EEM: a fund-of-funds approach

With 668 stocks from all over the world, the $40 billion MSCI Emerging Markets portfolio would be tough to replicate on your own. But how close can you get?

Buying all those stocks on your own would be folly, even if they were all publicly traded in ADR form. And since EEM (quote) is admirably diversified — even the most massive allocations are under 3% of the total portfolio — you cannot really just buy the top 10 holdings here and think you have it all.

However, while EEM is very granular on a holding-by-holding basis, it is surprisingly concentrated when it comes to the markets in which it invests. Yes, just 10 countries account for 88% of the total portfolio — and this is where a truly sophisticated investor can replicate and even fine-tune the global fund.

Let’s see what happens if we tinker with a few of the big weightings.


China accounts for 18.42% of EEM and this aspect of the portfolio is surprisingly concentrated, with only eight companies accounting for almost half that allocation.

If we break out China as a separate “mini-ETF,” then China Mobile (CHL, quote) takes up a whopping 10% of the “portfolio” and China Telecom (CHA, quote) represents another 6%.

The banks together account for 17% and China Life (LFC, quote) adds 6%. Oil giant CEO (quote) is weighted at 5% of the overall China allocation and PetroChina (PTR, quote) adds 4%.

Comparing this “mini-portfolio” to dedicated China ETFs reveals that it resembles a significantly less financial-heavy FXI (quote), only swapping CHU (quote) for CHT.

On the other side of the trade, GXC (quote) leans even further from the banks to add a few percentage points to oil and technology — EEM does not even list Baidu (BIDU, quote) among its holdings, while GXC gives it a 3% weight.


With 15.86% of the EEM portfolio, the Brazilian allocation is surprisingly concentrated. Counting preferred shares, Petrobras (PBR, quote) accounts for 19% of this “mini-portfolio” and banks ITUB (quote) and BBD (quote) add up to another 20%. Vale (VALE, quote) adds 10%.

EWZ (quote) is roughly the same animal, so simply swapping it in for EEM’s Brazil holdings should be a wash. However, because EWZ is so top-heavy, sophisticated investors may investigate replacing a slice of that 15.86% large-cap allocation with a smaller-cap offering like BRF (quote).

South Korea and Taiwan

Lumping these two countries together because of their relatively limited single-market ETF coverage and export focus, South Korea makes up 13.2% of EEM and Taiwan adds 10.4%.

ADR-deprived Samsung easily accounts for 20% of the Korean “portfolio,” followed by Posco (PKX, quote) at 9% , KB Financial (KB, quote) and Korea Electric (KEP, quote) at about 6.5% apiece.

EWY (quote), the dedicated Korea ETF, is somewhat less concentrated.

In Taiwan, Taiwan Semi (TSM, quote) is the heavyweight at 17% of the “portfolio,” followed distantly by Hon Hai Precision (HNHPF, quote) at 8%.

Note that the Taiwanese EEM holdings are a lot less concentrated than any of the countries we have seen before, but EWT (quote) is even less concentrated.

South Africa

Interestingly, South Africa, with a 7.91% weight, is more heavily represented in EEM than either Russia or India. EEM favors Sasol (SSL, quote) over MTN (MTNOY, quote) but the dedicated fund EZA (quote) basically switches the weightings.

The results

So what happens if you replace the big 88% in EEM with dedicated single-country funds and then maybe buy a frontier markets fund like FRN (quote) with the other 12%?

Using GXC, RSX (quote) and PIN (quote) where we have a choice of ETFs, we can replicate EEM’s country diversification.

Leaving fees and the hassle of building this granular portfolio out of it for the time being, the fund-of-funds approach would be up 5% so far this year. EEM itself is up only 1.5%.

Is EEM 350 basis points cheaper than any of the constituent funds we have been working with? No.

Is it more diversified? Not necessarily. Most of the funds in our fund-of-funds are just as diversified (or as concentrated), and FRN, the only real wild card, invests in exactly the same markets as the “little” 12% in EEM. In most cases, these are exactly the same stocks.

Is EEM just having a bad year? Maybe. But that is a question for index designers.