The School of Economics invites you to a School seminar by Dick van Dijk (Erasmus University Rotterdam).

Uncertainty and the macroeconomy: A real-time out-of-sample evaluation

Co-authors: 

Bart Keijsers (University of Amsterdam)

Abstract

Many economic uncertainty measures have been proposed during the last decade. They are all imperfect proxies of a latent entity and are supposed to be related to output variables. This paper identifies and classifies 15 monthly uncertainty measures based on how they measure economic uncertainty. We show that there is an underlying factor structure and identify two interpretable factors. First, a general economic uncertainty factor with a slight tilt toward financial conditions. Second, a consumer/media confidence index which remains elevated after recessions. Furthermore, we conduct a real-time out-of-sample forecasting analysis on the mean and quantiles of the four coincident variables (industrial production, employment, personal income, and manufacturing and trade sales) to assess whether the relationship with output variables holds out-of-sample. The uncertainty factors hold little predictive power for the mean, but are informative when forecasting the lower quantiles — the left tail — of industrial production and in particular employment. This suggests a nonlinear relationship between economic uncertainty and output.