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

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 p 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.

[1]  L. Reichlin,et al.  When Is Growth at Risk? , 2020, Brookings Papers on Economic Activity.

[2]  J. Galbraith,et al.  Asymmetry in unemployment rate forecast errors , 2019, International Journal of Forecasting.

[3]  BOARD OF GOVERNORS OF THE FEDERAL RESERVE SYSTEM , 2000 .

[4]  Enes Makalic,et al.  A Simple Sampler for the Horseshoe Estimator , 2015, IEEE Signal Processing Letters.

[5]  H. Kozumi,et al.  Gibbs sampling methods for Bayesian quantile regression , 2011 .

[6]  Florian Huber,et al.  Tail Forecasting with Multivariate Bayesian Additive Regression Trees , 2021, Working paper (Federal Reserve Bank of Cleveland).

[7]  Gianni De Nicoló,et al.  Forecasting Tail Risks , 2015, SSRN Electronic Journal.

[8]  Florian Huber,et al.  Approximate Bayesian inference and forecasting in huge-dimensional multi-country VARs , 2021 .

[9]  Michael Pfarrhofer.  Tail forecasts of inflation using time-varying parameter quantile regressions , 2021 .

[10]  L. Reichlin,et al.  Financial Variables as Predictors of Real Growth Vulnerability , 2020, SSRN Electronic Journal.

[11]  Dimitris Korobilis,et al.  Quantile regression forecasts of inflation under model uncertainty , 2017 .

[12]  A. Kottas,et al.  A Bayesian Nonparametric Approach to Inference for Quantile Regression , 2010 .

[13]  Ivan Petrella,et al.  Modeling and Forecasting Macroeconomic Downside Risk , 2020 .

[14]  Tae-Yong Doh,et al.  Assessing Macroeconomic Tail Risks in a Data-Rich Environment , 2019 .

[15]  Sebastiano Manzan,et al.  Forecasting the Distribution of Economic Variables in a Data-Rich Environment , 2012 .

[16]  Nina Boyarchenko,et al.  Vulnerable Growth , 2016, American Economic Review.

[17]  Dawit Zerom,et al.  Are Macroeconomic Variables Useful for Forecasting the Distribution of U.S. Inflation? , 2009 .

[18]  Gregor Kastner,et al.  Ancillarity-sufficiency interweaving strategy (ASIS) for boosting MCMC estimation of stochastic volatility models , 2014, Comput. Stat. Data Anal..

[19]  L. Lima,et al.  Constructing Density Forecasts from Quantile Regressions , 2012 .

[20]  Radford M. Neal.  Slice Sampling , 2003, The Annals of Statistics.

[21]  E. Ruiz,et al.  Growth in stress , 2019, International Journal of Forecasting.

[22]  Sheheryar Malik,et al.  The Term Structure of Growth-at-Risk , 2018, SSRN Electronic Journal.

[23]  T. Lancaster,et al.  Bayesian Quantile Regression , 2005 .

[24]  H. Chipman,et al.  Bayesian CART Model Search , 1998 .

[25]  Stefano Giglio,et al.  Systemic Risk and the Macroeconomy: An Empirical Evaluation , 2015 .

[26]  Dimitris Korobilis,et al.  The Time-Varying Evolution of Inflation Risks , 2021, SSRN Electronic Journal.

[27]  Marek Jarociński,et al.  Vulnerable Growth in the Euro Area: Measuring the Financial Conditions , 2020, Economics Letters.

[28]  T. Gneiting,et al.  Comparing Density Forecasts Using Threshold- and Quantile-Weighted Scoring Rules , 2011 .

[29]  H. Chipman,et al.  BART: Bayesian Additive Regression Trees , 2008, 0806.3286.

[30]  Todd E. Clark,et al.  Addressing COVID-19 Outliers in BVARs with Stochastic Volatility , 2021, Working paper (Federal Reserve Bank of Cleveland).