Should You Buy SpaceX Stock in September?

SpaceX shares have fallen 34% from a $225 peak to $147.95, despite reported Q2 2026 revenue growth of 92% year over year to $7.8 billion, led by AI revenue growth of 247% to $2.56 billion. The company is valued at roughly $2 trillion, equating to an 86.9x trailing price-to-sales ratio on $23 billion of last-twelve-month revenue, versus 6.2x for the Nasdaq-100. While Wall Street expects $105 billion of 2027 revenue and management targets up to $1 trillion of annual revenue by 2030, the article views near-term upside as constrained by valuation and dependent on highly ambitious execution.

Analysis

SPCX is being priced as a conglomerate growth option rather than a launch-services company, leaving the equity unusually exposed to execution slippage across three capital-intensive businesses simultaneously. The key valuation vulnerability is not merely revenue growth: investors need to underwrite sustained gross-margin expansion despite satellite replenishment, ground-network buildout, GPU depreciation, and potentially large orbital-infrastructure capex. A 2027 revenue miss or a lower-than-expected AI segment margin would likely compress the multiple sharply because little downside protection exists from mature cash-flow valuation.

The claimed AI opportunity has meaningful second-order implications for NVDA and GOOG, but neither should be treated as a direct read-through. NVDA benefits near term if SPCX is procuring conventional accelerated-compute systems; longer term, a proprietary space-based compute architecture could reduce the value of centralized terrestrial GPU clusters only if launch cadence, radiation-hardening, bandwidth economics, and in-orbit servicing prove viable. GOOG's exposure is two-sided: it could be a cloud/AI-capacity customer, but Starlink's enterprise distribution creates a credible edge-cloud and connectivity competitor in underserved geographies.

Over the next days to three months, SPCX will trade on momentum and narrative validation, not fundamental revisions. The investable catalyst is independently auditable disclosure around AI backlog, contracted versus internal compute revenue, segment gross margin, capex, and customer concentration; absent that, the current drawdown is not automatically a buying opportunity. Over 6-18 months, the decisive issue is whether connectivity cash generation can fund AI and orbital capex without recurring equity issuance or leverage expansion.

Contrarian view: consensus may be treating the AI revenue run-rate target as evidence of durable external demand, while spare-capacity rentals can be low-quality, concentrated, and cyclical. Conversely, a short is dangerous if verified backlog converts faster than expected and management demonstrates that satellite connectivity materially lowers data-delivery cost versus terrestrial alternatives. Thesis falsification for a cautious stance would be two consecutive quarters of externally verified AI revenue growth with stable or improving consolidated gross margin, alongside capex intensity below market expectations.

Key Decisions for Investors

• Do not initiate a directional SPCX long solely on the pullback. Place an alert for the next earnings release: consider a starter long only if disclosed external AI revenue/backlog and segment gross margin validate the growth narrative; size small because multiple compression remains the dominant 1-3 month risk.
• For a tactical bearish expression, prefer a defined-risk SPCX put spread 3-6 months out rather than an outright short, contingent on liquid listed options. Target a valuation reset following weak AI monetization or elevated capex; exit if management validates durable external backlog and margin progression.
• Maintain NVDA exposure independently of SPCX, but do not add on the premise that orbital AI infrastructure is an incremental near-term demand driver. Monitor disclosed accelerator procurement and depreciation/capex at SPCX; confirmed large third-party GPU purchases would be a 6-12 month NVDA demand tailwind.
• Watch GOOG for enterprise-network and cloud implications rather than trade it directionally on this item. A disclosed Starlink-plus-AI enterprise offering with major customer wins would warrant reassessing long-term cloud-network competitive risk, but current evidence is insufficient for a position change.