|
Lost Decades and P/E10
(Updated 6/6/10)
Introduction Over the long run, stocks have outperformed bonds. However, over shorter periods like 10 years, bonds have sometimes beat stocks. When stocks sport high valuations, they tend to generate sub-par returns. This article will study this subject in detail. Stocks and Bonds Large stocks, as represented by the S&P 500, have historically outperformed bonds. I use Intermediate-Term Government Bonds (5-year Treasuries) to represent bonds.
However, over shorter time periods like 10 years, bonds have sometimes outperformed large stocks. Starting with 1926 to 1935 and ending with 2000 to 2009, there have been 75 rolling 10-year periods. In twelve of those 10-year periods, bonds beat large stocks (the “Lost Decades” for stocks). These are shown as red bars in the following chart:
Stock Valuations P/E10 is a measure of S&P 500 valuations. It is derived from the average of CPI-adjusted earnings over the previous 10 years -- the “E10” part of the term. The “P” term is the current price of the S&P 500. This method attempts to smooth market cycles.
The chart above is from a spreadsheet on Robert Shiller’s website. Another handy website tracking P/E10 is www.multpl.com The following chart overlays P/E10 (blue line) onto the “Large Stocks minus Bonds” chart.
The above chart can be converted to a scatter plot:
The following table takes the data from the chart above and organizes it by valuations. “Plus Delta” counts the number of 10-year periods when stocks beat bonds. “Minus Delta” counts the number of 10-year periods when bonds beat stocks. And “Percent Plus” is the percentage of time when stocks beat bonds over 10-year periods.
And from the table, I created a histogram.
So while the number of data points thins out appreciably at P/E10 levels above 25, the charts nonetheless reveal that the odds of large stocks beating bonds over 10-year periods have diminished as valuations rose. Small Stocks While the preceding analysis applies to large stocks, results for other equity asset classes like small stocks may differ. The next chart shows the “Lost Decades” where bonds beat small stocks over 10-year periods.
So while small stocks had a similar number of “Lost Decades”, they were not the same decades as the ones for large stocks.
In addition, we see a weaker relationship between ‘small stocks minus bonds’ versus P/E10 (a smaller r-squared) than exists between ‘large stocks minus bonds’ versus P/E10. Conclusion Although stocks have outperformed bonds over the long term, there have been shorter periods (like 10 years) when bonds have beaten stocks. The charts presented in this article have shown that there has been a relationship between P/E10 and the number of times large stocks have beaten bonds over 10-year periods. Generally, the odds of large stocks beating bonds over 10-year periods have diminished as valuations rose. However, this relationship appears weaker for other equity asset classes like small stocks. Readers who are swayed by this analysis, may want to be cautious when P/E10 rises above 20. And may even consider lowering their stock allocation when P/E10 levels exceed 25. |
