Department of Economics and Business Economics

Timo Teräsvirta

Forecasting inflation with gradual regime shifts and exogenous information

Research output: ResearchWorking paper

Standard

Forecasting inflation with gradual regime shifts and exogenous information. / González, Andrés; Hubrich, Kirstin; Teräsvirta, Timo.

Aarhus : Institut for Økonomi, Aarhus Universitet, 2009.

Research output: ResearchWorking paper

Harvard

González, A, Hubrich, K & Teräsvirta, T 2009 'Forecasting inflation with gradual regime shifts and exogenous information' Institut for Økonomi, Aarhus Universitet, Aarhus.

APA

González, A., Hubrich, K., & Teräsvirta, T. (2009). Forecasting inflation with gradual regime shifts and exogenous information. Aarhus: Institut for Økonomi, Aarhus Universitet.

CBE

González A, Hubrich K, Teräsvirta T. 2009. Forecasting inflation with gradual regime shifts and exogenous information. Aarhus: Institut for Økonomi, Aarhus Universitet.

MLA

González, Andrés, Kirstin Hubrich, and Timo Teräsvirta Forecasting inflation with gradual regime shifts and exogenous information. Aarhus: Institut for Økonomi, Aarhus Universitet. 2009., 30 p.

Vancouver

González A, Hubrich K, Teräsvirta T. Forecasting inflation with gradual regime shifts and exogenous information. Aarhus: Institut for Økonomi, Aarhus Universitet. 2009.

Author

González, Andrés ; Hubrich, Kirstin ; Teräsvirta, Timo. / Forecasting inflation with gradual regime shifts and exogenous information. Aarhus : Institut for Økonomi, Aarhus Universitet, 2009.

Bibtex

@techreport{5a0d4200ed3911dd8f9a000ea68e967b,
title = "Forecasting inflation with gradual regime shifts and exogenous information",
abstract = "In this work, we make use of the shifting-mean autoregressivemodel which is a flexible univariate nonstationary model. It is suitablefor describing characteristic features in inflation series as well as formedium-term forecasting. With this model we decompose the inflationprocess into a slowly moving nonstationary component and dynamicshort-run fluctuations around it. We fit the model to the monthlyeuro area, UK and US inflation series. An important feature of ourmodel is that it provides a way of combining the information in thesample and the a priori information about the quantity to be forecastto form a single inflation forecast. We show, both theoretically and bysimulations, how this is done by using the penalised likelihood in theestimation of model parameters. In forecasting inflation, the centralbank inflation target, if it exists, is a natural example of such priorinformation. We further illustrate the application of our method byan ex post forecasting experiment for euro area and UK inflation. Wefind that that taking the exogenous information into account does im-prove the forecast accuracy compared to that of a linear autoregressivebenchmark model.",
keywords = "Nonlinear forecast, nonlinear model, nonlinear trend, penalised likelihood, structural shift, time-varying parameter",
author = "Andrés González and Kirstin Hubrich and Timo Teräsvirta",
year = "2009",
publisher = "Institut for Økonomi, Aarhus Universitet",
type = "WorkingPaper",
institution = "Institut for Økonomi, Aarhus Universitet",

}

RIS

TY - UNPB

T1 - Forecasting inflation with gradual regime shifts and exogenous information

AU - González,Andrés

AU - Hubrich,Kirstin

AU - Teräsvirta,Timo

PY - 2009

Y1 - 2009

N2 - In this work, we make use of the shifting-mean autoregressivemodel which is a flexible univariate nonstationary model. It is suitablefor describing characteristic features in inflation series as well as formedium-term forecasting. With this model we decompose the inflationprocess into a slowly moving nonstationary component and dynamicshort-run fluctuations around it. We fit the model to the monthlyeuro area, UK and US inflation series. An important feature of ourmodel is that it provides a way of combining the information in thesample and the a priori information about the quantity to be forecastto form a single inflation forecast. We show, both theoretically and bysimulations, how this is done by using the penalised likelihood in theestimation of model parameters. In forecasting inflation, the centralbank inflation target, if it exists, is a natural example of such priorinformation. We further illustrate the application of our method byan ex post forecasting experiment for euro area and UK inflation. Wefind that that taking the exogenous information into account does im-prove the forecast accuracy compared to that of a linear autoregressivebenchmark model.

AB - In this work, we make use of the shifting-mean autoregressivemodel which is a flexible univariate nonstationary model. It is suitablefor describing characteristic features in inflation series as well as formedium-term forecasting. With this model we decompose the inflationprocess into a slowly moving nonstationary component and dynamicshort-run fluctuations around it. We fit the model to the monthlyeuro area, UK and US inflation series. An important feature of ourmodel is that it provides a way of combining the information in thesample and the a priori information about the quantity to be forecastto form a single inflation forecast. We show, both theoretically and bysimulations, how this is done by using the penalised likelihood in theestimation of model parameters. In forecasting inflation, the centralbank inflation target, if it exists, is a natural example of such priorinformation. We further illustrate the application of our method byan ex post forecasting experiment for euro area and UK inflation. Wefind that that taking the exogenous information into account does im-prove the forecast accuracy compared to that of a linear autoregressivebenchmark model.

KW - Nonlinear forecast, nonlinear model, nonlinear trend, penalised likelihood, structural shift, time-varying parameter

M3 - Working paper

BT - Forecasting inflation with gradual regime shifts and exogenous information

PB - Institut for Økonomi, Aarhus Universitet

ER -