Advanced financial risk management

Introduction: Wall Street Lessons from Bubbles xxiiiPART ONE Risk Management: De finitions and Objectives CHAPTER 1 A Risk Management Synthesis: Market Risk, Credit Risk, Liquidity Risk,

Advanced Financial Risk Management

is globally committed to developing and marketing print and electronic products andservices for our customers’ professional and personal knowledge and understanding.The Wiley Finance series contains books written specifically for finance andinvestment professionals as well as sophisticated individual investors and theirfinancial advisors Book topics range from portfolio management to e-commerce,risk management,financial engineering, valuation, and financial instrument analysis,

as well as much more

For a list of available titles, visit our Web site at www.WileyFinance.com

Advanced Financial Risk Management

Second Edition

Tools and Techniques for Integrated

Credit Risk and Interest Rate

Risk Management

DONALD R VAN DEVENTER

KENJI IMAI MARK MESLER

Copyright ª 2013 by John Wiley & Sons Singapore Pte Ltd.

Published by John Wiley & Sons Singapore Pte Ltd.

1 Fusionopolis Walk, #07-01, Solaris South Tower, Singapore 138628

All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as expressly permitted by law, without either the prior written permission of the Publisher, or authorization through payment of the appropriate photocopy fee to the Copyright Clearance Center Requests for permission should be addressed to the Publisher, John Wiley

& Sons Singapore Pte Ltd., 1 Fusionopolis Walk, #07-01, Solaris South Tower, Singapore

138628, tel: 65–6643–8000, fax: 65–6643–8008, e-mail: enquiry@wiley.com.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and speci fically disclaim any implied warranties of merchantability or fitness for a particular purpose No warranty may be created or extended by sales representatives or written sales materials The advice and strategies contained herein may not be suitable for your situation You should consult with a professional where appropriate Neither the publisher nor the author shall be liable for any damages arising herefrom Other Wiley Editorial Of fices

John Wiley & Sons, 111 River Street, Hoboken, NJ 07030, USA

John Wiley & Sons, The Atrium, Southern Gate, Chichester, West Sussex, P019 8SQ, United Kingdom

John Wiley & Sons (Canada) Ltd., 5353 Dundas Street West, Suite 400, Toronto, Ontario, M9B 6HB, Canada

John Wiley & Sons Australia Ltd., 42 McDougall Street, Milton, Queensland 4064, Australia Wiley-VCH, Boschstrasse 12, D-69469 Weinheim, Germany

Library of Congress Cataloging-in-Publication Data

ISBN 978-1-118-27854-3 (Hardcover)

ISBN 978-1-118-27857-4 (ePDF)

ISBN 978-1-118-27856-7 (Mobi)

ISBN 978-1-118-27855-0 (ePub)

Typeset in 10/12pt Sabon-Roman by MPS Limited, Chennai, India.

Printed in Singapore by Ho Printing.

Introduction: Wall Street Lessons from Bubbles xxiii

PART ONE

Risk Management: De finitions and Objectives

CHAPTER 1

A Risk Management Synthesis: Market Risk, Credit Risk, Liquidity Risk,

Advances in Integrated Risk Management and Institutional

CHAPTER 2

Perils and Pitfalls in the Measurement of Risk: The Impact

Expected Losses on Tranches of Collateralized Debt Obligations 19

Introduction to Transfer Pricing: Extracting Interest Rate Risk

Perspectives on Measuring Risk: One Source of Risk or Many

vii

Option Risk Management Evolution 28

Risk and Strategy Management in a Complex Financial Institution 29

The Role of Capital in Risk Management and Business Strategy 32Capital-Based Risk Management in Banking Today: Pros and Cons 35History of Capital-Based Regulations in Commercial Banking 37

PART TWO

Risk Management Techniques for Interest Rate Analytics

CHAPTER 3

Background Information on Movements in the U.S Treasury

A Step-by-Step Approach to Analyzing Interest Rate Risk 55

CHAPTER 4

Calculating the Value of a Fixed Coupon Bond with

Calculating the Coupon of a Fixed Coupon Bond with

Principal Paid at Maturity When the Value Is Known 62

Calculating the Payment Amount of an Amortizing Bond

Calculating the Value of a Floating-Rate Bond or Loan with

Future Value of an Invested Amount Earning at a Simple Interest

Rate of y Compounded m Times per Year for n Periods 66Future Value of an Invested Amount Earning at a Simple Interest

Present Value of a Future Amount If Funds Are Invested at a

Simple Interest Rate of y Compounded m Times per

Present Value of a Future Amount If Funds Are Invested at a SimpleInterest Rate of y Compounded Continuously for n Years 67Compounding Formulas and Present Value Factors P(t) 67

Yield to Maturity for Long or Short First Coupon Payment Periods 69Calculating Forward Interest Rates and Bond Prices 69Implied Forward Interest Rates on Zero-Coupon Bonds 69

CHAPTER 5

Example A: Stepwise Constant Yields and Forwards vs Nelson-Siegel 77Deriving the Form of the Yield Curve Implied by Example A 79Fitting the Nelson-Siegel Approach to Sample Data 81Example D: Quadratic Yield Splines and Related Forward Rates 85Deriving the Form of the Yield Curve Implied by Example D 86Example F: Cubic Yield Splines and Related Forwards 94Deriving the Form of the Yield Curve Implied by

Example H: Maximum Smoothness Forward

Deriving the Parameters of the Quartic Forward Rate Curves

Comparing Yield Curve and Forward Rate Smoothing Techniques 111Ranking 23 Smoothing Techniques by Smoothness of the

Ranking 23 Smoothing Techniques by Length of the

Trading Off Smoothness vs the Length of the Forward Rate Curve 112The Shimko Test for Measuring Accuracy of Smoothing Techniques 116Smoothing Yield Curves Using Coupon-Bearing Bond Prices as Inputs 116Appendix: Proof of the Maximum Smoothness Forward Rate Theorem 117

