The Most Expensive Piece of Paper on Earth: Pricing the SpaceX IPO Honestly

Starship explodes during testing at Boca Chica -- the volatile reality behind the $1.75 trillion valuation.

On April 1, 2026, SpaceX filed a confidential draft registration statement with the SEC for what would be the largest initial public offering in the history of capital markets. Code-named "Project Apex," the deal targets a $1.75 trillion valuation and a $75 billion raise -- eclipsing Saudi Aramco's 2019 record by a factor of nearly three.

Twenty-one banks are fighting over the book like wolves over a carcass. Citigroup, Bank of America, Goldman Sachs, JPMorgan, and Morgan Stanley sit at the top of the syndicate, each desperate for their slice of what could be the richest fee pool in Wall Street history. Musk has reportedly allocated 30% of the offering to retail investors -- a structural choice that could juice day-one demand but destabilize institutional pricing discipline and turn the IPO into a casino.

The question every allocator, analyst, and LP is asking: Is $1.75 trillion real, or is this the most expensive piece of paper since the Dutch bought tulips?

Here's what the data says.

---

What the Prediction Markets Think

Before we get into the banking math, it's worth checking where real money is being placed. Polymarket has several active SpaceX IPO markets with telling odds:

The signal: prediction markets are very confident SpaceX clears $1 trillion but deeply skeptical of the $2T+ territory. The implied consensus is somewhere in the $1.2-1.8T range -- which tracks exactly with the pricing framework below.

The Musk trillionaire market at 73% is essentially a proxy for a successful SpaceX IPO. His ~42% stake means he crosses $1T net worth if SpaceX lists anywhere above ~$2.4T. The market thinks it's probable but far from certain.

---

The Financial Foundation

SpaceX generated approximately $8 billion in EBITDA on $15-16 billion in revenue in 2025, according to Reuters. That's roughly a 50% EBITDA margin -- territory typically reserved for monopolistic software businesses, not capital-intensive aerospace companies.

The revenue trajectory tells the growth story:

Revenue doubled between 2021 and 2022, nearly doubled again in 2023, and has since crashed to ~18-20% growth. This deceleration is the death knell for anyone trying to justify a triple-digit revenue multiple. You cannot ask public market investors to pay 109x revenue for a business growing at the same rate as Microsoft.

Elon Musk has noted that NASA now accounts for only 5% of SpaceX's revenue -- the vast majority flows from commercial Starlink subscriptions and government defense contracts.

---

Three Businesses in a Trenchcoat

SpaceX at IPO is really three distinct businesses stapled together, plus a money-losing AI subsidiary. Each has a radically different risk profile and valuation framework.

1. Starlink -- The Cash Engine

Starlink is the centerpiece. It generates an estimated 50-80% of total revenue depending on how you classify government Starlink contracts vs. consumer subscriptions. SpaceNews projected $11.8 billion in Starlink revenue for 2025, driven by strong consumer demand and growing U.S. military contracts under the Starshield program.

The network now serves over 10 million users globally with 9,500+ satellites deployed. This is a connectivity business with satellite infrastructure economics: high upfront capex, massive operating leverage once the constellation is deployed, and a near-monopoly on low-latency satellite broadband.

Comparable: The closest public analog is a blend of telecom infrastructure (think tower companies at 20-25x EBITDA) and high-growth connectivity (Palantir trades at ~75x EV/Sales). If you value Starlink's ~$12B revenue run-rate at 15-20x sales, you get $180-240 billion for Starlink alone.

2. Launch Services -- The Moat

SpaceX conducts more orbital launches annually than any other company on Earth. The Falcon 9 is the industry's workhorse, and the Starship mega-rocket is designed for full reusability -- a capability that, if achieved at scale, collapses the cost curve for every future mission.

NASA's Artemis program has tapped Starship as the lunar lander. The U.S. Space Force awarded SpaceX $739 million in launch contracts in January 2026 alone. Total federal contract awards across NASA, DOD, and Space Force now exceed $22 billion (Fed-Spend).

