What Happens When the Cost of Replicating Investment Expertise Approaches Zero?

Published on 05 November 2026
Investment Manager - Irene Bauer
Irene Bauer
Algo-Chain, Co-Founder

The demise of toll booth pricing models.

For much of the past three decades, the investment industry has benefited from a powerful economic reality: expertise was difficult to acquire, difficult to scale and even more difficult to replicate. Portfolio construction, manager research, risk management and asset allocation required experienced professionals, proprietary technology and substantial organisational investment. The complexity of the task created barriers to entry, while the scarcity of expertise provided the foundation for premium fees.

Today, however, those economics are beginning to change. Recent developments in artificial intelligence suggest that we may be approaching a point where replicating sophisticated intellectual processes becomes dramatically cheaper than at any point in history. While the discussion has largely focused on technology companies, the implications for investment management may prove equally significant.

A Lesson From Software

A useful illustration comes from the software industry. Recently, a small open-source initiative released a suite of AI-assisted alternatives designed to emulate several of Adobe's flagship applications. While the software remains far from matching the depth and reliability of Adobe's commercial products, the important point is that a company that once had a moat, requiring vast development teams, significant capital and years of engineering effort to replicate, can now be overcome by a relatively small group using modern AI development tools.

Whether these particular products succeed is ultimately beside the point. What matters is the broader trend they reveal. The cost of reproducing complex expertise is falling rapidly. Activities that previously demanded large teams of specialists are increasingly being supported, accelerated or partially replicated by AI systems.

Whenever the cost of production falls materially, pricing structures eventually come under pressure. History has demonstrated this repeatedly, whether in manufacturing, media, software or financial services. The question for investment management is whether the same process is now beginning to unfold within portfolio construction.

When Expertise Becomes Abundant

This debate is often presented as a question about artificial intelligence. In reality, it is a question about abundance. Historically, constructing a robust investment portfolio required access to research, analytics, optimisation tools and experienced practitioners. Such capabilities were not widely available and therefore carried economic value.

Today, many of the building blocks are becoming increasingly accessible. An investor can already use AI tools to assess asset allocation frameworks, compare ETF exposures, evaluate diversification characteristics, estimate factor sensitivities and model portfolio outcomes. The quality of these outputs will continue to evolve, but the broader direction of travel appears difficult to ignore.

As AI models become more capable, agentic workflows and computational costs continue to decline, the economics of portfolio construction begin to look increasingly similar to the economics of software development.

Processes that were once scarce become readily available, activities that once justified premium pricing become commoditised, the question is not whether this happens overnight. It will not. The question is whether the trend has already begun.

The Challenge to Model Portfolio Economics

This issue is particularly relevant for the model portfolio industry. Over the past decade, model portfolios have experienced remarkable growth, delivering scalable investment solutions and allowing firms to disseminate expertise across increasingly large client bases. Their success has been founded upon the ability to package research, portfolio construction and implementation into a repeatable and highly efficient framework.

However, if the underlying mechanics of portfolio construction become widely accessible through AI-driven tools, the industry's economics may change. This does not imply that model portfolios cease to add value. Rather, it suggests that investors may increasingly question the fee attached to the construction process itself.

If portfolio design, optimisation, risk analysis and implementation become readily available through intelligent systems, then fee structures based primarily upon access to those capabilities may face increasing scrutiny. Such fee pressure would not be unprecedented. Technology routinely compresses margins around activities that become easier to automate. In short, the investment industry should not assume immunity from the same forces.

What AI Cannot Easily Replicate

At this point, it is tempting to draw an overly simplistic conclusion. If AI can replicate portions of portfolio construction, then perhaps investment firms themselves become less relevant. Such a conclusion would, in my view, be mistaken.

The portfolio has never been the complete product. Indeed, sophisticated clients seldom purchase a portfolio for its own sake. What they seek is confidence that capital is being allocated appropriately to achieve specific objectives within clearly understood risk parameters.

This introduces a number of considerations that extend well beyond optimisation algorithms; investment philosophy, governance, behavioural discipline, organisational alignment and accountability matters. The list is long, and it does seem that an investment manager’s work is never done, particularly because of the ability to navigate periods of uncertainty, challenge consensus thinking and maintain conviction when markets become unsettled remains difficult to codify within any model. AI may become exceptionally effective at generating portfolios. It is less clear that it can assume fiduciary responsibility, provide institutional governance or build the trust required to guide clients through complex investment decisions. Those functions remain central to the role of the CIO.

Following the Migration of Value

Perhaps the most useful way to think about AI is not as a destroyer of value, but as a catalyst for its migration. When spreadsheets became commonplace, accountants did not disappear. Their role evolved. When electronic trading transformed markets, investors did not disappear. Competitive advantage moved elsewhere. The same process may now be taking place within portfolio management.

If portfolio construction becomes increasingly commoditised, value is likely to move higher up the decision-making chain. Competitive advantage may increasingly reside in investment philosophy, governance frameworks, behavioural coaching, client-specific customisation, tax intelligence, risk oversight and outcome design.

In other words, the activity of constructing portfolios may become less important than the ability to determine which portfolios should be constructed, why they should be constructed and how investors should behave once they own them.

The CIO Question for the Next Decade

The story emerging from the software industry is not really a story about software, it is a story about what happens when expertise becomes easier to replicate. For investment management, this raises a fundamental strategic question. If the cost of replicating investment expertise approaches zero, what remains scarce?

It is unlikely to be portfolio construction alone. Scarcity may instead be found in judgement, trust, governance and conviction. The real question is not whether artificial intelligence can build portfolios, it is what clients will continue to pay for once everybody can. The firms that answer that question correctly are unlikely to be the ones with the most sophisticated optimisation engine. They are more likely to be the firms that recognise where enduring value truly resides.

Because in an environment where portfolio construction becomes increasingly abundant, competitive advantage will not come from generating portfolios. It will come from generating outcomes.


Irene Bauer