Credit Spread Changes within Switching Regimes

Many empirical studies on credit spread determinants consider a single-regime model over the entire sample period and find limited explanatory power. We model the credit cycle independently from macroeconomic fundamentals using a Markov regime switching model. We show that accounting for endogenous credit cycles enhances the explanatory power of credit spread determinants. The single regime model cannot be improved when conditioning on the states of the NBER economic cycle. Furthermore, the regime-based model highlights a positive relation between credit spreads and the risk-free rate in the high regime. Inverted relations are also obtained for some other determinants.