Research On Credit Risk Assessment And Management Of P2P Network Lending

Today,financial innovation and traditional financial services serve the real economy,both of which have their own natural advantages.The wide application of various lending programs in P2 P Internet finance has rapidly changed the pattern of financial services.An important issue in the field of financial innovation is the correct and effective assessment of the credit risk of borrowers.In P2 P network loans,the unsecured nature of loans and the relative novelty of the platform make risk assessment a problem.This thesis firstly conducted a large number of research on the research literature of P2 P online loan,including online loans,credit risk factors and models,and summarized the current situation of P2 P and credit risk assessment management in China.In addition,it also focuses on the detailed analysis of China's P2 P network lending credit risk assessment and management,including the evaluation process,detailed analysis,and various methods that need to be used.From data selection,data organization,feature construction,feature processing,feature selection,and data set processing,to final quantitative assessment of credit risk.In the credit risk management analysis module,the management problems and the necessity of credit risk management are raised.Then,this thesis uses the borrowing data of a domestic P2 P lending platform as the basic data,and carries out empirical research on credit risk assessment with the purpose of quantifying credit risk.This study uses logistic regression algorithm to quantify the credit risk to identify loan objects with high probability of default and improve the quality of the overall loan.This thesis collects real data from a P2 P lending platform in China for feature processing and then performs modeling analysis.This analysis used 10,930 loan data,based on 64 fields,and finally selected 19 model variables to participate in credit risk quantification through feature structure and feature analysis.The final modeling test results show that the model has an AUC value of 0.81 and a K-S value of 0.44,indicating that the model has a good performance in quantifying the credit risk of borrowers.Finally,it puts forward relevant suggestions for the credit risk management of China's P2 P network lending.It is recommended to fully implement credit risk assessment to production and propose clear management policies and procedures.At the same time,it is required to adhere to the principle of innovation in the process of credit risk management,as well as compliance management,and comprehensively enhance the ability of corporate credit risk management.Finally,it is recommended that the P2 P network lending platform assist relevant departments to establish a credit information system and use it reasonably.