Comparable: United Launch Alliance (Boeing/Lockheed JV) and Rocket Lab (RKLB, $38B market cap at 62x EV/Sales). Launch is a high-barrier, government-reliant business. If SpaceX's launch segment does $3-4B in revenue, a 10-15x multiple puts it at $30-60 billion.

3. xAI + X -- The Wild Card

The February 2026 merger folded xAI (which owns X/Twitter and the Grok chatbot) into SpaceX as a wholly owned subsidiary. The merger valued SpaceX at $1 trillion and xAI at $250 billion. But xAI is burning cash: it posted a $1.46 billion net loss in Q3 2025 on just $107 million in revenue.

The stated vision is 1 million solar-powered AI data center satellites in orbit -- an extraordinarily ambitious bet on space-based compute. If this works, it transforms the valuation thesis entirely. If it doesn't, xAI is a $5-6 billion annual cash drain stapled to a social media platform worth maybe $15-25 billion.

Comparable: xAI's orbital AI ambition has no real public comp. X as a social platform is worth a fraction of what Musk paid for it. At best, the AI subsidiary adds optionality. At worst, it's a value destroyer that obscures SpaceX's core economics.

---

The Banker's Pricing Framework: Sum-of-the-Parts

If I were running the book, here's how I'd build the valuation:

Base Case (~$1.0-1.2 Trillion)

The "corporate premium" is real but hard to quantify. SpaceX is the only company on Earth that can land orbital rockets, operates the dominant satellite internet constellation, and has the launch cadence to service every government and commercial customer simultaneously. This is a structural monopoly premium.

PitchBook's independent analysis places fair value at $1.1 trillion -- right in line with this framework. Morningstar says the $1.5T target "is expensive but not irrational" IF Starship commercializes on schedule.

Bull Case (~$1.5-1.75 Trillion)

This requires you to believe:

• Starlink reaches $20B+ revenue by 2028 and earns a 20-25x multiple as a monopoly connectivity platform
• Starship achieves full reusability and unlocks $100B+ in addressable market (point-to-point transport, lunar logistics, Mars missions)
• xAI's orbital compute vision is technically viable and begins generating meaningful revenue within 3-4 years
• Government contract pipeline expands under the Golden Dome missile defense program ($13.4B proposed for FY2026)

Bear Case (~$600-800 Billion)

This is where gravity reasserts itself:

• Revenue growth continues decelerating toward 15-18%, making 109x EV/Revenue absurd for a maturing business
• xAI losses widen, consuming $6B+ annually with no clear path to profitability
• Starship development delays push lunar mission timelines out 2-3 years
• Key-man risk: Musk's attention is fractured across Tesla, SpaceX, xAI, X, Neuralink, Boring Company, and DOGE
• The 130x revenue multiple that MarketPulse flagged is unjustifiable relative to 20% growth

---

Sensitivity Analysis: What Growth and Multiples Imply

The table below shows the implied market cap ($B) at various 2-year revenue CAGR and EV/Revenue multiple combinations, using $16B as the 2025 revenue base and projecting forward to 2027.

To justify the $1.75T IPO valuation on 2027 forward revenue, you need one of:

• 15% growth + 130x multiple (extreme premium)
• 20% growth + 90x multiple (rich but defensible if margins hold)
• 25% growth + 70x multiple (requires growth re-acceleration)

The last scenario is the bull path -- and it's the one the syndicate is selling. If Starlink subscriber growth reaccelerates and Starship begins commercial operations, 25% growth is achievable. At 70x forward revenue, $1.75T falls right out of the model. But 70x is still a massive multiple. For reference, Palantir -- the most expensive large-cap defense/AI name -- trades at 76x today.

---

The Comp Table Reality Check

Here's where SpaceX sits relative to its public peers:

The spread is staggering. Traditional defense primes trade at 2-4x sales. The "new space" names (Palantir, Rocket Lab) trade at 60-75x. SpaceX is asking for 109x -- a premium to the premium names.

The justification is that SpaceX is none of these things individually. It's all of them simultaneously, with a growth rate and competitive position that no single public company can match. The question is whether "none of these things" deserves a premium or a discount.

The Public Market Conglomerate Discount

Here's the uncomfortable truth: public markets hate complexity.

