Will Strong Demand and Pricing Help MU Reach 86% Gross Margin in Q4?

Micron is targeting ~86% non-GAAP gross margin in Q4 FY2026, with Q3 already at a record 84.9% (vs 74.9% prior quarter and 39% a year earlier), driven mainly by higher DRAM/NAND pricing and favorable mix. Q3 revenue hit a record $41.46B as DRAM rose 67% sequentially to $31.3B (ASP up low-60s%) and NAND jumped 99% to $9.9B (ASP up mid-80s%). Segment margins also expanded sharply (e.g., core Data Center to 87% from 74%), and with AI demand keeping memory supply tight, the article argues Micron can meet or exceed the 86% goal.

Analysis

This is less a one-quarter beat than evidence that memory has become a scarce AI input, not a commodity. That matters because the first-order winner is MU, but the second-order winner is any supplier with the cleanest path to HBM and premium DRAM mix; the loser is whoever is forced to compete on capacity rather than specialization. In the next 1-3 months, the real catalyst is not the reported margin number itself but whether management can keep pushing pricing without tripping customer resistance or inventory normalization.

Relative to the group, MU looks higher quality than NAND-heavy exposure because the market is paying for scarcity rent, and DRAM/HBM should keep its pricing power longer than NAND if supply additions accelerate unevenly. SNDK can still work near-term, but its earnings power is more brittle if NAND bit supply catches up first. Hyperscalers such as AMZN and GOOGL are likely to absorb the cost in capex budgets, but sustained memory inflation can slow the pace of incremental AI buildouts at the margin, which is a second-order negative for the broader AI supply chain.

The contrarian risk is that the market is extrapolating a cycle peak as if it were structural; memory upcycles often look most durable just before ASPs flatten. The key falsifier is any sign that MU’s margin guide stops moving up or that ASP growth decelerates faster than shipments, which would force multiple compression quickly given how much optimism is already embedded in the stock. Over 6-18 months, watch capex announcements from memory producers and customer inventory commentary; those will tell you whether this is a sustained re-rating or just a very profitable phase of the cycle.

Key Decisions for Investors

• Long MU on pullbacks into the next earnings window; use a 6-12 month call spread if implied vol stays contained, because the risk/reward favors upside while estimate revisions remain positive.
• Pair trade: long MU / short SNDK for 1-3 months as a relative-value expression of DRAM/HBM durability versus NAND sensitivity; thesis breaks if NAND pricing outperforms DRAM on the next read-through.
• Set an alert to trim MU if management implies gross margin cannot hold near the current target or if sequential ASP growth slows materially; that would be the earliest sign the cycle is peaking.
• Avoid chasing NVDA as a direct beneficiary here; if memory stays tight, GPU economics face a modest input-cost headwind, but not enough evidence yet for a clean short.