History Says September Is Brutal for Stocks. Is This the Year AI Stocks Prove It Wrong?

The article argues September 2026 could be less bearish than usual because the 2026 rotation already eased crowded AI positioning (Russell 1000 Value up ~32% vs Growth up ~14% over the past year) while hyperscaler capex estimates rose sharply toward higher levels ($600B in Nov 2025 → $620B in Jan → ~$750B by mid-year). It cites AI demand strength (Google Cloud +63% YoY, and AWS fastest growth in 15 quarters) and expects a near-term catalyst from a packed tech earnings calendar (Broadcom Sept. 2, Oracle Sept. 8, Adobe Sept. 10, Micron later in the month). Main risk is AI monetization lag—AI services ~$25B in direct revenue versus heavy infrastructure build with capex/sales reaching 86% for Oracle and 47% for Microsoft—though the author frames any reckoning as a 2027–2028 issue, not September.

Analysis

Positioning is the key edge: the AI complex has already been partially de-risked, so incremental good news should have more price impact than it did earlier in the year. The cleanest beneficiaries are the names where capex is still being converted into visible order flow or backlog—AVGO, NVDA, MU—and the platforms that can keep spending without an immediate balance-sheet issue—MSFT, AMZN, GOOGL. The more fragile part of the trade is software with slower monetization proof; when spend outruns revenue, margin math becomes the first thing the market discounts, and ADBE is the clearest candidate for that skepticism.

The catalyst window is days to weeks around earnings, but the real reversal risk is 1-3 months if one or two large platforms stop raising capex or guide only in-line instead of up. That would shift the market from 'AI infrastructure is scarce' to 'AI spend is peaking,' which compresses multiples fast even if revenue still looks fine. The structural risk is later: if monetization lags for another 12-18 months, free-cash-flow conversion will matter more than headline growth, and the market will punish the most capex-intensive winners first. The bullish thesis is falsified if a major hyperscaler signals capex flattening, AI backlog growth decelerates, or near-term revenue guidance fails to keep pace with spending revisions.

Key Decisions for Investors

• Buy AVGO Nov/Dec call spreads into earnings; best reward/risk if guidance re-accelerates and the market rewards visible AI orders rather than broad sentiment.
• Own MSFT and AMZN on any post-earnings pullback; these are the highest-quality 'can spend and monetize later' expressions, with lower downside than the pure semiconductor names.
• Pair long AVGO / short ADBE for 4-8 weeks: long the infrastructure name with direct AI revenue exposure, short the software name most vulnerable to 'AI monetization lags spend' re-rating.
• Avoid chasing NVDA pre-print; use it as a confirmation trade only if the broader AI capex commentary stays upward and the stock holds gains after results.
• Set an alert for any hyperscaler capex guide revision down; that is the cleanest falsifier and would justify reducing AI exposure across semis and cloud.