The South Korean stock market is nearing the remarkable 76% gain it achieved in 2025, which led global equity performance. However, a key difference in 2026 is that this comparable gain has been realized in less than five months from the start of the year. As of Wednesday's close, the benchmark Korea Composite Stock Price Index (KOSPI) surged 6.5%, bringing its year-to-date increase to a stunning 75%. This marks the eighth time this year the index has posted a single-day gain exceeding 5%, a feat that occurred only once throughout the entire year of 2025.
Overseas retail and institutional investors purchased over $2 billion worth of Korean stocks on Wednesday, either directly or indirectly through cross-border ETFs, a figure just shy of the record set last October. This demonstrates the powerful influence of Korean trends, extending from the fashion scene in Seoul to global financial markets. The intense bullish sentiment surrounding Korean equities is primarily driven by a single, dominant investment theme: the artificial intelligence-fueled super cycle for memory chips.
Long overlooked by global investors, Korean tech giants, particularly Samsung Electronics and SK Hynix, are now experiencing significant re-rating due to their exclusive technological advantages across key segments of the AI computing supply chain. This has made their products, and consequently their stocks, increasingly sought-after. Propelled by the strong performances of the Korean and Taiwanese markets, global equities are witnessing a historic divergence, with Asian markets consistently setting new records and significantly outperforming US indices and developed market benchmarks.
The prevailing narratives of "AI anxiety trading" and "AI disruption" are reshaping global asset allocation strategies, prompting a flow of institutional and retail capital from the US towards Asian stock markets, which are seen as the most concentrated hubs for participants in the AI compute supply chain. Within this unprecedented AI infrastructure boom, often termed the "memory chip super cycle," the two Korean memory chip behemoths—Samsung and SK Hynix—which together account for nearly 50% of the KOSPI's weighting, serve as the primary engines attracting global capital. They are the core drivers behind the Korean market's record-breaking outperformance.
Wednesday marked a milestone for Samsung Electronics, whose shares have risen approximately 75% year-to-date, pushing its market capitalization above $1 trillion. This makes it the second Asian company, after Taiwan Semiconductor Manufacturing Company (TSMC), to reach this valuation level. SK Hynix, a dominant force in the High Bandwidth Memory (HBM) sector, has seen even stronger gains, with its stock price soaring over 130% since the start of the year.
Global investors are actively accumulating Korean chip stocks. The iShares MSCI South Korea ETF, traded in the US, has skyrocketed 80% year-to-date, outperforming the broader US market and the Philadelphia Semiconductor Index. Investors focused on the Hong Kong market are also aggressively buying leveraged ETFs tied specifically to the Korean chip sector. A double-leveraged ETF tracking SK Hynix listed in Hong Kong has surged 345% this year, while a similar ETF for Samsung is up 290%. Additionally, a China-listed ETF focusing on Sino-Korean semiconductors has gained 83%.
The underlying driver is the critical need for HBM memory systems, which are fully integrated with AI chips, within massive computing clusters like Google's TPUs or Nvidia's AI GPUs. This is compounded by tech giants accelerating the construction or expansion of AI data centers, requiring large-scale purchases of server-grade DDR5 memory and enterprise-level high-performance SSDs/HDDs. Samsung, SK Hynix, and Micron Technology are uniquely positioned across these three core memory domains: HBM, high-performance server DRAM (including DDR5/LPDDR5X), and high-end data center SSDs. They are the most direct beneficiaries of the "AI memory and storage stack," capturing a "super红利" (super dividend) from the AI infrastructure wave.
"The most remarkable aspect is that they still possess substantial upside potential," stated Chan H. Lee, managing partner at Seoul-based hedge fund Petra Capital Management. "This isn't just a Korean story; it's a global memory super cycle underpinned by the AI wave. The intrinsic value of Samsung Electronics is finally being recognized."
