Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 5

UDK: 336.774.5(497.6)
DOI: 10.2478/jcbtp-2021-0021

Journal of Central Banking Theory and Practice, 2021, 3, pp. 5-22

Received: 02 August 2020; accepted: 06 January 2021

Amila Žunić *, Kemal Kozarić **, *

School of Economics and Business,
Sarajevo, Bosnia and Herzegovina
Emina Žunić Dželihodžić ***,
E-mail:
[email protected]
Non-Performing Loan **
School of Economics and Business,
Determinants and Impact of Sarajevo, Bosnia and Herzegovina

COVID-19: Case of Bosnia and E-mail:

[email protected]
Herzegovina ***
Central Bank of Bosnia and
Herzegovina, Sarajevo,
Bosnia and Herzegovina
Abstract: The aim of this paper is to investigate the determinants E-mail:
of the movement of non-performing loans in the banking sector of [email protected]
Bosnia and Herzegovina, as well as the impact of the COVID-19
pandemic on them. For this purpose, secondary data from the bank-
ing sector of BiH were used, over which a multi-regression analysis
was performed. The variables non-performing loans, GDP, loan loss
provision, and dummy variable COVID-19 were used in the analysis.
The results of the analysis showed a significant influence of all three
mentioned variables. The existence of a significant positive link be-
tween non-performing loans and the state of the country's economy
has been proven. In addition to the above, it has been proven that the
variable COVID-19 has a delayed effect on NPLs, due to the current
application of the moratorium on loans.

Keywords: NPLs, GDP, COVID-19, loan loss provision.

JEL Code: G00, G21.

Introduction

The growing turbulence of the financial market inevita-

bly leads to the strengthening of banking risks. Against
such backdrop, achieving a profitable business becomes a
very difficult and challenging endeavour. Therefore, bank
6 Journal of Central Banking Theory and Practice

managers are tasked with creating an adequate plan and policy for controlling
overall risks. Management is responsible for establishing balance and ensuring
safe, stable, and profitable operations. There are various definitions of risk in the
literature, one of which was given by Vaughan and Vaughan (1995) who define
risk as: “A condition in which there is a possibility of a negative deviation from
the desired outcome we expect or hope for. Therefore, we can say that in order for
risk to exist in financial operations, it must: be possible, cause economic damage,
be uncertain, and be accidental."

Since loans are the riskiest part of bank assets, their quality is one of the most im-
portant determinants of business stability and success. Although there is still no
universally accepted definition of bad loans, the most commonly used definitions
are those of the International Monetary Fund, the Basel Committee on Bank-
ing Supervision, and the Institute of International Finance. As defined by the
Basel Committee on Risk Management, non-performing loans include all loans
that are not collected within 90 days from the due date. If we take into account
that one of the primary activities of all banks is the granting of loans, then the
importance of banks' exposure to credit risk and its management becomes clear
to us. Accordingly, several methods have been developed to manage this risk.
They have best proven their effectiveness/inefficiency during financial crises. In
this regard, there is more and more talk about bad loans and ways to solve them.
Banks are getting closer to restructuring bad loans and clearing their balance
sheets to create the preconditions for new credit growth.

The global spread of the coronavirus is a human tragedy that is taking place
around the world, and Europe is considered the current epicentre. The corona-
virus pandemic has caused an unprecedented shock to the economy and many
businesses are experiencing difficulties in maintaining business. Not all of these
businesses will survive nor will all households be able to repay the loans they
took out before or during the pandemic. This means that an increase in non-per-
forming loans is inevitable because a part of the loans will not be repaid in full.
In order to minimize this increase, the European Central Bank (ECB) constantly
emphasizes that banks should lend only to those clients who are likely to repay
loans, even in difficult times. It also reminds banks of the need to closely monitor
risks in order to identify non-performing loans early enough and start tackling
the problem.

The choice of this topic was driven by the importance of bad loans that proved
to be a major threat to the financial system. Namely, according to reports on the
financial stability of central banks, the biggest threat to the financial stability of
the banking sector is represented by non-performing loans. This paper analyses
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 7

the impact of the Covid-19 pandemic on the problem of non-performing loans

(NPLs), which includes the impact of loan loss provisions and real GDP growth
rate.

The initial assumption is that these variables will have a significant impact on the
level of non-performing loans. Therefore, this research should show whether and
how these variables are related. In fact, with the help of an appropriate econo-
metric methodology, it will be possible to find out whether these variables move
together. Therefore, from the following review of the theoretical and empirical
literature, it can be concluded that it is important to provide empirical evidence
on the impact of these variables on non-performing loans.

