Logo
Munich Personal RePEc Archive

Model Averaging in Markov-Switching Models: Predicting National Recessions with Regional Data

Guérin, Pierre and Leiva-Leon, Danilo (2014): Model Averaging in Markov-Switching Models: Predicting National Recessions with Regional Data.

This is the latest version of this item.

[thumbnail of MPRA_paper_60250.pdf]
Preview
PDF
MPRA_paper_60250.pdf

Download (983kB) | Preview

Abstract

This paper estimates and forecasts U.S. business cycle turning points with state-level data. The probabilities of recession are obtained from univariate and multivariate regime-switching models based on a pairwise combination of national and state-level data. We use two classes of combination schemes to summarize the information from these models: Bayesian Model Averaging and Dynamic Model Averaging. In addition, we suggest the use of combination schemes based on the past predictive ability of a given model to estimate regimes. Both simulation and empirical exercises underline the utility of such combination schemes. Moreover, our best specification provides timely updates of the U.S. business cycles. In particular, the estimated turning points from this specification largely precede the announcements of business cycle turning points from the NBER business cycle dating committee, and compare favorably with competing models.

Available Versions of this Item

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.