Copyright 2002 by
John Wiley & Sons Ltd,
Baffins Lane, Chichester,
West Sussex, PO19 1UD, England
(+44) 1243 779777
e-mail (for orders and customer service enquiries): email@example.com
Visit our Home Page on http://www.wileyeurope.com or http://www.wiley.com
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 under the terms of the Copyright, Designs and Patents Act 1988 or under the
terms of a licence issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London, UK
W1P 9HE, without the permission in writing of the publisher.
Joël Bessis has asserted his right under the Copyright, Designs and Patents Act 1988, to be identified
as the author of this work.
Other Wiley Editorial Offices
John Wiley & Sons, Inc., 605 Third Avenue,
New York, NY 10158-0012, USA
WILEY-VCH Verlag GmbH, Pappelallee 3,
D-69469 Weinheim, Germany
John Wiley & Sons (Australia) Ltd, 33 Park Road, Milton,
Queensland 4064, Australia
John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01,
Jin Xing Distripark, Singapore 129809
John Wiley & Sons (Canada) Ltd, 22 Worcester Road,
Rexdale, Ontario M9W 1L1, Canada
Library of Congress Cataloguing-in-Publication Data
[Gestion des risques et gestion actif-passif des banques. English]
Risk management in banking/Joël Bessis.—2nd ed.
Includes bibliographical references and index.
ISBN 0-471-49977-3 (cloth)—ISBN 0-471-89336-6 (pbk.)
1. Bank management. 2. Risk management. 3. Asset-liability management. I. Title.
HG1615 .B45713 2001
332.1′ 068′ 1—dc21
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
ISBN 0-471-49977-3 (cloth)
ISBN 0-471-89336-6 (paper)
Typeset in 10/12pt Times Roman by Laserwords Private Limited, Chennai, India.
Printed and bound in Great Britain by TJ International Ltd, Padstow, England.
This book is printed on acid-free paper responsibly manufactured from sustainable forestation, for which at
least two trees are planted for each one used for paper production.
SECTION 1 Banking Risks
Banking Business Lines
SECTION 2 Risk Regulations
SECTION 3 Risk Management Processes
Risk Management Processes
Risk Management Organization
SECTION 4 Risk Models
VaR and Capital
Risk Model Building Blocks
SECTION 5 Asset–Liability Management
The Term Structure of Interest Rates
Interest Rate Gaps
Hedging and Derivatives
SECTION 6 Asset–Liability Management Models
Overview of ALM Models
ALM and Business Risk
ALM ‘Risk and Return’ Reporting and Policy
SECTION 7 Options and Convexity Risk in Banking
20 Implicit Options Risk
21 The Value of Implicit Options
SECTION 8 Mark-to-Market Management in Banking
Market Value and NPV of the Balance Sheet
NPV and Interest Rate Risk
NPV and Convexity Risks
NPV Distribution and VaR
SECTION 9 Funds Transfer Pricing
26 FTP Systems
27 Economic Transfer Prices
SECTION 10 Portfolio Analysis: Correlations
28 Correlations and Portfolio Effects
SECTION 11 Market Risk
Market Risk Building Blocks
Standalone Market Risk
Modelling Correlations and Multi-factor Models for Market Risk
Portfolio Market Risk
SECTION 12 Credit Risk Models
33 Overview of Credit Risk Models
SECTION 13 Credit Risk: ‘Standalone Risk’
Credit Risk Drivers
Credit Risk: Historical Data
Statistical and Econometric Models of Credit Risk
The Option Approach to Defaults and Migrations
Credit Risk Exposure
From Guarantees to Structures
Credit Risk Valuation and Credit Spreads
Standalone Credit Risk Distributions
SECTION 14 Credit Risk: ‘Portfolio Risk’
Modelling Credit Risk Correlations
Generating Loss Distributions: Overview
Portfolio Loss Distributions: Example
Analytical Loss Distributions
Loss Distributions: Monte Carlo Simulations
Loss Distribution and Transition Matrices
Capital and Credit Risk VaR
SECTION 15 Capital Allocation
Capital Allocation and Risk Contributions
Marginal Risk Contributions
SECTION 16 Risk-adjusted Performance
Risk-adjusted Performance Implementation
SECTION 17 Portfolio and Capital Management (Credit Risk)
Portfolio Reporting (1)
Portfolio Reporting (2)
Credit Derivatives: Definitions
Applications of Credit Derivatives
Securitization and Capital Management
Risk management in banking designates the entire set of risk management processes
and models allowing banks to implement risk-based policies and practices. They cover
all techniques and management tools required for measuring, monitoring and controlling
risks. The spectrum of models and processes extends to all risks: credit risk, market risk,
interest rate risk, liquidity risk and operational risk, to mention only major areas. Broadly
speaking, risk designates any uncertainty that might trigger losses. Risk-based policies and
practices have a common goal: enhancing the risk–return profile of the bank portfolio.
