be no price reversal, as the
Overall, our research on seven markets in Islamic countries new price reflects changes in the distribution of security
in showed that investors’ reaction to the introduction of holdings in equilibrium13.
Shariah-compliant shares is positive. This is reflected in
improvement in the share price and market liquidity up to Information-based explanations include the information
150 days following the index addition. The positive outcome hypothesis and the liquidity hypothesis. Unlike the demand-
for six countries in MENA region is especially important based explanations, information-based explanations as­
because they were found from the data that became sume that index changes are not information-free events.
available by Dow Jones Indexes immediately following the Some studies, such those by Dhillon and Johnson (1991)
start of financial crisis, suggesting that Shariah compliant and Jain (1987), support the information hypothesis: they
investments in Islamic countries has been more resilient to showed that the addition of a stock to the index conveys
financial crisis than conventional investments. favorable news about the firm’s prospects and a permanent
price increase can result following this event. Amihud
In addition to Muslim countries, Islamic finance is practiced and Mendelson (1986), Beneish and Whaley (1996), and
outside the Muslim world without ties to any particular Hegde and McDermott (2003) contended that the price
jurisdiction. Shariah compliant investments are defined reactions can be explained by changes in market liquidity.
according to certain norms and conditions that can be According to the liquidity hypothesis, the price increase
applied anywhere in the world where there is a market and at index inclusion is caused by the increased liquidity due
people who wish to engage in financing transactions in a to the greater visibility of the shares, greater interest from
manner which is consistent with Shariah law. This progress institutional investors, higher trading volume, and lower
is specially facilitated by a form of reverse financial bid-ask spreads. Amihud and Mendelson (1986) suggested
engineering that reconstructs conventional financial that the increase in stock liquidity is positively related to
products into Shariah compliant instruments. This the firm’s value through a reduction in the cost of capital.
innovation has significantly increased Muslims investments Previous studies, such as Harris and Gurel (1986), and

Eds. Hatem A. El-Karanshawy et al. 83

 Sadeghi

Hegde and McDermott (2003) reported liquidity increases during the event windows can be interpreted as a measure
following index additions. of the effect of the event on the value of the firms, which is
reflected in their share price.
The topic of Shariah-compliant index revision is important
from two perspectives. First, the nature of companies’ Our event window extended from 10 days before to 25 days
activities and their capital structure makes them Shariah after the event. This asymmetric event window was chosen
compatible in the first place. Second, changes in investors’ to examine the extended effect of excess returns in the
demand result in subsequent market price reactions, post-event period15.
according to our earlier discussion. For example, reduction
in the level of debt in the capital structure can make a The normal returns of stocks are the expected returns
company Shariah-compliant, bringing about an increase if there are no events. The normal returns are estimated
in the demand from Muslims and higher share prices if over a period of time outside the event window (Peterson,
demand is not fully elastic. At the same time, the lower 1989). For applications in which the determinants of the
level of debt may move the capital structure of the company normal return are expected to change due to the event,
to a suboptimal level, at a higher cost of capital than in the estimation period can fall on both sides of the event
equilibrium. This may send negative signals to the market window. This period commences 125 trading days before
when shares are added to a Shariah-compliant index. and ends 125 trading days after the event dates, excluding
As a result, it is possible that the interaction of opposing the event period of day −10 to Day 25. As a result, the
market forces on index revision will bring about different estimation period consists of Day −135 to Day −11 and Day
outcomes compared with the effects of conventional index 26 to Day 150. We did not allow the event period to overlap
additions. Therefore, it is not possible to predict clearly with the estimation period, to avoid biasing the parameter
how the performance and liquidity of shares included in estimates in the direction of the event effect.
or excluded from the DJIM index will change, as it largely
depends on how the net effects of the influential factors are The following section describes the event study
revealed through our empirical investigation. methodology that we used in our study. MacKinlay (1997),
and Kothari and Warner (2004) have provided a survey
of event study methods, and we follow their papers to
3.  Data and methodology describe the models here.
To determine the impact of additions to and deletions from
the DJIM index, we applied several measures of both short
and long-term price and liquidity performance. We applied Liquidity effect
standard event study methodology to find the initial stock Market liquidity is an elusive concept and difficult to
price reaction of firms when an announcement of an measure. In this study, we use six proxies to evaluate changes
index change was made. We also applied several liquidity in market liquidity during post-event periods, compared to
measures to investigate the magnitude and direction of the corresponding control periods. The large number of
liquidity changes following the index revision. Data for this tests helps to confirm the robustness of our findings and
research has been collected through Dow Jones Indexes reduces the chance of making wrong inferences.
and Bloomberg.
These liquidity proxies include: 1) quoted spread, as the
simple difference between bid and ask prices; 2) percentage
Price effect spread, as the quoted spread normalized by the midpoint of
Our event-study methodology calculates the abnormal the bid and ask prices; 3) changes in the volume of trade as
returns. An abnormal return is the difference between the daily average of the transaction size, normalized on the
the realized return observed from the market and the average volume of trade in the control period; 4) changes
benchmark return. The return to the market portfolio is in volatility, measured by the standard deviation of returns;
estimated via both ordinary least square (OLS) and Scholes 5) the Amivest liquidity ratio, as the average ratio of share
and William (1977) procedures. The latter method is volume to absolute return over all days with non zero
usually used when stocks do not trade at the same level of returns; and 6) the proportion of zero daily returns. Zero
frequency as the market index and OLS may produce biased daily return is related to trading speed because the days
beta estimates. This problem is exacerbated for infrequently with zero return indicate delays or difficulties in executing
or thinly traded stocks as the sampling interval is reduced14. an order, interrupting the continuity of trading.
The advantages of these models are that they control for the
effect of market movements through the market portfolio, In calculating the percentage bid-ask spread and change
and also allow for an individual security’s responsiveness in the volume of trade, we largely follow Hegde and
as measured by beta. Return on the All Ordinaries index McDermott (2003). Changes in the volume of trade are
was used as a proxy for the market rate of return. directly related, and changes in the bid-ask spreads and
volatility are inversely related to the market liquidity.
We defined the event date as the day that a stock was added It is important to note that an increase in the volume
to or deleted from the DJIM index. For each event, the return accompanied by an increase in volatility can actually
time series data were divided into an estimation period impede market liquidity. The Amivest liquidity ratio is
and an event window. The estimation time series data are estimated according to Amihud and Mendelson (2002).
used to calculate the benchmark parameters, and the event This ratio measures the ability of a share to absorb changes
window period is used for computing prediction errors in trading volume without any significant change in
based on the estimated parameters. Abnormal returns are share price. Change in this variable is directly related to
represented by the prediction errors. The abnormal returns the liquidity. In estimating the proportion of zero daily

84 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study

40
DJIMSI
45
DJIMI
Jones Islamic Market Sustainability Index out performs Dow
DJGSI DJWSI Jones Global Sustainability Index by less than 1% during
35 40 this period. However, Dow Jones Islamic Market stock Index
35 shows a much higher return of 22% compare to Dow Jones
30 World stock Market Index. The superior performances of
30 Islamic Market indexes suggest that Shariah compliant
25 investment is more resilient and sustainable compare to
25 their counterparts within the family of Dow Jones Indexes
20
20 in the long term.
15
15
Cross country findings
10
10 A through presentation of our findings for all eight
5 5
countries studied so far is too long to report here. The
Results reported here is only from Egypt and Australia, in
0 0 order to show the contrasting nature of market reaction to
Index addition and deletion in two predominately Muslim
1/1/06
9/1/06
5/1/07
1/1/08
9/1/08
5/1/09
1/1/10
9/1/10
5/1/11

1/1/06
11/1/06
9/1/07
7/1/08
5/1/09
3/1/10
1/1/11
and non Muslim country.

Figure 1.  The Performance of Dow Jones Islamic Market Price effect

Sustainability Index (DJIMSI), Dow Jones Islamic Market Table 1 presents the estimated CARs for index additions in
Index (DJIMI), Dow Jones Sustainability Index (DJGSI), the pre- and post-event periods for Egypt. The coefficient
and Dow Jones World Stock Index (DJWSI) during for CARs, accumulated during the period (−10, 0), is
1/1/2006 to 1/5/2011. -1.71%. However, it is not statistically significant at the
conventional levels. The CARs coefficient estimated over
the shorter period (-5, 0) increases to 2.41% and becomes
statistically significant at the 0.05 level. CARs coefficients
returns, we follow Bekaret et al. (2004), as they found it for Day 0 (the event day) and for (0, 5) increase further
a reasonable proxy for a liquidity measure to use in their to 2.85% and 3.44%, respectively, and become highly
study. Change in this variable is inversely related to the significant at the 0.01 level. CARs for (0, 15) drop to 2.69%
market liquidity. and remain statistically significant at the 0.01 level. CAR
coefficients increase further to 6.31% during (0, 30) and
remain statistically significant at the 0.01 level.
4. Results
The prolonged effects of the index additions on CARs
Global findings in Table  1  indicate that these events are likely to contain
Figure  1 presents the Performance of Dow Jones Islamic information, thus sending signals about the features of the
Market Sustainability Index (DJIMSI), Dow Jones Islamic index additions to the market. To test this hypothesis, we
Market Index (DJIMI), Dow Jones Global Sustainability compared the cumulative returns (CRs) for the added firms
Index (DJGSI), and Dow Jones World Stock Index (DJWSI) with the cumulative return for the market over the period
during 1/1/2006 to 1/5/2011. Our findings show that Dow (-10, 150)16.

Table1. Cumulative abnormal returns and relevant statistics for stock additions to the DJIM
index in Egypt.
This table presents the cumulative abnormal returns (CARs) around the index addition for the 25 Egyptian firms
in our sample. Results are presented for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, 15), and (0, 30), where day 0
represents the addition date. The Generalized Sign Z-test is a test with the null hypothesis that the fraction of posi-
tive cumulative returns is the same as in the estimation period. The Positive/Negative column reflects how many
firms had positive cumulative abnormal returns in the window. The symbols $, *, **, and *** denote statistical
significance at the 10%, 5%, 1% and 0.1% levels, respectively, using a 1-tail test. The symbols), >, etc., correspond to
$,* and show the significance and direction of the Generalized Sign-Z test.

Scholes-Williams Market Model

Generalized Positive/
Intervals MCARs t-Statistics Sign Z-test Negative

(-10, 0) 1.71% 1.12 1.15  15/10

(-5, 0) 2.41% 1.75* 1.55$  16/9)
(0, 0) 2.85% 6.36*** 4.35***  23/2 >>>
(0, +5) 3.44% 2.63** 2.74**  19/6 >>
(0, +15) 2.69% 1.72* 2.35**  18/7 >>
(0, +25) 6.31% 2.69** 1.95*  17/8 >

Eds. Hatem A. El-Karanshawy et al. 85

 Sadeghi

4
3.5 Market Index
Shares
3
2.5
2
1.5
1
0.5
0
–0.5 1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154

Figure 2.  Cumulative firm return and market return around day −10 to Day 150
egyptian stocks addition to DJIM index.

40
35 Market Index
30 Shares

25
20
15
10
5
0
–5 1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154

Figure 3.  Risk adjusted cumulative firm return and market return around day −10
to day 150 Egyptian stocks addition to DJIM index

Figure  2 illustrates CRs for the portfolio of added stocks, accumulated during Day −10 to Day 0, is −1.22%, and is
compared with the market CRs during (−10, 150) for Egypt, statistically significant at the 0.05 level. When the CARs
showing the shares’ superior performance of 352% gain, coefficient is estimated over the shorter interval of Day
compared with less than 48% for the market by Day 150. −5 to Day 0, it increases slightly to −1.18% and remains
Figure 3 compares the performance of the same variables statistically significant at the 0.05 level. CARs for Day 0
on a risk-adjusted basis, calculated using the Sharpe Ratio. (the event day) and Day 0 to Day 5 increase to −0.32% and
According to this figure, the Sharpe Ratio for the shares 0.21%, respectively. However, they are not significantly
shows a value of 34 compared with a ratio of 1.4 for the different from zero at the conventional levels.
market.
CARs for the intervals Day 0 to Day 10 and Day 0 to Day
Table  2 and Table  3 present mean cumulative abnormal 25 decline continuously, dropping to −4.04% and become
returns (CARs) for the added and the deleted Australian highly significant at the 0.01 level. CARs for the entire
firms, respectively. To test the robustness of our findings, window (Day −10 to Day 25) is −4.94% and significant
we have used both the single-factor and Scholes-Williams at the 0.01 level. The temporary upward trend in CARs
market models as the benchmarks for estimating normal around the event day may have been caused by the positive
return. Our results show that the magnitudes of CARs and reactions of some Muslim investors to the additions news.
the level of their statistical significance from the application However, this reaction was perhaps not strong enough to
of the two methods are similar. Nevertheless, we report fully offset a negative response from the market as a whole.
and discuss the results from the Scholes-Williams model The coefficients for generalised sign tests are consistent
to avoid non-synchronous trading bias, as a considerable with the coefficients for t-statistics, although they are not
proportion of shares included in this study are likely to as strongly significant as the later ones. It is mainly the
trade less frequently. coefficient for Day 0 to Day 25 and the entire event window
(Day −10 to Day 25) that are statistically significant at the
Table  2 presents the estimated CARs for index additions conventional level, indicating that the significance of our
in the pre- and post-event periods. The coefficient for CARs findings is robust to both parametric and non-parametric

86 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study

Table 2.  Mean cumulative abnormal return and relevant statistics for stock additions to the DJIM index in Australia.
This table presents the mean cumulative abnormal returns (CARs) around the index addition for the 117 firms in our sample. Results are presented
for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, +10), (0, 15), (0, 25), and (-10, 25), where day 0 represents the addition date. The third column
is the precision-weighted cumulative mean abnormal return. The generalized sign Z is a test of the null hypothesis that the fraction of positive
cumulative returns is the same as in the estimation period. The symbols $, * and ** denote statistical significance at the 10%, 5%, and 1% levels,
respectively, using a 1-tail test.

Scholes-Williams Market Model

Cumulative Precision-
average abnormal weighted t- Generalized
Intervals return (CAAR) CAAR statistics sign Z-test

(-10, 0) -1.22% -1.06% -1.70* -0.65

(-5, 0) -1.18% -0.78% -2.24* -1.21
(0, 0) -0.32% -0.15% -1.16 -0.47
(0, +10) -1.11% -0.36% -1.24 -1.21
(0, +15) -1.62% -0.65% -1.55$ -1.39
(0, +25) -4.04% -2.03% -2.71** -1.58$
(-10, +25) -4.94% -2.94% -3.02** -1.76*

Table 3.  Mean cumulative abnormal return and relevant statistics for Australian stock deletions from the DJIM index.
This table presents the mean cumulative abnormal returns (CARs) around the index deletion for the 87 firms in our sample. Results
are presented for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, +10), (0, 15), (0, 25), and (-10, 25), where day 0 represents the
addition date. The third column is the precision-weighted cumulative mean abnormal return. The Generalized Sign Z is a test of the
null hypothesis that the fraction of positive cumulative returns is the same as in the estimation period. The symbols $, * and ** denote
statistical significance at the 10%, 5%, and 1% levels, respectively, using a 1-tail test.

Scholes-Williams market model

Cumulative Precision-
average abnormal weighted t- Generalized
Intervals return (CAAR) CAAR statistics sign Z-test

(-10, 0) 1.57% 1.36% 1.29$ 0.11

(-5, 0) 2.11% 1.62% 2.51** 2.60*
(0, 0) 0.47% 0.41% 1.35$ 1.19
(0, +5) 0.04% -0.04% -0.05 -0.31
(0, +10) 5.34% 2.24% 2.28* 1.40$
(0, +15) 6.05% 2.68% 2.33** 1.83*
(0, +25) 7.45% 3.82% 2.82** 2.05*
(-10, +25) 8.55% 4.77% 3.02** 2.06*

tests. Our findings are also consistent with the results of CARs increases further to 6.05% during the interval Day
Clarke and Russell’s (2008) study on Socially Responsible 0 to day 15, and to 7.45% during the interval Day 0 to Day
Investing (SRI): they found significant negative CARs for 25, respectively. Both coefficients remain highly significant
DS400 additions that persisted at least 30  days after the at the 0.01 level. CARs for the entire window (Day −10
events. to Day 25) is 8.55% and significant at the 0.01 level. The
temporary downward trend in CARs after the event day
Table 3 presents the estimated CARs for index deletion in may have been caused by the negative reactions of Muslim
the pre- and post-event periods for Australian shares. The investors to the deletion news. However, this reaction did
coefficient for CARs, accumulated during Day −10 to Day 0, not seem to be strong enough to fully offset the positive
is −1.57% and is marginally significant at the 0.10 level. The response from the market as a whole.
CARs coefficient estimated over the shorter interval of Day
−5 to Day 0, increases to 2.11% and becomes statistically Results in Table 2 and Table 3 show CARs of up to 25 days
significant at the 0.05 level. This coefficient for Day 0 (the after additions and deletion, respectively. Some studies in
event day) is 0.47% and statistically significant at the 0.10 the literature, such as one by Nesbitt (1994), suggest that
level. CARs for Day 0 to Day 5 is negative and statistically the value of socially responsible investing may be more
insignificant at the conventional levels. This coefficient apparent in the long-run. To examine whether DJIM Index
quickly rises to 5.34% during the interval Day 0 to Day additions and deletions have any prolonged information
10, and becomes statistically significant at the 0.05 level. effects on shares, we compared the cumulative returns

Eds. Hatem A. El-Karanshawy et al. 87

 Sadeghi

0
–0.02 1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154

–0.04
–0.06
–0.08
–0.1
–0.12 Market Index
–0.14 Shares

–0.16

Figure 4.  Cumulative return on a portfolio of added Australian shares, compared with cumulative return on the market
for the 160-day period from Day −10 to Day 150 around the event day.

0
–10 1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154

–20
–30
–40
–50
–60
–70 Market Index
–80 Shares

–90

Figure 5.  Cumulative risk adjusted return on a portfolio of added Australian shares, compared with risk-adjusted
cumulative return on the market for the 160-day period from Day −10 to Day 150 around the event day. Risk-adjusted
returns are estimated according to the Sharpe performance index.

Table 4. Measures of liquidity changes from pre- to post-stock additions to the DJIM index in Egypt.
This table presents the change of a variety of liquidity measures around the index addition day for an equally weighted portfolio of
25 Egyptian firms in our sample. Results are presented for the windows (1, 25), (1, 50), (1, 100), and (1, 150), compared with the
control periods (-35, -10), (-60, -10), (-110, -10), and (-160, -10), respectively. The bid-ask mean difference represents the difference
between average liquidity measures in each interval compared with the corresponding interval in the control period. The symbols $, *, **,
and *** denote statistical significance at the 10%, 5%, 1%, and 0.1% levels, respectively, using a 1-tail test.

Intervals (1, 25) vs. (1, 50) vs. (1, 100) vs. (1, 150) vs.
Liquidity measures (-35, -10) (-60, -10) (-110, -10) (-160, -10)

Standard Deviation (SD) 1.05% 0.85% 0.89% 0.85%

SD (control period) 1.49% 2.56% 2.05% 1.80%
SD change –0.44% –1.71%*** –1.16%*** –0.95%***
Relative bid-ask spread 1.53% 1.36% 1.35% 1.36%
Relative bid-ask spread 0.80% 0.71% 0.55% 0.45%
(control period)
Bid-ask mean difference 0.73% 0.65% 0.80% 0.91%
Average daily volume 40.24 53.97 130.73 202.37
Average daily volume 25 50 100 150
(control period)
Average daily volume change 60.95%* 7.94% 30.73%** 34.91%***
Amivest liquidity measure 13.85 13.84 13.68 13.69
Amivest liquidity measure 13.29 13.32 13.81 13.88
(Control period)
Amivest liquidity measure 0.56*** 0.52*** –0.13$ –0.19**
change

88 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study

Table 5. Measures of liquidity changes from pre- to post- Australian stock additions to the DJIM index.
This table presents the change of a variety of liquidity measures around the index addition for an equally weighted portfolio of 117 firms in
our sample. Results are presented for the intervals in days (1–25), (1–50), (1–100), and (1–150), compared with the control periods (-35
to 10), (-60 to -10), (-110 to -10), and (-160 to -10), respectively. Day 0 represents the addition date. The mean difference represents
the difference between average liquidity measures in each interval compared with the corresponding interval in the control period. The
symbols $, *, **, and *** denote statistical significance at the 10%, 5%, and 1%, and 0.1% levels, respectively, using a 1-tail test.

Intervals Liquidity Day 1 to 25 Day 1 to 50 Day 1 to 100 Day 1 to 150

measures (-35 to -10) (-60 to -10) (-110 to -10) (-160 to -10)

Absolute bid-ask spread 6.02¢ 6.17¢ 6.24¢ 6.48¢

Absolute bid ask (control period) 5.61¢ 5.63¢ 6.14¢ 5.84¢
Absolute bid-ask mean difference 0.41¢* 0.54¢** 0.10¢ 0.64¢***
Relative bid-ask spread 0.43% 0.45% 0.49% 0.51%
Relative bid-ask (control period) 0.47% 0.46% 0.47% 0.46%
Relative bid-ask mean difference –0.04% –0.01% 0.02%$ 0.05%**
Average volatility (SD) 3.60% 3.51% 3.61% 3.66%
Average volatility (SD) 3.45% 3.15% 2.87% 2.91%
(control period)
Average volatility ratio  0.15% 0.36%** 0.74%*** 0.75%***
Average daily volume 27.10 54.15 106.35 157.56
Average daily volume 25.00 50.00 100.00 150.00
(control period)
Average volume difference 8.40% 8.30%$ 6.35%* 5.04%*
Zero-return 12.12% 11.33% 12.79% 13.15%
Zero-return (control period) 8% 7.84% 12.77% 12.64%
Zero-return mean difference 4.12%*** 3.49*** 0.02% 0.51%
Amivest liquidity measure 12.44 12.38 12.40 12.33
Amivest liquidity measure 12.53 12.50 12.46 12.44
(Control period)
Amivest liquidity measure –0.09*** –0.12*** –0.06*** –0.11***
mean difference

(CRs) for the added and deleted firms with cumulative 5.  Concluding remarks
return for the market over the period from Day −10 to Current paper reports the outcome investigating
Day 15017. the sustainability and efficiency of Shariah –compliant
investment from the global and cross-country perspectives.
Figure  4 and Figure  5 provide long-term evidence of Our findings, thus far, suggest that Dow Jones Islamic
negative market reaction to the index addition. Figure  4 Market Sustainability Index out performs Dow Jones Global
illustrates CRs for the portfolio of added stocks, compared Sustainability Index by less than 1% during 1/1/2006–
with the market CRs during Day −10 to Day 150, showing 1/5/2011. However, Dow Jones Islamic Market stock Index
the market’s superior performance of −7.8% compared shows a much higher return of 22% compare to Dow Jones
with −13.1% for the shares by Day 150. Figure 5 compares World Stock Market Index during the same period. The
the performance of the same variables on a risk-adjusted superior performances of Islamic Market indexes suggest
basis, calculated according to the Sharpe performance that, relative to their counterparts within the family of
index (SPI). According to this figure, SPI for the market Dow Jones Indexes, Shariah compliant investments are
shows a figure of −42.9% compared with the SPI of −70.8% generally more resilient and sustainable in the long term.
for the shares.
In the cross country component of our study, we used data
Table  4 provides evidence of changes in liquidity from eight countries (only one is reported in this paper)
measures for Egypt. The results show a decline in the and an event study methodology to estimate cumulative
standard deviation of returns between 0.95% and 1.71%, abnormal returns in the days surrounding index additions
accompanied by an increase in the volume of trade from and deletions for testing the price effects of market reaction.
30.73% to 60.95%. Amivest liquidity measure changes also We also used several liquidity measures; including the bid–
suggest an increase in the market liquidity over the short to ask spread, the Amivest liquidity ratio, standard deviation
medium term and a decline over the medium to long term. of returns, and volume of trade to estimate changes in
The coefficients for changes in the bid-ask spread is positive; the liquidity of the added shares around these events. Our
however, they are not statistically significant. Overall, there results show that stock prices respond positively to index
is more evidence for improvement in the liquidity of the additions for Muslim countries and negatively for non
Egyptian stock market than for decline. Muslim countries, both in the short and long terms. Further

Eds. Hatem A. El-Karanshawy et al. 89

 Sadeghi

evidence from non Muslim countries suggests that stock 13. Refer to Beneish and Whaley (1996), Lynch and
market react positively to index deletions. Mendenhall (1997), Kaul et  al. (2000), and Wurgler
and Zhuravskaya (2002) for more details.
Observing negative abnormal return for index additions, 14. The frequency of trading declines with the reduction
and positive abnormal return for index deletions in in the sampling interval.
Australia suggests that market in this country perceives 15. This allowed for slow responses from overseas Muslims
these events as a value destroying, and value adding that might cause delays in the market reaction to the
exercises, respectively. This view is in line with Friedman index revision.
(1996) agency theory, perceiving any effort by companies 16. We believe that if index inclusion contains information,
to go beyond maximising their profit as a burden on their this information must have been reflected in share
return. These opposing reactions can also be explained prices earlier than the event day and should extend
by differences in both fundamental and socio-cultural for some time afterwards. As a result, we have used
factors in Muslim vs. non Muslim countries. For instance, a a sample of data that extends from 10 days before to
company in the West world can become Shariah compliant 150 days after the event.
by chance, or by force, not necessarily by choice. A low 17. We believe that if index inclusion and exclusion contain
debt/equity ratio in the capital structure of companies information, this information must have been reflected
in Western countries can make them Shariah-compliant. in share prices earlier than the event day and should
However, this may occur, perhaps, due to their inability to extend for some time afterwards. As a result, we have
borrow money if they are relatively small. While a company used a sample of data that extends from 10 days before
in a Muslim world may intentionally borrow less to comply to 150 days after the event.
with Shariah- principles.

These findings have important implications for the References

development and growth of Islamic finance around the Amihud Y, Mendelson M. 1986. Asset pricing and the bid-
world. For example, if Western countries plan to promote ask spread. Journal of Financial Economics. 17:223–49.
themselves as a centre for Islamic finance, they need
Amihud Y. 2002. Illiquidity and stock returns: Cross-section
to overcome certain impediments to be successful. This
and time-series effects. Journal of Financial Markets.
includes reduction in psychological barriers, as well as
5:31–56.
revision in taxation laws and non-taxation regularities to
ensure that they do not inhibit the development of Islamic Asyraf Wajdi Dusuki AW, Abdullah NI. 2007. Maqasid
finance. There is also need for a trained work force in al-Shari`ah, Maslahah, and Corporate Social
financial sector (education in Islamic economics, finance, Responsibility. The American Journal of Islamic Social
banking, insurance, accountancy, and business law), and Sciences. 24(1):25–45.
ability to market Islamic financial products overseas once
Becchetti LL, Ciciretti I, Hasan. 2009. Corporate social
they are developed.
responsibility and shareholder’s value: An empirical
analysis. Bank of Finland Research Discussion Papers. 1.
Notes Beneish M, Whaley R. 1996. An anatomy of the “S&P 500
1. HSBC Report, Islamic Banking and Finance Summit, Game: The effects of changing the rules. Journal of
Reuters’ Offices, Dubai, 2009. Finance. 51:1909–30.
2. S&P Press Release, 1st February 2010.
Boehmer E, Poulsen A. 1991. Event-study methodology
3. Daniel Mahler, A.T. Kearney, Inc. Report, titled Green
under conditions of event-induced variance. Journal of
Winners: The Performance of Sustainability-focused
Financial Economics. 30:253–72.
Companies in the Financial Crisis, 2009.
4. Transition Towards Sustainable Investing, World Chen HG, Noronha VS. 2004. The price response to
Economic Forum White paper, 2011. S&P 500 Index additions and deletions: Evidence of
5. DJIM Indices were introduced in 1999 as the asymmetry and a new explanation. Journal of Finance.
benchmarks to represent Shariah-compliant 59:1901–29.
portfolios.
Clarke S, Russell W. 2008. Stock price effects associated
6. Socially responsible fixed-income securities are found
with socially responsible index constituent changes.
in conventional financial markets, while, at least in
Paper presented at Midwest Finance Association
theory, they are banned by Shariah.
Conference. San Antonio, Texas, USA.
7. Refer to Sadeghi (2008) for more details.
8. Companies from Islamic countries were added to Dow Dhillon U, Johnson H. 1991. Changes in the Standard and
Jones Islamic market Index in 2009 for the first time. Poors 500 list. Journal of Business. 64:75–85.
9. Refer to Sadeghi (2010a) for more details.
Friedman M. 1996. The social responsibility of business
10. Refer to Sadeghi (2010b) for more details.
is to increase profits. In: Schwartz MS (Ed.). Beyond
11. This is a pragmatic compromise, rather than an ideal
Integrity: A Judeo-Christian Approach. Zondervan
situation from an Islamic perspective.
Publishing House. Grand Rapids.
12. The DJIM index constituents are screened from
around the globe and are mostly located in non- Harris L, Gurel E. 1986. Price and volume effects associated
Muslim countries. For instance, from 2403 DJIM index with changes in the S&P 500: New evidence for
constituents on 30th November 2009, 2204 originated the existence of price pressures. Journal of Finance.
in the non-Muslim world, especially in the West. 41:815–30.

90 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study

Hasan M, Dridi J. 2010. Islamic Banking: Put to the Test. Price Waterhouse Coopers International Limited Report.
Finance & Development. 47(4):45–47. 2009. Shariah-compliant funds: A whole new world of
investment.
Hegde S, McDermott J. 2003. The liquidity effects of
revisions to the S&P 500 Index: An empirical analysis. Sadeghi M. 2011a. Investment opportunities and stock
Journal of Financial Markets. 6:413–59. liquidity: evidence from DJIM index additions in the
Gulf States. Investment Management and Financial
HSBC Report. 2009. Islamic Banking and Finance Summit,
Innovations. 8(1):50–59.
Reuters’ Offices, Dubai.
Sadeghi M. 2011b. Shariah-compliant Investment and
Jain P. 1987. The effect on stock price of inclusion in or
Shareholders’ Value: An Empirical Investigation. Global
exclusion from the S&P 500. Financial Analysts Journal.
Economy and Finance Journal. 4(1):44–61.
43:58–65.
Sadeghi M. 2012. Are Faithful Investors Rewarded by
Karpoff J. 1987. The relation between price changes and
the Market Place? Evidence from Australian Shariah-
trading volume: A survey. Journal of Financial and
compliant Equities. Qualitative Research in Financial
Quantitative Analysis. 22:109–26.
Markets, forthcoming.
Kothari S, Warner J. 2004. Econometrics of event studies,
Sadeghi M. 2008. Financial Performance of Shariah-
Working Paper. (Tuck School of Business at Dartmouth,
Compliant Investment: Evidence from Malaysian Stock
Hanover, USA).
Market. International Research Journal of Finance and
Merton R. 1987. A simple model of capital market Economics. 20:15–26.
equilibrium with incomplete information. Journal of
Finance. 42:483–510.
Nesbitt S. 1994. Long-term rewards from shareholder
activism: a study of the CalPERS effect. Journal of
Applied Corporate Finance. 6:75–80.

Eds. Hatem A. El-Karanshawy et al. 91

The nexus between economic freedom and Islamic
bank performance: Empirical evidence from the
MENA banking sectors
Fadzlan Sufian1, Muhamed Zulkhibri, Abdul Majid2
1
Universiti Islam Antarabangsa Malaysia.
2
Economic Research and Policy Dept., Islamic Development Bank. Mailing address: 9th Floor, Economic Research and
Policy Department, Islamic Development Bank, P.O Box 5925, Jeddah, 21432, Saudi Arabia, e-mails:mzulkhibri@isdb.
org; [email protected], Tel: +966-2-646-6533; Fax: +966-2-506047132
All findings, interpretations, and conclusions are solely of the authors’ opinion and do not necessarily represent the
views of the institutions.

Abstract - The present study provides new empirical evidence on the impact of economic freedom
on Islamic banks’ performance. The empirical analysis focuses on Islamic banks operating in the
MENA banking sectors during the period 2000–2008. We find that the larger, more diversified,
and better capitalized Islamic banks tend to be relatively more profitable, while credit risk and
expense preference behaviour seem to exert negative impact. The findings suggest that greater
financial freedom positively influence the profitability of Islamic banks operating in the MENA
banking sectors. Interestingly, the impact of monetary freedom is negative implying that higher
(lower) monetary policy independence reduces (increases) Islamic banks’ profitability, providing
support to the benefits of government interventions.

Keywords: economic freedom, Islamic banks, profitability, panel regression analysis, MENA

JEL Classification: G21; G28

1. Introduction The Islamic banks operate in markets characterized by

competition-inhibiting government regulation and in a
Islamic banking is a relatively recent addition to the global
protected banking environment. Islamic banking, being a
financial markets. Its conventional brick and mortar root can
participatory type of banking system, has entered on the
be traced back to the early 1960s when Myt Ghamar Bank
global banking market in full force. In recent years, market
was formed in Egypt in 1963. Between 1963 and 1971 the
conditions in Islamic banking have undergone extensive
bank provided Muslims with a place to deposit their savings
changes from both the demand and supply sides. On the
in accordance to the Syari’a principles1. Despite its humble
demand side, customers have become more sophisticated,
beginning, Islamic banks have blossomed throughout the
value-oriented, and price sensitive, while on the supply
world and are looked upon as a viable alternative system
side, the globalization of financial markets has been
which has many things to offer.
accompanied by governmental deregulation, financial
innovation, and automation.
Although it was initially developed to fulfill the needs
of Muslims, Islamic banking has now gained universal
These two factors have resulted in an increase in the number
acceptance. According to El-Qorchi (2005), the number
of competitors, cost reductions, and profit declines. The
of Islamic financial institutions increased from a single
revolution in information technology, mainly in internet
institution in 1975 to approximately 486 financial
banking has enabled the larger financial institutions to
institutions operating in more than 75 countries
penetrate markets and to increase their market share
worldwide2. Total assets of Islamic financial institutions are
within both national and overseas markets by providing
estimated at US$250 billion and are tipped to be growing
competitive products at lower prices. Furthermore,
at 15% per year, three times the rate of conventional banks.
Islamic equity-type financial instruments are competing
The rapid growth rate confirms the growing importance of
with conventional banking products and now face strong
Islamic banking and finance in the global financial markets.

Cite this chapter as: Sufian F, Zulkhibri M, Majid A (2015). The nexus between economic freedom and Islamic bank
performance: empirical evidence from the MENA banking sectors. In H A El-Karanshawy et al. (Eds.), Islamic banking
and finance – Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar
Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Sufian et al.

competition from both banks and non-bank financial Stroup, 2000; Heckelman, 2000; De Haan and Siermann,
institutions. This also accentuates competition within the 1998), these studies have mainly examined the impact
financial services industry. of economic freedom on economic growth. On the other
hand, virtually nothing has been published to examine the
It is reasonable to assume that these developments posed impact of economic freedom on the performance of the
great challenges to Islamic banks as the environment in which conventional or Islamic banking sectors. This limitation
they operates in has changed rapidly. This could sensibly is somewhat surprising given the importance of bank
have an impact on the determinants of their performance. lending in promoting economic growth and development
Despite considerable development of the Islamic banking (e.g. Ben Naceur and Ghazouani, 2007; Beck and Levine,
sector, empirical works on Islamic banks’ performance is 2004; Rajan and Zingales, 1998) and given the impact that
still in its infancy. The knowledge of the underlying factors economic freedom is likely to have on the banking sector.
which influences the Islamic banking sector’s performance
is essential given the growing importance of Islamic banking The paper is divided into five sections. The following section
and finance in the global financial markets. It is therefore presents the literature review. Section 3 describes the data,
essential not only for the managers of the Islamic banks, sources, and empirical settings. In section 4 we present the
but for numerous stakeholders such as the central banks, results and finally, section 5 concludes.
bankers associations, governments, and other financial
authorities to help them identify and formulate policies
to improve the performance of the Islamic banking sector, 2.  Review of the literature
particularly in the MENA region3. The empirical evidence on the performance of the
conventional banking sectors is extensive. To date, the
On the perspective of economic freedom, economic theories numerous studies have mainly focused on the U.S. banking
suggest that economic freedom tend to affect incentives, sector (e.g. DeYoung and Rice, 2004; Stiroh and Rumble,
productive effort, and the effectiveness of resource use.4 2006; Hirtle and Stiroh, 2007; Tregenna, 2009) and the
Economists and economic historians have argued that banking sectors of the western and developed countries
since the time of Adam Smith, central ingredients for (e.g. Williams, 2003; Pasiouras and Kosmidou, 2007;
economic progress are the freedom to choose and supply Kosmidou et al. 2007; Hawtrey and Liang, 2008; Kosmidou,
resources, competition in business, trade with others, and 2008; Kosmidou and Zopounidis, 2008; Athanasoglou et al.
secure property rights (North and Thomas, 1973). Within 2008; Albertazzi and Gambacorta, 2008; Kasman et al.
the context of the MENA region, it can be observed from 2010). On the other hand, empirical works on the Islamic
Table 1 that the region has achieved modest improvement banking sector is still in its infancy. Typically, studies on
in economic freedom during the year 20105. It can be seen Islamic bank performance have focused on theoretical
from Table 1 that Bahrain retained the top ranking within issues and the empirical works have relied mainly on
the region and managed to be ranked in the world Top 10, the analysis of descriptive statistics rather than rigorous
while Qatar ranks in the world top 30. statistical estimation (El-Gamal and Inanoglu, 2005).

The ongoing transformations of innovative and reform- Hussein (2003) provides an analysis of the cost efficiency
oriented states such as Bahrain, Qatar, Kuwait, and Oman features of Islamic banks in Sudan. By using the stochastic
may pave the way for a more robust and dynamic regional cost frontier approach, he estimates cost efficiency for
economic growth in the region. On different scale, Jordan a sample of 17 banks over the period 1990 and 2000.
and Oman registered the highest gains in economic The results show large variations in the cost efficiency
freedom and Qatar’s improvement to 70.5, moved it of Sudanese banks with the foreign owned banks being
from the category of “moderately free” to “mostly free”, the most efficient, while the state owned banks being the
while Syria’s improvement lifted its designation from most cost inefficient. The empirical findings suggest that
“repressed” economy to “mostly unfree”. However, no the small banks are relatively more efficient compared
other MENA countries are rated as having “mostly free” to their large bank counterparts. In addition, banks with
economies. Nearly half of the region falls into the “mostly a higher proportion of musharakah and mudharabah
unfree” category and two countries, namely Libya and finance relative to total assets tend to exhibit efficiency
Iran, ranked among the world’s most repressed economies. advantages.
The institutional problems, such as lack of investment and
financial freedom and weak systems for protecting property In another study on the Sudanese Islamic banking sector,
rights and preventing corruptions continue to degrade the Hassan and Hussein (2003) examine the efficiency of the
region’s overall economic freedom and economic potential. Sudanese banking system during the period of 1992 and
2000. They employed a variety of parametric (cost and
The purpose of the present paper is to extend the earlier profit efficiencies) and non-parametric Data Envelopment
works on the performance of the Islamic banking sector in Analysis (DEA) methods to a panel of 17 Sudanese banks.
the MENA region and to establish empirical evidence on the They found that the average cost and profit efficiencies
impact of economic freedom. The paper also investigates to under the parametric method were 55% and 50%
what extent the performance of Islamic banks is influenced respectively, while it was 23% under the non-parametric
by internal factors (i.e. bank-specific characteristics) and method. During the period under study, they suggest that
to what extent by external factors (i.e. macroeconomic the Sudanese banking system has exhibited 37% allocative
conditions and economic freedom). Although studies on efficiency and 60% technical efficiency, suggesting that the
economic freedom is vast in the literature (e.g. Heckelman overall cost inefficiency of the Sudanese Islamic banks were
and Knack, 2009; Altman, 2008; Powell, 2003; Adkins mainly due to technical (managerially related) rather than
et al. 2002; De Haan and Sturm, 2000; Heckelman and allocative (regulatory).

94 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 1. Economic freedom index for MENA.

2011 Change Freedom

World Regional Overall from Business Trade Fiscal Government Monetary Investment Financial Property from Labour
Country Rank Ranking Score 2010 Freedom Freedom Freedom Spending Freedom Freedom Freedom Rights Corruption Freedom

Eds. Hatem A. El-Karanshawy et al.

Bahrain  10  1 77.7 1.4 77.4 82.8 99.8 80.2 74.0 75.0 80.0 60.0 51.0 97.0
Qatar  27  2 70.5 1.5 70.3 82.4 99.8 78.1 71.9 45.0 50.0 70.0 70.0 67.0
Oman  34  3 69.8 2.1 69.4 83.6 98.5 68.1 69.5 55.0 60.0 50.0 55.0 89.1
Jordan  38  4 68.9 2.8 65.8 78.8 92.7 60.9 81.4 70.0 60.0 55.0 50.0 74.2
Israel  43  5 68.5 0.8 66.1 87.8 62.3 44.8 78.4 80.0 70.0 70.0 61.0 64.3
UAE  47  6 67.8 0.5 67.3 82.6 99.9 79.1 76.5 35.0 50.0 50.0 65.0 72.4
Saudi  54  7 66.2 2.0 86.1 82.2 99.4 74.6 64.3 40.0 50.0 45.0 43.0 77.0
 Arabia
Kuwait  61  8 64.9 -2.8 64.4 81.6 99.9 69.7 69.3 55.0 50.0 50.0 41.0 67.9
Lebanon  89  9 60.1 0.6 57.5 80.5 91.0 64.9 77.7 60.0 60.0 25.0 25.0 59.0
Morocco  93 10 59.6 0.4 75.7 75.8 67.8 74.6 76.5 65.0 60.0 40.0 33.0 27.2
The nexus between economic freedom and Islamic bank performance

Egypt  96 11 59.1 0.1 64.5 74.0 89.6 65.3 60.8 65.0 50.0 40.0 28.0 53.6
Tunisia 100 12 58.5 -0.5 80.2 53.5 73.7 77.6 77.3 35.0 30.0 50.0 42.0 65.7
Yemen 127 13 54.2 -0.2 73.7 81.6 83.2 44.5 82.3 45.0 30.0 30.0 21.0 50.9
Algeria 132 14 52.4 -4.5 69.4 72.8 83.5 62.4 75.4 20.0 30.0 30.0 28.0 52.9
Syria 140 15 51.3 1.9 55.9 65.4 84.6 85.3 69.7 20.0 20.0 30.0 26.0 55.8
Iran 171 16 42.1 -1.3 69.4 44.8 81.1 76.0 60.7 0.0 10.0 10.0 18.0 50.7
Libya 173 17 38.6 -1.6 20.0 85.0 80.3 44.5 71.0 10.0 20.0 10.0 25.0 20.0

Note: Each one of the 10 freedom is graded using a 0 to 100 scale, where 100 represents the maximum freedom. A score of 100 signifies an economic environment or
set of policies that is most conducive to economic freedom. Many of the 10 freedoms are based on quantitative data that are converted directly into a score.
Source: The Heritage Foundation.

95
 Sufian et al.

El-Gamal and Inanoglu (2004) employ the stochastic 3.  Data and methodology
frontier approach to estimate the cost efficiency of Turkish
The present study employs an unbalanced annual bank level
banks over the period 1990–2000. The study compared the
data of all Islamic banks operating in the MENA countries
cost efficiencies of 49 conventional banks with four Islamic
covering the period 2000–2008. The financial statements
special finance houses (SFHs). The Islamic firms comprised
of Islamic banks operating in the MENA banking sectors
around 3% of the Turkish banking market. Overall, they
are collected from the Bankscope database of Bureau
suggest that the Islamic financial institutions to be the
van Dijk’s company. The macroeconomic variables are
most efficient. This could be explained by their emphasis
retrieved from the IMF Financial Statistics (IFS) and the
on Islamic asset-based financing which led to low non-
World Bank World Development Indicator (WDI) databases
performing loans ratios.
while economic freedom variables are extracted from The
Heritage Foundation.
The study by Hassan (2006) is among the few performed
to examine the efficiency of Islamic banks in a cross-country
setting. He employs both the parametric (Stochastic Measure of performance
Frontier Approach) and non-parametric (Data Envelopment
Analysis) methods to examine the efficiency of banks in the Following Ben Naceur and Goaied (2008), Kosmidou
sample. The findings indicate that during the period 1993– (2008), and Abbasoglu et al. (2007) among others, the
2001, Islamic banks have exhibited a relatively higher dependent variable used in this study is Return on Assets
profit efficiency compared to cost efficiency. He suggests (ROA). ROA shows the profit earned per dollar of assets
that the main source of inefficiency is allocative rather than and most importantly, reflects management ability to
technical. The results indicate that the overall inefficiency utilize banks financial and real investment resources
was output related. The results indicate that on average to generate profits (Hassan and Bashir, 2003). For any
the Islamic banking industry is relatively less efficient bank, ROA depends on the bank’s policy decisions as well
compared to their conventional counterparts. as other uncontrollable factors relating to the economy
and government regulations. Rivard and Thomas (1997)
While the above outlines the literature that employs advanced suggest that bank profitability is best measured by ROA,
modelling techniques to evaluate Islamic banks’ performance, since it is not distorted by high equity multipliers and
one should also note that there is a growing body of literature represents a better measure of the ability of firms to
that covers the general performance features of Islamic banks. generate returns on its portfolio of assets.
Such studies include those by Hassan and Bashir (2003)
who look at the determinants of Islamic banks’ performance
and show that Islamic banks to be just as efficient as their Internal determinants
conventional bank peers if one uses standard accounting The bank specific variables included in the regression
measures such as the cost-to-income ratio. Other studies models are LLP/TL (loans loss provisions divided by total
that followed similar approach are those by Sarker (1999) loans), EQASS (book value of stockholders’ equity as a
who examines the performance and operational efficiency of fraction of total assets), NIE/TA (total overhead expenses
Bangladeshi Islamic banks, while Bashir (1999) investigates divided by total assets), LOANS/TA (total loans divided by
the risk and profitability of two Sudanese banks. total assets), and LNTA (log of total assets).

Bashir (1999) and Bashir (2001) performed regression The ratio of loan loss provisions to total loans (LLP/TL) is
analyses to examine the underlying determinants of Islamic incorporated as an independent variable in the regression
banks’ performance. By employing bank level data from analysis as a proxy of credit risk. The coefficient of the LLP/
the Middle East, the results indicate that the performance TL variable is expected to enter the regression models with
of banks, in terms of profits, is mostly generated from a negative sign. In this vein, Miller and Noulas (1997)
overhead, customer short-term funding, and non-interest point out that the greater the exposure of banks to high
earning assets. Furthermore, Bashir (2001) claimed that risk loans, the higher would be the accumulation of unpaid
since deposits in Islamic banks are treated as shares, reserves loans and profitability would be lower. Miller and Noulas
held by banks propagate negative impacts such as reducing (1997) suggest that decline in loan loss provisions are in
the amount of funds available for investment. In essence, many instances the primary catalyst for increases in profit
the findings from this literature are that Islamic banks are margins. Furthermore, Thakor (1987) also suggests that
at least as efficient as their conventional bank counterparts the level of loan loss provisions is an indication of the bank’s
and in most cases are relatively more efficient. asset quality and signals changes in future performance.

The above literature reveals the following research gaps. The EQASS variable is included in the regression models
First, the majority of these studies have concentrated on to examine the relationship between profitability and bank
the conventional banking sectors and the banking sectors capitalization. Strong capital structure is essential for banks in
of the western and developed countries. Second, empirical developing economies, since it provides additional strength to
evidence on the developing countries banking sectors, withstand financial crises and increased safety for depositors
particularly the Islamic banking sectors are relatively scarce. during unstable macroeconomic conditions. Furthermore,
Finally, virtually nothing has been published to examine lower capital ratios in banking imply higher leverage and risk
the impact of economic freedom on the Islamic banking and therefore greater borrowing costs. Thus, the profitability
sector. In light of these knowledge gaps, the present paper level should be higher for the better capitalized bank.
provides new empirical evidence on the impact of economic
freedom on the performance of Islamic banks operating in The ratio of non-interest expenses over total assets, NIE/
the MENA countries banking sectors. TA, is used to provide information on the variations of bank

96 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance

operating costs. The variable represents total amount of of labour and indirect effects i.e. changes in interest rates
wages and salaries, as well as the costs of running branch and asset prices on the profitability of banks. Perry (1992)
office facilities. The relationship between the NIE/TA suggests that the effects of inflation on bank performance
variable and profitability levels is expected to be negative, depend on whether the inflation is anticipated or
because the more productive and efficient banks should be unanticipated. In the anticipated case, the profit rates
able to keep their operating costs low. Furthermore, the are adjusted accordingly resulting in revenues to increase
usage of new electronic technology, like ATMs and other faster than costs subsequently positive impact on bank
automated means of delivering services, may have caused profitability. On the other hand, in the unanticipated
expenses on wages to fall (as capital is substituted for case, banks may be slow to adjust their interest rates
labor). resulting in a faster increase of bank costs compared to
bank revenues and consequently negative effects on bank
An important decision that the managers of Islamic profitability6.
banks must take refers to the liquidity management and
specifically to the measurement of their needs related to the To examine the impact of concentration on Islamic banks’
process of deposits and loans. For that reason, the ratio of performance, the CR3 variable is introduced in the
total loans to total assets (LOANS/TA) is used as a measure regression models. The CR3 ratio is calculated as the total
of liquidity. Higher figures denote lower liquidity. Without assets held by the three largest banks in the country. The
the required liquidity and funding to meet obligations, variable is used to examine the impact of asset concentration
a bank may fail. Thus, in order to avoid insolvency in the national banking sector on the profitability of Islamic
problems, banks often hold liquid assets, which can be banks. The Structure-Conduct-Performance (SCP) theory
easily converted to cash. However, liquid assets are usually posits that banks in a highly concentrated market tend to
associated with lower rates of return. It would therefore collude and therefore earn monopoly profits (Molyneux et
reasonable to expect higher liquidity to be associated with al. 1996). Berger (1995) points out that the relationship
lower bank profitability. between bank concentration and performance in the U.S.
depends critically on what other factors are held constant.
The LNTA variable is included in the regression models as According to the industrial organization literature, a
a proxy of size to capture for the possible cost advantages positive impact is expected under both collusion and
associated with size (economies of scale). In the literature, efficiency views (Goddard et al. 2001).
mixed relationships are found between size and profitability,
while in some cases a U-shaped relationship is observed. The Z-Score (Z-SCORE) variable is used as a proxy of
LNTA is also used to control for cost differences related bank soundness. The index measures how many standard
to bank size and for the greater ability of the large bank deviations a bank is away from exhausting its capital base
to diversify. In essence, LNTA may lead to positive effects (a distance-to-default measure). The Z-Score is a popular
on bank profitability if there are significant economies of measure of soundness because it combines banks’ buffers
scale. On the other hand, if increased diversification leads (capital and profits) with the risks they face in a way that
to higher risks, the variable may exhibit negative effects. is grounded in theory (Cihak et al. 2009). A higher Z-Score
implies a lower probability of insolvency, providing a more
direct measure of soundness than, for example, simple
External determinants leverage measures (Cihak et al. 2009). This index combines
If analysis is done in a static setting, they may fail to capture in a single indicator: (i) profitability, given by a period
developments in the regulatory environment and in the average return on assets (ROA); leverage measure, given by
marketplace, which may have changed the underlying the period average equity-to-asset ratio (K) (equity here is
production technology and the associated production defined as total equity from the balance sheet of a bank); and
functions. Furthermore, different banking forms could return volatility, given by the period standard deviation of
demonstrate different reactions to environmental changes. ROA (Vol. (ROA)) i.e. Z = VolROA
.( ROA ) where ROA (profitability)
+K

Hence, the change in the financial landscape and structure, is a period average of ROA, K (leverage measure) is the
etc., may vary across banking groups (Saunders et al. period average equity-to-asset ratio, and Vol. (ROA) is the
1990; Button and Weyman-Jones, 1992; Berger, 1995). To return volatility given by the period standard deviation of
measure the relationship between economic and market ROA. A higher (lower) Z-SCORE indicates lower (higher)
conditions and Islamic banks’ performance, LNGDP, INFL, risk (De Nicolo et al. 2003).
CR3, and Z-SCORE variables are used.

