Micron Technology has already delivered the kind of rally that would normally make investors nervous about arriving late, yet one analyst now argues the opposite: the stock may still be inexpensive. Seeking Alpha analyst Julian Lin upgraded Micron to Buy, pointing to powerful demand created by the expansion of agentic AI, accelerating revenue across the business and a valuation of roughly six times earnings. The call lands only days after Micron delivered another extraordinary earnings report, putting the Micron stock forecast 2026 at the center of a much bigger question: has Wall Street still failed to appreciate how profoundly artificial intelligence is changing the economics of the memory industry?
The numbers explain why that question is becoming difficult to dismiss. Micron’s fiscal fourth-quarter revenue surged to $54.23 billion, while adjusted earnings reached $33.42 per share. Management then forecast fiscal first-quarter 2027 revenue of approximately $61.5 billion, comfortably above the $57.02 billion analysts had expected, with adjusted earnings projected at $38.15 per share. Those would have sounded almost absurd for Micron during previous memory cycles, yet AI data centers are consuming enormous quantities of high-performance memory at the same time that supply remains constrained. The result is a company producing record revenue, exceptional margins and massive cash flow while still trading at a multiple normally associated with a business whose profits are about to collapse.
That contradiction is the heart of the Micron story now.
AI Is No Longer Just an HBM Story for Micron
The obvious explanation for Micron’s transformation is high-bandwidth memory, or HBM, the specialized memory positioned alongside advanced AI accelerators to move enormous quantities of data quickly enough for modern artificial-intelligence workloads. HBM has become one of the most strategically important components in the AI infrastructure stack, and exploding spending on accelerators and data centers has tightened supply across the memory industry.
But Lin’s bullish argument goes further. AI demand is increasingly influencing all of Micron’s businesses, not simply the most obvious data-center products. The reason is that memory manufacturing capacity is interconnected. When producers dedicate more wafers and advanced packaging resources to high-value HBM, less capacity is available for conventional DRAM products. Meanwhile, AI servers themselves require substantial amounts of standard memory and storage in addition to HBM. That combination can tighten the broader market and give manufacturers greater pricing power even with customers that are not directly purchasing AI-specific memory.
Related articles
The latest industry data support that thesis. World Semiconductor Trade Statistics data cited by Barron’s showed memory sales rising 36% in August, with DRAM prices climbing 8% and NAND prices increasing 5%. Circular Technologies analyst Brad Gastwirth expects additional fourth-quarter increases, forecasting DRAM pricing growth of roughly 15%-20% and NAND increases of approximately 12%-20%. That matters enormously for Micron because DRAM represents roughly three-quarters of its revenue.
In other words, AI does not have to directly touch every Micron product for Micron to benefit. If AI absorbs enough industry capacity, pricing can strengthen throughout the memory hierarchy—and that is exactly the kind of operating environment capable of turning a historically cyclical manufacturer into an extraordinary cash generator.
Micron’s $54 Billion Quarter Shows How Extreme the Shift Has Become
Memory has traditionally been one of the semiconductor industry’s most violent boom-and-bust businesses. When demand exceeds supply, prices rise rapidly and manufacturers produce spectacular profits; when too much capacity enters the market, prices collapse and those same profits can disappear almost as quickly. Investors learned to treat peak earnings with suspicion because the better the numbers became, the closer the industry often appeared to its next downturn.
That historical experience helps explain the strange reaction to Micron’s latest results. Fiscal fourth-quarter revenue reached $54.23 billion, adjusted earnings increased to $33.42 per share and gross margins reached extraordinary levels, yet the stock did not respond with the explosive rally such numbers might normally suggest. Investors are already asking whether the memory cycle has become too good, and whether today’s earnings should therefore be treated as a peak rather than a sustainable base.
Micron insists the supply-demand environment could remain tight through fiscal 2027 and 2028. The company’s first-quarter guidance reinforces that argument, with expected revenue rising again to $61.5 billion. If achieved, Micron would continue growing from an already enormous fourth-quarter base instead of immediately slipping into the downturn skeptics fear.
That is where the six-times-earnings valuation becomes provocative. A low multiple makes perfect sense if profits are about to collapse. It looks very different if earnings remain elevated for several years.
Agentic AI Could Add Another Layer to the Memory Boom
Generative AI created the first wave of infrastructure spending as technology companies raced to train increasingly sophisticated models. Agentic AI could make the next phase even more memory intensive because AI systems capable of performing multi-step tasks, operating software and continuously interacting with external information may require substantially more inference compute.
Baird analyst Tristan Gerra recently raised his Micron price target to $1,520 from $1,280, citing the surge in agentic-AI demand, expectations for roughly 40% industrywide AI CPU unit growth in 2027 and slower DRAM supply-bit growth. Baird also expects a stronger HBM margin profile next year.