The innovation in this area is the gradual extension of new quantified risk measures to
all categories of risks, providing new views on risks, in addition to qualitative indicators
Current risks are tomorrow’s potential losses. Still, they are not as visible as tangible
revenues and costs are. Risk measurement is a conceptual and a practical challenge, which
probably explains why risk management suffered from a lack of credible measures. The
recent period has seen the emergence of a number of models and of ‘risk management
tools’ for quantifying and monitoring risks. Such tools enhance considerably the views
on risks and provide the ability to control them. This book essentially presents the risk
management ‘toolbox’, focusing on the underlying concepts and models, plus their practical implementation.
The move towards risk-based practices accelerated in recent years and now extends to
the entire banking industry. The basic underlying reasons are: banks have major incentives
to move rapidly in that direction; regulations developed guidelines for risk measurement
and for defining risk-based capital (equity); the risk management ‘toolbox’ of models
enriched considerably, for all types of risks, providing tools making risk measures instrumental and their integration into bank processes feasible.
THE RATIONALE FOR RISK-BASED PRACTICES
Why are visibility and sensitivity to risks so important for bank management? Certainly
because banks are ‘risk machines’: they take risks, they transform them, and they embed
them in banking products and services. Risk-based practices designate those practices
using quantified risk measures. Their scope evidently extends to risk-taking decisions,
under an ‘ex ante’ perspective, and risk monitoring, under an ‘ex post’ perspective, once
risk decisions are made. There are powerful motives to implement risk-based practices: to
provide a balanced view of risk and return from a management point of view; to develop
competitive advantages, to comply with increasingly stringent regulations.
A representative example of ‘new’ risk-based practices is the implementation of riskadjusted performance measures. In the financial universe, risk and return are two sides of
the same coin. It is easy to lend and to obtain attractive revenues from risky borrowers.
The price to pay is a risk that is higher than the prudent bank’s risk. The prudent bank
limits risks and, therefore, both future losses and expected revenues, by restricting business
volume and screening out risky borrowers. The prudent bank avoids losses but it might
suffer from lower market share and lower revenues. However, after a while, the risk-taker
might find out that higher losses materialize, and obtain an ex post performance lower
than the prudent lender performance. Who performs best? Unless assigning some measure
of risk to income, it is impossible to compare policies driven by different risk appetites.
Comparing performances without risk adjustment is like comparing apples and oranges.
The rationale of risk adjustment is in making comparable different performances attached
to different risk levels, and in general making comparable the risk–return profiles of
transactions and portfolios.
Under a competitive perspective, screening borrowers and differentiating the prices
accordingly, given the borrowers’ standing and their contributions to the bank’s portfolio
risk–return profile, are key issues. Not doing so results in adverse economics for banks.
Banks who do not differentiate risks lend to borrowers rejected by banks who better screen
and differentiate risks. By overpricing good risks, they discourage good borrowers. By
underpricing risks to risky customers, they attract them. By discouraging the relatively
good ones and attracting the relatively bad ones, the less advanced banks face the risk
of becoming riskier and poorer than banks adopting sound risk-based practices at an
earlier stage. Those banking institutions that actively manage their risks have a competitive advantage. They take risks more consciously, they anticipate adverse changes, they
protect themselves from unexpected events and they gain the expertise to price risks. The
competitors who lack such abilities may gain business in the short-term. Nevertheless,
they will lose ground with time, when those risks materialize into losses.
Under a management perspective, without a balanced view of expected return and risk,
banks have a ‘myopic’ view of the consequences of their business policies in terms of
future losses, because it is easier to measure income than to capture the underlying risks.
