SpaceX Filed the Largest IPO in History. Here's What the S-1 Actually Says.

On May 20, 2026, the most secretive major company in the world filed a public S-1 for the first time in its 24-year history. This is what $1.75 trillion actually looks like on paper.

SpaceX has spent two decades building businesses that would be individually transformational — and it is going public with all three at once.

On May 20, 2026, Space Exploration Technologies Corp. filed its S-1 registration statement on SEC EDGAR, targeting a Nasdaq listing under the ticker SPCX. The roadshow starts June 8. The company is aiming to raise $75 billion — a number that would more than double the previous record for any IPO in history. The reported valuation: $1.75 trillion, with some banks citing potential for $2 trillion or higher.

For the first time, the financial architecture of SpaceX is public record. The numbers are remarkable, the risks are real, and the package deal is more complicated than most investors realize.

Here is the full breakdown.

The Filing at a Glance

Three Companies. One Stock Price.

The most important thing to understand about the SPCX offering is that investors are not buying a rocket company. They are buying a package of three separate businesses that happen to share a CEO and a balance sheet.

The S-1 consolidates:

1. Starlink — a satellite internet constellation with 10.3 million subscribers across 164 markets
2. Space (Launch Services) — Falcon 9, Falcon Heavy, Starship; the world's dominant commercial launch provider
3. xAI — an AI infrastructure and LLM company that spent $12.7 billion on GPU clusters in 2025

Starlink is profitable and growing fast. Launch is stable and government-backed. xAI is burning cash at a rate that turned a profitable company into one losing nearly $5 billion a year. All three are in the offering. None are separable.

Segment 1: Starlink — The Cash Engine

Starlink is the reason this IPO is priced the way it is, and the reason the long-term bull case is defensible.

The Numbers

In 2025, Starlink accounted for 61% of total revenue and essentially all operating profit. In Q1 2026, that share climbed to 69%. The EBITDA margin on the segment is approaching 63% — software-like margins on what is, at its core, a capital-intensive infrastructure business.

Why the Moat Is Real

SpaceX currently controls approximately 75% of all maneuverable satellites on Earth. No competitor can close that gap without a decade of work and $50 billion or more in capital. The physics of low earth orbit are unforgiving — the first mover who saturates the orbital shells gets the best performance characteristics, and the best performance characteristics win the customers.

The subscriber base doubled in one year. At that trajectory, 20 million subscribers by 2027 is a realistic projection, not a stretch goal.

More importantly, the revenue lines that barely register in the current financials — Starlink Aviation, Maritime, and Direct-to-Cell in partnership with T-Mobile — are the next growth vector. Those three categories have not yet reached scale. The $11.4 billion is not the ceiling; it is the starting line.

Segment 2: Launch Services — The Monopoly Floor

Falcon 9 is the most reliable orbital rocket ever built. It has delivered a launch cadence no other provider comes close to matching.

Launch services generated approximately $4–6 billion in 2025 revenue (sources vary on the precise figure), primarily from:

• NASA — Commercial Crew (ISS transport), Commercial Resupply Services
• U.S. Space Force and DoD — classified national security launches
• NRO — National Reconnaissance Office payloads
• Commercial satellite operators — communications and earth observation constellations

Combined, U.S. government contracts represent roughly 35% of total company revenue. This creates both stability (long-term agreements, mission-critical contracts) and concentration risk (discussed in the risks section below).

Starship: The Optionality Ticket

SpaceX has spent $15 billion or more developing Starship — a fully reusable super heavy-lift vehicle that does not yet carry commercial payloads. The S-1 does not disclose a specific cost per Starship launch, but the company's stated target is $10–$100 per kilogram to orbit.

For context: Falcon 9 costs approximately $1,500–2,000 per kilogram. Achieving even the high end of the Starship target would represent a 15x cost reduction. Achieving the low end would be among the most consequential engineering outcomes of the decade — transforming everything from satellite deployment economics to the commercial viability of in-space manufacturing, deep space exploration, and point-to-point cargo delivery.

Starship is not currently in the revenue model. But it is absolutely priced into the $1.75 trillion valuation.

Segment 3: xAI — The $20 Billion Bet

The third business is the most contentious element of the offering, and the primary reason SpaceX swung from +$791 million net income in 2024 to -$4.94 billion net loss in 2025.

xAI is an artificial intelligence company operating the Colossus GPU clusters — currently 200,000+ NVIDIA GPUs — and developing the Grok series of large language models. SpaceX invested heavily in xAI and consolidates it within the S-1. The AI segment posted an operating loss of $6.35 billion in 2025 and $2.4 billion in Q1 2026 alone (a roughly $10 billion annualized burn rate).

