Micron Trades at 6 Times Next Year's Earnings. The $38 Billion of Capacity That Ends This Cycle Doesn't Open Until 2028.

Micron fiscal Q3 revenue more than quadrupled to $41.5B, with gross margin rising to 84.6% (from 74.4% in fiscal Q2) and operating cash flow up more than 5x to $25.4B. Management guided fiscal Q4 revenue to ~$50B with gross margin near 86% and EPS around $30.73, implying the stock trades near ~7x expected earnings. Supply timing is pushed out as SK Hynix approved ~$38B for two DRAM/NAND fabs with first clean rooms opening in Dec 2028 and Jun 2029, while Micron has 16 take-or-pay customer contracts through end-2030 covering ~20% of DRAM and ~1/3 of NAND volume—supporting durability of tight memory conditions beyond 2027.

Analysis

The market is still pricing MU like a cyclical that mean-reverts before the next wave of capacity arrives; the better read is that the cycle is being lengthened by customer contracting and long-dated fab timelines. That shifts the valuation framework from spot-ASP trading to forward gross margin durability, which should support a higher multiple than the current low-single-digit forward EV/EBIT. The biggest second-order winner is SKHYV, because delayed incremental supply in DRAM/NAND keeps both pricing power and utilization high longer than sell-side models likely reflect.

The losers are the memory-intensive end markets, not just the obvious device OEMs. If DRAM/NAND stay tight through 2027, server builders, handset assemblers, and storage hardware names face a margin tax that can show up with a lag in 1-3 quarter guidance cuts; that argues for caution on names with weak pricing pass-through rather than blanket bullishness on semis. NVDA is only a marginal beneficiary here through data-center demand continuity, but the bigger effect is that elevated component costs can slow AI server gross margin expansion for the ecosystem even while unit demand remains strong.

The main falsifier is not a new fab opening in 2028/29; it is an earlier demand inflection. If hyperscaler capex rolls over, DRAM spot pricing will compress first, then contracted pricing will matter less as renewals reset off weaker benchmarks; that is the 3-6 month risk window. Over 12-18 months, the structural bull case remains intact unless AI server order rates and cloud memory content per rack flatten materially.

Key Decisions for Investors

• Long MU on pullbacks over the next 2-4 weeks; use the current setup to add only if the stock consolidates rather than chases a vertical move. Risk/reward is favorable if the market re-rates it toward higher-quality semi multiples, but cut if management or peers imply a 2027 demand air pocket.
• Pair trade: long MU / short SOXX or SMH for 1-3 months. The thesis is that MU-specific contractual durability and supply timing can outperform the basket, while the ETF still carries weaker cyclicals and more beta to any semicap/autos slowdown.
• Long SKHYV versus short a weaker DRAM/NAND consolidator or legacy storage supplier if borrow/liquidity allows; otherwise express as MU over WDC/STX on any rally. The edge is that the market underprices how much delayed South Korean capacity extends tightness into 2028-29.
• Watch list, not a trade yet: short DELL/HPE/SMCI only if they start guiding to lower server gross margin from memory cost inflation. This becomes actionable only when customers cannot fully pass through bill-of-materials pressure.