Quantifying the effectiveness of VMI and integrated inventory management in a supply chain with uncertain lead-times and uncertain demands

This article considers the inventory management problem in a supply chain with uncertain replenishment lead-times and uncertain demands. The optimal integrated inventory management (IIM) policy is developed using stochastic dynamic programming theory. The IIM policy is contrasted with two pull-type vendor-managed inventory policies (VMI-1 and VMI-2) and a traditional retailer-managed inventory policy (RMI). Computational results show that in such stochastic supply chains, IIM performs about 23, 15, and 3% better than the optimised RMI, VMI-1 and VMI-2 policies, respectively, while two VMI policies are about 8 and 20% better than the best RMI. The basestock-based VMI-2 is a very good form of VMI. The ANOVA analysis reveals that the replenishment lead-times have the largest effect on the relative performance between IIM and other policies. Numerical examples demonstrated that the IIM policy has good structural properties and can be characterised by a set of switching curves.