A new method for forecasting annual flow macroeconomic aggregates using monthly indicators is developed. A continuous time dynamic model is employed in order to obtain the exact discrete error correction model. The FIML estimation method is applied on the exact discrete error correction model to derive parameter estimates that are invariant to the sampling interval. The monthly indicators can be either I(0) or I(1). In the last case the system error correction format has an advantage in the estimation by encompassing the case cointegration.

(pdf)