← Feed Deep Dive Matrix Subscribe

Micron Technology (MU) is Printing Cash. Can its Memory Momentum Last? - Yahoo Finance

finance.yahoo.com 2026-09-28 Yahoo Finance
Entities
Tags
Micron TechnologyMemory ChipsHBM4DRAMNANDFree Cash FlowShare BuybackAI ComputingWafer FabricationAdvanced PackagingCapital ExpenditureMemory CyclicalityStrategic Customer AgreementsWolfe ResearchSK Hynix
News Summary
Wolfe Research's reaffirmed Outperform rating and $1,500 price target on Micron underscores a structural shift in the memory industry's financial architecture. Rather than a cyclical upswing, the anal... Read original →
Industry Analysis
The memory industry is undergoing a structural break, not a cyclical turn. Micron's $22B in customer prepayments and take-or-pay agreements reclassify DRAM from commodity to compute-infrastructure input—more aggressive than Samsung's 2018 LTA with Apple, because the demand anchor is AI inference throughput, not handset shipments. On the technology chain, HBM4 stacking beyond 16 layers shifts the binding constraint from wafer output to advanced-packaging capacity. This elevates bargaining power of packaging houses in Taiwan, China, and compresses Micron's packaging-side margin even as wafer-side premium holds. The 1-beta EUV full-mask migration adds 15-20% per-wafer cost, but contractual price floors neutralize profit erosion from slowing bit growth. The real vulnerability is capital structure, not technology. At $7.1B quarterly capex plus HBM depreciation, the fixed-cost base sits roughly 2.3× the 2019 level. A six-month demand air pocket in 2027 would trigger a cascade: the $22B deposit default clauses amplify the shock far beyond a normal price correction. SK Hynix will almost certainly weaponize HBM4 timing to anchor pricing, while Samsung floods conventional DRAM with 1-gamma capacity. Micron's survival threshold: contractual coverage must exceed 60% of revenue by Q2 2027, or the de-cyclicalization narrative inverts into a liquidity trap.
Read Original Article →
Related
This page displays AI-generated summaries and metadata for research purposes. Original content belongs to the respective publishers.