The AI app ChatGPT is gaining ground in the financial world and, based on early results, it seems to justify the hype of the past few months. Two new recent studies have used ChatGPT to decipher the speech of the US Federal Reserve (FED) and predict stock movements from headlines, and in both cases, ChatGPT passed the tests, suggesting a major advance in using the technology to turn large amounts of text into trading signals.
Studies show that ChatGPT can predict the movements of the Fed
Anne Lundgaard Hansen and Sophia Kazinnik of the Fed conducted study #1 “Can ChatGPT decrypt Fedspeak?”, which showed that ChatGPT it went closer to humans to determine whether the central bank’s statements were dovish or aggressive. ChatGPT even explained its classifications of the Fed’s policy statements in a similar fashion to the central bank analyst, who acted as the human reference point for the study.
In the second study, “Can ChatGPT Forecast Stock Price Movements?”, from the University of Florida, Alejandro López-Lira and Yuehua Tang used ChatGPT to predict stock movements through the interpretation of headlines. corporate news. The results showed that ChatGPT responses were statistically correlated with stock movements, suggesting a correct interpretation of the news implications.
While it’s not surprising that machines can read almost as well as people, ChatGPT’s advancements seem poised to unlock new worlds of information and make the technology more accessible to a broader audience of financial professionals.
For Marinov, Man AHL’s head of machine learning, ChatGPT can potentially speed up the whole process. Nuance and context analysis technology has advanced, generating new data for financial decision making. The most sophisticated quants use NLP to measure the popularity of a stock on Twitter or incorporate the latest headlines about a company. ChatGPT’s advancements appear to make the technology more accessible to broader financial professionals.
Rate this news!
0 / 5 Number of votes: 1