Micron Stock Is Back Above $1,000. Is a Stock Split Coming?

Micron shares traded near $1,032, valuing the company at roughly $1.2 trillion, after AI-driven memory demand lifted fiscal Q3 revenue more than fourfold year over year to $41.5 billion and gross margin to 84.6% from 37.7%. Fiscal Q3 diluted EPS reached $24.67 versus $1.68 a year earlier, while fiscal Q4 guidance calls for $30.73 per share, implying about 25% sequential growth. Despite the four-figure stock price, management has not indicated a stock split ahead of Sept. 30 earnings, and the article argues a split would not alter Micron's valuation or fundamentals.

Analysis

The relevant signal is not a potential split but whether MU can sustain an AI-memory scarcity premium as customers qualify alternative HBM and high-capacity DRAM suppliers. At current profitability, even modest ASP normalization has outsized earnings sensitivity because the cost base is largely fixed: a 10-point gross-margin retracement would likely reduce annualized EPS materially and challenge the market's willingness to underwrite a low-single-digit forward multiple. The near-term setup is therefore asymmetric around fiscal 4Q guidance durability, not retail-accessibility optics.

Over the next 1-3 months, MU's report is a catalyst for the entire AI hardware complex, particularly NVDA and AVGO, because memory availability can become a system-level shipment constraint rather than merely a component-cost issue. Strong HBM bit-growth and firm contract-pricing commentary would support accelerator revenue conversion; any evidence that customers are double-ordering, building inventory, or seeing qualification progress at Samsung/SK Hynix would reverse that read quickly. A split announcement could create a brief retail-flow bid but should be treated as an opportunity to monetize volatility, not as a fundamental rerating catalyst.

The contrarian view is that the apparently depressed earnings multiple is not necessarily a floor: memory equities historically de-rate before earnings peak, and consensus may be extrapolating an unusually favorable mix and pricing environment too far forward. The thesis is falsified positively if MU delivers another guide-up while maintaining margins despite rising wafer starts; it is falsified negatively by sequential ASP deceleration, inventory-day expansion, or capex plans that imply material industry supply growth in calendar 2027.

Key Decisions for Investors

• Do not establish a directional position solely for a split; any post-announcement 3-5% retail-driven move without a concurrent estimate revision is a candidate to trim MU longs or fade tactically.
• For existing MU exposure, retain through the next earnings print only with a defined risk trigger: reduce if fiscal 1Q revenue guidance or gross-margin guidance implies sequential EPS below the current run-rate expectation. The central risk is multiple compression ahead of a memory-cycle peak, not an earnings miss in the reported quarter.
• Use a 1-3 month pair trade only if channel checks confirm continued HBM allocation tightness: long MU versus short SOXX or a diversified semiconductor basket. This isolates memory-specific pricing upside, while limiting broad AI multiple risk; exit if DRAM/NAND spot and contract-price indicators roll over for two consecutive weeks.
• Watch NVDA and AVGO customer commentary for memory-supply constraints. A confirmed constraint is supportive for accelerator backlog conversion and argues against shorting AI infrastructure; evidence of customer inventory digestion would instead favor reducing MU first, given its greater operating leverage to memory pricing.