Goldman Sachs has modelled SpaceX's financials ahead of its IPO, and the numbers are striking even by the standards of a company run by Elon Musk.
Revenue is expected to more than double this year to around $38bn, with cash burn also more than doubling to roughly $30bn. By 2030, Goldman projects revenue of $474bn and cumulative cash burn of $350bn.
The majority of that burn is not rockets. It is not Starlink. It is AI. Approximately 75% to 80% of SpaceX's projected capital expenditure is allocated to AI infrastructure, including data centres, the Colossus compute facility and the Terafab joint project with Tesla.
The company hopes to raise $75bn through the IPO. The models do not detail how much SpaceX will need to raise after that, but the implication is clear: the IPO is the beginning of the fundraising, not the end.
Musk has played down the prospect of further equity raises and the dilution they would bring. But a company burning $350bn over five years while generating revenue that only approaches breakeven in the later years of the forecast will need to access capital markets repeatedly.
The Anthropic wrinkle
Part of the revenue acceleration comes from SpaceX's deal with Anthropic, which is renting compute capacity at the Colossus data centre for $1.25bn per month through May 2029. That relationship is generating meaningful revenue but also highlights how dependent the financial model is on a single customer commitment.
The credibility question
One large buyside investor told reporters the numbers were "much larger than expected." The forecasts are based on verbal communications between analysts and bankers and are subject to multiple layers of assumption.
After the IPO, the banks will publish these projections and SpaceX will be measured against them. Goldman's models are not promotional material. They are the standard the company will be held to.
A $474bn revenue target by 2030 would make SpaceX one of the largest companies in the world by revenue. A $350bn cash burn would make it one of the most capital-intensive businesses in history.
Both numbers could be right. The question is whether investors are comfortable funding the gap between them.