Literature review

The banking sector is the dominant part of the financial system and as such rep-
resents a significant factor in the development of the overall economic system. In
the past three decades, studies of the banking sector of transition economies have
become more important due to banking reforms that included liberalization, pri-
vatization, and recapitalization of the banking sectors in the region (Grubišić,
Kamenković, & Kaličanin, 2020).

By analysing the available literature and works of other authors, a better insight
into the research state of this field was gained. It was investigated what the au-
thors of similar intentions wrote in the field of credit risk, non-performing loans,
cost efficiency, and global banking. It was found that there are three directions
in which the research of authors in this field moved, namely: analysis of micro-
economic determinants, analysis of macroeconomic determinants, and a com-
bination of micro- and macroeconomic determinants of non-performing loans.
In addition to the above, the impact of COVID-19 on the movement of non-
performing loans, and the reaction of commercial banks to these events were
presented.

One of the first papers in this area addresses the quality of loans and their impact
on credit growth. Two measures of credit quality were investigated: the ratio of
indebtedness to total loans and the ratio of non-performing loans and total loans.
These measures are good indicators of the likelihood of bad credit (Clair, 1992).
On the other hand, the authors investigated non-performing loans and efficiency
in commercial banks. They tried to prove the connection between credit quality,
cost efficiency, and capital. The results showed that non-performing loans lead to
lower-cost efficiencies, that better cost efficiencies reduce non-performing loans,
8 Journal of Central Banking Theory and Practice

and that low-capitalized banks have greater problems with non-performing loans
(Berger & DeYoung, 1997). Other authors investigated the impact of risk and
qualitative factors on efficacy. It was found that the level of financial capital has
the greatest impact on the possibility of financing non-performing loans. The
level of equity was used instead of the capital/asset ratio to monitor differences in
risk preferences (Altunbas, Lin, Moyneux, & Seth, 2000).

The following research is focused on the relationship between non-performing

loans, short-term deposit rates, and the impact of interest rates on loans. It has
been shown that the share of non-performing loans does not affect short-term
deposit interest rates. But that it does affect interest rates on loans, and in a posi-
tive way. The interest margin is also positively related to interest rates on loans.
The paper used the VAR (Vector Auto Regression) model for analysis on the ex-
ample of 28 banks from Korea (Chang, 2006). Furthermore, the authors inves-
tigated the critical factors responsible for non-performing loans on the example
of the Indian banking sector. Primary data collected from bank credit managers
were used. They showed that external factors have a stronger influence than in-
ternal ones (Sanjeev, 2007).

In addition, the hypothesis of poor management has been proven, which refers
to the fact that poor management leads to poorer cost efficiency, which results in
higher rates of non-performing loans (Podpiera & Weill, 2008). When analysing
the problem of high levels of non-performing loans, Johnson (2009) showed that
restructuring and clearing the balance sheets of large banks is the only sure way
out of the situation in which the entire financial system finds itself.

Subsequent research focuses on the conflict between bank managers and owners
over the topic of risk. It has been proven that risk-taking by banks correlates posi-
tively with the comparative strength of shareholders and the capital structure.
Regulations such as capital requirements or deposit insurance affect managers
and owners differently so the intensity of risk also depends on the strength of
the influence of one or the other. It has been shown that the relationship between
risk and regulation depends on the ownership structure of the bank (Leaven &
Levine, 2009).

Furthermore, the authors defined a model that has as its subject the cumulative
effects of macroeconomic shocks over a three-year time horizon and their impact
on non-performing loans. In this regard, the VAR model on the panel data was
used to verify the set model (Espinoza & Prasad, 2010). The effects of credit risk
spillover between banks in one banking system, in this case the Mexican banking
system, were further investigated. It has been proven that there is a spillover effect
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 9

between small and large banks, as well as that it is two-way, which is surprising
given that a one-way connection was expected (Herreriasa & Moreno, 2011).

In addition, the impact of risk management on non-performing loans and profita-

bility of Pakistan’s banking sector was investigated. It was found that an adequate
risk protection mechanism is not applied. It was concluded that non-performing
loans are growing due to the lack of an adequate prevention and control meth-
odology (Haneef et al., 2012). The following research focuses on non-performing
loans in CESEE (Central Eastern Southeastern Europe) countries and their mac-
roeconomic determinants.

