Bullwhip Effect Analysis in a Supply Chain

2. Background

The root of bullwhip effect can be traced back to Forester (1961). Although Forester described the bullwhip effect in some cases and pointed out management policies and decisions can generate variation in supply chain, however, the bullwhip effect owes its renown to Sterman’s studies in inventory management. Sterman (1989) reported his observations from bullwhip effect as “Beer Distribution Game” or simply “Beer Game”. In general, the researches in this area can be divided into three groups. In the first group, the researches focus on analyzing the creators (causes) of the bullwhip effect (for instance: Lee et al. (1997b), Bourland (1996), Miragliotta (2006)) whereas more detailed and specialized subjects are surveyed in the two last groups. The second group includes researches which measure or consider the bullwhip effect generated from creators known in the previous researches and the third group includes the researches which compare the created bullwhip effect generated from various methods or causes. Since this chapter aims to consider the impact of various forecasting methods on the bullwhip effect, we focus on those papers of the second and the third groups which considered the impact of forecasting methods on the bullwhip effect. Generally, four methods of forecasting have been considered so far: Exponential Smoothing, Moving Average, Linear Regression and ARIMA. In addition, it is assumed in these papers that a supply chain consists of three-echelons with one product and the same forecasting method in each echelon. Then, the chain was modeled and the bullwhip effect indices were analyzed. Some of these papers are summarized in Table 1.