Reuters and multiple outlets reported that SK Hynix on June 22, 2026, briefly overtook Samsung Electronics to become South Korea's most valuable listed company, driven by a surge in demand for HBM memory used in AI systems. Reuters reported SK Hynix's market capitalisation reached 2,080.4 trillion won after a 5.6% intraday rise, while Reuters reported Samsung's market value stood at 2,066.7 trillion won, excluding preferred shares. Reuters quoted Meritz Securities analyst Kim Sunwoo: "The emergence of customised AI memory fundamentally changed the industry's economics and allowed SK Hynix to establish itself as the market leader." Reuters also reported that including Samsung's preferred shares would raise its value to about 2,246.4 trillion won. Editorial analysis: This is a market-capitalisation milestone reflecting investor preference for firms tied to AI infrastructure rather than a disclosed change in either company's strategy.
What happened
Reuters and regional outlets reported that SK Hynix briefly overtook Samsung Electronics on June 22, 2026, to become South Korea's most valuable listed company. Reuters reported SK Hynix's market capitalisation rose to 2,080.4 trillion won after a 5.6% close, while Reuters reported Samsung's market value was 2,066.7 trillion won, excluding preferred shares. Reuters reported that including Samsung's preferred shares would increase its value to about 2,246.4 trillion won. Multiple outlets, including The Korea Herald and ARY News, attributed the rally to surging demand for high-bandwidth memory (HBM) used in AI accelerators and noted SK Hynix shares have climbed roughly 340% year-to-date in 2026.
Technical details (Editorial analysis - technical context)
Industry reporting places the stock move squarely in the context of AI infrastructure demand rather than a product launch. HBM, the specialised stacked DRAM SK Hynix supplies to customers such as Nvidia and Google/Alphabet, is repeatedly cited in coverage as a high-growth, constrained segment of the memory market. For practitioners: firms building large-scale AI training and inference systems continue to drive differentiated demand for memory bandwidth and capacity, which affects supplier revenue visibility and investor valuations across the memory supply chain.
Context and significance (Industry context)
South Korea's stock-market leadership has been dominated by Samsung Electronics since 2000, so the intraday top spot represents a notable market-perception shift reported across Reuters, The Korea Herald, ARY News, and Business Standard. Industry coverage frames this as part of a broader re-rating for memory suppliers as long-term contracts, product differentiation, and capacity tightness reduce traditional cyclicality. Observed patterns in comparable cycles show that when a single product class becomes critical to cloud and AI providers, supplier valuations can diverge sharply from diversified semiconductor groups.
What to watch
Reporting highlights two observable indicators readers should monitor: first, whether SK Hynix pursues a US ADR listing, which The Korea Herald and other outlets say analysts expect to broaden its investor base; second, how sustained order flow for HBM and long-term supply agreements evolve, which will affect revenue visibility for SK Hynix and peers such as Micron Technology. Editorial analysis: Market-cap milestones are sensitive to short-term flows; tracking end-customer procurement announcements, capex plans from major cloud/AI players, and announced supply agreements will give clearer evidence of structural valuation change rather than transient market sentiment.
Scoring Rationale #
A concrete AI-demand signal: SK Hynix overtaking Samsung in market cap for the first time is directly attributable to its dominant HBM3E position supplying Nvidia and other AI chipmakers. Well-sourced across Reuters, CNA, and Korea Herald. Relevant to practitioners tracking AI hardware supply chains and infrastructure valuations, though it is primarily a market event rather than a technical development.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.