An Assessment of Global Economic Forecasts based on Kalman Filter Techniques and a Low-Order Econometric Model

A global macroeconometric model which represents the combined economies of seven major western industrialized nations is assessed in terms of the accuracy of forecasts of inflation rate and production levels. The forecasts are generated using a Kalman Filter to estimate the current state of the composite economy, and then using the low-order, linear macroeconometric model to forecast future values of inflation and production. The accuracy of the forecasts are viewed in terms of ex post and ex ante comparisons and the Kalman Filter error covariance matrix as a function of the time interval of the forecast.