Bullwhip effect under substitute products

• Author: Yongrui Duan,Jiazhen Huo,Yuliang Yao
• Published date: 01 May 2015
• DOI: http://doi.org/10.1016/j.jom.2015.03.002
• Date: 01 May 2015

Journal of Operations Management 36 (2015) 75–89

Contents lists available at ScienceDirect

Journal of Operations Management

journal homepage: www.elsevier.com/locate/jom

Bullwhip effect under substitute products

Yongrui Duana, Yuliang Yao b,∗, Jiazhen Huoa

aSchool of Economics and Management, Tongji University, China

bCollege of Business and Economics, Lehigh University, United States

article info

Article history:

Received 14 May 2014

Received in revised form 2 March 2015

Accepted 3 March 2015

Available online 9 March 2015

Accepted by Daniel R Guide

Keywords:

Bullwhip effect

Substitute products

Supply chain management

Empirical analysis

abstract

Using a large-scale, product-level dataset collected from a supply chain dyad, we examine the effect of

own and substitute products on a focal product’s bullwhip effect and estimate the existence and mag-

nitude of the bullwhip effect at the product level. We find that, under substitute products, the bullwhip

effect is not only affected by a product’s own factors but also by those of its substitute products. An

increase in the number of own price changes is associated with a decrease in the bullwhip effect in terms

of the direct effect but with an increase in the bullwhip effect in terms of the total effect, and increases

in the number of price changes of substitute products and own stockouts are associated with increases

in the bullwhip effect. The potential effects for own price changes, price changes of substitute products

and own stockouts are as much as 59.51%, 95.06% and 66.11%. We also find that the bullwhip effect is

prevalent and very intensive at the product level. We discuss the theoretical and managerial implications

of the findings.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

In recent decades, a significant advancement in our under-

standing of supply chain management is the identification and

management of the bullwhip effect (de Kok et al., 2005). The bull-

whip effect refers to “the amplification of demand variability from

a downstream site to an upstream site” (Lee et al., 2004, p. 1887).

Research has shown that the bullwhip effect leads to tremendous

supply chain inefficiencies such as excessive inventory investment,

poor customer service, lost revenues, misguided capacity plans,

ineffective transportation, and missed production schedules (Lee

et al., 1997). While the bullwhip effect has generated attention

among researchers and practitioners alike, companies have not yet

succeeded in completely taming it (Lee et al., 2004; Songini, 2000;

Wheatley, 2004; Wong et al., 2007).

Extensive modeling literature has focused on this phenomenon

(e.g., Chen et al., 2000; Chen and Lee, 2009; Chen and Samroengraja,

2004; Lee et al., 1997; Metters, 1997). Building upon the theoretical

work,a growing body of empirical studies has made efforts to detect

and quantify the bullwhip effect and its driving forces (e.g., Bray

and Mendelson, 2012;Cachon et al., 2007; Fransoo and Wouters,

2000; Klug, 2013; Lai, 2005; Taylor, 1999). Still, our understand-

ing of the bullwhip effect is limited in several ways. First, empirical

∗Corresponding author.

E-mail address: yuy3@lehigh.edu (Y. Yao).

studies have used aggregate data or proxy data to measure the bull-

whip effect; for example, prior studies have examined the bullwhip

effect using aggregate data (e.g., Bray and Mendelson (2012) and

Shan et al. (2013) at firm level; Cachon et al. (2007) at industry

level), which may mask or dampen the estimation of the bullwhip

effect (Caplin, 1985; Fransoo and Wouters, 2000; Lai, 2005) and

shipment or delivery data to proxy order data because order infor-

mation is not readily available (i.e., material flow-based bullwhip

effects) (Chen and Lee, 2012). Second, analytical research on the

bullwhip effect has generally considered a model setting for a sin-

gle product (e.g., Chen et al., 2000; Lee et al., 1997); for example,

Lee et al. (1997) assume that a single product is being managed

and transacted between firms and compute the bullwhip effect for

the product. In practice, however, firms such as retailers manage

inventories and sales for multiple substitute products1simultane-

ously (e.g., kiwi juice vs. strawberry juice) (Kök and Fisher, 2007).

Research has demonstrated that a product’s demand is not only

influenced by its inherent characteristics but also by the behaviors

of its substitute products (Netessine and Rudi, 2003; Vulcano et al.,

2012). As a result, a product’s bullwhip effect may exhibit different

magnitudes and properties in the presence of multiple substitute

products. Few studies (including analytical and empirical studies)

have considered substitute products when exploring the bullwhip

1Substitute products are defined as those products where lowering the price of

one product leads to a decrease in the sales of another (Shocker et al., 2004).

http://dx.doi.org/10.1016/j.jom.2015.03.002

0272-6963/© 2015 Elsevier B.V. All rights reserved.