The Construction Of The Early-Warning System In Commercial Bank Credit Risk In China

The international financial crisis outbreaked in 2008 is rooted in U.S. sub-prime loans ,primarily the sub-prime mortgage lending institutions, and then the risk was enlarged and shifted through a variety of credit derivative products ,so that the impact of credit risk are continuously expanding, finally led to the sweeping world financial crisis . Although China's banking industry was less affected, despite the concerted efforts of governments and banks, China's non-performing loan ratio of banks has been greatly decreased, but because of the credit risk is the main risk facing the banking sector, involving economic entities over more, and its accumulated to a certain extent, may bring about disastrous consequences to the bank, and even affect the entire economic operation. Therefore, the establishment of a sound credit risk early warning system is necessary.In this paper, reviewed on the basis of historical documents, summarized the development process of credit risk, put forward five basic processes of early warning systems about credit risk and early warning system should include the basic elements. And then using discriminant function, Logistic model, factor analysis and k-means clustering methods to build a complete credit risk early warning system. In this system, the first part is the risk monitoring and tracking system. In the part, the author identified the key factors which is profitability factor and solvency factors affecting business categories through the Logistic model and factor analysis. Fixed these two factors and the relevant indicators as usual monitoring and tracking key indicators . Followed by the risk identification process, In this process, use of discriminant function to analysis the company's situation and to determine the critical value is set at -0.6.if the computed value is less than -0.6, then the company was identified as the crisis company which must be taken appropriate response measures to control, while If the calculated value is greater than -0.6, then keep track of the company. Then the degree of risk-rating and determine the crisis levels stage, In this part, based on the value calculated by the discriminant function, author used the clustering method to classify the company and determined the crisis levels. Ultimately, the companies were divided into eight categories. The company shall be regarded as crisis company, if the company's categories is below BBB categories. The crisis company is divided into four categories, namely BB level (mild crisis), B grade (moderately critical), C grade (severe crisis), and D grade (very severe crisis). Then depending on the different crisis levels established a degree of credit risk management system .The author has proposed several measures to deal with the stock of credit risk .Banks according to their circumstances and risk profiles of customers to choose. Finally, must make the evaluation of the results of early warning system and summary of experience .The experience will be introduced to the early warning system, constantly improve the early warning system, which is the system's self-update mechanism.