| As a kind of financial innovation model under the age of network financial.Pear-to-pear lending is an important tool to Improve the financial system,fill the credit vacancies,make up for SME financing gap,ease the demand for private capital investment.The platform has grown rapidly in the past decade,attracting the favor of many investors.However,the rapid development also is accompanied by a variety of problems and risks,one of the most fundamental problem is the borrower's credit risk.This paper starts from the credit risk of P2 P platform,and researches on the credit risk of borrower.The first step is collecting a number of experimental data from “renren loan” by reptile technology and referring to the personal risk assessment system of China Construction Bank and the United States FICO company.to select indexes.The second step is through date processing classification and quantitation methods to ensure the final modeling indexes.The third step is using Logistic regression methods to get a new credit risk assessment model.The last step is using another volume of testing data to test the accuracy.The accuracy is 87.6%.But the perfect risk system is better to take into account to the borrower's credit level.so we introduce AHP-analytic hierarchy process.The first step is using the value of information gain to roughly fit the subjective opinions of experts.then,the second step is constructing the judgment matrix to calculate the weight of each indicator,and the third step is referring to the rules of the major financial institutions,the borrower was diverted to the five levels.A grade overdue rate is only 1.9%,the worst grade HR overdue rate is 6.7%,but the default rate reaches 43%.we can see that the lower the level,the borrower's overdue rate is higher.The model is helpful to improve the accuracy of the P2 P platform to determine the solvency of the borrower,reduce the risk of the loan from the source,and contribute to the sustainable development of the P2 P industry. |
PDF Full Text Request