Elon Musk extracts a trillion-dollar valuation from public investors and keeps absolute control.
It is true, and worth saying plainly, that SpaceX has done things no private company had done before—orbital-class reusability, a working satellite-internet constellation that now carries traffic for the U.S. military, a heavy-lift rocket that is the envy of every state space program on Earth. The engineering is real. The rockets fly. The trouble, and it is the trouble that matters for anyone being asked to buy a piece of the company at a $1.75 trillion valuation, is that the price the bankers have set does not rest on the rockets. It rests on a story.
The numbers, drawn from the prospectus that the BBC reviewed and that MSI reported on when the share price was set at $135, are these. SpaceX lost nearly $5 billion last year. Its profitable Starlink business accounts for less than 20 percent of the $1.75 trillion target valuation. The remaining 80 percent—roughly $1.4 trillion in market capitalization—is pinned to a total addressable market the prospectus pegs at $28.5 trillion, of which $26.5 trillion is attributed to artificial intelligence services. The company’s own filing describes that market as including data centers in space, powered by solar energy and cooled by the vacuum of space, and “human-crewed bases on the Moon and Mars.” The prospectus itself acknowledges, in the careful language of securities lawyers, that capturing this market “requires building, commercialising and operating products and services … at a scale that has not been previously achieved.”
An engineer reading that sentence would recognise it for what it is: a disclaimer that the entire business case is speculative. The same engineer would note that SpaceX has never built a data center in space, has never demonstrated the capacity to cool one with vacuum, and has never landed a human being on the Moon, let alone Mars. The company launches rockets—superbly, at lower cost than any competitor—and operates satellites. The AI story is not an engineering plan; it is a narrative, and the narrative is what the bankers are selling.
The mechanism of the offering is the extraction architecture. SpaceX is initially placing only five percent of its equity into the public market—roughly $75 billion in shares—just enough to test demand and establish a mark-to-market benchmark for the ninety-five percent Musk will continue to hold behind closed doors. The equity split is secondary to the voting structure. Under the dual-class share architecture filed with regulators, Musk retains eighty-five percent of the shareholder voting rights despite owning forty-two percent of the stock. As the financial journalist Robert Armstrong observed when the terms were disclosed, this is not public ownership in any sense that confers agency. “Do you really own something you can’t control?” he asked. The answer, for the retail investor, is no. The offering is not designed to distribute ownership; it is designed to externalize the risk of capital expenditure onto retail index funds while the founder retains the throttle, the steering wheel, and the unilateral right to direct the company’s cash flow toward whatever subsidiary venture he chooses next.
One large institutional investor, quoted by the BBC, put the matter plainly: “The cult of Elon Musk requires disciples to pay a premium for the questionable privilege of having no real say in how the company they own is run. But people seem happy to do that.” Sinead O’Sullivan, an economist with experience at NASA, characterized the valuation as functionally indistinguishable from the brand premium attached to the chief executive: “When we look at the massive share price that they are trying to get here, you’re buying a share of the Elon Musk brand more than any kind of space industry.”
This is not a new pattern. Cory Doctorow, borrowing from John Kenneth Galbraith, calls the interval between the commission of a fraud and its discovery the “bezzle”—the period when the embezzler has his gains and the victim does not yet feel the loss. In formal protocol verification, you do not accept “the system will scale” as a sufficient specification. You require a proof obligation—a mathematical demonstration that the system satisfies its constraints under adversarial conditions. The SpaceX prospectus offers no proof obligation for its AI revenue. It’s a promise without a verification condition—the hallmark of a bezzle. The $1.75 trillion valuation is a bezzle of historic proportions. It converts a rocket company with a profitable satellite sideline into an AI-and-Mars conglomerate on the strength of a prospectus that, read carefully, admits the whole thing is a bet. The bet is that the public will buy the story, the stock will rise, and Musk—whose personal fortune already exceeds $700 billion, who has spent nearly $300 million electing the president whose administration awards his company multibillion-dollar contracts—will become the world’s first trillionaire, with index funds perhaps absorbing some of the supply even as a parallel flood of shares from AI competitors like Anthropic and OpenAI threatens a dot-com-era supply glut. If the bet goes wrong, the public shareholders will hold the bag. Musk will retain control of the company either way.
There is a remedy available. The Securities and Exchange Commission could require that companies with dual-class share structures that confer outsized voting control price those shares at a discount, or bar them from inclusion in major indices until governance is reformed. Congress could close the dual-class loophole. Neither is likely. The SEC, under the current administration, has shown no appetite for restraining Musk’s enterprises, and Congress is not about to pick a fight with a man who funded the winning campaign. The work of protecting public investors from the largest bezzle in market history is, as always, left to the investors themselves.
The prospectus is filed. The allocation windows for the initial five-percent offering are set for the market open on Friday. The regulator’s question is why dual-class super-voting structures of this magnitude are permissible on an exchange that is supposed to price risk rather than shield the insider from public capital. The investor’s question is simpler. Read the voting-rights schedule. The fine print is the whole deal, and you’re already paying for it.