Micron just posted the best quarter in its history. Revenue hit a record $41.5 billion, up 346% from a year ago. Earnings per share came in at $25.11, beating estimates by roughly 24%. Gross margin more than doubled to 84.9%. Guidance for the next quarter came in even stronger.

The stock has fallen 22% since that report.

So what actually knocked the stock down? Not the business. A handful of things landed at once: fears of an Iran ceasefire collapsing and reviving inflation worries, a South Korean regulator warning about leveraged memory-stock ETFs, a rival’s blockbuster Nasdaq debut, a well-known short-seller publicly betting against Micron, and plain old profit-taking after the stock ran up roughly 700–800% in a year. None of that touches Micron’s actual results. Pile it on top of a genuinely excellent quarter, and you get a stock in “bear market” territory days after reporting record numbers.

Why are the numbers this good? Micron makes memory chips — DRAM, NAND flash, and increasingly high-bandwidth memory (HBM) used in AI accelerators. Revenue here comes down to two levers: how many chips ship, and what price each one fetches. Right now pricing is doing almost all the work — shipment volumes were only up in the low single digits. That’s the signature of a supply-constrained market: AI workloads are consuming memory faster than the industry can produce it, and management expects that squeeze to last well past 2026.

The more important detail buried in the report: Micron has now locked in 16 multi-year Strategic Customer Agreements with defined pricing, rather than leaving revenue at the mercy of the spot market. Combined with its broader order backlog, the company is sitting on roughly $100 billion in contracted-but-undelivered business — and management says that figure likely understates what eventually gets booked as revenue.

Why isn’t this a debt story? Unlike some other AI-infrastructure names funding growth with leverage, Micron ended the quarter with $30.2 billion in cash against just $6.4 billion in debt — a real net cash position. Free cash flow came in positive at $18.3 billion, even after $7.1 billion in capital spending. Micron used that strength to retire $5.4 billion of older, higher-rate debt in March and raised its dividend by about 30% in the same stretch. Debt-to-equity sits at 0.06, among the lowest of any large chipmaker.

That matters because Micron’s real risk isn’t a balance sheet problem. It’s the same risk memory investors have lived with for decades: this industry booms and busts, hard. The open question is whether these new multi-year contracts genuinely soften that cycle, or just delay the next downturn.

Who’s actually ahead in this race? Global memory is a three-player game. Samsung leads DRAM share at about 38%, SK Hynix is second at roughly 29%, and Micron sits third around 22%. In the category that matters most right now — HBM, the memory feeding AI chips — Micron trails badly. SK Hynix controls 56–58% of HBM share and has reportedly locked up two-thirds of orders for Nvidia’s next-generation chips. Micron holds a respectable, improving position in the low-20% range, with HBM4 shipments ramping at roughly twice the prior generation’s pace. It’s fighting hard for share in the most prized part of this cycle — not leading it. Adding to the noise, SK Hynix just completed a massive U.S. Nasdaq listing, giving investors a second AI-memory stock to rotate into.

Is the price still right? Despite the enormous one-year run, Micron trades at just 6.3–6.8x forward earnings — about 82% below the semiconductor industry’s median of roughly 35x. Its trailing multiple of about 22x is actually in line with its own five-year average. In other words, earnings have grown even faster than the stock price. The market isn’t naively assuming these margins last forever; some normalization is already priced in.

Michael Burry’s disclosed short position, and short interest near 3.7% of float, show skeptics exist. But with a fortress balance sheet, a sub-7x forward multiple, and demand that’s contracted rather than hoped for, the risk here looks very different from companies funding AI buildouts with debt.

Bottom line: the fundamentals argue for real conviction — record results, raised guidance, a balance sheet getting stronger rather than weaker, and a valuation with genuine cushion. What keeps this from top-conviction territory is the size of the move already made, plus the fact that Micron isn’t the clear leader in HBM, the highest-margin part of this cycle. This is a multi-year thesis built for volatility, not a trade for the faint of heart.

For the full financial breakdown, valuation model, and competitive analysis, visit our Reports page for the complete Micron (MU) research report.