Micron Technology (MU)

Investment Review After Q2 FY2026 Earnings (2026.03.18)

TL;DR

Micron is the #3 global memory maker (DRAM ~25% share, NAND ~15%) — now an AI infrastructure enabler via HBM. Q2 FY2026 was a record: revenue $23.9B (3x YoY), DRAM $18.8B (+207% YoY), NAND $5.0B (+169% YoY), OM% 76%. Stock currently $366 at 9x P/E(ttm) — counter-cyclical low at peak earnings. Big news: HBM4 volume shipments begun for NVIDIA Vera Rubin, first 5-year strategic customer agreement signed; balance sheet swung from net debt to $6.5B net cash; 30% dividend hike. Q3 FY2026 guided at record $33.5B revenue (+40% QoQ), GM ~81%, EPS $19.15. Key risk: massive capex ramp ($25B+ FY2026, stepping up FY2027) creates execution and cycle exposure.

Summary

Finding	Assessment
5-Year Stock Return	Stock currently $366, up 326% from $86 (5 years ago). TTM OP surged from $3.8B to $51.8B. P/E(ttm) compressed from 92x to 9x as earnings dramatically outpaced stock.
Business Model	STRONG. #3 global memory maker (DRAM ~25% share, NAND ~15%). HBM leader with in-house design/packaging. 4 business units across data center, mobile, embedded, storage.
Competitive Moat	NARROW. Technology leadership (4 DRAM nodes), oligopoly structure (3 DRAM players), HBM qualification barriers. But memory remains commoditized and cyclical — advantages are temporal.
Past 5-YR Drivers	AI/HBM supercycle ($35B→$100B TAM by CY2028), 4-node technology leadership driving cost/margin expansion, supply discipline from consolidated industry.
Recent Stock Move	Stock currently $366, up 255% from $103 (Aug 2024). TTM OP 51x increase ($1.0B→$51.8B). P/E(ttm) compressed 92x→9x. R40 at 342% — highest in our coverage.
Revenue Trend	EXPLOSIVE. +266% YoY in Q3’26. HBM TAM ~40% CAGR to $100B by CY2028. Supply tight beyond CY2026. Q2 guided $18.7B (record). Server unit growth high teens %.
Margin Trend	EXPANDING rapidly. OM% from 20% to 76% in 7 quarters. Gross margin guided 68% for Q2’26 (record). Driven by pricing power, node transitions, and HBM mix. Expansion expected to be more gradual going forward.
TTM OP Trajectory	EXPLOSIVE. From loss (-$5.7B) to $51.8B in 10 quarters. QoQ +83%. FY2026 expects record across all metrics. But memory is cyclical — peak earnings compression risk.
Key Opportunities	Q3 guided at record $33.5B revenue, GM 81%, EPS $19.15. HBM4 volume shipments for NVIDIA Vera Rubin. AI memory demand structural — data center exceeding 50% of industry TAM. Record FCF, net cash position, 30% dividend increase.
Key Risks	Massive capex ramp ($25B+ FY2026, stepping up FY2027) creates execution and cycle risk. Trade/geopolitical impacts excluded from guidance. PC/smartphone unit declines in CY2026. FCF/OP at 58% still below 60% threshold despite record FCF.
Catalysts	Q3 FY2026 guided at record $33.5B ±$750M, GM ~81%, EPS $19.15. HBM4 production ramp for NVIDIA Vera Rubin. Tariff/trade policy clarity.
P/E(ttm) Valuation	LOW at 9x (cyclical range: 9x at peak earnings, 90x+ at trough). Counter-cyclical P/E — low P/E(ttm) at peak earnings is typical for memory stocks. At the bottom of the cyclical range. Current 9x implies skepticism that peak margins sustain.

Q2 FY2026 Key Data

Revenue $23.9B (3x YoY). Records across DRAM ($18.8B, +207% YoY), NAND ($5.0B, +169% YoY). HBM ramp accelerating; OM% expanded to 76%, R40 at 342% — highest in our coverage.
HBM4 volume shipments begun for NVIDIA Vera Rubin. 16-high product sampled at 48GB per cube (+33% vs HBM4 12H). First 5-year strategic customer agreement signed; data center >50% of industry bit TAM in CY2026.
Q3 FY2026 guided at record $33.5B revenue (+40% QoQ), GM ~81%, EPS $19.15. Q3 single-quarter guidance exceeds full-year FY2024 revenue. 30% dividend hike + $350M share repurchase.

