Correlated Defaults of UK Banks: Dynamics and Asymmetries

We document asymmetric and time-varying features of dependence between the credit risks of global systemically important banks (G-SIBs) in the UK banking industry using a CDS dataset. We model the dependence of CDS spreads using a dynamic asymmetric copula. Comparing our model with traditional copula models, we find that they usually underestimate the probability of joint (or conditional) default in the UK G-SIBs. Furthermore, we show that dynamics and asymmetries between CDS spreads are closely associated with the probabilities of joint (or conditional) default through the extensive regression analysis. Especially, our regression analysis provides a policy implication that copula correlation or tail dependence coefficients are able to be leading indicators for the systemic credit event.