I am excited about the SpaceX IPO for all the reasons investors shouldn’t be. Maybe it’ll be a real marquee moment for Silicon Valley, but I see the potential for a shitshow. After all, more than a decade ago, Musk said that SpaceX going public before going to Mars would be bad for the company.
Are private markets tapped out on cash to fund SpaceX ambitions? Elon Musk has been very clear about his feelings on publicly traded companies. Specifically: He doesn’t like them!
“I am hesitant to foist being public on SpaceX, especially given the long term nature of our mission.”
In 2013, Musk sent an email to SpaceX, which his biographer Ashlee Vance reprinted in his book, saying he didn’t want to take the company public until his Mars transport system is in place. Tesla went public because it “didn’t have any choice,” Musk wrote in a memo to SpaceX employees in 2013. “I am hesitant to foist being public on SpaceX, especially given the long term nature of our mission.”
An IPO can raise a lot of money for a company while letting longtime investors exit. But there’s a price. It’s possible the private market has overvalued the company or the financials don’t look as good as everyone hoped. And it’s much easier for investors to bail on a public company than a private one.
There are other things Musk cited in his 2013 memo, too. “Public companies are judged on quarterly performance,” he wrote. If SpaceX had a bad quarter as a public company, “short sellers would be hitting us over the head with a large stick.” The stock also would get beaten up every time something went wrong with a SpaceX rocket. (In non-Trump administrations, public companies are also scrutinized more closely by financial regulators than private ones — though Trump’s SEC seems to exist largely to get rid of old SEC cases.)
Data centers in space are the purported reason for acquiring xAI
Tesla has an inflated share price because of Musk’s fandom, which effectively precludes activists from taking over the board and ejecting its CEO; also, it is a mess. Given its business with the government, SpaceX could be worth taking a risk on. If SpaceX goes public, there’s a world in which a firm like Elliott Management buys shares and starts agitating. SpaceX’s sheer size may make that difficult for now, but if it has a couple bad quarters, that could change.
Musk’s positive case for going public right now is to put AI data centers in space. Well, Musk loves saying weird shit he won’t follow through on. Take, for instance, the part of the 2013 memo about SpaceX not going public until the Mars transport system is in place. It isn’t. In fact, Musk recast his Mars ambitions to the Moon, perhaps because he recently discovered the Moon is the celestial body actually shown on his “Occupy Mars” shirt.
Data centers in space are the purported reason for acquiring xAI, Musk’s tremendously money-losing home of the social media platform formerly known as Twitter and its boutique child porn generator Grok. Maybe adding X and xAI to SpaceX will “make a sentient sun to understand the Universe and extend the light of consciousness to the stars!” SpaceX alone would have been a pretty clean IPO, one that would be easy to make a case for. But SpaceX plus xAI is a mess, not least because a number of governments are very unhappy about the aforementioned child porn.
Musk has a history of buying companies that Elon Musk is the largest investor in; take Tesla’s 2016 Solar City acquisition. A number of Tesla investors sued, saying that the acquisition amounted to a bailout of the failing solar panel company — one meant to prop up Musk’s investment. Those investors ultimately lost the suit. But it does raise my eyebrows about the xAI acquisition of X, and the SpaceX acquisition of xAI.
So yes, I have some thoughts on why we are seeing this messy IPO, and all of them are more cynical than data centers in space and sentient suns.
I’m not sure about SpaceX’s overall cost basis but I know xAI is lighting money on fire
First, Musk needs to help out the investors in his debt-loaded Twitter takeover, which was notoriously overvalued even before it closed and became the worst merger for its banks since the financial crisis, until the Trump administration reversed their fortunes. So I don’t think it’s too far-fetched that Musk is now trying to pay off his equity investors. Smushing X into xAI gave those investors a slice of the more valuable company as a thank-you for their trouble. And, in turn, rolling the money-losing xAI into SpaceX gives them an even bigger return on their investment, with the IPO poised to give them an opportunity to sell, realize their profits, and put the whole thing behind them.
There’s also some maneuvering on SpaceX’s part that suggests it’s trying to prop up its value using a “fast entry” process to financial indexes such as the Nasdaq. Nasdaq is mulling a rule change that would allow entry into the index after 15 days — instead of months — if the company in question is large enough. SpaceX is targeting more than $1 trillion as its IPO valuation, which would qualify it to get added under the rule change, and getting added to the index means that SpaceX is automatically added to index funds. That makes it even easier for SpaceX insiders to unload on retail investors, and means Musk won’t have to rely on just his faithful cultists to boost shares.
