There is a reliable tell in markets. It is not a chart pattern or a Fed statement. It is the moment when someone completely outside the financial world — a racetrack acquaintance, a building guard, a colleague who grows things for a living — starts asking you, with total seriousness, how much of a specific pre-IPO company they should buy. That moment has arrived for SpaceX, and it arrived loudly, all at once, the way these things always do right before someone gets badly hurt.

SpaceX is, by any honest accounting, a genuinely extraordinary business. Its Falcon 9 booster recovery program has restructured the economics of orbital launch. Starlink, its satellite internet division, is operationally profitable and serves millions of subscribers across dozens of countries, including active military and emergency-response contracts. The company has achieved things that were, within living memory, the exclusive province of nation-states. None of that is in dispute. The question — the only question that matters to an investor — is what price those achievements justify.

And here the honest answer is: nobody knows, and the people acting most certain are the ones most likely to be wrong. Private market valuations pegged SpaceX at roughly $350 billion in its most recent secondary transactions, a figure that would make it, at listing, one of the largest public companies in the United States by market capitalization. For context, that valuation implies a multiple on revenue that dwarfs aerospace peers and sits comfortably in the territory usually reserved for the fastest-growing software companies in history. SpaceX is not a software company. It builds rockets and satellites. The margins, the capital requirements, and the competitive dynamics are categorically different.

The comparison that keeps surfacing among people paid to remember history is Amazon in 1999. Not the Amazon of today — the Amazon that was unambiguously a real business, genuinely disrupting retail, run by a visionary founder, and also priced at a level that meant early public investors waited more than a decade to break even. The company was right. The price was wrong. Those are not mutually exclusive outcomes, and conflating them is exactly how retail investors get destroyed in otherwise legitimate businesses.

OpenAI and Anthropic sit in a related but distinct category. Both are pre-revenue-at-scale in the sense that their cost structures — dominated by compute and talent — have not yet demonstrated the kind of durable margin profile that public market investors in mature companies expect. OpenAI, in particular, faces a governance structure that remains genuinely unusual: a nonprofit controlling entity sits atop a capped-profit commercial subsidiary, a structure that is currently under active legal and regulatory scrutiny as the company pursues a restructuring that would give that commercial entity more conventional corporate form. Anthropic, substantially backed by strategic investors with their own competitive interests, presents a different set of alignment questions for a purely financial investor.

The proxy-share problem is worth understanding clearly, because it is where retail enthusiasm most directly meets institutional reality. A number of funds and special-purpose vehicles were created specifically to offer indirect exposure to SpaceX's private equity. Some of those structures have redemption restrictions, gating provisions, or lock-up terms that mean investors who want out before an IPO may find the exit door bolted. The IPO itself, if it comes, will not automatically solve that problem — depending on the structure, proxy holders may not receive direct shares and may face continued lock-up periods post-listing. The enthusiasm of buyers on the way in was not matched by attention to the fine print.

There is also the index effect, which is being discussed less than it should be. A SpaceX listing at anything near current private valuations would trigger mandatory purchases by passive funds tracking major indices — a mechanical buying wave that has nothing to do with fundamental analysis. This is not a secret, and sophisticated participants know it. The question of who is positioned ahead of that wave, and how that positioning influenced the drumbeat of IPO speculation, is worth sitting with.

Of the three — SpaceX, OpenAI, Anthropic — the one with the most legible path to public-market valuation discipline is the one with the most concrete, auditable revenue: SpaceX, specifically its Starlink segment. A clean spinout of Starlink as a separate listed entity, which has been floated internally at various points, would give investors something they could actually model. A combined SpaceX listing bundles the profitable satellite business with the extraordinarily capital-intensive Mars and heavy-lift ambitions of a founder who has made clear those ambitions are not negotiable. Investors would be buying both, whether they want both or not.

The guard at the exchange, the racetrack executive, the gardening colleague — they are not stupid. They are responding rationally to a signal environment saturated with SpaceX hype. The hype is not fabricated; the underlying company is real. But markets have a long and well-documented history of being right about the company and catastrophically wrong about the price. When everyone already knows the name, the edge is gone. The question is not whether SpaceX is remarkable. The question is whether you are the last one buying someone else's remarkable position.