Recently, the first wave of research using ChatGPT for the financial sector has arrived and the results of the technology have lived up to expectations. Two new papers were released recently that applied ChatGPT to market-relevant tasks. One focused on deciphering Fed statements and the other looked into predicting stock moves from headlines. Both tasks were a success according to the early results, showing a great progress in the use of technology to consider nuance and context from news articles, tweets, and speeches when making trading decisions.
Slavi Marinov, head of machine learning at Man AHL, affirmed the positive results. Man AHL has been using technology known as natural language processing (NLP) to read texts such as earnings transcripts and Reddit posts for quite some time.
One of the papers, titled Can ChatGPT Decipher Fedspeak?, examined the accuracy of the AI chatbot in comparison to classifications based on dictionaries, as well as the commonly used model from Google called BERT. In this task, the chatbot was tested on its accuracy in recognizing if Fed statements were hawkish or dovish. It proved itself to be most accurate. What makes ChatGPT especially impressive is that it was able to explain its classifications of Fed policy statements in a way similar to the central bank’s own analyst.
In the other study, Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models, developers from the University of Florida challenged ChatGPT to act as a financial expert and interpret corporate news headlines from after 2021. This task was completed successfully as well – the answers it gave showed a statistical link to the stock’s subsequent movements. This shows that the technology is able to comprehend the implications of news accurately.
For many of Wall Street’s quants, this technology is nothing new as language models have been utilized for years to develop strategies. What ChatGPT brings to the table is a way to do this faster and more efficiently. Plus, it opens up new areas of analysis to a broader group of financial professionals.
OpenAI developed this artificial intelligence chatbot, which allows NLP to reach a new level of accuracy. Their successful programs have been a major advantage to algorithmic traders and data-driven investors who rely on language models in their strategies.
Man AHL is a London-based quantitative hedge fund that has been making use of NLP to understand texts like earnings transcripts and Reddit posts for some time. Slavi Marinov is the head of Machine Learning at Man AHL and he is enthusiastic about the success of ChatGPT. He believes that this progress can help users get more value from reams of data from tweets and speeches, turning them into meaningful signals for trading.
