Research On Credit Risk Assessment Of P2P Lending Platforms Based On Logistic Regression

P2P(peer-to-peer)network lending is a new way of Internet finance based on electronic network platform.As of the end of May 2017,China's P2 P platform,the cumulative transaction volume has reached 457.902 billion.However,due to the lack of effective supervision of the P2 P industry,resulting in an endless stream of various problems platform,the outbreak of credit risk to investors suffered a huge loss,P2 P credit risk has become an important factor restricting the development of P2 P industry.P2P network loan platform credit risk has long-term,hidden,destructive characteristics,so it is necessary to evaluate the evaluation.The principal component analysis method can effectively reduce the collinearity between variables.By using this method,the ten indexes such as registered capital,platform background and company size are reduced dimension,and two principal components M1 and M2 are obtained,which effectively reduces the variables Between multiple collinearity.Logistic regression model is very important for interpreting variables,which is suitable for the classification of variables.By using the model to introduce the correlation variables of 108 samples,the relationship between P and variables is obtained,and the credit risk is evaluated.The model results show that the platform credit risk is proportional to the expected return of the platform,which is inversely proportional to the registered capital and platform background,and the bank custody index coefficient is the largest and the actual situation.The model verification results show that the model accuracy is 88%,Indicating that the model of the sample network loan platform credit risk assessment capability is ideal.