Study On Credit Risk Of Listed Companies Based On Fuzzy Comprehensive Evaluation

In our country, the listed companies have developed for more than ten years and get conspicuous success. But with the increase in the number of listed companies, more and more listed companies face rough credit risk because of financial changes. "Yinguangxia incident" and "Kelong incident" are good examples. These credit problems stabbed investors especially the small and medium shareholders to the hearts, seriously affect the wholly credit image of listed companies and become the prominent problems which hold back the development in the market. So it is urgent to research and evaluate the credit risk of listed companies.There are five chapters in this paper. The fist chapter is introduction and introduces the research on credit risk in and out of our country, and the research route. The second chapter is modern credit risk evaluation models and comparison on some credit risk evaluation models, which have come into wide use. The third chapter is the construction of the fuzzy comprehensive evaluation model, expatiating on the process of the construction of the model. The Empirical research of the model and conclusion are respectively in Chapter Four and Chapter Five.On the construction of evaluation index system, considering the limitation of information, which enterprises or personal investors can get, only financial indexes are chosen. And according to correlation analysis and discriminability analysis, some indexes that have strong correlation with other indexes or have only little function to evaluate credit risk are eliminated. At last, the credit risk evaluation index system is set up from four aspects: credit capacity, operation performance, operation capacity and economic strength, with fifteen indexes.On the weighted coefficient, Entropy Weigh Calculation Method is used, as some ordinary methods such as AHP, have strong subjectivity.On the construction of the model, the traditional fuzzy comprehensive evaluation model is improved, and F function and the multidimensional linear regression model of several targets are combined to set up four multidimensional membership function. Just imputing the data of samples and using the principle of biggest degree of dependence, the sample's risk state can be evaluated. It simplifies the evaluation process of credit risk.In order to test the model, this article chooses some industrial listed companies in Shanghai Stock Exchange as testing samples evaluates their credit risk state and compares this result with the result of Lianhe Credit Rating Co., Ltd. The result shows that the tworesults are about and about, and the degree of fitting is 74.7%. Especially for natural and high-risk listed companies, the two results are the same. This shows that the fuzzy comprehensive evaluation model in this paper can effectively evaluate the credit risk of listed companies and provide decision support for enterprises or personal investors' investment or credit extension.