Research On Wind Control Model Of P2P Platform

P2P(peer to peer) network lending platform have developed rapidly since entering the country in 2006 for the first time. P2 P platform is more and more widely accepted by general public for its high trading efficiency, low barriers to entry and trading flexibility and other advantages. According to incomplete statistics, in January 2015, the turnover of the whole P2 P industry is 35.782 billion yuan, which is 204 percent higher than that of last year; the total transactions of February amounted to 33.514 billion yuan, which has an increase of 217% over last year. Beside the rapid growth, the platform also gradua lly expose some risks, there were total 76 platforms which has problems in 2014, while 72 platforms closed or led off due to operational problems, this reminded the participants and managers of the platform should be cautious about the risk, and take effective measures to minimize harm timely.In this paper, we will analyze the risk behavior of borrowers and investors which may bring to the platform from the perspective of the platform. First, cluster analysis is used to classify investors and borrowers generally, then we hope to get the optimum method of division by using Decision Tree to compare the different methods, and the exact cutting point which can divide borrowers better. Based on this, the borrower is divided into three types finally, namely the Ordinary Borrower, the Growing borrower and the Risk Borrower, and the average loan amount of the Risk Borrower normally exceed 148,000 yuan. Investors are also divided into three types, namely the Ordinary Investor, the Growing investor and the Risk Investor, and the average investment of the Risk Investor are higher than 1907 yuan. Then the Logistic regression model are used to predict the rate of overdue borrowers, in this model, we make overdue as the dependent variable, and the borrower’s age, income and education and so on are regarded as the independent variables, the result shows that the model have a good estimate of the borrower’s overdue probability. Finally, we build an investor dispersion index to give an reasonable investment guidance to the investorsBased on the above analysis and combined with the successful experience at home and abroad, we propose appropriate risk control measur es should be taken as follows: 1. we can guide platform participants’ behavior reasonably, be cautious about the Risk Borrower, develop punitive measures and recommend a rational investment according to the investment dispersion index; 2. the borrowers can be encouraged to improve their credibility consciously, the group mechanism and friends mechanism can also be introduced into the platform to restrict the behavior of the borrower, the borrowers’ subsequent repayment performance should also be supervised, in order to take timely me asures when there are problems. 3. the cooperation between platforms are recommended to enrich clients database and establish a more rational credit system, uniform rating standards and more rigorous examination of the borrower qualifications. 4. we may seek cooperation of insurance companies and establish risk-sharing mechanism, so that the investors will not bare the risk lonely.