MU Cracks Below $945 as Raymond James Flags an Early Memory-Cycle Peak

The Catalyst

The trigger was an analyst note, not a company event. Raymond James' Karl Ackerman told clients that DRAM and NAND average selling prices could peak in mid-2026 and start declining quarter-over-quarter into early next year. The Street's working assumption had been mid-2027 at the earliest, so this pulls the feared cyclical rollover forward by roughly a year — and for a memory maker, a price peak is the part of the cycle that historically does the most damage.

Worth being precise about what this is: Ackerman kept his Outperform rating and pointed to long-term supply contracts as a cushion. This wasn't a downgrade. But after a parabolic run, you don't need a downgrade — you need a credible voice naming the top, and a one-year pull-forward of the peak is exactly that. Investing.com framed the drop as pricing-peak anxiety meeting an over-owned name.

Why It Hit So Hard

The note didn't land in a vacuum. The same session, Broadcom reported strong numbers but declined to raise its $100 billion full-year AI revenue target, and the stock fell as much as 15% intraday. When the bellwether reiterates instead of raises, every AI-leveraged chip name gets re-rated on the spot — Micron, SanDisk, and Western Digital all dipped in sympathy.

The other half of the explanation is positioning. Micron had just pushed past a $1 trillion market cap for the first time and was up roughly 93% in a month, peaking near $1,089. A name that vertical doesn't need much to flush — the equity broke back under $1,000, and the HIP-3 perp tracking a single Micron share extended that to $945, a 10.54% move over 24 hours.

What to Watch

The real test is fiscal Q3 earnings on June 24. Expectations are still anchored to AI-driven HBM and DRAM strength, so the relevant question is no longer whether the quarter is good — it's whether management's pricing commentary confirms or rebuts Ackerman's mid-2026 peak. Guidance that validates the peak thesis turns this dip into a trend; guidance that pushes the peak back toward 2027 makes this look like a positioning shakeout in an otherwise intact supercycle.

Context also matters here: this is the third straight cycle-doubt episode for the perp, after the Korea AI-tax aftershock and the China H200 demand freeze. Each time the bear case has been a demand or pricing question rather than a balance-sheet one. That keeps the name fast in both directions and is why the perp is doing $263M of 24h volume on the move.