Gross domestic product (GDP) is among the most Economic freedom measurements
commonly used macroeconomic indicator to measure total In simple terms, economic freedom is a conceptual
economic activity within an economy. The GDP is expected measure of the private ownership and market allocation
to influence numerous factors relating to the supply and of resources, in lieu of government ownership and
demand for loans and deposits. Favourable economic control. Expressing the sentiment of many, including the
conditions will affect positively on the demand and supply originators of the economic freedom index, Berggren
of banking services, but will have either positive or negative (2003) defines economic freedom as “the degree to which
influence on bank profitability levels. an economy is a market economy—that is, the degree to
which it entails the possibility of entering into voluntary
Another important macroeconomic condition which may contracts within the framework of a stable and predictable
affect both the costs and revenues of banks is the inflation rule of law that upholds contracts and protects private
rate (INFL). Staikouras and Wood (2003) points out that property, with a limited degree of interventionism in the
inflation may have direct effects i.e. increase in the price form of government ownership, regulations, and taxes”.

Eds. Hatem A. El-Karanshawy et al. 97

 Sufian et al.

In regression model 2, OVER_FREE is introduced to examine et  al. (2009) suggest that potential endogeneity could be
the impact of overall economic freedom on the performance a problem when assessing bank profitability determinants.
of the Islamic banks operating in the MENA banking For instance, the more profitable banks may have sufficient
sectors. OVER_FREE is the overall economic freedom resources to provision for their non-performing loans. The
index and is defined by multiple rights and liberties. The more profitable banks may also find it easier to increase
index uses 10 specific freedoms, namely Business freedom, their customer base via successful advertising campaigns
Trade freedom, Fiscal freedom, Government size, Monetary and could hire the most skilled personnel, and therefore
freedom, Investment freedom, Financial freedom, Property enhances their profitability levels (Garcia-Herrero et al.
rights, Labor freedom, and Freedom from corruption. 2009).

Besides the overall economic freedom index, we have Arellano and Bond (1991) proposed an efficient
selected three other indices which are closely related to Generalized Methods of Moment (GMM) estimator that
the financial sector. These include BUSI_FREE, MONE_ uses instruments of which the validity is based on the
FREE, and FINA_FREE indices. BUSI_FREE is the business orthogonality between the lagged values of the dependent
freedom index. The index measures how free entrepreneurs variable and the errors. The technique eliminates the
are to start businesses, how easy it is to obtain licenses, and unobserved bank heterogeneity by estimating the equation
the ease of closing a business. Impediments to any of these in first-differences and to control for possible endogeneity
three activities are deterrents to businesses and therefore problem by using the model’s variables lagged by one
to job creations. MONE_FREE is the monetary freedom or more periods as instruments. We employ the GMM
index. The index combines a measure of price stability with estimator as proposed by Arellano and Bond (1991) to
an assessment of price controls. Both inflation and price ensure efficiency and consistency of the estimations.
control distorts market activity. Price stability without Therefore, a dynamic GMM model is adopted via the
microeconomic intervention is an ideal state of a free market. inclusion of a lagged dependent variable among the
FINA_FREE is the financial freedom index. The index is a regressors to capture the persistence of bank profitability
measure of banking security as well as independence from over time reflecting impediments to market competition,
government’s control. State ownership of banks and other informational opacity, and/or sensitivity to regional/
financial institutions such as insurer and capital markets is macroeconomic shocks (Berger et al. 2000).
an inefficient burden and political favoritism has no place
in a free capital market. All these indices have 0 to 100 The baseline model is formulated as follows:
scales, where 100 represents maximum freedom. A score
of 100 signifies an economic environment, or set of policies
π it = a + λπ i ,t-1 + ∑ b X it + ∑γ M t + ∑δ E t + µ t + υ i + e it (1)
that is most conducive to economic freedom.

Finally, CORR_FREE is introduced in regression model where i  =  1, 2, …, N (number of firms) and t   =  1, 2, …,

6 to assess the impact corruption on the performance of T (time period). In the specification, pit denotes the
Islamic banks. CORR_FREE is the freedom from corruption profitability of bank i at time t; pt−1 indicates a one period
index. The index is based on quantitative data that assess lagged profitability; Xi,t is vector exogenous bank-specific
the perception of corruption in the business environment, regressors: Mt is a vector of country specific variables; Ei is
including levels of governmental, legal, judicial, and a vector of country specific economic freedom variables; mt
administrative corruption. Similar to the BUSI_FREE, is a time fixed effect; vi is an unobserved banks’ fixed effect;
MONE_FREE, and FINA_FREE indices, the CORR_FREE ei,t is a serially uncorrelated error term.
index also takes a value of between 0 and 100, where
100 represent the maximum freedom. Table  1 contains We use several tests proposed by Arellano and Bond
the summary statistics of the variables used to proxy (1991) to check whether the instruments are properly
profitability and its determinants. chosen and the assumptions underlying the model holds.
Our estimations rely on the fact that the disturbances
follow an MA(1) process and there is no second order
Econometric specification autocorrelation (m2) together with Sargan/Hansen tests of
Since the panel data cover many heterogenous banks over-identifying restriction (J-test) to examine the validity
and time periods, the possible correlation between the of the instruments used in the regression models.
regressors and bank-specific effects, the endogeneity of
regressors with respect to idiosyncratic shock and the Extending Eq. (1) to reflect the variables as described in
heteroscedasticity of the disturbance term (idioscyncratic Table 2, the baseline model is formulated as follows:
shock) would result in a biased and inconsistent estimation
with Ordinary Least Square (OLS) estimation technique. ROA jt = b 0  + b 1 LOANS / TA jt + b 2 LNTA jt
The OLS estimator would result in an upward estimate of + b 3 LLP / TL jt + b 4 NIE / TA jt   + b 5 EQASS jt
the coefficient while the within-group estimator would
+ b 6 LNGDPt + b 7 INFLt + b 8 CR3t + b 9 Z - SCORE t
be downward biased (Blundell et al. 1992). A natural
+ b 9 OVER _ FREE t + b 10 BUSI _ FREE t
technique for dealing with variable that are correlated with
the error term is to instrument them. + b 11 MONE _ FREE t + b 12 FINA _ FREE t
+ b 13CORR _ FREE t + e jt (2)
Berger et al. (2000) points out that bank profitability
tend to persist over time reflecting impediments to market Table 3 provides information on the degree of correlation
competition, informational opacity, and sensitivity to between the explanatory variables used in the panel
macroeconomic shocks. Furthermore, Garcia-Herrero regression analysis. In general, the matrix shows that the

98 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 2. Descriptive of the variables used in the regression models.

Variable Description Mean Std. Dev. Sources/Database

ROA A proxy measure of bank profitability measured as the return on average  2.577  3.798 BankScope
total assets of the bank in year t.
Independent
Internal Factors
LLP/TL Loan loss provisions/ total loans. An indicator of credit risk, which shows  8.321 14.13 BankScope
how much a bank is provisioning in year t relative to its total loans.

Eds. Hatem A. El-Karanshawy et al.

EQASS A measure of bank’s capital strength in year t, calculated as equity/total 21.227 23.458 BankScope
assets. High capital asset ratio is assumed to be indicator of low leverage
and therefore lower risk.
NIE/TA Calculated as non-interest expense/ total assets and provides information  3.755  3.247 BankScope
on the efficiency of the management regarding expenses relative to the
assets in year t. Higher ratios imply a less efficient management.
LOANS/TA A measure of liquidity, calculated as total loans/ total assets. The ratio 48.583 23.706 BankScope
indicates what percentage of the assets of the bank is tied up in loans in
year t.
LNTA The natural logarithm of the accounting value of the total assets of the  8.117  2.666 BankScope
bank in year t.
The nexus between economic freedom and Islamic bank performance

External Factors

LNGDP Natural logarithm of gross domestic products.  6.057  4.237 IMF

International
Financial Statistics
INFL The rate of inflation.  2.246  2.154 IMF
International
Financial Statistics
CR3 The three largest banks asset concentration ratio.  0.730  0.151 IMF
International
Financial Statistics
Z-SCORE The Z-Score index. Is used as a proxy measure of the banking sector’s risk to 11.027  6.832 IMF
default. International
Financial Statistics
Economic Freedom
OVER_FREE Overall economic freedom is defined by multiple rights and liberties can 60.464 12.575 Heritage Foundation
be quantified as an index of less abstract components. The index uses 10 (www.heritage.org/index)
specific freedoms, some as composites of even further detailed and
quantifiable components.

99
(Continued)
 Sufian et al.

correlation between the bank specific variables are not

(www.heritage.org/index)
(www.heritage.org/index)

(www.heritage.org/index)

(www.heritage.org/index)
strong, implying that multicollinearity problems are not
severe. Kennedy (2008) points out that multicollinearity

Heritage Foundation
Heritage Foundation

Heritage Foundation

Heritage Foundation
is a problem when the correlation is above 0.80 which
is not the case here. However, it is worth noting that
the LNGDP variable is highly correlated to most of the
economic freedom variables. To address this concern, we
have also estimated all regression models by excluding
the macroeconomic variables. Furthermore, due to the
high correlation between the economic freedom variables,
the regression models are estimated by including the
each economic freedom indicator at a time, rather than
estimating all economic freedom variables concurrently.

4.  Empirical findings

26.329

24.140
14.763

12.318

The regression results focusing on the relationship

between bank profitability and the explanatory variables
are presented in Table 4. The reliability of our econometric
methodology depends critically on the validity of the
instruments, which can be evaluated with Sargan’s test
of overidentifying restrictions, asymptotically distributed
64.890

75.829

47.354

46.134

as c2 in the number of restrictions. A rejection of the null

hypothesis that instruments are orthogonal to the errors
would indicate that the estimates are not consistent (Baum
et al. 2010)7. We also present test statistics for the first and
second order serial correlations in the error process. In a
Monetary freedom combines a measure of price stability with an assessment

dynamic panel data context, second order serial correlation

Financial freedom is a measure of banking security as well as independence
Price stability without microeconomic intervention is the ideal state for the
Business freedom measures how free entrepreneurs are to start businesses,

should not be present if the instruments are appropriately

of price controls. Both inflation and price controls distort market activity.

perception of corruption in the business environment, including levels of

institutions such as insurer and capital markets is an inefficient burden,
from government control. State ownership of banks and other financial

uncorrelated with the errors (Baum et al. 2010).

Freedom from corruption is based on quantitative data that assess the
Impediments to any of these three activities are deterrents to business
how easy it is to obtain licenses, and the ease of closing a business.

It can be observed from Table 4 that for all the estimated

models, the Sargan statistics for overidentifying restrictions
and the Arrelano–Bond AR(2) tests shows that our
and political favoritism has no place in a free capital market.

governmental legal, judicial, and administrative corruption.

instruments are appropriately orthogonal to the error and

no second order serial correlation is detected respectively.
Furthermore, the highly significant of the lagged dependent
variable’s coefficient confirms the dynamic character of
the model specification, thus justifying the use of dynamic
panel data model estimation.

Concerning the liquidity results, the empirical findings

suggest a negative sign of the coefficient of the LOANS/
TA in the baseline regression model. As higher (lower)
figures of the ratio denote lower (higher) liquidity levels,
the results imply that the relatively less (more) liquid banks
and therefore to job creation.

tend to exhibit higher (lower) profitability levels. On the

other hand, the empirical findings also suggest that the
coefficient of the variable is positive when we control for
overall economic freedom, business freedom, financial
freedom, and freedom from corruption. However, the results
should be interpreted with caution since the coefficient of
free market.

the variable is not significant at any conventional levels in

any of the regression models estimated.

It can be observed from Table 4 that the coefficient of the

LNTA variable entered the baseline regression model with a
Table 2. (Continued)

negative sign and is statistically significant when we control

for economic and financial market conditions lending
support to Spathis et al. (2002), Dogan and Fausten (2003),
MONE_FREE

CORR_FREE

and Kosmidou (2008). Moreover, the earlier studies have

FINA_FREE
BUSI_FREE

concluded that marginal cost savings could be achieved by

increasing the size of the banking firm, especially as markets
develop (Berger et al. 1987; Boyd and Runkle, 1993; Miller
and Noulas, 1997; Athanasoglou et al. 2008). In this vein,

100 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 3. Correlation matrix for the explanatory variables.

The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total loans divided by total assets; LNTA is
a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure of bank risk calculated as the ratio of total loan loss provisions divided
by total loans; NII/TA is a measure of bank diversification towards non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy

Eds. Hatem A. El-Karanshawy et al.

measure for costs, calculated as non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the rate of inflation; OVER_FREE is the overall economic freedom index; BUSI_FREE is
the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial freedom index; CORR_FREE is the freedom from corruption index.

LLP/ LOANS/ OVER_ BUSI_ MONE_ FINA_ CORR_

TL EQASS NIE/TA TA LNTA LNGDP INFL CR3 Z-SCORE FREE FREE FREE FREE FREE

LLP/TL 1.000 0.493** 0.048 -0.369** -0.293** -0.169* -0.237** -0.097 -0.052 0.044 0.095 0.160* 0.031 0.059
EQASS 1.000 0.407** -0.026 -0.555** -0.479** -0.432** 0.007 -0.141* 0.484** 0.514** 0.419** 0.496** 0.387**
NIE/TA 1.000 -0.204** -0.265** -0.068 -0.129 0.077 -0.326** 0.148* 0.238** 0.058 0.202** 0.052
LOANS/TA 1.000 0.193 0.025 0.036 -0.015  0.109 -0.032 -0.094 -0.027 -0.117 0.067
LNTA 1.000 0.670** 0.619** -0.092 0.184** -0.624** -0.636** -0.622** -0.713** -0.476**
LNGDP 1.000 0.754** -0.130* 0.076 -0.841** -0.759** -0.791 -0.847** -0.657**
The nexus between economic freedom and Islamic bank performance

INFL 1.000 0.142* 0.046 -0.324** -0.094 -0.569** -0.135* -0.140*

CR3 1.000 -0.161** -0.084 0.002 -0.100 0.181** -0.177**
Z-SCORE 1.000 0.000 -0.142 0.045 -0.263** 0.220**
OVER_FREE 1.000 0.831** 0.815** 0.846** 0.869**
BUSI_FREE 1.000 0.639** 0.794** 0.732**
MONE_FREE 1.000 0.681** 0.618**
FINA_FREE 1.000 0.608**
CORR_FREE 1.000

Note: The table presents the results from Pearson correlation coefficients. ** and * indicates significance at 1% and 5% levels, respectively.

101
 Sufian et al.

Eichengreen and Gibson (2001) suggest that the effect of a Turning to the impact of banking sector’s concentration,
growing bank’s size on performance may be positive up to it can be observed from Table 4 that the coefficient of the
a certain limit. Beyond this point the effect of size could be three banks concentration ratio (CR_3) has consistently
negative due to bureaucratic and other reasons. exhibit a positive sign and becomes statistically significant
when we control for freedom from corruption (CORR_
As expected, the impact of credit risk (LLP/TL) is negative FREE). Within the context of the MENA Islamic banking
(statistically significant at the 10% level) suggesting that sector, the empirical findings clearly lend support to the
Islamic banks with higher credit risk tend to exhibit lower SCP hypothesis. The SCP hypothesis states that banks in a
profitability levels. The results imply that Islamic banks highly concentrated market tend to collude and therefore
should focus more on credit risk management, which earn monopoly profits (Short, 1979; Gilbert, 1984;
has been proven to be problematic in the recent past. Molyneux et al. 1996). It can be observed from Table 4 that
Serious banking problems have arisen from the failure the impact of banking sector risk (Z-SCORE) is positive
of financial institutions to recognize impaired assets and and highly significant. The result is in consonance with
create reserves to write off these assets. An immense help the findings of among others Boyd and De Nicolo (2006)
towards smoothing these anomalies would be provided by lending support to the stringent capital requirements
improving the transparency of the banking sector, which in of Basel II. From the policymaking point of view, the
turn will assist banks to evaluate credit risk more effectively empirical findings calls for a more effective policymaker’s
and avoid problems associated with hazardous exposure. role in reducing excessive bank risk exposures and at
the same time to induce more efficient risk management
Similarly, the empirical findings seem to suggest that practices by Islamic banks operating in the MENA banking
expense preference behaviour measured by NIE/TA has sectors.
consistently exhibit a negative relationship. The finding is in
consonance with the bad management hypothesis of Berger
and DeYoung (1997). Low measure of efficiency is a signal Does greater economic freedoms foster bank
of poor senior management practices, which apply to input- performance?
usage and day-to-day operations. Clearly, efficient cost To address the issue whether economic freedom matters in
management is a prerequisite to improve the profitability determining the performance of Islamic banks operating
of Islamic banks operating in the MENA banking sectors. in the MENA banking sectors, we re-estimate Eq. (2) to
Furthermore, most of the MENA countries banking sectors include the economic freedom indices variables discussed
have not reached the maturity level required to link quality in Section 3. The results are presented in columns 3 to 7
effects from increased spending to higher earnings. of Table 4. As observed, the empirical findings presented
in column 3 of Table  4 suggest that the coefficient of
Referring to the impact of capitalization, it can be observed the overall economic freedom (OVER_FREE) variable
from Table  4 that EQASS exhibits a positive relationship. is negative, but is not statistically significant at any
The result is consistent with the previous studies by conventional levels.
among others Isik and Hassan (2003), Goddard et al.
(2004), and Kosmidou (2008) providing support to the Concerning the impact of business freedom (BUSI_FREE)
argument that the well capitalized banks face lower costs on the profitability of Islamic banks, the empirical findings
of going bankrupt, thus lowers their cost of funding or presented in column 4 of Table 4 indicate that the coefficient
that they have lower needs for external funding resulting of the BUSI_FREE variable is positive. The results imply that
in a higher profitability level. Nevertheless, strong capital the greater ability to start, operate, and close businesses
structure is essential for banks in emerging economies fosters the performance of Islamic banks. Clearly, the
since it provides additional strength to withstand financial greater ability to set up new businesses in the MENA
crises and increased safety for depositors during unstable countries is a prerequisite to improve the performance of
macroeconomic conditions (Sufian, 2009). However, the Islamic banking sector.
it should be noted that the coefficient of the variable is
not significant at any conventional levels in any of the Referring to the impact of monetary freedom (MONE_
regression models estimated. FREE), it is interesting to note that the coefficient of the
variable is negative. If anything could be delved, the
The empirical findings seem to suggest that LNGDP has empirical findings indicate that higher (lower) government
positive and significant impact on the profitability of Islamic intervention in the market increases (reduces) the
banks operating in the MENA countries, lending support profitability of Islamic banks operating in the MENA banking
to the association between economic growth and banking sectors. A stable and reliable monetary policy is crucial to
sector’s performance. The high economic growth could business environment, as it may help firms and societies to
have encouraged Islamic banks to lend more and improve make investment, savings, and other long-term plans. High
the quality of their assets. The demand for financial services inflation rates not only confiscate wealth, but also distort
tends to grow as economies expand and societies become pricing, misallocate resources, and raise the cost of doing
wealthier. Likewise, it can be observed from Table 4 that the business. Furthermore, the value of a country’s currency
coefficient of the INFL variable exhibits a positive sign in the largely depends on the monetary policy of its government.
baseline regression model, implying that during the period A monetary policy that endeavors price stability and puts
under study the levels of inflation have been anticipated inflation at bay, enables firms to rely on the market prices
by Islamic banks operating in the MENA banking sectors. for their future investments plans.
This allows bank managements the opportunity to adjust
the profit rates accordingly and consequently earn higher As expected, the coefficient of the financial freedom
profitability. (FINA_FREE) variable entered the regression model with

102 Islamic banking and finance – Essays on corporate finance, efficiency and product development
 Table 4. Panel Generalized Methods of Moments (GMM) regression results.

ROA jt = b 0   + b 1 LOANS / TA jt + b 2 LNTA jt + b 3 LLP / TL jt + b 4 NIE / TA jt   + b 5 EQASS jt + b 6 LNGDPt + b 7 INFLt + b 8 CR3t + b 9 Z - SCORE t + b 9 OVER _ FREE t
+ b 10 BUSI _ FREE t + b 11 MONE _ FREE t + b 12 FINA _ FREE t + b 13 CORR _ FREE t + e jt

The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total loans divided by total assets; LNTA is a proxy measure
of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure
of bank diversification towards non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as non-interest expenses
divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a fraction of total assets; LNGDP is natural log of gross domestic products; INFL
is the inflation rate; OVER_FREE is the overall economic freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.

ONE STEP SYS-GMM

Eds. Hatem A. El-Karanshawy et al.

(1) (2) (3) (4) (5) (6) (7)

CONSTANT 5.073* -1.291 6.879 -12.792 5.575 -9.484 -4.259

(1.68) (-0.20) (0.42) (-0.98) (0.41) (-1.28) (-0.47)
Bank Characteristics

ROA t-1 0.126 0.155* 0.119* 0.152* 0.143** 0.107** 0.115

(1.19) (1.79) (1.65) (1.75) (2.32) (1.96) (1.43)
LOANS/TA -0.014 -0.016 0.004 0.004 -0.000 0.002 0.025
(-0.84) (-0.72) (0.19) (0.15) (-0.01) (0.09) (0.83)
LNTA -0.130 -0.871*** 0.256 -0.248 -0.337 -0.268 -0.059
(-0.68) (-2.60) (0.47) (-0.42) (-0.55) (-0.45) (-0.09)
The nexus between economic freedom and Islamic bank performance

LLP/TL -0.056* -0.026 -0.009 0.004 -0.002 0.001 0.006

(-1.67) (-0.97) (-0.34) (0.13) (-0.09) (0.04) (0.15)
NIE/TA -0.599** -0.807*** -0.529* -0.665* -0.736** -0.570* -0.641*
(-2.22) (-3.31) (-1.62) (-1.74) (-2.38) (-1.80) (-1.86)
EQASS 0.025 0.009 0.040 0.012 0.012 0.009 0.063
(0.90) (0.27) (1.08) (0.29) (0.28) (0.23) (1.49)
Economic Conditions

LNGDP 0.591** -0.771 0.428 -0.015 1.100* -0.593

(2.01) (-0.91) (0.81) (-0.04) (1.71) (-1.23)
INFL 0.026 -0.019 -0.006 -0.028 -0.112 -0.019
(0.29) (-0.19) (-0.07) (-0.32) (-0.86) (-0.26)
CR3 9.487 11.918 9.478 8.989 0.014 16.262*
(1.40) (1.37) (1.14) (1.12) (0.00) (1.87)
Z-SCORE 0.181** 0.280** 0.250** 0.251** 0.275*** 0.280**
(2.20) (2.47) (2.50) (2.38) (2.60) (2.53)
(Continued)

103
 Sufian et al.

a statistically significant positive sign, suggesting that

237.74***
-0.173***
banking security as well as independence from government

(-2.58)
control exerts positive impact on Islamic banks’ profitability.

0.027
0.243
0.999
The more banks are controlled by the government, the less

94
free they are to engage in essential financial activities that
facilitate private sector led economic growth.

Finally, it is observed from column 6 of Table  4 that the

183.61***
coefficient of the freedom from corruption (CORR_FREE)
0.127**

variable exhibits a negative sign (statistically significant

(1.97)

0.334
0.693
0.598
at the 1% level). The empirical findings from this study

94
clearly suggest that corruption (e.g. corruption in the
business environment, including levels of governmental,
legal, judicial, and administrative) has significant negative
impact on the profitability of Islamic banks operating in the
MENA banking sectors.
543.96***
(-1.20)

Robustness checks
-0.124

0.241
0.985
0.858

In order to check for the robustness of the results, we carry

94

out several sensitivity analyses. First, in light of Holmes

et al. (2008) arguments, we remove all the macroeconomic
and market conditions variables from the regression
models and repeat Eq. (2). The regression results are
presented in Table 5. All in all, it can be observed that the
1169.36***

coefficients of the baseline regression models stay mostly

the same: the sign and the order of magnitude remained
(0.71)
0.078

0.171
0.966
0.759

similar and significant as in the baseline regression models.

As observed, the empirical findings suggest that the
94

coefficient of the OVER_FREE, BUSI_FREE, and MONE_

Economic Freedom

FREE entered the regression models with a negative sign,

but are insignificant. From column 4 of Table  5 it can be
observed that the coefficient of the financial freedom
(FINA_FREE) retains its positive sign and is significant. On
999.69***

a similar vein, the empirical findings suggest that CORR_

(-1.16)
-0.264

FREE exhibits the same negative sign.

0.182
0.530
0.987
94

Second, it is also interesting to examine the persistence of

***, **, and * indicates significance at 1, 5, and 10% levels, respectively.

the explanatory variables over time. To do so, we lag all

the explanatory variables by one period and repeat Eq.
(2). The results are presented in Table  6. As can be seen,
the coefficients of the baseline variables remain stable as
in the baseline regression model. It is also worth noting
137.76***

that the coefficient of the LNTA variable is now significant

in four out of the six models estimated, while LLP/TL has
0.192
0.773
0.878

consistently exhibit negative and significant impact on

111

Islamic banks’ profitability levels. The empirical findings

suggest that the impact of capitalization (EQASS) is
positive in all of the regression models estimated.

It is also interesting to note that the impact of overall

Values in parentheses are z-statistics.

economic freedom (OVER_FREE) is now positive and

38.41***

significant. The empirical findings seem to support the

0.893
0.457
0.231

notion that economic freedom is a key to the creation

148

of an environment that allows a virtuous cycle of

entrepreneurship, innovation, and sustained economic
growth and development to flourish. Furthermore,
No. of Observationst-1
Table 4. (Continued)

economies with higher levels of economic freedom are

likely to enjoy higher living standards (Holmes et al.
2008). Holmes et al. (2008) points out that a higher level
Sargan p-value
AR(1) p-value
AR(2) p-value
MONE_FREE

CORR_FREE

of economic freedom is associated with a higher level of

OVER_FREE

FINA_FREE
BUSI_FREE

per capita GDP. They also suggest that countries which

Wald χ2

increase their levels of freedom tend to experience faster

growth rates and the freest economies also have lower rates
of unemployment and inflation. However, the coefficient

104 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance

Table 5. Panel Generalized Methods of Moments (GMM) regression results.

ROA jt = β 0 + β 1 LOANS / TA jt + β 2 LNTA jt + β 3 LLP / TL jt + β 4 NIE / TA jt + β 5 EQASS jt + β 6 LNGDPt + β 7 INFLt + β 8CR 3 t

+ β 9 Z - SCORE t + β 9OVER _ FREE t + β 10 BUSI _ FREE t + β 11 MONE _ FREE t + β 12 FINA _ FREE t + β 13CORR _ FREE t + e jt

The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total
loans divided by total assets; LNTA is a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure
of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure of bank diversification towards
non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as
non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the inflation rate; OVER_FREE is the overall economic
freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.

ONE STEP SYS-GMM

(1) (2) (3) (4) (5)

CONSTANT 4.095 4.410 10.278 -5.241 6.055

(0.45) (0.92) (1.45) (-1.23) (1.44)
Bank Characteristics

ROA t-1 0.151*** 0.140** 0.163*** 0.111** 0.194***

(2.98) (2.44) (3.04) (2.21) (3.07)
LOANS/TA -0.001 -0.002 -0.001 0.012 0.009
(-0.05) (-0.13) (-0.05) (0.85) (0.48)
LNTA 0.048 0.022 -0.132 0.544** -0.064
(0.16) (0.11) (-0.48) (2.07) (-0.27)
LLP/TL -0.041 -0.042 -0.038 -0.017 -0.040
(-1.26) (-1.33) (-1.37) (-0.55) (-1.42)
NIE/TA -0.827* -0.823** -0.911** -0.725* -0.975**
(-1.71) (-2.06) (-2.27) (-1.76) (-2.07)
EQASS 0.050** 0.050** 0.043* 0.057*** 0.068***
(2.50) (2.13) (1.77) (2.60) (3.44)
Economic Freedom

OVER_FREE -0.024
(-0.26)
BUSI_FREE -0.023
(-0.73)
MONE_FREE -0.075
(-1.39)
FINA_FREE 0.052**
(2.40)
CORR_FREE -0.058*
(-1.62)
Wald χ2 55.82 59.30 66.63 50.43 93.00
AR(1) p-value 0.556 0.584 0.654 0.934 0.505
AR(2) p-value 0.470 0.558 0.334 0.859 0.422
Sargan p-value 0.126 0.128 0.148 0.151 0.681
No. of Observationst-1 129 129 129 129 129

Values in parentheses are z-statistics.

***, **, and * indicates significance at 1, 5, and 10% levels, respectively.

of the variable is insignificant. Similarly, the empirical Third, we restrict our sample to banks with more than
findings seem to suggest that the coefficient of the BUSI_ three years of observations. All in all, the results remain
FREE is significantly related to the profitability of Islamic qualitatively similar in terms of directions and significance
banks operating in the MENA banking sectors. However, it levels. Finally, we address the effects of outliers in the
can also be observed from columns 5 and 6 of Table 6 that sample by excluding the top and bottom 1% of the
financial freedom and freedom from corruption loses their sample. The results continued to remain robust in terms of
explanatory power. directions and significance levels.8

Eds. Hatem A. El-Karanshawy et al. 105

 Sufian et al.

Table 6. Panel Generalized Methods of Moments (GMM) regression results.

ROA jt = β 0 + β 1 LOANS / TA jt + β 2 LNTA jt + β 3 LLP / TL jt + β 4 NIE / TA jt + β 5 EQASS jt + β 6 LNGDPt

+ β 7 INFLt + β 8CR3t + β 9 Z - SCORE t + β 9OVER _ FREE t + β 10 BUSI _ FREE t + β 11 MONE _ FREE t
+ β 12 FINA _ FREE t + β 13CORR _ FREE t + e jt

The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total
loans divided by total assets; LNTA is a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure
of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure of bank diversification towards
non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as
non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the inflation rate; OVER_FREE is the overall economic
freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.

ONE STEP SYS-GMM

(1) (2) (3) (4) (5) (6)

CONSTANT 1.290 -10.061* -6.400** 1.873 -4.750 0.559

(0.58) (-1.73) (-2.14) (0.21) (-1.25) (0.21)
Bank Characteristics

ROA t-1 0.530*** 0.405*** 0.492** 0.456*** 0.384** 0.484***

(3.34) (2.59) (2.55) (2.75) (2.14) (3.11)
ROA t-2 -0.156** -0.182*** -0.199*** -0.167*** -0.149** -0.160**
(-1.93) (-3.45) (-3.46) (-2.67) (-2.23) (-2.31)
LOANS/TA 0.017 0.006 0.024 0.012 0.012 0.015
(0.57) (0.22) (0.70) (0.45) (0.51) (0.52)
LOANS/TAt-1 -0.027 -0.003 -0.019 -0.011 -0.003 -0.011
(-0.89) (-0.12) (-0.64) (-0.43) (-0.11) (-0.40)
LNTA -2.492* -1.465 -2.223* -2.686* -1.361 -3.135*
(-1.64) (-0.98) (-1.84) (-1.84) (-0.62) (-1.77)
LNTAt-1 2.582* 1.910 2.649** 2.854** 1.814 3.334*
(1.63) (1.26) (2.05) (1.98) (0.87) (1.85)
LLP/TL -0.078*** -0.066** -0.061** -0.076*** -0.069*** -0.082***
(-2.84) (-2.35) (-2.31) (-3.53) (-3.87) (-4.38)
LLP/TL t-1 -0.019 -0.013 -0.026 -0.014 -0.001 -0.014
(-0.67) (-0.73) (-1.35) (-0.61) (-0.04) (-0.55)
NIE/TA -0.375 -0.044 -0.178 -0.096 -0.090 -0.124
(-1.09) (-0.16) (-0.56) (-0.30) (-0.31) (-0.37)
NIE/TA t-1 0.269 -0.020 0.172 0.032 -0.071 0.101
(0.82) (-0.05) (0.39) (0.07) (-0.15) (0.24)
EQASS 0.077** 0.076*** 0.074** 0.087*** 0.085*** 0.087***
(2.42) (2.89) (2.54) (2.98) (3.29) (2.97)
EQASS t-1 0.009 0.027 0.030 0.013 0.022 0.022
(0.24) (0.77) (0.81) (0.38) (0.68) (0.56)
Economic Freedom

OVER_FREE 0.113*
(1.61)
BUSI_FREE 0.057**
(2.03)
MONE_FREE -0.026
(-0.34)
FINA_FREE 0.031
(1.06)
CORR_FREE -0.023
(-0.80)
(Continued)

106 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance

Table 6. (Continued)
Wald χ2 567.29*** 1685.09*** 2902.11*** 3870.54*** 1471.55*** 1070.45***
AR(1) p-value 0.049 0.057 0.047 0.035 0.099 0.038
AR(2) p-value 0.465 0.152 0.202 0.097 0.378 0.181
Sargan p-value 0.219 0.229 0.306 0.061 0.049 0.114
No. of Observationst-2 111 98 98 98 98 98

Values in parentheses are z-statistics.

***, **, and * indicates significance at 1, 5, and 10% levels, respectively.

5.  Concluding remarks sector will be based on their operating performance.

Therefore, policy direction is expected to point towards
By using an unbalanced bank level panel data, the study
enhancing the resilience and performance of the banking
attempts to examine the impact of economic freedom on
institutions with the aim of intensifying the robustness
the performance of Islamic banks operating in the MENA
and stability of the Islamic banking sector.
banking sectors during the period 2000–2008. We find that
the larger, more diversified, and better capitalized banks are
Within MENA countries, enhancing economic freedom is of
relatively more profitable. The empirical findings seem to
an importance policy if the region is to attract more financial
suggest that efficient cost management is a prerequisite to
investments, improve weak financial infrastructure and
improve the profitability of Islamic banks operating in the
enhance the banking system performance. The current state
MENA banking sectors. Similarly, we find that higher credit
of the financial sector in MENA which mainly controlled
risk has negative and significant influence on the profitability
by state owned banks and dominate banking activities (up
of Islamic banks operating in the MENA banking sectors. The
to 95% of assets in several countries in the MENA region)
results suggest economic conditions exert negative impact on
resulting in poor services, high costs, and weak financing of
Islamic banks’ profitability levels when we control for overall
new investments and trade. As the MENA region competes for
economic freedom, monetary freedom, and freedom from
economic benefits for its citizens in the new global economy,
corruption. We also find that the level of inflation has positive
it is important that the policy makers in these countries
impact when we control for monetary and financial freedom.
to improve their quality of governance and transparency;
to promote a legal system that protects shareholders and
The findings from this study seem to suggest that greater
creditors rights; and to enhance their economic freedom.
financial and business freedom exerts positive impacts on
the profitability of Islamic banks operating in the MENA
banking sectors. The positive coefficient of the financial Notes
freedom variable indicate that higher (lower) freedom on 1. The basic tenets and principles of Islamic banking
the activities that Islamic banks could undertake increases are built upon the avoidance of usury (riba’) and
(reduces) their profitability, which is consistent with the the prohibition of impermissible activities as clearly
view that less regulatory control allows banks to engage in mentioned in the Quran, the Islam’s holy book and
various activities enabling banks to exploit economies of the traditions of Prophet Muhammad (sunnah):
scale and scope and generate income from non-traditional “Believers! Do not consume riba’, doubling and
sources. Furthermore, higher freedom on entrepreneurs redoubling…” (3.130); “God has made buying and
to start businesses is conducive to job creation and selling lawful and riba’ unlawful…” (2:274).
consequently increases Islamic banks’ profitability. 2. The estimates of the number of Islamic financial
Interestingly, the impact of monetary freedom is negative institutions vary considerably between institutions.
implying that higher (lower) monetary policy independence For instance, the International Monetary Fund
reduces (increases) Islamic banks’ profitability, providing (IMF) estimate that the number of Islamic financial
support to the benefits of government interventions. In institutions has increased to more than 300, while the
essence, although price stability without intervention is an Association of Islamic Banking Institutions Malaysia
ideal state for a free market, the empirical findings from this (AIBIM) suggests that there are around 486 Islamic
study clearly lend support to the benefits of government financial institutions around the world.
interventions in the markets. 3. MENA stands for Middle East and North Africa region
comprises of Algeria, Bahrain, Egypt, Iran, Iraq, Israel,
The findings of this study present considerable policy Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Qatar,
relevance. In view of the increasing competition Saudi Arabia, Syria, Sudan, Tunisia, UAE, and Yemen.
attributed to the more liberalized banking sector, bank 4. Islam has laid down some principles and prescribed
managements as well as the policymakers will be more certain limits for the economic activity of man so
inclined to find ways to obtain the optimal utilization that the entire pattern of production, exchange, and
of capacities as well as making the best use of their distribution of wealth may conform to the Islamic
resources, so that these resources are not wasted during standard of justice and equity.
the production of banking products and services. The 5. The Economic Freedom Index is released by The
findings pointed to the need for bankers to choose flexible Heritage Foundation and The Wall Street Journal.
operating environment and economic system favouring 6. Islamic banks income must not be uncontaminated by
the rapid development of a vibrant banking sector usury (riba’). Thus, in the case of the Islamic banking
to maximize their performance. From the regulatory sector, it is reasonable to assume that the interest rate
perspective, the performance of the Islamic banking to be the profit rate.

Eds. Hatem A. El-Karanshawy et al. 107

 Sufian et al.

7. Following Garcia-Herrero et al. (2009) among others, Berggren, N. (2003) “The Benefits of Economic Freedom: A
we instrument for all regressors. The macroeconomic Survey,” Independent Review, 8(2), 193–211.
characteristics are treated as exogenous (see among
Blundell, R. and Bond, S. (1998) “Initial Conditions and
others Baum et al. 2010).
Moment Restrictions in Dynamic Panel Data Models,”
8. To conserve space, we do not report the regression
Journal of Econometrics, 87(1), 115–143.
results in the paper but are available upon request.
Boyd, J. and Runkle, D. (1993) “Size and Performance of
Banking Firms: Testing the Predictions Theory,” Journal
References of Monetary Economics, 31(1), 47–67.
Abbasoglu, O.F., Aysan, A.F. and Gunes, A. (2007)
“Concentration, Competition, Efficiency, and Boyd, J.H. and De Nicolo, G. (2006) “The Theory of Bank
Profitability of the Turkish Banking Sector in the Post- Risk Taking and Competition Revisited,” The Journal of
Crisis Period,” Banks and Bank Systems, 2(3), 106–115. Finance, 60(3), 1329–1343.
Adkins, L., Moomaw, R. and Savvides, A. (2002) Button, K.J. and Weyman-Jones, T.G. (1992) “Ownership
“Institutions, Freedom and Technical Efficiency,” Structure, Institutional Organization and Measured
Southern Economic Journal, 69(1), 92–108. X-Efficiency,” The American Economic Review, 82(2),
439–445.
Albertazzi, U. and Gambacorta, L. (2009) “Bank Profitability
and the Business Cycle,” Journal of Financial Stability, ihák, M., Maechler, A., Schaeck, K. and Stolz, S. (2009)
5(4), 393–309. “Who Disciplines Bank Managers?” Working Paper,
International Monetary Fund.
Altman, M. (2008) “How Much Economic Freedom is
Necessary for Economic Growth? Theory and Evidence,” De Haan, J. and Siermann, C.L.J. (1998) “Further Evidence
Economics Bulletin, 15(2), 1–20. on the Relationship between Economic Freedom and
Economic Growth,” Public Choice, 95(3–4), 363–380.
Athanasoglou, P.P., Brissimis, S.N. and Delis, M.D. (2008)
“Bank Specific, Industry Specific and Macroeconomic De Haan, J. and Sturm, J.E. (2000) “On the Relationship
Determinants of Bank Profitability,” Journal of International between Economic Freedom and Economic Growth,”
Financial Markets, Institutions and Money, 18(2), 121–136. European Journal of Political Economy, 16(2), 215–241.

Bashir, A.H.M. (1999) “Risk and Profitability Measures DeNicolo, G., Bartholomew, P., Zaman, J. and Zephirin, M.
in Islamic Banks: The Case of Two Sudanese Banks,” (2003) “Bank Consolidation, Internationalization, and
Islamic Economic Studies, 6(2), 1–24. Conglomeration: Trends and Implications for Financial
Risk,” Financial Markets, Institutions & Instruments,
Bashir, A.H.M. (2001) “Assessing the Performance of 13(4), 173–217.
Islamic Banks: Some Evidence from the Middle East,”
paper presented at the American Economic Association DeYoung, R. and Rice, T. (2004) “Non-Interest Income
Annual Meeting, New Orleans, Louisiana. and Financial Performance at US Commercial Banks,”
Financial Review, 39(1), 101–127.
Baum, C.F., Caglayan, M. and Talavera, O. (2010) “Parlia­
mentary Election Cycles and the Turkish Banking Sector,” Dogan, E. and Fausten, D.F. (2003) “Productivity and
Journal of Banking and Finance, 34(11), 2709–2719. Technical Change in Malaysian Banking: 1989–1998,”
Asia-Pacific Financial Markets, 10(2–3), 205–237.
Ben Naceur, S. and Goaied, M. (2008) “The Determinants
of Commercial Bank Interest Margin and Profitability: Eichengreen, B. and Gibson, H.D. (2001) “Greek Banking
Evidence from Tunisia,” Frontiers in Finance and at the Dawn of the New Millennium,” Discussion Paper,
Economics, 5(1), 106–130. Center of Economic and Policy Research.

Berger, A.N. (1995) “The Relationship Between Capital El Qorchi, M. (2005) “Islamic Finance Gears Up”, Finance
and Earnings in Banking,” Journal of Money, Credit and and Development, 42(4), 46–49.
Banking, 27(2), 432–456. El-Gamal and Inanoglu (2004) “Islamic Banking in Turkey:
Berger, A.N. and DeYoung, R. (1997) “Problem Loans Boon or Bane for the Financial Sector,” Proceedings of
and Cost Efficiency in Commercial Banks,” Journal of the Fifth Harvard University Forum on Islamic Finance,
Banking and Finance, 21(6), 849–870. Cambridge: Center for Middle Eastern Studies, Harvard
University.
Berger, A.N. and Humphrey, D.B. (1997) “Efficiency
of Financial Institutions: International Survey and El-Gamal, M.A. and Inanoglu, H. (2005) “Efficiency and
Directions for Future Research,” European Journal of Unobserved Heterogeneity in Turkish Banking” Journal
Operational Research, 98(2), 175–212. of Applied Econometrics, 20(5), 641–664.

Berger, A.N., Bonime, S.D., Covitz, D.M. and Hancock, D. Garcia-Herrero, A., Gavila, S. and Santabarbara, D. (2009)
(2000) “Why Are Bank Profits So Persistent? The Roles “What Explains the Low Profitability of Chinese Banks?”
of Product Market Competition, Information Opacity Journal of Banking and Finance, 33(11), 2080–2092.
and Regional Macroeconomic Shocks,” Journal of Gilbert, R. (1984), “Bank Market Structure and Competition
Banking and Finance, 24(7), 1203–1235. – A Survey,” Journal of Money, Credit and Banking,
Berger, A.N., Hanweck, G.A. and Humphrey, D.B. (1987) 16(4), 617–645.
“Competitive Viability in Banking: Scale Scope Goddard, J., Molyneux, P. and J. Wilson (2004) “Dynamic
and Product Mix Economies,” Journal of Monetary of Growth and Profitability in Banking,” Journal of
Economics, 20(3), 501–520. Money, Credit and Banking, 36(6), 1069–1090

108 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance

Goddard, J., Molyneux, P. and Wilson, J.O.S. (2001) Miller, S.M. and Noulas, A. (1997) “Portfolio Mix and
European Banking: Efficiency, Technology and Growth. Large Bank Profitability in the USA,” Applied Economics,
New York: John Wiley and Sons. 29(4), 505–512
Hassan, M.K and Hussein, K.A. (2003) “Static and Dynamic Molyneux, P., Altunbas, Y. and Gardener, E.P.M. (1996)
Efficiency in the Sudanese Banking System,” Review of Efficiency in European Banking. John Wiley & Sons,
Islamic Economics, 14, 5–48. Chichester.
Hassan, M.K. (2006), “The X-Efficiency of Islamic Banks,” North, D., Thomas, R.P., (1973) The Rise of the Western
Islamic Economic Studies, 13(1–2), 49–77. World: A New Economic History. Cambridge: Cambridge
University Press.
Hassan, M.K. and Bashir, A.H.M. (2003) “Determinants of
Islamic Banking Profitability,” Paper Presented at the Pasiouras, F. and Kosmidou, K. (2007) “Factors Influencing
10th ERF Annual Conference, 16th–18th December, the Profitability of Domestic and Foreign Commercial
Morocco. Banks in the European Union,” Research in International
Business and Finance, 21(2), 222–237.
Hawtrey, K. and Liang, H. (2008) “Bank Interest Margins
in OECD Countries,” The North American Journal of Perry, P. (1992) “Do Banks Gain or Lose from Inflation,”
Economics and Finance, 19(3), 249–260. Journal of Retail Banking, 14(2), 25–40.
Heckelman, J.C. (2000) “Economic Freedom and Economic Powell, B. (2003) “Economic Freedom and Growth: The
Growth: A Short-Run Causal Investigation,” Journal of Case of the Celtic Tiger,” Cato Journal, 22(3), 431–448.
Applied Economics, 3(1), 71–91.
Rivard, R.J. and Thomas, C.R. (1997) “The Effect of
Heckelman, J.C. and Knack, S. (2009) “Aid, Economic Interstate Banking on Large Bank Holding Company
Freedom and Growth,” Contemporary Economic Policy, Profitability and Risk,” Journal of Economics and
27(1), 46–53. Business, 49(1), 61–76.
Heckelman, J.C. and Stroup, M.D. (2000) “Which Sarker, M.A.A. (1999) “Islamic Banking in Bangladesh:
Economic Freedoms Contribute to Growth?” KYKLOS, Performance, Problems, and Prospects,” International
53(4), 527-544. Journal of Islamic Financial Services, 1(3), 15–36.
Hirtle, B.J. and Stiroh, K.J. (2007) “The Return to Retail Saunders, A., Strock, E., and Travlos, N.G. (1990)
and the Performance of US Banks,” Journal of Banking “Ownership Structure, Deregulation, and Bank Risk
and Finance, 31(4), 1101–1133. Taking,” Journal of Finance, 45(2), 643–654.
Holmes, K.R., Feulner, E.J., and O’Grady, M.A. (2008) 2008 Short, B.K. (1979) “The Relation Between Commercial
Index of Economic Freedom. The Heritage Foundation: Bank Profit Rate and Banking Concentration in Canada,
Washington, D.C. Western Europe and Japan,” Journal of Banking and
Finance, 3(3), 209–219.
Hussein, K.A. (2003) “Operational Efficiency in Islamic
Banking: The Sudanese Experience,” Working Paper, Spathis, C., Kosmidou, K. and Doumpos, M. (2002)
Islamic Research and Training Institute (IRTI), Islamic “Assessing Profitability Factors in the Greek Banking
Development Bank. System: A Multicriteria Methodology,” International
Transactions in Operational Research, 9(5), 517–530.
Isik, I. and Hassan, M.K. (2003) “Efficiency, Ownership and
Market Structure, Corporate Control and Governance Staikouras, C. and Wood, G. (2003) “The Determinants
in the Turkish Banking Industry,” Journal of Business of Bank Profitability in Europe,” Paper presented at
Finance and Accounting, 30(9–10), 1363–1421. the European Applied Business Research Conference,
Venice, 9–13 June.
Kasman, A., Tunc, G., Vardar, G. and Okan, B. (2010)
“Consolidation and Commercial Bank Net Interest Stiroh, K.J. and Rumble, A. (2006) “The Dark Side of
Margins: Evidence from the Old and New European Diversification: The Case of US Financial Holding
Union Members and Candidate Countries,” Economic Companies,” Journal of Banking and Finance, 30(8),
Modelling, 27(3), 648–655. 2131–2161.
Kennedy, P. (2008) A Guide to Econometrics. Blackwell Sufian, F. (2009) “Determinants of Bank Efficiency During
Publishing: Malden, Massachusetts. Unstable Macroeconomic Environment: Empirical
Evidence from Malaysia,” Research in International
Kosmidou, K. (2008) “The Determinants of Banks’ Profits in
Business and Finance, 23(1), 54–77.
Greece during the Period of EU Financial Integration,”
Managerial Finance, 34(3), 146–159. Thakor, A. (1987) “Discussion”, Journal of Finance, 42(3),
661–663.
Kosmidou, K. and Zopounidis, C. (2008) “Measurement
of Bank Performance in Greece,” South Eastern Europe Tregenna, F. (2009) “The Fat Years: The Structure and
Journal of Economics, 6(1), 79–95. Profitability of the US Banking Sector in the Pre-Crisis
Period,” Cambridge Journal of Economics, 33(4), 609–632.
Kosmidou, K., Pasiouras, F. and Tsaklanganos, A.
(2007) “Domestic and Multinational Determinants Williams, B. (2003) “Domestic and International
of Foreign Bank Profits: The Case of Greek Banks Determinants of Bank Profits: Foreign Banks in
Operating Abroad,” Journal of Multinational Financial Australia,” Journal of Banking and Finance, 27(6),
Management, 17(2–3), 1–15. 1185–1210.

Eds. Hatem A. El-Karanshawy et al. 109

Efficiency of performance of banks in the
Gulf region before, during and after crises
(financial and political)
Abdel Latef Anouze
Department of Management and Marketing, College of Business and Economics, Qatar University, E: [email protected]

Abstract - Only a few cross-country empirical studies have been conducted to measure the perfor­
mance of commercial banks especially before, during, and after crises (financial or political). This study
makes an attempt to fill the gap in the literature by investigating the impacts of crises on Gulf Corporate
Council (GCC) commercial banks’ performance over the period 1997-2007. The rationale behind this
selection is that the GCC countries within this period witnessed two major crises: a political crisis (the
second Gulf war) and a financial crisis (the current global crisis). Clearly, it is important that a manager
recognizes the best bank policy in the face of each crisis that could help both bankers and regulators
in managing these crises. Also, the banking system within GCC countries comprises two different
operating banking systems, Islamic and conventional. As both are operating in similar environments, it
is of interest to examine whether one can make judgments concerning the success of their competitive
strategies, and other management-determined factors by using performance measures.

Two different evaluation methods are computed to measure bank performance: data envelopment
analysis (DEA), and classification and regression tree (CART). The overall results show that
conventional banks perform well during a political crisis, whereas Islamic banks performed better
during the financial crisis. However, this difference is not statistically significant, which means that
GCC commercial banks can be equally competitive when it comes to technical efficiency. Also, there
is no statistically significant relationship between bank geographical location and its efficiency score.
Moreover, the results confirm that large and small size GCC commercial banks are more efficient
than medium-sized banks. Out of the 24 environmental factors included in the study to investigate
the relationship between environmental factors (internal and external) and bank performance,
only 15 factors are considered to be important in predicting fully-efficient banks.

