Industry specific defaults

In this paper, the hidden common factor for a default correlation model is expanded to industry. By introducing industry-specific hidden factors as random effects, a comparison is made of the relative scale of within- and between-industries correlations. Empirical analysis is based on 14,249 U.S. public firms between 1990 and 2014. A comparison study among the without-hidden-factor model, the common-hidden-factor model, and our industry-specific common-factor model show that an industry-specific common factor is necessary for adjusting time and industry specific over- or under-estimation of default probabilities. The Monte Carlo EM algorithm is adopted for model estimation.