Staying Diverse With Emerging Market ETFs

by: Bennington Investment Ideas

There has been significant recent outcry over the under-performance of emerging markets relative to developed or mature markets. Over the past few months, emerging market returns have trailed mature markets. However, the disappointment is to some extent misplaced. In managing a portfolio, returns are obviously a good thing, but managing risk is also a critical aspect.

For this analysis, I will focus on ETFs since these are an easy way for casual investors to obtain exposure to different countries. Table 1 shows these ETFs as well as their recent returns from September 2010 through February 2011.

Table 1: Recent Returns for Country ETFs

ETF Ticker Market Type Name Return 9/1/10-2/1/2011 (%) Return 12/1/2010 - 2/1/2011 (%)
EWC Mature iShares MSCI Canada Index 19.4% 9.7%
SPY Mature SPDR S&P 500 18.5% 7.0%
EWG Mature iShares MSCI Germany Index 19.4% 6.8%
EWJ Mature iShares MSCI Japan Index 18.1% 6.7%
EWU Mature iShares MSCI United Kingdom Index 14.4% 6.3%
RSX Emerging Market Vectors Russia ETF 19.8% 2.3%
FXI Emerging iShares FTSE China 25 Index Fund 0.2% -0.8%
EWY Emerging iShares MSCI South Korea Index 13.4% -1.2%
IIF Emerging Morgan Stanley India Investment Fund, Inc. -8.9% -3.3%
EWZ Emerging iShares MSCI Brazil Index 0.4% -11.9%

Data is monthly data sourced from Yahoo!Finance Historical Data. Data used was the closing price adjusted for splits and dividends.

Clearly the recent trends favor the mature market ETFs.

Portfolio Theory

The concept of portfolio theory is essentially that diversification is a good thing. Diversification is most readily measured by the correlation of price returns between two securities. This correlation can be used to determine the volatility of two securities based on their individual volatility and correlation based on the following formula:

 \sigma_p^2 = w_A^2 \sigma_A^2 + w_B^2 \sigma_B^2 + 2w_Aw_B \sigma_{A} \sigma_{B} \rho_{AB}

where σ is the volatility of the portfolio of two equally weighted securities and ρ is the correlation coefficient between the two securities and w are the security weightings. The formula graphic is from Wikipedia. Note that σ2 is the variance and the volatility is the square root of the variance. Volatility is always a positive number.

Table 2 illustrates how this impacts two securities with varying correlations assuming two assets each with a 50% weighting and each asset has the same volatility - 10%.

Table 2: Example Calculation of Volatility of a 2 Security Portfolio

Correlation Coefficient Security Volatility Portfolio Volatility
100% 10% 10.0%
50% 10% 8.7%
0% 10% 7.1%
-50% 10% 5.0%
-100% 10% 0.0%

Portfolio theory is why we buy insurance (not withstanding various government requirements). To date, my lifetime return on my auto insurance is -100%, a terrible investment. However, if my car is destroyed for some reason, my insurance will pay out extremely well (I hope). The insurance return is negatively correlated to the auto return. This example is actually much clearer looking at houses and home insurance since cars are generally pretty lousy investments. Then again, housing has been a pretty lousy investment lately too, but you get the point.


The analysis shows correlations between the price returns for the ten country ETFs.

Table 3: Correlation Coefficients between Country ETFs

SPY 100%
RSX 74% 100%
EWZ 77% 85% 100%
FXI 68% 65% 79% 100%
IIF 67% 71% 76% 67% 100%
EWJ 85% 66% 68% 66% 57% 100%
EWU 92% 82% 85% 74% 74% 84% 100%
EWC 86% 86% 91% 74% 77% 77% 89% 100%
EWG 91% 75% 76% 73% 67% 86% 90% 83% 100%
EWY 81% 70% 75% 71% 63% 79% 77% 79% 84% 100%

Data is monthly data sourced from Yahoo!Finance Historical Data. Data used was the closing price adjusted for splits and dividends. Price histories are from May 2007 to present.

So despite the lower returns from the emerging markets they still show substantially lower correlations to SPY than the mature markets. However, it should be noted that these correlations are still reasonably high. Table 4 shows correlations over different time frames.

Table 4: Correlation to SPY over Different 2 Year Time Periods

ETF Ticker Market Type 3/2/2009 - present 8/1/2000 - 7/1/2002
SPY Mature 100.0% 100.0%
EWG Mature 92.6% 75.0%
EWU Mature 86.5% 78.9%
EWJ Mature 81.7% 59.8%
EWC Mature 81.4% 89.4%
EWZ Emerging 81.3% 67.0%
EWY Emerging 79.3% 73.5%
RSX Emerging 72.8% NA
FXI Emerging 69.3% NA
IIF Emerging 51.5% 55.4%

Data is monthly data sourced from Yahoo!Finance Historical Data. Data used was the closing price adjusted for splits and dividends.


This above table and overall analysis has a couple key implications:

  1. Correlations change over time - This is pretty obvious and pretty reasonable. Different markets and subject to their own countries regulations, economic policies, and government behavior that can substantially impact the markets. Note EWJ's correlation.
  2. Correlations seem to be increasing - This is also intuitive but unfortunate since as the world and financial markets become more interconnected it is more difficult to find opportunities not linked to everything else.
  3. Correlations between emerging markets and SPY seem to consistently lag correlations between mature markets and SPY - This is also pretty logical due to the structural differences between these markets and the types of companies that typically compose an emerging market stock market and a mature stock market.

The reason to invest in emerging markets is not just because they have historically provided good returns, but they also carry lower correlations to the U.S. markets than other mature markets offering additional risk reduction benefits.

While at times, emerging markets may suffer from negative macro economic issues and other issues not usually seen in mature markets, the overall global economic growth trend favors these markets. I think that individuals too often see the performance of a single security without considering its impact on their entire portfolio. I will continue to stick with my emerging markets investments.

I will also close with one final table showing these ETF returns to present from 7/2000 and from 10/2004. Investing is more of a marathon than a sprint.

Table 5: Gross Returns on Various Country ETFs

ETF Ticker Country Market Type Gross Return from 7/2000 Gross return from 10/2004
FXI China Emerging NA 169.9%
IIF India Emerging 490.1% 93.2%
EWZ Brazil Emerging 411.1% 375.0%
EWY South Korea Emerging 244.8% 144.7%
EWC Canada Mature 139.1% 117.0%
EWG Germany Mature 35.2% 70.3%
EWU United Kingdom Mature 31.2% 28.4%
SPY United States Mature 12.2% 33.8%
EWJ Japan Mature -1.5% 23.1%

Data is monthly data sourced from Yahoo!Finance Historical Data. Data used was the closing price adjusted for splits and dividends.

Disclosure: I am long SPY, EWZ, FXI.

Disclaimer: This article is for informational and educational purposes only and shall not be construed to constitute investment advice. Nothing contained herein shall constitute a solicitation, recommendation or endorsement to buy or sell any security.