Does Transparency Matter? The Moderating Effect of Algorithm Transparency on Social Influence in Investment Decisions

Abstract

With the rapid development of intelligent investment advisory platforms, the integration of algorithmic recommendations and social functions has become a prominent feature in the fintech field. Based on self-determination theory, this study explores how algorithmic transparency moderates the mechanisms by which social influence affects investment decisions. Social influence operates through two distinct pathways: social learning (inferring asset quality from others' choices) and social utility (deriving psychological value from coordinated investment behavior). Through two progressive online experiments (total sample size N=400), the study found that: social influence (friends' investment information) significantly increases individual investment intention; algorithmic transparency negatively moderates the social learning effect, reducing the positive effect of social influence on investment intention by approximately 35% under high transparency conditions; exploratory heterogeneity analysis suggests that the moderating effect of algorithmic transparency on the social utility mechanism is weaker, though conclusive evidence requires further investigation; perceived autonomy plays a partial mediating role in the above moderating effect, with a mediation proportion of 34.8%. This study reveals that algorithmic transparency, by influencing users' perceived autonomy, differentially moderates the underlying logic, providing empirical evidence for the design of fintech platforms and the formulation of regulatory policies.These findings are primarily applicable to nonprofessional investors, as robustness analyses reveal that the moderating effect diminishes among individuals with high financial literacy.