Understanding Market, Credit, and Operational Risk:
The Value-at-Risk approach
You may recall a
few years ago that Anthony Saunders published a largely non-technical book
looking at various commercial credit risk models. He teamed up with Linda Allen
to produce a (2002) second
edition. Now Jacob Boudoukh has joined the team, and they are out with a new
book. This one extends the discussion to market risk and operational risk.
Following an
introductory chapter, the authors start with market risk, focusing on
value-at-risk and stress testing. They devote two chapters to this but seem
torn between wanting to offer an elementary introduction and debating certain
technical issues. The first of the two chapters discusses inference procedures.
Almost immediately, it delves into a discussion of whether observed
leptokurtosis in unconditional returns can be explained by conditional
heteroskedasticity in those returns. The authors don't use the technical terms I
just did, but readers who are new to value-at-risk will be lost anyway. The
topic is technical. The discussion drags on for nine pages and is marred by
theoretical misstatements. Once it is complete, the authors return to more basic
topics—uniformly-weighted moving averages, exponentially-weighted moving
averages, GARCH in one dimension, etc. The next chapter offers a non-technical
explanation of linear, Monte Carlo, and historical transformation procedures. It
also offers a nice discussion of stress testing.
Contents
1. Introduction to Value at Risk
(VaR)
2. Quantifying Volatility in VaR
Models
3. Putting VaR to Work
4. Extending the VaR Approach to
Non-tradable Loans
5. Extending the VaR Approach to
Operational Risks
6. Applying VaR to Regulatory Models
7. VaR: Outstanding Research
Chapter 4 turns to
credit risk. It looks at traditional techniques of credit analysis and credit
scoring. It then explains the more modern Merton and reduced form models for
credit risk. The balance of the chapter focuses on the CreditMetrics and
Mark-to-Future commercial models.
A chapter
considers models for operational risk. It briefly considers a lot of different
models—plenty of breath but little depth. Another chapter
describes the modeling of market, credit and operational risk under the Basle II
capital requirements.
As with the
authors' earlier (and quite popular) book on credit risk modeling, discussions
are largely informal. Formulas are used. Some are quite technical, but they are
not central to the discussion. Non-technical readers can largely ignore the
formulas and still follow discussions. Don't plan on implementing any of the
models. This is an intuitive overview.
The one
significant shortcoming of the book is its brevity. At just 286 pages, it can't
possibly do justice to market risk, credit risk, operational risk, and Basle II.
A reasonable treatment of any one of these would exceed 300 pages. The book's
title suggests that value-at-risk might be a unifying theme tying the various
topics together, but this isn't the case. Topics are "siloed". If you want to
master them all, you might do better to buy a separate book on each. You will
spend more money, but you will learn much more.
.jpg)
.gif)