Even though risks remain a critical factor to all banks, they suffer from the limitations of
traditional risk indicators. The underlying major issue is to assign a value to risks in order
to make them commensurable with income and fully address the risk–return trade-off.
Regulation guidelines and requirements have become more stringent on the development of risk measures. This single motive suffices for developing quantified risk-based
practices. However, it is not the only incentive for structuring the risk management
tools and processes. The above motivations inspired some banks who became pioneers
in this field many years before the regulations set up guidelines that led the entire
industry towards more ‘risk-sensitive’ practices. Both motivations and regulations make
risk measurement a core building block of valuable risk-based practices. However, both
face the same highly challenging risk measuring issue.
RISK QUANTIFICATION IS A MAJOR CHALLENGE
Since risks are so important in banking, it is surprising that risk quantification remained
limited until recently. Quantitative finance addresses extensively risk in the capital markets.
However, the extension to the various risks of financial institutions remained a challenge
for multiple reasons. Risks are less tangible and visible than income. Academic models
provided foundations for risk modelling, but did not provide instrumental tools helping
decision-makers. Indeed, a large fraction of this book addresses the gap between conceptual models and banking risk management issues. Moreover, the regulators’ focus on
risks is still relatively recent. It dates from the early stages of the reregulation phase,
when the Cooke ratio imposed a charge in terms of capital for any credit risk exposure.
Risk-based practices suffered from real challenges: simple solutions do not help; risk
measures require models; models not instrumental; quantitative finance aimed at financial
markets more than at financial institutions. For such reasons, the prerequisites for making
instrumental risk quantifications remained out of reach.
Visibility on Losses is Not Visibility on Risks
Risks remain intangible and invisible until they materialize into losses. Simple solutions
do not really help to capture risks. For instance, a credit risk exposure from a loan is not
the risk. The risk depends on the likelihood of losses and the magnitude of recoveries
in addition to the size of the amount at risk. Observing and recording losses and their
frequencies could help. Unfortunately, loss histories are insufficient. It is not simple to
link observable losses and earning declines with specific sources of risks. Tracking credit
losses does not tell whether they result from inadequate limits, underestimating credit
risk, inadequate guarantees, or excessive risk concentration. Recording the fluctuations of
the interest income is easy, but tracing back such changes to interest rates is less obvious.
Without links to instrumental risk controls, earning and loss histories are of limited interest
because they do not help in taking forward looking corrective actions. Visibility on losses
is not visibility on risks.
Tracking Risks for Management Purposes Requires Models
Tracking risks for management purposes requires models for better capturing risks and
relating them to instrumental controls. Intuitively, the only way to quantify invisible risks
is to model them. Moreover, multiple risk indicators are not substitutes for quantified
measures. Surveillance of risk typically includes such various items as exposure size,
watch lists for credit risk, or value changes triggered by market movements for market
instruments. These indicators capture the multiple dimensions of risk, but they do not
add them up into a quantified measure. Finally, missing links between future losses from
current risks and risk drivers, which are instrumental for controlling risk, make it unfeasible to timely monitor risks. The contribution of models addresses such issues. They
provide quantified measures of risk or, in other words, they value the risk of banks.
Moreover, they do so in a way that allows tracing back risks to management controls
over risk exposures of financial institutions. Without such links, risk measures would
‘float in the air’, without providing management tools.
Financial Markets versus Financial Institutions
The abundance of models in quantitative finance did not address the issues that financial
institutions face until recently, except in certain specific areas such as asset portfolio
management. They undermined the foundations of risk management, without bridging the
gap between models and the needs of financial institutions.
Quantitative finance became a huge field that took off long ago, with plenty of
pioneering contributions, many of them making their authors Nobel Prize winners. In
the market place, quantification is ‘natural’ because of the continuous observation of
prices and market parameters (interest rates, equity indexes, etc.). For interest rate risk,
modelling the term structure of interest rates is a classical field in market finance. The
pioneering work of Sharpe linked stock prices to equity risk in the stock market. The
Black–Scholes option model is the foundation for pricing derivative instruments, options
and futures, which today are standard instruments for managing risks. The scientific
literature also addressed credit risk a long time ago. The major contribution of Robert
Merton on modelling default within an option framework, a pillar of current credit risk
modelling, dates from 1974.
