Non-stationary demand forecasting by cross-sectional aggregation

摘要: In this paper the relative effectiveness of top-down (TD) versus bottom-up (BU) approaches is compared for cross-sectionally forecasting aggregate and sub-aggregate demand. We assume that demand follows a non-stationary Integrated Moving Average (IMA) process order one Single Exponential Smoothing (SES) procedure used to extrapolate future requirements. Such processes are often encountered in practice SES standard estimators industry (in addition being optimal estimator an IMA process). Theoretical variances forecast error derived BU TD approach contrast relevant performances. The theoretical analysis supported by extensive numerical investigation at both level, empirically validating our findings on real dataset from European superstore. results demonstrate increased benefit resulting cross-sectional environment than stationary one. Valuable insights offered planners closes with agenda further research area.