Modeling Expected Loss

This paper develops a methodology for modeling and estimating expected loss over arbitrary horizons. We jointly model the probability of default and the recovery rate given default. Different model specifications are estimated using an extensive default and recovery data set that contains the majority of defaults between 1980–2004 of AMEX, NYSE and NASDAQ listed companies. We undertake extensive out-of-sample performance tests for both the default prediction models and recovery rate given default models. Under the joint model specification, we find that the probability of default and the recovery rate given default are negatively correlated out-of-sample, and that the magnitude of the correlation varies with the credit cycle. We also compare the accuracy of the out-of-sample one year ahead default predictions using quarterly and annual data. 1∗Texas A&M, Business School,∗∗London Business School and ∗∗∗Bauer College of Business