Research On Risk Assessment Of Individual Credit Loans Of A Internet Finance Company

With the deep integration of Internet technology and financial technology,China’s Internet finance industry has shown a trend of rapid development.It is considered by the public as a convenient,efficient and accessible financial service.Among the Internet financial products,consumer finance is growing in scale by virtue of its innate advantages such as easy operation and fast lending speed.However,behind the rapid development of consumer finance,many credit risk problems have been exposed.Due to the natural information asymmetry in consumer finance and the imperfection of the risk control system of consumer finance at this stage in China.This has led to a large number of default risks such as overdue,bad debt and fraud from various consumer financial institutions.At the same time,the low level of risk management and backward risk control means also directly affect the development of China’s consumer finance industry.Therefore,how to reduce the credit risk of each consumer finance platform in China is an issue that risk managers need to continuously focus on.Compared with China,developed countries such as Europe and the United States have an earlier start in risk assessment in the field of consumer finance,and the credit risk control system is more perfect.This paper firstly describes the advanced experience of domestic and foreign scholars on consumer finance risk control and the use of big data analysis,data mining and other technologies in consumer finance credit risk assessment through literature analysis method.It has implications for China’s consumer finance companies to improve their corporate credit risk management and reduce the default risk caused by information asymmetry.Secondly,this paper takes Company A as the research object,details the business process of Company A and the pain points in the industry risk management,and focuses on the current situation of credit risk control and information asymmetry problems of Company A.Finally,we focus on the establishment of the risk control model of Company A.Using Company A’s own customer data,the risk control modeling is conducted by logistic regression scorecard.Through data analysis and feature engineering,12 fields were finally selected for modeling.Considering the importance of model interpretability,this paper focuses on modeling based on logistic regression algorithm.The final modeling results show that the model KS(Kolmogorov-Smirnov)value is 0.285 and the AUC(Area Under Curve)value is 0.69.Through credit scoring,we have achieved the quantification of borrowers’ credit risk by using the theory related to quantitative assessment,which proves in practice that big data mining technology has the ability to be used in credit risk assessment.The data mining technology has proved its usefulness in credit risk assessment.Finally,based on the empirical analysis of this paper and the actual business situation of Company A,the recommendations of the risk control strategy in the future practical business operation are given.The risk level of each borrower of Company A is quantified,and the effectiveness and feasibility of the logistic regression scorecard model in solving the credit risk assessment problem of borrowers is also verified,which also provides reference for related research conducted in the same industry.