SpaceX filed its S-1 with the SEC on 20 May 2026 and applied to list on Nasdaq and Nasdaq Texas under the ticker SPCX. An offering size of USD 75 billion or more is expected, at market cap of approximately USD 1.75 trillion, with the roadshow and listing following closely. At that scale the deal will eclipse the previous record holders on both metrics that matter: gross proceeds and market capitalisation at offering, held by Saudi Aramco, which offered USD 26 billion in 2019. The S-1 puts hard numbers on a business that was, until last week, the world’s most-discussed unlisted company: USD 18.67 billion of 2025 revenue, a USD 4.94 billion net loss in 2025, more than 22,000 employees, and a controlling Class B stake that leaves Elon Musk with majority voting power post-IPO.

SpaceX S-1 projects a USD 28.5 trillion quantifiable total addressable market across Space, Connectivity and AI (China and Russia excluded), expecting Starship to commence payload delivery to orbit in the second half of 2026, with plans to begin deploying orbital AI compute satellites as early as 2028, and has set a long-term target of 100 GW of annual compute power deployed to orbit. That would require launching roughly one million metric tons to orbit per year, more than 130 times the cumulative mass SpaceX has launched in 22 years. The “Future Markets” list outlines a lot of different pathways, ranging from space tourism and in-orbit manufacturing to passenger and cargo transport to the Moon and Mars and asteroid mining.

SpaceX’s Estimated TAM by Segment

Source: SpaceX EDGAR S-1, InterCapital Research

Space is the operating moat, and the foundation on which everything else SpaceX does has been built. The company has delivered more than 80% of global mass to orbit each year since 2023, against a cumulative 7,400 metric tons launched across its 22-year history at a 99% mission success rate. The launch-cadence chart is the cleanest expression of the gap to peers: SpaceX flew more orbital missions in a single quarter of 2025 than Rocket Lab, the closest credible challenger, flew across all of 2025 (21 launches), and more than United Launch Alliance, Blue Origin and Arianespace combined over the same period. NASA’s own estimate placed the first Falcon 9 at USD 2,700 per kilogram to orbit against a historical industry baseline of USD 18,500 per kilogram. Starship V3, the fully reusable successor designed to carry 100 metric tons to orbit, is the next leg, and the company aims to take launch costs down by another 99% from that historical baseline.

SpaceX Space Segment Mass to Orbit and Launches (2023-2025, tons, number of launches)

Source: SpaceX EDGAR S-1, InterCapital Research

Space revenue grew from USD 3,557 million in 2023 to USD 4,086 million in 2025, a 7% three-year CAGR, while the operating result turned more negative as Starship R&D ramped from USD 1,538 million in 2023 to USD 3,004 million in 2025. Segment Adjusted EBITDA stayed positive at USD 653 million in 2025 but the operating line shows a USD 657 million loss, a direct consequence of expensing essentially all Starship development against current-period revenue. This is the segment most likely to look better on the income statement once Starship enters commercial cadence and the R&D line normalises.

SpaceX Space Segment Financials (2023 – 2025, USDm)

Source: SpaceX EDGAR S-1, InterCapital Research

Connectivity’s Starlink ended Q1 2026 with 10.3 million subscribers across 164 countries, territories and markets, compounding from roughly one million in early 2023 to ten times that in three years. ARPU has been the surprise of the ramp: contrary to the typical consumer-broadband pattern where unit pricing compresses as the user base scales, Starlink has held its blended ARPU broadly stable as it layered higher-tier business, maritime and aviation tiers on top of the residential base. The Starlink Mobile satellite-to-mobile network adds 7.4 million monthly unique devices across 30 countries on top of the broadband subscriber base. The combined subscriber and ARPU shape is the part of the Starlink story that distinguishes it from a typical consumer-broadband growth curve.

SpaceX Starlink Subscribers and ARPU (M, USD/month)

Source: SpaceX EDGAR S-1, InterCapital Research

The Connectivity segment’s three-year financial trajectory is the cleanest growth-into-profitability story in the prospectus. Revenue went from USD 3,869 million in 2023 to USD 11,387 million in 2025, a 2.9-times move in two years and a 72% three-year CAGR. The operating result moved from low single digits to USD 4,423 million in 2025, a 39% segment operating margin, and segment Adjusted EBITDA reached USD 7,168 million on a margin north of 60%.

SpaceX Connectivity Segment Financials (2023 – 2025, USDm)

Source: SpaceX EDGAR S-1, InterCapital Research

AI segment was formed by the early-2026 acquisition of xAI, which itself had previously absorbed the X social platform, so the historical revenue line is a composite that does not yet reflect the destination configuration. The infrastructure the capex line is building is the COLOSSUS and COLOSSUS II clusters, brought online in 122 and 91 days respectively against a roughly two-year industry benchmark for a 100 MW greenfield data centre. In May 2026 the company signed cloud services agreements with Anthropic for USD 1.25 billion per month through May 2029, and in March 2026 announced “Terafab” with Tesla and Intel, targeting one terawatt of compute hardware production per year.

SpaceX Nameplate Compute Draw (2023 – 2025, GW)

Source: SpaceX EDGAR S-1, InterCapital Research

AI segment revenue was effectively flat across 2023-2025 as the X platform passed through a 2024 ad-revenue trough and a 2025 recovery, but the operating result and Adjusted EBITDA moved sharply more negative as integration, training compute and headcount investment ramped. The bull case is that the third-party compute contract pipeline (Anthropic at the top of the list) converts USD 12,727 million of 2025 capex into infrastructure revenue (USD 3,201 million in 2025) with economics comparable to the Connectivity segment today. The bear case is that an internal frontier-model arms race absorbs the spend without producing returns. The first full year of post-IPO disclosure will let the market form a view.

SpaceX AI Segment Financials (2023 – 2025, USDm)

Source: SpaceX EDGAR S-1, InterCapital Research

Rolled up to the consolidated level, the three-year trajectory is a USD 10.4 billion to USD 18.7 billion revenue ramp, an operating loss that has widened to USD 2.6 billion as AI capex flowed through the income statement, and an Adjusted EBITDA line that has remained positive and grown to USD 6.6 billion in 2025 on the back of Connectivity. The price the market is being asked to pay for that profile, at the press-rumoured USD 1.75 trillion valuation, works out to roughly 93 times trailing-twelve-month revenue. That multiple is high for an industrial-and-connectivity composite but unremarkable for an AI infrastructure composite, and the segment weights of any institutional model will swing it materially.

SpaceX Consolidated Financials (2023 – 2025, USDm)

Source: SpaceX EDGAR S-1, InterCapital Research

The historical comparison sets the scale. The Saudi Aramco 2019 listing raised USD 25.6 billion on the Tadawul, USD 29.4 billion including the over-allotment. Alibaba’s 2014 NYSE deal raised USD 21.7 billion. The SoftBank Corp 2018 listing in Tokyo raised USD 21.3 billion, ICBC USD 19.1 billion, and Visa USD 17.9 billion in 2008. The SPCX press range at USD 75 billion would price the deal at roughly 2.5 times the largest IPO ever recorded, on a market capitalisation that would put it above every previous tech IPO ever brought to market.

Largest IPOs of all time by amount raised

Source: SpaceX EDGAR S-1, Bloomberg, InterCapital Research