Sometimes it is helpful to look at history. In evaluating what might be in store for the U.S. stock market, this is one of those times. I have selected the DJIA (Dow Jones Industrial Average) for analysis because data is available from before 1900.
The graph below shows the 10-year and 20-year moving averages for annual returns (without dividends) for the DJIA. 
Since 1950 the 20 year moving average has trailed the 10 year moving average very consistently. The average trailing time is 3-5 years except near extreme highs and lows where the two are more nearly coincident.
Before 1950 the two moving averages spent more than three times as long below the 107 year compounded average return of 5.3% than above. Since 1950 the 10 year moving average has been above the 5.3% average return for 41 years and 17 years below. This is almost the reverse of the first half of the twentieth century. From 1950 to 1965 we spent 16 years above the average; in 1966 and 1967 the 10 year moving average oscillated below and above the average; from 1968 to 1983 we spent 16 years below the average. From 1984 to 2007 the 10 year moving average has spent 24 years above the 5.3% level. Do we have a long period (say 24 years) with the 10 year moving average below the long-term 5.3% average annual return? If so, the coming years could look something the 1970’s for U.S. stocks. In the 1974 to 1982 period, there were six years with the 10 year moving average return was near zero or negative. There were only two years above zero with low single digit 10 year average returns.
A similar discussion can be held for the 20 year moving average with similar conclusions. A reversion to the mean for behavior of U.S. stocks would argue that for 2008 to 2065 we should reverse the behavior of 1950-2007 and spend more years with the 10 year moving average below 5.3% than above.
Although it is not plotted, the trend line for the 10 year moving average return has an upward slope for DJIA returns from 1900 through 2007. If that were to continue, projected returns would improve marginally over the next 57 years, but would not change the argument that reversion to the mean (now an upwardly sloping trend line) would produce below average returns for many years forward. Using this trend line rather than a level 107 year average would require that one accept that there are reasons why the average annual return trend line should continue the upward slope in the future. It is hard to argue that there are fundamentals to support this supposition.
Bottom line: be prepared for an extended period of many years with average annual returns that are less than 5%, with more than a few years of negative returns. There is an old saying: “Those that do not learn from history are doomed to repeat it.” Enjoy the good years, but the patterns of history indicate there are fewer good years coming than we have seen in the second half of the twentieth century.
