第 2 页
Credit Risk Management Basic Concepts: financial risk components, rating analysis, models, economic and regulatory capital
资料内容
第 4 页
Credit Risk Management Basic Concepts: financial risk components, rating analysis, models, economic and regulatory capital Dr. Ir. TONY VAN GESTEL Dexia Group, Risk Management Holding Prof. Dr. BART BAESENS Faculty of Business and Economics, Katholieke Universiteit Leuven, Belgium School of Management, University of Southampton, United Kingdom Vlerick Leuven Ghent Management School, Belgium 1
第 5 页
3 Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Bart Baesens and Tony Van Gestel 2009 The moral rights of the authors have been asserted Database right Oxford University Press (maker) First published 2009 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, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by Newgen Imaging Systems (P) Ltd, Chennai, India Printed in Great Britain on acid-free paper by CPI Antony Rowe, Chippenham, Wiltshire ISBN 978–0–19–954511–7 10 9 8 7 6 5 4 3 2 1
第 6 页
Preface Credit risk management is undoubtedly among the most crucial issues in the field of financial risk management. With the recent financial turmoil and the regulatory changes introduced by Basel II, credit risk analysis and risk assessment in general have been receiving even greater attention by the financial and banking industry. The ability to discriminate good customers from bad ones is a highly decisive element to be a successful player in the banking and credit industry. Predicting and mitigating default events is at the core of appropriate credit risk management and this can be greatly helped by employing suitable quan- titative models, without however precluding the reliance on human expert judgment. The optimal allocation of capital is also directly linked to appropriate credit risk models and has driven the interest of both academic and industrial communities. The recently established Basel II Capital Accord is an illustra- tion of how modern credit risk management techniques can be transformed in capital adequacy for banks. With their book, Tony Van Gestel and Bart Baesens provide newcomers to the field of risk management with a careful introduction to the different concepts of credit risk management, without entering into the technicalities often associated with this subject. This book is therefore appropriate for readers looking for a comprehensive and rigorous, yet accessible, descrip- tion of the various concepts underlying credit risk management techniques used in modern banking. Philippe Maystadt, president of the European Investment Bank August 2008
第 8 页
Contents Acknowledgements ix Introduction xi Chapter by chapter overview xiii 1. Bank risk management 1 1.1 Introduction 1 1.2 Banking history 2 1.3 Role of banks 9 1.4 Balance sheet 17 1.5 Sources of risk 23 1.6 Risk management 38 1.7 Regulation 52 1.8 Financial products 59 2. Credit scoring 93 2.1 Introduction 93 2.2 Scoring at different customer stages 95 2.3 Score types 105 2.4 Credit bureaus 109 2.5 Overrides 111 2.6 Business objectives 112 2.7 Limitations 113 3. Credit ratings 115 3.1 Introduction 115 3.2 Rating and scoring systems 117 3.3 Rating terminology 118 3.4 A taxonomy of credit ratings 121 3.5 Rating system architecture 143
第 9 页
viii Contents 3.6 Rating philosophy 145 3.7 External rating agencies 148 3.8 Rating system at banks 157 3.9 Application and use of ratings 162 3.10 Limitations 165 4. Risk modelling and measurement 168 4.1 Introduction 168 4.2 System life cycle 170 4.3 Overview of rating systems and models 174 4.4 Data definition and collection 201 4.5 Development 251 4.6 Implementation 264 4.7 Application and follow-up 265 4.8 Validation, quality control and backtesting 266 5. Portfolio models for credit risk 273 5.1 Introduction 273 5.2 Loss distribution 274 5.3 Measures of portfolio risk 278 5.4 Concentration and correlation 285 5.5 Portfolio model formulations 292 5.6 Overview of industry models 306 5.7 Basel II portfolio model 312 5.8 Implementation and application 325 5.9 Economic capital and capital allocation 327 6. Basel II 344 6.1 Introduction 344 6.2 Bank capital 350 6.3 Pillar 1 (minimum capital requirements) 354 6.4 Pillar 2 (supervisory review process) 418 6.5 Pillar 3 (market discipline) 431 6.6 Information technology aspects 441 6.7 Market impact 452 6.8 Future evolution 479 References 483 Index 519
