Español

Abstract

The direction in which economic activity can take is vital both for policy makers and private agents, however, the publication of the main macroeconomic indicators such as GDP, usually lags for up to 90 days from the end of a quarter. This work is an approximation to the projection of the rate of annual growth of the Real GDP of Ecuador using dynamic factor models and bridge equations, developing a nowcasting exercising that produces consistent estimates through the Principal Component Analysis and the Kalman filter, generating projections 50 days before the official publication. The evaluation of the models was performed under the approach of rolling windows, expanding windows, nowcast and Diebold-Mariano test, and it suggests that the model that provides more accurate forecasts is one that incorporates in the bridge equation, only the factors extracted in the dynamic factor model.