Can average skewness really predict financial returns? The euro area case