2021
DOI: 10.1080/07350015.2021.1990772
|View full text |Cite
|
Sign up to set email alerts
|

Fast and Flexible Bayesian Inference in Time-varying Parameter Regression Models

Abstract: We would like to thank the participants of the NBP Workshop on Forecasting (Warsaw, 2019), the European Seminar on Bayesian Econometrics (Madrid, 2021) and internal seminars at the University of Salzburg, the FAU Erlangen-Nuremberg and the ECB, four anonymous referees as well as Anna Stelzer, Michael Pfarrhofer and Paul Hofmarcher for helpful comments and suggestions.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
22
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
5
4

Relationship

6
3

Authors

Journals

citations
Cited by 25 publications
(22 citation statements)
references
References 65 publications
0
22
0
Order By: Relevance
“…There has been an explosion of interest in carrying out structural analysis and forecasting with time-varying parameter (TVP) regressions and Vector Autoregressions (TVP-VARs) in recent years, with many of the papers employing Bayesian methods. An incomplete survey of significant recent Bayesian contributions includes Cogley and Sargent (2005), Primiceri (2005), Chan et al (2012), Dangl and Halling (2012), Koop and Korobilis (2012), Korobilis (2013), D'Agostino et al (2013), Groen et al (2013), Nakajima and West (2013), Belmonte et al (2014), Kalli and Griffin (2014), Feldkircher et al (2017), Kowal et al (2017), Uribe and Lopes (2017), Koop and Korobilis (2018), Ročková and McAlinn (2018), Bitto and Frühwirth-Schnatter (2019), Hauzenberger et al (2019), Korobilis (2019), Paul (2019) and Huber et al (forthcoming).…”
Section: Introductionmentioning
confidence: 99%
“…There has been an explosion of interest in carrying out structural analysis and forecasting with time-varying parameter (TVP) regressions and Vector Autoregressions (TVP-VARs) in recent years, with many of the papers employing Bayesian methods. An incomplete survey of significant recent Bayesian contributions includes Cogley and Sargent (2005), Primiceri (2005), Chan et al (2012), Dangl and Halling (2012), Koop and Korobilis (2012), Korobilis (2013), D'Agostino et al (2013), Groen et al (2013), Nakajima and West (2013), Belmonte et al (2014), Kalli and Griffin (2014), Feldkircher et al (2017), Kowal et al (2017), Uribe and Lopes (2017), Koop and Korobilis (2018), Ročková and McAlinn (2018), Bitto and Frühwirth-Schnatter (2019), Hauzenberger et al (2019), Korobilis (2019), Paul (2019) and Huber et al (forthcoming).…”
Section: Introductionmentioning
confidence: 99%
“…The resulting model can be interpreted as a regime switching model with an unknown number of regimes and a diagonal transition probability matrix. This specification is closely related to the TVP regression model developed in Hauzenberger et al (2022), which uses sparse finite mixtures to model the time variation in the coefficients.…”
Section: How Bart Can Approximate Tvps: An Illustrationmentioning
confidence: 99%
“…Structural break models, Markov switching models (see Sims and Zha, 2006), and time-varying parameter (TVP) models are some prominent examples. TVP regressions and TVP-VARs, in particular, have been highly successful for structural macroeconomic analysis and forecasting (see, for example, Dangl and Halling, 2012;D'Agostino et al, 2013;Koop and Korobilis, 2013;Belmonte et al, 2014;Bitto and Frühwirth-Schnatter, 2019;Korobilis, 2021;Huber et al, 2021;Hauzenberger et al, 2022). In the TVP-VAR literature, parameters are assumed to evolve according to random walks or autoregressive processes.…”
Section: Introductionmentioning
confidence: 99%
“…Fortunately, there exist algorithms which allow us to avoid computing Cholesky factors altogether. The algorithm we use is based on Cong, Chen, and Zhou (2017), which has recently been applied to economic data in Hauzenberger, et al (2021).…”
Section: Fast Sampling Using Singular Value Decompositionsmentioning
confidence: 99%