Focused information criterion for locally misspecified vector autoregressive models

Jan Lohmeyer, Franz Palm, Hanno Reuvers, Jean-Pierre Urbain

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

This paper investigates the focused information criterion and plug-in average for vector autoregressive models with local-to-zero misspecification. These methods have the advantage of focusing on a quantity of interest rather than aiming at overall model fit. Any (suxfb03;ciently regular) function of the parameters can be used as a quantity of interest. We determine the asymptotic properties and elaborate on the role of the locally misspecified parameters. In particular, we show that the inability to consistently estimate locally misspecified parameters translates into suboptimal selection and averaging. We apply this framework to impulse response analysis. A Monte Carlo simulation study supports our claims.

Original languageEnglish
Pages (from-to)763-792
Number of pages30
JournalEconometric Reviews
Volume38
Issue number7
Early online date8 Feb 2018
DOIs
Publication statusPublished - 9 Aug 2019

Keywords

  • Focused information criteria
  • frequentist model averaging
  • impulse responses
  • local misspecification
  • model selection
  • model uncertainty
  • vector autoregressive models
  • IMPULSE-RESPONSE ANALYSIS
  • ORDER SELECTION
  • PREDICTION

Cite this

Lohmeyer, Jan ; Palm, Franz ; Reuvers, Hanno ; Urbain, Jean-Pierre. / Focused information criterion for locally misspecified vector autoregressive models. In: Econometric Reviews. 2019 ; Vol. 38, No. 7. pp. 763-792.
@article{6a9b7a2d3f214c429bccc922ed0519f1,
title = "Focused information criterion for locally misspecified vector autoregressive models",
abstract = "This paper investigates the focused information criterion and plug-in average for vector autoregressive models with local-to-zero misspecification. These methods have the advantage of focusing on a quantity of interest rather than aiming at overall model fit. Any (suxfb03;ciently regular) function of the parameters can be used as a quantity of interest. We determine the asymptotic properties and elaborate on the role of the locally misspecified parameters. In particular, we show that the inability to consistently estimate locally misspecified parameters translates into suboptimal selection and averaging. We apply this framework to impulse response analysis. A Monte Carlo simulation study supports our claims.",
keywords = "Focused information criteria, frequentist model averaging, impulse responses, local misspecification, model selection, model uncertainty, vector autoregressive models, IMPULSE-RESPONSE ANALYSIS, ORDER SELECTION, PREDICTION",
author = "Jan Lohmeyer and Franz Palm and Hanno Reuvers and Jean-Pierre Urbain",
note = "No data used. Article online : December 11, 2017.",
year = "2019",
month = "8",
day = "9",
doi = "10.1080/07474938.2017.1409410",
language = "English",
volume = "38",
pages = "763--792",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Routledge/Taylor & Francis Group",
number = "7",

}

Focused information criterion for locally misspecified vector autoregressive models. / Lohmeyer, Jan; Palm, Franz; Reuvers, Hanno; Urbain, Jean-Pierre.

In: Econometric Reviews, Vol. 38, No. 7, 09.08.2019, p. 763-792.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Focused information criterion for locally misspecified vector autoregressive models

AU - Lohmeyer, Jan

AU - Palm, Franz

AU - Reuvers, Hanno

AU - Urbain, Jean-Pierre

N1 - No data used. Article online : December 11, 2017.

PY - 2019/8/9

Y1 - 2019/8/9

N2 - This paper investigates the focused information criterion and plug-in average for vector autoregressive models with local-to-zero misspecification. These methods have the advantage of focusing on a quantity of interest rather than aiming at overall model fit. Any (suxfb03;ciently regular) function of the parameters can be used as a quantity of interest. We determine the asymptotic properties and elaborate on the role of the locally misspecified parameters. In particular, we show that the inability to consistently estimate locally misspecified parameters translates into suboptimal selection and averaging. We apply this framework to impulse response analysis. A Monte Carlo simulation study supports our claims.

AB - This paper investigates the focused information criterion and plug-in average for vector autoregressive models with local-to-zero misspecification. These methods have the advantage of focusing on a quantity of interest rather than aiming at overall model fit. Any (suxfb03;ciently regular) function of the parameters can be used as a quantity of interest. We determine the asymptotic properties and elaborate on the role of the locally misspecified parameters. In particular, we show that the inability to consistently estimate locally misspecified parameters translates into suboptimal selection and averaging. We apply this framework to impulse response analysis. A Monte Carlo simulation study supports our claims.

KW - Focused information criteria

KW - frequentist model averaging

KW - impulse responses

KW - local misspecification

KW - model selection

KW - model uncertainty

KW - vector autoregressive models

KW - IMPULSE-RESPONSE ANALYSIS

KW - ORDER SELECTION

KW - PREDICTION

UR - http://www.tandfonline.com/doi/full/10.1080/07474938.2017.1409410

U2 - 10.1080/07474938.2017.1409410

DO - 10.1080/07474938.2017.1409410

M3 - Article

VL - 38

SP - 763

EP - 792

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

IS - 7

ER -