Key Takeaways from Earnings Call

HBM is the structural growth pillar: HBM4 volume shipments have begun for NVIDIA's Vera Rubin platform, with 16-high product sampled and the first 5-year strategic customer agreement signed. Management framed AI memory demand as durable, with DRAM and NAND supply-demand conditions tight through CY2026 and beyond.
Balance sheet transformation: from net debt to net cash position of $6.5B in 3 quarters, with $5B+ debt reduction and record FCF of $6.9B in Q2. Capital returns ramping — 30% dividend hike, $350M repurchase. Net cash position enables aggressive capex ($25B+ in FY2026, stepping up in FY2027) without leverage risk.
Capex and macro caveats: FY2027 capex stepping up meaningfully with $10B+ YoY increase in construction-related spend alone; Idaho, New York, Japan, Singapore, India fabs ramping simultaneously. Management expects PC and smartphone units to decline low-double-digits in CY2026 due to memory supply constraints. Trade/geopolitical impacts explicitly excluded from guidance.

Key Segments and Revenue Mix (FY2025)

Segment	FY25 Rev	Mix	Products	End Markets	Key Competitors
Compute & Networking (CNBU)	~$16B	44%	HBM, server DRAM, networking DRAM	AI data centers, cloud, networking	Samsung, SK Hynix
Mobile (MBU)	~$8B	22%	LPDRAM, mobile NAND	Smartphones, AI PCs, tablets	Samsung, SK Hynix
Embedded (EBU)	~$4B	11%	Automotive DRAM/NAND, industrial memory	Automotive (ADAS, L2+/L3), industrial IoT	Samsung, SK Hynix, Winbond
Storage (SBU)	~$8B	22%	Data center SSDs, client SSDs, QLC NAND	Cloud storage, enterprise, consumer	Samsung, SK Hynix, Western Digital, Kioxia
Total FY2025	$36.2B	100%	+46% YoY. HBM TAM CAGR ~40% through CY2028. Record revenue across all BUs.

Business Model

SUPPLIERS

Equipment makers (ASML, Applied Materials, Lam Research)
Silicon wafers & specialty chemicals
Packaging materials
Advanced metalization suppliers

Equipment / Materials →

MICRON

DRAM manufacturing (1-gamma node)
NAND manufacturing (G9 node)
HBM production (HBM4 ramping)
Global fabs (US, Japan, Singapore, Taiwan)
~$20B CapEx FY2026

Memory / Storage → $ Revenue

CUSTOMERS

NVIDIA (GPU/HBM integration)
Hyperscalers (data center servers)
Smartphone OEMs (Samsung, Apple)
PC OEMs (Dell, HP, Lenovo)
Automotive (Tesla, BYD)

Moat Assessment: NARROW

Technology leadership with 4-node advantage: Leading in DRAM (1-gamma), NAND (G9), and HBM (HBM4). In-house design, manufacturing, and advanced packaging creates differentiation. But technology advantages are temporary in memory — competitors can close the gap within 1-2 nodes.
Oligopoly market structure: Only 3 major DRAM producers (Samsung ~40%, SK Hynix ~30%, Micron ~25%) and 5 NAND producers. Consolidated structure supports supply discipline and pricing power — but memory remains more commoditized than logic chips.
HBM qualification barriers: HBM requires deep co-design with GPU makers (NVIDIA). Qualification cycles are long (6-12 months). Micron has completed CY2026 HBM supply agreements. But SK Hynix pioneered HBM and Samsung is catching up — Micron’s HBM position is earned, not locked in.

Stock vs. Earnings — 5-Year Performance

Stock currently $366, up 326% from $86 (5 years ago), driven by the AI-fueled memory supercycle — HBM demand transforming Micron from a cyclical commodity player to an AI infrastructure enabler.
TTM OP surged from $3.8B to $51.8B, as operating margins expanded to 76% — the most dramatic margin expansion in the semiconductor industry.
P/E(ttm) compressed from high double digits (92x at FY2024 start) to 9x as earnings growth outpaced stock appreciation. R40 at 342% is the highest of any company we cover.

MU Stock vs TTM Operating Profit — 5-Year Performance

Past 5-Year Key Drivers

AI/HBM supercycle: HBM revenue went from near-zero to multi-billion run rate. HBM TAM forecast to grow ~40% CAGR from ~$35B (CY2025) to ~$100B (CY2028), arriving 2 years earlier than prior outlook. Micron’s HBM4 leads industry in speed (>11 Gbps) with in-house CMOS and packaging.
Technology leadership driving margin expansion: 4 consecutive DRAM node leadership generations. 1-gamma DRAM ramping as majority of bit output in H2 CY2026. G9 NAND with record QLC mix. Cost reductions from node transitions compound into massive margin expansion (OM% from negative to 68%).
Supply discipline transforming cyclicality: Industry supply constrained beyond CY2026. Micron pursuing multi-year contracts with specific customer commitments. CapEx disciplined at ~$20B FY2026 (up from $18B) — investing in cleanroom but pacing supply to demand. Memory industry consolidation (3 players: Samsung, SK Hynix, Micron) reduces boom-bust severity.

