Article ID: | iaor201524814 |
Volume: | 24 |
Issue: | 2 |
Start Page Number: | 219 |
End Page Number: | 236 |
Publication Date: | Feb 2015 |
Journal: | Production and Operations Management |
Authors: | Huchzermeier Arnd, Breiter Andreas |
Keywords: | forecasting: applications, demand, marketing, time series: forecasting methods, inventory |
Demand forecast errors threaten the profitability of high–low price promotion strategies. This article shows how to match demand and supply effectively by means of two‐segment demand forecasting and supply contracts. We find that demand depends on the path of past retail prices, which leads to only a limited number of reachable demand states. However, forecast errors cannot be entirely eliminated because competitive promotions entail some degree of random (i.e., last‐minute) pricing. A hedging approach can be deployed to distribute demand risk efficiently over multiple promotional campaigns and within the supply chain. A retailer that employs a portfolio of forward, option, and spot contracts can avoid both stockouts and excess inventories while achieving the first‐best solution and Pareto improvements. We provide an improved forecasting method as well as stochastic programs to solve for optimal production and purchasing policies such that the right amount of inventory is available at the right time. By connecting a stockpiling model of demand with the supply side, we derive insights on optimal risk management strategies for both manufacturers and retailers in a market environment characterized by frequent price promotions and multiple discount levels. We employ a data set of the German retail market for a key generator of store traffic–namely, diapers.