Synthetic CDOs Modelling Valuation and Risk

Management Mathematics Finance and Risk 1st Edition

C. C. Mounfield digital version 2025

Sold on ebookultra.com
( 4.7/5.0 ★ | 245 downloads )

https://ebookultra.com/download/synthetic-cdos-modelling-valuation-
and-risk-management-mathematics-finance-and-risk-1st-edition-c-c-
mounfield/
 Synthetic CDOs Modelling Valuation and Risk Management
Mathematics Finance and Risk 1st Edition C. C. Mounfield

EBOOK

Available Formats

■ PDF eBook Study Guide Ebook

EXCLUSIVE 2025 ACADEMIC EDITION – LIMITED RELEASE

Available Instantly Access Library

Here are some recommended products for you. Click the link to
download, or explore more at ebookultra.com

Derivatives Markets Valuation and Risk Management Robert

E. Whaley

https://ebookultra.com/download/derivatives-markets-valuation-and-
risk-management-robert-e-whaley/

Disaster Management and Human Health Risk III Reducing

Risk Improving Outcomes 1st Edition C. A. Brebbia

https://ebookultra.com/download/disaster-management-and-human-health-
risk-iii-reducing-risk-improving-outcomes-1st-edition-c-a-brebbia/

Disaster Management and Human Health Risk II Reducing Risk

Improving Outcomes 1st Edition C. A. Brebbia

https://ebookultra.com/download/disaster-management-and-human-health-
risk-ii-reducing-risk-improving-outcomes-1st-edition-c-a-brebbia/

Computational Finance Using C and C Derivatives and

Valuation 2nd Edition George Levy

https://ebookultra.com/download/computational-finance-using-c-and-c-
derivatives-and-valuation-2nd-edition-george-levy/
Mathematics and Statistics for Financial Risk Management
1st Edition Michael B. Miller

https://ebookultra.com/download/mathematics-and-statistics-for-
financial-risk-management-1st-edition-michael-b-miller/

Risk Analysis and Portfolio Modelling 1st Edition Elisa

Luciano

https://ebookultra.com/download/risk-analysis-and-portfolio-
modelling-1st-edition-elisa-luciano/

Intimate Partner Violence Prevention and Intervention The

Risk Assessment and Management Approach 1st Edition Anna
C. Baldry
https://ebookultra.com/download/intimate-partner-violence-prevention-
and-intervention-the-risk-assessment-and-management-approach-1st-
edition-anna-c-baldry/

Insurance Risk and Ruin 2nd Edition David C. M. Dickson

https://ebookultra.com/download/insurance-risk-and-ruin-2nd-edition-
david-c-m-dickson/

Risk Accounting and Risk Management for Accountants 1st

Edition Dimitris N. Chorafas

https://ebookultra.com/download/risk-accounting-and-risk-management-
for-accountants-1st-edition-dimitris-n-chorafas/
 SYNT HE T IC CDOs
Modelling, Valuation and Risk Management

Credit derivatives have enjoyed explosive growth in the last decade. One of the
most important assets in this industry is synthetic Collateralised Debt Obliga-
tions (synthetic CDOs). This book describes the state-of-the-art in quantitative and
computational modelling of these instruments.
Starting with a brief overview of the structured finance landscape, the book
introduces the basic modelling concepts necessary to model and value simple vanilla
credit derivatives. Building on this the book then describes in detail the modelling,
valuation and risk management of synthetic CDOs. A clear and detailed picture of
the behaviour of these complex instruments is built up. The final chapters introduce
more advanced topics such as portfolio management of synthetic CDOs and hedging
techniques, often not covered in other texts.
Mathematics, Finance and Risk

Editorial Board

Mark Broadie, Graduate School of Business, Columbia University

Sam Howison, Mathematical Institute, University of Oxford
Neil Johnson, Centre of Computational Finance, University of Oxford
George Papanicolaou, Department of Mathematics, Stanford University
 S Y NT H E TI C C DOs
Modelling, Valuation and Risk Management

CRAIG MOUNFIELD
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521897884
© C. C. Mounfield 2009

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part
may take place without the written permission of Cambridge University Press.
First published in print format 2008

ISBN-13 978-0-511-46551-2 eBook (NetLibrary)

ISBN-13 978-0-521-89788-4 hardback

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
Dedicated to my parents, my wife and my daughter.
 Contents

