Theresearch On Credit Risk Assessment Of Enterprises Basedon Logistic Model

Along with the gradual openness of financial market,how to manage financial risk effectively has become a common topic around the world.As one of the most important and commonly faced financial risks,credit risk has been paid more and more attention.A large number of studies showed that many models used in enterprise credit risk assessment had performed very well,but most of them didn't adequately consider the impact of the industry division.Therefore,on the basis of existing research,this paper conducted an empirical comparative study by establishing three principal component Logistic models for three industries respectively,which were manufacturing industry,construction industry and information technology industry.Principal component analysis and Logistic regression model were used jointly and all models were tested with data of second year.Empirical research showed that all three Logistic models had good validity and stability.Among them,Logistic model was the most suitable to measure the credit risk of construction enterprises industry with overall judgment accuracy rate of 87.1%.What's more,the significance level of three models were different.The Hosmer-Lemeshow test showed that the Logistic model of manufacturing industry had the highest level of significance,followed by construction industry.According to empirical research results above,some suggestions are made to the assessment of credit risk.In practice,enterprises should first be classified according to their industries,then applied with different credit risk assessment models accordingly.In addition,scientific and perfect databases should be established.Dynamic credit risk assessment models should also be established combining both qualitative and quantitative research methods,thus,the accuracy of credit risk assessment could be improved greatly.