The credit premium

In this section, we review the theory and empirical evidence of the credit premium. The credit premium is the excess return that an investor obtains for holding bonds issued by entities other than governments. A natural starting point for this objective is to discuss the so-called “credit spread puzzle” and different attempts to resolve it.

18 March 2011

Main findings

  • The spread of BBB-rated (three- to five-year maturity) corporate bonds over Treasuries (the credit spread) averaged 170 basis points p.a. during the period 1997-2003, while the total loss from default for the same bonds averaged 20 basis points p.a. for the same period. Clearly, the credit spread has historically compensated for more than the expected loss from default.

  • The credit spread puzzle refers to the observation that structural models such as the one proposed by Merton (1974) have failed to explain the high excess returns received by corporate bondholders historically. Key assumptions of the structural models from which the puzzle arose were time-invariant default probabilities and recovery rates. The puzzle suggests either that the static assumptions of the Merton model may be too restraining or that components other than default and recovery risk affect the credit spreads.

  • Broadly, the academic literature can be classified into approaches based on reduced-form models and approaches based on structural models.

  • Reduced-form models have attempted through statistical analysis to identify factors, including non-default-related factors, that can account for the observed credit spreads, such as liquidity, tax, equity volatility and interest rate structure. • Structural models combine economic theory, measurements and identification to quantitatively account for observed credit spreads.

  • Recent structural models argue that credit spreads can be accounted for by extending the standard models along the same dimensions that have previously been used to account for the equity premium puzzle. These structural models incorporate time-varying reward-to-volatility ratios and can capture both the level and time-variation of historical spreads. This strand of literature suggests that credit spreads are not primarily driven by credit-specific idiosyncratic risk, but rather by the same common systematic risk factors that drive other security prices like equity.  

  • The academic literature finds that there is a positive and time-varying credit premium. This premium is typically seen as compensation for two main types of risk: default risk and recovery risk. The former refers to the risk of an issuer defaulting, while the latter is the risk of receiving less than the promised payment if the issuer defaults.

Download the discussion note (PDF)