TOKYO: Japanese and South Korean shares plunged on Friday (Nov 6) after US shares tumbled overnight on worries that midterm election results could slow or derail the artificial intelligence data centre investment boom.
Both Tokyo’s benchmark Nikkei 225 index and Seoul’s Kospi were down more than 6 per cent as of 10am Singapore time.
At one point, the Kospi index dived 8.5 per cent to 4,923 – triggering a circuit breaker which suspended trading for 20 minutes.
In Singapore, the Straits Times Index was down more than 2 per cent. Local semiconductor solutions company AEM Holdings plunged as much as 13 per cent in early trade, while semiconductor equipment supplier plunged as much as 15.2 per cent.
Investors had initially cheered Tuesday’s Democratic drubbing of US President Donald Trump’s Republican Party.
But they have since taken a more anxious stance, centred on worries that the electoral shakeup will hit key AI investments seen as important to economic growth.
Of the major US indices, the most impacted was the tech-focused Nasdaq, which finished down 5.5 per cent in a sell-off that battered Nvidia and other semiconductor stocks.
It was the Nasdaq’s worst session in terms of percentage loss since April 2025.
In South Korea on Friday, memory chip makers – whose advanced chips are used alongside other data-crunching semiconductors to power AI tools – also suffered.
Samsung Electronics plunged nearly 10 per cent while SK hynix fell more than 11 per cent.
The tech rout also hit chip-related shares in Japan, with Tokyo Electron shares trading down around nine per cent and Kioxia shedding 13 per cent.
“Following the Democratic victory in the midterm elections, concerns about stricter regulations on AI-related technologies have led to a sell-off across the entire (US) market, particularly in technology stocks, causing all three major US market indices to decline,” online brokerage Monex said.
“In particular, there are concerns that stricter regulations on data centre construction will slow down construction projects.”
Governments and tech giants worldwide are racing to build data centres to train and run AI models – part of a global construction boom costing hundreds of billions of dollars.
