| In recent years,P2 P lending attracts a large number of domestic borrowing users with its flexible,convenient,unsecured small loan model.Although the high demand of the domestic market to promote the P2 P loan market continues to grow,but the lack of law and regulation makes the P2 P loan industry chaos.Some business difficulties and other issues have occurred.P2 P lending platform is facing many risks,such as borrower credit risk,platform operational risk,etc.However,the fundamental,the borrower’s credit risk is the P2 P loan platform is the most core risk.The effective management and control of the borrower’s credit risk will affect the long-term development of P2 P lending platform.Looking back at the previous theoretical studies,we can see that in the P2 P lending transactions,on the one hand,the information asymmetry between the borrowers is the main reason for the credit problems such as the adverse selection of borrowers and moral hazard.As much as possible to fully collect and disclose information on borrowers to ease the information asymmetry of the borrowing process;On the other hand,there is a conflict between personal information disclosure and privacy protection,and there is a need for a balance between the two.This article is based on this,to explore the P2 P loan platform borrower credit risk management methods.Through the analysis of the actual operation platform,this paper summarizes the domestic P2 P loan platform operating mode is divided into pure information intermediary,debt transfer and third party guarantee mode.And the P2 P operating platform under different operating modes has two problems in managing the borrower’s credit risk: Ignore pre-loan management and lack of quantitative models.Therefore,this paper proposes to construct the borrower to apply the risk assessment model.Through the model of the means to strengthen the borrower before the audit efforts and to control the borrower’s credit quality from the source.This paper builds the borrower to apply the risk assessment model P2 P network lending platform.First of all,this article randomly sampled 4515 personal credit data,which contains the borrower’s 27 credit indicators from the domestic representative of the P2 P network loan platform.To improve the quality of the modeling data to clean the sample data,including missing values,abnormal values and continuous variables of the WOE group.Secondly,by comparing the use of Stepwise_Logistic regression and Lasso_Logistic regression model in feature selection and risk assessment,the results show that the Stepwise_Logistic regression model has a slightly higher prediction accuracy and stability.Therefore,15 significant variables were selected from 27 independent variables using the Stepwise_Logistic regression model and passed the test.The model results show that the borrower’s gender,education level,marital status,business type,working life,company size,job type,whether the car,the number of contacts,borrowing purposes,the borrower’s credit results play an important role in the credit.Finally,this paper applies the risk assessment index to the borrower to score the conversion,and applies the score to the P2 P loan platform loan approval decision and the interest rate pricing management.The conclusion of this paper and the proposed borrower’s application of the risk assessment model have practical application value in the credit risk management of the borrower of the P2 P loan platform.At the same time,it provides a certain theory for the quantification of the financial risk of the Internet and the management of financial information basis. |
