Abstract
In this work, we have constructed a monthly Brazilian GDP measure based on a state space model that was estimated by Kalman filter imposing a time aggregation constraint: the monthly GDP instantaneous growth rate must add up to the quarterly GDP instantaneous growth rate published by IBGE. The explanatory variables of the model were selected based on statistical criteria and in view of a structural model, which is based on the Asset Pricing Equation. The monthly GDP forecast performance showed better results than that of IBC-Br (BACEN) and IAE (FGV).