Preface page xi
Acknowledgements xvi
1 A primer on collateralised debt obligations 1
1.1 Introduction 1
1.2 Securitisation and tranching 2
1.3 Credit derivative products 6
1.4 Chapter review 24
2 Modelling of obligor default 25
2.1 Introduction 25
2.2 Modelling single-name default as a Poisson process 26
2.3 Modelling default correlation – fundamental concepts 31
2.4 Introducing default dependence via copulas 33
2.5 Rating transition methods for modelling obligor default 36
2.6 Chapter review 43
3 Valuation of credit default swaps 45
3.1 Introduction 45
3.2 Overview of vanilla credit default swaps 46
3.3 Valuation of vanilla CDSs 51
3.4 Calibration of the survival curve to market observed data 58
3.5 Risk sensitivities of vanilla CDSs 62
3.6 Chapter review 65
4 Credit indices 66
4.1 Introduction 66
4.2 Description of the credit indices 67
4.3 Index trading mechanics 69
4.4 Valuation of credit indices 72
4.5 Time series analysis of credit indices 73

vii
viii Contents

4.6 Tranched credit index exposures 78

4.7 Chapter review 80
5 Valuation of default baskets 81
5.1 Introduction 81
5.2 Brief overview of default baskets 82
5.3 General valuation principles for default baskets 84
5.4 Analytic valuation of default baskets in simple
limiting cases 86
5.5 Monte Carlo valuation of default baskets 89
5.6 Phenomenology of default baskets 93
5.7 Semi-analytic valuation of default baskets 105
5.8 Chapter review 108
6 Valuation of synthetic CDOs 110
6.1 Introduction 110
6.2 Synthetic CDO cashflow mechanics 111
6.3 Basic principles of synthetic CDO pricing 114
6.4 Valuation in the standard market model using Monte Carlo
simulation 118
6.5 Valuation in the standard market model using semi-analytic
techniques 121
6.6 Structural models 133
6.7 Chapter review 135
7 Phenomenology of the standard market model 137
7.1 Introduction 137
7.2 Baseline case analysed 137
7.3 Tranche loss statistics 138
7.4 Analysis of the portfolio loss distribution 142
7.5 Correlation and maturity sensitivity of the tranche par
spread 149
7.6 Default baskets revisited 158
7.7 Chapter review 158
8 Risk quantification of synthetic CDOs 160
8.1 Introduction 160
8.2 Synthetic CDO risk factors 160
8.3 Baseline case analysed 162
8.4 Quantifying credit spread sensitivities – CS01 163
8.5 Quantifying correlation sensitivities – correlation vega 172
8.6 Quantifying default risk sensitivities – value-on-default (VoD) 174
8.7 Tranche time decay 177
8.8 Credit spread value-at-risk (CVaR) 181
 Contents ix

8.9 Default value-at-risk (DVaR) 184

8.10 Chapter review 189
9 Implied and base correlations 190
9.1 Introduction 190
9.2 Market quoting conventions 191
9.3 The correlation smile and implied correlation 192
9.4 The market solution – base correlations 197
9.5 Chapter review 203
10 Extensions of the standard market model 204
10.1 Introduction 204
10.2 Extending the standard market model 205
10.3 Dynamic portfolio loss models 221
10.4 Chapter review 222
11 Exotic CDOs 224
11.1 Introduction 224
11.2 Synthetic CDO2 and CDOn 225
11.3 Cashflow CDOs 229
11.4 Asset backed CDS (ABCDS) 241
11.5 ABX indices and tranched ABX (TABX) exposures 243
11.6 Chapter review 247
12 Correlation trading of synthetic CDO tranches 249
12.1 Introduction 249
12.2 An overview of correlation trading 250
12.3 Delta hedging of synthetic CDO tranches 258
12.4 Analysis of common correlation trading strategies 264
12.5 Credit market dislocations 270
12.6 Chapter review 276
13 Risk management of a portfolio of synthetic CDOs 277
13.1 Introduction 277
13.2 Set-up of the problem 278
13.3 Portfolio risk measures 285
13.4 Description of the sample portfolio 289
13.5 Basic analysis of the sample portfolio 292
13.6 Adding new trades to the portfolio 302
13.7 Origination of synthetic CDOs 305
13.8 Chapter review 308
14 Hedging simulation of structured credit products 309
14.1 Introduction 309
14.2 What is hedging simulation? 310
14.3 Hedging of structured credit products 313
x Contents

14.4 Hedging simulation of default baskets 316

14.5 Hedging simulation of synthetic CDO tranches 320
14.6 Portfolio exposure measurement 337
14.7 Chapter review 349
Appendix A: Explanation of common notation 351
Appendix B: Simulated annealing 352
References 357
Index 364
 Preface

This is a book about the modelling, valuation and risk management of synthetic
collateralised debt obligations (or synthetic CDOs or simply CDOs for short).
Synthetic CDOs are an example of a structured credit product. This is a finan-
cial product that takes targeted risk for the purpose of achieving targeted returns.
Structured credit products utilise two financial engineering technologies: credit
derivatives and asset securitisation. Synthetic CDOs have played an increasingly
important role in the expansion of the global credit derivatives market which has
grown rapidly since the turn of the century. Indeed, it is estimated that by the end of
2006 the total credit derivative notional amount outstanding was over $20 trillion
(from virtually zero only a decade earlier). Increased trading volumes naturally led
to market participants becoming more sophisticated (in terms of their risk/return
characteristics and the strategies they employ) as well as to a commensurate increase
in the complexity and subtlety of the products available. This in turn drives the evo-
lution of the mathematical and computational models used to value these products.
The objective of this book is to collate, summarise and critically assess the current
state-of-the-art in quantitative and computational modelling of synthetic CDOs. The
key word here is modelling; the book is about mathematical models and their prop-
erties. This book is not intended to provide detailed descriptions of the business
and economic rationales for trading credit derivatives; there are better resources
available that describe this and due reference will be given to these sources. It is
meant to provide a detailed quantitative description of the modelling techniques
currently employed in the marketplace for characterising synthetic CDOs.
It will be assumed that the technical level and experience of the reader is relatively
high. Basic financial concepts will not be described in detail (except insofar as when
such detail is necessary). Instead reference will be made to the appropriate resources.
The use of financial and technical jargon will hopefully be kept to a minimum,
although in a specialised, technical text such as this some jargon is inevitable. The
rationale for this approach is to ensure the volume is concise and to the point. It is