These contributions fostered major innovations, from pricing market instruments and
derivatives (options) that serve for investing and hedging risks, to defining benchmarks and
guidelines for the portfolios management of market instruments (stocks and bonds). They
also helped financial institutions to develop their business through ever-changing product
innovations. Innovation made it feasible to customize products for matching investors’
needs with specific risk–return bundles. It also allowed both financial and corporate
entities to hedge their risks with derivatives. The need for investors to take exposures
and, for those taking exposures, to hedge them provided business for both risk-takers
and risk-hedgers. However, these developments fell short of directly addressing the basic
prerequisites of a risk management system in financial institutions.
Prerequisites for Risk Management in Financial Institutions
The basic prerequisites for deploying risk management in banks are:
• Risks measuring and valuation.
• Tracing risks back to risk drivers under the management control.
Jumping to market instruments for managing risks without prior knowledge of exposures
to the various risks is evidently meaningless unless we know the magnitude of the risks
to keep under control, and what they actually mean in terms of potential value lost. The
risk valuation issue is not simple. It is much easier to address in the market universe.
However, interest rate risk requires other management models and tools. All banking
business lines generate exposures to interest rate risks. However, linking interest income
and rates requires modelling the balance sheet behaviour. Since the balance sheet generates both interest revenues and interest costs, they offset each other to a certain extent,
depending on matches and mismatches between sizes of assets and liabilities and interest
rate references. Capturing the extent of offsetting effects between assets and liabilities
also requires dedicated models.
Credit risk remained a challenge until recently, even though it is the oldest of all banking
risks. Bank practices rely on traditional indicators, such as credit risk exposures measured
by outstanding balances of loans at risk with borrowers, or amounts at risk, and internal
ratings measuring the ‘quality’ of risk. Banking institutions have always monitored credit
risk actively, through a number of systems such as limits, delegations, internal ratings and
watch lists. Ratings agencies monitor credit risk of public debt issues. However, credit
risk assessment remained judgmental, a characteristic of the ‘credit culture’, focusing on
‘fundamentals’: all qualitative variables that drive the credit worthiness of a borrower.
The ‘fundamental’ view on credit risk still prevails, and it will obviously remain relevant.
Credit risk benefits from diversification effects that limit the size of credit losses of a
portfolio. Credit risk focus is more on transactions. When moving to the global portfolio
view, we know that a large fraction of the risk of individual transactions is diversified
away. A very simple question is: By how much? This question remained unanswered until
portfolio models, specifically designed for that purpose, emerged in the nineties. It is easy
to understand why. Credit risk is largely invisible. The simultaneous default of two large
corporate firms, for whom the likelihood of default is small, is probably an unobservable
event. Still, this is the issue underlying credit risk diversification. Because of the scarcity
of data available, the diversification issue for credit risk remained beyond reach until new
modelling techniques appeared. Portfolio models, which appeared only in the nineties,
turned around the difficulty by modelling the likelihood of modelled defaults, rather than
This is where modelling risks contributes. It pushes further away the frontier between
measurable risks and invisible–intangible risks and, moreover, it links risks to the sources
of uncertainty that generate them.
PHASES OF DEVELOPMENT OF THE REGULATORY
Banks have plenty of motives for developing risk-based practices and risk models. In
addition, regulators made this development a major priority for the banking industry,
because they focus on ‘systemic risk’, the risk of the entire banking industry made up
of financial institutions whose fates are intertwined by the density of relationships within
the financial system.
The risk environment has changed drastically. Banking failures have been numerous in
the past. In recent periods their number has tended to decrease in most, although not all, of
the Organization for Economic Coordination and Development (OECD) countries, but they
became spectacular. Banking failures make risks material and convey the impression that
the banking industry is never far away from major problems. Mutual lending–borrowing
and trading create strong interdependencies between banks. An individual failure of a large
bank might trigger the ‘contagion’ effect, through which other banks suffer unsustainable
losses and eventually fail. From an industry perspective, ‘systemic risk’, the risk of a
collapse of the entire industry because of dense mutual relations, is always in the background. Regulators have been very active in promoting pre-emptive policies for avoiding
individual bank failures and for helping the industry absorb the shock of failures when
they happen. To achieve these results, regulators have totally renovated the regulatory
framework. They promoted and enforced new guidelines for measuring and controlling
the risks of individual players.
