Buffalo, NY - Can ChatGPT pick stocks better than the market? Canisius University Assistant Professor of Economics and Finance Marc LoGrasso, PhD, set out to answer that question, and his research is earning international recognition.
LoGrasso is the recipient of the 2026 Best Paper Award from Modern Finance, an international peer-reviewed journal, for his research exploring whether ChatGPT could have generated abnormal returns through stock recommendations. His paper, "Could ChatGPT Have Earned Abnormal Returns? A Retrospective Test from the U.S. Stock Market," examines the increasingly important intersection of generative artificial intelligence (AI) and financial markets.
The answer, LoGrasso found, is nuanced.
Testing ChatGPT as a Stock Picker
In a retrospective analysis spanning nearly four decades, he asked GPT-4 to select 10 U.S. stocks each year from 1985 through 2021. LoGrasso limited the model to information that would have been available at the time of each selection, allowing him to test how the AI's recommendations would have performed without the benefit of hindsight.
The findings, however, were not consistent from year to year. While the portfolios produced positive abnormal returns overall, individual portfolio alphas were statistically significant in only about one out of every four years. The study found no years in which the portfolios significantly underperformed the market.
What ChatGPT's Stock Picks Revealed
LoGrasso's research grew out of a practical question about how everyday investors might use a technology that has quickly become part of the way people find and evaluate information.
“I started thinking about people who want to invest in the stock market but don’t know where to begin and wouldn’t go to a financial advisor for various reasons,” LoGrasso said. “These individuals might turn to a large language model like ChatGPT for guidance. The question became: would that actually work?”
His study suggests that ChatGPT took a relatively cautious approach to stock selection, tending to favor large, growth-oriented companies, particularly in the technology and health care sectors.
The research also illustrates some of the challenges of using generative AI for financial decision-making. LoGrasso found that precise prompting was important to obtaining specific stock recommendations from GPT-4.
What this research means for AI and investing
“This study shows both the promise and limits of AI as a financial advisor,” LoGrasso added. “ChatGPT may not replace human expertise or sophisticated portfolio optimization, but it demonstrates how emerging technologies can help everyday investors make informed decisions while minimizing downside risk.”
LoGrasso teaches undergraduate and graduate courses in finance at Canisius, including asset pricing, risk factors, market anomalies, derivatives, corporate finance, and new and alternative investments. His broader research interests include the role of information in investing and the use of data to better understand financial markets and investment decisions.
He earned his PhD in finance from the University at Buffalo and holds bachelor's degrees in mathematics and economics from Canisius. Before returning to the classroom in 2022, LoGrasso spent more than a decade working in higher education analytics.
The Modern Finance Best Paper Award recognizes an outstanding article published in the journal during the preceding calendar year. For the 2026 award, five papers published in Modern Finance during 2025 were identified based on citation counts and readership before being evaluated by the journal’s Best Paper Award Committee. LoGrasso’s paper was selected as the winner.
