The biggest IPO in history
On 12 June 2026, Space Exploration Technologies Corp. — SpaceX — sold shares to the public for the first time, on Nasdaq, under the ticker SPCX. It priced at $135, raised about $75 billion — the largest initial public offering ever completed — jumped 19% on its first day — to about $161 — and has since settled around $153. At that price the company's equity is worth roughly two trillion dollars.
Put that number next to the financials in the prospectus and it should give you pause. Two trillion dollars is about 107 times the company's 2025 revenue of $18.7 billion, and the company did not earn a profit — it lost $4.9 billion. For the first time, though, we can stop guessing: a just-listed SpaceX comes with an audited S-1, real segment financials, and a live market price to value against. So we will do the exercise properly.
You are an analyst the morning after the IPO. The filing is on your desk; the tape says $153. Is the stock cheap, about right, or expensive on the fundamentals? And if it is expensive — what, precisely, would you have to believe to justify the price? Hold an answer in mind; we will return to it with numbers.
One filing, three businesses
A sum-of-the-parts valuation needs parts. The filing hands us three.
After xAI was merged into the company early in 2026, SpaceX reports three operating segments, and they could not be more different from one another. Connectivity is Starlink — satellite broadband — and it is the cash engine: $11.4 billion of 2025 revenue, +$4.4 billion of operating income, an EBITDA margin around 63%, and 10.3 million subscribers as of March 2026, up 105% in a year. Space is the launch business — Falcon, Dragon, Starshield — the world's dominant launch provider, yet near break-even on $4.1 billion of revenue, because most of its launches are internal Starlink missions flown at cost. AI is xAI and Grok: $3.2 billion of revenue, an operating loss of $6.4 billion, and $12.7 billion of capital expenditure — a business burning money at scale against an enormous claimed market.
The consolidated cash picture follows directly. In 2025 SpaceX generated $6.8 billion of operating cash flow and spent $20.7 billion on capital expenditure, for a free cash flow of negative $14.0 billion. Starlink prints the cash; the AI segment consumes it and then some; the IPO refilled the tank with about $74 billion of net proceeds, whose first stated use is "AI compute infrastructure." Three businesses, three risk profiles — so we will value each at its own discount rate, then add them up.
How to value it
For each segment we project ten years of unlevered free cash flow — the cash the operating business throws off, before financing — discount it at the segment's cost of capital, and add a terminal value for everything beyond year ten:
The discount rate r is where two things enter. First, risk: a stable subscription business deserves a lower rate than a speculative one. We build each rate from the standard capital-asset-pricing logic — a risk-free rate plus a beta times the equity premium — and lever it lightly, because SpaceX carries almost no debt relative to its equity. Second, ESG: we nudge each rate by a small amount to reflect the environmental, social and governance character of the segment, in the direction the asset-pricing literature implies.
| Segment | Beta | ESG nudge | WACC | rationale |
|---|---|---|---|---|
| Connectivity | 1.00 | −20 bp | 8.98% | stable subscriptions; a small social discount for digital access |
| Space | 1.30 | +30 bp | 10.99% | lumpy; an environmental premium for launch emissions and orbital debris |
| AI | 1.55 | +20 bp | 12.18% | speculative; a governance premium for concentration and related parties |
The sign of the ESG adjustment is not arbitrary. Pástor, Stambaugh and Taylor show that when investors have a taste for green assets, those assets earn a lower cost of capital and carry a higher valuation; the carbon-premium evidence of Bolton and Kacperczyk points the same way for "browner" assets, which command a higher discount rate. So Starlink earns a small discount and the riskier segments a small premium. Notice the magnitudes, though — twenty or thirty basis points. Hold that thought; it becomes the lesson.
Valuing the parts
The anchorStarlink
Starlink is the segment a discounted-cash-flow model handles most comfortably, because it already generates cash. We let subscribers keep growing into the tens of millions while ARPU keeps falling — the filing is explicit that revenue per user has slid from $99 a month in 2023 to $66 by early 2026 and will keep declining as growth shifts to cheaper plans — and let the EBITDA margin expand toward a utility-like 70% as the constellation matures. Discounting at 8.98% with 3% terminal growth, the Connectivity segment is worth about $408 billion. That is roughly seven times a much larger 2035 EBITDA: rich, but not heroic, for what is becoming a global broadband utility. It is the crown jewel, and it anchors everything that follows.