Originally, regulations were traditional conservative rules, requiring ‘prudence’ from
each player. The regulatory scheme was passive and tended to differentiate prudent rules
for each major banking business line. Differentiated regulations segmented the market
and limited competition because some players could do what others could not. Obvious
examples of segmentation of the banking industry were commercial versus investment
banking, or commercial banks versus savings institutions. Innovation made rules obsolete,
because players found ways to bypass them and to compete directly with other segments
of the banking industry. Obsolete barriers between the business lines of banks, plus failures, triggered a gradual deregulation wave, allowing players to move from their original
business field to the entire spectrum of business lines of the financial industry. The corollary of deregulation is an increased competition between unequally experienced players,
and the implication is increased risks. Failures followed, making the need for reregulation
obvious. Reregulation gave birth to the current regulatory scheme, still evolving with new
guidelines, the latest being the New Basel Accord of January 2001.
Under the new regulatory scheme, initiated with the Cooke ratio in 1988, ‘risk-based
capital’ or, equivalently, ‘capital adequacy’ is a central concept. The philosophy of ‘capital
adequacy’ is that capital should be capable of sustaining the future losses arising from
current risks. Such a sound and simple principle is hardly debatable. The philosophy
provides an elegant and simple solution to the difficult issue of setting up a ‘pre-emptive’,
‘ex ante’ regulatory policy. By contrast, older regulatory policies focused more on corrective actions, or ‘after-the-fact’ actions, once banks failed. Such corrective actions remain
necessary. They were prompt when spectacular failures took place in the financial industry
(LTCM, Baring Brothers). Nevertheless, avoiding ‘contagion’ when bank failures occur
is not a substitute for pre-emptive actions aimed at avoiding them.
The practicality of doing so remains subject to adequate modelling. The trend towards
more internal and external assessment on risks and returns emerged and took momentum
in several areas. Through successive accords, regulators promoted the building up of
information on all inputs necessary for risk quantification. Accounting standards evolved
as well. The ‘fair value’ concept gained ground, raising hot debates on what is the ‘right’
value of bank assets and how to accrue earnings in traditional commercial banking activities. It implies that a loan providing a return not in line with its risk and cost of funding
should appear at lower than face value.
The last New Basel Accord promotes the ‘three pillars’ foundation of supervision: new
capital requirements for credit risk and operational risks; supervisory processes; disclosure
of risk information by banks. Together, the three pillars allow external supervisors to audit
the quality of the information, a basic condition for assessing the quality and reliability of
risk measures in order to gain more autonomy in the assessment of capital requirements.
Regulatory requirements for market, credit and operational risk, plus the closer supervision
of interest rate risk, pave the way for a comprehensive modelling of banking risks, and a
tight integration with risk management processes, leading to bank-wide risk management
across all business lines and all major risks.
FROM RISK MODELS TO RISK MANAGEMENT
Risk models have two major contributions: measuring risks and relating these measures to
management controls over risks. Banking risk models address both issues by embedding
the specifics of each major risk. As a direct consequence, there is a wide spectrum of
modelling building blocks, differing across and within risks. They share the risk-based
capital and the ‘Value at Risk’ (VaR) concepts that are the basic foundations of the new
views on risk modelling, risk controlling and risk regulations. Risk management requires
an entire set of models and tools for linking risk management (business) issues with
financial views on risks and profitability. Together, they make up the risk management
toolbox, which provides the necessary inputs that feed and enrich the risk process, to
finally close the gap between models and management processes.
Risk Models and Risks
Managing the banking exposure to interest rate risk and trading interest rate risk are
different businesses. Both commercial activities and trading activities use up liquidity that
financial institutions need to fund in the market. Risk management, in this case, relates to
the structural posture that banks take because of asset and liability mismatches of volumes,
maturity and interest rate references. Asset–Liability Management (ALM) is in charge of
managing this exposure. ALM models developed gradually until they became standard
references for managing the liquidity and interest rate risk of the banking portfolio.
For market risk, there is a large overlap between modelling market prices and measuring
market risk exposures of financial institutions. This overlap covers most of the needs,
except one: modelling the potential losses from trading activities. Market risk models
appeared soon after the Basel guidelines started to address the issues of market risk.
