searcher propose either myopic heuristics or numerical solution approaches for a short planning horizon (28) Popescu finds the optimal policy under variable ordering cost,backorder, and linear perishing cost, and shows that the optimal expected cost function is a convex function with respect to the available and future capacity ordered L periods in advance, for L being one or two periods. Furthermore, she shows that the optimal policy is a stationary policy when far enough from the end of the horizon