SpaceX is going public in what will be the largest IPO in history at a $1.75 trillion valuation, making Elon Musk the world’s first trillionaire on paper. The company is hard to analyze with traditional financial frameworks because it bundles together a dominant rocket launch business, a fast-growing satellite internet service (Starlink), a struggling social media platform (X/Twitter), an AI company (xAI), and a chip manufacturing venture (Terrafab) — all under one roof. The hosts break down what each piece does, why it matters, and whether the price tag makes any sense.
The core business: launches
SpaceX builds rapidly reusable rockets and dominates the launch market, handling roughly 80–85% of all payloads sent to orbit.
They drove the cost of launching a kilogram to space down by roughly 50–100x compared to the pre-SpaceX era, and expect to cut costs in half again once their next-generation rocket, Starship, is operational.
About 40% of their launches carry their own Starlink satellites, meaning the launch business and the internet business are deeply intertwined.
Starlink: the cash cow
Starlink is a satellite internet service designed for places with poor or no connectivity — rural areas, remote regions, war zones, airplanes, and boats.
It has 10 million paying subscribers, generates about $11 billion in annual revenue, runs at roughly 40% EBITDA margins, and is recurring revenue with no real competitors.
The service is only about four years old. It was nearly killed off when Musk, frustrated by lack of progress from the Seattle-based team, effectively gutted the project and restarted it from scratch.
Growth is starting to slow in the US, and expansion into developing markets (Africa, India) will mean lower average revenue per user since those customers can’t pay US-level prices.
A newer product called “direct to cell” lets Starlink connect directly to ordinary cell phones without a dish, partnering with carriers like T-Mobile to fill in dead zones for a small monthly add-on fee. The total addressable market for cellular and home internet is roughly $2 trillion.
Starship: the big bet
Starship is SpaceX’s next-generation rocket, designed to carry 7–10 times more payload than the Falcon 9. It has not yet successfully flown and has been in development and testing for years.
If it works and becomes rapidly reusable — Musk has described a future of multiple launches per day, like airport operations — it would be one of the greatest engineering feats in human history.
Betting against Musk’s technical execution has historically been a losing bet, even when timelines slip by years.
Data centers in space: the next unlock
SpaceX’s investor presentation points to space-based data centers as the next major growth frontier. The idea is to put GPU clusters in orbit, powered by solar energy and cooled by the natural cold of space, then beam AI compute tokens back to Earth.
The pitch is that this could be dramatically cheaper than building data centers on land, where permitting, regulation, and community opposition (“red tape”) make construction slow and expensive — Musk has quipped that it’s easier to build in space than to get a county to approve a data center.
The physics question is whether this actually works. Musk’s track record is to take “no market risk” — meaning if he can build it, demand is guaranteed — and figure out the engineering.
If SpaceX becomes the lowest-cost provider of AI compute, the analogy is that they’d be the “Saudi Arabia of compute” the way Saudi Arabia dominated oil.
X (Twitter) and xAI: the awkward additions
X’s ad revenue is about $1.8 billion, down 40% from pre-acquisition levels and $100 million lower than last year. Total revenue including subscriptions and payments is about $2.8 billion, down from $4.5 billion when Musk bought it.
Musk has tried to make X’s data advantage feed into Grok (xAI’s chatbot), but Grok is well behind ChatGPT and Anthropic in users — roughly 100 million versus ChatGPT’s 1 billion.
xAI built Colossus, the largest GPU cluster for training AI, bigger than anything Google or Meta has. The problem is not enough users to fill it.
So Musk pivoted: he’s now renting out Colossus capacity to Google and Anthropic in deals reportedly worth a combined $20+ billion (roughly $1 billion per month each), on short-term contracts with 90-day cancellation notices.
SpaceX also has an option to buy Cursor (an AI coding tool doing $3–4 billion in revenue) for $60 billion, another attempt to play catch-up in AI through acquisition.