Second, Musk wants to go to market before Sam Altman does. Do I need to recap this entire stupid slapfest for you? (I am about to drag my ass to an Oakland courtroom to watch them bicker expensively at the end of April). Honestly, rich people should have better hobbies, like building spectacular architecture that will outlast their miserable families.
And third, it’s reasonable to conclude that SpaceX needs money. That’s what forced Musk’s hand into going public with Tesla, after all. Now, you may be suggesting that Musk really needs those funds for his data centers in space, in which case, I’d advise you to continue waiting for the Hyperloop to get built. I don’t think that’s the issue. I’m not sure about SpaceX’s overall cost basis but I know xAI is lighting money on fire.
“No one has ever scaled like that outside of SpaceX.”
Let’s think about SpaceX’s big moneymaker Starlink by itself for a minute, actually; maybe that will clarify if the company needs money even without an xAI acquisition. Starlink was a surprising advance on the old satellite model, says Caleb Henry of market research firm Quilty Space. Once upon a time, you parked a satellite in geostationary orbit. On paper that satellite lasts 15 years; in reality, it might last as much as 20. You make back your costs in seven years and then you just print money.
Starlink emerged as the streaming wars upended TV broadcast — and, indirectly, the satellite business. TV broadcast revenue was drying up for satellite providers, and the geostationary style of satellite business came under competition from constellations like Starlink.
SpaceX has also had impressive cost reductions when it comes to its equipment, Henry notes. The user terminal was initially $3,000 to build, but as production increased, the price dropped — and now SpaceX says it’s no longer taking a loss on those. “No one has ever scaled like that outside of SpaceX,” Henry says. “When they hit production in a week that the rest of the industry did in a year, the cost calculation changed.”
Henry says he thinks Starlink is making money, and will make more money every year. Starlink has expanded the markets it’s in — from consumers, to militaries (especially in Ukraine), to aircraft and maritime markets.
Starlink is assumed to be the major revenue driver for SpaceX
But as Starlink has grown, it’s also slashed prices — and possibly its margins, according to The Information. Meanwhile, a new competitor is about to enter space: Amazon, which has partnered with AT&T on Leo, a new satellite internet service. Amazon, of course, is very, very good at customer acquisition, and AT&T is already an internet provider. This is to say nothing of potential competition from China and the EU. Musk’s train wreck of a personality may give his competitors an edge, especially in countries outside the US. And while SpaceX has an impressive edge on the competition now, Musk has squandered his impressive head start before.
Starlink is assumed to be the major revenue driver for SpaceX, with a 2024 Morgan Stanley report projecting that the company’s revenue would reach $19 billion in 2025, on the strength of 6 million Starlink subscribers, The Information reported. But SpaceX’s revenue last year was only $16 billion, despite 9.2 million Starlink subscribers. Oops!
In 2023, Musk said Starlink had “breakeven cash flow,” but Bloomberg reported in 2024 that “people with knowledge of Starlink’s balance sheet say money-making quarters have been less consistent than Musk suggested to investors.” Quilty Space estimates that Starlink generated about half a billion in free cash flow in 2024, its first cash flow positive year. I’m saying “cash flow” instead of “profits” because those numbers don’t take into account lifetime expenses for Starlink in terms of, say, R&D and so on. Incidentally, this is exactly the kind of thing we’re all going to find out about during the IPO process.
There are kind of a lot of questions around Starship because it’s now a choke point
The crucial question is Starlink’s costs. Henry figures that satellite constellations like Starlink could cost between $5 billion and $20 billion. SpaceX’s V2 satellites, the current version, cost between $800,000 and $1 million to make, according to Quilty Space estimates. V3, which has not yet been deployed, is estimated to cost $1.2 million per satellite. An FCC filing suggests the V3 will weigh 2,000kg, or about 4,400 pounds — about three and a half times as much as V2. The weight difference matters because it suggests that SpaceX must rely on its new Starship, rather than the Falcon 9, to launch its heavier satellites.
The Falcon 9’s launch costs are probably below $20 million per launch, according to Quilty Space. Starship has more question marks around it.
Starship’s lifetime R&D costs are likely $10 billion, according to analysis from Payload Pro. Manufacturing the full stack for it probably costs $90 million, but launch costs could plausibly drop to less than $10 million per launch if Starship is fully reusable, according to the same report. There are kind of a lot of questions around Starship because it’s now a choke point: There are V3 satellites just sitting around waiting for their launch, says Henry. Meanwhile, Starship, also on V3, hasn’t yet shown it can launch a hundred or more metric tons. “For Starship to be the answer to all of SpaceX’s problems, it needs to be as powerful in reality as has been contemplated on paper,” Henry says.
