Bootstrapping Structural Change Tests

Otilia Boldea, Adriana Cornea-Madeira, Alastair R. Hall

Research output: Working paperOther research output

Abstract

This paper analyses the use of bootstrap methods to test for parameter change in linear models estimated via Two Stage Least Squares (2SLS). Two types of test are considered: one where the null hypothesis is of no change and the alternative hypothesis involves discrete change at k unknown break-points in the sample; and a second test where the null hypothesis is that there is discrete parameter change at l break-points in the sample against an alternative in which the parameters change at l + 1 break-points. In both cases, we consider inferences based on a sup-W ald-type statistic using either the wild recursive bootstrap or the wild fixed bootstrap. We establish the asymptotic validity of these bootstrap tests under a set of general conditions that allow the errors to exhibit conditional and/or unconditional heteroskedasticity, and report results from a simulation study that indicate the tests yield reliable inferences in the sample sizes often encountered in macroeconomics. The analysis covers the cases where the first-stage estimation of 2SLS involves a model whose parameters are either constant or themselves subject to discrete parameter change. If the errors exhibit unconditional heteroskedasticity and/or the reduced form is unstable then the bootstrap methods are particularly attractive because the limiting distributions of the test statistics are not pivotal.
Original languageEnglish
Place of PublicationIthaca
PublisherCornell University Library
Publication statusPublished - 13 Nov 2018

Publication series

NameArXiv
Volume1811.04125

Fingerprint

Structural change
Bootstrapping
Inference
Heteroskedasticity
Bootstrap
Bootstrap method
Two-stage least squares
Test statistic
Sample size
Statistics
Macroeconomics
Limiting distribution
Reduced form
Simulation study
Bootstrap test

Keywords

  • multiple break points
  • instrumental variables estimation
  • two-stage least squares
  • wild bootstrap
  • recursive bootstrap
  • fixed-regressor bootstrap
  • heteroskedasticity

Cite this

Boldea, O., Cornea-Madeira, A., & Hall, A. R. (2018). Bootstrapping Structural Change Tests. (ArXiv; Vol. 1811.04125). Ithaca: Cornell University Library.
Boldea, Otilia ; Cornea-Madeira, Adriana ; Hall, Alastair R. / Bootstrapping Structural Change Tests. Ithaca : Cornell University Library, 2018. (ArXiv).
@techreport{1d969dc0891342df89c74e189eec8aae,
title = "Bootstrapping Structural Change Tests",
abstract = "This paper analyses the use of bootstrap methods to test for parameter change in linear models estimated via Two Stage Least Squares (2SLS). Two types of test are considered: one where the null hypothesis is of no change and the alternative hypothesis involves discrete change at k unknown break-points in the sample; and a second test where the null hypothesis is that there is discrete parameter change at l break-points in the sample against an alternative in which the parameters change at l + 1 break-points. In both cases, we consider inferences based on a sup-W ald-type statistic using either the wild recursive bootstrap or the wild fixed bootstrap. We establish the asymptotic validity of these bootstrap tests under a set of general conditions that allow the errors to exhibit conditional and/or unconditional heteroskedasticity, and report results from a simulation study that indicate the tests yield reliable inferences in the sample sizes often encountered in macroeconomics. The analysis covers the cases where the first-stage estimation of 2SLS involves a model whose parameters are either constant or themselves subject to discrete parameter change. If the errors exhibit unconditional heteroskedasticity and/or the reduced form is unstable then the bootstrap methods are particularly attractive because the limiting distributions of the test statistics are not pivotal.",
keywords = "multiple break points, instrumental variables estimation, two-stage least squares, wild bootstrap, recursive bootstrap, fixed-regressor bootstrap, heteroskedasticity",
author = "Otilia Boldea and Adriana Cornea-Madeira and Hall, {Alastair R.}",
year = "2018",
month = "11",
day = "13",
language = "English",
series = "ArXiv",
publisher = "Cornell University Library",
type = "WorkingPaper",
institution = "Cornell University Library",

}

Boldea, O, Cornea-Madeira, A & Hall, AR 2018 'Bootstrapping Structural Change Tests' ArXiv, vol. 1811.04125, Cornell University Library, Ithaca.

Bootstrapping Structural Change Tests. / Boldea, Otilia; Cornea-Madeira, Adriana; Hall, Alastair R.

Ithaca : Cornell University Library, 2018. (ArXiv; Vol. 1811.04125).

Research output: Working paperOther research output

TY - UNPB

T1 - Bootstrapping Structural Change Tests

AU - Boldea, Otilia

AU - Cornea-Madeira, Adriana

AU - Hall, Alastair R.

PY - 2018/11/13

Y1 - 2018/11/13

N2 - This paper analyses the use of bootstrap methods to test for parameter change in linear models estimated via Two Stage Least Squares (2SLS). Two types of test are considered: one where the null hypothesis is of no change and the alternative hypothesis involves discrete change at k unknown break-points in the sample; and a second test where the null hypothesis is that there is discrete parameter change at l break-points in the sample against an alternative in which the parameters change at l + 1 break-points. In both cases, we consider inferences based on a sup-W ald-type statistic using either the wild recursive bootstrap or the wild fixed bootstrap. We establish the asymptotic validity of these bootstrap tests under a set of general conditions that allow the errors to exhibit conditional and/or unconditional heteroskedasticity, and report results from a simulation study that indicate the tests yield reliable inferences in the sample sizes often encountered in macroeconomics. The analysis covers the cases where the first-stage estimation of 2SLS involves a model whose parameters are either constant or themselves subject to discrete parameter change. If the errors exhibit unconditional heteroskedasticity and/or the reduced form is unstable then the bootstrap methods are particularly attractive because the limiting distributions of the test statistics are not pivotal.

AB - This paper analyses the use of bootstrap methods to test for parameter change in linear models estimated via Two Stage Least Squares (2SLS). Two types of test are considered: one where the null hypothesis is of no change and the alternative hypothesis involves discrete change at k unknown break-points in the sample; and a second test where the null hypothesis is that there is discrete parameter change at l break-points in the sample against an alternative in which the parameters change at l + 1 break-points. In both cases, we consider inferences based on a sup-W ald-type statistic using either the wild recursive bootstrap or the wild fixed bootstrap. We establish the asymptotic validity of these bootstrap tests under a set of general conditions that allow the errors to exhibit conditional and/or unconditional heteroskedasticity, and report results from a simulation study that indicate the tests yield reliable inferences in the sample sizes often encountered in macroeconomics. The analysis covers the cases where the first-stage estimation of 2SLS involves a model whose parameters are either constant or themselves subject to discrete parameter change. If the errors exhibit unconditional heteroskedasticity and/or the reduced form is unstable then the bootstrap methods are particularly attractive because the limiting distributions of the test statistics are not pivotal.

KW - multiple break points

KW - instrumental variables estimation

KW - two-stage least squares

KW - wild bootstrap

KW - recursive bootstrap

KW - fixed-regressor bootstrap

KW - heteroskedasticity

M3 - Working paper

T3 - ArXiv

BT - Bootstrapping Structural Change Tests

PB - Cornell University Library

CY - Ithaca

ER -

Boldea O, Cornea-Madeira A, Hall AR. Bootstrapping Structural Change Tests. Ithaca: Cornell University Library. 2018 Nov 13. (ArXiv).