Spend enough time looking across institutional portfolios and a curious pattern emerges: different firms, different brands and different investment philosophies often lead to remarkably similar portfolio construction.
The phenomenon may seem easy to dismiss as a coincidence. It isn’t.
Asset management has developed an extraordinarily effective system for producing consensus, and it begins with the talent.
Many investment professionals come from the same relatively concentrated group of undergraduate institutions and then attend the same leading MBA programs. Then comes professional training: essentially one globally recognized CFA curriculum and examination framework. Through it, thousands of investment professionals are taught to analyze markets, value securities, and think about risk.
None of this is inherently bad. Shared standards create competence. The problem arises when a system designed to establish common foundations begins to produce common conclusions.
The result is a form of intellectual herding. Managers may describe their strategies differently, but they frequently work from similar frameworks, examine similar securities, and ultimately arrive at similar portfolios. The industry is exceptionally good at identifying what is attractive “now.” It is less consistently rewarded for imagining what might matter next.
Now, artificial intelligence may be making this problem even more pronounced.
The issue is not that AI lacks depth or discipline. It’s that investment professionals are increasingly using the same tools, trained on overlapping information, to answer similar questions.
Mercer’s 2026 survey of 131 asset managers found that 55% had already integrated AI into at least one investment process, while 91% expected to increase their use over the next year. Sixty-three percent reported using off-the-shelf AI tools, while 58% use some vendor-provided data.
If thousands of professionals ask similar models similar questions, using similar datasets, we should not be surprised if the resulting insights begin to converge even more.
That creates an uncomfortable paradox. AI promises to democratize intelligence, but widespread adoption of the same intelligence infrastructure could also reduce differentiation.
As a top-quartile CIO and portfolio manager for nearly a decade, I credited much of my performance to a solid process and a great team. However, I also brought a different perspective to the table. I did not attend any of those schools, and although I paid for the entire team to earn their CFAs, I never studied for or took the exams myself. As a leader, I made sure that individual thought and observation were valued as highly as formal training.
The strategic question for asset managers, therefore, should not simply be, “How do we use AI?” It should be: “How do we use AI without becoming more like everyone else?”
The answer starts with deliberately cultivating intellectual diversity. Investment organizations need people who have been trained differently, think differently, and are willing to challenge the consensus.
Analysts and portfolio managers need research processes designed not merely to identify what is working today, but to explore second- and third-order consequences: What is changing beneath the surface? Which behaviors are emerging? What assumptions embedded in today’s models may prove wrong?
The best investment organizations of the future won’t necessarily be those that use AI most aggressively. They’ll be the ones that use AI to expand the range of questions they ask while retaining enough independent human judgment to recognize an idea or trend that everyone else has missed.
The industry’s competitive advantage has historically been framed in terms of information, technology, and analytical horsepower—resources that are becoming increasingly commoditized.
What’s becoming a scarce resource now might just be original thought.