The importance of conducting stress tests to support the monitoring of financial

stability and interconnections in the financial system was further explored. This
paper points out those short-term and long-term goals of stress tests that need
to be addressed (Bookstaber et al., 2014). According to Cucinelli (2015), where
banks avoid failure, NPLs impact negatively on a bank’s cost structure and ef-
ficiency and their willingness to lend. Furthermore, according to Balgova, Nies,
& Plekhanov (2016), “as high levels of NPLs at some point arrest credit growth,
the NPL ratio starts rising rapidly, while in “no NPL problem” countries it re-
mains stable due to the healthy growth of credit and nominal GDP.” In their
research, Radivojević & Jovović (2017) suggest that “the GDP has a crucial de-
terministic role for the NPLs ratio unveiling that the state of the economy of
(selected) emerging countries is linked to bank asset quality.” On the other hand,
some authors have found out that prompt corrective action has a major impact on
banking sector stability and decrease in the level of non-performing loans (Jiang,
Kanas, & Molyneux, 2018).

Unemployment, public debt, credit growth, lagged values of non-performing

loans, the cost to income ratio and financial crises has positive effect on nonper-
forming loans (Bayar, 2019). In their paper, Kozarić & Žunić Dželihodžić (2020)
analyse the impact of macroeconomic conditions on non-performing loans and
financial stability in Bosnia and Herzegovina`s banking sector. The results have
shown that improvement in macroeconomic conditions causes improvement in
credit quality. Also, it was disclosed that better macroeconomic conditions ensure
better conditions for the maintenance of the banking sector`s financial stability.

During the COVID-19 pandemic NPLs identification has been prolonged by

banks in an effort to delay recognition in the profit and loss statement, and to
conceal the loss of capital. This behaviour may lead to continued financing of
non-viable firms, the so-called zombie lending, and delay much-needed restruc-
turing efforts at the firm level with negative consequences for economic growth.
10 Journal of Central Banking Theory and Practice

If forbearance measures are kept in place for too long, this issue is further exac-
erbated and more desirable measures, such as internal workouts or the transition
to more market-based solutions are prevented (Kasinger et al., 2021). Effect of
changing asset quality on risk weights is present for the internal ratings-based
approach, in line with our expectations based on regulatory standards (Brož &
Pfeifer, 2021). The following paper reaches a conclusion that climate change can
adversely affect balance sheets of financial institutions. Therefore, climate change
is a source of financial risk and thus a part of the mandate of central banks and
supervisors in preserving financial stability (Fabris, 2021). The findings of the
following paper reveal the presence of significant positive long-run relationship
between ROA, ROE, trade openness, and human capital, while government ex-
penditures have negative impact on financial stability. Trade openness, human
capital and government expenditures can keep the financial system stable as a
whole (Zeqiraj, Iskenderoglu, & Sohag, 2021).

Figure 1: NPLs movement The figure 1 is a graphical representa-

tion of non-performing loans in the
period from 2003 until the end of
2020. After the growth of NPLs re-
corded after the outbreak of the Glob-
al financial crisis, their decline fol-
lowed from 2015 until the end of the
analysed period. This movement of
NPLs is one of the reasons for analys-
ing the causes of the decline in their
level s in the banking sector of Bosnia
and Herzegovina.
Source: Authors` interpretation

Hypothesis development

A review of the literature was conducted to identify the hypotheses relevant to

the achievement of the objectives of this research. Therefore, based on the early
literature and variables of study, the following hypotheses are formulated:

H0: Non-performing loans level is determined by economic conditions and it is

affected by crisis occurrence

Historically, the occurrence of banking crises has often been associated with a
massive accumulation of non-performing loans which can account for a sizable
share of total assets of insolvent banks and financial institutions, especially dur-
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 11

ing episodes of systemic crises. Deterioration in banks’ loan quality is one of the
major causes of financial fragility. Past experience shows that a rapid build-up of
bad loans plays a crucial role in banking crises (Demirgüç-Kunt & Detragiache,
1998, and González-Hermosillo, 1999). In their research, Turan & Koskija (2014)
examine the continuous increase of NPLs in Albania resulting from problems
caused by the economic crisis and the decrease of emigrant incomes. Namely, it
has been shown that non-performing loans began to grow immediately after the
outbreak of the financial crisis in 2008, but sharp growth occurred a year later
when GDP in most CESEE countries experienced a contraction. In addition to
these macroeconomic factors, the relatively high variability in non-performing
loans indicates the existence of a negligible contribution of bank-specific factors
(Klein, 2013).

H1: Better macroeconomic conditions measured by GDP lead to lower NPL level

Among the macroeconomic determinants, the results suggest that higher unem-
ployment rate, exchange rate depreciation (against the euro), and higher infla-
tion contribute to higher NPLs, while higher GDP growth results in lower NPLs
(Klein, 2013). Also, Babouček & Jančar found evidence of a positive correlation
between NPLs, unemployment rates, and consumer price inflation, while GDP
growth decreases the NPL rate.

H2: Higher riskiness of credit placements measured by credit loss provisions (res-
ervations) leads to a higher level of NPLs

Keeton (1999) showed that rapid credit growth, which was associated with lower
credit standards, contributed to higher loan losses in certain states in the US.
The result in the research by Seyoum, Nigussie, & Tesfay (2016) depicts that weak
credit risk management and easily admitted borrowers cause occurrences of non-
performing loans.

H3: COVID-19 crisis has postponed effect on NPL levels

An increase in non-performing loans, deterioration of the quality of the loan

portfolio, and possible extreme cases of bank runs are mentioned as the main
negative consequences of COVID-19 for banks (Goodell, 2020). It can be expect-
ed that many of the company loans will become non-performing in the coming
months. “What we have seen after the Global Financial Crisis and the Banking
and the Sovereign Debt Crisis in the Eurozone was that the share of non-per-
forming loans tends to be higher in countries with inefficient insolvency systems
and tends to be lower in countries with more efficient insolvency systems” (De-
mary, 2021).
12 Journal of Central Banking Theory and Practice

Research methodology

In this scientific research we used formulation and presentation of research re-

sults, combinations of numerous general scientific methods such as analysis and
synthesis methods, inductive and deductive methods, abstraction and concre-
tization methods, specialization and generalization methods, and comparison
methods. In the data collection, we used analysis of the content which is used to
research scientific - theoretical knowledge, relevant literature, and modern busi-
ness practice. For the analysis, the following variables were used: non-perform-
ing loans, loan loss provisions, GDP, dummy variable COVID-19. The analysis
will be performed on quarterly data on these variables for the period 2012-2020.
The sources of data were websites of the Banking Agencies and the Central Bank
of Bosnia and Herzegovina. Namely, the goal was to determine the impact of
the loan loss provisions, nominal GDP, as well as the impact of the COVID-19
pandemic on the level of non-performing loans. The data used in this study are
presented quarterly. This time series is representative because it covers the pre-
pandemic and pandemic periods. Thus, it was possible to detect the exact effect
in moment of COVID-19 appearance. The statistical program SPSS (Statistical
Package for the Social Sciences) was used for quantitative analysis.

Although we primarily planned to include the above variables in the model, pre-
liminary analysis showed that some of them are not significant for modelling in
the banking sector of Bosnia and Herzegovina, given their specifics. Therefore,
the following variables are included in the model: nominal value of GDP, loan
loss provisions, and impact of the COVID-19 pandemic.

Before carrying out the econometric analysis, descriptive data analysis had been
provided as well as the verification of normality and symmetry distribution and
the existence of multicollinearity problem. Since the model involves time series,
the problem of autocorrelation may exist, but previous studies have shown that
this approach is suitable for NPL time series analysis as long as the problem of au-
tocorrelation is solved. Therefore, linear regression model in statistical literature
is defined as follows (Somun-Kapetanović, 2008):

, (1)

Where y is the explained variable and, in this case, we will have model with de-
pendent variable: non-performing loans. α is constant term and x is nxp matrix
of independent variables, in this case: non-performing loans, loan loss provi-
sions, nominal GDP, dummy variable COVID-19, and ε is error term.
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 13

Furthermore, the correlation between the variables was determined by applying

the Pearson's Coefficient Correlation with the following rules:
• Very strong positive / negative correlation (+/- 0.81 to +/- 1)
• Strong positive / negative correlation (+/- 0.61 to +/- 0.80)
• Moderate positive / negative correlation (+/- 0.41 to +/- 0.60)
• Weak positive / negative correlation (+/- 0.21 to +/- 0.40)
• No connection (+/- 0.00 to +/- 0.20)

If the sign of the correlation coefficient was positive, then a positive correlation
would have existed. If the sign of the correlation coefficient was negative, then a
negative correlation would have existed (Sedgwick, 2012). Coefficient correlation
can be measured as follows (Somun- Kapetanović, 2008):

, (2)

where:
(3)

(4)

(5)

Multiple regression analysis will be applied to the collected data to obtain a spe-
cific model. The general form of the multiple regression model can be presented
as follow:

, (6)

where:

y - dependent variable,
x1, x2, ..., x k - independent variable,
β0, β1, β2, ..., βk - estimated parameters,
ε - random error (Bahovec & Erjavec, 2009)

Based on these two statistical methods, a final judgment will be made on the im-
pact of independent variables on the dependent one. Using the synthesis method,
all individual conclusions will be combined into one global conclusion. The um-
brella method that will unite everything that has been done so far is the method
of generalization.
14 Journal of Central Banking Theory and Practice

Research results

At the beginning of the analysis, normality was tested to determine whether the
data followed a normal distribution. After we determined the normality of the
data by the Shapiro-Wilk test (p = 0.043, confidence level 0.1), we can perform
multiple regression.

