On The Predictive Capacity Of The CAPE Ratio To Determine Future Prices

Includes: CAPE
by: Paul Boland


Within this article we look at several regression models that show how to utilize Cyclically Adjusted Earnings over various time-frames to help predict the future price of the S&P 500.

We find that the current state of the market has historically been more correlated with 12-month trailing earnings, but that future prices are better explained through more long-term based metrics.

An investor utilizing the cyclically adjusted earnings paradigm can benefit through this asymmetrical information gap and consequently benefit greatly.

A link to the regression models in excel is also provided for your own download and for your own use and analysis.

Generally, the S&P 500 is valued by conventional investors through the utilization of the standard 12-month trailing earnings P/E ratio. This ratio for the S&P 500 currently stands at roughly 19 - 20 which does not indicate an extreme high for the general pricing of the index. However more recently there has been various economists whom have indicated that it would be better to analyze the market through a 10 year cyclically adjusted P/E ratio otherwise known as the CAPE Ratio.

The concern for most CAPE Ratio adherents is that the CAPE shows the market as somewhat overvalued as the CAPE for the S&P 500 is around 26 (and should be interpreted in the same way as the classical P/E).

So I decided to do several regression models in order to see what the price of the market ought to be based on both 1 year of earnings and 10 years of earnings adjusted for inflation (or Cyclically Adjusted). I utilized Dr. Robert Shiller's data set which you can download from his website right here. With this as the original data set, I proceeded to insert columns for both 10 year earnings, and 1 year earnings. I then proceeded to add columns for 1 year future average price for the S&P 500 and 3 years, 5 years, and 10 years.

At the following link you can download our analysis in excel for more of your own analysis. However here are a few highlights:

Does an average of the previous 10 years of earnings help explain the current price of the S&P 500 better then just a twelve month trailing earnings?

The answer is no as you can see from this chart:

The R^2 of the 10 year averaged earnings regression model helps explain the price of the S&P 500 at a rate of about 76.5% accuracy while the twelve month trailing earnings helps explain the price action of the S&P 500 about 78.5% of the time. This can be explained through the conventional investors usage of the 12 month trailing earnings metric versus the more unconventional 10 year averaged earnings.

How well does 10 year averaged real earnings help explain the future price action of the S&P 500 versus twelve month trailing earnings?

The short answer - over the long-term the 10 year averaged earnings help explain future prices better.

Here is the chart:

As could probably be expected, during the first year the R^2 for one year of earnings as an explanation of the future price action for the S&P 500 was better versus 10 years of earnings. However the average price of the S&P 500 for the subsequent 3-year, 5-year, and 10-year period was better explained by the 10 year average earnings versus 1 year.

From these two charts we can gather two separate pieces of information:

A) That the markets current pricing or valuation is better explained through earnings during the previous 12 months.

B) That the markets future average price is better explained by 10 years of averaged earnings adjusted for inflation.

This supports the central thesis of most people whom invest based on the CAPE Ratio as the CAPE Ratio utilizes Cyclically Adjusted Earnings from the previous 10 years as the denominator in the commonly used P/E ratio. The idea behind it, is that it explains the future price action of the S&P 500 better then one year of earnings. This should probably be rather intuitive since a single year is hardly indicative of value in the market. However since an investor operating under the CAPE Ratio paradigm is investing today against market participants whom are investing under the paradigm of the single year metric, they stand to gain as long as they invest for the long haul.

In any case, the two types of regression models actually reflect rather differently on where the S&P 500 will be in the future. See the following table:

So we have 10 year averaged earnings currently at 76.12 and with a CAPE value of over 26 with the S&P 500 at around the mid 1900's. Within the historical context there is very little chance of investing in the S&P 500 and seeing a return during the next 10 years. Even with a single year of earnings data, we are not looking at much of a return over the long-term either. So despite the discrepancy in the level of accuracy between the two methods, they are essentially telling us the same thing. That the market is getting overvalued and if the S&P 500 continues to perform as it has since 1871, then we are looking at small returns over the next 10 years.

With all that in consideration, I tend not to be a pessimist. The fact that the market has been more volatile lately over the last 20 years and that the average CAPE has been quite high during the last 20 years would seem to indicate that the market might perform in a manner inconsistent with the historical relationship of earnings and price (although doubtful, we can always hope). However I would be very careful investing currently. I would look at investing more in individually undervalued stocks and possibly globally versus an S&P 500 index.

You are more then welcome to download the excel and take a look at it all for yourself. Feel free to do your own analysis over shorter time frames to see if the market is overvalued or undervalued within the context of only 25 or 50 years.

Happy Investing!

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.