Research On The Method Of Credit Risk Measurement Of Listed Companies Based On Fisher Discriminant Analysis

Finance is the central of modern economy,and the listed company is the trunk of contemporary finance.Credit risk of listed companies affects not only the enterprise itself,but also the stable operation of commercial banks and even the whole macro economy,which may cause serious financial crisis and social crisis.Taking the new standard of international financial supervision "Basel Capital Accord III" as the guide,from the point of view of the serious credit risk environment that Chinese financial circle are faced with,and making the credit risk metric method as the main line,the paper summarizes the domestic and external research results about credit risk measurement.And on the basis of these,the paper does some research as follows:First of all,this paper summarizes the definition of credit risk from three levels and analyze three common performances about the credit risk of listed companies:financial risk,fluctuations in the market value and the poor internal management of the company.It sums up the causes of credit risk of listed companies from two aspects of macro economic environment and enterprise's operation and management.Then it outlines the current situation about credit risk of listed companies in China and summarizes the prevailing management standards of credit risk measurement in the international community.Secondly,the paper summarizes the classification of the credit risk measurement method(traditional method and modern method).It probes into the merits and faults,applicability in Chinese market of two kinds of commonly used credit risk measurement methods(KMV model and credit scoring method)and make a comparison between them.KMV model takes into account more risk factors,and performs better than credit scoring method in terms of predictability and dynamism.However,a large amount of credit data is needed to operate the model,while China is unable to provide complete data support for the time being.The credit scoring method is simple and feasible,and the operability is strong and requirement for data is not too much.But there are also some defects that are too strict in modeling conditions and may be difficult to achieve in reality.Aiming at the shortcomings of the credit rating method,this paper chooses the Fisher discriminant method which is unrestricted to distribution and variance,widely used,simple to test and accurate.It corrects the credit rating method and constructs a Zeta model based on Fisher discriminant method.Thirdly,this paper does empirical research about the measurement of credit risk in listed companies,based on the Fisher method of discriminant analysis.Based on the relevant data of the domestic part of A-share enterprises in 2015,the experimental group is composed of credit default group and non-default group according to the credit situation.Using SPSS software to do factor analysis and comparison about factor contribution rate,the parameter index system is composed of the factors which are independent and the rate of cumulative contribution is high.and the Fisher discriminant analysis is carried out to the experimental samples and the Zeta model is constructed.The test group was used to test the Zeta model with the related data of the A-share enterprise in 2016.The test results show that some of the credit situation in Chinese securities market can be better explained by the quantity model to a certain extent.Finally,the main content and conclusions of the paper are summarized.And on the basis of these,it puts forward some policy recommendations about how to strengthen the credit risk management after the global financial crisis for the financial regulatory authorities,commercial banks,and listed companies,and points out some shortcomings in the study.