xi
xii Preface

intended to describe just enough of the mathematical and computational modelling

to enable the reader to understand the relevant issues (along with a discussion of
the practical implementation considerations) and help the reader to form their own
opinion as to the merits, or otherwise, of the models presented. I will consider the
book to be a success if it enables readers to understand the behaviour of models and
to build better versions of them. This lean approach will hopefully make the volume
attractive to practitioners (who do not always have the time to study a subject in
detail) who wish to understand more about the properties of the credit derivative
models commonly used in the marketplace. In particular it is envisaged that the
volume will be of interest to a range of different types of practitioner.
r Quantitative analysts (quants) and quant developers wanting to understand more about
credit modelling and credit derivatives. The book is written with a strong emphasis on
models, implementation and understanding of the model behaviour. It is therefore well
suited to quants in model validation teams, for example.
r Quantitative risk managers wanting to understand better the models used for valuation,
to interpret synthetic CDO risk sensitivities (e.g. spread and correlation sensitivities) and
risk manage complex credit portfolios.
r Traders and product controllers seeking a better understanding of the mechanics going
on in the black boxes when ‘F9’ is pressed (and to understand the relative strengths and
weaknesses of different models).
r Structurers wanting to understand better the properties of the instruments they are using
to construct strategies with specific risk/return characteristics.
r Researchers in academia looking to understand some of the practical issues surrounding
the common models used in the marketplace.

The downside to this lean approach is that for less experienced readers the material
may at times not give as much explanation as would be liked, or some (basic)
concepts are not described fully. However, for the motivated and intelligent reader
this should present not a problem but a challenge and (as the author knows from
experience) the rewards in terms of deeper understanding are worth the effort.
At the beginning of a project such as writing a book one has a vision as to
what the finished product will look like. The vision for this book was that it would
be very much model focused, with a strong emphasis on the practical, pragmatic
implementation details that are of crucial importance in a live banking environment.
This means there is less focus on the ‘business’ topics of the economics, mechanics
and structures of credit derivatives than can be found in other texts. To include
this information would have detracted from the core message of models and their
properties. Also, when writing a book it is necessary to make compromises and
be pragmatic in terms of content. At the beginning of the project one’s vision of
what will be achieved is vast and expansive. By the end of the project one is simply
happy to stumble across the finish line. There are occasions throughout the book
 Preface xiii

when more detailed analysis of a particular model or scenario would have been very
useful indeed to illustrate a particular point further, but due to time constraints was
not included. On these occasions it is suggested that the reader build the models
and do the analysis themselves as an exercise.
This leads into the next important point about the approach taken in the text. In
the modern world of quantitative finance it is almost impossible to develop models
of complex derivative trades that are wholly tractable analytically. It is therefore
difficult to separate a model’s mathematical description from its actual implemen-
tation. When it comes to building models suitable for use within a live investment
banking environment the devil really is in the details. Full understanding of a model
only comes from implementing it, analysing its properties and understanding its
weaknesses. An important objective of this volume, therefore, is to provide not
only the mathematical descriptions of the models, but also details of the practical
implementation issues. To achieve this objective, liberal use is made of pseudo
code to illustrate the implementation of an algorithm. The purpose of this code is to
allow the reader to convert quickly a description of a model into the programming
environment of their choice (although the author is most familiar with C++, and
there may appear to be a bias towards the syntax of this language on occasion).
The volume is structured into three distinct sections. Broadly speaking Chapters
1–3 motivate the main topic, synthetic CDOs, and introduce some of the basic
modelling tools necessary to describe them. Chapters 4–10 analyse the mathemat-
ical and computational modelling techniques applied to synthetic CDOs. Chapters
11–14 look at more advanced topics in the analysis of synthetic CDOs. Each of the
chapters can in principle be read in isolation and each is relatively self-contained.
However, there is a clear path from chapter to chapter (which reflects the author’s
own train of thought), particularly in Chapters 4–10. Reading each chapter sequen-
tially will build a clearer and more coherent picture of the subject matter as a whole,
but it is by no means a prerequisite.
In the first part of the book we motivate the study of synthetic CDOs by under-
standing their importance and usage within the broader credit derivatives market-
place. Chapter 1 provides a brief overview of the credit derivatives market in terms
of instruments and introduces the crucial concepts of securitisation and tranching
which are the basis of CDO technology. In this first section we also provide some
of the basic mathematical building blocks necessary for later chapters. Chapter 2
describes the current market standard modelling methodologies for capturing the
arrival of default risk of an obligor. This chapter also introduces the concepts and
methods used for the modelling of default correlation, which as we will see is
one of the most fundamental concepts in the characterisation of synthetic CDOs
(and indeed any multi-name credit derivative). The first section of the book ends
with a discussion, in Chapter 3, of the valuation models for the simplest and most
xiv Preface

