Global brokerage Jefferies has raised India's allocation in its Asia-Pacific ex-Japan portfolio to 14 per cent, up two percentage points from 12 per cent, while trimming its exposure to South Korea and Taiwan, as its strategists warn that the unprecedented boom in artificial intelligence investment may be running out of steam.
The change comes in the latest edition of the firm's widely followed GREED & fear report, written by Christopher Wood, Jefferies' global head of equity strategy, and dated 5 November. India now carries a two-percentage-point overweight against its 12 per cent benchmark weight, while South Korea has been cut by one percentage point to 18 per cent and Taiwan by one percentage point to 24 per cent, leaving both underweight against their benchmarks, by one percentage point and 2.8 percentage points respectively.
The report does not explicitly attribute India's higher allocation to domestic earnings growth, valuations or economic fundamentals. Instead, the shift is set squarely against growing concern about AI-related investments and their impact on technology-heavy markets, with AI-linked semiconductor stocks in sharp correction and investors reassessing whether the massive spending on technology infrastructure can be sustained.
Wood's report points to a significant correction in Chinese semiconductor stocks as evidence that enthusiasm for AI-related companies is weakening. The China AI Semiconductor Top 10 Index surged 109 per cent in dollar terms during the first half of 2026, reaching a peak in June, but has since fallen 50 per cent, compared with a much smaller 1 per cent fall in the broader MSCI China index over the same period.
The damage extends beyond China. An index of memory-chip companies Micron, Samsung Electronics and SK Hynix has fallen approximately 40 per cent from its late-June peak, according to the report. The declines matter most for South Korea and Taiwan because of their heavy exposure to semiconductor manufacturing and the global technology supply chain, the two markets Jefferies has just trimmed.
The deeper warning in the report is that the AI capital expenditure cycle itself may have peaked, or at least entered a mid-cycle correction. Financial markets are increasingly signalling reluctance to fund AI-related investments, with rising financing costs and doubts over the returns on the enormous sums being spent weighing on technology investment in the coming year.
The scale of that spending is extraordinary. The four major US hyperscalers, Microsoft, Amazon, Alphabet and Meta, increased their combined capital expenditure by 108 per cent year on year to 202 billion dollars in the third quarter of 2026, and consensus estimates suggest they could collectively spend 1.04 trillion dollars on capital expenditure in 2027. Wood questions whether financial markets will continue to support investments of that magnitude.
The stakes go well beyond technology shares. The report notes that AI-related capital expenditure contributed 1.39 percentage points, or 67 per cent, of US real GDP growth of 2.1 per cent in the four quarters to September 2026, so a sharp slowdown in that spending would have broader implications for economic growth and global equity markets. Jefferies' research has also pointed to global AI stocks' earnings growth slowing from around 100 per cent in 2026 to 30 per cent in 2027 and 22 per cent in 2028, while their valuations remain almost twice those of non-AI stocks.
For Indian investors, the decision signals a relative preference for India within Jefferies' regional portfolio as AI-related risks intensify, and it underlines the importance of monitoring global technology exposure, particularly through international funds and exchange-traded funds. But the report's own caveats matter: these are changes to a recommended model portfolio, not actual foreign institutional investor inflows into Indian equities, and a higher model allocation does not guarantee stronger Indian equity returns or increased foreign investment.
What happens next will depend on whether the correction in AI-heavy markets deepens or stabilises, and whether it encourages broader portfolio diversification towards markets such as India. Traders will be watching the December Federal Reserve decision, the next round of hyperscaler spending plans and whether the financing squeeze on AI infrastructure eases or tightens further, before judging whether Wood's call marks a durable rotation or only a pause in the AI trade.
Details in this post were checked against the Business Today and The Hindu BusinessLine reporting at about 6:00 am Pakistan time on 8 November 2026.
This post is market news information only, not financial advice.
