Gaussian Process Vector Autoregressions and Macroeconomic Uncertainty

2021/12/03 by Niko Hauzenberger, Florian Huber, Hauzenberger, Niko +5 · 3 citations
Computer Science · Decision Sciences · Engineering · #Econometrics (econ.EM) #FOS: Economics and business #Fault Detection and Control Systems #Forecasting Techniques and Applications #Gaussian Processes and Bayesian Inference

paper · pdf · doi:10.48550/arxiv.2112.01995

openalex publication_date 2021/12/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We develop a non-parametric multivariate time series model 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). A flexible stochastic volatility specification is used to provide additional flexibility and control for heteroskedasticity. Markov chain Monte Carlo (MCMC) estimation is carried out through an efficient and scalable algorithm which can handle large models. The GP-VAR is illustrated by means of simulated data and in a forecasting exercise with US data. 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.

Cited by

• Bayesian Markov-Switching Partial Reduced-Rank Regression
• A Bayesian Gaussian Process Dynamic Factor Model
• SABR-Informed Multitask Gaussian Process: A Synthetic-to-Real Framework for Implied Volatility Surface Construction

Related

• Vector autoregression models with skewness and heavy tails
• BVARs and Stochastic Volatility
• The Time-Varying Multivariate Autoregressive Index Model
• Bayesian Vector Autoregressions
• Origins and Nature of Macroeconomic Instability in Vector Autoregressions