Can Investors on P2P Lending Platforms Identify Default Risk?

ABSTRACT

Using data from Renrendai, one of the largest peer-to-peer (P2P) lending marketplaces in China, we found an asymmetric relationship between the borrowing rates and the default risks of borrowers; specifically, orders with the same interest rate may have different default risks. A counterintuitive result is that the higher a borrower’s income, the greater the default risk. Furthermore, it is found that investors may be ignorant of the relationship between certain information (income, age, education, etc.) of the borrowers and the default risk, but they pay more attention to borrower creditworthiness, loan amount, and loan term, which turn out to be the key factors in borrowers’ default risks; because they have a good knowledge of the relationship between these three pieces of information and the default risk, investors are able to identify default risk. Finally, we find that investors can learn to identify default risk.

KEY WORDS AND PHRASES:

• Interest rate
• default risk
• financial asymmetry
• risk identification
• P2P lending

Acknowledgments

This research was supported by the National Natural Science Foundation of China (71601075, 71401157), the Humanity and Social Science Foundation of Ministry of Education of China (15YJC790071), the Natural Science Foundation of Hunan Province, China (2016JJ3047) and China Postdoctoral Science Foundation funded project (2016M592416, 2017T100594).

Notes

¹ Represented by the probability of default of the borrower.

² We choose $yea{r}_i$ rather than bank rates as the covariate, because$yea{r}_i$ could control the effect of the change of the supply and demand of the fund with the development of the platform and the effcts of some policy factors on the interest.

³ For credit certification orders that have been seriously overdue (i.e., overdue for more than 30 days), Renrendai will pay the investor the remaining principal of the loan and mark the status of the order to be “advanced by the platform.” At this time, the investor will suffer a loss of the interest or even some of the principal. For credit certification orders that have been overdue for 30 days, there is no way to tell whether the investor will bear loss.

⁴ Although heteroscedasticity exists in the models, we find that the conclusions are robust using robust standard errors.

⁵ The rate ranks third, indicating that investors pay more attention to earnings when default risks of borrowers is fixed.

⁶ Although the coefficient of p^(2A) in 2015 is not significant at 10 percent level, it is significant at an 11 percent level. In addition, the coefficient of p^(2A) in the first two years is much larger than that in the last three years, which can be accounted for by the fact that the number of investors increases rapidly with the development of the platform.

⁷ We do not present the results for robustness test here because of the limited space. They are available upon request from the authors.

Additional information

Notes on contributors

Rongcai Hu

RONGCAI HU ([email protected]) is currently an Associate Professor of Econometrics at Hunan University. He received his Ph.D. degree from Renmin University in quantitative economics in 2008. His research interests include econometrics and financial engineering.

Meng Liu

MENG LIU ([email protected]) is currently working at Shanghai Xinqu Sub-branch, Bank of Communications. He obtained his B.Sc. and M.Sc in finance from Hunan University in 2014 and in 2017, respectively. His research interest is P2P lending.

Pingping He

PINGPING HE ([email protected]) is currently an Associate Professor of Finance at Hunan University. He received his Ph.D. degree from Beihang University in management science and engineering in 2007. His research focuses on internet finance.

Yong Ma

YONG MA ([email protected]; corresponding author) is currently an Associate Professor of Financial Engineering at Hunan University. He received his Ph.D. degree in management science and engineering from South China University of Technology in 2014. His research interests include credit risk management and financial derivatives pricing.