| Since the default of the first credit bond in 2014,the number and amount of the default of credit bonds have increased exponentially.In 2018 alone,the default amount of credit bonds has reached 120.961 billion yuan,exceeding the sum of the default amount of credit bonds from 2014 to 2017.Under the background of the continuous decline of macro economy and the gradual normalization of bond defaults,it is of great theoretical and practical significance to study the default risk evaluation of Chinese credit bonds.First,this paper summarizes the background,research significance and relevant literature,and expounds the status and influencing factors of China's credit bond defaults.Second,the default risk evaluation model of Chinese credit bonds is constructed.According to the applicability and shortcomings of KMV and Logistic models,the model was modified and the parameters were defined.Third,empirical analysis and test.This paper selects 25 default credit bonds of listed companies in 2014-2018 as default samples,and 75 non-default bonds of the same industry in the same year as non-default groups for credit evaluation.The empirical study shows that: when using KMV model to estimate,the volatility effect of stock return estimated by GARCH(1,1)model is stronger than that of stock return calculated directly by using samples,and the mean T test of independent samples can better distinguish default samples from nondefault samples.When the default distance is calculated by selecting(short-term debt + longterm debt)as the default point,and combining the financial indicators and corporate governance indicators of the enterprise with regression analysis,the improved KMV-logistic model is superior to the original traditional model and has the stronger ability to judge whether the credit bonds of Chinese listed companies are in default.Finally,three Suggestions are put forward :to improve the early warning index of default risk of credit bonds,to improve the application of modern credit risk measurement model and to improve the forward-looking credit rating system.The main innovative work of this paper is reflected in three aspects:First,the innovation of research perspective.Existing credit risk evaluation studies mainly focus on the discrimination and prediction of ST and non-ST companies,while the research on the real default risk evaluation of credit bonds only stays in the multiple discrimination model,Logistic model and KMV model.This paper takes the Chinese listed companies whose credit bonds are in default from 2014 to 2018 as the sample,and uses the improved KMV-Logistic model for empirical analysis to verify an evaluation method applicable to the default risk of Chinese credit bonds.Second,the improvement of research methods.The traditional KMV-Logistic model is used to evaluate the default of credit bonds.The volatility of stock yield has the characteristic of sharp peak and thick tail,which does not conform to the standard normal distribution.The default point of KMV model is only set for the sample of American companies,and it is not applicable to Chinese listed companies.This paper uses GARCH(1,1)model to calculate the volatility of the stock return rate,increasing four default point in short-term liabilities and longterm liabilities of different values.Through sample test to find the best representation of China's credit bonds default point and default distance,and uses the optimized default distance as an independent variable in Logistic regression,improving the discriminant of KMV-Logistic model and its accuracy of prediction.Third,introducing corporate governance factors for Logistic regression.This paper analyzes four aspects: the macro factors,industrial factors,corporate governance factors,and financial factors,based on the influence factors of Chinese bond defaults,17 financial indicators and 6 corporate governance indicators were constructed to distinguish significant indicators for Logistic regression,and to make up for the inadequacy of the original model of financial indicators. |
PDF Full Text Request