Keywords: data envelopment analysis, classification and regression tree, bank performance, Islamic
bank, GCC countries

1. Introduction compared with those made by their counterparts during

same period. On the other hand, it also helps policy
In the light of the on-going international financial crisis,
makers to develop a strong and healthy environment for
and the large costs generated for national and international
the banking sector by examining the impact of economic
financial systems, it is essential to assess the performance of
and financial reforms that have taken place. Meanwhile,
the financial sector in order to avoid the financial disaster
investors want to see how well a bank is performing before
becoming more complicated. Assessing banks’ performance
potentially investing in it. A high stock price alone is not
would help managers examine the success of managerial
a sufficient measure to use; they have to see how well a
decisions that they have taken before, during and after the
bank is performing too. Therefore, if a bank is to survive
crisis; to better understand their management effectiveness,
and succeed, it should learn the status of its efficiency and
and it would provide them with valuable reference for
how it compares with counterparts in same country or
improving their performance. Also, such assessment would
other countries. Hence, to identify appropriate financial
help managers to measure the success of these decisions

Cite this chapter as: Anouze A L (2015). Efficiency of performance of banks in the Gulf region before, during and after
crises (financial and political). In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate
finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Anouze

decisions that will achieve better allocation of financial studies of the impact of financial or political crises on
resources in a more efficient and effective manner, it is banking performance. To our knowledge, this is the first
important to assess bank performance at the country and/ study to explore the combined effect of financial and
or international level. A number of international empirical political crises on banking performance. However, our
studies have been conducted to measure the performance of work contributes and relates closely to several branches of
the banking sector before, during, and after crises (Mercan literature on bank performance, including studies of the
et  al. 2003; Jeon and Miller 2004 and 2005). However, impact of the financial crisis, bank health, and financial
all of these studies among others were carried out prior regulations on banking performance.
to the current global financial crisis. Therefore, this study
attempts to fill the gap in the literature by assessing GCC Few research studies have explored the impacts of the
commercial banks’ performance before, during and after current financial crisis on of bank performance. Xiao
the crises to guide bank managers and other stakeholders, (2009) used qualitative and quantitative tools to examine
such as policy makers and investors, in their decisions. the performance of French banks during 2006–2008.
The findings showed that French banks were not immune
There is a substantial body of literature discussing different but proved relatively elastic to the global financial crisis.
methods applied to evaluate the performance of banks Beltratti and Stulz (2009) studied the bank stock return
(e.g., Anouze 2010; Fethi and Pasiouras 2010; Berger across the world during the period from the beginning of
and Humphrey 1997). Reviewing 130  studies of the July 2007 to the end of December 2008; they found that
efficiency of financial institutions, Berger and Humphrey large banks with more deposit financing at the end of 2006
(1997) classified these methods according to the technical exhibited significantly higher returns during the crisis.
approach employed into parametric, such as the stochastic Cornett, McNutt and Tehranian (2010) analyzed the internal
frontier approach (SFA), and nonparametric, such as data corporate governance mechanisms and the performance
envelopment analysis (DEA). Application of these methods of US banks before and during the financial crisis; they
alone to evaluate banks’ performance determines efficiency found that the largest banks faced the largest losses during
scores but gives no details of factors related to inefficiency, the crisis. Dietrich and Wanzenried (2011) examined
especially if these factors are in the form of non-numeric how bank-specific characteristics, industry-specific, and
variables such as the operating style of the banking sector macroeconomic factors affected the profitability of Swiss
(Emrouznejad and Anouze 2010). This study proposes a commercial banks over the period from 1999 to 2009; their
comprehensive performance evaluation framework based results provide some evidence that the financial crisis did
on managerial, financial, and macroeconomic indicators have a significant impact on banks’ profitability.
to measure and predict banks’ performance. It allows
exploration and discovery of meaningful, previously These studies, among others, used a regression analysis
hidden information from given data. It integrates Data and limited bank performance to a single indicator—such
Envelopment Analysis (DEA) with the Classification and as profit, capital, and deposit to assets ratio—to measure
Regression Tree (CART) technique. DEA is a nonparametric bank performance during or after the crisis. Although
method for measuring the performance of Decision Making regression analysis is a useful tool, it tells nothing about
Units (DMUs) such as banks, hospitals, universities, or how to improve the performance, nor which is the best
services. It groups data into inputs and outputs to produce practice during or after the crisis. Also, it only counts for
a productive efficiency frontier against which an individual a single indicator, whereas banks could aim to maximize
bank or the banks of an entire country can be benchmarked. more than one indicator during their financial transactions.
Input variables within the DEA context are resources to be Furthermore, none of these studies have investigated the
minimized, while output variables are product or services performance of the GCC banking sector during the financial
to be maximized in order to achieve a high efficiency score. crisis.
The DEA efficiency score is a relative measure, which is
derived for each bank from the DEA based on the quality of Other researchers paid particular attention to the impact
transforming the inputs into outputs. CART, on other hand, of financial regulation on bank performance. Policy
is a nonparametric data-mining technique which allows makers introduce such regulations to develop a healthy
meaningful information to be explored and discovered from environment that increases competition and improves
a given data set. Unlike the DEA model, in which each case banking sector efficiency. Although are numerous studies
needs to be compared, CART produces results that can easily have examined the impact of financial regulations on banks’
be applied to determine the efficiency of a bank. A unique performance, the overall impact of financial regulation is
feature of CART is that it illustrates the data in the form of a ambiguous. Huang, Hsiao, Cheng and Change (2008);
decision tree so that the results can be presented in the form Brissimis, Delis and Papankiolaou (2008); Koutsomanoli-
of diagrams that are easy to understand. Integration of the Filippaki, Margaritis and Staikouras (2009); Hsiao, Chang,
two techniques would help stakeholders to assess, predict Cianci and Huang (2010); and Zhao, Casu and Ferrari
and identify the banks that are most likely to be troubling (2010), among others, found that deregulation improves
or, on the other hand, outperforming. Hence, stakeholders banking performance and stimulates competition in the
would have an overall understanding of banks’ performance financial market. In contrast, findings from other studies
and, consequently, better improvement policies could be such as those of Fukuyama and Weber (2002); Halkos and
developed for unsuccessful banks. Salamouris (2004); Park and Weber (2006): and Fu and
Heffernan (2009) show a decline in bank efficiency during
a period of financial reform.
2.  Literature review: Banking performance
Although there is a huge volume of published research on Finally, others researchers studied the real effects on bank
banking efficiency, little effort has been made to conduct performance of deterioration in bank health or competition

112 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

during the financial crisis. Almeida, Campello, Laranjeira,

and Weisbenner (2009) and Duchin, Ozbas and Sensoy Model 1a: Standard Input Model 1b: Standard Output
(2010) studied the effect of the recent financial crisis on Oriented DEA – VRS Oriented DEA – VRS
corporate investment. Their results show that corporate
investment declined significantly following the onset of Min θ n Max φ n

the crisis. Berger and Bouwman (2010) examined the subject ∑l x j ij ≤ θ x ij0 ; ∀i subject ∑l x j ij ≤ x ij0 ; ∀i
effect of pre-crisis bank capital ratios on banks’ ability to to j =1
n
to j =1
n
survive financial crises, market shares, and profitability ∑l j y rj ≥ y rj0 ; ∀r ∑l j y rj ≥ φ y rj0 ; ∀r
during the crises. Their findings show that capital helps j =1
n
j =1
n
banks of all sizes during banking crises; possession of
higher capital helped banks to increase their probability ∑l
j =1
j =1 ∑l
j =1
j =1
of survival, market shares, and profitability. Gryglewicz l j ≥ 0; ∀j, θ free l j ≥ 0; ∀j,φ free
(2011) studied the impact of both liquidity and solvency
concerns on corporate finance, and showed how changes To reach to the CRS from model 1a and 1b one can remove
in solvency affect liquidity, and also how liquidity concerns the following constraint from the above model ∑ j=1 l = 1
n

affect solvency through capital structure choice.

j

These studies, among others, provide a comprehensive

examination of the effects of a financial crisis on bank
efficiency (see table A1 in appendix for summary of previous In the input and output oriented DEA models (model 1a and
literature). Evaluation of this literature that presented in 1b, respectively) bankj0 is assessed under variable returns
table helped us to select the appropriate variables to be to scale, where the efficiency of bankj0 is the optimal value
included in our analysis. However, the impact of the crisis on of ? in Model 1a and 1/? in Model 1b (Thanassoulis 2001).
bank performance has not yet been fully analyzed. Moreover,
most of the reviewed research made use of statistical models One of the key concerns when we have a variable that takes
to study the impact of the financial crisis on firms. Statistical positive values for some banks and negative values for
models make some assumptions about statistical distribution others is that its absolute value should rise or fall for the
or propensities of the data, but most financial data do bank to improve its performance, depending on whether
not meet the statistical requirements of certain statistical the bank concerned has a positive or negative value on that
models; also, statistical tests are sensitive to sample size. variable (Emrouznejad, Anouze, and Thanassoulis 2010a).
On the other hand, an operations research technique makes For example, in the case of an output variable, if the bank
no assumptions about statistical distribution, and it is more has a positive value (profit) the output should rise to
accurate when testing complex or large samples (Demyank improve further but it should fall in absolute value as long as
and Hasan 2010). Therefore, use of such a technique to study it continues to be negative (loss). To overcome this problem
the impacts of the financial crisis on banking performance Emrouznejad and Anouze (2009) and Emrouznejad et al.
tends to be more appropriate in practical situations. (2010a and 2010b) treated each variable that has positive
value for some banks and negative for others as consisting
of the sum of two variables, and proposed a semi-oriented
3. Methodology radial model (SORM).
Data Envelopment Analysis (DEA)
DEA is a nonparametric relative performance evaluation Classification and Regression Tree (CART)
method developed by Charnes, Cooper and Rhodes in CART is the commonly used decision tree in data mining
(1978) considering constant return to scale (CRS). The CRS that was developed by Breiman, Friedman, Olshen and
model compares banks’ performance based only on overall Stone (1984) and further improved by Ripley (1996).
efficiency assuming constant returns to scale; however, it In principle, CART is similar to regression analysis since
ignores the fact that different banks could be operating both are used for prediction. However, CART has some
at different scales. To overcome this drawback Banker, advantages over the regression model:
Charnes and Cooper (1984) introduced the variable returns
to scale (VRS) model that is similar to the CRS Model, but it • A model generated by a CART is easier to understand
ensures that an efficient bank is only benchmarked against and relatively simple to interpret for non-statisticians
banks of similar size, while in the CRS model a bank may be (Breiman et al. 1984; Torgo 1997; Han and Kamber
benchmarked against banks which are substantially larger 2001)
or smaller than it. Subsequently, the original DEA models • There are no assumptions to be made regarding the
(CRS and VRS) have undergone many modifications and underlying distribution of values of the predictor
developments. Most of these developments occurred when variables as it is a nonparametric technique
the deficiencies of the original model were exposed during • CART can handle numerical data, as well as categorical,
its application to solving real life problems (Thompson, with either ordinal or non-ordinal structure.
Singleton, Thrall and Smith 1986).
These are important features of CART as they will
To introduce the DEA-VRS model, assume there are n banks eliminate analyst time which would otherwise be spent
(j = 1,…, n) using m inputs (xij i = 1,…m) and producing determining whether variables are normally distributed
s outputs (yrj, j  =  1,…s). DEA measures the technical and making transformations if they are not; specifically, it
efficiency of bank j0 compared with n peer group of banks, is important to use CART with DEA since DEA scores are
as illustrated in model 1a and 1b. skewed to one side.

Eds. Hatem A. El-Karanshawy et al. 113

 Anouze

To validate the results generated by CART, the dataset on the predicted CART results. The 3660 banks were
is partitioned into two datasets, training and validation divided into the two datasets, training and validation, by
(Han and Kamber 2001). The data then go into two major the ratio of 7:3 (Zhou and Jiang 2003; Emrouznejad and
phases of process: growth and pruning (Kim and Koehler Anouze 2010).
1995). In the growth phase, CART constructs a tree from
the training dataset. In this phase, either each leaf node
is associated with a single class, or further partitioning of Data description and analysis
the given leaf would result in the number of cases in one Banking industries in Gulf state countries
or both subsequent nodes being below some specified The early banking sector in the GCC countries experienced
threshold. In the pruning phase the CART generated in much foreign ownership primarily by British banks with
the growth phase is improved in order to avoid over-fitting. branches extending across all six GCC countries. Local
Also in this phase, the CART result is evaluated against the banks were uncommon as there was insufficient experience.
validation dataset in order to generate a sub-tree with the Subsequently, governments adopted central banking
lowest error rate. systems to strength local banks and to eliminate foreign
involvement. Today there are 68 local banks operating in
There are several criteria for measuring CART results. The GCC countries. These banks can be grouped according
predictive accuracy of a CART is commonly measured by to their operating style (mode of running financial
R-squared (average squared error); however, simplicity transactions) into two groups: Islamic and conventional
and stability are also important measures for a CART. banks. Unlike conventional banks, Islamic banks run their
Simplicity refers to the interpretability of the CART and is financial transactions free of interest (i.e., no interest
often based on the number of leaves in the CART. Stability rate is taken or given against any financial transaction).
of a CART refers to obtaining similar results for the training Among the 68 local banks, 18 are Islamic banks and 50 are
and validation datasets. One way to assess the stability of conventional banks.
the CART can be by comparing the predicted mean value of
the target variable (based on the training dataset) and the Figure  2 illustrates the share of Islamic and conventional
corresponding value for the validation dataset for each rule banking assets within each country. Saudi Arabia is the
of the CART (Han and Kamber 2001). largest investor in the GCC, holding 32% of the total bank
assets, with nine conventional banks and two Islamic
banks, and had total assets of US$ 239,095  million in
Proposed methodology (DEA with CART) 2007. The UAE, with fifteen conventional and five Islamic
Figure  1 illustrates the proposed analysis, that is, DEA banks, and total assets of US$ 224,542  million is the
and CART. The DEA stage is to compute the efficiency second largest investor in the area. Bahrain follows, with
score of each bank using DEA. Accordingly, the banks are nine conventional and six Islamic banks and total assets of
categorized into two groups: efficient banks (target  =  1) US$ 108,307 million, along with Kuwait, which has seven
and inefficient banks (target  =  0). In the CART stage the conventional and three Islamic banks and total assets of
classified efficiency score (0 or 1) is used as the target of US$ 108,174 million. Qatar is in fourth position, with four
CART while the environmental (explanatory) variables conventional and two Islamic banks and total assets of US$
is used as an input. However, an accurate CART requires 56,429 million, which represents only 7% of the total GCC
a large dataset, whereas our sample was limited to 60 assets. Finally, Oman has only six conventional banks and
banks. Therefore, a new stage was introduced before total assets of US$22,259  million, this representing only
the CART stage to increase the original dataset using the 3% of the total assets. Although our study aimed to include
bootstrapping technique. Thus, we randomly selected 60 all GCC commercial banks, eight banks are excluded on
banks (by replacement) and repeated this sampling 61 the basis of lack of availability of data, the remainder
times to achieve 3660 banks, so ensuring better accuracy comprising 48 conventional and 12 Islamic banks.

Figure 1.  Integrated data envelopment analysis and classification and regression tree.

114 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

Figure 2.  GCC commercial banks: Share of assets, (2007 data).

Table 1.  Descriptive analysis of input and output variables (2007).

Inputs/Outputs Variables Mean Std Dev Min Max

Inputs US$M Fixed Assets  7.28  24.16   0.03   413.34

Non-earning assets  21.86  55.08   0.00   609.61
Deposits 424.11 940.28   0.00 11,161.00
Outputs US$M Investments 226.01 525.43   0.00 5,766
Loans 256.26 531.37   1.27  7,528.63
Off-balance sheet 166.87 423.91   0.00  4,619.70
Net profit  8.70  21.52 -289.01   195.97

Data Description Data Envelopment Analysis (DEA)

Selection of proper input and output variables to define and
measure bank performance is always an extremely important Overall performance of GCC commercial banks
decision (Mercan et al. 2003). It is especially so when using To study bank performance before, during and after the
DEA, as different results may obtain from different sets of crises one grand-frontier (common-frontier) is computed
variables. Traditional bank behaviour theories described for all banks in all countries. The grand-frontier provides
banks as accepting deposits from households, and making a trend in the efficiency of banks, which would not be
loans to investors (Diamond and Dybvig 1983; Diamond available if we computed the efficiency of banks using a
1984; Gorton and Winton 2003). Yet, the change in bank separate frontier for each year. The approach employed,
involvement in markets and bank behaviour during crises therefore, provides variations in the efficiency of banks
requires a new theory of financial intermediation. In over both time and space. This comparison across time and
this study, the input variables include fixed assets, non- countries is on the same principle as the global frontier
earning assets, and deposit, while the output variables are used by Portela and Thanassoulis (2010). A VRS output-
investments, loans, off-balance sheet, and net profit. oriented model is used to measure banks’ efficiency,
since the CRS model is not possible in technologies
The selected input and output variables varied over where negative data can exist (Portela, Thanassoulis
the study period. It can be seen from table  1 that the and Simpson 2004). The efficiency score obtained for
minimum value of fixed assets—which is one of the all GCC commercial banks at the individual bank level is
inputs—is US$ 0.03 million whereas the maximum value aggregated to obtain the annual average efficiency scores
is US$ 413.34  million, with average US$ 7.28  million of all banks, and this is thenaggregated at country level
and standard deviation of US$ 24.16  million. Similarly and at operating style level.
for other variables, for example, the net profit, the
minimum net profit (loss) is US$ −289.01  million and For better capturing of bank performance during the
the maximum value is US$ 195.97 million, with average crises, the study period (1998–2007) is divided into four
of US$ 8.70  million and standard deviation of US$ periods:
21.52  million. Given the long time period analysed,
such variation would be expected; nonetheless, since 1. Before the political crisis (second Gulf war, 1998–2002).
DEA models are sensitive to observations it is likely that 2. Political crisis (2003).
significant levels of variation would also be found in 3. After political crisis (2004–2006).
banks’ performance. 4. During the financial crisis (2007).

Eds. Hatem A. El-Karanshawy et al. 115

 Anouze

It can be seen from able 2 that the overall average efficiency that the technical efficiency remains relatively stable over
score is 85.6% for all banks (60 banks); this suggests that the period 1998–2003, then improved a little to reach
with the same level of inputs and by adopting best practices, its highest level (92%) during 2004, while the period
GCC commercial banks can, on average, increase their 2005–2007 witnessed a volatility of the efficiency score,
outputs by 14.4% (i.e., 100–85.6%). However, the potential reach 79% at the end of the period. The year 2005, that
increment in outputs from adopting best practices varies is, two years after the Gulf war (political crisis) exhibits
from bank to bank. In general, GCC commercial banks have a decreased technical efficiency (77%) across all banks
the scope of producing 1.17 times (i.e., 1/0.856) as much studied. It seems that, over time, banks were wasting more
outputs from the same level of inputs. resources on average, relative to best practice technical
frontiers for the industry.
The literature on technical efficiency provides no consensus
on how efficiency in banking varies through time in response To find out whether the efficiency scores show a particular
to market forces (Berger 1993). However, since the study trend during the period 1998–2007, the question is
period covers a long and turbulent time (including the whether the mean efficiency score increased since 1998. In
second Gulf war in 2003 and the 2007 financial crisis), fact, Figure 3 shows that the trend of mean efficiency scores
it is expected that the political and financial crises will decreased over time. It moved in the same direction over
dominate the market forces. the period 1998–2002 (before the political crisis), then
declined a little to reach 86% during the second Gulf war.
It can also be seen from table 2 that, of the 60 commercial It fluctuated over the study period 2004–2006, reaching
banks covered in this study, there are ten banks which are its highest level in 2004, and deteriorating to the lowest
fully-efficient over the entire study period. The overall efficiency level in 2005–2006. The mean efficiency score
results show relatively low average efficiency scores; further declined in 2007 (the year of the financial crisis)
nevertheless, it is possible to detect a slight improvement to reach 79%. Although 2004seemed to be an atypical
in the efficiency levels over the study period (+2.2% year, it is important to note that the performance of GCC
between 1998 and 2004). In general, the table shows commercial banks varied over the study period.

Table 2.  Summary of banks’ technical efficiency.

Bank Code Efficiency score

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Average

Average 89 88 89 88 87 86 92 77 81 79 85.6
No of efficient banks 27 30 26 26 27 27 29 24 26 26 10

Figure 3.  Technical efficiency of GCC banks during study period (1998–2007).

116 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

Table  3. Statistical descriptive of average overall efficiency score of all banks at the individual bank level is
technical efficiency. aggregated at the operating style level to obtain the annual
average efficiency scores of Islamic and conventional banks,
1998– 2003 2004– 2007 as illustrated in the figure below.
2002 2006
Figure  4  shows that the Islamic banks outperformed the
Mean 88.02 86.20 83.37 79.30 conventional banks for the first four years (1998–2001),
Std Dev 14.04 16.50 20.03 24.60 and thereafter their performance declined. It reached its
lowest level of the study period (78.6%) by 2003 (second
Another appropriate way to study the trend is by looking at Gulf war). The performance improved to reach 88% by
the mean and the standard deviation of technical efficiency. 2004; however, it was still below the performance of the
If the banking markets of the GCC became more alike during conventional banks. Subsequently, the Islamic banks
the ten-year period under consideration, an increase in appeared to be ahead of the GCC commercial banks, with
mean technical efficiency and a decrease in the spread of an average efficiency score of around 89.3%.
technical efficiency would be expected. Table 3 shows that
the exact mean technical efficiency was relatively stable for For further analysis and comparison between the
the period 1998–2003, and then reached its highest level performance of Islamic and conventional banks over the
in 2004. The lowest efficiency score was exhibited during study period, a Mann-Whitney rank sum test was applied.
the 2005, which is two years after the second Gulf crisis, The Mann-Whitney test, which is an alternative to the
then fluctuated below the average for the last two years. independent group t-test, is a nonparametric (distribution-
The standard deviation was relatively stable for the period free) test for testing whether the number of times scores
1998–2003, and then reached its lowest level in 2004. The from one sample are ranked significantly higher than
standard deviation tends to be low when average technical scores from another, unrelated, sample. Similar to many
efficiency is high, and vice versa. These results strongly non-parametric tests, it uses the ranks of the data rather
support the view that traditional efficiency techniques than their raw values to calculate the statistic. For this test,
based on pooled frontier efficiency scores tend to estimate the efficiency score is considered as the group variable and
the actual efficiency levels of each bank. the bank operating style as the test variable.

The results of the Mann-Whitney test reveal that there is no

Islamic and conventional banks performance significance difference in bank efficiency performance due
before and during the crises to differences in operating style. Hence, the null hypothesis
To compare commercial bank performance based on their that the two efficiency scores have the same value of median
operating style, whether Islamic or conventional, the is rejected at the 5% level of significance.

Figure 4.  Performance of Islamic and conventional banks before and during the financial crisis.

Eds. Hatem A. El-Karanshawy et al. 117

 Anouze

Table 4.  Mann-Whitney test for 2007 results. Table 5.  Results of Kruskal-Wallis test.

Bank Type Sample Mean Mann-Whitney Z-value Bank N Mean x2 d.f. Asymp.
Size Rank U Location Rank Sig.

Islamic 12 29.6 245.5 -0.82 Bahrain 11 34.32 6.952 5 0.224

Conventional 48 34.04 Kuwait  9 19.94
Oman  5 39.20
Qatar  6 32.00
Saudi Arabia  9 36.00
UAE 20 28.05
Efficiency of commercial banks across GCC
countries
To measure commercial bank technical efficiency across
countries the efficiency score for all banks is aggregated
at country level to get the annual average efficiency scores
for each country. Figure  5  shows that the Kuwaiti banks information in order to assist them in developing a strong
outperform other countries banks before and during the and healthy environment. Similarly, investors want to know
second Gulf crisis. Thereafter, Kuwaiti banks decline, where to invest their money in a way that will maximize
becoming the worst performers of all GCC countries, their return.
and then become even worse during the financial crisis.
Although there is tough competition between Saudi and UAE Classification and Regression Tree (CART) analysis
commercial banks as they appear to be following the same
pattern before and during the second Gulf war, UAE banks’ The first stage results show the differences in
performance deteriorated during and after the financial inefficiency among banks in the six countries. In this
crisis. stage an efficient score is treated as a target variable,
while the internal and external environmental factors
Qatari banks performed badly before and during the are considered as predictors for the CART algorithm.
second Gulf war; however, performance increased rapidly These factors were identified from the related literature
after the crisis, but declined again during the financial and include economical, financial and political factors.
crisis. The performance of Bahraini and Omani banks Data of 24 factors was collected and tested to determine
followed the same behaviour before, during and after the the appropriate factors to include in the CART analysis.
second Gulf war, whereas the Omani banks outperformed Correlation tests showed a high correlation between
all other GCC commercial banks during the financial numbers of factors. For example, number of branches
crisis. and number of employees were highly correlated so we
included only a number of branches to reflect the size
For further investigation of the efficiency score across of banks. Also, the price/book value and price earnings
GCC countries, we adopted the Kruskal-Wallis rank test ratio was highly correlated so we included only a price/
(Sueyoshi and Aoki 2001) to examine whether or not scores book value factor to reflect the size of the stock market
vary among countries. The Kruskal Wallis X^2  statistics price for each bank. Therefore, fifteen factors were
are 6.952 (p = 0.224), meaning that there is no statistically considered as input factors for the CART algorithm (see
significant relationship between the geographical location table 6).
of a bank and its efficiency scores.
All factors as an input of CART algorithm
Previous analyses have been directed mainly at bank
managers; however, regulators may require different We built different CART models with a different selection
of input factors for CART with the efficiency score as
target. First, we included all factors as inputs and efficiency
classification as output. Figure  6  shows the importance
of variables. The fifteen environmental factors were
considered to be important in predicting the fully-efficient
banks; only seven of these factors are considered as primary
splitters for the decision tree. Assets structure is the most
important factor (100%), followed by financial strength
(92%) and ROA (91%), whereas operating style, population
density, size, and support rating have low importance. This
suggests that banks should give more importance to their
assets structure as it is one of the most important factors for
the efficiency of banks.

Figure  7  shows the predicated accuracy of the generated

tree:

Figure 5.  Bank performance across GCC countries Out of 3,660 cases, 1586 cases are actually efficient and
before and during the financial crisis. predicted to be efficient; and 2074 cases are inefficient

118 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

Table 6.  Statistical description of environmental factors.

Variable Descriptive Statistics

Variable type Minimum Maximum Mean Std. Deviation

Establish Date Categorical 1 5

Country Categorical  1.00   6.00
Inflation Numerical  3.60  14.00  8.61  4.71
Population Density Categorical  0.70  23.60
Operating Style Categorical  1.00   2.00
Internal Growth Numerical  0.27  45.15  14.93  8.74
GDP Growth Numerical  1.90   8.40  6.34  1.98
Bank Size Categorical 1 3
Return on Assets (ROA) Numerical   -2.53   8.28  2.76  1.53
Return on Equity (ROE) Numerical -34.18  33.37  17.79  8.86
Financial Strength Numerical  1.00  13.00  7.90  4.36
Support Rating Categorical  1.00   4.00
Loan to Deposit Ratio Numerical  28.50 1,904.35 138.76 263.59
Market Share Numerical  0.00   8.44  1.67  1.80
Asset Structure Numerical  0.02 3,534.00 209.70 518.82

Established date: Banks are grouped according to their established date into 5 groups to capture the age affect: group
5 banks established before 1960; group 4 (1960–1970); group 3 (1970–1980); group 2 (1980–1990) and group 1
(1990–2000). It is expected to have strong positive relationship between bank performance and the established date; the
older are the more efficient.
Country: Although, GCC countries mostly have the same regime, it is expected to have a variation in efficiency score
according to the bank geographical location due to differences in each country regulations.
Inflation: is an indicator of macroeconomic stability, and is directly related to the interest rate levels and, thus, interest
expense and revenue.
Population density: is measured as a ratio of country population to the GCC countries total populations. It is believed that
banks in heavily populated countries are more likely to operate closer to their optimal size than banks in less populated
country. Hence it is easier for bank management to sustain higher efficiency levels in heavily populated areas than in less
populated.
Operating style: to capture the efficiency of Islamic rule and regulations.
Internal growth rate: is calculated as the percentage of retained profits of the year on the equity at the beginning of the
year.
Bank size: is measured by the bank total assets, which classified into three groups hence, the larger banks (with total
assets more than US $15,000 Million), medium size (with total assets between US $5,000 – 15,000 Million) and small
size (total assets less than US $5,000 Million).
Profitability ratios: we measure this variable using return on assets (ROA) and return on equity (ROE).
Financial strength rating: it provides an opinion of a bank’s intrinsic safety, soundness and risk profile (Arab banking and
finance, 2007). It takes a scale from AAA (extremely strong finance and highly attractive operating environment) to D
(extremely weak financial condition and untenable position).
Support rating: it assesses the possibility that the bank will receive enough financial assistance from the government or
private owners in the event of difficulties to enable them to meet their financial obligations. It takes a scale from 1 (very
likely) to 5 (very unlikely) (Arab banking & finance, 2007).
Loan/Deposit: loan-to-deposit ratio is a measure of the extent to which banks are able to transform deposits into loans. It
is mainly used to measure the loan and deposit fund utilization of banks.
Market Share: is the ratio of total deposit of each bank to total deposit of all banks.
Asset structure: is the ratio of tangible assets to the total assets.

and predicted to be so. This means that the accuracy in 1. Financial strength is greater than or equal 4.0, ROA is
predicting the efficient and inefficient banks is 100%, greater than or equal to 2.59, and country is less than
which represents a high level of confidence. Certain of 4 (122 cases).
the rules extracted for efficient and inefficient banks are 2. Financial strength is greater than or equal 4.0, ROA is
as follows: greater than or equal to 2.59, country is greater than
or equal to 4, and internal growth is greater than or
equal to 4 (61 cases).
4.  Rules for efficient banks 3. Financial strength is greater than or equal to 4.0,
Banks are efficient (total of 1586 cases) if: ROA is less than 2.59, internal growth is greater than

Eds. Hatem A. El-Karanshawy et al. 119

 Anouze

Figure 6.  Factors importance in predicting fully-efficient banks.

or equal to 5.66, and established date is greater than

or equal 4 (100 cases).
6.  Rules for efficient banks
Banks are efficient if:

5.  Rules for inefficient banks 1. Established date is greater than or equal to 5, GDP
Banks are inefficient if: growth is less than 7.95%, inflation is less than 5.72,
country less than or equal to 4, support rating is
1. Financial strength is greater than or equal 4.0, ROA is greater than or equal to 2.5 but less than or equal to
greater than or equal to 2.59, country is greater than or 3.5, and operating style is 1 (61 cases).
equal to 4, and the internal growth is less than 4.44. 2. Established date is greater than or equal to 5, GDP
2. Financial strength is greater than or equal 4.0, ROA is growth is less than 7.95%, inflation is less than 5.72,
less than 2.59, and internal growth is less than 5.66 country less than or equal to 4, support rating is greater
(854 cases). than or equal to 2.5 but less than or equal to 3.5, and
operating style is 2 (61 cases that represent 16.7%).
3. Established date is greater than or equal to 5, GDP
External Factors as a Single Input of CART growth is less than 7.95%, inflation is less than 5.72,
Algorithm country less than or equal to 4, support rating is
To investigate the impact of the economic and political factors greater than or equal to 2.5 but less than or equal to
(external) on bank performance, CART is drawn by including 3.5, and operating style is 2 (183 cases).
only the external factors. All the external environmental
factors are considered to be important in setting the rules for 7.  Results and Discussion
fully-efficient banks. Operating style and established date are The overall technical efficiency for all GCC commercial
the most important factor, followed by inflation (89.14%). banks over the study period is 85.6%. It reaches its highest
Support rating and GDP growth seems to have medium level in 2004, which is one year after the second Gulf
importance whereas country and total population density crisis. The reason behind this unexpected improvement
have low. The predictive accuracy of the generated tree is in performance could be due to the injection of more
92%, which represents a high level of confidence. The rules money into the market through policy makers and
of efficient and inefficient banks that extracted as follow: regulators deciding to produce more oil in order to avoid
failure of the banking sector or bankruptcy after the Gulf
crisis. Therefore, the banking sector performed well,
until the regulators stopped the injection of funds, when
performance declined to reach its lowest level over the
study period. It is worth noting that the performance of the
banking sector in countries like Saudi Arabia (the largest
oil producer), Qatar (the largest gas producer), and Oman
all improved after the second Gulf war. The performance of
banks in all GCC countries deteriorated during the financial
crisis, except for Omani commercial banks, which reached
Figure 7.  Predicated accuracy of tree. their highest performance level during the crisis.

120 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

The highest average efficiency score is for the Saudi banks, banking sector and to establish a set of rules for the efficient
at around 89.8%, followed by banks of UAE, which have operation of banks.
an efficiency score of 86.3%. There seems to be tight
competition between Omani and Bahraini commercial
banks, which have average efficiency scores of 85.7% and 8. Conclusion
85.1%, respectively. Banks operating in Qatar are the least This paper investigates the performance of banks in the Gulf
efficient banks, with a score of around 81.3%. states before, during and after crises (political and financial).
The study period (1998–2007) includes two crises: the second
Although, not really comparable as they differ in terms of Gulf crisis (2003) and the global financial crisis (2007).
frontier, inputs and output variables, and the study period, This period allowed us to take look deeply into each bank’s
these results are in the line with the research of Al Shammari performance under two different situations. The results show
(2003), who found that the banks of Saudi Arabia and UAE that the overall technical efficiency of all GCC commercial
are ahead of those in the other GCC countries, while Qatar banks is relatively stable over time. The commercial banks of
and Bahrain have the poorest performing banks. Saudi Arabia appear to be ahead of the GCC countries, followed
by banks of UAE, whereas Qatar has the least efficient banks.
When the GCC commercial banks efficiency scores are However, there is no reason to believe that bank performance
compared with those of their counterparts in other countries differs from a statistical perspective according to location.
(e.g., Singapore banks −95%; Japan −87%; Germany −92%; Also, different regulations (if any) that have been put in place
Peru −98%), the results show that, on average, those of the within GCC countries during the crises have had more or
GCC banks are lower. Nevertheless, the results are relatively less the same impact on banks’ performance. Furthermore,
similar to the average efficiency for banks in industrial conventional banks performed better during the second
countries like France (84.3%), US (83%), UK (83.9%), Gulf crisis, whereas it was the Islamic banks that performed
Spain (82–84%), or other developed countries such as better during the global financial crisis. Nevertheless, from a
statistical perspective,Islamic and conventional banks rank
Lebanon (84%) and China (85%) (Ariss 2008; Avkiran more or less are same.
2009; Burki and Niazi 2009; Emrouznejad and Anouze
2010; Emrouznejad and Anouze 2009; Hermes and Nhung Out of the 24 environmental factors, fifteen were tested
2008; Huang et al. 2010; Ismail, Davidson and Frank 2009; and considered to be important, and only seven of them
Koetter 2008) are viewed as primary splitters for the decision tree. Assets
structure is the most important factor, followed by financial
The results suggest that, even though it is possible to detect a strength and ROA. The operating style, population density,
slight improvement in the overall efficiency scores, there are size and support rating all have low importance. Testing only
marked insignificant differences in bank efficiency levels across for the external environmental factors, operating style and
GCC countries. Islamic banks seemed to be more affected by established date are the most important factors, whereas
the Gulf war than were conventional banks, whereas, during country and total population density have low importance.
the international financial crisis, Islamic banks seemed to be
the more resistant. This could be due to the level of involvement Finally, this study contributes to the theory in developing
of the banks in the international financial institutes: Islamic a comprehensive framework for measuring bank
banks might have less involved than conventional ones and, performance, and in identifying the most important factors
hence, they were less affected. Also, it could be due to the that improve bank performance. The study also makes a
differences in the relation between bank and clients, which practical contribution as it is the first to assess the impact of
is based on profit-loss sharing in Islamic banks and based on financial and political crises on banks of the Gulf states, and
fixed rate (interest rate) in conventional banks, so the latter it provides useful information for banks managers, investors
made less profit compared with Islamic banks and policy makers for tracking banks’ efficiencies in order
to maintain a sustainable growing sector, and in providing
Assets structure, followed by financial strength and ROA early warning signals of a bank that is potentially at risk.
were most important, whereas, operating style, population
density, size, and support rating were found to be of low The results of this study are limited to the selected banks
importance. Considering only the external factors, the set and study period. Researchers are therefore encouraged
of efficiency rules that allow the prediction that a bank is to study the performance of the GCC banking sector after
fully-efficient indicate that it should be old, and operate in a the current global financial crisis; also, to compare the
country with high GDP growth and a lower level of inflation. performance of Islamic and conventional banks as the
Such rules benefit regulators or policy makers in their quest different financial tools used by each of them may lead to
to establish a healthy environment that will help their differences in performance.
banking sector to achieve a high level of efficiency as well as
be a regional financial hub. Managers could also benefit from
this analysis in working to improve their bank performance. References
DEA produces information to guide an improvement policy
Al Shammari, S., Structure-conduct-performance and effici­
for inefficient banks, and any such improvement may result
ency in Gulf Cooperation Council (GCC) banking markets,
in them being considered fully-efficient banks. Furthermore,
Unpublished PhD thesis, Wales University, 2003.
investors will find such results in their interest as they will
want to invest their money in such a way as to maximize Almeida, H., Campello, M., Laranjeira, B., and Weisbenner,
returns. Therefore, managers, policy makers, investors and S., Corporate debt maturity and the real effects of
researchers are encouraged to use the proposed methodology the 2007 credit crisis, Unpublished working paper,
to gain more information about the performance of the University of Illinois, 2009.

Eds. Hatem A. El-Karanshawy et al. 121

 Anouze

Anouze, AL., Evaluating productive efficiency: comparative Das, A., and Ghosh, S., Financial Deregulation and Profit
study of commercial banks in Gulf countries, Unpublished Efficiency: A Nonparametric Analysis of Indian Banks.
PhD thesis, Aston University, 2010. Journal of Economics and Business, 61, 2009. 509–528.
Ariss, R., Financial liberalization and bank efficiency: Demyank, Y., and Hasan, I., Financial crisis and bank
evidence from post-war Lebanon, Applied Financial failures: a review of prediction methods, Omega, 38,
Economics, 18, 2008. 931–946. 2010.315–324.
Avkiran, N., Removing the impact of environment with Diamond, D., and Dybvig, P., Bank runs, deposit insurance and
units-invariant efficient frontier analysis: An illustrative liquidity, Journal of Political Economy, 91, 1983. 401–419.
case study with intertemporal panel data, Omega, 37,
Diamond, D., Financial intermediation and delegated
2009. 535–544.
monitoring, Review of Economic Studies, 51, 1984.
Banker, R., Chang, H., and Lee, S-Y., Differential impact of 393–414.
Korean banking system reforms on bank productivity,
Dietrich, A., Wanzenried, G., Determinants of bank
Journal of Banking and Finance, 34, 2010. 1450–1460.
profitability before and during the crisis: Evidence
Banker, R., Charnes, A., and Cooper, W., Some models from Switzerland, J. Int. Financ. Markets Inst. Money,
for estimating technical and scale inefficiencies in data doi:10.1016/j.intfin.2010.11.002, 2011.
envelopment analysis, Management Science, 30, 1984.
Duchin, R., Ozbas, O., and Sensoy, B., Costly external
1078–92.
finance, corporate investment, and the subprime
Beltratti, A., Stulz, R., Why did some banks perform better mortgage credit crisis, Journal of Financial Economics,
during the credit crisis? a cross-country study of the 97, 2010. 418–435.
impact of governance and regulation, ECGI’s Finance
Emrouznejad, A, Anouze, AL, and Thanassoulis, E., A semi-
Working Paper No. 254/2009, 2009.
oriented radial measure for measuring the efficiency of
Berger, A., and Humphrey, D., Efficiency of financial decision making units with negative data, using DEA,
institutions: International survey and directions for EJOR, 200, 2010a. 297–304.
future research, European Journal of Operational
Emrouznejad, A. and Anouze, AL., A note on the modeling
Research, 98, 1997. 175–212.
the efficiency of top Arab banks, Expert Systems with
Berger, A., Distribution-free estimates of efficiency in Applications, 36, 2009. 5741–5744.
the U.S. banking system and tests of the standard
Emrouznejad, A. and Anouze, AL., Data envelopment
distributional assumptions, Journal of Productivity
analysis with classification and regression tree: a case of
Analysis, 4, 1993. 261–292.
banking efficiency, Expert Systems, 27, 2010. 231–246.
Breiman, L., Friedman, J., Olshen, R. and Stone, C.,
Emrouznejad, A., Amin, G., Thanassoulis, E. and Anouze,
Classification and regression trees, Pacific Grove,
AL., On the boundedness of the SORM DEA models with
Wadsworth-Monterey, USA, 1984.
negative data, EJOR, 206, 2010b. 265–268.
Brissimis, S. Delis, M. and Papankiolaou, N., Exploring the
Fethi, M. and Pasiouras, F., Assessing bank performance
nexus between banking sector reform and performance:
with operational research and Artificial Intelligence
Evidence from newly acceded EU countries, Journal of
techniques: A survey, EJOR, 204, 2010. 189–198.
Banking & Finance, 32, 2008. 2674–2683.
Fu, X. and Heffernan, S., The effects of reform on China’s
Burki, Abid and Niazi, G.S., Impact of financial reforms on
bank structure and performance, Journal of Banking
efficiency of state-owned, private and foreign banks in
and Finance, 33, 2009. 39–52.
Pakistan, Applied Economics, 1–14 URL: http://dx.doi.
org/10.1080/00036840802112315, 2009. Fukuyama, H. and Weber, W., Estimating output allocative
efficiency and productivity change: application to
Casu, B., and Girardone, C., integration and efficiency
Japanese banks, European Journal of Operational
convergence in EU banking markets, Omega, 38, 2010.
Research, 137, 2002. 177–190.
260–267.
Fukuyama, H., and Weber, W., A Slacks-based inefficiency
Charnes, A., Cooper, W., and Rhodes, E., Measuring the
measure for a two-stage system with bad outputs,
efficiency of decision making units, EJOR, 2, 1978.
Omega, 38, 2010. 398–409.
429–444.
Gorton, G., and Winton, A. Financial intermediation, In G.
Chiou, C-C., Effects of financial holding company act on
Constantinides, M. Harris, and R. (Stulz, Handbook of
bank efficiency and productivity, Neurocomputing, 72,
the Economics of Finance North Holland-Amsterdam,
2009. 3490–3506.
2003. 431–552.
Chiu, Y-H., and Chen, Y-C., The analysis of Taiwanese bank
Grifell-Tatjé, E., Profit, Productivity and distribution:
efficiency: incorporating both external environment,
differences across organizational forms—The case of
Economic Modelling, 26, 2009. 456–463.
Spanish banks, Socio-Economic Planning Sciences, 38,
Cornett, M., McNutt, J. and Tehranian, H., The financial 2010. 1–12.
crisis, internal corporate governance, and the
Gryglewicz, S., A theory of corporate financial decisions
performance of publicly-traded U.S. bank holding
with liquidity and solvency concerns, Journal of
companies, unpublished working paper, Boston College,
Financial Economics, 99, 2011. 365–384.
2010.

122 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)

Halkos, G. and Salamouris, D., Efficiency measurement Park, K. and Weber, W., A note of efficiency and productivity
of the Greek commercial banks with the use of growth in the Korean banking industry, 1992–2002,
financial ratios: a data envelopment analysis approach, Journal of Banking & Finance, 30, 2006. 2371–86.
Management Accounting Research, 15, 2004. 201–24.
Portela S, Thanassoulis, E, and Simpson, G., A Directional
Han, J and Kamber, M., Data Mining Concepts and Techniques, distance approach to deal with negative data in DEA:
San Francisco: Morgan Kaufmann Publishers, 2001. an application to bank branches, JORS, 55, 2004.
1111–1121.
Hermes, N. and Nhung, T., The impact of financial
liberalization on bank efficiency: evidence from Latin Portela, S. and Thanassoulis, E., Malmquist-type indices in
America and Asia, Applied Economics, URL: http:// the presence of negative data: An application to bank
dx.doi.org/10.1080/00036840802112448, 2008. branches, Journal of Banking & Finance, 34, 2010.
1472–1483.
Hsiao, H-C, Chang, H, Cianci, A, and Huang, L-H., First
financial restructuring and operating efficiency: Ray, S., and Das, A., Distribution of Cost and Profit
Evidence from Taiwanese commercial banks, Journal of Efficiency: Evidence from Indian Banking, European
Banking & Finance, 34 (2010), pp. 1461–1471. Journal of Operational Research, 201, 2010. 297–307.
Hsiao, H-C., Chang, H., Cianci, A., and Huang, L.-H., Ripley, B., Pattern recognition and neural networks,
First financial restructuring and operating efficiency: Cambridge University Press, Cambridge, 1996.
evidence from Taiwanese commercial banks, Journal of
Shleifer, A. and Vishny, R., Unstable banking, Journal of
Banking and Finance, 27 (2010), pp. 1461–1471
Financial Economics, 97, 2010. 306–318.
Huang, L-H., Hsiao, H-C., Cheng, M-A., and Chang, S-J., Effects
Siriopoulos, C., and Tziogkidis, P., How do Greek banking
of financial reform on productivity change, Industrial
institutions react after significant events? – A DEA
Management & Data Systems, 108, 2008. 867–886.
Approach, Omega, 38, 2010. 294–308.
Huang, T-H, Liao, Y-T and Chiang, L-C., An examination
Staub, R., Souza, G., and Tabak, B., Evolution of bank
on the cost efficiency of the banking industry under
efficiency in Brazil: A DEA approach, European Journal
multiple output prices’ uncertainty, Applied Economics,
of Operational Research, 202, 2010. 204–213.
42, 2010. 1169–1182.
Sueyoshi T. and Aoki, S., A use of a nonparametric statistic
Ismail A, Davidson, I, and Frank, R., Operating performance
for DEA frontier shift; the Kruskal and Wallis rank test,
of European bank mergers, The Service Industries
Omega, 29, 2001. 1–18.
Journal, 29, 2009. 345–366.
Thanassoulis, E., Introduction to the Theory and application
Jeon, Y. and Miller, M., Performance of domestic and
of data envelopment analysis: a foundation text with
foreign banks: The case of Korea and the Asian financial
integrated software, Kluwer Academic Publisher,
crisis, Global Economic Review, 34, 2005. 145–165.
Massachusetts, USA, 2001.
Jeon, Y. and Miller, M., The effect of the Asian financial
Thompson, R. Singleton, F. Thrall, R. and Smith, B.,
crisis on the performance of Korean nationwide banks,
Comparative site evaluations for locating a high-energy
Applied Financial Economics, 14, 2004. 351–360.
physics lab in Texas, Interfaces, 16, 1986. 35–49.
Kim, H. and Koehler, G., Theory and Practice of Decision
Torgo, L. (1997), Functional Models for regression tree
Tree Induction, Omega, 23, 1995. 637–652.
leaves, Proceedings of the Fourteenth International
Koetter, M., The stability of bank efficiency rankings Conference on Machine Learning (ICML 1997),
when risk preferences and objectives are different, The Nashville, Tennessee, USA, July 8–12, 1997.
European Journal of Finance, 14, 2008. 115–135.
Xiao, Y., French Banks Amid the Global Financial Crisis,
Koutsomanoli-Filippaki, A., Margaritis, D., and Staikouras, IMF Working Paper, WP/09/201, 2009).
C., Efficiency and productivity growth in the banking
Zhao, T., Casu, B., and Ferrari, A., The impact of regulatory
industry of Central and Eastern Europe, Journal of
reforms on cost structure, ownership and competition
Banking & Finance, 33, 2009. 557–567.
in Indian banking, Journal of Banking & Finance, 34,
Lozano-Vivas, A., and Pastor, T., Do performance and 2010. 246–254.
environmental conditions act as barriers for cross-
Zhou, Z. and Jiang, Y., Medical diagnosis with C4.5 Rule
border banking in Europe? Omega, 38, 2010. 275–282.
preceded by artificial Neural network ensemble, IEEE
Mercan, M., Reisman, A., Yolalan, R. and Emel, A., The Transactions on information Technology in Biomedicine,
effect of scale and mode of ownership on the financial 7, 200337–42.
performance of the Turkish banking sector: results of a
DEA-based analysis, Socio-Economic Planning Sciences,
37, 2003. 185–202.

Eds. Hatem A. El-Karanshawy et al. 123

 Anouze

Appendix

Table A-1.  Summary of selected studies.

Study
Study Country Period Approach Inputs Outputs

Banker, Korea 1995–2005 Intermediation (i) interest expense and (i) interest revenue, &
Chang, & (ii) other operating expense (ii) other operating
Lee (2010) revenue
Casu & European 1997–2003 Intermediation (i) Personnel expenses, (i) total loans and
Girardone, Countries (ii) other administrative (ii) other earning assets.
(2010) expenses,
(iii) interest paid,
(iv) non-interest expenses.
Chiou Taiwan 1999–2004 Intermediation (i) staff, (i) Provision of loan
(2009) (ii) fix asset, services (business &
(iii) bank deposits (including individual loans),
current deposits, savings (ii) investments
deposits, time deposits, check (iii) interest revenue and
deposits, & other deposits), & (iv) non-investment
(iv) salary expense. revenue.
Chiu & Taiwan 2002–2004 Intermediation (i) Number of employees, (i) Total amount of loans,
Chen (2009) (ii) total deposits, (ii) total investment,
(iii) fixed assets (iii) non-interest revenue.
Das & India 1992–2004 Intermediation (i) deposits, (i) Loans & advances,
Ghosh (ii) labor, (ii) investments,
(2009) (iii) capital/fixed assets (iii) other income.
(iv) equity
Fukuyama & Japan 2000–2006 Production and 1st stage: 1st stage:
Weber intermediation (i) labor, (i) deposits
(2010) (ii) physical capital, 2nd stage:
(iii) financial capital (i) loans, and
2nd stage: (ii) securities investments,
(i) deposits and
(iii) other business
activities.
Grifell-Tatjé Spain 1994–2004 Intermediation (i) Real operating profit (i) financial expense
(2010) from intermediation activities, (interest on deposits,
(ii) real gross loan and loans, labor expense)
financial income, and
(iii) average value of loans &
financial investments.
Hsiao et al. Taiwan 2000–2005 Intermediation (i) interest expenses, (i) interest revenue,
(2010) (ii) non-interest expenses, and (ii) non-interest
(iii) total deposits. revenue, and
(iii) total loans.
Lozano- European 2004 Production (i) labour, (i) loans, and
Vivas & Countries (ii) funds and (ii) other earning assets
Pastor (iii) physical capital
(2010)
Ray & Das India 1996–2006 Intermediation (i) deposits, (i) investments,
(2010) (ii) labor, (ii) earning advances, and
(iii) capital/fixed assets (iii) other income
(iv) equity & reserves
Siriopoulos & Greece 1995–2003 Intermediation (i) Personnel expenses, (i) Financial claims,
Tziogkidis (ii) provisions, (ii) operational income,
(2010) (iii) operational expenses. (iii) net income before
taxes.
Staub, Brazil 2000–2007 Intermediation (i) Labor, (i) outputs,
Souza, & (ii) capital, & (ii) loans and
Tabak (iii) purchased funds. (iii) investments,
(2010)

124 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency
and stock market performance: Evidence from
GCC countries
Samir Srairi1, Imen Kouki2, Nizar Harrathi3
1
Associate Professor, Finance, Faculty of Law, Economics and Management of Jendouba, (LAREQUAD), Tunisia,
E-mail: [email protected]
2
Assistant Professor, Finance, Higher Institute of Management, (LAREQUAD), Tunisia, E-mail: [email protected]
3
Assistant Professor of Quantitative Methods, Faculty of Economic Sciences and Management of Nabeul,
(LAREQUAD), Tunisia

Abstract - Using data envelopment analysis (DEA), this paper estimates the efficiency of 25 Islamic
banks operating in Gulf Cooperation Council (GCC) countries during the period 2003–2009.
It also examines the relationship between the efficiency of Islamic banks and the performance of
their stock. The results suggest that efficiency measures, particularly technical and pure technical
efficiency, have increased over the period of study but remain low as compared to conventional
banks. The inefficiency of Islamic banks can be attributed to pure technical inefficiency rather than
to scale inefficiency. We also find that large and small banks are more efficient than medium banks
in terms of overall technical efficiency. Furthermore, the empirical findings show that both technical
and pure technical efficiency changes are positively related to share returns, while changes in scale
efficiency have no impact on stock performance. Finally, the regression also indicates a significant
and positive association between market return and the book-to-market equity ratio with share
prices.

