Asymptotic behaviour of temporal aggregates of time series

Abstract

The problem of modelling and forecasting temporal aggregates of time series is discussed. Let z t be the basic series and x T be the m -component temporal aggregatos. Forecasts of future x T may be constructed from data on (i) z t or (ii) x T . It is shown that, for large m , there is no gain in using the basic data if z t is stationary, but considerable gain can be obtained when z t is nonstationary.