Empirical analysis of structural change in Credit Default Swap volatility

The purpose of this paper is to study the structural change in Credit Default Swap volatility. We use statistical properties and a network approach to better understand the behavior of CDS volatility. We hypothesize that structural change occurs in CDS index during a financial crisis and it requires subperiod analysis, rather than full period analysis, to investigate properly. Our results show that the probability of large volatility is related to the structure of volatility but it is more significantly related to the size of volatility. Both the memory property and the size of volatility are confirmed to have dependence on the structure of volatility. The linked degree of CDS volatilities is highly related to the probability of large volatility and its predictability, regardless of structural change in volatility. Another interesting result is that the CDS volatility of a country is more related to the behavior of other volatilities, not the geographical location.