Many well-known stock market pundits to include the likes of Henry Blodgett point to Professor Robert Shiller’s cyclically adjusted PE ratio or CAPE to make a case that the stock market (S&P 500) is overvalued. According to Prof. Shiller, the long-term average PE ratio (CAPE adjusted) for the S&P 500 is approximately 16. And today, according to CAPE (cyclically adjusted PE) the S&P 500 sits at a CAPE PE ratio of approximately 24 or so. Therefore, according to the statistical analysis, the S&P 500 is 20% to 30% overvalued.
The thesis of this article is that although Prof. Shiller may eventually end up being correct, and we emphasize “may”, his hypothesis is vague and imprecise, and therefore, too general to be of value. On the other hand, we emphatically hold that he is absolutely wrong regarding specific blue-chip US corporations. Furthermore, we also contend that on an absolute basis (precise calculation of earnings) that the S&P 500 is undervalued based on historical norms.
Therefore, we are concerned that these statistically inferred measurements of so-called overvaluation could deter investors from investing in or owning our best companies at precisely the time when their valuations are low and the long-term rewards for investing in them are better than they have been in decades. Furthermore, thanks to these low valuations the risk of owning blue-chip stocks is simultaneously also lower than it has been in a long time.
What CAPE Really Says
Prof. Shiller's cyclically adjusted PE ratio (CAPE) calculates an average of 10 years of S&P 500 earnings, which is used as the level of earnings with which to divide into the current price of the S&P 500 in order to determine the CAPE PE ratio. This calculation, which theoretically takes into consideration the cyclicality of the S&P 500's earnings, is promoted as being superior to forecasting future earnings. However, we will soon demonstrate that this is precisely what the CAPE measurement is doing; predicting or forecasting future earnings.
Furthermore, the only way that it can be of true value is if its implied forecast is correct in future time. Of course, it should also be recognized that this is no different than any other forecast methodology. The veracity of the assumptions underlying the hypothesis can only be proven if they produce an accurate future level of earnings, in this case for the S&P 500. Also, to be of any real value to investors, CAPE needs to offer a precise point in time when earnings are allegedly going to fall.
Mathematically, the actual forecast behind CAPE in the vast majority of cases is forecasting a lower-level of future earnings on the S&P 500. This is because the S&P 500 earnings are generally increasing and only interrupted by the occasional recession. Therefore, CAPE is suggesting that the S&P 500 is overvalued based on a vague notion about future earnings, not current earnings.
Another claim that CAPE makes as being a superior methodology is that it uses as reported earnings in contrast to operating earnings like most other prognosticators use. According to Standard & Poor's website, the following definitions to these earnings measurements are as follows:
"Operating earnings: income from product (goods and services), excludes corporate (M&A, financing, layoffs) and unusual items."
"As reported earnings: income from continuing operations, also known GAAP (Generally Accepted Accounting Principles) and As Reported
We do not intend to debate which of these earnings measurements represent better reflections of a company's profitability. However, we will emphatically state, that after decades of producing earnings and price correlated F.A.S.T. Graphs™ on individual companies, that the market places a higher correlation on price and earnings with operating earnings than it does with as reported earnings. Therefore, our research tool calculates growth rates and earnings justified valuation lines based on operating earnings.
With the above in mind, and in the spirit of fairness, we have calculated the S&P 500 CAPE valuation based on the 10-year time period 12/31/2001 to 12/31/2010 using both operating earnings and as reported earnings. We chose this time frame because it is the most recent completed 10-year period so all numbers are actual, therefore, no estimates are used. We pulled both earnings numbers directly from the Standard & Poor's 500 index data. Our calculations are as follows:
On an as reported basis, the 10-year average S&P 500 earnings would be $52.03, and by dividing that number into the S&P 500 price on 12/31/2010 of $1257.64, we came up with a CAPE of 24.17. On an operating basis, the 10-year average S&P 500 earnings would be $64.41, and by dividing it into the same S&P 500 price we came up with a modified operating earnings CAPE of 19.5. Both of these CAPE calculations are above the historical normal 16 PE ratio according to Prof. Shiller.
There are several important points to consider here. First, this time frame includes the two recessions of 2001 and 2002. Second, future S&P 500 earnings must fall to CAPE calculated $52.03 (an implicit forecast) in order to validate the overvaluation thesis behind them. That would require an earnings drop in excess of 54%, which is more than even the great recession of 2008. Next, this calculation would only be relevant to an index investor. Finally, this refers to an index that contains numerous cyclical stocks; however as we will later illustrate, not all S&P 500 stocks are cyclical in nature.
The PE Ratio Based on Actual Earnings
The following earnings and price correlated graph on the S&P 500 was selected to illustrate valuation based on actual earnings. There are two price earnings ratio lines that are drawn on the graph. The blue line (normal PE ratio) represents the widely accepted normal PE ratio of 20 that applies to approximately the last 50 years. The orange earnings justified valuation PE line represents the longer term historical PE ratio of 15, which is generally accepted as fair value for the average company and approximates the PE of 16 that Prof. Shiller embraces. The starting year 1995 was chosen because it was the most recent year (excluding the recession of 2008) where the S&P 500 was actually fairly valued with a PE of 15. The graph ends in 2010 because it is the last completed calendar year.
We believe that this graph provides a much more relevant depiction of valuation based on a clearer correlation between earnings and price. By reviewing this graph, it becomes crystal clear that the S&P 500 has, as Prof. Shiller currently contends, been mostly overvalued since 1995. The 20 PE ratio blue line is a trimmed average PE ratio where the highest and lowest PEs have been discarded. We believe this presents a clear picture of a major fallacy behind relying on statistical analysis.
Although the 50-year PE of 20 calculates and validates the statistical calculation, from the graph, it is clear that the market has only occasionally traded at the 20 PE. Over much of the time, the S&P 500 traded both, above and below, the statistical reference point. Perhaps validating the old adage that statistics don't lie, but statisticians are darn liars.
However, we believe the real value that the following graphic provides the perspective investor is the lesson on the importance of valuation and earnings growth to shareholder returns. This graphic shows that the S&P 500 was fairly valued with a PE ratio of 15 (price touched the orange earnings justified valuation line) at the beginning of 1995. However, equally as importantly, the price also ended in calendar year 2010 at precisely the same fair value PE of 15. Although there was a lot of overvaluation, and a couple of recessions in between, the earnings growth rate of the S&P 500 averaged 6.1%.
The 16-year performance results associated with the above earnings and price correlated graph of the S&P 500 establishes the importance of valuation and earnings growth. The closing annualized rate of return of 6.4% correlates almost perfectly with the 6.1% earnings growth rate. The additional $66,000 worth of dividends paid (assumed spent, not reinvested), kicks the total annualized return up to a respectable 7.9% (see red circle) per annum. Attractive and fair valuation at both the beginning and the end of this time period validates sound investing practices.
The next graph covers the period 1992 to October 20, 2011. The focal point should be the current absolute low valuation of the S&P 500 relative to the last 20 years. The indicated fair value based on a 15 PE times the estimated operating earnings of $97.88 indicates fair year-end 2011 value of approximately 1470 (see flag on graph), which is approximately 23% higher than current. In other words, on an absolute basis we believe the S&P 500 currently is 23% undervalued, not 30% overvalued.
The following estimated earnings and return calculator is based on Standard & Poor's own current estimate for calendar year 2011, followed by Bloomberg's estimates for fiscal year 2012 and 2013. In a report published on October