Micron Stock Has Soared Nearly 700% in a Year. Is a Stock Split Imminent?

Micron shares are up 698% over the past year to nearly $1,000 as the memory-chip supply shortage is expected to last beyond 2027. The article cites Q3 FY2026 EPS of $25.11 (+1,200% y/y) and current-quarter EPS guidance of $31.00 (implying ~10x y/y growth), arguing the valuation is compelling at 23x earnings vs 35x for the Nasdaq-100 (forward: 7x vs 27x). It frames a potential stock split as a possible retail-demand tailwind, but also notes fractional-share buying and concludes the stock is a “buy” regardless.

Analysis

The actionable signal is not the stock-split chatter; it is that memory remains one of the few semis where pricing power can still outrun capacity additions. That makes MU the cleanest expression of an AI infrastructure bottleneck, but it also means the market will keep re-rating the stock on each incremental indication that supply is tighter for longer. The first-order winner is MU; the second-order winners are memory equipment and test/packaging names if capex extends, while the potential losers are AI system integrators and accelerator buyers if HBM constraints keep pushing BOM costs higher.

The key risk is timing: a split could support sentiment for days to weeks, but it does nothing to protect margins if Samsung/SK Hynix or China capacity normalizes faster than expected. Over 1-3 months, the catalysts are memory pricing trends, HBM allocation commentary, and any change in gross margin guidance; over 6-18 months, this is a classic cycle where peak earnings can coexist with peak multiples and then de-rate quickly once supply catches up. NVDA is the cleanest indirect watch item: if memory remains the gating factor for AI server shipments, its unit growth can be constrained even if end-demand is healthy.

Consensus is likely underweighting how much of MU’s valuation is now a duration bet on scarcity, not just earnings growth. If the supply shortage truly lasts beyond 2027, the stock can stay expensive longer than skeptics expect; if not, the current multiple on forward earnings is vulnerable to compression even without a demand shock. The market is probably overstating the importance of a split and underestimating the risk that the cycle’s tail gets shorter than management implies.

Key Decisions for Investors

• Buy MU on a 5-8% pullback or after a post-rumor fade; target 3-6 months; use a tighter stop if DRAM/HBM pricing softens or gross margin guidance inflects lower.
• Pair trade: long MU / short SOXX for 1-3 months to isolate memory pricing power versus broader semis; thesis breaks if semiconductor breadth expands on a sustained capex-upcycle.
• If you want convexity, use MU 6-12 month call spreads instead of outright equity; the payoff is best if pricing remains tight into the next two earnings prints.
• Watch NVDA as a relative short-term hedge only if AI shipment growth slows on memory bottlenecks; otherwise avoid forcing a short until there is evidence of HBM constraint bleeding into guidance.
• Set an alert on any commentary from Samsung/SK Hynix capex or inventory buildup; that is the earliest sign the cycle is rolling over and the trade should be reduced.