Robustness of Bootstrap in Instrumental Variable Regression

Lorenzo Camponovo, Taisuke Otsu

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This paper studies robustness of bootstrap inference methods for instrumental variable (IV) regression models. We consider test statistics for parameter hypotheses based on the IV estimator and generalized method of trimmed moments (GMTM) estimator introduced by Čížek (2008, 2009), and compare the pairs and implied probability bootstrap approximations for these statistics by applying the finite sample breakdown point theory. In particular, we study limiting behaviors of the bootstrap quantiles when the values of outliers diverge to infinity but the sample size is held fixed. The outliers are defined as anomalous observations that can arbitrarily change the value of the statistic of interest. We analyze both just- and overidentified cases and discuss implications of the breakdown point analysis to the size and power properties of bootstrap tests. We conclude that the implied probability bootstrap test using the statistic based on the GMTM estimator shows desirable robustness properties. Simulation studies endorse this conclusion. An empirical example based on Romer's (1993) study on the effect of openness of countries to inflation rates is presented. Several extensions including the analysis for the residual bootstrap are provided.

Original languageEnglish
Pages (from-to)352-393
Number of pages42
JournalEconometric Reviews
Volume34
Issue number3
DOIs
Publication statusPublished - 2015 Mar 7
Externally publishedYes

Fingerprint

Instrumental variables regression
Bootstrap
Robustness
Statistics
Outliers
Generalized method of moments estimator
Breakdown
Bootstrap test
Quantile
Finite sample
Simulation study
Bootstrap inference
Approximation
Sample size
Inflation rate
Instrumental variables estimator
Test statistic
Regression model
Openness

Keywords

  • Bootstrap
  • Breakdown point
  • Instrumental variable regression
  • Robustness

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Robustness of Bootstrap in Instrumental Variable Regression. / Camponovo, Lorenzo; Otsu, Taisuke.

In: Econometric Reviews, Vol. 34, No. 3, 07.03.2015, p. 352-393.

Research output: Contribution to journalArticle

Camponovo, Lorenzo ; Otsu, Taisuke. / Robustness of Bootstrap in Instrumental Variable Regression. In: Econometric Reviews. 2015 ; Vol. 34, No. 3. pp. 352-393.
@article{4c93389e25d1489f8f09dd92a2c8b8ed,
title = "Robustness of Bootstrap in Instrumental Variable Regression",
abstract = "This paper studies robustness of bootstrap inference methods for instrumental variable (IV) regression models. We consider test statistics for parameter hypotheses based on the IV estimator and generalized method of trimmed moments (GMTM) estimator introduced by Č{\'i}žek (2008, 2009), and compare the pairs and implied probability bootstrap approximations for these statistics by applying the finite sample breakdown point theory. In particular, we study limiting behaviors of the bootstrap quantiles when the values of outliers diverge to infinity but the sample size is held fixed. The outliers are defined as anomalous observations that can arbitrarily change the value of the statistic of interest. We analyze both just- and overidentified cases and discuss implications of the breakdown point analysis to the size and power properties of bootstrap tests. We conclude that the implied probability bootstrap test using the statistic based on the GMTM estimator shows desirable robustness properties. Simulation studies endorse this conclusion. An empirical example based on Romer's (1993) study on the effect of openness of countries to inflation rates is presented. Several extensions including the analysis for the residual bootstrap are provided.",
keywords = "Bootstrap, Breakdown point, Instrumental variable regression, Robustness",
author = "Lorenzo Camponovo and Taisuke Otsu",
year = "2015",
month = "3",
day = "7",
doi = "10.1080/07474938.2014.944803",
language = "English",
volume = "34",
pages = "352--393",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Taylor and Francis Ltd.",
number = "3",

}

TY - JOUR

T1 - Robustness of Bootstrap in Instrumental Variable Regression

AU - Camponovo, Lorenzo

AU - Otsu, Taisuke

PY - 2015/3/7

Y1 - 2015/3/7

N2 - This paper studies robustness of bootstrap inference methods for instrumental variable (IV) regression models. We consider test statistics for parameter hypotheses based on the IV estimator and generalized method of trimmed moments (GMTM) estimator introduced by Čížek (2008, 2009), and compare the pairs and implied probability bootstrap approximations for these statistics by applying the finite sample breakdown point theory. In particular, we study limiting behaviors of the bootstrap quantiles when the values of outliers diverge to infinity but the sample size is held fixed. The outliers are defined as anomalous observations that can arbitrarily change the value of the statistic of interest. We analyze both just- and overidentified cases and discuss implications of the breakdown point analysis to the size and power properties of bootstrap tests. We conclude that the implied probability bootstrap test using the statistic based on the GMTM estimator shows desirable robustness properties. Simulation studies endorse this conclusion. An empirical example based on Romer's (1993) study on the effect of openness of countries to inflation rates is presented. Several extensions including the analysis for the residual bootstrap are provided.

AB - This paper studies robustness of bootstrap inference methods for instrumental variable (IV) regression models. We consider test statistics for parameter hypotheses based on the IV estimator and generalized method of trimmed moments (GMTM) estimator introduced by Čížek (2008, 2009), and compare the pairs and implied probability bootstrap approximations for these statistics by applying the finite sample breakdown point theory. In particular, we study limiting behaviors of the bootstrap quantiles when the values of outliers diverge to infinity but the sample size is held fixed. The outliers are defined as anomalous observations that can arbitrarily change the value of the statistic of interest. We analyze both just- and overidentified cases and discuss implications of the breakdown point analysis to the size and power properties of bootstrap tests. We conclude that the implied probability bootstrap test using the statistic based on the GMTM estimator shows desirable robustness properties. Simulation studies endorse this conclusion. An empirical example based on Romer's (1993) study on the effect of openness of countries to inflation rates is presented. Several extensions including the analysis for the residual bootstrap are provided.

KW - Bootstrap

KW - Breakdown point

KW - Instrumental variable regression

KW - Robustness

UR - http://www.scopus.com/inward/record.url?scp=84908434456&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84908434456&partnerID=8YFLogxK

U2 - 10.1080/07474938.2014.944803

DO - 10.1080/07474938.2014.944803

M3 - Article

VL - 34

SP - 352

EP - 393

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

IS - 3

ER -