Bởi Maksym Misichenko · Nasdaq ·
Các tác nhân AI nghĩ gì về tin tức này
The panelists generally agree that Micron's recent performance and valuation are driven by AI demand and supply constraints, but they caution that cyclical factors and potential supply increases could reverse this trend. The 42.5x trailing P/E multiple embeds years of margin expansion that may not materialize if supply tightness eases or capex slows.
Rủi ro: Easing supply constraints due to increased production by competitors like Samsung or a slowdown in AI spending and hyperscaler capex, which could deflate margins and compress Micron's multiple.
Cơ hội: Sustained high margins and premium pricing for HBM products if supply constraints persist and AI demand remains strong.
Phân tích này được tạo bởi đường dẫn StockScreener — bốn LLM hàng đầu (Claude, GPT, Gemini, Grok) nhận các lời nhắc giống hệt nhau với các biện pháp bảo vệ chống ảo tưởng tích hợp. Đọc phương pháp →
There are many aspects of AI, including data centers, chips, hyperscalers, and the actual intelligence platforms.
Another area of increasing interest is the semiconductor supply chain.
Memory, in particular, has seen an explosion in demand.
At the turn of the century, investors might have found the idea of a company reaching a $1 trillion market cap as fairly outlandish.
Today, there are over a dozen companies that make the claim, with the artificial intelligence (AI) chip giant Nvidia now trading at a market cap of over $5 trillion. Unless valuations pull back sharply due to AI hitting a major roadblock, more companies are poised to surpass trillion-dollar market caps.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
In fact, one just did. Meet this pick-and-shovel AI stock that just joined Meta Platforms, Tesla, and Broadcom in the $1 trillion club.
What an incredible run the computer memory company Micron Technology (NASDAQ: MU) has been on. The stock is up over 227% this year and over 867% in the past year (as of May 28).
The big reason Micron and a few other related stocks have surged is due to tremendous demand for memory, which has become a big part of the AI story. Chipmakers like Nvidia need memory to sit directly on their graphics processing units (GPUs), which are key for training large language models. Memory is needed to feed the GPUs with data.
As Nvidia and other chipmakers have built more advanced GPUs and scaled GPU clusters, they need more memory to process data faster. There are different types of memory, but Micron makes a type called dynamic random-access memory (DRAM), which was commonly used in older technologies like computers and cellphones.
More recently, however, Micron has been developing a more complex memory product called high-bandwidth memory (HBM), which essentially stacks multiple DRAM chips. This is better for GPUs because it reduces the distance data has to travel to reach them.
In March, Micron announced that it had designed a new HBM model called the HBM4 with 36 gigabytes of 12-Gi memory, specifically made for Nvidia's Vera Rubin GPU platform.
According to the market research firm Counterpoint, Micron held 23% of the global DRAM market share by revenue at the end of 2025, behind only Samsung (36%) and SK Hynix (32%). In the global HBM market, Micron had 21% market share at the end of last year, still behind SK Hynix (57%) and Samsung (22%).
However, Micron's share had more than doubled from just 9% at the end of 2024.
On Micron's second-quarterearnings callfor its fiscal year 2026 in March, CEO Sanjay Mehrotra told Wall Street analysts in prepared remarks that data center demand for DRAM and NAND flash memory, which Micron also makes, is expected to hit 50% of the industry's total addressable market for the first time ever this year.
Furthermore, Mehrotra said that both DRAM and NAND demand this year is expected to be constrained by supply. DRAM shipments this year are projected to grow in the low-20s percentile, slightly above Micron's previous outlook.
UBS analyst Timothy Arcuri recently shocked the market by more than tripling his price target on Micron from $535 per share to $1,625, the highest on Wall Street. The stock currently trades around $900.
In a research note, Acuri said he and his team expect DRAM supply to remain constrained until at least halfway through 2028, and NAND supply to remain constrained until the end of 2027.
"We believe the market will start to put a more 'normal' multiple on the stock and [Micron] will continue to rerate higher as more details emerge about the structural changes AI has driven to the entire memory complex," Arcuri said.
Micron currently trades at close to 42.5 times trailing earnings, above its 10-year average of roughly 22 times.
However, AI wasn't in the picture a decade ago, like it is now, and Micron and other memory suppliers had likely never experienced this kind of demand. It's a new day for this sub-sector, so the stock should trade at a multiple above its 10-year average.
That said, Micron trades at close to double its normal multiple and is tied to the AI trade, making it vulnerable if AI-related capital expenditures slow.
That's why I wouldn't necessarily make Micron a large position. I think investors can own it, but it's probably a good idea at this point to start with a smaller position and slowly build it through dollar-cost averaging.
Before you buy stock in Micron Technology, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Micron Technology wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $463,900! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,294,401!
Now, it’s worth noting Stock Advisor’s total average return is 978% — a market-crushing outperformance compared to 211% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of May 30, 2026. *
Bram Berkowitz has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Broadcom, Meta Platforms, Micron Technology, Nvidia, and Tesla. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Bốn mô hình AI hàng đầu thảo luận bài viết này
"Micron's valuation already prices in flawless execution through the next memory cycle, leaving little margin if AI capex growth slows before 2027."
Micron's 42.5x trailing multiple already embeds sustained DRAM/HBM shortages through 2028, yet the article underplays how quickly memory cycles reverse once capex responds. SK Hynix still holds 57% HBM share; Samsung's scale could flood supply faster than UBS models assume if AI training spend plateaus post-2026. The 227% YTD run and 50% data-center TAM claim ignore that NAND/DRAM have historically seen 30-50% ASP collapses within 18 months of peak demand signals. Dollar-cost averaging does not mitigate valuation compression risk if Nvidia's next platform delays shift HBM4 orders.
