China's Commercial Banks' Internal Rating Model And Empirical Test

With the improvement of market degree of Chinese economy operation and the continuous intensification of reform, the credit risk resulted from asymmetrical information between the commercial banks and the firms will become the main risk that the commercial banks will confront with .The accurate measurement of the credit risk is the prerequisite to control and dissolve the credit risk. Recently, under the trend to strengthen the credit risk management of commercial bank, especially, after the Basel Committee, the international organization of bank supervision, has issued the "New Basel Capital Accord", many researchers and bank administrators from different countries have commonly focused their attention on how to accurately measure the credit risk by means of new approaches to value at credit risk and implementing of the Internal Rating Based Approaches(IRB).IRB is the basic of credit risk rating system, which is used to dissolve the credit risk in our commercial banks. But we should admit the fact that there are so many defects in the risk management system under using, as in the rating indexes, rating method, the process of rating result checking-out and so on. In order to keep up with the advanced approaches which international banks use nowadays, in this paper we present a new credit risk rating model, based on Factor Analysis and Cluster Analysis.The paper on the bases of analysis of fundamental theories about credit risk rating model and IRB approaches in commercial banks, which casts back the involvement of the IRB in commercial banks and the new approaches to value at credit risk along with their applications in IRB, so that we could form a comprehensive understanding and evaluation of IRB. After that, considering the shortcoming of recent credit risk rating on rating indexes and rating method in our commercial banks credit rating system, we put forward a new model to IRB based on Factor Analysis and Cluster Analysis, It turns the matter of measuring credit risk to the measuring of financial condition, which is shown by financial indexes, leading to compute borrower's default rate. Via this the rating indexes become more scientific and suitable, which means less credit information iterative and missing, it also leads to improving the accuracy of the rating system. At last we verify the model by sample to confirm the availability of the new model.