IMPACT OF SIMPLE FORECASTING METHODS ON SUPPLY CHAIN PERFORMANCE UNDER TEMPORAL DEMAND HETEROSCEDASTICITY

摘要

This paper demonstrates that moving average and double exponential smoothing forecasting methods should be avoided if demand exhibits heteroscedastic behavior. Instead, an advanced GARCH model should be employed in order to improve the supply chain performance under temporal demand heteroscedasticity. Moreover, capacity tightness plays a role in affecting the supply chain performance via different forecasting methods. Therefore, supply chain managers should be aware of underlying assumptions which necessitate using advanced forecasting models to achieve greater cost savings for the supply chain relative to savings achieved using simple forecasting methods.