Keywords: banking, technical efficiency, stock performance, Islamic banks, data envelopment
analysis, GCC countries

1. Introduction world. During the last decade, the Islamic banking industry
has grown at a remarkable pace, at 20–30% per year being
An Islamic bank is an institution that mobilizes and invests
three times the rate for conventional banks. According to
financial resources according to Shariah. Islamic banking
many reports, the rapid and continued growth of Islamic
transactions are based on six basic principles: prohibition
banking is driven by multiple factors such as: increasing
of interest, risk sharing, money as potential capital,
demand from a large number of Muslims; increasing oil
prohibition of speculative behaviour, sanctity of contracts,
wealth of Muslim countries; low banking penetration in
and Shariah approved activities (Iqbal 1997).
Muslim majority nations; increasing demand from non-
Muslim customers and countries; and the support of
Islamic banking, which started to operate from the 1960s,
government and regulatory bodies for the development
exists today in all regions of the world, particularly in the
and promotion of Islamic banking.
Middle East and Southeast Asia. According to the report of
the Blominginvest bank, which was established in February
Furthermore, the Islamic financial system has been less
2009, more than 390 Islamic financial institutions are
affected than the traditional system by the latest economic
spread across 75 countries with total assets estimated
and financial crisis (2008), due mainly to its profit-loss
to be close to $1 trillion by 2010. The rating agency,
sharing principle, and also because of its strict prohibition
Moody’s Investors’ Service, forecast that Islamic bank
of investments in risky instruments, such as toxic assets and
assets worldwide will reach $4 trillion within five years.
derivatives. In addition, according to an IMF survey (2010)
The Islamic financial system is considered to be one of
and Chapra (2009), Islamic banks have contributed to
the fastest growing financial and economic sectors in the

Cite this chapter as: Srairi S, Kouki I, Harrathi N (2015). The relationship between Islamic bank efficiency and stock
market performance: Evidence from GCC countries. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Srairi et al.

financial and economic stability during the global financial comprehensive database on the GCC Islamic banking
crisis. The strong performance of Islamic banks over recent industry. Also, to the best of our knowledge, this is the
years has encouraged several universal banks in developed first study that relates the efficiency of Islamic banks in
countries to add Islamic products to their conventional GCC countries to their stock prices. Finally, our paper also
banking industry, through Islamic banks windows or attempts to study the impact of the recent economic and
Islamic banking subsidiaries. financial crisis on the performance of GCC Islamic banks,
and compares the efficiency of large, medium and small
In view of the rapid growth of Islamic banks, several issues banks.
are revealed about the performance of these financial
institutions. In addition, as Islamic banking was introduced
as a parallel system of conventional banks in the majority of 2.  Literature review
countries, the performance of the new form of banking may Two streams of literature are discussed in this study, the
have an impact on the soundness and stability of the banking first concerning the efficiency of Islamic banks, the second
system as a whole (Mariani 2010). Moreover, the last being relevant to the relationship between bank efficiency
economic and financial crisis has turned the focus towards and share performance.
Islamic financial institutions which, according to many
sources, have showed stronger resilience than conventional
banks (e.g., Moody’s; IMF working paper 2010). Despite Studies on Islamic bank efficiency
the strong position of Islamic banks, several studies (Iqbal While there is wide discussion in the literature on bank
2007; Iqbal and Van Greuning 2007) have identified efficiency within the conventional bank sector, particularly
weaknesses and vulnerabilities among Islamic banks in the for the developing countries and, to a smaller degree, the
areas of risk management (operational risk; weak internal transition economies, the work on Islamic banks remains
control processes) and human resource issues (quality limited. Even with the development of the Islamic banking
of management; technical expertise; professionalism). sector in several regions of the world, few studies have
Therefore, it will be interesting to analyse the performance evaluated the efficiency of the new form of banking, and
of Islamic banks during the last decade in order to provide noneconcern the relationship between bank efficiency and
some guidelines for managers, investors and policy makers share performance.
to improve the efficiency of these banks and to formulate
managerial strategies and public policies. Therefore, the According to Bashir (2007) and Sufian et  al. (2008), the
aim of this study is to investigate the efficiency of Islamic majority of studies on Islamic banks have focused on
banks operating in Gulf Council Cooperation (GCC) the concept issues describing the underlying principles
countries during the period 2003–2009, and to examine (Al-Omar and Iqbal 2000; Zahar and Hassan 2001; Lewis,
the relationship between the efficiency of Islamic banks 2008) and performance measures using the traditional
and the performance of their stock. To our knowledge, this financial ratios of these type of banks (Bashir 2001;
is the first study which analyses the relationship between Olson and Zoubi 2008; Srairi 2009). A few studies have
efficiency and share performance in the context of Islamic utilized frontier analysis techniques rather than traditional
banks in GCC countries. methods to estimate the efficiency of Islamic banks. Using
both the stochastic frontier approach (SFA) and the DEA
To gain a better understanding of the Islamic banking sector models, Hassan (2007) estimated a variety of parametric
in GCC countries, our analysis is conducted in two steps. techniques (cost, profit efficiency, and productivity)to a
First, by employing Data Envelopment Analysis (DEA) as panel of 43 Islamic banks operating in 22 countries during
a non-parametric approach, we estimate the technical the period 1993–2001. He found that Islamic banks are
efficiency of 25 GCC Islamic banks under the profit-oriented relatively more efficient in generating profits compared
method which defines cost variables as inputs, and revenue with control costs. In fact, the score of profit efficiency was
variables as outputs. In addition, to analyse the sources of found to be about 84%, while for cost the efficiency was
inefficiency of these banks, we calculated pure technical only 74%. The results also indicated that the major source
efficiency and scale efficiency as two components of of inefficiency was allocative inefficiency rather than
technical efficiency. We chose a period of six years between technical inefficiency.
2003 and 2009 in order to investigate the evolution of the
efficiency of Islamic banks over time. Moreover, in this Mokhtar et al. (2008) used a non-parametric DEA technique
study we attempt to compare the efficiency measures of and an intermediation approach to estimate the technical
Islamic financial institutions according to their size in terms and cost efficiency of the fully-fledged Islamic banks as
of total assets. Following several studies concerning the well as Islamic windows in Malaysia from 1997 to 2003.
conventional banking industry (e.g., Haddad et  al. 2010; The main results of the study revealed that, although the
Pasiouras 2008; Beccali et al. 2006), in the second stage of fully-fledged Islamic banks were more efficient than the
this paper we investigat the potential association between Islamic windows, the two types of Islamic banks were still
Islamic banks’ efficiency and their share prices. To meet this less efficient than the conventional banks. This finding also
objective, we regress annual stock returns calculated as the showed that the average efficiency of the overall Islamic
sum of daily share returns on efficiency scores obtained in banking sector increased over the survey period.
the first step, adding some control variables.
Employing the DEA model, Sufian et al. (2008) examined
This paper presents some interesting points compared the technical efficiency and its components (pure technical
with some other studies on Islamic banking efficiency efficiency and scale efficiency) of 37 Islamic banks
in GCC countries. First, our sample comprises more than operating in 16 MENA and Asian countries during the
90% of GCC Islamic banks assets, which makes it the most period 2001–2006. The results suggest that pure technical

126 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries

inefficiency dominated scale inefficiency of Islamic banks three efficiency levels: technical efficiency under constant
during all years except for the year 2006. On the other hand, returns to scale (CRS); technical efficiency under variable
the authors found that the MENA Islamic banks exhibited returns to scale (VRS); and scale efficiency. The results
higher technical efficiency compared to their Asian Islamic indicated that annual changes in technical efficiency (under
bank counterparts. CRS or VRS) were positively related to stock returns, while
changes in scale efficiency had an insignificant impact on
A more recent study concerning GCC countries was share performance. Erdem and Erdem (2008) used a DEA
conducted by Srairi (2010), who employed a SFA model with intermediation approach, and found no association
with country-specific environment variables and estimated between stock price returns and change in economic
the cost and profit efficiency of 71 commercial banks efficiency for Turkish banks.
during  the period 1999–2007. The empirical results
indicated that, on average, the conventional banks are Across international financial markets, Beccali et al. (2006)
more efficient in terms of cost and profit than the Islamic used both SFA and DEA approaches to estimate cost
banks. This study also revealed that both conventional and efficiency for a sample of banks operating in five European
Islamic banks in Arab Gulf countries are relatively more countries (France, Germany, Italy, Spain and United
efficient in generating profits than in controlling costs. Kingdom) in the year 2000. The results suggested that
the change in the prices of bank shares reflects percentage
changes in cost efficiency, particularly those derived from
Bank efficiency and share performance DEA. More recently, Liadaki and Gaganis (2010), who
While there is an extensive literature examining several employed a larger sample (15 EU countries and 171 banks)
issues on bank efficiency, such as the impact of liberalization and a longer time period (2002–2006) than Becalli et  al.
on the efficiency of banks (e.g., Chen et al. 2005; Das and (2006), estimated the cost and the profit efficiency by using
Ghosh 2006; Paul and Kourouche 2008), the sources the SFA model and taking into account the macroeconomic
of bank inefficiency (e.g., Grigorian and Manole 2006; and other country-specific characteristics. The main result
Pasiouras 2008; Sufian 2009), the comparison of the of this study showed higher profit inefficiency (21%) than
efficiency of banks according to country (e.g., Fries and cost inefficiency (10%). This means that European banks
Taci 2005; Kasman and Yildirim 2006; Inui et  al. 2008), are more efficient in controlling costs than in generating
ownership structure (e.g. Isik and Hassan 2003; Bonin profits. However, Srairi (2010) found that profit efficiency
et  al. 2005; Kyj  and Isik 2008), and comparison of type scores are more informative to shareholders and investors
of bank (foreign and domestic: Havrylchyk 2006; new in Gulf Arab countries. In fact, changes in profit efficiency
and old: Canhoto and Dermine 2003; conventional and have a positive and significant effect on stock returns, while
Islamic: Srairi 2010), only a limited number of papers have there is no association between changes in cost efficiency
investigated the impact of the efficiency of banks on stock and stock returns.
performance, and none of these papers have concerned
Islamic banks. The relationship between the efficiency of
banks and stock performance within the conventional 3.  Methodology and data
banking sector has been studied both on the basis of an In this study, we employ a three-stage procedure to analyse
individual country and for a cross-section of countries. the efficiency of Islamic banks and the relation to share
price performance:
Haddad et al. (2010) estimated the monthly efficiency and
productivity of 24 listed Indonesian banks and their market 1. A non-parametric approach (DEA technique) is used
performance using the non-parametricSlack-Based Model to estimate efficiency scores with an input-oriented
(SBM) approach over the period January 2006 to July model.
2007. They found that the stock market values of the banks 2. Annual stock returns are calculated on the basis of daily
were in accordance with their performance. The results share returns in order to measure the share performance
also indicated a positive correlation between the index of for each bank.
the Indonesian stock exchange (JCI) and bank efficiency. 3. The relationship between bank efficiency and stock
On the other hand, the findings suggest that Indonesian performance is examined by regressing the annual
banks with foreign ownership tend to be less efficient than return on stock against the yearly change of efficiency
their domestic counterparts. levels.

Using both DEA and SFA methods, Xiang and Shamsudding

(2009) calculated the technical, cost and profit efficiency DEA model
of nine publicly-listed Australian banks over the period From the literature, it is apparent that two models are used
1997–2007, and analyzed the potential link between these to examine the efficiency of banks. Parametric techniques,
efficiency scores and stock returns. They observed that such as Stochastic Frontier Analysis (SFA), Thick Frontier
an improvement in cost and profit efficiency, calculated Approach (TFA), and Distribution Free Approach (DFA),
using the SFA model, increased bank stock performance. use econometric tools and specify the function form for
However, the DEA efficiency scores were uncorrelated with the cost or profit function. On the contrary, non-parametric
stock returns. approaches, such as Data Envelopment Analysis (DEA)
and Free Disposable Hull Analysis (FDHA), do not
Pasiouras et  al. (2008) examined the association make an assumption concerning the functional form of
between the efficiency of ten Greek banks and their share the frontier, and use a linear program to calculate the
performance between 2000 and 2005. The authors used the efficiency level. The small size of our sample pushed us to
DEA technique (profit-oriented approach) and computed adopt the DEA technique, which was first introduced by

Eds. Hatem A. El-Karanshawy et al. 127

 Srairi et al.

Charnes et al. (1978). According to Avkiran (1999), DEA is assumption (TE  =  PTE*SE). To  calculate these efficiency
thought to work well with fewer data, fewer assumptions, scores, we employed the software DEAP version 2.1
and limited sample sizes. Furthermore, DEA does not developed by Coelli (1996).
require any specification of the functional form on the data
to construct the production frontier, and the distribution
forms of errors (Bauer et  al. 1998). However, DEA has Specification of inputs and outputs
some limitations. This technique is very sensitive to To estimate the efficiency frontier using the DEA techni­
outlying observations, and all deviations from the frontier que, we needed measures of inputs and outputs. In the
indicate inefficiency (Havrylchyk 2006). Moreover, the literature, there has been little consensus over which
DEA approach does not allow for any error in the data and, inputs and outputs should be used with the DEA model
in consequence, it may overstate the true levels of relative and how they could be measured (Berger and Humphrey
inefficiency for some entities (Drake and Hall 2003; Berger 1992). Consequently, several approaches are used in
and Mester 1997). Despite its limitations, we propose that bank efficiency studies: the production approach, the
DEA is a robust tool for examining the efficiency of Islamic intermediation approach, the operating approach, and the
banks in GCC countries. profit approach.

DEA is a deterministic model that can be used to examine the Following recent studies on bank efficiency (e.g., Drake et al.
relative efficiency of a number of entities (decision-making 2006; Pasiouras 2008; Sturm and Williams 2004), in this
units: DMUs) in the sample having the same multiple inputs study we adopt the profit-oriented approach. This method
and multiple outputs. To calculate the efficiency scores, a focuses on revenues as well as costs. It also has the advantage
linear programming model is solved for each bank. The DEA of allowing a better understanding of the strategies used by
model measures the efficiency of each DMU relative to all banks to respond to the changes in environment. Accordingly,
other DMUs, with the simple restriction that all DMUs lay three inputs and two outputs are selected to estimate
on, or below, the efficiency frontier (Das and Ghosh 2006). efficiency levels. Hence, the vector of inputs comprises:
If a DMU lies on the frontier, it is referred to as an “efficient employee expenses (x1), other operating expenses (x2) and
unit”. Otherwise, it is DEA-inefficient. The value of the loan loss provisions (x3). The vector of outputs includes two
efficiency score for each DMU is ranged between zero and variables: net interest income (y1 = interest income- interest
one. To define the best practice frontier, DEA can run under expense) and other operating income (y2).
either constant returns to scale (CRS), or variable returns
to scale (VRS). The main difference between these two
models is the treatment of returns to scale. The VRS model, Bank efficiency and share performance
which was defined by Banker et al. (1984), compares each Once the efficiency scores (TE, PTE, SE) and the annual
bank only with other banks operating in the same region share returns are computed,in the third stage of this study
of return to scale (banks of similar size). However, the CRS we examine the impact of the efficiency of Islamic banks
assumption is only justifiable when all banks are operating on performance (e.g., Liadaki and Gaganis 2010; Sufian
at an optimal scale. It means that a rise in inputs results and Abdul-Majid 2009; Erdem and Erdem 2008). The
in a proportionate rise in outputs. On the other hand, a relationship is checked using the following linear model:
DEA model can be constructed using an input-orientation
(minimizing inputs) or output-orientation (maximizing RSit = a + b1CEit + b2MRjt + b3BSFit + eit (1)
outputs) approach.
where RSit is the annual return on bank i’s stock in year t.
The input-orientation approach is defined as the ability CEit represents the annual percentage change in bank
of the bank to obtain a given level of outputs by utilizing efficiency and includes the technical (TE, model 1) or pure
a minimum combination of inputs; the opposite approach technical (PTE, model 2) or scale efficiency (SE, model 3)
analyzes the ability of banks to produce the maximum for bank i in year t. MRjt is the market return for the banking
level of outputs, given the current level of inputs (Cooper sector j in year t, and BSFit concerns some specific factors
et al. 2000). In this study, we adopt an input-oriented DEA and includes two variables, LTAit, which is the size of
technique because of the expressed interest of the Islamic bank i in year t measured as the natural logarithm of total
banking sector in more control costs. Many studies (e.g., assets, and BMit, which is the book-to-market equity ratio
Archer and Abdel-Karim 2002; Kamaruddin et  al. 2008) calculated as the ratio of the book value of a bank’s equity
conclude that the cost of funds and labour in Islamic banks to its market value. The a intercept represents the constant
is higher compared with those in conventional banks. of the model, bi is the parameters to be estimated and eit is
the disturbance term calculated as follows:
The DEA approach permits calculation for each bank of
the overall technical efficiency (TE) and its two components, eit = uit + vi
pure technical efficiency (PTE) and scale efficiency (SE).
PTE, also called “managerial efficiency”, represents the Since we have a panel regression combining cross section
failure of the bank to extract the maximum output from its and time series data, we estimate this model by using a
adopted input level and, hence, it relates to the ability of the fixed effect model (ni which represents bank specific effect
manager to utilize the firm’s given resources (Drake and Hall is fixed over time) and a random effect model (in the case
2003; Pasiouras 2008). SE, another indicator of efficiency, ni is considered as an error term). The fixed effect model is
measures the proportional reduction in input usage if the tested by the Fisher (F) test, while the random effect model
bank can operate at a point where the production exhibits is examined by the Lagrange Multiplier (LM) test. If  the
CRS (Kyj and Isik 2008). It can be computed by dividing null hypothesis of heteroscedasticity residual variance
TE under the assumption of CRS to the TE under the VRS is rejected, the ordinary least square (OLS) regression

128 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries

Table 1.  Summary statistics of dataset used in the study (average values).

2003 2004 2005 2006 2007 2008 2009

Panel A: inputs and outputsa

–  Employee expenses (x1) 23.47 25.21 32.7 48.35 67.63 73.89 70.27
–  Other operating expenses (x2) 19.66 21.26 36.47 43.77 52.78 58.83 61.22
–  Loan loss provision (x3) 19.95 19.05 17.71 13.16 18.92 61.73 120.47
–  Net interest income (y1) 103.92 116.50 168.35 200.29 272.66 275.72 262.78
–  Other operating income (y2) 27.86 38.62 77.86 113.12 157.13 132.67 95.82

Panel B: control variables and stock returnb

–  Total assets (US$ Millions) 2707 3065 3835 5200 7124 9062 9739
–  Book-to-market equity – 2.15 0.90 1.35 1.45 2.23 1.53
–  Annual stock return – 44.82 67.84 -27.33 16.56 -81.25 -15.33

 = variables in US$ million; b = all variables are in percentages, except where indicated.
a

is favored. To choose between these two models, we regression over 2003 to 2009. The table shows a great
calculated the Hausman test (H). increase of all inputs and outputs during the period of study.
In fact, we note that employee expenses, the other operating
expenses, the net interest income, and other operating
Data income have risen about 200%, 211%, 153%, and 243%,
Our sample comprises 25 Islamic banks operating in five respectively. The loan loss provision was constant during
Gulf Arab countries (GCC) with six banks in Bahrain, eight 2003–2007 and grew rapidly during the two last years of
banks in Kuwait, two banks in Qatar, two banks in Saudi the study period (2008 and 2009). It is interesting to note
Arabia, and seven banks in the United Arab Emirates, over that the crisis did not have the same effect on Islamic banks
the period 2003–2009. The choice of region is justified as is reported for conventional banks (Blominvest bank
for many reasons: first, the GCC countries, which comprise report 2009), the income of Islamic banks exhibiting only a
six states (Bahrain, Kuwait, Qatar, Saudi Arabia, the United small decrease of 4%. Finally, we note an increase of more
Arab Emirates, and Oman) hold the largest share (about than 25% of the average rate of assets.
61.6%) of Islamic bank’ assets in the world ($263  billion
in 2008). Saudi Islamic banks occupy the first place in
terms of GCC Shariah-compliant assets (35%), followed by 4.  Empirical results
Kuwait (24%), the United Arab Emirates (19%), Bahrain The analysis of the empirical findings on the efficiency of
(14%), and Qatar (8%). During the last decade, the Islamic Islamic banks in GCC countries is structured in two main
banking sector in GCC countries had achieved strong parts. First, we estimate the overall technical efficiency and
growth in term of total assets (over 35%). Also, since 2002, its components, measured by DEA method, and evaluate
the GCC region has been in a relatively strong position its evolution over time. Further, we attempt to examine
(7% growth between 2002 and 2008) and is expected to the efficiency of Islamic banks according to their size. In
continue at the same pace and to launch huge projects of the second part, we extend the analysis by examining the
more than $1 trillion during the next decade. Finally, while relationship between efficiency scores of Islamic banks and
the GCC states provide opportunities in many sectors and their share performance
offer ample liquidity in the banking sector, Islamic banks
are expected to further diversify their products and services
and so attract a wider clientele. In addition, the Islamic DEA efficiency measures
financial system will continue to spread to investment In this section, we examine the efficiency scores of Islamic
banking, project finance, capital markets, insurance, banks calculated under the profit-oriented approach and
wealth management and micro-finance (Iqbal 2007). obtained by the DEA technique. In order to analyse the
evolution of the efficiency of Islamic banks between 2003
The annual data of Islamic banks (financial statements) and 2009, we chose to construct a common frontier for
used to calculate the efficiency scores are collected from all banks in the sample; the implicit assumption was of
Bankscope Database of Bureau Van Dijk’s Company. The an absence of technical change during the period of study.
daily stock prices and market index are obtained from In this approach, the efficiency of each bank observed
Datastream. Since Gulf countries have different currencies, in different years is estimated in relation to a common
all the annual financial values are converted into US dollars benchmark technology (Canhoto and Dermine 2003).
using appropriate average exchange rates for each year.
Also, to ensure comparability of data across countries, all Table 2 provides a summary of annual means of efficiency
values are deflated to the year 2003 using each country’s indexes over 2003–2009 classified by year (panel A) and
consumer price index (CPI). by size (panel B). As can be seen from this table, overall
technical efficiency scores exhibit an upward trend from
Table  1  summarises the mean of inputs and outputs 2003 to 2009. The mean of TE varies from 61.2% (2003)
employed in the DEA model and also presents the average to 68.5% (2009) with an average equal to 65.5%. This
value of stock returns and control variables used in the result appears to show an improvement of the efficiency of

Eds. Hatem A. El-Karanshawy et al. 129

 Srairi et al.

Table 2.  Efficiency scores by year and size (average values).

TE PTE SE

Mean SD Mean SD Mean SD

Panel A: by year
2003 0.612 0.147 0.718 0.136 0.855 0.116
2004 0.643 0.195 0.738 0.149 0.864 0.138
2005 0.650 0.141 0.751 0.143 0.883 0.200
2006 0.642 0.112 0.778 0.141 0.839 0.150
2007 0.671 0.115 0.799 0.106 0.847 0.131
2008 0.681 0.162 0.813 0.128 0.838 0.127
2009 0.685 0.085 0.817 0.138 0.852 0.128
Panel B: by size
Small banks 0.669 0.156 0.676 0.153 0.990 0.157
Medium banks 0.653 0.118 0.762 0.123 0.840 0.140
Large banks 0.686 0.159 0.779 0.147 0.885 0.142
Overall 0.655 0.140 0.773 0.137 0.855 0.144

TE = technical efficiency; PTE = pure technical efficiency; SE = scale efficiency.

Islamic banks during the period of study. Indeed, efficiency the inefficiency in Islamic banks could be attributed to
scores, particularly TE and PTE, increased by 12% and 13% pure technical inefficiency (29.3%) rather than to scale
on average, respectively, while scale efficiency remained inefficiency (17%). It means that Islamic banks in GCC
constant. However, during 2008 and 2009, these measures countries are managerially inefficient in controlling costs
are constant but slightly changed and increased by 1.5% but manage their inputs efficiently. This finding of the
and 1.7%, respectively, compared to 2007. It is apparent dominant impact of managerial inefficiency over scale
that the last financial and economic crisis has affected the inefficiency is also reported in other studies, for example,
performance of Islamic banks, but to a lesser extent than Sufian et al. (2008) for Islamic banks in MENA and Asian
for conventional banks. According to Hasan and Dridi countries; Kyj and Isik (2008) for the Ukrainian banking
(2010), “the initial impact of the crisis on Islamic Banks’ industry; and Zaim (1995) for Turkish banks. According
profitability in 2008 was limited. However, with the impact to several studies (e.g., Bashir 2007; Iqbal 2007), the
of the crisis moving to the real economy, Islamic Banks inefficiencies in Islamic banks can also be attributed
in some countries faced larger losses compared to their to many other causes such as: limited number of short-
conventional peers”. term instruments; shortage of products for medium and
long term maturities; portfolios of Islamic banks being
Despite the increase in efficiency of Islamic banks between concentrated on equity and non-interest based financing,
2003 and 2009, the average of the input waste is large and especially focused on trade financing; small size
and equal to 34.5%. Therefore, there is still room for of banks; weak management; and lack of proper risk-
improvement in the performance of these banks through monitoring systems.
more efficient use of resources. Indeed, the efficiency scores
of Islamic banks in GCC countries are low compared not Furthermore, we attempt in this study to identify the
only to conventional banks (Srairi, 2010; Rosly and Abu nature of scale inefficiency, which can be due to increasing
Baker 2003) but also to Islamic banks in other countries. For returns to scale (IRS) or decreasing returns to scale (DRS).
instance, Kamaruddin et al. (2008) found that the average Table  3 displays statistics for the number of banks in the
of technical efficiency of the Malaysian Islamic banks is different categories of scale economies, and also presents
93% for the period 1998–2004. In a recent study of Islamic the returns to scale of banks classified by size. According to
banks in MENA and Asian countries, Sufian et  al. (2008) the figures in this table, only 19% of Islamic banks operate
found that Islamic banks in Indonesia during the period at their optimal scale (CRS) and the majority of banks are
2001–2006 are the most efficient from the Asian region, scale-inefficient (58% at DRS and 23% at IRS). It is also
exhibiting a mean technical efficiency of 92.3%. However, interesting to note that the share of the banks experiencing
several studies (e.g., Mohammed et al. 2008; Hassan et al. economies of scale (IRS) and diseconomies of scale (DRS)
2009) suggested that there are no significant differences are relatively constant during the sample period. The
between the overall efficiency results of conventional results confirm those shown in Table  2, relative to the
compared with Islamic banks. stability of scale efficiency of Islamic banks over the period
of study. Panel B of Table 3 also indicates that the majority
The decomposition of overall technical efficiency into PTE of Islamic small banks (83%) exhibited IRS (53%) or CRS
and SE components provides information on the source (30%), while the medium and large banks operated at DRS
of technical inefficiency. Table  2  reveals that the pooled (80%). It means that increasing the activities and size of
means for PTE and SE during the period analyzed are Islamic small banks may bring significant cost savings and,
of 77.3% and 85.5%, respectively. The result shows that in consequence, improve the technical efficiency of these

130 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries

Table 3.  Return to scale in Islamic banks by year and size.

DRS IRS CRS Total

of
Years Nb. of banks % share Nb. of banks % share Nb. of banks % share banks

Panel A: by year
2003 15 62 6 25 3 13 24
2004 15 62 5 21 4 17 24
2005 14 58 6 25 4 17 24
2006 13 52 6 24 6 24 25
2007 12 48 7 28 6 24 25
2008 15 60 5 20 5 20 25
2009 15 60 5 20 5 20 25
Total 99 58 40 23 33 19 172
Panel B: by size
Small banks 10 17 32 53 18 30 60
Medium banks 44 79 3 5 9 16 56
Large banks 45 80 5 9 6 11 56

DRS: decreasing returns to scale; IRS: increasing returns to scale; CRS: constant returns to scale.

banks, in contrast to the case of expansion by the medium practice, while Islamic medium banks need to increase
and large banks. A similar finding has been made for other their scale efficiency.
countries such as Singapore (Rezvanian and Mehdian
2002), Turkey (Isik and Hassan 2002) and India (Rezvanian
et al. 2008). Efficiency and share performance
To assess the relationship between the efficiency of
In order to compare the efficiency scores of banks according Islamic banks and their share prices, we regress annual
to their size, we categorized the sample banks into three stock price returns on annual percentage change of
groups based on their total assets, with an approximate efficiency scores, derived from DEA analysis, with other
number of banks in each category. The first group comprises explanatory variables. Models 1, 2 and 3  in Table  4
nine small banks with an asset size of less than $3 billion. present the regression results estimated by the fixed-
The second group includes medium banks (eight banks) effect model for technical, pure technical, and scale
whose assets are between $3 and $5 billion, while, the last efficiency changes respectively. The results indicate that
group comprises large banks (eight banks) whose assets both technical and pure technical efficiency changes
exceed $5 billion. have a positive and statistically significant (1% for
TE and 5% for PTE) effect on stock returns. Indeed,
In terms of overall technical efficiency, panel B of Table 2 the share prices of Islamic banks respond positively
shows that large (68.6%) and small (66.9%) banks are towards improvement in managerial efficiency. Hence,
the most efficient, while the medium banks presented the it seems that information regarding the efficiency of
lowest mean TE of 65.3%. This is consistent with several banks is reflected in the stock prices of banks. In fact,
studies which reported a significant positive association in an efficient market, share prices incorporate all
between size and efficiency (e.g., Drake and Hall 2003; publicly available information (Fama 1970). Thus,
Chen et  al. 2005; Pasiouras 2008; Srairi, 2010). Large according to Beccalli et al. (2006) and others, efficient
banks present some advantages over small and medium banks can better improve their share price performance
banks. According to Kyj and Isik (2008), “large banks than inefficient banks. So, our results are in line
may be able to hire a better management team, utilize with several studies in other countries which found
better technology, be located in larger, more competitive a positive association between technical efficiency
markets, and have more diversified loan portfolio. Large change and share performance (e.g., Pasiouras et  al.
banks, thus, may have lower default risk, and lower 2008 for Greek banks; Xiang and Shamsudding 2009
borrowing costs”. However, other studies found a negative for Australian banks; Sufian and Abdul Majid 2009 for
(e.g., Christopoulis et  al. 2002; Bonin et  al. 2005) or no China banks). However, other researches (e.g., Liadaki
significant (e.g., Berger and Hannam 1998; Girardone and Gaganis 2010 for European banks; Ioannidis et  al.
et  al. 2004) relationship between size and efficiency. On 2008 for Asian and Latin American banks; Chu and
the other hand, the result indicates that large (77.9%) Lim 1998 for Singapore banks) show that changes in
and medium (76.2%) banks are more pure technically stock returns reflect changes only in profit efficiency
efficient than small banks (67.6%). However the latter rather than in cost efficiency. According to Liadaki and
display a superior measure for scale efficiency, this being Gaganis (2010), these results can be explained by the
10.5% and 15% higher than for medium and large banks, fact that rational shareholders and investors are more
respectively. Consequently, it seems that Islamic small interested by the profit of banks as an indicator of the
banks need more improvement in terms of managerial future dividends. Moreover, cost efficiency reflects the

Eds. Hatem A. El-Karanshawy et al. 131

 Srairi et al.

Table 4.  Regression results of equation (1).

RSit = a + b1CEit + b2MRit + b3BSFit + eit
RSit is the annual return on bank i’s stock in year t. CEit represents the annual percentage change in bank efficiency and
includes the technical (TE, model N°1) or pure technical (PTE, model N°2) or scale efficiency (SE, model N°3) for bank
i in year t. MRjt is the market return for banking sector j in year t and BSFit concerns some specific factors and includes
two variables: LTAit is the size of bank i in year t measured as the natural logarithm of total assets and BMit is the book-to-
market equity ratio calculated as the ratio of the book value of a bank’s equity to its market value.

Variables Model 1 Model 2 Model 3

Constant: α 2.15 2.03 2.11

(2.75)* (2.62)* (2.70)*
Annual change in efficiency scores
Technical Efficiency: TE 3.14 – –
(2.30)**
Pure Technical Efficiency: PTE – 7.07 –
(2.61)*
Scale Efficiency: SE – – 0.83
-0.83
Control Variables
Market return: MR 0.84 0.86 0.87
(9.05)* (8.88)* (8.96)*
Size of bank: LTA 0.71 0.56 0.51
(1.61) (0.91) (1.60)
Book to market equity ratio: BM 7.31 7.35 7.12
(2.55)** (2.49)** (2.41)**
Adjusted R2 0.617 0.597 0.499
F value 38.66* 35.29* 35.60*
Nb. Observations 125 125 125
F value 48.3 46.4 43.2
LM 385.2 391.3 380.4
Hausman test 36.1 33.8 31.2

T-statistics are between parentheses; *, and ** indicate statistical significance at 1%, and 5% respectively.

capability of managers but it is not directly observed in of BM, implying a possibility of market expectation of
the stock market. systematic risk.

From Table 4 (model 3), it is also noted that the estimated

coefficient of scale efficiency change is positive but it is not 5. Discussion
statistically significant. It means that scale efficiency does Many policy implications and recommendations can be
not have any impact on a bank’s share returns. This finding derived from the results of this paper. First, since Islamic
is also confirmed by the coefficient of bank size which is banks in GCC countries exhibited a lower level of efficiency
insignificant in all of the regression models. A similar result compared to conventional banks, it is necessary for these
was also found by Pasiouras et al. (2008) and Sufian and institutions to promote and enhance their functioning in
Abdul Majid (2009). several areas (Bashir 2007; Iqbal 2007). Islamic banks are
still operating with a limited number of instruments for the
With regard to the control variables and their influence short-term, and there is a shortage of products for medium-
on stock returns, Table  4  indicates that market return in to long-term maturities. In this regard, Islamic banks have to
all models has the expected sign and a significant power offer new products and modes of finance that enhance risk
to explain the variation in stock prices. This result, which management and portfolio diversification. Due to limited
is consistent with previous studies (e.g., Xiang and size and resources, Islamic banks are unable to reap the
Shamsuddin 2009; Erdem and S.Erdem 2008), shows that benefits of economies of scale and are also unable to afford
stock price returns of Islamic banks are positively related to high cost management information systems to assess and
the overall performance of the market. On the other hand, monitor risks. Accordingly, Islamic banks have to perform
the association between the ratio of book-to-market value strategic alliances with other Islamic financial institutions,
(BM) and share performance is positive and significant and collaborate with conventional banks which are more
at the 5% level for all models. However, our results are sophisticated in terms of financial engineering. Further,
different from the study of Xiang and Shamsuddin (2009) to better control and reduce their costs, Islamic banks
concerning Australian banks which found a negative sign need to invest more in technology, to develop innovative

132 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries

methods in terms of risk management, and to increase Using the efficiency scores of Islamic banks, we analysed
the efficiency of their staff by investing in training and the link between efficiency change and stock returns.
development. The results also show that there has been an The results derived from the fixed effect model show that
improvement in efficiency of Islamic banks over the period percentage changes in the prices of bank stocks reflect
of liberalization in Gulf countries. Therefore, authorities in percentage changes in both technical and pure technical
this region should continue to reinforce financial reforms, efficiency. However, we do not find any significant
increase economic integration between countries, and relationship between scale efficiency and stock returns.
undertake constructive policy actions to develop Islamic Thus, our results seem to support the argument that
capital markets which help to integrate Islamic financial stock returns respond positively towards improvement in
institutions into regional and international financial managerial efficiency, but do not react towards changes
systems. Finally, while there is a positive association in scale efficiency (Sufian and Abdul Majid 2009). Hence,
between the performance of Islamic banks and their stock the efficiency of a bank’s operation provides significant
price returns, it appears that efficiency measures contain information about its share price performance, which is not
important and helpful information which could be used by explained by market movements.
managers of banks, shareholders and investors.
One implication of the findings is that managerially-efficient
banks should be more profitable and therefore generate
6. Conclusions greater shareholder returns. This is in line with the efficient
Islamic banking is viewed as competitive and an alternative market theory that, in an efficient market, a change in cost
to the conventional banking system in many states of the efficiency should be incorporated in the price formation
world, particularly in GCC and some Asian countries. In process. Finally, the study also revealed that market return
addition, during the last decade, Islamic banking assets and ratio of book-to-market value have a positive impact on
have been growing at a faster pace (an average annual stock returns.
growth of 20%) than the overall banking system, with the
expectation that it will play an increasing important role
in the coming years. Moreover, the Islamic financial system References
has proved to be the least affected by the last economic
Al-Omar, F.A., Iqbal, M. Some strategic suggestions for
and financial crisis. In the light of these considerations, it
Islamic banking in the 21st century, Review of Islamic
is important to assess and analyse how Islamic banks have
Economics, 9, 2000. 37–56.
performed during the past few years.
Archer S, Abdel-Karim R. Islamic Finance. Euromoney
In the present study, we estimate the efficiency of Books, UK. 2002.
25 GCC Islamic banks over the period 2003–2009. By
Avkiran, N. K The evidence on efficiency gains: The role
using a non-parametric DEA technique, under the profit
of mergers and the benefits to the public, Journal of
oriented approach, we calculate technical, pure technical
Banking and Finance, 23, 1999. 991–1013.
and scale efficiencies to study the evolution of these
efficiency measures across time and to analyse the size Bashir, A.H.M. Assessing the performance of Islamic banks:
efficiency relationship. Additionally, this paper attempts Some evidence from the Middle East, in American
to investigate the influence of the performance of Islamic Economic Association Annual Meeting, New Orleans,
banks in terms of efficiency on their stock prices. Several USA, 2001.
important findings emerge from this present study. The
Bashir, A.H.M. Islamic banks participation, concentration
results indicate that the average technical efficiency was
and profitability: Evidence from MENA countries.
equal to 66% and that there was a rising trend for both TE
Working Paper 0402, 2007.
and PTE, suggesting that Islamic banks in GCC countries
improved their efficiency during the survey period. This Banker, R. Charnes, A. Cooper, W “Some models for
was the period where the processes of liberalization of estimating technical and scale inefficiencies in data
the GCC financial system were realised at an accelerated envelopment analysis”. Management Science, 1984.
pace. Overall, we also find that inefficiency in Islamic banks 1078–1092.
is attributed mainly to pure technical inefficiency (29%)
Bauer, P. W. Berger, A. N. Ferrier, G. D. Humphrey, D.
rather than scale inefficiency (17%). Thus, it seems that
B. Consistency conditions for regulatory analysis of
Islamic banks are managerially inefficient in controlling
financial institutions: A comparison of frontier efficiency
their costs and their inputs. It is interesting to note that the
methods, Journal of Economics and Business, 50,
majority of Islamic banks are scale-inefficient and are either
(1998). 85–114.
small- or medium-sized, which implies that these banks
can achieve cost savings and improve their efficiency by Becalli, E. Casu, B. Giraradone, C. Efficiency and stock
increasing their size and scale of operations. Furthermore, performance in European banking, Journal of Business
our findings regarding the impact of size on the efficiency Finance and Accounting, 33, 2006. 245–262.
of Islamic banks suggest that, while large banks are more
Berger, A. N., Humphrey, D. B. Measurement and Efficiency
managerially and technically efficient than small banks,
issues in commercial banking, in Z. Griliches (Ed.)
they are also less scale-efficient than the smaller banks. In
Measurement issues in the service sectors, National
terms of pure technical efficiency, large-sized Islamic banks
Bureau of Economic research, University of Chicago
seem also to be the most efficient ones, followed by the
Press, 1992. 245–279.
medium banks. In this regards, it appears that small banks
need to improve their managerial practices, while medium Berger, A. N. Hannam, T. H. The efficiency cost of market
banks have to increase their scale efficiency. power in the banking industry: A test of the quiet life and

Eds. Hatem A. El-Karanshawy et al. 133

 Srairi et al.

related hypotheses, Review of Economics and Statistics, Fields, J. A. Murphy, N. B. Tirtiroglu, D. An International
80, 1998. 454–465. comparison of scale economies in banking: Evidence
from Turkey, Journal of Financial Services Research, 4,
Berger, A. N. Mester, L. J. Inside the black box: What explains
1993. 157–168.
differences in the efficiencies of financial institutions,
Journal of Banking and Finance, 21, 1997. 895–947. Fries S, Taci A. Cost efficiency of banks in transition: Evidence
from 289 banks in 15 post-communist countries. Journal
Blominvest Bank Islamic banking in the MENA region. 2009.
of Banking and Finance, 29, 2005. 55–81.
Bonin J. P., Hassan I., Wachtel P. Bank performance,
Girardone C, Molyneux P, Gardener E P M. Analyzing
efficiency and ownership in transition countries. Journal
the determinants of bank efficiency: the case of Italian
of Banking and Finance, 29, 2005.31–53.
banks, Applied Economics, 36, 2004. 215–227.
Canhoto, A. Dermine, J. A note on banking efficiency in
Grigorian, D. A. Manole, V. Determinants of commercial
Portugal: New vs. old banks, Journal of Banking and
bank performance in transition: An application of data
Finance, 27, 2003. 2087–2098.
envelopment analysis, Comparative Economic Studies,
Chapra, M. U. The global financial crisis: Can Islamic 48, 2006. 497–522.
finance help, NewHorizon, January-March, 170, 2009.
Hadad, M. Hall, M. Kenjegalieva, K. Santoso, W. Satria,
Charnes, A. Cooper, W. Lewin, A. Seiford, L. Data R. Simper, R. Bank efficiency and stock Market
envelopment analysis: Theory, methodology, and performance: An analysis of listed Indonesian banks,
application, Kluwer Academic Publishers, Boston, 1994. Review Quantitative Finance and Accounting,
DOI.10.1007/s11156–010–0192–1, 2010.
Charnes, A. Cooper, W. Rhodes, E. Measuring the
efficiency of decision making units, European Journal Hassan, K. M. The X-efficiency of Islamic banks, Islamic
of Operational Research, 2, 1978. 429–444. Economic Studies, 2, 2007. 49–77.
Chen, X. Skully, M. Brown, K. Banking efficiency in China: Hasan, M. Dridi, J The effects of the global crises on
Application of DEA to pre- and post-deregulation Islamic and conventional banks: A comparative study,
era: 1993–2000, China Economic Review, 16, 2005. International Monetary fund, Working Paper /10/201,
229–245. 2010.
Christopoulos, D.K., Lolos, S.E.G. Tsionas, E.G. Efficiency Hassan, T. Mohamed, S. Bader M. Efficiency of conventional
of Greek banking system in view of the EMU: A versus Islamic banks: Evidence from the Middle East,
heteroscedastic stochastic frontier approach, Journal International Journal of Islamic and Middle Eastern
of policy model, 24, 2002. 813–829. Finance and Management, 2, 2009. 46–65.
Chu, S. Lim, G. Share performance and profit efficiency Havrylchyk, O. Efficiency of the Polish banking industry:
of banks in an oligopolistic market: Evidence from Foreign versus domestic banks, Journal of Banking and
Singapore, Journal of Multinational Financial Finance, 30, 2006. 1975–1996.
Management, 8, 1998. 155–168.
Inui, T. Park, J. Hyun-Han, S International comparison
Coelli, T. A guide to DEAP version 2.1, A data envelopment of Japanese and Korean banking efficiency: Comments
analysis, (Computer Program), CEPA, Working Paper and discussion, Seoul Journal of Economics, 21, 2008.
96/08, University of New England, Armidale, 1996. 1–16.
Available at: www.une.edu.au/econometrics/cepa.htm.
International Monetary Fund Survey Islamic banks: More
Cooper, W. W. Seiford, L. M. Tone, K. Data envelopment resilient to crisis, Washington, D.C., 2010.
analysis. Kluwer Academic publishers, Boston, 2000.
Ioannidis, C. Molyneux, P. Pasiouras, F. The relationship
Das, A. Ghosh, M. S. Financial Deregulation and efficiency: between bank efficiency and stock returns: Evidence
An empirical analysis of Indian banks during the post from Asia and Latin American, University of Bath,
reform period, Review of Financial Economics, 15, School of Management, Working paper 2008.10, 2008.
2006. 193–221.
Iqbal, Z. Islamic financial systems, Finance and
Drake, L. Hall, M. J. B. Efficiency in Japanese banking, Development, 34, 1997. 42–45.
Journal of Banking and Finance, 27, 2003. 891–917.
Iqbal, Z. Challenges facing Islamic financial industry,
Drake, L. Hall, M. J. B. Simper, R. The impact of Journal of Islamic Economics, Banking and Finance, 3,
macroeconomic and regulatory factors on bank 2007. 1–14.
efficiency: A non-parametric analysis of Hong Kong’s
Iqbal, Z. Van Greuning, H Analyzing banking risk for
banking system, Journal of Banking and Finance, 30,
Islamic banks, World Bank Publications, Washinghton,
2006. 1443–1466.
USA, 2007.
Erdem C, Erdem M.S. Turkish banking efficiency and its
Isik, I. Hassan, K. M. Efficiency, ownership and market
relation to stock performance. Applied Economics
structure, corporate control and governance in Turkish
Letters, 15, 2008. 207–211.
banking Industry, Journal of Business Finance and
Fama, E F. Efficient capital markets: A review of theory and Accounting, 30, 2003. 1363–1421.
empirical work. Journal of Finance, 25, 1970. 383–417.
Kamaruddin B H, Safa M S, Mohd R. Assessing production
Fare, R. Grosskopf, S. Lovell, C.A The measurement of efficiency of Islamic banks and conventional bank
efficiency of production, Kluwer Academic Publishers, Islamic windows in Malaysia. International Business
Boston, 1985. Management Research, 1, 2008. 31–48.

134 Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries

Kasman A, Yildirim C. Cost and profit efficiency in transition Rezvanian, R., Rao, N., Mehdian S. M. Efficiency change,
banking: the case of new EU members. Applied technological progress and productivity growth of
Economics, 38, 2006. 1079–1090. private, public and foreign banks in India: Evidence
from the post-liberalization era, Applied Financial
Kyj, L. Isik, I. Bank x-efficiency in Ukraine: An analysis
Economics, 18, 2008. 701–713.
of service characteristics and ownership, Journal of
Economics and Business, 60, 2008. 369–793. Rosly, SA. Abu Baker, MA. Performance of Islamic and
mainstream banks in Malysia, International Journal
Liadaki, A. Gaganis, C. Efficiency and stock performance
Social Economics, 30, 2003. 1249–1265.
of EU banks: Is there a relationship? Omega, 38, 2010.
254–259. Srairi, S. A comparison of the profitability of Islamic and
conventional banks: The case of GCC countries, Bankers,
Lewis, M (2008). In what ways does Islamic banking differ
Markets, Investors, 98, 2009. 16–27.
from conventional finance, Journal of Islamic Economics
Banking and Finance, 4, 9–22. Srairi, S. Cost and profit efficiency of conventional and
Islamic banks in GCC countries, Journal of Productivity
Mariani, A.M. The input requirements of conventional and
Analysis, 34, 2010. 45–62.
Shariah-compliant banking, The International Journal
of Banking and Finance, 7, 2010. 51–78. Sufian, F. Determinants of bank efficiency during unstable
macroeconomic environment: Empirical evidence
Mohamad, S. Hassan, T. Bader M. Efficiency of conventional
from Malaysia, Research in International Business and
versus Islamic banks: international evidence using
Finance, 23, 2009. 54–77.
stochastic frontier approach, journal of Islamic
Economics, Banking and Finance, 24, 2008. 107–130. Sturm, J. E., Williams B. Foreign bank entry, deregulation
and bank efficiency: Lessons from the Australian
Mokhtar, H. S. Abdullah, N. Alhabshi, S. M. Efficiency
experience, Journal of Banking and Finance, 28, 2004.
and competition of Islamic banking in Malaysia,
1775–1799.
Humanomics, 24, 2008. 28–48.
Sufian, F. Abdul Majid, M.Z. Bank efficiency and share
Moody’s (2008). Islamic banks in the GCC: A comparative
prices in China: Empirical evidence from a three-stage
analysis.
banking model, International Journal of Computational
Olson, D. Zoubi, T. Using accounting ratios to distinguish Economics and Econometrics, 1, 2009. 23–47.
between Islamic and conventional banks in the GCC
Sufian, F. Noor, M.A Abdul Majid, M.Z. The efficiency of
region, The Intermediation Journal of Accounting, 43,
Islamic banks: Empirical evidence from the MENA and
2008. 45–65.
Asian countries Islamic banking sectors, The Middle
Pasiouras, F Estimating the technical and scale efficiency East Business and Economic Review, 20, 20081–19.
of Greek commercial banks: The impact of credit risk,
Xiang, D. Shamsuddin, A. Efficiency and stock market
off-balance sheet activities, and international operation,
performance of Australian banks, paper presented at
Research in International Business and Finance, 22,
the proceedings of the Asian Finance Association 2009
2008. 301–318.
Conference. Brisbane, June 30 to July 3, 2009.
Pasiouras, F. Liadaki, A. Zopounidis, C. Bank efficiency
Zahar, T.S. Hassan, K.H. A comparative literature survey
and share performance: Evidence from Greece, Applied
of Islamic finance and banking, Finance Markets
Financial economics, 18, 2008. 1121–1130.
institutions and instruments, 10, 2001155–199.
Paul, S. Kourouche, K. Regulatory policy and the efficiency
Zaim, O. The effect of financial liberalization of the
of the banking sector in Australia, The Australian
efficiency of Turkish commercial banks, Applied
Economic Review, 41, 2008. 260–271.
Financial Economics, 5, 1995. 257–264.
Rezvanian, R. Mehdian, S. (2002). An examination of cost
structure and production performance of commercial
banks in Singapore, Journal of Banking and Finance,
26, 79–98.

Eds. Hatem A. El-Karanshawy et al. 135

Conventional banks versus Islamic banks:
What makes the difference?
Huseyin Aytug1, Huseyin Ozturk2
Department of Economics, University of California at Santa Cruz
1

[email protected]
2

Abstract - This paper investigates the determinants of banking profitability in the Turkish banking
sector between 2003 and 2011. In addition, we calculate the effect of being an Islamic bank on
banking profitability, which allows us to differentiate conventional and Islamic banks. We introduce
the method of propensity score matching to the banking literature in order to estimate the average
treatment effect (ATE) of being an Islamic bank in Turkey where there exists a dual banking system.
The results show that in terms of return on asset (ROA) and return on equity (ROE), being an Islamic
bank does not create any difference. However, being an Islamic bank turns out to have a significant
and negative effect on net interest margin (NIM). These results have many policy implications in the
Turkish banking industry where Islamic banks mimic others to be one of the leading examples.

Keywords: average treatment effect, propensity score matching, Islamic bank, profitability

JEL Classification: C51, G15

1. Introduction networks promises a lucrative and stable source of funding.

Despite its growing popularity, Islamic banking remains a
Islamic finance is a source of funding that complies with
small part of the total financial sector and will likely remain
Islamic jurisprudence. This source of funding has already
so due to lack of penetration in the market and strong
been in practice in countries where the majority of the
competition that is challenged from the conventional
population is Muslim; however, the importance of Islamic
banking system.
finance has prevailed in the global financial system recently.
Although the distinction between Islamic and conventional
The most explicit distinction of Islamic finance from
finance needs deeper understanding, the main difference
conventional finance, where the return is based upon
within Islamic finance can be attributed to the Islamic
interest, is the prohibition of riba (interest based lending)
idealism of creating a moral economy where profits come
and gharar (speculation or uncertainty). The underlying
from commerce or real transactions not from money
contracts of Islamic products differ from their conventional
lending or speculative transactions. While conventional
counterparts in structure as well. For example murabaha (a
finance helps direct the flow of capital to investment
sale-based instrument), which is similar to a conventional
opportunities that are supposed to provide the highest
loan, involves the purchase of an asset by a bank and
return in the market, Islamic finance allegedly seeks socio-
its sale to a client at a cost plus a pre-determined profit.
economic optimality. Another principle of Islamic finance
The structure of a sale has important legal implications
is that investment is expected to produce an optimal socio-
according to the Islamic rules, which extensively dictate the
economic outcome in line with Islamic norms.
terms of risk and mutual consent. Other sale-based products
include the financing of commissioned manufacturing or
The globalization has affected the scope and breadth of
construction (istinaa) and the forward-sale (salam), and
Islamic capital markets in a way that its pervasiveness
these types of products require structural differences from
has reached a global scale. Moreover, the financial crisis,
their conventional equivalents. Lease-based instruments
which has recently erupted and hit the global finance and
(ijara) are similar to the traditional leasing with certain
economy severely, has paved the way for other means
distinctions and equity-based financial intermediation,
of financial schemes. Islamic finance, which has stayed
which is known as mudaraba, takes place through profit
relatively resilient, became a focus of the global financial
and loss arrangements (Warde, 2010). Another product
system since then.1 The bulk of Islamic funds staying in
that recently became a regular budget financing instrument,
financial hubs of many Gulf States and in global financial

Cite this chapter as: Aytug H, Ozturk H (2015). Conventional banks versus Islamic banks: What makes the difference?
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Aytug and Ozturk

sukuk, is an asset-based instrument of which tangible

Growth: Growth
Loans

25.51
33.67
45.87
41.84
49.86
43.62
36.32
28.00
35.51
22.91
7.30
assets are specified according to the type of the issuance.

