Blog - Investing Notes

June 27, 2006 - Another Adjustment to the Real Earnings Yield Model (and Discarding the Average of Models)

In our blog entry of 6/21/06, we describe an adjustment to our Real Earnings Yield (REY) Model that increases its volatility but makes it more current with respect to inflation rate shocks. In this entry, we describe another adjustment, implemented as of today, that dampens the volatility of the model's output without sacrificing data currency.

ADJUSTMENT TO REAL EARNINGS YIELD MODEL

As described in the updated description of the REY Model, we have added a parameter to the model that mutes inflation rate volatility. Setting the parameter to zero eliminates inflation rate effects and produces results very similar to those of our Reversion-to-Value (RTV) Model (driven solely by an average price-to-earnings ratio). Setting the parameter to one produces no muting of inflation rate volatility. At the risk of data dredging, we set the parameter empirically at 0.6. The following charts compare the long-term and short-term REY Model-generated mockups of the S&P 500 index for inflation rate volatility muting parameter values of 0, 0.6 and 1:

This change affects our Stock Market Status because the peaks and troughs of the REY Model are now less pronounced.

DISCARDING THE AVERAGE OF THE MODELS

With the change described above, we will no longer track the averaged outputs of our REY Model and RTV Model. Averaging had the net effect of counting earnings shocks twice and inflation rate shocks once. Parameterization of the REY Model accomplishes close to the same thing. The adjusted REY Model essentially takes the place of the averaged models.