$MU Just Broke Memory Economics — 86% Gross Margins Are Not Normal and Wall Street Is Finally Noticing
A week of chipmaker carnage, meme stock chaos, and a Micron print that rewrote the memory playbook

Ticker Ratings
| Ticker | Rating | Entry Price | Current | $ Gain | % Gain |
|---|---|---|---|---|---|
| MU MICRON TECHNOLOGY INC | buy | $1213.97 | — | — | — |
| SNDK Sandisk Corp | hold | $2150.02 | — | — | — |
| NVDA NVIDIA CORP | buy | $200.69 | — | — | — |
| GOOGL Alphabet Inc. | hold | $342.10 | — | — | — |
| WEN Wendy's Co | hold | $8.23 | — | — | — |
| HTZ HERTZ GLOBAL HOLDINGS, INC | sell | $3.09 | — | — | — |
| IBM INTERNATIONAL BUSINESS MACHINES CORP | buy | $261.51 | — | — | — |
| QCOM QUALCOMM INC/DE | buy | $223.00 | — | — | — |
| GEV GE Vernova Inc. | buy | $1080.10 | — | — | — |
| AFRM Affirm Holdings, Inc. | hold | $78.25 | — | — | — |
| AMZN AMAZON COM INC | buy | $233.12 | — | — | — |
| XOM EXXON MOBIL CORP | buy | $136.37 | — | — | — |
| FDX FEDEX CORP | hold | $316.40 | — | — | — |
Let's set the scene: the Philadelphia Semiconductor Index dropped ~8% mid-week after SK Hynix and Samsung each cratered 10%+ on fears of slowing AI memory expansion. $MU itself fell 13% in a single session. Reddit was in full panic mode, X was declaring the AI bubble officially popped, and every permabear with a YouTube channel dusted off their 'dot-com 2.0' slides.
Then Micron reported. Q3 EPS came in at $25.11 vs. $20.49 expected. Q4 revenue guidance? $49-51B vs. street estimates of $43.24B. Gross margins hit 84.9% in Q3 with Q4 guided at 86% — a figure that, as Bloomberg noted on air, memory makers have never posted in history. Shares ripped ~9% after hours. The crowd that sold the dip on Monday looked very silly by Thursday.
Meanwhile, $WEN squeezed 28% on a Reddit 'save Wendy's' campaign, $HTZ face-planted 38% on a profit warning, SpaceX launched a $20B bond sale, and the Iran-Hormuz soap opera kept oil traders stress-eating. It was a week that reminded everyone: sentiment moves first, fundamentals catch up — or don't.