Can we trust artificial intelligence with asset management? A consumer preference analysis of robo-advisor services

Abstract

Mathematical algorithms and AI-based robo-advisors are expected to offer more efficient and cost-effective financial management compared to traditional human advisors. However, for widespread adoption, a clear understanding of consumer perspectives is essential. This study conducted a discrete choice experiment (DCE) with a stratified sample of 1,000 participants between November and December 2023 to assess preferences regarding key attributes of robo-advisors. Specifically, the study examined which investor characteristics are associated with a willingness to delegate asset management to AI, and how trust in AI influences these preferences. Results indicated that expected returns, minimum investment thresholds, and management fees were the most valued attributes. Participants preferred high returns, low entry barriers, and minimal fees, but many expressed skepticism toward AI, favoring a hybrid model where AI shares decision-making authority with human advisors. In addition, by using factor analysis to incorporate variables identified in the Technology Acceptance Model (TAM) into the DCE, this study explored how investment behavior and trust in AI influence consumer choices related to the functions of robo-advisors. These insights inform strategic recommendations for policymakers and industry stakeholders to improve the acceptance and effective integration of robo-advisor services, emphasizing the need to build consumer trust and offer flexible, user-centered solutions.