Research On CreditRisk+ Model Under The Volatile Default Rate

Now, commercial banks'risk management capabilities have become core competencies based the market. But how to improve the risk management level of Chinese commercial banks? Under these conditions that commercial banks still lack many things such as risk management technology, databases and so on, Chinese commercial banks should design precise credit risk measurement models according to their actual situation and in the light of the international banking industry's best practices, using Basel New Capital Accord as the guide. In this context, this dissertation aims at providing a more precise and more effective method to measure unexpected losses under the volatile default rate for Chinese commercial banks.Under the volatile default rate, the original CreditRisk+ model has several deficiencies, for example that original bands division method, the inherent shortcomings of Panjer algorithm, as well as non-coherent of the VaR method all affect the accuracy of the model. This dissertation analyses the corresponding revised methods against some inherent deficiencies of this original model.In this dissertation, we consider the effect of systematic risk factors and pairwise correlation to debtors'default rate, which is much closer to the real word; The standard deviations of obligors'default rates are estimated through the loan default table method, which is more precise,and using this method we have solved the measuremet problem of the default rate and its related design parameters in the CreditRisk+ model; The combining use of the weighted average method of band division, Panjer algorithm and Saddlepoint approximation can increase the calculated accuracy and robustness of the CreditRisk+ model; And in this paper, we use VaR and ES to describe unexpected loss, which is more precise.This dissertation empirically analyses using the corporate loan data of a city commercial bank in China,and the study result shows that the revised CreditRisk+ model can measure unexpected losses of loan portfolios more precisely and more effectively for commercial banks.This dissertation puts forward some suggestions for Chinese commercial banks: The revised CreditRisk+ model under volatile default rate combined with the RAROC model can help to improve the credit risk management system of commercial banks, such as measurement of the economic capital, economic capital allocation, performance evaluation, pricing loans, and loan approval limits and so on.