On Measurement And Management Of Credit Risk Of Commercial Bank

How to measure and manage credit risk is an eternal subject concerning the commercial bank management. In recent years, the credit risk faced by commercial banks has been becoming bigger and more complex, which results in more attention from the bank system, the economic subjects and even the supervising and managing authority. This paper examines the most important part of risk management,that's, the risk measurement and management of the enterprises with loans, from the aspect of commercial banks, with a desire to provide the technology and methods for our commercial banks in development.The concept of credit risk and its characteristics are defined firstly. To express our research structure and direction we make a comprehensive induction and reorganization of the risk management background theories, the measurement methods of credit risk, as well as the domestic and foreign findings. To focus on the improvement of the risk measurement assessment system and empirical research, the index assessment system of the credit risk is established based on financial information disclosed by listed companies, and the fuzzy neural network method to measure the credit risk is proposed. In cooperation with a commercial bank in Shanghai, studying their loan details and their corporate finance data between 1999 and 2005, 8 financial indexes which can reflect the enterprise's credit condition most are selected by utilizing rough sets, and the credit appraisal is carried on with the neural network method again. Empirical research shows that the proposed method appears quite precise.With the listed companies'stock quotation data and the financial data provided by CCER, the empirical research of KMV method is conducted to contrast and improve its calculating formulas for violation distance, and operating procedure that suits to the Chinese national condition is obtained. To measure dynamic credit risk of non-listed companies, we study the PFM model developed from the KMV model, where the neural network is used to estimate non-listed companies'assets value and undulation rate and the value-added rate is used to replace continuous rate of returns to carry on the violation distance calculations. The result shows that the method has good credit risk assessment and forecast ability for both listed and non-listed companies. We also discuss the optimization of the organization structures of credit risk management, the improvement of the information system, and the significance of credit risk management culture and idea to the quantitative management. Then concrete reform paths for banks are suggested.Finally, we also summarize our main results and point out its future research direction.