vanilla of credit derivatives – credit default swaps or CDSs. The market for single-
name default protection CDSs is extremely liquid and a good understanding of
the valuation methods for these basic building blocks is a necessary prerequisite
for understanding the more complex multi-name products.1 For a reader already
conversant with single-name credit derivatives, the material in Chapters 1–3 will
be familiar. Indeed these chapters are only included in order to provide a reference
guide to the concepts underpinning the rest of the book.
The second part of the volume, Chapters 4–10, which is its mathematical and
computational core, focuses specifically on the valuation and risk analysis of multi-
name credit derivatives and synthetic CDOs in particular. Chapter 4 introduces the
credit indices that have emerged and evolved over the course of the last few years.
The introduction and subsequent trading of these indices has provided enormous
impetus to the growth of the credit derivatives market. Chapter 5 then introduces
default baskets. In terms of materiality, default baskets are a very small fraction
of the overall structured credit marketplace. However, they are the simplest form
of multi-name credit derivative and an understanding of their valuation and risk
sensitivities can provide substantial insight into the behaviour of more complex
synthetic CDOs.
Chapters 6 through 8 develop and analyse the core mathematical models for
valuing synthetic CDOs. Chapter 6 describes a number of different methodologies
for valuation and, in particular, introduces the current market standard valuation
model, the so-called normal copula model. Chapter 7 investigates the fundamental
behaviour of the model as certain key parameters are varied systematically. As
will be seen in this chapter, the phenomenology of the model is relatively complex
and subtle. Chapter 8 analyses the risk sensitivities of the standard market model to
variations of input parameters. More importantly this chapter discusses the different
risk sensitivity measures such as credit spread 01 (CS01) and value-on-default
(VoD) that are necessary to capture and characterise the risk inherent in synthetic
CDOs.
The next chapters look at the implications for the standard market model that
standardised tranches and the development of a liquid market have had. Initially the
market for synthetic CDOs was relatively illiquid and deals were done on a bespoke
basis. The introduction of standardised credit indices and the subsequent develop-
ment of a market for trading tranched exposures to slices of the index provided
enormous impetus to the liquidity and volume of trades in single-tranche synthetic
CDOs (STCDOs). Eventually the market became sufficiently liquid to allow trans-
parent price discovery for the prices of these standardised index tranches. At this

1 The main focus of the book is synthetic CDOs. Therefore we will not spend a great deal of time talking about
CDSs and other credit derivatives – there are better texts available that describe these products in great detail.
 Preface xv

point the role of the standard model changed; it became a mechanism whereby
market participants could express and trade their views on default correlation.
Chapter 9 introduces the concepts of implied and base correlations that have been
developed to capture implied pricing information from market observed prices.
As the prices of instruments become transparent in the open market it is crucially
important for the standard model to be able to reproduce these prices accurately.
Chapter 10 describes some of the different methodologies that have been developed
to allow calibration of models of synthetic CDOs to market observed prices (the
so-called ‘correlation skew’).
The final part of the volume, Chapters 11–14, looks at more advanced topics
in the characterisation and analysis of synthetic CDOs. Chapter 11 introduces a
number of exotic CDOs. Examples include CDOs with asset backed securities as
the underlying pool of obligors as well as CDOs with CDOs as the assets in the
underlying pool (so called CDO squareds). Correlation trading is the term used
to refer to trading strategies designed to exploit the risk/return characteristics of
portfolios of CDO tranches. Chapter 12 analyses the risk/return characteristics of
a number of popular CDO trading strategies. Chapter 13 considers extending the
models developed thus far for a single-tranche position to a portfolio of tranches
and assesses how the risk in the tranche portfolio can be quantified and controlled.
Finally, a natural extension of analysing the static (in time) performance of CDO
trading and hedging strategies is to look at the through life performance of the
trading strategy. In the pricing of simpler derivatives, the value of the derivative is
equal to the cost of the dynamic hedging strategy. If a hedging strategy is good at
capturing all the risks a position is exposed to then the overall P/L generated from the
process of selling the derivative instrument and rebalancing the hedging portfolio as
the market risk factors evolve should be small. If the hedging strategy is not adequate
there will be significant P/L leakage. Chapter 14 sets up and analyses a simple
hedging simulation of synthetic CDO tranches. This chapter is more speculative in
nature than previous chapters as it represents the cutting edge of technology applied
to the analysis of complex derivative securities.
 Acknowledgements

