Chapter 6 Forecasting with VARMA Models

摘要: Abstract Vector autoregressive moving-average (VARMA) processes are suitable models for producing linear forecasts of sets time series variables. They provide parsimonious representations data generation processes. The setup these in the presence stationary and cointegrated variables is considered. Moreover, unique or identified parameterizations based on echelon form presented. Model specification, estimation, model checking forecasting discussed. Special attention paid to issues related contemporaneously temporally aggregated VARMA Predictors alternatively past information disaggregated compared.