Forecasting Aggregated Vector ARMA Processes (Lecture Notes in Economics and Mathematical Systems)

Is concerned with forecasting aggregated time series variables. Investigates impact of the level of aggregation on the forecast efficiency. Temporal aggregation over several time periods and contemporaneous aggregation over regions, firms, etc. are considered. The data on which the forecasts are based are assumed to be generated by vector ARMA processes. The consequences of the estimation and specification uncertainty for the forecast efficiency are analyzed.