SpaceX revenue jumps 92% but $18.3bn spending drives a loss

By Nathan Cross · 2026-08-04

Revenue nearly doubles, but the bill arrives first

SpaceX posted $7.8 billion in revenue for the second quarter of 2026, a 92% increase on the same period a year earlier, the largest revenue figure of any company in a debut quarterly report as a public company.^[1] The number confirmed what private fundraising rounds had long signalled: the underlying business is genuinely enormous. What it also confirmed is that SpaceX is spending faster than it is earning.

Capital expenditure for the quarter reached $18.3 billion, a rise of more than 550% year-on-year, turning what would otherwise have been a profitable quarter into a $541 million net loss.^[1] For the first half of 2026 combined, the net loss reached roughly $2 billion.^[1] Investors are now sitting with a company growing at near-double speed while burning capital at a rate that dwarfs its losses.

Three segments, three different stories

The earnings release, published after market close on 4 August 2026, gave public markets their first full view of SpaceX's segment structure.^[1] SpaceX reports across three divisions: space launch, connectivity and artificial intelligence, and each tells a materially different story about where the business stands and where it is headed.

The space launch division, which covers Falcon 9, Falcon Heavy and Starship operations, generated $962 million in revenue against a $542 million segment loss.^[1] Starship development costs, which remain substantial, sit inside this segment alongside the capital demands of running heavy-lift systems for commercial, civil and government customers.

Connectivity is carrying the business right now. Starlink generated $4.29 billion in revenue and $1.6 billion in operating profit, making it the only division currently delivering meaningful positive returns.^[1] The satellite internet service now counts 12 million subscribers, double the total from a year earlier, a growth rate few connectivity businesses of any kind have matched at this scale.^[1]

The AI segment is the sharpest contradiction in the report. It produced $2.56 billion in revenue while recording a $1.2 billion operating loss, a gap that reflects the extreme cost of building out the compute infrastructure required to run the Grok AI model and other AI applications SpaceX is developing.^[1] The segment is early, expensive and scaling fast.

Starlink carries the weight

Starlink's subscriber doubling is not incidental to the investment case; it is the investment case, at least for now. A connectivity segment producing $1.6 billion in profit on $4.29 billion in revenue is a high-margin subscription business growing at a pace most telecoms would struggle to model.^[1] The results do not directly answer how much of the $18.3 billion in capex flows into Starlink constellation expansion versus AI infrastructure versus Starship development.

Strip out Starlink's $1.6 billion in operating profit and the rest of the business posted a combined operating loss of roughly $1.74 billion for the quarter, before accounting for the capital expenditure figure.^[1] Starlink is not just a product line; it is the financial engine making the broader programme viable in the near term.

CFO flags novel territory, Musk sets a trillion-dollar target

Bret Johnson, SpaceX's head of finance, addressed the capital question directly, telling the call that capital spending would stay at a "very similar" level for the rest of the year.^[3] Johnson's framing is a candid acknowledgment that the spending trajectory is not a near-term aberration but a structural feature of how SpaceX is building its AI and infrastructure divisions.

Elon Musk, speaking on the earnings call, set the ambition at a scale that invites scepticism and attention in equal measure. Musk said SpaceX would likely reach $1 trillion in revenue by 2030, a year earlier than he had put it six weeks before.^[2] Going from roughly $31 billion annualised at the current run rate to $1 trillion in four years implies a compound annual growth rate that would rank among the fastest in corporate history.

Musk's track record of setting targets that arrive late but do eventually arrive gives the forecast more credibility than it might otherwise carry, though public market investors are accustomed to pricing the gap. The CFO's description of capital spending as a feature rather than a bug is consistent with how SpaceX has always operated: treat near-term losses as the cost of capturing a market that does not yet fully exist.^[3]

Share price punishes the gap between growth and profit

SpaceX completed what was described as the largest initial public offering in history on 12 June 2026, listing its Class A common stock on the Nasdaq under the ticker SPCX at $135 per share and raising $75 billion.^[1] Shares peaked at $176 in June before sliding as concerns over the scale of AI and infrastructure spending mounted through July.

After the earnings release, shares fell roughly 9% in after-hours trading, extending a decline that has kept the stock below the IPO price for weeks.^[1] The market's reaction is a clear read on investor priorities: 92% revenue growth is impressive, but an $18.3 billion capex bill against $7.8 billion in revenue means SpaceX is spending at more than twice the rate of its income.

SpaceX is simultaneously running three capital-intensive programmes: constellation expansion, Starship development and AI infrastructure, at a pace that requires it to keep demonstrating that each will eventually produce returns comparable to what Starlink is already delivering. The first quarterly report as a public company confirms the scale of the ambition; the timeline for when that ambition pays out in profit remains unconfirmed.

Nathan Cross

Nathan Cross writes about big technology companies and the economics of a generation coming up behind them. He is interested in scale, and in who pays for it.