CHAPTER 6

Key Implications and Notation of the HJM Approach 129

Building the Bushy Tree for Zero-Coupon Bonds

Building the Bushy Tree for Zero-Coupon Bonds

Valuation of a Zero-Coupon Bond Maturing at Time T5 4 139Valuation of a Coupon-Bearing Bond Paying Annual Interest 140Valuation of a Digital Option on the One-Year U.S Treasury Rate 140

CHAPTER 7

Key Implications and Notation of the HJM Approach 146

Building the Bushy Tree for Zero-Coupon Bonds

Building the Bushy Tree for Zero-Coupon Bonds

Valuation of a Zero-Coupon Bond Maturing at Time T5 4 155Valuation of a Coupon-Bearing Bond Paying Annual Interest 156Valuation of a Digital Option on the One-Year U.S Treasury Rate 158

CHAPTER 8

Probability of Yield Curve Twists in the U.S Treasury Market 161

Introducing a Second Risk Factor Driving Interest Rates 163Key Implications and Notation of the HJM Approach 167

The Formula for Zero-Coupon Bond Price Shifts with

Building the Bushy Tree for Zero-Coupon Bonds

Valuation of a Zero-Coupon Bond Maturing at Time T5 4 183Valuation of a Coupon-Bearing Bond Paying Annual Interest 184Valuation of a Digital Option on the One-Year

Replication of HJM Example 3 in Common Spreadsheet Software 186

CHAPTER 9

Probability of Yield Curve Twists in the U.S Treasury Market 190

Risk Factor 1: Annual Changes in the One-Year U.S

Alternative Specifications of the Interest Rate Volatility Surface 200Key Implications and Notation of the HJM Approach 201

The Formula for Zero-Coupon Bond Price Shifts with

Building the Bushy Tree for Zero-Coupon Bonds

Valuation of a Zero-Coupon Bond Maturing at Time T5 4 219Valuation of a Coupon-Bearing Bond Paying Annual Interest 225Valuation of a Digital Option on the One-Year

CHAPTER 10

How Many Risk Factors Are Necessary to Accurately

Risk-Neutral and Empirical Probabilities of Interest

Common Pitfalls in Interest Rate Risk Management 238Pitfalls in the Use of One-Factor Term Structure Models 238Common Pitfalls in Asset and Liability Management 243Summarizing the Problems with Interpolated Monte Carlo

CHAPTER 11

Political Factions in Interest Rate Risk Management 251

Life Insurance Companies and Property and Casualty Insurance

Making a Decision on Interest Rate Risk and Return:

Assessing the Risk and Return Trade-Offs from a

CHAPTER 12

Legacy Rate Risk Tools: Interest Rate Sensitivity Gap Analysis 258

Modeling the Maturity Structure of a Class of Assets 267

Comparing a Duration Hedge with Hedging in the HJM Framework 271

Yield to Maturity for Long or Short First Coupon Payment Periods 274Applying the Yield-to-Maturity Formula to Duration 275

The Perfect Duration Hedge: The Difference between the

CHAPTER 13

What Is an Academic Term Structure Model and Why

Using Ito’s Lemma to Build a Term Structure Model 287

Conclusions about the Use of Duration’s Parallel Shift Assumptions 290

The Merton Term Structure Model: Parallel Yield Curve Shifts 293

Appendix A: Deriving Zero-Coupon Bond Prices in the

Appendix B: Deriving Zero-Coupon Bond Prices in the

Appendix C: Valuing Zero-Coupon Bonds in the Extended

CHAPTER 14

Fitting Zero-Coupon Bond Prices and Volatility Parameters Jointly 317Steps in Fitting the Interest Rate Volatility Assumptions 318Example 1: Fitting Interest Rate Volatility When

Interest Rate Parameter Fitting in Practical Application 330

PART THREE

Risk Management Techniques for Credit Risk Analytics

CHAPTER 15

An Introduction to Credit Risk: Using Market Signals in Loan

Increased Precision in Measuring the Safety and

Credit Default Swaps: The Dangers of Market Manipulation 341

Daily Nondealer Trading Volume for 1,090 Reference Names 347Credit Default Swap Trading Volume in Municipals and

Credit Default Swap Trading Volume in Sovereign Credits 353

CHAPTER 16

The Jarrow Model and the Issue of Liquidity in the Bond Market 364The Jarrow-Merton Put Option as a Risk Index

Fitting the Jarrow Model to Bond Prices, Credit Derivatives

Fitting to Current Price Data and Historical Price Data 366Fitting the Jarrow Model to Credit Derivatives Prices 366Fitting the Jarrow Model to a Historical Database of Defaults 366Fitting the Jarrow Model to Retail, Small Business, and

The Jarrow and Jarrow-Turnbull Models: A Summary 373

Measuring Ordinal Ranking of Companies by Credit Risk 379The Predictive ROC Accuracy Ratio: Techniques and Results 380The Predictive Capability of the Jarrow-Chava Reduced

Performance of Credit Models vs Nạve Models of Risk 386ROC Accuracy Ratios for Merton Model Theoretical

Testing Credit Models: The Analogy with Interest Rates 388Market Data Test 1: Accuracy in Fitting Observable

Market Data Test 2: Tests of Hedging Performance 389Market Data Test 3: Consistency of Model Implications

Market Data Test 4: Comparing Performance with Credit

Appendix: Converting Default Intensities to Discrete

Converting Continuous Instantaneous Probabilities of

Default to an Annual Default Probability or Monthly

Converting Continuous Default Probability to an

Converting Continuous Default Probability to a

Converting an Annual Default Probability to a Continuous

Converting a Monthly Default Probability to a Continuous

CHAPTER 17

A Better Convention for Credit ModelIndependent Credit Spreads 398Deriving the Full Credit Spread of a Risky Issuer 399Credit Spread Smoothing Using Yield CurveSmoothing Techniques 404Setting the Scene: Smoothing Results for the Risk-Free Curve 404

