Online peer-to-peer (P2P) lending grows remarkably in recent years. Supported by a unique P2P lending dataset, this study empirically investigated the determinants of delinquency and default behaviors of borrowers with a two-stage probability model. The results suggested that special collection actions could help reduce bad debt. Historical repayment performance of borrowers indicated by many factors would have significant impact on both delinquency and default. If the required repayment due date was during weekend, borrowers were less likely to be delinquent and default. But the effects were opposite for other holiday-related factors. The results by latent class model also demonstrated two types of borrowers, 35% of which were found to be of higher credit and quality. The determinants of delinquency and default affect somewhat differently between the two segments of borrowers. As one of earliest empirical studies on delinquency in P2P lending, this study offers several implications for research and practice.
Zhang, Chenghong; Lu, Tian; and Xu, Yunjie (Calvin), "Assessment of Borrowers' Delinquency and Default Behaviors in Online P2P Lending: A Two-stage Model" (2017). PACIS 2017 Proceedings. 51.