Application Of Value-at-Risk Measures On Evaluation Credit Risk Based On Data Mining Technology

As an important part of financial risks, credit risk has been paid more and more attentions in economic life. Research to credit risk changes greatly both in financial theory and practice. At present, evaluation of credit risk is shifting from qualitative analysis to quantitative analysis, this trend will emphasize on evaluating credit risk by mathematical models and speed up the revolution of credit risk management. So that, testing and supervising the credit risk is crucial to finance in China.Recent years, with high-speed development of market economy and deep competition in China, policymakers of enterprises and financial offices begin to realize that it's not enough to make decisions by traditional experience. Poor understanding of mathematical methods and models has made theories disjointed with practices. On the other hands, with the renewing of computer science, it has been used widely. It's a trend to evaluate credit risk intelligently by computer techniques. This thesis will undertake by follow researches and exploration about the construction of a computer assistant "Evaluating Credit Risk Model": (1) Comprehensively described some terms and contents in credit risk management, andcompared credit risk management situation in China to foreign institutions. Putforward certain defects in China from different views of enterprises and banksseparately. (2) Analyzed and compared several credit risk models in common use and their virtuesand weaknesses. (3) Applying the technique of Data Ware and Data Mining to the construction of creditrisk models to further improve the intellectualized degree of this model. (3) This article will use the typical VaR Measure into evaluation of credit risk andcalculate the Value-at-Risk by Extreme Value Method, overcoming the weakness oftraditional models.The starting point of this thesis is to find a powerful way to quantify credit risk combining with possible credit problems. This article emphasizes on the application of Data Ware and Data Mining techniques in the construction of credit risk model and evaluation of credit risk by Extreme Value Method. Effectively overcame defects in traditional models and greatly improved the intellectual degree of this model. At last, based on quantitative analysis, the author advanced some suggestions, which aimed at present credit risk management situation of our country, and provided certain theoretical basis for the career of credit risk supervision.