摘要
Using a robust textual analytic method, we decompose the P2P loan description into common and distinctive contents, which contain general and unique information provided by borrowers. We then investigate the role of the distinctive content in affecting P2P lending decisions and out-comes. Controlling for loan/borrower characteristics, loan applications with more distinctive content are more likely to be funded, and have larger amounts and longer maturity, but these loans carry higher interest rates and default probability. Overall, the evidence suggests that borrowers use self-reported loan descriptions to hype their loan applications rather than provide soft information to reduce information asymmetry.
学科与关键词
关键词
Peer-to-peer lendingDistinctive contentHypeLoan descriptionDefault probability
- Predicting loan default in peer坼to坼peer lending using narrative dataYufei, Xia; Lingyun, He; Yinguo, Li; Nana, Liu; Yanlin, DingJournal of Forecasting, 2020, 39(2): 260-280.被引 19
- Credit risk evaluation model with textual features from loan descriptions for P2P lendingZhang, Weiguo; Wang, Chao*; Zhang, Yue; Wang, JunboElectronic Commerce Research and Applications, 2020, 42: 100989.被引 20
- Analyzing credit risk among Chinese P2P-lending businesses by integrating text-related soft informationLiang, Kun*; He, JunElectronic Commerce Research and Applications, 2020, 40: 100947.被引 31
