A condition is derived that must be satisfied by a continuous function h: X → X in order to be the optimal policy function of a strictly concave deterministic dynamic programming problem which is ...