A Nạve Approach: Smoothing ABC Yields by Ignoring

The Credit Risk Premium: The Supply and Demand for Credit 416

CHAPTER 18

Through the Cycle vs Point in Time, a Distinction

Moral Hazard in “Self-Assessment” of Ratings Accuracy

Comparing the Accuracy of Ratings and Reduced Form Default

Problems with Legacy Ratings in the 2006 to 2011 Credit Crisis 431

Valuing Multipayment Bonds with the Merton Model of Risky Debt 444Estimating the Probability of Default in the Merton Model 445Implying the Value of Company Assets and Their Return Volatilityσ 446Mapping the Theoretical Merton Default Probabilities to

Copulas and Correlation between the Events of Default of

Simulating the Future Values of Bonds with No Credit Risk 454Current and Future Values of Fixed Income Instruments:

Valuation of a Straight Bond with a Bullet

Simulating the Future Values of Bonds with Credit Risk 471

CHAPTER 20

Collateralized Debt Obligations: A Worked Example of

Collateralized Debt Obligations: Practice 486The Copula Method of CDO Valuation: A Postmortem 487

PART FOUR

Risk Management Applications: Instrument by Instrument

CHAPTER 21

Example: European Call Option on Coupon-Bearing Bond 501Example: Coupon-Bearing Bond with Embedded

HJM Special Case: European Options in the One-Factor

CHAPTER 22

Futures on Zero-Coupon Bonds: The Sydney

Futures on Coupon-Bearing Bonds: Dealing with the

CHAPTER 23

Valuing Options on Forwards and Futures:

European Options on Forward Contracts on Zero-Coupon Bonds 532

European Options on a Eurodollar Futures-Type Forward Contract 540European Options on Futures on Coupon-Bearing Bonds 546European Options on Money Market Futures Contracts 546Defaultable Options on Forward and Futures Contracts 546

CHAPTER 24

Caps as European Options on Forward Rate Agreements 550

Value of a Loan with a Cap and a Floor 563

Measuring the Credit Risk of Counterparties on Caps and Floors 565

CHAPTER 25

CHAPTER 27

An Overview of Numerical Techniques for Fixed

An Example of Valuation of a Callable Bond with a

An Example of Valuation of a Rationally Prepaid

HJM Valuation of American Fixed Income Options

CHAPTER 28

Analysis of Irrationality: Criteria for a Powerful Explanation 623

A Worked Example Using an Amortizing Loan with

Rational and Irrational Prepayment Behavior 626

CHAPTER 29

Transactions Costs, Prepayments, Default, and Multinomial Logit 640Legacy Prepayment Analysis of Mortgage-Backed Securities 643Legacy Approaches: Prepayment Speeds and the

Constant Prepayment Speeds Are Simply a Principal

Logistic Regression, Credit Risk, and Prepayment 648Mortgage-Servicing Rights: The Ultimate Structured Product 648

An Introduction to the Valuation of Mortgage-Servicing Rights 649Comparing Best Practice and Common Practice in

The Role of Home Prices in Defaults and Prepayments 652Other Sources of Cash Flow Related to

Specifying the Rate and Balance Movement Formulas 659The Impact of Bank Credit Risk on Deposit Rates and Balances 669Case Study: German Three-Month Notice Savings Deposits 672

CHAPTER 31

Setting the Stage: Assumptions for the Domestic and

Numerical Methods for Valuation of Foreign Currency Derivatives 677Legacy Approaches to Foreign Exchange Options Valuation 678Implications of a Term Structure Model-Based FX Options Formula 680The Impact of Credit Risk on Foreign Exchange Risk Formulas 681

CHAPTER 32

Impact of Collateral on Valuation Models: The Example of

The Impact of Changing Home Prices on Collateral Values

The Impact of Collateral Values on a Rationally Prepaid Mortgage 684Conclusions about the Impact of Collateral Values 693

CHAPTER 33

Fluctuating Credit Risk and Revolving Credit Drawdowns 696Incorporating Links between Credit Quality and Line Usage 697

Is a Line of Credit a Put Option on the Debt of the Issuer? 697

CHAPTER 34

Modeling Equities: The Traditional Fund Management Approach 701

Modeling Equities: A Credit RiskAdjusted Approach 703Options on the Common Stock of a Company That Can Go Bankrupt 704Convertible Bonds of a Company That Can Go Bankrupt 706

CHAPTER 35

Life Insurance: Mortality Rates vs Default Probabilities 708Cyclicality in Default Probabilities and Mortality Rates 711

PART FIVE

Portfolio Strategy and Risk Management

CHAPTER 36

Value-at-Risk and Risk Management Objectives Revisited at the

The Jarrow-Merton Put Option as a Measure of Total Risk:

A Four-Question PassFail Test for Financial Institutions’

VaR vs the Put Option for Capital Allocation 728Why Are the VaR and Put Approaches So Different:

Calculating the Jarrow-Merton Put Option Value and

Valuing and Simulating the Jarrow-Merton Put Option 732

Liquidity, Performance, Capital Allocation, and

CHAPTER 37

Liquidity Risk Case Studies from the Credit Crisis 735

Consolidated JPMorgan Chase, Bear Stearns, and

Implications of the Credit Crisis History for Liquidity

Measuring Liquidity Risk as a Line of Credit in the

Integrating Managerial Behavior and Market Funds Supply

CHAPTER 38

Transaction-Level Performance Measurement vs

Plus Alpha Benchmark Performance vs Transfer Pricing 767Why Default Risk Is Critical in Performance Measurement

“Plus Alpha” Performance Measurement in Insurance and Banking 769Decomposing the Reasons for Plus or Minus Alpha in a Fixed

A Worked Example of Modern Fixed Income Performance Attribution 772

Using the Jarrow-Merton Put Option for Capital Allocation 780

Using the Jarrow-Merton Put Option Concept for Capital Allocation 780

Extending the Jarrow-Merton Capital Allocation

CHAPTER 39

Controlling the Probability of Failure through the Credit Cycle 789Hedging Total Risk to Maximize Shareholder Value 790Implications for Basel II, Basel III, and Solvency II 791

CHAPTER 40

Common Practice in Risk Management Systems: Dealing with

Upgrading the Risk Infrastructure: The Request for Proposal Process 795

CHAPTER 41

Master the Politics and Exposition of Risk Management:

Lessons from Bubbles

The credit crisis that began to unfold in the United States and Europe in 2006contains a treasure trove of lessons for risk managers that we have tried toincorporate into this book Since we have each worked in Japan, we felt strongly thatthe collapse of the Japanese bubble, which peaked in late 1989, contained equallyuseful lessons for risk managers As you’ll note in the “key fallacies in risk man-agement” discussed below, many ignored the lessons of the Japanese bubble because

of the common fallacy that“if it hasn’t happened to me yet, it won’t happen to me,even if it’s happened to someone else.”

