Weak Correlation: Another Reason to Have Global REIT

The other day I read a paper on the October issue of Financial Planning magazine. The article, Think Globally by Bruce Eidelson, discussed the appeals of global REIT (real estate investment trust) to investors who seek diversifications and steady income streams, when the US real estate market struggles. One of the reasons to go global can be illustrated by the following table, which shows the correlations of real estate markets between several developed countries, including US, Canada, UK, Netherlands, France, Japan, and Australia.

  US Canada UK Netherlands France Japan Australia
US 1.00 0.77 0.43 0.47 0.57 0.31 0.46
Canada 0.77 1.00 0.46 0.48 0.59 0.37 0.57
UK 0.43 0.46 1.00 0.65 0.67 0.31 0.48
Netherlands 0.47 0.48 0.65 1.00 0.82 0.37 0.56
France 0.57 0.59 0.67 0.82 1.00 0.32 0.49
Japan 0.31 0.37 0.31 0.37 0.32 1.00 0.23
Australia 0.46 0.57 0.48 0.56 0.49 0.23 1.00

As we know, a well diversified portfolio can lower the risk because with different asset classes that don’t move in the same direction at the same time, the chance of either sharp rise or steep decline of the portfolio can be reduced. How well two asset classes (or securities) move in relation to each other is measured statistically by the correlation between the two. A correlation of 1 means the two asset classes move, either up or down, at the same time. To minimize the risk, we want asset classes in a portfolio to have weak correlations (it would be perfect if two asset classes have 0 correction, but that’s unlikely in real life).

The above table shows how the US real estate market correlates with other markets. With many correlations below 0.50 (the correlations could be even weaker if developing markets are included), the benefits of investing in global real estate markets are clear, as pointed out by the article:

Investing in global REIT structures offers investors the benefit of diversification. Real estate markets around the globe are affected by supply and demand factors that are highly localized. Because the correlations of returns between international markets are relatively low, even lower than for global equities, there are significant diversification advantages of investing in REITs on a global basis.

Currently, there are several exchange-traded funds (ETFs) that invest in global real estate markets. Among them, FFR, which I have covered before, has the largest portion in the US market, while other funds invest mainly in overseas markets.

First Trust FTSE EPRA/NAREIT Global Real Estate Index Fund (FFR, ER: 0.60%):

  1. United States: 37.92%
  2. Japan: 12.72%
  3. Australia: 11.96%
  4. Hong Kong: 9.52%
  5. United Kingdom: 9.45%
  6. France: 3.77%
  7. Canada: 3.54%
  8. Singapore: 2.75%
  9. Austria: 2.02%
  10. Netherlands: 2.01%
  11. Other: 4.34%

SPDR Dow Jones Wilshire International Real Estate ETF (RWX. ER: 0.60%)

  1. Australia: 20.46%
  2. Japan: 18.76%
  3. United Kingdom: 16.17%
  4. Canada: 7.57%
  5. Singapore: 7.00%
  6. Hong Kong: 6.98%
  7. France: 6.90%
  8. Austria: 4.99%
  9. Netherlands: 4.70%
  10. Sweden: 1.29%
  11. Switzerland: 0.91%
  12. Germany: 0.84%

WisdomTree International Real Estate Fund (DRW, ER: 0.58%)

  1. Australia: 35.64%
  2. Hong Kong: 24.12%
  3. Japan: 10.46%
  4. Singapore: 7.66%
  5. United Kingdom: 6.54%
  6. France: 6.15%
  7. Sweden: 2.60%
  8. Spain: 1.64%
  9. Italy: 1.27%
  10. Netherlands: 0.93%

iShares S&P World ex-U.S. Property Index Fund (WPS, ER: 0.48%)

  1. Japan: 22.48%
  2. Australia: 19.99%
  3. Hong Kong: 15.28%
  4. United Kingdom: 12.49%
  5. Singapore: 6.47%
  6. France: 5.63%
  7. Austria: 4.50%
  8. Canada: 3.49%
  9. Netherlands: 1.89%
  10. Germany: 1.59%
  11. Sweden: 1.53%

The following chart shows the year-to-date performances of these four funds. So far in 2007, the Dow Jones Equity All REIT Index (^DJR) declined 8.26% and the iShares Dow Jones US Real Estate (IYR) dropped 11.05%.


This article was originally written or modified on . If you enjoyed reading this post, please consider subscribing to my full RSS feed. Or you can also choose to have free daily updates delivered right to your inbox.

Author Info

This post was written by Sun You can find out more about Sun and his activities on Facebook , or follow him on Twitter .

8 Responses to “Weak Correlation: Another Reason to Have Global REIT”

  1. shadox |  Oct 23, 2007 at 11:34 pm

    I looked at some of these a while back. DRW and RWX both have a fairly limited history, which bothered me a bit. You are absolultely right about the low correlation. I recently had a post on the subject of low correlation asset classes. Turns out that global bonds and commodities are two other asset classes which consistently exhibit low correlations vs. the S&P.

  2. Presh Talwalkar |  Oct 24, 2007 at 5:52 pm

    Regarding, “To minimize the risk, we want asset classes in a portfolio to have weak correlations (it would be perfect if two asset classes have 0 correction, but that’s unlikely in real life).”

    I really enjoy the technical analysis on this blog, but just to be nit-picky, isn’t a perfectly negative correlation of -1 the best in theory?


  3. Sun |  Oct 26, 2007 at 4:55 pm

    Presh: Yes, in theory a correlation of -1 would be perfect. It means the two will go the opposite directions every time by the same amount. But I think that may not be what we want when it comes investment. If that’s the case, then at the end we will have no gain and no loss either. With the a correlation of 0, the two will go totally random. When one goes up, the other would go either up or down, but has nothing to do with the movement of the first one. I think this is what we want to have for our investments :)

  4. Presh Talwalkar |  Oct 26, 2007 at 5:13 pm

    Sun: Thanks for the reply. I guess this is a point of theory since perfectly negatively correlated assets don’t exist. Plus, they would not do much good since arbitrage would make a portfolio of them return the risk-free rate. So yes, I ultimately agree that high returning investments with as weak correlations are the way to do.