Impact of inventory allocation policies on downstream supply variability in a two-echelon system

No: 011-0319 Abstract Title: Impact of Inventory Allocation Policies on Downstream Supply Variability in a Multi-echelon System Authors Information: SUMAN NIRANJAN School of Management, State University of New York Binghamton, Vestal, NY 13902 607-777-4120 [email protected] FRANK CIARALLO Biomedical, Industrial, and Human Factors Engineering, Wright State University, Dayton, OH 45435 937-775-5024 [email protected] POMS 20th Annual Conference Orlando, Florida U.S.A. May 1 to May 4, 2009 Impact of Inventory Allocation Policies on Downstream Supply Variability in a Multi-echelon System SUMAN NIRANJAN* and FRANK CIARALLO** *School of Management, State University of New York Binghamton, Vestal, NY 13902 **Biomedical, Industrial, and Human Factors Engineering, Wright State University, Dayton, OH 45435 Abstract In this paper we study the impact of inventory allocation policy on the downstream supply variance in a two-echelon system with multiple sources of demand in the upper echelon. Allocation policies play a crucial role in deciding the amount of inventory assigned to every source of demand. The cause of the supply variance is always attributed by the researchers as demand & capacity variance, or capacity utilization, or lead time. But we find that the nature of the inventory allocation policy is equally important as other exogenous variables, and has a direct impact on the downstream supply variance, in turn affecting the optimal safety stock cost and total cost in a supply chain. We introduce a new metric that will aid managers in the selection process of inventory allocation policies in presence of multiple sources of demand. Further the relation between the inventory allocation policy and the new metric is explored. We find that the variance in the supply increases/decreases significantly on the basis of inventory allocation polices used at the upstream node. We further find that the optimal inventory levels at each node now in a multi-echelon system are not only a function of the demand, capacity, and lead time, but also a function of the allocation policy employed. A Simulation based Inventory Optimization (SIO) algorithm is developed to obtain the best base-stock levels for a two-echelon system based on derivative estimates from an Infinitesimal Perturbation Analysis (IPA) framework.In this paper we study the impact of inventory allocation policy on the downstream supply variance in a two-echelon system with multiple sources of demand in the upper echelon. Allocation policies play a crucial role in deciding the amount of inventory assigned to every source of demand. The cause of the supply variance is always attributed by the researchers as demand & capacity variance, or capacity utilization, or lead time. But we find that the nature of the inventory allocation policy is equally important as other exogenous variables, and has a direct impact on the downstream supply variance, in turn affecting the optimal safety stock cost and total cost in a supply chain. We introduce a new metric that will aid managers in the selection process of inventory allocation policies in presence of multiple sources of demand. Further the relation between the inventory allocation policy and the new metric is explored. We find that the variance in the supply increases/decreases significantly on the basis of inventory allocation polices used at the upstream node. We further find that the optimal inventory levels at each node now in a multi-echelon system are not only a function of the demand, capacity, and lead time, but also a function of the allocation policy employed. A Simulation based Inventory Optimization (SIO) algorithm is developed to obtain the best base-stock levels for a two-echelon system based on derivative estimates from an Infinitesimal Perturbation Analysis (IPA) framework.