Micron Is 1.6 Points From Taking the No. 2 Spot in DRAM

Micron's Q2 2026 DRAM revenue surged 66% sequentially to $36.0 billion, raising its global market share to 23.3% and narrowing its gap with No. 2 SK Hynix to 1.6 percentage points from 7.5 points in Q3 2025. The gain was driven primarily by sharply higher conventional DRAM prices amid AI-led memory shortages, while Micron's HBM share fell to 18% from 21%. Micron guided for roughly 86% gross margin in fiscal Q4, but conventional DRAM price increases are expected to slow to 13%-18% in Q3, leaving sustained upside dependent on HBM4 execution, where it remains third behind SK Hynix and Samsung.

Analysis

MU’s apparent share momentum is a price/mix phenomenon rather than evidence of durable technology leadership. That distinction matters because conventional DRAM price spikes create extraordinary operating leverage for MU’s largely fixed-cost fab base, but they also tend to mean-revert once customer inventory rebuilding slows and incremental supply is released. The relevant near-term question is not whether MU overtakes SKHY in an aggregate ranking, but whether contract-price deceleration is slower than the market currently discounts in MU’s forward earnings multiple.

Over the next 1-3 months, MU should retain stronger earnings revision momentum if server DRAM contract pricing remains positive and utilization stays constrained. However, HBM4 qualification and yield data are the more consequential 6-18 month determinant of relative valuation: SKHY and SSNLF have greater exposure to the AI-memory profit pool, while MU needs sustained design-win conversion to prevent its conventional-memory windfall from being valued as a classic cyclical peak. NVDA is a secondary beneficiary if additional qualified HBM4 capacity reduces memory bottlenecks, though greater supplier competition could shift some economics back to GPU customers.

The contrarian setup is that a low earnings multiple does not necessarily imply MU is cheap; it may reflect market skepticism that peak margins and pricing are repeatable. Conversely, if HBM4 shipments, yields, and customer qualifications improve concurrently with conventional DRAM remaining tight, MU could experience a material multiple re-rating because investors would need to underwrite a less cyclical earnings base. Before acting on reported market-share shifts, reconcile third-party revenue definitions with company disclosures, particularly whether contract-price inflation, product mix, and recognized revenue periods are comparable.

Key Decisions for Investors

• Tactically long MU versus short SKHY through MU’s next earnings and DRAM contract-price update, sized as a 1-3 month relative-value trade. Thesis requires continued positive conventional DRAM pricing and MU maintaining gross-margin guidance; exit if quarterly contract-price commentary turns flat/down or MU cuts margin guidance. Target 10-15% relative outperformance; risk 6-8% if HBM4 qualification headlines favor SKHY.
• Add MU only on post-results confirmation of HBM4 volume shipments, customer qualification breadth, and stable-to-upward fiscal margin guidance; avoid chasing a share-ranking headline alone. A credible HBM mix inflection could justify multiple expansion beyond a pure memory-cycle valuation over 6-12 months.
• Maintain SSNLF as the preferred hedge against an HBM4-driven reversal in the MU/SKHY relative trade. Samsung’s incremental HBM capacity and improving yields could pressure HBM pricing but also make aggregate DRAM share less informative; reassess if Samsung signals aggressive conventional DRAM supply additions.
• Watch NVDA supplier commentary and lead-time indicators as a read-through rather than a direct trade. Faster HBM4 availability is bullish for AI-system shipment volumes, but a sharp decline in HBM pricing without offsetting GPU unit growth would signal supplier margin competition and weaken the long-MU thesis.