Short term forecasting with the survey of business intentions of the Spanish Ministry of Industry
Visualitza/Obre
Estadístiques de LA Referencia / Recolecta
Inclou dades d'ús des de 2022
Cita com:
hdl:2099/4315
Tipus de documentArticle
Data publicació1977-12
EditorUniversitat Politècnica de Barcelona. Centre de Càlcul
Condicions d'accésAccés obert
Llevat que s'hi indiqui el contrari, els
continguts d'aquesta obra estan subjectes a la llicència de Creative Commons
:
Reconeixement-NoComercial-SenseObraDerivada 2.5 Espanya
Abstract
This gives an account of an study conducted by the authors to empirically asses the predictive value and causal relationships of the Monthly Survey of Business Intentions of the Spanish Ministry of Industry.
The approach implies some extensions of the BoxJenkins procedures for time series modelling and forecasting. After giving some economic and intuitive justification for our model parametrization, models for the individual series were'estimated. Then two forecasts were obtained for each series: a model based one (offline) and an automatic forecast (with out series model) 'selftuning' predictor. The 'selftuning' predictor provides additional evidence on the 'optimality' of BoxJenkins forecasting methods. Finally a crosscorrelation analysis, with the innovations of pairs of series was used, to establish the relationships between series. When that was óf a feedback free nature, inputoutput models were estimated, and a thir forecast with these models were obtained. The three forecasts were compared in performance.
The main conclusions are: the suitability of the methods used for the study¡ the predic tive value of the Surbey was confirmed, and the 'consistency' of the empresarios answeres, allowing for further econometric analysis. In addition we ferified the need for specification analysis from the raw data, and the lack of efficiency of 'long AR' models.
ISSN0210-8054 (versió paper)
Fitxers | Descripció | Mida | Format | Visualitza |
---|---|---|---|---|
article.pdf | 1,021Mb | Visualitza/Obre |