Picture a fire hose attached to the stock market. Now picture three of them, all turned on at once.

Today, Brad Gerstner - founder of Altimeter Capital and an early investor in companies like Snowflake, Databricks, and SpaceX - sat down on CNBC's Halftime Report to talk through what he called a "flood" of mega-IPOs hitting the market this year. Not one, not two, but three of the most anticipated public listings in recent memory are set to arrive within months of each other: SpaceX in June, OpenAI likely in September, and Anthropic targeting October.

The combined capital those three offerings could absorb from investors? Somewhere north of $200 billion - a figure that would exceed the total value of every U.S. IPO completed between 2022 and the first quarter of 2026 put together. That is not a typo.

Gerstner's framing was direct: "Today is not the day that you shove all in on the market." What that means for the stocks people already own - and the ones they are about to be offered - is the story.

The Background

An IPO - initial public offering - is the moment a private company sells shares to the public for the first time. Before an IPO, ownership of the company is held by founders, employees, and investors who got in early, sometimes a decade before the rest of the world had heard of the product. An IPO is, in essence, the moment those early owners get to sell some of what they built to anyone with a brokerage account.

For most of 2022, 2023, and 2024, the IPO market was close to dormant. Rising interest rates - the cost of borrowing money, set by central banks - made investors skittish about paying high prices for fast-growing companies that were not yet profitable. The math that had made tech valuations soar during 2020 and 2021 stopped working when money was no longer practically free.

But conditions shifted. Rates began to ease, the AI boom drove a wave of genuine business results across the tech sector, and the pipeline of companies that had been waiting for the right window started moving. 2025 saw a partial thaw. Now in 2026, the dam appears to be breaking - and the first wave is a very large one.

The three companies about to go public are not obscure. SpaceX, Elon Musk's rocket and satellite internet company, has been the most valuable private company in the world for years. OpenAI built ChatGPT, the product that arguably launched the current AI moment, and has reportedly grown its annualized revenue from roughly $2 billion in 2023 to over $20 billion by end of 2025. Anthropic, the AI safety company behind the Claude family of models, raised $30 billion in February 2026 at a private valuation of $380 billion and reportedly hit an annualized revenue run rate of $19 billion in March. These are not small bets.

What Is Actually Happening

The SpaceX IPO is the opening act. The company filed its S-1 prospectus with the SEC on May 20, targeting a Nasdaq listing under the ticker SPCX with pricing expected on June 11 and shares available as early as June 12. The target is to raise approximately $75 billion at a valuation of roughly $1.75 trillion - which would make it the largest IPO in history by deal size. The previous record belongs to Saudi Aramco, which raised $29 billion in 2019.

What Gerstner discussed on air today makes the SpaceX story more complicated than rockets and Starlink. SpaceX merged with Elon Musk's AI company xAI in February 2026, and that S-1 filing disclosed something unexpected: Anthropic has agreed to pay xAI $1.25 billion per month through May 2029 for access to the Colossus 1 data center in Memphis, Tennessee - the entire 300-megawatt output of the facility. At $15 billion a year, this single contract is nearly equal to SpaceX's total annual revenue of around $18 billion. xAI had built out the data center to train its own models; Grok, its flagship AI assistant, saw usage drop significantly in recent months, freeing servers the company is now selling to a direct competitor.

Gerstner framed this as proof of a thesis he has been building publicly: that data infrastructure companies - not the AI model companies themselves - are the clearest financial beneficiaries of the current AI moment. "Data is the oil that runs the AI engine," he said. He used Snowflake as the most immediate example. Snowflake reported yesterday that revenue grew 33.5% year-on-year to $1.39 billion in the first quarter, beating analyst consensus expectations of 27% growth. The stock surged more than 30% in after-hours trading.

Following SpaceX, OpenAI is targeting a public listing as early as September 2026 at a valuation of roughly $852 billion to $1 trillion, working with Goldman Sachs and Morgan Stanley on a confidential S-1 filing. Anthropic is eyeing October 2026 at its $380 billion private valuation, having engaged law firm Wilson Sonsini for IPO preparations. According to analysts at Indmoney, the combined proceeds from these three offerings alone could approach $200 billion - more than the entire U.S. IPO volume from 2022 through Q1 2026.

The Money Trail

Here is what Gerstner was really explaining when he said "today is not the day you shove all in."

When a new stock arrives at this scale, the money to buy it has to come from somewhere. Investors - whether individuals, pension funds, or hedge funds - do not conjure fresh capital from nothing. They sell what they already own to buy what is newly available. "When Anthropic comes public in the fall or OpenAI comes public in the fall, I'm going to have to sell something else to buy this," Gerstner said directly, talking about his own fund's position. "We don't have unlimited capital."

