Credit Risk Calibration Based on CDS Spreads

Abstract

As observed in the financial crisis, CDS spreads tend to increase simutaneously as a reaction to common shocks. Focusing on the spillover effects triggered by extreme events, we propose a credit risk analysis tool by applying credit default swap spread returns to the concept of 4CoVaR suggested by Adrian and Brunnermeier (2011). The interconnection and mutual impact on credit spreads are investigated based on CDS spreads of the biggest derivative dealers in the market. By including factors identified as determinants of CDS spreads to the set of explanatory variables such as equity return and equity volatility and implementing the variable selection technique least absolute shrinkage and selection operator (LASSO), the results demonstrate an improved performance in CDS spread VaR calculation. The enhancement is more significant in pre-crisis period but both methodologies tend to overestimate risk in turbulent period. Further, non-linear effects between CDS spreads in extreme events are captured by the introduction of a partial linear model in the CoVaR calculation.