Stocks started the week with promise but ended in despair. Our take on the big reversal

After rising more than 2% through Wednesday, the Nasdaq fell 1.4% Thursday and nearly 3% Friday, ending the week down 2% and October down 5.3% as disappointing tech results and concerns about AI spending drove a rotation into defensive stocks. Alphabet fell about 12% after missing earnings and revenue expectations, while Intel plunged more than 28% on a weak outlook; Microsoft slipped just 0.5% for the week. Healthcare led sectors, with Eli Lilly up more than 11% for the week, while expectations of a possible December Fed rate hike added pressure to technology shares.

Analysis

AI is shifting from a thematic trade to a capital-allocation test. That is a dispersion signal, not yet proof that compute demand is rolling over: the market is penalizing uncertain payback and weak cash conversion more than cloud growth itself. Higher rates amplify this distinction by raising the hurdle for long-dated projects and making capex-heavy balance sheets less forgiving. Second-order, elevated memory costs can squeeze PC demand and Intel while supporting memory suppliers such as SK hynix; that benefit is vulnerable if AI infrastructure orders slow. Grid equipment and industrial-gas suppliers such as GE Vernova and Linde offer exposure to power constraints, but are not insulated from hyperscaler capex cuts.

Over days, crowded semiconductor and mega-cap positions may remain vulnerable to de-risking; a sharp post-earnings drop alone is not a reliable entry signal. Over 1–3 months, watch hyperscaler capex guidance against cloud growth and free-cash-flow conversion, plus evidence that ad weakness is company-specific or broadening. Over 6–18 months, power infrastructure may capture spend even if AI compute economics disappoint, but order conversion and project timing matter. The contrarian risk: investors may extrapolate one quarter’s cash-flow deterioration into a permanent impairment, while underestimating how quickly capex can be reprioritized. Conversely, strong cloud growth is not sufficient if incremental returns stay opaque.

Key Decisions for Investors

• Consider a measured long MSFT / short GOOG pair as a relative-execution expression, not a broad AI bet. MSFT’s reported combination of cloud growth and tighter-than-expected spending contrasts with GOOG’s weaker ad results and cash-flow concerns. Reassess after the next results: exit or reduce if MSFT’s cloud growth decelerates materially, its capex rises without better cash conversion, or GOOG shows ad stabilization and improving free cash flow. Valuation and positioning should be checked before sizing.
• Avoid chasing a broad semiconductor rebound. Keep INTC on a watchlist rather than buying solely after its selloff: verify PC CPU demand, memory-cost effects, and execution in subsequent guidance. SK hynix may benefit from memory demand, but reduce exposure if hyperscaler capex guidance or memory pricing weakens; this is a cyclical demand exposure, not a hedge against an AI-spending slowdown.
• Prefer a staged GEV position over a near-term AI-chip proxy for power-infrastructure exposure, conditional on order conversion and project economics. The thesis is falsified by further deterioration in gas-turbine order growth or evidence that data-center power projects are delayed; do not treat the acquisition announcement as proof of near-term earnings uplift.
• For the next 1–3 months, prioritize cash-flow and capex disclosures over headline AI growth. A renewed rate-hike signal or weaker free-cash-flow conversion across hyperscalers would favor defensives and pressure long-duration tech; a pause in rate expectations alongside improving conversion would weaken that downside case.