ChatGPT was successful in decoding Federal policy statements like the central bank analyst. Here’s all you need to know about the AI bot’s new capacities.
ChatGPT gains new abilities
The first rush of academic research applying ChatGPT to the finance sector is almost here. Two new papers that used the AI bot in market-relevant tasks provide insights into its capacities. One study looked at its ability to decipher if Federal Reserve statements were hawkish. Another study focused on determining the effects of headlines on a stock. The AI bot aced both tests, showing a possibly major step in the use of tech for turning text from articles, speeches, and tweets into trading signals.
Nevertheless, this is not a new process on Wall Street. For a long, quantitative points have been using language models for underpinning the chatbot to inform strategies. However, the study reveals OpenAI developed technology is reaching a new level for parsing nuance and context. “It’s one of the rare cases where the hype is real,” stated Slavi Marinov. Marinov is the head of machine learning at Man AHL. The firm has been using natural language processes to read texts for years. “Previously you had to label the data yourself. Now you could complement that with designing the right prompt for ChatGPT,” added Marinov.
More on the new study
In the study titled “Can ChatGPT Decipher Fedspeak?,” researchers from the Fed found that the AI bot was coming close to humans figuring out central bank statements. It was also successful in explaining classifications of the Fed polity in a style resembling the central bank’s analyst. For example, take the following statement from May 2013: “Labor market conditions have shown some improvement in recent months, on balance, but the unemployment rate remains elevated.” The AI bot clarified the line is dovish since it suggests the economy is not yet completely recovered.
