DRAM ETF Issuer: Samsung, SK Hynix, and Micron—all surpassing $1 trillion in market cap—the AI era for memory chips has only just begun
TechFlow Selected TechFlow Selected
DRAM ETF Issuer: Samsung, SK Hynix, and Micron—all surpassing $1 trillion in market cap—the AI era for memory chips has only just begun
Sellers remain bullish, exercising caution while adopting the best elements.
Navigating Web3 tides with focused insightsAuthors: Dave Mazza, Thomas DiFazio
Translated by: TechFlow
TechFlow Intro: The market capitalizations of the world’s three largest memory chip manufacturers have all surpassed $1 trillion. In response, Morningstar published a blog post urging investors in memory ETFs not to overlook fundamentals. Roundhill Investments—the issuer of the DRAM ETF—replied with a point-by-point rebuttal: AI infrastructure is reshaping the supply-demand dynamics of the memory industry; the manufacturing barriers for High Bandwidth Memory (HBM) are so high that new entrants simply cannot break in; and the combined projected profits of the three giants for 2027 are expected to reach $704 billion. Note that the authors of this article manage the DRAM ETF and thus hold an inherently bullish stance.
The world’s three largest memory chip manufacturers—Samsung Electronics (005930 KS), SK Hynix (000660 KS), and Micron Technology (MU)—have all crossed the $1 trillion market cap threshold, joining an exceptionally exclusive club. Yet this milestone has also drawn scrutiny.
Morningstar recently published a blog post reminding investors in memory ETFs not to ignore fundamentals, raising several pointed questions:
- Historical patterns warrant caution: The memory industry has repeatedly cycled through boom-and-bust phases—investors may be overlooking this history.
- Memory companies lack moats: Memory is fundamentally a commodity business; new capacity can always enter the market and erode pricing power, leaving firms without real barriers to protect profitability.
- Rally may be momentum-driven—not fundamentals-driven: Enthusiasm for memory stocks reflects excitement about AI more than sober analysis of earnings, margins, and supply-demand dynamics.
- Valuations have surged: Memory stocks have risen sharply—prices may already be ahead of fundamentals.
Caption: Overview of the memory chip industry
Roundhill’s position is that this time is different. To understand the future of the memory industry, one must first look back at its past.
History Is Indeed Cautionary—But Is It Still Relevant?
Boom-and-bust cycles in the memory chip industry are factual. The most classic episode occurred in the mid-1990s. In August 1995, Microsoft launched Windows 95, transforming personal computers from enterprise-only tools into mass-market consumer products. DRAM capacity per PC jumped fourfold—from 1–2 megabits to 4–8 megabits. Manufacturers were caught off guard by this sudden demand surge and rushed to build new fabs, ultimately overbuilding and triggering a price collapse.
A similar story repeated in the mid-2010s. When Apple upgraded the base storage of the iPhone 7 from 16GB to 32GB—a seemingly minor change—demand exploded at scale, prompting another wave of aggressive investment, followed again by oversupply and falling prices.
These cycles share a common pattern: technological breakthrough → demand surge → capacity expansion → oversupply → price collapse.
The question is: Does this pattern still apply today?
The memory chip industry has undergone structural change. Memory demand is no longer tied to consumer electronics upgrade cycles but instead anchored to the expansion of AI infrastructure compute capacity. This market dwarfs any single smartphone upgrade wave—and offers vastly greater growth potential.
Since January 2024, DRAM and NAND prices have surged over fivefold. Hyperscalers have begun demanding long-term supply agreements to lock in bandwidth. Historically, long-term supply agreements in the memory industry were loose frameworks subject to market fluctuations. That model has now changed. During its January 2026 earnings call, SK Hynix stated that current agreements reflect a “strong mutual commitment” between customers and suppliers—driven by the extreme capital intensity of cutting-edge memory manufacturing. Micron has reported similar long-term agreement terms.
Caption: DRAM and NAND price trends
The Moat in Memory Chips: Manufacturing Complexity
Not all memory chips are created equal. The memory powering today’s AI systems is called High Bandwidth Memory (HBM), which differs entirely from the memory used in smartphones and PCs. HBM is purpose-built for AI workloads and requires extremely stringent manufacturing conditions.
