Problems & Solutions: Probabilities of Default

Michael Ong

Abstract

1. Probabilities of Default

One peculiarity of credit markets is the large difference between probabilities of default calculated from historical data and probabilities of default implied from bond prices or from credit default swaps. For example, using historical data Moody's estimates that about 0.9% of companies that are initially rated A default within a sevenyear period, while the bond market indicates that about 8.6% of these companies are expected to default in the next seven years. Why are these two estimates so different and which one should be used in practice?

Editor's note: John Hull, Mirela Predescu and Alan White will present a potential answer to this conundrum in the Credit Risk Forum section in the next issue.

2. Credit Spreads

Credit Spreads Credit investors need a measure to determine how much they are being paid to compensate them for assuming the credit risk embedded within a security. A number of such measures exist and are commonly known as "credit spreads", since they attempt to measure the return of the credit asset relative to some benchmark with higher credit quality. What are these commonly used credit spreads and what exactly do they measure?

Editor's note: In the next issue, Dominic O'Kane and Saurav Sen present a variety of suggestions to these questions using fixed-rate bonds, floating-rate notes and credit default swaps.