The Day Space Went Public
Elon Musk woke up a trillionaire on June 12, 2026. Not through some crypto scheme or a viral tweet — though he's had his share of those — but because SpaceX priced its Nasdaq debut at $135 a share and the market decided, almost immediately, that it was worth far more.
The opening bell brought an 11% pop to $150. By close, shares were at $160.95 — a 19% gain on day one. The greenshoe overallotment option pushed total capital raised from the initial $75 billion target to a staggering $85.7 billion. That's not a typo. Eighty-five point seven billion dollars, pulled from public investors in a single transaction.
Within days, the stock climbed another 20% on June 15 and over 8% on June 16, pushing SpaceX's market capitalization to $2.7 trillion. Amazon had been the last company to hold that title, and now it was history.
What made this pop feel almost inevitable was the mechanics behind it. The IPO was oversubscribed roughly four times post-listing, but the publicly available float sat at just 4% of outstanding shares — index-eligible shares, meaning institutional holders couldn't easily dump stock into the market. Underwriters Goldman Sachs and Morgan Stanley walked away with roughly $500 million in fees alone. Robinhood reported what it called record-breaking traffic on the day.
This wasn't a tech IPO where early backers quietly cashed out. This was a company that had spent two decades proving it could do what governments couldn't — launch satellites, deliver cargo to the International Space Station, land boosters — and then decided it was time to let the public share in the upside.
Who Actually Got Rich
Let's talk about the money, because that's what this story is really about.
Musk retained 85.1% of SpaceX's voting power after the IPO. Eighty-five percent. The rest — the publicly traded shares — were a rounding error in his net worth calculation, but they created something remarkable for everyone else.
Approximately 4,400 SpaceX employees became millionaires overnight. Four hundred of them crossed into centimillionaire territory — that's nine figures, $100 million-plus net worth. These aren't executives in corner offices. These are engineers who spent years working on rocket engines, flight software, ground station operations. People whose day jobs involved making sure a $100 million rocket didn't explode on the pad.
The venture capital returns were even more staggering. Founders Fund, which held roughly 3% of the company from an initial $600 million investment, saw that stake balloon to over $50 billion. Sequoia's position reached more than $20 billion. Andreessen Horowitz came in at over $10 billion. Combined, those three firms extracted more than $80 billion in returns from investments made when SpaceX was still a scrappy startup trying to prove it could land a first stage.
There's something almost obscene about those numbers, but there's also a logic to them. SpaceX didn't just raise money — it built the most capable launch system in human history, repeatedly, at costs that made every competitor look like they were operating on a different planet. The IPO was the market's way of catching up to reality.
The Compute Money Machine
Here's where the story gets interesting — and a little unsettling.
SpaceX didn't just go public to fund more rockets. It went public to monetize something most people never thought the company owned: massive AI compute infrastructure.
Google signed a deal to lease 110,000 NVIDIA GPUs and CPUs from SpaceX for $920 million per month. The contract runs from October 2026 through June 2029 — that's roughly $31 billion in total commitment. The hardware lives at Colossus 1, SpaceX's data center facility in Memphis, Tennessee. Google needs the capacity to support its Gemini Enterprise products as demand for AI inference surges.
Anthropic followed with an even larger arrangement: $1.25 billion per month through 2029, also anchored at Colossus 1. That's $47 billion over the contract period.
Combined, these two deals represent something close to $78 billion in committed revenue before SpaceX has even launched a single new payload from the facility. The math is brutal in the best way: SpaceX spent years building GPU clusters to train its own AI models, and now it's converting what would have been capital expenditures into high-margin recurring leases.
The timing is no accident. These agreements were signed in the months leading up to the IPO, giving SpaceX something public investors love to see — contracted revenue streams that make the valuation feel less like speculation and more like a utility business. You don't buy a rocket company. You lease compute from one.
The $60 Billion Bet on Software
Days after the IPO closed, SpaceX announced it was acquiring Cursor — the popular AI-powered coding editor that had been gaining serious traction among developers.
The deal was valued at $60 billion and structured as an all-stock swap. That's not a typo either. Sixty billion dollars for a software tool.
On the surface, this looks like a pivot. SpaceX has always been a hardware company — rockets, satellites, starships. Cursor is code. But the logic holds if you think about vertical integration differently.
SpaceX has been building AI capabilities in-house through xAI, its large language model effort. The Cursor acquisition suggests a strategic repositioning: rather than competing directly in the consumer AI tool space, SpaceX is buying an established product and folding it into its broader software ecosystem. Enterprise customers who already use Cursor get a seamless path to SpaceX's compute infrastructure. Developers get better tooling backed by more powerful hardware.
It also raises an uncomfortable question about Musk's attention. He's now running Tesla, SpaceX, xAI (or at least the AI division that Cursor may be replacing), and a public company with $2.7 trillion in market value. SpaceX COO Gwynne Shotwell has hinted that a potential merger between SpaceX and Tesla could ease the management strain — speculation that sent ripples through investor circles.
Whether that merger happens or not, the Cursor deal signals that SpaceX is thinking about software the way it used to think about launch: as a moat, not an afterthought.
The Risks Nobody's Talking About
Every IPO prospectus contains warnings. SpaceX's S-1 filing was no exception, and it included specific language about future equity dilution risks.
Here's what that means in practice: as SpaceX issues more shares to pay for acquisitions, retain employees, and fund operations, each existing share becomes a smaller slice of the pie. The $60 billion Cursor deal, paid in stock, will absolutely dilute current holders. So will any future acquisitions or capital raises.
Then there's the revenue reality check. SpaceX reported $18 billion in revenue for 2025 — impressive by almost any standard, but it also posted a $4.9 billion loss that year. The company is profitable on launches, sure, but the AI infrastructure buildout and software ambitions require sustained investment that could keep losses elevated for years.
The compute leasing deals help, but they're also commitments. Google and Anthropic are locking in pricing for three years. If AI demand softens, if new chip architectures arrive that make SpaceX's existing GPU clusters look dated, those contracts become liabilities rather than assets.
And then there's the Musk factor. No matter how well-managed SpaceX has become under Shotwell's operational leadership, the company remains inextricably linked to its founder. Every headline about Tesla, every political controversy, every distraction pulls attention away from the core business. Investors who bought into SpaceX's independence from its CEO's other ventures may be in for a rude awakening.
The IPO was historic. The next chapter might be harder.