The Stock-Return Predictor reveals the effect of the stock valuation level that applies at the time of a stock purchase on long-term returns.
This new calculator tells you what return you can reasonably expect at various time-periods from an investment in the S&P stock index, presuming that stocks perform in the future much as they have in the past. The results are expressed in terms of real, annualized total (that is, with dividends reinvested and without additions or subtractions to principal) returns.
The purpose of the calculator is to help you set your stock allocation strategy. Stocks generally offer a stronger value proposition at times of low or moderate valuation than they do at times of high valuation. By comparing the long-term returns likely to follow from stock purchases made at today’s stock valuation level with the long-term returns likely to follow from stock purchases made at a lower or higher valuation level, you can put together a stock purchase strategy likely to aid you in becoming a true Buy-and-Hold stock investor. Predicting stock returns by making reference to stock valuation levels prepares you for what is to come in the years ahead.
AUDIO: Rob’s Financial Freedom Insight #2 — Knowing Your Stock Return in Advance Takes the Risk and Emotion Out of Stock Investing
The Stock-Return Predictor results are based on John Walter Russell’s regression analysis of what the historical stock-return data shows regarding the effect of stock valuation on long-term stock returns, published at the www.Early-Retirement-Planning-Insights.com web site. Russell’s research grew out of The Great Safe Withdrawal Rate Debate, an ongoing exploration of the stock valuation question that has generated intense controversy at half a dozen Financial Freedom Community discussion boards for four years now, beginning with a post put to a Motley Fool board by Rob Bennett on May 13, 2002. Those discussions also led to development of the Valuation-Informed Indexing approach to investing, examined in articles posted by Bennett to the www.PassionSaving.com web site. The new calculator taps into the insights developed during those discussions in an effort at predicting stock returns. The default results set forth above show the returns you can expect for stocks purchased at the stock valuation levels that now apply (the default setting is adjusted monthly). To see how the value proposition of a stock purchase changes with upward or downward movements in stock valuation, please move the sliding control button (the “slider”) to higher or lower S&P price levels. For example, if you want to see what your expected returns over various time-periods would be if the price of the S&P 500 were to fall by 20 percent, move the slider to a price level 20 percent below the level that applies for the default results.
Moving the slider to a new S&P price level brings up the results that apply for stocks purchased at the P/E10 (P/E10 is the price of the index divided by the average of the past 10 years of earnings–Robert Shiller has long argued in favor of P/E10 as an effective means of assessing valuations, and Russell’s research offers support for Shiller’s position) level that applies at that S&P price level. An alternate way of viewing the results that apply for different stock valuation levels is to directly move the slider to different P/E10 levels.
The article “Valuation Ratios and the Long-Run Stock Market Outlook: An Update,” by John Y. Campbell and Robert J. Shiller, provides background on the P/E10 tool and its use in making reference to stock valuation levels in predicting stock returns. P/E10 fell to between 5 and 6 in 1921 and 1932, and to between 6 and 7 in 1922 and 1982 (the starting-point of the huge bull market). The highest recorded P/E10 value is 44 (reached in early 2000, perhaps the ending point of the huge bull market). A moderate P/E10 value is 14. P/E10 exceeded 24 in 1928, 1929, 1930, 1966, and in all the years from 1995 through early 2006. The calculator shows that the stock valuation level that applies at the time a purchase is made is an important factor in predicting stock returns obtained in 10 years, 20 years or 30 years.
There is a 50 percent chance that the real-world return will be less and a 50 percent chance that the real-world return will be more than the returns identified by the calculator as “Most Likely,” according to Russell’s regression analysis of the historical stock-return data. There is only a 20 percent chance that the real-world return will be greater than returns identified as “Lucky” and only a 20 percent chance that the real-world returns will be less than returns identified as “Unlucky.” There is only a 5 percent chance that the real-world returns will equal or exceed the returns identified as “Best Possible” and only a 5 percent chance that the real-world returns will equal or fall short of the returns identified as “Worst Possible.”
Predicting stock returns is an inexact business. The Stock-Return Predictor results are the product of research resting on the assumption that stocks are likely to perform in the future much as they have in the past. It is of course not a certainty that this will prove to be the case. In any event, there are almost certain to be at least some ways in which stocks will perform in the future in ways different from how they have in the past. Moreover, it is important to understand that, while changes in stock valuation affect long-term returns as a matter of “mathematical certainty” (those are the words of William Bernstein, author of The Four Pillars of Investing ), predicting stock returns in the short-term (time-periods of less than 10 years) is extremely difficult, if not outright impossible.
The calculator is intended to serve as a guide to how changes in stock valuation may affect long-term stock returns. The authors of the calculator believe that it is a useful tool for predicting stock returns. But they also stress that the study of how changes in stock valuation affect long-term stock returns is very much an ongoing effort. The calculator’s authors (John Walter Russell and Rob Bennett) very much do not want any users of the calculator to read into the “predictions” offered any more certainty than is warranted by the nature of the research project that produced it.