S&P 500 Profits Are on Track for a Third Straight Quarter of 25%+ Growth. The Index Hasn't Kept Up.

FactSet projects S&P 500 Q3 earnings growth of 28.9%, which would mark a third consecutive quarter above 25%, while next-12-month earnings estimates have risen 8.8% since June 30 versus only a 1.8% increase in the index. The forward P/E declined from 20.4x to 19.1x (about 19.4x after Monday's 1.5% rally), below the five-year average of 19.8x, as earnings expectations—not price appreciation—have driven the market near record highs. Growth is concentrated in energy (+110%), technology (+63%), chipmakers (+126%), and Meta-supported communication services, creating downside risk if AI spending or oil prices weaken.

Analysis

The apparent broad-index derating is less durable than it looks because incremental earnings revisions are highly concentrated in semis, energy, and META. That concentration raises the index's effective cyclical beta: a modest miss in hyperscaler capex, memory/foundry utilization, or crude prices can remove a disproportionate share of expected EPS growth while leaving the rest of the market still valued near long-run averages. The key near-term question is not whether estimates rose, but whether the revisions survive earnings calls and calendar-2027 guidance; estimate dispersion is likely to widen sharply over the next 1-3 months.

NVDA and the semiconductor complex retain operating leverage to AI infrastructure spend, but the market should distinguish revenue growth from the return on customers' AI capex. If cloud operators signal slower deployment or longer payback periods, equipment and upstream semiconductor names should de-rate before reported chip earnings weaken. Conversely, sustained capex guidance would support a second leg higher in suppliers with less crowded positioning than NVDA, including ASML and AMAT, though these are more exposed to order-cycle volatility.

META is unusually important to index-level upside because its earnings sensitivity combines ad pricing, engagement, and AI-driven efficiency. The contrarian risk is that consensus treats AI spending as margin-accretive by default: incremental depreciation and data-center commitments could cap operating-margin upside even if advertising demand remains healthy. Energy's contribution is the least reliable component of the revision story; oil-price-driven EPS gains can reverse quickly if global demand indicators soften or supply disruptions ease.

The actionable conclusion is not a blanket long VOO. The index can grind higher if revisions hold, but its narrow earnings engine makes downside asymmetric around guidance season: an estimate reset would likely produce multiple compression rather than merely slower EPS growth.

Key Decisions for Investors

• Maintain a modest long SPY/VOO only as a 1-3 month earnings-revision trade; add on confirmation that aggregate forward EPS continues rising through the first two weeks of reporting season. Reduce if forward S&P 500 EPS revisions turn negative for two consecutive weekly FactSet updates or if the forward multiple re-expands above 21x without further estimate growth.
• Prefer a quality AI-capex pair: long AMAT or ASML versus short SMH in equal beta-adjusted dollars over 3-6 months. The thesis is that equipment names offer catch-up order optionality if capex persists, while the short hedges crowded semiconductor-beta exposure; exit if major foundry/customer order commentary weakens or the pair underperforms by 10%.
• Treat META as an earnings-event long only with a defined hedge: long META versus short XLC through the next results, sized to limit a 15% single-name drawdown. Upside requires ad-revenue durability plus operating-margin preservation; falsification is AI-capex guidance that exceeds revenue growth or a material deceleration in ad impressions/pricing.
• Avoid adding broad energy beta after the earnings revision impulse. For existing XLE exposure, use a 1-3 month trailing stop tied to crude: reduce if Brent falls below its 50-day moving average and refinery margins weaken, as energy estimate cuts would remove an important index-level support.
• Watch hyperscaler capex guidance and semiconductor inventory/order metrics rather than headline index EPS. A sequential reduction in AI infrastructure budgets is the highest-probability catalyst for a 5-10% semiconductor de-rating and would argue for rotating index exposure toward defensives such as XLV and XLP.