P2P网贷借款人逾期率的实证研究——以“人人贷”为例

An Empirical Study on the Overdue Rate of P2P Lending——Take renrendai as an example

借款人逾期行为是P2P网贷投资人面临的主要风险,而平台信息披露与信用评级是投资人能够直接获取的有效信息。收集2016年9月至2017年9月人人贷上已完成交易的借款标的数据,通过构建Probit回归模型和Logistic回归模型,比较二者预测拟合度进而选择预测更准确的回归模型,来研究影响借款人发生逾期行为的因素,并建立借款人逾期率的概率模型。帮助投资人根据此模型对借款人的逾期行为进行初步判断,从整体上减少资金回笼不利的局面,并提高网贷平台的风险控制能力,促进网贷平台的持续发展。

Borrowers＇ overdue behavior is the main risk faced by investors on P2P lending.Platform information disclosure and credit rating are effective information that investors can get directly.Therefore,the collection of completed transactions borrowing target data in September 2016-September 2017 on renrendai through the construction,Probit regression model and Logistic regression model,and then choose the fitting degree of regression model prediction more accurate prediction compared with the two,to study the influencing factors of occurrence of overdue borrowers behavior,and establish the probability model of overdue rate.It helps investors to make a preliminary judgement of the borrowers＇ overdue behavior based on this model,reduce the adverse situation of capital withdrawal from the whole,and improve the risk control ability of the network loan platform,and promote the sustainable development of the network loan platform.