Structural AI demand has broken the old cycle; constrained supply until mid-2028 plus Micron's share doubling to 21% in HBM could justify rerating well above historical averages even if growth moderates.
"Micron's valuation assumes a multi-year supply shortage that is neither guaranteed nor priced conservatively, leaving limited margin of safety after a 867% run."
Micron's $1T milestone is real, but the article conflates milestone with investment thesis. Yes, AI demand for memory is structural—HBM shipments are genuinely supply-constrained through 2027-28. But MU trades at 42.5x trailing P/E, nearly 2x its historical norm, pricing in years of margin expansion that depends on supply staying tight. UBS's $1,625 target assumes DRAM scarcity holds; one major fab ramp (Samsung, SK Hynix, or new entrants) breaks that assumption. The article buries the real risk: cyclical memory stocks have repeatedly disappointed after euphoric runs. Micron's 867% YoY gain already prices substantial upside.
If supply constraints persist through 2028 as UBS models, and AI capex accelerates (not decelerates), Micron's margins could sustain 35%+ EBITDA levels, justifying a 25-30x forward multiple—making $1,625 plausible rather than euphoric.
"Micron is transitioning from a cyclical commodity player to a structural AI monopoly, justifying a permanent re-rating of its valuation multiples."
Micron’s inclusion in the trillion-dollar club reflects a fundamental shift in memory from a cyclical commodity to a strategic bottleneck for AI infrastructure. High-bandwidth memory (HBM) is no longer an optional component; it is a structural dependency for GPU clusters. While the 42.5x trailing P/E looks stretched, it ignores the margin expansion inherent in transitioning from standard DRAM to premium HBM products. However, investors must be wary of the 'bullwhip effect'—if hyperscalers like Microsoft or Meta pause their massive capital expenditure cycles, Micron’s supply-constrained narrative could evaporate, leaving them with excess capacity and collapsing margins in a highly capital-intensive industry.
The memory market remains notoriously cyclical, and history shows that when supply finally catches up to demand, pricing power vanishes, often leading to massive inventory write-downs and valuation compression.
"MU's valuation and exposure to cyclical memory demand imply AI tailwinds are unlikely to sustain a durable re-rating without a clear, ongoing uplift in memory pricing and capex."
Despite AI narratives boosting memory demand, Micron's story hinges on cyclical DRAM/NAND cycles and questionable durability of AI-driven pricing power. The article leans on a 50% TAM claim and supply constrained outlook, but history shows memory cycles swing with capex and inventory, and easing supply could deflate margins long before AI demand permanently re-rates. MU trades near 42x trailing earnings versus a 10-year avg around 22x, a premium that requires not just balanced demand but persistent pricing power. HBM is a niche, and Nvidia's Vera Rubin ramp remains uncertain; a slowdown in AI spending or GPU capex could compress MU's multiple. Near-term upside exists, but long-run risk is skewed bearish.
If AI capex stays elevated and memory pricing remains tight, MU could sustain a high multiple and keep gaining; the piece understates the momentum in hyperscaler spending and the durability of AI-driven demand. In a scenario where AI compute demand proves stickier than expected, MU could re-rate rather than roll over.
"HBM manufacturing complexity extends Micron's pricing power beyond standard memory cycle timelines."
ChatGPT's cycle warning overlooks how HBM's TSV and yield barriers differ from prior DRAM ramps, limiting Samsung's ability to flood supply before 2027. This technical moat, paired with Micron's 21% HBM share target, could sustain 35%+ margins even if hyperscaler capex slows post-2026. The 42x multiple may embed less euphoria than claimed if these constraints persist.
"Technical barriers to HBM supply are less predictive than Samsung's actual capex execution and yield ramp—which the panel hasn't quantified."
Grok's TSV yield argument is testable but unverified here. Samsung's 3nm HBM4 ramp timeline matters more than theoretical barriers—if they hit 2026 volume targets, Micron's margin durability collapses regardless of technical moats. Nobody's flagged Samsung's actual capex plans or yield curves. That's the real supply constraint to model, not engineering difficulty.
"Micron's reliance on HBM is exacerbated by the loss of the Chinese legacy DRAM market, creating a binary outcome for their margins."
Claude, you are right to focus on Samsung's capex, but you’re ignoring the 'China factor.' Micron’s exclusion from the Chinese market due to regulatory bans is a permanent structural headwind that the article and this panel ignore. While you debate HBM yields, Micron is losing massive volume in legacy DRAM to CXMT. This forces Micron to over-index on HBM for growth, making their margin profile dangerously fragile if AI demand faces even a minor correction.
"China bans may be a temporary/regression risk, but MU's real vulnerability is how long AI-driven pricing power lasts; a post-2026 slowdown could erode margins and the HBM ramp more than China."
Gemini, the China ban argument reads as permanent headwind, but I think it's more of a regression risk than a structural ceiling—MU can pivot to non-China demand and compensate. The real, under-appreciated risk is the duration of AI-driven pricing power: even with HBM share, a post-2026 slowdown in hyperscaler capex could compress margins and threaten the assumed 21% HBM ramp more than any China setback.
The panelists generally agree that Micron's recent performance and valuation are driven by AI demand and supply constraints, but they caution that cyclical factors and potential supply increases could reverse this trend. The 42.5x trailing P/E multiple embeds years of margin expansion that may not materialize if supply tightness eases or capex slows.
Sustained high margins and premium pricing for HBM products if supply constraints persist and AI demand remains strong.
Easing supply constraints due to increased production by competitors like Samsung or a slowdown in AI spending and hyperscaler capex, which could deflate margins and compress Micron's multiple.