Multiply that logic across every major fund manager in the country, and the arithmetic becomes interesting. If institutions need to raise tens of billions of dollars to participate in SpaceX, and then more tens of billions for OpenAI, and more again for Anthropic, they will be trimming existing holdings to do it. The stocks that are most likely to get trimmed are the ones that have already run the hardest - the so-called Magnificent 7 (Apple, Microsoft, Nvidia, Amazon, Alphabet, Meta, Tesla), which dominate most technology portfolios and are the easiest source of liquidity.

This does not mean those stocks will crash. The U.S. equity market is large: Gerstner himself noted that global capital markets are measured in hundreds of trillions relative to $75 billion for SpaceX alone. But it does mean the timing and sequencing of when investors receive cash back matters enormously.

There is a second layer to the money trail, specific to how stock indices work. A stock index - like the S&P 500 or Nasdaq 100 - is a list of companies weighted by their market value. Index funds, which manage trillions of dollars in retirement savings, automatically buy companies when they are added to the list. When SpaceX joins an index at a $1.75 trillion market cap, every passive fund that tracks that index is mechanically required to buy SPCX shares. To buy them, those funds sell a proportional slice of everything else. Gerstner flagged that SpaceX was likely to receive early inclusion in the indices, which he sees as a source of structural demand - but it is also a source of structural selling pressure on existing holdings.

The SpaceX listing also has an unusual twist on supply. Its dual-class share structure will give Elon Musk 85.1% of the voting power, and the company will retain a majority of its own stake. That means the float - the actual percentage of shares available to trade - will be small relative to the headline valuation. Small float plus index inclusion plus institutional demand is the setup Gerstner acknowledged was "a little too complex" to game, but which several market commentators have noted historically tends to push prices up until the lock-up period expires and more supply enters the market.

The xAI losses disclosed in the S-1 filing are also material context: xAI lost $2.4 billion in the first quarter of 2026 alone, up from $936 million a year earlier, having spent $12.7 billion on AI infrastructure in 2025 and already $7.7 billion in Q1 2026. The Anthropic compute deal is not incidental to the IPO story - it is the revenue line that makes the numbers hold together.

What People Are Doing About It

Active fund managers like Gerstner are trimming positions that have run far above their price targets and building cash reserves ahead of the listings. He described the logic in terms of portfolio sizes: moving from a fully invested posture toward something more moderate when the market has climbed fast, keeping "dry powder" - cash that has not been deployed - ready for moments when volatility creates opportunities.

Retail investors are watching the SpaceX listing closely. Comment sections on financial videos published today show a split between enthusiasts forecasting enormous returns and skeptics warning that early investors will use the IPO as an exit. One commenter on the CNBC video this morning wrote that Gerstner's framing about selling other positions to buy SpaceX "was for the rest of you, telling you what to do, so that he can sell SpaceX which he probably was into from years ago" - which is, notably, not inaccurate. Early investors in companies that go public at trillion-dollar valuations do tend to use the liquidity event.

Institutional investors are also running the index-inclusion math with particular attention to the float mechanics Gerstner declined to fully explain on air. The gap between SpaceX's headline market cap and the actual tradable float will determine how much passive fund buying occurs at launch and how much price pressure results.

Meanwhile, the data infrastructure companies that sit upstream of AI - Snowflake, Databricks, ClickHouse - are attracting capital from investors who want AI exposure without betting directly on which model company wins. Gerstner made this framing explicit: as more AI "tokens" - the units of computation that AI models consume to generate responses - are processed, the companies that store and move the underlying data benefit regardless of whether it is GPT, Claude, or Grok doing the processing. Snowflake's results today are being read as the first hard evidence that this thesis is working at scale.

The Bottom Line

The arrival of SpaceX, OpenAI, and Anthropic on public markets within roughly five months of each other is the largest simultaneous demand for fresh investor capital in the history of the U.S. stock market. The money to buy these offerings has to come from somewhere, and the somewhere is likely the stocks that have already had the strongest run. That does not make these bad investments. It does mean the timing of entry matters more than usual, and that the downstream effects on everything already in a portfolio - index funds included - will be real, if hard to predict precisely. The AI wave is not over. It is being priced.

Timeline

Summary

Who: SpaceX (Elon Musk), OpenAI, and Anthropic - the three largest private AI and tech companies in the United States, alongside institutional investors, retail investors, and index funds worldwide

What: A wave of three major IPOs set to arrive between June and October 2026, targeting a combined capital raise of over $200 billion; the largest simultaneous public offering in U.S. market history

When: SpaceX targets June 12, 2026; OpenAI targets September 2026; Anthropic targets October 2026

Where: Nasdaq (SpaceX), with OpenAI and Anthropic's exchange listings not yet confirmed; underlying assets in the U.S., including SpaceX's Colossus data center in Memphis, Tennessee

Why: All three companies have reached the scale at which private capital markets no longer provide sufficient liquidity for employees and early investors; AI-driven revenue growth has made the public market window viable after three years of limited IPO activity