"사지 않으면 다른 이들이 사들일 것"…애널리스트, 반도체주 200% 상승 가능성 언급

Investing.com -- DA Davidson analyst Gil Luria raised its price target on Micron stock to $3,000 from $2,100, the highest on Wall Street, arguing that the broader market has fundamentally underestimated Micron's multi-year growth runway as artificial intelligence reshapes memory demand. Micron stock is up roughly 1,500% since April last year. Before you buy NVIDIA: our AI model's stock picks are beating the market by 120%+. See the list now » The call, from DA Davidson analyst Gil Luria, rests on a 19x market multiple applied to the firm's FY27 EPS estimate, according to DA Davidson's research note. Luria's central thesis is straightforward: AI infrastructure requires vastly more memory than prior technology cycles, supply cannot keep pace, and investors have not yet priced in what that imbalance means for Micron over the next three to five years. "We concluded that investors are early in their journey of understanding MU's value and believe that journey will lead them to assigning a far higher multiple," Luria wrote, following investor meetings with Micron's senior IR leadership, including Samir Patodia and Jeff Grattan. The demand logic Luria presents is grounded in how AI models actually consume memory. "Memory is a lever for better AI performance. AI models generate better results with more memory, run faster with more memory, and have longer context windows with more memory," he wrote. DA Davidson sees this as a structural shift rather than a cyclical uptick, with the firm asserting that demand will outstrip supply in both 2027 and 2028, a view it says Micron itself validated during its most recent earnings report, where the company guided to both volume gains and price gains through 2028. One of the more counterintuitive arguments in the note concerns the trend of memory "de-specing" by customers such as Nvidia, which some investors have read as a bearish signal. Luria rejects that interpretation. Not finding the right stocks? Our AI's list of investor-grade stock picks for October just dropped » Using an automotive analogy, he wrote: "If Tesla thought they would sell 1 million cars with 100KW battery at $50,000 and ordered 100MW of battery only to find out they could sell 2 million cars because demand was surging, they would sell 2 million 50KW battery cars at $40,000 and make more money." In Luria's framing, customers reducing per-unit memory specs to move more product actually creates pent-up demand, because degraded performance pushes them to buy more memory in future product cycles. He argues that targeting long-term supply contract agreements (SCAs) at roughly 50% of capacity is the right strategic balance, comparing it to airline yield management. "This removes the downside while keeping half the upside," he wrote. Crucially, Luria identifies Micron's core SCA counterparties — Amazon, Microsoft, Google, Nvidia, and Apple — as among the least credit-risky buyers in any industry, noting these are "not companies that tend to breach contracts." He further argues that as competitors in the memory industry also shift toward long-term contracting, the structural volatility of the traditional memory cycle is being reduced across the board. On China, Luria is direct in dismissing the risk of domestic Chinese memory producers displacing Micron globally. "China has been in the math, and is still in the math," he wrote, arguing that domestic Chinese AI demand is itself so large that Chinese memory producers will continue to absorb their own capacity internally rather than flooding export markets. Micron is the primary publicly traded expression of the HBM and advanced DRAM buildout tied to AI infrastructure, alongside SanDisk and South Korea-based giants Samsung Electronics and SK Hynix. Original Article Chip stock could soar nearly 200%, analyst says: "If you don't buy it, they will" Want to be a stock market bear? BCA says watch potential U.S. tax increases Rare Earths Americas CEO weighs in as race to secure rare earth supplies outside China intensifies