Atactical implication of predictability: fighting the FED model

This paper confirms that high earnings yield portend high equity returns. Absolute valuation levels of equity have predictive power over future long run equity returns. The predictability is far less powerful in the short term. On a tactical investment horizon, investors tend to rely on the relative valuation of equity versus bonds to gauge whether equity markets are attractive. The FED model, which compares earnings yield and bond yield, is the preferred yardstick in the finance profession. First, this paper examines the FED model and shows that is not only theoretically flawed, but also not able to predict equity returns over long sample periods. Second, we improve the model by adding corrections for perceived risk enabling a better fit of the data. Third, the main innovation is testing a tactical asset allocation model constructed on the basis of the improved model. A model portfolio taking advantage of the short-term deviation in relative value, corrected for risk, leads to superior performance.