In just a couple of weeks, SpaceX will IPO. If you've been planning to buy SpaceX stock, it might be a good idea to peruse the company's prospectus before investing -- if you have the time.
Granted, at 380 pages in length (including footnotes), the SpaceX IPO prospectus is not an easy read -- or a quick one. The prospectus also spends a lot of time telling us things we already know about the most famous Elon Musk company that isn't named Tesla (TSLA 2.75%).
Image source: Getty Images.
Things I already knew
Not everything in the SpaceX IPO prospectus is actual "news" -- not to those of us who've been following the company for years. SpaceX leads off the document with a series of boasts about how SpaceX has launched more than 650 missions to space, 85% of which used reusable rockets, and put 78 humans in orbit.
Boasts about the wildfire success of Starlink, which began launching satellites in 2019, went live in 2020, and today possesses 9,600 satellites in orbit, serving 10.3 million customers in 164 countries.
SpaceX even boasts about the artificial intelligence (AI) business that it has shoehorned into the SpaceX IPO, a business that has accumulated 550 million monthly active users since 2023 -- and that lost $6.4 billion last year.
But if you keep reading and push past the braggadocio, some of the things you find in the prospectus just might surprise you. Here, at least, are three things that surprised me.
1. Tesla isn't part of the SpaceX IPO (except that it kind of is)
Tesla, per se, is not part of the SpaceX IPO. But the more you read, the more you realize that Tesla is somehow integral to SpaceX, being mentioned 87 times throughout the prospectus.
A glossary of terms used in the prospectus, for example, mentions "Megapack" as "a containerized, utility-scale lithium-ion battery energy storage system produced by Tesla and designed to stabilize power grids, store renewable energy, and replace fossil fuel" in the SpaceX AI division.
Later, the prospectus refers to "Terafab, a chip manufacturing initiative with Tesla and Intel," to an "agentic AI platform" called Macrohard that SpaceX is developing in cooperation with Tesla, to certain Tesla chips being "optimized for the space environment," and in general to collaboration through "shared engineering resources, intellectual property, and infrastructure across Tesla and SpaceX." The prospectus even hints at plans to "deepen our strategic collaboration with Tesla."
So no, Tesla isn't part of the SpaceX IPO... except that it kind of is.
NASDAQ: TSLA
Key Data Points
2. Marvel has a Hulk. SpaceX has the Algorithm.
SpaceX famously offers the lowest launch prices per kilogram to Earth orbit. Partly, it can do this because SpaceX possesses the only reusable rockets on Earth. Partly, it's because SpaceX practices what it calls "the Algorithm," a "five-step iterative process" the company uses to rapidly innovate and optimize, emphasizing:
- Making the requirements "less dumb"
- Deleting unnecessary processes or parts
- Optimizing necessary processes or parts
- Accelerating cycle timesteps
- Automating proven processes only after the first four steps are completed
While the effects may be hard to quantify, building the Algorithm into everything SpaceX does does seem likely to drive down costs, helping SpaceX to underprice rivals even as it grows profit by reducing unnecessary costs.
The Algorithm also generated some unexpected good PR for SpaceX when, in 2024, then-United Launch Alliance CEO Tory Bruno complimented the resulting ultra-clean design of the SpaceX V3 Raptor engine, saying SpaceX had "done an excellent job making the assembly simpler and more producible" -- enough so that he struggled to believe photos of the engine depicted a complete product.
3. Starlink is even more important than we thought
Until the prospectus arrived, the best data on SpaceX came from industry experts such as Payload Space. As recently as January, Payload estimated Starlink/Connectivity generated $10.4 billion of SpaceX's $15 billion in 2025 revenue (not counting xAI and X, which weren't really considered part of SpaceX at the time).
The prospectus shows that Starlink is even more important than that, generating $11.4 billion of the $15.5 billion in revenue attributed to "Space" and "Connectivity." (SpaceX's now-included AI business, comprised of X and Grok, added $3.2 billion to yield $18.7 billion in total revenue across all three business segments.)
|
Segment |
2024 Revenue |
2025 Revenue |
2024 Operating |
2025 Operating |
|---|---|---|---|---|
|
Space |
$3,796 |
$4,086 |
$21 |
($657) |
|
Connectivity |
$7,599 |
$11,387 |
$2,006 |
$4,423 |
|
AI |
$2,620 |
$3,201 |
($1,561) |
($6,355) |
|
Total |
$14,015 |
$18,674 |
$466 |
($2,589) |
Data source: SpaceX IPO prospectus, all numbers in millions. SpaceX earned a net profit of $791 in 2024. SpaceX's net loss in 2025 was $4,937.
Clearly, Connectivity is SpaceX's most important part and will become more so once Starship comes online and begins launching ever-larger Starlink satellites. In late 2026, SpaceX will begin deploying next-generation V3 satellites designed to offer one Tbps download speeds. In 2027, SpaceX will begin deploying its next-generation V2 Mobile satellites to offer mobile "broadband data and IoT connectivity."
This will grow the company's Direct-to-Cell business, which currently uses 650 V1 Mobile satellites to serve 7.4 million monthly unique devices across 30 countries, further strengthening SpaceX's lead over rivals in DTC services such as AST SpaceMobile (soon) and Blue Origin (presumably).
Had SpaceX opted to IPO Starlink alone -- as it once promised to do -- there'd be no question this is a good business worth buying. But even with the money-losing AI business attached, I still think the SpaceX IPO is worth a look.