CB
In comparison with conventional debt instruments,
these products do not pledge fixed income or an interest-
based income stream. Yet, it collects returns based on the

Loans

46.11
61.93
30.30
60.20
67.69
51.89
67.95
24.33
38.31
30.15
28.40
collection of lease or sale of certain assets that are specified

IB
beforehand. These types of products are predominantly
used in the Middle East, but their prevalence is visible in East
Asian countries as well as developed European markets.

Market
Share:
Loans

95.16
95.94
94.38
96.25
96.62
95.01
95.07
95.82
94.24
94.02
94.26
Islamic finance in Turkey is yet a more recent issue. Changes

CB
in domestic financial systems and public sensibilities have
allowed participation banking to gradually become more
visible. They have also emerged resilient in the context of
two periods of economic turmoil: the domestic financial

Market
Share:
Loans

4.84
3.38
4.06
5.62
4.18
3.75
4.99
4.93
5.76
5.98
5.74
crisis of 2001 and the global financial crisis of 2008. The

IB
severe banking crisis in 2001 did not have as much inverse
effects on participation banks as the conventional banks.
The contagious effect of the 2001 crisis had a limited effect

Growth:
on participation banks due to lack of interbank activities

Equity

18.89
19.94
26.97
21.00
12.96
14.91
22.24
12.42
22.66
25.56

 2.83
CB
by participation banks. The growing financial capacity of
the religiously conservative public has also been another
factor that made participation banks attractive (Hardy,
2012). These types of banks became the sole option for
those people who are resistant to conventional banking.
Growth:
Equity

40.33
50.19
30.47
16.62
48.25
54.38
81.50
25.37
51.83
17.79
26.90
The increasing level of associated client portfolios and

IB
deposits have enabled participation banks to reach over 4.5
percent of market share in total assets from nearly 1 percent
in 2001. Recently a new legislation passed to facilitate
Table 1.  Market share and growth in assets, equity and loans of Islamic and commercial banks.

Islamic banking in the private sector, and government

Market
Share:
Equity

96.95
97.90
96.00
97.63
98.51
98.45
96.11
97.02
95.98
95.98
96.11
officials have indicated interest in issuing sovereign sukuk
CB

(rent certificate), comparable to bonds, for funding central

government budget requirements.

Table  1–2 gives an idea about the market share and the
Market
Share:
Equity

4.00
3.05
1.55
2.10
2.37
1.49
3.89
2.98
4.20
4.02
3.89

growth of conventional and Islamic banks between 2004

IB

and 2011. It is striking that the market share of Islamic

banks in assets, equity and loans has grown from 1% to 5%
approximately and the growth rate of these fundamentals
Growth:

have been higher than the conventional banks. Despite the

Assets

21.90
23.34
20.46
24.80
24.26
15.87
28.42
16.14
23.98
24.68
17.05

fact that the growth rate has declined after the crisis in
CB

2008, it is still higher than its counterparts.

Islamic banking in Turkey followed a parallel path to

the history of the Turkish economy. Turkey had a highly
Growth:
Assets

31.24
39.12
43.65
45.21
47.00
38.68
60.47
31.32
37.18
31.25
25.21

centralized economy, whereby state institutions owned and

IB

managed most important industries until the beginning of

the 1980s (Ozturk et al., 2010). However, the 1980s saw
a period of liberalization of the tightly controlled Turkish
economy. As part of a plan to attract more foreign direct
Market
Share

96.78
97.55
96.01
97.87
98.14
96.84
97.35
96.08
96.52
95.61
95.82

investment from the Arab Gulf states, a decree passed in

CB

1983 legalized the operation of special finance houses to

provide interest-free banking. These institutions were
highly regulated but they did not gain the same status as
conventional banks, e.g., they were not covered by the
Market
Share

3.99
3.22
1.86
2.45
2.65
2.13
3.48
3.16
4.39
4.18
3.92

insurance scheme the other banks in the system utilized and

IB

could not invest in government securities by its nature.2

Following the Asian financial crises in the late 1990s, the

Turkish economy experienced tremendous volatility, which
caused some consolidation of the special finance houses.
2004–2008
2008–2011
2004–2011

Poor regulation, the accumulation of public debt, and

politically driven lending habits contributed to a severe
2004
2007
2006
2005
2009
2008
2011
2010

financial crisis in Turkey in 2001. This affected all strands of

the domestic banking sector, although conventional banks

138 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?

Table 2. Banks’ size and market share (*share of assets/**share of equity/***share of loans).

2011-Q4* 2007-Q4* 2011-Q4** 2007-Q4** 2011-Q4*** 2007-Q4***

Banks IB/CB Overall IB/CB Overall IB/CB   Overall IB/CB Overall IB/CB Overall IB/CB Overall

Akbank 11.51  10.98 12.15  11.75 12.68  12.14 14.43  13.98 10.58   9.97 13.20 12.55
Denizbank  3.10   2.96  2.66  2.57  2.85   2.73  1.98   1.92  3.38   3.18  3.71   3.53
Finans Bank  3.98  3.80 3.72  3.60  4.11   3.94  3.57   3.46  4.56   4.29  5.05   4.81
HSBC  2.08  1.98 2.39  2.31  2.02   1.93  2.75   2.66  2.08   1.96  3.33   3.17
ING Bank  1.81  1.73 2.23  2.16  1.71    1.63  1.73   1.68  2.31   2.18  3.03   2.89
Sekerbank   1.24  1.18 1.08  1.05   1.06    1.01  1.18   1.14  1.28   1.21  1.29   1.23
TEB  3.28  3.13 2.10  2.03   3.04   2.91  1.24   1.20  3.86   3.64  2.45   2.33
TR Ziraat 13.84  13.21 14.42  13.94   9.52  9.11  9.82   9.52 10.75   10.13  7.70   7.32
TR Garanti 12.63  12.05 12.04  11.64 12.70 12.15  9.37   9.07 12.62   11.88 13.27   12.62
TR Halk   7.85  7.49 7.17  6.93   6.24  5.97  5.96   5.78  8.46   7.97  6.46   6.14
TR Is 13.93  13.39 14.29  13.81 12.94 12.39 14.43 13.98 13.79   12.99 12.12   11.52
TR Vakiflar  7.68  7.33 7.56  7.30  6.72   6.43  7.11   6.89  8.63   8.13  8.37   7.96
Yapi Kredi  9.31  8.88 8.97  8.67  8.45   8.09  6.67   6.47 10.20   9.61 10.17   9.66
Albaraka Turk 18.65  0.86 18.99  0.64 16.22   0.69 22.58   0.70 17.71   1.03 18.56   0.91
Bank Asya 30.65  1.41 32.21  1.08 34.51   1.48 36.12   1.13 32.06   1.86 29.74   1.46
Kuveyt Turk 26.57  1.22 19.90  0.67 23.22   0.99 16.44   0.51 25.01   1.45 20.37   1.02
Turkiye Finans 24.12  1.11 28.90  0.97 26.05   1.12 24.86   0.77 25.22   1.47 30.97   1.52

suffered more than the special finance houses since they deposit insurance fund for participation banks as well. In
had a much larger role in the overall economy. Although doing so, the insurance scheme began to cover the whole
the crisis had a major negative effect on the entire banking banking system.
sector, the special finance houses also suffered. Turkish
holding company Ulker purchased Faisal Finans in 2000, These changes may represent a shifting paradigm in the
changing its name to family Finance House. Then, Ihlas level of acceptance for participation banking in Turkey.
Finans filed for bankruptcy in 2001 as the liquidity crisis in Participation banking emerged stronger after each of these
Turkey reached its peak. This showed that the participation periods of instability. Evidently, participation banking’s
banks were not immune to crises even tough they role in the economy will also likely grow as Turkey
functioned with a different business model. The reasons considers options for attracting investors from the Gulf
for this were twofold; the participation banks were not region who are currently highly liquid in terms of capital
decoupled from the whole financial system due to business and also religiously conservative. Therefore, the recent
connections with other banks and the economic crisis developments in Islamic banking bring old debates to
hit the overall economy. Yet, it is worth mentioning here discussion again whether they are really different from the
that the prohibition of holding public debt protected the conventional banks or not.
participation banks from a worse shock than what could
have otherwise ensued. In the initial stages of the financial The current literature on Islamic banking addresses the
crisis, Ihlas and other participation banks, that could issue of whether the distinction between conventional
have held liquid government securities, were thus not as banks and Islamic banks is only their names or the
greatly affected as conventional banks. However, when the distinction also appears in their business model. As a bold
liquidity crisis hit, Ihlas exposure led it to a collapse, and it example, Kassim et al. (2009) questions the argument
experienced a traditional bank run on its deposits. whether Islamic banks are not susceptible to the interest rate
changes as compared to their counterparts where both of
The 2001 crisis led to a rehabilitation of Turkey’s financial them operate in tandem. Their research question emanates
system, and the parliament passed a new law in order to from the basic proposition whether they differentiate in
discipline the overall banking system (Law No. 4389). In behavior to common macro-economic shocks. We aim to
addition to strengthening banking regulations and creating contribute to the current literature by putting Turkey into
new oversight bodies for conventional banks, Law No. 4389 the center and investigating the profitability issue with a
founded the Union of Private Finance Houses in order to focus on Islamic and conventional bank differentiation.
address common issues among participants and provide Although the determinants of profitability in conventional
a level of state control for the sector. All special finance banks in Turkey have been a subject of some research, there
institutions were required to become members of this have been a few studies conducted regarding profitability
association, but they still lacked many of the privileges that of the participation banks in the literature (Macit,
conventional banks had, such as the provision of deposit 2012). Islamic banking in Turkey is under an interesting
insurance. In 2006, banking law No. 5411 officially replaced transformation that is reflected in their asset and equity
the term “special finance institutions” with the name growth. This transformation opens a new debate: what
“participation banking.” Participation Banks Association of are the determinants of this change? Do Islamic banks
Turkey was established and employed with the unification really differ from their counterparts? To what extent do
of Private Finance Houses. The new law created a savings the dynamics associated with the performance of Islamic

Eds. Hatem A. El-Karanshawy et al. 139

 Aytug and Ozturk

banks differs from their counterparts? These questions rates. In other words, Islamic banks are expected to be more
will be addressed in this study. stable than the conventional banks where Islamic banking
is not influenced by interest rates. Stability in demand for
We use a broad set of data to investigate the determinants of money holds some positive effects in terms of efficiency in
bank profitability. The bank specific variables that may have monetary policies and the financial stability in the system.
an impact on bank profitability are selected in accordance On the other hand, Kia and Darrat (2007) refer to two
with the current literature (e.g., Athanasoglou et al. (2008); major reasons why interest-free Islamic banks contribute to
Garca-Herrero et al. (2009); Dietrich and Wanzenried the stability more than the others. A first reason is related
(2011). The macroeconomic and industry specific variables with the demand for money whereas the second one is an
that are considered to be influential on bank efficiency are assessment with the balance sheet perspective. Of these
also included. The recent financial crisis is also considered factors determining demand for money, interest rates
by separating the whole period as two sub-periods: pre crisis appear to be the most effective component for speculation.
and post crisis. This approach is similar to Hasan and Dridi Thus, interest free banking reduces stability in the banking
(2011); Dietrich and Wanzenried (2011), who separate the system since Islamic banks shy away from interest rate.
whole period as pre crisis and post crisis. One of the main As for the balance sheet perspective, along with changes
contributions of this paper is that we introduce the method of in the interest rates, the banks revalue their assets before
propensity score matching to the banking literature in order liabilities. The loan interest rates respond to a change in the
to estimate the average treatment effect (ATE) of being an interest rate much earlier than the savings interest rates.
Islamic bank where there exists a dual banking system. The In this case, the revaluation of balance sheet entries, a key
results are compared with the ordinary least squares (OLS) component for profit maximization, makes the impact of
estimation results. We make use of a unique and up-to-date interest rate changes more vulnerable. However, in the
database by combining quarterly conventional bank and Islamic banking system, there is no need for revaluation of
participation bank data. The results provide insightful policy balance sheet entries because there is no risk of interest rate.
making implications and will be discussed in the upcoming All in all, for these two primary reasons, it is anticipated
sections. All in all in our study seeks to examine a series of that the banking system, dominated by conventional
questions. First, the issue of being an Islamic bank on bank banking, is more unstable. To test for the stability issue in
profitability will be mainly discussed. This issue deserves the Islamic banking system, there are also plenty of studies
particular attention due to motivation discussed above. investigating the country experiences from interest rate
Beyond this main scope, the determinants of profitability of change perspective. All these studies present empirical
the Turkish banking industry will be explored. This is also findings suggesting that demand for money is more stable in
crucial to study considering the limited number of studies banking systems where Islamic banks are the key players.3
specific to Turkish banking. Last but not the least, the effect of
recent financial crisis on profitability of banks will be explored. Since the early works by Short (1979), Kwast and Rose
The recent financial crisis that caused great havoc in the (1982) and Bourke (1989), a considerable amount of recent
banking sector of many countries – e.g., many defaults or bail- studies have investigated some of the major determinants of
outs have taken place very recently especially in developed bank profitability. The empirical studies have focused their
country space – reasonably might affect the profitability. analyses either on cross-country evidence or on the banking
system of individual countries. The studies by Molyneux
The remainder of the study is organized as follows: the and Thornton (1992), Demirguc and Huizinga (1999),
second section will explain the data. The data set we compile Goddard, et al. (2004), Micco, et al. (2005), and Pasiouras
is the largest data set as to the best of our knowledge. The and Kosmidou (2007) investigate a panel data set. Studies
third section will discuss the methodology in some detail. by Berger, et al. (1987), Berger (1995), Michelle (1997),
The ATE methodology and the model specifications will be Bennaceur and Goaied (2008), Athanasoglou, et al. (2008)
discussed in this section. The fourth section will discuss the and Garca-Herrero, et al. (2009) center their analyses on
results and policy implications. The policy implications and single country cases.
policy recommendation will be provided together specific
to the Turkish case. The fifth section will conclude. Bank profitability is vastly measured as a function of return
on asset (ROA) or return on equity (ROE).
2.  Literature review Some works also add net interest margin (NIM) as a
In the studies, which investigate the Islamic banking and complementary measure that it related with profitability.
conventional banking dualism, the recent findings reveal The literature classifies determinants as being internal and
that there is not a fundamental difference in terms of external determinants. The internal determinants include
their routine activities. In other words, they show similar bank-specific variables of which the intrinsic features of
responses to basic impulses, e.g., their profitability measures individual banks compose this bloc. The external variables
respond to market interest rates in a similar way. Ergec reflect factors that are expected to affect the profitability
and Arslan (2011) contend that Islamic banks, relying on of banking industry although banks are not capable of
interest-free banking, shall not be affected by the interest controlling them.
rates; however, in concurrence with the previous studies,
the article finds that the Islamic banks in Turkey are visibly In most studies, variables such as bank size, risk, capital ratio
influenced by interest rates. This study differentiates from and operational efficiency are used as internal determinants
Kassim et al. (2009) in one respect. Kassim et al. (2009), of banking profitability. Pasiouras and Kosmidou (2007)
claim that the primary reason why Islamic banking may find a positive and significant relationship between the size
become more stable compared to conventional banks is and the profitability of a bank. This is due to the fact that
that they are not affected by the fluctuations on interest larger banks are likely to have a higher degree of product

140 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?

and loan diversification than smaller banks. Furthermore, Most studies have shown a positive relationship between
large banks benefit from economies of scale. Berger, et al. inflation, central bank interest rates, GDP growth, and
(1987), provide evidence that costs are reduced only slightly bank profitability (e.g., Bourke (1989), Molyneux and
by increasing the size of a bank and those very large banks Thornton (1992), Demirguc and Huizinga (1999),
often encounter scale inefficiencies. Micco, et al. (2005) find Athanasoglou, et al. (2008), Albertazzi and Gambacorta
no correlation between the relative bank size and the return (2009)). As per the effects of macroeconomic variables, the
on assets for banks, i.e., the coefficient is always positive but effect of inflation rate on bank profitability depends upon
never statistically significant. Therefore the impact of being whether the inflation is anticipated or unanticipated. In the
a big bank in size on profitability is mix. case of an anticipated inflation bank profits may improve
as the banks may adjust the price of lending according to
If it is the risk that is concerned in this literature, Abreu and the inflation rate. However, an unanticipated inflation
Mendes (2002) who examined banks in some European may have negative effects. Bourke (1989) and Molyneux
countries, find that the loans-to-assets ratio, as a proxy for and Thornton (1992) find that a higher inflation rate is
risk, has a positive impact on the profitability of a bank. associated with better profitability indicators. Recently,
Bourke (1989) and Molyneux and Thornton (1992), among Macit (2012) has recorded similar findings.
others, find a negative and significant relationship between
the level of risk and profitability. This result reflects the To measure the effects of market structure on bank
fact that banks that are exposed to high risk also have a profitability, the structure conduct and performance
higher accumulation of non-performing loans, and non- (market-power) hypothesis states that increased market
performing loans lower the returns of the affected banks. power yields monopoly profits. The inverse relation
between the degree of market concentration and degree
Another bank feature that is suggested to be effective on of competition has been the underlying assumption of
the profitability of banks is the asset composition of the the structure conduct performance hypothesis. The bank
banks. Empirical evidence by Bourke (1989), Demirguc concentration is discussed in some studies. The Herfindahl
and Huizinga (1999), Abreu and Mendes (2002), Goddard, Hirschman index was used to proxy the level of competition
et al. (2004), Bennaceur and Goaied (2008), Pasiouras and in the industry (see Dietrich and Wanzenried (2011)).
Kosmidou (2007) and Garca-Herrero, et al. (2009) indicate According to the results of Bourke (1989) and Molyneux
that the best performing banks are those that maintain and Thornton (1992), the bank concentration ratio shows
a high level of equity relative to their assets. The authors a positive and statistically significant relationship with the
explain this relation with the observation that banks with profitability of a bank and is, therefore, consistent with the
higher capital ratios tend to face lower costs of funding structure conduct performance paradigm. In contrast, the
due to lower potential bankruptcy costs. I.e., equity has the results of Demirguc and Huizinga (1999) and Staikouras
lowest order to be paid during liquidation. and Wood (2002) indicate a negative but statistically
insignificant relationship between bank concentration and
One more bank-specific variable is the ownership of a bank profits. Likewise, the estimations by Berger (1995)
bank. Private or state owned banks are the most visited and E.C. and P.C. (2003) contradict the structure-conduct
separation in the literature. This separation has many performance hypothesis. As briefly discussed above, the
insightful findings. For instance, the owner-ship structure determinants of bank profitability can be defined with three
plays an important role in explaining banking profitability. blocks of variables briefly discussed above. The literature
Micco, et al. (2007) found that the separation of banks is more or less similar in terms of the data selection. The
as privately owned or state-owned provides insights in variation of data employed in the analysis is rather sparse.
examining bank performance. According to their results, Therefore the existing literature provides a comprehensive
state-owned banks operating in developing countries tend examination of the effects of bank-specific, industry-specific,
to have a lower profitability, lower margins, and higher and macroeconomic determinants on bank profitability.
than comparable privately owned banks. In industrialized In this study, we take being an Islamic bank as the centre
countries, however, this relationship is found to be much and investigate the effect of being an Islamic bank on
weaker. Iannotta, et al. (2007) point out that government profitability. This contribution is vital in investigating the
owned banks exhibit a lower profitability than privately on-going transformation in the banking sector of Turkey.
owned banks. Therefore, bank ownership structure is also Considering the lack of studies on banking profitability in
important regardless of being developed or developing, the case of Turkey, especially the special attention given to
yet the analysis suggests a different degree of significance. Islamic banks, this study bridges an important gap. Dietrich
Another strand of bank ownership is being a foreign and Wanzenried (2011), underlines the relative scarcity of
bank. The international connection of a bank may have literature that discusses the effect of the recent financial
a significant impact on profitability. This is plausible in crisis on bank profitability. In our study, we also address
a sense that foreign banks have easy access to a pool of this issue by making pre-crisis and post-crisis analysis. The
funds abroad that domestic banks cannot easily reach. data employed in our analysis is in line with the current
Yet in terms of profitability issues, being a foreign bank is literature with small variations that will be discussed in the
found to have differing impact on profitability. Demirguc next section.
and Huizinga (1999) suggest a significant relationship, on
the other hand Bourke (1989) and Molyneux and Thornton
(1992) find this relationship insignificant. 3. Data
There is a well-established set of determinants available
Many studies in profitability literature take—such as to investigate the profitability of the banking system in
central bank interest rate, inflation, GDP growth etc.— Turkey. To examine the effect of being an Islamic bank
another bloc of variables that affect bank profitability. on banking profitability, we rely on our data set in line

Eds. Hatem A. El-Karanshawy et al. 141

 Aytug and Ozturk

with current literature. The data is gathered from the growth, level of foreign exchange rates, consumer inflation,
quarterly unconsolidated balance sheets of banks that and real interest rate. Turkey has experienced great growth
operated between 1994Q1 and 2011Q4. The balance during the period subject to analysis. Therefore, GDP
sheets are obtained from The Banks Association of growth is expected to have a positive impact on profitability
Turkey and Banking Regulation and Supervision Agency. regardless of being conventional or Islamic.
There are 29 conventional banks and 4 Islamic banks in
1994Q1–2011Q4. Macroeconomic variables are from Industry-wise we do not include concentration measures
Central Bank of Turkey and the Undersecretariat of in our analysis. The Turkish banking industry constitutes
Treasury who are responsible for economy management a competitive market without a dominant group of
in Turkey. As per the effects of macroeconomic variables, banks or single bank. Dietrich and Wanzenried (2011)
the real interest rate on government bonds used to proxy take concentration into their analysis. Yet, their focus
interest rate. The daily interest rate of government bonds is is on the Swiss banking industry, where there exists a
not available. Therefore we use one-year T-bill rates at the huge concentration. As per industry specific variables
data issue. These will proxy for the interest rates for each we use growth measures. The reason why we employ
and every quarter. The level of foreign exchange rate is the growth measures is that the growth in the banking
USD/TRY rate and an increase in exchange rate implies industry was visible in the Turkish case in the last decade.
depreciation in Turkish Lira. The recapitalisation was the major theme of the banking
industry (see Yeldan, 2007).
Islamic banks constitute a small portion of the banking
system in Turkey but potential to grow, may be in size rather We define state bank as the base and present three dummies
than number. Of these four banks, two of them are open to for private bank, state banks and Islamic banks. In doing so,
public and are daily traded in the stock market. Three of we aim at controlling for industry specific effects on bank
these four banks are foreign and only one bank is domestic. profitability. The variables that are used in our analysis are
In terms of bank specific determinants, we look at various detailed in Table 3.
different variables, namely the ratio of equity to total assets,
the ratio of net loans to total assets, log of real assets, and the Table  4 provides descriptive statistics for our sample
ratio of non-performing loans to total loans. To control for showing the observations, means and standard deviations
macroeconomic determinants of profitability, we use GDP of all variables. Observations are divided into two groups

Table 3. Variables used in the Empirical Analysis

Variable Notation Measure

Bank-Specific Variables Return on asset ROA Net income/total assets

Return on equity ROE Net income/total equity
Net interest margin NIM Net interest income/total
assets
Capital adequacy ETA Equity/total assets
Asset quality LTA Loans/total assets
Asset size NPLTA—log (TA) Non-performing loans/total
assets (Natural logarithm of
total assets)
Liquidity LQD Liquid assets/total assets
Deposits DPTA Deposits/total assets
Credit risk CR Loans loss provision/loans
Liquidity risk LR Loans/deposits
Market-Specific Variables Total asset growth TAG Growth of total assets
Total equity growth TEG Growth of total equity
Total loans growth TLG Growth of total loans
Macroeconomic Variables GDP growth GDPGR Growth of GDP
Inflation rate INF Inflation rate
Exchange rate FX Exchange rate between
TL and $
Real interest rate IR Real interest rate
Dummy Variables Islamic bank dummy IB 1=Islamic banks,
0=Normal banks
Ownership dummy PD 1=Privately owned;
0=Publicly owned
FD 1=Foreign owner;
0=Domestic owner
Crisis dummy CD 1=After the crisis;
0=Before the crisis

142 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?

Table 4. Descriptive Statistics

Conventional Banks Islamic Banks

Variable Obs Mean Std. Dev. Min Max Obs Mean Std. Dev. Min Max

ROA 1044 0.013 0.027 –0.176  0.322 144 0.007 0.003 0.001 0.024
ROE 1044 0.074 0.128 –1.786  0.465 144 0.058 0.033 0.007 0.236
NIM 1044 0.030 0.028 –0.0217  0.169 144 0.017 0.009 0.005 0.063
ETA 1044 0.174 0.140  0.037  0.916 144 0.118 0.024 0.073 0.183
LTA 1044 0.413 0.212  0.001  0.847 144 0.694 0.089 0.436 0.835
NPLTA 1044 0.023 0.022  0.000  0.177 144 0.036 0.022 0.005 0.135
LOGTA 1044 6.574 0.890  4.346  8.231 144 6.569 0.391 5.664 7.236
LQD 1044 0.217 0.182  0.009  0.812 144 0.080 0.050 0.016 0.235
DPTA 1044 0.513 0.235  0.000  0.903 144 0.786 0.068 0.322 0.880
CR 1044 0.029 0.146  0.000  3.095 144 0.008 0.002 0.001 0.012
LR 1044 1.138 1.883  0.000 34.532 144 0.897 0.186 0.518 2.012

as conventional banks with 1044 observations and Islamic by calculating the propensity score that is defined as the
banks with 144 observations. Simple inspection of the table conditional probability of receiving a treatment despite the
shows that conventional banks are likely to perform better unavailability of experimental data. In the present context,
in profitability since the means of return on asset, return on while Islamic banks constitute the treatment group,
equity and net interest margin are higher. For conventional conventional banks constitute the control group, and being
banks, ROA is 1.3%, ROE is 7.4% and NIM is 3% on average. an Islamic bank is defined as receiving the treatment.
For Islamic banks, ROA is 0.7%, ROE is 5.8% and NIM is 1.7%.
On the other hand, credit risk and liquidity risk are lower for IBjt is defined as a dummy variable whether bank j is an
Islamic banks as one may expect this result because of the Islamic bank at time t. While, y 1jt denotes the profitability
risk-sharing principle. While credit risk and liquidity risk are of bank j that is an Islamic bank at time, y 0jt denotes the
2.9% and 114% for conventional banks, 0.8% and 89% for profitability of bank j at time t that is a conventional bank.
Islamic banks. Another issue that is worth underlining is the The average treatment effect of being an Islamic bank is
total cumulative asset growth, which is higher for Islamic defined as:
banks than conventional banks, being 31.28% and 25.64%
respectively. Correlation matrix for independent variables τ jt = y 1jt − y 0jt ,  (1)
is presented in Table  5. Correlations among most of the
variables are quite low, signalling multi-collinearity does not
create significant bias in any of our analyses. ( )
If both states of the world, y 1jt and y 0jt , were ( )
observable, the average treatment effect would be
estimated with no trouble. Nevertheless, due to
unobservability, the average treatment effect (τ) would be
4. Methodology equal to the difference of mean outcomes ( y 1 − y 0 ) . Since
In the banking literature, it is quite common to use ordin­ary
least square (OLS) methods. The recent studies also employ
( ) ( )
either of states of world, y 1jt and y 0jt , are observable,
we use propensity score matching to calculate the average
a generalized method of moments (GMM) since the profit treatment effect. As argued in Rosenbaum and Rubin
persistence is the common feature of banking data. The (1985), a vector of covariates, Z, can be used to compare
main focus of our study is to observe the differing behavior Islamic and conventional banks. Z is defined as:
of Islamic banking in profitability. To observe the differing
behavior, the sample can be split into two sub-samples, or y 1jt , y 0jt ⊥ IB|Z Pr ( IB = 1|Z ) ∈(0, 1) ,  (2)
pooled estimation can be done through assigning a dummy
variable to examine the “being an Islamic bank” effect.
However, one of the main problems in such analysis is the where ⊥ denotes independence. However it is not possible
selection bias or non-random selection that may have been to find observations with identical values for all covariates
produced by OLS or GMM estimators. Specifying some in Z. In order to eliminate this problem, they suggest using
certain banks as Islamic banks and investigating the effect propensity score matching, which uses probability of bank
of being an Islamic bank is a self-selection choice, therefore pairs that receive the treatment (IB) on the characteristics
it is possible to have self-selection bias. of the pair. I estimate the probability of a bank-pair that
receives the treatment (IB) using the following logit
In order to overcome this possibility, we introduce the specifications.
meth­od of propensity score matching developed by
Rosenbaum and Rubin (1985) to the banking literature
( )
p IB jt = 1 = F ( BSF , MSF , MAC ) ,  (3)
in order to estimate the “average treatment effect” (ATE)
of being an Islamic bank in Turkey where there exists where BSF is a vector of bank-specific variables includes
a dual banking system. The purpose of this method is capital adequacy, asset quality, credit risk, etc. MSF is a
to create a control group that is similar to a treatment vector of market specifics variables, including total asset
group, and the similarity among banks will be assessed growth, total loans growth and total equity growth, and

Eds. Hatem A. El-Karanshawy et al. 143

144
Table 5. Correlation matrix

ROA ROE NIM ETA LTA NPLTA LOGTA LQD DPTA CR LR IB PD FD TAGR TEGR TLGR GDPGR INF FX IR

ROA 1.000
ROE 0.643 1.000
NIM 0.246 0.167 1.000
ETA 0.420 0.016 0.226 1.000
LTA −0.189 0.046 0.037 0.464 1.000
NPLTA 0.174 -0.003 0.248 0.376 0.024 1.000
LOGTA 0.016 0.265 0.000 0.421 0.379 −0.033 1.000
LQD 0.136 0.078 0.113 0.424 0.549 0.019 −0.507 1.000
DPTA −0.254 0.006 0.081 −0.518 0.581 0.000 0.357 −0.409 1.000
CR 0.131 0.024 0.074 0.424 0.258 0.137 −0.113 0.199 −0.197 1.000
LR 0.066 0.136 0.004 0.062 0.020 −0.086 −0.081 0.036 −0.375 −0.066 1.000
IB 0.084 0.043 0.163 0.137 0.415 0.192 −0.002 −0.253 0.373 −0.051 −0.045 1.000
PD 0.024 0.200 0.000 0.162 0.118 −0.044 −0.421 0.161 0.055 0.042 −0.082 0.117 1.0000
FD 0.026 0.098 0.033 0.077 −0.164 −0.091 −0.490 0.366 −0.185 −0.071 0.087 −0.011 0.326 1.0000
TAGR 0.029 0.035 0.030 −0.063 0.141 −0.006 0.007 −0.090 0.127 −0.026 −0.008 0.281 0.033 −0.003 1.0000
TEGR 0.081 0.104 0.115 0.006 0.050 0.036 −0.036 −0.062 0.081 −0.024 −0.018 0.207 0.024 −0.002 0.111 1.000
TLGR −0.027 0.034 0.059 −0.067 0.050 −0.042 −0.038 −0.094 0.058 −0.021 0.010 0.142 0.017 −0.002 0.770 −0.182 1.000
GDPGR −0.076 0.077 0.108 −0.043 −0.065 −0.019 −0.050 −0.051 −0.006 0.033 −0.004 0.000 0.000 0.000 0.184 −0.117 0.501 1.000
INF −0.017 0.013 0.017 0.006 −0.256 0.068 −0.199 0.006 −0.026 −0.011 −0.042 0.000 0.000 0.000 −0.308 0.052 −0.288 0.015 1.0000
FX −0.050 0.034 0.009 0.001 0.022 0.066 0.098 0.069 0.014 0.051 −0.048 0.000 0.000 0.000 −0.219 −0.103 −0.359 −0.124 0.048 1.000
IR −0.028 0.034 0.019 −0.027 −0.252 0.005 −0.220 −0.014 −0.030 −0.013 −0.022 0.000 0.000 0.000 −0.295 0.017 −0.242 −0.028 0.824 −0.105 1.000

Islamic banking and finance – Essays on corporate finance, efficiency and product development
Aytug and Ozturk
Conventional banks versus Islamic banks: What makes the difference?

MAC is a vector of macroeconomic variables, including GDP Tables  6,7 and 8 present regression results for three
growth, inflation rate, foreign exchange rate between the different dependent variables (ROA, ROE and NIM).
Turkish lira and the US dollar and real interest rate. The logit In our analysis, the coefficients are estimated for the
model is used here to identify the effect of Islamic banking entire, pre-crisis and post-crisis periods. For each time
on banking profitability. By using the logit model we can period, we estimate three regression models while the
compare banking profitability of conventional and Islamic first one only includes bank specific factors, the second
banks since both types of banks are similar in terms of their one includes both bank and market specific factors, and
propensity scores. Next step is to use a matching technique in the third one includes bank-specific, market-specific and
order to estimate missing counterfactuals by using obtained macroeconomics factors. Table  6 depicts the regression
(
propensity scores, y 1jt , y tjt
1
) 0
( 0
)
or y jt , y tjt . There are results when ROA is the dependent variable. According
several methods used as matching technique but we use to Table  6, the determinants of profitability vary over
three of them in our analysis, Nearest-Neighbor Matching, the periods, and being an Islamic bank does not have a
Stratification Matching and Kernel Matching. Thanks to statistically significant effect on ROA except in the third
these techniques we will be able to observe whether different model for the full sample. Capital adequacy, asset size,
matching algorithms result in different treatment effects. credit risk, liquidity risk, total asset growth, total equity
The average treatment effect of Islamic banking is given by growth, foreign exchange rate have statistically significant
effects on ROA. The results are in line with the recent
τ TT = E {E [ y 1|IB = 1, z ] − E [ y 0|IB = 0, z ]} ,  (4) literature (Athanasoglou et al., 2008).

Table  7 shows the results when banking profitability is
measured as return on equity (ROE). It is obvious that
5. Results the determinants of profitability do not vary over periods
To test the relationship between bank profitability and as much as they do for ROA. However for ROE, being an
the bank specific, market specific and macroeconomic Islamic bank never plays a significant role, even though
determinants described earlier, we first estimate a linear the coefficients are higher compared to the coefficients in
regression model in the following form: Table 6. On the other hand, being a private bank is one of
the determinants of profitability over periods. Interestingly,
liquidity risk is an important factor for ROE while it is not
ROA = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it  (5) for ROA.

ROE = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it  (6) Table  8 treats NIM as a dependent variable and presents
regression results. The determinants of NIM and ROE are
very similar as they are revealed. ETA, LTA and NPLTA
NIM = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it  (7) are positively correlated with NIM and ROE. The most

interesting result is being an Islamic bank is negatively
Table 6. Regression results: ROA is the measure of bank profitability

Dependent Variable
ROA Full sample (1994–2011) Before the crisis (1994–2007) After the crisis (2008–2011)
Independent
Variables Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

IB −0.0053 −0.0081 −0.0091  0.0098 −0.0139 −0.0140 −0.0021 −0.0036 −0.0033

ETA  0.0661  0.0631 0.068  0.0759  0.0709  0.0674  0.0898  0.0916  0.0934
LTA 0.014  0.0143  0.0187  0.0316  0.0325  0.0312 −0.0076 −0.0022  0.0000
NPLTA −0.0329 −0.0250 −0.0161  0.0025  0.0170  0.0128 −0.0673 −0.1036 −0.1071
LOGTA  0.0028  0.0027 0.006  0.0053  0.0042  0.0026  0.0071  0.0079  0.0097
LQD  0.0043  0.0044  0.0071  0.0164  0.0155  0.0122  0.0043  0.0066  0.0101
DPTA −0.0016 −0.0028 −0.0035  0.0082  0.0063  0.0056  0.0075  0.0059  0.0051
CR −0.0235 −0.0229 −0.0211 −0.0059 −0.0053 −0.0040 −0.0214 −0.0189 −0.0162
LR  0.0003 −0.0003 −0.0004 −0.0004 −0.0005 −0.0005  0.0027  0.0028  0.0031
PD −0.0045 −0.0042 −0.0024 −0.0069 −0.0073 −0.0082  0.0007  0.0011  0.0023
FD  0.0003  0.0002  0.0025  0.0033  0.0026  0.0018  0.0001  0.0008  0.0018
TAGR  0.0579  0.0533  0.1075  0.0826  0.0642  0.1019
TEGR  0.0314  0.0318  0.0141  0.0118  0.0238  0.0098
TIGR −0.0209 −0.0006 −0.0558 −0.0188 −0.0879 −0.1709
GDPGR −0.0002 −0.0016  0.0002
INF  0.0003  0.0003  0.0003
FX −0.0086  0.0284 −0.0155
IF  0.0001 −0.0004 −0.0001
R-SQ 0.18 0.18 0.19 0.14 0.15 0.16 0.39 0.43 0.45
PROB>CHI2 (F test) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Eds. Hatem A. El-Karanshawy et al. 145

 Aytug and Ozturk

Table 7. Regression results: ROE is the measure of bank profitability

Dependent Variable
ROA Full sample (1994–2011) Before the crisis (1994–2007) After the crisis (2008–2011)

Independent
Variables Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

IB −0.0116 −0.0256 −0.0306 −0.0023 −0.0177 −0.0220 −0.0170 −0.0267 −0.0260

ETA 0.1389 0.1204 0.1424 0.2284 0.2026 0.1702 0.1966 0.2042 0.2298
LTA 0.062 0.0643 0.0877 0.11778 0.1244 0.1119 −0.1393 −0.1109 0.0999
NPLTA −0.5497 −0.5206 −0.5001 0.486 −0.4218 −0.4526 −0.1330 −0.3459 −0.2325
LOGTA 0.0233 0.0228 0.0400 0.0422 0.038 0.0298 0.0623 0.0694 0.0926
LQD 0.0070 0.0068 0.0188 0.0130 0.0065 −0.0181 0.0248 0.0354 0.05655
DPTA −0.0005 −0.0072 −0.0113 0.0137 −0.0238 −0.0301 0.1387 0.1318 0.1216
CR −0.0509 −0.047 −0.0369 −0.0120 −0.0090 −0.0015 −0.0157 −0.0057 0.0169
LR 0.0075 −0.0075 −0.0077 −0.0105 −0.0107 −0.0115 0.0452 0.0458 0.0465
PD −0.0806 −0.0794 −0.0694 −0.0894 −0.0907 −0.0936 0.0004 0.0057 0.0206
FD 0.0104 0.0099 0.0220 0.0354 0.0331 0.0297 0.0028 0.0083 0.0225
TAGR 0.2695 0.2350 0.5297 0.3850 0.2357 0.3929
TEGR 0.1898 0.1901 0.0979 0.1059 0.1601 0.0953
TIGR −0.1402 −0.0217 −0.3472 −0.0486 −0.3472 −0.647
GDPGR −0.0011 −0.0099 0.0003
INF 0.0030 0.0013 0.0061
FX −0.0452 0.0421 −0.0718
IF 0.0001 −0.0008 −0.0010
R-SQ 0.12 0.13 0.15 0.14 0.15 0.17 0.23 0.26 0.26
PROB>CHI2 (F test) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Table 8. Regression results: NIM is the measure of bank profitability

Dependent Variable
ROA Full sample (1994–2011) Before the crisis (1994–2007) After the crisis (2008–2011)

Independent
Variables Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

IB −0.019 −0.0238 −0.0244 −0.0110 −0.0176 −0.0185 −0.0229 −0.0272 −0.0272

ETA 0.0263 0.0197 0.0229 0.0383 0.0293 0.0226 0.0203 0.0239 0.0239
LTA 0.0165 0.0165 0.0182 0.0136 0.0139 0.0056 0.0233 0.0355 0.0385
NPLTA 0.1572 0.1651 0.1754 0.0977 0.1258 0.1313 0.3258 0.2664 0.2905
LOGTA −0.0011 −0.0017 0.0002 0.0033 0.0017 −0.0026 0.000 0.0012 0.0033
LQD 0.0167 0.0157 0.0168 0.0211 0.0203 0.0134 0.0231 0.0259 0.0315
DPTA 0.0084 −0.0062 0.0059 0.0026 −0.0007 −0.0014 0.0025 −0.0003 −0.0012
CR −0.0163 −0.0153 −0.0142 −0.0077 −0.0065 −0.0059 −0.0076 −0.0025 0.0019
LR −0.0004 −0.0005 0.0005 −0.0006 −0.0007 −0.0008 0.0032 0.0033 0.0038
PD −0.0021 −0.0020 −0.0008 −0.0004 −0.0001 −0.0017 0.0003 0.0005 0.0018
FD −0.0027 −0.0032 −0.0018 0.0045 0.0056 −0.0081 0.0024 0.0038 0.0050
TAGR 0.1326 0.1123 0.1311 0.1098 0.1681 0.3036
TEGR 0.0472 0.0520 0.0401 0.0453 0.0447 0.0249
TIGR −0.0685 −0.0329 −0.0510 −0.0128 −0.185 −0.4159
GDPGR −0.0003 −0.0025 0.0008
INF 0.0001 0.0009 0.0023
FX −0.0006 0.0289 −0.0190
IF 0.0002 −0.0003 −0.0001
R-SQ 0.11 0.14 0.17 0.13 0.15 0.14 0.22 0.30 0.35
PROB>CHI2 (F test) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

146 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?

correlated with NIM and it is always significant at a 99% diminishing bank profitability. However the effect on NIM
level for the full sample and after the crisis. Nevertheless, reveals a different picture since the coefficients are always
being an Islamic bank lowers the profitability when it positive and significant.
is measured as net interest margin (NIM). The effect of
Islamic banks was pretty limited or not significant before The sector specific variables, the asset and equity growth
the crisis. positively affect profitability in Turkey. In aggregate terms,
as the banking system as a whole scales up, the asset and
Capital adequacy (ETA) has a positive and significant equity size, individual banks tend to be more profitable.
impact on ROA and ROE in all periods with different This also implies that asset and equity growth is distributed
models. The findings imply that banks that have higher evenly among banks. Therefore, no specific bank or
capital adequacy are more profitable. This may be related banking group dominated the others. In contrary to this,
to the fact that banks with higher capital adequacy tend total loans growth has a negative and significant effect on
to be more credible and operate with lower costs. The profitability.
specific finding for the Turkish case constitutes a different
dimension in explaining the solvency risk (capital In both estimated regressions where ROA and ROE is the
adequacy) on profitability. Pervan, et al. (2012) find that dependent variable, asset size is found to have a significant
capital adequacy ratio is negatively related with return in and positive impact on profitability. Hauner and Peiris
their analysis. They conclude that higher capital adequacy (2005) suggest two potential explanations for this impact.
implies lower profitability. They point out that a higher First, if it relates to market power, large banks should pay
level of bank capital provides safety and over-caution in less for their inputs, i.e., lower cost as mentioned in capital
the banking business and it reduces profitability in the adequacy case. Second, there may be increasing returns to
Macedonian banking system. Likewise, our results show scale gains through the allocation of fixed costs.
that capital adequacy doesn’t have a significant effect on
NIM after the crisis. While it has a positive and significant The deposit to total assets ratio has a negative but
effect on NIM for the full sample, the relationship between insignificant impact on bank profitability. This might be
capital adequacy and NIM is blurry before and after the an indication of the fact that the link between deposit and
crisis with different models. lending is not efficiently operated. Turkish banking systems
have suffered from the short term character of its deposit
In terms of liquidity risk, the findings are quite specific to base. The maturity of deposit is mainly short and cannot
the Turkish case and are not in accordance with studies in be efficiently converted into lending, i.e., higher income
the literature (Pervan et al., 2012). While liquidity risk has earnings.
a positive and significant effect on ROA and ROE only after
the crisis, it has a negative effect on ROE before the crisis Our results regarding the impact of ownership on
and for the full sample. The banks with less liquidity risk profitability support the findings of Micco, et al. (2007)
are deemed to be more profitable, since the banks that have and Iannotta, et al. (2007), who point out that state banks
higher loan to deposit ratios are expected to produce more exhibit a lower profitability than privately owned banks.
returns due to interest revenues. However, the Turkish The case also holds for foreign banks. Foreign banks are
case proves to be different than the theory, and Turkish also more profitable than state banks. The findings shed
banks are more tempted to invest in government assets light on the inefficiency of state banks in the past where
that proposed higher yields (Aysan and Ceyhan, 2007). state banks were mandated as the lender of unprofitable
Therefore, the liquidity risk implies a positive relationship and politically driven projects. For instance, the duty losses
between liquidity and profitability as it is in some extent that were one of the main causes of the 2001 banking
not in line with recent findings. crisis were an indication of how state banks operated with
political bias. One of the main purposes of this paper is to
In the model where ROA is dependent, it is interesting to find the average treatment effect (ATE) of being an Islamic
note that the coefficient of non-performing loans to asset bank. Regression results give us some insights about how
reveals a negative relationship with bank profitability the ATE might look for different measures of profitability.
and is statistically significant only after the crisis. The According to the regression results Islamic banks play an
estimated coefficients are negative for all the samples. insignificant role when the measure is ROA and ROE. In
When the ROE is a dependent variable, the coefficients addition to this, Islamic banks lower the profitability when
are still negative, but they are statistically significant only the measure is NIM.
for the full sample. The empirical finding is in contrast
with the skimping hypothesis of Berger and DeYoung Estimates of treatment effects are presented in Table 9. As
(1997). Berger and DeYoung (1997) suggest that under we mentioned above we use the different algorithms to
the skimping hypothesis, a bank maximizing the long calculate the treatment effect for the sake of robustness.
run profits may rationally choose to have lower costs First, the treatment effect proves that there is no relationship
in the short run by skimping on the resources devoted between Islamic banks and ROA. The coefficient is always
to underwriting and monitoring loans but bear the zero. Second, Islamic banks are positively correlated
consequences of greater loan performance problems. with ROE and the magnitude is higher. However, OLS
Therefore, cost minimization in the short run may coefficients are insignificant. Third, the ATE on NIM
not bring long-term profitability. Yet, the findings are varies over periods. It is always positive before the crisis
plausible in Turkey’s case where the non-performing and negative after crisis. For the entire period, different
loans are low. Therefore, holding more funds kept for algorithms prove the negative relationship between Islamic
potential losses in the future increases banks cost, thus banks and NIM.

Eds. Hatem A. El-Karanshawy et al. 147

 Aytug and Ozturk

Table 9. Estimates of treatment effect

1994–2011 1994–2007 2008–2011

RE NM S OLS NM S OLS NM S
ROA Islamic Banking −0.005 0 0 −0.014 0 0 0.005 0.001 0.001
No. Obs 1188 160 1080 660 84 600 528 68 478
Treated 144 144 144 80 80 4 64 64 62
Controls 1044 16 936 580 4 596 464 4 416
1994–2011 1994–2007 2008–2011

RE NM S OLS NM S OLS NM S
ROE Islamic Banking −0.010 0.004 0.003 −0.006 −0.007 0.012 0.017 0.015 0.013
No. Obs 1188 160 1080 660 84 600 528 68 478
Treated 144 144 144 80 80 4 64 64 62
Controls 1044 16 936 580 4 596 464 4 416
1994–2011 1994–2007 2008–2011

RE NM S OLS NM S OLS NM S
NIM Islamic Banking −0.023 −0.003 −0.004 −0.018 0.001 0.011 −0.023 −0.005 −0.005
No. Obs 1188 160 1080 660 84 600 528 68 478
Treated 144 144 144 80 80 4 64 64 62
Controls 1044 16 936 580 4 596 464 4 416

6. Conclusion investigated. Second, the results of the Turkish case have

to be compared with other country cases by identifying
This paper provided evidence on the implications of Islamic
more proper econometric methodologies as introduced
banking in Turkey. As a baseline analysis and in parallel
in this paper. The issue of “self-selection bias” needs to
with the literature, the determinants of profitability in
be tested in other country cases as well. As a continuation
Turkish banking have been found to be familiar with the
of the concurrent problem of “self-selection bias,” the
previous studies. However, our main difference comes
profitability of Islamic banks should be compared with the
with the role of being an Islamic bank. In our study
conventional banks in a synthetic control group framework
we investigated the effect of being an Islamic bank on
introduced by Abadie and Gardeazabal (2003).
bank performance in terms of various bank profitability
measures. Specifying some certain banks as Islamic bank
and investigating the effect of being an Islamic bank is a Notes
conscious “self-selection” choice, therefore have a potential
1. See Hasan and Dridi (2010) for differing performance
to have “self-selection bias”. In order to overcome this
of conventional and Islamic banks during and after the
possibility, we introduced the method of propensity score
financial crisis in countries where conventional banks
matching to the banking literature in order to estimate the
and Islamic banks jointly operate.
average treatment effect (ATE) of being an Islamic bank in
2. After the decree, initial institutions were Bahrain-
Turkey where there exists a dual banking system. In order
based Al Baraka Turk and Saud-based Faisal Finans
to estimate the average treatment effect propensity score
Kurumu, which each opened subsidiaries in 1985. The
matching has been conducted. The results are compared
Kuwaiti-based Kuveyt Turk Kurumu began operations
with the OLS results. According to the OLS results, Islamic
in 1989. Eventually, special finance houses began
banks have a negative relationship with NIM and a positive
lending with mainly domestic capital, including:
relationship with ROA. The effect of being an Islamic bank
Anadolu Finans (1991), Ihlas Finans (1995) and Asya
on ROE is insignificant on the other hand. Propensity
Finans (1996).
score matching technique confirms some of OLS results.
3. See Darrat (1988), Zuberi (1992), Darrat and
The ATE of being an Islamic bank, on ROA and ROE, are
Salaaming (1990), Kia and Darrat (2007).
positive, yet it is blurry but negative on NIM. The results
provided evidence of increasing performance of Islamic
banks in the Turkish banking system. Therefore, although References
not in total asset size – these banks signal the efficacy of
Abadie A, Gardeazabal J. (2003) “The Economic Costs of
Islamic banks in the near future.
Conflict: A Case Study of the Basque Country. American
Economic Review. 93(1):113–132.
Several potential research themes emerge as well. Firstly,
the blurry results of being an Islamic bank on NIM have to Abreu M, Mendes, V. (2002). Commercial bank interest
be identified in some detail. This discrepancy could provide margins and profitability: evidence from EU countries.
further insights about Islamic banks operations if deeply Working Paper, Porto.

148 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?

