摘要：This paper proposes a new version of the Spain-STING (Spain, Short-Term INdicator of Growth), a dynamic factor model used by the Banco de España for the short-term forecasting of the Spanish economy. The extended and revised version of the Spain-STING presented in this document includes a forecast for each of the demand components of the National Accounts. In order to select the indicators that best estimate the Spanish GDP and its demand components, several models are considered. Following this strategy, the selected models are those in which the common factor explains the highest proportion of the variance of the GDP. These models allow us to forecast GDP, private consumption, public expenditure, investment in capital goods, construction investment, exports and imports in a consistent way. We assess the predictive power of the models for GDP and its demand components for the period 2005–2017. With regard to the GDP forecast, we find some improvement of the predictive power compared to the previous version of Spain-STING. As for the demand components, we show that our proposal has better predictive power than other possible time series models.
其他摘要：This paper proposes a new version of the Spain-STING (Spain, Short-Term INdicator of Growth), a dynamic factor model used by the Banco de España for the short-term forecasting of the Spanish economy. The extended and revised version of the Spain-STING presented in this document includes a forecast for each of the demand components of the National Accounts. In order to select the indicators that best estimate the Spanish GDP and its demand components, several models are considered. Following this strategy, the selected models are those in which the common factor explains the highest proportion of the variance of the GDP. These models allow us to forecast GDP, private consumption, public expenditure, investment in capital goods, construction investment, exports and imports in a consistent way. We assess the predictive power of the models for GDP and its demand components for the period 2005–2017. With regard to the GDP forecast, we find some improvement of the predictive power compared to the previous version of Spain-STING. As for the demand components, we show that our proposal has better predictive power than other possible time series models.
