Live

Watch CBSN Live

Why Concerns About Diversification Are Overblown

The New York Times recently ran an article referencing a study done by HSBC and Oxford University showing increasing correlation between major asset classes. The implication is that thanks to globalization, diversification across asset classes no longer works.

My first thought on reading the article was that this was another example of why one of my favorite sayings is that what you don't know about investing is the investment history you don't know.

Students of investment history know that correlations aren't constant -- they drift. Correlations rise when the news (meaning new information) is dominated by events that affect the equity and bond markets of different countries in similar ways (such as oil embargoes, wars, the events of September 11, 2001 or the financial crises surrounding the failures of Long-Term Capital Management and Lehman Brothers). When crises are resolved, correlations tend to return to their historical norms as returns are more impacted by the idiosyncratic risks of various countries, regions, etc.

Another key point is that the most important diversification is to high-quality fixed income, making sure your allocation is high enough to dampen the risk of your overall portfolio to an acceptable level -- because you can be virtually certain there will be future crises that will cause correlations of all risky assets (not just equities) to rise toward one. Risky fixed income investments -- such as high-yield bonds, convertible bonds and emerging market bonds -- see their correlations to equities dramatically increase during crises. When that happens, the correlation of the safest fixed income investments tends to fall.

For example, for the period 1926-2010, the annual correlation of the S&P 500 Index to long-term Treasury bonds was 0.03. However, when the correlations of risky equities were rising from 2000 through 2010, the correlation of long-term Treasuries to the S&P 500 fell to -0.80. This explains why making sure your portfolio has a sufficient allocation to the highest quality fixed income assets is the most important diversification of all. This point can't be emphasized enough.

The Data To test the hypothesis that globalization has led to a reduction in the benefits of global diversification we can examine the historical correlations of returns between the S&P 500 Index (a domestic large-cap index) and the MSCI EAFE Index (an international large-cap index). I ran the data from 1970 through 2007, to see if there was this trend to higher correlations before the financial crisis of 2008 hit. I then split the 38-year period into two equal 19-year periods to see if there is a trend toward rising correlations.

  • For the first period, 1970-88, the annual correlation of the S&P 500 to the MSCI EAFE was 0.623.
  • For the following period, 1989-2007, the annual correlation actually fell slightly to 0.614.
We can also examine the correlation data of international small-cap stocks over the same periods. Historically, international small-cap stocks have had lower correlation to U.S. stocks, making them superior diversifiers of the risks of domestic equities.
  • For the first period, 1970-88, the annual correlation of the S&P 500 to international small-cap stocks was just 0.459 -- quite a bit lower than the 0.623 correlation of the MSCI EAFE for the same period.
  • For the following period, 1989-2007, the annual correlation fell to 0.374 -- again, much lower than the 0.614 of the MSCI EAFE for the same period.
As you can clearly see, there hasn't really been a trend toward rising correlations -- the benefits of global diversification are as strong as ever. It's just that the market's movements over the most recent past have been driven mostly by systematic risks that impact global stock and bond markets in similar ways. I thought it was interesting that the study referenced by the New York Times article covered all of five years. In other words, we can explain all the noise about a new paradigm as just another case of recency -- the tendency to give too much weight to recent experience, while ignoring the lessons of long-term historical evidence.

More on MoneyWatch:
First Quarter Update on 2011's Sure Things Can You Find the Future Winners? John Bogle: Can His Son Beat the Market? What Would Happen if Everyone Indexed? Quest for Alpha: What You Need to Consider When Handling Your Own Investing
Three ways I can help you become a wiser investor: