The next wave of AI mega-IPOs may arrive on Wall Street before they arrive in the index fund portfolios most investors actually own.
SpaceX can raise a historic amount of money, trade at a valuation usually reserved for the biggest public companies in the world, and still be kept outside the S&P 500. That is the point S&P Dow Jones Indices made on June 4 when it decided not to relax its entry rules for very large newly public companies.
The decision matters because SpaceX, OpenAI and Anthropic are not ordinary IPO candidates. They are the companies around which much of the current AI market narrative is being built. If they list at the valuations being discussed, they will immediately look too important to ignore. But the S&P 500 is not just a list of the largest companies in America. It is a benchmark with gates, and one of those gates is profitability.
Under the rules S&P left in place, an IPO must trade on an eligible exchange for at least 12 months before it can be considered for inclusion. A company must also show positive GAAP net income in the most recent quarter and across the sum of the most recent four quarters. The index provider also kept its investable weight factor requirement, which looks at the amount of stock actually available to public investors.
That sounds procedural. It is not. It changes the economics of going public for the biggest private technology companies on the planet.
SpaceX is expected to go public this month in what could become the largest IPO ever. Reuters reported that the company is seeking to raise about $75 billion at a valuation near $1.75 trillion, a level that would place it among the most valuable listed companies in the United States on day one. The same report said SpaceX generated $18.67 billion in revenue in 2025, up 33%, but posted a $4.94 billion net loss.
That last number is the problem. A company can be strategically important, politically visible and central to multiple markets, from launch services to Starlink to AI infrastructure, while still failing the financial viability test used by the S&P 500. Evercore ISI analysts, cited by Bloomberg, do not expect SpaceX to reach annual net profitability until 2027. If that proves right, the company may not become a serious S&P 500 candidate until 2028.
OpenAI may face a longer road. The company has enormous revenue momentum and the kind of consumer visibility most public companies would envy, but it is also spending heavily on compute, talent, model training and infrastructure. Anthropic is in a similar position. It has moved closer to the public markets after filing confidential IPO paperwork, but a single profitable quarter would not be enough if the broader 12-month result remains negative or future spending wipes out the gain.
This is where the IPO calculus changes. For years, founders and bankers could present index inclusion as a second-stage catalyst. First came the listing. Then came the passive buying. When a company entered the S&P 500, index funds had to buy it, whether the valuation looked comfortable or not. That demand helped support prices and gave early public investors a visible path to liquidity.
Now the largest AI listings may have to live without that support during the period when price discovery is most fragile.
Retail investors get access, but not through the usual door
The split between S&P and other index operators makes the situation more complicated. Nasdaq has moved to allow faster entry for large new listings into the Nasdaq 100, while FTSE Russell has also created faster routes for certain large IPOs. S&P did make changes for broader indexes such as the S&P Total Market Index and Dow Jones U.S. Total Stock Market Index, effective before the market open on June 8, but it held the line on the S&P 500, S&P MidCap 400 and S&P SmallCap 600.
That means some passive products may buy these companies quickly, while the biggest and most closely watched benchmark waits. For ordinary investors, the difference is not academic. Millions of people own S&P 500 exposure through retirement accounts, index mutual funds and ETFs. If SpaceX, OpenAI and Anthropic stay outside the benchmark, those investors will not automatically own them through their core S&P 500 funds.
Bloomberg Intelligence has estimated that fast S&P 500 inclusion could have triggered about $14 billion of passive buying for SpaceX, more than $8 billion for OpenAI and about $4.6 billion for Anthropic. Those are not small flows. They can affect trading, valuation and the confidence of investors who buy IPOs expecting index funds to arrive behind them.
The result is a cleaner test for the AI trade. Without immediate S&P 500 inclusion, these companies will need to find buyers who want the stock on its own terms. That puts more weight on revenue quality, losses, governance, public float and the credibility of management forecasts. It also reduces the chance that a trillion-dollar valuation alone can force the market to absorb a company quickly.
Founders may still pressure index providers to modernize their rules, especially if the next generation of strategic companies stays private longer and enters public markets at unprecedented scale. But S&P has just made clear that size is not enough. For AI companies preparing to list, the message is simple: Wall Street may welcome the story, but the S&P 500 still wants proof that the business can earn money.
Also read: Africa's startups are learning to fund growth closer to home • Andrew Bailey says AI may soon run into a power ceiling • OpenAI is bringing Lockdown Mode to ChatGPT users as security risks grow
