Micron Technology (MU) — Stock Analysis & Q2 FY26 Preview

Philip A. Fisher | March 2026

The Central Question

Micron Technology presents the investor with a question that I have grappled with throughout my career: can a company that has historically been subject to brutal commodity cycles become, through management skill and technological leadership, something more durable? My answer, after careful study, is that Micron is in the process of earning that transformation — but the transformation is not yet complete, and the investor must be clear-eyed about what remains cyclical and what has structurally changed.

What This Company Does

Micron is one of three companies on earth capable of manufacturing advanced memory semiconductors at scale — the others being SK Hynix and Samsung. It produces DRAM (79% of revenue) and NAND flash (20%), serving data centres, mobile devices, automotive systems, and cloud computing infrastructure. The company's recent trajectory has been defined by a single product: High Bandwidth Memory (HBM) — the stacked DRAM modules that sit directly on AI accelerator chips like Nvidia's GPUs. HBM has transformed Micron from a commodity memory maker into a technology-differentiated supplier with pricing power.

The Latest Quarter: Q1 FY26 (Reported December 2025)

The results were exceptional by any standard I have encountered in the semiconductor industry:

The margin trajectory deserves particular emphasis. Two years ago, in Q1 FY24, Micron posted a GAAP gross margin of negative 0.7%. Today it is 56.0%. This is not a gradual recovery — it is a structural transformation driven by product mix shift toward HBM, supply discipline, and pricing power that the company has never before sustained at this magnitude.

Revenue by business unit reveals where the growth is coming from:

DRAM contributed 79% of revenue (10.8B), withpricinggainsexceeding502.7B).

Scuttlebutt Assessment

Customers

Micron's customer position has fundamentally shifted. The company has moved from being an interchangeable commodity supplier to a qualified technology partner. This distinction matters enormously. When Samsung failed Nvidia's HBM3E qualification tests and Micron passed, it was not merely a commercial victory — it was evidence that Micron's engineering had reached a level where the most demanding customer in semiconductors was willing to depend on it. SK Hynix holds 53% HBM market share, Samsung 35%, and Micron 11% — but Micron's share has grown from approximately 4% just two years ago.

The company has finalized all CY2026 HBM price and volume agreements, and capacity is sold out through the year. Micron claims it can only meet 50-67% of key customer demand in the medium term. When demand persistently exceeds supply, and the supplier has qualification moats protecting its position, pricing power follows naturally.

For data centre SSDs, Micron achieved the #2 market share position for three consecutive quarters — a first in the company's history. The 9550 Gen5 SSD and 60TB 6550 ION demonstrate genuine product innovation, not merely capacity scaling.

Employees

Glassdoor shows 3.9/5 with 77% recommending the company — a score I find encouraging for a manufacturing-intensive semiconductor firm. Great Place to Work certified at 79% in 2025. Process engineers report work-life balance challenges (3.0/5), which is common in fab environments where equipment runs continuously. Software engineers rate the company 3.5/5 for work-life balance. The overall impression is of a company where people work hard but believe in the mission. I note reports of management turnover at middle levels, which merits monitoring but is not yet alarming.

Micron's expansion plans — Idaho Fab 1 (first wafers mid-2027), Idaho Fab 2 (accelerated with $1.2B redirected CHIPS Act funding), New York megafab (construction late 2026, online circa 2030) — will create over 26,000 direct jobs. This level of long-range capital commitment speaks to management's conviction in sustained demand. The CHIPS Act provides $6.4B in federal funding, reducing execution risk.

Competitors

SK Hynix: The clear HBM leader at 53% share, with sold-out capacity through 2026 and a technology lead in HBM3E yield optimisation. SK Hynix has been first to market with HBM4 samples and maintains the closest relationship with Nvidia.

Samsung: Samsung's HBM journey has been troubled — it failed initial Nvidia HBM3E qualification and has been playing catch-up. However, Samsung plans a 50% HBM capacity surge in 2026 and is expected to gain share as its HBM3E qualifications complete and HBM4 enters production. Samsung has won AMD qualification as an alternative customer.

The competitive picture is nuanced. In HBM, Micron has earned its position through technical excellence but remains the #3 player. In conventional DRAM and enterprise SSDs, Micron is gaining share through supply discipline and product innovation. The memory shortage of 2024-2026 benefits all three players but particularly advantages those who pivoted early to HBM — which Micron did, including the strategic exit from consumer NAND (Crucial brand) in December 2025.

