Raises India’s recommended weighting by 2 percentage points to 14%, maintaining an overweight position compared to its 12% weighting in the MSCI benchmark

New Delhi

Therecommended allocation to India in Jefferies Asia-Pacific ex-Japan relative-return portfolio has been increased while exposure to South Korea and Taiwan dropped, over concerns that the global artificial intelligence capital expenditure cycle may be approaching a peak, a report has said.

The report from Global Head of Equity Strategy at Jefferies Christopher Wood raised India’s recommended weighting by 2 percentage points to 14 per cent, maintaining an overweight position compared to its 12 per cent weighting in the MSCI benchmark. Meanwhile, Jefferies cut South Korea and Taiwan by 1 percentage point each to 18 per cent and 24 per cent, respectively, leaving both markets underweight.

The change raises India’s overweight to 2 percentage points, while Korea remains 1 percentage point underweight and Taiwan 2.8 percentage points underweight. Both these markets are closely tied to memory chips, foundries and semiconductor equipment.

Jefferies’ positioning reversed from the stance it took earlier in 2026 when AI hardware stocks rallied, which prompted lifting of index weights of Korea and Taiwan. India is now being positioned not primarily as a domestic-growth overweight, but as relatively less exposed to a potential slowdown in AI-related capital spending.

Though India is not insulated from a broad emerging-market selloff, its lower direct exposure to hyperscaler spending and semiconductor hardware felt attractive to the analyst. Wood said that financial markets are more cautious to fund the next phase of AI investment, even though capital-spending plans remain elevated.

Consensus estimates cited by Jefferies project nearly $1.04 trillion of capital expenditure by Microsoft, Amazon, Alphabet and Meta in 2027, while credit spreads and credit default swap spreads have widened for many of these companies, pointing to signs of stress.

Memory stocks saw 40 per cent dip from their late-June peak, while the China AI Semiconductor Top 10 Index has halved from its mid-year high. Another report said that India’s benchmark index Nifty 50 to reach 28,000 points by the end of 2027, supported by a strong earnings outlook, lower oil prices and the prospect of slower global artificial intelligence-related capital expenditure.