The Global Credit Spread Puzzle

Abstract

We examine the ability of structural models to predict credit spreads using global default data and security-level credit spread data in eight developed economies. We find that two representative, pure default-risk models tend to underpredict the average credit spreads on investment-grade (IG) bonds, especially their spreads over government bonds, thereby providing evidence for a “global credit spread puzzle.” However, a model incorporating endogenous liquidity in the secondary debt market helps mitigate the puzzle. Furthermore, the model captures certain determinants of corporate bond market frictions across the eight economies and substantially improves the cross-sectional fit of individual IG credit spreads.