If you’ve done any reading about investing, you know that stock market returns are unpredictable over short periods and predictable over long periods.
And you also know that–over long enough periods (however long those may be)–stocks outperform bonds.
However, until I started reading Jeremy Siegel’s Stocks for the Long Run, I’d never heard anyone make the case that–over extended periods–stocks not only earn greater returns than bonds, but more predictable returns as well.
In other words, not only do stocks earn more than bonds, they are actually safer investments for long-term investors.
How does Siegel reach this conclusion?
In short, he compares the range of after-inflation returns of stocks to the range of after-inflation returns of bonds over periods of various lengths. Unfortunately, the data in the edition I have only goes through 1997, and I was curious to see how the conclusion would hold up after updating for the last decade.
Time to pull out the spreadsheet and plug in some numbers!
Updated for 2008: Are stocks less risky than bonds?
The chart below shows us the best and worst stock market returns (after inflation) over periods of various lengths. As you can see, the real return from stocks becomes much more predictable as you look at longer periods.
For example, the worst 1-year real return for stocks was -37.7%, but the worst 10-year real return for stocks was a compounded -4.6%.
The next chart shows the same thing, but for bond returns (as measured by the total return on 10-year U.S. Treasury bonds). As with stocks, the returns become more predictable as we look at longer and longer periods.
The following chart essentially combines the previous two, allowing us to compare the range of bond returns (from best to worst) to the range of stock returns over periods of varying lengths.
Over short periods, the range of stock returns is much greater than the range of bond returns. For example, the worst 1-year real return for stocks was more than 90% worse than the best 1-year real return for stocks. In contrast, the worst 1-year return for bonds was just over 40% worse than the best 1-year return. (This is–in part–why people refer to stocks as “risky.”) But look at those 30-year ranges! The range of after-inflation returns for stocks actually becomes lower than the range for bond returns. In other words, stocks are more predictable (ie, safer) than bonds over periods of 30-years.
(If you compare our first chart to our second chart, you can verify this on your own: Over 30-year periods, stock returns fall into a narrower range than bond returns.)
And for any statistically-inclined readers, our final chart shows the standard deviation of real returns for stocks and bonds. Again, we see that 30-year real returns for stocks are more predictable than 30-year real returns for bonds.
What can we learn here?
If you’ve got a 30-year investment time frame a heavy allocation to stocks just makes sense–not just from a total return point of view, but from a risk point of view as well.
The catch: You can’t bail out and sell when the market drops. Otherwise you don’t get those nice, predictable 6-7% after-inflation returns.
Notes on the data:
The period considered is 1928-2008. The yearly returns are for calendar years.
Bond data and inflation data comes from the Federal Reserve Bank of St. Louis’ research site.
Stock return data is from Aswath Damodaran, Professor of Finance at the NYU Stern School of Business.