They appeared sufficiently reliable to allow internal usage by banks, under supervision of
regulators, for defining their capital requirements.
For credit risk, the foundations exist for deploying instrumental tools fitting banks’
requirements and, potentially, regulators’ requirements. Scarce information on credit
events remains a major obstacle. Nevertheless, the need for quantification increased over
time, necessitating measuring the size of risk, the likelihood of losses, the magnitude
of losses under default and the magnitude of diversification within banks’ portfolios.
Modelling the qualitative assessment of risk based on the fundamentals of borrowers
has a long track record of statistical research, which rebounds today because of the
regulators’ emphasis on extending the credit risk data. Since the early nineties, portfolio
models proposed measures of credit risk diversification within portfolios, offering new
paths for quantifying risks and defining the capital capable of sustaining the various levels
of portfolio losses.
Whether the banks should go along this path, however, is no longer a question since the
New Basel Accord of January 2001 set up guidelines for credit risk-sensitive measures,
therefore preparing the foundations for the full-blown modelling of the credit risk of
banks’ portfolios. Other major risks appeared when progressing in the knowledge of
risks. Operational risk became a major priority, since January 2001, when the regulatory
authorities formally announced the need to charge bank capital against this risk.
Capital and VaR
It has become impossible to discuss risk models without referring to economic capital and
VaR. The ‘capital adequacy’ principle states that the bank’s capital should match risks.
Since capital is the most scarce and costly resource, the focus of risk monitoring and
risk measurement follows. The central role of risk-based capital in regulations is a major
incentive to the development of new tools and management techniques.
Undoubtedly a most important innovation of recent years in terms of the modelling
‘toolbox’ is the VaR concept for assessing capital requirements. The VaR concept is a
foundation of risk-based capital or, equivalently, ‘economic capital’. The VaR methodology aims at valuing potential losses resulting from current risks and relies on simple
facts and principles. VaR recognizes that the loss over a portfolio of transactions could
extend to the entire portfolio, but this is an event that has a zero probability given the
effective portfolio diversification of banks. Therefore, measuring potential losses requires
some rule for defining their magnitude for a diversified portfolio. VaR is the upper bound
of losses that should not be exceeded in more than a small fraction of all future outcomes.
Management and regulators define benchmarks for this small preset fraction, called the
‘confidence level’, measuring the appetite for risk of banks. Economic capital is VaRbased and crystallizes the quantified present value of potential future losses for making
sure that banks have enough capital to sustain worst-case losses. Such risk valuation
potentially extends to all main risks.
Regulators made the concept instrumental for VaR-based market risk models in 1996.
Moreover, even though the New Accord of 2001 falls short of allowing usage of credit
models for measuring credit risk capital, it ensures the development of reliable inputs for
The Risk Management Toolbox
Risk-based practices require the deployment of multiple tools, or models, to meet the
specifications of risk management within financial institutions. Risk models value risks
and link them to their drivers and to the business universe. By performing these tasks, risk
models contribute directly to risk processes. The goal of risk management is to enhance
the risk–return profiles of transactions, of business lines’ portfolios of transactions and
of the entire bank’s portfolio. Risk models provide these risk–return profiles. The risk
management toolbox also addresses other major specifications. Since two risks of 1 add
up to less than 2, unlike income and costs, we do not know how to divide a global risk
into risk allocations for individual transactions, product families, market segments and
business lines, unless we have some dedicated tools for performing this function. The
Funds Transfer Pricing (FTP) system allocates income and the capital allocation system
allocates risks. These tools provide a double link:
• The top-down/bottom-up link for risks and income.
• The transversal business-to-financial sphere linkage.
Without such links, between the financial and the business spheres and between global
risks and individual transaction profiles, there would be no way to move back and forth
from a business perspective to a financial perspective and along the chain from individual
transactions to the entire bank’s global portfolio.
Risk Management Processes
Risk management processes are evolving with the gradual emergence of new risk
measures. Innovations relate to:
• The recognition of the need for quantification to develop risk-based practices and meet
risk-based capital requirements.
• The willingness of bankers to adopt a more proactive view on risks.
• The gradual development of regulator guidelines for imposing risk-based techniques,
enhanced disclosures on risks and ensuring a sounder and safer level playing field for
the financial system.
