Measurement And Warning Mechanism On Default Risk Of Credit Bonds

In recent years,China's credit bond defaults have occurred frequently,and even serial defaults have occurred,which has had a great impact on the stable development of the credit bond market.The measurement,early warning and risk control of credit bond default risk have important theoretical and practical significance for the stable and healthy development of the credit bond market.When the traditional Logistic model studies the risk of default,the input dependent variable usually uses the sample real default data.In order to solve the problem of lack of data,this paper improves the Logistic model,uses the KMV model to calculate the default distance of the sample bond,.Based on the bond default data,a two-class logistic regression model is established to quantify the default risk of credit bonds.This paper takes the main body of the credit bond that defaulted in China in 2018 as a sample.The sample default distance calculated by the KMV model is used as the dependent variable,sample indicators including solvency,profitability,and equity pledge ratio as independent variables.An improved two-class logistic regression model was constructed to measure the risk of default on credit bonds.And verify the validity of the improved model,compare the calculated probability with the actual default of the sample group,and then compare it with the traditional Logistic model for goodness of fit test and Hosmer-Lemeshow test.Finally,the system is clustered and analyzed according to the probability of default of the empirical results,the warning threshold is set for different risk levels of bonds,and an early warning mechanism for default risk is established.Corporate solvency and equity pledge ratio are the most important factors.The validity test results show that the new model can accurately determine the risk of default,and the improved model results are highly consistent with the actual default data.The prediction accuracy is 95.3%.It proves that the improved model has a better fitting effect.Finally,the warning threshold is set for different risk levels of bonds,and the early warning mechanism of default risk is established.When the bond default probability is lower than the green light threshold of 0.34,the bond is in the green security zone;Between 0.34-0.54,it means that the bond enters the high-risk area and will emit an orange warning signal;once it exceeds the red light threshold of 0.54,the bond has entered the dangerous area,the warning mechanism sends a red warning signal,and the regulatory agency must adopt mandatory wind control.The innovation of this paper is that in the traditional two-class Logistic model modeling,it is necessary to judge whether the subject will default in the future,which is inconvenient to obtain.The improved logistic model is a default calculated by the KMV model;most existing domestic research only establishes models to measure default risk,and does not combine risk management with early warning analysis.In this research,the warning threshold is set for different risk levels of bonds,and the early warning mechanism of default risk is established.