Implications of return predictability for consumption dynamics and asset pricing

We use the evidence on predictability of returns at dierent horizons to discriminate among competing asset pricing models. Specically, we employ predictors-based variance bounds, i.e. bounds on the variance of the Stochastic Discount Factors (SDFs) that price a given set of returns conditional on the information contained in a vector of return predictors. We document that consumption-based asset pricing models such as the classical long-run risk and habit models do not produce SDFs volatile enough at the one-year horizon. When we look at long-horizons our evidence shows that it is the habit model, not the long-run risk model, that satises our bounds. The rare disasters model satises our predictors-based bounds at each horizon. As a consequence, the investment horizon and the use of conditioning information emerge as fundamental ingredients that permit either to set models apart, or to select the common behavior among apparently dierent models.