The cacophony of “the end is near” cries from the doom and gloomers is much with us as stocks have marched unrelentingly higher over the last few years. The doom and gloomers do a lot of talking and shouting, but they don’t seem to do much actual research. For if they did, they would have discovered a relationship between earnings yields (the inverse of P/E) and inflation that suggests that at a P/E of 21.7x stocks are selling just about where they should be.
The picture below is a screenshot from my Bloomberg terminal showing the historical spreads of the S&P 500’s earnings yield and the personal consumption deflator, the measure of inflation that the Federal Reserve says is most accurate. Earnings yield, which I define as S&P 500 operating earnings divided by price, or the inverse of P/E, is quoted as a percentage, as is inflation. This produces a visual representation between the two data series that is easy to analyze.
I have written about this relationship as the most important driver of the S&P 500 and Dow Jones 30 price-to-earnings ratios for many years. I first discovered the relationship between earnings yield and inflation in the 1990s via some data that Value-Line provided their subscribers. I was surprised that of all the data that I studied, inflation was the most highly correlated with earnings yield. I would have bet almost anything that interest rates would have proven to have had the best correlation. (I will show the current picture of earnings yield and interest rates later.).
In the upper left-hand side of the picture, you can see the movements of inflation versus earnings yield charted on a quarterly basis going back to 1973. The difference between the two data series is shown in green. During this time, you will note that the two lines have moved almost on a tit for tat basis and now stand at about 4.6% for earnings yield and 1.4% for inflation. The data box at the top right of the picture shows that the average difference between earnings yield and inflation during this 43- year period has been about 3.4%. Right above that data-point is the current difference of about 3.3%. There is a lot of other data on this picture, but let me direct your attention to two other important indicators. I have drawn a red arrow (lower right) pointing at R^2 (Correlation). You can see that the R^2 between the two data series is very high at .705. R^2 is a statistical measure of fit between two or more data series and calculates that quarterly movements in inflation have been able to predict just over 70% of the movements of earnings yield and thus P/E. At our firm, we have more complex models using a wide range of other data series that can raise the R^2 up as high as 95%.
The second important indicator to consider is contained in the red circle I have drawn around a faint red asterisk on the lower left-hand side of the picture. The red asterisk shows the current differential between earnings yield and inflation, and it is sitting just about where we would hope it would be -- on the fair value line.
In summary, trading at 21.7 times operating earnings, stocks are not wildly overvalued, using inflation as the measure of value, and there is no other single data series that I know of that is able to pinpoint P/E with such statistical accuracy. I think the doom and gloomers are howling in the wind and their doom and gloom will continue to build as stocks go higher. In my judgement, if the current earnings growth continues, inflation would have to rise nearly one percent to roughly 2.5% before stocks would begin to feel headwinds.
Below I have copied a picture of the relationship between earnings yield and interest rates. You will note the fit between the two data series is much less convincing than that of earnings yield and inflation. Indeed, the two data series flip flopped positions in the late 1970s and thus have an R^2 of only about .42.