A significant rotation is underway in equity markets, with capital moving from AI-heavy technology into healthcare, industrials, and traditional value sectors. The shift has accelerated over the past month as questions about AI revenue timelines mount, and it now has enough duration to be a trend rather than a squall.

The arithmetic driving it is not mysterious. Hyperscalers have committed hundreds of billions to AI infrastructure while revenue from AI applications has not matched the outlay — a gap the paper's business desk quantified this week when the combined free cash flow of the major cloud providers turned negative for the first time in two decades (as the paper reported). Markets can forgive spending if revenue follows on schedule. What they forgive less readily is a moving schedule, and the schedules have been moving.

Healthcare has been the primary beneficiary, and the composition of the bid matters more than its size. The paper's market desk found medical device makers — Medtronic, Boston Scientific, Edwards Lifesciences among them — outperforming the broader tape by 8-12% over the past quarter (per the market-desk analysis). Those are not lottery tickets; they are companies with predictable revenue, regulatory moats, and aging-demographics demand. When money rotates into them, it is pricing certainty, not speculation. Industrials tell the same story in different clothing: order books tied to grid buildout and reshoring pay whether or not chatbots get smarter.

The X interpretation is blunter than the flow data. On trading timelines, the rotation reads as a vote of no confidence in AI's near-term revenue promises — the market finally auditing the gap between capex announcements and invoices (and Friday's record close sharpened the contrast). That reading is half right. Flows measure where money went, not why, and attribution after the fact flatters narrative over noise. But the persistence of the move through a week that ended with the Dow at records suggests something structural: portfolio managers repositioning for a scenario in which AI returns arrive later than priced, without betting against the technology itself.

Because that is the distinction worth holding onto. This is not 2000-style capitulation; no serious participant argues the AI buildout is valueless. The hyperscalers still generate enormous operating cash; their shareholders are lending it to the future rather than losing it. Nvidia remains supply-constrained. What has changed is the marginal dollar's question — from "how do I get more AI exposure" to "what am I being paid to wait" — and sectors with visible earnings answer that question better than sectors with promises.

The stampede scenario deserves its own caution. Rotations that begin as prudence can end as momentum, and momentum does not check fundamentals on the way through. If healthcare multiples stretch far enough, the trade becomes crowded exactly as AI revenues surprise positively, and the whipsaw will be sold as wisdom in both directions. The flow data says reassessment; only time says whether the reassessment was early or merely contagious.

For now, read the tape as the collective judgment it is: AI's promise remains real, its timeline uncertain, and other sectors' cash flows newly cheap relative to both. The market is not abandoning the future. It is charging the future interest.

-- THEO KAPLAN, San Francisco