We implement a long-horizon static and dynamic portfolio allocation involving a risk-free and a risky asset. This model is calibrated at a quarterly frequency for ten European countries. We also use maximum-likelihood estimates and Bayesian estimates to account for parameter uncertainty. We find that for most European countries the dividend-price ratio and inflation have predictive power. For countries where returns are predictable, we demonstrate out-of-sample economic signi cance for the long-horizon allocation. Parameter uncertainty plays a second-order role, dominated by strong variation in the dynamic allocation itself induced by large variations in the state variables. The market timing appears economically relevant for many countries.