An Empirical study of the relation between equity implied probabilities of default and fundamental information

The risk of credit exposures to entities whose equity is not exchange traded cannot be estimated directly using Black-Scholes-Merton (BSM) default probabilities requiring the market value of equity. We examine the extent to which firm level accounting variables; aggregate market data and other observable macroeconomic variables capture default related information from empirical estimates of BSM default likelihoods. Our findings indicate that close to half of the variation in BSM default probabilities is captured by accounting fundamentals used in prominent credit scoring models and fluctuations in industry level default expectations. We also show that much of the time variation in the underlying risk of default associated with the state of the macro-economy, inferred from aggregated estimates of BSM, can be usefully summarized with a parsimonious set of market variables that are commonly used to measure time variation in asset returns. Taken together, our results suggest that much of the timely information in equity implied probabilities of default are recoverable using a combination of firm level accounting information and aggregate market data. These findings have practical implications for modeling the point in time risk of default associated with exposures to private firms using fundamental information. Further, by calibrating our fundamentals based model to expectations we circumvent the usual modeling requirement of historical default observations.