Quarterly Key Metrics

	Q2 2024	Q3 2024	Q4 2024	Q1 2025	Q2 2025	Q3 2025	Q4 2025	Q1 2026	Q2 2026	Q3 2026(E)
Quarter End	Feb 2024	May 2024	Aug 2024	Nov 2024	Feb 2025	May 2025	Aug 2025	Nov 2025	Feb 2026	May 2026
Stock Price	$118	$131	$103	$84	$87	$123	$167	$285	$338	$366
Stock QoQ	+39%	+11%	-21%	-18%	+4%	+41%	+36%	+71%	+19%	+8%
Rev YoY	+58%	+82%	+93%	+84%	+38%	+37%	+46%	+57%	+196%	+266%
OM%	-1%	10%	20%	25%	22%	24%	33%	45%	68%	76%
R40 (Rev YoY + OM%)	+57%	+91%	+113%	+109%	+60%	+60%	+79%	+102%	+264%	+342%
TTM OP ($M)	-3,985	-1,799	1,021	4,336	6,149	7,721	9,908	13,874	28,255	51,765
TTM OP QoQ	-35%	-55%	-157%	+325%	+42%	+26%	+28%	+40%	+104%	+83%
FCF/OP ttm	—	—	12%	13%	10%	24%	17%	34%	36%	58%
Debt/EBITDA	3.3x	2.1x	1.5x	1.1x	1.0x	1.0x	0.8x	0.5x	0.3x	0.3x
P/E(ttm)	—	—	207x	26x	21x	22x	22x	26x	15x	9x

Recent Stock Performance & Drivers

Stock currently $366, up 255% from $103 (Aug 2024). The most dramatic appreciation in our coverage — driven by explosive earnings growth as AI memory demand surged.
TTM OP exploded from $1.0B to $51.8B — a 51x increase in 8 quarters, driven by HBM ramp, DRAM pricing power, and massive margin expansion (OM% 20% → 76%).
P/E(ttm) compressed from 92x to 9x as earnings growth dramatically outpaced stock appreciation. R40 at 342% is the highest we’ve ever recorded.
Cash flow risk: FCF/OP at 58%, below the 60% threshold — massive capex ($20B guided for FY2026, up from $18B) consuming most operating cash flow. Micron is investing in cleanroom expansion to meet demand that exceeds supply.

Opportunities

Q3 FY2026 guided at record $33.5B revenue (+40% QoQ), GM ~81%, EPS $19.15. Q2 revenue of $23.9B tripled YoY with records across DRAM ($18.8B, +207% YoY), NAND ($5.0B, +169% YoY), and HBM. Q3 single-quarter guidance exceeds full-year FY2024 revenue. AI-driven memory demand is structural — data center bit TAM exceeding 50% of industry for first time in CY2026.
HBM4 volume shipments begun for NVIDIA Vera Rubin platform. HBM4 16-high product sampled with 48GB per cube (+33% vs HBM4 12H). HBM4E development underway for CY2027 volume ramp. First 5-year strategic customer agreement signed. DRAM and NAND supply-demand conditions tight through CY2026 and beyond.
Record FCF of $6.9B in Q2 with net cash position of $6.5B. Debt reduced by $5B+ in 3 quarters. 30% dividend increase. $350M share repurchase. Balance sheet transformation from net debt to net cash enables aggressive capex ($25B+ in FY2026, stepping up in FY2027) without leverage risk.

Challenges

Massive capex ramp creates execution risk. FY2026 capex >$25B, FY2027 stepping up meaningfully with >$10B YoY increase in construction-related capex alone. New fabs in Idaho, New York, Japan, Singapore, India all ramping simultaneously. Any construction delay or demand softening could create overcapacity.
PC and smartphone unit declines in CY2026. Management expects low-double-digit percentage declines in both PC and smartphone units due to DRAM/NAND supply constraints. While content growth per device offsets unit weakness, a macro downturn could compound unit and pricing pressure.
Trade/geopolitical risk explicitly excluded from guidance. Management noted “any impacts from trade or geopolitical developments are not included in our guidance.” Tariffs on memory imports or China restrictions could materially impact demand or supply chain.

