Credit Risk Measurement In and Out of the Financial Crisis: New Approaches to Value at Risk and Other Paradigms, Third Edition

CHAPTER 9

The VAR Approach: CreditMetrics and Other Models

INTRODUCTION

In the previous two parts of this book, we provided objective empirical models that can be used to estimate the critical parameters of credit risk assessment: PD and LGD, and their correlations. Now, we can put them all together in an integrated model that incorporates these parameters in order to assess credit risk.

Since 1993, when the Bank for International Settlements (BIS) announced its intention to introduce a capital requirement for market risk, great strides have been made in developing and testing value at risk (VAR) methodologies. The incentive to develop internal VAR models was given a further boost in 1996, when the BIS amended its market risk proposal and agreed to allow certain banks to use their own internal models, rather than the standardized model proposed by regulators, to calculate their market risk exposures. Since the end of 1996 in the European Union and 1998 in the United States, the largest banks (subject to regulatory approval) have been able to use their internal models to calculate VAR exposures for their trading book and, thus, capital requirements for market risk.¹

In this chapter, we first review the basic VAR concept and then look at its potential extension to nontradable loans and its use in calculating the capital requirement for loans on the bank’s books. Considerable attention will be paid to CreditMetrics, originally developed by J.P. Morgan in conjunction with several ...