The Impact of Generative Artificial Intelligence on Individual Manual Investment Decisions: Empirical Evidence from Mutual Funds

Abstract

The rapid ascent of generative artificial intelligence (GAI) has led individual investors to seek guidance from GAI-based consulting tools such as GAI-based investment consultants (GAICs). Yet, there is scant empirical effort examining the business impact of such GAI tools on individual investors' financial market investments. To fill the critical gap, we collaborate with Ant Fortune, Alibaba’s leading investment arm, and analyze data related to the rollout of Ant Fortune’s GAIC, Zhi Xiaobao. Our analyses provide the first empirical evidence showing that the use of GAIC positively influences investment decisions, redemption activities, and overall returns. Interestingly, contrary to the common belief that novice investors could benefit from AI investment tools for accessible investment information, we find that experienced investors harness more benefits from GAIC, utilizing their existing financial acumen. In addition, we document that the platform’s influence on returns is more significant for risk-seeking investors, suggesting that GAIC could amplify their market decision making efficacy. Despite that novice and risk-averse investors engage more redemption behaviors, they do not attain the equivalent investment return relative to experienced and risk-seeking counterparts, highlighting the role of financial literacy in harnessing the economic benefits of GAICs. In summary, GAICs enhance decision making for experienced and risk-tolerant investors but offer limited advantages to novice and risk-averse investors. Our research not only provides essential managerial insights for platform managers considering GAIC applications, but also sheds light for policy makers in understanding how to improve the use of GAICs for vulnerable investor segments.

Keywords: fintech, generative AI, financial investment, algorithm and human, quasi-natural experiment

JEL Classification: G15,D14