AI Capital Spending Faces a Slower Growth Test as Investors Reassess Risk

The market’s latest AI stock selloff sharpened a question that matters more than any single day’s price move: can companies keep funding an enormous infrastructure buildout while producing enough cash to justify it? The investors interviewed at Schwab IMPACT 2026 in Boston saw reasons to expect continued spending, but also signs that its pace and financing could become less comfortable.

On Thursday, Nov. 5, 2026, the Nasdaq Composite fell 5.4% in afternoon trading, entering correction territory as chip and data-center stocks dropped. A rebound on Friday pared some of that decline. The episode exposed how quickly sentiment can shift when investors weigh the cost of AI infrastructure against the growth it may eventually support.

Cash generation is a key difference

Omar Aguilar, CEO and CIO of Schwab Asset Management, cautioned against treating today’s biggest AI spenders as direct stand-ins for companies in earlier market bubbles. He said that comparison can reflect “anchoring,” or the habit of interpreting a new period through the pattern of a past one.

His distinction is the source of the money. Aguilar said 60% to 70% of capital spending comes from mega-cap companies with “very strong balance sheets.” He described these firms as strong generators of free cash flow, unlike businesses that have no earnings and rely heavily on borrowing. That financial capacity gives the current spending cycle a different footing, though it does not remove the risk that spending could outrun returns.

Aguilar sees a change in financing as a point to watch. More companies are turning to debt markets to fund expansion, he said, and “that transition starts to get a little bit worrisome.” Debt can extend a buildout, but it also raises the importance of future cash generation. If returns take longer to arrive, borrowers may have less room to keep expanding at the same pace.

Spending can rise while its growth slows

Tim Nyland, co-CIO of Synergy Asset Management, said his team believes the growth rate of AI capital spending peaked in the third quarter of 2026. That does not mean total investment is set to fall. As Nyland put it, “It’s not that we’re going to see a decline in CapEx spend. Rather, what we’re going to see is the rate of increase or the growth rate itself is going to start declining.”

The distinction matters for investors assessing companies tied to chips, data centers and other infrastructure. A slower rate of growth can still leave spending at a high level, while changing expectations for the companies supplying the buildout. The key question becomes whether demand and cash generation can support the spending already planned, even if each new period brings a smaller increase.

Nyland pointed to the size of the bill and cited “some instances of negative free cash flow,” including at Meta. He said similar pressure may be emerging at other hyperscalers. The outcome could also widen performance differences among the largest technology companies. In his words, “The Magnificent 7 might fragment a little bit. We may no longer have a Magnificent 7. It might be a Mag 3, might be a Mag 5. We don’t know.”

A halt would carry a different risk

Nyland drew a line between a gradual slowdown and a complete stop. “The only way we’re going to have a market crash or a wipeout is if CapEx spending comes to a complete halt,” he said, adding that such a halt would “drive us into a pretty deep recession.” He does not expect that outcome. His base case is “a gradual slowdown in the AI CapEx commitment and actual spend.”

Aguilar also expects economic growth to decelerate and said some technology companies are already cutting back. “They’re not stopping it, but they’re reducing their CapEx.” Together, the comments point to a cycle that may cool without ending. For firms building infrastructure, investors will be watching both the level of spending and whether operating cash can carry a growing share of the cost.

Portfolio concentration remains in focus

Tom Skrobe, head of RIA business and portfolio solutions at WisdomTree, said advisory firms were already reviewing exposure to the S&P 500, where AI-related stocks account for a substantial concentration. He described the interest as portfolio maintenance rather than a response to fear of a major downturn. “We’re seeing a lot of RIAs asking us to take a look at their portfolio and possibly diversify a little bit away from the S&P, not totally away from it,” he said.

Skrobe said dividend-weighted products can add diversification and argued that it makes sense to review exposure to a single stock sector or investment theme at different points in the cycle. Before the selloff, he said he did not see a crash or major correction as a large concern. The market’s abrupt decline has since put concentration and the durability of AI spending back in view.

The useful test from here is whether the largest spenders can maintain infrastructure investment while preserving cash generation, and whether returns become broad enough to support the companies financing the buildout. Continued spending would not, by itself, settle that question; its pace, funding mix and results will shape how sharply the market distinguishes among AI companies.