SpaceX isn't just asking investors to underwrite a satellite internet business. It's asking them to bet on:

• Telecom infrastructure (Starlink)
• Defense contracting (Launch)
• Venture-stage AI moonshots (xAI orbital compute)
• Social media turnarounds (X/Twitter)
• Capital-intensive manufacturing (Starship)
• Regulatory capture risk (government dependence)

Berkshire Hathaway works as a conglomerate because it's capital-light and cash-generative across every segment. SpaceX bundles high-capex hardware businesses with volatile AI cash burn and regulated defense exposure.

The result? A conglomerate discount, not a premium.

General Electric traded at a 20-30% discount to sum-of-parts for decades before breaking up. Honeywell, 3M, and United Technologies all faced the same penalty. Public market investors pay less for complexity, not more.

If SpaceX traded at a 25% conglomerate discount to the $1.2T base case, you're looking at $900 billion -- not $1.75 trillion. The xAI merger made this worse, not better.

---

What I'd Actually Price It At

If I were the lead left banker on Project Apex, I'd set the IPO price range to imply a $1.1-1.3 trillion valuation.

Here's my reasoning:

1. Leave money on the table. The cardinal rule of large-cap IPOs is that the stock needs to trade up 15-25% on day one. If you price at $1.75T and the market thinks fair value is $1.1T, you get a broken deal. If you price at $1.1T and the market takes it to $1.4T, you're a hero.

2. The xAI merger muddies the water. Institutional investors can underwrite Starlink + Launch. They cannot underwrite an AI subsidiary losing $6B/year with $400M in annualized revenue. The merger creates a conglomerate discount, not a premium. A clean Starlink spinoff IPO would command a higher multiple than the combined entity.

3. The deceleration curve matters. Going from 100% growth to 18% growth in three years changes the multiple framework entirely. At $16B in revenue growing 18%, this is a 60-80x revenue business, not a 109x revenue business. That puts you at $960B-$1.28T.

4. The $75 billion raise is enormous. This is not a standard IPO. Raising $75B means selling roughly 4-5% of the company at $1.75T. The market has to absorb that. Setting the range lower ensures oversubscription and strong aftermarket performance.

5. Polymarket agrees. Prediction markets give 88% odds SpaceX clears $1T on day one, but only 8% odds it hits $2.5T+. The implied consensus sits in the $1.2-1.8T band. Smart money is pricing the same range I am.

My realistic pricing: $1.1-1.3 trillion at IPO, with a first-day pop to $1.4-1.5 trillion. That's roughly a 70-80x EV/Revenue multiple on 2025 numbers, which is rich but defensible given the monopoly position, 50% EBITDA margins, and government contract pipeline.

The $1.75 trillion figure is where the stock trades six months post-IPO if Starship hits milestones and Starlink subscriber growth reaccelerates. It's a target, not a launch price.

---

The Bottom Line

SpaceX is a generational company. It has no real competitor in its core launch business, Starlink is approaching escape velocity as a connectivity platform, and the government contract pipeline provides a floor that most tech companies would kill for.

But $1.75 trillion at IPO is delusional pricing. It assumes perfection across every business line, no execution risk on Starship, and that public market investors will pay a higher multiple than the last private round. History suggests they won't -- at least not on day one. This is the kind of valuation that gets laughed out of allocation committees and becomes a cautionary tale at Wharton.

The smart money will buy at $1.1-1.3T and hold for the convergence toward $1.5-2T as the business proves itself in public markets. That's where the real alpha is.

Key sources:

• Reuters: SpaceX generated ~$8B profit on $15-16B revenue
• Reuters: SpaceX business and finances overview
• CNBC: SpaceX lines up 21 banks for Project Apex
• Al Jazeera: SpaceX files to go public
• Fortune: SpaceX sets $800B valuation, confirms 2026 IPO
• SpaceNews: Starlink $11.8B revenue projection
• SpaceNews: $739M Space Force contract
• Fed-Spend: $22B in government contracts
• MarketPulse: 130x revenue red flags
• Morningstar: SpaceX valuation analysis
• Polymarket: SpaceX IPO market cap
• Polymarket: Musk trillionaire odds