Can AI provide better financial analyses than human experts? This question was explored by US researchers, who arrived at an intriguing conclusion.
Researchers at the University of Chicago conducted a fascinating study on AI and financial analysis. They found that large language models (LLMs) can perform financial analyses with a level of accuracy that matches, and in some cases surpasses, that of professional analysts.
AI’s Potential to “Revolutionize the Industry”
The findings, published in a working paper titled “Financial Statement Analysis with Large Language Models,” could have significant implications for the future of financial analysis, according to VentureBeat. The technology website noted that AI has the potential to “revolutionize the industry.”
The researchers tested ChatGPT-4’s ability to analyze financial statements to predict future earnings growth. Remarkably, the OpenAI model outperformed human analysts, even when provided only with standardized, anonymized balance sheets and income statements without any contextual text.
Study: AI Capable of “Intuitive Thinking”
The language model generated “useful narrative insights into a company’s future performance,” the study’s authors wrote. ChatGPT-4 achieved an accuracy rate of 60 percent in predicting future profits or losses, significantly higher than the 53 to 57 percent range of human analysts’ forecasts.
The results suggest that “LLMs could play a central role in decision-making,” the researchers concluded. They inferred from their investigations that large language models are capable of “intuitive thinking” even with incomplete information.
Numerical Analysis: A Challenge for Large Language Models
These findings are particularly noteworthy because numerical analysis has traditionally been a challenging area for large language models. “One of the most demanding areas for a language model is the numerical domain, where the model must perform calculations as well as human-like interpretations and make complex judgments,” said Alex Kim, one of the study’s co-authors.
While it is unlikely that human expertise and judgment will be entirely replaced in the near future, powerful tools like ChatGPT-4 could significantly enhance the work of analysts and “potentially reshape the field of financial reporting analysis in the coming years,” according to VentureBeat.