Now that the United States and much of Europe are experiencing the collapse of

a bubble much like that which burst in Japan, the lessons from each of these bubblesseem much more relevant to risk managers around the world

We have worked hard in the second edition of this book to severely emphasize the discussion of financial models that are obviously inaccurate, mis-leading, or clearly inferior to another modeling approach We make this judgment onthe basis of cold hard facts (via model testing) or because of assumptions that areknown to be false The list of models that failed under the duress of the credit crisis is

de-a long one, de-and we mde-ake no de-apologies for reflecting those fde-ailures in this book.We’ve also worked hard to explain which models performed well during the creditcrisis Again, we base that judgment on model testing and the logical consistency andaccuracy of the assumptions behind those models

We believe in a“multiple models approach” to risk management That doesn’tmean, however, that all models used are equally accurate Nothing could be furtherfrom the truth The use of a multiple models approach, however, makes it clear when

a better model has been brought to the risk management discipline and when it’s timefor an older model to be removed from the toolkit One of our British friends pro-vided the elegant observation that “I don’t think it’s gentlemanly to compare theaccuracy of two models.” The authors, by contrast, believe that such comparisonsare a mandatory part of corporate governance and best practice risk management.With that brief introduction, we turn to a short summary of common fallacies inrisk management that have been exposed as fallacies very starkly in the wake of therecent credit crisis

KEY FALLACIES IN RISK MANAGEMENT

Summarizing the dangerous elements of conventional wisdom in risk management isn’teasy We’ve restricted ourselves to the seven most dangerous ways of thinking about risk.Each of them has much in common with this famous quote of John Maynard Keynes from

xxiii

“The Consequences to the Banks of the Collapse of Money Values” in 1931: “A soundbanker, alas, is not one who foresees danger and avoids it, but one who, when he is ruined,

is ruined in a conventional way along with his fellows, so that no one can really blamehim.” We summarize them here and discuss each briefly in turn:

n If it hasn’t happened to me yet, it won’t happen to me, even if it’s happened tosomeone else

n Silo risk management allows my firm to choose the“best of breed” risk modelfor our silo

n I don’t care what’s wrong with the model Everyone else is using it

n I don’t care what’s wrong with the assumptions Everyone else is using them

n Mathematical models are superior to computer simulations

n Big North American and European banks are more sophisticated than otherbanks around the world and we want to manage risk like they do

n Goldman says they do it this way and that must be right

We discuss each of these fallacies in turn

Happened to Someone Else.

The perceived risk in any given situation is often a function of the age, the location,and the personal experience of the person making the risk assessment There are lots

of examples Some motorcycle riders don’t wear helmets Some smokers think cancerhappens to other people In risk management, perhaps the biggest mistake in 2005and 2006 was the assumption that“home prices don’t go down” in the United States,

in spite of considerable evidence in a major feature on the home price bubble fromThe Economist (June 16, 2005) Exhibit I.1 compares, from the perspective of Jan-uary 2006, the relative home price appreciation during the Japanese bubble (since1982) and Case-Shiller home prices indices for Los Angeles and a 10-city composite(since 1996) U.S market participants, in spite of evidence like this, continued toassume U.S home prices would rise

Of course, this assumption was dramatically wrong It’s best illustrated byRichard C Koo (“The World in Balance Sheet Recession: Causes, Cure, and Poli-tics,” Nomura Research Institute, 2011) in Exhibit I.2, showing how similar thecollapse in U.S and Japan home prices has been

Obviously, just because something hasn’t happened to you yet doesn’t mean that

it won’t Why that isn’t obvious to more risk managers is a mystery to us

Risk Model for Our Silo.

Alan Greenspan, former Chairman of the Board of Governors of the Federal Reserve,made an important confession in this quote from the Guardian, five weeks after thecollapse of Lehman Brothers, on October 23, 2008:“I made a mistake in presumingthat the self-interests of organizations, specifically banks and others, were such thatthey were best capable of protecting their own shareholders and their equity in thefirms.” In the same manner, management of large financial institutions often ignores

1 Per m 2 , 5-month moving average.

Japan: Osaka Area Condo Price1

Japan: Tokyo Area Condo Price1

Composite Index Futures U.S 10 Cities Composite Home Price Index

(U.S Jan 2000  100, Japan: Dec 1985  100)

EXHIBIT I.2 U.S and Japan—Collapse in Home Prices

Sources: Bloomberg, Real Estate Economic Institute, Japan, S&P, S&P/Case Shiller s Home Price Indices,

EXHIBIT I.1 Feeling Homesick?

the fact that their staff acts in their own best interests, with the best interests of theorganization being a secondary consideration at best How else to explain whyinstitutions often use separate and inconsistent risk systems for market risk, creditrisk, liquidity risk, interest rate risk, capital allocation, and so on? Using today’sfinancial and computer technology, there is no systems or financial theory reason forsuch systems to be separate and distinct.

Disparate risk management units almost never cooperate in consolidating risksystems, even when there is an obvious increase in accuracy in risk assessment, unless

it is forced from the top down With the recent decline in U.S home prices driving all

of these silo risks, using a suite of separate risk systems (most of which could notanalyze home prices as a risk factor) does nothing for the shareholders, but it buildsbarriers that preserve the independence of silo risk management organizations

One of us gave a presentation to the European meeting of the International ciation of Credit Portfolio Managers in Zurich in May 2007, just as the credit crisiswas beginning to become obvious to more people The thrust of the presentation wasthat the commonly used copula method for valuation of collateralized debt obliga-tions gave dramatically more optimistic valuations than an alternative approach (the

Asso-“reduced form approach”), in which a series of key macroeconomic (macro) factorsare common drivers of default probabilities for each reference name After the pre-sentation, two bankers illustrated the accuracy of Keynes’s previous quote by saying,

“That was a good presentation, but there were 100 people in this room and 99 ofthem think you’re wrong.” Alas, the presentation was right on the mark, but thepower of conventional wisdom blinded many to their model risk

Using Them.