Product & Technology

Four technology vectors are worth examining:

1. HBM3E: Qualified at Nvidia, shipping to 4+ customers, 12-high crossover completed. Power efficiency leadership claimed. Revenue run rate approximately $8B annualised.

2. HBM4: This is the thesis test. Conflicting reports exist: some sources indicate Micron is excluded from initial Nvidia Vera Rubin builds, while others suggest qualification has been achieved. Micron claims HBM4 samples exceed 11 Gbps with 30% lower power consumption. Management states volume production begins Q2 2026. The resolution of this ambiguity — expected within weeks — will significantly affect sentiment and competitive positioning.

3. 1-gamma DRAM: The leading-edge process node, moving from sampling to volume production. Micron claims leadership for four consecutive DRAM nodes, which if accurate represents sustained R&D effectiveness under Point 3.

4. Enterprise SSD: The 9550 Gen5 (world's fastest data centre SSD) and 6550 ION (60TB, 12 GB/s) demonstrate that Micron is not merely riding HBM tailwinds but genuinely innovating across the product line.

Management

Sanjay Mehrotra has led Micron since 2017, bringing 35+ years of semiconductor experience as co-founder of SanDisk. His strategic repositioning of Micron — from commodity memory to AI-critical infrastructure — has been decisive and well-executed. The consumer NAND exit, the HBM investment ahead of competitors, the aggressive fab expansion with federal backing — these are the actions of a CEO with both long-range vision and operational discipline.

Mehrotra's guidance cadence is conservative. Q1 FY26 beat guidance by 9.1% ($1.1B). The trailing four-quarter average earnings surprise is 14.4%. This systematic underpromise/overdeliver pattern is precisely what I look for in management communication.

However, I note a concern on Point 8 (management depth). Mehrotra is clearly the driving force. Reports of middle-management turnover and limited visibility into his succession bench give me pause. This is not a disqualifying concern, but it means the investment carries key-man risk.

Analyst Sentiment

29 analysts cover Micron with a Strong Buy consensus. Price targets range from $107 to $500, with a median around $368. This extraordinary range reflects genuine uncertainty about the duration and magnitude of the HBM supercycle. I do not invest based on analyst targets, but I note that the consensus is overwhelmingly positive.

Industry Context

The memory industry is experiencing what IDC terms a "global memory shortage crisis." Gartner forecasts DRAM prices to increase 47% in 2026 due to undersupply. PC DRAM prices in Q1 2026 surged over 100% QoQ — a record. The structural driver is HBM's disproportionate wafer consumption: each gigabyte of HBM requires approximately 3x the wafer capacity of DDR5. Every wafer allocated to AI memory is a wafer denied to consumer products. This zero-sum dynamic creates pricing power across Micron's entire product portfolio, not just HBM.

Fifteen Points Assessment

Score: 8 Strong/Exceptional, 4 Adequate, 1 Uncertain, 2 Clean. This is a respectable Fifteen Points profile, held back primarily by management depth uncertainty and the inherent cyclicality of the business.

The Cyclicality Problem: What Fisher Would Say

I must be candid about the fundamental tension in this investment. Micron has demonstrated, as recently as FY23, that it can post negative 33% gross margins. The business is capital-intensive, subject to supply/demand cycles, and its products — despite increasing differentiation — remain closer to commodities than to the proprietary products I typically favour.

What has changed is the mix. HBM is not a commodity. It requires qualification at the customer level. Once qualified, switching costs are substantial — a hyperscaler will not risk its AI training cluster by changing memory vendors mid-cycle. This creates the kind of customer stickiness that, in my experience, transforms a company's competitive dynamics. The question is whether HBM represents a permanent structural shift or a temporary supercycle that will eventually normalise.

My view: the structural shift is real but the magnitude of the current margin expansion is cyclically inflated. Gross margins of 68% (Q2 FY26 guidance) are likely unsustainable over a full cycle. But margins in the 40-50% range through a full cycle — compared to the historical 25-35% average — represent a genuine, durable improvement driven by mix and qualification moats.

What to Expect in Q2 FY26 (Reporting ~March 18, 2026)

I believe the upcoming quarter will be another extraordinary result, likely exceeding guidance on every metric. Here is my reasoning:

Revenue: Guided at $18.7B ± $400M. Analyst consensus is $19.15B. TrendForce reports that actual contract negotiations in January 2026 showed DRAM and NAND pricing gains exceeding 50%, with some categories reaching triple-digit increases — substantially above the ~30% ASP growth Micron contemplated in guidance. I expect revenue of $19-20B, representing another significant beat.

