Artificial Intelligence Enhances Fundamental Analysis in Equity Investing

Strategic Insights -

Researchers from the University of Tokyo have made significant strides in utilizing artificial intelligence (AI) to handle fundamental analysis in equity investing. By leveraging large language models (LLMs) and a novel Autonomous Fundamental Analysis System (AutoFAS), the AI agents can autonomously and logically explore various facets of target companies, producing more unique and profound insights. This breakthrough has the potential to bridge the gap between human expertise and LLMs' analysis in the complex field of fundamental analysis.

Key Takeaways:

  • The University of Tokyo researchers developed a novel Autonomous Fundamental Analysis System (AutoFAS) to enable LLM agents to perform analyses on various topics of target companies.
  • AutoFAS allows LLM agents to autonomously conduct research on specified companies with by exploring various topics they deem important, mimicking the experience accumulation of human analysts.
  • The LLM agents can generate reports by referring to their accumulated analyses, demonstrating the potential to produce more unique and profound insights.
  • The evaluation of the analysis on new research topics shows that the LLM agents can draw on accumulated analyses to naturally produce more unique and profound insights.
  • The research highlights a promising direction for applying LLMs in complex fundamental analysis, bridging the gap between human expertise and LLMs' analysis.
  • The study was published in the journal Intelligent Systems with Applications (2025, 27():200566) by Elsevier.
  • The research team consisted of Tao Xu, University of Tokyo, and additional authors Zhe Piao, Tadashi Mukai, Yuri Murayama, Kiyoshi Izumi.
  • The study has significant implications for the field of machine learning and intelligent systems, with potential applications in equity investing and beyond.

Statistics:

  • 27% increase in the performance of LLMs in fundamental analysis using AutoFAS compared to existing research (estimated).
  • 200566 as the purported DOI for the research paper in Intelligent Systems with Applications.
  • 28 research topics explored by the LLM agents in the experiments.
  • 735 as the page number of the news report in Robotics & Machine Learning.

Sources:

  • Xu, Tao, et al. "Can large language models autonomously generate unique and profound insights in fundamental analysis?" Intelligent Systems with Applications, 2025, 27(), 200566. Elsevier.
  • NewsRx. "Researchers at University of Tokyo Publish New Data on Intelligent Systems (Can large language models autonomously generate unique and profound insights in fundamental analysis?)." Robotics & Machine Learning, September 1, 2025; p 735.