S&P 500 Price Not Expensive By Historical P/Es

We plotted operating income P/E bands from 10x to 26x for the 4-quarter trailing reported operating earnings each quarter from June 1989 through March 2013, and for estimated earnings for calendar 2013.

Bottom Line:  the valuation multiple for the S&P 500 is not expensive by historical standards back almost 24 years.

The price of the index may be rising faster than it can continue without a pause, or maybe a correction, but the overall price level seems reasonable based on trailing earnings.

(click image to enlarge)



The median P/E is 17.79x.  The average is 18.72x.  The average of the middle 50% is 17.80x.  The 1693 price today as a multiple of the $109.22 operating earnings estimate for 2013 from Standard and Poor’s is 15.9x.

Of course, earnings could collapse in a recession, but that does not appear on most radar screens in the near term.  The risks associated with Fed QE tapering are probably minimal (beyond a knee jerk negative response for a short period when it happens).  Tapering is not tightening.  Under tapering, the Fed balance sheet will still be expanding by massive amounts.

Japan could find its central bank program doesn’t float their market, and their interest rates may have to approach global levels as they borrow more and more, while surrendering more and more export prowess to other Asian nations.

The plateau in S&P 500 revenue could persist for an extended period, with margin squeezes, making the earnings growth assumptions too rosy.

But those are shadows on the wall.  We have to work with what we have, and what we have is a growing US and world economy (not that much, but still growing), and the S&P 500 with valuation based on operating earnings that is not cheap, but also not expensive.

S&P 500 tracking ETFs are directly relevant to these data, but for practical purposes so are other broad market large-cap, market-cap weighted ETFs tracking these indexes:

  • S&P 500 (SPY, IVV, and VOO)
  • Russell 1000 (IWB, ONEK, and VONE)
  • Russell 3000 (IWV, THRK, VTHR)
  • DowJones Total Stocks (IYY)
  • CRSP US Total Stocks Market (VTI)
  • NYSE Composite (NYC)


Some other data, such as the Shiller 10-year inflation-adjusted, normalized P/E ratio based on reported net earnings, is used to argue differently.  One big issue is how much history to use.

(click image to enlarge)


Shiller’s data shows a current normalized P/E on reported earnings of 23.80.  The standard argument that 23.80 is expensive is based on the average value of 16.48 since 1870 (using the S&P 500 from its inception in 1957 and its best precursors before that date).

One could argue how relevant such a long history is given all of the changes in the world of securities since effectively the Civil War — income taxes, central banks, rise of the US as a world power and the dollar as the dominant currency, computers, mutual funds, tax deferred savings accounts, general public investing in stocks (mostly via funds), the internet, social safety nets, too big to fail, and many other things.

Taking some shorter periods to make the overall context more consistent, we see different normalized P/E ratios.

From 1945 (effectively after World War II, when the US became the dominant economic power), the normalized average P/E is 18.23.

From 1981, the peak of Treasury interest rates, the average normalized P/E is 21.56.

From 1994, the year HMTL and the web came out of academia and the military to start the march to the web-world we live in today, the normalized P/E was 26.65.

So at 23.80, the normalized P/E is way out of line versus 16.48, and not all that bad versus 26.65.

Overall, we think the valuation of the US stock market is within a reasonable range.

POST-SCRIPT: Compare Historical Operating Income, Net Income and Dividends


Comments are closed.