Micron Technology Inc. (NASDAQ: MU) surged more than 12% on Tuesday, leading a memory-led rally on Wall Street as Bank of America reiterated a Buy rating with a price target of $1,550. Analyst Vivek Arya argued that the proliferation of open-source, lower-cost Chinese AI models will fuel greater memory demand rather than curb it, countering concerns that these models threaten hardware sales. The report highlights that while pricing for access may be low, the hardware requirements for running massive models remain substantial, driving enterprise investment in memory infrastructure.

The Core Argument

Arya posits that while closed models like OpenAI's GPT-5.6 consolidate memory demand in centralized data centers, open models multiply it. Every enterprise or government downloading open weights creates a new customer-side memory socket. The analysis follows the release of Kimi K3 by Moonshot AI, a 2.8-trillion-parameter system described as the largest open-weight model ever built. Despite Chinese labs charging significantly less for access—Kimi K3 runs at $3 per million input tokens compared to $15 for Anthropic's Claude Opus 4.8—BofA emphasizes that pricing is a business-model choice rather than a reflection of hardware costs. Running Kimi K3 still requires approximately 1.4 terabytes of high-bandwidth memory (HBM) across 64 or more accelerators per serving instance.

Competitive Landscape

The report addresses the risk of Chinese competition, specifically from ChangXin Memory Technologies (CXMT). BofA notes that CXMT, China's fastest-growing DRAM maker, competes primarily in commodity chips rather than the HBM3E and HBM4 components essential for advanced AI systems. Additionally, it remains uncertain whether US equipment makers will receive approval to sell to CXMT in the near term. This suggests that Micron remains insulated from direct competition in the highest-margin segments of the AI memory market.

Financial Outlook and Analyst Ratings

Beyond AI demand, BofA identifies the potential expiration of Micron's share repurchase restrictions under the CHIPS Act around December 2026 as a key catalyst. Arya estimates the company could generate $120 billion to $130 billion in annual free cash flow over the next several years. A hypothetical 40% payout policy could translate to $50 billion to $60 billion in annual share repurchases, representing roughly 5% to 6% of projected market capitalization.

The consensus rating on Micron remains a Buy, with an average price target of $1,316.79. Individual targets range from a low of $385 to a Street-high of $2,000.

UBS’s June warning to reduce risk in AI-linked semiconductor stocks has proven directionally correct regarding near-term volatility, though it was early on absolute returns. The trading desk identified stretched positioning and a binary winners-versus-losers framework in a market treating the AI super-cycle as a linear progression. While the call did not mark a regime change, it successfully flagged the risk of fat left tails for investors running tight limits or trading on margin.

Directionally Right on Near-Term Risk

On the risk front, UBS accurately assessed the setup. AI chip leaders, including SanDisk Corp (NASDAQ: SNDK), Micron Technology, Inc. (NASDAQ: MU) and Advanced Micro Devices, Inc. (NASDAQ: AMD), did not crash but began trading like a crowded momentum basket rather than a one-way factor. The cohort experienced meaningful drawdowns from late-June peaks, with the most significant air pockets visible in the most parabolic names. The warning captured the reality that incremental upside would come with higher risk than the prior six months implied.

Early on Absolute Returns

Where UBS’s mid-June call appeared early was in absolute price terms. An investor acting on the June 18 note immediately would have missed further gains as the group pushed higher into late June. This was particularly evident in Micron, which continued to squeeze as the "winner" label hardened. The super-cycle narrative had one more leg before gravity took effect, illustrating the frustration of timing risk notes in secular stories.

Trade Repriced, Leaders Did Not Rotate

UBS’s most significant oversight was regarding positioning and leadership. The desk framed the AI complex as an extreme winners-versus-losers regime, a dynamic that remains unchanged. Following the pullback, the same small set of semiconductor and memory names continue to dominate performance tables, index impact, and options activity. The trade has repriced but not rotated; flows stepped back and volatility rose, yet capital did not decisively migrate toward laggards or value semis. The note served as a useful risk flag, but the AI chip trade is now less complacent rather than less concentrated.