To which extent can non-default components explain CDS (Credit Default Swap) spreads is under debate in the literature. Unlike other research applying conventional structural or reduced- form models, this study investigates this issue by conducting a principal component analysis anda non-parametric local linear regression using corporate CDS data during the period from 2001 to 2011, which includes the recent global financial crisis. A model with two market state factors, approximated by the firs