Is $1.75 trillion too much?
SpaceX targets the largest IPO valuation in history. At ~94× trailing revenue and a $4.9B FY25 net loss, the multiple has no clean precedent at the mega-cap scale. Bloomberg has reported demand strong enough to push pricing toward $2T; analysts have separately warned that any post-IPO multiple compression has further to fall than usual. This is the bull case, the bear case, and a sum-of-the-parts that lets you set your own number.
~94× sales. No public-market comp.
The target valuation divided by FY2025 revenue produces a trailing price-to-sales multiple of approximately 93.7×. There is no positive-revenue mega-cap currently trading at anywhere near that level. The closest reference points are pre-revenue biotech and very-early-stage technology IPOs — names that trade on optionality alone — but those companies typically have market caps in the billions, not the trillions.
Where would $1.75T put SPCX on a multiples scatter? Here are the trailing P/S figures for the most-watched names in mega-cap technology, sorted from cheapest to most expensive:
| Ticker | P/S (trailing) | Net margin | Multiple commentary |
|---|---|---|---|
| AMZN | ~3–4× | ~6% | Retail dominates revenue mix; profit comes from AWS |
| GOOGL | ~6–7× | ~24% | Mature advertising business + cloud growth |
| AAPL | ~7–9× | ~26% | Services mix shift driving multiple expansion |
| TSLA | ~10–12× | ~10% | Auto + energy storage; multiple reflects AI and FSD optionality |
| MSFT | ~12–14× | ~36% | Recurring SaaS + Azure; most defensible mega-cap multiple |
| NVDA | ~30–35× | ~55% | Highest P/S in mega-cap; backed by ~60% net margin |
| SPCX target | ~94× | (26)% | Unprofitable, growing 33% YoY |
P/S ranges are approximate trailing values across May 2026. Source: company filings, exchange data. Net margin from latest 10-K.
Two observations matter. First: the SPCX multiple is ~3× Nvidia's — the most expensive mega-cap on a sales basis — but where Nvidia trades on a profitability mountain, SPCX trades on a profitability promise. Second: this multiple looks much less alarming on a sum-of-the-parts basis, which we work through below.
Bull and bear, side by side.
Bull case
- Starlink unit economics are inflecting. 38.6% segment operating margin in FY2025 on $11.4B revenue is materially better than what bears and bulls were modeling at the last private-market mark. Each incremental subscriber is closer to pure margin than the prior one.
- Starship halves launch cost again. Falcon's reusability already disrupted launch pricing. Starship, fully reusable, takes that another step. The Starlink V2 deployment economics improve materially the day Starship reaches operational cadence.
- xAI optionality is free. The merger structure means the public is paying for Starlink and Starshield economics; the xAI line is priced as an option, not as a discounted cash flow. Even modest commercialization of orbital AI compute would re-rate the company.
- Scarcity premium. SPCX is the only space pure-play of mega-cap scale and the only company with a credible Mars architecture. Index inclusion will be automatic; passive bid is structural.
- The Musk premium. The historical pattern (Tesla 2010–present) is that any Musk-led public company trades at a structural premium to discounted-cash-flow fair value. The S-1 acknowledges key-person risk; the market priced it as a positive on Tesla for fifteen years.
Bear case
- The multiple has nowhere to expand and a long way to fall. A re-rating to even the Nvidia P/S of ~32× takes the price down ~66% from $1.75T. Multiple compression from 94× to 40× takes off ~57%. There is no historical mega-cap precedent for sustaining 94× trailing sales.
- Starship is binary and slipping. Every Starship R&D dollar shows up as a loss. The S-1 confirms $3.0B FY25 + $930M Q1 2026. Each quarter without first commercial payload to orbit raises the question of when, not whether, schedule expectations re-set.
- Regulatory risk is concentrated. FCC spectrum, FAA launch licenses, ITU coordination, and U.S. State Department export controls all intersect the SpaceX business. Any one of those agencies can stand the company down on short notice.
- Customer concentration in U.S. government. The S-1 names NASA, NRO, DoD, and Space Force as material customers. U.S. budget cycles, continuing-resolution gaps, and political shifts on space spending all create concentrated exposure.
- Key-person risk is real. The S-1 lists it. Musk is also CEO of Tesla and X — public-company peers have already noted the divided attention. A health event, a regulatory enforcement action, or a controlling-shareholder dispute could revalue the equity overnight.
- xAI integration is unproven. The merger added a money-losing segment. Compute capex, model training cycles, and the X platform's monetization rebuild all run independent of the launch and connectivity businesses. Integration overhead is rarely the multiple-expanding story bankers describe at IPO.
