Research On Medium-sized Enterprises Credit Rating Of Commercial Banks

When commercial banks are processing management of credit risk, the first step is,according to the information of subject to audit borrowers, to identify the probability of default. Measurement of probability of default is listed as an important part of Basel.But there are a variety of ways and means applied in the banks over the world, even in China,there exist no standard model.The models which are used worldwidely including:Logistic regression, Discriminant Analysis, Nearest Neighbor Methods, Decision Tree. In addition, more and more emerging methods are applied in credit rating, one of the most famous is Artificial Neural Network.And the old ones have continually been given new vitality.This paper will make combination of quantitative credit rating model used in medium-size capital-intensive enterprises of a large state-owned bank, according to the same variables, use the above-mentioned methods respectively, select the data of120listed companies, consider the impact of economic cycle, analysis the goodness of fit and accuracy of judgement of each method, to get the most appropriate method for credit rating in medium-size capital-intensive enterprises and provide a reference for the development of medium-size enterprises and reduction of bank losses.The first chapter of this essay is an introduction of background and objective as well as significance of research; The second chapter is about literature review,this paper reviews the theories of relative fields and makes some comment on their contributions and limitations;In the third chapter,this essay will take a large state-owned bank as an example to introduce the status of the credit rating of banking industry in our country; Chapter4and chapter5together,through theoretical study and empirical research, will study the applicability of all the above-mentioned models;The last chapter concludes and analyze the result and provide some suggestions for development of credit risk management.