An Empirical Study On The Credit Risk Of Manufacturing Listed Companies Based On The Modified KMVLogit Model

Manufacturing industry occupies an important share in the macro economy of China.It has developed rapidly.Its growth rate is faster than the average growth of the world.However,at this stage,China's manufacturing industry is facing some challenges.The internal and external dividends have subsided,making the enterprise upgrading slowly and financing difficulties.This environment has resulted in the deterioration of the operating conditions of the manufacturing enterprises,the poor ability to resist risk and the existence of certain credit risk problems.Therefore,the credit risk of the listed companies in the manufacturing industry has become one of the hot topics of widespread concern.This paper attempts to reference and study domestic and foreign credit risk management methods.According to the specific situation of China,model should be modified for the actual situation of China's manufacturing industry listed companies.First,the KMV model is improved and applied to the manufacturing industry.The improved approach includes a higher default point modified by long-term debt weight,and a more flexible GARCH(1,1)amendment to the volatility of equity value.It is found that the prediction effect of the model is not ideal,and the early warning time is only one year.But the distance of default can be used as a separate variable to measure the possibility of default of the listed company.Secondly,through the comparative analysis of Logit model and KMV_Logit model,we get a better prediction result of KMV_Logit model.It shows that the DD does improve the effect of risk estimation,and the early warning time is one to three years.From the selection of the model,the KMV_Logit model is the first choice.The weaker effect of Logit model,but compared with the KMV model is better.When predicting three years ahead of time,the Logit model is better than the KMV-Logit model,indicating that the default distance is not enough to find the risk problem in T-3 years.In the longer time,the Logit model is a more suitable model.Finally,from the perspective of prediction time,whether the KMV model,Logit model or KMV_Logit model,their prediction effect is getting worse with time gone,and T-1 is the best.