第 10 页
Acknowledgements It is a great pleasure to acknowledge the contributions and assistance of many individuals to the writing of this book. This book is the result of many years of academic research and industry experience, throughout which we have been involved in various social networks that all contributed to this writing in some way. First of all, we would like to thank our editor, Oxford University Press. We are grateful to Alison Jones for contacting us three years ago with the idea of writing a book, which in fact turned out to be a book series. We would like to acknowledge the direct and indirect contributions of the many colleagues and friends with whom we collaborated during the last ten years. Olivier Creignou, Marc Itterbeek, Christian Latouche, Geert Kindt, Daniel Saks, Erik Verstraeten, Pierre Vérot, Damien Delas, Christine Lensel-Martinat, Marc Declerc, Naji Freiha, Romain Arend, Olivier Vimard, Patrick Tetu, Anne-Claire Balençon, Alain Probst, Yannick Vandewynckel, Joris Van Helleputte, Huseyin Uyar, Mustafa Aydin, Sinan Gül, ¸Senol Kahraman, Tom Dewyspelaere, Ronny Langendries, Marc Ronvaux, Laurent Balthazar, Joao Garcia, Thomas Alderweireld, Olivier Debliquy, Nicolas Van Oystaeyen and Nico Demuijnck shared their financial expertise with the interpretation of the statistical results. Lyn Thomas, Jos De Brabanter, Bart Hamers, Gerd Castermans, Karlien Vanden Branden, David Martens, Johan Huysmans, Joachim Jacobs, Steven Monsaert and Hugues Jan gave us useful feedback with their expert knowledge on statistics, implementations, computations, project and database management. We sincerely thank Claude Piret, André Delasnerie, Henri Bonaque, Eric Hermann, Rudi Sneyers, Luc Leonard, Jan Vanthienen, Johan Suykens, Joos Vandewalle and Bart De Moor, for the opportunities they gave us to work on the many challenges of Basel II and for their moral support to complete this book series. Last but not least, we gratefully acknowledge our families and parents for their love, support and encouragement.
第 12 页
Introduction Credit risk, or the risk that money owed is not repaid, has been prevalent in banking history. It is a principal and perhaps the most important risk type that has been present in finance, commerce and trade transactions from ancient cultures till today. Numerous small and large failures, combined with the cor- responding economic and social impact, further accelerated the importance of credit risk management throughout history. Credit risk management is a process that involves the identification of potential risks, the measurement of these risks, the appropriate treatment, and the actual implementation of risk models. Efficient credit risk management tools have been vital in allowing the phenomenal growth in consumer credit during the last 50 years. Without accurate automated decision tools, credit lending would not have allowed banks to expand the loan book with the speed they have. Nowadays, effective credit risk measurement and management is recognized by many economic actors, not in the least because of financial failures of banks themselves. The recent Basel II capital accord articulates new market standards for credit risk management and capital adequacy for banks. The level of capital, a cushion to absorb credit and other losses, is matched to the portfolio risk depending on the risk characteristics of individual transactions, their concentration and correlation. All organizations, including banks, need to optimally allocate capital in relation to the selective investments made. Hence, efficient tools and techniques for risk measurement are a key cornerstone of a good credit risk management. Where retail credit scoring has been one of the earliest successful financial risk management tools, developed initially by large US retailers, at the same time as the development of portfolio risk management, its success made it an excellent ambassador of quantitative modelling techniques for use in other asset classes. Electronic data availability and computation power, which increased exponentially over time, enabled the development and application of advanced statistical modeling techniques to support credit risk measure- ment, management and decision making in various types of asset classes.
