Tracing the impact of non-uniform forecasting methods on the severity of the bullwhip effect in two- and three-level supply chains

Abstract

In this study we have investigated the bullwhip effect and then we have modelled demand forecasting in two- and three-level supply chains by using exponential smoothing and moving average methods. Concerning chains, we have considered two- and three-level chains, and 300-period and 500-period horizons. Demand patterns fall into two major groups: first, a raising pattern with low variability; and second, a diminishing demand with high variability. Our goal is to examine and trace the effect of forecasting methods on the intensity of the bullwhip effect when the applied forecasting methods are not equal, and each layer of the chain may use his/her own method without any information on how other nodes forecast their own demands. Reviewing the literature, there are only a few studies such this one that consider dissimilar forecasting methods in one chain. Because of a lack of information sharing, and a large difference in forecasting parameters in this kind of modelling, there should be a large amount of demand variance and this will lead to a strong impact of the bullwhip effect.