Micron Technology, the largest U.S. maker of computer memory chips, delivered an extremely upbeat outlook for the current quarter—well above consensus estimates that analysts had been repeatedly raising in recent days. The company also unexpectedly increased its capital expenditure plan for fiscal year 2026.
Micron’s CEO indicated that this round of the memory-chip supercycle could extend into 2027, further signaling that under the unprecedented global push to deploy AI, surging demand for memory chips—along with a long-term supply shortfall—has enabled the company to charge higher prices for its DRAM and NAND products.
The strong results helped Micron’s shares jump more than 8% in U.S. after-hours trading. At a time when “AI bubble” narratives are spreading globally, Micron’s performance has helped sustain the AI bull-market storyline and investor conviction in AI.
Fueled by the powerful “memory supercycle” narrative, Micron’s share price has entered a rocket-like mode since the second half of this year, driving a year-to-date surge of more than 170%.
To concentrate memory capacity on supplying storage chips for the wave of newly built large-scale AI data centers, Micron even announced in early December that it would stop selling certain memory products to individual consumers in the PC/DIY market—underscoring that as global AI infrastructure construction accelerates, enterprise demand for high-performance, data-center-grade DRAM and NAND products continues to climb.
Wells Fargo said that as overall DRAM industry revenue— including HBM memory systems—is expected to grow more than 100% year over year in 2026, Micron will be one of the biggest beneficiaries. Wells Fargo reaffirmed a $300 price target for Micron and maintained an overweight rating.
Market interpretation:
Micron has strategically realigned its memory capacity toward AI. As AI infrastructure continues to drive explosive growth in memory demand—and demand rises for the key components—Micron is positioned to be one of the biggest winners capable of supplying those memory components.
