SpaceX is planning a monster initial public offering (IPO). Elon Musk is reportedly seeking to raise some $75 billion, at a valuation of $1.75 trillion, next month. It will be both his birthday and a moment when Venus and Jupiter will be in alignment. It would be the biggest IPO in history by far, utterly dwarfing Saudi Aramco’s $29 billion figure back in 2019.

Musk might sell those shares, too. One reason why, despite the ludicrous valuation, is that stock indices are changing their rules to allow SpaceX to join almost right away and with fewer conditions, thus forcing investors who follow a passive strategy, like index funds and many pension funds, to buy the company’s shares.

More from Ryan Cooper

The NASDAQ stock exchange used to require a company to have at least 10 percent of its stock publicly traded, and see it traded for at least three months, before being able to join. Those rules are now gone; SpaceX will be eligible after just 15 trading days and with less than 5 percent of its stock available to the public. The S&P 500 reform now under discussion is removing the requirement that companies be profitable (as SpaceX is not, as will be seen below)—but only for the largest 100 companies.

“Historically, the function of IPOs has been to allow insiders to cash out,” said economist J.W. Mason, co-author of the recent book Against Money (for which, full disclosure, I provided a blurb). Thanks to these rule changes, a large fraction of the bag holders will be passive investors—which is to say, basically everyone saving for retirement. “It will be impossible to not own them … This will be really helpful for demand,” one adviser to the deal told the Financial Times.

Stock indices are changing their rules to allow SpaceX to join almost right away and with fewer conditions.

Let’s start with SpaceX’s business fundamentals. The company can reasonably argue that it is by far the most important rocket company, accounting for more than 80 percent of all space launches in the world, with both commercial and government contracts. It is also the biggest player in the space-based internet business—and the two sides are interlinked, with a reported 80 percent of SpaceX’s recent launches being for Starlink satellites.

It is questionable just how valuable this business model is, however. Space-based internet comes in handy in many circumstances, like the war in Ukraine or in rural locations. But any person living in a developed country in or near a city very likely already has access to fiber broadband that essentially has to be better and cheaper than Starlink, or could easily be made so. A wired connection will always be much more reliable and have far less latency than a wireless one, particularly if the signal has to go all the way into space and back. And while SpaceX has gotten the cost of rocket launches down by a lot, it’s not anywhere close to the price of one guy and a van full of tools to hook your house up to the local trunk line.

Conversely, while Starlink would be most useful for people in poorer nations without broadband, such people are generally not going to be able to afford high subscription costs. There is also competition afoot, both from European and Chinese firms abroad, and Jeff Bezos’s Blue Origin company domestically.

Even if SpaceX did manage to outcompete Verizon, Comcast, and everyone else, broadband internet is not some Facebook-esque geyser of profits. Verizon made $17.6 billion in profits last year, across its entire internet, cable, and phone business—a respectable sum, but less than a third of what Meta pulled in.

But wait, it gets worse! Last year, on its core operations, SpaceX supposedly made $8 billion in profit—about as much as Unilever—on $15 to $16 billion in revenue. But because Musk merged SpaceX with xAI (formerly known as Twitter), which is torching money developing and running his Grok AI model—its main apparent selling point, aside from calling itself “MechaHitler,” is generating deepfake nudes and child pornography—the company as a whole actually lost about $5 billion. Indeed, despite the enormously expensive (and polluting) build-out of the “Colossus” data center in Memphis, demand for Grok is so low that Musk has been renting out some of its compute capacity to Anthropic instead. People often poke fun at Tesla’s ludicrous price-to-earnings ratio, but in the case of SpaceX such a calculation is mathematically impossible.

The deal gets worse still when we see how Musk has structured it. Shares available to the public will not actually convey any ownership rights whatsoever. As with Mark Zuckerberg and Meta, Musk has drawn up a two-class share structure giving him permanent power over the company.

So how on earth is Musk justifying this valuation? The same way he always does: highly suspect promises about utopian sci-fi schemes just over the horizon. With Tesla, it was humanoid robots that will turn the company into a $25 trillion behemoth; with SpaceX, it’s selling Starlink to a billion-plus people, a colony on Mars (which also gets him another $1 trillion payday), and data centers … in spaaace! With a gazillion new data centers apparently needed to power our wonderful AI-generated future, and companies running into NIMBY resistance here on the ground, let’s just put them in orbit instead. It’s that easy!

This one might be the most ridiculous of all of Musk’s many promises, and that is saying something. All the problems with space-based internet are orders of magnitude worse with data centers. Space is full of highly energetic radiation and micrometeorites, meaning either expensive shielding or a very short life span for the equipment. It is very hard to disperse heat in a vacuum, meaning heavy heat sinks. Repairs or upgrades will be nearly impossible. You can generate solar power 24 hours a day with the right orbit, but with ground-based solar plus batteries now the cheapest form of full-time power in most of the world, and likely to get cheaper in future, this isn’t much of an advantage.

AI Is Supercharging the Surveillance State

Then there is the fact that the AI boom looks very much like a classic investment bubble—meaning Big Tech companies might be radically overestimating the need for more data centers.

It’s obvious what is happening here, because we’ve seen it dozens of times before. Musk has to say something wild to build hype in the market, and at least for the moment, it seemingly doesn’t matter how many of his past promises have not come true.

People party to this deal will admit the economics of it are nuts. “From a strict corporate finance perspective, the valuation makes no sense. But Elon is great at getting people to dream,” another adviser to the deal told the FT. This combination of greed and hype likely explains the rush to get SpaceX into the important indices. Musk has produced several hot stocks in the past, and everyone wants in on the latest one.

The difference with stock manias of the past is that this one is significantly driven by the mechanics of the market itself. The point of index funds is that because almost everybody is provably bad at picking stocks, the smart investor simply buys a representative sample of the market. Except it turns out that doing that cedes a great deal of power to the people who make the rules about what is allowed into the index. If you’re a cynical CEO with a notoriously credulous cult of retail investors, you might trust that enough of them will buy your latest pig in a poke to make the price pop, and as a result the immense passive investment players—BlackRock oversees assets worth $12.5 trillion—will have to buy as well, pushing the price up.

Though no one can say when, however, nobody can escape the accounting gods forever. “These giant deals often happen at the very top of the market,” said Mason. Perhaps someday, financial investment will once again consider questions like “Does the business make any money?” A man can dream.

Read more

Ryan Cooper is a senior editor at The American Prospect, and author of How Are You Going to Pay for That?: Smart Answers to the Dumbest Question in Politics. He was previously a national correspondent for The Week. His work has also appeared in The Nation, The New Republic, and Current Affairs.