Tinbergen Institute Discussion Paper No. 15-029/III/DSF87
Tinbergen Institute, Amsterdam and Rotterdam
This paper investigates the dynamic properties of systematic default risk conditions for firms from different countries, industries, and rating groups. We use a high-dimensional nonlinear non-Gaussian state space model to estimate common components in corporate defaults in a 41 country sample between 1980Q1-2014Q4,covering both the global financial crisis and euro area sovereign debt crises. We find that macro and default-specific world factors are a primary source of default clustering across countries. Defaults cluster more than what is implied by shared exposures to macro factors, indicating that other factors are of high importance as well. For all firms, deviations of systematic default risk from macro fundamentals are correlated with net tightening bank lending standards, implying that bank credit supply and systematic default risk are inversely related.
systematic default risk credit portfolio models frailty-correlated defaults international default risk cycles state space methods