Last Updated May 11, 2009 8:39 PM EDT
Let's use the same 20-year period as Patrick (ending March 2009), and start you with a 100 percent equity portfolio, represented by the S&P 500. During this period, the portfolio returned 7.4 percent per year. It's true that long-term government bonds outperformed the S&P 500 during this time, but why would you have guessed that? In the previous 20 years, stocks outperformed bonds nearly two to one -- 6.8 percent versus 3.5 percent.
Instead, consider the effects of diversifying the portfolio. By making the portfolio 60 percent S&P 500 and 40 percent long-term government bonds, you'd raise the return to 8.7 percent per year while lowering the portfolio's volatility by nearly one-third (14.9 percent to 10.0 percent).
Stretching the time period to 30 years gives you an even better perspective of the benefits of diversification. For the 30-year period ending March 2009, the S&P 500 returned 10.3 percent per year, and long-term government bonds returned 10 percent per year. However, a portfolio of 60 percent S&P 500 and 40 percent long-term government bonds did better than either alone, with an average return of 10.6 percent per year. And it even did this with less volatility than government bonds alone, with a standard deviation of 11.1 percent versus 11.5 percent for the bonds.
This is why it's important to never consider the returns of an asset class in isolation. Instead, asset classes should also be judged on their diversification benefits and the effects additions would have on your portfolio.
Further reading: Another MoneyWatch colleague, James Picerno, gives you the full Dish on Diversification.