Knightian Uncertainty and Stock-Price Movements: Why the REH Present-Value Model Failed Empirically