SpaceX Is Going Public.
Three Companies.
One Stock. One Question.

Three Companies. One Stock. A $1.75 Trillion Question.

SpaceX filed its S-1 on May 20, 2026 — the largest IPO registration in history by targeted capital raise ($75 billion) and implied valuation ($1.75 trillion). The roadshow begins June 5. The listing target is June 12 on Nasdaq under the ticker SPCX.

What investors are actually buying is not one company. It is three fundamentally different businesses stapled together under a single ticker — each with different economics, different risk profiles, and different timelines to value creation.

Connectivity (Starlink) is the spine: $11.4 billion in revenue, $4.4 billion in operating income, 10.3 million subscribers growing 2× year-over-year across 164 countries. Strip Starlink out, and the rest of the business is deep in the red.

Space (Launch + R&D) is a deliberate loss center, burning roughly $3 billion per year on Starship development that underpins every downstream plan. Eleven test flights complete. Commercial operations targeted for 2H 2026.

AI (xAI + Grok + X) — acquired in February 2026 — is the wildcard. It consumed $12.7 billion in capex in 2025 and lost $6.4 billion from operations. It is the reason SpaceX needs $75 billion from public markets right now.

Three Very Different Financial Profiles

The consolidated picture hides the segmented reality. Starlink funds everything. Space is being built deliberately at a loss. AI is where the capital is going — $12.7 billion in 2025 capex alone, rising to an annualized pace approaching $31 billion in Q1 2026.

The Golden Goose — and Its Quiet Warning Sign

Starlink is the financial spine of SpaceX. 10.3 million subscribers, doubled in 12 months, across 164 countries with 9,600+ satellites in orbit. Connectivity adjusted EBITDA reached $7.2 billion in 2025 — up 86% year-over-year. This segment alone would be one of the most valuable standalone businesses in the world.

But there is a quiet warning sign buried in the subscriber narrative: ARPU is compressing fast.

Where SpaceX Actually Makes Things

The S-1 is a financial document, but buried inside it is the most detailed public accounting SpaceX has ever given of its manufacturing footprint. The picture that emerges is not of a rocket company. It is a vertically integrated industrial complex — quietly one of the largest advanced manufacturers in the United States.

The two Gigabays under construction represent the most important capital investment in the S-1 that nobody is discussing publicly. The Florida Gigabay is 11 times larger than SpaceX’s current biggest integration facility. When both come online at end of 2026, SpaceX can stack and refurbish Starships in parallel at a scale that permanently changes the unit economics of orbital launch.

Inside the Redmond Satellite Factory

SpaceX built the Redmond facility specifically because Starlink required a manufacturing philosophy the aerospace industry had never attempted: mass production of complex spacecraft on consumer-electronics timelines. Before Starlink, the entire global satellite industry launched roughly 400 satellites per year. SpaceX can match that in 10 weeks.

Every satellite is designed for a 5-year operational life. The fleet runs at over 99% on-orbit reliability. When a satellite shows signs of becoming non-maneuverable, SpaceX proactively deorbits it — accepting the loss of an otherwise healthy satellite rather than letting it become debris. They chose aluminum fuel tanks over composite overwrap pressure vessels specifically so the satellite will fully demise on atmospheric reentry.

The on-board collision avoidance system triggers at a 1-in-100,000 probability — ten times stricter than the 1-in-10,000 industry standard. When a conjunction is detected, the satellite can execute a “duck maneuver”: rotating edge-on toward the potential collision to reduce its cross-sectional area by 4–10×. SpaceX also shares all satellite ephemeris and covariance data publicly on Space-Track.org three times per day.

The choice to deploy at 350km rather than the industry-standard 550km is deliberately conservative. At 350km, atmospheric drag will deorbit any satellite that fails initial checks within weeks. Competitors operating above 1,000km face natural deorbit times measured in centuries. SpaceX’s constellation self-cleans in 5–6 years.

