N°24-67: A New Test for an Old Puzzle

Traditional tests of the credit spread puzzle are imprecise. We show that a key input of these tests, the equity Sharpe ratio, is difficult to estimate in available empirical samples. Instead, we show that the difference between the Sharpe ratios of debt returns and equity returns provides a more precise test. Using the new test we find no evidence for a credit spread puzzle in the corporate bond market: after controlling for the unexpected decline in interest rates beginning in the 1990s, the Sharpe ratios of equity and bonds of high quality firms are nearly identical. However, we uncover a new puzzle in the market for bank debt: loan returns have higher Sharpe ratios than equity returns.