Download This PaperAI and Perception Biases in Investments: An Experimental Study
68 Pages Posted: 17 Apr 2024 Last revised: 11 Apr 2025
Date Written: April 8, 2024
Abstract
AI has the potential to broaden access to investment advice. But can it replicate the investment preferences and rationales of investors that have been historically underrepresented? We ask 1,272 human and 1,350 AI respondents to rate stocks, bonds, and cash investments. First, default AI-generated responses overrepresent the preferences of young high-income individuals. However, algorithmic bias disappears with demographically seeded prompts. Second, AI-generated free-form responses closely capture human rationales: risk and return, financial knowledge, and past experiences. Third, AI can help identify where a lack of financial knowledge induces uncertainty about investment, as shown in our textual analysis of transitivity violations.
Keywords: Investment preferences, Large language models, Behavioral biases, Experimental economics, Financial surveys, Generative AI. JEL Classification: C1, G10, G11, G12
JEL Classification: C1, G10, G11, G12.
Suggested Citation: Suggested Citation
Anastassia Fedyk
University of California, Berkeley - Haas School of Business ( email )
545 Student Services Building, #1900
2220 Piedmont Avenue
Berkeley, CA 94720
United States
Peiyao Li
University of California, Berkeley - Haas School of Business ( email )
545 Student Services Building, #1900
2220 Piedmont Avenue
Berkeley, CA 94720
United States
University of California, Berkeley - Haas School of Business ( email )
545 Student Services Building, #1900
2220 Piedmont Avenue
Berkeley, CA 94720
United States
Institute for the Study of Labor (IZA)
P.O. Box 7240
Bonn, D-53072
Germany
HOME PAGE: http://www.iza.org/en/webcontent/personnel/photos/index_html?key=918
European Corporate Governance Institute (ECGI)
c/o the Royal Academies of Belgium
Rue Ducale 1 Hertogsstraat
1000 Brussels
Belgium