Demand Information Sharing in Online Retail Marketplaces

Abstract

In this paper we make use of game theory tools to address a specific type of problems that arise in inventory management in multi-agent markets. In particular, we study which market characteristics are able to generate incentives for agents to buy and share information that help them to improve the prediction of their demands and, thus, to improve their inventory policies. We modeled the retailers’ decisions regarding demand information acquisition and sharing as a biform game. Our findings demonstrate that, for any set of demand information acquisition strategies, there are clear incentives for cooperation among retailers. Moreover, no coalition of retailers has an incentive to abandon cooperation in pursuit of a better outcome independently. We introduce a family of benefit distribution rules that are coalitionally stable. Additionally, we show that demand information acquisition can always reach equilibrium, irrespective of the value of the information. We further analyze the equilibria in scenarios with specific characteristics. Finally, we compare the equilibria derived from our model with those obtained in a centralized setting.

Keywords: Newsvendor, Game Theory, biform game, information sharing