$72 a share is the cleanest way to understand SpaceX’s IPO premium: Morningstar frames that gap as the market’s call option on orbital data centers that don’t yet exist.
XOOMAR Intelligence
Analyst Take
SpaceX is reportedly coming to market on Friday with a $75 billion stock offering that is deeply over-subscribed, with some institutional investors seeking $10 billion blocks, according to TechCrunch. The bankers’ implied valuation is nearly $1.8 trillion. Morningstar puts fair value around $825 billion. Aswath Damodaran, the New York University finance professor, lands at $1.2 trillion.
That valuation gap is the story. SpaceX’s mature businesses already look rare: high-margin launch, Starlink, and a strategic grip on access to space in the U.S. and Europe. But the richest IPO upside appears tied to something much harder to underwrite: space AI compute, built around orbital data centers.
$135 Per Share Prices More Than Rockets
Morningstar’s analyst draws the sharpest line. The firm’s fair value estimate is $63 a share, versus the offering price of $135. That leaves $72 as the implied premium for SpaceX delivering orbital data centers at the pace and capability Musk says the company will attempt.
XOOMAR analysis: that makes this IPO less a clean aerospace listing and more a hybrid of launch monopoly, satellite communications, AI infrastructure, and hard-tech venture financing. Public investors aren’t just buying Falcon launches or Starlink subscriptions. They’re being asked to price a market that SpaceX still has to create.
There are obvious reasons for caution in the source material. TechCrunch notes that big IPOs tend to sink, SpaceX is losing money, and Musk’s public behavior would raise governance questions at any other technology company. Yet demand appears intense. That tension says investors are not mainly paying for the last decade of SpaceX execution. They’re paying for the next industrial leap.
The IPO has already become a gravitational event for adjacent space-market speculation. XOOMAR has tracked that pull in $1.2B Quantum Space SPAC Chases SpaceX IPO Cash Wave, while pre-IPO volatility showed up in SpaceX Pre-IPO Hype Cracks as SPCX Dumps 27% in 3 Weeks.
Three Hard-Tech Hurdles Sit Behind the $1.8 Trillion Pitch
TechCrunch identifies three near-impossible engineering feats inside the orbital data center plan: a reusable rocket, a new American chip foundry, and a faster-than-ever satellite production ramp.
| SpaceX moonshot | Why it matters to the IPO | Source-backed pressure point |
|---|---|---|
| Reusable Starship | Lowers the cost of placing heavy compute hardware in orbit | Rapid reusability is not yet proven, and SpaceX is under an FAA mishap investigation |
| AI satellite factory | Turns orbital compute from demo into capacity | Musk’s target implies 6,666 satellites a year, about 556 a month |
| Terafab chip foundry | Feeds later-stage scaling toward massive compute output | Chip fabs typically cost billions and can take as long as a decade |
Starship is the foundation. Without cheap, repeatable heavy-lift launch, orbital data centers become a capital sink. A recent test flight “went well enough,” TechCrunch reports, but did not show rapid reusability is close. SpaceX may reuse only the booster at first, which would raise rollout costs.
The second hurdle is manufacturing. Musk’s math, based on 150 kW of maximum power delivery per satellite, implies 6,666 satellites a year to reach an annualized gigawatt per year of space AI compute. That is roughly twice the reported current Starlink production rate of 70 satellites a week.
The third hurdle is Terafab, SpaceX’s planned chip foundry. This is the least forgiving piece. Rockets can fail upward through testing. Satellite manufacturing can improve through iteration. Chip fabs demand capital discipline, process control, and time.
Musk’s Gigawatt Target Is the IPO’s Hardest Number
Musk gave investors a number that is easy to remember and hard to model.
“This is not a promise of what we’ll do,” Musk said in the video. “This is what we are going to try to do, and think we probably can do, which is to get to roughly an annualized rate of a gigawatt per year by the end of next year, in terms of space AI compute.”
That caveat matters. The target is not presented as a promise. It is a stated attempt. Public markets often struggle with that distinction when a founder has a record of turning improbable manufacturing goals into investor narratives.
SpaceX’s own market framing adds another complication. In its S-1 market analysis, the company puts its largest opportunity in enterprise AI, estimating that market at $22.7 trillion. It compares that with $2.4 trillion for AI infrastructure and just under $2 trillion for its space efforts.
