Steve Eisman Says Cutting AI Capex Would Sink the Market. Spending Too Much Could Be Just as Dangerous for Hyperscaler Stocks.

The artificial intelligence (AI) boom has become one of the most powerful forces driving the stock market, fueling massive gains in semiconductor companies, data-center suppliers, utilities, and the tech giants leading the spending race. At the center of that boom are hyperscalers like Alphabet (GOOGL), Meta Platforms (META), Microsoft (MSFT), and Amazon (AMZN). These companies are pouring hundreds of billions of dollars into chips, servers, networking equipment, and data centers to support rapidly growing AI demand.

Yet the scale of that spending has also become one of the market’s biggest concerns. Investors are increasingly questioning whether these firms can generate sufficient returns on their enormous capital expenditures, particularly as AI investments consume a growing share of operating cash flow and place pressure on free cash flow. Each new increase in spending guidance now risks reinforcing fears that Big Tech may be investing too aggressively without enough near-term revenue to justify the cost.

Investor Steve Eisman of The Big Short fame, however, believes the market faces an equally serious danger if hyperscalers begin cutting their AI budgets. Eisman recently argued that the entire market has become dependent on continued capex from the largest tech companies and warned that stocks could fall sharply if that spending suddenly slows.

That raises an important question for investors. If cutting AI capital expenditures could send the market sharply lower, could continuing to raise spending indefinitely become just as dangerous? Let’s take a closer look.

Big Tech’s AI Capex Supercycle Faces a New Investor Reality

The narrative surrounding AI and so-called hyperscalers — Alphabet, Meta, Microsoft, and Amazon — has shifted dramatically in recent years. The AI boom ignited with ChatGPT’s launch in late 2022, unleashing a historic capex supercycle and prompting Big Tech to spend heavily on data centers packed with advanced chips, servers, networking equipment, and cooling systems. AI hyperscalers monetize these investments by selling cloud computing capacity, charging for access to AI models and premium software tools, and using AI to strengthen advertising, search, and other core businesses.

Initially, the story was pretty simple. Every time one of these firms boosted its capex guidance, the stock would surge. Investors viewed this as evidence of healthy demand and an expanding revenue backlog. Another important point is that, in the early years of the AI boom, major tech companies comfortably funded their AI investments with internal operating cash flow, effectively eliminating concerns about the scale of the spending. Moreover, Microsoft, Alphabet, and Meta generated enough free cash flow in their latest fiscal years to cover dividends and share buybacks.

However, investors' mindsets toward AI capex budgets have changed dramatically over the past year. A steadily rising ratio of capital expenditures to operating cash flow has prompted hyperscalers to tap debt markets to finance their AI investments, with some even turning to equity markets. That marked a turning point, prompting investors to scrutinize AI investments more closely.

In particular, investors have sought more convincing evidence that hyperscalers’ massive AI spending is driving new growth rather than weighing on profits. Moreover, instead of piling into every hyperscaler, investors have begun distinguishing between winners and losers in the AI race and rewarding only those delivering relatively stronger returns on their AI investments. This was most clearly demonstrated when the four Big Tech hyperscalers reported their first-quarter results. Despite all four reporting strong financials, only Alphabet and Amazon gained after earnings, as investors viewed their performances as more impressive.

Now, in the midst of the Q2 earnings season, the narrative appears to be shifting even further. Alphabet, the first major tech company to report Q2 results, was hammered last week after raising its full-year capex guidance and posting negative free cash flow for the quarter. That’s despite reporting 82% year-over-year (YOY) revenue growth in its cloud division. If Alphabet is any indication, investors are no longer going to celebrate strong results while AI spending continues to rise into the stratosphere.

The latest shift in the narrative creates a challenging backdrop as earnings come in from Microsoft, Meta Platforms, and Amazon. More importantly, it raises a key question. In this new environment, where all eyes are on AI capex, is less spending better?

AI Capex Guidance Cuts vs. No Change: What It Could Mean for the Market

That question can be interpreted in two ways: hyperscalers either maintain their capex forecasts or lower them. Let’s take a closer look at both of these scenarios.

I think Steve Eisman summed up the latter case perfectly. Eisman said in a recent CNBC interview that the stock market would “go straight down” if any major hyperscaler cuts its AI capex guidance. He noted that the immediate impact on Nvidia (NVDA) would spark a sharp market selloff, and argued that the market has become a one-trade environment, with its direction almost entirely dependent on the success or failure of AI. Even sectors that appear unrelated to the trade — such as major U.S. banks that recently reported strong results — are closely tied to AI because, as Eisman put it, “they’re doing a lot of AI financing.” Eisman also pointed to the growing concentration of the tech and AI trade not only in equities but in the bond market. “Of the 40% in bonds, most of the new issuance of bonds is AI-related," he said. “So even people who think they’re diversified because they’re investing in bonds are not really that diversified. What scares me is that it’s all one trade.”

At this point, I’d like to add my two cents. I fully agree that such a move would sink NVDA stock, but I believe other AI-related stocks — particularly memory chipmakers — would be hit even harder. The key point is that earnings projections — and valuations by extension — for chip and AI infrastructure companies depend heavily on hyperscalers’ capital expenditures. While there have been no signs of capex cuts or a slowdown, concerns about the sustainability of the AI spending boom have periodically resurfaced, triggering sharp selloffs in chip stocks. So, if mere concerns have been enough to spark wild swings, imagine how sharply these stocks could fall if they have to be revalued following an actual cut in spending forecasts. With that, just as chipmakers propelled major stock indices higher on AI optimism, they could drag them sharply lower on AI pessimism. I don’t think lower capex would help hyperscalers, either, as they would most likely be swept up in a broader market selloff.

So, what could happen in the second scenario in which hyperscalers maintain their capex forecasts? I actually view it as a “Goldilocks” scenario, as it would allow investors to focus on financial results without fueling additional concerns about AI spending. In other words, strong earnings combined with unchanged capex guidance would drive post-earnings pops in hyperscalers' shares. For example, I believe GOOGL stock would have surged after earnings last week if Alphabet had maintained rather than raised its full-year capex guidance, allowing investors to look past the free cash flow issue. In fact, we already saw such a precedent three months ago, when Amazon reported negative free cash flow for Q1 but maintained its full-year capex guidance, allowing investors to focus on its strong results.

With that, it will be interesting to see capex updates from Microsoft, Meta, and Amazon — and more importantly how the market reacts to them — as they report earnings this week.