Stocks saw new highs and big declines: How the volatile AI trade moved last week's market

The S&P 500 rose 1.2% for the week, closing above 7,800 for the first time, while the Nasdaq gained 0.6%, despite a sharp Thursday sell-off in AI stocks and a Friday rebound. Reports on OpenAI revenue expectations drove the volatility; SpaceX reportedly is seeking $40 billion for Nvidia chips and AI compute, and announced a spectrum acquisition that pressured telecom stocks. Starbucks customer data showed improving spending intentions, though reported exploration of a Chipotle acquisition added uncertainty.

Analysis

The AI read-through is less about one customer’s reported spending plan than whether compute demand can be financed and monetized. A chip order is not proof of attractive returns: utilization, lease pricing, power availability, and customer concentration determine whether spending sustains or merely pulls forward demand. That distinction matters for NVDA and the wider supplier complex; AVGO may also trade with the theme despite different exposure, so Thursday-style indiscriminate selling could create relative-value opportunities rather than invalidate AI investment broadly. Verify financing terms and realized compute revenue before treating reported plans as durable demand.

SpaceX’s spectrum ambitions create a long-dated competitive overhang for U.S. carriers, not an immediate earnings shock. Spectrum access alone does not establish coverage economics, handset performance, or a rapid buildout. Regulatory approvals and deployment milestones are key; partnerships with carriers remain a possible outcome, which could blunt a simple disruption thesis.

For SBUX, customer-intent survey gains are encouraging but are not transaction data. The more consequential near-term risk is management attention and capital allocation if deal exploration advances: buying CMG could dilute the standalone turnaround narrative and introduce integration risk. In the next 1–3 months, track comparable sales, traffic, and margin commentary, alongside any confirmation or denial of deal talks. Over 6–18 months, sustained operating improvement—not survey sentiment—would support a rerating. The contrarian risk is that investors overread a favorable survey while underpricing deal uncertainty.

Key Decisions for Investors

• Keep NVDA exposure, but add only on volatility and after checking evidence of customer utilization and financing; reduce the thesis if spending plans fail to convert into recurring compute revenue or suppliers guide to weaker demand.
• Do not short U.S. telecom stocks solely on the spectrum headline. Reassess only after regulatory progress and credible deployment milestones; partnership announcements would weaken the disruption case.
• Treat SBUX as a conditional watch, not a survey-driven buy. Consider adding only if standalone comparable-sales and traffic trends improve and deal speculation recedes; worsening operating metrics or confirmed acquisition pursuit would invalidate that setup.
• Avoid broad AI-trade extrapolation from a single disputed customer-growth report. Monitor supplier guidance and realized capex; a deterioration across both would be a stronger bearish signal than sentiment-driven daily reversals.