Table 1: Tests of Normality

Kolmogorov-Smirnova Shapiro-Wilk
Statistic Df Sig. Statistic Df Sig.
NPL .115 36 .200* .938 36 .043
Source: Authors` interpretation

Furthermore, the table below shows a description of the statistics for the observed
variables. Table 2 shows the mean values and standard deviations as the average
linear deviation around the mean values for each model variable.

Table 2: Descriptive Statistics

Mean Std. Deviation
NPL 11.468 2.943
LOAN LOSS PROVISIONS 4906.436 377.208
GDP 7375251.416 599163.708
COV_19 .111 .319
Source: Authors` interpretation

Namely, the purpose of the next part of the analysis reveals the potential correla-
tion between dependent and independent variables. To determine the correlation
between the variables of this model, we used the Pearson correlation coefficient.
The following table shows the results of the correlation analysis.

Table 3: Correlation matrix

NPL LOAN LOSS PROVISIONS GDP COV_19
NPL 1.000 .032 -.700 -.606
Pearson LOAN LOSS PROVISIONS .032 1.000 .339 -.176
Correlation GDP -.700 .339 1.000 .232
COV_19 -.606 -.176 .232 1.000
NPL . .042 .000 .000
Sig. LOAN LOSS PROVISIONS .042 . .022 .0153
(1-tailed) GDP .000 .022 . .087
COV_19 .000 .015 .087 .
Source: Authors` interpretation
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 15

The results of the correlation analysis indicate a positive and statistically signifi-
cant correlation between the level of non-performing loans and loan loss provi-
sions. This link was expected as riskier placements required higher provisions
for loan losses. Consequently, it is more likely that these loans will go into the
category of non-performing. The variable GDP has a strong negative correlation
with non-performing loans. This connection was expected since better economic
conditions and stronger economic activity create conditions for better business,
which is reflected in the creditworthiness of customers and, ultimately, in the
level of non-performing loans. Although the growth of non-performing loans
is expected during a crisis, in our case the variable that indicates the COVID-19
pandemic has a negative connection with non-performing loans. Thus, although
there has been an evident slowdown in economic activity and the emergence of
many problems in business cycles caused by COVID, this effect is still not visible
on non-performing loans. Moreover, according to the current data, it is the op-
posite of what was expected. The reason for these results is the fact that the pos-
sibility of a moratorium was introduced, which “saved” many loans from being
classified as low-quality. Therefore, a delayed effect is expected for this variable
after the moratorium expires, which represents the potential for future research.

Table 4: Model Summary

Model R R Square Std. Error of the Estimate Durbin-Watson
1 .852 .726 1.612 1.130
Source: Authors` interpretation

Furthermore, the Model summary and ANOVA are presented. In Table 4 we can
see that the correlation coefficient (R) is 0.852, and on the other hand, the deter-
mination coefficient (R Square) is 0.726. The obtained results tell us that 72.6% of
the change of the dependent variable (NPLs) can be explained by the change of
independent variables (GDP, LOAN LOSS PROVISIONS, COVID-19). Further-
more, the Durbin-Watson coefficient is within the normal range (between 1.5
and 2.5) and amounts to 1.130, which means that we do not have a problem with
autocorrelation.

Table 5: ANOVA
Model Sum of Squares Df Mean Square F Sig.
Regression 219.958 3 73.319 28.228 .000
1 Residual 83.117 32 2.597
Total 303.075 35
Source: Authors` interpretation
16 Journal of Central Banking Theory and Practice

As it can be seen, Table 5 – test shows us that the model is good and significant.
Table 6 shows the coefficients of the independent variables and the VIF model. As
we can see, all the variables of this model are statistically significant. Also, VIF
amounts to 1.1224, 1.254, and 1.145, which means that the model has no problem
with multicollinearity.

