SpaceX Stock Now Has a Price to Sales Ratio Over 115x. Is It Worth Buying Anyway?

SpaceX is valued at about 116x trailing 12-month sales on ~$19.3B revenue, which the article calls “incredibly high.” It argues AI compute deals (Anthropic, Google, Reflection AI) could add ~$2.32B per month—~$27.8B annualized starting October—potentially driving 2027 revenue to roughly $59B ($22–24B Starlink, $30B xAI, ~$6B rocket launches) and implying a lower forward P/S near 38. Despite the revenue upside, the key caveat is profitability uncertainty for 2027, so the stock remains expensive even on improved assumptions.

Analysis

The market is likely overfocusing on the headline sales multiple and underpricing the difference between revenue quality and revenue quantity. If the new compute business is real, it is not automatically software-like: it can be capital-intensive, power-constrained, and margin-dilutive, so the equity value created per incremental dollar of sales may be far lower than the top line suggests. The more obvious second-order beneficiaries are the picks-and-shovels layer—GPU, networking, and power/cooling vendors—rather than the private platform itself.

The consensus miss is sequencing risk. Even if contract value is large, cash conversion can lag by several quarters and can reverse quickly if capacity is leased, customer concentration rises, or pricing weakens. Over the next 1-3 months, the key catalyst is not narrative expansion but proof of margin structure and backlog durability; over 6-18 months, any slippage in launch cadence or capex discipline would compress the premium fast.

Alphabet being named as a counterparty is not a clean negative for the stock; for a hyperscaler, external capacity is often a flexibility trade rather than an existential one. The more fragile view is the private-market extrapolation: investors may start treating every AI-adjacent platform as if revenue optionality alone justifies triple-digit sales multiples. That is where the setup looks most overdone.

Key Decisions for Investors

• Do not chase SPCX on the headline; treat it as a watch item until audited gross margin and FCF conversion are visible. If secondary liquidity is available, fade rallies that imply >80-100x forward sales without proof of durable margin.
• Long NVDA on a 3-6 month horizon as the cleaner second-order beneficiary of external AI compute demand; buy on 5-7% pullbacks, target 10-15% upside, and cut if datacenter capex guidance rolls over.
• No meaningful directional trade in GOOG/GOOGL from this alone; the customer relationship is more signal than P&L impact. Use any dip in Alphabet only if broader AI capex trends remain intact and ad/search fundamentals are unchanged.
• Set a catalyst alert for the next disclosure on compute margin and backlog duration: if revenue recognition slips or margins come in below hosting-like economics, the thesis is falsified and the valuation multiple should compress sharply.