Investigating Growth at Risk Using a Multi-country Non-parametric Quantile Factor Model

TL;DR

This paper introduces a Bayesian non-parametric quantile panel regression model combining BART and latent factors to analyze growth at risk across multiple countries, enhancing flexibility and tail observation accuracy.

Contribution

It develops the QF-BART model that integrates non-parametric BART with a factor structure for improved growth at risk analysis in multi-country panels.

Findings

Enhanced flexibility in modeling growth at risk.

Improved tail risk estimation through cross-country data.

Effective Bayesian MCMC estimation methods.

Abstract

We develop a Bayesian non-parametric quantile panel regression model. Within each quantile, the response function is a convex combination of a linear model and a non-linear function, which we approximate using Bayesian Additive Regression Trees (BART). Cross-sectional information at the pth quantile is captured through a conditionally heteroscedastic latent factor. The non-parametric feature of our model enhances flexibility, while the panel feature, by exploiting cross-country information, increases the number of observations in the tails. We develop Bayesian Markov chain Monte Carlo (MCMC) methods for estimation and forecasting with our quantile factor BART model (QF-BART), and apply them to study growth at risk dynamics in a panel of 11 advanced economies.