Finite sample properties of tests based on prewhitened nonparametric covariance estimators

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Standard

Finite sample properties of tests based on prewhitened nonparametric covariance estimators. / Preinerstorfer, David.

In: Electronic Journal of Statistics, Vol. 11, No. 1, 2017, p. 2097-2167.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Harvard

Preinerstorfer, D 2017, 'Finite sample properties of tests based on prewhitened nonparametric covariance estimators', Electronic Journal of Statistics, vol. 11, no. 1, pp. 2097-2167. https://doi.org/10.1214/17-EJS1281

APA

CBE

MLA

Vancouver

Author

Preinerstorfer, David. / Finite sample properties of tests based on prewhitened nonparametric covariance estimators. In: Electronic Journal of Statistics. 2017 ; Vol. 11, No. 1. pp. 2097-2167.

Bibtex

@article{4a8c6825d4244c9fb4c6d4642ce3f154,
title = "Finite sample properties of tests based on prewhitened nonparametric covariance estimators",
abstract = "We analytically investigate size and power properties of a popular family of procedures for testing linear restrictions on the coefficient vector in a linear regression model with temporally dependent errors. The tests considered are autocorrelation-corrected F-type tests based on prewhitened nonparametric covariance estimators that possibly incorporate a data-dependent bandwidth parameter, e.g., estimators as considered in Andrews and Monahan (1992), Newey and West (1994), or Rho and Shao (2013). For design matrices that are generic in a measure theoretic sense we prove that these tests either suffer from extreme size distortions or from strong power deficiencies. Despite this negative result we demonstrate that a simple adjustment procedure based on artificial regressors can often resolve this problem.",
keywords = "Autocorrelation robustness, prewhitening, size distortion, power deficiency, artificial regressors, AUTOCORRELATION ROBUST-TESTS, MATRIX ESTIMATION, HETEROSKEDASTICITY, SIZE, SELECTION, SERIES, POWER",
author = "David Preinerstorfer",
year = "2017",
doi = "10.1214/17-EJS1281",
language = "English",
volume = "11",
pages = "2097--2167",
journal = "Electronic Journal of Statistics",
issn = "1935-7524",
publisher = "nstitute of Mathematical Statistics",
number = "1",

}

RIS

TY - JOUR

T1 - Finite sample properties of tests based on prewhitened nonparametric covariance estimators

AU - Preinerstorfer, David

PY - 2017

Y1 - 2017

N2 - We analytically investigate size and power properties of a popular family of procedures for testing linear restrictions on the coefficient vector in a linear regression model with temporally dependent errors. The tests considered are autocorrelation-corrected F-type tests based on prewhitened nonparametric covariance estimators that possibly incorporate a data-dependent bandwidth parameter, e.g., estimators as considered in Andrews and Monahan (1992), Newey and West (1994), or Rho and Shao (2013). For design matrices that are generic in a measure theoretic sense we prove that these tests either suffer from extreme size distortions or from strong power deficiencies. Despite this negative result we demonstrate that a simple adjustment procedure based on artificial regressors can often resolve this problem.

AB - We analytically investigate size and power properties of a popular family of procedures for testing linear restrictions on the coefficient vector in a linear regression model with temporally dependent errors. The tests considered are autocorrelation-corrected F-type tests based on prewhitened nonparametric covariance estimators that possibly incorporate a data-dependent bandwidth parameter, e.g., estimators as considered in Andrews and Monahan (1992), Newey and West (1994), or Rho and Shao (2013). For design matrices that are generic in a measure theoretic sense we prove that these tests either suffer from extreme size distortions or from strong power deficiencies. Despite this negative result we demonstrate that a simple adjustment procedure based on artificial regressors can often resolve this problem.

KW - Autocorrelation robustness

KW - prewhitening

KW - size distortion

KW - power deficiency

KW - artificial regressors

KW - AUTOCORRELATION ROBUST-TESTS

KW - MATRIX ESTIMATION

KW - HETEROSKEDASTICITY

KW - SIZE

KW - SELECTION

KW - SERIES

KW - POWER

U2 - 10.1214/17-EJS1281

DO - 10.1214/17-EJS1281

M3 - Journal article

VL - 11

SP - 2097

EP - 2167

JO - Electronic Journal of Statistics

JF - Electronic Journal of Statistics

SN - 1935-7524

IS - 1

ER -