The Case for (and Against) Buying SpaceX Stock Right Now

SpaceX is projected to reach a $100 billion ARR run rate by year-end, supported by reported monthly AI-compute contracts of $1.25 billion from Anthropic, $920 million from Google Cloud, and a further $1.11 billion hosting deal expected in December. However, the company remains unprofitable, trades near 20x 2027 forward sales despite forecasts for revenue to rise from $18.7 billion last year to $108.3 billion next year, and faces substantial lock-up-related share supply. Up to 328.4 million shares may become eligible for sale on each of Oct. 9 and Oct. 24, followed by as many as 1.3 billion shares after Q3 results; $25 billion of newly issued bonds at 5.35%-6.65% also raises funding risk as interest rates increase.

Analysis

SPCX is being valued as a scarcity AI infrastructure asset, but its economics should be underwritten as a leveraged project-finance hybrid: recurring compute revenue carries meaningful power, cooling, depreciation, network, and customer-concentration costs before it converts into free cash flow. The key non-obvious issue is that internal vertical integration can reduce launch and network unit costs while simultaneously obscuring transfer pricing between launch, satellite, and compute segments; consolidated ARR is therefore a weak indicator of standalone profitability. GOOG is a more credible indirect beneficiary if third-party capacity helps it meet AI demand without committing incremental owned capex, although this would also signal that hyperscaler capacity remains structurally constrained.

The next several trading days are dominated by technical supply rather than fundamentals. With a limited public float, each unlock can create outsized volatility and reduce the scarcity premium; the larger post-earnings release is the more important 1-3 month event because it coincides with the first opportunity to test whether reported revenue growth is matched by gross-margin and cash-flow disclosure. A rise in long-end yields would be particularly damaging: it raises refinancing costs while reducing the present value of cash flows that remain distant and execution-dependent.

Consensus appears too focused on the binary success of orbital compute. The nearer risk is conventional terrestrial AI capacity becoming less scarce before orbital infrastructure has commercial proof, compressing hosting prices and leaving SPCX with an expensive buildout whose utilization is uncertain. Conversely, a sustained capacity shortage, independently disclosed contract duration/backlog, and improving operating cash conversion could justify continued premium valuation; absent those data, the post-unlock weakness is more likely a financing signal than a clean dip-buying opportunity.

Key Decisions for Investors

• Avoid initiating a directional long in SPCX before the October unlock windows and first post-unlock earnings release. Reassess only if management provides segment gross margin, capex, contracted backlog duration, and a credible path to positive operating cash flow; ARR alone is insufficient.
• For a tactical bearish expression over the next 1-3 months, prefer SPCX put spreads or a defined-risk put fly rather than an outright short, subject to option liquidity and borrow confirmation. The thesis is float normalization plus rate sensitivity; exit if shares absorb unlock volume and close above the post-earnings high on expanding volume.
• Pair a small long GOOG against SPCX downside exposure over 3-6 months. GOOG has a stronger balance sheet and can monetize AI demand while retaining the option to diversify capacity supply; invalidate the pair if Google materially increases owned-data-center capex without corresponding cloud margin stability.
• Monitor 10-year Treasury yields, SPCX bond spread performance, and quarterly capex-to-revenue. A widening credit spread or capex accelerating faster than revenue would confirm financing risk; stable spreads plus improving free-cash-flow conversion would negate the bearish setup.