Global AI Stock Selloff: Should Indian Investors Be Worried?

A bond, not a chip, lit the fuse. In mid-October, stress in the debt of AI data-centre operator CoreWeave helped turn a quiet worry into a global tech rout: South Korea's Kospi fell about 7.6% in a single session, Japan's Nikkei about 4.9%, and the Nasdaq had its worst week in months. For an Indian investor the useful question is not whether AI is a bubble. It is which parts of your money are actually wired to this.

What actually happened

The selloff began in Asia's chip complex. SK Hynix and Samsung Electronics were among the hardest hit, and CNBC reported that 20 of the world's most valuable AI and semiconductor stocks lost more than $1 trillion in value in a week. The Nasdaq Composite dropped 1.9% on October 14, its biggest one-day fall since July, and roughly 4.5% over the week by one count.

Reported figures vary between outlets, especially for the Kospi's weekly decline, so treat the exact percentages as approximate. The direction is not in dispute.

Why financing, not demand, is the story

AI data centres are built largely with borrowed money. A company like CoreWeave raises debt, buys chips and builds capacity, then hopes customer contracts pay the bills before the loans come due. Reports said its 2032 bonds, issued only months earlier, fell below 80 cents on the dollar and yields briefly spiked past 15%. The company had flagged billions in principal payments due over this year and next.

Then there is the rate backdrop. The US 10-year Treasury yield topped 5.3% earlier this month, a level not seen since 2002. When safe government bonds pay that much, speculative borrowers must pay far more, and investors demand more proof that today's spending will earn returns. Analysts quoted by Bloomberg note there is little evidence yet that chip orders have fallen. The fear is about funding, and it can hit share prices well before it hits order books.

Why Seoul and Tokyo fell hardest

Concentration. TSMC, Samsung and SK Hynix together make up nearly a third of the MSCI Asia ex-Japan index, so a wobble in AI spending expectations moves whole markets rather than single stocks. Other reporting adds that US inflation data and competition from Chinese chipmakers also weighed, so financing is not the only explanation.

India looks different on this measure. It has no equivalent of a Samsung or an SK Hynix at the centre of its benchmark indices, which is why the direct earnings link is thin.

The three routes into an Indian portfolio

The first is foreign money. Foreign portfolio investors had already been net sellers of Indian equities in October, with data through October 16 showing roughly Rs 54,600 crore sold. Global risk-off episodes tend to pull funds out of emerging markets broadly. Yet this year has also shown the opposite pattern. In late July, when an AI correction hit South Korea, Indian markets rallied as analysts said global investors were looking at markets with less AI dependence. Neither pattern is a rule, but both show that the flow channel can cut either way.

The second is Indian IT services. The sector has had its own AI story. Foreign investors pulled about Rs 16,949 crore out of Indian IT stocks in February, and the IT index fell 19.5% that month, its worst since September 2008, on fears that AI tools would erode outsourcing demand. A funding squeeze at AI infrastructure firms is a different worry from AI replacing services work, and the two should not be blended into one story. Still, clients' technology budgets are where the two meet.

The third is the one most people overlook: your own overseas funds. If you hold a Nasdaq-linked or global technology fund of funds, you carry the chip and AI concentration directly, plus rupee movement on top. Those funds are also awkward to top up. Indian mutual funds operate under an industry-wide overseas investment ceiling, and several fund houses reportedly paused or restricted fresh inflows in April and May 2026. Check the fund house before assuming you can add money, and confirm the current status of any specific scheme.

What this isn't

This is not a repeat of a banking crisis, and nothing in the reporting says AI demand has collapsed. Equally, a drop of this speed in a market as large as Korea's shows how crowded the trade had become. Crowded trades unwind quickly and sometimes stop just as fast. Some coverage already described the tech rout as losing steam within days.

Domestic factors still matter more for most Indian portfolios. Crude oil near or above $100 earlier in the autumn, a rupee weakening toward the mid-90s per dollar and heavy foreign selling have been pressing on Indian equities independent of AI. The Nifty's year-to-date decline predates this episode.

A practical check you can do today

Open your fund factsheets and look at the top ten holdings and the sector split. Many flexi-cap and large-cap funds hold Indian IT firms; international funds hold the US and Asian chip names directly. Add up the overlap. If you did not know you owned that much technology, that is the finding worth acting on, through your asset allocation and time horizon rather than the headlines.

If your money is for goals more than five years out and you invest through a SIP, a few bad weeks in someone else's chip stocks is a reason to review, not a reason to stop.