AI Stocks Drive Nearly All of S&P 500’s Gains, Data Reveals

The S&P 500 has reached fresh record highs in 2026, but the rally is narrowly concentrated in AI-linked stocks, which now represent a record 45% of the index’s market cap. The US 500 Excluding Artificial Intelligence Enablers Price Return Index (SPXXAI) is down 1.84% since its February launch and only about 5.07% off its lows, underscoring how little the broader market has participated. Goldman’s prior work showed the headline S&P 500 returned 76% over three years versus 32% for the ex-AI version, highlighting rising reliance on the AI trade amid geopolitical uncertainty.

Analysis

This is a narrow-leadership tape, not a broad risk-on regime. When a small cluster of mega-cap AI names reaches this kind of index weight, the marginal buyer of the benchmark is effectively underwriting one thematic earnings cycle; that makes the whole market more vulnerable to any disappointment in capex, cloud monetization, or model-competition optics over the next 1-2 quarters. The immediate second-order effect is crowding: passive inflows and systematic trend-following will continue to concentrate into the same winners until breadth deterioration triggers de-risking.

The biggest hidden risk is that AI beneficiaries themselves may be front-loading demand. If hyperscaler capex growth moderates even modestly, the rerating can compress quickly because the market is already paying for multi-year durability, not just current profits. That creates an asymmetric setup: downside is driven by multiple compression across the most index-heavy names, while upside requires an increasingly heroic pace of earnings beats to justify today’s concentration.

For the rest of the market, the implication is a funding and relative-performance squeeze rather than a macro collapse. Industrials, small caps, and financials are being implicitly starved of incremental attention; if AI leadership pauses, these segments can outperform mechanically on mean reversion without needing a strong economic catalyst. Geopolitical uncertainty adds a separate fragility: if risk premia rise, investors will likely sell the less-liquid laggards first, but the market’s actual beta resides in the AI complex, so that is where the next drawdown would likely be born.

The contrarian view is that this is not necessarily a bubble top so much as a legitimate earnings concentration trade. The consensus mistake is assuming breadth must improve for the bull market to continue; in reality, index-level gains can persist for months as long as AI cash flow, capex, and guidance keep beating. But the trade has become increasingly binary: it is now more sensitive to one or two reporting cycles than to the broader macro narrative.

Key Decisions for Investors

• Trim equal-weight exposure to the S&P 500 and add a tactical underweight to the AI-heavy megacap basket via QQQ vs. RSP over the next 4-8 weeks; the risk/reward favors relative short breadth if leadership narrows further.
• Buy downside hedges on the AI complex through 1-2 quarter-dated put spreads on QQQ or SMH ahead of the next major earnings season; limited premium outlay against a potentially sharp multiple reset if capex guidance softens.
• Go long a basket of lagging breadth beneficiaries — IWM, XLI, and XLF — versus short a concentrated AI proxy; the setup is a mean-reversion trade with better upside if market leadership pauses even briefly.
• For hedging existing long books, use call-spread financed collars on AI winners rather than index puts; if the concentration trade keeps working, you retain upside while capping the most crowded tail risk.