Bond Default Risk Based On KMV-RF Portfolio Model: A Case Study Of Yango Group

With the development of China’s bond market,the market scale is gradually huge,and the number of bond issuers and types are also increasing.The expansion of financing channels and conditions makes the bond market more active.However,with the rapid development of the market,the risk problems that have been accumulating gradually erupted.In addition,the diversified development of China’s financial instruments is not perfect,the existing financial supervision system is still insufficient,and the awareness of risk prevention is not high enough.As a result,the problems in the financial market are gradually exposed.Therefore,more and more scholars have applied machine learning and other methods to the research of risk identification to establish bond default risk early warning models,in order to identify enterprises and their bonds with greater risks in advance and achieve the effect of avoiding risks.Through data search and research,this thesis selects the mature or default bonds issued by listed companies in the domestic bond market from 2014 to 2022 as a sample,and collects indicator data based on the time one year before the maturity date or default date of each bond,covering three major aspects: the financial indicators of enterprises,the attributes of bonds,and the macroeconomic situation of the country.It innovatively combines the default distance generated by the KMV model,A total of 25 characteristic indicators are generated to construct the indicator system,and the Random Forest model is used for training and learning.Considering the large difference between the number of normal bonds and the number of defaulted bonds in the sample,this thesis uses the method of oversampling and undersampling to deal with the problem of sample imbalance.The results show that the KMV model can be combined with the Random Forest model,and the prediction effect is further improved,indicating that the KMV-RF combination model has certain theoretical and practical significance.Finally,this thesis introduces Yango Group as a case to further verify the availability of the portfolio model,and analyzes the default risk of enterprises to a certain extent,and gives reasonable suggestions and countermeasures.