Analyzing the effect of consumer returns in a multi-period inventory system

Return of a sold item by a customer becomes tremendously common situation in many industries. Increase in the amount of returned items promotes return information to be a critical factor for inventory control. Undoubtedly another critical parameter for an inventory system is the length of the review period. Effect of the review period or length of the time-bucket is amplified with returned items, because available return information at a decision point is related to the frequency of the review. In this study, we analyze the effects of these two parameters over a multiperiod inventory system where the length of a time horizon is fixed. Dynamic programming approach is used to calculate the optimal inventory positions. In dynamic programming, it is assumed that a fixed proportion of sold items are returned. Computational results are obtained to compare the effects of return information under different return proportions and period lengths. These results are used to conduct various analyses to explore the level of the advantage gained by using return information.