The EM diversification pitch is now a three-chip bet

The 2026 emerging-markets rally has a diversification problem dressed as a performance story: funds tracking the broad MSCI benchmark have outpaced U.S. equities this year on AI strength, but the same strength has remade the index into something far narrower than the EM category's diversification pitch implies.

ETF Trends ran the comparison for its sector review, using IEMG for the MSCI Emerging Markets Investable Market Index and SPYM for U.S. large caps. In that tally, Taiwan and South Korea, home to TSMC, Samsung, and SK Hynix, comprise more than 45 percent of core EM index weight, and adding China takes the three countries to about 70 percent of geographic exposure; technology is roughly 40 percent of sector weight, three hardware names a quarter of the portfolio, and five names a third.

The return numbers built that concentration, which makes it easy to overlook as a problem. It is one. A VettaFi advisor poll cited in the review, run in a webcast with Pictet Asset Management, found diversification remains a key reason investors allocate to EM — a rationale already out of sync with a benchmark whose country and sector weights are this narrow.

The EM concentration argument now looks like the U.S. one. Investors spent years being told that American benchmarks were too dependent on mega-cap technology and that emerging markets offered an escape; the broad EM index has since developed a single-sector problem of its own, centered on technology, with a top-of-book dependency that leaves a quarter of the portfolio in three companies.

None of that makes the rally counterfeit: AI-driven earnings at TSMC, Samsung, and SK Hynix have been strong enough to pull a broad index into double digits, and the reason is supply-chain demand out of the AI cycle. The issue is labeling, not performance. An index with this structure will behave much more like an AI-hardware wager than like the all-country diversifier that has anchored EM allocations for years.

The constructive response is deciding which bet the sleeve is supposed to make, not exiting emerging markets. If the goal is participation in AI capital spending, the current benchmark delivers it without a separate theme fund; if the goal is diversification, the benchmark is a poor instrument for it, and no amount of valuation appeal changes that.

Last month's quality-screened international exposure strategy starts from the same premise: when a cap-weighted index stops diversifying, fund construction has to carry the burden. The EM version of that decision is coming into view as country and sector weights concentrate. An investor who still wants the diversification benefit can seek an index or active approach with country caps and single-name limits, or simply accept that a broad EM ETF is now a deliberate concentration trade.

ETF Trends notes EM equities continue to trade at a significant discount to the developed world even after a strong 2026, which is a real attraction; but a discount on an index carrying a 40 percent technology weight is a discount on a concentrated supply-chain position, not on a broad sampling of developing economies. It can be owned, and it may do well; the label will not change the exposure.