José Antonio Carrasco, UAI

“Costly multi-unit search”

Coautoreado con Rodrigo Harrison (UAI)

Abstract: We solve a costly multi-unit search problem of a seller liquidating inventory. There is no organized competitive market. Instead, buyers arrive with capped demands, aiming to buy single or multiple units at random prices, upon which the seller can partially act. We inductively characterize the optimal strategy and
specify the search strategy (when to search) and the selling strategy (how much to sell). An escapable fixed flow cost reverses the diminishing returns to optionality property of inventory and translates into non-monotone endogenous selling costs. Since reservation prices are guided by total revenue considerations, our model makes sense of strategies that combine purchase premiums, volume discounts and bundling. We show that bundle pricing is optimal when the seller faces homogeneous demands. Otherwise, the optimal strategy combines selling additional units at a premium when inventory is large with “sales” in the form of bundling and volume discounts, or even premiums, when inventory is low. Our theory explicitly accounts for the effect of escapable fixed costs on optimal liquidation strategies.