Key Takeaways:
A synchronized selloff in global artificial intelligence stocks erased more than $200 billion in market value as overcrowding in the sector's most popular trades collided with shifting narratives around AI infrastructure spending.
The selloff swept across markets from Seoul to New York, with the Philadelphia Semiconductor Index falling 3.2% and South Korea's KOSPI dropping 2.8% as memory chipmakers led declines. The Nasdaq 100 slid 2.1%, its worst single-day drop in three weeks, as investors rotated out of the AI infrastructure complex that had powered much of this year's gains.
"The market is repricing the timeline for AI monetization relative to the scale of capital being deployed," said Michael Wilson, chief equity strategist at Morgan Stanley. "When positioning gets this concentrated, any narrative disruption triggers a disproportionate unwind."
The move coincided with three catalysts: a 12-basis-point rise in the 10-year U.S. Treasury yield to 4.38%, profit-taking ahead of mid-year earnings reports that begin this week, and growing debate over whether hyperscaler capital expenditure can sustain its current trajectory. The VIX jumped 2.6 points to 18.4, its highest level in a month, as options market pricing reflected increased demand for downside protection.
Sector rotation was stark. Information technology was the worst-performing S&P 500 sector, falling 2.8%, followed by communication services at minus 1.9%. Utilities and consumer staples were the only sectors in positive territory, each gaining less than 0.5%, as defensive positioning re-emerged. Within tech, semiconductor names bore the brunt: Nvidia Corp. fell 4.1%, Advanced Micro Devices Inc. dropped 3.7%, and ASML Holding NV declined 2.9%. Memory stocks were hit hardest, with Samsung Electronics Co. sliding 4.5% and SK Hynix Inc. falling 5.2%.
Korea's semiconductor-heavy index bears watching
The KOSPI's 2.8% decline was its steepest since April, when the Iran conflict roiled energy markets. South Korea's benchmark is now down 1.6% year-to-date, trailing the MSCI ACWI ex-U.S. Index by more than 10 percentage points, as the memory-driven rally that propelled it earlier this year shows signs of exhaustion. The won weakened 0.4% against the dollar to 1,508, adding to pressure on foreign investor flows.
In China, the CSI 300 fell 1.1%, while Hong Kong's Hang Seng Index dropped 1.8%. A-share technology names showed relative resilience, with the STAR 50 index declining just 0.6%, as domestic investors stepped in to absorb some of the selling pressure from offshore funds.
Trading volumes surged. The Nasdaq Composite recorded 14.2 billion shares changing hands, 22% above its 20-day average, while the S&P 500 saw 5.8 billion shares traded, 18% above the norm. The elevated volumes confirmed the breadth of the selloff: declining stocks outnumbered advancers by more than 3-to-1 on the New York Stock Exchange.
What comes next
The mid-year earnings season, which kicks off this week with major bank results, will test whether fundamental data can arrest the momentum-driven unwind. For AI stocks specifically, the key question is whether companies like TSMC, which reports July 18, can deliver guidance that justifies the multiples the sector commands. The Philadelphia Semiconductor Index trades at 28 times forward earnings, above its five-year average of 22 times.
Analysts at Goldman Sachs noted in a research note Monday that the AI trade's crowding had reached levels last seen in the 2021 ARK Innovation unwind, with the top five stocks in the S&P 500 now accounting for 26% of the index's total market capitalization. "Historical parallels suggest that such concentration tends to resolve through broadening, not collapse," the note said, "but the adjustment period can be volatile."
For portfolio managers, the selloff presents a tactical dilemma. The three signals to watch are whether Korean and U.S. semiconductor stocks can stabilize in the coming sessions, whether A-share tech names can maintain their relative strength, and whether trading volumes sustain at elevated levels — a sign that institutional repositioning is still underway rather than complete.
This article is for informational purposes only and does not constitute investment advice.
