GPT-4 outperforms human analysts in forecasting corporate profits.

Researchers from the University of Chicago have found that large language models (LLMs) can analyze financial statements with accuracy that not only matches but often exceeds the results of professional analysts, reports Venture Beat.

The study focused on GPT-4, a large language model developed by OpenAI, and tested its ability to analyze corporate financial statements and forecast future revenue growth. Notably, GPT-4 outperformed human analysts even when provided only with standardized, anonymous balance sheets and profit and loss statements without any textual context.

In the study, GPT-4 surpassed human analysts, who typically forecast with accuracy in the range of 53-57%. Researchers used a new method that includes "thought chain" prompts to guide GPT-4's reasoning, helping it simulate the analytical process of financial analysts by identifying trends, calculating coefficients, and synthesizing information to form forecasts.

💵 If you found this helpful, consider showing your support by tipping me through the Binance Tipping feature. Your generosity will help me continue to provide high-quality content. Thank you!