VOLUME XXV
FALL 2017

INTEGRATED MODEL OF SHORT-TERM FORECASTING OF THE SPANISH ECONOMY (MIPRED MODEL)

 
ÁNGEL CUEVAS
Macroeconomic Research Department Independent Authority for Fiscal Responsibility and UNED Programa de Doctorado en Economía y Empresa
GABRIEL PÉREZ-QUIRÓS
Macroeconomic Analysis Unit Bank of Spain
ENRIQUE M. QUILIS
Macroeconomic Research Department Independent Authority for Fiscal Responsibility
 

This paper presents a methodology for predicting in real-time Gross Domestic Product (GDP) and its demand components (private consumption, public consumption, investment in equipment, investment in construction, exports and imports) simultaneously. The model, on the one hand, consists of a set of dynamic factor models for both GDP and its demand components, which will provide individual forecasts for each. On the other hand, a balancing procedure is incorporated to ensure the transversal consistency of these forecasts, thus providing a consistent set of estimates based on the statistically most useful indicators about current economic activity and demand developments. The methodology is applied to the Spanish economy, presenting real-time quarterly forecasts of GDP and its demand components.

 

Key words: dynamic factor models, short term economic analysis, spanish economy, Kalman filter, forecasting, nowcasting, national accounts, balancing.
JEL classification: C22, C53, C82, E27.


TO DOWNLOAD THIS PAPER