The stochastic inventory models in Chapters 4–6 assume that inventory is the only tool for mitigating uncertainty. In contrast, this chapter examines uncertainty mitigation using other means. In all of the strategies covered here, the idea is to “pool” multiple demand streams in some way, and to share some resource—inventory or capacity—among them. Because not all of the demand streams will need all of the resources at all times, there is no need to dedicate whole resources to each stream. By pooling them, we can reduce the amount of safety stock required to meet a given service level (or increase the service level attained by a given level of safety stock).
Section 7.2 deals with risk pooling, in which we physically combine the inventories used to satisfy multiple demand streams, by storing them together in the same warehouse. Section 7.3 discusses a strategy called postponement, in which we differentiate products later in their manufacturing process. This allows a reduction in inventory since multiple demand streams (from different end products) are sharing inventory of the the undifferentiated product. The cost savings from postponement is due to the risk pooling effect.
Another way that inventory can be pooled is by allowing transshipments—“lateral” transfers of inventory from one retailer to another when one has extra inventory and the other has a shortage. In Section 7.4, we discuss a model for deciding how much ...