Abstract
We investigate optimal pricing and capacity planning decisions for product-line settings such as introducing a new product or dropping an existing one. We consider a two-product, two-period model with stochastic demands, where price and capacity decisions are made at the outset. Investment in capacity must be traded-off against the possibility of buying at higher spot market prices due to shortage in capacity or charging a higher price to manage the demand. Prior studies argue that introducing an additional product to the product-line strains capacity, resulting in an increase in the price of an existing product. In contrast, we find that introducing a new product can also result in a drop in price of an existing product, enabling strategic pricing by firms. The necessary condition for this to occur is that the demand uncertainties for the products are of similar magnitude and negatively correlated. Similar insights are obtained for the setting where an existing product is dropped from the product-line. Hence, product-market decisions and contextual factors play a role in capacity planning, capacity cost allocation and pricing.
Notes
1. A third setting could be ‘Replacing an Existing Product’ where product 1 and product 2 are available in period 1, but product 3 replaces one of the existing products in period 2. This generates similar insights and hence, we do not discuss this setting in the paper for brevity. Detailed results are available from the authors. We treat the product-line decision settings in our paper as exogenous and do not model events leading to them, similar to Banker and Hughes (Citation1994) and Banker, Hwang, and Mishra (Citation2002), to focus on the capacity and pricing decisions.
2. Observe that our value of γi in Corollary 1 (i) is different from that of Banker, Hwang, and Mishra (Citation2002) because is different from
due to the new product’s availability only in period 2. In fact, in their scenario both products are available in both periods. Hence, the joint distribution of demand uncertainties remains the same even if the two products are correlated.