Editor’s Note: My team has been building a new AI trading algorithm and seeing some amazing results.
113% in NEXTERA in 2 days. 100% in Gilead Sciences in 3 days. 151% in Microsoft in 2 days.
You can start getting these trades directly in your inbox.
SpaceX priced its IPO at $1.77 trillion, making it one of the most valuable companies in the world on its first day of trading.
A new analysis from Milk Road AI shares the roadmap for SpaceX to become the first company valued at $100 trillion.
I know that sounds crazy. A 56x increase – after the richest IPO in history. So let me actually walk through the argument – and break it down.
The one big idea
Every once in a generation, a company emerges that doesn't compete within an industry — it owns the input cost of an entirely new economy.
Standard Oil owned the cost of energy, and everything that ran on energy became a customer. TSMC owns the cost of computation, and every company building on computation pays a toll. Owning the input cost, rather than competing inside an industry, is the rarest compounding advantage in business history.
SpaceX owns the cost per kilogram to low Earth orbit. And it has been cutting that cost by roughly one order of magnitude per decade for twenty years:
Space Shuttle: ~$54,500 per kilogram
Pre-Falcon expendable rockets: ~$18,000
Reused Falcon 9 today: as little as $600
Starship target, fully reusable at scale: ~$100
That's a 99.8% reduction from the Shuttle era — with the final leg still ahead.
Want to get this week’s urgent AI trade alerts? Simply go here right now.
Why the cost curve matters more than any product
At $600 per kilogram, you can launch a communications satellite. At $100 per kilogram, you can launch things that were never viable businesses before: an orbital data center, a pharmaceutical platform, a solar array that beams power to Earth, a cargo route to the Moon.
Every order-of-magnitude drop doesn't just make existing industries cheaper. It creates new ones.
The launch business is bigger than it looks
The S-1 reports three segments: Space at 22%, Connectivity at 61%, and AI at 17%. Read quickly, that makes launch look like a mid-sized contributor to a satellite-internet company. That's backwards.
In 2025, 122 of SpaceX's 165 Falcon 9 launches deployed Starlink satellites. Those launches generated zero Space-segment revenue — the cost gets capitalized into Connectivity. The reported Space number only captures what SpaceX charged outside customers.
So the most strategically critical thing the company does — launching its own constellation — is essentially invisible in the reported numbers. Without launch there's no Starlink. Without Starlink there's no cash flow. Without cash flow there's no Starship.
And the lead is hard to copy. In 2025 SpaceX flew 165 orbital missions — roughly half of every orbital launch on Earth — on reusable boosters, its most-flown booster completing 34 flights. China's entire state and commercial industry accounted for 11% of global mass to orbit. Even Amazon's Kuiper is years behind.
Starlink is the cash engine
Starlink generated an estimated $4–5 billion in EBITDA on $11.4 billion of revenue in 2025 — a 35–45% margin on a business that grew 105% year over year. As of 2026 it had 10.3 million subscribers across 164 countries, up from 5 million a year earlier, running about 9,600 satellites.
The economics legacy telecom can't match are rarely spelled out. Verizon, AT&T, and Comcast are effectively construction companies — fiber alone often costs $500 to $1,500 per household passed. Three billion people lack reliable internet not because no one wants to serve them, but because the math never worked.
Starlink's cost per home passed, once the constellation is built, is functionally zero. Every satellite covers every home in its footprint, and the only marginal cost is the user terminal.
That's the flywheel: Starlink EBITDA funds Starship, Starship deploys next-gen satellites far cheaper, that drives more subscribers and more EBITDA, which funds the next Starship. The base case has Starlink at 240 million subscribers and over $600 billion in annual revenue by 2040 — bigger than any telecom company on Earth today. For context, the combined market cap of every telecom company in the world is about $3.1 trillion.
Two bets that change the size of the prize
Direct-to-Cell. As of 2026, Starlink Mobile already served 7.4 million monthly devices across 30 countries — text, light data, and voice on ordinary smartphones, no hardware change. In May 2026 the FCC approved SpaceX's $17 billion acquisition of EchoStar's spectrum, the missing piece for full mobile service from orbit. That moves the target from underserved households to 5.5 billion global mobile subscribers.
AI in orbit. This is the strangest segment. The S-1 disclosed two cloud deals signed fifteen days apart:
Google: $920 million per month
Anthropic: $1.25 billion per month
Combined, that's about $26 billion annualized — roughly 80% of SpaceX's 2025 revenue base — from two clients, both cancelable on 90 days' notice. Treat that concentration as a real risk.
But read it the other way: Anthropic has access to every compute option on Earth, and it decided orbital capacity was worth $15 billion a year. Power and cooling are now the binding constraint on AI — US grid generation has grown under 3% a year since 2023 while data-center demand roughly tripled. Orbital data centers solve it from first principles: continuous solar, free cooling in vacuum. SpaceX targets 2028 for its first deployment.
The government moat
About 20% of revenue comes from NASA, the Space Force, the NRO, and the growing Starshield program. It's the same pattern that kept SpaceX alive in 2008 — government contracts fund a capability, then SpaceX scales it commercially.
The honest part — where the bull case lives and dies
None of this works without Starship reaching reliable, reusable cadence. SpaceX has spent more than $15 billion on it. The most recent V3 flight deployed mock satellites and splashed down its upper stage on target — but the Super Heavy booster failed its landing and went into the Gulf. Musk himself lists failure to reach bi-weekly cadence as a genuine bankruptcy risk in the S-1.
The estimated odds of reaching operational cadence by 2030 are about 75%. Which means 25% of the time, you own the bear case. And here's what makes the setup unusual: even the bear case models out to roughly a $35 trillion company by 2060 — about a 9% annual return from the IPO price.
Does the math to $100 trillion actually hold?
It requires a 26.5% revenue compound rate sustained for 35 years. That rate isn't unprecedented:
Amazon: 27% for 25 years
Apple: 26% market-cap compounding for 20 years
TSMC: 23% for 30 years
What's novel is the duration — and the fact that most of the terminal revenue comes from industries that don't exist yet.
Will SpaceX become a $100 trillion company in my lifetime? Nobody knows.
Given Elon Musk’s massive success – with Tesla and SpaceX – I wouldn’t bet against him. And given SpaceX ambitious plans – I think the stock is a “must own” after the IPO.
That being said, the stock will be volatile after the IPO. It’s always prudent to dollar cost average into positions. And position sizing is an important way to manage risk.
Have you been using AI to select trades?
Wall Street’s biggest banks and most profitable hedge funds are making huge investments in AI. Yet they’re not making their profitable trading algorithms available to regular investors.
That’s why my team is building the AI Profit Predictor. It’s a new way to find trades with a high probability of profit. And I’d like to share it with you.
Ian Wyatt
Editor, Daily Profit