Gross Margin: Guided at 68% ± 100bps (non-GAAP). Given the pricing environment, I would not be surprised to see 69-70%. This would represent an all-time high by a wide margin.

EPS: Guided at $8.42 ± $0.20 (non-GAAP). Consensus is $8.69. Given the pattern of 14.4% average earnings surprise, a result in the $9-10 range is plausible.

HBM4 Update: This is the most important qualitative disclosure. Management must address the conflicting reports about Nvidia Vera Rubin qualification. If Micron confirms qualification and volume timing, the stock will likely react positively. If qualification is delayed or limited, the competitive narrative shifts toward SK Hynix.

Guidance for Q3 FY26: The market will focus on whether the margin expansion continues or shows signs of peaking. Any guidance suggesting gross margins above 65% for Q3 would confirm the durability of the cycle.

The Growth Arc: "Fortunate and Able" — with a Qualification

Micron is both fortunate and able. It is fortunate in that the AI infrastructure buildout has created insatiable demand for the exact products it manufactures. It is able in that its engineering teams qualified HBM3E at Nvidia when Samsung could not, its management had the foresight to exit consumer NAND and redirect capacity to enterprise, and its fab investment program positions the company for a decade of growth.

However, I distinguish this from the companies I consider "fortunate because they are able" — those where management creates the market through product innovation and category definition. Micron is supplying a market that Nvidia and the hyperscalers are creating. This is an important distinction. When the AI capital expenditure cycle eventually moderates — and it will, though I cannot predict when — Micron's growth rate will moderate with it. The company does not control its own demand destiny to the degree that, say, a Rubrik or an Axon does.

This does not make it a poor investment. It makes it a different kind of investment — one where the investor must be willing to hold through the inevitable downcycle, confident that the structural margin improvement and competitive positioning will produce returns over a full cycle that justify the price paid.

Verdict

Thesis: Intact. Conviction: Moderate-High.

Micron is executing exceptionally in an extraordinarily favourable environment. The transformation from commodity memory to AI-critical infrastructure supplier is genuine, supported by qualification moats, supply discipline, and a $200B multi-decade investment commitment. Management under Mehrotra is candid, conservative in guidance, and strategically decisive.

The concerns that prevent me from moving to Very High conviction are:

1. Cyclicality is not vanquished. Gross margins of -33% just three years ago remind us that this business can turn brutally. The current 68% guided margins are cyclically inflated.

2. HBM4 qualification ambiguity. Conflicting reports about Nvidia Vera Rubin create genuine uncertainty that only the Q2 FY26 earnings call or an explicit qualification announcement can resolve.

3. Key-man risk. Mehrotra is outstanding, but I lack confidence in the management bench beneath him.

4. Customer concentration. Four HBM buyers. Any share shift to SK Hynix in HBM4 would be high-impact.

For the upcoming Q2 FY26 report, I expect another substantial beat — revenue of $19-20B against $18.7B guidance, gross margins of 69-70%, and EPS well above the $8.42 guide. The critical disclosure will be HBM4 qualification status with Nvidia.

The stock, trading near $280 at approximately 7x annualised Q1 EPS, is not expensive for a company growing revenue at 57% with 47% operating margins. But the investor must accept that they are paying for a cycle — an extraordinary cycle, but a cycle nonetheless. If the next downturn brings margins to 35% (my estimate of structural floor, up from the historical 25%), the investment case remains intact. If it brings margins below 30%, the structural thesis is weaker than I believe.

I would hold, add on weakness, and watch the Q2 FY26 call with particular attention to HBM4 qualification and the Q3 FY26 margin guidance.

Sources:

• Seeking Alpha — Micron Q2 Earnings Preview
• TrendForce — Memory Price Outlook Q1 2026
• TrendForce — HBM4 Mass Production Delayed
• Enertuition — Micron Out of Initial Nvidia Rubin Builds
• TechStock01 — SK Hynix and Micron HBM4 Qualification
• Samsung Warns of Memory Shortages
• IDC — Global Memory Shortage Crisis
• Glassdoor — Micron Technology Reviews
• Tom's Hardware — Micron NY/Idaho Fab Updates
• Astute Group — HBM Market Share