But the company is also selling compute to Anthropic and Google, which TechCrunch describes as ostensible competitors in the model business. That creates a strategic split: is SpaceX trying to own AI models, sell compute to AI labs, or do both?
XOOMAR analysis: the orbital data center story is designed to collapse that trade-off. If SpaceX can create enough compute in orbit, it could support its own AI ambitions while selling capacity to others. If it can’t, the company risks looking more like a neocloud with extraordinary launch assets than a full-stack AI platform.
The Existing Businesses Are Stronger Than the Orbital Compute Proof
The most attractive parts of SpaceX, in both the Morningstar and Damodaran analyses, are the high-margin space launch business and the satellite internet network. That matters because it gives the IPO a floor that most speculative hard-tech offerings do not have.
Starlink is not just a revenue story in this framing. It gives SpaceX satellite manufacturing experience, orbital operations knowledge, and a communications network that could support future space infrastructure. The catch is scale. TechCrunch says the AI satellites are simpler in architecture than Starlink satellites, per Musk, but the production facility for the new satellites has not been built yet.
The NASA comparison is another warning signal. NASA has a nearly $4 billion contract with SpaceX to use Starship as a moon lander, but TechCrunch reports that the agency still is not ready to commit to a test mission with the vehicle scheduled for late 2027. That does not mean Starship fails. It means even major government partners are treating the timeline with caution.
Four Investor Stories Are Competing Inside One Stock
Growth investors may see SpaceX as a rare infrastructure platform with multiple shots on goal. Value-oriented buyers will see a company losing money, asking public markets to fund a capital-heavy expansion into AI compute before the core technology stack is proven.
Cloud and AI buyers have a different lens. The source confirms deals to sell compute to Anthropic and Google, but it does not show whether orbital compute has paying customers at scale. That distinction is everything. A contract for compute is not proof that space data centers can run with repeatable economics.
Regulators sit closer to the launch side of the story. The source specifically names the FAA mishap investigation into why the Starship booster stage failed to make a controlled reentry as planned. If orbital compute depends on high launch cadence, regulatory interruptions become investor-relevant, not just operational noise.
Rivals are less visible in the supplied reporting. TechCrunch does not document competitive responses. The safer read is that SpaceX’s own execution pace, not rival reaction, is the first test.
The Post-IPO Test Is Evidence, Not Narrative
Three evidence points will decide whether the premium holds.
Starship economics: investors need to see reuse move from test objective to operating reality. Reusing only the booster at first would weaken the cost case for orbital data centers.
Satellite production: the implied target of 556 AI satellites a month needs a real factory, real throughput, and repeatable quality. That is a manufacturing proof point, not a slide-deck milestone.
Compute customers: SpaceX must clarify whether the AI business is model-building, compute sales, enterprise agents, or some durable mix of the three. The $22.7 trillion enterprise AI market estimate is only useful if SpaceX can show how value flows to its products.
SpaceX’s IPO may be extraordinary because the existing company is already strategically important. But the valuation premium belongs to a future market that still has to be built in orbit. The next watch item is simple: whether SpaceX can turn Musk’s gigawatt target from ambition into measurable capacity, with launch cadence, chips, satellites, and paying compute demand all moving at the same time.
The Bottom Line
- SpaceX’s IPO price appears to depend heavily on unproven orbital data center ambitions.
- The gap between Morningstar’s fair value and the offering price highlights how much future growth investors are underwriting.
- Strong institutional demand suggests public markets may be willing to fund hard-tech moonshots at unprecedented scale.
SpaceX IPO valuation benchmarks
| Benchmark | Valuation | Per-share figure | Implication |
|---|---|---|---|
| IPO implied valuation | $1.8 trillion | $135 offering price | Market pricing includes major future upside |
| Damodaran estimate | $1.2 trillion | Not provided | Below IPO valuation but above Morningstar |
| Morningstar fair value | $825 billion | $63 fair value | Suggests a $72 per-share premium |
SpaceX valuation estimates
Primary Sources & Disclosures
Written by
XOOMAR
Data desk
XOOMAR is a capital markets software and data company. Every brief on this site starts from a dataset the company collects itself from primary sources (CFTC, SEC EDGAR, FINRA, the Federal Reserve, exchange APIs) and names the numbers it is built on, with a link to the data page so you can check them. Briefs are reviewed before they go out and corrected in place when the data is revised.
