Credit Spreads and Regime Shifts

The Nonlinear Dynamics of Corporate Bond Spreads: Regime-Dependent Effects of their Determinants

“…This finding corroborates our choice of only two regimes, which is consistent with the number of regimes set or identified in the majority of other studies (e.g. Davies, 2008;Maalaoui Chun et al, 2014;Pavlova et al, 2015).…”

The Time-Varying Impact of Systematic Risk Factors on Corporate Bond Spreads

“…Moreover, the paper confirms that the impact of key corporate bond spread determinants alters in time and can, for some variables, even change the direction. Pavlova et al (2015) also apply a regime switching model to US bond data. However, unlike Davies (2004), Chun et al (2014a), and our paper, they analyze daily data which limits the explanatory variables to term structure variables, stock market returns, and measures of market volatility.…”

“…Chun, Dionne, and François (2014a) reveal that corporate credit spreads can be explained much better by regime switching approaches than by simple constant parameter models. Interestingly, prior work on regime dependence in corporate bond spreads focuses exclusively on the US market (Davies (2004), Chun, Dionne, and François (2014b), Chun et al (2014a), Pavlova, Hibbert, Barber, and Dandapani (2015)). Given different market characteristics 3 in distinct regions, it seems unreliable to simply transfer the results for the US to other markets.…”

“…Among our most significant variables are changes in the unemployment rate and in an index for economic sentiment, nothing of which is used in the previous literature on regime switching in corporate bond spreads. The other way around, industrial production and implied stock market volatility, which are employed by Davies (2004) and Pavlova et al (2015), do not enter any of our specifications. Furthermore, our models have significantly higher explanatory power compared with these studies although we use a comparable number of exogenous variables.…”

“…Another aspect differentiating our paper from existing studies is that we apply a Markov switching model with time-varying transition probabilities as proposed by Diebold, Lee, and Weinbach (1994) in order to shed light on the economic forces driving the shifts in regimes. Chun et al (2014b) and Pavlova et al (2015) use two-step approaches where they, first, extract regimes and, then, try to explain these regimes by economic variables. This is methodologically unsound and may bias the results as the inclusion of these variables in the estimation process may, of course, affect the determination of regimes.…”