Default forecasting considering correlation between business and credit cycles

Bank’s major approach in her internal rating system is credit scoring valuation which focused on corporates’ idiosyncratic risks and based on their financial indexes. Hence, an influence on corporates’ credit risks by business variation is not considered in her system. We model the effect on corporates’ credits by macroeconomic variables and analyze it. Firstly we model a corporate’s credit variable by the credit cycle index and her idiosyncratic risk factor, and consider the correlation between the business and credit cycles. And we decompose the business cycle into a trend and a cycle using Hodrick-Prescott filter and show we can build the more explanatory model than one based on macroeconomic variables themselves. Secondly we optimize the weights of credit cycle index by some distance measures for Japanese corporates and quantify the severity on historical Japanese corporates’ credits by macro economy. -- business cycle ; credit cycle index ; Hodrick-Prescott filter ; probit model ; distance measure