The first chart is histogram of the quarterly, trailing 12-month (TTM) P/E ratios of the S&P 500 since 1950. It shows the frequency with which the market has traded around a given level relative to earnings. As is shown, the market has traded right around its current P/E ratio 30% of the time since 1950. The chart also shows that when the market trades at twenty or more times trailing earnings, it does not trade there for long. Relative to the past 50+ years, the valuation level does not seem to be excessive.
The next chart shows the trend of the S&P 500’s P/E ratio since 1950, along with its mean value over that time period. The P/E ratio has fluctuated wildly around the average, but it appears the S&P is currently near its average P/E of the past 57 years. This average is, of course, biased upward by the lofty multiples of 1999-2000.
The third chart shows the relationship between the S&P 500 earnings yield and the 10-year Treasury note, along with the average of each value over the period from 1953 through the second quarter of 2007. I am using the constant maturity 10-year Treasury as a proxy for long-term interest rates. I was able to get data from 1953, so this chart runs from that point to the end of 2Q 2007.There is a clear relationship, as there should be, between the earnings yield on stocks and long-term interest rates. As the chart shows, the average spread between the earnings yield on stocks and the rate on Treasuries over this 54-year period is just 0.3%. Currently, that spread is 1.0%; stocks yield 5.7% on a trailing basis and the 10-year T-note is trading at 4.7%. What this basically means is that if the S&P 500 experienced zero growth over the next 10 years and paid out all earnings as dividends it would yield 1% more than Treasuries. Looking from this perspective, on a relative basis stocks appear to be less expensive than average. Yet 10-year T-notes have traded as high as 5.3% in the past couple months, which would bring the spread down to only 0.4%, about average. And just because stocks are relatively more attractive does not mean they couldn't get more attractive (read: continue to fall). As the chart reveals, the earnings yield on stocks was substantially higher than long-term rates for the better part of the 1970s. Such a situation would present an incredible buying opportunity.
I do not focus solely on the rear view mirror when I drive, and it is hazardous to do so when investing. But as Mark Twain said, “History doesn’t repeat itself, but it rhymes.” It helps to have a historical perspective, especially about the broad market. And on a broad market basis, I don’t think stocks are expensive. But I don’t know where they will trade next month or next year. Sure, they can go down (which is what they’re doing now) but over long periods the trend is up. In my view, the lower one can buy the market, the better, because the implied future rate of return rises as the market falls.A couple caveats about these charts: the P/E ratio – and the earnings yield – is very dependent on the E(arnings). If recessionary conditions come back into the fore, corporate earnings will fall (the earnings yield will drop, absent stock price declines). This analysis is meant to provide some perspective on what the market looks like relative to low-risk alternatives such as long-term bonds (favorable). Also, the time period I chose (largely as a result of the data set that was available) could also have an impact on the averages I’ve calculated. They are merely illustrative and should not be construed as precise calculations. It is better to be approximately right than precisely wrong, as Warren Buffett says.