On May 20, 2026, SpaceX filed its S-1 with the SEC. For the first time, a company that had operated in deliberate opacity for 24 years volunteered to put its finances in writing, under oath, for the world to read.
The numbers are remarkable. The structure is even more remarkable. And the implications for anyone thinking seriously about the space economy are more significant than conventional coverage has acknowledged.
Start with what this IPO is not.
The IPO That Isn’t
In the old model, an IPO was a capital event. Founders needed money to build. Public markets provided it. Shareholders got equity and, in exchange, governance rights: a board that could hire and fire the CEO, proxy advisors with leverage, institutional shareholders with standing to object.
SpaceX launched in 2002. Post dot-com crash, post 9/11, into one of the worst capital environments in a generation for big, audacious, long-duration bets. They built it anyway. Twenty-four years later, they don’t need the money. The $75 billion raise at a $1.75 trillion valuation, the largest IPO in history if it closes, is not about capital. It’s about creating a currency.
A public listing with 23 investment banks means a liquid, globally recognised share price. That share price is how you attract and retain talent. It’s how you make acquisitions. It’s how you signal permanence to governments, regulators, and customers who are betting infrastructure decisions on your survival.
The shareholders are mostly along for the ride. The S-1 makes that plain. Respect.
In Musk We Trust
The governance structure of this IPO is, in a word, unprecedented for a company of this size and boldness of mission.
Class A shares, the ones the public will buy, carry one vote each. Class B shares, held by Musk, carry ten. Post-IPO, Musk retains majority voting control. He controls the board. The S-1 states it without ambiguity: Mr. Musk will be able to elect, remove, or fill any vacancy among the Class B Directors.
“For so long as he beneficially owns more than 50% of the voting power of our common stock, Mr. Musk will control the voting power over the selection of our board.”
In the traditional model, the board’s job is to hold the CEO accountable. Here, Musk controls the board. There’s no independent check. If the vision is right, this structure is an extraordinary feature. It’s precisely why SpaceX has been able to pursue a 24-year mission without being pressured for short-term returns. If the vision is wrong, there’s no mechanism to course-correct.
Buying SPCX is not a vote on the strategy. It’s a bet on the man. In Musk We Trust is not a marketing slogan. It’s the actual legal structure of the investment.
Follow the Money: Space, Communication, AI
The S-1 reports three segments. Reading them in order of where the money goes tells a different story than the headline numbers.
Space: The Credibility Engine
The rocket business, Falcon 9, Dragon, Starship, Starbase, generated $4.1 billion in revenue in 2025 and an operating loss of $657 million. Of that, $3 billion went to Starship R&D alone. Q1 2026 added another $930 million.
This isn’t where growth is coming from. It’s where credibility comes from. The track record of launching rockets is what gives investors permission to believe the rest. Space is the founding myth that makes everything else fundable. Without it, xAI is just another frontier AI company burning cash. With it, it’s part of a coherent civilisational project.
SpaceX has spent over $15 billion developing Starship, more than originally budgeted. Twelve test missions. Payload delivery to orbit expected in H2 2026. When that happens, the credibility engine becomes the cost-reduction engine: the mechanism by which everything else in the stack gets cheaper.
Communication: The One That Pays
Starlink generated $11.4 billion in revenue in 2025, 61% of total consolidated revenue, with an adjusted EBITDA of $7.2 billion, up 86% year-over-year. With 10.3 million subscribers across 164 countries and 9,600 satellites deployed, it’s the only segment making serious money.
Starlink is doing two things simultaneously. It’s the cash engine that funds everything else, with Starlink’s profitability effectively subsidising xAI’s losses. And it’s the connectivity layer of the entire stack. Orbital AI compute is useless without low-latency global connectivity to reach it. Starlink provides that. It’s infrastructure in the most literal sense.
The S-1 is explicit: Starlink satellites in Sun-synchronous orbit will provide the low-latency, global connectivity linking orbital AI systems to people around the world. Communication isn’t a standalone business. It’s the nervous system of the organism.
AI: Where the Money Goes and the Losses Live
xAI, now called SpaceXAI after the February 2026 merger, generated $3.2 billion in revenue in 2025 and lost $6.4 billion doing it. In Q1 2026 alone, it lost $2.5 billion on $818 million in revenue. Capital expenditure was $12.7 billion in 2025, more than SpaceX spent on Starlink and rockets combined.
This is where the majority of investment goes and where the majority of the TAM lives. The S-1 names it clearly: 100 gigawatts of AI compute deployed to orbit, space-based solar power, the everything app. These aren’t sideshows. They’re the primary thesis.
X, the everything app, acquired via xAI in March 2025, is the consumer layer of this stack. Payments, banking, communications, investing, media, identity. The S-1 describes it as integrating real-time information, communications, media, payments, banking, commerce and more within one consumer experience. Off-planet, X isn’t a social network. It’s the economic operating system.
SpaceX gets you to Mars. X is how you buy a coffee there.
Also inside the AI segment: Grok, the frontier model competing with GPT-4o and Claude. And Optimus, the humanoid robot. If you’re building a civilisation on another planet, you need physical labour that doesn’t require life support. Optimus is that.
In May, Anthropic, one of xAI’s chief competitors, agreed to pay xAI $1.25 billion per month to access compute through the Colossus data centre. That’s not a failing business. That’s early-stage infrastructure with a real revenue model.
The Full Stack: Why Every Block Has to Hold
Vertical integration is the thesis and the risk simultaneously.
The orbital economy SpaceX is building requires all of the following to work:
Remove any one component and the others are diminished. This isn’t three companies bolted together. It’s one organism with six critical systems. The S-1 is the first time that organism has had to be described in public, in legal language, with audited numbers attached.
Think of it like those building block games where every piece supports the others. Pull one out and the whole structure wobbles. The bet isn’t on any single component. The bet is on the integration.
The Imagination Premium
$18.7 billion in revenue. $4.9 billion net loss. $41.3 billion accumulated deficit. A $1.75 trillion valuation.
The gap between those numbers requires imagination to bridge. Specifically, it requires an inspirational TAM that’s not yet in the model: orbital data centers, space-based solar, lunar economy, multi-planetary commerce. The S-1 names these explicitly. It also notes, with admirable candour, that the timeline for certain of our initiatives involving unproven or new innovations may be difficult or impossible to determine.
That’s not a red flag. That’s an honest description of what frontier infrastructure investment looks like. Amazon lost money for years. So did Tesla. The question isn’t whether losses exist but whether the capital is building something real. The success of rocket launches gives credibility that shareholder patience will be rewarded. That’s especially important for the talent the mission needs to attract.
The adjusted EBITDA picture is more encouraging: $6.6 billion in 2025, with Starlink alone generating $7.2 billion. The losses are being driven almost entirely by xAI capex, a deliberate investment decision, not operational failure. Starlink and the legacy Space business are profitable on an operating basis.
Whether the imagination premium is justified depends entirely on #2. On whether you trust the man with complete control of all six systems to execute a 24-year track record into the next century.
What This Means for the Orbital Economy
Issue 1 of The View from Space asked what unlocks at $100/kg. The answer was: almost everything. Orbital manufacturing, data centers, asset custody, the infrastructure of a serious space economy.
This issue answers who profits when it does. The S-1 makes that clear. One company, vertically integrated across the full stack, with one man in complete control, has spent 24 years and $41 billion building toward a moment that, if the $100/kg threshold is crossed, will be the most significant economic inflection point since the internet.
You don’t have to agree with the governance structure to recognise the strategic logic. You don’t have to like the control concentration to track what it produces.
That’s exactly what The View from Space tracks.