The Most Violent Momentum Unwind in 45 Years: Why Chip Stocks Crashed While the Index Held Up

TL;DR The market didn't crash — one crowded corner of it did. From their June highs: SanDisk -53%, Intel -39%, Micron -32%, and the chip index -24%. Over the same stretch the S&P 500 is up more than 7% for the year and sitting near a record high.

The index looks fine. Inside it, there's a massacre

Let me clear up the biggest misconception first: the overall market is not falling apart.

The Nasdaq is up over 5% this year. The S&P 500 is up more than 7%. The Dow is up close to 10%. Judging by the indexes alone, this is a boringly calm bull market. But underneath that calm surface, one specific group has been obliterated — the momentum names that went straight up over the past year and a half, and almost all of them are semiconductor and AI chip companies.

One recent session captures it. The Dow finished higher while Micron fell 9%, AMD fell 8%, and Applied Materials fell nearly 8%. Same day, same market.

What stands out most to me is how surgical the selling was. It hit the crowded trade and left nearly everything else untouched — not just untouched, but higher. That's not a market breaking down. That's a position being liquidated. Those are two very different events, and they call for very different responses.

The most violent unwind in 45 years

One number from Goldman Sachs explains the character of this move better than anything else. Goldman tracks a "momentum group" — the basket of stocks that have been climbing fastest — and over the past three months that group posted its wildest, most violent swings in 45 years.

The wording matters here. This is not the biggest drawdown in history; it is the most violent one. It's a story about speed and the sheer force of the reversal, not about depth. When something that has been rocketing straight up flips direction this hard, an enormous number of people get caught completely flat-footed.

The surrounding data tells the same story. More than $1 trillion of value has been wiped out of US tech stocks. The Nasdaq logged four losing sessions in a row. And one well-known Wall Street mood gauge, the bull and bear indicator, hit 9.6 out of 10 — a reading that only shows up when investors have crowded far too heavily into one corner of the market. We all know which corner that was.

What the damage actually looks like

Abstractions don't land, so here are real prices on real dates.

SanDisk closed at $233.50 on June 25. By July 28 it was around $109.60 — a 53% crash in roughly one month. Intel went from nearly $141 on June 22 to $86, down 39%. Micron fell from well over $1,200 to $820, down 32%.

The rest differ only in degree.

Stock	Decline from June high
SanDisk	-53%
Marvell	-45%
Intel	-39%
Western Digital	-38%
Lam Research	-38%
KLA	-37%
Applied Materials	-34%
Micron	-32%
Seagate	-32%
AMD	-22%
Broadcom	-21%
Dell	-16%

The main index tracking chip stocks fell about 24% from its June 2026 high. That is officially a bear market for the entire sector — at the exact moment the S&P 500 was up 7% on the year and near an all-time record. Two completely different worlds, sitting side by side.

Put it in dollars and it gets sharper. Split $120,000 evenly across those 12 names at their June peaks and one month later you'd have roughly $79,000. About a 34% loss in a single month.

Here's the frightening part. You could have been completely right that AI and memory demand were booming and still lost that money — because the price you paid already assumed perfection would continue indefinitely.

This didn't happen in a single day

The timeline tells the whole story.

June 2 — The Nasdaq and many of the biggest tech names hit their highs. The top of the mountain.

June 23 — The chip index drops nearly 8% in a single session, with Micron alone falling 13% that day. The first crack in the ice.

Mid-July — The damage spreads fast. In a single week, investors pulled roughly $11 billion out of chip funds — reportedly the largest weekly exit of this entire century.

July 28 — The selling accelerates again. SanDisk -14% in one day, Micron nearly -9%, AMD -8%. And yet the S&P 500 and the Dow both finished that same day just fine.

That last line is the whole point. The market didn't break. One crowded seat did.

They crashed right after reporting spectacular news

This is the part that really messes with people's heads. A lot of these companies didn't fall on bad news. They fell immediately after reporting extraordinary news.

Take Micron. It posted its fifth record quarter in a row, with revenue over $41 billion — up from $9 billion a year earlier — and enormous profits. The stock is down more than 32%. Micron actually hit its highest price the very day after it reported those record numbers.

SanDisk is similar. Its recent quarter showed revenue nearly doubling, with the data center business up 233%. Spectacular. The stock still fell 53%. To be fair and precise, SanDisk's crash didn't come immediately on the back of a blowout report — but it clearly wasn't the business deteriorating either.

The cleanest signal came on July 7 from Samsung. Samsung announced massive blowout numbers and chip stocks around the world fell anyway. Micron dropped. SanDisk dropped. The whole sector dropped, on a day of fantastic business news.

To me, that is the most honest indicator of where you are in a cycle. Early in a run — when a stock is cheap and unloved — merely good news sends it flying. Near the top, when everyone already owns it and the price assumes perfection, even jaw-dropping news isn't enough. Everyone who was going to buy has already bought.

It's probably not over

One Goldman strategist warned that AI and tech positioning is "far from cleared," meaning there could be more forced selling ahead.

The retail numbers are especially painful. The basket of stocks retail investors love most fell about 13% in July alone, and one tracker of the most heavily retail-owned names is down roughly 25% since June. Momentum looks like a strategy, but it functions as a risk. It's a rubber band that stretches and stretches until it snaps back the other way — and when it snaps, fear of missing out flips overnight into fear of holding. Even large professional hedge funds were reportedly running for the exit at breakneck speed.

My takeaway is straightforward. None of this says AI was wrong. It says a great story and a great investment are two different things. A wonderful company bought at the wrong price is still an awful investment. That one sentence explains the last two months entirely.

For more context, I'd also read Micron and the AI memory bottleneck and my SanDisk vs Intel comparison.