Next 12 Months

Q3 FY2026 guided at record $33.5B ±$750M, GM ~81%, EPS $19.15. MU has beaten guidance high-end for 4 consecutive quarters. A beat above $34B with GM >81% would confirm the AI memory supercycle and could push stock toward $400-450. A miss or weak Q4 guidance due to tariff/macro would trigger sharp correction given 9x P/E.
HBM4 production ramp for NVIDIA Vera Rubin. Volume shipments have begun. Qualification and yield trajectory will determine share of NVIDIA’s next-gen platform. Any share gain vs Samsung/SK Hynix would be a major catalyst.
Tariff/trade policy clarity. Guidance explicitly excludes trade impacts. Any tariff announcements on memory imports or tightening of China restrictions would be the #1 risk to stock in next 6 months.

Next 12-24 Months

FY2027 capex step-up is the key test. Construction capex increasing >$10B YoY plus higher equipment spend. If AI memory demand sustains, this investment drives long-term earnings power. If demand softens, capex overshoot compresses FCF/OP and triggers the memory cycle downturn fears that have historically crushed MU’s P/E.
HBM4E and 1δ node transitions. HBM4E ramping in CY2027 with EUV adoption at 1δ node. These transitions determine Micron’s competitive position vs Samsung and SK Hynix for the next 2-3 years. Technology leadership = pricing power = margin sustainability.
Robotics as 20-year growth vector. Management identified humanoid robots as requiring memory comparable to a high-end L4 autonomous vehicle. Early but could become a meaningful demand driver beyond data center by FY2028+.

Core Metrics Scorecard

	Q4'21	Q1'22	Q2'22	Q3'22	Q4'22	Q1'23	Q2'23	Q3'23	Q4'23	Q1'24	Q2'24	Q3'24	Q4'24	Q1'25	Q2'25	Q3'25	Q4'25	Q1'26	Q2'26	Q3'26
Quarter End	Aug21	Nov21	Feb22	May22	Aug22	Nov22	Feb23	May23	Aug23	Nov23	Feb24	May24	Aug24	Nov24	Feb25	May25	Aug25	Nov25	Feb26	May26
R40 > 40%	✅	✅	✅	✅							✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
R40 accelerate								✅	✅	✅	✅	✅	✅			✅	✅	✅	✅	✅
TTM OP QoQ > 10%	✅	✅	✅						✅	✅				✅	✅	✅	✅	✅	✅	✅
TTM OP QoQ accelerate									✅					✅			✅	✅	✅
TTM FCF/OP > 60%	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	❌	✅	✅	✅	✅	✅	✅	✅
Debt/EBITDA < 3.5	✅	✅	✅	✅	✅	✅	✅	✅	❌	❌	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
Total passes	4	4	4	3	2	2	2	3	4	3	4	4	3	5	4	5	6	6	6	5
Stock QoQ	-16%	+30%	-16%	-29%	-9%	+0%	+22%	+5%	+8%	+25%	+39%	+11%	-21%	-18%	+4%	+41%	+36%	+71%	+19%	+8%
P/E(ttm)	13x	13x	9x	6x	6x	9x	40x	—	—	—	—	—	207x	26x	21x	22x	22x	26x	15x	9x

Stock QoQ tracks changes in total passes through the memory cycle. During the downturn (Q1’23–Q3’23), passes fell to 2 as R40 and TTM OP QoQ both failed, and the stock declined –9% to +0%. When passes rose to 4 by Q3’24, the stock rallied +39%. The strongest runs came as passes climbed from 5 to 6 across Q1’26–Q3’26 (stock +36%, +71%, +19%). Conversely, when passes dropped from 4 to 3 in Q1’25 (FCF/OP failed), the stock fell –21%.
R40 and TTM OP QoQ are the primary swing metrics, while FCF/OP and Debt/EBITDA rarely drag. R40 above 40% and TTM OP QoQ above 10% both pass in the strongest quarters (Q4’21 at 6 passes, Q1’26–Q3’26 at 6 passes). When both fail (Q1’23–Q3’23, 2 passes each), the stock flatlines. FCF/OP failed only once (Q1’25) and Debt/EBITDA failed only twice (Q1’24–Q2’24), so balance-sheet metrics almost never cap the score — the cycle is driven by growth.
Notable pattern: TTM OP QoQ acceleration is the rarest pass — clearing in only 6 of 22 quarters. It coincides with the sharpest stock moves: when it fires alongside 5–6 total passes (Q2’25 at 5 passes, stock +41%; Q1’26 at 6 passes, stock +36%), the stock surges. Its absence even with high pass counts (Q3’26 at 6 passes but no acceleration, stock +19%) marks the late stage of the cycle.

Chart: R40 vs Stock Price (Since 2020)

Chart: Scorecard Total Passes vs Stock Price (Since 2020)

Source: SEC EDGAR annual and quarterly filings, company earnings conference calls and presentations.

Disclaimer: This report is for educational purposes only. NO investment advice.