Structural break inference using information criteria in models estimated by two stage least squares

Alastair R. Hall, Denise R. Osborn, Nikolaos Sakkas

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper makes two contributions in relation to the use of information criteria for inference on structural breaks when the coefficients of a linear model with endogenous regressors may experience multiple changes. Firstly, we show that suitably defined information criteria yield consistent estimators of the number of breaks, when employed in the second stage of a two-stage least squares (2SLS) procedure with breaks in the reduced form taken into account in the first
stage. Secondly, a Monte Carlo analysis investigates the finite sample performance of a range of criteria based on BIC, HQIC and AIC for equations estimated by 2SLS. Versions of the consistent criteria BIC and HQIC perform well overall when the penalty term weights estimation of each break point more heavily than estimation of each coefficient, while AIC is inconsistent and badly
over-estimates the number of true breaks.
Original languageEnglish
Pages (from-to)741-762
Number of pages22
JournalJournal of Time Series Analysis
Volume36
Issue number5
Early online date22 Jan 2015
DOIs
Publication statusPublished - 1 Sep 2015

Fingerprint

Two-stage Least Squares
Structural Breaks
Information Criterion
Consistent Estimator
Coefficient
Inconsistent
Penalty
Linear Model
Term
Model
Estimate
Range of data
Inference
Information criterion
Two-stage least squares
Coefficients
Structural breaks
Experience
Form
Monte Carlo analysis

Cite this

Structural break inference using information criteria in models estimated by two stage least squares. / Hall, Alastair R.; Osborn, Denise R.; Sakkas, Nikolaos.

In: Journal of Time Series Analysis, Vol. 36, No. 5, 01.09.2015, p. 741-762.

Research output: Contribution to journalArticle

@article{58425e5c234e4a22b52c8871318e6ffc,
title = "Structural break inference using information criteria in models estimated by two stage least squares",
abstract = "This paper makes two contributions in relation to the use of information criteria for inference on structural breaks when the coefficients of a linear model with endogenous regressors may experience multiple changes. Firstly, we show that suitably defined information criteria yield consistent estimators of the number of breaks, when employed in the second stage of a two-stage least squares (2SLS) procedure with breaks in the reduced form taken into account in the firststage. Secondly, a Monte Carlo analysis investigates the finite sample performance of a range of criteria based on BIC, HQIC and AIC for equations estimated by 2SLS. Versions of the consistent criteria BIC and HQIC perform well overall when the penalty term weights estimation of each break point more heavily than estimation of each coefficient, while AIC is inconsistent and badlyover-estimates the number of true breaks.",
author = "Hall, {Alastair R.} and Osborn, {Denise R.} and Nikolaos Sakkas",
year = "2015",
month = "9",
day = "1",
doi = "10.1111/jtsa.12107",
language = "English",
volume = "36",
pages = "741--762",
journal = "Journal of Time Series Analysis",
issn = "0143-9782",
publisher = "Wiley-Blackwell",
number = "5",

}

TY - JOUR

T1 - Structural break inference using information criteria in models estimated by two stage least squares

AU - Hall, Alastair R.

AU - Osborn, Denise R.

AU - Sakkas, Nikolaos

PY - 2015/9/1

Y1 - 2015/9/1

N2 - This paper makes two contributions in relation to the use of information criteria for inference on structural breaks when the coefficients of a linear model with endogenous regressors may experience multiple changes. Firstly, we show that suitably defined information criteria yield consistent estimators of the number of breaks, when employed in the second stage of a two-stage least squares (2SLS) procedure with breaks in the reduced form taken into account in the firststage. Secondly, a Monte Carlo analysis investigates the finite sample performance of a range of criteria based on BIC, HQIC and AIC for equations estimated by 2SLS. Versions of the consistent criteria BIC and HQIC perform well overall when the penalty term weights estimation of each break point more heavily than estimation of each coefficient, while AIC is inconsistent and badlyover-estimates the number of true breaks.

AB - This paper makes two contributions in relation to the use of information criteria for inference on structural breaks when the coefficients of a linear model with endogenous regressors may experience multiple changes. Firstly, we show that suitably defined information criteria yield consistent estimators of the number of breaks, when employed in the second stage of a two-stage least squares (2SLS) procedure with breaks in the reduced form taken into account in the firststage. Secondly, a Monte Carlo analysis investigates the finite sample performance of a range of criteria based on BIC, HQIC and AIC for equations estimated by 2SLS. Versions of the consistent criteria BIC and HQIC perform well overall when the penalty term weights estimation of each break point more heavily than estimation of each coefficient, while AIC is inconsistent and badlyover-estimates the number of true breaks.

U2 - 10.1111/jtsa.12107

DO - 10.1111/jtsa.12107

M3 - Article

VL - 36

SP - 741

EP - 762

JO - Journal of Time Series Analysis

JF - Journal of Time Series Analysis

SN - 0143-9782

IS - 5

ER -