A study published by the University of Florida found that using ChatGPT for sentiment analysis of publicly traded companies yielded investment returns well above the market average.
ChatGPT, an artificial intelligence chat application developed by OpenAI, is a large-scale language model, a type of artificial intelligence algorithm, but not explicitly trained for financial analysis.
The study used headline data for various stocks from October 2021 to December 2022 and tracked actual stock prices over the same period. The study omits headlines, daily stock movement reports, and duplicate headlines where companies are passively mentioned in an article about another subject. In total, the survey extracted his 67,586 headlines related to 4,138 companies.
During the study period, the researchers found that a strategy of investing $1 in the market, buying on good news, selling on bad news (identified by ChatGPT itself) yields more than $5.50 when transaction costs are not taken into account. I discovered. Even though the market average decreased over the same period. This strategy was repeated daily.
Considering transaction costs, investments based on ChatGPT’s positive or negative ratings outperformed the market average if transaction costs were 25 basis points or less per trade.
The study also found that a strategy of selling only bad news outperformed the opposite strategy of buying only good news. Researcher Alejandro López Lira explained that bad news drives down stock prices more than good news drives them up, and ChatGPT was able to detect this pattern.
Additionally, companies with smaller market caps are more sensitive to headline sentiment than larger companies. On average, positive headlines can lift small- and medium-sized stocks by 60bps, but only 20bps for large-cap stocks, according to López Lira.
While the study wasn’t meant to explain that gap, López Lira said that less is known about small caps, which could make investors more sensitive to headlines than large caps. suggests one of the possible explanations. López Lira argues that investors can better balance their existing knowledge of headlines and big companies and be less impacted by headlines.
Lópezrilla said AI digests headline data and performs sentiment analysis much faster than humans, so the main advantage of using AI for trading is that it can trade faster than competitors who rely on human expertise. He explains that it is the ability to act on public sentiment.
Tags: AI, artificial intelligence
