Hedge funds record decade-high weekly buying in Korea, Japan, Taiwan equities, Morgan Stanley says

Global hedge funds posted the heaviest weekly buying of South Korean, Japanese and Taiwanese stocks in more than 10 years, with notional buying hitting a decade high for the week ended May 7. Morgan Stanley said hedge fund net exposure to Japan, South Korea and Taiwan rose to the highest level since 2010, now about 19% of global positioning, as investors chased AI-linked semiconductor and hardware names. Goldman Sachs also said April saw the largest monthly hedge fund buying inflows into Asian equities in a decade.

Analysis

This is less a one-day flow story than a regime shift in global factor leadership: Asia semis and hardware are becoming the marginal destination for crowded growth capital that previously had to find expression almost exclusively in U.S. megacap tech. When positioning in Japan, Korea, and Taiwan becomes a larger share of global hedge fund books, the second-order effect is multiple expansion in the best balance sheets and sharper underperformance in domestic cyclicals that do not directly participate in the AI capex cycle.

The key mechanical risk is that the trade is self-reinforcing until it isn’t. Once these markets become consensus long, the next catalyst is not fundamentals but positioning friction: a stronger USD, a rate scare, or any sign that AI capex growth is normalizing could trigger rapid de-grossing because crowded Asia tech is now the cleanest liquid expression of the theme. That means near-term upside can persist for weeks, but the path likely gets more fragile over months as implied expectations for earnings beats and guidance rise.

TSM is the highest-quality vehicle, but the better risk/reward may be in the second-order beneficiaries: toolmakers, substrate, advanced packaging, and select Korean/Japanese hardware names that still trade at a discount to the headline AI winners. The mispricing is that investors are paying for scarce AI exposure while still underappreciating the supply-chain bottleneck; that should support pricing power and utilization longer than consensus expects, but it also means any capex delay would hit equipment and materials names first.

The contrarian point is that this may be less a valuation catch-up than a crowded underownership trade that is already partially done. If hedge funds are now at a decade high in net exposure, upside from further multiple re-rating is likely smaller than the crowd assumes, and the next leg probably requires real earnings revision momentum rather than just flow. In that setup, chasing the index is lower quality than expressing the theme through relative-value or options structures with defined downside.

Key Decisions for Investors

• Long TSM vs short a U.S. AI mega-cap basket for 1-3 months: favor the Asia supply-chain catch-up trade over the crowded U.S. expression; target modest relative outperformance if flows persist, but cut if USD strength or U.S. rates reprice higher.
• Buy call spreads on TSM or a Taiwan semis basket into any 2-5% pullback over the next 2 weeks: use defined risk to capture momentum while avoiding outright chase; the trade works if positioning remains supportive, but premium should be capped because crowding is already elevated.
• Overweight semiconductor equipment and advanced packaging names in Japan/Korea for a 3-6 month horizon: these are the cleaner second-order winners if AI capex broadens beyond the headline foundries, with better asymmetry than the most obvious large caps.
• Pair long TSM/ASML-equivalent supply-chain beneficiaries against short domestic cyclical indices in Korea/Taiwan/Japan: the market is paying for AI beta, while non-participating local cyclicals should lag if flows stay concentrated in tech.
• Set a tactical stop on all Asia tech longs if global hedge fund flows reverse for two consecutive weeks: this is a positioning-driven trade, so flow inflection matters more than near-term fundamentals once consensus ownership normalizes.