Since OpenAI’s ChatGPT was unleashed over three years ago, most investors have focused on pickaxe plays to invest in the boom: chipmakers like Nvidia, fabs like TSMC, and companies that build equipment to facilitate the rush, like ASML. The list goes on and on. Then, simultaneously, there has been excitement around the companies undertaking the data center buildout: Microsoft, Google parent Alphabet, and Oracle, among others.
This trend of investment has continued to permeate outward to every pocket of the AI economy. In recent months, it has reached key equipment producers; storage companies Western Digital and Seagate have been among the best performers in the S&P 500 in recent months. Then, there’s the memory producers: Micron Technology and South Korean giants SK Hynix and Samsung.
But later this year, for the first time, investors will get a bid/ask on the companies taking advantage of all of this new hardware: the AI labs themselves. Sure, we’ve long had investing access to Google parent Alphabet, the creator of Gemini. But the rising crop of big tech has the makings of something altogether different. It could also represent a hazard to America’s largely passive wealth machine, which contains trillions of dollars of our money.
The threat of new technology
Today, Wall Street offers investors limited ways to gain direct exposure to the actual models that should be facilitating this technological revolution. And ultimately, it might be for the best. While the models are technically impressive, it’s no secret that they are money pits.
Research from The Information shows that OpenAI had a net cash burn of $8.5 billion in 2025, despite achieving a milestone $20 billion in annual recurring revenue. Anthropic burnt $5.2 billion on its way to $9 billion ARR. Elon Musk‘s xAI, embracing the most aggressive AI strategy, burned over $10 billion last year to achieve less than half a billion in ARR; it was acquired by SpaceXin some imaginative dealmaking to shore up its sizable losses.
Here, the growth from various AI labs is astronomical, so too are the losses and valuations. Even more recent figures suggest even more growth has taken place, with Anthropic more than doubling its ARR run rate to $19 billion, per new data published.
They are exhausting every resource that they can to extend their cash runway until they can make AI models more affordable or efficient — or create new justifications for increasing token costs (e.g: more capable models).
They’re sparing no expense. OpenAI, Anthropic, and xAI raised over $277 billion from investors at steep valuations, largely based upon the promise that AI could be disruptive and lucrative. But from here, they’re left with few other options than to test the public markets.
And while it might work out, therein lies a worry.
The $2 trillion problem that could affect your retirement
In recent weeks, all three of the aforementioned big tech names have ramped up IPO talks. They are all now looking to hit the markets this year, with unreal valuations in tow.
OpenAI’s $730 billion post-money valuation, Anthropic’s $380 billion valuation, and a newly merged SpaceX-xAI valued at over $1.75 trillion will make them among the biggest private companies to debut on public markets. Despite that, none of them are profitable and none of them expect to be profitable for several more years.
Perhaps that is not worrisome in itself, but it is if these unprofitable companies land in your portfolio at their extremely advanced valuations. You might not necessarily know you’re buying these hot assets at all-time highs, either.
The majority of wealth in the U.S. is passively invested. That is to say, most of their money is on autopilot; mutual funds and exchange-traded funds (ETFs) following cookie cutter portfolios. Many of these portfolios are market-cap-weighted. In other words, they invest relative to the market at large. Over $2 trillion in new tech listings could cause quite a splash in these popular strategies.
AI leaders don’t want to wait to be added, either. Elon Musk’s SpaceX, which recently acquired the encumbered xAI has already asked to be fast-tracked into “major stock indexes” when it tees up its IPO later this year. Nasdaq has even proposed a new “fast entry” rule for the index in anticipation of the new listings.
Fast tracking might not even be necessary for many popular total market funds like the S&P Total Market Index or indexes issued by MSCI, which automatically add new listings.
The wisdom is that these passive index strategies are effective because they are simple and cheap; they follow the momentum and way of the market at a low fee. But a splash from these three listings, plus perhaps a handful more behind them, could turn your passive index fund into a concentration nightmare.
Suddenly, it won’t be the boring fund that demonstrated promise as a “get rich slow” scheme. Many will be heavily concentrated in technology, buying these still-unprofitable AI firms at all-time highs.
Beware of dump
Independent of how you feel about the promise of AI, companies testing the public market at these valuations and premiums disproportionately favors one group: the private investors.
As these companies line up to test the market, they will be able to exploit their hefty valuations to raise gobs of money from initial public offerings (IPOs). And then, long-time investors who benefited from the fantastical growth of the businesses will have an opportunity to exit as indexes buy the stock relative to their size.
In short: they get money now, your portfolio absorbs the new risk, and they can continue spending billions on their ambitious business ventures.
It might work out for them, in which case, it would work out for investors in some way. But mind you, if we are “still at the beginning” of AI as a generational technology, the valuations sure don’t seem to support that. At this stage, the private market’s price discovery seems to favor an alternative.
That’s why we might ask: what if it doesn’t work? As the public markets were largely left out of the price discovery, Wall Street might have second thoughts about the valuations of these firms. This is not to say that technologists can’t do a lot with today’s AI models, but it’s to say that the use cases currently don’t pay the bills.
Perhaps those second thoughts might not come immediately. There is ample appetite for this new wave of big tech companies, even despite a pullback seen among the Mag7 crowd in recent weeks. Investors might consider today’s valuations and premiums to be reasonable because of the promise of future profitability.
But at the core, this is an admission that many investors are buying into a story — the only thing stronger than fundamentals. That is, they will be able to sell, supported by tall tales they’ve spun about solving “the last great problem.” (We’ve talked about that, as well as why Wall Street might misunderstand the gravity of solving it in a previous TheStreet Daily article.)
Only, fundamentally, it remains a story until it actually happens. And like all stories, it’s open to interpretation, or worse, repricing. In the event that the story transforms or the profit window moves, American portfolios might be saddled with the losses.
The system is designed in this way. It has absorbed impacts in the past, but this time is verifiably different. And as we learn more about the impacts on major index funds favored by 401(k) or IRA portfolios, it might be wise to adjust positions based on where you land on a scale from “AI will transform everything” to “AI is not for real and this is all crazy.”
Are you worried or excited aboutOpenAI, Anthropic, and SpaceX’s public debut? If you’d like to share, email [email protected] to tell us how you’re thinking about the IPO boom.