Micron Guided a Single Quarter to $50 Billion. The Stock Sits 24% Below Its High.

Micron guided fiscal Q4 revenue to $50.0B (±$1.0B) with ~86% gross margin and non-GAAP EPS of $31 (±$1), implying roughly $36B of adjusted profit potential for the quarter. FY2026 revenue could reach about $129B (~3.5x FY2025) if the quarter hits the midpoint, but the article notes the stock is ~24% below its 52-week high (~$950 vs $1,255), reflecting investor skepticism about sustainability of cycle/AI-driven pricing. The demand mix is heavily skewed to AI data centers (cloud memory $13.8B; data center $11.5B), while take-or-pay strategic customer agreements are cited to support durability.

Analysis

The market is still pricing MU like this is a one-quarter peak, but the key variable is whether management can keep the first fiscal 2027 guide from rolling over on Sept. 30. The longer quarter flatters the headline step-up, so the equity should not be valued on the reported run-rate alone; the real tell is whether weekly revenue and gross margin hold once the calendar benefit disappears. If that cadence persists, the multiple can expand even with memory-cycle scar tissue.

Second-order winners are the hyperscale buyers that locked in supply, but their economics get less attractive if DRAM/HBM stays tight; BOM inflation usually shows up first in server OEM margins and procurement delays, not in NVDA’s gross margin. That makes MU the cleaner expression than the broad semiconductor basket, while more commodity-heavy memory exposure should lag if the mix is shifting toward AI-differentiated supply rather than a generic upcycle.

The contrarian risk is that take-or-pay contracts only smooth the downturn until the next renegotiation window; if AI capex pauses, the earnings air pocket arrives with a lag, which is why this can still be a better short 6-12 months out than today. What would falsify the bull case is a September guide implying margin compression back into the low-80s, or any sign that hyperscale demand is being met with inventory rather than true consumption. Near term, the pullback may already reflect some of that skepticism.

Key Decisions for Investors

• Long MU / short WDC for 1-3 months into the Sept. 30 guide: express continued AI-memory pricing strength versus a more commodity-memory balance sheet and mix. Best if MU holds above the post-earnings gap and WDC lags on weaker mix; exit if MU guides below current run-rate or gross margin slips meaningfully.
• Buy a 3-6 month MU call spread 5-10% OTM to capture upside into the first fiscal 2027 guide with defined risk. This is a cleaner trade than outright stock if you want exposure to a further re-rating without assuming the cycle is permanent.
• Do not short NVDA on this print alone; if anything, use MU strength as a hedge against overexposure to the AI basket rather than a thesis on GPU demand. The better short expression, if you need one, is broader semiconductor beta via SMH after a rally, not NVDA idiosyncratically.
• Set a post-Sept. 30 alert: if MU’s forward gross margin guide falls back below the low-80s or management tones down 2027 demand, reduce exposure aggressively. That would be the first credible sign the market’s peak-cycle skepticism is correct.