Goldman Sachs data shows SK Hynix, Samsung, and Micron collectively control nearly all global HBM supply. Decades of industry consolidation have accumulated manufacturing expertise that cannot be replicated overnight. Manufacturing complexity itself constitutes a moat—and is precisely why these three companies have reached their current dominance.
Caption: Global HBM market share distribution
This logic differs fundamentally from prior cycles. In the past: demand rose → new capacity entered → prices collapsed. Today, the bottleneck is neither capital nor willingness—it’s technical capability. SK Hynix currently controls ~58% of global HBM supply and announced on June 2 that it plans to double wafer capacity over the next five years—while warning that supply shortages will persist until 2030. Building a new fab takes at least three years; locating and building one from scratch takes over five.
Additionally, ASML—the world’s sole manufacturer of extreme ultraviolet (EUV) lithography machines, essential for producing advanced memory chips—entered 2026 with a backlog of €38.8 billion, exceeding its full-year projected revenue. Delivery lead times for individual EUV tools exceed 12 months. This bottleneck cannot be resolved in the near term.
Fundamentals: Memory Makers Are Poised to Join the World’s Most Profitable Companies
The earnings, revenue, and margin forecasts for Samsung, SK Hynix, and Micron reflect the secular wave of AI adoption. Bloomberg Consensus Estimates project that by 2027, these three firms will rank among the world’s top ten most profitable companies.
Caption: Projected 2027 ranking of the world’s most profitable companies (Bloomberg Consensus Estimates)
Combined net income for the three companies is forecast to reach $704 billion in 2027, with total revenue exceeding $1 trillion.
Caption: Revenue forecasts for the three memory manufacturers
Caption: Profit forecasts for the three memory manufacturers
In terms of margins, operating gross margins for Samsung, SK Hynix, and Micron have hit historic highs—surpassing previous peaks set in 2018.
Caption: Historical gross margin trends for the three memory manufacturers
These figures are unprecedented in memory industry history. Even if growth moderates, the memory sector stands to settle at an unprecedentedly high baseline amid the ongoing integration of generative AI across the global economy.
Revaluation in a New Profitability Era
Historic stock price performance, coupled with substantial upward revisions to fundamentals, signals a major revaluation driven by earnings growth and margin expansion.
SK Hynix and Samsung serve as two representative cases. For nearly a decade, both stocks’ NTM (next-twelve-month) price-to-book ratios have oscillated within a narrow band—constrained by the cyclical profit profile characteristic of the memory industry. But that ceiling may no longer apply. Both companies’ projected ROE (return on equity) has surged to levels never before seen in memory industry history, requiring investors to reassess the valuation frameworks they’ve long used to evaluate these stocks.
Caption: SK Hynix NTM P/B ratio and ROE trends
Caption: Samsung Electronics NTM P/B ratio and ROE trends
Despite recent extraordinary price gains, the median NTM P/E ratio of holdings in the DRAM ETF stands at just 8.37x—making valuations attractive relative to broader tech equities. Meanwhile, the median EPS growth rate for the current fiscal year across the portfolio is 632%. To claim memory stocks are overvalued is to apply outdated metrics to a fundamentally transformed industry. In Roundhill’s view, the gap between historical valuation conventions and current fundamental performance represents opportunity.
Caption: Valuation and earnings growth overview of DRAM ETF holdings
Conclusion: Why Roundhill Isn’t Worried
Skepticism toward sharp price surges is reasonable—fundamentals remain paramount over the long term. Yet in this case, fundamentals are precisely what’s driving the rally.
The old cycle was defined by: explosive demand with no ceiling → excessive capacity expansion → inevitable price collapse. Today’s situation is structurally different: manufacturing barriers restrict new entrants; industry leaders themselves state supply shortages will persist until 2030; and the profit cycle has only just begun reflecting the scale of AI infrastructure buildout.
Roundhill believes the market is not pricing in a bubble—but rather the emergence of a new era for an industry that has struggled through decades of boom-and-bust cycles.
⚠️ Editor’s Note: Authors Dave Mazza and Thomas DiFazio are both members of Roundhill Investments, the issuer and manager of the DRAM ETF (Roundhill Memory ETF). The article’s perspective is inherently bullish. Readers should consider this alongside third-party views such as those from Morningstar. The original article’s ETF risk disclosures and legal disclaimers have been omitted; full details are available via the original source link.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@roundhill
