Trump Says Micron Is One of the 'Hottest' Stocks. Does That Make MU a Buy Here?

Micron reported fiscal Q2 revenue of $23.86B (vs $8.05B year-ago) and expects gross margins to exceed 80% in fiscal Q3, yet MU shares are down ~27% from 52-week highs. Management raised fiscal 2026 capex to >$25B (from $20B), citing a new Taiwan fab and U.S. expansion, prompting investor concern about future returns. Company says DRAM and NAND supply is far short of demand and Gen6 SSD demand exceeds supply, while reports of potential Google memory-compression and a ~280% YTD run-up raise competitive and valuation risks.

Analysis

Micron sits at an inflection where manufacturing cadence, yield ramps and customer procurement cycles matter more than quarterly prints; margins and cash conversion are highly elastic to utilization over 12–36 months, so any change in effective fab throughput will move free cash flow multiples materially. That elasticity means equity moves will be driven more by forward signals (yield updates, customer purchase patterns, equipment delivery schedules) than by current bookings, creating rapid re-rating opportunities when visibility improves.

The firm's recent industrial choices create concentrated exposure to two vectors: execution risk on new capacity and geopolitical/regulatory timing for cross-border fabs. Those vectors generate asymmetric outcomes — on-time yields compress payback to 3–5 years and justify rich multiples, while multi-quarter delays or tariff/subsidy reversals can take IRR to single digits and force asset-level write-downs.

The competitive “compression” narrative from large AI software vendors is a credible long-term moderation risk but is a slow, adoption-limited process; even aggressive per-model memory reductions are overwhelmed if model count and parameter budgets continue growing at 50–100%+ year-over-year across hyperscalers. Practically, this makes meaningful downside a multi-year story unless a reproducible, hardware-agnostic compression method is validated at scale within 6–12 months.

Tactically, monitor three binary catalysts: independent third-party memory-per-inference benchmarks, quarterly fab yield commentary, and any public shifts in national subsidy or export-control timelines. These are the fastest paths to re-rating; absent them, the path to upside is gradual and asymmetric to the downside on execution slips, so position sizing and tail protection should be prioritized.

Key Decisions for Investors

• Core long (MU equity + tail protection): Establish a 1–2% portfolio long in MU for a 12–36 month time horizon. Size 60% as outright shares and buy 9–12 month puts ~15–20% OTM to cap downside to ~15% (cost funded by selling 1–2 near-term covered calls 20–30% OTM). R/R: asymmetric — steady FCF re-rate can deliver 40–80% upside, protected to low-double-digit downside.
• Leverage the AI stack (paired options): Express convex exposure to end-demand by buying a 12–18 month MU call-forward (e.g., buy LEAPS call spread 1:1) financed by a small NVDA vertical put spread instead of cash, horizon 12–18 months. R/R: if hyperscaler GPU rollouts accelerate, MU re-rates and NVDA support increases call value; limited premium outlay with 2–3x upside potential vs defined downside from the put spread.
• Event hedge against compression (GOOG tail): Buy 6–12 month GOOG puts ~20% OTM (small position, <0.5% portfolio) to hedge the low-probability, high-impact scenario where a broadly adopted memory-compression breakthrough materially reduces server memory demand within a year. R/R: small insurance cost protects core MU exposure from a fundamental demand structural break.
• Tactical volatility play around yield updates: If independent yield commentary or equipment delivery guidance is due in the next 30–90 days, buy a strangle (calls + puts) with expiries bracketing the update sized to 0.5% portfolio notional. R/R: captures a sharp directional move from new clarity; exit once implied volatility re-prices post-announcement.