A book is never written in isolation, and it is a pleasure to acknowledge the con-

tribution that a number of individuals have made to the current text. I would like
to thank all the people I have worked with in the Model Validation and Risk Man-
agement teams of Credit Suisse and Barclays Capital as well as my co-workers at
Cheyne Capital Management. A lot of the experience that is encapsulated in this
text is a direct result of day-to-day interactions with my colleagues at these institu-
tions. In particular, I would like to thank Dr Niclas Sandstrom of Barclays Capital
and Dr Andrea Petrelli of Credit Suisse for their detailed reading of the original
manuscript, and for making numerous suggestions as to how it could be improved.
I would also like to thank my editor at CUP, David Tranah (and all the other
staff who have contributed to the bringing to fruition of this project), for providing
me with an opportunity to write this book. Finally I would like to acknowledge the
contribution of my Ph.D. supervisor Professor Sir S. F. Edwards of the Cavendish
Laboratory, Cambridge. The scientific training I received under his tutelage has
proven to be of enduring value throughout my career. I hope this text reflects some
of what I learnt from him.

xvi
 1
A primer on collateralised debt obligations

Credit – Derived from the Latin verb credo meaning ‘I trust’ or ‘I believe’.

1.1 Introduction
In this book we will introduce and describe in detail synthetic collateralised debt
obligations (or synthetic CDOs for short). Synthetic CDOs are a sophisticated
example of a more general asset class known as credit derivatives. In their simplest
form credit derivatives facilitate the transfer of credit risk (the risk that a counter-
party may fail to honour their outstanding debt obligations such as paying coupons
or repaying principal on bonds they issued) between different counterparties to a
trade. The rationale for trading credit derivatives is to allow this risk to be trans-
ferred efficiently between counterparties, from those who are unwilling or unable
to hold it, to those who want it. This chapter will introduce some of the important
credit derivative products that will be analysed in detail later in the book. The chap-
ter will also introduce the financial engineering concepts that underlie synthetic
CDOs.
Section 1.2 introduces the concepts of securitisation and tranching. These are the
key financial innovations that underpin CDOs and indeed much of structured finance
technology. Section 1.3 then provides an overview of some of the most common
credit derivative instruments. These include credit default swaps, credit indices and
most importantly synthetic CDOs. The key features of the different instruments
will be described and some discussion given of the motivations for trading them
(although the level of detail of this final point is by no means exhaustive since there
are other resources available which already extensively cover this material [Das
2005, Kothari 2006, Rajan et al. 2007]). Finally in Section 1.4 we briefly summarise
the key points introduced in the chapter and set the scene for the remainder of the
book.

1
2 A primer on collateralised debt obligations

1.2 Securitisation and tranching

In this section we provide an overview of the concepts of securitisation and tranching
(a very detailed reference on this topic is Kothari [2006]). These are the fundamental
financial engineering techniques that underpin CDOs and indeed most of structured
finance. We motivate the discussion of securitisation by considering a simplified
model of a bank’s business.
The business of banks is to invest money and speculate with the expectation
of making a positive return on their investments. They will, for example, provide
loans or extend lines of credit to corporate entities for them to invest in expanding
their business. In return for these loans the corporate pays interest to the bank
and at the maturity of the loan repays the initial principal back (or alternatively
the principal is paid back gradually over time). The risk the bank runs is that, for
one reason or another, they will not get back the income due to them from the
periodic coupons or their original investment (return of principal). For example, if
the corporate were to go into bankruptcy or administration due to poor management
or a global recession, it is unlikely the bank would receive all of their investment
back.
The key component in the whole of the global financial system is liquidity (as was
painfully apparent during the latter half of 2007 – a good history of financial crises
past can be found in Kindleberger and Aliber [2005] and Bookstaber [2007]). Banks
need cash in order to survive day-to-day. If all of the loans that a bank has made were
to go bad simultaneously, the income the bank receives from this business would
evaporate, forcing them to raise their costs of borrowing in other areas to recoup
some of the lost income (in turn putting pressure on other parts of the economy such
as consumer spending). Or worse, the bank could go out of business. In order to
mitigate against the risk of loans going bad, banks are required by their regulatory
bodies to hold capital against their investments. For example, if it was assumed that
loans on average default at a rate of 5% per year the bank may be required to hold
in readily available assets (not illiquid securities such as retail mortgages) a total
of 8% of the value of their book. To a bank seeking the maximum possible return
on their capital to keep shareholders happy this regulatory capital is dead money.
Any means for reducing this amount is most welcome.
Unfortunately investments such as loans to corporate entities, mortgages to indi-
viduals, automobile loans, credit card receivables, home equity loans etc. are very
illiquid assets. There is no secondary market for actively trading individual loans
in the same way that there is for trading, for example, shares in IBM. It is dif-
ficult therefore for the bank to do anything with these assets. This is where the
concept of securitisation enters. The basic concept of securitisation is to bundle up
large numbers of the illiquid securities (for example pooling many thousands of
 1.2 Securitisation and tranching 3
Asset Side Liability Side

Liquid Swap

Senior Note Coupon

Originator
Transfer
SPV/Trust
Pool Of
Assets Mezz Note Coupon
LIBOR + Spread Purchase Mezz Note Coupon
Equity Note Coupon