Albertazzi U, Gambacorta L. (2009) Bank Profitability Hasan M, Dridi J. (2011) The Effects of the Global Crisis on
and the Business Cycle. Journal of Financial Stability. Islamic and Conventional Banks: A Comparative Study.
5(4):393–409. Journal of International Commerce, Economics and Policy
(JICEP). 2(02):163–200.
Athanasoglou PP, Brissimis SN, Delis MD. (2008)
Bank-Specific, Industry-Specific and Macroeconomic Hauner D, Peiris SJ. (2005) Bank Efficiency and Com­petition
Determinants of Bank Profitability. Journal of in Low-Income Countries: The Case of Uganda. IMF
International Financial Markets, Institutions and Money. Working Papers 05/240. International Monetary Fund.
18(2):121–136.
Iannotta G, Nocera G, Sironi A. (2007) Ownership
Aysan AF, Ceyhan SP. (2007) Market Disciplining Role Structure, Risk and Performance in the European
of Crisis on the Restructuring of the Turkish Banking Banking Industry. Journal of Banking and Finance.
Sector. Technical Report. 31(7):2127–2149.
Bennaceur S, Goaied M. (2008) The Determinant of Kassim S, Majid M, Yusof R. (2009) Impact of Monetary
Commercial Bank Interest Margin and Profitability: Policy Shocks on the Conventional and Islamic Banks in
Evidence from Tunisia. Frontiers in Finance and Dual Banking System: Evidence from Malaysia. Journal
Economics. 5(1):106–130. of Economic Cooperation and Development. 30(1):41–58.
Berger AN. (1995) The Profit-Structure Relationship Kia A, Darrat AF. (2007) Modelling Money Demand Under
in Banking-Tests of Market-Power and Efficient- the Profit-Sharing Banking Scheme: Some Evidence on
Structure Hypotheses. Journal of Money, Credit and Policy Invariance and Long-Run Stability. Global Finance
Banking. 27(2):404–31. Journal. 18(1):104–123.
Berger AN, DeYoung R. (1997) Problem Loans and Cost Kwast ML, Rose JT. (1982) Pricing, Operating Efficiency,
Efficiency in Commercial Banks. Journal of Banking and and Profitability Among Large Commercial Banks.
Finance. 21(6):849–870. Journal of Banking and Finance. 6(2):233–254.
Berger AN, Hanweck GA, Humphrey DB. (1987) Macit F. (2012) Bank Specific and Macroeco­ no­
mic
Competitive Viability in Banking: Scale, Scope and Determinants of Profitability: Evidence From Partici­
Product Mix Economies. Journal of Monetary Economics. pation Banks in Turkey. Economics Bulletin. 32(1):
20(3):501–520. 586–595.
Bourke P. (1989) Concentration and Other Determinants Micco A, Panizza U, Yaez M. (2005) Bank Ownership and
of Bank Profitability in Europe, North America and Performance Does Politics Matter? Working Papers
Australia. Journal of Banking and Finance. 13(1):65–79. Central Bank of Chile 356. Central Bank of Chile.
Darrat A, Salaaming M. (1990) Islamic Banking: An Outline Micco A, Panizza U, Yanez M. (2007) Bank Ownership and
of Some Conceptual and Empirical Aspects. Savings and Performance – Does Politics Matter? Journal of Banking
Development. 19:185–192. & Finance. 31(1):219–241.
Darrat AF. (1988) The Islamic Interest-Free Banking Michelle NC, Wheelock D. (1997) Why Does Bank
System: Some Empirical Evidence. Applied Economics. Performance Vary Across States? Review. (Mar):27–40.
20(3):417–425.
Molyneux P, Thornton J. (1992) Determinants of European
Demirguc A, Huizinga H. (1999) Determinants of Bank Profitability: A Note. Journal of Banking and
Commercial Bank Interest Margins and Profitability: Finance. 16(6):1173–1178.
Some International Evidence. World Bank Economic
Ozturk H, Gultekin-Karakas D, Hisarciklilar M. (2010)
Review. 13(2):379–408.
The Role of Development Banking in Promoting
Dietrich A, Wanzenried G. (2011) Determinants of Bank Industrialization in Turkey. Region et Developpement.
Profitability Before and During the Crisis: Evidence from 32:153–178.
Switzerland. Journal of International Financial Markets,
Pasiouras F, Kosmidou K. (2007) Factors Influencing
Institutions and Money. 21(3):307–327.
the Profitability of Domestic and Foreign Commercial
EC M, PC R. (2003) Determinants of Greek Commercial Banks in the European Union. Research in International
Banks Profitability, 1989–2000. Spoudai. 53(1):84–94. Business and Finance. 21(2):222–237.
Ergec EH, Arslan BG. (2011) Impact of Interest Rates on Pervan M, Curak M, Poposki K. (2012) Industrial Con­
Islamic and Conventional Banks: The Case of Turkey. centration and Bank Performance in an Emerging
Market: Evidence from Croatia. In: Advances in Finance
Garca-Herrero A, Gavil S, Santabrbara D. (2009) What
and Accounting. Tomas Bata University, Czech Republic.
Explains the Low Profitability of Chinese Banks? Journal
of Banking and Finance. 33(11):2080–2092. Rosenbaum PR, Rubin DB. (1985) Constructing a Control
Group Using Multivariate Matched Sampling Methods
Goddard J, Molyneux P, and Wilson JOS. (2004) Dynamics that Incorporate the Propensity Score. The American
of Growth and Profitability in Banking. Journal of Statistician. 39(1):33–38.
Money, Credit and Banking. 36(6):1069–90.
Short BK. (1979) The Relation Between Commercial Bank
Hardy L. (2012) The Evolution of Participation Banking Profit Rates and Banking Concentration in Canada,
in Turkey. Online Journal on Southwest Asia and Islamic Western Europe, and Japan. Journal of Banking and
Civilization. Winter 2012. Finance. 3(3):209–219.

Eds. Hatem A. El-Karanshawy et al. 149

 Aytug and Ozturk

Staikouras C, Wood G. (2002) The Determinants of Yeldan E. (2007) Patterns of Adjustment Under the Age
European Bank Profitability. International Business and of Finance: The Case of Turkey as a Peripheral Agent of
Economics Research Journal. 3:57–68. Neoliberal Globalization. Technical report.
Warde I. (2010) Islamic Finance in the Global Economy. Zuberi HA. (1992). Interest-Free Banking and Economic
2nd Edition. Edinburgh University Press. Edinburgh. Stability. The Pakistan Development Review. 31:1077–1087.

150 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba
time deposits of Islamic banks
Zeynep Topaloglu Calkan
Adjunct Assistant Professor, Georgetown University – Qatar, Phone: +974 5568 9211, [email protected]

Abstract - On the deposit side, Islamic banks work on a mudaraba (partnership) contract, where
the depositor and the bank are business partners. While in conventional banks the depositor is
provided with a fixed interest rate, in Islamic banks the depositor can only discover his return when
the investment period is over. This fundamental distinction brings forth a disadvantage for Islamic
banks while competing with their conventional counterparts in the market.

On the other hand, most of the credits extended by Islamic banks follow a murabaha (cost-plus
sale with deferred repayment) contract, and the banks specify profit rate on the credits from the
beginning. Using this information we have developed a forecast model to quote the depositors their
expected returns on mudaraba time deposits within a 95% confidence interval at the beginning of
the investment term. Besides increasing competitive advantage, estimating expected returns will
assist Islamic banks in their risk management and asset-liability management.

Keywords: Islamic banking, estimating returns on Islamic time deposits

1. Introduction financial instruments place Islamic banks in an unfavorable

position compared to their competitors.
The financial crisis of 2007–2008 triggered an extended
global recession, which still distresses the economic
One of the structural factors that cause difficulties for
activities. This crisis also evoked questions about morality
Islamic bankers is the inability to provide a specific rate of
of the current global financial system. The research after
return to depositors. Depositors are used to knowing the
the crisis shows that financial institutions of the West,
interest rate and are most of the time comparing different
with the help regulators and the rating agencies, deceived
investment alternatives based on their rate of return. This
their clients and the public.1 Islamic finance and its ethical
becomes challenging especially when banks are interested
principles emerged as an alternative model because most
in attracting customers to long-term savings accounts.
of the elements that caused and extended the crisis are not
Depositors of Islamic banks need to wait up until the end
permissible by Shariah rule.2 Although the virtue of the
of their investment period in order to observe their return,
Islamic financial system was invigorated after the crisis,
whereas in conventional banks depositors know how much
the scholars and practitioners of Islamic finance have been
they will receive at the beginning of their investment.
emphasizing its value long before. Considering its merit,
the insignificant size of Islamic financial assets compared
The reason behind this lack of information is the underlying
to global financial assets makes us question what can be
Shariah contract in Islamic savings accounts. Typically two
and should be improved by practitioners of the area of
types of contracts, wadia (safe-keeping) and mudaraba
Islamic finance to reach its deserved levels. Here, in this
(profit-loss sharing), are used in these accounts. Briefly,
paper, we propose a tool to improve competitiveness of
wadia is a contract where the bank acts as a safe keeper
Islamic banks. This is an attempt to facilitate the process
to depositor’s funds with a permission to utilize them at
of reflecting intrinsic value of Islamic finance to statistical
its own risk, and mudaraba is a type of profit-loss sharing
figures.
contract where a depositor invests money into a partnership
with the bank, where the bank provides its labor/services
Except a few countries3 where the entire internal financial
to the partnership in an agreement to share proceedings –
system has been transformed, Islamic banks have to
either profit or loss – later on. In both of these contracts it
compete with their conventional counterparts in attracting
is forbidden to specify a rate of return in the beginning as
customers. Factors like lower penetration4 and fewer

Cite this chapter as: Calkan Z T (2015). Estimating expected returns on Mudaraba time  deposits of Islamic banks.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Calkan

it nullifies the essence of these contracts. Besides Shariah No.  66 defines PSIAs as a new category between liability
prohibition, in reality it is not possible to know the value of and owners’ equity since PSIA investors are not regular
ingenerated profits/losses. depositors receiving a fixed return but they are profit/
loss-sharing partners. On the other hand PSIA investors
Currently, the banks that use these two contracts in their do not have managerial and voting rights as owners of
profit sharing investment accounts (PSIA) do not share the bank. Akacem and Gilliam (2002), Kahf (2005),
a profit rate with their deposit clients. They only share Sultan (2006), Ayub (2007), Ibrahim (2007), Shubber and
historical information, mainly previous term’s profit Alzafiri (2008), in agreement with the definition provided
distribution. In this paper, we argue it is possible for Islamic by AAOIFI, states PSIAs are an equity-like instrument
banks to provide an anterior rate to their clients without compared to conventional deposits. This attribute of the
breaking Shariah rules by sharing already available PSIAs brings challenges in terms of accounting treatment
information in their accounts.5 of them. Atmeh and Ramadan (2012) critically evaluate
accounting treatment of mudaraba returns by AAOIFI, and
Although it is not permitted and possible to specify they explain how some of its implications deteriorate the
the profits in advance, it is still possible to make precise reliability and fairness of the financial statements and how
estimations. The reason behind it lies in the way PSIA funds they compare AAOIFI standards with IFRS in that respect.
are being utilized in the Islamic banks. Scholars argue that In application we observe not all banks stick to the AAOIFI
Islamic banks are initially established in order to promote standards. Rating Agency Malaysia (RAM) in its Research
sharing risk not only on the depositor side but also on the Report (October–December 2007) comparing Malaysian
financing side through profit/loss sharing system. However and Middle Eastern IFIs, points out reporting PSIAs as
due to reasons like mismatch between asset and liability a liability on the balance sheet as one of the differences
duration, moral hazard and adverse selection issues, today of Malaysia.
the main mode of financing in Islamic finance is not one of
the profit/loss sharing contracts (musharaka or mudaraba) Besides accounting treatment, PSIAs equity-like structure
but one of the trade contracts namely murabaha. In brings out different risk implications compared to
murabaha contract the Islamic financial institution (IFI) conventional risk management. Since PSIA investors are
acts as a tradesman. It purchases the items from the vendor supposed to bear the associated loss in these accounts,
and sells them to its client with a mark-up in a deferred capital adequacy calculation, and the approach towards
payment plan. There are several requirements of murabaha interest rate risk and credit risk differs in IFIs. Khan and
contract that have to be satisfied by the bank and the client; Ahmed (2001) provide a detailed analysis of the risk
however, for our research, the significance of this contract management in the Islamic financial industry. Archer
is its feature to provide ex-ante profit rate. Both parties and Abdel Karim (2009) especially call attention to the
know and agree on how much profit will be charged by the regulatory challenges faced in Europe and North America,
bank on the item well before the transaction takes place. where there is no special regulation for IFIs. Ariffin and
Most of the time Islamic banks follow conventional loan Kassim (2011) through a survey on selected banks uncover
rates when quoting their mark-up price since they are in the areas for improvement in the risk management practices
competition with them. within IFIs.

Therefore even if IFIs cannot specify the profit rate that On the other hand, Sundararajan (2005) with his cross-
they will distribute to PSIA holders, on the financing country study argues PSIAs are subject to a considerable
side they know how much money they will make when amount of smoothing on their return, which in turn implies
they sell the murabaha contract. It is possible to use this the investment risks of the banks that are not fully shared
knowledge to generate a reliable estimation about their by the PSIA investors. He challenges whether IFIs really
profit distribution. need a distinct risk treatment compared to conventional
banks.
In the following sections of the paper the research proceeds
as follows: In the second section, the main discussions Second major issue about PSIAs is the return equalization
about profit loss sharing accounts in the literature have activities of IFIs. As pointed out by Sundararajan (2005),
been presented. In the third section the methodology to and Archer et al. (2010), in order to compete in the
estimate PSIA return has been developed, and in the last market and avoid withdrawal risk by providing market-
section the results and conclusion have been shared with related returns to PSIAs, IFIs employ a variety of return
the readers. smoothing activities. Additionally, in Jordan, Malaysia and
Qatar, the central bank requires IFIs to manage PSIAs in a
way not to reflect losses to the investors and to “smooth”
2.  Literature review returns. Therefore due to regulatory requirements or
Profit sharing investment accounts (PSIA) have been a topic market pressure, IFIs are driven to use a combination of
of interest among scholars in the field of Islamic finance. methods like conservative investment strategies, profit
There are mainly two issues, one of which is accounting equalization reserves (PER), (a reserve account formed
treatment and risk management implications of PSIAs, out of profits of PSIA to smooth profits), investment risk
while the other is the profit smoothing applications of reserves (IRR), (another similar account to cover periodic
Islamic Financial Institutions (IFI) against the risk of losing losses) or a donation to PSIA holders from the share of the
PSIA investors, in other words displacement risk. owners. Besides Sundararajan (2005), many scholars like
Zuobi and Al-Khazali (2007), Taktak et al. (2010), Farook
Accounting and Auditing Organization for Islamic Financial et al. (2012) confirm that IFIs pursue income-smoothing
Institutions (AAOIFI) in its Financial Accounting Standard activities to compete with market-based deposit interest

152 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks

rates. Taktak et al. (2010) demonstrate unlike conventional value. It represents the participation share of PSIA holder to
banks, IFIs do not use Loan Loss Provisions (LLP) but use the total funds. There is an account value for each individual
PER and IRR to provide steady returns. On the other hand account holder. The total of all depositors’ account values
Zuobi and Al-Khazali (2007) report GCC banks that use make the account value for the total funds.
LLP to smooth their returns.
After defining the basic terminology, we will now define
The evidence for profit distribution management revealed the formula to calculate unit value. As stated, unit value
recently by Farook et al. (2012) represents the strongest is basically an index to track the performance of PSIAs.
support in the literature. They have utilized an extended It can be said that by converting PSIAs to an index value,
dataset covering 37 banks in 17 countries. They have shown the Turkish Banking Authority requires Islamic banks to
most IFIs do really manage profit distributions, with IFIs remove the size effect and concentrate on the performance.
in Brunei, Malaysia and UAE showing lower average profit In our study we use the following simplified formula for
distribution management, while in Bahrain, Indonesia, unit value:
Pakistan, Saudi Arabia, IFIs presenting a higher average
profit distribution management. Percentage of Muslim ut+1 =
uat + Rt - (ct + y t ) (2)
population, financial development, market concentration,
at
depositor reliance and age of the Islamic bank are the most
significant factors to answer the question why IFIs engage where ut+1 represents unit value for time t + 1, the next day,
in income smoothing. uat represents unit account value for time t, Rt represents
revenues to PSIA for time t, Ct represents costs to PSIA for
The existence of income smoothing activities by itself time t, yt represents reserves for time t and at represents
proves the competitive pressure on Islamic banks. Literature account value for time t.8
shows competing with the conventional banks is one of the
most important challenges of IFIs and they utilize various Rt, revenues to PSIA, is defined as a combination of profits
methods to overcome this challenge. In this study we will accrued to the PSIA account within current business day
provide a model to forecast PSIA returns in the short-run to plus any annulment of former provisions and/or reserves.
provide a marketing tool for IFIs. Ct, costs to PSIA, is defined as a combination of provisions
imposed by regulations, plus payments to deposit insurance
3. Methodology fund and loan loss provisions. Revenues and costs are
reflected on the calculation starting with the first day of
In order to define PSIA returns, we will use a simplified the loan. Therefore even if the bank does not receive any
version of the framework mandated by the Turkish Banking installments, the PSIA account will start recording a profit
Authority to the IFIs in Turkey. The reasons behind choosing when the loan is extended. For the purposes of simplicity
the Turkish framework follow: we will assume all loans follow a murabaha contract.9
1. PSIA funds are not mixed with bank equity capital After calculating unit value, we define daily returns to our
2. Profit/loss distribution is done on a daily basis PSIA funds with the following formula:
3. Instead of PER and IRR, LLP is used to smooth returns
which makes it comparable to conventional banks
ut+1 − ut
rt+1 = × 100 (3)
Before stating the formula for PSIA returns, we need to ut

define some of the terminology that will be used in our
specification. In this calculation for unit value, PSIA costs and reserves
are constant percentages of PSIA revenues given by
Unit Value (ut): It is assumed to be equal to 100 on the the regulatory authority. Therefore the only variants in
first day IFI accepted deposits to its PSIAs. It changes daily this formula are unit account value, which represents
based on the distribution of profit and loss to the account. the amounts deposited to the bank and PSIA revenues,
We can consider unit value as an index to tell us how our representing the returns from the loans extended by the
funds are performing.7 There is one unit value for each day bank. Unit account value fluctuates as the depositors
of operation. The change from one day to another reflects withdraw their money or deposit new funds to their
the percentage of profits or losses made by the bank during accounts. PSIA revenues increase as new murabaha loans
one day. are extended to the clients. Therefore in our analysis we
will simulate these two variables and the unit value and
Unit Account Value (uat): It defines the current total daily returns to PSIA funds will adjust.
monetary value of PSIAs for the bank, which can be
calculated by multiplying unit value and account value.
For every individual account holder, unit account value is
equal to the amount of money they have deposited in the
4.  Results and conclusion
first day. Unit account value will grow each day based on
bank’s performance.
Data
In our analysis, we will use the formula generated in
uat = ut × at (1) the methodology section to simulate the daily average
PSIA returns of Turkish participation banks.10 Currently
Account Value (at): It is the value calculated by dividing there are four participation banks operating in Turkey.
the amount of money deposited to PSIAs by that day’s unit They publish daily returns to their PSIA funds for the

Eds. Hatem A. El-Karanshawy et al. 153

 Calkan

Figure 1.  Average PSIA returns for 4 Turkish participation banks.

trailing one-month, three months, six months and one Assuming a unit account value of TL 6 billion and unit value
year for different currency classes in their websites. We of 100, account value is calculated to be TL 60 million.13
have taken four banks’ average data for trailing one- With the separation of unit value and account value, it
month (31 days) returns to Turkish Lira PSIA funds for is now possible to track changes in deposits in two ways.
the period between January 15, 2012, and February 15, Unit value captures the changes due to the investment
2012. In our simulation we try to estimate returns during returns (profits and losses), and account value captures
this one-month period. We have chosen one-month the fluctuation in the funds deposited to the bank (deposits
maturity, since these accounts hold the highest amount and withdrawals).
of funds compared to other maturities. The date selection
is arbitrary. Turkish lira has been chosen as the selected The percentages 15% and 5% for PSIA costs and reserves
currency since the analysis is being done with Turkish assumed in our simulation are an approximation based
data. Figure 1 presents the path followed by PSIA returns on the rules imposed by the Turkish regulatory authority.
for our selection. Turkish regulatory authority requires banks to set aside
reserves under the names of special provision, general
First simulation provision, federal deposit insurance premium and
In our simulation we have assumed the following: precautionary provision. Based on Turkish monetary and
macroeconomic policy, the rates on these provision expenses
At t = 0 At t >0 can be altered; a total of 20% is a fair approximation to
Unit account ua0 : TL6 billion Endogenous the reality. The ratios for PSIA costs and reserves are kept
value constant in all simulations since they represent the best
legal approximation.
Unit value u0 : 100 Endogenous
Account value a0 : 60 million Assumed constant For PSIA revenues our selection of TL 2 million daily
PSIA revenues R0 : TL 2 million Will be estimated return is derived from 6.75%, with a starting monthly
PSIA costs C0 : 15% of revenues Exogenous PSIA return given in Figure 1. TL 2 million daily revenue
Reserves y0 : 5% of revenues Exogenous is equal to 0.03% daily return on TL 6 billion pool of
funds. Using 70/30 profit sharing ratio between the
bank and the depositor, it refers to a 6.72%14 monthly
Unit account value represents the deposits collected by PSIA return.
our average participation bank. Since we are using the
data from the beginning of 2012 for TL accounts, we have In our first simulation, we have assumed the account
referred to TL deposits collected at the end of 2011. Total values to be constant at t = 0 value, meaning that there are
value of TL deposits collected has been TL 24.04 billion.11 no deposits or withdrawals in PSIA funds. Therefore we
In our simulation we have taken the average of this value only need to simulate the trend in PSIA revenues. In 2011,
and used TL 6 billion12 as unit account value. participation banks have extended a total of TL 41.14
billion15 in loans. This will approximately refer to daily TL
Unit value is assumed to be 100. Since it is an index to 40 million16 of loans. The revenue on these loans depends
follow the performance of the funds, instead of its absolute on the profit rate charged by the bank to the borrower.
value, the changes from day to day are significant for our There is going to be a negative correlation between the
analysis. rate and the loan amount. As the rate on the loan goes

154 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks

Figure 2.  Estimating the trend in PSIA revenues: Panel A gives the area for mean-standard deviation combination,
Panel B uses 4 sample points from Panel A to construct confidence interval for PSIA returns.

down there will be more demand for loans and vice versa. in Turkey was 1.27%17. Thus, if the bank extends TL 40
Also the profit rate has to be set in accordance with market million of loans at a rate of 1.27% everyday, PSIA revenues
conditions. Islamic banks using their internal data can are increasing around TL 17,000 per day. Of course this is
make analysis on this negative correlation and find out a rough estimation. Therefore in a way to provide the 95%
the relationship between the profit rate and the amount of confidence interval to PSIA returns curve given in Figure 1,
loans they can extend in detail. However in our analysis we have constructed a range for the trend of PSIA returns. If
we will use market interest rate as a reference point since mean and standard deviation of PSIA revenue change stays
participation banks are competing with conventional banks in the area given in Figure 2 Panel A, the bank can provide
in attracting customers. Average monthly commercial loan a profit rate quote that is within 95% confidence interval of
rate for the period January 15th –February 15th 2012 the actual realized value. In order to demonstrate this, we

Eds. Hatem A. El-Karanshawy et al. 155

 Calkan

have plotted confidence interval for the actual PSIA return Therefore our assumptions in this simulation are as
for four different points in the area given. follows:

The result for the first simulation represents the possibility At t = 0 At t >0
of providing a profit rate quotation to the deposit holders Unit account ua0 : TL6 billion Endogenous
in Islamic banks. As our analysis suggests the banks can do value
this with a level of flexibility in their revenue estimations. Unit value u0 : 100 Endogenous
Here in our analysis we have used market interest rate and Account value a0 : 60 million Linked to deposit
four banks’ average for other variables. Islamic banks on the growth
other hand have access to historical data about individual PSIA revenues R0 : TL 2 million Will be estimated
loans with profit rates, loan values, payment information PSIA costs C0 : 15% of revenues Exogenous
and so on. Using this extra information it is even possible
Reserves y0 : 5% of revenues Exogenous
for them to provide a better estimate for PSIA returns.

Second simulation With the introduction of PSIA funds variation, mean

In the second simulation, we release our assumption of and standard deviation combinations for revenues have
constant account value, in other words now there are been updated. Figure  4 Panel A displays the new range
deposits to and withdrawals from PSIA funds. In the first for revenues that makes it possible for the Islamic bank
simulation account value was assumed constant at TL to provide a profit rate estimate within 95% confidence
60 million along the estimation period. Islamic banks interval of the actual value. In Panel B, four different
in applying this framework can use their internal data mean and standard deviation combinations from Panel A
about the fluctuations in the PSIA funds. However for our have been used to plot confidence intervals for the return
analysis we will use market growth rate as the trend. We estimate. Actual value lies within 95% confidence interval
have extracted the trend in TL deposits for participation in each of them.
banks for a period two months prior to our estimation
period. Weekly deposits data is available at Central Bank’s This second simulation demonstrates a more accurate
website18. Deposits data is equivalent of unit account value estimate for realized values. We have incorporated
in our framework. However we calculate unit account variations in PSIA funds due to new deposits and
value endogenously using equation 1. Therefore we will withdrawals. Although we have used market growth rate
use the trend in deposits as an approximation to trend in as an approximation to PSIA variations, recognizing it as
account value. Figure  3 displays the path followed by TL an exogenous variable, in reality banks have some level of
deposits of Participation Banks. As we have been doing all control over their deposits. They are affected from market
along, here again we have divided total TL deposits value developments, but by charging a higher rate they can
by four to reach an average value for one of the four banks attract more deposits and they are always free to refuse
in the market. new deposits. We have assumed in our analysis that they

Figure 3.  Average TL deposits of participation banks.

156 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks

Figure 4.  Estimating the trend in PSIA revenues: Panel A gives the area for mean-standard deviation combination,
Panel B uses 4 sample points from Panel A to construct confidence interval for PSIA returns.

do not have any control. Figure 5 demonstrates the effect area provides a more conservative estimate, which suits
of change in deposit assumption from one simulation to both growth assumptions.
another. It plots Panel A of Figure 2 and 4 on top of each
other. As can be seen, when we assume growing deposits, In both of the analyses we have considered PSIA revenues
the bank needs to receive more credit return meaning they as a policy tool. The bank can change the profit rate they
will provide more murabaha credits to the clients. This distribute to deposit holders by increasing or decreasing the
figure also shows that Islamic bank can use deposit growth mark-up rate they charge on murabaha credits. Although
as another policy tool. The changes in deposit growth will the demand for credit depends on the rate they charge,
move the policy region up or down. Also, orange colored there is still room for policy making. Therefore, using this

Eds. Hatem A. El-Karanshawy et al. 157

 Calkan

Figure 5.  Changes in revenues.

tool, Islamic banks can also compete with each other and IFIs can improve their fund management by controlling the
with conventional banks in attracting deposits, besides amount of funds deposited to PSIAs and setting their mark-
providing a profit rate quotation. up rates according to their PSIA return targets. Furthermore
these simulation methods can be used to better manage
risks associated with asset liability management and rate of
6. Conclusion return.
In this study we have studied the possibility of providing
an ex-ante return rate quotations to deposit holders at
Islamic banks. Simulation results suggest, it is possible to
offer a reliable forecast within a 95% confidence interval of
Appendix: Detailed unit value calculation
upcoming actual PSIA returns to the clients. The benefits The information below is extracted from Annex 1
of this study are three-fold. The most important benefit regulation on the principles and procedures for accepting,
of providing this information is increasing Islamic banks’ withdrawal of deposits and participation funds as well as
competitive advantage compared to their conventional the prescribed deposits, participation funds custody and
counterparties. Secondly, using these simulation techniques receivables in Turkish banking legislation.

UA + R - (C + Y)
New U
A
Details of Numerator
UA Unit Account Value C PSIA Costs (a+b+c+d)
R PSIA Revenues (a+b+c+d+e) a Special Provision Expenses
a Participation Share of Dividend Incomes b General Provision Expenses
Dividend Incomes Procured from Loans Deposit Insurance Fund (DIF) Premium
a.1 Extended Arising from PSIA c Expenses
a.2 Profit Equivalent to Extended Fund Surplus d Precautionary Provision Expenses
-
b Collections Made from Loans Cancelled
Amounts/ Reserves Allocated from Profit
c Cancellations of Special Provisions Y to be Distributed to PSIA
d Cancellations of General Provisions
Provision Cancellations Set Aside from Profits
e to be Distributed to PSIA
Denominator
A 0 Account Value

158 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks

PSIA Revenues (a + b  + c  + d): general provisions from the total amount of income
items stated in (b), (c) and (d) sub paragraph of the
a) Participation Share of Dividend Incomes: Dividend PSIA revenues explanation. These provisions set aside
amount equivalent to extended fund surplus is deducted are recorded to the account of amounts set aside from
from dividend incomes procured from extended loans profit to be distributed to PSIAs included in communiqué
arising from participation account. The amount found on uniform chart of account and its explanation to be
is separated on a currency type basis according to its implemented by participation banks.
weight in total participation accounts. The amount
found by multiplying the separated amount by the Amounts Allocated from Profit to be Distributed to PSIAs:
ratio of account owner’s participation in profit defines
the amount in dividend income falling to the share of It is the provision amount allocated within the scope of
participation accounts. the provision of the article 14(3) of the regulation on
1. Dividend Incomes Procured from Extended Loans principles and procedures relating to determination of
Arising from PSIA: This is the dividend income qualifications of loans and other receivables by banks
procured from funds extended arising from PSIA on and provisions to be set aside from profit amounts to be
currency basis. Whether or not delay funds collected distributed to participation accounts by calculation date of
for those not paid in their maturity among those unit values.
funds or dividends deprived of as well as income from
required reserves shall be taken into consideration as
dividend income in the unit value calculation of PSIA Notes
shares are determined in PSIA contracts.  1. Wall Street and Financial Crisis: Anatomy of
2. Profit equivalent to extended fund surplus: the Financial Collapse, report by US Senate Permanent
amount found by multiplying by the ratio calculated Subcommittee on Investigations, 2011, Available
by dividing the dividend income procured from at: http://www.hsgac.senate.gov//imo/media/
loans extended arising from PSIA on a currency basis d o c / Fi n a n c i a l _ C r i s i s / Fi n a n c i a l C r i s i s Re p o r t .
to the sum of funds extended, with extended fund pdf?attempt = 2.
surplus.   2. Kayed and Hassan (2011), Al Mamun and Mia (2012).
b) Collections Made from Loans Cancelled: The amount   3. Iran, Sudan and Pakistan.
falling to the share of participation accounts, from the   4. Average market share of Islamic banks in MENA region
collections made concerning cancelled loans from loans is 14%.
extended arising from PSIA.   5. Some of the banks might be using this information in
c) Cancellation of Special Provisions: The amount relating their internal policy making at the moment.
to PSIA, among cancelled amounts of special provisions  6. AAOIFI Statement of Financial Accounting No. 2:
set aside for loans arising from PSIA classified as Concepts of Financial Accounting for Islamic Banking
non-performing loans pursuant to the regulation and Financial Institutions.
on principles and procedures for determination of   7. Turkish Banking Authority’s approach to PSIA return
qualifications of loans and other receivables by banks calculation is similar to return calculation in a fund.
and provisions to be set aside.   8. See appendix for detailed calculation of unit value.
d) Provision Cancellations Set Aside from Profits to be  9. As majority of the loans in the Islamic banking
Distributed to PSIAs: It is the amount cancelled for are provided with murabaha system, we will assume
meeting SDIF premium and special and general all loans given follows murabaha contract for the
provisions of provisions monitored in amounts set aside simplicity.
from profits to be distributed to PSIAs. 10. In Turkey, Islamic Financial Institutions are called
participation banks.
PSIA Costs (a  + b  + c  + d  +e): 11. Data from the website of Participation Banks
Association of Turkey: www.tkbb.org.tr.
a) Special Provision Expenses: It is the part fall to the share 12. 24 billion/4 banks = 6 billion.
of PSIA of general provisions set aside for PSIA emanated 13. 60 million = 6 billion/100.
loans classified as NPL pursuant to the regulation on 14. 0.03% × 360days = 9.60% total return; 70% × 9.60% =
principles and procedures relating to for determination 6.72% depositors share.
of qualifications of loans and other receivables by banks 15. Data from the website of Participation Banks
and provisions to be set aside. Association of Turkey: www.tkbb.org.tr
b) General Provision Expenses: It is the part that falls to the 16. 41.14 billion/4  =  10.28 billion per bank; 10.28/250
share of participation accounts of general provisions set (business days in a year)~40 million per day.
aside for PSIA emanated loans pursuant to the regulation 17. From Turkish Central Bank Electronic Data Distri­
on principles and procedures relating to determination bution System.
of qualifications of loans and other receivables by banks 18. From Turkish Central Bank Electronic Data Distri­
and provisions to be set aside. bution System.
c) DIF (Deposit Insurance Fund) Premium Expenses: It is the
part that falls to the share of DIF premium participation
accounts. References
d) Precautionary Provision Expenses: It is the amount of Akacem M, Gilliam L. (2002) Principles of Islamic Banking:
precautionary provision to be used in meeting the part Debt Versus Equity Financing. Middle East Policy.
fall to the share of DIF premium PSIA and special and 9(1):124–138.

Eds. Hatem A. El-Karanshawy et al. 159

 Calkan

Al-Mamun Md, Mia MAH. (2012) Origin of & Solution Shubber K, Alzafiri E. (2008) Cost of Capital of Islamic
to Global Financial Meltdown: An Islamic View. Banking Institutions: An Empirical Study of a Special
International Journal of Business and Management. Case. International Journal of Islamic and Middle Eastern
(7):12. Finance and Management. 1(1):10–19.
Archer S, Abdel Karim RA. (2009) Profit-Sharing Investment Sultan SAM. (2006) An Overview of Accounting Standards
Accounts in Islamic Banks: Regulatory Problems and for Islamic Financial Institutions. Finance Bulletin.
Possible Solutions. Journal of Banking Regulation. Jan-Mac.
10:300–306.
Sundararajan V. (2005) Risk Measurement and Disclosure
Archer S, Abdel Karim RA, Sundararajan V. (2010) in Islamic Finance and the Implication of Profit
Supervisory, Regulatory, and Capital Adequacy Sharing Investment Accounts. Paper presented at the
Implications of Profit-Sharing Investment accounts 6th International Conference on Islamic Economics,
in Islamic Finance. Journal of Islamic Accounting and Banking and Finance. Jakarta, November 22–24.
Business Research. 1(1):10–31.
Taktak NB, Zouari SBS, Boudriga A. (2010) Do Islamic
Ariffin NM, Kassim SH. (2011) Risk Management Practices Banks Use Loan Loss Provisions to Smooth Their
and Financial Performance of Islamic Banks: Malaysian Results? Journal of Islamic Accounting and Business
Evidence. Paper presented at The 8th International Research. 1(2):114–12.
Conference on Islamic Economics and Finance, Doha,
Turkish Banking Legislation: Annex 1 to Regulation on
December 19–21.
the Principles and Procedures for Accepting, Withdrawal
Atmeh MA, Ramadan AH. (2012) A Critique on Accounting of Deposits and Participation Funds as Well as the
for the Mudarabah Contract. Journal of Islamic Prescribed Deposits, Participation Funds Custody and
Accounting and Business Research. 3(1):7–19. Receivables. Available at: <http://www.bddk.org.tr/
WebSitesi/english/Legislation/9656engdepositandpar
Ayub M. (2007) Understanding Islamic Finance. John Wiley
ticipation_fundannex_08_06_2010.pdf>. Accessed: 10
& Sons, Ltd. England.
May 2013.
Farook S, Hassan MK, Clinch G. (2012) Profit Distribution
Turkish Banking Legislation: Regulation on Procedures
Management by Islamic Banks: An Empirical
and Principles for Determination of Qualifications of
Investigation. The Quarterly Review of Economics and
Loans and other Receivables by Banks and Provisions
Finance. 52(3):333–347.
to be Set Aside. Available at: <http://www.bddk.org.tr/
Ibrahim SH. (2007) IFRS vs. AAOIFI: The Clash of Standard? WebSitesi/english/Legislation/8836eng_provisions_
Munich Personal RePEc Archive (MPRA). No. 12539. to_be_set_aside_13_06_2011.pdf>. Accessed: 10 May
2013.
Kahf M. (2005) Basel II: Implications for Islamic Banking.
Paper presented at the 6th International Conference Turkish Banking Legislation: Communiqué on Uniform
on Islamic Economics and Banking, Jakarta, November Chart of Account and its Explanation to be Implem­
22–24. ented by Participation Banks. Available  at: <http://
www.bddk.org.tr/WebSitesi/english/Legisla­t ion
Kayed RN, Hassan MK. (2011) The Global Financial
/8802eng_par ticipationbank s_unifor m_c har t_
Crisis and Islamic Finance. Thunderbird Int’l Bus Rev.
ofaccounts_08_06_2011.pdf>. Accessed:10 May 2013.
53:551–564.
Zoubi TA, Al-Khazali O. (2007) Empirical Testing of the
Khan F. (2010) How ‘Islamic’ is Islamic banking? Journal
Loss Provisions of Banks in the GCC region. Managerial
of Economic Behavior & Organization. 76(3):805–20.
Finance. 33(7):500–511.
Khan T, Ahmed H. (2001) Risk Management – An
Analysis of Issues in Islamic Financial Industry. Islamic
Development Bank-Islamic Research and Training
Institute. Occasional Paper No. 5. Jeddah.

160 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah
financing
Volker Nienhaus
Honorary Professor, University of Bochum, ICMA Centre, University of Reading (United Kingdom),
Tel. +49 201 8695750, [email protected]

Abstract - Banks avoid participatory financing due to serious information asymmetries, adverse
selection and moral hazard problems resulting in negative impacts for the return on capital
provided. Even financing instruments with a participatory legal form such as musharakah sukuk
have been stripped of their risk sharing substance and become functional equivalents of interest-
bearing bonds. Several authors have addressed these issues, but some proposals are applicable only
for (listed) joint stock companies, while others imply Shariah compliance issues. To overcome these
limitations, a “self-adjusting profit sharing ratio” is proposed, based on building blocks found in
AAOIFI Shariah standards for musharakah financing and musharakah sukuk. These building blocks
allow a (surprisingly) wide range of discretionary adjustments of participatory contracts, provided
the contracting parties come to an agreement in re-negotiations of the contractual terms. This
requires an agreement on a fair distribution of profits. What the parties consider a fair distribution
is already known when the contract is initially concluded: It determines the parties’ profit shares
based on their profit expectations at this point in time. The AAOIFI building blocks allow the
structuring of a formula for the profit sharing ratio, which automatically adjusts to changes in the
expected or actual profit. It thus ensures continuously a profit distribution in line with the initially
agreed-upon principles of fairness. The formula can be calibrated such that the financing party gets
under “normal” circumstances a return in line with a predetermined benchmark (e.g., the market
rate of fixed term financings plus a risk mark-up) while the financed party has the advantages of an
“insurance” against losses and unrestricted upside gains. Thus, financing instruments or sukuk with
new risk/return profiles and some participatory elements could be structured so as to overcome the
problems caused by information asymmetries in “pure” PLS financings.

Keywords: Islamic finance, musharakah, profit and loss sharing, information asymmetries

1.  Discrepancies between theory and moral hazard issues in PLS financing (where the
practice of Islamic finance ratio of profit sharing is fixed in advance and not
changed afterwards) can explain the abstinence
Islamic economists consider finance based on profit and
from participatory modes of financing.
loss sharing (PLS) (participatory finance) as the genuine
• Banks apply the PLS principle only in the deposit
Islamic mode of finance and the major factor distinguishing
business, i.e., in contracts of mudarabah-based
Islamic from conventional finance. Indeed, an economic
investment accounts, but even there the practice
system where PLS is the dominant mode of financing would
was quite different from the model: Fluctuations
have different qualities with regard to efficiency, stability
of investment returns were not passed on to the
and distribution compared to a conventional interest- and
account holders but rather smoothed out by
debt-based system. However, the practice of Islamic finance
recourse to reserves (investment risk reserves and
(which was factually Islamic banking until the early 2000s)
profit equalization reserves) and, in worst cases,
was and is very different from this ideal model.
by interest-free loans or even “voluntary” gifts of
the shareholders to the account holders. This was
• PLS financing hardly ever takes place in Islamic
done to avoid massive withdrawals by disappointed
banks. Bad experiences of pioneering banks and
investment account holders which would have
theoretical explanations of adverse selection and

Cite this chapter as: Nienhaus V (2015). Self-adjusting profit sharing ratios for Musharakah financing.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Nienhaus

created serious problems for the bank. With Islamic economists have recognized that the practice
ex ante announced “anticipated” returns which were of Islamic finance deviates substantially from the ideal,
typically congruent with the realized ex post returns, but they did not simply criticize practitioners for this
and with the coverage of investment accounts by unfortunate development. They analyzed the reasons
capital protecting deposit insurance schemes in for the observable discrepancy (mainly agency problems
some jurisdictions, mudarabah-based investment in anonymous markets), and they came forward with a
accounts got the “look and feel” of conventional number of proposals suggesting how to solve the identified
interest-bearing deposits. problems and promote participatory finance.
• With the growing popularity of sukuk in the 2000s, a
new option for PLS financing emerged in the Islamic • The second chapter summarizes and comments on a
capital market: musharakah sukuk. But again, the number of such contributions. Looking at the rapid
practice converted this participatory instrument into growth of musharakah sukuk in the 2000s, it seemed
an Islamic bond with factually fixed returns for the that at least the Islamic capital market had overcome
sukuk holders. The applied techniques are explained the agency problems and moved toward the PLS
in detail later in this paper. ideal. Unfortunately, this was not the case.
• The third chapter takes a closer look at the
Unfortunately, Islamic bankers and Shariah scholars never structuring of these sukuk and explains how
shared the enthusiasm of Islamic economists for PLS equity-based instruments could be converted into
finance in practice. It is not that Shariah scholars did not functional equivalents of debt instruments with
allow PLS arrangements: Mudarabah and musharakah predetermined costs and capital guarantees. After
contracts (and modern derivatives thereof) are explicitly a critique of prevailing practices by a prominent
approved as Shariah compliant. However, the approval was Shariah scholar in 2007, and a resolution of the
done in a way which opened the door widely for Islamic Accounting and Auditing Organisation of Islamic
bankers to convert the participatory concepts into close Financial Institutions (AAOIFI) in 2008, the issuance
functional equivalents of conventional interest-based and of musharakah sukuk dropped sharply from then
factually risk-free modes of financing. In practice, Shariah until today. Despite the recovery of the sukuk market
scholars have approved the conversion of “equity-based” in recent years, the once dominant form of sukuk has
sukuk (participatory instruments with a variable return) become marginal by today. This is a deplorable signal
into “Islamic bonds” (debt instruments with fixed costs). because mudarabah and musharakah sukuk were the
only financing instruments of significant quantitative
The growth of Islamic finance over the last decade was driven weight in the Islamic finance markets which upheld
by the inroad of Western financial institutions into this new at least the PLS form.
and seemingly lucrative segment of the global financial • The fourth chapter outlines the mechanics of a
industry. Many Islamic financial institutions are run by musharakah sukuk concept, which is based on the
CEOs and management teams who have been socialized PLS principle but uses building blocks found in the
in conventional finance before they converted to Islamic Shariah standards of AAOIFI to overcome agency
finance. In addition, many executives and staff of Islamic problems. The concept allows the structuring of a
financial institutions were hired from the conventional financial instrument that brings differing commercial
sector. Individuals and teams who were successful in interests of contracting parties in balance and
conventional finance before they joined the Islamic finance keeps this balance by a self-adjusting profit sharing
industry were familiar with strategies and instruments for ratio. The automated ratio adjustment takes place
a good or even outstanding performance of their financial whenever new information on the expected or
institution. This was probably the reason why they were actual performance (=profit) of the financed venture
lured away for their previous employment, and shareholders becomes available.
expect a continuation of such a management performance
in the new Islamic environment. Thus, it should not come
as a surprise that Islamic bankers who were socialized in 2.  The agency problem of participatory
the conventional system tried to replicate those strategies finance in anonymous markets
and instruments which they had applied successfully in Participatory finance is a generic term for different forms of
their previous position in conventional finance. financing on the basis of profit and loss sharing, for example
mudarabah and musharakah bank financing or mudarabah
The Islamic economics literature was often too theoretical, and musharakah sukuk.
abstract, macro-oriented or prescriptive to be of much
use for practitioners who were looking for more effective Bacha (1997) sees mudarabah financing as a hybrid of
instruments in commercial, for-profit financial institutions. elements of debt and equity financing. Somehow the
The Islamic economists have not been able to convey their agency problems of both are combined in this hybrid:
enthusiasm for PLS instruments to Islamic bankers – and
probably also not to Shariah scholars. While participatory • The equity agency problem is that the mudarib has a
finance is often seen by Islamic economists as the core strong incentive to “produce” costs which accrue to
of Islamic finance—as sale- and rent-based modes of him as benefits. This goes on as long as the marginal
finance are mentioned only on the sidelines in their utility from fringe benefits or perks exceeds the
models—it is exactly the opposite from the perspective corresponding reduction in the mudarib’s share of
of Shariah scholars: sale- and rent-based contracts have the profit. In addition, since mudarabah financing
been elaborated over centuries in extensive detail, while is for a specified project of an existing firm, the firm
financing based on profit and loss sharing were sidelined. may have various possibilities to shift overhead and

162 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing

other costs to the mudarabah project, thus reducing in all AAOIFI standards dealing with mudarabah and
the profits which have to be shared with the bank musharakah, and later the IFSB took the same position.
(without reducing the overall profit of the firm). A protection against capital losses can only be given by an
• The debt agency problem is one of moral hazard: independent third party, not by the mudarib.
Debt (or more general: external capital at costs
lower than the return on total capital) can leverage Ahmed (2002) also addresses agency problems. He
the return of equity substantially (even if a certain proposes a contractual arrangement that deals with the
percentage of the profit goes to the provider of the moral hazard problem arising from an underreporting
external funds). External mudarabah funds have not of profits. It shall be overcome by an incentive-compatible
only a factual but even a contractual loss absorbing contractual arrangement. The financing bank will have
quality. Consequently, a high leverage by mudarabah the right to undertake a costly audit if the profit reported
funding would create incentives for a firm to embark by the financed firm falls short of the expected profits (on
on projects with high profit potentials and high risks which the parties have to agree). The auditing expenses
because the upside chances surpass the downside will be shared ex ante by both parties and become a co-
risk for the equity holders. determining factor for the calculation of the profit sharing
ratio. But they materialize only if the bank has reason to
Bacha concludes that mudarabah financing has more undertake the audit. If the audit shows that the firm has
agency problems than a pure debt or a pure equity financing reported correctly, the bank will refund the firm’s share
and therefore is an inferior option for the capital providing of the audit costs (as a reward for honesty). If the audit
party. To overcome the agency problems of mudarabah in its uncovers false reporting, i.e., if it becomes apparent that the
genuine form, Bacha proposed the introduction of “equity actual profit is higher than the reported profit, the firm has
kickers:” specified events or outcomes to trigger an equity- to bear the full auditing costs, calculate the bank’s profits
related provision in a financing contract. The equity kicker share on the basis of the actual profit and pay a significant
in a mudarabah contract would be a clause “whereby in the fine as a (additional) penalty for the false reporting. This
event of losses in the Mudarabah financed project, the Rab- contractual arrangement shall reduce (if not eliminate) the
Ul-Mal absorbs the losses but is ‘reimbursed’ for the amount moral hazard incentives for the financed firm.
of losses thru issuance of new equity by the Mudarib to
him.” (Bacha 1997, 18). By this clause the mudarib transfers By resolving the firm’s incentive to swindle, the model
benefits to the rab al-mal so that in the end the rab al-mal’s removes only one of several obstacles that deter Islamic
will be shielded against “avoidable” losses (caused by profit financial institutions to engage more in “true” PLS
reducing management practices or excessive risk taking), financings. If the bank has fixed the profit sharing
and his risk will decrease. For the mudarib the opportunity ratio, and the correctly reported profits fall short of the
costs of profit compression or high risk ventures would expectations, the bank may not receive the benchmark
increase, and this should solve or at least mitigate the of, for example, murabaha or ijarah financings with low
agency problems. risk and predetermined returns (especially not if the
bank insisted on an audit and has to bear the full auditing
A major drawback of the equity kickers approach is that it costs). This, however, may be to the advantage of the firm
can be applied only for the mudarabah financing of joint because the PLS financing costs would be less than the
stock companies. Start-ups and small and medium sized costs of sale/rent-based modes of financing with low risk
enterprises (SMEs) usually do not have this legal form. But and predetermined returns for the bank. This, by the way,
by far the largest number of enterprises are SMEs, which creates another problem – a kind of an ex ante replacement
play a prominent role for employment, income generation of the discouraged ex post moral hazard: the higher the
and poverty reduction in many Muslim (and non-Muslim) profit expectations of the bank, the lower the profit sharing
countries. ratio necessary to meet a given benchmark. If the firm
is able to present its project in such a way that the bank
Another problem is the Shariah compliance of the equity becomes more optimistic about the future profit than the
kickers. Shariah principles prohibit a protection of the firm itself, and if the firm’s expectations are correct, then
rab al-mal’s capital through a guarantee by the mudarib. the results for the firm are similar to those of false reporting
Bacha claims that this is not the case because the equity – but without any risk of penalties.
kickers are no guarantee against losses. The rab al-mal
“will make losses if the project makes losses – although Hasan (2002) presented a model for the determination
it will be much less than under existing Mudarabah.” of the equilibrium profit share for a bank (as provider of
Ignoring the problem of a gharar-free determination of mudarabah capital to a firm) in a mixed system (= where
the value of the transferred equity in cases of loss, the fact the firm has the option of an interest-based loan financing)
remains that the equity has some value and thus is a partial under alternative constellations of the externally-set
compensation for the loss. This could be seen as equivalent market rates of interest (which is a kind of benchmark),
to a partial guarantee of mudarabah capital by the mudarib a risk premium (to be paid in case of loan financing), the
– which would be a violation of a Shariah stipulation. so-called leverage ratio (=  share of externally provided
But Bacha himself sees another violation of a Shariah capital on interest or mudarabah base in the total capital
requirement: “The one Shariah requirement that would of the firm), and the total return on capital of the firm.
not be met by the proposed arrangement is the requirement The equilibrium profit sharing ratio of the bank will be less
that in Mudarabah, the financier should absorb all the than its loss sharing ratio, and it will vary inversely with
losses. Any proposal that seeks to overcome the problems the return on total capital and directly with the leverage
of existing Mudarabah would invariably come up against ratio. The model clarified some interdependencies
this injunction.” This injunction is repeated over and again between variables, but it did not explain (without

Eds. Hatem A. El-Karanshawy et al. 163

 Nienhaus

additional exogenous hypotheses) why the use of mud­ A weakness of this concept that it is applicable only for
arabah financing is so low or what could be done to joint stock companies and it requires the readiness of the
increase its use. Therefore Hasan supplemented his model actual shareholders to accept a certain dilution of their
with contemplations on a change of Muslims’ attitudes ownership rights by the issuance of new or the transfer of
towards very high-risk aversion in business matters (risk existing stocks to new shareholders. This cannot be taken
understood as uncertainty, which cannot be measured and for granted, for example, for family-run businesses.
insured). Since banks are no exceptions to this behavioral
pattern, they are very reluctant to finance risky ventures Further, one may challenge the fairness or balance of a
in which both the earnings and the repayment of the conversion clause for cases where the underperformance
provided capital are uncertain. To improve the situation, is not due to profit skimming of the financed firm but
Hasan makes two proposals: to market forces which lead to an unexpectedly poor
performance of the financed project. In such a situation
• The contracting parties should agree on a clause in the conversion option may be felt as an undue and very
the mudarabah contract “to treat the bank among severe penalty for something that was beyond the control
the preferential creditors of the borrowing firm in of the management and not caused by its “misbehavior.”
case of insolvency … The penalty is severe because the conversion gives the
• … alternatively, the bank can be allotted redeemable financing party a perpetual claim on parts of all future
preference shares for the money advanced.” (Hasan profits, which may exceed the amount of the initially
2002, 49–50). provided capital by a multiple. While this may be very
attractive for the capital provider, it is far from obvious that
It is very questionable whether Shariah could accommodate it would also be acceptable for the financed firm, especially
these provisions. For example, AAOIFI had dealt with if the murabahah or musharakah was only for a short to
the Shariah rules and requirements in the Financial medium term.
Accounting Standard No. 3 on mudarabah financing and
No. 4 on musharakah financing, both adopted in 1996. Finally, a very similar result for overcoming the moral
Neither the treatment of the bank as a preferential creditor hazard issues, but without the requirement of a particular
nor the allotment of redeemable preference shares, as a legal form of the financed firm (joint stock company) and
kind of security, are compatible with the restrictive Shariah the “penalty character” of the conversion clause (under
principles spelled out by AAOIFI. AAOIFI explicated the adverse market conditions), could be achieved by longer
principles in more detail again in the Shariah Standards term or even a perpetual sukuk with a redemption clause.
No. 12 on musharakah and No. 13 on mudarabah, both The redemption clause would be a promise (wa’d) of
adopted in 2002. the issuer to redeem sukuk certificates at their fair value
on the demand of the sukuk holders. An increase of the
Diaw, Bacha and Lahsasna (2012) address agency value of the firm would be reflected in the fair value of the
problems of participatory finance not in banking but certificates.
for musharakah sukuk. They diagnose a fundamental
incongruence between (formal) requirements of Shariah It has to be mentioned here that Diaw, Bacha and Lahsasna
contracts and they aim “to reproduce the substance of a (2012) offer another approach for the analysis of equity-
financial instrument that is repugnant to their nature and based sukuk which is independent from the equity kickers
to the Islamic paradigm in finance” (p. 45). The proposed proposal. It is based on the idea of variable profit sharing
solution takes inspirations from convertible bonds and ratios. A few remarks on differences between their analysis
develops the idea of equity kickers (Bacha 1997) further. and the model outlined below can be found at the end of
Basic agency problems arise in mudarabah or musharakah this paper.
arrangements (bank financings or sukuk) because the
financing contract covers only a limited period of time.
It is in the interest of the mudarib or the managing party 3.  The lack of participatory financing
of a musharakah and the owners of the financed firm to in banking and capital market
keep as much of the realised profits within the firm during While the agency problems of participatory modes of
the period of the participatory financing. A wide range finance can hardly be ignored, it seems that none of the
of possible measures to compress the reported profit that solutions developed in the academic literature have been
has to be shared with the capital provider – from cost applied in practice. While mudarabah financing is virtually
allocation to false reporting – has already been indicated. non-existent in banking, musharakah sukuk expanded at
These techniques reduce the value (payouts) of sukuk an extraordinary rate over the first half of the 2000s and
and increase the value of the firm. This is a problem for became the most widely used form of sukuk by 2007 (see
the capital provider if he cannot participate in the increase chart below).
of the value of the firm. A solution would be a conversion
clause (which is like an embedded option in the financing
contract): If the sukuk performance falls short of a Profit and loss sharing in banking
benchmark (for example, if the return on the sukuk is less The virtual non-existence of participatory (=  mudarabah
than a stipulated minimum), the sukuk holders have the or musharakah) bank financing may, on the one hand, be
right (but not the obligation) to convert their sukuk into explained by the aforementioned agency problems and an
shares of the financed firm. The exercise of the option adverse selection problem that is summarized in the box
would not only prolong the initially temporary partnership below (The “lemon problem” in mudarabah financing).
to an indefinite period but would also grant ownership These issues in their totality have not been solved in
rights to the previously silent partner. theoretical models.

