An essay in scrolls

Escape
Velocity

SpaceX's IPO prepares to defy gravity — physical and financial.
Scroll

The Offering

The Largest IPO Ever.

SpaceX priced its debut at a $1.77 trillion valuation, edging Saudi Aramco's $1.7T for the largest in history. Each block is $25 billion. Aramco was the only company that ever came close; the famous tech IPOs barely register.

The Offering

But You Can't Buy Much.

Of that $1.77 trillion, only about 4% (roughly $75 billion) is sold to the public. The other 96% stays locked with insiders. Hold that thought; it matters later.

The Claim

Real, Fast Growth.

SpaceX's revenue is climbing fast: $4.6B in 2022, $8.7B in 2023, $14.0B in 2024, $18.7B in 2025. Roughly quadrupling in three years, even as the growth rate cools.

The Claim

Then This.

To justify the $1.77 trillion price, Morgan Stanley (a co-lead underwriter) points at 2040: $3.4 trillion. Zoom both axes out to fit it, and the last four years collapse to a sliver. It's 182× what SpaceX sold last year.

The Naive Read

41.5% a Year. For Fifteen Years.

Compounded, that's the rate that turns $18.7 billion into $3.4 trillion. Aggressive — but is it actually unprecedented?

The Naive Read

Tesla Climbed Steeper.

Plot the rates as curves and Tesla's is the steeper one: 62% a year versus the 41.5% SpaceX needs. By rate alone, SpaceX is the tamer story. So the rate is not the tell.

The Reframe

A Lower Bar.

Drop each climb to a single point: its fifteen-year growth rate. SpaceX's lands below Tesla's, 41.5% against 62%. On rate, SpaceX is the lower bar, not the higher one. So how does it sit against the field?

The Reframe

Against the Field.

Stack the other great compounders by the same measure: fifteen-year growth, every point still at the same starting line. A handful cleared SpaceX's rate: early Tesla, Amazon, Cisco. Most never did. By rate alone, 41.5% is high, but not impossible.

The Reframe

From 160× the Base.

But not every rate is earned from the same place. Give the points a second axis (starting size) and they fan out. The ones that out-grew SpaceX were all small; Tesla ran 62% from $117 million. SpaceX needs 41.5% from 160 times that base.

The Frontier

Growth Has a Speed Limit.

Plotted against size, a shape appears, and it bends down. The bigger you start, the slower you're allowed to grow. The curve holds whether you measure the 1980s or the 2020s.

The Shape

The Growth Frontier.

Fit a curve and it's blunt: starting size explains about half of who grows fast (R² ≈ 0.53). Not a hard law, a stubborn trend. Growing fast from a giant base is like climbing out of a deeper gravity well — the heavier the body, the more velocity it takes just to break orbit. Size becomes a rate-limiting factor to sustained velocity.

The Outlier

Measure the Gaps.

Measure how far each company sits from the frontier: its actual growth ÷ what the trend predicts. That gap is the residual, how much it beat the speed limit or fell short.

The Outlier

Rank Them.

Lift the gaps off the cloud and set them aside, then line every company up, largest overshoot to smallest.

The Outlier

Most Sit Near 1.0×.

Ranked, almost every company lands close to 1.0×, having grown about as fast as its size predicted. A handful clear the bar; one sits in a class of its own.

The Outlier

Off the Manifold.

Collapse those scores into a distribution. Tesla's 1.49× is the record — just inside the outlier fence. SpaceX's required 2.15× falls beyond it: a statistical outlier, ~44% past the best the data has ever produced.

The Tell

79% Margins. At $3.4 Trillion.

The forecast also assumes a 79% EBITDA margin. Aramco, pumping the cheapest oil on Earth, manages 55%. Software tops out near 45%. And $3.4 trillion would be five times Walmart, the most revenue any company has ever booked. One firm. Building rockets.

The Scale

Six Percent of Everything.

Run the headline forward: SpaceX at 41.5% while the U.S. economy grows at its consensus ~3.7% nominal. By 2040 a single company's revenue would equal ~6% of U.S. GDP, beyond the share Walmart commands today. A hundredfold rise in fifteen years.

The Machine

Forced to Buy.

Nasdaq deleted its float minimum and built a 15-day fast track for the 40 largest firms. SpaceX floats under 5%. Funds tracking QQQ, IWM and FTSE are forced to buy it, roughly $60 billion by Goldman's estimate, squeezed through a tiny float, setting a price before the market finds one.

The Machine

Someone Sells Into It.

That forced bid needs a seller. When the lock-ups expire 90–180 days later, insiders sell into the demand the index created. The cash flows out to them. The public is left holding the position, bought at a price set for it.

The Thesis

Coherence Isn't Truth.

Coherence is cheap — a roadshow is built to manufacture it. Real businesses, a vast market, a growth rate with precedent, arranged until the story stops feeling like a forecast and starts feeling like a fact. But a coherent story isn't a true one, and this one doesn't have to be. It only has to hold long enough for the index to buy and the lock-ups to lift. The value was never in 2040. It was always in the rebalance.

The Twist

The Elon Frontier.

Two companies sit far above everyone else's frontier, both Musk's. Tesla earned its place with revenue it delivered; SpaceX has only been priced to match. Connect them and a line of their own appears. Two points don't make a law, but it's the line the market has drawn. So the question stands: is Elon Musk on a frontier of his own — or is the market extrapolating one proven outlier onto an unproven one?

A line is not a law.