Responsive Pricing Newsvendor Networks with Discretionary Commonality

Discretionary commonality is a form of operational flexibility used in multi-product manufacturing environments. Consider a case where a firm produces and sells two products. Without discretionary commonality, each product is made through a unique combination of input and production capacity. With discretionary commonality, one of the inputs could be used for producing both products, and one of the production capacities could be used to process different inputs for the production of one of the products. In the latter case, the manager can decide, upon the realization of uncertainty, not only the quantities for different products (outputs) but also the means of transforming inputs into outputs. The objective of this paper is to understand how the firm's value, its inventory levels for common and unique inputs, and capacity levels for flexible and dedicated resources are affected by the demand characteristics and market conditions faced by the firm. In pursuing this research, we extend Van Mieghem and Rudi (2002)'s newsvendor network model to allow for the modeling of product interdependence, demand functions, and random shocks, and for the modeling of firm's ex-post price setting. Applying the general framework to the network with discretionary commonality, we discover that inventory and capacity management can be quite different compared to a network where commonality is non-discretionary. Among other results, we find that as the degree of product substitution increases, the relative need for discretionary commonality increases; as the market correlation increases, while the expected firm value may increase for complementary products, the discretionary common input decreases but the dedicated input increases. Numerical study shows that discretionary flexibility and responsive pricing are strategic substitutes