Predictive Density Combination Using a Tree-Based Synthesis Function

60 Pages Posted: 22 Nov 2023

See all articles by Tony Chernis

Tony Chernis

Bank of Canada

Niko Hauzenberger

University of Salzburg

Florian Huber

University of Salzburg

Gary Koop

University of Strathclyde, Glasgow - Strathclyde Business School - Department of Economics

James Mitchell

Federal Reserve Bank of Cleveland

Date Written: November 21, 2023

Abstract

Bayesian predictive synthesis (BPS) provides a method for combining multiple predictive distributions based on agent/expert opinion analysis theory and encompasses a range of existing density forecast pooling methods. The key ingredient in BPS is a “synthesis” function. This is typically specified parametrically as a dynamic linear regression. In this paper, we develop a nonparametric treatment of the synthesis function using regression trees. We show the advantages of our tree-based approach in two macroeconomic forecasting applications. The first uses density forecasts for GDP growth from the euro area’s Survey of Professional Forecasters. The second combines density forecasts of US inflation produced by many regression models involving different predictors. Both applications demonstrate the benefits – in terms of improved forecast accuracy and interpretability – of modeling the synthesis function nonparametrically.

Keywords: Forecast density combination, Bayesian nonparametrics, Bayesian predictive synthesis

JEL Classification: C11, C32, C53

Suggested Citation

Chernis, Tony and Hauzenberger, Niko and Huber, Florian and Koop, Gary and Mitchell, James, Predictive Density Combination Using a Tree-Based Synthesis Function (November 21, 2023). FRB of Cleveland Working Paper No. 23-30, https://doi.org/10.26509/frbc-wp-202330, Available at SSRN: https://ssrn.com/abstract=4640161 or http://dx.doi.org/10.2139/ssrn.4640161

Tony Chernis

Bank of Canada ( email )

234 Wellington Street
Ontario, Ottawa K1A 0G9
Canada

Niko Hauzenberger

University of Salzburg ( email )

Akademiestraße 26
Salzburg, Salzburg 5020
Austria

Florian Huber

University of Salzburg ( email )

Akademiestraße 26
Salzburg, Salzburg 5020
Austria

Gary Koop

University of Strathclyde, Glasgow - Strathclyde Business School - Department of Economics ( email )

100 Cathedral Street
Glasgow G4 0LN
United Kingdom

James Mitchell (Contact Author)

Federal Reserve Bank of Cleveland ( email )

East 6th & Superior
Cleveland, OH 44101-1387
United States

HOME PAGE: http://https://www.clevelandfed.org/en/our-research/economists/james-mitchell.aspx

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
23
Abstract Views
165
PlumX Metrics