Estimation and forecasting in models with multiple breaks

G.M. Koop, S. Potter

Research output: Contribution to journalArticle

92 Citations (Scopus)

Abstract

This paper develops a new approach to change-point modelling that allows the number of change-points in the observed sample to be unknown. The model we develop assumes that regime durations have a Poisson distribution. It approximately nests the two most common approaches: the time-varying parameter (TVP) model with a change-point every period and the change-point model with a small number of regimes. We focus considerable attention on the construction of reasonable hierarchical priors both for regime durations and for the parameters that characterize each regime. A Markov chain Monte Carlo posterior sampler is constructed to estimate a version of our model, which allows for change in conditional means and variances. We show how real-time forecasting can be done in an efficient manner using sequential importance sampling. Our techniques are found to work well in an empirical exercise involving U.S. GDP growth and inflation. Empirical results suggest that the number of change-points is larger than previously estimated in these series and the implied model is similar to a TVP (with stochastic volatility) model.
LanguageEnglish
Pages763-789
Number of pages26
JournalReview of Economic Studies
Volume74
Issue number3
DOIs
Publication statusPublished - 2007

Fingerprint

Change point
Multiple breaks
Markov chain Monte Carlo
GDP growth
Modeling
Empirical results
Importance sampling
Time-varying parameters
Time-varying parameter model
Stochastic volatility model
Inflation
Poisson distribution
Exercise

Keywords

  • economic estimation
  • economic forecasting
  • modelling
  • multiple breaks

Cite this

@article{787772e4cd6c47089b3fc49b09af981e,
title = "Estimation and forecasting in models with multiple breaks",
abstract = "This paper develops a new approach to change-point modelling that allows the number of change-points in the observed sample to be unknown. The model we develop assumes that regime durations have a Poisson distribution. It approximately nests the two most common approaches: the time-varying parameter (TVP) model with a change-point every period and the change-point model with a small number of regimes. We focus considerable attention on the construction of reasonable hierarchical priors both for regime durations and for the parameters that characterize each regime. A Markov chain Monte Carlo posterior sampler is constructed to estimate a version of our model, which allows for change in conditional means and variances. We show how real-time forecasting can be done in an efficient manner using sequential importance sampling. Our techniques are found to work well in an empirical exercise involving U.S. GDP growth and inflation. Empirical results suggest that the number of change-points is larger than previously estimated in these series and the implied model is similar to a TVP (with stochastic volatility) model.",
keywords = "economic estimation, economic forecasting, modelling, multiple breaks",
author = "G.M. Koop and S. Potter",
year = "2007",
doi = "10.1111/j.1467-937X.2007.00436.x",
language = "English",
volume = "74",
pages = "763--789",
journal = "Review of Economic Studies",
issn = "0034-6527",
number = "3",

}

Estimation and forecasting in models with multiple breaks. / Koop, G.M.; Potter, S.

In: Review of Economic Studies, Vol. 74, No. 3, 2007, p. 763-789.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Estimation and forecasting in models with multiple breaks

AU - Koop, G.M.

AU - Potter, S.

PY - 2007

Y1 - 2007

N2 - This paper develops a new approach to change-point modelling that allows the number of change-points in the observed sample to be unknown. The model we develop assumes that regime durations have a Poisson distribution. It approximately nests the two most common approaches: the time-varying parameter (TVP) model with a change-point every period and the change-point model with a small number of regimes. We focus considerable attention on the construction of reasonable hierarchical priors both for regime durations and for the parameters that characterize each regime. A Markov chain Monte Carlo posterior sampler is constructed to estimate a version of our model, which allows for change in conditional means and variances. We show how real-time forecasting can be done in an efficient manner using sequential importance sampling. Our techniques are found to work well in an empirical exercise involving U.S. GDP growth and inflation. Empirical results suggest that the number of change-points is larger than previously estimated in these series and the implied model is similar to a TVP (with stochastic volatility) model.

AB - This paper develops a new approach to change-point modelling that allows the number of change-points in the observed sample to be unknown. The model we develop assumes that regime durations have a Poisson distribution. It approximately nests the two most common approaches: the time-varying parameter (TVP) model with a change-point every period and the change-point model with a small number of regimes. We focus considerable attention on the construction of reasonable hierarchical priors both for regime durations and for the parameters that characterize each regime. A Markov chain Monte Carlo posterior sampler is constructed to estimate a version of our model, which allows for change in conditional means and variances. We show how real-time forecasting can be done in an efficient manner using sequential importance sampling. Our techniques are found to work well in an empirical exercise involving U.S. GDP growth and inflation. Empirical results suggest that the number of change-points is larger than previously estimated in these series and the implied model is similar to a TVP (with stochastic volatility) model.

KW - economic estimation

KW - economic forecasting

KW - modelling

KW - multiple breaks

UR - http://dx.doi.org/10.1111/j.1467-937X.2007.00436.x

U2 - 10.1111/j.1467-937X.2007.00436.x

DO - 10.1111/j.1467-937X.2007.00436.x

M3 - Article

VL - 74

SP - 763

EP - 789

JO - Review of Economic Studies

T2 - Review of Economic Studies

JF - Review of Economic Studies

SN - 0034-6527

IS - 3

ER -