Micron At $1,750 Is Now My Base Case

Micron reported a record Q3 FY26 with revenue of $41.5B (+346% YoY) and non-GAAP gross margin up to 84.9%. AI data center demand and long-term take-or-pay contracts now drive ~50% of revenue, reducing cyclicality, while management guided to $50B revenue next quarter. Micron also forecasts $30B free cash flow for Q4 FY26, fully funding aggressive CapEx from operating cash generation—supporting a strongly positive outlook that should move the stock.

Analysis

The market should treat this less as a one-quarter earnings beat and more as a change in the asset class. Contracted revenue plus self-funded capex reduces the usual memory-cycle discount, which means MU can migrate from being valued on peak/trough earnings to something closer to a scarce AI infrastructure supplier. That rerating is the real equity story; if investors accept it, the multiple expansion can matter more than another increment of earnings revision.

The second-order winners are the memory complex and the upstream tool chain. If buyers are willing to pre-commit, it validates tighter supply discipline across HBM and advanced DRAM, which should support pricing power for the highest-spec suppliers and keep utilization high for equipment vendors like AMAT, KLAC, and LRCX. The losers are AI platform buyers and GPU vendors only at the margin: memory cost per rack rises, but near term that is a budget issue, not a demand killer, because AI buildout remains compute-constrained.

The main risk is extrapolation. If competitors add capacity too quickly, or if hyperscaler demand pauses, the current margin profile can normalize faster than the market expects; the contract structure reduces volatility, but it does not eliminate renewal and pricing risk. The thesis weakens if forward gross margin stops holding in the 70s/80s range or if free cash flow falls short of capex needs on the next guide, because that would suggest scarcity economics rather than durable contracted returns.

Over 1-3 months, the stock can continue to re-rate if the market starts comparing MU to quasi-utility AI infrastructure instead of cyclical semis. Over 6-18 months, the key test is whether contract coverage can persist without triggering industry overbuild; that is where the bear case lives. For now, the move looks directionally right but likely invites volatility if investors chase it too aggressively after the print.

Key Decisions for Investors

• Buy MU on a post-event consolidation rather than into strength; use a 3-6 month call spread to express the rerating thesis with defined downside if AI memory demand cools.
• Set up a relative-value long MU / short SOXX position for the next 1-3 months to isolate the contracted-cash-flow premium from broad semiconductor beta; cover if MU underperforms SOXX by ~8-10% or if memory pricing data rolls over.
• Add AMAT, KLAC, or LRCX on pullbacks as a second-order beneficiary trade over the next 2-4 quarters; the risk is that tool demand lags if MU’s capex cadence flattens sooner than expected.
• Watch for a take-profit signal if MU’s next guide implies margin normalization or weaker cash conversion; that would be the first evidence that the market is pricing scarcity too aggressively.