The Anthropic Deal

The single most surprising disclosure in the filing is a compute contract between Anthropic and xAI's Colossus infrastructure: $1.25 billion per month for dedicated access to SpaceX's GPU clusters through May 2029. The total deal value, if run to term, is approximately $45 billion — one of the largest AI infrastructure contracts ever signed.

The significance: Colossus is large enough that xAI can rent capacity to the biggest model builders in the world while running its own training workloads. This converts SpaceX from a rocket company with an AI hobby into an AI infrastructure landlord — a fundamentally different and potentially very valuable asset class.

But xAI is still spending faster than it earns. Capital expenditures on AI infrastructure hit $12.7 billion in 2025 and $7.7 billion in Q1 2026 — suggesting an annualized capex run rate closer to $30 billion. The Anthropic contract generates roughly $15 billion per year. The math requires additional large compute tenants or dramatically accelerated Grok revenue to reach breakeven.

The Full Financial Picture

Revenue & Profitability

Balance Sheet

Valuation Math

At the reported $1.75 trillion valuation:

• Price-to-Starlink Revenue: ~153x ($1.75T / $11.4B)
• Price-to-Starlink EBITDA: ~243x ($1.75T / $7.2B)
• Price-to-Total Revenue: ~94x ($1.75T / $18.7B)

These are not value multiples. They price in full execution across Starship, xAI profitability, Direct-to-Cell penetration, and a global satellite internet TAM that is still being built. For comparison, Nvidia — widely regarded as one of the best businesses in the world — has peaked at roughly 30x revenue. Buying SPCX at IPO means paying three times Nvidia's peak multiple for a company that is currently losing $5 billion a year.

The bull case requires believing those losses are temporary and that the underlying earnings power of the three businesses, fully scaled, justifies the price. The bear case does not require disbelieving SpaceX — it just requires believing the current price already reflects the bull case.

The Bull Case: Why $1.75 Trillion Is Defensible

1. Starlink Has No Credible Competitor

There is no Western entity capable of deploying a competing LEO broadband constellation at comparable scale within the next five years. OneWeb has fewer than 700 satellites and no consumer product. Amazon's Project Kuiper has the financial backing but is years behind on deployment and has yet to prove market traction. The structural moat here is real and durable.

2. The Direct-to-Cell Opportunity Is Barely Counted

SpaceX's partnership with T-Mobile to beam Starlink connectivity directly to standard smartphones — no dish required — is only in early commercial deployment. If the technical performance scales, this becomes a global connectivity product that bypasses every terrestrial network. The addressable market is not 10 million Starlink subscribers. It is 6 billion smartphone users.

3. Starship Changes the Entire Launch Economics Model

If Starship reaches operational status with costs anywhere near the stated targets, SpaceX will have built the equivalent of the first commercially viable steamship in an industry that was previously operating on sail power. The competitive position in launch services would become effectively permanent.

4. The AI Infrastructure Build Is Real and Monetizable

The Colossus cluster exists. The Anthropic deal is signed. The infrastructure-as-a-service model for AI compute is a $40+ billion annual market growing at 50%+ per year. SpaceX is already in it.

5. Vertical Integration at Global Scale Is Unreplicable

SpaceX designs, builds, launches, operates, and monetizes the entire satellite internet stack — rocket, satellite, terminal, software, ground network. No competitor controls all of those layers simultaneously. The cost advantages and speed-of-iteration advantages that flow from vertical integration compound over time.

The Bear Case: 38 Pages of Risk Factors Exist for a Reason

1. The Elon Musk Problem

The S-1 states plainly that the company is "highly dependent on the continued service and performance of Mr. Musk, whose leadership, vision, and expertise are critical to the development of our technologies and the execution of our business strategy."

Musk simultaneously runs Tesla, X, The Boring Company, and xAI, in addition to holding a formal government role at DOGE. He controls 85% of voting power through super-voting shares while owning approximately 42% of equity. Public shareholders cannot vote to change strategic direction, cannot remove him, and cannot redirect capital allocation — including the xAI burn rate. If Musk is incapacitated, politically disqualified from holding government contracts, or simply distracted, the business faces risk that cannot be hedged or insured against.

2. xAI Is Currently a Multi-Billion Dollar Annual Drain

Before the AI bet, SpaceX was profitable. The $20 billion-plus bet on AI infrastructure turned the company into a cash-burning enterprise. The Anthropic deal provides meaningful revenue, but xAI is still spending roughly twice what it earns. Buyers of SPCX own this exposure whether they intend to or not. There is no way to buy "just Starlink."

