Mixed-Frequency Models for Tracking Short-Term Economic Developments in Switzerland

We compare several methods for monitoring short-term economic developments in Switzerland. Based on a large mixed-frequency data set, the following approaches are presented and discussed: a factor-based information combination approach (a dynamic factor model based on the Kalman filter/smoother and estimated by the EM algorithm), a model combination approach resting on MIDAS regression models, and a variable selection approach using a specific-to-general algorithm. In an out-of-sample GDP forecasting exercise, we show that the considered approaches clearly beat relevant benchmarks such as univariate time-series models and models that work with just one indicator. Moreover, we find that the factor model is superior to the other two approaches under investigation. However, forecast pooling of the three methods turns out to be even more promising.