On July 14, 2026, the stocks that carried the market all year became the reason it fell. Micron dropped about 4.7%, KLA, Marvell, Broadcom and AMD slid alongside it, and the VanEck Semiconductor ETF fell more than 3%. When the leaders crack, the whole tape feels it.
By Ruslan Averin.
This is Ruslan Averin's read on a chip selloff that was really about crowding.
The damage
| Name | Same-session move |
|---|---|
| Micron | ~−4.7% |
| VanEck Semiconductor ETF | ~−3%+ |
| KLA / Marvell / Broadcom / AMD | Broadly lower |
| Nasdaq Composite | ~−1.6% |
Why the winners are the fault line
Concentration is a two-way street. The same characteristics that made semiconductors the market's engine — high multiples, huge weightings, universal ownership — make them the fault line when sentiment shifts. Everyone owns them, so everyone sells them at once. The move was amplified by IBM's warning that enterprises are rerouting budgets toward hardware and away from software: a reminder that even inside the AI trade, the flows are uneven and the leadership can turn on a single headline.
My read
I separate the noise from the signal. The signal, from TSMC's accelerating sales in the same window, is that chip demand is still strong. The noise is a crowded, expensive group de-risking in a single session. Both are true. What the day exposed is fragility: when a market's earnings growth leans this heavily on one sector, a bad session in that sector is a bad session for everyone. That is not a reason to abandon the theme — it is a reason to respect position sizing and to expect these air pockets as a feature, not a bug, of a concentrated market.
Bottom line
The chip rout was crowding unwinding, not the demand story breaking — but it showed how much of the market now rests on one group. I am not telling anyone to buy or sell; this is analysis, not advice.