Franchise + optionSpace, and Starship
The launch business is harder to love on today's numbers. As a near-break-even operation flying mostly internal missions, a straight DCF values it at only about $8 billion — strikingly, less than its $28 billion order backlog, a reminder that backlog is revenue, not value, and that thin margins cap what a launch dollar is worth. The real upside is not in today's Falcon; it is in Starship — point-to-point transport, heavy lift, lunar and Mars cargo — which is not yet in the cash flows. So we value it separately, as a real option, by scenario: $0 if it never pays off, $80 billion in a base case, $250 billion if it transforms launch economics.
The swing factorAI
The AI segment is where honest valuation gets uncomfortable. It earns $3.2 billion, loses $6.4 billion at the operating line, spends $12.7 billion on capacity, and points at an issuer-claimed total addressable market of $26.5 trillion. Tiny today; potentially enormous; genuinely unknowable. So rather than pretend to a single estimate, we run three scenarios — and even the optimistic one is sobering.
| Scenario | what you believe | AI segment EV |
|---|---|---|
| Bear | never earns its capital; the burn continues | −$24bn |
| Base | scales to solid profitability by about 2031 | −$2bn |
| Bull | captures real share of its market | +$83bn |
Read the bottom row carefully. Even in the Bull case — hyper-growth and a healthy terminal margin — a ten-year DCF values the entire AI segment at under $100 billion. Hold that figure against the $26.5 trillion market the segment is chasing, and against the price we are about to test.
Adding it up
Summing the parts, bridging from enterprise to equity value, and dividing by the post-IPO share count gives a value per share. In the base case it looks like this.
| Connectivity / Starlink | $408bn |
| Space (base launch business) | $8bn |
| Starship real option | $80bn |
| AI / xAI | −$2bn |
| Σ Enterprise value | $494bn |
| + cash & IPO proceeds − debt | +$61bn |
| Equity value ÷ 13.08bn shares | $555bn |
| Value per share | $42 |
| Market price | $153 |
And because Connectivity and Space form a fixed anchor while only the moonshot — AI plus the Starship option — moves with the scenario, the whole valuation collapses to a simple picture: a football field of value per share against the price.
A disciplined sum-of-the-parts puts SpaceX between roughly $35 and $62 a share across bear and bull. The market says $153. No reasonable set of cash-flow assumptions closes a gap that large. So the right question is no longer "is it expensive?" — it plainly is — but "what is the market seeing that the cash flows are not?"
What must you believe?
When a DCF and a price disagree by a factor of three, the most useful move is to invert the model.
We cannot reach $153 from the cash flows. So instead of forcing the assumptions, we hold the parts we are confident about — Starlink, the base launch business, the Starship option — at our disciplined marks, and ask the model a single question: what value must the market be assigning to the AI / moonshot bucket for the equity to be worth $153 a share? The answer is the residual that makes the arithmetic work.
The number is ≈ $1.45 trillion. To pay $153 for SpaceX you must believe the AI-and-moonshot piece — the bit a ten-year DCF cannot see — is worth roughly one-and-a-half trillion dollars, against a disciplined estimate of essentially zero, and against the roughly $200 billion that private markets last attached to xAI (press-reported).
That $1.45 trillion is not a precise valuation of the AI business. It is a residual — it absorbs every conservative choice we made elsewhere, and it moves if you change Starlink's terminal growth or the Starship scenario. What it measures cleanly is the size of the bet: the value the market assigns to everything a disciplined DCF cannot capture — Starship at scale, Mars, orbital compute, AI market capture.
A narrative stock is priced on the option, not the cash flows.
This is the lesson, made concrete, that Damodaran tells about Tesla and every story stock since: when narrative and numbers diverge this far, the discounted-cash-flow model has not failed — it has done its job, which is to tell you exactly how large a story you are being asked to buy.
Where ESG actually bites
We put real ESG factors into the discount rate. Starlink's mission to connect three billion unconnected people is a genuine social good, so it earns a small discount; the launch business carries an environmental premium for emissions, for the National Environmental Policy Act reviews that can delay launches, and for the orbital-debris and Kessler-syndrome risk of a crowding low-Earth orbit; the AI segment carries a governance premium. Now ask the question that matters for an investor: how much does any of this move the value?
