Research On Credit Risk Assessment Model Of P2P Lending Borrowers

P2P lending provides convenient and fast financing channels for SMEs and individuals.It has developed rapidly after entering the market.However,due to the imperfection of the credit information system and the limitation of space,P2P lending has serious information asymmetry in the transaction process,resulting in frequent credit risk crisis of borrowers such as overdue and bad debts.It hurts the interests of investors,increases the operational burden of the lending platform,and also has a negative impact on the P2P lending industry.Therefore,studying the credit risk of borrowers has positive significance for the healthy development of China's P2P lending industry.Based on the theoretical combing and literature reading,this paper analyzes the basic situation of China's P2P lending industry and the credit risk of borrowers in the P2P lending industry.Then,based on the characteristics of P2P lending industry and existing research,we build the P2P lending borrowers' credit risk evaluation index system which consists of 29 indicators.Finally,according to the P2P lending borrowers' credit risk evaluation index system constructed in this paper,this paper build the neural network model which constructed based on BP algorithm.Then,this paper trained the BP neural network model by LM algorithm,quantitative conjugate gradient method and Bayesian normalization method.By comparing the performance of the three training methods,the input layer node is selected as 9,the hidden layer node is 11,the output layer node is 1,and the training method is the BP neural network structure of LM optimization algorithm.This industry will provide practical guidance for the establishment and improvement of borrower credit risk management system for China P2P lending platform.