About a year ago, I decided to look at a projected S&P price based on average earnings and compare that number with the actual S&P 500 Index price. The exercise produced some interesting observations (available here). Simply by tracking the forecasted earnings, we can "picture" where the S&P 500 would trade out to December 2014 using the new forecasted earnings available from S&P and based on historical data. Using the latest data as of July 11th, we can update the charts.
The following data derived from Standard & Poor's earnings estimate Excel spreadsheet shows the correlation over several time periods between the actual S&P 500 quarterly close and what the "projected price" would have been using a proprietary formula.
For example, in the table above, the 3 year average operating earnings possessed a .84 correlation with the actual quarterly closing price of S&P 500 (1 would equal 100% correlated, -1 would equal 100% opposite correlation). By simply averaging the last 3 years (12 quarterly earnings reports), the change in the S&P 500 index over time (1989 to present) nearly matched the change in operating earnings over the same time period (.84 correlation). The second column calculates what the S&P should have been based on the 3 year average (S&P 500 Projected Price) and correlated with the actual S&P at .85 (still – very highly correlated). However, moving into the 1 year average earnings and 1 year S&P equivalent produced a supercharged 97.5 percent correlation.
And the next question that always occurs is "What the heck do I do with the information?" Here is the chart showing where the S&P would trade predicated on the forecasted earnings.
While the indicator resembles a moving average – the S&P projected price (Red Line) tracks the S&P 500 quarterly price relatively well. The calculation involves using an average Price to Earnings ratio for the same time period and multiplying that by the average earnings (for the mathematically oriented – insert the word "mean" into the "average" spots). To extend the projection out to December 2014, we use the average earnings growth of 1% per quarter (calculated by using 1989 to present quarter over quarter change in growth). However, one must be leery using "earnings forecast" because these are normally downwardly revised and exogenous events (like 2008) may dramatically change earnings.
Additionally, by calculating when to be "in the market" (buy the S&P 500 Index) and when to exit (be in cash), we see that an initial investment of $10,000 would be worth just over $61,301 (excluding S&P 500 dividends) VS simply investing in the S&P 500 and getting $46,687. This relatively simply strategy will not make one rich, but may help in identifying when to buy and when to sell.
Forecasted S&P Index prices generally mean nothing and choosing to predict the future remains futile (and yet we still do it -- the definition of "insane" perhaps?). But, employing this simple concept on a quarterly basis after earnings are "in the books" may assist in decisions on whether to be "in the market" or "out."
Here are a couple past charts for comparison purposes.