AI and Capital Markets Frontier Forum Held at Shanghai Jiao Tong University, Exploring Investment Decisions and Investor Education in the AI Era 2026-11-07

The AI and Capital Markets Frontier Forum was held on November 1 at Antai College of Economics and Management (ACEM), Shanghai Jiao Tong University (SJTU), drawing more than 100 participants. Jointly hosted by the Department of Finance at ACEM, the SJTU-BOC Institute of Technology and Finance, the Institute of Industrial Innovation, SJTU and the Financial Powerhouse Laboratory at ACEM, and co-organized by the Ruiyuan Fund Investor Education Base, the forum took 'Investment Decisions and Investor Education in the AI Era' as its theme, bringing together scholars and financial practitioners from China and abroad to examine how artificial intelligence is reshaping capital market operations, investment research and investor education.

The forum was co-initiated and co-chaired by Prof. Jun Tu of ACEM's Department of Finance and Prof. Jun Zhao, Dean of Institute of Industrial Innovation, SJTU and Professor at ACEM.

The opening ceremony was moderated by Prof. Wenfeng Wu. Noting that AI is profoundly changing how investment research is conducted and how investors access information, he said the forum was designed to combine academic research with institutional practice around three themes—investment decisions, information quality and investor education—and to address whether gains in analytical efficiency can translate into investment value, and what new risks model application may bring.

Prof. Feng Gu, Vice Chair of the University Council of SJTU and Professor at ACEM, addressed the forum. With the University marking its 130th anniversary this year, he noted that AI has drastically shortened the time needed to gather materials, read documents and form judgments, substantially lowering the threshold for obtaining analytical conclusions—but it has not eliminated the uncertainty inherent in investing. 'Whether an answer reads smoothly and whether it is correct are two entirely different things.' For investors, he stressed, recognizing the limits of AI tools matters far more than receiving a buy-or-sell conclusion, and education must answer the core question of 'how to cultivate people stronger than artificial intelligence.'

Prof. Shaoxuan Liu, Vice Dean of ACEM and Executive Dean of the SJTU-BOC Institute of Technology and Finance, introduced the College's pioneering work in technology finance, including the country's first professional master's program in technology transfer (MTT) launched in 2021, and called on universities to break down disciplinary and industry barriers. 

Prof. Jun Zhao outlined Institute of Industrial Innovation's priorities—supporting the science and technology innovation district around the university, advancing the integration of AI with finance, technology transfer and business administration, and pooling technological, industrial and capital resources. Prof. Jun Tu explained the forum's rationale: as AI becomes embedded in investment decisions, academia should help retail investors identify information, perceive risk and understand the limits and model risks of AI—because pricing in the secondary market shapes capital allocation in the primary market and, in turn, technological innovation.

Moderated by Prof. Jun Tu, the morning keynote was delivered by Prof. Liyan Yang of the Rotman School of Management at the University of Toronto on ‘Open Banking and Data Quality.’ Open banking is meant to promote competition and lower interest rates through data sharing, but his analysis reveals an overlooked mechanism: once data are opened up, banks have weaker incentives to clean their data while users have stronger incentives to manipulate it, which can degrade data quality and defeat the policy’s purpose. The ‘endogenous data quality’ framework offers a new lens for governing financial data.

Two academic sessions followed. Chaired by Prof. Qi Zhang, Vice Dean of ACEM, the session on ‘Institutional Trading and Asset Pricing’ featured Assoc. Prof. Jianfeng Hu of Singapore Management University’s Lee Kong Chian School of Business, who drew on three decades of daily hedge fund trading data to show that institutions’ information advantage is shifting from a ‘tool dividend’ to a ‘data dividend’ and a ‘compute dividend’; and Prof. Xi Wang of Peking University’s School of Economics, who introduced the Wasserstein distance into asset pricing models to improve the estimation of stochastic discount factors.

Chaired by Assoc. Prof. Xiangwei Wan, the session on ‘Information Production, Dissemination and Price Discovery’ featured Asst. Prof. Wu Zhu of Tsinghua University, whose quasi-natural experiment built on the launch of FactSet Mercury found that generative AI broadens analysts’ information sources but reduces forecast accuracy and weakens market reactions, because information input exceeds analysts’ cognitive processing capacity; and Research Fellow Zhipeng Zhang of ACEM, who identified factual and framing deviations in 861 earnings briefings from 398 STAR Market companies and found that markets react to verifiable deviations without promptly correcting prices.

The afternoon sessions opened with a dialogue between industry and academia that brought the Class of 2026 Master of Finance students into direct contact with front-line practice. Moderated by Prof. Yun Feng, the first keynote was delivered by Prof. Dacheng Xiu of the University of Chicago’s Booth School of Business on ‘How AI Is Changing Investment Research.’ He outlined three core capabilities—machine learning, large language models and agents—and argued that the bottleneck for AI replacing humans in finance is an extremely low signal-to-noise ratio: markets are noisy, data are limited and feedback shifts with strategy, unlike mathematics or programming where feedback is precise and immediate. Institutions’ future edge, he said, will lie in ‘unique proprietary data plus a reusable research system plus workflow-level engineering capability,’ with investment research becoming a collaboration between people and large numbers of agents—humans always bearing final judgment and responsibility.

Moderated by Prof. Jun Zhao, the second keynote was delivered by Yu Jin, Managing Director of Ruiyuan Fund Management. Describing the AI industry as ‘capital in the second half, adoption in the first half,’ he noted that capital expenditure intensity has approached that of the largest railway construction cycles in history while enterprise adoption remains early, and that the industry’s ‘five-layer cake’ structure is currently inverted—overinvestment upstream, with value at the application layer yet to be released. For active investment research, AI can enhance data processing and analysis, but the core value of investing still lies in deep understanding of intrinsic value, critical thinking and first-principles judgment.

Moderated by Assoc. Prof. Wenjian Xu, the third keynote was delivered by Hailin Hu, Chief Operating Officer of TokenRhythm (Shanghai Jiyuan Lvdong Technology), on financial AI agent systems. He introduced the concept of an ‘agent harness’—if the large model is the horse, the harness organizes context, invokes tools, verifies results and reuses evaluations—and distilled three computational constraints for financial markets: comparable evidence, layered expectations and updated judgments. Moving from a personal assistant to an organizational agent, he said, hinges on traceable research trails and reusable organizational memory. TokenRhythm is an incubated company of the Institute.

The forum closed with a roundtable on ‘Investment Practice and Investor Education in the AI Era,’ moderated by Assoc. Prof. Haibing Shu. Six panelists from banking risk control, fintech, public funds, trust and academic research—Jun Sheng, Miao Li, Fanglan Wu, Hui Xie, Junwei Tian and Hui Bu—shared their practices. In a quick poll on the ‘trust boundary of AI,’ all agreed that AI’s most reliable capabilities today remain information retrieval and data processing, while investment decisions and final judgments must remain in human hands.

Prof. Yi Tan, Executive Dean of the International School of Finance at SJTU and Vice Dean of the Graduate School, delivered the closing address. He affirmed the forum’s academic depth and practical value and noted that, as AI takes over basic and standardized work, the industry faces shrinking entry-level positions and a gap in talent pipelines, reshaping the traditional path of advancement through accumulated proficiency. He looked forward to deeper cooperation with financial institutions and stressed that ‘choosing SJTU means choosing responsibility.’

Participants agreed that artificial intelligence is profoundly reshaping the operating logic and investment paradigm of capital markets, and that academia and industry should work together to explore a healthy path for technology-enabled finance.