Supply Chain Contracts that Prevent Information Leakage

Abstract

This paper determines two groups of contracts -- downside-protection and upside-protection contracts -- that facilitate vertical information sharing in a supply chain while precluding horizontal information leakage among competing newsvendors. In particular, we consider a supply chain in which retailers replenish inventory from a common supplier to satisfy uncertain demand and are engaged in newsvendor competition. Each retailer has imperfect demand information. Yet, one of the retailers (the incumbent) has a more accurate demand forecast than the other (the entrant). Substantial anecdotal evidence and academic research have shown that information leakage among such competing retailers precludes vertical information sharing and is a reason for retailers to abandon collaborative forecast-sharing initiatives, leading to sub-optimized supply chains. We show that whether a contract can prevent information leakage is due only to how the supply chain's expected supply-demand mismatch costs are distributed among the supplier and retailers, but does not depend on each party's expected revenue. In addition, we show that wholesale-price contracts and two-part tariff contracts, which are extensively used in practice, cannot prevent information leakage. By using the wholesale-price contract as a benchmark contract, we determine that downside-protection contracts and upside-protection contracts can prevent information leakage. We define a downside-protection contract as one that effectively reduces retailers' cost of excess inventory by shifting some of their overage cost to the supplier. Examples of such contracts include buy-back and revenue-sharing contracts. We define an upside-protection contract as one that effectively increases retailers' cost of inventory shortage by shifting some of the supplier's underage cost to retailers. Examples of such contracts include penalty and rebate contracts. We show that these two groups of contracts can prevent information leakage. We identify necessary and sufficient nonleakage conditions and mechanisms for each group to prevent information leakage in the supply chain.