Nowcasting de inversión: Una estimación en tiempo real con indicadores de alta frecuencia

In this paper I present a real-time estimation of the evolution of the Investment, constructed from a broad set of high frequency economic indicators: known in the literature as Nowcasting. The Nowcast exercise was developed considering three groups of monthly indicators throughout dynamic factor models to forecast Investment growth. Additionally, I conducted a forecast-pooling exercise. Using the Giacomini and White test it was possible to conclude that factor models and the pooling exhibit a better relative predictive capacity than an AR (1) model considered as a benchmark. Furthermore, the inclusion of more indicators does not necessarily improve the predictive capacity.