Managing Overconfident Newsvendors: A Target-Setting Approach

Abstract

Experimental results support that overconfident newsvendors deviate from the optimal inventory levels to place orders closer to the mean of demand distributions, thereby lowering profits. To improve performance, we investigate whether nonstandard preferences based on prospect theory can manage overconfident newsvendors. We develop a framework where a manager determines a profit-based target, which serves as a reference point in the newsvendor's utility function. We show that target-setting can induce optimal orders for low-margin products. We extend our model to a duopoly and show that the type of competition determines whether target-setting improves performance beyond the monopolist setting. Decentralized competition, where competing newsvendors have different managers, reduces the effectiveness of target-setting. However, centralized competition, where competing newsvendors share the same manager, increases the range of profit margins within which target-setting can achieve the system-optimal order quantities. We compare profit-margin targets with gross-profit targets throughout our analysis. Although the targets offer similar performance, we find that profit-margin [gross-profit] targets dominate for extremely [moderately] low-margin products when the manager is uncertain about the newsvendor's overconfidence level. However, when managing a newsvendor with an uncertain degree of loss aversion, we find that profit-margin targets offer greater benefits regardless of the product's profitability.