Selecting the regularization parameters in high-dimensional panel data models: Consistency and efficiency

Tomohiro Ando, Jushan Bai

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This article considers panel data models in the presence of a large number of potential predictors and unobservable common factors. The model is estimated by the regularization method together with the principal components procedure. We propose a panel information criterion for selecting the regularization parameter and the number of common factors under a diverging number of predictors. Under the correct model specification, we show that the proposed criterion consistently identifies the true model. If the model is instead misspecified, the proposed criterion achieves asymptotically efficient model selection. Simulation results confirm these theoretical arguments.

Original languageEnglish
Pages (from-to)1-29
Number of pages29
JournalEconometric Reviews
DOIs
Publication statusAccepted/In press - 2016 Mar 15

Fingerprint

Regularization
Predictors
Common factors
Simulation
Information criterion
Model selection
Model specification
Principal components

Keywords

  • Endogeneity
  • factor models
  • heterogeneous coefficients
  • information criterion
  • penalized method
  • smoothly clipped absolute deviation (SCAD)

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Selecting the regularization parameters in high-dimensional panel data models : Consistency and efficiency. / Ando, Tomohiro; Bai, Jushan.

In: Econometric Reviews, 15.03.2016, p. 1-29.

Research output: Contribution to journalArticle

@article{504a0436cf04486bb88aa4d904039d5c,
title = "Selecting the regularization parameters in high-dimensional panel data models: Consistency and efficiency",
abstract = "This article considers panel data models in the presence of a large number of potential predictors and unobservable common factors. The model is estimated by the regularization method together with the principal components procedure. We propose a panel information criterion for selecting the regularization parameter and the number of common factors under a diverging number of predictors. Under the correct model specification, we show that the proposed criterion consistently identifies the true model. If the model is instead misspecified, the proposed criterion achieves asymptotically efficient model selection. Simulation results confirm these theoretical arguments.",
keywords = "Endogeneity, factor models, heterogeneous coefficients, information criterion, penalized method, smoothly clipped absolute deviation (SCAD)",
author = "Tomohiro Ando and Jushan Bai",
year = "2016",
month = "3",
day = "15",
doi = "10.1080/07474938.2015.1092822",
language = "English",
pages = "1--29",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Selecting the regularization parameters in high-dimensional panel data models

T2 - Consistency and efficiency

AU - Ando, Tomohiro

AU - Bai, Jushan

PY - 2016/3/15

Y1 - 2016/3/15

N2 - This article considers panel data models in the presence of a large number of potential predictors and unobservable common factors. The model is estimated by the regularization method together with the principal components procedure. We propose a panel information criterion for selecting the regularization parameter and the number of common factors under a diverging number of predictors. Under the correct model specification, we show that the proposed criterion consistently identifies the true model. If the model is instead misspecified, the proposed criterion achieves asymptotically efficient model selection. Simulation results confirm these theoretical arguments.

AB - This article considers panel data models in the presence of a large number of potential predictors and unobservable common factors. The model is estimated by the regularization method together with the principal components procedure. We propose a panel information criterion for selecting the regularization parameter and the number of common factors under a diverging number of predictors. Under the correct model specification, we show that the proposed criterion consistently identifies the true model. If the model is instead misspecified, the proposed criterion achieves asymptotically efficient model selection. Simulation results confirm these theoretical arguments.

KW - Endogeneity

KW - factor models

KW - heterogeneous coefficients

KW - information criterion

KW - penalized method

KW - smoothly clipped absolute deviation (SCAD)

UR - http://www.scopus.com/inward/record.url?scp=84962110206&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84962110206&partnerID=8YFLogxK

U2 - 10.1080/07474938.2015.1092822

DO - 10.1080/07474938.2015.1092822

M3 - Article

AN - SCOPUS:84962110206

SP - 1

EP - 29

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

ER -