KOSPI Falls More Than 4% Intraday Despite Rebound in U.S. Semiconductor Stocks
The KOSPI traded around 5,616, down about 2.6%. It was down more than 4% at around 10:10 a.m. before recovering some ground. The KOSDAQ also remained in the 790s, down more than 2%, while the number of declining stocks rose sharply on both the Korea Exchange and the KOSDAQ.
Foreign investors were net sellers in both the cash and futures markets. Selling in the exchange’s cash market continued at the trillion-won level, while futures selling exceeded KRW 500 billion intraday. The won strengthened, with the won-dollar exchange rate in the 1,329-won range, but the stronger currency did not offset weakness in equities.
Foreign ownership of Korean semiconductor stocks stood at 48.2%, below the 59% recorded in 2019. In a market heavily weighted toward large-cap semiconductor stocks, declines in Samsung Electronics and SK hynix can deepen losses across the broader index if other stocks fail to rise enough to provide support. Analysts said a vicious cycle could emerge as falling stock prices prompt selling by index-tracking passive funds.
Factors cited behind foreign selling included waning preference for AI hardware, profit-taking, and limited upside potential in the Korean market. Semiconductor stocks also faced continued selling in Taiwan, while Kioxia was reported to have fallen sharply from its peak. Some observers, however, said much of the foreign selling may already have taken place, potentially reducing the capacity for further sales.
In the United States, the Philadelphia Semiconductor Index rebounded 1.6%, and memory stocks including Micron rose, but the gains did not extend to the Korean market. Wilson, a strategist at Morgan Stanley, said the U.S. equity correction appeared to be nearing its end and that investors should prepare to increase exposure to risk assets, focusing on quality stocks and companies with improving earnings. UBS interpreted selling in AI stocks as the unwinding of leveraged positions rather than damage to corporate fundamentals, while Howard Marks argued that market caution about risk made it difficult to conclude that an AI bubble was imminent.