Book review: Measuring and managing credit risk

Michel Crouhy

Abstract

Measuring and managing credit risk by Arnaud de Servigny and Olivier Renault McGraw-Hill, 2004. 388pp. Hardcover, US$65.00. (ISBN: 0071417559)

Credit risk is the largest yet most fundamental risk faced by banks. Credit risk is also a significant risk faced by other nonbank financial institutions and by non-bank corporations as well. Assessing credit risk is a complicated task, since many uncertain elements are involved in determining both how likely it is that an event of default will happen and how costly default will turn out to be if it does occur. It is therefore not surprising to find many different approaches to the credit risk problem. Some of the newest approaches employ equity market data to track the likelihood of default in public companies, while other approaches have been developed to assess credit risk at the portfolio level using mathematical and statistical modeling. Other approaches to the credit risk conundrum are more traditional and are based on credit risk assessments within an overall framework known as a credit rating for corporate obligors or credit scoring for retail customers.

The book by de Servigny and Renault addresses all these complex issues at a level and in a language that both risk professionals and academics can appreciate. Professionals will not be overwhelmed by advanced mathematics, as is the case with some recent excellent books on credit risk pricing. Both academics and professionals will find an excellent and well-documented overview of current issues in credit risk measurement, credit risk management, capital allocation, credit risk mitigation and credit risk pricing.

Empirical evidence is extensively discussed whenever it is available, especially on topical issues such as loss given default (LGD), default correlations and credit spreads. This is a distinguishing feature of the book compared with other more technical contributions that focus mainly on credit risk pricing or credit portfolio modeling. The book benefits greatly from the experience of both authors as senior professionals with Standard & Poor's Risk Solutions group and as lecturers in finance programs and speakers at conferences. The book is well structured and the topics are explained very clearly. It will be extremely helpful to those concerned with the implementation of Basel II, which is slated to become effective for international banks at the end of 2007.

The ten chapters cover all aspects of credit risk management. The first chapter is a general introduction to the microeconomic foundations of banking, with a focus on the three major intermediation functions fulfilled by financial institutions in modern economies, ie, liquidity intermediation, risk intermediation and information intermediation. There is also an interesting discussion on the adverse selection and moral hazard that result from the asymmetric relationship which prevails between the borrower and the lender.

Chapter 2 discusses external and internal rating systems. Readers involved in the implementation of Basel II will find critical information about the design of an internal rating system and a high-level presentation of the methodologies followed by the major rating agencies, such as Standard & Poor's and Moody's, in assessing the credit-worthiness of a corporate obligor. Controversial issues such as "point-in-time" vs. "throughthe-cycle" rating approaches and "procyclicality"- which are still being debated within
the industry and the Basel Committee - are discussed. This chapter is complemented by excellent appendices on how to derive transition matrices in discrete time and continuous time, and how to estimate risk-neutral transition probabilities for pricing purposes.

Chapter 3 goes deeper into the quantitative methodologies for assessing default risk. It first reviews the shortcomings and merits of the structural models of credit risk which are based on the well-known Merton model. KMV, a widely used vendor application of the Merton model that produces estimates of the expected default probabilities of public companies, is briefly reviewed. Then statistical scoring models such as the well-known Z-score model by Professor Altman and other statistical approaches are presented in some detail. Model performance measurement, which is a very important topic in the context of Basel II, is also discussed.

Chapter 4 covers another important topic related to Basel II - the estimation of loss given default. The chapter reviews the determinants of LGD, including jurisdiction (an appendix reviews the insolvency regimes for bonds and loans in the US, the UK, France and Germany). There is also a detailed discussion of various approaches to fitting the LGD distribution.

Chapter 5 covers one of the thorny issues in credit risk modeling, the modeling of default correlations. The chapter reviews equity and asset-based models, as well as the approach using copulas, which is the most popular approach among practitioners for pricing single-tranche CDOs and first-todefault credit derivatives. Correlations in an intensity framework, as originally proposed by Duffie-Singleton and Jarrow-Turnbull, are
also discussed.

Chapter 6 is a classic that can be found in many contributions on credit risk, eg, the review of the industry-sponsored credit portfolio models: CreditMetrics, KMV-Portfolio Manager, CreditRisk+ and other models such as Portfolio Risk Tracker from Standard & Poor's and CreditPorfolioView. Still, a major challenge with these models is to incorporate complex credit products such as ABSs, CDOs and CDSs. The chapter also offers an introduction to less well known approaches, such as the saddle-point method and the fast Fourier/Monte-Carlo approach. It provides the reader with an excellent introduction to economic capital attribution, the risk contribution of individual facilities to the overall portfolio and performance measurement (RAROC - risk-adjusted return on capital, and EVA - economic value added). Chapter 7 elaborates further on these concepts.

Chapter 8 discusses the current literature on the modeling of yield spreads, with an excellent appendix on reduced-form models. Chapter 9 gives an overview of structured credit products and credit derivatives, with a section on Standard & Poor's and Moody's approaches to the rating of CDO tranches. The last chapter reviews bank regulation - and especially the new Basel Capital Accord, with its strengths and shortcomings.

Personally, I find this book extremely useful as a comprehensive introduction to credit risk management. Each topic is well covered and an extensive bibliography provides the reader with more advanced references so that he or she can delve further into technical aspects. This book should satisfy several audiences, such as practitioners in banks involved in the implementation of Basel II, bank regulators, practitioners in credit risk departments of financial and non-financial institutions, and academics looking for an introduction to credit-related issues. With the enactment of Sarbanes-Oxley, boards of directors and senior management cannot hide behind the defense of ignorance any more; this book is a must for them and should belong to their library.

Michel Crouhy, Head of Financial Engineering, IXIS Corporate and Investment Bank, Paris