Spring 2000

ME/IE 8773-8774

CONSUMER DEMAND WITH STOCK-OUT BASED SUBSTITUTION: ESTIMATION AND EMPIRICAL APPLICATION

by

Ravi Anupindi, Ph.D.
Assistant Professor of Operations Research
Managerial Economics and Decision Sciences Department
Northwestern University
Evanston, IL 60208

Wednesday, January 26, 2000
3:35 - 4:25 p.m.
Room 108 ME
Broadcast on UNITE Channel A
Coffee will be available in 152 ME following the seminar

This presentation will explain how we develop a dynamic model of demand in which consumers may substitute one product with the other when faced with a stock-out. Under the assumption that customer arrivals are described by a Poisson process, and that inventory is replenished at regular intervals, we derive expressions for maximum likelihood estimators for relevant parameters of the choice process. Small sample properties of the estimators are illustrated using simulated data, while the model's potential for implementation is illustrated using an experiment based on a "Virtual Vending Machine". Comparison with some benchmark models reveals that ignoring stock--outs can lead to substantial bias in estimating the demand rates under full availability of all items in the category. Furthermore, since the possibility of substitution is not considered these benchmark models cannot provide estimates of the substitution rates. When the choice process may be assumed to be subject to the Independence of Irrelevant Alternatives (IIA) property, the model provides estimates of the rate of "no purchase" even though in the data we do not observe consumers who leave without purchasing. In this case the modeling of consumer response to stock--outs provides additional, valuable information about the magnitude of lost sales.

Ravi Anupindi is a faculty member in the Operations Management group of the Managerial Economics and Decisions Sciences Department at the Kellogg Graduate School of Management. He teaches a required course in Operations Management and an elective in Logistics and Supply Chain Management. He has also taught in several executive education programs at the Kellogg Graduate School of Management. He is the co-author of a textbook Managing Business Process Flows, Prentice Hall, 1999, which is widely used in the U.S. to teach the core Operations Management course for MBAs. His main research area is supply chain management with specific focus on the analysis and design of supply contracts, decentralized distribution systems, and stochastic models in retail and manufacturing operations. He spearheaded the development of BeerNET: Remote Group Software for Studying Supply Chain Dynamics - a platform for experimental research in supply chains. His work has appeared in leading journals like Management Science, Operations Research, Marketing Science, IIE Transactions. He also has contributed a chapter on supply contracts for Quantitative Models in Supply Chain Management by S. Tayur, M. Magazine, and R. Ganeshan (Eds.), Kluwer Academic Publishers, 1998. His consulting experience and speaking engagements include IBM, Digital Knowledge Associates, RealTimeData, Inc., McDonald's Corp., Budget Group Inc., Deloitte & Touche, McKinsey, Kellogg Alumni Association of Chicago. He is an Associate Editor of IIE Transactions: Scheduling and Logistics. He was a Guest Editor for Sadhana: Special Issue on Competitive Manufacturing Systems and was invited to be a Specialist Referee for the Hong Kong Research Grants Council (RGC). He is also a recipient of the 1996 Operations Research Meritorious Service Award. Professor Anupindi received a Ph.D. in Management of Manufacturing and Automation from Carnegie Mellon University in 1993, an M.E. in Automation from the Indian Institute of Science, Bangalore, India, and a B.E.(Hons.) in Electrical and Electronics Engineering from Birla Institute of Technology and Science, Pilani, India in 1982.

Faculty Host: Prof. Diwakar Gupta (625-1810) guptad@me.umn.edu