S&P 500 Earnings Breadth Masks AI Concentration Risk
All 11 S&P 500 sectors were expected to increase earnings in 2026, even as AI-related profits stood out from the rest of the index. That diffusion call, reported on Oct. 1, sits at the center of the current equity debate. Breadth in profit growth coexists with extreme narrowness in price performance. Reuters
The index had gained about 12% in 2026 as of Sept. 8. It was down about 1% from its record-high close on Aug. 13. The pullback was modest. It left the market still levered to a small cohort of winners. Reuters
Concentration was reaching historic proportions as of May amid the accelerating AI boom, not only in the U.S. By September, AI chipmakers dominated year-to-date share gains among S&P 500 companies. Leadership had not broadened in price terms even as earnings expectations had broadened in principle. Reuters Reuters
The earnings math explains the reliance. Analysts on average expected S&P 500 aggregate second-quarter earnings to jump 39% from a year earlier, with AI-related stocks accounting for much of the growth. Earlier in the cycle, the base was already strong. As of April, 81.3% of S&P 500 companies had posted earnings above analysts' expectations, and overall S&P 500 earnings were expected to be up 16.1%. Reuters Reuters
Volatility around that exposure has been sharp and rotational. Investors showed nervousness over the AI-led rally after industry leaders called for reining in the pace of AI development. The AI scare trade spread from software to wealth management and logistics, with cybersecurity firms among the latest hit. Wall Street then ended higher on a Friday in late September, lifted by Microsoft and other AI-related technology stocks. Reuters Yahoo Finance Reuters
The broader context here is a market pricing two different things at once. Earnings diffusion across all 11 sectors supports index-level multiples. Price concentration in AI chipmakers and megacap platforms concentrates tracking error, factor exposure and drawdown risk.
In my view, the distinction between earnings beta and price beta matters more than usual. A 39% second-quarter aggregate jump driven by a narrow group implies high operating leverage to AI capex, cloud consumption and semiconductor cycle duration. The April beat rate above 80% and the 16.1% expected growth rate suggest fundamentals were not narrow at that point. Prices were.
Looking at what this means for portfolio construction, concentration changes the behavior of passive exposure. Cap-weighted indices carry larger single-name and single-theme deltas. Dispersion between equal-weight and cap-weight outcomes widens. Hedging through sector diversification is less effective when earnings growth is broad but return contribution is not, because the marginal driver of index variance remains the AI complex.
The scare trade rotation is instructive for risk desks. Movement out of software into pressure on wealth management, logistics and cybersecurity points to a de-rating of duration-like equities on AI displacement or capex concerns, rather than a simple semiconductor inventory worry. The late-September rebound led by Microsoft and other AI-related names confirms the two-way nature of that positioning. Crowding amplifies both legs.
For credit and equity analysts, the forward question is earnings persistence versus multiple persistence. If all 11 sectors deliver growth, index earnings can hold even if AI contributors decelerate. Index prices may not hold in the same way, given where year-to-date gains sit. That gap between profit diffusion and return concentration is the fault line to monitor into year-end.
