Idiosyncratic Downside Risk and the Credit spread Puzzle

The puzzle is that spreads on corporate bonds are about twice as large as can be explained by defaults, taxes and illiquidity. The higher a bond’s rating and the shorter its maturity, the greater is the puzzle. We use a large dataset of bonds to identify the relevant risk factors. Systematic factors fail to generate large spreads, regardless of whether they are conventional (market covariance, size and book-to-market) or have a downside character (co-skewness and downside covariance). Idiosyncratic factors are much more successful in generating plausible spreads, not only idiosyncratic equity volatility but especially idiosyncratic bond value-at-risk. Our explanation of this result is twofold. First, idiosyncratic risk matters because a bond is priced from the risk-neutral distribution of firm value. Second, bond value-at-risk is a good proxy for the fat left-hand tail of that distribution, which reflects a large risk-premium for extreme returns on the downside.