V3 Starlink: A Step-Function Leap in Capacity

The next-generation V3 Starlink satellite can only be deployed by Starship — Falcon 9 cannot carry it. Each Starship V3 launch adds more than 20× the constellation capacity of a Falcon 9 V2 launch. That is not a marginal improvement. It is a step function.

xAI, Grok & X: The $6.4B Loss That Changes Everything

The February 2026 merger folded xAI — and X, which xAI had absorbed in March 2025 — into SpaceX under common-control accounting, recasting all historical financial periods. The AI segment is where the financial story breaks.

In 2025, AI consumed $12.7 billion in capex — 61% of total company capex. It generated $3.2 billion in revenue growing at 22% year-over-year. That growth rate is below every frontier AI peer. It lost $6.4 billion from operations. In Q1 2026 alone, AI capex reached $7.7 billion — an annualized pace approaching $31 billion.

This is the reason the IPO must happen. Cash position fell $8.8 billion in Q1 2026 alone.

The Cursor deal is not a binding acquisition. It is an option. If SpaceX exercises, it pays $60 billion in stock for a coding assistant in a category where every major tech company is competing. If it walks away, it pays $10 billion in fees and foregone services. This is an expensive hedge, not a strategic moat.

Shattering Every Record in History

At $75 billion targeted raise, this would be the largest IPO in history — nearly doubling Saudi Aramco’s previous record of $29.4 billion.

Underwriting syndicate: Goldman Sachs, Morgan Stanley, BofA, Citi, J.P. Morgan, Barclays, Deutsche Bank, RBC, UBS, Wells Fargo, Allen & Company, Cantor, Needham, Raymond James.

“The Largest Actionable Market in Human History”

SpaceX claims a $28.5 trillion total addressable market — a number that spans from satellite launches to AI enterprise software. Ninety-three percent of that figure sits in AI, a business that is currently losing $6.4 billion per year.

What the $1.75T Valuation Is Actually Pricing In

Beyond Starlink, investors are paying for a stack of venture-scale bets inside a public company wrapper. Here’s an honest assessment of each.

The Risks Every Investor Needs to Know

Sum-of-the-Parts: Breaking Down $1.75 Trillion

To value SpaceX honestly, you must separate the three segments. The consolidated number obscures a cash-generative satellite business funding a money-losing AI venture.

You Are a Passenger. Musk Is the Pilot.

The dual-class share structure means Class B shareholders — effectively Musk — elect a majority of the board and control all matters requiring shareholder approval. SpaceX qualifies as a “controlled company” under Nasdaq rules and explicitly opts out of independent board requirements.

There is also the Mars Compensation Package: a $1.3 billion grant tied to Mars-related milestones, disclosed in the S-1 in addition to Musk’s existing equity position. The company’s ultimate performance target is multi-planetary civilization — a mission horizon that no public equity market has ever been asked to price on a quarterly basis.

SPCX: Speculative Buy — But Know What You Own

SpaceX is simultaneously the most operationally capable aerospace company ever built, a cash-generative satellite monopoly, a nascent AI infrastructure play, and a Mars colonization venture — all being offered to public investors at 94× trailing revenue with a founder who holds 85% of the votes.

The investment case rests on three things happening: Starlink continuing to grow subscribers faster than ARPU compresses; Starship achieving commercial operations in 2026–2027; and the AI segment eventually justifying its $12.7 billion per year capex spend through revenue growing at something closer to 100% than the current 22%.

If those three things happen — and Musk’s track record suggests they might — the $1.75 trillion valuation looks reasonable in five years. If Starship slips, ARPU collapses, and xAI fails to achieve model leadership, the public equity could trade at a severe discount to the private market that funded it.

Retail access through Fidelity, Schwab, and Robinhood ensures massive day-one demand. But this is the first time in SpaceX’s history that every quarter of AI losses and every point of ARPU decline will be scrutinized publicly by analysts who did not sign up to fund a Mars colony.

This analysis is based on the SpaceX S-1 registration statement filed May 20, 2026, publicly available financial disclosures, and third-party analyst data. It does not constitute investment advice. Estimates are analyst projections, not SpaceX guidance. Consult a licensed financial advisor before making investment decisions.