The financial picture
SpaceX generated $18 billion in revenue but lost $2.5 billion, with $6.6 billion in adjusted EBITDA. It burned $8 billion in cash last year and spent roughly $20 billion on capital expenditure.
The hosts are skeptical of the “adjusted EBITDA” figure, noting it adds back real costs like depreciation and stock-based compensation — the kind of adjustments Warren Buffett and Charlie Munger famously criticized.
The company also holds about $2 billion in Bitcoin on its balance sheet.
Ownership and who gets rich
Musk owns 42% of the company with 85% voting control — remarkable for a capital-intensive business that’s been raising money for 20 years. For comparison, Box’s Aaron Levy owned about 4% at IPO.
The second-largest individual shareholder is Antonio Gracias (~7%), a longtime friend and operational advisor who helped Musk with production bottlenecks at Tesla and SpaceX and once loaned Musk $1 million personally when Tesla was on the brink. His stake is worth roughly $90 billion.
Gigafund, founded by Luke Nosek (a PayPal co-founder) and Steve Jurvetson, was built on the simple thesis of backing every Elon company. Jurvetson was also an early investor in Tesla and Bitcoin and never sold his SpaceX shares.
The Ontario Teachers’ Pension Fund invested in 2019 and is set to make $12 billion — roughly $33,000 per teacher in the fund — despite Ontario’s public boycotts of Musk.
The IPO is expected to create over 4,000 new millionaires among SpaceX employees, including some cafeteria workers who had stock options.
Sam Bankman-Fried had an early SpaceX stake that would have been worth $15 billion today; his total portfolio of illegal investments would have been worth $114 billion, which would have made him one of the greatest investors of all time. Those shares were liquidated in bankruptcy.
Musk’s compensation package
Musk’s base salary is zero. His compensation comes through two massive stock grant awards:
The Mars Award: 1 billion shares (~$135 billion at IPO price) if SpaceX reaches $7.5 trillion in market cap AND establishes a self-sustaining colony of at least 1 million people on Mars.
The AI CEO Award: 300 million shares if the company reaches $6.5 trillion in value AND delivers 100 terawatts of compute per year from non-Earth data centers — which would be 100 times the entire current US electrical grid.
The total potential grant value is roughly $750 billion. The hosts note that the biggest risk to shareholders isn’t whether the engineering works — it’s whether Musk lives long enough to see it through.
The mission
SpaceX’s stated mission is “to make life multiplanetary and extend the light of consciousness to the stars,” creating species-level redundancy so humanity doesn’t share the fate of the dinosaurs.
The hosts compare this to Meta’s mission of “connecting people” and note that for most companies, the real mission is simply to move money from customers’ bank accounts to the founders’. SpaceX’s mission is either inspiring or evidence of a god complex, depending on your perspective.
How to think about the valuation
At $1.75 trillion, the company is priced at roughly 100x revenue — a multiple that makes traditional investors uncomfortable.
To justify the price, you’d need to believe that: Starlink grows from $11B to $30–50B in revenue within five years; Starship works and scales; space-based data centers become real; and Musk continues to execute on seemingly impossible engineering goals.
The hosts invoke Charlie Munger’s framework: “a wonderful business at a silly price.” You can admire it from afar without needing to own it.
They also note that “Elon assets” trade at a premium that defies conventional analysis — the “price-to-Elon ratio” — and that betting against his eventual success has been one of the most unprofitable bets in modern investing.
Nuggets and trivia
SpaceX paid a vendor in shares with a strike price of $420 (or similar numbers referencing 420), and the 409A valuation of Twitter was set at $42 billion — Musk’s well-documented obsession with the number 420.
The hosts ran their analysis through Claude, asking it to explain SpaceX as if they were Charlie Munger: “Is it in my circle of competence? How do I lose money here? Show me the incentives.”
Steve Jurvetson, an early SpaceX and Tesla investor, maintains a Flickr photo album documenting early rocket launches, Bitcoin mining rigs, and other technology milestones with detailed captions — described as an underrated resource for technology nerds.
