Credit Risk Modeling Using Excel and VBA with DVD

7

Measuring Credit Portfolio Risk with the Asset Value Approach

A credit portfolio risk model produces a probability distribution of losses that can arise from holding a portfolio of credit risky instruments. A financial institution can use such models to answer questions such as: ‘What is the probability that losses on my loan portfolio exceed 100m over a horizon of one year?’

The annus mirabilis of portfolio credit risk models is 1997, which saw the publication of three different approaches; a fourth approach was developed at about the same time.¹

Even though extant models are similar in underlying structure, it is beyond the scope of this chapter to provide thorough implementations of each. Accordingly, we cover just one approach – the asset value or latent variable approach exemplified by CreditMetrics. In this approach, the portfolio loss distribution is obtained through a Monte Carlo simulation. Computing time is thus an important implementation issue. To keep focused on this issue, we start with a simplified framework in which we just consider losses from default (but not from changes in market value). We then show how to speed up simulations, and conclude with some generalizations.

A DEFAULT-MODE MODEL IMPLEMENTED IN THE SPREADSHEET

We can split portfolio credit risk modeling into four main steps. In the following, we describe those steps for a general model and for a specific approach – a default-mode model in which we consider only losses from default:

1. Specify probabilities ...