Foreign investors withdrew a record $137.36 billion from Asian equities across seven markets in the first half of 2026, marking the fastest six-month outflow since 2010. South Korea and Taiwan bore the brunt, losing $70.8 billion and $29.6 billion respectively. This exodus occurred despite significant market rallies in these regions, with South Korea's KOSPI nearly doubling and Taiwan stocks rising 62%, largely driven by just three major chipmakers: TSMC, Samsung, and SK Hynix. This concentration spurred investors to reduce exposure to these high-performing assets amid concerns about concentration risk and to seek opportunities in less-valued markets.

Analysts attributed the withdrawals to factors like currency hedging and benchmark rebalancing, rather than a straightforward "risk-off" move. Funds sold outperforming stocks to manage concentration risks in their portfolios. For instance, in June alone, foreign investors sold $27.08 billion of regional equities, including $12.63 billion from South Korea and $8 billion from Taiwan. Bank of New York Mellon data showed mutual funds sold $7.50 billion of South Korean equities, pension funds sold $4.35 billion, and hedge funds sold an additional $1.87 billion, indicating rebalancing and profit-taking rather than a broad rejection.

In India, a significant portion of past inflows has been withdrawn to fund investments in the global AI trade. Nearly 60% of the $20 billion attracted by India-focused funds between March 2023 and October 2024, approximately $12 billion, has been redeemed. Investors pulled $9 billion from India-focused funds in 2026 alone, with $7 billion from long-only funds and $2 billion from exchange-traded funds. Redemptions were highest from Luxembourg at $3.5 billion, followed by the United States at $2.4 billion, and Japan at $2.1 billion, as the AI trade became increasingly concentrated in a few direct beneficiaries. Investors are now looking for value in other parts of the region, with Southeast Asia being considered "very, very cheap" and possessing long-term structural tailwinds.