Forecasting The Market: A Thought Experiment Revisited

Includes: SPY
by: Doug Short

By Chris Turner

With 98% of companies reporting for the Q3 2012 earnings season, here is the latest update of my ongoing "thought experiment" for forecasting the S&P 500 price based on earnings fundamentals.

The chart below is based on the latest trailing 12 months earnings (TTM) data published on the Standard & Poor's website as of Dec. 16. The numbers are from the spreadsheet maintained by senior analyst Howard Silverblatt. (See this monthly valuation update for instructions on downloading the spreadsheet.)

Here are the key assumptions in the calculations:

  • The 10-year average of nominal TTM earnings is 55.22 at the end of 2011, rising to 87.78 by the end of 2012, based on "as reported" earnings forecasts.
  • The average nominal cyclical P/E 10 is currently 18.15.
  • The S&P 500 historic prices used in the calculations are monthly averages of daily closes.
  • Standard & Poor's estimates of TTM earnings for Q1 2012 through Q4 2012 are $88.54, $87.92, $86.50, and $87.78, respectively (as of the latest spreadsheet).
  • The months between the quarterly earnings estimates are linear interpolations.

The blue line represents Standard & Poor's TTM forecast earnings by month, multiplied by the historical nominal 10-year P/E ratio. At 2012 year-end earnings of $87.78 and an average nominal P/E of 18.15, we would see the S&P 500 at 1593. At this level, the nominal P/E 10 would be 26.89, and the index would be about 47% above a hypothetical price multiple of the extrapolated 10-year earnings average.

The red line represents a hypothetical S&P 500 price that is a multiple of the average nominal P/E 10 of 18.15 and the 10-year average earnings of 55.22 for December 2011. The monthly index price estimates thereafter are linear extrapolations based on average 10-year earnings growth.

The optimistic view (blue line) would put us around 1593 in the S&P 500 by the end of November, the assumptions being that the Standard & Poor's earnings forecasts are correct and the nominal P/E 10 ratio is the multiple we see.

The pessimistic view (red line) is a reversion to the historic earnings and nominal P/E 10 multiple.

But history shows us that, regardless of your preferred earnings divisor (nominal or real, TTM or the 10-year average TTM), the P/E ratio has never hovered around the average. The market swings above and below its long-term average valuation in erratic arcs that can last for many years. For a long-term perspective on valuation extremes, see "Four Market Valuation Indicators" and the compelling research of Ed Easterling on the history of earnings per share.