On perfect foresight and uncertainty in economic models: Two monetary examples