Tech stocks shed $2.5 trillion as US midterms deepen AI spending fears
Stock markets in the United States and Asia fell sharply as investors sold off shares in technology and artificial intelligence companies, wiping hundreds of billions of dollars from the value of major firms.
The tech-heavy Nasdaq Composite index closed 5.5% lower, shedding an estimated $2.5 trillion in market value, according to Dow Jones data. Other leading US indices also recorded steep losses in what was their weakest session in more than a year.
The sell-off followed the US midterm elections, in which the pace and direction of AI development featured prominently in campaign debate, and came amid longer-running questions over whether the enormous sums being invested in AI infrastructure will eventually produce returns.
In Asia, technology-focused indices fell sharply on Friday morning as companies that make the chips and other equipment underpinning AI saw their shares slide.
Japan's Nikkei 225 was 6% lower in early trading. Memory chip maker Kioxia lost more than 12.6%, while rival Tokyo Electron slipped 9.5%.
In South Korea, chipmaker SK Hynix fell more than 11.5% and Samsung was down nearly 10%. The declines pushed the Kospi index down more than 8.5%, triggering a circuit breaker, a mechanism that temporarily halts trading to curb panic selling.
Taiwan's Taiex dropped more than 8.3%, with chipmaking giant TSMC falling nearly 8.5%.
Britney Nguyen, a chips analyst at MarketWatch, described the move as a "continuation of worries" over the level of spending on AI infrastructure.
Circuit breakers are exchange-level safeguards designed to pause trading briefly when losses cross set thresholds, giving investors time to absorb information and reducing the risk of cascading selling. Their use does not by itself indicate a market has bottomed or that trading will resume smoothly.
The companies that fell hardest sit at the upper end of the AI supply chain. Chipmakers and equipment manufacturers have benefited from heavy demand for the hardware that powers data centres and AI models, and their share prices have risen steeply in recent years. That same exposure makes them particularly sensitive when investors reassess how quickly that demand will translate into profits.
Analysts have for months debated whether the capital expenditure committed by large technology firms is proportionate to the revenue that AI products are likely to generate in the near term. The debate is not settled. Some argue the investment will underpin a durable shift in how computing is done; others warn that spending has run ahead of demonstrated commercial returns.
Market participants also caution against reading too much into a single trading session. Equity markets can recover quickly, and sharp one-day falls are not reliable predictors of longer-term economic trends. Technology companies have accounted for a large share of gains in global indices in recent years, which makes those indices especially sensitive to shifts in sentiment towards the sector.
The figures cited here reflect trading at the time of reporting, and Asian markets were continuing to trade. How the losses settle, and whether policy signals from Washington affect the outlook for AI investment, will become clearer in the sessions ahead.
