A study on default prediction of Chinese online lending: based on the analysis of mobile phone usage data

Abstract

ABSTRACT Information asymmetry between online financial lenders and borrowers may lead to a high risk of overdue repayment. To identify the influencing factors of the default rate in Chinese online lending, the presented study was conducted using a dataset constructed by 5108 borrowers in an online lending platform between July 1 and September 30 in the year of 2019. This research adopted a logistic regression approach to investigate the influencing factors of borrowers’ defaults based on loan application time and borrowers’ mobile phone usage. The results show that the default rate is relatively low when the loan application is made during the daytime. Meanwhile, the study indicates that the default rate is negatively correlated with the most of the investigated parameters, which include the number of loan applications, the number of contacts, the duration of the mobile phone using the internet, and the number of one person’s multiple phone numbers. Based on observations of the presented research, it could be found that mobile phone usage data possesses a significant impact on the default prediction, which is capable of providing constructive guidance to effectively reduce default risk for the borrowers and the online lending platform.