164 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing

The “lemon problem” in mudarabah financing

An entrepreneur can always compare the costs of Shariah-compliant funding based on the expected profit and a
negotiated profit-sharing ratio on the one hand, and the costs of a fixed mark-up sale/rent financing for his project
on the other hand. Less religious-minded entrepreneurs could also add riba-based financing alternatives offered by
conventional banks. The mark-up is determined by competition (within the Islamic banking sector and/or between
Islamic and conventional banks) and becomes the benchmark rate to which the profit sharing ratio has to be adjusted
for a given expected profit. Roughly-speaking, both for the bank and the entrepreneur the mark-up multiplied by the
amount of financing should equal the expected profit multiplied by the respective profit-sharing ratio. If the entrepreneur
were able to convince the bank to become optimistic about his project and to expect a profit that is higher than the
profit he himself expects realistically (without communicating this to the bank), then the bank would agree on a profit
sharing ratio, which is too low (compared with the benchmark or the ratio based on a “realistic” profit expectation).
In this setting, the entrepreneur would benefit from profit sharing financing. The bank has to take into account that in
principle all customers who ask for a mudarabah financing have a strong incentive to present overly optimistic profit
projections. The bank could protect itself to some degree against wrong profit projections by a thorough evaluation
of business plans. But that requires human resources with a profound knowledge of the markets of their customers,
and the experts of the bank should, on average, be better than the entrepreneurs themselves in predicting financial
outcomes of business plans. Expert staff with such qualifications is hard to find, very expensive and probably even
harder to retain (because these employees have all the qualities to become entrepreneurs by themselves). Therefore,
the bank may take recourse to a less expensive protective mechanism, namely a simple “safety margin” on all profit-
sharing ratios. But this will be anticipated by the entrepreneurs. If an entrepreneur presents a realistic profit projection,
the “safety margin” on the bank’s profit sharing ratio will make the mudarabah financing more expensive for him than
the mark-up financing. Entrepreneurs with good projects may not like to enter into the troubles of debating with the
bank the credibility of their profit projections (in order to eliminate the safety margin). Instead, they prefer fixed-
cost financing from the outset. In contrast, for entrepreneurs with weak projects, the mark-up financing may be too
expensive, and they have a strong incentive to present an acceptable profit projection to the bank. In the end, the
bank will have more weak than strong projects in its mudarabah portfolio, and it is highly probable that a number of
weak projects will go bust so that the realised profit will fall short of the expectations. To avoid this, it is probably the
best option not to enter into participatory financing at all. The market for mudarabah and musharakah financing will
collapse (or never emerge).

On the other hand, Hasan’s reference to an exaggerated the issuance of mudarabah and musharakah sukuk offers
risk aversion in Muslim countries points to another a simple explanation: In spite of their participatory form
issue in Islamic banking: The most widely used Shariah and their classification as “equity-based” sukuk, most of
compliant alternative for interest-bearing savings and term these sukuk have never had a participatory substance.
accounts are mudarabah-based unrestricted investment Hence, they did not suffer from the agency problems
accounts. Conceptually, losses from the investment of the discussed in the academic literature. Instead, most of
investment account holders’ funds should be passed on to the musharakah sukuk were intentionally structured as
the investment accounts. The investment account holders functional equivalents of conventional bonds, i.e., as debt
are most probably risk averse, although they have signed a instruments with predetermined returns. On the other
profit sharing and risk bearing contract: Muslims who were hand, “equity-based” sukuk are very flexible and allow (in
looking for an alternative for conventional savings and term contrast to most other types of sukuk such as murabahah
deposits which could give them a Shariah compliant return or ijarah sukuk) the issuing of a security, which is not tied
have hardly an alternative to a mudarabah-based contract, to the true or beneficial ownership of an existing specific
which exposes their funds to a risk of loss. But to accept tangible asset. Instead, mudarabah and musharakah
such a contract because no Shariah compliant alternative sukuk create joint ventures for the investment of the sukuk
is available, does not imply that the account holders would capital in profit-generating Shariah approved assets, but
ever want this risk to materialize. Most probably they these assets must not yet exist when the joint venture is
expect from the bank that all conceivable forms of risk formed. The assets can be created by the employment of
mitigation and risk avoidance are applied. Some Islamic the sukuk resources (i.e., the money paid by the sukuk
bankers articulate such expectations very forcefully. They subscribers), and the composition of the assets held by
defend their banks’ policies of virtually risk free mark-up the joint venture can change over the life of the sukuk.
techniques only and the abandonment of mudarabah or
musharakah in the financing business with their fiduciary This flexibility regarding the underlying asset was the
duties towards the risk-averse investment account holders. main attraction for practitioners and can explain the rapid
growth of musharakah sukuk. Their popularity was not due
to their equity structure which, in theory, brought them
Mudarabah and Musharakah Sukuk closer to the Islamic economists’ ideal of participatory
In view of the serious agency problems of participatory finance. On the contrary, the equity elements, in particular
financing, the boom of musharakah sukuk is much more the possible volatility of returns and the downside risk of a
surprising than the lack of mudarabah or musharakah in capital loss, were somewhat disturbing and have effectively
bank financing. However, a closer look at the practice of been removed by contractual engineering.

Eds. Hatem A. El-Karanshawy et al. 165

 Nienhaus

Debt character of equity-based Sukuk: special purchase undertakings (PUs), which did not
Capital guarantees only guarantee the face value of the certificate at
The debt character of equity-based sukuk was achieved maturity, but factually also the expected profit. These
by a (binding) promise of the obligor to repurchase the PUs were not only triggered by the maturity of the
sukuk certificates at maturity (or in the event of a default) sukuk but also when the venture was not performing
at their issuing price, respectively their face value. This well and the obligor failed to pay the expected profit.
eliminates contractually the risk of a capital loss of the rab The exercise of the PU obliged the obligor to pay
al-mal, and it is a functional equivalent to the guarantee the outstanding principal plus any so far accrued
of the capital of one party by the other. Effectively, losses but unpaid profit. Mokhtar (2011, p. 33) points out
are not borne in proportion to the capital contributed to the that “accrued profit is not necessarily actual profit
venture, or even not borne at all by the capital providers. earned. Profit accrued is the expected profit that is
An arrangement with such a consequence can hardly meet earned by the investors as time passes by.” Given that
the Shariah principle that only risk justifies return. The the prospectus indicated an expected profit accrued
principle of loss sharing or loss bearing has been stated over over the life of the sukuk, then the early redemption
and again—from classic legal manuals to contemporary clause for the PU was effectively a guarantee of a
AAOIFI standards (for example, Financial Accounting (minimum) predetermined return. This converts the
Standard No. 4 on musharakah financing, adopted in 1996, substance of an equity certificate into the equivalent
Shariah Standard No. 5 on guarantees, adopted in 2001, of a conventional bond. It is obvious that this very
No. 12 on sharika (musharakah), adopted in 2002, and No. special form of a “face value plus accrued profit PU”
17 on investment sukuk, adopted in 2003). violates Shariah principles even more than a “plain
face-value PU” at maturity.
There is but one escape from the rule of loss bearing by • A more widely used technique for the conversion
the capital provider, namely a voluntary guarantee by an of a participatory instrument into a close equivalent
independent third party. The case study of the 2005 IDB of a bond with (nearly) predetermined returns
sukuk by Mokhtar (2011, pp. 34–35) reveals that the criteria for the sukuk holders in “good” years (where the
for the definition of an “independent third party” can be mudarabah or musharakah generates a return that
very formalistic and limp in economic substance. IDB had meets or surpasses an articulated profit expectation
set up for its sukuk issuances the IDB Trust Services Limited (=benchmark) were “incentive fees”: The profit share
in Jersey, a SPV with an authorized share capital of £10,000 of the sukuk holders is set, for example, at 99%.
and an issued share capital of £2. All the assets underlying Should the actual profit fall short of the expectation
the sukuk issuances were transferred from IDB to its SPV, (benchmark), (nearly) all of the profit goes to the
and the prospectus of the 2005 sukuk advertised the fact sukuk holders. But as soon as the actual profit exceeds
that IDB was the unconditional and irrevocable guarantor the expected profit (which should be the normal
of the sukuk issuance. The Shariah Board of IDB declared: situation), the amount of the actual profit that
“As it [IDB] is not the issuer of the Trust Certificates and is exceeds the benchmark is given to the sukuk manager
not a manager or participant, IDB can enter into contractual as an incentive fee for “good management.” Suppose
obligations which have the effect of guaranteeing the that the agreed upon expected profit is calculated on
Aggregate Nominal Amount of the Trust Certificates and the basis of LIBOR plus a risk factor as the benchmark.
any Periodic Distribution Amounts in respect of the Trust Then this arrangement implies that sukuk holders
Certificates.” Unfortunately, the Shariah resolution does will receive under “normal circumstances” the
not disclose whether the Shariah Board considered his equivalent of the risk-adjusted market rate of interest.
decision in accordance with AAOIFI standards, and if not, It may come as a surprise, but such a technique is in
how it would justify its different position. harmony with AAOIFI standards and thus should be
considered Shariah compliant.
Predetermined returns for equity-based Sukuk In November 2007 the chairman of AAOIFI’s Shariah
For achieving predetermined returns, sukuk engineers used Board criticized the prevailing practices that changed
at least three different techniques: the substance of musharakah sukuk from an equity-based
instrument (as it was conceived) into a functional equivalent
• When actual profits fall short of the expected profits, of an interest-bearing bond, and in 2008 AAOIFI issued a
sukuk managers often provided interest-free loans resolution on sukuk. This resolution is a pointed summary
(which should be recovered later) in order to meet of what was already contained in the AAOIFI standards.
the expectations of the sukuk holders and to beef-up
the payouts to them. AAOIFI made it clear that this
practice is not Shariah compliant: “It is not permissible Building blocks of “AAOIFI compliant” Sukuk
for the Manager of Sukuk, whether the manager acts as However, AAOIFI did not summarize in this resolution
Mudarib (investment manager), or Sharik (partner), a number of remarkable provisions in AAOIFI standards,
or Wakil (agent) for investment, to undertake to offer which could be used as “building blocks” for the structuring
loans to Sukuk holders, when actual earnings fall short of more “AAOIFI compliant” sukuk:
of expected earnings. It is permissible, however, to
establish a reserve account for the purpose of covering • The profit sharing formula has to be fixed at the
such shortfalls to the extent possible, provided the beginning of a mudarabah or a musharakah: “It is a
same is mentioned in the prospectus.” requirement that the mechanism for distributing profit
• Another technique that violates Shariah principles must be clearly known in a manner that eliminates
was applied in some musharakah sukuk with very uncertainty and any possibility of dispute.”

166 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing

• The profit distribution does not have to be based on a pronouncement on musharakah sukuk, the practice moved
profit sharing ratio which is a simple percentage. The away from equity-based sukuk and switched to ijarah and
ratio can also be based on a more complex formula: murabaha sukuk since 2008.
“It is permissible for the partners to agree on the
adoption of any method of allocation of profit, either The trends captured in the graph until 2009 continued: In
permanent or variable, for example, by agreeing that 2012, the share of mudarabah and musharakah sukuk in
the percentages of profit shares in the first period are global sukuk issuances (in US$) had declined to 15% while
one set of percentages and in the second period are murabaha and similar sale-based sukuk accounted for 65%
another set of percentages, depending on the disparity and ijarah sukuk for 16%.
of the two periods or the magnitude of the realised
profit. This is allowed provided that using such a The growing popularity of sale- and rent-based sukuk could
method does not lead to the likelihood of a partner be explained as follows:
being precluded from participation in profit.”
• Irrespective of the complexity of the initially accepted • Murabaha sukuk are based on short term debt creating
formula, it is “permissible for the parties to change the sale transactions, and a sufficiently large portfolio of
ratio of distribution of profit at any time and to define the such transactions should provide effectively a built-
duration for which the agreement will remain valid.” in protection (albeit not a formal guarantee) for the
• The changing of the distribution scheme can even be capital invested by the sukuk holders and accrued
made when the actual profit of the venture is known profit shares.
at the end of the life of the sukuk: “The parties may • Ijarah sukuk have a longer maturity and a significantly
bilaterally agree to amend the percentages of profit- higher market risk that could lead to losses. However,
sharing on the date of distribution. Also, a partner may AAOIFI allows for this type of sukuk a straightforward
relinquish, on the date of distribution, a part of the guarantee of the capital by a purchase undertaking
profit that is due to him in favor of another party.” of the issuer at face value.
• “It is permissible for the issuer or the certificate
holders to adopt permissible methods of managing Another reason for the departure from musharakah sukuk
risk, of mitigating fluctuation of distributable profits may be that the aforementioned options for more flexible
(profit equalisation reserve), such as establishing profit distribution rules and re-negotiations were not
an Islamic insurance fund with contributions of practicable in the business environment in which sukuk
certificate holders, or by participating in Insurance flourished, namely the market of institutional investors
(Takaful) by payment of premiums from the income and the interbank market. The market participants are
of the shares of Sukuk holders or through donations mostly financial institutions which are hardly interested in
(tabarru’at) made by the Sukuk holders.” participatory components in their contractual arrangements
because that would imply an additional rate of return risk.
Further, if predetermined returns are the objective, it may
The changing Sukuk landscape be less complicated and less risky to structure sale- or lease-
Instead of using these building blocks to modify sukuk based deals than to insert more complex profit distribution
in such a way that they meet the 2008  AAOIFI Shariah formulas into musharakah contracts. Finally, a readjustment

Figure 1.  Trend in types of Sukuk issued.

Source: Mokhtar 2011, p. 5

Eds. Hatem A. El-Karanshawy et al. 167

 Nienhaus

of the profit sharing ratios at the day of distribution, i.e., by commercial banks in many Muslim countries, in
when the distributable profit is known, is commercially particular in those countries where the transformation
equivalent to the fixing of the distribution of the profits in of the economic system is on the political agenda. On
absolute amounts. the other hand, SMEs are usually individual or family
run enterprises which do not have the legal form of a
More important, a profit sharing ratio agreed upon at the joint stock company and cannot get funding from the
beginning of the contract will generate one allocation of floating of shares.
amounts (in absolute figures) on the distribution day, and a • Two phases in the life of a SME can be of particular
revised ratio will generate a different allocation of amounts. interest to sukuk-funded finance companies: the
To accept the revised ratio implies a definite gain for one start-up phase and the expansion phase (after
party and a definite loss for the other. It is extremely unlikely a SME has become established and grows in its
that banks or institutional investors would voluntarily give market) with different risk/return profiles for funds
up profits which are legally due to them in favour of another provided. Both could be quite attractive for finance
bank or institutional investor. Against this background companies compared with financings in areas where
it is not surprising that the additional building blocks for stiff competition (by banks) has compressed margins.
mudarabah and musharakah sukuk were not utilized in the However, sufficient expertise in the industries of the
typical market environment for this kind of sukuk. financed SMEs is required for the assessment of the
business plans for start-up or expansion.
But the situation could be different under a different • The finance companies could provide sale/rent-
market setting and for institutions with somewhat different based financings (murabaha, istisna’, ijarah) to their
objectives. For example, “Modarabah Companies” were target group, but they could also apply equity-based
established in the 1980s in Pakistan. They can have a similar types of financing, for example mudarabah for start-
function as an SPV (set up, for example, by a financial holding ups, musharakah for growth financing. Participatory
company or a bank): As non-banks, they could collect funds modes of finance could generate higher returns, but
through the issuance of mudarabah sukuk and invest these they are also associated with higher risks: general
funds in profitable projects, for example in the leasing of business risks, but also the risks resulting from the
equipment to manufacturing firms on the basis of ijarah various agency problems outlined above.
contracts or in seed or growth financing on a musharakah
basis. “Initially, there was a desire for the mudarabah sector Where SMEs are not joint stock companies, the agency
to concentrate on funding SMEs in Pakistan, which were at problems in financing cannot be solved by those proposals
times neglected by the banks. Unfortunately, due in part to which are based on a conversion of temporary external
the profit motive, it has steered more toward medium-sized participatory capital to permanent equity by providing
enterprises and big-ticket funding, an area where banks stocks to mudarabah or musharakah investors. Some SMEs
are already involved. Rather than competing with banks, may have the legal form of joint stock companies, but if
it may prove a competitive advantage to focus on smaller they are owned by individuals or families who do not want
enterprises in a microfinance manner, as was originally to dilute the ownership structures through the issuance
envisioned.” (Khwaja 2009, 245). of new (or transfer of existing) stocks to other parties, they
will hardly opt for finance contracts with equity kickers or
Non-bank finance companies could utilize for their own similar conversion clauses.
funding sukuk structures. Their sukuk issuances will have
similarities but also marked differences to mudarabah and
musharakah sukuk of investment banks and actors in the 4.  A Musharakah Sukuk with a self-
interbank market. adjusting profit sharing ratio
Since AAOIOFI has provided an interesting but underutilized
• Finance companies which do not take deposits but toolbox for flexible mudarabah and musharakah sukuk
issue sukuk would not need a full banking license structures, it is possible to reconcile the main interests of
in most jurisdictions, and they would not be subject sukuk holders and financed entrepreneurs, namely:
to the strict Basel III capital adequacy and liquidity
rules for banks. • the interest of risk-averse sukuk holders in risk-
• Their sukuk could be subscribed by institutional mitigated predictable and stable returns,
investors, but it is also possible to envisage a retail- • the interest of entrepreneurs
oriented marketing. Malaysia has recently taken • to get some financial relief in “bad times,” i.e.,
steps towards the creation of a retail sukuk market. to have some risk-sharing elements in financing
• Institutional investors have already shown their arrangements in a year of unexpectedly poor
strong preference for predetermined returns. On performance
the other hand, it seems reasonable to assume that • to have shared but uncapped upside potentials in
also retail investors – similar to investment account “goods times,” i.e., to avoid a total skimming-off
holders – would be risk averse even if they purchase of profits in years of above-average performance
equity-type securities.
• The profit generating projects of sukuk-funded The basic idea of the musharakah sukuk with a self-
finance companies will be smaller than the big ticket adjusting profit sharing ratio is quite simple: If it is
transactions in the actual sukuk market. Small and permissible to adjust the profit-sharing ratio at any time
medium sized enterprises (SME) could become a during the life of a mudarabah or musharakah sukuk by
particular target group for the finance companies. consent of the contracting parties, this must not be done in
On the one hand, SMEs may be still underserved a discretionary manner (by separate negotiations after new

168 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing

performance information become available). Instead, this profits will remain with the other party in good years, the
can be automated by a simple formula which links the profit value of its profit share would be reduced in bad years.
sharing ratio to the actual performance (in the simplest The simple model outlined below indicates, inter alia, that
case to the actual profit) of the financed project or venture. the distribution parameters could also be calibrated with
respect to leverage effects of external capital (provided
To calculate performance-dependent profit sharing ratios for the parties agree on such a pattern). These are only a few
a musharakah structure between an SPV and an enterprise, examples for a wide range of conceivable arrangements for
it must be known how much capital is provided externally the reconciliation of the revealed preferences and interests
by the SPV (respectively the sukuk holders) and how much of the contracting parties in a participatory finance setting
internally by the enterprise (originator). At the beginning such as a musharakah (sukuk or bank financing). The
of the joint venture, a certain total profit is expected which preferences and interests are contractually recorded and
has to be distributed to the external and the internal capital translated into a structure that protects the distributional
provider. The profit sharing ratio denotes the share of profit preferences of the contracting parties by automatic
allocated to the external capital provider. This profit sharing alignments of the distribution parameters (in particular
ratio will not be fixed numerically directly but computed by the profit sharing ratio) whenever new performance
a formula on which the parties have agreed. This formula is information of the project or joint venture become available.
based on a distribution pattern for the profits about which The automatic adjustment obviates or replaces ex post
the contracting parties have achieved a consensus. It could re-negotiations on redistributions which are permissible
be, for example, a fixed relation between the rate of return for according to AAOIFI, but very difficult to realise when the
the external and for the internal capital, or it could tie one of gains of one party are the losses of the other party.
the rates of return to an external benchmark. The distribution
pattern, i.e., a particular relation between the rates of return The following is an illustration of the basic mechanism
or the link of one rate to a benchmark, has to be determined underlying the idea of an adjustable profit sharing ratio.
at the beginning of the joint venture, and this pattern Assume that the contracting parties had agreed on one of
(= distribution formula) should not be modified afterwards. the following distribution rules (“objective functions” in
the model):
A model in which only ratios are agreed upon does not fix
ex ante the absolute value (in currency units) of the profit The profit sharing ratio should be such that either
for the contracting partners that would violate Shariah
principles. It only determines the relative positions of the • the return on external capital is equal to the return
parties. The absolute volumes depend on the realized on total capital, or
profits. To keep the relative positions of the parties in the • the return on external capital is the same as the
agreed-upon proportion, it is necessary that the profit return on internal capital, or
sharing ratio is adjusted automatically whenever new • the return on external capital equates the benchmark
information on the expected or actual profit of the joint β, or
venture become available. This is automatically achieved • the return on internal capital is a multiple or fraction
by a simple formula. For the computation of the adjustable α of the return on external capital
profit sharing ratios, the following variables are used:
The appropriate profit sharing ratios are determined as
Ke = external capital (provided by the SPV) follows:
Ki = internal capital (provided by the entrepreneur)
K = Ke + Ki = total capital Cases (1) and (2), profit sharing ratio for re = ri = r:
P = total distributable profit Case (3), profit sharing ratio for re = β:
Pe = s ∙ P = profit allocated to the external capital Case (4), profit sharing ratio for ri = α ∙ re:
Pi = (1 − s) P = profit allocated to the internal capital
s  =  Pe/P  =  profit sharing ratio  =  share of total profits A numerical example is given in the following table. It
allocated to the external capital illustrates the influence of different benchmarks, of a
β = benchmark rate of return surcharge on the relative return for one party, and of
re = Pe/Ke = rate of return for external capital different (actual or expected) profits. The table does not
ri = Pi/Ki = rate of return for internal capital change the relation between external and internal capital,
r = p/K = rate of return on total capital but the relevance of the capital structure is clearly visible
from the profit sharing formulas above.
Next an “objective function” has to be defined. The
contracting parties may discuss various alternatives. For The above sample did explain the basic mechanism of the
example, they may consider it as a just distribution that model only for three simple objective functions. In practice,
both parties achieve the same rate on return on their contracting parties might agree on more complex formulas
invested capital. A variant of this approach could be to give (for example, with caps or equivalents of “incentive fees”).
one party a predetermined bonus over the share of the other Admittedly, it would also be possible to define an objective
partner (for example, as a compensation for management in such a way that one party receives a fixed amount as
efforts). Another plausible scenario could be that the profit share (provided the volume of the total profit is at
provider of the external capital prefers a profit distribution least as large as the fixed amount) – which would not be
which reduces the volatility of his own profit share and permitted under Shariah. But discretionary re-negotiations
gives him a more stable revenue stream that meets a certain can achieve the same result. Insofar the automated system
benchmark; for example, the return from an investment in is not better or worse than discretionary practices regarding
fixed-income securities (such as ijarah sukuk); exceeding a possible “misuse.”

Eds. Hatem A. El-Karanshawy et al. 169

 Nienhaus

Profit Sharing Ratios for Different Objective Functions

ri = αre, ri = αre,
Objective function: re = r ( = ri) re = β re = β α = 1.1 α = 0.8

Ke Assumption 400 400 400 400 400 400

Ki Assumption 100 100 100 100 100 100
K = Ke+Ki Definition 500 500 500 500 500 500
P Assumption 50 50 50 50 50 50
s = Pe/P Definition
Pe = s∙P Def. & result 40 48 32 39 42
Pi = (1-s)P Def. & result 10 2 18 11 8
β and α Assumption 0,120 0,080 1,100 0,800
re = Pe/Ke Def. & result 0,100 0,120 0,080 0,098 0,104
ri = Pi/Ki Def. & result 0,100 0,020 0,180 0,108 0,083
r = P/K Definition 0,100 0,100 0,100 0,100 0,100 0,100

Solutions:

s = Ke/K re = r (= ri) 0,800

s = β(Ke/P) re = β 0,960 0,640
s = 1/(1+αKi/Ke) ri = αre 0,784 0,833

ri = αre, ri = αre,
Objective function: re = r ( = ri) re = β re = β α = 1.1 α = 0.8

Ke Assumption 400 400 400 400 400 400

Ki Assumption 100 100 100 100 100 100
K = Ke+Ki Definition 500 500 500 500 500 500
P Assumption 75 75 75 75 75 75
s = Pe/P Definition
Pe = s⋅P Def. & result 60 48 32 59 63
Pi = (1-s)P Def. & result 15 27 43 16 13
β and α Assumption 0,120 0,080 1,100 0,800
re = Pe/Ke Def. & result 0,150 0,120 0,080 0,147 0,156
ri = Pi/Ki Def. & result 0,150 0,270 0,430 0,162 0,125
r = P/K Definition 0,150 0,150 0,150 0,150 0,150 0,150

Solutions:

s = Ke/K re = r (= ri) 0,800

s = β(Ke/P) re = β 0,640 0,427
s = 1/(1+αKi/Ke) ri = αre 0,784 0,833

It is clear that the profit sharing mechanism will only work as in a musharakah structure. They present a general equation
long as profits are generated. It does not provide an effective that indicates the directions in which different parameters
protection against the need of “loss sharing” in an individual influence the profit sharing ratio. They do not transform
musharakah setting. In the interest of risk mitigation and loss their general equation in such a way that it would become
avoidance, the management of a musharakah sukuk should, an objective function and could be used (after calibration)
for example, diversify investments by financing not only one in negotiations on a distribution pattern. Instead, Diaw,
entrepreneur but a number of firms in different markets Bacha and Lahsasna present a Monte Carlo simulation,
with uncorrelated market trends, and it may also invest which gives a feeling of implied dependencies and possible
some funds in risk-minimized fixed-return instruments. dynamics. In addition, they offer a backtesting, which
But risk management in Islamic finance in general is a topic shows relations between the average return on capital,
which goes beyond the scope of this paper. an average benchmark (indicated opportunity costs or
benefits), the average return to sukuk, return to equity and
The idea of a variable profit sharing ratio can also be found the profit sharing ratio. However, their input data for this
in the paper of Diaw, Bacha and Lahsasna (2012). But their exercise were taken from mudarabah sukuk, i.e., sukuk
perspective is a theoretical analysis and not the outline of which do not incorporate a profit sharing ratio. Insofar it
an implementable recommendation for contracting parties is not clear what the results of the backtesting can show.

170 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing

It may be read such that it shows (on the basis of observed For the financed entrepreneur it is beneficial that
returns on equity and returns on sukuk capital) what profit a participatory financing implies a kind of “embedded”
sharing ratios would have been necessary in each period to insurance against unexpected downside risks (loss sharing).
achieve a certain benchmark return. However, the loss absorbing qualities of a mudarabah or
musharakah contract will not come for free but will be
Diaw, Bacha and Lahsasna do not explicitly integrate reflected in the costs of funds (e.g., by a risk premium
changes of the expected (or actual) profit into their added to a benchmark rate by the financier). On the other
analysis. Changes of profits and those adjustments of the hand, the contract can be calibrated such that both parties
profit sharing ratio during the life of a participatory finance can enjoy upside gains. This is in contrast to the widespread
are the main focus here, which are required to sustain the practice of a complete skimming off of gains by one party
initially stipulated profit distribution pattern. in musharakah sukuk. The financier can trade in upside
opportunities for predictable and stable returns based on a
Islamic economists have made many attempts to overcome risk-adjusted benchmark rate.
agency problems in participatory finance in order to make
them more acceptable to market participants, in particular It is obvious that arrangements based on self-adjusting profit
to providers of funds. This is because many see participatory sharing ratios are no “real” profit and loss sharing contracts
finance as the ideal form of Islamic finance and as the core or “full” risk sharing partnerships. Insofar the approach is
element of a genuine Islamic financial system. Even if one not the first best solution in an ideal world. However, under
would not go that far, participatory modes of finance such real-world circumstances first best models do not work,
as mudarabah or musharakah could fill a gap, for example, and the technical and Shariah merits or shortcomings of
in the start-up and growth financing of SMEs. In general, the approach should be discussed in relation to other real-
participatory finance gives entrepreneurs some financial world proposals to overcome the inherent agency problems
relief in times of an unexpectedly poor performance of their of PLS or risk sharing arrangements.
business. This risk-reduction can enhance their willingness
and ability to ramp up innovations (new products, processes
or technologies) or to enter into new markets. This should References
not only generate private profits but also social benefits AAOIFI. (2008) Guidance Statement on Accounting for
in terms of employment and income opportunities. The Investments and Amendment in FAS 17. Available at:
model presented here outlines a technique which could http://www.aaoifi.com/aaoifi_sb_sukuk_Feb2008_
make participatory modes of financing more attractive to Eng.pdf.
financiers as well as entrepreneurs seeking finance.
AAOIFI. (2010a) Accounting, Auditing and Governance
Standards for Islamic Financial Institutions. 1432 H –
5. Conclusion 2010. Manama: AAOIFI.
AAOIFI has allowed a remarkable wide range of “corrective AAOIFI. (2010b) Shari’a Standards for Islamic Financial
measures” in mudarabah and musharakah contracts Institutions. 1432 H – 2010. Manama: AAOIFI.
which are deemed Shariah compliant: The contracting
Abdel-Khaleq AH, Crosby T. (2009) Musharakah Sukuk:
parties can re-negotiate factually all commercially relevant
Structure, Legal Framework and Opportunities. In:
aspects of their contracts at any time. A main reason for
Abdulkader T (Ed.). Sukuk. Sweet & Maxwell Asia.
such re-negotiations will be new information about the
Petaling Jaya. 187–222.
performance (profit) of the financed project or venture. The
crux of re-negotiations is that they will be successful only if Ahmed H. (2002) Incentive-Compatible Profit-Sharing
one party is willing to give up advantages (financial gains) Contracts: A Theoretical Treatment. In: Munawar I,
of the contractual status quo. This voluntary redistribution David T. Llewellyn DT (Eds.). Islamic Banking and
is not very likely. Finance – New Perspectives on Profit-Sharing and Risk.
Edward Elgar. Cheltenham. 40–54.
The proposal of a self-adjusting profit sharing ratio obviates
Bacha OI. (1997) Adapting Mudarabah Financing to
the need for discretionary re-negotiations. It uses elements
Contemporary Realities: A Proposed Financing Structure.
of the AAOIFI toolbox to structure a contractual arrangement
The Journal of Accounting, Commerce and Finance. 1(1).
that maintains the distribution pattern, which was initially
http://mpra.ub.uni-muenchen.de/12732/.
agreed upon by the contracting parties. It does so by an
automatic adjustment of the profit-sharing ratio whenever Casey P. (2012) Regulatory Lessons on Sukuk Financial
new information on the expected profit becomes available. Products, an Opinion. In: Ariff M, Munawar I, Shamsher
This technique facilitates a solution of fundamental agency M (Eds.). The Islamic Debt Market for Sukuk Securities:
problems in participatory finance. It does not require The Theory and Practice of Profit Sharing Investment.
a particular legal form of the financed firm (such as a Edward Elgar. Northampton, MA. 99–118.
joint stock company) and can be calibrated such that risk
Diaw A, Obiyathulla IB, Ahcene L. (2012) Incentive-
aversion of the financier can be factored in. A musharakah
Compatible Sukuk Musharakah for Private Sector
contract with a self-adjusting profit sharing ratio is incentive
Funding. ISRA International Journal of Islamic Finance.
compatible insofar as it does not provide the incentives for
4(1):39–80.
profit compression by underreporting, by allocation of fixed
cost, by fringe benefits, or by shifting profits into periods Hasan, Z. (2002) Mudarabah as a Mode of Finance in
after the termination of the musharakah contract. All this Islamic Banking: Theory, Practice and Problems. Middle
is achieved without the need for a sophisticated accounting East Business and Economic Review. 14(2):41–53. http://
system of the financed enterprise. mpra.ub.uni-muenchen.de/2951/.

Eds. Hatem A. El-Karanshawy et al. 171

 Nienhaus

Ibn R. (1996) The Distinguished Jurist’s Primer, Vol. 2 Demetriadies DG, Tyser CR, Effendi IH (Translators).
(Bidayat al-Mujtahid). Translated by Imran Ahsan Khan (2001) The Mejelle – Being an English Translation of
Nyazee. Garnet. Reading. Majallah El-Ahkam-i-Adliya and a Complete Code on
Islamic Civil Law. The Other Press. Kuala Lumpur.
Kapetanovic H, Muhamed B. (2009) Mudharabah Sukuk:
Essential Islamic Contract, Applications and Way Saeed A, Salah O. (2012) History of Sukuk: Pragmatic and
Forward. In Abdulkader T (Ed.). Sukuk. Sweet & Idealist Approaches to Structuring Sukuk. In: Ariff M,
Maxwell Asia. Petaling Jaya. 223–247. Iqbal M, Shamsher M (Eds.) The Islamic Debt Market
for Sukuk Securities: The Theory and Practice of Profit
Khwaja M. (2009) Mudharabah Sector Report – Pakistan.
Sharing Investment. Edward Elgar. Northampton, MA.
In Abdulkader T (Ed.). Sukuk. Sweet & Maxwell Asia.
42–66.
Petaling Jaya. 245–246 (=  Annex to Kapetanovic and
Becic 2009).
Mokhtar S. (2011) Application of Wa’ad in Equity Based
Sukuk: Empirical Evidence. ISRA Research Paper 20.
ISRA. Kuala Lumpur.

172 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk
sharing mechanism for government financing
in Muslim countries
Syed Aun R. Rizvi1, Shaista Arshad2
1
INCEIF, Lorong Universiti A, Kuala Lumpur, Malaysia, Phone: +60136145752, Email: [email protected]
2
IIUM, Institute of Islamic Banking and Finance, Kuala Lumpur, Malaysia, Phone: +60133771370,
Email: [email protected]

Abstract - Over the past decades much effort and research has gone into establishing a viable set
of Islamic financial institutions. An area of utmost importance, which still has gaping holes, is the
development of instruments for government financing on a global level. Most Muslim countries,
with the exception of a few Gulf countries, are heavily indebted with high reliance on multilateral
financing primarily based on high interest rates. This vicious cycle of interest rates and debt have
stunted the growth of these nations and worsened the conditions of the masses. This paper is
an attempt at introducing the concept of risk sharing instrumentation in the form of GDP-linked
Papers for Muslim governments for their financing through multilateral and supranational bodies.
Without dwelling much on the Shariah technicalities, we attempt to discuss the potential benefits
of transferring government debt in these forms of risk sharing instruments. The authors have
attempted to represent the economic benefits of these risk-sharing mechanisms for a set of four
indebted Muslim countries with empirical proof. Through this paper, we endeavour to initiate a
thought provoking and practical discussion for further development of these instruments for the
betterment of Islamic countries.

Keywords: risk sharing instruments, sovereign debt, GDP-linked paper, Muslim countries

JEL Classification Codes: F34, H63

1. Introduction Yet despite this impressive record, Islamic finance only

accounts for 1% of the global financial system. Some
Islamic finance has progressed phenomenally over the past
scholars are of the opinion that this is due to the substantial
few decades, moving from a small market in the Arab world
indebtedness of Muslim nations. Narrowing in on the
to a much larger global presence. Marked with impressive
Organization of Islamic Centre (OIC) member countries, a
growth, the presence of Islamic finance is estimated to be
study revealed 19 from 39 Heavily Indebted Poor Countries
US$1.6 trillion and is expected to grow at the rate of 15%
(HIPC) are part of the OIC (IMF, 2010). Furthermore, a mere
over the next few years.
6 out of 57 OIC member countries, in the Gulf Cooperation
Council (GCC) are prepared to meet financial shocks of
The viability of Islamic finance as an alternative was put
sorts. The remainder are heavily reliant on borrowing from
to test in the wake of the recent financial crisis, where
the international market to pay off their debts.
in the midst of bank failures and stock market crashes,
institutions practicing Islamic finance managed to stay
Additionally, these countries are dependent on Western
afloat and rather provided a buffering effect to the crisis.
financial intermediaries with high interest rates to meet
Several economists argue that due to the innate laws of
their debt requirements, transacting mainly in dollars.
Islamic finance it was able to withstand the financial crisis
During times of economic instability, liability dollarization
much better than its conventional counterpart was.
(see Calvo et al., 2003; Calvo and Reinhart, 1999; Durdu

Cite this chapter as: Rizvi S A R, Arshad S (2015). Indexing government debt to GDP: A risk sharing mechanism for
government financing in Muslim countries. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on
corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Rizvi and Arshad

and Mendoza, 2005) and relying heavily on high interest countries was discussed by Hasan and Siddiqui (1992) who
borrowings can backfire for these countries, as they are more highlighted that an equity financed deficit is not necessarily
likely to have long-term shocks on the economy leading to unstable and that the output of the economy can be
higher cost of borrowings. Alternatively, they may need to enhanced by the government expenditure policies.
restructure or renegotiate the terms of borrowing or lastly,
be forced to employ pro-cyclical policies that would lead to Keeping in mind the evident lack of research on this
disastrous economic and social outcomes. pertinent topic, this paper aims to assess the benefits of
GDP-linked Sukuk, and tries to empirically show some of
One way to avoid any of the above possibilities is to index the key benefits these instruments can offer. This paper also
their debt against real economic variables to reduce attempts to address some of the concerns from the investor
financial distress. Several economists advocate the value angle, to provide a case for these instruments as alternative
of indexing government debt reasoning that it would allow to sovereign debt for Islamic countries.
for an improved risk sharing among debtor countries and
international creditors. The present paper is structured as follows: Succeeding the
introduction is the motivation and justification for the study,
Amongst the key proponents of this instrument is Bailey which is followed by an insight into the need for GDP-linked
(1983) who had initially suggested transferring the sovereign Sukuks, benefits and fiqhi aspect in Section  3. Section  4
debt into claims on the exports of the country. Delving provides some simple empirical results on how the economies
further on this, Krugman (1988) and Froot et  al. (1989) will benefit with these instruments, which is followed
studied the virtues of indexing sovereign debts to variables by discussion on the viability for investors in Section  5.
that are controlled by the country, partially or completely. Subsequently, the authors provide concluding remarks.

Shiller (1993) first proposed the concept of a true GDP-

linked bonds, which fundamentally were perpetual claims 2.  Motivation and justification of study
on a fraction of a country’s GDP. Similarly, Borensztein In this study, the primary focus of the authors is to
and Mauro (2004) also sought to review the case for emphasis on the benefits of employing GDP-linked papers
GDP growth-linked bonds, and Griffith-Jones and Sharma for Muslim countries that are emerging but debt ridden.
(2006) in their paper summarize the benefits of introducing With the substantial growth of Islamic finance over the
GDP-linked bonds. Working on Shiller’s concept, Obstfeld years, it becomes imperative to introduce instruments that
and Peri (1998) propositioned members of the European would benefit Muslim economies at large. In particular, the
Union to issue GDP-linked euro denominated securities. reliance of Muslim nations on high-interest rate borrowing
Likewise, Haldane and Quay (1999) argue in favour of has largely been neglected in Islamic finance, with little to
indexing sovereign debt to commodity prices. Dreze (2002) no research available on alternatives.
advocated using GDP-indexed bonds as an integral part of
restarting the debt of the poorest countries strongly. The motivation for this paper arises from the need of
evidence to support the practicality of linking real economic
While the idea of GDP-linked debt has been implemented variables to sovereign debt for OIC countries. This will
to a limited extent, it received renewed impetus after a further allow us to examine the efficacy of such instruments
wave of financial and debt crisis that flooded emerging on the economies of Muslim countries.
markets in the 1990s. Caballero (2002) has suggested the
issuance of copper indexed bonds during Chile’s financial This paper is amongst the few that have discussed the use
upheaval. He further recommends making the returns of GDP-linked instruments, primarily discussed by Diaw
of debt instruments dependent on external shock variables and Bacha (2011) who proposed Sukuk structures indexed
that trigger crises (such as exports, regional economic with real economic variables. The use of real economic
growth, etc.) arguing that most emerging markets are variable linked instruments, as proposed in this study, aims
affected by external shocks that they are not well-equipped to quell instability in Muslim countries and help stabilize
to withstand. government spending while reducing the impact of interest
borrowings.
Reverting back to our issue at hand, as most of Muslim
nations are severely debt ridden, a fundamental question Islamic finance theoretically has the capacity to provide
arises: Does Islamic finance have the aptitude to provide for an alternative to these Muslim countries that would allow
the borrowing and financial needs of Muslim governments? them to restructure their debt and continue growing as
Much has been written about the development and an emerging economy. Islamic finance should aim to be
innovation of Islamic finance over the years. Mirakhor et al. the principal financial system for Muslim countries, hence
(2011 and 2012) believes that risk sharing is the foundation raising the need for the development of new instruments
of Islamic finance and is instrumental in the development and products to meet the needs of Muslim economies at
of any financial products within the ambits of the Shariah. both a micro- and macro-level.

A dearth in the present literature found concerns GDP- This study is novel in this area as it is a humble attempt to
linked papers using the principles of Islamic finance. The initiate an empirical research-based argument to provide
potential benefit of using GDP-linked Sukuk was discussed policy makers and economists an alternative Shariah
by Diaw et  al. (2012) where he proposed structuring the compliant instrument as a replacement of the conventional
instrument based on forward ijarah. Mirakhor (2011) sovereign debt.
highlights the benefits of using macro-markets and similar
GDP-linked papers in his works. Similarly, the prospective In the current literature in conventional finance this area
benefits from using an equity financed budget for Muslim has not been studied in detail. Regarding Islamic finance,

174 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries

it is very rarely studied area. But considering the structure tightening and tax reforms and removal of subsidies provide
of Muslim economies, which are primarily indebted a further negative impact on growth. A fiscal tightening
economies, the issue of sovereign debt is of immense by the government would reduce the expenditure in the
importance. The objective of this study is to provide country influencing the industrial sector’s growth.
empirically evidence on the viability and economic benefits
of GDP-linked Sukuks. Similarly, an increase in tax and removal of subsidies would
reduce the disposable income of the people, thus reducing
the consumption in the country. Both of these results
3.  GDP-linked Sukuk end in amplifying the economic downturn. Calvo (2003)
highlights some of these issues in his literature on sudden
Need for GDP-linked Sukuk stops, where he argues that sovereign debt fails to serve as
GDP-linked Sukuk aim to provide a breathing room to a device for inter-temporally smoothing the impact of such
economies when they go into recession and on the other slowdowns.
side provide an automatic slowing-in mechanism when the
economy goes into high growth zones and into overheat.
Analysing the advantages of this instrument, we can find Key benefits of GDP-linked Sukuk
two sets of clear advantages from an economic point of In this section we try to highlight the key benefits GDP-
view for the Muslim countries. linked Sukuk offer the issuing country from an economic
perspective:
Firstly, since the payment of profit is linked to the actual
growth of the economy, it can reduce the likelihood of • In case the economy of the issuing country
crises. In comparison to the normal borrowing mechanism experiences a persistent low growth phase, the GDP-
that Muslim economies are subjected to via plain vanilla linked papers provide a cushion against mounting
sovereign bonds or multilateral agency debts, the payments debt as in conventional borrowing. Since the payment
are fixed, and any slowdown in economy results in a due on these papers would vary with the growth
ballooning of debt. rate, the Debt/ GDP ratio would only increase via a
deficit borrowing of the country. This would lead to
In existing literature, there is an immense amount of a less probability of default, and provide breathing
evidence on the sovereign debt’s position of dependence room for policy makers to expand fiscally or through
on the economic growth. A glance at the history shows tax cuts to boost the economy.
that slow growth has primarily underlined debt crises, • Usually emerging countries and HIPC countries
like the Latin American crisis of 1980’s and the near (where most Muslim economies fall) face extreme
complete default of the Heavily Indebted Poor Countries difficulty for borrowing in times of crises due to weak
of 1990’s. In economic theory and its applications, the economic numbers. With GDP-linked papers, since
ratio of external debt to GDP is a significant predictor of their liabilities would reduce, that would provide room
impending crises. Detragiache and Spilimbergo (2001) in for counter cyclical policies to boost the economy.
their study empirically show that an increase of 10% in the • As Barro (1995) highlights, a mechanism of linking
debt/GDP ratio has an impact of nearly 20% increment in GDP with the liability of the country can provide ease
the probability of the crisis. In another study by Easterly of maintaining smoother tax rates and provide basic
(2001), he argues with proof that a hundred basis point infrastructure to the population in times of crises
decline in GDP growth rate is associated with 1.5 times without increasing tax rates.
more debt rescheduling in the fifteen years that follow. • It has been observed, in the case of many Islamic
countries, as well as emerging economies, the
In recent times, Argentina in the global economic system tendency to go into frivolous expenditure in the times
has flirted with the concept of GDP-linked borrowings. of high economic growth, which may lead to overheat
Although the conventional GDP-linked bonds differ from of the economy followed by a recessionary phase to
the conceptual Sukuk proposed in this paper, the mechanics bring the economy back to equilibrium level. Amongst
of economics can be argued for a benchmark and as an the key Islamic countries, the case of Indonesia and
example. There appears to be opposing views on how much Malaysia, in the early 1990s is an example. The
the Argentinian government benefited from GDP-linked economic crash of 1997 affected the region bringing
borrowing, but economic numbers show a stable growth the growth number to negative in one year after a
and recovery for the economy. decade of double-digit growth. If the government is
using GDP-linked paper, in times of high growth the
Secondly, GDP-linked Sukuk, as an alternative to the payment on Sukuk will be higher as well, which will
conventional debt, reduce the need of procyclical policies, keep a check on excessive fiscal expansion.
which have stunted the growth of most Muslim countries
post crises. It acts as automatic stabilizers in the case of Any instrument is not viable for financial markets until it
an economic slowdown and in boom periods. Considering provides benefits to both sides involved in the transaction.
the example of Pakistan, post 2008stock market crash, Although in this paper we aim to provide a concept of a
and civil unrest owing to terrorism, the economy slowed possibility of using GDP-linked papers in Muslim economies,
down. Having immense amounts of external debt, which why it may work is provided in the following passage by
had a fixed interest payment due, the country had to revert listing the potential benefits to the buyer of this instrument.
to an IMF bailout plan, thus incurring another 10  billion
dollar loan to meet its prior commitments. The increase in • Diversification of investments: The economic cycles
debt and the bailout plan came with specific conditions of of different countries are far from being correlated
fiscal tightening and tax reforms. These policies of fiscal perfectly. Although there exists regional correlation

Eds. Hatem A. El-Karanshawy et al. 175

 Rizvi and Arshad

to some extent, but global perfectly correlated Islamic world comprise nearly 60% of the GDP of Islamic
economic cycles is a utopian scenario. The investors ex GCC countries. These four countries have experienced
in these papers can diversify their portfolio while economic crashes and have been under external debt for
keeping their investments in the real sector of the the past few decades.
economy instead of financial papers only.
• An application of GDP-linked papers will provide a For a simple review of the benefits of a GDP-linked
smoothing of the economic growth of the Muslim sovereign can be illustrated by taking the example of our
and emerging economies, reducing the probability sample countries. Suppose a simple musharakah based
and frequency of defaults. From an investor GDP-linked Sukuk was used for all financing of these four
perspective, a default is a high cost of litigation and countries from 1985 onwards, including conversion of the
restructuring. existing debt to this instrument. The country would have to
• While in developed countries, the stock markets pay a profit rate of whatever the growth of the underlying
provide an investment opportunity into the future business is – in this case the economy of the country. In any
prospects of the real sector, in Muslim economies, the year where the business (economy) does not grow, there
stock markets are still at nascent stage. Issuance of GDP- would be a zero payment of profit.
linked Sukuk will provide an opportunity to investors
to invest in the potential of the real economy. For the purpose of this example, we make the strong
assumption that the composition of the debt and the changes
in the debt levels does not have any impact on the behavior of
Insight into structure and Fiqhi issues any of the other variables in the economy during 1985–2010.
The proposal for a GDP-linked Sukuk has been raised in What repayments the countries would have had to make
the corridors of Islamic finance previously in a paper by and the interest cost savings is illustrated in the Figure  1.
Diaw et  al. (2011), where they have proposed an Ijarah The graphs show, what was the actual interest payment and
based GDP-linked Sukuk for public sector funding and debt forgiveness in each year as a representative of the cost
debt management. The proposal of this Sukuk takes into the country actually had to bear due to conventional mode
account this earlier study and is an attempt to further the of financing. In addition, the payments for GDP-linked
idea. Whereas the previous study proposes an ijarah-linked Sukuk are shown for the same period and the savings as a
paper, we propose a pure musharakah linked paper. percentage of the national income each year.

