General News

AI’s Forced Slowdown, Not a Demand Collapse

By Luke Lango2 min readInvesting.com
AI’s Forced Slowdown, Not a Demand CollapseAI’s Forced Slowdown, Not a Demand Collapse

The AI trade is facing a real slowdown. But it is a slowdown caused by supply chains, Wall Street and Washington, not by a drop in demand. And that is the super-critical distinction between the “AI Boom dying” and the “AI Boom transitioning.”

If demand were the problem, the spending wouldn’t just slow. It would outright fall or even stop, and we’d have a dot-com bust all over again. But if the demand is still there, and the slowdown is forced by issues unrelated to demand, then spending just slows. It doesn’t fall or stop. It just gets stretched out over a longer period of time.

We have the latter right now. We are looking at a non-demand-related slowdown in AI capital spending in 2027 and beyond, which means the spending gets stretched out but does not fall off a cliff. This is a transition period, not an ending.

Now, of course, that means AI stocks could fall further, because all of their revenue and EPS estimates over the next few years were modeled on the probably-now-false assumption that AI capex will rise in 2027, and 2028, and 2029, and so on. The new consensus assumption will shift to AI capex stabilizing in 2028 and beyond, which means lower capex estimates and lower EPS estimates for 2028 and beyond all throughout the AI supply chain. This is the transition period.

But on the other side of this transition period is a very bullish landscape: More durable spending. Fewer supply shortages. A capex build that becomes almost entirely self-funded with hyperscaler cash, because they won’t need to raise more debt if annual spending stalls around here (which is right at cash flow levels). Maybe we even get some public support for data centers if we can start building them more responsibly. All the risks that have stalled the AI trade since April essentially go away on the other side of this transition.

Which is why, you could argue, we are actually more bullish than ever on the three- to five-year outlook for the AI trade. We’re just also cautious on the three- to six-month outlook.