SpaceX IPO: Musk's Gain, Your Pain?

The impending public offering, heralded as the largest in history, signifies what some observers are calling the ultimate manifestation of financial nihilism.

For a considerable period, no IPO document appeared as ill-conceived as WeWork's, until Elon Musk's filing to take SpaceX public. While WeWork was largely perceived as a misstep, SpaceX presents a more profound concern. Should Musk and his financial advisors succeed, investors risk becoming unwitting stakeholders in a venture built on questionable premises.

Key financial details surfaced well before the official S-1 filing. Rumors suggested a valuation exceeding $1 trillion, despite the company reporting nearly $5 billion in losses last year. SpaceX's self-proclaimed total addressable market (TAM) for AI applications was stated as an astonishing $28.5 trillion. For context, the entire gross domestic product of the United States was just over $24 trillion, according to the St. Louis Fed.

One might facetiously accept Musk as a visionary leader for a collection of abstract digital entities.

While these figures appear nonsensical, their impact might be negligible in a market driven by influence. Musk, a pioneer among financial influencers, oversees Tesla, an electric car company that trades at over 300 times earnings, starkly contrasting Ford and Toyota, which trade at approximately 11 times earnings, or even Nvidia, a highly profitable company, at 33 times earnings. Tesla's status as a "meme stock" positions SpaceX to potentially follow suit. The fact that SpaceX combines a space company, an AI company, and a social network into one entity is irrelevant in the context of a meme stock, which often defies conventional financial logic.

The narrative surrounding the offering suggests a return to a 2015-era perception of Musk's altruistic motivations, particularly his ambition to advance humanity into space. The S-1 filing heavily emphasizes Musk's "messianic mission to 'extend the light of consciousness to the stars'," a phrase appearing seven times, with "light of consciousness" mentioned three additional times. The document includes an artist's rendition of "Life on Mars," depicting inhabitants as polygons, prompting a wry comparison to the design of the Cybertruck.

The grandiosity of this mission makes "the power of We," as espoused by WeWork's Rebekah Neumann, seem rather modest in comparison.

Musk's strategy leverages the fervent dedication of his retail investor base, evidenced by 30 percent of the IPO being reserved for them. For institutional investors, the situation presents a "Keynesian beauty contest": if one anticipates that a loyal investor base will purchase shares regardless of underlying value, and that Nasdaq rule changes might expedite index inclusion, participating in the IPO could be rational. The expectation is that the stock price will rise, irrespective of fundamentals. As economic theory suggests, "The market can stay irrational longer than you can stay solvent."

This offering transcends a typical company IPO; it embodies the broader trend of financial nihilism. While platforms like Robinhood capitalized on this phenomenon, SpaceX is distinct and arguably more problematic, raising concerns about its widespread market implications.

Despite its reputation for rocket building, the S-1 filing for SpaceX, known for its extensive references to the Moon (74 times), Mars (63 times), and "and beyond" (13 times), reveals a surprising emphasis on AI. By its own admission, $26.5 trillion of its $28.5 trillion TAM is attributed to AI applications. Notably, these optimistic estimates exclude the Russian and Chinese markets.

In 2025, approximately $13 billion, or two-thirds, of SpaceX’s capital expenditure was directed towards AI development. However, the company’s AI division reported a $6 billion operating loss on revenues of just $3.2 billion. This stands in stark contrast to Anthropic, which is projected to achieve an operating profit of $559 million in the second quarter of this year.

Regarding Grok, the S-1 describes it as a "truth-seeking AI model ... which has emerged as one of the world’s most advanced frontier models." This assertion is somewhat tempered by Musk's own statement in March that "xAI was not built right first time around, so is being rebuilt from the foundations up." Is it truly one of the most advanced models? It appears to be "distilled from them," at best.

Further complicating matters are the terms of xAI’s recent deal with Cursor. SpaceX announced a potential acquisition of the AI coding company in April, aiming to compete with enterprise AI products from Anthropic and OpenAI. If the deal proceeds, existing shareholders face a dilution of $60 billion. If it fails, SpaceX must pay Cursor $1.5 billion and grant access to over $8 billion in compute resources, indicating a weak negotiating position for SpaceX.

Greg Brockman and Ilya Sutskever, co-founders of OpenAI, expressed in 2018 that Musk "really hasn't done his homework [on] AI / AGI."

The filing also discloses ongoing investigations into Grok's generation of nonconsensual sexualized images, including those of children, with three specific lawsuits mentioned, two of which seek class-action status.

The timing of this filing is noteworthy, coming shortly after Musk lost his lawsuit against Sam Altman and OpenAI, a legal battle that primarily underscored his perceived deficiencies in AI. The 2018 assessment by OpenAI co-founders Greg Brockman and Ilya Sutskever—that Musk "really hasn't done his homework [on] AI / AGI"—appears to remain relevant.

SpaceX reported that its AI unit, encompassing X and xAI, generated only $818 million in revenue in the first quarter of 2026. This is significantly less than Twitter's revenue of $1.2 billion in the first quarter of 2022, prior to Musk's acquisition, highlighting a considerable decline in performance under his stewardship.

Shifting focus to space, the company's grand plans largely hinge on Starship, a rocket prototype known for its susceptibility to "unexpected explosions." Starship is critical for launching heavier versions of Starlink satellites, some of which are currently awaiting deployment, and for fulfilling various NASA and other government contracts.

