SK Hynix and Samsung Electronics shares extended their dramatic sell-off on Wednesday, with SK Hynix falling more than 8% during Seoul trading and Samsung dropping nearly 5%, deepening a rout that has wiped out roughly $600 billion in SK Hynix’s market value since June. The slide followed Tuesday’s historic session, when SK Hynix tumbled 14.7%, Samsung lost 14.4%, and South Korea’s KOSPI index sank 10.84%.
The crash came despite record second-quarter earnings from SK Hynix. The memory-chip giant posted revenue of 79.32 trillion won and operating profit of 60.54 trillion won—a 557% surge from a year earlier—driven by booming demand for high-bandwidth memory (HBM) used in AI systems. Mass shipments of next-generation HBM4 began during the quarter. However, both figures missed elevated consensus forecasts by roughly 5% to 7%, triggering a sharp repricing after months of near-perfect execution had been baked into valuations.
Investor anxiety centers on two main themes. First, doubts are growing about the pace of hyperscaler spending on AI infrastructure, with some fearing that cloud giants may slow capital expenditure. Second, Chinese memory-chip maker CXMT’s strong stock-market debut in Shanghai has intensified worries about new supply flooding the market and pressuring chip prices. Reports of Chinese advances in deep-ultraviolet lithography technology added to the jitters, suggesting domestic production could accelerate faster than anticipated.
Analyst reactions reflected the uncertain outlook. Barclays analyst Simon Coles maintained a Buy rating on SK Hynix with a $330 price target—implying roughly 130% upside—citing expected supply constraints through 2027. Mirae Asset Securities retained a Buy but slashed its target to 2.8 million won from 4.2 million won, pointing to lower sector valuations and possible NAND weakness. NH Investment & Securities’ Shawn Oh called the stock a compelling buy, attributing some of the selling pressure to Korean retail deleveraging rather than fundamentals. Meanwhile, SK Hynix noted it has finalized multi-year agreements with around 10 major customers, providing some visibility into structural AI-memory demand.
The spillover into cryptocurrency markets could be significant. AI-themed tokens and broader risk assets often correlate with tech equities during periods of heightened volatility. As the semiconductor rout fuels a wider risk-off mood—Nasdaq futures declined, and the S&P 500 edged lower—crypto traders may reassess positions in high-beta coins. The episode underscores how AI-sentiment shocks can transmit across asset classes, potentially weighing on tokens tied to decentralized compute and artificial intelligence narratives.