3. Government Revenue Concentration Carries Political Risk

Approximately 35% of revenue comes from U.S. government contracts. These are critical relationships, but they are also contracts that renew on political cycles. A change in NASA's Artemis priorities, a DoD budget cut, or a regulatory action stemming from Musk's political activities could curtail this revenue stream with six to eighteen months' notice. The same political relationships that currently benefit SpaceX could become a liability under different conditions.

4. The Valuation Leaves No Margin for Error

At 94x total revenue, the stock is priced for a scenario where everything works: Starship goes commercial, Starlink hits 50+ million subscribers, xAI reaches profitability, and Direct-to-Cell scales globally. Any one of those failing — or taking five years longer than modeled — compresses the valuation materially. There is no "good but not great" outcome at these multiples.

5. You Have Zero Governance Rights

Dual-class share structures are common in founder-led tech IPOs. The degree of control here is extreme even by those standards. Public shareholders are buying a non-voting economic interest in Elon Musk's allocation of capital. The $30 billion annual AI capex run rate was not put to a shareholder vote. The next large bet will not be either.

6. The Balance Sheet Is Stretched

$29.1 billion in debt against $15.9 billion in cash is a net debt position of $13.2 billion. With capex running at approximately $10 billion per quarter, the $75 billion raise is primarily fuel for the build-out — not a new growth catalyst on top of an already healthy balance sheet. Without the raise, the balance sheet becomes strained within 18 to 24 months.

7. Starship Still Has to Work

More than $15 billion has been spent on Starship's development. The vehicle has not yet carried a commercial payload. Each major test flight that ends in a RUD (Rapid Unscheduled Disassembly) is priced into the valuation as a solvable engineering problem. If the timeline slips by three or more years, or the cost targets prove unreachable, the $1.75 trillion valuation loses one of its core pillars.

8. Orbital Congestion and Spectrum Risk Are Growing

As the Starlink constellation expands toward its full build-out of 40,000+ satellites, orbital congestion, collision risk, and spectrum interference become real operational constraints. International regulators — particularly in the EU and through the ITU — are pushing back on SpaceX's spectrum dominance. A regulatory cap on constellation size would directly limit the Starlink subscriber ceiling.

Who Else Benefits: Companies Named in the Filing

These are companies identified as direct customers, contract partners, suppliers, or sector beneficiaries of the SpaceX filing.

Direct Relationships

Hardware Supply Chain

Space Sector Sentiment Lift

A SpaceX IPO at $1.75 trillion normalizes the "space as infrastructure" investment thesis and draws significant institutional capital into the category. These are the public companies best positioned to benefit from that rerating.

Companies With Competitive Headwinds

The Bottom Line

Starlink alone might justify $300–500 billion. The launch business is a profitable, government-backed monopoly worth real money in its own right. The Colossus/xAI bet is either the smartest infrastructure build of the decade or a $30 billion annual distraction.

What investors are actually being asked to buy at $1.75 trillion is all three, consolidated, with one person controlling 85% of the votes and no ability to unbundle them.

The quality of the core businesses is not in question. The price demands that they all succeed simultaneously, on an aggressive timeline, without execution failures on Starship, without political disruption to government contracts, and without xAI's cash burn proving structural rather than transitional.

Starlink is a generational infrastructure business. The stock is priced like it already won.

Track the Space Sector in Real Time

The SpaceX IPO will move its entire sector — from Rocket Lab and Intuitive Machines to T-Mobile and NVIDIA. Set alerts on the stocks that matter before the roadshow hits.

Start monitoring space stocks with Stock Alarm Pro →

Or screen the space and defense sector directly with live fundamentals and real-time quotes:

Open the Screener →

Disclaimer: This article is for educational and informational purposes only and does not constitute financial or investment advice. Stock Alarm Pro may hold positions in securities discussed. All figures are sourced from publicly reported S-1 data as covered by CNBC, Bloomberg, TechCrunch, and multiple financial publications. Always conduct your own research before making investment decisions.

Sources:

• SpaceX S-1 Registration Statement — SEC EDGAR
• CNBC: SpaceX is Heavily Reliant on Starlink for Growth and Profit
• CNBC: SpaceX Historic IPO Plans — Billions in Losses
• TechCrunch: Anthropic Will Pay xAI $1.25B Per Month for Compute
• TechCrunch: The SpaceX IPO Filing — AI Bets, Starship Dreams, Elon Musk
• SatNews: Inside SpaceX's S-1 — Three Companies. One Profit. $1.75 Trillion.
• MoneyWise: SpaceX IPO S-1 Risk Factors
• The Next Web: SpaceX Has Spent More Than $15 Billion on Starship
• Tesery: SpaceX IPO S-1 Full Breakdown
• 247 Wall St: SpaceX $2T IPO — These 6 Stocks Will Ride the New Space Economy
• Hawkmont Research: SpaceX IPO Supply Chain Stocks 2026