A recent study conducted by researchers at Sheffield Hallam University reveals that large language models, such as ChatGPT, can effectively predict future interest rate decisions made by central banks.
The research team from Sheffield Business School analyzed speeches delivered by members of the Bank of England’s Monetary Policy Committee shortly before interest rate votes. By utilizing ChatGPT, they classified each speech as dovish, neutral, or hawkish based on its tone and content. This classification was subsequently integrated into an econometric model, which successfully forecasted how each member would likely vote in the following one or two policy meetings.
The findings indicated that the sentiment analysis performed by ChatGPT on the speeches was a statistically significant factor in determining future voting behavior. Specifically, committee members who presented more neutral speeches tended to favor interest rate hikes in subsequent meetings.
Dr. Drew Woodhouse, a senior lecturer in economics at Sheffield Business School and the lead author of the study, remarked, “Our findings highlight the predictive potential of tools like ChatGPT for processing human beliefs and expectations, which has major implications for forecasting policy decisions and modeling economic expectations.”
By employing natural language processing and textual analysis, ChatGPT demonstrated its ability to classify the intricate language used by central bankers, adeptly connecting speech content with eventual policy actions.
The research suggests that this methodology could be applied to investigate other facets of central bank communications, such as forward guidance. It also showcases how accessible AI tools like ChatGPT can enhance economic analysis and inform financial decision-making.
In April, it was reported that JPMorgan has developed a ChatGPT-based language model to analyze Federal Reserve statements and speeches in order to identify potential trading signals. In their note, the bank stated that “preliminary applications are encouraging,” and the model has already been expanded to encompass the European Central Bank and the Bank of England, with plans to include additional central banks in the future.