The Co-movement of the Czech Republic, Hungary and Poland Sovereign Credit Default Swaps Spreads

Abstract

We use a VEC DCC M-GARCH model to investigate the daily co-movement of the Czech Republic, Hungary and Poland one-, five- and ten-year sovereign credit default swap (CDS) spreads in the period Jan 2009–May 2018. To control for a systemic risk stemming from the EU and other international markets, we nest the analysis within a four-variate system including the Germany CDS spread and the CBOE VIX. The latter serves us as a proxy for the exogenous driver of spreads. The analysis shows that the long-run dependence among the logs of CDS spreads is rare. It is only the Czech Republic and Germany one-year CDS spreads that exhibit a common stochastic trend. The remaining spreads do not co-integrate. Each country maturity time t log change in the spread depends upon that of time \(t - 1\) and earlier. The dynamics of spreads are country specific. The Hungary CDS spreads Granger cause almost all their counterparts. The causality running other way round is incidental. The median of pairwise conditional correlation estimates among the countries of interest differs across the maturities in the way indicating that the Czech, Hungarian and Polish markets are better integrated among one another than any single with the German market.