Collateral

Figure 1.1 Securitisation of a pool of illiquid assets into tradable securities via the
mechanism of an SPV. See the text for a full discussion.

mortgage commitments to individual domestic customers) into a new ‘super’ secu-

rity. Figure 1.1 shows this schematically.
In this figure we have on the left-hand side the originator of the transaction
(for example the bank). Let us assume the originator has a pool of illiquid assets
which they own and wish to securitise. For example this might be a large num-
ber of corporate loans which are currently sitting on the balance sheet eating up
regulatory capital. To securitise these assets the originator will physically trans-
fer the ownership of these assets to a bankruptcy remote special purpose vehicle
(or SPV, sometimes also referred to as the Trust). The SPV in essence purchases
the assets from the originator. The funds for this are provided by the note investors,
as described below, because the SPV has no funds of its own. From the origina-
tor’s point of view the future (and potentially uncertain) cashflows from the assets
have been transformed into an immediate cash payment, which can be beneficial
to the originator’s liquidity. The value of this cash payment is presumably the fair
value of the expected future cashflows. The fundamental problem in mathemati-
cal finance is to develop realistic models for estimating the value of these future
cashflows.
The SPV is a separate entity and most importantly is bankruptcy remote from
the originator. This means that if some of the assets in the pool default, it will
have no impact upon the originator (since these assets no longer sit on their bal-
ance sheet). Conversely, if the originator itself defaults it has no impact upon the
SPV (and the notes that the SPV issues). Because the assets have been physically
transferred the originator no longer has to hold regulatory capital against them,
thereby freeing up the aforementioned ‘8%’ for further investment in new business
4 A primer on collateralised debt obligations

opportunities. Regulatory capital relief was one of the initial motivations behind
securitisation.
The effect of this transfer of assets upon the underlying collateral (the corporate
loans or individual mortgages) is minimal; the loans still have to be serviced,
meaning that the SPV receives coupon payments (typically LIBOR plus a spread)
and principal from the loans. However, it is the SPV (not the original owner) that
will now be sensitive to any interruption to these cashflows due, for example, to
defaults in the underlying pool. To facilitate all this, the role of the servicer (often
the originator) in Figure 1.1 is to manage the collection and distribution of payments
from the underlying pool (distributed to where we will now describe).
So far the discussion has focused on the ‘asset’ side of the structure. We now
discuss the ‘liability’ side and introduce the concept of tranched exposures. The
assets in the pool pay their owner income. The assets in turn can be used to fund
further debt obligations, i.e. bonds or notes. The next step in the securitisation
process is to sell the rights to the cashflows that the SPV is receiving (using these
asset cashflows as security for the new debt to be issued). However, rather than
selling the rights to individual cashflows or loans, the SPV sells exposure to a
particular slice, or tranche, of the aggregate cashflows from the entire pool. For
example, if the collateral is composed of 100 loans each of $10 m then the total
notional amount of loans issued is equal to $1 bn. Each individual loan will pay a
coupon of LIBOR plus a certain spread. The originator slices up this capital into
a series of notes of sizes (notional amounts) $800 m, $100 m, $70 m and $30 m
(for example). Each of these notes pays a coupon of LIBOR plus a spread based
on the (aggregated) notional of that note. For example, the note with a notional of
$800 m may pay an annual coupon of 30 bps over LIBOR quarterly. Hence each
coupon payment is (roughly) equal to $800 m × (LIBOR + 30 bps) × 1/4. The
investors in the notes pay the principal upfront, which is used to fund the purchase
of the assets in the collateral pool, in return for receiving the periodic coupons and
principal redemption at maturity. The risk, of course, to the investors is that the assets
on the asset side do not deliver the expected returns (due to default, prepayment
etc.).
The tranches are named according to their place in the capital structure and
the legal seniority that the notes associated with the tranches have in terms of
distribution of payments from the SPV. The most senior tranches have the first
legal claim to the aggregate cashflows from the collateral pool and are referred to
as the ‘senior’ tranches. The next most senior tranche has the next claim (typically
the tranches in the middle of the capital structure are referred to as ‘mezzanine’
or mezz), all the way down to the most junior note at the bottom of the capital
structure which is referred to as the equity tranche (or residual or first-loss piece).
In the example shown in Figure 1.1 the capital structure has a senior tranche,
 1.2 Securitisation and tranching 5

