Keynotes
Keynote 1: Michinori (Mitch) Kanokogi
Keynote Speaker — September 10, 2026
Affiliation: Nippon Life Asset Management / CFA Society Japan
Talk Title: Analysts with Their Agents: How Agentic AI is Redefining Investment Management
Abstract: Generative AI has evolved from chatbots into agents capable of working autonomously over long horizons. Inside an asset management firm, we are building AI agents that work alongside human investment professionals. This talk reports on the possibilities and the limits we have encountered in developing these agents in production: where AI succeeds, where it fails, and where humans still need to be involved. It closes by proposing a research agenda for establishing human-AI feedback loops that improve investment decision making.
Bio: Michinori (Mitch) Kanokogi is Head of Data, AI & Quant Research at Nippon Life Asset Management, leading AI R&D for investment processes. He previously led AQR’s Japan operations and worked at Russell Investments, UBS, PwC, and Deloitte. He holds a BEng from the University of Tokyo, an MEng from Kyoto University, and an MBA from INSEAD. A CFA charterholder and Vice President of CFA Society Japan, he authored “How Generative AI is Transforming Asset Management” and translated key finance and AI books, including “Advances in Financial Machine Learning”.
Keynote 2: Kiyoshi Izumi
Keynote Speaker — September 11, 2026
Affiliation: The University of Tokyo
Talk Title:
What Does Generative AI Generate in Financial Markets?
- Frontiers of Generative AI Research for Finance -
Abstract: Beyond summarization and translation, what can generative AI uniquely produce in financial markets? This lecture presents four pillars. Coupling large language models with causal discovery over a million cause-effect pairs extracted from Japanese earnings reports generates economic relationships that improve forecasts. CoFinDiff, a wavelet-based conditional diffusion model, generates synthetic price paths under specified trend and volatility, strengthening deep hedging in rare regimes. A multi-agent framework combining explicit know-how with tacit heuristics learned via Direct Preference Optimization generates the analytical process itself, improving post-earnings stock-move prediction. Finally, persona agents generate digital twins of markets and societies for counterfactual policy simulation.
Bio: Kiyoshi Izumi received his Ph.D. from Graduate School of Arts and Sciences, the University of Tokyo, in 1998. From 1998 to 2010, he worked at the National Institute of Advanced Industrial Science and Technology, AIST. In 2010, he joined Graduate School of Engineering, the University of Tokyo, as an Associate Professor, and has been a Professor there since 2015. His research interests include financial informatics, market simulation, and financial data mining. He is a member of IEEE and Vice President of the Japanese Society for Artificial Intelligence.
Stay Updated
Follow @ieee_cifer_infoJoin Google Group
Important Dates (AoE)
Paper submission deadline is firm and has been finalized (no-extension).
- Paper Submission: May 15, 2026
- Notification: July 14, 2026
- Camera Ready: August 14, 2026
- Early Registration: August 1, 2026
- Conference: September 10-11, 2026