European chip stocks jump after another Micron blockbuster quarter

Micron delivered a blowout fiscal Q3 with revenue of $41.46B, more than 4x year-ago levels and well above the $35.84B consensus, while adjusted EPS of $25.11 also beat the $20.78 estimate. The company guided Q4 revenue to around $50B versus $43.58B expected and said data center revenue surged more than sevenfold to $11.5B, underscoring AI-driven demand. European semiconductor stocks rallied 2.3% to 6.8% in response, with Micron shares up more than 18% in after-hours trading.

Analysis

This is less a one-day sympathy rally than a repricing of the entire AI memory demand curve. The key second-order effect is not just better pricing for memory vendors, but higher confidence that hyperscaler capex is shifting from compute-only to a more balanced compute-plus-memory stack, which should keep DRAM/HBM pricing tight for several quarters. That favors the highest-quality capacity holders and exposes weaker peers that cannot ramp advanced nodes fast enough.

The market is also underappreciating how strong guidance changes bargaining power across the supply chain. If long-duration supply agreements are being signed with meaningful committed spend, memory suppliers gain partial visibility on utilization and pricing into next year, which should support margin durability and reduce the probability of the usual late-cycle inventory air pocket. The beneficiary set extends beyond the memory name itself into tools and lithography, but the magnitude will be uneven: equipment suppliers with real exposure to advanced packaging and memory intensity should see estimate revisions, while commodity-sensitive analog/industrial names benefit far less.

The contrarian risk is that this becomes a crowded "AI infrastructure everywhere" trade and the market extrapolates peak margins too early. Memory is historically the most cyclical part of semis, and any moderation in cloud capex, a re-acceleration in supply, or a customer mix shift away from premium HBM could compress the forward multiple quickly. On a 1-3 month horizon the trend should persist on estimate revisions; over 6-12 months, the trade becomes much more sensitive to capex digestion and whether vendors overbuild in response to current strength.

Key Decisions for Investors

• Stay long MU into the next 1-2 earnings cycles; risk/reward remains favorable while consensus is still likely too low on FY guidance, but trim into any further 15-20% squeeze as the move may price in near-term perfection.
• Add tactically to ASML on 3-6 month horizon only if memory spending translates into order commentary; use it as a lower-beta AI capex proxy, but expect less upside than MU because the market already views it as a quality compounder.
• Pair long MU / short STM or a broader weaker-analog basket for 1-3 months: memory upside is driven by AI infrastructure scarcity, while STM is more exposed to auto/industrial digestion and should lag if the market rotates toward the highest incremental earnings revisions.
• Buy short-dated call spreads on a memory/equipment basket rather than outright long stock if you want to express momentum without paying full multiple expansion; the asymmetry is best over the next 4-8 weeks while revisions are still being digested.
• Avoid chasing SMCI/APP on this print alone; the article supports the AI infrastructure theme, but these names are more vulnerable to multiple compression if investors start demanding evidence of sustained end-market demand rather than headline growth.