As charts indicate, the KOSPI has soared this year amid the AI fervor. Memory chip manufacturers have become among the world's most coveted stocks due to persistent supply constraints and expectations of sustained strong growth, potentially insulated from broader economic cycles. Even after significant price increases and record-breaking performances this year, analysts believe shares of Samsung and SK Hynix have further room to grow, citing seemingly endless demand for memory chips and relatively low valuations.
Policies promoted by South Korean President pushing for stock assets to become a cornerstone of household wealth are also supporting the market, through reforms aimed at strengthening corporate governance and boosting shareholder returns. While rising oil prices amidst geopolitical tensions add risk for some sectors, the situation has also refocused attention on the quietly accumulated global competitiveness of Korean shipbuilders and international defense companies.
Despite these converging positive factors, concerns are growing that the rally may be becoming excessive. The KOSPI closed Wednesday at 7,384.56 points, far surpassing a target of 5,000 points once considered ambitious. The benchmark index is also approaching a year-end target of 8,000 set by Wall Street giant Goldman Sachs last month. Notably, highlighting the narrow breadth of the rally, despite the sharp rise in the key index, over 600 of the 835 constituents in the KOSPI composite index declined on Wednesday.
Short sellers are increasing bets that a significant short-term correction is inevitable after such a rapid and explosive rise, according to Ihor Dusaniwsky, head of predictive analytics at S3 Partners. "I would maintain a cautious stance from here as well, because the macro backdrop is no longer entirely favorable," said Jung In Yun, CEO of Fibonacci Asset Management Global. "The market is increasingly reliant on continuous earnings revisions from a very small number of leading semiconductor companies. If global liquidity tightens further or AI spending expectations begin to normalize, volatility could rise quite rapidly given how concentrated this rally has become."
As shown in accompanying analysis, Korea leads global earnings revision trends, driven by AI chip demand. Although the unprecedented surge might appear precarious, many analysts argue it is well-supported by earnings growth—consensus estimates project overall earnings for KOSPI constituents to grow over 200% in the next 12 months. Samsung currently trades at a forward P/E ratio of just 6x, and SK Hynix at 5.3x, meaning their valuations are a fraction of Nvidia's 22x multiple, the global AI chip superpower with a market cap nearing $5 trillion.
"The Korean market's rise is supported by multiple tailwinds," said Stanley Tang, senior portfolio manager at Sumitomo Mitsui DS Asset Management Co. "Memory chip manufacturers' profits are at historic highs, driven by seemingly limitless strong AI-driven computing demand; while shipbuilders benefit from a shipping industry bull cycle amid geopolitical crises and relatively low steel prices."
The logic is stark: without sufficient memory chips, AI models must recalculate from scratch. HBM, DDR5, and SSD technologies are collectively unleashing a memory chip super cycle. Regarding DRAM/NAND price trends, Goldman Sachs recently significantly upgraded its forecasts. The investment bank now expects DRAM prices to rise 250-280% in 2026, up from a prior estimate of around 150%, and NAND prices to increase 200-250%, up from approximately 100%. Goldman argues this is not a typical inventory restocking cycle, but a "super supply shortage cycle" caused by unprecedented AI-driven demand surges, HBM's complex manufacturing process crowding out capacity, and limited supply elasticity for general-purpose DRAM/NAND.
In this architecture, GPUs generate intelligence, HBM/DRAM feeds data at high speed, enterprise NAND/eSSD handles hot data and caching, and HDDs manage long-term storage for massive cold/warm data. Goldman posits that the AI computing arms race led by cloud giants is transforming memory chips from cyclical commodities into scarce strategic assets. The 2026 price increases are not the end, but potentially the initial phase of a super cycle.
The essence of Korea's market surge—the KOSPI's 75% gain, Samsung's $1 trillion valuation, SK Hynix's record highs—is not merely a domestic bull market. It represents a global bet on the "AI-driven memory super cycle." Samsung and SK Hynix now comprise about 50% of the KOSPI's total market cap. Foreign investors' single-day purchase of approximately 3.1 trillion won ($2.13 billion) underscores that the Korean market is becoming a "high-beta expression" of the global AI memory/storage chain.
