Michael Pfarrhofer

Michael Pfarrhofer

Assistant Professor

Investigating growth-at-risk using a multicountry non-parametric quantile factor model

less than 1 minute read

Published:

“Investigating Growth-at-Risk Using a Multicountry Non-parametric Quantile Factor Model” [DOI] with Todd E. Clark, Florian Huber, Gary Koop and Massimiliano Marcellino is now forthcoming in the Journal of Business & Economic Statistics.

We develop a non-parametric quantile panel regression model. Within each quantile, the response function is a combination of linear and nonlinear parts, which we approximate using Bayesian Additive Regression Trees (BART). Cross-sectional information is captured through a conditionally heteroskedastic latent factor. The non-parametric feature enhances flexibility, while the panel feature increases the number of observations in the tails. We develop Bayesian methods for inference and apply several versions of the model to study growth-at-risk dynamics in a panel of 11 advanced economies. Our framework usually improves upon single-country quantile models in recursive growth forecast comparisons.

Codes and replication files are available at mpfarrho/qf-bart.

You May Also Enjoy

Bayesian nonparametric methods for macroeconomic forecasting

less than 1 minute read

Published:

Here is a [draft] of our chapter “Bayesian nonparametric methods for macroeconomic forecasting” (joint work with Massimiliano Marcellino) prepared for the Handbook of Macroeconomic Forecasting which is edited by Mike Clements and Ana Galvao.

General Bayesian time-varying parameter vector autoregressions for modeling government bond yields

less than 1 minute read

Published:

“General Bayesian time-varying parameter vector autoregressions for modeling government bond yields” [DOI], with Manfred M. Fischer, Niko Hauzenberger and Florian Huber, on flexibly modeling structural breaks in the dynamic evolution of government bond yields is now out in print in the January/February 2023 issue of Journal of Applied Econometrics.

Nowcasting in a pandemic using non-parametric mixed frequency VARs

less than 1 minute read

Published:

“Nowcasting in a pandemic using non-parametric mixed frequency VARs” [DOI], with Florian Huber, Gary Koop, Luca Onorante and Josef Schreiner, on mixed-frequency econometrics for multivariate regression trees is now out in print in the January 2023 issue of Journal of Econometrics.