Impact of information sharing on bullwhip effect in a non-serial supply chain with deterministic lead time

In this paper, we study the impact of retailers’ forecasting methods on the bullwhip effect. We consider three forecasting methods: Moving average (MA), exponential smoothing (ES), and minimum mean squared error (MMSE). We model retailers’ demand by a first-order autoregressive process. We also assume a base-stock policy for retailers’ stock replenishment. Based on Zhang (2004), we derive expressions for the bullwhip effect for each of the forecasting methods. We considered deterministic as well as stochastic lead time. We also consider multiple independent retailers. Our findings indicate that the impact of demand correlation and lead time on bullwhip effect changes with the forecasting method. However, the patterns seem to remain unaffected by the nature of lead time and the number of retailers. We also observe MMSE to outperform MA and ES when the bullwhip effect is the main concern.