Table 6: Coefficients (Statistical significance of variables)

Collinearity Statistics
Model T value Sig.
Tolerance VIF
(Constant) 6.853 .000
LOAN LOSS PROVISIONS 1.789 .083 .817 1.224
1
GDP -6.411 .000 .797 1.254
COV_19 -4.239 .000 .873 1.145
Source: Author interpretation

The results of the performed regression analysis suggest the presence of the mod-
el with the analysed variables as follows:

NPLs = 28.96 + 0.01LLP - 14.87GDP - 3.88COV

The obtained results tell us that if the loan loss provision increases by 1 unit, we
expect that the level of non-performing loans will increase by 0.01% if other vari-
ables remain unchanged. Furthermore, if GDP increases by 1 unit, we expect that
the level of non-performing loans will decrease by 14.87%. On the other hand,
the level of non-performing loans decreases with the advent of COVID-19. This
unexpected connection is a consequence of the introduction of a moratorium on
loans, which prevented the growth of the level of non-performing loans and led
to the delay effect.

Effects of Covid-19 on Western Balkan financial systems

With the advent of the coronavirus, almost all world economies have seen a de-
cline in economic activity, especially small and open economies such as coun-
tries of Western Balkans. Tourism-oriented countries such as Montenegro and
Albania have seen a significant contraction with the onset of the corona crisis,
as the corona crisis has hit the sector hardest. Countries such as BiH, Serbia,
and Northern Macedonia, whose exports are part of the global supply chain to
Western Europe, registered a slightly milder decline in economic activity. As the
recovery is long and uncertain, and the crisis creates new opportunities, the joint
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 17

action of the countries of the region is one of the ways to adequately respond to
the crisis. One of the most important requirements for creating a common eco-
nomic market is the coherence of investment policies and better promotion of the
region as a whole to attract foreign direct investments, and creating a regional
industrial area that will allow better integration into global supply chains for
Western Europe.

The pandemic has led to a shock on both the supply and demand side, and the full
effect, duration, and economic consequences are yet to be seen and cannot yet
be fully predicted. Although the financial systems in countries in the region are
mostly stable, sufficiently capitalized and liquid, the pandemic still poses a sig-
nificant threat. Lower profitability, poorer asset quality, and lower credit growth
will have negative effects on the economy and pose additional challenges to the fi-
nancial stability of countries in the region. Therefore, regulators across the region
have launched a series of activities to prevent a decline in economic activity and
to maintain financial stability. The focus of central banks in this challenging time
is to monitor liquidity and credit trends. In order to maintain financial stability
throughout the region, regulatory easing measures have been introduced such as
moratoriums on loan repayment, a change in the classification of credit risk, and
a change in the terms after which a loan is considered uncollectible.

What poses additional risk to most countries in the region are the restrictions
on monetary regimes in force in these countries, especially in BiH, Kosovo, and
Montenegro. In this regard, the central banks face additional challenges in man-
aging liquidity of the financial systems and providing liquidity support itself. To
create additional capital space for banks, regulators have introduced a number
of measures such as the release of countercyclical capital reserves, the release of
capital reserves to limit profit distribution, a temporary ban on the payment of
bonuses to directors, and the like. When we talk about standard monetary policy
measures in the region, the central banks in Albania, Northern Macedonia, and
Serbia cut their reference interest rates. However, as a significant contraction in
economic activity is expected as a result of measures to curb the spread of coro-
navirus, which is a consequence of both supply and demand shocks, the impact
of monetary policy on this type of shock is limited. In this regard, the joint action
of fiscal and monetary policies is becoming increasingly prominent in order to
contribute to the stabilization of economic activities in these countries.

The coronavirus pandemic will have long-term consequences for general social
and economic activity. In the current year, a record contraction was recorded,
while positive expectations for the recovery of economic activity in the coming
period are associated with more favourable development of the epidemiological
18 Journal of Central Banking Theory and Practice

situation and high reliability of the vaccine. The latest official projections of the
European Commission and central banks in the region indicate a moderate re-
covery next year while reaching the pre-pandemic level is expected only at the
end of 2022. In the coming years, very uneven economic growth is expected,
i.e. a rapid recovery of industrial activities and stagnation or slower growth of
service activities. Although there is a heterogeneous economic structure among
the countries of the region, in all countries the service sector has recorded an in-
creasingly important share in the overall structure of economic activity in recent
years. Bearing in mind that the corona crisis will have the greatest impact on
tourism, transport, and other service activities, the recovery will thus be slower
in countries that are heavily dependent on these activities. The period of the great
financial crisis in the EU was marked by a long recovery of economic activity,
which returned to the pre-crisis level only after six years, among other things,
followed by a strong impact of the public debt crisis on individual stagnation. At
the same time, the recovery in the USA was achieved after three years, indicating
that policy coherence, along with technological advances, largely determine the
speed of recovery from crisis periods.

