Research On The Credit Risk Identification Of Chinese P2P Lending On Logistic Regression Model

Peer-to-Peer Lending is one of an important forms of internet finance.It is not only the organic combination of private lending and modern network information technology,but also the deepening of traditional e-commerce in the field of the private lending.P2P lending relies on the Internet technology to realize the direct financing between the borrowers and the lenders,breaking the credit model of traditional finance,effectively raising social idle funds to provide loans for those who need money,and gaining profits to realize the rational allocation of resources.P2P lending's convenient,efficient,low threshold,low cost and high-yield features greatly improve the financing efficiency,broaden the financing channels and meet the credit needs for the small and medium-sized micro-enterprises and the personal loan.With the foundation of China's first online credit platform Pai Pai Dai in 2007,P2P relying on its unique characteristics and advantages becomes more and more popular in the field of Internet finance,and its market scale has been expanding.However,due to the information asymmetry of the market,the imperfection of laws and regulations and the lack of supervisors,the adverse selection and moral hazard always happen.The platform risk is increasing,and P2P network credit market is facing a huge credit risk and trust crisis threats.Therefore,how to improve the level of credit risk recognition of investors to the markets is directly related to the safety of investors and the healthy development of P2P lending market.In this regard,the state issued a series of laws and regulations to rectify the Internet financial industry which converted it from "promoting development" into "standardized development." P2P lending industry will gradually return to reason.At present,the problem of credit risk identification and management has aroused more and more attention of scholars at home and abroad,and becomes the core of P2P research.Based on the theory of credit theory and information asymmetry,this paper constructs a signal game model between the borrowers and the lenders and study the credit risk identification method of the investors based on the basic characteristics,credit characteristics and borrowing characteristics of the borrower.Then,using "Renren Loan" as an example,the paper uses the crawler program to capture 35,649 transaction records from January to June in 2016,and uses binary logistic regression model to regress all samples and sub-samples to study the effect of investors' credit risk identification of borrowers and to learn how to improve the success rate of network lending.The study shows that the lenders will do the credit risk identification by valuing the characteristics of the loan,the basic characteristics of the borrower and the credit characteristics of the borrower.The borrower's credit rating and whether or not there is a local certification have a great influence on the lender's investment will.Besides,loan amount,Interest rate,the number of full scale,the performance of historical loans and financial conditions also have an important impact.The innovation of this paper is as follows:(1)Applying the signal game model to the study of P2P network lending behavior,the paper analyzes how the lenders and the borrowers can maximize their own profit by making the credit risk identification.(2)Based on 35649 transaction records from January to June in 2016,this paper uses binary logistic model to explore the impact of credit risk identification on investor credit risk,to verify the feasibility of the theory,and to propose some policy recommendations to improve China's P2P market.