In this book, we devote a lot of time to pointing out the false assumptions of variousmodels as an important part of assessing model risk Surprisingly, many risk man-agers are unconcerned about false assumptions, especially if many others are usingthe same false assumptions Two of the most spectacular falsehoods have to bementioned The first is the use of the Black model to value interest rate caps andfloors, even though the Black model assumes interest rates are constant! The second

is the assumption of the most common form of copula model for CDO valuation,that every pair of reference names has the identical correlation in default risk Thelack of anxiety about the false assumptions in these two cases in particular hasresulted in a lot of job losses at major financial institutions Beware

Mathematical Models Are Superior to Computer Simulations.

Consistent with Alan Greenspan’s earlier quote, it would not be accurate to assumethat the academic builders of financial models have the shareholders’ interests atheart They don’t They have their own interests at heart, and primary among thoseinterests is to be published in prestigious academic journals as often as possible Eventhe most casual reader of such journals will observe that the overwhelming majority

of articles about valuation models presents a formula for an answer, not a

simulation In order to get a mathematical formula for an answer, even the bestacademic financial theorists are forced to make simplifying assumptions that are notaccurate Among the many such assumptions are assumptions that a variable isnormally distributed when it’s not, assumptions that returns from period to periodare independent when they’re not, and the assumption that the number of risk factorsthat determine risk are small in number (1, 2, or 3) when in fact that’s not the case.

A common theme in this book is that such simplifications have caused manyinstitutions to fail and many risk managers to lose their jobs A realistic simulationthat has no“closed form” mathematical solution is usually the only accurate way todescribe risk accurately Sadly, such a simulation will normally not be published in aprestigious academic finance journal

Big North American and European Banks Are More Sophisticated

Than Other Banks around the World and We Want to Manage Risk

Like They Do.

In spite of the second $1 trillion bail-out of U.S financial institutions in the last 25years, many persist in the belief that large North American and European financialinstitutions are the most skilled risk managers in the world This hallucination per-sists in spite of a mass of public evidence to the contrary Two quotes with respect toCitigroup in the aftermath of the credit crisis illustrate the point quite well TheNew York Times (November 22, 2008) suggests the risk management expertise oftop management at Citigroup:

“Chuck Prince going down to the corporate investment bank in late 2002was the start of that process,” a former Citigroup executive said of thebank’s big C.D.O push “Chuck was totally new to the job He didn’t know

a C.D.O from a grocery list, so he looked for someone for advice andsupport That person was Rubin And Rubin had always been an advocate

of being more aggressive in the capital markets arena He would say,‘Youhave to take more risk if you want to earn more.’

In Fortune magazine (November 28, 2007), Robert Rubin makes the point evenmore strongly:“I tried to help people as they thought their way through this Myself,

at that point, I had no familiarity at all with CDOs.”

Besides this evidence, there are the public records of the bailouts of Citigroup,Bank of America, Merrill Lynch, Bear Stearns, Royal Bank of Scotland, HBOS, and ahost of European institutions Why does this belief—a hallucination, perhaps—in therisk management expertise of North American and European institutions go on? Wethink that management teams in many countries are much more sophisticated aboutrisk management on a day-to-day basis than U.S banks were during the credit crisis.Working with risk managers in 32 countries, we still see this differential in expertise

on a daily basis

Goldman Says They Do It This Way and That Must Be Right.

We continue to be astonished that many financial market participants believe themodeling approach described to them by the big Wall Street firm’s local sales rep The

use of financial model discussions in the mid-1980s by Salomon Brothers in Tokyo was

a legendary example of how to do what Wall Street does, fleecing clients by building up

a “relationship” about financial models in which nạve buy-side clients reveal theirproprietary modeling approach in trade for hearing about Salomon’s “proprietary” yenfixed-income models In reality, the modeling conversation was strictly to exploit theclients’ naiveté by model arbitrage Total profits in Tokyo for Salomon during this timeaveraged more than $500 million per year for more than a decade Model arbitrage was

at the heart of the CDO losses on Wall Street, with an interesting twist Many tradersarbitraged their own firms and then moved on when the game ended It is important toremember that Wall Street and Wall Street traders have the same“relationship” withclients that you might have with a hamburger at McDonald’s

SELECTED EVENTS IN THE CREDIT CRISIS

As the credit crisis recedes into history, leaving only U.S and European governmentdeficits behind, it is important to record the credit crisis history before it’s lost Thissection summarizes recent events to confirm the accuracy of the assertions above Inthe chapters that follow, we describe an approach to risk management designed toprotect our readers from the kinds of events that unfolded in the credit crisis.The credit crisis chronology below was assembled from many sources:

n A contemporaneous set of press clippings and news articles maintained byKamakura Corporation from the very early dates of the credit crisis

n A credit risk chronology maintained by the Federal Reserve Bank of St Louis

n A credit risk chronology maintained by the University of Iowa We would like tothank @bionicturtle via twitter for bringing this chronology to our attention

n The“Levin report,” released by Senator Carl Levin on April 13, 2011

The full citation for the Levin report is as follows:

Wall Street and the Financial Crisis: Anatomy of a Financial Collapse, Majority andMinority Staff Report, Permanent Subcommittee on Investigations, United States Senate,April 13, 2011

This list is an abridged version of a longer chronology on www.kamakuraco.comlisted in the bibliography

September 2, 2004 Washington Mutual chief risk officer Jim Vanasek sends

internal Washington Mutual memo stating“At this point

in the mortgage cycle with prices having increased farbeyond the rate of increase in personal incomes, thereclearly comes a time when prices must slow down orperhaps even decline.” (Levin report, p 66)

May 31, 2005 “Fed Debates Pricking the U.S Housing Bubble,” New York

Times (Levin report, p 272)June 3, 2005 “Yale Professor Predicts Housing Bubble Will Burst,”

National Public Radio (Levin report, p 272)June 16, 2005 “The global housing boom: In come the waves The

worldwide rise in house prices is the biggest bubble in

history Prepare for the economic pain when it pops.”The Economist.