Conclusion

With their lending activity, banks are one of the most important participants in
the modern financial market. Given that the placement of loans is the main activ-
ity of banks and the largest item in the assets of banks' balance sheets, it is very
important that bank management successfully manages loans that banks grant
to customers. The quality of approved loans has proven to be one of the main
factors influencing the bank's profitability. And in recent decades bank manage-
ments have put a special focus on monitoring and managing the quality of loans
and preventing the occurrence of non-performing loans.

Although countries define the term non-performing loan differently, its defini-
tion in BiH does not differ significantly from the definition used in developed
countries. In the period following the Global financial crisis, non-performing
loans were also a problem in the BiH banking sector.

Due to the Global financial crisis, there was a decrease in economic activities,
and thus a decrease in the purchasing power of individuals and legal entities, so
banks faced an increase in the number of NPLs. However, with the emergence
of the coronavirus pandemic, credit risk is expected to increase, and according
to some research, the banking sector in our region could face one of the most
difficult years after the global financial crisis of 2007. Due to the outbreak of the
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 19

coronavirus pandemic, the Banking Agency in Bosnia and Herzegovina passed

decisions on temporary measures for recovery fromnegative economic conse-
quences. The results of the measures remain to be seen.

The analysis of the collected data from the banking sector of Bosnia and Herze-
govina identified the relations between the assumed determinants and non-per-
forming loans. Thus, it was determined that the improvement of non-performing
loans is largely due to the improvement of economic conditions in the country.
On the other hand, their growth is related to the growth of loan loss provisions as
one of the indicators of quality of the loan portfolio. The impact of the COVID-19
pandemic on NPLs in the banking sector of Bosnia and Herzegovina is not yet
adequately noticeable given the introduced moratorium on loans. Namely, these
moratoriums have largely prevented loans from moving into the category of non-
performing loans, and the real effect of COVID-19 on this variable is expected
only after the expiration of the moratorium period. Therefore, this may be inter-
esting for further research, as well as some other macroeconomic determinants
that could have implications for NPLs.

Although the recovery is uncertain and long, it needs to be used to create a re-
gional economic market as a path to stronger and more sustainable growth for
all Western Balkan countries. Thus, as is the case with the world's leading central
banks, the central banks in the region are increasingly turning to non-standard
monetary policy measures to adequately respond to the challenges posed by the
COVID-19 pandemic. In this regard, the joint action of fiscal and monetary pol-
icy is becoming increasingly prominent in order to contribute to the stabilization
of economic activities in these countries. A rapid recovery of activities in the EU,
as the most important foreign trade partner, is crucial for the countries in the
region as that will have a positive impact on exports, investments, and personal
consumption.
20 Journal of Central Banking Theory and Practice

References

1. Altunbas, J., Lin, M. H., Moyneux, P., & Seth, R. (2000). Efficiency and risk
in Japanese banking. Journal of banking and finance.
2. Babouček, I.; Jančar, M. Effects of Macroeconomic Shocks to the Quality of
the Aggregate Loan Portfolio; Working Paper Series, No. 1; Czech National
Bank: Praha, Czech, 2005.
3. Bahovec, V. & Erjavec, N. (2009). Uvod u ekonometrijsku analizu. Element,
Zagreb.
4. Balgova, M., Nies, M., & Plekhanov, A. (2016). The Economic Impact of
Reducing Non-Performing Loans. SSRN Electronic Journal. doi:10.2139/
ssrn.3119677
5. Bayar, Y. (2019). Macroeconomic, Institutional and Bank-Specific
Determinants of Non-Performing Loans in Emerging Market Economies: A
Dynamic Panel Regression Analysis. Journal of Central Banking Theory and
Practice, 8(3):95-110
6. Berger, A. N., & DeYoung, R. (1997). Problem loans and cost efficiency in
commercial banks. Journal of banking and finance.
7. Bookstaber, R., Cetina, J., Feldberg, G., Flood, M., & Glasserman, P. (2014).
Stress tests to promote financial stability: Assessing progress and looking to
the future. Journal of Risk Management in Financial Institutions, 16-25.
8. Brož, V. & Pfeifer, L., *2021), Are risk weights of banks in the Czech
Republic procyclical? Evidence from wavelet analysis, Journal of Central
Banking Theory and Practice, 113-139
9. Chang, Y. T. (2006). Role of NPLs and capital adequacy in banking structure
and competition. IMF Working paper.
10. Clair, R. T. (1992). Loan growth and loan quality: some preliminary
evidence from Texas banks. Economic Review, 9-22.
11. Cucinelli, D. (2015) “The impact of non-performing loans on bank lending
behavior: Evidence from the Italian banking sector”, Eurasian Journal of
Business and Economics, 8: 59-71.
12. Demary, M. (2021). Will COVID-19 cause insolvencies, zombification or
debt deleveraging? (No. 3/2021). IW-Kurzbericht.
13. Demirgüç-Kunt A. & E. Detragiache .1998. The Determinants of Banking
Crises: Evidence from Developed and Developing Countries. IMF Staff
papers, Vol. 45, No. 1.
14. Espinoza, R., & Prasad, A. (2010). Nonperforming loans in the GCC
banking system and their macroeconomic effects. IMF working paper.
15. Fabris, N. (2021). Financial Stability and Climate Change. Journal of Central
Banking Theory and Practice, 27-43.
 Non-Performing Loan Determinants and Impact of COVID-19: Case of Bosnia and Herzegovina 21

