Research On Influencing Factors Of Default Risk Of Credit Bonds Based On KMV-LOGIT Hybrid Model

With the focus of the 12th five-year plan on China's economic structural transformation and the main task of supply-side reform in recent years,the domestic financial market has experienced a period of high risks.On March 5,2014,"11 Chaori bond" took the lead in breaking the payment bottom line,becoming the first domestic default bond in the history.By the end of 2018,a total of 245 defaults involving 205.836 billion yuan had occurred in the credit bond market.In particular,the number of defaults on credit bonds exploded in 2018 was extraordinarily huge.A total of 123 credit bonds were defaulted during the whole year,involving 119.95 billion yuan,accounting for half of the number and scale of defaults on credit bonds in recent years.In the context of the gradual normalization of bond default in China,it is of practical significance to study the influencing factors of the default risk of credit bonds and optimize the quantitative method of default risk for the companies issuing bonds,investors and regulators to identify and prevent credit risks.This article is based on the real default events of credit bonds happened in 2018.Through combining KMV model and LOGIT model,the KMV-LOGIT hybrid model is built based on various factors of credit default risk measurement,including indicators of risk,profit ability,debt paying ability,operation ability,cash flow and growth ability.And then the article uses the data sets,including default samples and contrast samples to do empirical research based on KMV model,LOGIT model and KMV-LOGIT hybrid model.The results showed that:(1)using KMV model to calculate the default distance of samples in China's credit bond market can better distinguish the credit risks of default samples and contrast samples.At the same time,it is found that the value of corporate assets does not obey the normal distribution,which proves that there is a phenomenon of peak and thick tail in the reality.Therefore,it is not suitable to use KMV model to calculate the theoretical expected default frequency in this context.(2)the classical LOGIT model constructed by using the actual default data and financial indicators of samples can fit the financial situation of listed companies in China well,and the overall discrimination rate of the model reaches 97.2%.The indicators of return on total assets,long-term debt to capital ratio and operating cash to income ratio are important factors affecting the default risk of the credit debt.(3)the KMV-LOGIT hybrid model constructed in this paper combines the market indicator output by KMV model--default distance and the explanatory variables of financial indicators in the classic LOGIT model.Empirical results show that default distance,return on total assets,long-term debt to capital ratio and operating cash to income ratio are the influencing factors of default risk.In addition,the goodness of fit of the hybrid model is higher than that of the classical LOGIT model,and the overall discrimination rate of the model is increased to 98.1%.Therefore we can come to the conclusion that the KMV-LOGIT hybrid model is helpful to thoroughly measure the factors influencing the default risks of China's domestic credit bonds.