Graphic & Text | Sister Tang

SpaceX, the company carrying Elon Musk's vision of 'landing on Mars,' has finally submitted its prospectus, laying out nearly all its details in public for the first time. Out of curiosity—and with a critical eye—I downloaded and reviewed the original document as soon as it was posted on the SEC website on May 20th.

I must say, Musk has truly outdone himself in terms of transparency—it's refreshing.

For the first time in over two decades of commercial spaceflight, a major player of this scale has laid bare its financials to the SEC. With a valuation nearing $400 billion, SpaceX is the world's highest-valued private company and led the industry in rocket launches last year. While the market has long had a sense of its scale, details were pieced together only through fragmented sources like valuation rumors, Starlink subscriber numbers, and Musk's ownership stake. A complete financial picture has never been public ly availablebefore.

Even fresher is xAI. While U.S. frontier AI companies have dominated headlines in recent years, none had submitted a full financial disclosure to the SEC—until now. This time, xAI is included in SpaceX's consolidated financials, making this prospectus the first to reveal exactly where a cutting-edge AI company is burning cash and to what extent.

With that in mind, I had several specific questions before diving in: How large is Starlink's satellite internet business now? Is the rocket division profitable? How much is xAI burning? What is Musk's exact ownership stake? And so on.

After reviewing the document, it's clear that these seemingly disparate business lines are woven together by Musk through various capital maneuvers. The first test for this behemoth post-IPO may not be whether liquidity can support financing, but rather how the market should value it in the first place.

01 Not Just Rockets

From an overall perspective, SpaceX is not an especially large company—not on the scale of the 'Magnificent Seven.' In 2025, its consolidated revenue reached $18.7 billion. The prospectus breaks the company down into three segments: aerospace launch (Space), satellite internet (Connectivity), and artificial intelligence (AI).

The aerospace launch segment covers rockets and spacecraft: Falcon 9, Falcon Heavy, Dragon spacecraft, Starship R&D, plus government contracts with NASA, the U.S. Space Force's NSSL, and the U.S. Department of Defense. The satellite internet segment includes Starlink broadband and the military satellite business called Starshield. The AI segment houses xAI's Grok large model, the X platform (formerly Twitter), and two computing centers.

It's worth noting that the AI business was only consolidated into SpaceX in February 2026. Prior to that, xAI had first merged X into itself in March 2025, and then SpaceX merged xAI in February 2026. Since all three companies were under Musk's control, this was accounted for as a 'reorganization under common control,' requiring the financial statements to be restated retroactively for the past three years.

The most visible impact is on the 2023 AI business: while reported revenue was about $3 billion, a footnote reveals that nearly all of it came from advertising, subscriptions, and data licensing on X (i.e., Twitter)—hardly what we now consider AI revenue.

Among the three segments, only Starlink is profitable. In 2025, the satellite internet business generated $4.4 billion in operating profit, with a margin of nearly 39%, accounting for all of SpaceX's operating profit. Aerospace launch had a $660 million operating loss in 2025, while AI incurred a $6.4 billion loss. The AI segment's annual operating loss far exceeded the aerospace launch segment's full-year revenue of $4.1 billion.

Combining the three, SpaceX reported a $2.6 billion operating loss and a $4.9 billion net loss for 2025. This pattern is not short-term: from its founding through March 2026, the company's cumulative deficit (net losses minus profits over the years) reached $41.3 billion. Only in 2024 did it record a net profit of $791 million, compared to losses of $4.6 billion and $4.9 billion in 2023 and 2025, respectively.

Let's first look at Starlink, the mature business. With nearly 9,600 satellites in orbit, it accounts for about 75% of all maneuverable satellites globally. Subscribers nearly doubled in a year, from 5 million in Q1 2025 to 10.3 million in Q1 2026. The business is also actively reducing revenue per user: ARPU fell from $91/month in 2024 to $81 in 2025. According to the prospectus, the company prioritizes scale expansion and efficiency gains over ARPU growth.

The other two segments are still burning cash, but in different ways. Aerospace launch losses stem mainly from Starship. While Falcon 9 remains profitable, of its 165 launches in 2025, only 43 were paid external customer missions—the remaining 122 were for Starlink deployments. Starship, the next generation, is entirely in the investment phase: SpaceX has self-funded over $15 billion in R&D for Starship, but as of Q1 2026, all flight tests have been internal missions without any paid payloads.

AI's cash burn is primarily due to infrastructure investment. In Q1 2026 alone, the AI segment spent $7.7 billion on capital expenditures—more than three times the $2.4 billion combined for aerospace launch and satellite internet in the same period.

It's also worth mentioning that even the brief profitability in 2024 had little to do with core operations. Of the $791 million net profit, $955 million came from unrealized gains on SpaceX's 18,712 bitcoins, which had not moved in three years since acquisition. Another $549 million was a book gain from tax adjustments based on expected future profits—unrelated to operations.

Together, these two items totaled $1.504 billion, exceeding the 2024 net profit itself. Excluding them, the company was still operating at a loss of over $700 million in 2024. In 2025, these adjustments reversed, bringing the consolidated net loss back to $4.9 billion.

