Mixed bundling strategies and multiproduct price competition

This paper deals with price competition among multiproduct firms. We consider a model with n firms and one representative buyer. Each firm produces a set of products that can be different or identical to the other firms' products. The buyer is characterized by her willingness to pay -in monetary terms- for every subset of products. To handle the combinatorial complexity of this general setting we use the linear relaxation of an integer programming package assignment problem. This approach allows to characterize all the equilibrium outcomes. We look for subgame perfect Nash equilibrium prices in mixed bundling strategies, i.e., when firms offer consumers the option of buying goods separately or else packages of them at a discount over the single good prices. We find that a mixed bundling subgame perfect Nash equilibrium price vector always exists. Also, the associated equilibrium outcome is always efficient, in the sense that it maximizes the social surplus. We extend the analysis to a model with m buyers and offer the conditions under which the equilibrium outcome set is non-empty.