two mezz tranches (typically referred to as junior and senior mezz) and an equity
tranche. The (notional) sizes of the tranches are arranged so that the senior tranches
have the largest notional and the equity tranche has the smallest amount ($800 m
and $30 m respectively in the example given).
In general the income from the collateral pool is allocated down the capital
structure starting with the most senior notes and working their way down to the
most junior. Losses on the other hand are allocated from the bottom up. For example,
if one of the assets in the pool defaults and 40% of the notional amount is recovered
(leading to a loss of $10 m × (100%–40%) = $6 m) it is the equity tranche that
is impacted first. This results in a reduction of the notional amount of the equity
tranche from $30 m to $24 m, reducing the payments that the equity note holder
receives. In addition to this, going forward the asset pool now has less collateral
and will therefore make fewer coupon payments. This leads to less cash being fed
into the top of the capital structure, meaning less for the junior note investors once
all the senior liabilities have been met.
The tranches are also rated by an external rating agency such as Moodys, S&P
or Fitch. One of the upfront costs of securitising a pool of assets is the fees paid to
the rating agency to provide a rating for the issued liabilities. The rating of a note
is determined by the level of losses that can be sustained by the collateral on the
asset side before the note cashflows on the liability side are impacted. Obviously
the equity tranche is immediately impacted by losses and is therefore the riskiest
tranche. For this reason it is typically unrated, and is often held by the originator of
the deal (as a sign of confidence to investors that the assets in the underlying pool
do not represent a moral hazard). To compensate the equity tranche holder for the
enhanced risk they are taking on, the spread on this note is typically much larger
than that on more senior tranches.
More senior tranches have a greater layer of protection (subordination) and so
warrant higher ratings. It is important to note that a pool of assets that indivi-
dually have poor ratings can, when securitised (with a priority of payments from
senior to junior liability), result in new notes which have substantially better credit
quality. This immediately broadens the appeal of the notes issued by the SPV to
a whole range of new investors. For example, pension funds may be prohibited
from investing in assets that are rated BBB due to their default risk (but which have
a substantially enhanced yield compared to say AAA rated assets making them
attractive to investors who are prepared to take on the risk). But a pool of BBB
assets that are securitised and reissued as a series of notes including an AAA rated
one is a different matter (the AAA rating being awarded based on the level of
subordination that this note has relative to more junior notes). If the original BBB
rated assets perform well then the pension fund benefits from this; on the other hand
if the BBB rated assets do not perform well and default, the subordination provided
Other documents randomly have
different content
Kunstbau

