Starlink Prints Money, Starship Burns It, and AI Eats the Rest — Decoding SpaceX

SpaceX just became the most valuable company to ever hit public markets. Three weeks later, SPCX is trading at $145.30 — barely above its $135 IPO price, and a full 35.6% below the post-IPO high of $225.64. Everybody’s asking: what the hell happened?

The answer starts with a simple truth Wall Street doesn’t love to say out loud. SpaceX isn’t one company. It’s three. And they’re all fighting for oxygen from the same balance sheet.

Let’s start with the good news — because there’s plenty of it. Starlink generated $11.4 billion in revenue last year, with $4.4 billion in operating income. That’s a 39% operating margin. Ten point three million subscribers, 85% on recurring contracts. This is the closest thing the space industry has to a SaaS business, and it’s minting money.

Segment two: Launch services. Falcon 9 is the undisputed workhorse of global launch — 166 missions in 2025 alone, at roughly $67 million per flight. Competitors charge $200 million or more for comparable payloads. That pricing advantage is a moat you could see from orbit. The launch business generates about $4.7 billion in revenue. It’s a mature, reliable cash engine, but it’s not growing fast enough to justify a $1.9 trillion valuation.

So where’s the growth? That brings us to segment three — the xAI complex. Grok, the X platform, the recently announced $60 billion acquisition of Cursor. Combined revenue is roughly $3.2 billion. Sounds impressive until you realize this division is burning through about $14 billion a year. AI infrastructure is expensive — data centers, GPUs, power, talent. SpaceX’s $20.9 billion in capital expenditures last year was largely driven by this buildout. Free cash flow came in at negative $14.121 billion. That’s not a rounding error — that’s a hole.

Here’s where the math gets uncomfortable. SpaceX’s total revenue was $18.674 billion last year. Market cap is roughly $1.914 trillion. That’s 102 times sales. The company posted a net loss of $4.937 billion in FY2025 — despite Starlink’s operating profits. Operating cash flow was positive at $6.785 billion, but that’s before the capex monster. Forward P/E: 167.6 times. For a company that has never reported a GAAP-profitable year.

None of this was a secret when SPCX debuted on June 12. The IPO priced at $135, and the stock ripped to $225.64 on pure momentum. The quiet period ended on July 7, and ten banks initiated coverage with an average price target of $242 — 67% upside from here. That same day, SPCX joined the Nasdaq-100, triggering $4.3 billion in forced buying from index funds. The stock fell 6.83%. Welcome to “sell the news” at its finest.

But the quiet period and the index inclusion are warmup acts. The real event is the lock-up expiration, expected in late July or early August. There are more than 12 billion insider shares — roughly 96% of the total float — that could theoretically hit the market at once. Only 281 million shares are publicly traded right now. That’s a 4.3% float. When lock-up expires, the supply dynamics shift in a way that makes most IPO lock-ups look like a weekend yard sale.

Then there’s Starship. COO Gwynne Shotwell has publicly stated the vehicle hasn’t reached orbit yet, with a target of “end of 2026.” Starship is the entire bull case for SpaceX’s next phase — Mars cargo, Starlink V3 deployment, deep space missions. If it doesn’t fly this year, the narrative shifts from “pre-revenue growth driver” to “delayed R&D project with no near-term ROI.” Dangerous territory when your stock trades at 102 times sales.

So what does all this mean for an investor looking at SPCX at $145? Starlink is an absolute machine, generating 39% margins on $11.4 billion in recurring revenue. The launch business prints cash at industry-leading costs. The moat around SpaceX’s core operations is wider than any competitor’s by a mile — 10,200 Starlink satellites in orbit versus the hundreds anyone else can claim, boosters flying twelve-plus missions, vertical integration from silicon to satellite to network. Those advantages aren’t going anywhere.

But here’s the other side. The AI/xAI division is a cash incinerator at a scale that could stress even SpaceX’s balance sheet. The lock-up overhang is real and it’s coming within weeks. Starship hasn’t reached orbit yet. The first earnings report as a public company isn’t expected until September 2026 — that’s when the market gets its first real look under the hood. S&P 500 eligibility doesn’t come until June 2027. Between now and then, there are a lot of potential speed bumps.

Three companies, one stock, zero earnings. The Starlink business alone is worth a fortune. But the AI play is an open question, and the lock-up is a loaded gun. For now, SPCX at $145 might be a bet on Starlink’s cash flow winning out over the xAI burn rate. Just know what you’re buying — and what you’re not.

Disclosure: The Signal holds no position in SPCX. Positions may change. This is not financial advice.