Mimicking Finance
Public Lecture in the Dean’s Distinguished Speaker Series | Hybrid, 15 January 2026, NBS Auditorium, Wee Cho Yaw Plaza, and via Zoom
As artificial intelligence becomes more capable of recognising patterns, questions are emerging about what machines can replicate, and where human judgment continues to matter. This public lecture examined how AI can learn from past financial behaviour, what it can and cannot predict, and how this may reshape decision-making, incentives, and careers in finance.
The session featured Lauren H. Cohen, L.E. Simmons Professor of Business Administration at Harvard Business School. It was moderated by Timothy Tay, Chief Investment Officer for APAC Credit at UBS Global Wealth Management and hosted by the Division of Banking and Finance at Nanyang Business School.
The following is an edited transcript of the Q&A segment.
Mr Tay:
Your research shows that a large portion of a fund manager’s actions can be predicted by a model trained only on their past behaviour. Does this mean that being predictable should be penalised?
Prof Cohen:
I would not frame this as punishment. The key point is pricing. In well-functioning labour markets, we do not pay a premium for tasks that are easily replicated. If a portion of behaviour can be mimicked at low cost, it should be compensated accordingly. The remaining part, the decisions that cannot be inferred from past actions, is where skill and value creation tend to reside.
Mr Tay:
Some might argue that what you describe as the “unpredictable” portion is simply randomness. How do we distinguish skill from luck?
Prof Cohen:
That concern is valid, which is why we adjust for risk. What we find is that the harder-to-predict decisions generate higher risk-adjusted returns consistently over long horizons. These results hold across decades. This suggests that we are not just observing noise, but systematic differences in how managers add value.
Mr Tay:
If markets learn from these patterns, wouldn’t today’s innovation eventually become predictable as well?
Prof Cohen:
Yes, and that is exactly how markets evolve. Once a pattern is identified and exploited, it comes under pressure. The boundary of predictability moves. Innovation is not a fixed category. It is a moving frontier that forces participants to adapt continuously.
Mr Tay:
There is growing concern about over-reliance on AI, especially during periods of market stress. Are institutions delegating too much decision-making authority to machines?
Prof Cohen:
At this point, not really. Most organisations are still cautious. AI is used primarily to support routine tasks. The analogy I often use is autopilot in aviation. It works well under stable conditions, but you still want a human pilot when unexpected shocks occur. Finance is no different.
Mr Tay:
You mentioned that AI struggles with true novelty. Can you elaborate on that limitation?
Prof Cohen:
AI systems are trained to reward answers that resemble past patterns. They are very good at completing “the cat sat on the mat”, but innovation often comes from answers that do not fit historical templates. Many breakthroughs, whether in finance or medicine, arise precisely from what has not been said or done before. That remains difficult for pattern-based systems.
Audience Member:
What does this mean for jobs and careers, especially for students entering finance?
Prof Cohen:
I actually think this is an exciting time. Finance has always been an early adopter of new technologies because the stakes are high. AI will change the nature of work, but it will also create new roles. The emphasis will shift toward judgment, problem framing, and innovation, rather than execution of routine processes
Download the slides here, and watch the webinar here:
