Researcher ORCID Identifier
0009-0004-2143-7268
Graduation Year
2024
Date of Submission
12-2023
Document Type
Open Access Senior Thesis
Degree Name
Bachelor of Arts
Department
Economics
Reader 1
Eric Hughson
Terms of Use & License Information
Rights Information
© 2023 Ian J Larson
Abstract
Abstract
This paper investigates whether ChatGPT has any predictive power when it comes to the stock market. Since its public release in December 2022, OpenAI’s ChatGPT has become a popular tool for millions of users across the world. With new use cases being found every day, I explore whether the system has any utility at predicting where stock prices will go. Using a 10-year study window, I prompt ChatGPT to give U.S. equity picks on a one-year holding period basis, segregating its picks into 3 different category sizes across 8 directed prompts. I examine the 24 equal weight portfolios of the stocks and I determine whether the tool is successful in outperforming the market. In the end, I find that 21 of the 24 portfolios have average alphas which are positive and statistically significant. With other insights into which prompts and categories are the most successful, this paper has preliminary findings which indicate that ChatGPT does have some predictive ability in the financial markets.
Recommended Citation
Larson, Ian J., "AI-nvesting: An Empirical Analysis with Sector Categorization and Prompt Complexity Considerations Assessing the Predictive Power of ChatGPT in Stock Market Forecasting" (2024). CMC Senior Theses. 3491.
https://scholarship.claremont.edu/cmc_theses/3491
