When OpenAI’s confidential IPO filing hit the news on May 20 — timed, reportedly, to step on the imminent public unveiling of SpaceX’s own S-1 — the initial reaction was giddy. Three of the most consequential private companies on earth, all racing toward the public markets within months of each other. Combined valuations north of a trillion and a half dollars. The sort of moment that makes investment bankers weep with joy and financial journalists reach for superlatives.

The Economist, in a piece published yesterday, framed the question neatly: can the stockmarket actually swallow these companies? SpaceX at $800 billion. OpenAI reportedly targeting something in the same postal code. Anthropic not far behind. These aren’t IPOs in the traditional sense; they’re entire asset classes showing up at the front door asking for a ticker symbol.

But the question is wrong. The market can swallow anything if the price is right. The real question — the one buried in the footnotes of the S-1 filings and the byzantine lockup agreements that accompany them — is whether the structure of these deals turns the ordinary investor into a patsy on a schedule.

The Figma Dress Rehearsal

Figma went public in late 2025 at $33 a share. On day one, it closed at $143, a market cap near $95 billion. Eight months later, it was at $22. That’s not a correction. That’s a design feature.

The mechanism isn’t complicated, but it’s deliberately obscured. These companies aren’t just going public — they’re going public with a carefully engineered float. Small. Scarce. The shares available to trade on day one represent a fraction of the total outstanding. Insiders, early investors, and employees are locked up under agreements that stagger their ability to sell over months or years. Some of those releases are triggered not by dates but by stock-price performance thresholds. The float expands only after the price has been run up high enough to benefit the people who wrote the terms.

What this means in practice: the first wave of public investors — retail, ETFs, momentum funds — buys into artificial scarcity. The price spikes. The narrative writes itself. “Market validates $800B valuation.” Then the lockups expire in tranches, the float triples, and suddenly there’s a lot more supply than anyone anticipated. The price slides. The insiders, who structured their own exit windows to coincide with the peak scarcity, are out before the slide begins.

“The lockup architecture is the trade,” a trader at a mid-sized hedge fund told me Monday, during a lull while the desk waited on SpaceX pricing details. “You don’t trade the company. You trade the unlock schedule. And unless you’re in the room when the schedule is written, you’re on the wrong side of it.”

The Bank Overlap Tells You Everything

Axios reported that OpenAI and SpaceX are working with many of the same banks: Goldman Sachs, Morgan Stanley, JPMorgan Chase. That’s not a conflict of interest — it’s a concentration of it. The same syndicate desks that will price and allocate these offerings are the ones who have spent the last five years advising these companies on their private fundraising rounds, collecting fees at every tier of the capital structure.

Ask yourself what incentive those banks have to price an IPO aggressively. The answer is none. Their relationship is with the issuer — the company, the insiders, the early venture backers who will take their future banking business elsewhere if the opening trade isn’t a pop. The buy-side institutions that get allocations at the IPO price are part of the same ecosystem. They’ll flip for a quick gain or hold through the first lockup release, depending on which version of the game they’re playing.

The retail investor who buys on day three at $147 because CNBC ran a segment about the “historic IPO wave” is not in the same ecosystem. They’re the exit.

The Thing Nobody Wants to Say About the Float

There’s a version of this story where these companies go public with 25% or 30% of shares outstanding, the way big IPOs used to work. The price is set by actual supply and demand, not by manufactured scarcity. The opening trade is boring. No one gets a 300% pop. No one gets wiped out six months later when the real supply arrives.

That version is not on offer. SpaceX’s S-1, according to those who have reviewed it, contemplates a float far smaller than the historical norm for a company of its size. OpenAI’s confidential filing almost certainly does the same. The banks have learned that a thin float generates headlines, and headlines generate retail flow, and retail flow generates the liquidity that lets insiders exit gracefully over the subsequent quarters.

None of this is illegal. It’s not even particularly novel — the lockup-as-trade has been around for years, refined through the SPAC boom, the 2021 IPO wave, and the Figma implosion. What’s new is the scale. When a company with an $800 billion valuation goes public with a 5% float, the dollar amount of retail capital that can be drawn into the initial spike is enormous. The amount that can be lost when the lockups release is larger still.

Don’t Confuse Participation With Access

The Economist’s framing — can the market swallow these companies? — presumes the market is a unified thing, a digestive system that absorbs new securities and finds their proper price. But the market isn’t one thing. It’s a set of nested games with different rules for different players. The insiders play the lockup game. The banks play the allocation game. The ETFs play the index-inclusion game. And the retail investor, finally offered a piece of the companies they’ve read about for years, plays the game where they walk into a casino that has already calculated their expected loss.

The IPO wave of 2026 will almost certainly be historic. The companies are real, the technology is real, the revenue trajectories are real. But the question worth asking isn’t whether the stock market is big enough. It’s whether the people selling these shares have any interest in the people buying them getting a fair deal.

They don’t. The S-1 makes that clear, if you know where to look.

Sources