第 13 页
xii Introduction Complementary to the existing judgmental risk management processes, risk quantification evolved to become an indispensable foundation of modern risk management. It is the objective of this book series to provide an overview of all aspects, steps, and issues that should be considered when undertaking credit risk management. The book series is written for both practitioners (e.g. financial managers, consultants, auditors, regulators, . . .) as well as academics (lec- turers, students, researchers, . . .). All three books aim at providing a solid technical basis without losing the focus on practical implementation. The first book lays the foundation for the next two by defining and reviewing some basic nomenclature and risk management concepts. This book is very useful for readers requiring a high-level understanding of the various con- cepts. Book II goes into the technical details of how to develop credit risk systems in a quantitative way. This is especially useful for those responsi- ble for implementation or academics doing quantitative research. It provides both the introduction to the techniques and practical examples to guide young and experienced practitioners and academics in the fascinating, but complex world of modelling. Book III then discusses model risk control and follow- up. This will be especially targeted towards model validators, auditors, regulators and/or people doing research on model monitoring and follow-up. Book I is primarily intended for newcomers in the field who need a global overview of the different concepts of risk management, measurement and modelling, without knowing the technical details discussed in the other two books. It introduces financial risk management and measurement, with the focus on credit risk. Default, loss and exposure risk, defined at a certain maturity, are the risk components that define the credit risk of a single trans- action. The various types of credit scores and ratings to indicate these risk components are discussed first. The entire process to construct scoring and rating systems to predict, monitor and measure credit risk at the counterpart and transaction level, is discussed next. This is followed by an overview of portfolio models that calculate how the risk of a whole portfolio depends on the risk levels of the individual products, the concentration of large expo- sures and the correlation between the risk levels of the different products. Book I concludes with an overview of the Basel II capital accord and a discussion of the practical business impact. Tony Van Gestel Bart Baesens Brussels, Belgium August 2008
第 14 页
Chapter by chapter overview Chapter 1 introduces risk management and defines some basic banking con- cepts needed in the subsequent chapters. It starts by giving an overview on banking history (section 1.2). Next, the economic role of banks as financial intermediaries and brokers is reviewed (section 1.3). The bank’s organi- zation and balance sheet reflect the different activity types (section 1.4). These activities are not without risk: banks are exposed to different risk types, of which the most important are credit, market and operational risk (section 1.5). The key concepts of sound risk management practices to con- trol these risks are then reviewed (section 1.6). Because banks have a central role in the economy, they are regulated to make sure that their capital cushion is aligned to their risk profile (section 1.7). This capital cushion serves to absorb losses and to protect the depositors’ funds. The chapter is concluded with an overview of financial products (section 1.8), each with its different risk characteristics. Chapter 2 provides an introduction to credit scoring. The different types of scoring that are nowadays being observed in industry, and how they relate to one another, are discussed first. Scores are used at different stages of the cus- tomer life cycle: marketing, application, performance, collection and profit scoring (section 2.2). Another classification is based upon the properties of the score (section 2.3): the purpose, risk type and risk entity, and inter- nal/external score author. Credit bureaus are a popular external reference source for credit scores (section 2.4). The practical use of scoring systems is discussed in the remainder of this chapter. When scores are assigned auto- matically, human expert judgment may adjust the score outcome and credit decisions. The reason and number of overrides provides useful information on the quality perception of the scoring system (section 2.5). Credit scores are used for many purposes. In a credit risk context, their use concerns credit
第 15 页
xiv Chapter by chapter overview decisions and portfolio risk monitoring, to advise in the pricing and to sup- port securitization transactions (section 2.6). Their limitations are discussed in section 2.7. Chapter 3 on credit ratings complements the chapter on scoring. Like scores, credit ratings order the credit risk. In addition, ratings represent the risk level into a limited number of categories. These and other differences between scores and ratings are reviewed in section 3.2. Ratings are so impor- tant that a specific terminology has developed. Concepts like rating actions, qualifiers, watchlist and outlook are introduced (section 3.3). Next, an exten- sive overview of rating types is given (section 3.4): ratings express credit quality on different maturities; on issuer and issue level; in terms of probabil- ity of default, loss given default, exposure, or expected loss; in terms of local or foreign currency; in terms of stand-alone risk, support or sovereign risk. . . With these elements, the generic architecture of a rating system is described (section 3.5). The ratings