The Dark Side of Algorithms? The Effect of Recommender Systems on Online Investor Behaviors

Despite the widespread adoption of recommender systems by online investment platforms, empirical research into their impact on online investors' behaviors is scarce. Using data from a global e-commerce platform, the authors of this study adopt a regression discontinuity design to causally examine the effects of recommender systems on online investor behaviors, specifically in a mutual fund investment context. The results show that funds featured by recommender systems prompt significantly more purchases. This effect is especially salient among unsophisticated investors, who appear more likely to follow system-provided recommendations. Further analysis also reveals that these investors tend to suffer significantly worse investment performance after purchasing the recommended funds. Thus, recommender systems threaten to amplify wealth inequality among investors in financial markets.