September 30, 2005 Case-Shiller home price index for Boston, MA peaks

(Standard & Poor’s)April 30, 2006 In an April memo discussing Countrywide’s issuance of

subprime 80/20 loans, which are loans that have nodown payment and are comprised of afirst loan for 80percent of the home’s value and a second loan for theremaining 20 percent of value, resulting in a loan tovalue ratio of 100 percent, Countrywide CEO AngeloMozilo wrote “In all my years in the business I havenever seen a more toxic pr[o]duct.” (Levin report,

p 232)May 2, 2006 Ameriquest Mortgage closes retail branch network and lays

off 3,600 employees (Orange County Register)May 5, 2006 Merit Financial Inc.files for bankruptcy (SeattlePI)July 31, 2006 Case-Shiller home price index for 20-city Composite Index

peaks (Standard & Poor’s)August 31, 2006 Case-Shiller home price index for Las Vegas peaks (Standard

& Poor’s)September 30, 2006 Case-Shiller home price index for Los Angeles peaks (Standard

& Poor’s)October 31, 2006 S&P director in an internal e-mail“note also the ‘mailing in

the keys and walking away’ epidemic has begun.” (Levinreport, p 268)

December 13, 2006 Dan Sparks informed the Firm-Wide Risk Committee of

Goldman Sachs that two more subprime originators hadfailed in the last week and that there was concern aboutearly payment defaults, saying,“Street aggressively puttingback early payment defaults to originators therebyaffecting the originators business Rumors around morefailures in the market.” (Levin report, p 478)

January 31, 2007 By January 2007, nearly 10 percent of all subprime loans

were delinquent, a 68 percent increase from January 2006.(Levin report, p 268)

February 2, 2007 Dan Sparks reported to senior Goldman Sachs executives,

“The team is working on putting loans in the deals back

to the originators (New Century, Wamu, and Fremont,all real counterparties) as there seem to be issuespotentially including some fraud at origination ”(Levin report, p 484)

February 22, 2007 HSBC fires head of its U.S mortgage lending business as

losses reach $10.5 billionApril 2, 2007 New Century, subprime lender, declares bankruptcy

(Bloomberg; Levin report, p 47)June 17, 2007 Two Bear Stearns subprime hedge funds collapse (Levin

report, p 47)

July 6, 2007 UBSfires CEO and the heir apparent for Chairman of the

Board Peter Wuffi (Financial Times)July 9, 2007 Credit Suisse releases a report that shows overall market

CDO losses could total up to $52 billion (Bloomberg)August 2, 2007 Bailout of IKB Deutsche Industriebank AG due to losses of

up to h1 billion on mortgage-related CDOs (Bloomberg)August 3, 2007 AXA rescues money market funds after subprime losses

(Reuters)August 9, 2007 BNP freezes $2.2 billion of funds over subprime (Reuters,

Bloomberg)August 14, 2007 17 Canadian structured investment vehicles fail when

commercial paper is denied by Canadian banks.(Bloomberg)

August 16, 2007 Countrywide taps $11.5 billion commercial paper backup

line (Bloomberg)September 14, 2007 Bank of England rescues Northern Rock over UK mortgage

losses (Reuters)October 1, 2007 UBS announces a $3.7 billion write-down and, after the

announcement, the chief executive of its investmentbanking division, Huw Jenkins, was replaced (FinancialTimes)

October 5, 2007 Merrill Lynch writes down $5.5 billion in losses on

subprime investments (Reuters)October 16, 2007 Citigroup announces $3 billion in write-offs on subprime

mortgages (Financial Times)October 18, 2007 Bank of America writes off $4 billion in losses (Bloomberg)October 24, 2007 Merrill Lynch writes down $7.9 billion on subprime

mortgages and related securities (Bloomberg; FinancialTimes)

October 30, 2007 Merrill Lynch CEO O’Neal fired (Reuters)

November 5, 2007 Citigroup CEO Prince resigns after announcement that

Citigroup may have to write down as much as $11 billion

in bad debt from subprime loans (Bloomberg)November 7, 2007 Morgan Stanley reports $3.7 billion in subprime losses

(Bloomberg)November 13, 2007 Bank of America says it will have to write off $3 billion in

bad debt (BBC News)November 15, 2007 Barclays confirms a $1.6 billion write-down in the month

of October on their subprime holdings The bank alsoreleased that more than d5 billion in exposure tosubprime loan packages could lead to more write-downs in the future (Bloomberg)

November 21, 2007 Freddie Mac announces $2 billion in losses from mortgage

defaults (Financial Times)December 6, 2007 The Royal Bank of Scotland announces that it expects to

write down d1.25 billion because of exposure to the U.S.subprime market (BBC News)

December 10, 2007 UBS announces another $10 billion in subprime

write-downs, bringing the total to date to $13.7 billion UBSalso announced a capital injection of $11.5 billion fromthe government of Singapore and an unnamed MiddleEast investor (MarketWatch.com)

December 12, 2007 Federal Reserve establishes Term Auction Facility to

provide bank funding secured by collateral (Levinreport, p 47)

December 19, 2007 Morgan Stanley announces $9.4 billion in write-downs from

subprime losses and a capital injection of $5 billion from aChinese sovereign wealth fund (Financial Times)January 15, 2008 Citigroup reports a $9.83 loss in the fourth quarter after

taking $18.1 billion in write-downs on subprimemortgage-related exposure The firm also announced itwould raise $12.5 billion in new capital, including $6.88billion from the Government of Singapore InvestmentCorporation (Financial Times)

January 17, 2008 Merrill Lynch announces net loss of $7.8 billion for 2007

due to $14.1 billion in write-downs on investmentsrelated to subprime mortgages (BBC News)