can be expressed using different philosophies: point-in-time ratings take into account the current economic conditions for rather short-term horizons, while through-the-cycle ratings take into account conservative assumptions during the whole business cycle on a longer-term perspective (section 3.6). Next, external rating agencies and their rating pro- cesses are discussed in section 3.7, while internal rating systems in banks are discussed in section 3.8. These ratings are applied for investment decisions, credit management on issue and portfolio level, for regulatory capital cal- culations, pricing and performance measurement, structured products, and credit risk reporting (section 3.9). The chapter is concluded with a critical appraisal of the limitations of credit ratings (section 3.10). Chapter 4 on risk modelling and measurement gives an overview of the entire model development life cycle, discussing every step in detail, without focusing on mathematical and technical aspects. This chapter is especially useful for financial practitioners who are involved in the development, man- agement, implementation and use of risk models. An overview is provided of the different aspects of risk measurement and modelling: data, modelling techniques and implementation for use. The life cycle of a rating system is presented in section 4.2. Next, the general overview of credit scoring models introduces the different classes of rating systems: structural, statistical and expert models, and human expert ratings (section 4.3). Most models rely on data, for risk measurement, model use and for model development. Risk data collection involves a definition of default, and a calculation methodology for exposure at default and loss given default for the identified defaults. In addi- tion, the explanatory variables, that drive the differences in risk levels, need
第 16 页
Chapter by chapter overview xv to be identified and collected (section 4.4). Once the data is collected, the model development starts and involves the model choice, the score function construction, the segmentation into risk classes or ratings, and the calibration into quantified risk levels (section 4.5). Next, implementation aspects are discussed (section 4.6). Credit scoring models are not static applications, but dynamic instruments that are used in continuously evolving and changing environments. Therefore, model maintenance and follow-up are required (section 4.7). Because the quality of the rating models has an important impact on the bank’s operations, models are subject to strong internal and external control before being put into operation, and during operation. The different, but also partially overlapping aspects of model validation, quality control and backtesting are reviewed in section 4.8. Chapter 5 introduces portfolio models. Whereas scores and ratings con- cern the risk of individual transactions, portfolio models analyze the loss risk on a portfolio of credits. The loss distribution of a portfolio is the aggregated result of the risk of different securities in the portfolio, each with individual risk components: exposure at default, loss given default and probability of default (section 5.2). Common risk measures for portfolio risk are the expected loss, the loss standard deviation, the value-at-risk, and expected shortfall (section 5.3). The portfolio risk depends not only on the risk of the individual facilities, but also on concentration and correlation (section 5.4). Although high concentrations do not impact the expected loss, they increase the value-at-risk. Joint credit quality downgrades and defaults due to cor- relation also increase the portfolio risk. Portfolio model formulations are either based on simplified mathematical models or apply Monte Carlo simu- lations to generate joint losses due to correlated defaults and possibly also market losses (section 5.5). Popular industry formulations like CreditMet- rics, Portfolio Manager, Portfolio Risk Tracker, Credit Portfolio View and CreditRisk+ are reviewed (section 5.6). Next, the Basel II portfolio model for regulatory capital calculation is explained (section 5.7). Implementation and application aspects are reviewed (section 5.8). The chapter is concluded with the concepts of economic capital calculation, capital allocation and risk-adjusted performance measures (section 5.9). Chapter 6 concludes this book with a detailed overview on the Basel II Capital Accord. The capital accord consists of three mutually reinforcing pillars. First, the components of bank capital are described (section 6.2). Pillar 1 defines the minimum capital requirements for credit, market and operational risk (section 6.3). Pillar 2 describes the supervisory review pro- cess to verify whether the bank holds sufficient capital to cover all its risks
第 17 页
xvi Chapter by chapter overview (section 6.4). Pillar 3 defines the market disclosure to catalyze prudential risk management and sufficient capitalization (section 6.5). After the description of the new capital accord, the practical impact is discussed. It has impor- tant implications for the bank’s information and communication technology: data needs to be collected on various levels: risk information, exposure, loss measures; computation engines calculate the risk on transactions and portfo- lios; data needs to be transferred correctly between different levels; and risk reports need to be communicated to regulators, senior management and the financial markets (section 6.6). The Basel II rules make capital requirements more risk sensitive, which will impact, a.o. (amongst other), the credit pric- ing and capital needs for banks with different risk profiles (section 6.7). A discussion on future evolutions concludes this chapter (section 6.8).
