CMStatistics 2022: Start Registration
View Submission - CFE
A1929
Title: Gaussian process vector autoregressions and macroeconomic uncertainty Authors:  Niko Hauzenberger - University of Strathclyde (United Kingdom) [presenting]
Florian Huber - University of Salzburg (Austria)
M. Marcellino - Bocconi University (Italy)
Nico Petz - University of Salzburg (Austria)
Abstract: A non-parametric multivariate time series model is developed that remains agnostic on the precise relationship between a (possibly) large set of macroeconomic time series and their lagged values. The main building block of our model is a Gaussian Process prior on the functional relationship that determines the conditional mean of the model, hence the name of Gaussian Process vector autoregression (GP-VAR). We control for changes in the error variances by introducing a stochastic volatility specification. To facilitate computation in high dimensions and to introduce convenient statistical properties tailored to match stylized facts commonly observed in macro time series, we assume that the covariance of the Gaussian Process is scaled by the latent volatility factors. We illustrate the use of the GP-VAR by {comparing it with other nonlinear and time-varying models in a forecasting exercise. Moreover, we use the GP-VAR to analyze} the effects of macroeconomic uncertainty, with a particular emphasis on time variation and asymmetries in the transmission mechanisms.