Good News for S&P 500 Earnings: 86% of Companies Beat Expectations in 2026

Bloomberg analysis found that 86% of S&P 500 companies have exceeded analyst earnings expectations in 2026, suggesting AI-related capital spending is supporting profits beyond mega-cap technology. The article cautions that the index remains concentrated in AI-linked stocks and presents diversification into small-cap value as a hedge; Vanguard forecasts U.S. small-cap and value equities will outperform large-cap growth over the next 10 years. Vanguard's S&P 500 ETF has returned about 15% annualized over 16 years, while the iShares Russell 2000 Value ETF returned roughly 9.9% annually over the past decade and has only 7.1% allocated to technology.

Analysis

The unusually high beat rate is more informative about estimate-setting behavior than durable breadth unless accompanied by rising full-year EPS and free-cash-flow guidance. The key near-term question is whether non-tech beats reflect AI-linked demand and capital spending or true end-demand acceleration; the former supports NVDA and data-center supply chains but leaves the index exposed to a synchronized hyperscaler capex reset. Watch forward S&P 500 EPS revisions excluding the largest technology cohort over the next 1-3 months: breadth is credible only if those revisions remain positive.

A rotation into small-cap value is not a clean anti-AI hedge. IWN carries materially higher refinancing, wage, and domestic-cycle sensitivity than the S&P 500; its relative performance historically needs falling real yields, easier credit, or accelerating nominal growth. If long-end yields remain elevated, small-cap balance-sheet fragility can overwhelm valuation support, while profitable large-cap platforms retain financing and margin advantages.

The more actionable divergence is between AI infrastructure beneficiaries with visible order coverage and companies whose earnings are only indirectly supported by data-center construction. NVDA remains vulnerable to a multiple reset if customer capex growth decelerates even while reported earnings beat; upside requires upward revisions to 2027 demand rather than another backward-looking beat. Consensus appears too willing to treat broad earnings strength as evidence that AI productivity has already diffused through the economy, when much of the transmission may still be capex-driven and margin-dilutive for customers.

Key Decisions for Investors

• Maintain a tactical long NVDA only while hyperscaler capex guidance and NVDA forward revenue estimates are rising; use a 10-15% trailing risk limit. The 1-3 month upside case is continued estimate revision, while the principal downside is a capex-growth deceleration that compresses the multiple despite an earnings beat.
• Express a conditional breadth rotation via long IWN / short QQQ only after the Russell 2000 relative trend improves and 10-year real yields decline; target a 5-8% relative move over 3-6 months. Exit if real yields rise materially or high-yield spreads widen, which would expose small-cap refinancing risk.
• Avoid treating IWN as a defensive hedge against AI. For an immediate de-risking sleeve, reduce concentrated semiconductor exposure rather than substituting into leveraged small caps; reassess after the next earnings cycle establishes whether ex-megacap EPS guidance is being revised upward.
• Set an alert on hyperscaler capex guidance and semiconductor equipment order commentary during the next reporting season. A broad reduction in capex growth expectations is the falsifier for AI-infrastructure longs and would favor accelerating the NVDA underweight versus a market-neutral technology basket.