Whether credit rating agencies provide investors with private credit risk information is a central yet open empirical question. I use the introduction of a new rating type by Standard & Poor's reflecting expected recovery outcomes to isolate the effect of the information component in credit ratings on security prices. Studying the relationship between bond yield spreads and rating-implied recovery assessments, I find that bondholders welcome positive news about expected recovery rates. On the contrary, positive news about expected recovery rates lead to negative stock returns and an increase in implied volatilities of equity options. The results suggest that information in credit ratings is valuable to market participants.
Distress Risk, Expected Shareholder Losses, and the Cross-Section of Expected Returns, with Robert F. Dittmar
We analyze the impact of financial distress on expected stock returns accounting for the severity of default. Whereas we find that shareholders' average losses amount to 32% of equity value from the day before to the day after a bankruptcy filing, outcomes differ substantially among firms. We specify and estimate a model that predicts these losses for individual firms. When sorting stocks into portfolios according to their predicted bankruptcy losses, we find that a long-short strategy buying stocks with high predicted bankruptcy losses and selling stocks with low predicted bankruptcy losses earns a monthly premium of 0.5%. We also sort stocks independently into quintiles according to predicted bankruptcy loss and failure probability. We find that the distress-risk puzzle is not present among stocks in the highest quintile of predicted bankruptcy loss. On the contrary, the long-short returns of stocks sorted by predicted bankruptcy losses are strongest for stocks in the highest quintile of failure probability.
We develop a new methodology to extract market expectations of recovery rates that uses information from credit default swap spreads on debt instruments of different seniorities, incorporates information on the firm-specific liability structure and allows for deviations from the absolute priority rule. In our empirical analysis, we find that expected recovery rates exhibit a large cross-sectional and time-series variation and that recovery rates of financial institutions are on average larger than those of non-financial companies. Using a panel regression, we show that anticipated government support increases the market expectations of recovery rates of financial companies and therefore helps to explain these differences.