AI capex scrutiny hits hyperscalersThere’s no single clean thread, but the dominant one is clear: the AI capex bill is c…

AI capex scrutiny hits hyperscalers

There’s no single clean thread, but the dominant one is clear: the AI capex bill is coming due and the market is demanding clearer ROI. Nasdaq futures are down ~2% into the open after $Amazon(AMZN.US) and $Intel(INTC.US) compounded $Alphabet(GOOGL.US)’s earlier miss, accelerating a rotation out of the AI complex into energy, healthcare, and other cash generators. Oil is bid on geopolitics and refining strength. Bonds are rallying on the risk-off bid. Month-end flows and the Employment Cost Index at 8:30 a.m. ET are the near-term macro overlays. Don’t force a broader thesis beyond the AI spend scrutiny.

AI Capex Reckoning: $Amazon(AMZN.US), $Intel(INTC.US), $Alphabet(GOOGL.US) Drive the Selloff

$Amazon(AMZN.US) missed on both revenue and EPS, posted a trailing-twelve-month free cash flow outflow of $42.4B (driven by AI-related property and equipment spend), and held 2026 cash capex guidance at ~$220B. AWS grew 39% to a $184B annualized run rate (fastest in nearly five years) with management projecting >40% growth in 2027 and 2028 and a path to a trillion-dollar AWS business, but the cash burn and e-commerce softness (online sales only +2% after Prime Day timing) dominated. Shares were down ~7% premarket. Analyst and market commentary frames this as the cost of capacity that is still constrained, not a demand problem—yet the market is no longer willing to ignore the FCF hit or the debt funding component.

$Intel(INTC.US) beat Q3 (adj. EPS $0.52 vs. ~$0.40–0.41 expected; revenue $16.51B roughly in line, +21% y/y) with data center/AI revenue +78% to $7.3B, but guided Q4 revenue of only $15–16B (well below ~$17.1B consensus) and adj. EPS of $0.25. CFO David Zinsner pointed to PC market softness lasting at least through 1H 2027; an Israel fab incident added ~$300M of the revenue shortfall and ~200 bps of gross margin pressure. Shares plunged ~23–25% premarket. The turnaround narrative took a hit as traditional PC exposure still funds the foundry bet.

$Alphabet(GOOGL.US) (prior session) delivered its first top- and bottom-line miss in nearly four years, negative FCF of ~$9.1B, Q3 capex of $51.4B, and maintained 2026 capex of $195–205B while flagging a significant increase in 2027. Google Cloud surged 103% to $30.7B, but search/ad softness and the spend trajectory triggered the worst day in years (roughly –12–13%, erasing hundreds of billions in market value). The common thread across the three: strong cloud/AI demand is no longer sufficient; investors want evidence that the massive outlays convert into durable margins and free cash flow rather than further balance-sheet strain.

FinTwit and sell-side commentary is bifurcated. Some see a conversion phase where only those proving near-term revenue pull-through get a pass; others view the pullback as an opportunity if AWS-style growth persists and capacity remains the bottleneck. Downstream beneficiaries (power, networking, memory, optics) remain in focus on any sustained reacceleration, but the near-term tape is punishing the hyperscalers themselves.

Premarket Movers and Rotation

$Intel(INTC.US) –23–25%, $Amazon(AMZN.US) –6–7%, $Reddit(RDDT.US) –22%, $Roblox(RBLX.US) –16–17%, $Twilio(TWLO.US) –17–19%, $Sandisk(SNDK.US) weaker on peak-pricing concerns despite a beat. Nasdaq-100 futures –2%, S&P –1%, Dow roughly flat. Breadth on Thursday was constructive outside tech (Russell 2000 +1%, healthcare and real estate leading), so the open is a continued rotation rather than a full risk-off flush. Small-cap premarket gainers are mostly low-quality/volume names; focus remains on the large-cap tech reaction and the energy/pharma prints.

Macro and Agency Backdrop

Fed held the target range at 3.75–4% on Oct. 28 (10–2 vote; Cook and Hammack preferred a hike). Statement noted strong growth, resilient spending, robust investment, and elevated inflation; more data needed before further moves. Recent GDP (+3.4% annualized) and PCE (headline 0.5% m/m, y/y elevated) keep December hike odds alive in the 75–80% range per market pricing. BOJ held at 1.25% and raised inflation forecasts, declaring underlying inflation consistent with target. Oil is firm (Brent in the low $110s, WTI near $103–104) on Strait of Hormuz noise and refining disruptions from conflicts; energy majors are the clear beneficiary. Treasuries rallied on the equity risk-off. Midterms next week remain a volatility overlay.

Friday Earnings Offsets: Energy and Pharma

$ExxonMobil(XOM.US) and $Chevron(CVX.US) both beat on refining strength. Adjusted EPS: $ExxonMobil(XOM.US) $4.23 (vs. ~$4.04–4.12), $Chevron(CVX.US) $5.28 (vs. ~$4.83–4.89). Downstream profits hit records or multi-year highs as wars and refinery outages lifted diesel/gasoline margins; both are using the cash to pay down debt rather than accelerate buybacks. Shares modestly higher premarket. This is a clean geopolitical-plus-margin story that can support the rotation trade. $Abbvie(ABBV.US), $Regeneron Pharma(REGN.US), $General Dynamics(GD.US) and others also reporting before the open; early indications on some have been constructive. Oil and refining names remain the highest-conviction relative strength into the print.

Other Catalysts

Binance suffered a ~$6.3B hack (largest on record; North Korean actors linked by multiple firms). Roughly $3B frozen on-chain; the exchange says it will cover losses via SAFU and company funds, deposits are open, withdrawals remain paused pending security work. $Grayscale Bitcoin Mini Trust ETF(BTC.US) under pressure near $78k. Not a systemic equity event but adds to month-end risk-off noise.

Bottom line for swing traders: The AI complex is being stress-tested on cash conversion. Favor names with visible cash generation or clear demand bottlenecks outside pure hyperscaler spend. Energy ($ExxonMobil(XOM.US) $Chevron(CVX.US)) and selective defensive/healthcare names offer relative strength on volume into the open. Watch the Employment Cost Index reaction and whether the rotation holds after the open. Size smaller into the tech reaction; the drawdown is cleaning froth but not yet a full regime change.

NFA / DYOR — This is not financial advice. Do your own research. Markets can reverse quickly on catalysts.