• The emergence of new techniques of managing risks (credit derivatives, new securitizations that off-load credit risk from the banks’ balance sheets) serving to reshape the
risk–return profile of banks.
• The emergence of new organizational processes for better integrating these advances,
such as loan portfolio management.
Without risk models, such innovations would remain limited. By valuing risks, models
contribute to a more balanced view of income and risks and to a better control of risk
drivers, upstream, before they materialize into losses. By linking the business and the risk
views, the risk management ‘toolbox’ makes models instrumental for management. By
feeding risk processes with adequate risk–return measures, they contribute to enriching
them and leveraging them to new levels.
Figure 1 shows how models contribute to the ‘vertical’ top-down and bottom-up processes, and how they contribute as well to the ‘horizontal’ links between the risk and return
views of the business dimensions (transactions, markets and products, business lines).
Comprehensive and consistent set of models for bank-wide risk management
THE STRUCTURE OF THE BOOK
The structure of the book divides each topic into single modular pieces addressing the
various issues above. The first section develops general issues, focusing on risks, risk
measuring and risk management processes. The next major section addresses sequentially
ALM, market risk and credit risk.
The structure of the book is in 17 sections, each divided into several chapters. This
structure provides a very distinct division across topics, each chapter dedicated to a major
single topic. The benefit is that it is possible to move across chapters without necessarily
following a sequential process throughout the book. The drawback is that only a few
chapters provide an overview of interrelated topics. These chapters provide a synthesis of
subsequent specific topics, allowing the reader to get both a summary and an overview
of a series of interrelated topics.
Risk Management Processes
Asset–Liability Management Models
Options and Convexity Risk in Banking
Mark-to-Market Management in Banking
Funds Transfer Pricing
Portfolio Analysis: Correlations
Credit Risk Models
Credit Risk: ‘Standalone Risk’
Credit Risk: ‘Portfolio Risk’
Portfolio and Capital Management (Credit Risk)
Building Block Structure
The structure of the book follows from several choices. The starting point is a wide
array of risk models and tools that complement each other and use sometimes similar,
sometimes different techniques for achieving the same goals. This raises major structuring
issues for ensuring a consistent coverage of the risk management toolbox. The book relies
on a building block structure shared by models; some of them extending across many
blocks, some belonging to one major block.
The structuring by building blocks of models and tools remedies the drawbacks of
a sequential presentation of industry models; these bundle modelling techniques within
each building block according to the model designers’ assembling choices. By contrast,
a building block structure lists issues separately, and allows us to discuss explicitly the
various modelling options. To facilitate the understanding of vendors’ models, some chapters provide an overview of all existing models, while detailed presentations provide an
overview of all techniques applying to a single basic building block.
Moreover, since model differentiation across risks is strong, there is a need to organize the structure by nature of risk. The sections of the book dealing directly with risk
modelling cross-tabulate the main risks with the main building blocks of models. The
main risks are interest rate risk, market risk and credit risk. The basic structure, within
each risk, addresses four major modules as shown in Figure 2.
Risk drivers and transaction
Top-down & bottom-up
Risk & return measures
The building block structure of risk models
The basic blocks, I and II, are dedicated by source of risk. The two other blocks, III and
IV, are transversal to all risks. The structure of the book follows from these principles.
The focus is on risk management issues for financial institutions rather than risk modelling
applied to financial markets. There is an abundant literature on financial markets and financial derivatives. A basic understanding of what derivatives achieve in terms of hedging or
structuring transactions is a prerequisite. However, sections on instruments are limited to
the essentials of what hedging instruments are and their applications. Readers can obtain
details on derivatives and pricing from other sources in the abundant literature.
We found that textbooks rarely address risk management in banking and in financial
institutions in a comprehensive manner. Some focus on the technical aspects. Others focus
on pure implementation issues to the detriment of technical substance. In other cases, the
scope is unbalanced, with plenty of details on some risks (market risk notably) and fewer
on others. We have tried to maintain a balance across the main risks without sacrificing
The text focuses on the essential concepts underlying the risk analytics of existing
models. It does detail the analytics without attempting to provide a comprehensive
coverage of each existing model. This results in a more balanced view of all techniques
for modelling banking risk. In addition, model vendors’ documentation is available
directly from sites dedicated to risk management modelling. There is simply no need to
replicate such documents. When developing the analytics, we considered that providing a