At the time of the filing, Starship prototypes had barely achieved more payload capacity than SpaceX's lowest-end rocket, the Falcon 9—a detail curiously absent from direct disclosure. Instead, the S-1 proclaims: "Starship V3 is designed to deliver 100 metric tons to space in a fully reusable configuration while enabling rapid turnaround times." The feasibility of "fully reusable" is questioned, as it typically implies capacity limitations.

Rather than providing concrete details on Starship's actual achievements, the S-1 offers forward-looking statements: "We expect to commence deploying our next-generation V3 satellites, designed to offer one Tbps of downlink capacity per satellite, using Starship in the second half of 2026. We expect that a single Starship launch will be capable of deploying up to 60 V3 satellites."

A "sort-of successful" V3 launch did occur two days after the filing, on May 22nd. While it avoided explosion and reached space, one engine failed, and the booster exploded upon return. The primary challenge remains satellite deployment. This test flight deployed only 20 dummy satellites, falling short of the stated goal of 60. The lack of clear disclosure to investors regarding the actual operational status of such a pivotal rocket is a significant concern.

An irksome detail emerges upon closer inspection of the S-1's timeline and figures. Assuming the S-1's timeline is accurate—a questionable assumption given SpaceX's history of delays—the V3 satellites, weighing 2,000kg each according to an FCC filing, would total 120 metric tons for 60 units. If Starship performs as advertised with a 100-metric-ton capacity, it would still be insufficient to launch all 60 V3 satellites simultaneously.

"It’s Musk math," commented Chris Quilty, co-founder and co-CEO of Quilty Space, an analytics firm. While Quilty suggests the V3 satellites might not weigh precisely 2,000kg, and efficient stacking could make 60 satellites feasible, the lack of detailed information in the filing is problematic, especially given Starship's central role in the company's plans.

The launch business saw a decline in revenue by over a quarter in the first three months of this year, primarily due to fewer customer launches. The premise that cheaper rocketry would lead to a surge in demand for space launches appears challenged, with the S-1 suggesting that SpaceX itself is its largest launch customer.

This section also delves into speculative ventures, including Musk's 2017 vision of Starship as a point-to-point transportation system, in-space manufacturing, space tourism, and asteroid mining. Starship's first crewed flight, initially projected for two years ago, has yet to materialize. Such aspirational elements, despite their unproven nature, are included in the S-1, perhaps to appeal to a market segment drawn to futuristic concepts.

Starlink, while a cost center, has performed better than the AI division. Expenditures for Starlink were $3 billion in 2025 and $930 million in the first quarter of 2026. The success of Starship is intrinsically linked to Starlink's profitability and, by extension, SpaceX's overall financial health.

The sole viable business identified within this IPO is Starlink, the satellite internet provider, which generated over $11 billion in revenue last year. Quilty confirms, "The reality is that Starlink is the cashflow machine that will fund the xAI and SpaceX Starship business."

Starlink is indeed a robust business, even though its reported revenue in the S-1 falls short of both Morgan Stanley’s projections and earlier figures provided to potential investors. This discrepancy may be attributed to heavy discounting, which led to a roughly 25 percent decline in revenue per subscriber. The filing offers no insight into the potential impact on subscriber retention.

Starlink's emergence was opportune, coinciding with the decline of traditional TV and its impact on geostationary satellite TV providers. Quilty notes its successful expansion from consumer markets to enterprise customers, securing a significant lead over competitors. While some critics point to the absence of deorbiting costs in the S-1, Quilty explains that at Starlink's current lower operational orbit (380km above Earth), satellites are "functionally self-cleaning" due to faster atmospheric decay, which takes weeks to months at 300km, compared to years above 700km.

SpaceX has strategically integrated its nascent AI ambitions with its successful Starlink business, envisioning space-based data centers that beam data via Starlink. The core argument for this integration is that space offers easier access to solar energy, thereby addressing AI's energy consumption bottleneck. (Presumably, the vacuum of space would also facilitate cooling, often a contentious issue on Earth due to water usage, similar to how the James Webb Space Telescope radiates heat.) The filing states an ambitious goal: "Our goal over time is to launch 100 gigawatts of compute to space each year."

Notably, the S-1 features variations of the phrase "The Sun contains approximately 99.8% of the solar system’s energy" four times. This level of precise detail, while seemingly scientific, can be perceived as an affected attempt to project an engineering-driven image, rather than simply stating "almost all" or "more than 99 percent."

This rhetorical flourish highlights a cynical interpretation of the filing: Musk's strategy of presenting grand, futuristic promises that are difficult for external parties to scrutinize for plausibility, unlike more conventional Earth-based data center proposals.

Let's consider the terrestrial implications.

Despite his stated mission to save humanity through interstellar expansion and his car company's foundation in renewable energy, Musk oversaw the acquisition of an additional $2.8 billion in polluting gas turbines to power SpaceX’s data centers, a detail relegated to a footnote. SpaceX is currently facing litigation over its turbine usage, operating 46 units despite permits for only 15. These turbines are even listed in the risk factor section, indicating that an injunction or permit revocation could severely impact the company's struggling AI division.

Space analyst Chris Quilty holds a more optimistic view on space-based data centers,