TLDR
- MU shares have declined approximately 20% following Q2 results released March 18, fueled by concerns that Google’s TurboQuant technology could diminish memory chip demand
- Vijay Rakesh from Mizuho reaffirmed his Outperform stance with a $530 price objective, viewing the decline as an attractive entry point
- The company’s DRAM average selling prices jumped mid-60% in Q2, while NAND ASPs climbed high-70%, demonstrating robust pricing leverage
- Wall Street remains divided: some analysts identify irrational market behavior, while others highlight concentration risks and questions about sustained pricing strength
- MU shares have surged 324% in the past year, surpassing performance from Nvidia, AMD, TSMC, and Broadcom
The past several weeks have been turbulent for Micron. Following an exceptional rally in the semiconductor space — gaining 324% over twelve months — this memory chip manufacturer encountered significant resistance. The trigger came from Google’s TurboQuant algorithm, a lossless compression innovation that triggered investor anxiety about potential decreases in future DRAM and NAND requirements. Markets responded swiftly.
Following the company’s second-quarter financial release on March 18, shares have retreated approximately 20%. This represents a substantial correction for an organization that recently commanded attention within AI-focused investment circles.
The downturn stems from a basic concern: if Google’s TurboQuant delivers more efficient data compression while maintaining model precision, cloud providers might require less physical memory for AI operations. Reduced DRAM and NAND demand translates to diminished pricing leverage for Micron. However, multiple analysts are questioning this narrative.
Vijay Rakesh from Mizuho responded forcefully. He preserved Outperform recommendations for both Micron and Sandisk (SNDK), establishing price objectives at $530 and $710 respectively. Rakesh referenced the Jevons paradox — an economic principle suggesting efficiency gains frequently drive increased consumption rather than reduction. His illustration: when DeepSeek emerged in 2025 and disrupted GPU equities, AI infrastructure investment actually intensified subsequently.
Rakesh additionally highlighted Google’s TurboQuant research itself as a potential catalyst for expanded models and accelerated inference, which would continue demanding significant memory resources. He characterizes the present selloff as excessive market reaction.
Examining the Financial Reality
Micron’s second-quarter performance painted an encouraging picture. DRAM bit volumes increased mid-single digits sequentially, while ASPs expanded in the mid-60% range. NAND bit volumes advanced low-single digits, with ASPs rising in the high-70% range. This represents substantial pricing premium, propelled by constrained supply rather than volume expansion.
Seeking Alpha analyst Oliver Rodzianko identified this pattern. He noted Micron currently faces greater supply constraints than demand limitations, and that DRAM and NAND supply-demand dynamics should remain tight past 2026 based on management guidance. His apprehension centers not on technology capabilities — but rather on the sustainability of price-driven versus structurally enduring earnings strength.
Should pricing revert toward historical norms, profit margins could contract. Rodzianko also emphasized concentration vulnerability: Micron maintains heavy exposure to hyperscaler capital expenditure, and any deceleration in that investment cycle would impact shares quickly and severely.
Optimistic Analysts Emphasize AI Infrastructure Growth
Analyst Dmytro Lebid adopted a more constructive perspective. He characterized the decline as resulting from “irrational investor behavior” and suggested markets are overweighting deceleration probabilities. From his assessment, hyperscaler demand for HBM3E memory remains intact, and Micron’s supply-limited position sustains healthy margins.
He contended that demand from Nvidia alone should continue expanding, establishing a sustainable foundation beneath Micron’s pricing structure.
The company is also expanding production at locations in Idaho, Tongluo, and Singapore through 2027–2028 — representing a long-horizon commitment that AI-driven memory consumption will maintain upward trajectory.
As of early April 2026, MU shares traded near $366, reflecting a market capitalization approaching $413 billion with a 52-week trading range spanning $61.54 to $471.34.
