The SpaceX IPO: The Latest Offering from Mr. 10x
We have covered the Buffett method for valuing businesses in a prior piece: The Most Misunderstood Phrases in Buffett-Speak: Free Cash and the Float.
Elon Musk is taking SpaceX public. The ask is $1.75 trillion.
Let’s talk about what that number means. And what it doesn’t.
SpaceX is a remarkable company. The Falcon 9 is the most-flown active orbital rocket and the backbone of modern launch cadence. Starlink has over 10 million subscribers and dominates space-based internet traffic. These are real accomplishments. Nobody serious disputes them.
But a business in an IPO is worth exactly one thing: the free cash flow it generates till the end of its life, discounted back to today. Not revenue. Not narrative. Not the founder’s Twitter following. Free cash flow, discounted.
Free cash is cash you could give out as a dividend if you wanted to. It’s never needed for future expenses or growth. It belongs to the owners, free and clear. Everything else — earnings, EBITDA, adjusted this, pro forma that — is accounting. Free cash is reality.
Discounted means adjusted for the cost of money over time. A dollar received ten years from now is worth less than a dollar today — because you could have invested that dollar and grown it in the interim. Inflation eats the value of money too. The longer you wait for the cash, and the riskier the bet that it ever arrives, the less it is worth today.
Think of an IPO — or any investment — as a gold mine. The gold is the product you make that generates free cash for you. Eventually you mine it all. The gold runs out, or nobody needs your gold anymore, or someone can make it for less.
The discounted free cash the business earns must be more than what the IPO was valued at.
No business lasts forever. A competitor can make something just as good or better, flood the market, and kill the price of yours. Technology can make your mine obsolete before you’ve dug halfway through it.
Your gold mine might have ten billion dollars worth of gold in it but if you spend a hundred billion dollars mining and processing it, you just lost ninety billion dollars. The point is not to just break even. You need to make a profit — and one that accounts for the money you could have earned with the cash you paid for the IPO over that same lifetime.
Wall Street does not like to show you this math. They prefer forward revenue multiples and TAM-based narratives — frameworks that sound rigorous but avoid the inconvenient question of how much free cash actually escapes the machine. Usually they also have a greater fool story — that momentum will make someone else buy it from you for more.
A prudent investor ignores that entirely. For purposes of value calculation, they assume they can never sell it again. In the end, all they will ever have is the free cash the business generated. That is the only number that matters.
When Warren Buffett considers buying a company, he first calculates its value and then looks at the stock price. If the value is much more than the stock price, he might buy it. Not the other way around.
When you run the free cash math on SpaceX, the number is not $1.75 trillion without a lot of help from the tooth fairy.
Mr. 10x
You might have thought I call Elon Mr. 10x because he has so many companies with the name X in it. SpaceX. xAI. Twitter, which he renamed X. A reasonable hypothesis. Nope. That’s not what the X means.
Musk has earned a nickname. Call him Mr. 10x. The 10x is the multiplier you attach to everything about the business — the timelines, the capabilities, the valuation. It is always there. It is always wrong in the same direction.
The pattern is consistent enough to be a physical constant. It takes ten times longer to deliver than he promises. He promises ten times more capability than he delivers. The valuation he gets is ten times the true valuation.
I covered the 10x pattern in detail for Tesla in The True Value of Tesla.
The SpaceX IPO is the 10x pattern applied to valuation — and most likely also to schedule and capabilities. He’s asking you to pay today for a company that needs to produce Apple-scale free cash flow — $100 billion annually, sustained — to justify the price. Free cash is cash you can pay as a dividend. Apple proves it — they do exactly that, billions every quarter. At $1.75 trillion, SpaceX would need to match Apple’s output for nearly twenty years just to return the purchase price. Undiscounted.
Smoke and Mirrors
AI is hot. The promise of transformational technology is in the air. SpaceX bundled rockets, satellites, a bleeding AI startup, and a cancelable compute contract into one whale’s tale and filed it with the SEC. The goal was a 10x valuation multiplier — tech multiples on an aerospace business. It worked. For now.
The IPO prospectus now pitches a combined entity with a $28.5 trillion total addressable market anchored on AI infrastructure.
The S-1 includes an Anthropic deal: Anthropic agreed to pay xAI $1.25 billion per month for compute capacity through May 2029 — a contract that could exceed $40 billion. One can assume it’s there to buttress the future earnings story.
But what are we really talking about here? SpaceX is renting excess capacity to Anthropic. There is nothing to stop Anthropic from building its own data centers or getting them from someone else. The contract carries a 90-day termination clause for either party. This is not likely a permanent revenue source. The implication that it is some kind of continuing revenue source is there (?) — since $40 billion as a one-time revenue item would be noise as part of a $1.75 trillion valuation.
The Math Nobody Is Doing
You can buy a piece of the future. That is what equity is. But just because the future is big does not mean it is a good investment at the price being offered. There has to be math. You have to estimate the size of the gold deposit, the cost to extract it, the timeline, the competition, the discount rate. Then you look at the price being asked and you either accept or reject the assumptions baked into it. Most SpaceX bulls skip that step. They see a big future and stop there. That is not investing. That is hoping.
At $1.75 trillion, with a 10% discount rate and modest terminal growth, SpaceX needs to generate roughly $120 billion in annual free cash flow. Essentially permanently.
