民营企业债券违约风险预警和防范体系研究

Research on EarlyWarning and Prevention System of Bond Default Risk of Private Enterprises

目前,我国债券市场信用风险开始有序释放,债券违约现象明显增多,如何识别、度量和防范债券违约风险变得尤为重要。本文将实际发生债券违约的上市公司作为违约组,配对抽取对照组,开展实证分析。首先,基于KMV模型得出违约距离适用于预测上市企业债券违约风险的结果。其次,在选取传统财务因素的基础上,加入股权质押比例、商誉占比,以及违规行为、并购和对外大额投资等非财务指标,进行Logit回归。结果发现传统财务指标、股权质押比例、商誉占比等都对债券违约风险有显著影响,且非财务信息也能起到很好的预警作用。最后,结合KMV和Logit模型,构建Logit-KMV混合模型。结果表明Logit-KMV混合模型在模型拟合效果和民营企业债券违约风险度量及预警方面均更为有效。

At present,the credit risk of China's bond market has begun to be released in an orderly manner,and the phenomenon of bond default has increased significantly.This paper treats the listed companies with default of bonds as the default group,and the paired sampling method is adopted to select the control group.The comparison results show that the average default distance of enterprises in the default group is smaller than that in the control group.Firstly,the default distance can be used to predict the default risk of listed companies'bonds,which based on KMV model.Secondly,on the basis of the traditional financial factors affecting bond default.Logit re-gression analysis was carried out by adding such indicators as the proportion of equity pledge,the proportion of goodwill and other non-financial information such as violation penalties,large foreign investment,mergers and acquisitions,etc.The results show that the traditional financial indicators and the proportion of equity pledge and the proportion of goodwill have significant impact on bond default risk.Non-financial information also play a good early warning role for bond default.Finally,the results show that the Logit-KMV hybrid model is more effective in model fitting effect and private enterprise bond default risk measurement and early warning.