The conditional dependence structure of insurance sector credit default swap indices

• We studied the dependence between the US, EU, and UK insurance sector CDS indices.
• We adopted a copula-GARCH approach to capture complex interrelationships.
• Rank correlation measures increased substantially during the financial crises.
• We found asymmetric tail dependence for the US–UK pair prior to the crises.
• The Frank copula with zero tail dependence was the best fit during the crises.

This study assesses the dependence structure of insurance sector credit default swap indices, using a copula-GARCH approach. We use daily data of the US, EU, and UK insurance sectors, covering the period from January 2004 to June 2013. We find substantial increases in dependence during the financial crisis periods. Prior to the crises, various copulas are found to best fit each pair; specifically, asymmetric tail dependence is found for the UK–US pair, suggesting the possibility of large simultaneous losses. However, during the crisis periods, the Frank copula fits best, with no significant tail dependence detected, implying low systemic risks.