The program has had some hiccups — most notably, a series of explosions
Starship doesn’t just matter for Starlink; it also matters for NASA and other contracts. There should be a test flight of Starship in the next “four to six weeks,” according to SpaceX president Gwynne Shotwell. SpaceX estimates Starship will be ready to launch a new constellation of Starlink satellites in “mid-2027,” said Michael Nicolls, SpaceX’s senior vice president said at Mobile World Congress. This all, of course, assumes that Starship development goes according to the plan. The program has had some hiccups — most notably, a series of explosions.
We may very well find out more with the financial document known as the S-1, which gives basic business information on the company and is required for a public offering. Will Starship be fully reusable? How much risk is there for continued explosions? These are pretty significant variables to consider when it comes to trying to figure out SpaceX’s future trajectory; if only part of Starship is reusable, launch costs will be higher. Also, explosions are generally bad.
These are pretty big things to be uncertain about, and if I were an investor, I’d be taking Musk’s projections with a pillar of salt. I mean, this is the guy who said “demand is off the charts” about the Cybertruck, a vehicle that didn’t meaningfully contribute to Tesla’s 2024 revenue and then flatlined in 2025. Musk once projected that he’d sell 250,000 Cybertrucks annually; as of last October, Tesla had sold an estimated total of less than 60,000 in 2024 and 2025. Investors have given Musk a massive amount of leeway on what appear to be fairly routine overestimates, bad projections, and sometimes outright lies. I wonder if that will extend to SpaceX, too.
But there’s one other factor working against the IPO: Musk’s ego
Still, despite the risks, there’s reason to think that SpaceX has some real possibility for rewards. I bring all this up because SpaceX is the good part of the business. xAI, meanwhile, burned through almost $10 billion in cash in the first nine months of last year to achieve… $210 million in revenue. This is to say nothing of the $17.5 billion in debt (which may incur penalties if it is paid back early), the fun new regulatory risks, and the lawsuits that xAI also tacks on to SpaceX. Maybe xAI is Musk’s version of a poison pill to keep activist investors away, who knows!
Sure, beating the competition to market, especially when that competition is Sam Altman, is fun. But that seems like a bonus. From where I’m standing it sure looks like SpaceX, now that it has xAI weighing it down, needs money. SpaceX executives expect to raise $50 billion in the IPO, making it twice the size of the previous largest-ever IPO. As with Tesla, Musk is probably relying on the enthusiasm of retail investors — but this time, he’s trying to shore up SpaceX with index fund money, too.
Whether the IPO will be successful is an open question. Separately from any doubts investors may have about SpaceX’s finances, Musk has alienated a lot of people over the last two years through his political ventures — doubtless another reason to target index funds, which are a basket of investments designed to mirror a benchmark such as the Dow Jones Industrial Average, or the Nasdaq — and investors’ moods may sour on AI. But there’s one other factor working against the IPO: Musk’s ego.
Whatever number SpaceX actually needs is going to be large, but whatever Musk is going to want will be larger
Remember the WeWork IPO? That S-1 was magnificently insane: a tech-company valuation of $96 billion on what was essentially a landlord, plus a bunch of other fun stuff that benefitted only the company’s CEO and founder. It was insane because WeWork was insane. “WeWork was interviewing bankers to lead its IPO, and the bankers all came in and said words to the effect of ‘we think you are great, we understand your story and want to be the ones to tell it, and we think you are worth a lot of money,’ in order to convince WeWork to hire them,” explained Bloomberg’s Matt Levine. And that’s how the $96 billion valuation happened — the bankers had no incentive not to make their client, WeWork, happy. Anyway, investors got the S-1, laughed until they cried, and didn’t invest; WeWork had to yank its IPO, and then file for bankruptcy. Not ideal.
Consider for a moment Musk’s megalomania. What numbers do you suppose Bank of America, Goldman Sachs, JPMorganChase, and Morgan Stanley gave Musk on SpaceX? Whatever number SpaceX actually needs is going to be large, but whatever Musk is going to want will be larger. Will it make sense?
Maybe, and maybe SpaceX’s finances will be great, and I will write something about how surprisingly normal and sensible the SpaceX S-1 is, and we will all move on with our lives. But given all the agita around the extremely overvalued Twitter acquisition, Musk’s “first buddy” status, the Starship explosions, and the AI bubble fears, plus whatever the fuck the Iran action does to energy markets, Musk may have seen the future back in his 2013 memo. To borrow a phrase, SpaceX may soon be getting hit with very large sticks.
