Where Will SpaceX Stock Be in 5 Years?

Goldman Sachs projects SpaceX AI revenue could increase 100x to $322 billion by 2030, supported by neocloud compute demand, a reported $30 billion Alphabet capacity agreement through 2029, and approximately $15 billion of annual Anthropic spending. Starlink, which accounts for about 61% of revenue and is described as the company’s only profitable business, has 12 million subscribers, while U.S. customers are estimated to rise fivefold to 15 million by 2030. However, capex rose 308% to $26.5 billion in the first half of the year, with more than 82% directed to AI, creating profitability, financing-cost, and stock-volatility risks.

Analysis

The investable implication is less about SPCX than whether incremental neocloud demand is real, financed, and contracted. If credible, the first beneficiaries are AI infrastructure vendors with near-term revenue recognition—NVDA, AVGO, VRT and ETN—while the ultimate compute lessor bears utilization, power, depreciation, and customer-credit risk. A capacity provider can report rapidly rising bookings while still destroying equity value if GPU useful lives shorten faster than contract duration or financing costs exceed contracted gross margin.

The article's commercial assertions require unusually high skepticism: several cannot be reconciled with readily established public-company disclosures or industry ownership structures. Until counterparty filings, debt documents, or audited revenue disclosures validate these commitments, SPCX should be treated as non-actionable promotional flow rather than a fundamental catalyst; an alleged customer commitment is not equivalent to take-or-pay cash flow.

Over 1-3 months, the relevant read-through is whether hyperscalers sustain capex guidance despite rising rates and whether GPU lead times remain tight. Over 6-18 months, excess neocloud capacity is the key tail risk: if enterprise AI monetization lags, smaller operators will cut rental rates first, impairing residual values and pressuring suppliers' order visibility. Consensus is likely underweighting this asymmetry—semiconductor vendors capture the build phase, but compute lessors absorb the utilization cycle.

A genuine low-cost launch breakthrough would eventually pressure terrestrial satellite and launch incumbents, but that is a multi-year regulatory, insurance, and technical-execution question rather than a 2026 earnings catalyst. Falsify the cautious view with independently disclosed multi-year take-or-pay contracts, utilization above 80%, and positive data-center contribution margins after depreciation and interest expense.

Key Decisions for Investors

• Do not initiate a directional SPCX position on this article alone; first require confirmation that the security is investable and that reported AI contracts appear in customer filings or audited disclosures. Treat any sharp retail-driven move as a liquidity-risk event, not a fundamental rerating.
• Maintain a 1-3 month quality-AI-infrastructure basket long NVDA/VRT versus a short on a diversified high-beta software or unprofitable AI-hosting basket only after next hyperscaler capex guidance confirms demand; target 10-15% gross spread upside with a 5% stop if aggregate hyperscaler capex guidance is cut.
• Watch GOOG disclosures for incremental committed compute capacity and AI infrastructure capex. A verified long-duration external capacity agreement would be modestly positive for GOOG by easing internal capacity constraints, but is more immediately positive for the disclosed supplier only if pricing, term, and minimum-volume commitments are known.
• Set an alert around NVDA and AVGO order commentary: reduce infrastructure longs if lead times normalize and customers signal lower 2027 capex, as this would shift the market from supply constraint to utilization-risk valuation.