Does Diversification Still Work?

2008 saw all risky assets fall in value. You may have wondered if diversification still works. The answer is yes -- though not every year. To illustrate that point, let's take a trip back in time.

It's Jan. 1, 1978, and you've been given the gift of perfect foresight -- you can forecast returns and volatility with 100 percent accuracy. Your crystal ball provides the following information on four major asset classes -- U.S. stocks, international stocks, real estate (through REITs) and commodities -- through 2009. (Please note that we're using the 31-year time frame because that's how far back the REIT index goes.)

However, before jumping to those conclusions, let's examine how an equally weighted portfolio of the three equity asset classes (S&P 500, MSCI EAFE and REITs) would have performed. With annual rebalancing, such a portfolio would have returned 11.82 percent per year with a standard deviation of 16.08. The result was a Sharpe ratio of 0.459, 0.041 (or 10 percent in relative terms) higher than the Sharpe ratio of the best individual performing asset class. Clearly, diversification worked.

Adding Commodities
Now let's examine how adding commodities impacted the portfolio. With annual rebalancing of the equally weighted portfolio (25 percent in each of the four asset classes), it would have provided an annualized return of 11.43 percent with volatility of just 14.68 percent. The result was a Sharpe ratio of 0.465, 0.047 (or 11 percent) higher than the Sharpe ratio of the best individual performing asset class. It was also an improvement over the Sharpe ratio of the all-equity portfolio.

Other Combinations
The equally weighted four-asset-class portfolio produced a higher Sharpe ratio than any equally weighted combination of the other asset classes (either 50/50 of any two asset classes or 33 1/3 of any three asset classes), with one exception. (An equally weighted, 50/50 portfolio of the S&P 500 and REITs produced both the highest return (12.14) and the highest Sharpe ratio (0.486).)

Unfortunately, no one in the real world has the perfectly clear crystal ball that would tell you in advance which portfolio would produce the best risk-adjusted return. But even this exception shows the benefits of diversification -- while the S&P 500 Index produced lower returns than did REITs, its addition to an all-REIT portfolio produced higher returns, exhibited less volatility and produced a higher Sharpe ratio than did REITs.

What lessons can we learn? First, we shouldn't make the mistake of considering the risk and returns of an asset class in isolation. Second, while diversification doesn't work every year, it works over the long term.