Default Likelihood under Regime-Switching
by Andreas Milidonis of University of Cyprus, and
Feb 02, 2012
Abstract: Default risk can be estimated from equity returns using Merton's (1974) default risk (MDR) model. Changes in default risk (bond ratings) have been associated with preceding changes in equity mean returns and volatility. We develop the regime-switching default risk (RSDR) model that supports an expanded range of asset probability density functions, and can accommodate skewness, excess kurtosis and sudden changes in asset mean returns and volatility. The significance of our model lies in both the level and changes of estimated default probabilities. First, we show using simulation that the RSDR model incorporates sudden changes in asset values faster than the MDR model, especially in times of deteriorating creditworthiness. Second, we empirically implement the RSDR, MDR and an extension of the MDR model with jumps, using maximum likelihood on the sample of downgrades by a rating agency using only publicly available information in its ratings. We find that the RSDR model uses structural breaks preceding downgrades to produce higher estimated default probabilities, faster, than both benchmark models.
Keywords: Structual Default Risk Model, Regime Switching, Bond Ratings, MLE.