The Diversification Mirage Inside Today’s Indexes

Why It Matters

Concentration risk undermines the protective premise of passive investing, exposing portfolios to sector‑specific shocks. Recognizing and correcting this bias is critical for advisors and institutional investors seeking resilient, long‑term performance.

Key Takeaways

• AI mega‑caps dominate major US indexes
• Passive funds mask underlying concentration risks
• Pave's engine evaluates 88 risk and 55 return factors
• Commoditization could compress AI pricing power
• Intentional diversification needed beyond index exposure

Pulse Analysis

The rise of artificial‑intelligence megacap stocks has reshaped the composition of flagship benchmarks, pushing weightings well beyond traditional sector limits. While passive vehicles still offer low‑cost market exposure, their underlying holdings now mirror a narrow technological narrative, eroding the diversification cushion investors assume. This concentration amplifies vulnerability to AI‑related earnings volatility and to a potential future where AI services become commoditized, squeezing margins and unsettling price expectations.

To counteract the illusion of safety, Pave has built a data‑intensive platform that scores roughly 9,000 securities across 88 risk and 55 return factors. By mapping sensitivities to macro variables, capital flows, and geopolitical shifts, the engine surfaces hidden overlaps among ETFs and highlights where capital is truly concentrated. Advisors can then re‑weight portfolios, integrate alternative factors, or impose client‑specific constraints without sacrificing benchmark alignment, effectively turning passive exposure into a more nuanced, risk‑aware strategy.

The broader market implication is clear: as AI infrastructure spending matures, the once‑high‑growth narrative may give way to marginal‑cost pricing, pressuring valuations across the board. Investors who rely solely on index tracking risk missing early signals of this transition. A deliberate, factor‑driven approach—leveraging tools once reserved for large institutions—offers a path to preserve genuine diversification, manage sector‑specific tail risk, and maintain performance resilience amid evolving technology cycles.

The diversification mirage inside today’s indexes

By Manal Ali · Feb 17, 2026

Market concentration is not a new phenomenon. Every long equity cycle eventually narrows around its most dominant companies and themes. What is new is how deeply it now sits inside the index itself.

Peter Corey, chief market strategist at [Pave], has spent more than four decades building and managing portfolios across multiple market regimes, from running derivatives desks since the 1980s, to launching macro‑driven [hedge fund] strategies and advising institutional investors. That perspective informs his view of today’s environment. Indexes remain efficient vehicles for market exposure, he argues, but their internal composition has shifted in ways that fundamentally change how [diversification] works.

At the center of the issue is the growing dominance of a narrow group of mega‑cap companies whose valuations and earnings expectations are tied to artificial intelligence and related infrastructure. Their weight within major benchmarks has reached levels where index exposure increasingly reflects a concentrated view on one technological and capital‑spending cycle. Investors may still hold hundreds of securities through passive funds, yet a significant share of performance and vulnerability rests on the same drivers.

“Passive has worked extraordinarily well,” Corey says. “But you have to recognise what you actually own. When concentration reaches this level, the index is no longer delivering the diversification people assume it is.”

A familiar cycle, with different consequences

Artificial intelligence is following that path. Investment in data centers, cloud infrastructure and advanced computing has channelled capital toward a handful of dominant firms. As those firms grow, they increasingly compete across overlapping business lines.

“Now they’re all converging into one,” Corey says. “There are going to be winners and losers. They’re not all going to be as great in the new business as they were in their legacy business.”

That convergence introduces a different kind of risk. If AI capabilities become widely accessible and difficult to differentiate, pricing power could compress over time. Corey draws a parallel to the early internet era, when once‑distinct platforms gradually became commoditized.

“If AI ultimately becomes a commoditized product -- even an extraordinary one -- pricing tends to move toward marginal cost,” he says. “Data centers have high fixed costs, but once they’re built it doesn’t take much to keep them running. At some point investors will start asking whether the projected returns justify the capital being deployed.”

That does not mean the technology will fail to deliver productivity gains. Corey expects AI to drive substantial efficiency and profitability across the global economy over time. The challenge lies in the transition period between heavy investment and broad monetization. When expectations are embedded at scale, markets can react sharply if returns take longer to materialize than anticipated.

“Investors get impatient,” he says. “You can have tremendous long‑term potential, but if expectations get too far ahead of reality, you can get a lightning bolt in the market.”

Reintroducing intention into diversification

For Corey, the implications are less about abandoning passive strategies than about redefining diversification. When the largest index constituents are driven by similar macro forces and capital cycles, traditional approaches may not provide the balance investors assume.

At [Pave], the company’s focus has been on giving advisors and investors the tools to see those exposures clearly and respond more deliberately. The firm’s portfolio engine evaluates thousands of securities globally each week, ranking roughly 9,000 stocks across dozens of return and risk factors that extend well beyond traditional metrics such as size, leverage and price momentum.

The model incorporates 88 risk factors and 55 return factors, measuring sensitivities across macro variables, capital flows and global market dynamics. When tariffs or geopolitical developments redirect investment between regions, for example, the system identifies which companies are most exposed to those shifts and adjusts rankings accordingly.

“We wanted a system that adapts as capital actually starts flowing, not just when the idea appears,” Corey says. “It allows us to identify where capital is actually moving, not just where people think it should move.”

The goal is not simply to express macro views but to wait for confirmation before positioning portfolios. Corey notes that traditional strategies may hold those positions regardless of performance, leaving portfolios exposed if the thesis takes time to materialize or proves incorrect.

“You can have the right idea and still underperform for a long time,” he says.

Technology once reserved for institutions

Much of this analytical capability was historically limited to large hedge funds and institutional managers with the resources to maintain dedicated data‑science teams and proprietary models. The objective at [Pave] has been to make similar tools accessible to independent advisors and their clients.

The platform’s optimization engine constructs portfolios around individual constraints and preferences while evaluating risk holistically across existing holdings. Rather than treating each ETF or strategy in isolation, the system looks through underlying securities to measure true exposure.

“A lot of investors think they’re diversified because they own multiple ETFs,” Corey says. “When you look under the hood, they often have a significant concentration in the same names across all of them.”

That transparency allows portfolios to remain aligned with benchmark exposure while adjusting weightings, incorporating alternative factors or accommodating individual constraints. The aim is not to replace passive investing but to restore genuine diversification around it.

“You have to increase your level of sophistication and analytical capability,” Corey says. “The easy way out is to hide in the index. But when the structure of the market changes, that approach becomes more vulnerable.”

Passive exposure will continue to play a central role in portfolios, he adds. What has changed is the assumption that it alone provides sufficient balance.

“Diversification doesn’t come automatically with the index anymore,” Corey says. “You have to build it deliberately.”

This article has been produced in partnership with Pave.