Micron at 300 Percent: Supercycle, or the Top Nobody Wants to Call

Micron just capped a quarter that will be studied for years. The stock is up roughly 300 percent in 2026, it printed record quarterly revenue north of 40 billion dollars, and it sits at the top of every AI memory watchlist. And right on cue, the doubts arrive.

The lawsuit tells you how good the pricing is

This week Micron, Samsung and SK Hynix got hit with an antitrust class action alleging they held back DRAM supply to inflate prices during the AI boom. Set aside whether it has legal merit. The more useful signal is what it implies: memory pricing has run so hard that regulators and plaintiffs now assume it must be collusion rather than a genuine shortage. When your pricing power draws lawsuits, the near term demand picture is not the problem.

Why this cycle looks structural

I have traded enough memory cycles to be naturally skeptical. What keeps me long is HBM. High bandwidth memory is hard to make, the yields are punishing, and capacity takes years to build. $Micron Tech(MU.US) has its allocation locked under binding contracts, and the read across the sector agrees: Sandisk is up over 700 percent this year on the same NAND tightness, and Bernstein just put a 3,000 dollar target on it. Three names all shouting the same thing about supply is not a coincidence.

The SK Hynix listing is the capstone

On July 10, SK Hynix lists ADRs on the Nasdaq targeting around 29 billion dollars, with 58 percent of the global HBM market behind it. That raise exists to fund more HBM capacity for one reason: Nvidia and the hyperscalers cannot get enough. It is the loudest confirmation yet that this is a multi year buildout, not a quarter of hype.

How I am playing it

I am staying long Micron and Sandisk as my core memory exposure. I trimmed a little into the run because discipline matters and nobody goes broke taking profit. The real risk is not the lawsuit, it is hyperscaler capex. If even one big buyer signals a pause, these names are priced for perfection and will correct fast. Until the data turns, I let the winners run with a hand near the exit.

Not financial advice, just how I read the setup.