Another theory of predicting future stock market returns states that there are three main components to long-term stock market performance. Amongst many others, I learned this from authors and investors Jack Bogle and William Bernstein.

Part 1: Dividend Yield
If your stock distributes 2% in dividends each year, then you will have a 2% contribution towards of return. This is what dividend investors love to see coming in each quarter, and is relatively easy to track for a large group of companies. Here it is over time for the S&P 500, courtesy of Multpl.com:

Part 2: Earnings Growth
If earnings stay constant, then all other things equal, one would expect the share price of your company to stay constant as well. If the earnings grow by 5% every year, then your share price will grow by 5% per year. Thus, earnings growth rate is a vital component of total return.

If your portfolio was all of the stocks traded in the United States, like that of a broad-based index fund, this would create a connection between the growth rate of the nation’s Gross Domestic Product and the earnings growth rates of all US companies. In other words, the fundamental return is based on GDP growth. In turn, the GDP growth rate is connected to population growth and productivity per person.

These two parts added to together are coined the fundamental return:

Some bad news: Now, from 1950-2000, fundamental returns were 10%: 4% dividend yield and a 6% earnings growth rate. These days, the S&P 500 has a dividend yield of only about 2%. Earnings growth rate estimates are subject to debate, but they hover around 5-6%.

Part 3: Changes in P/E Ratio
The price-to-earnings (P/E) ratio is the price per share divided by earnings per share. In other words, it is how much investors are willing to pay for each unit of earnings. If they are willing to pay 20 times annual earnings, the share price of the stock will be twice as high as if they only paid 10 times earnings. This part is denoted the speculative return, as it has changed throughout history. Here it is again for the S&P 500:

In 1950, the P/E ratio was less than 10. As of right now in mid-2010, it is 20. It is very unlikely that this more than doubling of price-per-share will happen again, with the historical average being around 15. (During the dot-com bubble, the P/E ratio was over 40. In 2008, it was over 25.) This will lead to a zero, and quite possible negative, future speculative return!

Summary

Source: Little Book of Common Sense Investing, Exhibit 7.1

When predicting future returns, you have to look at all the sources of those expected returns. Fundamental return is still a solid reason why stock prices will go up on the long-term, especially if you are not investing only in one country or economy. Some people call it a belief in capitalism, that economic growth will continue and GDP will continue to increase. I simply believe in the passion and motivation of all the people out there, from Sweden to China to Brazil. However, there is good evidence that you might not be getting 10% historical returns due to P/E ratio contraction.

In a recent column, Larry Swedroe shares that the forecasts that he has read are predicting a 5% total annual growth in earnings and 2% dividends for a total return of 7% (similar to above). Inflation is predicted at 2.5%. However, he points out the current minimal-risk return is pretty low as well, so you need consider the big picture:

The bottom line is that while the expected nominal return to stocks is lower than the historical return, so is the expected return to Treasury bonds. You should decide if the expected risk premium for stocks is sufficient given your unique ability, willingness and need to take risk.

Every year, the low-fee investment advisor Portfolio Solutions, LLC founded by Rick Ferri provides a 30-year market forecast based on their analysis of several factors. In their own words:

Each year, we analyzed the primary drivers of asset class long-term returns including risk as measured by implied volatility, expected earnings growth based on expected long-term GDP, market implied inflation based on the spread between long-term Treasury Bonds and TIPS, and current cash payouts from interest and dividends on bond and stock indexes. These factors plus others are used in a valuation model to create an estimate for risk premiums over the next 30 years. In a sense, we believe these expected returns reflect what the market is estimating will be a fair payment for each asset class over T-bills over the long-term.

You can view all the asset classes on their site, but I have included some of the major ones below for preservation and reference. Another set of estimates to throw into the mix.

*The estimate of risk is the estimated standard deviation of annual returns.

As part of gathering the data needed to go beyond net worth, I’ve shown ways to track find your personal savings rate by tracking your current spending with your current after-tax income. For now, I’m skipping ahead to estimating your portfolio’s long-term investment returns.

There are a lot of ways to estimate future stock returns. You’ve probably heard of the P/E ratio, which is usually the price divided by last year’s earnings. This is one measure of “value”. Here is a plot of historical values of the inflation-adjusted S&P 500 index, along with its annual earnings (source).

Historical Price & Earnings Separately

Historical P/E Ratio

As you can see, there is a lot of volatility in P/E ratio. The “P/E 10” ratio is the share price divided by the average earnings over the last 10 years. By taking a long-term average, you smooth out the noise and bumps.

Professor Robert Shiller of Yale University, which provided the above data as well, spoke about the usefulness of this ratio in his book Irrational Exuberance, and provided the data for the following chart. The x-axis shows the real “P/E 10” of the S&P Composite Stock Price Index (inflation adjusted price divided by the prior 10-year mean of inflation-adjusted earnings). The y-axis shows the geometric average real annual return of the same index, reinvesting dividends, and selling 20 years later.

Price-Earnings Ratios as a Predictor of Twenty-Year Returns

You can definitely see the relationship that a higher P/E10 usually leads to a lower future return. However, there is still a great deal of scatter for any given P/E10 ratio.

What about today? As of June 2, 2010 the P/E10 is 19.99 with the S&P 500 index at around 1,098. Historical average is about 15. Back in March 2009 when everything was looking bleak, the P/E10 was 13.32. (P/E10 is also referred to as Cyclically Adjusted Price Earnings (CAPE) ratio.)

The Early Retirement Planning Insights website provides a calculator that forecasts future returns based on the current value of P/E 10, again using historical returns. I’m not sure exactly how they did all the regression. Here’s their return forecast for a P/E 10 ratio of 20.

Future Returns Prediction (P/E10 = 20)

For a 20-year forecast, it shows an average outlook of 4% returns, on a real (after-inflation) basis. Of course, the range indicates that the actual returns could look much worse. As the time-horizon lengthens, the range gets smaller due to the phenomenon of reversion to the mean. Kind of scary to look 60 years ahead!

Warnings

To me, this stuff is useful as a general Big Picture planning tool, not as a short-term trading tool. These are just predictions based on the past, which is often the best we can do, but still far from perfect. Many people use P/E10 as a tool for market timing, shifting their asset allocation as it rises and falls. Shiller himself states that his plot above:

confirms that long-term investors—investors who commit their money to an investment for ten full years—did do well when prices were low relative to earnings at the beginning of the ten years. Long-term investors would be well advised, individually, to lower their exposure to the stock market when it is high, as it has been recently, and get into the market when it is low.

That may be true, but market timing systems can really test an investor’s faith when they seem to be wrong for a long period of time. On a personal basis, I’d probably limit any asset allocation moves – if any – to if the P/E10 ratio moved to an extreme, for example dropped below 10 or above 25.