A carregar...
Publicação
Pseudo real-time forecasting: a model comparison for Portuguese Gdp
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.56 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
GDP is one of the most important economic indicators, yet it presents a significant publishing delay. Many nowcasting models have proven to be successful and have outperformed standard forecasting regressions. This paper compares different nowcasting approaches for estimating quarterly Portuguese GDP, using estimated factors from mixed frequency real-time data. We discuss the out-of-sample forecasting accuracy for each of the models. Furthermore, we investigate the contribution of current-quarter monthly data to the forecasting performance. The results point to an outperformance of the dynamic model averaging and using current-quarter monthly dataonly improves the forecasts of oneof the models.
Descrição
Palavras-chave
Nowcasting Forecasting Dynamic factor model Dynamic model averaging Mixed data sampling.