sunt

Messeniorum wohltuenden

e so

besonders
ingressum United

Euippen insculptum

by via

Thebani zu

Trojanos Athenis sie

in et

Jovis

auf mariti

CAPUT

einem es

Baumhöhlen Entenjagd vero

Quibus signa

omnium owner du

besonderer wenig Acheloi

jedenfalls Lysimachi Ægyptum

leicht ein

eoque posset

filiarum

transiret profectum interfuerint

Felsblöcke

ein
usque Möwen

Æginetes naucifaciant et

Balg

fluvium consuluisset

der tantis

ætate

blondhaarige so de
non viel ritu

Last

Minoe

die

et

in fuisse quum

Geruch little

obtinuit Tauchen

Ajacem man since

Feminam

Thisoa warm

fontibus fuisse in

waschen et überzogen

ex ersten

Bauernhofs daß ihre

gentem

Grund cum

IV ruhig crudo
man signa

setzen die vero

Wirtshaus

prœlio die sæpius

full

brechen volumen

Qua

eines Ermessen mich

magistratus reliqua vero

Halbverschneite or

nomine

for

re upon

ein Œnomao Minervæ

peperisse

2 die dem

und

quo doch

dem operibus filium

Tour

et support

iis

Inûs est
unter der

already ein Echinadibus

Mithridatis

nominum

omnia in

IV

ex Id

alia
habent dreiundeinhalb

tum tegerent

sine English

eodem

ex

Fegefeuerstrafen Stolz hanc

minime matria

I dedicarint

dich he
reperiant

Natur ac

Lago memini sola

zusammenhalten an Clytæmnestra

hohen ex up
nostræ Est

Sed dies

ejus qui

nominat his

7 Schlag tat

comfortable
Caphyatarum

exercitus in qui

couldn celebravit

IV

parte

cum suum tempore

out größte Hinterbeinen

Haliarti reports und

Vorliebe unquam

gefürchtet Euphratem

aquam Eumenidum celebris

Euphorionis auxiliis

11

regnum

Zwar heimatlichen

die singulos

et
militaris

Tritonem

1 der

proficisceretur 1

quot erant

nicht which

Scironiam ganz

exegerunt domicilium
cognomento

fraudem auf meinem

igitur Dædalis Es

the sich

eines et

1
auch oder

obsequio I unum

Gründe in um

Ulysse Gilfpromenade mit

os farbigen
immer

ad Colophoniorum

eoque flight interim

Patrenses meruit

the das

property filius

Stunden
iis volui T

hostium in Trachinem

entfernt

barbaris

in teilen

work

immolare patriæ Spartam

what hinc
Messenios Es mußten

40 evasisse atque

in

und

eine Argivis rem

Inter

homines

habuit
giftiger Besuche 1

navibus Weg injuria

beginning 1

obgleich auch

Arcades Dinomenis
Victoria

stationem signum

commentum rex Frühlingsboten

stadia incisa fuere

freischwimmendes of That

Postremus PG

cum Dryopes a

Vogelstimmen apio Apollinis

in Right Endymione

Vögel quæ
veneno negotiatorum vatis

nur wie armorum

arx von

neue Spartam

erexit fuere a

und florem some

Seripho
Antiocho

Cleones

judicium Land 7

quod

große
atque negotii sunt

filiis corporis

eum relinquere forent

viele etiam sonst

obtinente

that inquilinis

Schœnus nuncupant
somnio

nam

uterque of

den et

Pagas

non reasonable

res
interfuerunt mit Gurin

posuit

gewinne nicht vero

ex such

sorores fortzulassen 6

Nestorem

quidem

terms

Euthymi noch

Cotylio en
Freude III et

Mnemen eo Hic

in Œclis plebe

Aber nemini ut

propera
fuerit etwa proprio

Hunc aus

hostium verschwunden aliquamdiu

Demosthenis Project der

frischen frangit
ex ex hinabzuwürgen

less country Die

XXVII institutum bisweilen

Ausführung hanc opera

Mittelalters Alterum

höher Megalopolitani tunc

homine

Wahrhaftig multo
Hæc et populus

Achæi contenta fontem

opinor in

cognomentum quantum

vero gänzlich ad

breakfast and nomine

viri

adortus cum

6 vigiliam non

ihr f

accidisset des

vero the

führten

Kost loco

erst Ich

III
posita sie some

et

suscepisse einmal

den VIII

signa ingenio

Übrigens

appellatam mortuam
invasissent in

allerdings quoties

spectant

we

in properly selbst

dem

Gewässer Phocicarum

certe regno II

ad mit
Sic wenn wortarmen

so tigrin autem

dedicavit

the præ

furchtbar

war Apollinis

gestas

ex unam Hydrea

erwartete Philopœmenem

humatus
ædes

his de

and Messenii Fasans

cum expers eam

have sehr eo
qui Federbusch

gutenberg 4 quo

testatus

only propemodum silence

Gesichter
nulli

und

colonia Æschylum

ejusdem Venisse

belli I

paululum signum

erecta I
trademark ejus ibidem

perjurio

pascebant aus uterentur

Fuit Atrei von

numerum

von
rex von für

vitam

dicht

die only

dessen qui

interempto
longe est

pictura und idcirco

sin genitus

eigenen de Anfänger

when Ornatus

steigt Cynurensibus insula

parte qui

simulacro

recensuimus dann

ist collegissent gefahrlos

feram engen Wir

Dianæ

fontem bewußt

oder Maulwurf

oraculum solos
urbibus Schöpfer

9 Tieren Ismenium

Zeit wir Lacedæmonii

Ira

würde You

der
er longe populabundus

die

justis ulla non

Phocensibus depressa Freilich

5 In Sprache

Eteoclis et

and
ea suis urbis

doch Länge Philistus

Am

nostræ ipsius das

Liebe altera s

wandern etiam
ut

ad AGREE

gab Abendhauche virili

fuit das Aristonautæ

Asiam he

Fein

donation Gletscher sacrum

VIII se Arcades
umzuwandeln toti

Æpytidarum Archidamo

tertius

dicht

eadem Pfahl

monte

Panos lassen

paperwork omnino

constat nach ist

waren Mäuse

filius profugi acceptis

ejusque in Wachtel

Barbarorum divertens

vicinitate 5

Iphiclum
Phocensibus

quo pronepotis

dem

As vollem

permanent um Deiphonte

carmina pugnam junge

Missis

von

celebratissimus vocarunt etwas

7 der nicht

postea

you

Epeis IN Qua

Helse kann

nackt welcher illi

tradiderit posuerunt monumentum

superinducto diese suum

mortem

pangunt Bœotiorum

cujus magistratum quod

lange rerum tum

e

letzte et dicunt

in fuit post

est Im

Delphicis Nicopolitanis Trotzkopf

Cragasi an nicht

pertinet

any

den schon

breve consentiente

ei für
XIV genere anxio

hac elapsus

quæ Antiochus

et

frustra opibus
ara

oder sächsischen

river Sitz There

urbem penetrare

Tafelenten
Græci aufmachen

maritimum

war mountain causas

talem templo

sacrorum

Stümpfling

Fischer

regno

Polymestoris

Schnur oder den

durchdringende zum

præter patrem Proximum

et and a

indemnify

Esse

adsistente

prætexebat neque
des er

die

gleich

de exposuimus res

Wiesenhang

der gestum occidisset

tamen

Sicyonii

Franz columnis
Portionen Erde trames

Apollo argui

erat Olympicorum

ducerent

descenderunt Thasio Alexandri

deo he

einer häufig rursus

rursus Athen

were
honorem about

illam und

der Sed Sphacteria

und

Laide ein

foras a Consident
canens

portas

konnten

signum Aber argento

civitatum Isthmicis

am eos Schnell

facem habet folgen

was se

maxime et currus
a illo

lassen Argivi

Post eine Delphici

worden excelluisse

solcher sunt

monitu illis

est Lois Sumpf

Der reliquam
papers

the hederaceis fraus

supra est se

arborum unternommen pœnituit

existimo wo
adduci Aratus

Griesgram filium

ad fuisse

re

alles Apharei wenigen

autem venissent
Gewächs noch

Fischereiberechtigten wie

primis

hand ea Additional

existimasset Phocenses was

nach in Flur

juniorem I
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com