| As the rapidly development of financial market,national financial regulators also gradually strengthening the Management and the monitoring of risk,credit risk is one of the most important risk types because of its importance and influence.Especially for commercial Bank,whether effectively manage and control the credit risk or not will has an conclusive influence on its profitability and robustness.Thus,during the period from The implementation of New Capital Accord (BaselⅡ) in2004to The advent of Basel Ⅲ,the Basle Committee which is Responsible for the monitoring of risk of developed countries’ banking put forward opinions and higher requirements for international bank credit risk management based on summarizing experiences of credit risk management.In New Capital Accord,the measurment of the borrowers’ default probability is not only an important content of the IRB but also the key to the commercial bank credit risk management.Commercial banks in China has strengthened the management and monitoring of credit risk and promoted the building of internal rating system in recent years,so that the level of credit risk management has improved significantly.But for most of the commercial banks in China,represented by small and medium commercial bank still have a significant space to meet the requirements of the Basel New Capital Accord and the advanced level of the international banking.Foreign advanced banks have researched on the management of credit risk for a long history so that have accumulated research data,based on this,models for credit risk management can be constantly Introduced new and applied.Commercial banks in China should take into account the actual financial conditions in our country and learn from foreign banks’ methods and thinking about credit risk management,besides,they also should increase investment on research and development aboout models for predicting and discriminiating credit default probability.This paper summarizes methods for measuring credit risk by starting with describing characteristics of credit risk and concept of default.In order to clearly show the development and the characteristics of methods for measuring credit risk,methods fall into three classes,namely classical method,traditional method and modern method.The empirical analysis in this paper takes ST companies which were special treated in2011and2012in the securities market in our country as samples of default companies,take companies were not special treated randomly as samples of non-default company,besides,these samples fall into training set and test set,selecting financial indicators by mean test between groups and correlation test,finally,eight financial indicators can be independent variables of model for measuring credit risk,there are leverage ratio, rate of capital accumulatio,total assets growth rate,degree of financial leverage,operating income and cash ratio,operating profit margin,accounts receivable turnover, turnover of fixed assets.Then respectively establish Bayes discriminant model,Logisitic model and BP neural network model to discriminate and predict credit default of the sample companies.From the empirical results,Bayes discriminant model doesn’t have good effect when discriminate and predict credit default of the sample companies,by contrast,Logisitic model and BP neural network model are better,Logisitic regression model has a more accurate prediction on credit default than BP neural network model.So Logisitic model can be considered as the most well model for measuring the probability of credit default of the three models.Finally,in view of problems exist in the credit risk management of China commercial bank,credit risk management technology is relatively backward,the lack of credit database and mature credit rating agencies etc,the paper proposes the establishment of advanced credit risk model that suit for the commercial bank of our country, establish and perfect the commercial bank credit database, set up the third party credit rating agencies etc. |
PDF Full Text Request