Encyclopedia of Financial Models I

An Introduction to Credit Risk Models

DONALD R. VAN DEVENTER, PhD

Chairman and Chief Executive Officer, Kamakura Corporation

Abstract: Credit risk technology has evolved with advances in computer science and information technology. Traditional credit ratings date back to 1860, an era when the cost of collecting and analyzing corporate credit information was high. The commercial advantages of a central provider of credit risk analysis were high. With the advent of better computer technology and databases of corporate financial information and stock prices, quantitative approaches to credit risk assessment have become more popular and increasingly accurate. Credit scoring is a quantitative approach to retail credit assessment, but, in the corporate world, more and more credit analysts prefer a default probability with an explicit maturity to a “credit rating” or “credit score.”

This entry introduces the topic of credit risk modeling by first summarizing the key objectives of credit risk modeling. We then discuss ratings and credit scores, contrasting them with modern default probability technology. Next, we discuss why valuation, pricing, and hedging of credit risky instruments are even more important than knowing the default probability of the issuer of the security. We review some empirical data on the consistency of movements between common stock prices and credit spreads with some surprising results. Finally, we compare the accuracy of ratings, the Merton model of risky ...