第 18 页
1. Bank risk management 1.1 Introduction Banks and banking activities have evolved significantly through time [96, 105, 216, 246, 416, 508]. With the introduction of money, financial services like deposit taking, lending money, currency exchange and money transfers became important. Because of the central role of money, banks had and still have an important role in the economy. Banks act as brokers between supply and demand of securities, and they transform short-term deposits into medium- and long-term credits. Specialized information on financial products is gathered by banks to improve investment decisions and to manage the risk. Like any other firm, banks are exposed to classical operational risks like infrastructure breakdown, supply problems, environmental risks, etc. More typical and important for a bank are the financial risks it takes by its trans- formation and brokerage function [246]. A bank raises funds by attracting deposits, borrowing on the interbank market or issuing debt instruments on the financial market [96]. Essentially, the bank’s main activity is to buy and sell financial products with different profit and risk characteristics. This transformation from supply to demand side is not without risk. Banks are exposed to credit, market, operational, interest rate and liquidity risk. The appropriate management of these risks is a key issue to reduce the earnings risk of the bank, and to reduce the risk that the bank becomes insolvent and that depositors cannot be refunded. In this introductory chapter, a broad overview of banking and risk man- agement is given. A nutshell overview on banking history is reported in section 1.2. The key role of banks in the economy as brokers and financial intermediaries is reviewed in section 1.3. From the structure of the bank balance sheet discussed in section 1.4, it becomes clear that banks use a high leverage to generate an acceptable level of profit. The high leverage requires
第 19 页
2 Bank risk management a proper understanding of the financial risks a bank takes, which are enu- merated in section 1.5. The main issues concerning the management of these risks are discussed in section 1.6. Because of their central role in the econ- omy, banks are subject to international and national regulation as explained in section 1.7. section 1.8 concludes the chapter with an overview of financial products. For more extensive information, the reader is referred to general books on the banking activity, like [98, 110, 174, 311, 342, 412, 419]. 1.2 Banking history The banking industry has a long history and has had an important influence on the economy and even politics [105, 216, 416, 508]. The development of banks is strongly related with the development of money in any form. Barter was the earliest form of trade. Goods were produced by those who were good at it and they exchanged their surplus with others to the benefit of both. A crucial element for barter is the double coincidence of wants: both counterparts in the trade must be interested in the exchange of goods. Other problems are the indivisibility of some goods and the lack of a clear unit to calculate whether one makes a profit with barter. Man invented money as a means to make payments and to account for debts and credits [128]: it served to settle juridical disputes, to give religious tributes, and to be an intermediate commodity for exchange and trade. The indirect exchange with money overcomes the great difficulties of barter and was a great step forward in the economic development. Money is a highly divisible and durable commodity. In most countries and civilizations, gold and silver have been dominant commodities for money. Other types of money that have been used in past are, a.o., amber, cattle, cowries, eggs, grain, ivory, precious metals, rice, salt and seeds. The Inca society was unique in the sense that it reached a high standard, but did not use any kind of money. Wealth in terms of gold and silver money is an easy target for thieves. In early civilizations, like Mesopotamia, the temple was considered as a safe place to store money: there were constantly people around, and the sacred place may have discouraged thieves. But while the money was stored in the temple, the government or other people active in trade may need it to finance projects. During the reign of the Babylonian emperor Hammurabi in the eighteenth century BC, records exist1 of loans made by priests of the 1 A stone tablet dated to around 2000 Before Christ (BC) has the inscription “Two shekels of silver have been borrowed by Mas-Schamach, the son of Adadrimeni, from the Sun priestess Amat-Schamach,
第 20 页
Banking history 3 temple. Banking had become important enough to define laws to regulate banking operations. Banking further developed during the Ancient Greek and Roman civi- lization [20]. Financial transactions like deposit taking, lending money and currency exchange were made by private entrepreneurs, temples and public organizations. Book transactions allowed customers to deposit money in one city and retrieve it in another city. Customers avoided the risky transport of large amounts of gold and silver coins. Normal interest rates were at around 10%, while risky operations had higher interest rates. Silver and gold coins were used as money. The Romans further improved and regularized Greek banking practices and improved administration, but still kept a preference for cash money. The downfall of the Roman empire and the collapse of trade made bank- ing less important. Coin money was largely abandoned in the first centuries after the Roman’s defeat. In addition, interest rate charging was seen as controversial by the powerful Christian church. A similar interpretation is made in Islam, where the basic principle is the sharing of profit and loss, and where interest rate charging is forbidden. Banks do not explicitly charge interest, instead they are paid by profit margins on deferred pay- ments and other mechanisms to compensate for the risk and the time value of money [331, 360, 364, 446]. In many other religions, interest taking was legitimate, especially when cattle or agricultural products were lent. The banking industry kept simmering, until it revived in the twelvth cen- tury, when Jews started to provide banking services to finance the economic welfare. Moneychangers issued documents similar to bank notes that were exchangeable at other medieval trade fairs. The Order of Knights Templar provided banking services to kings and powerful families. Their demand notes were redeemable in local currency at their castles across Europe and the Holy Land. The success and religion of both groups made them vul- nerable to envy and repression. Common people took over banking. North Italian bankers called Lombards took the place of Jewish bankers in the thirteenth century. Note that the term Lombard loan is still used today to indicate loans collateralized by, e.g., securities. Their commercial instinct and the invention of double-entry bookkeeping made their banking industry successful. They avoided the Christian sin of usury by creative bookkeeping the daughter of Warad-Enlil. He will pay the Sun-God’s interest. At the time of the harvest, he will pay back the sum and the interest upon it.” [330, 475]
显示前 20 页。下载资料查看完整内容。
评论 (0)
暂无评论。快来抢沙发吧！