Efficient Lateral Transshipment Policy for Multi-Retailer System

Abstract

Recent development in inventory management shows that the use of lateral transshipment to pool inventory promotes the customer service. Inspired by the existing works, this paper studies the lateral transshipment policy in an inventory system consisting of multiple retailers. We model this problem as Markov Decision Process (MDP), and several desirable properties of the optimal lateral transshipment policy are developed. The primary contribution of this study lies in its effort to extend the state-of-the-art results on two retailers to arbitrary number of retailers. Since the computational capacity required for solving large-scale MDP is prohibitive, a comprehensive heuristic lateral transshipment policy is proposed. The effectiveness of the heuristic is demonstrated through numerical experiments. This study brings new perspectives on seeking a balance between benefits and risks of lateral transshipment actions taken by retailers, and provides practical insights on inventory management problem involving lateral transshipment.