Detecting Regime Shifts in Corporate Credit Spreads
by Georges Dionne of HEC Montreal,
Abstract: Studies about credit spread switching regimes typically make assumptions about the number of regimes for in-sample regime detection. This is because exploratory regime detection techniques are lacking in the literature. We employ a real time sequential technique to detect possible breakpoints in the mean and the variance of credit spreads. Our evidence shows that regime shifts are closely related to systematic shocks. Detected shifts in the mean and the variance have different patterns that provide new insights on the relation between economic and credit cycles. We also show that the employed out-of-sample detection technique can be valuable for market timing.
Keywords: Credit spread regimes, shifts in the mean and the variance, credit cycle, economic cycle, market timing.