%ChatGPT and other generative %ArtificialIntelligence (A.I.) platforms might be able to predict the movements of stocks, according to a study by the University of Florida.
Alejandro Lopez-Lira, a finance professor at the University of Florida, says that an analysis he undertook shows that large language A.I. models may be able to forecast stock prices.
The professor used ChatGPT to review news headlines for whether they were positive or negative for a stock and found that ChatGPT’s ability to predict the direction of the next day’s share price movement was extremely accurate.
Specifically, Lopez-Lira had ChatGPT examine more than 50,000 headlines about public stocks on the New York Stock Exchange and Nasdaq index, and then predict the next day’s share price movement.
With more power and better datasets, A.I. models may be able to predict the movement of stock prices even more accurately, said Lopez-Lira, noting that the chatbot has capabilities that weren’t originally planned when it was developed.
In March, investment bank %GoldmanSachs ($GS) issued a report saying that 35% of financial jobs are at risk of being automated by artificial intelligence.
However, the University of Florida experiment did not include target prices or have ChatGPT do any math. It simply reviewed headlines from the financial press.
Still, Lopez-Lira said he was surprised by the results, adding they suggest the future ability to incorporate A.I. and machine learning into stock trading strategies.
Lopez-Lira said in media interviews that several hedge funds have already reached out to him to learn more about his research.