This dissertation addresses many fundamental questions regarding the equilibrium behavior of decentralized supply chains with competing retailers. We confine ourselves to two-echelon supply chains with a single supplier servicing a network of competing retailers. The supplier distributes closely substitutable products to a set of competing firms that in turn sell these to the consumers. Each firm's sales volume depends on the prices charged by him as well as those charged by all other retailers, according to a general demand function. In this setting, we investigate various monotonicity properties under the model's (Nash) equilibrium, with respect to specific exogenously specified parameters. Alternatively, the price each retailer can charge for its product depends on the sales volumes targeted by all the retailers. The supplier incurs manufacturing and distribution costs that depend on the retailers' purchase volumes.
A major challenge for these chains is to optimize their aggregate performance while maintaining decentralized decision making. The challenge for the supply chain therefore consists of creating an appropriate coordination mechanism to structure the costs and rewards of all of its members so as to align their individual objectives with the aggregate system-wide objective. The mere existence of such coordination mechanism is in question for most supply chain models, in particular in the presence of price competition between its retailers.
We then analyze a variant of this model, in which the firms' costs consist of inventory holding costs and setup costs associated with production runs and procurement orders. In this setting, coordination is achieved if the supplier offers a discount from a basic wholesale price, based on the annual sales volume, the order quantity, and the order frequency.
We next turn our attention to stochastic models that incorporate demand uncertainty. We investigate, once again, the equilibrium behavior of decentralized supply chains with competing retailers and the specification of coordination mechanisms. We address our investigation in a single period setting, useful for products with a single sales season, and then proceed with more general multi-period settings in which retailers carry inventories and possible backlogs from one period to the next.