SpaceX's S-1 Shows 76% of Q1 Capital Going to AI Infrastructure, Not Rockets

SpaceX filed its S‑1 with the SEC on May 20, 2026, requesting a dual listing on Nasdaq and Nasdaq Texas under the ticker “SPCX.” The prospectus quantifies a vision the company has pursued since acquiring xAI three months earlier: a single balance sheet that integrates orbital launch, satellite broadband, and AI compute.

SpaceX's own S-1 quantifies the ambition: a "quantifiable" total addressable market of $28.5 trillion, split into $370 billion for space, $1.6 trillion for connectivity, and $26.5 trillion for AI, according to KraneShares' analysis of the S-1 filing. Against that framing, the company reported $18.7 billion in 2025 revenue and a net loss of $4.9 billion, per Wikipedia's SpaceX entry, while proposing a valuation of $1.75 trillion to $2 trillion — a multiple GAB Growth's S-1 deep dive calculates at roughly 94 times 2025 sales.

What the filing actually bundles together

The S-1 does not describe SpaceX as an aerospace company with a side project. Layered on top is the AI business, built around the February 2026 acquisition of xAI, whose effective date appears in the S-1 filed with the SEC.

The clearest sign of where that capital is actually going is not the launch manifest. It is the AI infrastructure line. In the first quarter of 2026, SpaceX directed 76% of its capital deployment — $7.72 billion — into AI infrastructure and data-center buildouts, according to Futurum Group's analysis of the S-1 by analyst Brendan Burke. The same filing disclosed a compute-leasing arrangement with Anthropic worth $1.25 billion a month, which Burke's analysis says establishes SpaceX "as a major hyperscale landlord" — a rocket company renting out data-center capacity to a frontier AI lab, funded in part by a satellite-broadband business, all inside one entity trading under one ticker.

Where the capital is actually going

Sources: Futurum Group's S-1 analysis; KraneShares' S-1 breakdown; Wikipedia's SpaceX entry; CNBC's report on the Louisiana spaceport.

Three-quarters of a quarter's capital going to AI infrastructure, in a company whose S-1 still frames launch as the "foundation," is the anomaly worth sitting with. Whether that capital produces AI revenue anywhere near what has been projected is exactly the question the market has not resolved: Bernstein's forecast, cited in CNBC's report on SpaceX options strategies ahead of Starship's next flight, has AI revenue rising from $24.6 billion in 2026 to $115 billion in 2027 — a projection CNBC itself calls "a true hockey stick proforma."

The Louisiana bet running in parallel

Five days before that CNBC piece, SpaceX and the Louisiana Economic Development office jointly announced a spaceport in Vermilion Parish carrying an investment of up to $100 billion, intended to support thousands of launches annually. Per CNBC's report on the announcement, the project is projected to bring roughly 3,000 direct jobs and 8,100 indirect jobs to the state at an average salary near $92,600. Musk said in a video accompanying the announcement that the project would bring "probably 10,000 really exciting jobs," per the same CNBC report. Construction is slated to begin by the end of 2027, with first launch targeted for as soon as 2029 — a decade-long physical build running concurrently with the AI infrastructure buildout, funded from the same balance sheet that just took public money for the first time.

That balance sheet has not traded calmly. The stock then rose to a post-IPO high of $225.64 on June 16 before falling to $104.83 by August 3 — a decline of roughly 53% from peak to trough — before recovering to close at $147.95 on September 4, per CNBC's SPCX quote page. CNBC's own options-strategy piece describes traders spending the summer "treating SPCX as a lockup-and-capex problem" before the narrative "flipped" toward the integrated-stack thesis in the 48 hours before publication — language that describes a stock being repriced on story rather than on a resolved change in fundamentals.

The bet underneath all of it has not yet had its test flight. Starship, the vehicle meant to deploy the next generation of Starlink satellites and anchor the connectivity segment's growth, had not completed a fully orbital flight as of early September 2026; CNBC's piece describes Flight 14, targeted for September 15, as "the first fully orbital attempt." A trillion-dollar valuation resting on three businesses bundled into one filing is, for now, resting in part on a rocket that has not yet finished the flight its own prospectus assumes.