The Mega-IPO Wave Was Supposed to Drain the Market. Instead It Is Pacing Itself.

Two weeks ago, the loudest fear in the market was supply.

SpaceX had just priced the largest IPO in history, raising roughly $75 billion at $135 a share and valuing the company near $1.77 trillion. Behind it sat OpenAI and Anthropic, both widely expected to follow. The bearish version of the story spread quickly: a wave of mega-listings could pull a trillion dollars of value out of the broader market as investors sold existing holdings to fund new shares.

This week, the wave did something this thesis did not really account for: it started slowing itself down.

OpenAI is reportedly weighing a delay of its IPO to 2027. According to reports that moved markets on Friday, advisers presented the company with a choice: list sooner at a valuation below $1 trillion, or wait for conditions that support the number it wants. Sam Altman reportedly has little interest in taking the discount. SoftBank, one of OpenAI’s largest outside backers, fell sharply on the news. The caution, according to the same reporting, had a very visible reference point sitting on the tape: SpaceX.

That is the more interesting signal. The issue is no longer whether demand exists for these listings. The issue is whether that demand will validate the private-market number.

What SpaceX Actually Showed

SpaceX launched priced at $135. It opened at $150, closed its first day near $161, and then ran to an intraday high of $225.64 on June 16. After that it faded. By Friday it was back near $153, trading inside its opening-day range and roughly 30% below the peak.

Nothing fundamental about the company changed over those two weeks.

What changed was the market’s willingness to pay a scarcity premium on a relatively thin float. The early spike reflected a familiar mix of constrained supply, index expectations, and retail enthusiasm. The fade reflected something simpler: the market working out what it was actually willing to pay once the initial scarcity premium started to ease.

That matters because SpaceX was supposed to be the strongest possible setup for the mega-IPO case. If the most anticipated listing in years cannot hold its peak for even two weeks, the next issuer does not just have to ask whether buyers will show up. It has to ask whether buyers will pay the private-market number the seller wants.

For a company like OpenAI, coming from an $852 billion last-round valuation and reportedly aiming for $1 trillion, that is not a side question. It is the main one.

Where the Crash Thesis Starts to Break

The original “trillion-dollar withdrawal” argument was not unserious.

It leaned on real market mechanics, especially the inelastic-markets work arguing that capital flows can have an outsized effect on aggregate equity values. The basic point is intuitive: if a large enough amount of money is pulled out of existing equities to fund new supply, the overall market can feel that stress disproportionately.

The problem is that the argument is often applied asymmetrically. The same mechanism that amplifies outflows on the way down is also part of what inflated valuations on the way up. It cannot only count when it is convenient for the bearish case. More importantly, the strongest version of the model depends on rigid, mandate-bound capital being forced to rebalance in ways that are relatively insensitive to price.

That is not what an IPO does immediately.

Before a new name enters a benchmark, much of the capital moving into the deal is discretionary rotation rather than the kind of forced, mechanical flow the theory depends on. The more dramatic crash arithmetic tends to treat the entire raise as though it were a clean subtraction from the rest of the equity market. That is the assumption doing most of the work, and it is also the weakest one.

OpenAI’s reported delay is the market resolving that tension in a more ordinary way. If the supply cannot clear at the valuation the seller wants, the supply waits. That is not a system mechanically draining itself toward collapse, but a system refusing to validate a price it does not yet want to pay.

The Signal Underneath the IPO Story

There is a more important signal underneath all of this, and it has less to do with IPO mechanics than with how the broader AI buildout is being financed.

The capital intensity behind these valuations is unlike anything the market sees very often. Hundreds of billions in infrastructure commitments are sitting underneath these names. Some of that ecosystem is also highly interlocked, with major suppliers investing in customers who then become major buyers of their products.

That kind of structure deserves attention.

Markets have seen versions of this before. The late-1990s telecom boom ran in part on vendor financing, where large suppliers helped fund the same customers whose spending later showed up as demand for the suppliers’ own products. The businesses were real. The technology was real. That did not stop the financing structure from obscuring how much demand was actually durable.

That is not the same as saying today’s AI leaders are fake businesses or obvious bubbles. They are not. The revenue growth is real, and rare. The more limited point is that when valuations rely this heavily on a small set of interlocking commitments, the public market is justified in wanting disclosed financials before paying up to the private-market mark.

In that sense, a delay can be read less as weakness than as clarification. Choosing to remain private rather than open the books at a discount says something important about where the negotiation now sits between private pricing and public validation.

What This Adds Up To

The bearish framing had part of the risk right, but it pointed at the wrong mechanism.

The danger was never that these listings would mechanically drain a closed system. The system is not closed, and large global pools of capital can still come in from the outside. The more immediate friction is simpler than that.

Private markets have already run some of these companies from near-zero to near a trillion-dollar mark. Public markets are now being asked to validate that move in a single print.

That is a very different kind of challenge from simply absorbing new supply. This week, one of the biggest names in the queue looked at that challenge and chose to wait. That says more about where the market is than any dramatic crash scenario does.

The mega-IPO wave is not draining the market. It is discovering its own price, and finding, for now, that the number private sellers want is not the number public buyers are willing to pay.

High risk. DYOR. Not financial advice. See tessera.pe/terms