Workshop Report

Professor Katsuhiko Okada of Kwansei Gakuin University's Institute of Business and Accounting presented his research and engaged in active discussions at the Bank of Japan's Institute for Monetary and Economic Studies (IMES) Finance Workshop titled "Applications of Machine Learning and AI to Finance Analysis," held at the Bank of Japan headquarters on November 28, 2025.

This workshop provided a groundbreaking opportunity for experts seeking to apply large language models (LLMs) to the financial sector to gather in one venue. With over 300 participants, the workshop set a record for the highest attendance in the history of the IMES Finance Workshop series, demonstrating strong interest in AI and machine learning within the financial industry. Participants included R&D specialists from Preferred Networks (PFN), Japan's leading AI company; portfolio managers from financial institutions; and econometricians and macroeconomists, spanning a diverse range from AI developers to practitioners. This breadth enabled multifaceted discussions from various perspectives.

Professor Okada's presentation, "From Words to Returns: Sentiment Analysis of Japanese 10-K Reports using Advanced Large Language Models," reported research findings on extracting sentiment from Japanese securities reports using state-of-the-art LLMs, including GPT-4, Claude, and Gemini, and analyzing the relationship with stock returns. In addition to expert commentary from designated discussant Professor Kei Nakagawa (Osaka Metropolitan University), vigorous Q&A sessions from the floor addressed machine learning interpretability and the financial applications of LLMs.

Participation in this workshop provided valuable insights into the cutting edge of AI-powered financial analysis from both academic research and practical application perspectives. We plan to advance our research on financial applications of LLMs further going forward.

Workshop: Bank of Japan IMES Finance Workshop "Applications of Machine Learning and AI to Finance Analysis"
Presentation Title: From Words to Returns: Sentiment Analysis of Japanese 10-K Reports using Advanced Large Language Models
Presenters: Katsuhiko Okada (Professor, Kwansei Gakuin University)
Designated Discussant: Kei Nakagawa (Professor, Osaka Metropolitan University)