Font Size: a A A

Credit Bond Default Risk Measurement Based On KMV-random Forest Model

Posted on:2021-08-23Degree:MasterType:Thesis
Country:ChinaCandidate:J ChenFull Text:PDF
GTID:2510306302953339Subject:Master of Finance
Abstract/Summary:PDF Full Text Request
After more than 40 years of development,China's reform and opening up have achieved great results,the total economic volume has leapt to the second place in the world,the comprehensive national strength and international influence have achieved a historic leap,the level of science and technology has been continuously improved,a number of indicators are at the forefront of the world,and the quality of life has significantly improved.Financial market,as a barometer of the national economy,with the development of the economy and the promotion of market-oriented reforms,China's finance has undergone profound changes in terms of scale,structure,format,function,competitiveness,and international influence..While booming,the problems accumulated in the long-term development of financial markets cannot be ignored.China's financial industry has entered the tackling stage of deep-level structural reforms.The outstanding problems include the inadequate financial instruments,the problem of the regulatory system,and the imperfect risk prevention measures and so on.In recent years,there have been frequent default events in the bond market,and China's corporate bond credit risk has received extensive attention and supervision.As an important part of the financial market and an important financing method for the company,research on credit bond default risk helps companies compete fairly,regulate market behavior,and protect investors' interests.The first part analyzes the default phenomenon of credit bonds in China's bond market.Based on domestic and foreign research on credit bond default risks and domestic development status,through studying relevant theories and analysis,it is found that KMV model has a quantitative effect on foreign listed company bond risk,combined with China's national conditions,by using a random forest model to quantify China's credit debt risk,so as to make a reasonable measurement and prediction.In the second part,this article will select ST,ST * companies and normal companies in the same industry as the experimental group and the control group from the listed companies on the Shanghai and Shenzhen Stock Exchanges in 2010-2019 to ensure that the data are sufficient and available.The company has been losing money for two consecutive years,so the annual report data of 2008-2017 was selected for analysis and prediction,and the data of 2008-2014 was selected as the training setand the data of 2015-2017 was used as the test set.First,calculate the default distance according to the KMV model,and find that there is a significant gap between the default distance and the control group,indicating that the KMV default distance can be used as one of the factors to distinguish ST companies from normal companies.In the third part,the random forest model is used to screen the traditional financial indicators,the default distance from the representative market information and the nature of firm to obtain a random forest model evaluation index system: ROE,ROA,net sales margin,ROIC,quick ratio,cash flow interest Guarantee multiple,EBITDA/interest expenses,year-on-year growth rate of operating profit,year-on-year growth rate of net profit,year-on-year growth rate of net assets,cash content of net profit,net cash flow from operating activities/operating income,asset-liability ratio,tangible assets/Total assets,default distance.Using the data from 2015 to 2017 as a test set,observe the evaluation effect of the above model on the default risk of listed company bonds,and compare it with the LOGIT model and distinguish analysis model to confirm that the random forest model has a better effect on the measurement of default risk.Finally,in the empirical study,we also remove the KMV default distance indicator for comparison and observe that the prediction effect of the random forest model has decreased to a certain extent,which shows that the combination of the KMV model and the random forest model has theoretical and practical significance.Finally,this article puts forward corresponding suggestions and countermeasures for the avoidance of the default risk of listed company bonds from the perspectives of regulators,listed companies and investors,and explain the shortcomings of the paper.
Keywords/Search Tags:credit debt default risk, KMV model, default distance, random forest model
PDF Full Text Request
Related items