Abstract
This study examines the predictive power of self-disclosed social media information on borrowers’ default in peer-to-peer (P2P) lending and identifies social deterrence as a new underlying mechanism that explains the predictive power. Using a unique data set that combines loan data from a large P2P lending platform with social media presence data from a popular social media site, borrowers’ self-disclosure of their social media account and their social media activities are shown to predict borrowers’ default probability. Leveraging a social media marketing campaign that increases the credibility of the P2P platform and lenders disclosing loan default information on borrowers’ social media accounts as a natural experiment, a difference-in-differences analysis finds a significant decrease in loan default rate and increase in default repayment probability after the event, indicating that borrowers are deterred by potential social stigma. The results suggest that borrowers’ social information can be used not only for credit screening but also for default reduction and debt collection.
Acknowledgments
The authors are grateful to participants of the 2017 Hawaii International Conference on System Sciences (HICSS), particularly the cochairs of the “Strategy, Information, Technology, Economics and Society” minitrack: Eric Clemons, Rajiv Dewan, Robert Kauffman, and Thomas Weber, seminar participants at City University of Hong Kong, Arizona State University, Shanghai Jiaotong University, and the coeditors and anonymous reviewers of the JMIS special section for their valuable comments on earlier versions of this manuscript.
Funding
Juan Feng acknowledges support from the General Research Fund (GRF 9042133) and City University SRG (Grant 7004566). Bin Gu acknowledges support from the National Natural Science Foundation of China (Grant #71328102).
Notes
1. According to our definition, Facebook and Twitter are social media, while WhatsApp is not.
2. Although in social media users’ home pages can be accessed by anyone, it is difficult to locate the home page of a certain user without his username or home-page address.
3. We calculated the percentage improvement as 0.657 – 0.5 / 0.623 − 0.5 = 1.28, where 0.5 is subtracted from both AUCs because 0.5 is the AUC under a noninformative (random) system.
4. We also calculated the AUC values by using 60 percent of randomly-selected samples as training data and the rest 40 percent of samples as test data. The predictive powers of the proposed model and the integrated models are 34 percent and 39 percent higher than the benchmark model, respectively.
Additional information
Funding
Notes on contributors
Ruyi Ge
Ruyi Ge ([email protected]) is an associate professor in the department of electronic commerce at Shanghai Business School, China. She received her Ph.D. in management information systems from Shanghai Jiao Tong University. Her research interests focus on P2P lending and crowdsourcing. She has published in a variety of journals, and in the proceedings of conferences such as the Hawaii International Conference on System Sciences. She has also presented her research at the Workshop on Information Systems and Economics (WISE).
Juan Feng
Juan Feng ([email protected]) is an associate professor in the Department of Information Systems, College of Business, City University of Hong Kong. She received her Ph.D. in business administration with a dual title in operations research from Pennsylvania State University. Her research focuses on the economics of information systems. She has published in Management Science, Information Systems Research, Production and Operations Management, Marketing Science, and other journals.
Bin Gu
Bin Gu ([email protected]; corresponding author) is a professor and associate dean at the W.P. Carey School of Business at Arizona State University. He received his Ph.D. in operations and information management from the Wharton School at the University of Pennsylvania. His research interests focus on online social media, user-generated content, mobile commerce, and online platforms. His work has appeared in Management Science, MIS Quarterly, Information Systems Research, Production and Operations Management, and other journals. He serves as senior editor for MIS Quarterly and an associate editor for Information Systems Research.
Pengzhu Zhang
Pengzhu Zhang ([email protected]) is Chair Professor and Head in the Department of Management Information Systems, Antai College of Economics and Management, Shanghai Jiaotong University. He received his Ph.D. from Xi’an Jiaotong University. His research focuses on decision support in innovation and health management. His research has been published in many academic journals including MIS Quarterly, Decision Support Systems, Information & Management, Computers in Human Behavior, PLos One, Journal of the American Society for Information Science and Technology, Journal of Nanoparticle Research, International Journal of Information Technology & Decision Making, Government Information Quarterly, Journal of Management Analytics, as well as some important journals in Chinese.