Micron Technology Has Fantastic News for This Artificial Intelligence (AI) Infrastructure Stock That Has More Than Doubled in 2026

Micron reported fiscal Q3 revenue up 346% YoY to $41.5B, with EPS up 1,215%, and said DRAM/NAND demand continues to significantly exceed supply. Micron plans capex of $27B in fiscal 2026 (nearly 2x fiscal 2025) and forecasts $10B in capex next quarter, implying fiscal 2027 spending will exceed $40B. Lam Research shares jumped over 7% after Micron’s results, helped by Lam’s exposure to memory-capital spending (39% of revenue from memory manufacturing equipment) and expectations that the memory supply shortage persists into 2030.

Analysis

The cleanest winner here is the equipment layer, not the memory makers. Capex is the fastest monetization path in a supply shortage because tool orders hit revenue before bit supply normalizes, so LRCX should keep converting industry anxiety into bookings while MU carries the capital-intensity burden and later faces the risk of its own supply response compressing pricing power. The second-order beneficiary set is broader than the article implies: any foundry-capex proxy tied to node transitions should see a halo, while downstream AI hardware and consumer OEMs face a temporary component-cost tax if memory stays tight.

Near term, the trade is about revisions, not the headline narrative. The next 1-3 months should be dominated by whether memory ASPs and LRCX bookings keep inflecting higher; if they do, the market will likely pay up for the visible order book, not just the theoretical 2030 shortage. Over 6-18 months, however, this is a classic cycle where aggressive capex seeds future oversupply—if the current spending wave comes in too strong, 2028/29 margins could normalize faster than sell-side models assume, which is the main falsifier for a long-only memory thesis.

The consensus may be underweighting the asymmetry between MU and LRCX. MU needs the shortage to persist; LRCX mainly needs customers to keep spending to relieve it, which is a lower bar and usually a better multiple story. The contrarian risk is that the market is extrapolating a multi-year scarcity regime from a few quarters of panic buying; if DRAM/NAND prices plateau or AI server demand pauses, MU can de-rate even while earnings look good, because the capex step-up starts to look like a peak-margin signal rather than a durable moat.

Key Decisions for Investors

• Long LRCX on any 2-4% pullback over the next 1-3 weeks; use a 3-6 month horizon and target a 15-20% upside move if bookings and capex commentary keep accelerating. Falsify if memory capex guidance rolls over or LRCX order growth decelerates two quarters in a row.
• Pair trade: long LRCX / short MU to isolate the capex beneficiary from the commodity exposure. Best entry is after MU-driven sympathy rallies in the next 1-2 weeks; thesis works if capex keeps rising while memory ASP upside starts getting fully reflected in MU's multiple.
• If you want convexity, buy LRCX call spreads 4-6 months out rather than outright MU calls; this expresses the lower-risk view that tool demand leads the cycle by several quarters. Risk/reward is better if the market keeps rewarding backlog visibility.
• Use SOXX/SMH as a tactical hedge around the long LRCX exposure: if memory news triggers a broad semi pullback, the hedge should cushion beta while preserving the idiosyncratic capex story. Exit if SOXX underperforms LRCX by more than ~5% after the next earnings cycle, which would signal the thesis is already priced.