U.S. stock futures rise after AI rout batters Wall St; payrolls data in focus

Thursday’s AI-led selloff drove the Nasdaq Composite down 5.5%, the S&P 500 down 4.4% and the Dow down 3.4%; the Philadelphia Semiconductor Index fell 12.25%, with Nvidia off nearly 10% and AMD down 17%. U.S. index futures were mixed Thursday evening: S&P 500 futures rose 0.26% and Nasdaq 100 futures gained 0.83%, while Dow futures fell 0.15%. Investors were weighing political and regulatory concerns about AI and data-center expansion, alongside September nonfarm payrolls due Friday and the risk that resilient jobs data could give the Fed more room to raise rates after its September 25-basis-point hike.

Analysis

The selloff looks less like evidence that AI demand has vanished than a repricing of the time, certainty, and permitted scale of returns on data-center capital. That distinction matters: chip suppliers can be hit before cloud platforms materially change spending, while power, grid, cooling, and construction projects face a slower but potentially broader permitting overhang. Political rhetoric is not enacted policy; watch for actual permitting limits, utility interconnection delays, and customer capex revisions before underwriting a structural demand reset.

Near term, payrolls and Treasury yields may dominate the AI narrative. A resilient report could pressure long-duration growth again; a weak report could stabilize multiples but raise concerns about enterprise demand. Defensive-sector outperformance is not automatically durable: utilities may benefit from load growth, yet higher yields and delayed data-center connections complicate that hedge.

Over 1–3 months, the key test is whether hyperscalers maintain capex plans while showing better monetization. If spending holds but returns remain unclear, supplier earnings expectations and AI multiples can fall together. Over 6–18 months, permitting and grid capacity—not just chip availability—could become binding constraints. Contrarian point: the market may be extrapolating a political headline into immediate project cancellations, while the more durable risk is slower returns on capital. Anthropic’s reported IPO uncertainty is a sentiment signal, not proof of weaker demand across the sector.

Key Decisions for Investors

• Avoid buying the first rebound in high-beta AI. Wait through payrolls and for stabilization in Treasury yields and semiconductor price action; reassess if hyperscaler capex guidance or supplier order commentary is cut.
• Consider a staged relative-value position: long AAPL against short AMD, sized conservatively, to express lower direct data-center-capex exposure versus higher semiconductor beta. Falsify if Apple’s relative resilience reverses or AMD reports materially stronger demand/guidance; the one-day move alone is not a valuation signal.
• Do not treat utilities or staples as a blanket hedge. Monitor real yields, utility interconnection queues, and enacted state or federal permitting rules; actual project delays would strengthen the case for underweighting data-center-linked infrastructure exposure.
• For the next 1–3 months, track hyperscaler capex guidance alongside AI-related revenue or productivity disclosures. Sustained spending without clearer monetization supports continued multiple risk; credible monetization and no material policy constraints would argue against extending the selloff.