That combination matters because the AI investment story is evolving beyond GPUs alone. CPUs, memory, storage, networking and power infrastructure all become increasingly important as AI moves from model training toward large-scale deployment and inference. Every additional server needs memory, and increasingly sophisticated workloads can require more memory per system.
Micron therefore occupies an unusually attractive position in the AI supply chain: it does not have to predict which individual AI application ultimately wins. It needs the overall quantity of compute—and the amount of data moving through that compute—to keep increasing.
So far, customers appear confident enough to make commitments extending years into the future.
$150 Billion of Contracted Business Could Change Micron’s Old Boom-Bust Model
Perhaps the most important development in the Micron investment thesis isn’t this quarter’s revenue at all. It is what the company is doing to make future revenue less unpredictable.
Micron has accumulated 26 strategic customer agreements, some extending through 2031. D.A. Davidson analyst Gil Luria estimates that these agreements represent at least $150 billion in remaining performance obligations, with pricing frameworks and floors that could make the ultimate revenue value even higher if market prices remain strong.
Reuters separately reported that Micron’s customer commitments backed largely by cash deposits increased to $32 billion from $22 billion in June, while remaining performance obligations jumped from $100 billion to $150 billion.
That is potentially transformative because one of the biggest reasons investors historically assigned low multiples to memory companies was poor visibility. Customers could overorder during shortages, manufacturers could overbuild capacity and pricing could suddenly collapse when the cycle turned. Long-term agreements cannot eliminate cyclicality, but binding commitments, price floors and better demand visibility could make the next downturn less destructive than previous ones.
If that happens, the market may eventually decide Micron deserves a higher earnings multiple.
And that could matter almost as much as earnings growth itself.
The Biggest Micron Risk Is That Today’s Incredible Numbers Really Are the Peak
The bullish case is powerful, but the skepticism isn’t irrational. Memory remains a commodity-like industry, and enormous profits inevitably encourage investment. Micron itself is expanding capacity in the United States and Japan, while competitors are also motivated to capture the AI-memory opportunity. Additional supply expected from new facilities later in the decade could eventually loosen today’s extraordinary conditions.
Competition from Chinese memory manufacturers is another long-term concern. Bank of America analyst Vivek Arya has highlighted the potential threat from ChangXin Memory Technologies, particularly if lower-priced alternatives become increasingly attractive to electronics manufacturers. Arya has also questioned whether future AI-chip architectures could alter the amount of HBM required per accelerator, creating uncertainty around long-term demand assumptions.
Margins deserve attention too. Deutsche Bank has flagged elevated costs even while maintaining a Buy rating, illustrating how rapidly expectations have risen. When a company reaches extraordinary profitability, simply maintaining those economics becomes harder because investors begin measuring every quarter against near-perfect conditions.
That explains why the stock can post record results without automatically exploding higher. Wall Street isn’t debating whether Micron’s current business is strong. It is debating how long strength this extreme can possibly last.
The Micron Stock Forecast 2026 Comes Down to Whether This Cycle Is Different
Calling a cyclical semiconductor stock cheap at peak earnings has historically been one of the easiest mistakes investors could make. A company trading at six times earnings is not genuinely inexpensive if those earnings are about to be cut in half. That is precisely why Micron’s valuation remains so low despite spectacular results.
But this cycle contains ingredients that previous cycles did not have at the same scale. AI infrastructure spending is absorbing huge amounts of advanced memory, HBM manufacturing consumes significantly more wafer capacity than conventional products, customers are signing multi-year commitments, pricing remains firm and Micron expects supply-demand tightness to persist well beyond the current quarter. Meanwhile, its first-quarter revenue guidance suggests the business is still accelerating rather than rolling over.
That makes Julian Lin’s upgrade particularly interesting. The argument is not simply that Micron’s earnings are enormous. Everyone can see that. The argument is that the market may still be valuing those earnings as though they are about to disappear while evidence increasingly suggests the cycle could remain stronger for longer.
Investors should therefore watch memory pricing, HBM supply, hyperscaler capital spending and the execution of Micron’s long-term customer agreements far more closely than the next day’s stock-price reaction. If DRAM pricing continues strengthening while AI infrastructure investment expands, today’s roughly six-times-earnings valuation could become increasingly difficult to justify.
If memory supply catches demand faster than expected, however, that low multiple could turn out to have been warning investors all along.
Micron has already proved AI can transform its earnings.
The next question—and the one that could decide MU stock’s next major move—is whether AI has transformed the memory cycle itself.
Disclaimer
This article is for informational purposes only and does not constitute financial or investment advice. Readers should conduct their own research and, where appropriate, consult a qualified financial advisor before making investment decisions. This article was researched and drafted with the support of AI, then reviewed, fact-checked, and edited by the editorial team before publication.
