The purpose of this paper is to discuss issues relating to stress testing methods for credit risks in banks. Also, it suggests a solution to bank supervisors on evaluating stress test results.
Discussion is based on cases analysis on a stress period of the Hong Kong banking sector.
The paper finds that econometric modeling does not work well modeling stress scenarios. The stressed probability of default (PD) provided by Basel II would be much higher than stressed PD observed in the history.
Bank supervisors should develop cost‐effective methods to monitor the stress test results reported by banks.
The paper addresses the issues of stress testing and provides a practical solution for bank supervisors to monitor stress test results reported by banks.
