Whose reviews are most valuable for predicting the default risk of peer-to-peer lending platforms? Evidence from China

Online reviews of a firm may come from diverse sources including real customers, competitors, and the firm itself. Review manipulation by posting fake negative reviews about competitors or fake positive reviews oneself has major impacts on product sales and firm reputation. This study aims to answer the question of whose reviews are most valuable for predicting a firm's default risk. To uncover the value of manipulated and authentic reviews in firm default risk prediction, we conduct an empirical analysis using unique weekly panel data from a third-party portal of online peer-to-peer lending platforms in China. The results indicate that firm default probability increases with the number of manipulated positive reviews in the short term but decreases with the number of manipulated positive reviews posted over the long term. In addition, the signaling role of manipulated positive reviews is stronger when the peer-to-peer lending platform experiences more intense pressure such as downturn of business performance, stricter financial regulation policies, or aggressive attacks from competitors. Manipulated negative reviews are harmful for peer-to-peer lending platforms, which will increase the probability of platform default. Finally, authentic positive reviews are positively associated with default due to the overconfidence effect in the online lending context, and the authentic negative reviews in the short term work as a significant signal for fraud risk.