Let’s look at the real business here — SpaceX minus the AI spin. The AI spin is just spin at this time. xAI is not Anthropic or OpenAI. xAI was carried at an implied valuation of roughly $250 billion in the merger. Whether that number is justified is its own leap of faith — Mistral, a genuine AI company with real enterprise revenue, is valued at $13.8 billion. Twitter — now X — is a perennial money loser. In 2025, xAI generated $3.2 billion in revenue and lost $6.4 billion. That is the AI business being bundled into a $1.75 trillion valuation.
Starlink at maturity — 30 to 40 million subscribers, $66 per month average (and falling as it expands into lower-income global markets), facing real competition from Amazon Leo and Blue Origin’s TeraWave (with a whole list of additional suitors on the way) — generates maybe $26 billion in revenue. 35 million subscribers at $66 per month is roughly $27 billion in revenue. At a 15% free cash flow margin — generous for a business that must continuously replace satellites — that’s $4 billion in free cash annually from Starlink.
Launch services at maturity: $12 to 15 billion in revenue. At a 20% free cash flow margin — better than satellite because reusable rockets don’t fall out of orbit — that is roughly $3 billion in free cash flow.
Government and defense contracts: $1 billion.
Total, at peak, on optimistic assumptions: $8 billion in annual free cash flow.
And that is at maturity — and assuming no significant competition compresses margins. Both are generous assumptions.
Now discount it. SpaceX doesn’t reach that state until 2032 at the earliest. Ten years out at 10% cuts the present value roughly in half. Assign a terminal multiple (FCF lingo) of 10x. You land somewhere between $50 and $150 billion in fair value, depending on how generously you treat the assumptions.
Let’s consider a somewhat comparable company, SiriusXM. After 25 years, one near-bankruptcy, and a forced merger between the only two competitors, SiriusXM generates $1.26 billion in free cash flow on $8.5 billion in revenue. The market values it at $9 billion. That is what a mature satellite subscription business actually looks like when the hype is gone.
The market is asking $1.75 trillion. That is Apple prices for a company with SiriusXM economics and xAI risk.
The Capex Treadmill
SpaceX is not See’s Candies. It’s Sun Microsystems with rockets. Sun reported billions in earnings for years. Every dollar was already committed to the next hardware cycle. Skip one cycle and you’re dead. Oracle bought the corpse for $7 billion. SpaceX’s constellation replacement cycle is that treadmill. It never stops.
Is There a Moat or a Puddle That Can Be Stepped Over?
Reusability is proven. Every serious launch company is now building reusable rockets because SpaceX demonstrated it works. Blue Origin has New Glenn. Rocket Lab is building Neutron. The head start is real. The exclusivity is gone.
The one genuine moat is orbital slot and spectrum allocation. SpaceX filed first, deployed first, and locked up premium low-earth orbital positions. Competitors face interference constraints that SpaceX doesn’t. That’s a real advantage.
It’s also a regulatory advantage, which means a government body can modify it. That has happened before in spectrum history.
Consider the benchmark. Tesla is Musk’s best execution — twenty years of manufacturing, iteration, and genuine engineering achievement. It survived only because a government tariff wall keeps Chinese competitors out of the American market. In the markets where the wall doesn’t exist, Xiaomi — a phone company — replicated Tesla’s core product in under two years and is now outselling it in China. Xiaomi didn’t invent anything. They assembled largely off-the-shelf components and applied their manufacturing expertise. Two years.
Mr. 10x is not some uber genius. Except for the rocket business, he is not even the best at anything I am aware of. Well, one thing — he is best at getting 10x valuations.
It is amazing how his fan boys seem to see something else.
If his strongest company, built over two decades, can be duplicated by a phone manufacturer in under two years the moment the market is open, what does that tell you about the durability of SpaceX’s advantages? Orbital slots are real. First-mover scale is real. Neither stopped Amazon, Blue Origin, and AST from entering the same market with serious capital. The moat is narrower than the price implies.
What the S-1 Actually Says
Read the filing and you find two things: a real rocket company, and an episode of the Twilight Zone.
The S-1 glossary — a legal document filed with the SEC — defines Kardashev Type II civilizations, lunar mass drivers, orbital AI data centers, and something called “Macrohard.” SpaceX filed under SIC code 7370: Computer Programming and Data Processing. Twenty-four years building rockets, and the moment they go public, they’re an AI company. The glossary isn’t accidental. It filters out investors who would ask uncomfortable questions about free cash flow and attracts believers willing to buy the dream at any price.
The investors lining up for this IPO are like the passengers in the Twilight Zone episode “To Serve Man” — boarding the rocket to a better life, certain they’ve been chosen for something extraordinary. “To Serve Man” turned out to be a cookbook.
Elon Musk’s S-1 isn’t a business plan. It’s a cookbook. And you’re the main course if you are buying into this valuation.
This also reminds me of something else — the people who waited in matching outfits for a spaceship that never came. Heaven’s Gate was also built on absolute conviction that the ship was coming. The ship did not come.
The S-1 also discloses that Musk retains 85.1% voting control through a super-voting share structure. You own the risk. He owns the company.
At $1.75 trillion, you’re not buying the business. You’re buying the dream — Mars colonization, point-to-point Earth transport, and orbital manufacturing.
Bon appétit.