January 22, 2008 Ambac reports a record loss of $3.26 billion after

write-downs of $5.21 billion on its guarantees of subprimemortgage-related bonds (Financial Times)

February 14, 2008 UBS confirmed a 2007 loss of $4 billion on $18.4 billion in

write-downs related to the subprime crisis

February 17, 2008 Britain announces the nationalization of Northern Rock,

with loans to Northern Rock reaching 25 billion poundssterling (Financial Times)

February 28, 2008 AIG announces a $5.2 billion loss for the fourth quarter of

2007, its second consecutive quarterly loss AIGannounced write-downs of $11.12 billion pretax on itscredit default swap portfolio (Financial Times)

March 3, 2008 HSBC, the largest UK bank, reports a loss of $17.2 billion in

write-downs of its U.S mortgage portfolio (BBC News)March 14, 2008 Federal Reserve and JPMorgan Chase agree to provide

emergency funding for Bear Stearns Under theagreement, JPMorgan would borrow from the FederalReserve discount window and funnel the borrowings toBear Stearns (Forbes; DataCenterKnowledge.com)March 16, 2008 JPMorgan Chase agrees to pay $2 per share for Bear

Stearns on Sunday, March 16, a 93 percent discount tothe closing price on Friday March 14 JPMorgan agreed

to guarantee the trading liabilities of Bear Stearns,effective immediately (New York Times)

April 1, 2008 UBS, whose share price fell 83 percent in the last year,

reports it will write down $19 billion in thefirst quarter

on its U.S holdings (Financial Times)

April 1, 2008 UBS CEO Ospel resigns after announcement that UBS total

losses are almost $38 billion (Bloomberg)April 18, 2008 Citigroup reports $5.11 billion infirst quarter losses and $12

billion in write-downs on subprime mortgage loans andother risky assets The bank plans to cut 9,000 jobs inaddition to the 4,200 jobs cut in January (BBC News)May 9, 2008 AIG reports $7.81 billion infirst quarter losses and $9.11

billion of write-downs on the revaluation of their creditdefault swap portfolio AIG Holding Company was alsodowngraded to AA2 by two major rating agencies.(SeekingAlpha.com)

June 2, 2008 Wachovia CEO Thompson is ousted following large losses

that resulted from the acquisition of a big mortgagelender at the peak of the housing market (Reuters)June 25, 2008 Shareholders of Countrywide, a troubled mortgage lender,

approve the acquisition of the company by Bank ofAmerica (Financial Times)

July 11, 2008 IndyMac Bank, a $30 billion subprime mortgage lender

fails and is seized by FDIC after depositors withdraw

$1.55 billion (Levin report, pp 47 and 234)July 17, 2008 Merrill Lynch writes down $9.4 billion on mortgage related

assets (Financial Times)July 19 2008 Citigroup lost $5.2 billion and had $7.2 billion of write

downs in the second quarter (Financial Times)July 22, 2008 Marking the largest loss in the history of the fourth largest

U.S bank, Wachovia loses $8.9 billion in the secondquarter (Financial Times)

September 7, 2008 U.S takes control of Fannie Mae & Freddie Mac The two

companies currently have $5,400 billion in outstandingliabilities and guarantee three-quarters of new U.S.mortgages The government agreed to inject up to $100billion in each of them and will buy mortgage-backedbonds (Levin report, p 47; Financial Times)

September 11, 2008 Lehman Brothers reports its worst ever third quarter as it lost

$3.9 billion total and $7.8 billion in credit linked downs The company plans to shrink its size (FinancialTimes)

write-September 13, 2008 The U.S Treasury and Federal Reserve refuse to provide

public funds to help with a rescue takeover for LehmanBrothers as they did for Bear Sterns Bank of Americabacks out of negotiations with Lehman Brothers because

of the lack of government funds (Financial Times)September 15, 2008 Lehman Brothers bankruptcy (Levin report, p 47)September 15, 2008 Merrill Lynch announces sale to Bank of America (Levin

report, p 47)

September 16, 2008 Federal Reserve offers $85 billion credit line to AIG; Reserve

Primary Money Fund NAV falls below $1 (Levin report,

p 47)September 18, 2008 Lloyds rescues HBOS, the largest mortgage lender in the

UK (Reuters)September 21, 2008 Goldman Sachs and Morgan Stanley convert to bank

holding companies (Levin report, p 47)September 25, 2008 Washington Mutual fails, subsidiary bank is seized by the

FDIC and sold to JPMorgan Chase for $1.9 billion.JPMorgan Chase immediately wrote off $31 billion inlosses on the Washington Mutual assets (The Guardian;Levin report, p 47)

October 3, 2008 Congress and President Bush establish Troubled Asset

Relief Program (TARP), which is created by theEmergency Economic Stabilization Act (EESA) Therevised bailout plan allows the Treasury to restorestability to the financial system by buying $700 billion

in toxic debt from banks (Levin report, p 47; SIGTARPreport, p 2; CNN)

October 4, 2008 Wells Fargo offers $15.1 billion to buy Wachovia,

outbidding Citigroup’s $2.2 billion bid (Financial Times)October 6, 2008 Germany announces h50 billion bail-out of Hypo Real

Estate AG (USAToday)October 6, 2008 Germany announces unlimited guarantee of h568 billion in

private bank deposits (USAToday)October 8, 2008 The UK government implements d400 billion rescue plan

that includes government investing in banking industry,guaranteeing up to d250 billion of bank debt, and addingd100 billion to the Bank of England’s short-term loanscheme (Financial Times)

October 13, 2008 The UK government injects d37 billion in the nation’s three

largest banks, kicking off the nationalization process.(Financial Times)

October 14, 2008 The U.S government announces capital injection of $250

billion, of which $125 billion will go to nine large banks

as part of the Capital Purchase Program (CPP) inexchange for more government control on items such

as executive compensation (Financial Times; SIGTARPreport, p 1)

