Below the SurfaceEmerging Market Diversification

Analysis of iShares MSCI Emerging Markets ETF (EEM) concentration — Year-End 2025 through June 16, 2026.

Executive Summary

It is common knowledge that the S&P 500 and Nasdaq 100 indexes became quite concentrated in the Mag 7 names over the past decade, primarily through the relative outperformance of these Mega Cap stocks relative to the rest of the index constituents. Dramatic relative outperformance will do that in capitalization weighted indexes.

I wanted to take that thought process to the Emerging Markets space, where a few spectacular semiconductor companies in Korea and Taiwan are creating a very similar dynamic with the iShares MSCI Emerging Markets ETF (EEM).

EEM now has a much more material tie into the AI buildout theme. After a remarkable run in memory and semiconductor stocks, three names — Taiwan Semiconductor (TSMC), Samsung Electronics, and SK Hynix — have grown from roughly 18% to nearly 29% of the fund, in less than six months.

Looking into the factors causing upward weight drift

The index is free-float market-cap weighted with no cap on individual constituents. In addition to cap weighting, which concentrated the US large cap stock indexes, it also has some free-float adjustments which can further increase the weights of successful companies over time.

• Free-float increases. EEM weights by market cap of the free-floating shares. Tightly held shares not in the free-float are excluded because they are less accessible. Benchmarking on shares available prevents undue pressure on large companies with a small float. As large companies gain a global following, US-listed ADRs and other shares outside the home market exchange also increase float available.
• Foreign Inclusion Factor (FIF) increases. Korea and Taiwan have historically had foreign ownership limits on some stocks; when foreign “room” expands (or MSCI raises the FIF), the investable weight steps up at the next review.

This rebalance based on float adds considerable complexity to some of this otherwise simplistic exercise of showing upward drift of persistent outperformers in a cap-weighted index in a market where free float is less volatile (such as the US).

Table 1 — Weight Drift & Genuine Rebalancing (YE 2025 – Jun 16, 2026)

All three Dec-31-2025 start weights are official iShares EEM fund weights (TSMC 2330, Samsung 005930 common, SK Hynix 000660). Component figures are rounded to one decimal and may not sum exactly to the totals shown.

Because these names carried a large and rising weight while massively outperforming, they drove an outsized share of EEM’s +25.5% year-to-date price return. The estimate ranges from ~39% on a beginning-weight basis (a lower bound, since it ignores the rising weight) to ~67% on a rebalance-aware average-weight basis (an upper estimate). This return contribution is an estimate, not a precise calculation. The point is that a material share of the EEM return in 2026 (as of mid-June) came from three very large, very hot stocks tied to the AI buildout theme.

Table 2 — Contribution to EEM’s YTD Return (EEM total: +25.5%)

Samsung and SK Hynix returns are USD, converted from local KRW returns (~+150% and ~+130%) at an approximate −8% won/dollar headwind. TSMC uses its USD ADR return. MSCI quarterly index reviews were enacted on Feb 27, 2026 and May 29, 2026 — at which time increased float adjustments also bumped up index weighting in EEM.

Conclusion — and the reason I dove down this rabbit hole

There is nothing wrong with the EEM ETF. It is liquid, and the return it has captured from increasing dependence on the new Emerging AI 3 stocks has made it more attractive to more investors and further increased its AUM. My point is that you should understand what you are buying. Most advisors do not understand this nuance, so I went below the surface.

This is not your father’s Emerging Market fund. Three outstanding, world-class companies that would be darling companies wherever they were located just happen to reside in two countries that MSCI happens to still call Emerging Markets. They are global powerhouses, and not what would have historically been thought of as an Emerging Market exposure. Index concentration in a popular sector after dramatic relative outperformance is one that I have less often seen applied to Emerging Markets.

Be aware, and do not unknowingly double down on the AI buildout theme under the guise of trying to diversify in what you believe is another asset class.

Methodology & Sources

• Contribution method: Brinson-style attribution. Contribution = weight × security return; share of EEM return = contribution / EEM total return.
• Re-anchoring: Implied weight if no rebalance = start weight × (1 + security return) / (1 + EEM return). Price drift = implied − start. Genuine rebalancing effect = actual − implied.
• Weights: Year-end 2025 weights for all three (TSMC 11.85%, Samsung 3.85%, SK Hynix 2.41%) are official iShares EEM dated holdings as of 12/31/2025. Current weights from StockAnalysis.com (data via Finnhub), as of June 16, 2026.
• Returns: Estimated YTD 2026 figures; Korean names converted from KRW to USD. EEM total return +25.5% (price return, excludes dividends).
• Caveats: Several inputs are estimates; the “genuine rebalancing effect” residual also absorbs any remaining imprecision in returns. Component figures may not sum to totals due to rounding. A precise figure would require official EEM constituent weights on each MSCI rebalance date.