The idea of authors is to issue a musharakah paper by Looking at the Malaysian case, the Malaysian economy
the government as the wakil of the country, where the after a short decline in the mid-1980s bounced back and
underlying business is the economy as a whole. Keeping in grew at a robust speed for the second half of the eighties
mind the musharakah nature, where profit is paid on actual and first part of nineties. This was followed by the infamous
profit is the business, which in this case is the growth of Asian financial crisis of 1997, which brought the ASEAN
the economy. This idea is subject to Shariah debate and economies down to their knees in a matter of a year. With a
approval, but in the understanding of the authors if we couple of years of negative economic growth, the economy
consider the economy as one business, there should not be was propped back up with fiscal support policies and
a problem in sharing of the risk of the business (economy). capital controls. Indonesia experienced a similar crisis but
decided to take the multilateral agency bailout plan, and
A key insight of this model is that in case the economy goes both economies have experienced a revival in economies
to zero growth or negative growth in any year, the investors in the last decade.
would share in the losses, which would mean a reduction in
their principle amount invested in the business (economy). With liberalization policies in Indonesia, a massive influx
This sharing in risk of the economy via a musharakah paper of foreign investment supported by steady infrastructure
is different from indexation of the debt, since a debt is not development has resulted in high economic growth figures.
created when a musharakah Sukuk is shared, but is similar Within these two economies, in our example for using
to issuance of equity shares in the business (economy) of GDP-linked Sukuks, the graphs show the crisis would have
the country. resulted in zero payments during the crisis years, providing
room to the government for fiscal expenditure, to propel the
The issuance of indexation is a much-debated topic under economy further. Governments using GDP-linked Sukuk
the Shariah law, but is primarily focused on “debt” and not would have obtained a large reduction in the interest bill,
much has been discussed and debated on the issue of Equity leaving more room to avoid procyclical fiscal measures.
shares linked to GDP. Bacha and Mirakhor (2012), provide Large interest savings would also have applied during the
a strong argument in favor of development of similar GDP- sudden slowdown of 2008–2009. In the case of Malaysia,
linked securities, linking it to welfare of Ummah through the average profit payment over the 25-year period amounts
the concept of macro markets. to 5.99% as compared to 6.54% average interest payments
that were paid by the Malaysian government. In addition
to the lower average payment, over the period, Malaysia
4.  Empirical analysis would have saved nearly 14% of its national income. In the
case of Indonesia the difference between profit payment on
Empirical evidence for some Muslim economies GDP-linked Sukuk and interest payment is a mere 30 basis
The empirical section shows some primary basic economic points annually, but the savings as a percentage of national
numbers to analyze what benefit GDP-linked Sukuk may income would have been 5%.
provide to the Muslim economies. To make our analysis,
we take into account four countries, Malaysia, Indonesia, On the other hand, Turkey has had a unique economic
Iran and Turkey. These four countries throughout the growth case, where the real GDP has experienced high

176 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries

12% Malaysia 1985-2010 Savings in National Income 12%

GDP Linked
Interest Paid
10% 10%

8% 8%

6% 6%

4% 4%

2% 2%

0% 0%

–2% –2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
12% Turkey 1985-2010 Savings in National Income 12%
GDP Linked
10% Interest Paid 10%

8% 8%

6% 6%

4% 4%

2% 2%

0% 0%

–2% –2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
12% Indonesia 1985-2010 Savings in National Income 12%
GDP Linked
Interest Paid
10% 10%

8% 8%

6% 6%

4% 4%

2% 2%

0% 0%

–2% –2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

14% Iran 1985-2010 Savings in National Income 14%

GDP Linked
12% Interest Paid 12%

10% 10%

8% 8%

6% 6%

4% 4%

2% 2%

0% 0%

–2% –2%
2010
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009

Figure 1.  Interest savings over the economic cycle.

Eds. Hatem A. El-Karanshawy et al. 177

 Rizvi and Arshad

volatility as compared to other countries. With its strategic Table 1.  Standard deviation of Debt/GDP.
location as the bridging country between Europe and Asia,
Turkey has experienced internal economic crises during Iran Indonesia Malaysia Turkey
our sample period.
Without 0.087 0.267 0.122 0.061
The 1980–88 Iraq-Iran war benefitted the Turkish economy. GDP-linked
At that time, the Turkish economy increased trade with papers
both countries and became the supply route for Iraqi oil With 0.084 0.256 0.110 0.061
exports. The end of war in 1988saw a sudden slowdown GDP-linked
in the economy, which stabilized with the economic papers
policy of import substitution of the then government. The
economy again received a major blow during the Persian
Gulf War, with a nearly $3 billion loss due to reduced trade. Tables 2 and 3 represent the maximum and minimum value
Saudi Arabia, Kuwait, and the United Arab Emirates (UAE) of the Debt/GDP ratio for a comparative analysis of actual and
moved to compensate Turkey for these losses, and by 1992, simulated using GDP-linked Sukuk. For our sample countries,
the economy again began to grow rapidly, before it was we observe that during the 25 year period under study, the
plunged again into crisis in 1994, owing to government maximum levels of Debt/GDP would have been substantially
borrowings. Post crisis, the economy stabilized but lower for Indonesia and Malaysia by nearly 300 basis and 400
experienced sharp downturns due to military assault in basis respectively. In the case of Turkey and Iran, the maximum
Iraq, the economic slowdown of Europe and the financial level of Debt/GDP would have been considerably lower as
crisis of 2007. With this volatile nature of the economy, well. Same trend holds for minimum level of Debt/GDP ratio,
GDP-linked Sukuk, would have provided savings as a where the minimum levels would have been much less than
percentage of national income to the tune of 24% during the actual Debt/GDP ratios these countries experienced.
this period, while the average payment per annum for GDP-
linked paper is 4.77% compared to 6.83% average annual
interest paid by Turkey. Table 2.  Maximum value of Debt/GDP.
The fourth country in the sample, Iran, with its decade-long Iran Indonesia Malaysia Turkey
war with Iraq in the 1980s and subsequent international
sanctions owing to its nuclear research has experienced Without 39.11% 158.69% 77.48% 57.62%
an economic situation marked by dogged growth, where GDP-linked
it has been impacted by sanctions and restrictions on its papers
oil export. Iran has the largest Islamic financial assets in With 38.63% 155.50% 73.43% 55.78%
the world, and using the Musharakah based GDP-linked GDP-linked
Sukuk would have benefitted in aggregate savings of 6% papers
over national income while the average profit payments on
GDP-linked papers would have been 4.07% as compared to
actual interest payments of 5.30% annually that Iran paid
to external lenders. Table 3.  Minimum value of Debt/GDP.

Iran Indonesia Malaysia Turkey

Debt/GDP stabilization effect
To illustrate the benefits of the GDP-linked Sukuk, in Without 2.93% 27.57% 29.30% 32.80%
stabilizing the Debt/GDP ratio of a country we perform a GDP-linked
simple exercise. For our sample countries using the well- papers
established identity in fiscal economics: With 2.67% 28.34% 29.04% 31.35%
GDP-linked
Dt D  papers
= (1 + r - g t ) t-1  - St
Yt  Yt-1 
The earlier results highlight the potential benefits of having
where Dt is government debt, Yt is output, st is the primary GDP-linked Sukuk as a replacement of conventional debt.
surplus as a share of GDP, gt is the growth rate, and r is the Nevertheless, the illustrations in earlier tables have been
interest/profit rate on the debt. provided under strong assumption of keeping all factors
constant. In reality, the potential benefits may be large
Using actual data for the time period under consideration, owing to interacting nature of economic variables. We have
holding all other values of growth as constant (a very assumed that the primary surplus remained the same as
strong assumption), we simulate the new Debt/GDP levels actual every year in our sample period, but it is important
for these four economies. Table  1 reports the standard to note that while using this Sukuk fiscal policy economic
deviation as a volatility measure of the Debt/GDP ratio growth and in savings/increments in profit payment may
during that period. It is evident that using the GDP-linked respond. This would change the primary surplus and thus
Sukuk would reduce the volatility in the Debt/GDP ratio may amplify the benefits achieved by these instruments in
for all the countries minus turkey, where it stays similar. stabilizing the Debt/GDP ratio.
Although the difference in the standard deviation doesn’t
seem huge, but if it is translated in absolute dollar terms, it Primarily in years of sluggish or negative growth with
goes in excess of a US$1 billion for these countries. debt sustainability as a major concern, governments

178 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries

Table 5.  Correlation of GDP growth rates amongst selected few Islamic countries.

Bangladesh Egypt Indonesia Iran Malaysia Nigeria Tunisia Turkey

Bangladesh 1.000 0.369 (0.125) 0.321 (0.127) 0.070 0.431 0.225

Egypt 1.000 0.003 0.043 0.013 0.214 0.336 0.001
Indonesia 1.000 0.101 0.778 0.122 (0.096) 0.177
Iran 1.000 0.182 0.304 0.559 0.063
Malaysia 1.000 (0.009) 0.027 0.296
Nigeria 1.000 (0.020) 0.074
Tunisia 1.000 0.142
Turkey 1.000

tend to move towards an increase in the primary surplus, Secondly, the issue of risk exposure and diversification while
effectively implying that fiscal policy is procyclical. In our using GDP-linked Sukuk is an implementation issue. Studying
sample countries, this is evident when we analyze the the economic structure of different regions and considering
primary surplus during economic downturns. In the case that Muslim economies are spread far and wide across the
of Malaysia and Indonesia it was evident during the Asian globe, the economic cycles are not perfectly synchronized
financial crisis when Malaysia with an average deficit of and have low or negative correlations. If a mass issuance of
6% in 1997–98  moved to a 6% surplus in the succeeding the GDP-linked Sukuk is done by Muslim countries, from
2 years. This is a common occurrence in Islamic countries, an investor perspective a well-diversified portfolio can
which are primarily liquidity constrained and have to be constructed to hedge against over exposure to a single
impose these policies to maintain their credibility in the business cycle. Table  4, provides the correlation matrix for
international credit market. GDP growth rate amongst a selected few Islamic countries.

Our sample countries are highlighted in grey. It can be

5.  Viability for investors observed from the specific countries that most of them
By illustrating some of the ways GDP-linked Sukuk have a varying correlation in growth with each other.
can benefit the Islamic countries, within our sample Malaysia and Indonesia, being in the same region and
countries, the authors do realize that implementation having similar economic fundamentals, tend to have
of these instruments require a much more thorough and a high correlation while these two countries with Iran
rigorous empirical work. In addition to the fine-tuning of have a very low correlation. Looking at the selected few
these instruments, a major obstacle that can be foreseen countries, we observe that there are traces of negative
is the creation of markets for new instruments globally. correlation between Islamic countries as well, for instance,
Past experiences with the introduction of innovative in the case of Bangladesh with Malaysia/Indonesia, and
instruments in world credit markets has been met with Nigeria with Malaysia. This bodes well from an investor
skepticism and resistance. perspective for diversification and builds a good case for
these instruments.
Since GDP-linked Sukuk based on pure musharakah
structure is a novel idea in the Islamic financial landscape,
not much debate has happened on potential obstacles. 6. Conclusion
Nevertheless, owing to the global economic environment, This study focuses on the predicament of several Muslim
these instruments would have to be introduced in the countries that are debt-ridden and struggling to finance
world financial markets, which are dominated by the their debt. Most commonly, these countries turn to western
conventional investors. Due to this reason, we borrow financial intermediaries offering high interest rates to meet
some of the concerns of global investors from literature their debt requirements. Unfortunately, the reliance of
of conventional finance that seem to hold valid for these Muslim nations on high-interest rate borrowing has largely
instruments as well. been neglected in Islamic finance, with little to no research
available on alternatives.
A serious question is based on whether international
investors would be willing to directly expose their The development of financial instruments for sovereign
investments to volatility of GDP in Muslim and emerging financing has only just begun whereby a few seminal
countries. In the understanding of the authors, international studies started to show how Shariah compliant instruments
investors are already exposed to the economic growth can help alleviate Muslim nations from their current
of the country implicitly via the investments in the stock predicament.
markets and the standard debt contracts. If a country goes
into a recessionary phase, and debt sustainability levels Owing to fact that 51 out 57 OIC member countries are
are breached, the borrower country is likely to default or debt-ridden, the authors propose a solution to this problem
go into a restructuring program, which is a heavy legal found within the ambit of Islamic finance. Stemming from
and financial cost. The GDP-linked Sukuk tends to provide conventional finance, the authors recommend the use of
a buffer against the possibility of default and its costs by GDP-linked papers using the principles of Musharakah.
providing the investors a steady return that is bound to pay The paper highlights the key benefits of using GDP-linked
a higher amount in the long term as economies pick up. Sukuk arguing that such an instrument will allow for a

Eds. Hatem A. El-Karanshawy et al. 179

 Rizvi and Arshad

lower probability of default, which permits counter cyclical Barro RJ. (1995) Optimal Debt Management. Working
policies to boost the economy. Similarly, it leads to an ease Paper. NBER, 5327.
of smoother tax rates and curbs excessive spending during
Caballero RJ. (2002) Coping with Chile’s External
times of economic growth.
Vulnerability: A Financial Problem. MIT.
Covering Malaysia, Indonesia, Iran and Turkey, the Calvo GA, Reinhart CM. (1999) When Capital Inflows Come
authors attempt to further emphasis the benefits of GDP- to a Sudden Stop: Consequences and Policy Options.
linked Sukuks. In the case of Malaysia, using GDP-linked Center of International Economics, Department of
sovereign paper would have provided an average profit Economics. University of Maryland.
payment of 5.99% as compared to a 6.54% interest payment
Calvo GA, Izquierdo A, Talv E. (2003) Sudden Stops,
that was paid by the Malaysian government over the 25-
The Real Exchange Rate, and Fiscal Sustainability:
year period. This would provide more room for Malaysia
Argentina’s Lessons. NBER Working Paper.
to avoid procyclical fiscal measures. In addition, Malaysia
would have saved nearly 14% of its national income. Detragiache E, Spilimbergo A. (2001) Crises and Liquidity—
Evidence and Interpretation. IMF Working Papers 01/2.
Indonesia, on the other hand, only shows a difference of 30 International Monetary Fund.
basis points between the profit payments and interest rate
Diaw A, Bacha OI, Lahsasna A. (2011) Public Sector Funding
payments on GDP-linked sovereign paper and borrowings.
and Debt Management: A Case for GDP-Linked Sukuk.
Nonetheless, the savings on its national income would
Paper presented at the 8th International Conference
have amounted to 5%. The case of Turkey showed that due
on Islamic Economics and Finance, 19–21 December,
to the volatile nature of the economy, GDP-linked Sukuk
Doha, Qatar.
would have provided a savings of around 24% as a fraction
of national income. The average payment per annum for Diaw A, Bacha OI, Lahsasna A. (2012) Incentive Compatible
GDP-linked paper is 4.77% compared to 6.83% average Sukuk Musharakah for Private Sector Funding. ISRA
annual interest paid by Turkey. Lastly, Iran’s saving would International Journal of Islamic Finance. 4(1).
have been 6% over national income while profit payments
Dreze JH. (2002) Loss Reduction and Implicit Deductibles
would have been only 4.07% compared to the 5.30% that
in Medical Insurance. CORE Discussion Papers 2002005.
was paid out to lenders.
Université catholique de Louvain, Center for Operations
Research and Econometrics (CORE).
The authors further highlight that using GDP-linked Sukuk
would reduce the volatility in the Debt/GDP ratio for Durdu CB, Mendoza EG. (2006) Are Asset Price Guarantees
the countries. The maximum value of Debt/GDP ratio was Useful for Explaining Sudden Stops? The Globalization
considerably lower for all four countries, with Malaysia and Hazard-Moral Hazard Trade of Asset Price Guarantees.
Indonesia recording a reduction of nearly 400 bases and IMF Working Paper. WP/06/73.
300 bases respectively. Parallel trends apply for minimum
Easterly W. (2001) The Effect of International Monetary
levels of Debt/GDP ratio, where the minimum levels would
Fund and World Bank Programs on Poverty. Policy
have been much less than the actual Debt/GDP ratios that
Research Working Paper Series, 2517. The World Bank.
these countries experienced.
Froot K, Keen MJ, Stein, J. (1989) LDC Debt: Forgiveness,
While this study lays out an introductory emphasis Indexation, and Investment Incentives. Journal of
on the benefits of a GDP-linked Sukuk, it is not without Finance. 44(5):1335–1350.
limitations. The implementation of this study has been
Haldane A, Quah D. (1999) UK Phillips Curves and
restricted to only four countries; further research calls for a
Monetary Policy. Journal of Monetary Economics.
much larger sample study. Furthermore, this study makes
44(2):259–278.
an assumption that all interactions of economic variables
remain the same, while in reality a more detailed analysis Hasan MA, Siddiqui AN. (1992) Is Equity Financed Budget
would be better to see how using these instruments will Deficit Stable in an Interest Free Economy? Journal of
change the behaviour of growth and other economic Islamic Economic Studies. 1(2).
variables.
Krugman P. (1988) Financing vs. Forgiving a Debt Overhang.
Journal of Development Economics. 29:253–68.
This preliminary study is an emphasis on the benefits of
Islamic finance as an alternative for government financing. Obstfeld M, Peri G. (1998) Regional Nonadjustment and
The empirical evidence provided by this study is promising Fiscal Policy: Lessons for EMU. Center for International
and provides an avenue for further research. The plight and Development Economics Research (CIDER) Working
of Muslim nations with respect to their debt structure has Papers C98–096. University of California at Berkeley.
largely been ignored in Islamic finance. It is the opinion of
Shiller RJ. (1993) Aggregate Income Risks and Hedging
the authors that such instruments are vital for the support
Mechanisms, NBER Working Papers 4396. National
and development of Muslim nations and should be further
Bureau of Economic Research, Inc.
researched.
Zamir I, Abbas M. (2012) An Introduction to Islamic Finance:
Theory and Practice. 2nd Edition. Wiley Publications.
References
Zamir I, Abbas M, Hossein A, Noureddine K. (2011) Risk
Bailey N. (1983) A Safety Net for Foreign Lending. Business
Sharing in Finance: The Islamic Finance Alternative.
Week. January 10.
Wiley Publications.

180 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation
and financial engineering
Nadi Serhan Aydin, FRM1, Martin Rainer2
1
Organization of Islamic Cooperation (OIC), Ankara Centre – SESRIC, Institute of Applied Mathematics,
Middle East Technical University, Ankara, Turkey, Email: [email protected]
2
ENAMEC Institute, Institute of Applied Mathematics, Middle East Technical University, Ankara, Turkey,
Email: [email protected]

Abstract - Starting from the fundamental principles, whether a designed contract implies a clean
bay’ rather than ribā is central to our discussion. We argue that ribā and gharar may easily arise
through neglect of risk or inappropriate valuation methods for value and risk of assets and financial
instruments. This possibility, in turn, strongly necessitates an estimation of expected forward
values, market risk, and default risk, which is consistent with Islamic principles of avoiding ribā
and gharar. Based on consistent estimatation of risk and return, the expected costs of risks should
be quantified for all parties to the contract in order to judge, whether a contract is free from usury
(ribā) and evitable risk (gharar) and whether it the remaining inevitable risks are distributed fairly
between the counter-parties, for example between the investor (rabbu l-māl) and the entrepreneur
(mu-ārib) within a mu-āraba contract. Therefore, an unbiased quantitative estimation of the risks
and returns is desirable so as to put Islamic principles to work.

We argue that Islamic principles, in particular the avoidance of riba¯ and gharar, should be
applied to real economic value in the first place, and not a priori to a monetary value in terms of
conventional currency. In order to reconcile monetary value with economic value, we propose a
reference currency linked to an appropriate commodity basket, reflecting the common economic
realities and needs of the respective monetary union. Based on this currency, real economic value
can be computed in analogy with conventional financial engineering methods.

In order to reflect global economic needs and realities, a global reference currency should be linked
to a basket of commodities including in particular the natural resources necessary to ensure both,
sustainable survival of mankind and a sustainable living standard above poverty.

Referring to the recent financial crisis of the European Union, we argue that apart from the common
economic realities and needs within a given socio-political union, such as the OIC countries, also the
different realities and needs should be honoured appropriately. We propose a 3-level construction
of reference currencies, reflecting the economic realities and needs globally – for each region, and
for each country.

We compare conventional financial engineering, based on zero bond numéraires computed from
fixed income forward contracts, with Islamic financial engineering based on numéraires computed
from bay’u l-salam or/and forward contracts on the basis of the reference currencies relevant for
the counter-parties of the contract.

We propose that contract valuation and risk management should be performed on the basis of
Islamic financial engineering rooted on the reference currencies reflecting the economic realities
and needs relevant for the counter parties. Considering the benefits of such a risk management
for social economy, particularly when the Organization of Islamic Cooperation (OIC) countries
are considered, we argue that an implementation of the described Islamic financial engineering

Cite this chapter as: Aydin N S, FRM, Rainer M (2015). Concept and mathematics of Islamic valuation and financial
engineering. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation

Developing Inclusive and Sustainable Economic and Financial Systems

 Aydin et al.

enables Islamic risk management making transparent expected return and risks and enables their
fair distribution between the counter-parties.

Islamic financial industry following the guideline of such principles of Islamic financial eng­
ineering  would be able to contribute to sustainable development by (i) more risk-(and-return)-
consciousness reflected in the participatory structure of financial contracts, and thus, (ii)
encouraging Islamic financial institutions to innovate financial products consistent with real
implementation Islamic principles, reflecting the real necessities of modern business and economy.

1. Introduction appears: Iceland had its own currency, which helped to

soften the consequences, but Greece is currently trapped
Recent decades have witnessed an explosion in financial
within the EUR monetary union. Recently enterprises and
innovation and engineering of novel contracts. First of all,
banks already simulate scenarios for the reintroduction of
from the so-called “conventional banking system” emerged
the Drachma1 in Greece. The point that was missed, when
increasing the need to adapt to the challenges from the
Greece entered 2001 for the second time2 into a monetary
demand side. Secondly, particularly also within the so-
union, is that a common currency can be sustainable only
called “Islamic finance industry,” the incentive to innovate
within a political region that is operating sufficiently within
around prohibited and disadvantaged transactions has
its economic realities and needs. If some important factor
been unfortunately high. A clash between abused financial
such as economic productivity differs largely among the
engineering and Islamic principles seems to prevail, not
participating countries, an artificially introduced monetary
only since the renaissance of Islamic finance, and not in
union creates artificial fluxes and processes that may be
the Muslim world only. In medieval Europe canonical law
in contradiction to the economic reality of some country.
prohibited usury, i.e., ribā. European merchants, however, This recently became evident with Greece again. The idea
used a combination of contracts called contractum trinius that a monetary union of a region like Europe is a priori
to circumvent prohibition of usury. The construction going to strengthen its political union has proven again to
uses put-call parity to synthesize a conventional interest- fail, because in the case of Greece, the EUR did not reflect
bearing loan. During the recent renaissance of Islamic Greek economic reality, with economic mismatch finally
finance, this construction prevailed in practices of some so- resulting in political instability. We will come back to this
called “Islamic” banks, when they include into murābaha point below when we discuss the construction of reference
contracts a combination of conditions similar to contractum currencies reflecting real value of economy.
trinius, thus effectively ensuring and concealing a risk-free
profit, which might also be called “regulatory arbitrage.” On the background of financial crises in Europe, what
Exceeding considerably the inflation rate, such a so-called can be learned from Islamic finance? On the Banque de
“risk-free return” is nothing else but usury, i.e., ribā. Such France conference, March 4, 2011, Kenneth Rogoff of
a combination of contracts in order to circumvent Islamic Harvard University commented: “Western policymakers
principles may be considered as product structuring that is and economists often portray Islamic financial systems, with
a part of conventional financial engineering. their emphasis on shared risk and responsibility in lending,
as less efficient than western systems that put no strictures
Here and below, we will be concerned however not so much on debt. Yet one can equally argue that Western financial
with product structuring but instead with quantitative intermediation is far too skewed towards debt, and as a
valuation of financial instruments. Careful computation consequence generate many unnecessary risks.”
of expected value, expected return, and expected risks is a
prerequisite not only for risk management, but also already In the Muslim world, conventional agreed interest loans are
for thorough valuation of compliance of contacts with commonly objected for the fact that they avoid a sharing
respect to regulatory requirements including compliance participation in the default risk related to the purpose of
with Islamic law. A clean product structuring, therefore, the loan. For the same reason, transfer of risk by selling
depends on a serious valuation of value and risk as the more it to another party is often objected. Within the Islamic
fundamental and more demanding section of financial finance community, this sometimes appears to create a
engineering, which we will deal with subsequently. far spread impression that a passive attitude towards risk
would be more ethical rather than an active management
In Europe in particular, the growing public debt of several of the risks involved within a project and related contracts.
European Union (EU) member countries has surfaced We would like however to argue that both, transparency
already in several financial crises, which were escalated by about all risk involved, and clear agreements how they are
salvation programs for financial institutions. According to to be shared should be part of any contract. Also we believe
IMF 2010, the public debt of Italy and Greece is significantly that for a single agent, entrepreneur or investor, it should
exceeding annual GDP, while others (like Portugal, Ireland, be legitimate to use his portfolio of contracts with different
and Belgium) are following with over 90% of GDP. Continuing counter-parties in order to minimize his own exposures
conventional public loan policy, it is likely that European to different types of risks. In fact, avoidance of gharar is
public debt will exceed all possible salvation funds in the a fundamental request according to Islamic discourse,
long run. Comparing the 2008 financial crisis of Iceland e.g., numerous hadiths forbidding gharar sales. Having
with the 2010 crisis of Greece, some important difference accepted this, it is becoming clear that, modern tools of risk

182 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering

management and hedging should be applied appropriately providing the required capital for small and medium size
using all the knowledge we currently have in order to reduce companies in order for them to remain operational and
risk exposures. Quantitative evaluation of risks involved continue making essential contributions for an innovative
therefore is a necessary precondition in order to enable and vital economy. Following the financial crisis of 2007,
the counterparties to first obtain transparency about the new regulatory frameworks such as Basel III encourage
existing risks and, on the basis of this information, to reach credit institutes to impose more severe conditions and
a fair agreement about the mutual distribution of risks. As tougher rating conditions on entrepreneurs for credits. The
far as Islamic finance is concerned, we currently perceive regulatory requirements now became the pretence of credit
commonly loose interpretation of quantitative evaluation, institutes for harsher credit conditions, contradicting to
and last but not least, avoidance of mathematics as being too generally decreasing interest rates in Europe, particularly
sophisticated. However we’d like to remind that striving for in Germany. Government tried to intervene on this with
knowledge is any Muslim’s duty, rather than condemning several measures to enable and to push financial institutions
it plainly as misleading since the mathematics involved to follow up their duty of providing liquidity for enterprises.
appears too complicated for many non-experts. However, by the time being, Europe’s financial industry
reacts only very reluctantly. In this aspect, governments
One historically grown reason for the adverse attitude of of most OIC countries should be in a better position not
many Muslims towards financial mathematics and, only because the influence of governments on the domestic
in particular, the fair value approach comes from the financial sector has traditionally been more powerful, but
conventionally common practices of discounting future cash much more also because the basis of understanding between
flows with so-called “risk-free” interest rates often derived governments and financial industry is derived from the
from forward rates of inter-bank markets and certain common rules of Shariah. In this context, a very different
government bonds. Indeed these rates are deceptive. First culture of entrepreneurship has shaped the Islamic finance
because they are not really “risk-free” as their name suggests. industry with the commonly accepted participatory means
Secondly, their level is considerably higher than sustainable, of financing, fixed income products being obsolete.
because they are indeed driven by and related to fixed interest
loans. Moreover, the high level of these rates also contributes In the following, we will sketch some essential mathematical
to push inflation rates up. Such objections against the common aspects of Islamic finance enabling financial engineering
discount curves, based on the aforementioned interest rates, similarly easy and rigorous as in the conventional interested-
are fully legitimate. Below we will argue indeed for a more based finance.
flexible and adequate construction of discount curves. We
will also point out that the mathematical framework of
fair value—based on relative prices with respect to some 3.  The myth of risk-free interest
reference asset—does not require at all that the reference and fixed income
asset should be given by an artificial zero-coupon bond In this section we will argue that the notion of a risk-free
linked to “risk-free” discount rates, which, in turn, relate to fixed income is a myth kept up by the conventional banking
fixed interest loans and the conventional inter-bank markets sector, rather than an economic reality. Corresponding rates
for forward rates. The mathematical framework of fair value are set mainly by the agreement of an inter-bank market
works perfectly also with a universal commodity-linked among conventional creditors. To calibrate a current credit
currency or even equity as a reference asset. contract to such rates may be questionable particularly
for the situations where one or more counterparties of
the contract have limited or no access to this inter-bank
2.  Islamic finance for sustainable market.
development
The role of financial institutions is to provide the capital After we have seen that fixed interest rates by themselves are
for projects, in particular, for those projects that are useful not suitable as a basis for a risk-neutral measure, we conclude
or even vital for development of societies with sustainable that a new basis for such a measure must be sought. This
infrastructures. Notwithstanding, many financial institutions should be done by taking into account the position of the agent
(e.g., in Europe) that had been shaken by the last financial in the market, i.e., by carefully investigating her exposure
crisis, have been observed to refuse credits or offer credits to the various risks of the relevant markets and economies
only with unbearably high risk premiums. This attitude involved. Furthermore, credit spreads should be considered
has severely threatened the existence of in particular local more flexible, i.e., not constant, and not necessarily always
small- and medium-size family businesses and enterprises, positive. A negative credit spread for some period would
since these suffer from additional discrimination by the reflect the possibility that the credibility of the considered
traditional rating systems favoring large international enterprise is in fact better than that of the reference.
players instead. As it is known, however, the economic
and innovative power of the society is driven very much by It is one main issue in Islamic finance that a supposed
locally rooted small and medium size enterprises, rather risk-free interest rate should be close to zero. The reference
than the big transnational players. This is known to hold rates of some countries like Switzerland and Japan have
true in Europe, as well. And it is very likely to hold similarly already come very close to this, since many years. Also in
for the group of OIC member countries that have for long Europe, after the financial crisis of late 2000s, interest rates
been striving similarly for their sustainable development, dropped drastically. This indicates that the economical
while facing similar challenges along the way. reality in interest-based financial markets might essentially
honour – sooner or later – the fact that there is no free lunch
As a consequence, it is the task of each government to take in any market and that the assumption of a risk free return
care that the financial institutions follow up their duty of is a myth.

Eds. Hatem A. El-Karanshawy et al. 183

 Aydin et al.

4.  Inflation of currencies, time-value successive decoupling from their reference commodities
and time-less value simultaneously with their devaluation is described, e.g., in
El Diwany (2010).
To the extent that inflation of any currency is inevitable,
a time-value that compensates the expected inflation rate
Within the EU, after some period of fixed cross currency
should be disputable. If we decide that full participation
rates, the Euro was introduced. Less known however is that
in the inflation risk is not bearable to the investor, at least
an early predecessor of the Euro was defined already in
a partial compensation of hedging against inflation risk
the early 1930s. The universal European currency intended
should be admitted. On the other hand, one might argue
as a “currency for peace” was called “l’Europa” (Le Fédériste
that investor and entrepreneur perhaps share some common
(1933)). It was defined as a basket of several valuable
risks, such as the risks of everyday life, and accordingly, the
commodities. Much later, Lietaer (2001), introduced a
inflation risk – as it is inevitable to both parties – should
global Trade Reference Currency (TRC), dubbed also as
also be shared among them rather than being put on one
“Terra,” comprising a basket of a dozen internationally
party’s shoulders.
circulated currencies.
Real contracts usually involve several cashflows and/or
One advantage of the general concept of a basket of
depend on asset values at different times. A fair contract
commodities underlying to a reference currency is the
evaluation requires the ability to compare cashflows and/or
increased stability. With its currency linked to a basket
asset values at different times. The nominal value of assets
of commodities and its monetary authority backed 100
is usually measured in units of a certain currency. However,
percent by a sufficiently large reserve of these commodities,
the real economic value of this currency may changes with
artificial depreciation (appreciation) of a country’s local
time, e.g., due to inflation. For our fair valuation, we are
currency (domestic prices) would come to an end, together
nevertheless interested in the intrinsic real value of the
with an improvement in its immunity to the associated
currency, and of our assets. E.g., in the face of inflation, this
monetary and fiscal challenges facing its economy.
real value of assets might be measured by inflation-adjusted
prices. These are obtained by adjusting future cash flows by
An inevitable effect of any commodity-based currency
discount factors and account just for the expected inflation
is the increase of efforts for production of or mining for
of the currency.
the underlying commodities. Taking into account the
challenges of 21st century and beyond, previous choices
Viewed superficially, discount factors might be objected for
of baskets did not yet account sufficiently for the aspect
introducing undesirable time-value to cash flows. However,
of sustainability and desirability of the production of the
if they are chosen just such to compensate for the time-
commodities chosen for the basket.
dependency of the nominal value of the currency, they in
fact may yield in fact just the desired timeless real value
According to Islam, at times of the Prophet Muhammad
standard. We would like to emphasize that the timelessness
(pbuh), exploration and production of gold was still
of value as suggested by Islam has to be requested for the
desirable as the most precious commodity known at that
real value rather than the nominal value in the first place.
time. It was not yet challenged by an excessive mining
industry with all its social and ecological problems. From our
Islam advocates in fact the use of time-independent measures
current modern point of view, in our opinion, the negative
for value. At times of the prophet (pbuh) gold was much more
effects related to the intensive industrial production, such
rare than today, since there was no mining industry yet. It
as ecological damages and exploitation of workers, cannot
was a fairly good inflation-free currency. Its value was stable
be ignored. In essence, what is true for gold and other
and hardly to be influenced. Only exceptional political events
metals is also true for oil and gas as well as agricultural
such as conquests and sieges could trigger sudden regional
commodities. For any commodity, consideration should
gains and losses of huge gold treasures, which then could
me made whether and to what extent its production is
indeed change locally the real value. Except such extreme
desirable. So from an Islamic perspective, we would like to
events, the real value of gold was stable. Today, however,
argue in favor of a historically conscious reading that the
developments of the mining industry, demands from high
“gold” mentioned in the Holy Qur’an should be read just as
technology, and different market pressures apply almost
synonymous for the “most precious commodity.” If we ask
continuously on the value of gold in different directions.
ourselves – from the perspective of the global challenges
Although gold might be still more stable than most paper
ahead of us – what are these “most precious commodities”,
money, its real value has become much more volatile than
we might identify such commodities as cleanly produced
at times of the prophet (pbuh).
renewable energy, clean drinking water, and agricultural
products produced according to ecological standards.
Hence the challenge is to find the reference assets, which
represent a timeless stable standard of value, similar to
gold in previous times. 6.  Euro lesson: A multi-dimensional
currency system reflecting diversity
and community appropriately
5.  The gold of 21st century: Back to The Euro currency was introduced well-intended, with
commodity-based currency the idea to enhance the political and economical solidarity
From the very early times of Islam, up to the 20th century, between the countries involved and the ultimate goal to make
the commodities of gold and silver have played an important them together more competitive in the world markets arena.
role in defining the modern currencies. The histories of However, the recent Euro crisis has shown that as long as
the GBP and USD in this respect and, in particular, their the economic foundations of the constituents are basically

184 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering

too different, the hard entry of one or more countries into The fraction βw should be negotiated on a world conference,
a currency union may be a problem. If productivity of the common for all regions. It should be as high as possible
countries highly differs, the stronger countries have to aid in order to meet global challenges. However, the case
the weaker ones in order to avoid a potential collapse in the where βw < 1 is desirable in order to admit the differences
common currency. As a result, the political disputes going according to the particular situation of different regions of
along with this process not only put the common currency the world.
union at risk, but even worse may damage the political union,
as such concerns have been central to the Euro crisis. Similarly, the fraction βr should be negotiated on a regional
conference, common for all countries. It should be as high
For example, the EC currently is split about whether to as possible in order to meet common regional challenges.
introduce Euro bonds or not. The introduction of Euro However, βr < 1 is desirable in order to admit in this case
bonds would be consequent since there exists already a the differences according to the particular situation of the
common currency. The opponents however argue in view different countries of the region.
of the different economic realities within the member
countries. The dilemma is that, any currency has to reflect By construction, the higher the value of βw, the more
the economic realities, but the EUR member countries now strongly any two world currencies are correlated. The
discover that they will not be able to sufficiently converge same is true for  the currencies of the countries within a
on their economic realities. region: the higher the value of βw + βr, the stronger the
correlation between the currencies of any two countries
In our opinion, the choices for countries that belong to within a region will be. In particular, the negotiable weight
a certain political entity (e.g., the EU) to either share a βr may give to a region, e.g., Europe, the flexibility to
common currency 100 percent or withdraw from the regulate a partial entry into a currency union with some
monetary partnership, are insufficient in number in the face weight βr, realistically reflecting the degree of economical
of the current economic and political realities. The currently unification.
practiced “0 or 1” hard entries into a currency system
apparently ignore the risk that a similar hard exit from it
may endanger the whole system. Therefore, we propose a
structured system of world currencies, starting from a global 7.  Fighting usury: The risk-neutral
reference currency Cw. It is defined by a basket of commodities fair value
S1 … Sn whose productions are globally desirable, i.e.: The value of an asset can equivalently be quoted in the
form of a price or in the form of a rate, which is relative
n to a reference price. Hence, a quoted rate per se does not
C w = ∑ w i Si (1) yet imply ribā, while, vice versa, quoted prices may conceal
i=1
ribā. The form of the quotation, whether it is given price St,
or rate, Rt, is irrelevant to the question of usury.
Similarly, we may consider a basket Cr of commodities, the
production of which is desirable just within a certain region, The equivalence relation is:
e.g., the political entity of the OIC member countries. Cr
acts as the currency of this region. ∆Cr: = Cr − Cw accounts
S(t)
for structural differences of the region from the rest of the = 1 + R(t)⋅ t (5)
world. It may also give additional (non-negative) weight to S(0)
some of the commodities in Cw, e.g., clean drinking water,
which is already within the global world basket. It may also When St is unknown and subject to risk, it will be stochastic
contain new commodities specific for the region. Hence: and one has to consider its expected value EQ[St] under a
certain risk measure Q. The conventional present value (at
n m
time 0) of the stochastic variable S is then given by:
ΔC r = ∑ w(i r )Si + ∑w S (2)
(r )
i i
i=1 i=n+1
1
S(0) = E Q[ S(t)] (6)
Finally, we admit an extra basket for individual countries 1 + tR(t)
in order to take into account their particular situation,
differing from that of their region as well as the world at The spot rate R(t) may equivalently be converted into a
large: discount factor, corresponding to the initial value P(0, t) of
a zero-coupon bond with a nominal value of P(t, t) = 1 and
m l
time-to-maturity t, i.e.:
ΔC c = ∑ w (i c )Si + ∑w (c)
i S (3)
i
i=1 i=m+1
1
P(0, t) = (7)
The currency Cc of any country will be composed from a 1 + tR(t)
worldwide defined fraction βw of the world currency, the
regionally-defined fraction βr of its regional commodity The relation (6) above may then be written as:
basket, and the remaining local country fraction (1-βw-βr)
of its local commodity basket, i.e.:
S(0) E [ S(t)]  S(t) 
= Q = EQ   (8)
C c = β w C w + β r ΔC r + (1 - β w - β r )ΔC c (4) P(0, t) P(t , t)  P(t , t) 

Eds. Hatem A. El-Karanshawy et al. 185

 Aydin et al.

This is simply the martingale expression for the relative Note that this relation holds, independent of the auxiliary
price of S with respect to the zero-coupon bond P(·,t), currency units within which S and Ccountry are evaluated.
whose price is inversely proportional to the spot rate R(t). This invariance is an advantage of relative prices. In general,
the relative price in (8) will be stochastic. However, the
In order to determine whether a certain quoted price or relative price of Ccountry to itself becomes trivially constant
rate implies usury, it is rather important to have at hand an equal to 1.
independent method to determine a fair price with a risk-
neutral position. The expected price should be computed
under a measure Q, which in particular compensates 8.  Conventional versus Islamic valuation
any risk that is out of the responsibility of the considered Let us give first a simple description of a forward contract
counterparty. The computed value is supposed to match as and its conventional valuation. A forward contract agrees to
close as possible the realistic economic expectation. a fixed price: the fair forward price F to be paid at delivery
time T in exchange for an asset S (say a commodity) having
Conventional finance usually assumes depreciation of present value S(0) and unknown value S(T) at delivery.
cash flows, e.g., according to S(t), in a certain currency A fair value forward contract on some traded asset S should
during time. It further assumes that the appropriate way always have value 0 at time 0, the time of contraction.
to account for this is to discount with a particular spot rate Using relative prices w.r.t. zero bonds having price P(t, T),
curve R(t) derived from forward rates Fi (t) = F(t; Ti-1 , Ti ) conventional financial engineering demands:
corresponding to quoted interest rates from certain inter-
bank markets (e.g., LIBOR). S(0) F S(0)
0= - , or equivalently F = . (10)
From an Islamic finance perspective, the questionable P(0, T ) P(T , T ) P(0, T )
point here is the market from which the quotes are taken.
The method of calibration of the expected spot value to the Hence P(0, T) acts as a discount factor. Note that (10) implies
forward curve by itself agrees with the principle of fair value. a nominal value concept with respect to some conventional
currency but tries to compensate this by taking zero bonds
Even within the still very conventional setting of R(t) as numéraire. Above, (10) is postulated directly, with
and P(·,t) and relations (6) and (8), there is already the deterministic P(0, T). Note that, for independent unknown
flexibility to adjust the rate R(t) (and equivalently the stochastic prices S(t) of the asset and B(t,  T) of the zero
zero-coupon bond) to the any expected time-dependent bond, it holds
performance. It is not at all required that it is linked to the
conventional inter-bank markets for deposits or swaps.  S(t)  E P[ S(t)] E [ S(t)]
EP  = = P , (11)
If it is linked to performance of equity, commodity, or (as  B(t , T )  E P[ B(t , T )] P(t , T )
we will consider below) to currency, it may be consistent
with principles of Islamic finance. Also compensation of with deterministic prices P(t,  T)  =  EP[B(t,  T)], from
inflation of a currency may be considered as legitimate which (10) follows immediately, assuming F  =  EP[S(T)].
under circumstances that the counterparty cannot bear to The independency assumption may hold for equity or
be exposed to inflation risk, e.g. because it is too small and commodity assets, when it appears plausible that such
does not have the possibility to protect itself. In this case assets develop essentially independent from inter bank
R(t) could be linked to inflation rate. markets determining B(t,  T). When asset prices are non-
trivially correlated with the stochastic discount factors, the
Concerning relation (2) above, remarkably, in mathematical above derivation of (10) has to be replaced by:
finance the meaningful quantities are mostly relative.
In particular, price dynamics is investigated mainly with using  S(T )   S(t)  S(0)
the relative price Sˆ := S/S0 with respect to the price S0 of a F = E P[ S(T )] = E P   = EP  = , (12)
reference asset. Several desirable properties for the reference  B(T , T )   B(t , T )  P(0, T )
asset are usually postulated. Best stability of its value is one of
them. The more the price is stable, the lower its volatility, i.e., whence again the fair price of the future contract is simply
the less risk of change is connected to the asset price. In the S(0)/P(0, T).
past, the commodity gold was considered as the most stable
physical asset with the least volatility of its price. Let us now consider the valuation of a forward contract
within an Islamic financial engineering framework relative
Note, however, that considering “risk-free” zero bonds or to a reference currency C. According to (9) above, we
rates is unrealistic.3 Therefore, we propose, instead of the replace (12) by:
zero-bond linked to a deterministic rate R(t) to consider
another reference asset instead, reflecting the stochastic  S(T )   S(t)  S(0)
risk. We propose the commodity-linked currency (4) as a EQ 
C(T )  = E Q  C(t)  = C(0) . (13)
natural reference asset for this purpose. With this choice,    
relation (8) will be replaced by:
Unlike (12), where B(T, T)  =  1 by definition of the zero
bond, here C(T) is in general still unknown, stochastic.
S(0)  S(t)  Furthermore the stochastic forward price S(t) and the
= EQ   (9)
C country (0)  C country (t)  stochastic currency C(t) are unlikely to become either
independent or fully dependent at maturity. Then only

186 Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering

value that can be estimated is their relative value. The the 2007 financial crisis, it has been realized now within
fair relative value w.r.t. to the given currency is constant. the conventional banking sector that, because of the basis
For  the future contract hence the fair price should be spread included in quoted forward prices, the discount
specified as relative to the reference currency, simply as curve which is supposed to yield fair arbitrage-free prices,
S(0)/C(0). Remarkably, the fair price is quite analogous to can no longer be equal to the (inter bank) market forward
the conventional future price, which also can be viewed as curve, rather a 2 curve approach becomes necessary.5 The
the relative price w.r.t. a zero bond. interpretation is that the conventional inter bank markets
are biased, rather than risk-neutral.
Important from the perspective of Islamic principles is that,
in the conventional case the zero bond numeraire is a pure
interest based asset, which is linked essentially to inter 9.  Risk profiles and equivalent
bank markets, while in the Islamic case the numeraire is martingale measures
a currency, which we propose to be linked to a commodity Let us now consider the dynamics of the relative price,
basket supposed to reflect a more appropriate and relevant Sˆ(t) = S(t)/S0 (t). Since a future price S(t), t > 0, is subject
economic reality. to the risk of change, its expected value E Q Sˆ(t) is of
particular importance, since it is the most objective value
Now let us consider the special case of full dependence, one can assign to them. The expected value EQ depends
S(t) = a  · C(t), i.e., the traded asset is proportional to the on the measure Q related to the risk of change. Under the
reference currency (with constant factor a), and hence assumptions of no arbitrage and a complete market, there
proportional to the commodity basket defining C. Then exists a unique measure Q with a risk-neutral expectation,
(13) yields: i.e.:

 S(T )  S(0) E [S(T)] F E Q Sˆ(t) = Sˆ(0). (15)

EQ  = =a= Q = , (14)
 C(T )  C(0) E Q[C(T)] E Q[C(T)]

This is called the martingale measure.
and therefore F = a · EP[C(T)].
However, if the market is not complete, e.g., due to
Comparing the conventional with the Islamic approach, a illiquidity, extreme events, or other reasons, the martingale
practically important difference is that the zero bond prices measure Q and the derived arbitrage-free value are no
are computed conventionally purely deterministically from longer unique. In the case of an incomplete market,
available quotes of money market, future, and swap prices; it makes even more sense to consider also alternative
while for the reference currency and the commodity basket reference assets. In particular, the reference curve R(t) may
their forward prices are currently not easily available from be chosen according to a particular tailor-made dynamical
quotes. Their estimation is however possible via calibrated risk-aversion profile, agreed by the counterparties of a
stochastic processes. When the asset S(T) is non-trivially contract. A particular risk aversion profile corresponds to a
correlated with the reference C(T), then stochastic simulat­ certain utility function, which in turn may be used to select
ion of their relative value may be necessary in order to a particular martingale measure among several equivalent
determine the expected value in (13). This shows the need of ones. One might even consider time periods within which
an analogous stochastic forward rate model for commodities R(t) becomes negative.
in analogy to the conventional LIBOR market model.
The important issue from the Islamic perspective is the
We finally emphasize that, the numeraires in (12) and (14) mutual agreement about the risk profile implied by the
are not equivalent numeraires, i.e., they are not resulting choice of R(t).
from transformation of each other by the change of
numeraire theorem.4 They are rather the different choices Determining R(t) rather freely according to a mutually
for the primary numeraire, defining the martingale measure agreed risk-profile, rather than determining it from the
P, respectively Q, i.e., defining what is meant by fair value. corresponding reference currency, might be interesting
particularly for direct contracts between investor and
This fair value according to (14) becomes more than entrepreneurs, where both of them wave developed a
another arbitrary definition exactly when the reference common understanding of the risks involved in the project,
currency C reflects a real value linked to the real economy. and correspondingly have consciously both agreed on a
Only in this way the Islamic choice (14) becomes really particular non-standard form of the R(t) curve. Because
superior to the conventional choice (13). Otherwise, for the project under consideration, this choice is more
“value” would remain just a conventional concept. This suitable than the plain comparison with the value of the
is the case in the conventional case, since the interbank currency.
interest markets behind zerobond curves for conventional
currencies do in most cases not really reflect the economic The last mentioned alternative approach of direct
reality to which the counterparties are exposed. This contractual agreement on a risk-profile and corresponding
becomes evident particularly in situations, when the finance R(t) requires outmost transparency about the risks involved
industry meets some of their homemade crises. Then, the in the contract. Therefore, in cases where there is no detailed
conventional zero curves tend to exhibit artefacts from the understanding and clarity about risks and the counter party
conventional banking sector. In particular the bias towards attitudes appear to be rather indifferent, then it might be
short maturities introduces basis spreads into the forward preferable to stick to a standard choice for the reference
rates traded and quoted in the conventional sector. Since asset, namely our proposed commodity-linked currency.

Eds. Hatem A. El-Karanshawy et al. 187

 Aydin et al.

10.  Consequences for Islamic derivatives Notes

as hedging instruments 1. From 1831 to 2001 the Greek currency was the
We exemplify the consequences of our modified evaluation Drachma.
framework for the design of Islamic derivatives. It is 2. From 1869 to 1914 the Drachma was effectively coupled
commonly agreed that such derivatives should be permitted 100% to the union monétaire latine.
only for the purpose of hedging. 3. Even a conventional fixed income market in fact is not
risk-free, but suffers in particular also from the risk of
For example, in Jobst 2010 three examples of synthetic changing market interest rates.
instruments from asset-based investment finance have 4. See e.g., Brigo & Mercurio 2007.
been given. The valuation nevertheless has been based on 5. See e.g., Bianchetti 2009 for the theoretical background,
conventional measures of value using risk-free discount and Ametrano & Bianchetti 2009 for practial
factors. As a consequence, the call-put parity could be used implementation of the 2 curve approach.
to synthesize a conventional loan.

If the present value would be computed using our proposed

References
reference asset-based currency, the value of a certain Ametrano FM, Bianchetti M. (2009) Bootstrapping the
amount of this currency would be constant without any Illiquidity: Multiple Yield Curves Construction for Market
interest. Call- put parity can no longer be used to synthesize Coherent Forward Rate Estimation. SSRN Working Paper.
a conventional interest-based loan. Hence motivation for Anonymous. (1933) L’Europa – M nnaie de la Paix. Le Fédériste.
such constructions will become void. 1 January.
Al-Bashir M, Al-Amine M. (2005) Commodity Derivatives:
An Islamic analysis. In: Iqbal M, Khan T. (2005) (Eds.).
11.  Concluding remarks Financial Engineering and Islamic Contracts. Palgrave
We have shown that the mathematical framework of risk- MacMillan. New York. 58–98.
neutral valuation and arbitrage free contract prices may Brigo D, Mercurio F. (2007) Interest Rate Models–Theory
be applied for Islamic finance, similarly as in conventional and Practice, 2nd Edition. Springer Verlag, Heidelberg.
finance, provided some modification on some conventional
inputs. In particular, the conventional discount rates drawn Bianchetti M. (2009) Two Curves, One Price – Pricing and
from inter-bank markets forward interest rates have to be Hedging Interest Rate Derivatives Using Different Yield
replaced. Curves for Discounting and Forwarding. SSRN Working
Paper.
A rather mild modification for this goal consists in El Diwany T. (2010) The Problem with Interest, 3rd Edition.
replacing them by performance rates of underlying equity Kreatoc, Ltd. London.
or commodity. Shifting the reference is from fixed income
interest to equity or commodity, compatibility with Islamic Hamzawy A. (2003) Hatar al-’Aulama: Die Arabische
principles might be achieved. Globalisierungsdebatte – Eine Neuauflage der
Kontroverse über Moderne und Authentizität? Welt des
As a more challenging method, we propose to replace Islams. 43(2):173–213.
the conventional interest-based zero rate curve by an IMF. (2010) Data and Statistics. World Economic Outlook
alternative reference asset. We propose in particular a Data Base. July 2010. Available at: http://www.imf.
commodity-linked multi-dimensional reference currency org/external/pubs/ft/weo/2010
system. Globally negotiated fraction βw of the world part,
and regionally agreed fraction βr may provide easy political Jobst A. (2010) Derivatives in Islamic Finance. Paper
instruments to regulate the degree of required homogeneity presented at SSRN. Preprint.
of currencies globally or within a defined region such as Latif AM, Janahi AR. (2005) Shariah Alternatives to
Europe respectively. Government Bonds. In: Iqbal M, Khan T (Eds.). Financial
Engineering and Islamic Contracts. Palgrave MacMillan.
The most challenging approach is nevertheless the one that New York. pp. 99–122.
leaves also the most contractual freedom and responsibility
at the same time to the contracting counterparties. Provided Lietaer B. (2001) The Future of Money. Random House.
all counterparties have both, the necessary capability for London.
risk analysis and commitment for risk transparency then Qutb S. (1949) Al-Adala al-Ijtima’iyya fi’l-Islam. Translation
they might agree on a particular zero rate curve freely, by Hardie JB (1980). Social Justice in Islam. Octagon
according to their commonly-agreed risk aversion profile. Press. New York.
The zero rates in this case may be linked closely to their
common judgement of risk for the project underlying to the Rogoff K. (2011) Summary of Comments. Banque de France
contract. The zero rates in this case may be even negative, Conference. March 4.
meaning that a later cash flow is strategically preferable to Schacht J. (1982) An Introduction to Islamic Law. Clarendon
an earlier one. Press. Oxford.

188 Islamic banking and finance – Essays on corporate finance, efficiency and product development