In summary, SpaceX today is a company where 'Starlink supports all profits while the other two segments burn cash significantly.' Its only accounting profit in the past three years was propped up by bitcoin gains and one-time accounting adjustments. But understanding this company requires looking beyond its business operations—its web of relationships with Musk's other ventures is another critical issue.

02 Musk's Empire Ambitions

Musk is such an exceptional entrepreneur that many investors pay little attention to the fundamentals of his companies. That said, there are clear internal synergies between the first two segments of SpaceX's business: aerospace launch delivers satellites for Starlink, which in turn provides stable launch demand for the aerospace division. The synergy between AI and the other two segments, however, is far more tenuous—so tenuous that only a grand vision of 'space energy + computing power' can loosely connect them.

But such 'tenuous connections' are par for the course in Musk's business empire.

From a governance perspective, SpaceX has no built-in balancing mechanisms. Musk personally holds a 12.3% economic interest but controls 85.1% of the voting rights after weighting. This means that upon listing, SpaceX will directly invoke Nasdaq's 'controlled company' exemption, eliminating the need for a majority of independent directors or independent nomination and compensation committees.

The prospectus also discloses two key points: the company explicitly will not purchase key-person life insurance for Musk, meaning no insurance payout would occur if he were incapacitated; it also clarifies that Musk does not dedicate all his time and energy to SpaceX—he remains CEO of Tesla and can hold any other position he desires.

Betting entirely on Musk is the most critical gamble for this company post-IPO. Moreover, SpaceX has another characteristic: relatively blurry boundaries with Musk's other companies.

Take Tesla as an example. While Tesla has never directly invested in SpaceX, it is now a SpaceX shareholder: in January 2026, Tesla committed $2 billion to xAI's Series E funding. When SpaceX merged xAI in February, Tesla's xAI equity was automatically converted into SpaceX Class A shares at the merger consideration.

The company's charter also includes a specific waiver for business opportunities: it forgoes certain business opportunities that may arise for Musk and certain directors, who are under no obligation to present these opportunities to SpaceX first. If a business opportunity arises for both Musk and SpaceX, the charter allows Musk to legally direct it to another company under his control, with no statutory right for SpaceX to be informed first.

Together, these two points mean that SpaceX has no traditional independent boundaries with Musk's other companies—they form a network where capital and opportunities flow internally.

This network effect makes valuation highly nuanced. When valuing an ordinary public company, investors typically look for external market transactions as reference anchors (recent M&A prices in the industry, unit prices of third-party customer contracts, discounts in marketed equity offerings, etc.). Among the specific valuation reference points in SpaceX's prospectus, only one truly qualifies as a 'market price anchor.'

In September 2025, SpaceX signed a spectrum acquisition agreement with U.S. satellite communications company EchoStar to secure more spectrum resources for Starlink. The total consideration was $19.6 billion, with $11.1 billion paid in SpaceX Class A shares at an issue price of $42.40 per share (post-split). Backing into the pre-IPO fully diluted share count, this implies a total valuation of just over $500 billion—higher than the 'nearly $400 billion' private valuation previously attributed to SpaceX.

Thus, at this stage, there are roughly three viable approaches to valuing SpaceX:

1. Directly use EchoStar's anchor. The $42.40/share price implies a total valuation of about $500 billion, serving as a relatively hard reference for the IPO's initial pricing—even though the price came from negotiations between SpaceX and EchoStar, not a public market auction.

2. Use a sum-of-the-parts (SOTP) valuation. Starlink, being mature, can be valued using multiples from consumer internet or telecom service providers; aerospace launch can be compared to traditional space contract companies; and the AI segment will likely be tied to xAI's previous private valuation. However, substantial intersegment transactions (Falcon 9 launches for Starlink, GPU leasing through related parties for AI, etc.) complicate this approach—valuing each segment separately and adding them up may not be more accurate than a holistic valuation.

3. Abandon independent company valuation frameworks. Treat SpaceX as the core node of Musk's business network, valuing it similarly to Tesla: much of the valuation reflects Musk's execution ability, the capital operation capabilities of his network, and whether the grand narratives of 'Mars colonization + orbital computing' can materialize.

Currently, the market will likely lean toward the third approach as the mainstream valuation method.

03 Conclusion

From selling PayPal to eBay, taking Tesla from near-bankruptcy to the world's most valuable automaker, reducing human launch costs by an order of magnitude with SpaceX, to Starlink achieving global coverage in just a few years—a string of 'science fiction turned reality' accomplishments justifies the market's bet on Musk.

But valuing Musk and valuing an independent company are two different matters. Valuing an independent company, however difficult, still has relatively certain cash flows and business logic as anchors. Valuing an individual essentially means betting on Musk's execution ability over the next decade, his attention allocation, political risks, health status, and even far-off promises like 'whether Mars will ever be built.'

The uniqueness of SpaceX's IPO lies here: it moves the task of 'valuing Musk's business network' from the private markets to the public sphere for the first time. Henceforth, valuing SpaceX will be closer to valuing Musk himself. That said, public market patience may not match that of private investors. If delivery timelines fall short of expectations, valuation corrections could be far more violent than operational adjustments.

Thus, this IPO is also a new test for Musk himself.