Micron (MU) Q2 FY2026 Earnings: Record $23.9B Quarter — So Why Did the Stock Fall?

Published on: Wed Mar 18 2026

Micron Technology ($MU) just posted the largest quarterly revenue beat in its history — $23.86 billion against a consensus of $18.9 billion, a 26% upside — and the stock fell 4%.

That’s the paradox at the center of Micron’s Q2 FY2026 report. A quarter that featured record revenue, record gross margins, and record free cash flow was met with a selloff rooted in one fear: can this possibly continue? The short answer, based on Q3 guidance of $33.5 billion, is yes — and likely faster than the market expected.

Results at a Glance

Q3 FY2026 Guidance: Revenue $33.5B ± $750M, gross margin ~81%, EPS $19.15 — implying yet another record quarter.

Revenue: The Memory Cycle Bites Back

Micron’s revenue history is a masterclass in cyclicality. The memory industry swings violently between supply gluts and demand booms, and Micron’s financials reflect every turn. FY2023 saw revenue collapse 50% year-over-year as a chip inventory hangover hammered pricing across DRAM and NAND. Then AI infrastructure spending arrived and rewrote the story entirely.

The V-shaped recovery from the FY2023 trough is as sharp as any in semiconductor history. TTM revenue of $58.1 billion nearly quadruples the FY2023 level in just three years — driven by HBM (High Bandwidth Memory) pricing power, data center DRAM demand, and the beginning of a structural shift where memory is consuming an ever-larger share of the AI server bill of materials.

Free Cash Flow: From Trough to Record in Three Years

The FCF story is equally dramatic. Micron generated negative $6.1 billion in free cash flow in FY2023 as it continued building out manufacturing capacity into a brutal pricing environment. Two years later, a single quarter (Q2 FY2026) generated $6.9 billion — more than the entire FY2023 operating cash flow.

The TTM FCF of $10.3 billion represents a cash generation machine that simply did not exist two years ago. With capex running at $5B per quarter and revenue scaling faster than infrastructure costs, the operating leverage is compounding rapidly.

Key Operating Metrics

HBM (High Bandwidth Memory): The real story behind the margin surge. HBM commands a 5–8x pricing premium over conventional DRAM and is the critical enabler of AI accelerators like NVIDIA’s H100/H200/Blackwell chips. Micron’s entire 2026 HBM supply is already committed — this is not a demand story, it’s a supply story.

Gross Margin: 74.6% in Q2, up from 56.8% last quarter. Guided to ~81% in Q3. A 25-percentage-point expansion in two quarters reflects both HBM mix shift and manufacturing cost improvements from the 1-gamma DRAM and G9 NAND process nodes.

Segment Mix:

• DRAM: $18.8B revenue (79% of total), up 207% YoY and 74% sequentially
• NAND: $5.0B revenue (21% of total), up 169% YoY

Data Center / Cloud: Cloud memory revenue rose over 160% year-over-year to $7.75B. Mobile and client combined for $7.71B.

Why Did the Stock Fall?

The 4% post-earnings decline — despite guidance of $33.5B next quarter — comes down to one concern: growth rate sustainability.

The bull case has been so thoroughly priced in that even a historic beat triggered profit-taking. Key worries:

1. HBM supply constraints are self-limiting. Once Micron’s committed supply is delivered, can they grow HBM capacity fast enough to sustain momentum through FY2027?
2. Memory cycles end. FY2023 is a fresh reminder of what happens when pricing turns. Bulls argue this cycle is structurally different due to AI; bears say every cycle says that.
3. Concentration risk. A handful of hyperscalers — Amazon, Microsoft, Google, Meta — represent the bulk of the incremental demand. If any large AI capex programs slow, Micron feels it immediately.

William Blair’s Sebastien Naji framed it well: Micron is trading at just 6x their 2026 EPS estimates — below its historical multiple — making it a “core beneficiary of the AI supercycle.” The selloff may be less about fundamentals and more about fear of cycle duration.

Strategic Highlights

HBM4 Launch: Micron began volume shipments of HBM4 36GB 12-high modules in Q1 calendar 2026, aligned with NVIDIA’s Vera Rubin GPU platform roadmap. This positions Micron as a day-one supplier for the next generation of AI accelerators — a critical competitive advantage over Samsung and SK Hynix.

1-Gamma DRAM & G9 NAND: These next-generation process nodes are driving meaningful cost-per-bit reductions, which will sustain margin expansion even as HBM pricing normalizes over time. The manufacturing technology gap between Micron and competitors has narrowed significantly over the past two years.

Memory’s Growing BOM Share: Memory now accounts for a larger slice of the total server bill of materials than at any prior point in industry history. As AI model sizes grow and inference workloads scale, this structural shift is durable regardless of near-term pricing cycles.

Valuation

The bull case: At 6x forward earnings for a company growing revenue at 200%+ YoY with 75%+ gross margins, Micron is cheap — if the cycle holds. Q3 guidance of $33.5B suggests it is holding, and HBM supply commitments give multi-quarter visibility. Cantor Fitzgerald has a $700 price target, reflecting a scenario where the AI memory cycle extends meaningfully into FY2027.

The bear case: Memory is famously cyclical, and the street has been burned before by extrapolating peak-cycle earnings. The stock at $444 has already moved from ~$80 in 2023, embedding significant optimism. If HBM pricing softens in H2 FY2026 — either from competitor capacity additions (Samsung, SK Hynix) or demand normalization — the P/E multiple will look very different.

Analyst consensus: Strong Buy skew — RBC ($525), Bernstein ($510), TD Cowen ($500), Wedbush ($500), Rosenblatt ($500). Median target ~$510, or ~15% upside from current levels.

What to Watch Next Quarter

1. HBM capacity expansion cadence — can Micron grow HBM supply enough to meet FY2027 commitments from next-gen NVIDIA platforms?
2. Gross margin at 81%+ — Q3 guidance implies further expansion. If actual margins come in below guidance, the pricing thesis cracks.
3. NAND recovery — NAND remains the weaker segment. A recovery in NAND pricing (enterprise SSD, mobile) would be incremental upside to consensus models.

Micron’s Q2 was not just a beat — it was a paradigm shift in what a memory company’s financials can look like in an AI-driven infrastructure buildout. The 4% post-earnings dip looks more like an opportunity than a verdict. The market will decide whether the AI capex cycle is secular or cyclical — but Micron’s own Q3 guidance is making the bull case for them.

This article is for informational purposes only and does not constitute investment advice. Always do your own research before making investment decisions.