Can the Performance of Structural Corporate Bond Models Be Improved?
Craig W. Holden
Indiana University - Kelley School of Business - Department of Finance
March 9, 2010
We analyze variations of three risk determinants of the Extended Merton structural corporate bond model. We consider three alternative non-Gaussian distributions, varying recovery rates, and the possibility of early default or default at maturity. We test a sample of 79 corporate bonds from 1987 to 1998. We obtain four empirical findings. First, for the full sample we find that our variations of Extended Merton dominate the Black and Cox first passage time benchmark in out-of-sample performance. Second, we test whether the risk of early default is priced. We find that it is not priced for high credit rating, short maturity bonds and that it is priced for all other bonds. Third, we test whether the risk of a low recovery rate in severe default is priced. We find that it is priced for low credit rating, long maturity bonds and it is not priced for all other bonds. Fourth, we obtain a negative relation between cumulative default rates and recovery rates. Overall, we document for the first time how risk determinants vary to generate credit spreads for different rating/maturity categories.
Number of Pages in PDF File: 58
Keywords: credit risk modeling, structural model, default risk, corporate bond pricing, non-Gaussian distribution, jump-diffusion, stochastic volatility
JEL Classification: C51, C52, C53, G12, G13
Date posted: March 19, 2010