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

SpaceX has fallen roughly 50% from its post-IPO peak and now trades near its $2 trillion opening-day valuation, despite management targeting a $100 billion ARR run rate by year-end. New AI-compute contracts could add about $3.28 billion of monthly revenue, including $1.25 billion from Anthropic and $920 million from Google Cloud, but the company remains unprofitable and trades near 20x 2027E sales. Near-term risks include up to 1.3 billion shares becoming eligible for sale after Q3 earnings and $25 billion of new debt carrying 5.35%-6.65% coupons amid rising rates and continued capital-intensive expansion.

Analysis

The key underwriting issue is not headline ARR but conversion to durable, high-margin cash flow. Large AI hosting commitments may be capacity reservations with customer concentration, pass-through power costs, cancellation rights, or heavy upfront network-equipment obligations; until disclosed, they should not receive hyperscaler-like revenue multiples. SPCX is effectively being valued as a blended AI infrastructure, telecom, aerospace, and launch monopoly, yet its capital intensity and depreciation profile make free-cash-flow breakeven—not revenue growth—the relevant 6-18 month rerating catalyst.

Near term, the asymmetric risk is technical: a small initial float makes incremental unlocked supply disproportionately price-setting, particularly if early holders lack a fundamental reason to retain exposure after a volatile post-listing period. The resulting weakness would spill over less to NVDA than consensus implies—GPU demand remains supply constrained—but could pressure AI-infrastructure comparables such as ORCL, CRWV, and NBIS if investors reassess whether contracted compute revenue earns software-like multiples. GOOG is a modest beneficiary if its agreement represents a flexible alternative to internally funded capacity, but its exposure is immaterial; the more relevant read-through is whether external capacity is genuinely cheaper than building internally.

Contrarian view: the market may be over-focusing on the lock-up calendar while underpricing execution correlation across the business. A launch-delay, data-center power/interconnect bottleneck, or funding-market disruption would simultaneously defer the three narratives supporting the valuation. Conversely, verified customer prepayments, disclosed gross margins, and evidence that compute demand is incremental rather than cannibalistic could absorb supply quickly; absent those disclosures, this is a trading vehicle rather than a core long.

Key Decisions for Investors

• Avoid initiating a directional SPCX long ahead of the remaining unlock and first post-IPO earnings release; reassess 1-3 months after supply normalizes, using disclosed compute gross margin, customer concentration, capex, and free-cash-flow guidance as gating data.
• For tactical mandates, express a bearish SPCX view via defined-risk put spreads spanning the next unlock/earnings window rather than an outright short. Target a 15-25% downside move; exit if the stock closes above the pre-unlock high or management discloses material prepaid, non-cancellable compute commitments.
• Pair long GOOG / short SPCX on a 3-6 month horizon: GOOG retains AI-demand exposure with a diversified cash-generative balance sheet, while SPCX bears funding, dilution, and execution risk. Falsify if SPCX demonstrates positive free cash flow before planned expansion or GOOG materially increases external-capacity commitments without corresponding AI monetization.
• Maintain NVDA exposure independently of SPCX. Do not treat a potential SPCX selloff as a broad GPU-demand signal unless it coincides with reduced accelerator orders, lower utilization, or customer capex cuts from multiple AI infrastructure buyers.