Higher Volatility with Lower Credit Spreads: The Puzzle and Its Solution

This dissertation explains the puzzling negative relationship between changes in stock volatility and credit spreads of corporate bonds. This relationship has been encountered in some empirical studies but has remained unexplained in the theoretical literature, which unanimously suggests the opposite relationship. This dissertation shows that this negative relationship can be produced by the dynamic endogenous asset composition of borrowing firms. On the one hand, higher asset volatility corresponds to lower future volatility of the firm's investments and lower credit spreads if the firm can reallocate resources optimally. On the other hand, short-term stock volatility corresponds to the current allocation of resources and thus increases with asset volatility. The combination of these two effects produces the negative relationship between changes in stock volatility and credit spreads. The empirical part of the dissertation shows that the relationship between changes in stock market volatility and credit spreads of long-term, high-quality corporate bonds (controlling for other variables) is negative, robust, and economically significant. Consistent with the predictions in this dissertation, the corresponding regression coefficient is a U-shaped function of the credit quality of the bonds. In addition, the dissertation shows that the relationship changes its sign in distressed market conditions and that a combination of normal and distressed market conditions can produce erroneous results