Investment Returns over Multi-Year Periods

It’s all about your time horizon.

In the last few posts, I have quoted some statistical returns for stocks, bonds and inflation, but only looking at single-year periods. It turns out those figures are not very helpful, since we all invest our retirement assets for some longer-term period, our “time horizon” in investment industry jargon. How many years will it be before you will have to cash in your investment and spend the proceeds? The longer the time horizon, the more stocks have done better than bonds.

Over longer periods of time, inflation poses a greater risk than depreciation in stock value. Conversely, the shorter your time horizon, the less appropriate stocks are. The risk of a short run loss in value over one or two years is too great to make it worth trying to strive for their greater returns.

The following graph shows how stocks and bonds have fared historically over two, five, 10, 20, and 30 year periods. The blue bars show the range of annualized real (i.e., after-inflation) returns experienced by stocks (as represented by total return on the S&P 500 index) and bonds (as represented by total return on intermediate term Treasury bonds) during the period 1926-2008. Inflation is represented by changes in the Consumer Price Index. The black bar shows the median return for all periods during the 83 years studied.

This graph paints an important picture. Notice that stocks have performed well over longer periods of time. In fact, they have outperformed bonds in every 20-year and 30-year period from 1926 to 2008. According to author and finance professor Jeremy Siegel, in his book Stocks for the Long Run, not since 1831-1861 have fixed income investments outperformed stocks over a 30-year period. And that remains true for the 20- and 30-year periods punctuated by the positively horrible 2008.

Another important thing to notice from the graph is how the wild swings in returns, in both stocks and bonds, are increasingly muted as your time horizon expands. Time erases your losses! Unfortunately, time also erases your outsized gains. The longer the period, the smaller the range of annualized returns. For example, annualized real returns in stocks over short (2-year) periods ranged from a whopping gain of 43% to a sickening loss of 29%; a spread of 72%. But over 20-year periods, the range of annualized real returns in stocks shrank considerably—to a high of 13% and a low of 0.85% (a mere14% range).

A third thing to notice is how the median return stays roughly the same no matter how long or short the period. Moreover, as the range of returns shrinks with ever-longer time horizons, the median becomes much more useful as a tool for projecting the future, since your future is more likely to be closer to the median than with shorter periods.

After you have experienced a year like 2008, it is a bit comforting to remember that time can heal your outsize losses. But it is equally critical to remember that if you invest your retirement assets in stocks, you can always expect to experience many individual one-year periods of significant real losses (one out of three calendar years between 1926 and 2008, as seen in yesterday’s post). And often those one-year periods follow on each others’ heals, making for quite the sickening ride. It takes courage and conviction to get through the short-term downs in order to benefit from long-term gains.

Now here’s the ugly part. When it comes to retirement savings, defining your time horizon is particularly complex. That’s because you won’t be spending all your money at once. Rather, if for example you are ten years away from retirement, you have a ten-year time horizon on a small piece of your investments, an 11-year time horizon on another small piece of your investments, and so on and so on. Nothing’s easy, is it?