As revealed by Jeremy Werner, senior vice president of Micron's Data Center Business Unit, from a data flow engineering perspective within AI data centers, the core driver is more complex than "AI needs more compute chips." The era of AI inference, dominated by AI agents, is turning memory/storage from supporting components into system bottlenecks. While AI training relies heavily on massive parallel computing, inference—especially involving long context, multi-turn conversations, and agentic AI workflows—requires persistent storage of KV Cache, context states, and intermediate results. Insufficient memory/storage forces models to recompute historical states, reducing GPU utilization and increasing token generation costs.
Consequently, HBM, DDR5, LPDDR, enterprise SSDs, and even HDDs/data lakes are forming an "AI memory chain" from GPU-proximal to distal storage, determining an AI system's throughput, latency, concurrency, and per-token economics. This explains the synchronized surge in stocks like Micron, Samsung, SK Hynix, SanDisk, and Western Digital—demand is not confined to HBM but is spilling over across the entire AI server architecture to DRAM, NAND, SSD, and HDD.
Critically, AI CPUs are opening a second demand curve. While the market has largely equated AI compute with GPUs and HBM, as inference workloads grow more complex, CPUs are evolving from "GPU supporting actors" into "AI coordinators" that schedule multiple agents, manage context, and orchestrate workflows. This significantly boosts demand for server DDR5 and data center-grade SSD configurations. Simultaneously, with HBM capacity largely locked by AI GPUs, squeezing available capacity for general-purpose DRAM, DDR5 and DDR4 price trends are diverging. The shortage is spreading from high-end HBM to the broader DRAM/NAND supply chain.
TrendForce, citing Micron's CEO, reported strong demand for both traditional and AI servers, constrained by DRAM and NAND supply tightness. Samsung and SK Hynix have recently warned that AI-driven memory shortages could persist until 2028 or beyond.
Amidst Middle Eastern conflicts, the "AI bull market" narrative remains potent. The AI explosion has彻底 ignited a "memory super cycle," leading to comprehensive shortages from HBM to NAND. The KOSPI's relentless climb to new highs despite geopolitical pressure, coupled with record highs for the Taiwanese market led by TSMC—a major AI beneficiary—a record 18-day winning streak for the Philadelphia Semiconductor Index, and five consecutive weeks of gains for the S&P 500, reinforce investor belief that the "AI compute investment theme" can overpower market noise, particularly related to Middle East geopolitics.
Driven by seemingly endless demand for memory chips from the global AI data center construction frenzy, Samsung Electronics, the world's largest DRAM/NAND supplier, reported earnings that far exceeded expectations, with profits skyrocketing 48-fold. SanDisk, a leader in SSD storage, reported fiscal third-quarter revenue of $5.95 billion, a 97% year-over-year surge, vastly exceeding analyst estimates of $4.7 billion. Adjusted EPS was $23.41, nearly 1.6 times the expected $14.54, with its data center business revenue more than tripling year-over-year to $1.47 billion.
An analyst team led by star analyst Ben Reitzes at Melius Research recently published a report stating the AI boom will drive robust memory chip demand growth through the end of the decade (2030). According to Counterpoint Research, the memory market has entered a "super bull market" or "super cycle" phase, with current supply, demand, and pricing conditions far surpassing previous peaks seen during the 2018 cloud computing boom.
With the concentrated emergence of powerful AI agent tools in 2026, the AI wave is rapidly sweeping the globe. The bottleneck in AI computing architecture is shifting from GPU-centric matrix multiplication throughput to the "full-stack AI system driven by AI agents." In this evolving narrative, data center CPUs and memory chips could emerge as the biggest winners. Essentially, the AI compute bull market is expanding beyond "computing systems centered on AI GPU/ASIC chips" to encompass central processors and the foundational "data storage base."
Comments