How the Stock Market Projection Model Works

Is it possible to forecast the stock market return for the next 10 years? If so, how do you do it? In Projecting Stock Market Returns we saw the results of a simple forecasting model that seemed to produce promising results. Below is the resulting graph introduced and discussed in that post (click to expand). In this post, I'll describe how the model estimates future returns.

Projecting 10-Year Stock Market (Dow) Returns

Stock Market (DJIA) Forecast Methodology

We have consistently seen evidence that
there is a strong relationship between normalized price/earnings ratios and subsequent stock market returns (see links at the bottom of this post).  Therefore, it is not surprising that p/e ratios play a central role in this forecasting methodology.

The methodology depends primarily on two critical assumptions:

• The long-term growth rate of corporate earnings is relatively consistent.
• The normalized price/earnings ratio tends to revert to the mean.

Given the above, the model simply

• projects normalized earnings for the next 10 years,
• prices those earnings at the average normalized price/earnings ratio (NPE), and
• calculates the return for a theoretical investor who buys at the current price and sells 10 years from now at the forecasted price, reinvesting dividends in the interim.

To predict future returns accurately, the model would need to accurately predict the earnings multiple (NPE) investors will be willing to pay 10 years from now. I doubt that is possible. As a result, I think it is more realistic to think of the results not so much as stock market predictions as stock market projections. There is no real expectation that earnings growth will be exactly the same as the historical average growth rate. Nor is there an expectation that the normalized price/earnings ratio 10 years from now will exactly match the historical average NPE.  Instead, the methodology simply says that IF earnings grow at the average rate experienced in the past, and IF the NPE ratio applied by investors is the average NPE experienced in the past, THEN this is the return investors would receive.

In effect, the model establishes a benchmark reflecting what might be called average performance given the current circumstances.  The benchmark can be used "as is" -- as a simple-minded, mechanical forecast.  Or, it can be used as the starting point for a more refined estimate incorporating qualitative judgments.

The Current Projection

My objective is not to predict short-term stock market direction or performance. My database suggests that short-term, year-to-year DJIA (Dow Jones Industrial Average) price movements are dominated by changes in the normalized price/earnings ratio. Trying to predict those changes would be incredibly difficult, if not impossible. On the other hand, growth in dividends and normalized/sustainable earnings is normally the primary determinant of performance over the long term. Trying to forecast those appears to be a much more manageable task.

Click here for the most recent projection.

Related Posts

Projecting Stock Market Returns: the first post in this series.
The most recent Stock Market Projection

Following are links to the most important posts contributing to this methodology:
100 Years of Stock Market History: 100-year graph of Dow closing prices seems to show long periods of excitement alternating with "long flat periods."
Analyzing & Understanding 100 Years of Stock Market History: Introduces spreadsheet/model that decomposes long-term price changes into contributions from a) normalized earnings (NE) growth, b) dividends, and c) change in the normalized price/earnings ratio (NPE).
About P/E, Normalized Earnings and Normalized P/E Ratios: Describes how I normalize earnings and calculate normalized price earnings ratios.
Stock Market Earnings Growth History and a Worst Case Scenario: Derives a long-term earnings growth rate of 5.5 - 6%.
Dow P/E Ratio History since 1929: Demonstrates the long-term volatility of NPE compared to the relatively stable growth of dividends and normalized earnings, and suggests significant impact of NPE on price.  Shows that earnings & dividends continue to increase; NPE does not.
The Extraordinary Impact of Price/Earnings Ratios: Suggests that year-to-year changes in stock market prices are dominated by changes in NPE rather than changes in dividends or normalized earnings growth.
Earnings, Dividends Determine Long-Term Returns: shows that for 50-year periods the impact of dividends and earnings growth dominates the impact of NPE.
Rolling Returns vs. P/E Ratio Graphs: Strongly suggests an inverse relationship between 10-year return and NPE; high NPEs are associated with low 10-year returns, and vice versa. Also shows that NPE ratios are cyclical.
Starting P/E Ratio vs. 10-Year Returns: Shows the 10-year returns that resulted from each initial P/E ratio; shows returns decrease as P/E increases.
Borrowing Returns from the Future: long periods of high returns are followed by low returns.
Dow P/E Ratio Impact on Future Returns - A Summary: Compares 10-year returns of investors buying in high p/e ratio markets with those buying in low p/e ratio markets.
Major Bull and Bear Markets since 1900: Demonstrates that secular market peaks have occurred when NPEs were in the top 20 percentiles; secular troughs have occurred when NPEs were in the bottom 10 percentiles.

For a complete list of stock market posts, see the subject index in the sidebar on the left

Share This Article

Bookmark this on Delicious
See below to share via Facebook, Twitter, etc.

This work is licensed under a Creative Commons Attribution 3.0 unported license.

Last modified: 8/13/2011

Note: In 2011 I made a slight modification to the methodology that tends to moderate the highest and lowest projections. For example, the original 1999-2009 projection was changed from -2.5% to -1.1%. The graph from the original backtest is reproduced below.