Building $1.75T from the segment up.
This is an illustrative SOTP, not a recommendation. It applies segment-appropriate multiples to FY2025 segment revenue, then aggregates. The point is to show where the implied valuation can come from — and which segments carry the most subjective weight.
| Segment | FY25 revenue | Multiple framework | Low | Mid | High |
|---|---|---|---|---|---|
| Starlink | $11.4B | SaaS-like P/S on recurring connectivity revenue | $340B (30×) | $570B (50×) | $910B (80×) |
| Falcon | ~$4.2B | Defense + commercial launch earnings multiple | $40B | $80B | $130B |
| Starshield | ~$1.8B | Defense prime contractor multiple | $25B | $50B | $90B |
| xAI / Other | ~$1.3B | Private AI comp (Anthropic, OpenAI implied) | $80B | $150B | $300B |
| Starship | n/a (R&D) | Optionality — binary launch milestone | $0 | $150B | $400B |
| SOTP total | $18.67B | — | $485B | $1.00T | $1.83T |
Multiples are illustrative ranges. SaaS comps near the low end reflect mature broadband connectivity multiples; the high end reflects highest-growth SaaS comps. Defense earnings multiples follow the 15–22× P/E range typical of prime defense contractors. AI comps reflect implied private-round valuations for Anthropic and OpenAI at FY25 revenue runrates.
What the SOTP tells you
The $1.75T headline is reachable only in the high case — and the high case requires (a) Starlink getting a SaaS multiple normally reserved for highest-growth software, and (b) Starship being valued at $300–400B of optionality, which in turn implies the market is willing to underwrite a launch milestone that hasn't happened yet. The midpoint SOTP lands much closer to $1.0 trillion, the standalone SpaceX valuation at the time of the xAI merger. From a pure SOTP perspective, the $1.75T target is asking the market to credit roughly $750B of upside that is real but not yet observable.
How peers actually trade.
| Ticker | Company | Stage | Approx P/S |
|---|---|---|---|
| TSLA | Tesla | Mature, profitable | ~10–12× |
| NVDA | NVIDIA | Mature, hyper-profitable | ~30–35× |
| RKLB | Rocket Lab | Growth, partial profitability | ~10–18× |
| ASTS | AST SpaceMobile | Pre-revenue scale, optionality stock | n/a (~$0 revenue ramp) |
| IRDM | Iridium Communications | Mature, slow-growth satellite | ~4–6× |
| LUNR | Intuitive Machines | Early-stage lunar contractor | ~8–14× |
The cleanest narrow comp is Rocket Lab (RKLB) — the only other publicly traded launch + space-systems operator at meaningful scale. Even there, RKLB trades on a P/S in the high-single-digit to high-teens range, an order of magnitude below the SPCX target multiple. The gap reflects two real differences: Starlink's $11B+ recurring revenue (RKLB has nothing comparable), and the binary Starship optionality (RKLB's Neutron is a similar story but at a much smaller absolute scale).
See our full peer comparison page for live charts of each peer and a more detailed side-by-side.
Four outcomes that would make SPCX cheap.
- First commercial Starship payload to orbit. The S-1 targets H2 2026. The day Starship delivers a customer payload, Starlink V2 deployment economics step down and Starship optionality compounds into the segment's valuation.
- Starlink gross margin steps to 50%. Currently ~38.6% operating; segment gross margin is materially higher. A reported gross-margin print north of 50% would change the SaaS comp peer set the segment is valued against.
- Starshield contract step-up. A multi-year, multi-billion DoD prime contract for surveillance/early-warning would re-rate Starshield from a margin story to a unit story.
- xAI compute monetization. Material API revenue or enterprise commitments from the xAI segment changes "free optionality" into an earnings line. Even modest paid traction would expand the multiple at the consolidated level.
Four outcomes that would make SPCX expensive.
- Starship slips to 2027 or later. Every quarter without first commercial payload to orbit compresses the optionality multiple and increases the modeled R&D burn. A formal program-delay disclosure in a 10-Q would likely trigger a step-down.
- Starlink growth decelerates below 30% YoY. The SaaS-like multiple depends on persistent 30%+ growth. A print at 20% would force a peer-set re-evaluation toward mature broadband multiples.
- Regulatory enforcement. An FCC spectrum dispute, an ITU coordination breakdown, an export-control enforcement action, or an FAA stand-down would each reset the operating model.
- Lock-up cliff. Heavy insider selling at the December 9, 2026 lock-up expiry could create a multi-month overhang that compresses the multiple regardless of operating fundamentals.