The Great Rotation: the Dow prints a record while tech logs a rare four-day slide

1What happened this week

On Thursday, the split was stark: the Dow touched an intraday record of 52,655.66 (above its June 16 peak of 51,999.67), while the Nasdaq closed down 0.46% at 25,358.60 — its first four-day losing streak since February.

The mega-caps led lower: Apple −6%, Microsoft −3.5%, Amazon −3.1%, with Nvidia, Meta and Alphabet all off too. What lifted the Dow was the cohort with no tech in it — Johnson & Johnson +1%, Caterpillar +6%, as healthcare, financials and industrials carried the load. The tell: chipmaker Micron delivered crushing earnings and jumped 15%, and it still couldn't rescue the sector's mood.

2Why — the "Great Rotation" is real money, not just mood

When two indexes move opposite ways for several sessions, market watchers read it as genuine institutional sector rotation: capital leaving high-multiple, long-duration growth names for value, defensive and cyclical stocks with more predictable cash flows. The data backs it up:

Equal-weight is beating cap-weight for the first time in 2026 — after trailing by more than 10 points in 2023. Three threads: ① "AI capex fatigue" in the mega-caps, raising debt to build but facing return questions; ② money moving into energy, industrials, staples and utilities; ③ widening earnings breadth, with a 27.1% blended Q1 EPS growth rate — profits are no longer the property of a few.

3The core debate: is AI a bubble?

Underneath the rotation is one question Wall Street is shouting about: can the enormous AI spend ever earn its money back?

The most honest reminder is historical: Cisco was profitable in March 2000 and still fell ~85% from its peak — it took 24 years to recover. A revolution can be real and the price can still be wrong. The question isn't "is AI real?" but "can the capital structure built around it survive a repricing without breaking the real thing underneath?"

4What it means· what to watch next

For an ordinary investor, the week's signal isn't "flee tech" — it's "don't put every egg in seven baskets." Broadening breadth is usually the healthier setup. What many institutions are doing: a core position plus a sleeve of equal-weight / value as a catch-up trade.

• Next Thursday: June nonfarm payrolls — early forecasts see growth easing sharply from May's 172,000, which feeds the rate-cut path.
• July 28: Alphabet's Q2 report — the first big test of whether its $175–190B 2026 capex is starting to convert into revenue.
• The deciding variable: whether revenue catches up to capex. If it does, the rotation stays gentle; if it doesn't, the overbuild becomes a reckoning.