October 16, 2008 Switzerland government gives UBS a $59.2 billion bailout

(Bloomberg)October 28, 2008 U.S uses TARP to buy $125 billion in preferred stock at nine

banks The nine banks held over $11 trillion in bankingassets or roughly 75 percent of all assets owned by U.S.banks (Levin report, p 47; SIGTARP report, p 1)

November 23, 2008 U.S gives government bailout to Citigroup, agreeing to

cover losses on roughly $306 billion of Citigroup’s riskyassets (Reuters)

December 18, 2008 President George W Bush reveals plan to rescue General

Motors and Chrysler by lending them a total of $17.4billion (Financial Times)

January 15, 2009 The U.S government gives Bank of America an additional

$20 billion as part of TARP’s Targeted InvestmentProgram (TIP), which allows the Treasury to makeadditional targeted investments than what was givenunder TARP’s Capital Purchase Program Furthermore,the government agrees to guarantee nearly $118 billion

of potential losses on troubled assets (SIGTARP report,

p 1; Financial Times)March 20, 2009 German parliament passes Hypo Real Estate Nationalization

bill (Deutsche Welle)September 12, 2010 German government adds another h50 billion in aid to

Hypo Real Estate AG bringing total support to h142billion (Business Standard)

OneRisk Management: Definitions and Objectives

A Risk Management Synthesis Market Risk, Credit Risk, Liquidity Risk, and Asset and Liability Management

The field of risk management has undergone an enormous change in the past 40years and the pace of change is accelerating, thanks in part to the lessons learnedduring the credit crisis that began in late 2006

It hasn’t always been this way in the risk management field, as FrederickMacaulay must have realized nearly 40 years after introducing the concept ofduration in 1938 The oldest of the three authors entered the banking industry inthe aftermath of what seemed at the time to be a major interest rate crisis takingplace in the United States in the years 1974 and 1975 Financial institutions werestunned at the heights to which interests rates could rise, and they began looking forways to manage the risk Savings and loan associations, whose primary asset classwas the 30-year fixed rate mortgage, hurriedly began to offer floating-rate mort-gages for the first time In the midst of this panic, where did risk managers turn? Tothe concept of mark to market and hedging using Macaulay duration? (We discussthese in Chapters 3 to 13.) Unfortunately, for many of the institutions involved, theanswer was no

During this era in the United States, a mark-to-market approach came naturally

to members of the management team who rose through the ranks on the tradingfloor In this era, however, and even today, chief executive officers who passedthrough the trading floor on their way to the top were rare Instead, most of thesenior management team grew up looking at traditional financial statements andthinking of risk on a net income basis rather than a mark-to-market basis This ispartly why Citicorp CEO Charles Prince was described in the Introduction as anexecutive who“didn’t know the difference between a CDO and a grocery list.”

As a result, the first tool to which management turned was simulation of netincome, normally over a time horizon of one to two years Given the Wall Streetanalyst pressures that persist even to this day, it is not a surprise that net incomesimulation was the tool of choice What is surprising, however, is that it was oftenthe only choice, and the results of that decision were fatal to a large number of U.S.financial institutions when interest rates rose to 21 percent in the late 1970s andearly 1980s One trillion dollars later, U.S financial institutions regulators hadbailed out hundreds of failed financial institutions that disappeared because of the

3

unhedged interest rate risk and the credit risk that was driven by the high interestrate environment.

From the perspective of 2012, risk management has taken two steps forwardand one step backward The Federal Reserve’s Comprehensive Capital Analysis andReview 2012 (CCAR, 2012) appropriately focuses on a lengthy list of macroeco-nomic factors that contributed heavily to the credit crisis of 2006–2011 Thosemacro factors represent the two steps forward The step backward was that thereporting of the CCAR 2012 results is still heavily oriented toward financialaccounting and net income measures instead of mark-to-market risk measures.From a perspective in the 1980s, the failure of net income simulation-focusedrisk management must have been only mildly satisfying to the fans of FrederickMacaulay and his analytical, mark-to-market approach to risk management Even inthe mid-1970s, when a moderate crisis had just occurred and hints of the crisis tocome were getting stronger, resistance to mark-to-market concepts in risk manage-ment were strong and the costs to its advocates were high in career terms The eldest

of the three authors, encouraged by the U.S Comptroller of the Currency, pushedhard for the adoption of mark-to-market-based risk management at the sixth-largestbank in the United States from 1977 to 1982 The chief financial officer, one of thebrightest people in banking at the time, listened carefully but did not buy the concept,his PhD from Stanford University notwithstanding He was so programmed byaccounting concepts by that stage of his career that the mark-to-market approach torisk management was a foreign language that he would never feel comfortablespeaking The advocate of the mark-to-market approach to risk management had tochange firms

This type of financial accounting–focused executive is still in the majoritytoday The departed Charles Prince at Citigroup is a prime example, but the formerCEOs at failed firms like Lehman Brothers, Bear Stearns, Northern Rock PLC,Royal Bank of Scotland PLC, Merrill Lynch, Bank of America, Wachovia, andWashington Mutual all suffered from the same fate The risk that CEOs face fromignoring mark-to-market risk measurement is much larger and subtler now than

it was during the savings and loan crisis Trading activities are larger, financialinstruments are more complex, and the compensation system at large finan-cial institutions has caused the interests of traders and the interests of managementand the shareholders to diverge wildly

On the one hand, CEOs understand how critical risk management can be.Lehman CEO Dick Fuld was quoted by McDonald and Robinson (2009) in AColossal Failure of Common Sense, as saying,“The key to risk management is neverputting yourself in a position where you cannot live to fight another day” (317) Onthe other hand, the view held by the staff of management’s expertise showed a clearunderstanding of what management did and did not understand McDonald andRobinson note, “The World War I British Army was once described by a Germangeneral as‘lions led by donkeys.’ Mike Gelband’s [head of fixed income] opinion ofthe chain of command in Lehman’s little army could scarcely have been more suc-cinctly phrased” (234)

Every once in a while, the donkeys at the top realized that there are both lionsand donkeys below Business Insider quoted John Thain after he was brought in torescue Merrill Lynch in 2008: