The real time ability of financial and real variables in nowcasting

In this paper we develop an approach to choose the variables useful for the nowcasting (short-term forecasting) of growth rate, that is based on the real time ability of financial and real variables to reproduce reference time series. We assess, for Germany and Italy, the impact of real and financial variables in estimating smoothed gross domestic product (GDP) from 2003 to 2012. In synthesis, the innovation of our work is the one to focus both on a limited number of series both on a large dataset of economic variables, by implementing the dynamic factor model.