16. Goodell, J. (2020). COVID-19 and finance: Agendas for future research.
Finance Research Letters, vol 35.
17. González-Hermosillo, B. 1999. Determinants of ex=-ante banking system
distress: A macro empirical exploration of some recent episodes. IMF
Working Paper, 33.
18. Grubišić, Z., Kamenković, S., & Kaličanin, T. (2021). Comparative Analysis
of the Banking Sector Competitiveness in Serbia and Montenegro. Journal
of Central Banking Theory and Practice, 1, 75-91.
19. Haneef, S. (2012). Impact of risk management on non-performing loans and
profitability of banking sector of Pakistan. International Journal of Business
and Social Science.
20. Herreriasa, R., & Moreno, J. O. (2011). Spillovers and long-run diffusion of
non-performing loans risk. Midwest finance association.
21. Jiang, C., Kanas, A. and Molyneux, P., 2018. Public policy and financial
stability: The impact of PCA and TARP on U.S. bank non-performing loans,
International Journal of Finance & Economics, 23(4): 376-392
22. Johnson, S. (2009). The Quiet Coup. The Atlantic Online, 1-11.
23. Kasinger, J., Krahnen, J. P., Ongena, S., Pelizzon, L., Schmeling, M., &
Wahrenburg, M. (2021). Non-performing loans-new risks and policies? NPL
resolution after COVID-19: Main differences to previous crises (No. 84).
SAFE White Paper.
24. Keeton, W. (1999). Does faster loan growth lead to higher loan losses?
Economic Review, vol 84, pp 57-75
25. Klein, N. (2013). Non-performing loans in CESEE: Determinants and
impact on macroeconomic performances. IMF Working paper.
26. Kozarić, K., Žunić Dželihodžić, E. (2020). Effects of Macroeconomic
Environment on Non-Performing Loans and Financial Stability: Case of
Bosnia and Herzegovina. Journal of Central Banking Theory and Practice,
pp. 5-17.
27. Leaven, L., & Levine, R. (2009). Bank governance, regulation and risk
taking. Journal of financial economic.
28. Podpiera, J., & Weill, L. (2008). Bad luck or bad management? Emerging
banking market experience. Journal of financial stability.
29. Radivojević, N., & Jovović, J. (2017). "Examining of Determinants of
Non-Performing Loans," Prague Economic Papers, Prague University of
Economics and Business, vol. 2017(3), pages 300-316.
30. Sedgwick, P. (2012). Pearson’s correlation coefficient. BMJ, 345(jul04 1),
e4483–e4483.
31. Seyoum, A., Nigussie, H., & Tesfay, T. (2016). Factors affecting non-
performing loans: case study on development bank of Ethiopia central region.
22 Journal of Central Banking Theory and Practice

32. Somun - Kapetanović, R., 2008. Statistika u ekonomiji i menadžmentu.

Sarajevo: School of economics and business Sarajevo, pp.65
33. Turan, G., Koskija, A. 2014. Nonperforming Loans in Albania. Academic
Journal of Interdisciplinary Studies. MCSER Publishing, 3(3), 491-500.
34. Vaughan, E., & Vaughan, T. (1995). Osnovi osiguranja - Upravljanje rizicma.
Zagreb: Mate.
35. Zeqiraj, V., Iskenderoglu, O., & Sohag, K. (2021). Dynamic Impact of
Banking Performance on Financial Stability: Fresh Evidence from
Southeastern Europe, Journal of Central Banking Theory and Practice, 165-
181.