Abstract
We develop a nonparametric 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 heteroscedasticity. Markov chain Monte Carlo (MCMC) estimation is carried out through an efficient and scalable algorithm which can handle large models. The GP-VAR is used to analyze the effects of macroeconomic uncertainty, with a particular emphasis on time variation and asymmetries in the transmission mechanisms.
| Original language | English |
|---|---|
| Pages (from-to) | 27-43 |
| Number of pages | 17 |
| Journal | Journal of Business and Economic Statistics |
| Volume | 43 |
| Issue number | 1 |
| Early online date | 29 Mar 2024 |
| DOIs | |
| Publication status | Published - 2 Jan 2025 |
Funding
Niko Hauzenberger gratefully acknowledges financial support from the Austrian Science Fund (FWF, grant no. ZK 35) and the Jubilaumsfonds of the Oesterreichische Nationalbank (OeNB, grant no. 18718 and 18763), Florian Huber acknowledges financial support from the Austrian Science Fund (FWF, grant no. ZK 35) and the Jubilaumsfonds of the OeNB (grant no. 18304), and Massimiliano Marcellino thanks MUR- Prin 2022- Prot. 20227YZ9JK for partial financial support.
Keywords
- bayesian nonparametrics
- non-linear vector autoregressions
- asymmetric uncertainty shocks