Credit Risk Measurement Of P2P Networks Loans In China

Online Peer to Peer Lending refers to the borrowers and lenders on the line between lending transaction,rather than the financial intermediaries.It is a new model of financial transactions which takes the Internet as a medium.It was first made by United Kingdom Zopa online trading site,founded in 2005,this new mode of trade developed rapidly after appeared as United States Prosper P2 P network with global influence lending platform.Starting from 2007 the P2 P network lending platform.Has experienced its infancy,now it is in a consolidation period.That phase of the P2 P network to expand lending at a very alarming rate,number of P2 P network lending platform on the market by April 2017 reached 5,613.But because of the lack of effective supervision,resulting in risk of lending platform over P2 P networks rife,and there is a lot of the platform in question,which has seriously affected the country's financial order.As China continues to deepen system reform,economic growth slowed down,online lending industry from the pressure of economic downturn,the credit risks are growing.Controlling credit risk of P2 P lending platform of network management is a very important task,it is related to the platform's core competitiveness and development potential.So based on the P2 P network platform for the study of credit risk of loans,by comparing the model selection Logistic regression models of credit risk measurement credit risk measurement models as a P2 P network lending platform.With credit for all transaction data as we build a model of the original data,we selected and the borrowers ' own information and loan information-related variables,such as age,education,marital status,loan amount,interest rate and term.These are the main factors determining risk of borrower default.Through the method of factor analysis to the original dimension variables to determine the number of common factors,It will be the common factor built into the model out of the logistic regression model.Empirical results show that the there is a negative relationship between default probabilities and education,marital status,property mortgages monthly income and other indicators,while there is a positive relationship between default probability and loan amounts,loan terms and other indexes.In addition,by modeling data set and verify the data group has also verified the results we build the Logistic model has strong predictive power.