Financial Risk Management: Models, History, and Institutions

CHAPTER 8

Portfolio Credit Risk

In this chapter, we extend the study of credit risk to portfolios containing several credit-risky securities. We begin by introducing the most important additional concept we need in this context, default correlation, and then discuss approaches to measuring portfolio credit risk.

A portfolio of credit-risky securities may contain bonds, commercial paper, off-balance-sheet exposures such as guarantees, as well as positions in credit derivatives such as credit default swaps (CDS). A typical portfolio may contain many different obligors, but may also contain exposures to different parts of one obligor's capital structure, such as preferred shares and senior debt. All of these distinctions can be of great importance in accurately measuring portfolio credit risk, even if the models we present here abstract from many of them.

In this chapter, we focus on two approaches to measuring portfolio credit risk. The first employs the factor model developed in Chapter 6, the key feature of which is latent factors with normally distributed returns. Conditional on the values taken on by that set of factors, defaults are independent. There is a single future time horizon for the analysis. We will specialize the model even further to include only default events, and not credit migration, and only a single factor. In the CreditMetrics approach, this model is used to compute the distribution of credit migrations as well as default. One could therefore label the approach ...