Business

Why AI hyperscalers are now the epicenter of a bear case for stocks

CNBC June 08, 2026 1 views
Why AI hyperscalers are now the epicenter of a bear case for stocks

Advertisement

If the facts change, then I have to change. I learned that from reading the legendary British economist John Maynard Keynes, and it has never steered me wrong. I was a huge bull on this market because of the terrific things happening in artificial intelligence. And because we had a new Federal Reserve chair in Kevin Warsh, who wanted to cut interest rates to support growth, given the frozen housing market and an underclass struggling with higher gas prices, a drop in health care coverage, and the gratuitous cut in food stamps. I made this judgment, as I always do, by going to the source: the top twenty public U.S. retailers, who know far more than the government does about spending habits. Those include the auto and the auto parts companies, as well as ancillary home care businesses. The idea that we don't need rate cuts is fanciful. But then we got the monthly employment report on Friday, which showed that job growth unexpectedly surged in May, with nonfarm payrolls jumping a seasonally adjusted 172,000, far above the Dow Jones consensus estimate of 80,000. Of course, that figure never distinguishes between those who are employed and doing well and those who are employed and trying to hold it together. A job means self-sufficiency in this country, and there's no counterargument to that line of thinking, despite my attempt to make one when I interviewed Kevin Hassett, President Donald Trump's chief economic adviser. But it wasn't worth the effort, even though the nature of the poor working class is obvious to anyone except those who only have wealthy friends and acquaintances. Knowing people of all levels of the economic spectrum is like knowing the cost of milk and gasoline. It helps you to do your job well. The stronger-than-expected report wiped out the chance of rate hikes, or even a single one, this year, and that had been a major prop to my bull case. I could have looked past the macro if not for the dramatic change in the data center buildout, where costs have risen sharply on everything from labor and construction materials to power and site development. We went from thinking there would be a payback in the near future to having no sense of when it would occur. As recently as a month ago, I felt that Amazon's confidence about a return on invested capital was so definitive that there was no reason to worry. Now I worry that Amazon may need to make an equity offering because the payoff from AI is more elusive than definite. I also thought that we would only have to deal with equity offerings from a few private companies: OpenAI, Anthropic, and SpaceX. But after Alphabet announced plans to raise $80 billion through stock sales to fund its AI efforts, it looks like the biggest tech companies, including Amazon, Microsoft , and Meta , may need to raise huge sums for AI by selling stock. There is no way this market can handle that much equity and stay at these levels. So we find ourselves going from the prospect of a rate cut in 2026 and a smooth glide path to hyperscaler profitability, to a possible rate increase this year and an endless series of gigantic fund raises. Now, let's take a break for a moment and let me don my hedge fund hat. I don't like to do so for many reasons. One is if I were going to do it, I might as well be a hedge fund manager, and I did that and retired because, well, go read "Confessions of a Street Addict," and you will know why. Another is that I can't do what a hedge fund manager does. When I saw the Alphabet offering succeed, I should have realized that we would then face a deluge of deals without the money to do them. I have said over and over to you that when we have so many deals, we can't think straight, and it's time to hit the road. Hit the road in hedge fund lingo means selling everything you were long and going short those same stocks, while pivoting to stocks unrelated to the data center, à la Johnson & Johnson and Procter & Gamble . It also means that you blow out of stocks like AI stocks like Arm , Qnity , and even Corning , as well as Marvell and Nvidia . You would do that because you would realize that you are going to get multiple stories about how the data center is hated all over the country, too hard to build, too expensive, and a never-ending source of more equity. The hyperescalers, which were the source of the greatest stock story ever told, are now at the epicenter of the bear case. I can't do any of that now. I can't trade around that much, and I am always talking about my positions. That's why the Morning Meeting (at 10:20 a.m. ET) is so important: you hear what I would do, not what I can't do. The rules protect everyone. I am in a position to influence prices, and I don't take advantage of that personally. So what do we make of a case where we believe we could get an offering from Microsoft or Amazon at any moment? We have to first posit what happens when the SpaceX deal gets priced next Friday. There are many people who don't have any link to E-Trade or Robinhood (or any other firm with the stock) who will buy some shares at the market price. They will be the ones who determine the opening. The stock should be worth $4 trillion unless SpaceX founder and CEO Elon Musk says that the institutions can take profits. Right now, because I have no idea where the supply comes from, the opening should be like the spring of 1999. That kind of opening actually creates some liquidity that might be helpful for the next five deals. Or the company just sops up every available dollar and the edifice crumbles, which is a very realistic possibility. The novelty of the deal does not help its case, because nobody seems to know what will happen. I don't care about all these tranches; I just care about the first batch of shares coming to market. I can't expect it to be good. I just want it to be on a day when Trump doesn't create some manufactured turmoil. The real issue here is how much money is allocated to tech and how much to non-tech stocks. Growth investors used to the Rule of 40 — a software metric that posits a company's revenue growth rate and profit margin should equal at least 40% — are fleeing tech for health-care and consumer companies, among others, that still have strong organic growth. So you have to reposition yourself. We've tried. We have more to do. I like all of our health care, especially Cardinal, which had no flies to begin with. I am mindful that we started our Intel position too early with our initial buy last week . But if the market absorbs all of this supply, Intel will be the next Micron . Let's talk about that possibility. At some point, an Nvidia client will have to step up and say how lucrative this all is. I believe that will be Anthropic. When that deal is done, the worst may be over. If I were a hedge fund manager, I would know to rebuild the whole shooting match. But I am not. So it is too difficult, but I will do my best to let you know. So what will my posture be tomorrow? Simple: a chastened bull, not an aggressive bear. Chastened because I kept saying that when we get too many deals, we have to skedaddle. But a bull because the long-term impact of the Fourth Industrial Revolution isn't so long-term; it's happening now. It's just that Microsoft doesn't want to be marginalized. Amazon wants Amazon Web Services to remain the leading cloud platform. Meta might want to have a robust Meta Web Services. And Alphabet wants to preserve its frontrunner status. They all have to keep spending heavily or be left behind. But in the near term, there are a lot of mouths to feed and not enough to eat. The market could get overwhelmed. That's where we find ourselves. It's a suboptimal place to be. (See here for a full list of the stocks in Jim Cramer's Charitable Trust.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.

<small>Source: CNBC</small>

How did this make you feel?

Advertisement

Category
Business

Advertisement

🌙