Using Optimal Policy Surfaces to Analyze Aggregate Inventory Tradeoffs

The marginal cost information needed to implement traditional inventory models is not likely to be available in practice. The most important inventory management issues in practive involve aggregate objectives and constraints while the richest theoretical models deal with single item management. To help resolve these problems, the authors propose that inventory decisions be conceived as policy tradeoffs on a three dimensional response surface showing the optimal relationships among aggregate customer service, workload, and investment. We show that any optimal management decision must result in a point located on the surface. Computational results show that the methodology suggested can make improvements in management policy in four inventories that total more than 78,000 line items.