SpaceX Posts Record Revenue, Half-Billion Loss, Stock Shrugs

SpaceX reported $7.8 billion in revenue for the second quarter of 2026, up 92 percent year-over-year and comfortably ahead of the $6.81 billion analyst consensus. The company also posted adjusted EBITDA of $3.5 billion, up 191 percent from the prior year. And then it reported a $541 million net loss.

For context, that loss narrowed considerably from the $1 billion loss SpaceX recorded in the year-ago period. Progress, in other words. The kind of progress that typically gets celebrated in growth-stage companies where the math still involves red ink. But SpaceX is no longer a private company. It raised $85.7 billion in its June IPO—the largest in history—at $135 per share, which means it now trades in a market where investors have long memories and spreadsheets that work.

Yet here is where the story gets interesting, and by interesting I mean genuinely weird: shares climbed roughly 10 percent during regular trading on Tuesday following the earnings announcement, marking their strongest performance since the IPO. This occurred despite the company burning through $18.4 billion in capital expenditure during the quarter alone, well above the $13 billion analyst forecast. The stock only reversed course once the full context of that spending became apparent in after-hours trading, falling as much as 8 percent on Wednesday as the market collectively decided that perhaps doubling capex above guidance was not, in fact, a feature.

The capex surge reflects AI. SpaceX disclosed $16 billion in AI-related capital expenditure in the quarter, double the prior quarter, with plans to sustain that rate for at least two more quarters. The company's AI segment revenue jumped 247 percent year-over-year to $2.6 billion, driven by heavy infrastructure investment that, you will note, has not yet translated to profitability. The connectivity segment did perform better, with $4.3 billion in revenue and 1.7 million net subscriber additions, but even there the company is clearly plowing everything back into growth.

So why did the stock climb at all? The likely answer is that $7.8 billion in quarterly revenue, with margins expanding at the EBITDA line, carries psychological weight. SpaceX projects a $100 billion annualized revenue run rate by the end of 2026. It has moved internal forecasts of a $1 trillion revenue milestone forward to 2030 from 2031. It has contracted an additional $6.7 billion in cloud services revenue beginning in October. The numbers, if you squint at them the right way, suggest a company that is not merely growing but accelerating.

But here is the part that actually requires attention. On Tuesday and Wednesday, 911.5 million shares unlocked following the earnings release, flooding approximately $100 billion in SpaceX stock into tradable circulation. More than doubled the available float. This is the moment when market conviction typically gets tested. Institutional investors who bought at the IPO suddenly have the ability to exit. Employees with option grants suddenly face a decision. The lock-up expiry is historically the moment when the math finally catches up to the narrative.

The stock climbed anyway. Then it fell. Then it will probably climb again.

What is remarkable is not that SpaceX is burning capital on AI infrastructure—that is what frontier-stage AI companies do. What is remarkable is that the market locked in a 10 percent gain before even processing what the company plans to spend next quarter. The loss narrowed. The revenue accelerated. The capex guidance effectively doubled. And the stock's initial instinct was to celebrate the revenue number and ignore the rest, at least until the spreadsheets started working.

CEO Elon Musk, for his part, declared Starship's heat shield problem 'solved,' while COO Gwynne Shotwell announced plans to build a terrestrial mobile network to compete with AT&T, T-Mobile, and Verizon. Neither announcement was new. Both were offered as forward guidance with the casual certainty of a man who has spent the last two decades proving that certainty, in aerospace and automotive, often outlasts skepticism.

The earnings report itself is defensible. The stock's immediate reaction to it is harder to defend. Markets are supposed to hate surprises, especially negative ones. SpaceX beat on revenue and missed on capex by $5.4 billion in a single quarter. The adjusted EBITDA was spectacular. The net loss was still a loss. The unlock event more than doubled available shares.

And somehow the stock still climbs. Either the market has exceptional conviction in SpaceX's ability to monetize $16 billion in quarterly AI spending before it becomes a problem, or it has exceptional amnesia about how previous tech companies have responded to similar capital calls. Neither possibility is reassuring.