What Drives the Cross-Section of Credit Spreads?: A Variance Decomposition Approach

I decompose the cross-sectional variation of the credit spreads for corporate bonds into changing expected returns and changing expectation of credit losses with a model-free method. Using a log-linearized pricing identity and a vector autoregression applied to micro-level data from 1973 to 2011, I find that the expected credit loss component and the excess return component each explains about half of the variance of the credit spreads. Unlike the market-level findings in Gilchrist and Zakrajsek (2012), at the firm level, the expected credit loss is volatile and affects the firms' investment decision more than the expected excess returns.