$133 Billion From AI. Not Rockets

Tomorrow, SpaceX reports Q3.

The bar is high.

UBS expects $13.8 billion in revenue—well above the Street's best guess. That's 77% more than last quarter.

Let me explain why that number matters.

SpaceX went public in June. Its first report showed $7.8 billion in revenue—a 92% jump from a year ago. That beat Wall Street by close to $1 billion.

Starlink crossed 12 million users. Cash profits hit $3.5 billion.

And yet… the stock sits near $148.

Sounds like a lot. At that price, SpaceX is worth about $1.8 trillion.

But Evercore just raised its target to $275. That's 85% upside. Analyst Kutgun Maral sees SpaceX earning at a $104-billion-a-year pace by December.

So what could push it there?

Two letters. AI.

In February, SpaceX bought xAI—Elon Musk's AI firm—for about $250 billion in stock. Since then, it is seeking to raise $40 billion in new debt for Nvidia chips… and started leasing compute to Google and Anthropic.

TD Cowen projects that AI lease revenue alone hits $66 billion by 2027. And $133 billion by 2028. Those aren't typos.

That's the bull case. A rocket company that earns most of its cash from AI.

Now, I know what you're thinking. “What's the catch?”

There are a few.

SpaceX lost $541 million last quarter. It burned $18.4 billion in capital. It needs Starship to fly at scale to launch its next-gen V3 Starlink birds. And lock-up tranches keep dumping insider shares onto the market through November.

At $1.8 trillion, you're paying for everything to go right. Every AI deal has to close. Every flight has to land. Every user has to stay.

So tomorrow's call—3:30 p.m. Central—matters.

Five things to listen for: total revenue, AI compute capacity, any clues on 2027 pricing, capital spending pace, and the count of lock-up shares still ahead.

It's the most important earnings call in the history of the space economy.

We'll see.

A little-known meeting in 1910 cost the average American family their entire financial freedom.

They met in secret, used fake names, and told absolutely no one where they were going.

On November 22nd, 1910, seven of America's most powerful bankers gathered on Jekyll Island to design the Federal Reserve.

Their goal?

To create a system where they control the money supply and you foot the bill.

Fast forward to today…

That system has created $38 trillion in debt, destroyed 98% of the dollar's purchasing power, and now they're preparing the final move:

Central bank digital currencies (CBDCs).

Don't care about protecting your retirement savings? Skip this.

But if you have at least $100,000 saved and you're concerned about what's coming, you need to hear this.

The same people who met on Jekyll Island have one fatal weakness they didn't plan for…

Decentralized finance.

It's the one thing they can't control, can't inflate, and can't confiscate.

I've prepared a free important broadcast that shows you exactly how to use this technology to escape their system before they force everyone into digital dollars.

You'll learn the three-phase protocol our members use to become their own bank and why the window to do this legally is closing faster than most people realize.

Watch now before this presentation is taken down →

Watch the spread before the print. Wall Street consensus sits near $12.9 billion in revenue for the quarter, with a first post-IPO profit expected; UBS is higher at $13.8 billion. Analysts roughly split the total between about $6.4 billion from AI compute and $4.9 billion from Starlink, with connectivity subscribers around 14 million. The gap between the Street and the bulls… is almost entirely about how fast the AI business is ramping. Watch whether the actual AI segment number lands closer to the cautious or the aggressive estimate — that single line sets the tone for the stock.

Hold a counterweight to the hype. During the first orbital Starship test, one of the engines shut down early, and Morgan Stanley flagged what it called recurring propulsion issues on the vehicle. The flight still reached orbit and deployed its satellites, but the whole AI-and-Starlink growth story leans on Starship flying often and reliably to loft the heavy V3 satellites. An engine that quits early is exactly the kind of thing that can slow a cadence ramp. Watch for any management detail on the engine issue and the next Starship flight date; reliability is the quiet gating factor.

Follow the supply of stock, not just the demand. Since the June IPO, a staggered series of lock-up expirations has released insider shares in tranches, and more are scheduled through the fall. Each wave adds sellable supply, which can cap the stock even when the news is good — a dynamic that has pressured SPCX on several unlock days already. A strong earnings print has to fight that mechanical overhang. Watch how many lock-up shares remain ahead after this report; the fewer left, the cleaner the path if the business delivers.

Watch the 4:30 p.m. Eastern call tomorrow, and listen past the headline revenue number.

Here's why the headline won't be the whole story. Evercore said plainly that this print will be judged less on the quarter itself — where the bulls are already ahead of the Street — than on what management signals about the medium term. In other words… the guidance matters more than the result.

Three forward signals carry the most weight. First, AI compute capacity: SpaceX has pointed toward more than two gigawatts online by year-end, and any hard number updates the whole lease-revenue thesis. Second, Starship cadence and the engine issue, which gate the V3 Starlink rollout. Third, any hint on 2027 pricing for compute, which tells you whether the sky-high AI forecasts are realistic.

The risk cuts both ways. The stock has run up into the report, and expectations are high — a mere in-line quarter could disappoint, even if the business is healthy. A pre-earnings run-up is its own hazard.

There's also the first-profit question. Consensus expects SpaceX's first profitable quarter since listing; missing that, even narrowly, would hand the skeptics a headline.

So watch two things tomorrow: the AI segment revenue and the compute-capacity guidance. Those, far more than the top-line beat, decide how the stock trades into 2027.

Let's keep this simple.

There's a rule in markets that trips up new investors again and again: a stock trades on expectations, not results.

Picture a student everyone knows is brilliant. The whole class expects a 98 on the test. She scores a 94 — an excellent grade by any measure. And yet people are faintly disappointed… because the bar was set at 98.

That's exactly the trap waiting for SpaceX tomorrow.

The company will almost certainly post enormous growth — revenue up sharply, possibly its first profit since going public. On its own, that's a blockbuster. But the stock has already climbed in anticipation, and analysts are modeling a $100-billion-plus future. The bar is set sky-high.

So the danger isn't a bad quarter… it's a merely good one. If SpaceX grows fast but not quite as fast as the most aggressive estimates, the stock can fall — not because the business failed, but because reality landed a notch below the dream.

Here's the deeper point for any investor. When a company is priced for perfection, good news is already baked in. Only a surprise moves the price — and the more optimistic the crowd, the harder it is to surprise them to the upside.

That's why the smart move tomorrow isn't to react to the big revenue headline. It's to compare the result against what was already expected, and to listen to the guidance about next year, where the real debate lives.

A number that sounds huge can still be a letdown. A number that sounds modest can still be a thrill. It all depends on the bar.

So don't just ask how SpaceX did. Ask how it did against what everyone already assumed.

Follow the gap between results and expectations — that gap, not the raw number, is what actually moves a stock.