Micron just printed a gross margin of 67.7 percent, and if you have followed this company for more than one cycle, you know exactly how rare and how dangerous that number is.
Micron just printed a gross margin of 67.7 percent, and if you have followed this company for more than one cycle, you know exactly how rare and how dangerous that number is. Let me tell you what this filing actually says.
Transcript
Micron just printed a gross margin of 67.7 percent, and if you have followed this company for more than one cycle, you know exactly how rare and how dangerous that number is.
Let me tell you what this filing actually says.
Start with the top line. Thirty-seven and a half billion in revenue against twelve billion in cost of goods. That spread is almost obscene for a commodity memory manufacturer. Micron's entire history is margin compression and cyclical agony — DRAM and NAND are brutal businesses where you get rich and then you give it all back. So when you see a gross margin approaching seventy percent, the first question is not celebration. The first question is: where are we in the cycle, and how much of this is structural versus temporal.
The trailing twelve-month gross margin sits at 58.4 percent. That means this single quarter is running nearly ten points above the trailing average. The business is accelerating into the print, not decelerating. That kind of intra-cycle margin expansion typically means one thing in memory: pricing power driven by constrained supply meeting a demand spike. Right now that demand spike has a name, and it is HBM — high bandwidth memory for AI infrastructure. Micron's HBM3E is reportedly sold out well into calendar 2025 and beyond. The conventional wisdom already knew this. The filing tells you the magnitude is larger than the conventional wisdom priced.
Operating margin of 59.4 percent. Net margin of 50.7. On thirty-seven billion in revenue. The operating margin running eleven points above the trailing twelve-month gross margin means operating leverage is compounding on top of gross leverage. OpEx is not scaling with revenue. That is the hallmark of a company moving up the product stack — when you sell high-value HBM instead of commodity DRAM, your incremental revenue does not require proportional incremental engineering or sales cost.
Now here is the number most people scroll past. Forward PE of 10.75 against a trailing PE of just over fifty. That is not a typo and it is not a rounding artifact. It means the Street is modeling an earnings number roughly five times the current trailing figure over the next twelve months. Alternatively — and this is the bear case embedded in that same multiple — it means the market is discounting that these margins are ephemeral, that the cycle turns, and earnings collapse back toward something more historically normal. A ten-times forward multiple on a semiconductor name is the market essentially saying it does not believe the earnings base is durable. That tension is the entire story.
The 52-week range tells you everything about sentiment volatility: 103 dollars on the low, 1213 on the high. That is not a stock. That is a referendum. The market has fundamentally disagreed with itself about this company's value by a factor of twelve within a single year. That range is a reminder that memory semiconductors reprice faster and more violently than almost any other asset class when the cycle turns.
EPS growth of 412 percent year-over-year on a trailing twelve-month basis. Revenue up 85.5 percent. The EPS leverage is running nearly five times the revenue growth rate — that is operating leverage at a level that looks extraordinary until you remember that eighteen months ago this company was burning cash and writing down inventory. The base effect is doing real work here, but the absolute margin levels confirm it is not purely base effect. This is genuine pricing power in a product — HBM — where Micron is one of three suppliers on the planet.
What the Street likely mispriced is the duration of the HBM constraint. Most cycle analysts modeled memory the old way: DRAM is fungible, pricing normalizes in four to six quarters. HBM is not fungible. The process complexity and yield challenges mean supply does not respond to price signals the way conventional DRAM does. The filing's margin profile is the numerical argument that this cycle has a different shape.
That is the teardown. The numbers were always there — most people just do not look. See you at the next filing.