What could reduce the market's increasing dependence on AI leaders?

It’s no secret the equity market continues to be led by a small group of large companies: This year, it has been more concentrated among the top 10 components than at any time in the past 100 years (Exhibit 1).

As mega-cap growth and AI-linked stocks have widened their lead over most S&P 500 components in recent weeks, the broadening that we expected in 2026 has yet to materialize in a durable way. Investors have repeatedly expected market leadership to broaden, only to see AI-related leadership reasserted, supported by strong earnings growth.

Is the high market concentration really a problem? It hasn’t been. As we’ve written many times, leadership cycles in the past started with a small group of companies tied to a major economic transition (railroads, internet platforms, and more). It’s historically common.

Here’s the potential rub: What happens if AI companies become less dominant contributors to future returns?

As of early October, a string of positive earnings surprises from a handful of companies have supported the market. The opposite could also become true at some point: Earnings that surprise less to the upside could negatively affect the S&P 500.

Also, valuation pressure from higher rates (absent an improving economy) or earnings growth in sectors not as closely tied to AI could reduce the market’s dependence on the mega-cap AI leaders over time.

What could reduce the market’s dependence on AI leaders?

In no way are we arguing that investors should abandon their core equity allocations in a concentrated market.

However, they may need to think more deliberately about diversification. We think a broadening of the historically concentrated market—something we’ve anticipated this year—may still happen for three key reasons:

1. Strong earnings growth could eventually expand beyond AI-focused technology.
2. Higher rates absent an improving economy could create valuation pressure for the market’s most richly valued companies.
3. AI investments could start to shift from infrastructure providers to AI adopters, which include more small cap, mid cap, and international value plays.

What actions should investors take?

If investors haven’t already, it’s likely time to review portfolio diversification with their advisors. Market concentration is increasingly thematic, not just equity-specific. Capital spending, earnings growth, and performance across multiple asset classes are becoming more closely tied to the AI theme. These dynamics extend across public investment-grade fixed income, private equity, credit markets, and real estate.

Even a portfolio that appears well diversified by asset class may still have meaningful exposure to the same underlying AI-related theme.

For extreme portfolios, advisors may want to:

• Trim outsized client positions back toward intended weights.
• Complement market-cap-weighted exposure with more balanced approaches.
• Consider active strategies where dispersion is rising.
• Look at allocation balance by sector, style, and region, not just by asset class.
• Review aggregate AI exposure across client portfolios.

Within equities, advisors might consider whether small and mid cap allocations remain appropriately represented in client portfolios. Many of those components are underrepresented in mega-cap-heavy benchmarks, including regional banks, industrial suppliers, health care services, consumer cyclicals, and domestically focused businesses.

Many of these stocks may also be resilient valuation plays. Dozens of them enter the fourth quarter with lower price-to-earnings ratios, greater sensitivity to financing conditions, and more room for operating leverage if demand remains resilient.

Conclusion

Today's market concentration reflects the extraordinary business success of a small group of companies. Their leadership has been supported by strong earnings growth and continued customer spending.

Could the market broaden from here? It might.

Even if it doesn’t, investors may be underestimating the need for diversification.