A few dollars a share. The ESG adjustments are twenty or thirty basis points on the discount rate; the gap to the market price is about $110. ESG moves cents; the moonshot moves everything.
That is not an argument that ESG is irrelevant — it is a sharper lesson about where ESG lives in a valuation. For a narrative stock, a basis-point repricing of risk is second-order to the growth beliefs baked into the cash flows. ESG's real bite here is not in the discount rate at all; it is in the cash-flow scenarios — would debris regulation cap the constellation? does Starshield concentrate geopolitical and reputational risk? — and, above all, in governance, to which we now turn. This is the morning's machine P = E[m·x] in a single stock: ESG mostly in the discount rate (cents), the moonshot in the cash flows (everything). Knowing where each force lives is itself the integration result worth teaching.
The governance overlay
SpaceX has two classes of common stock. Class A, held by the public, carries one vote; Class B, held by Elon Musk and insiders, carries ten. The arithmetic of that structure is stark: with about 5.70 billion Class B shares against 7.38 billion Class A, Musk controls 82.4% of the vote while holding roughly 46% of the economics. SpaceX is, in Nasdaq's language, a "controlled company," exempt from the requirement that a majority of its board be independent. Musk is chief executive, chief technology officer, and chairman.
Then there is the pay award. Musk's 2026 performance grant is some 1.30 billion Class B shares — about 9% dilution — that vest only on a combination of market-capitalisation milestones reaching into the multiple trillions and operational moonshots: a permanent Mars colony of at least one million people, and non-Earth data centres delivering 100 terawatts of compute a year. The dilution, in other words, is contingent on precisely the upside that would trigger it.
The same 82.4% control that prevents minority holders from disciplining strategy, capital allocation, or the $20.2 billion of related-party compute leases with a board member's fund is also what let SpaceX make decade-long, capital-destroying bets — reusable rockets, a satellite constellation — that a quarterly public market would have killed in the cradle. The G in ESG is not a checklist here. It is the single question of whether you trust the one person who controls the moonshot.
What SpaceX teaches about valuation
Four lessons survive the exercise. First, a sum-of-the-parts disciplines a story: splitting SpaceX into a cash engine, a franchise, and a swing factor — each at its own discount rate — is far more honest than one blended multiple on a loss-making whole. Second, when a DCF and a price diverge by three times, you invert the model; the reverse DCF turns "it's expensive" into a falsifiable claim about a $1.45 trillion moonshot. Third, ESG lives somewhere specific — a few basis points in the discount rate, but first-order in the cash-flow scenarios and in governance. And fourth, governance is priced: 82.4% control is at once the central risk and the reason the audacious bets were ever made.
So return to the question from the prologue. The disciplined number is somewhere around forty dollars, a bull case in the low sixties; the market pays $153. You do not have to call the market wrong. You only have to decide whether you are the analyst — who values the cash flows — or the market — which is buying a call option on Mars. The honest valuation does not settle that for you. It tells you, to the dollar, what you are being asked to believe.
References & sources
Finance citations are drawn from leading journals and foundational works. All financial figures are extracted from the audited SEC filing and computed in the accompanying workbook.
- Pástor, Ľ., Stambaugh, R. F., & Taylor, L. A. (2021). Sustainable investing in equilibrium. Journal of Financial Economics, 142(2). — green tastes → lower cost of capital, higher valuation, lower expected return.
- Bolton, P., & Kacperczyk, M. (2021). Do investors care about carbon risk? Journal of Financial Economics, 142(2). — the carbon premium: browner assets, higher discount rate.
- Cochrane, J. H. (2011). Discount rates. Journal of Finance, 66(4). — why valuations move on the denominator; the discount-rate lens this case applies.
- Damodaran, A. (2017). Narrative and Numbers: The Value of Stories in Business. Columbia University Press. — how story and DCF meet for high-growth firms.
- U.S. Securities and Exchange Commission. Space Exploration Technologies Corp., Form S-1/A (Amendment No. 2), June 2026, CIK 0001181412. — all base-year financials, share structure, governance and risk factors.
- SpaceX ESG valuation workbook (2026). SpaceX_ESG_Valuation_2026.xlsx. — the sum-of-the-parts model; figures here are its outputs.