A key tenant of modern portfolio theory is that a diversified portfolio of non-correlated asset classes when regularly rebalanced has long term performance that is superior to the compounded returns of any individual asset class in the portfolio.
This means that a mixture of stocks and bonds can beat the long term compounded performance of stocks or bonds individually. Portfolios with even more asset classes perform even better.
How does this work? Rebalancing. At the end of a year, if one asset class performs very well - perhaps exceeding your target allocation percentage by 20%, you sell some of it and buy more of your other asset classes until your target allocation has been met. You have effectively sold one asset class at a high price and bought the under-represented classes at low prices. How do you make money? Sell high and buy low! Rebalancing does just that.
In simple portfolio allocations, bonds are the asset class that usually, but not always, is outstripped by equities performance. Historically, we have been told to subtract our age from 120 to determine the % of bonds our portfolio should have. With young people often having little bonds - after all, at that age, who wants 5-7% average returns?
Unfortunately, the young investor and those that make recommendations to them confuse average returns with cumulative returns. They are simply not the same.
I recommend investors think of bonds as a form of stored investment capability - like a battery. When the market goes down, you have your bonds - often stable in value - as a source of capital to buy into a cheap market.
Bonds and the entire concept of non-correlated asset classes takes advantage of volatility. When markets push one asset class down, another asset class retains enough purchasing power to buy the depressed asset class at a discount. It's the equivalent of buying food or clothing on sale. Think value investing. However, most of the investment media and many brokerages make recommendations based on average annual performance - they ignore or fail to clearly discuss volatility. The result is that average returns are used to project growth -- the benefit from rebalancing is lost. Furthermore, cumulative investment returns fail to match average annual returns - it's a bar that cannot be met. If a fund drops 10% in one year and 20% the next, it's cumulative value will not equal 100% - it will equal 108%. Rebalancing helps bring the return back to 110%.
Without bonds or other non-correlating asset classes, there would have been no reserve money to buy more when the stock was cheap.
Perhaps the formula to determine the percentage of bonds should not be 120-your age, but rather, what is the expected volatility of your non-correlated investments - is there enough of your non-correlated investment (e.g.) bonds, to rebalance the rest of your portfolio if? If stocks drop 25%, will you have enough allocated in bonds to cover the 25% loss in stock?
Incidentally, the average rebalancing benefit is usually 0.5%. In a low expense mutual fund it could make a very big difference.
I'm sure there are modelling tools available to recommend the ideal % of bonds compared to the volatility of the rest of the portfolio.
My gut says that it should be at least 20% but possibly higher. What is your bond allocation?