Standardization and estimation of factor numbers for panel data

Ryan Greenaway-McGrevy, Chirok Han, Donggyu Sul

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Practitioners often standardize panel data before estimating a factor model. In this paper we show an example that the standardization leads to inconsistent estimation of the factor number. When the common component exhibits strong heteroskedasticity, the conventional eigenvalue-based decompositions are consistent but standardization does not necessarily result in consistent estimation. To overcome this issue, we recommend using a "minimum-rule" whereby the minimum factor-number estimated from both the conventional and standardized panel is used. Monte Carlo studies and an empirical application are provided.

Original languageEnglish
Pages (from-to)79-88
Number of pages10
JournalJournal of Economic Theory and Econometrics
Volume23
Issue number2
Publication statusPublished - 2012 Jun 1

Fingerprint

Factors
Standardization
Panel data
Monte Carlo study
Decomposition
Heteroskedasticity
Eigenvalues
Common component

Keywords

  • Bai-Ng criteria
  • Factor model
  • Panel data
  • Principal components estimator
  • Selection criteria
  • Standarization

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Standardization and estimation of factor numbers for panel data. / Greenaway-McGrevy, Ryan; Han, Chirok; Sul, Donggyu.

In: Journal of Economic Theory and Econometrics, Vol. 23, No. 2, 01.06.2012, p. 79-88.

Research output: Contribution to journalArticle

Greenaway-McGrevy, Ryan ; Han, Chirok ; Sul, Donggyu. / Standardization and estimation of factor numbers for panel data. In: Journal of Economic Theory and Econometrics. 2012 ; Vol. 23, No. 2. pp. 79-88.
@article{1c3e7b15b2534153aefcbe7d5e1c01d3,
title = "Standardization and estimation of factor numbers for panel data",
abstract = "Practitioners often standardize panel data before estimating a factor model. In this paper we show an example that the standardization leads to inconsistent estimation of the factor number. When the common component exhibits strong heteroskedasticity, the conventional eigenvalue-based decompositions are consistent but standardization does not necessarily result in consistent estimation. To overcome this issue, we recommend using a {"}minimum-rule{"} whereby the minimum factor-number estimated from both the conventional and standardized panel is used. Monte Carlo studies and an empirical application are provided.",
keywords = "Bai-Ng criteria, Factor model, Panel data, Principal components estimator, Selection criteria, Standarization",
author = "Ryan Greenaway-McGrevy and Chirok Han and Donggyu Sul",
year = "2012",
month = "6",
day = "1",
language = "English",
volume = "23",
pages = "79--88",
journal = "Journal of Economic Theory and Econometrics",
issn = "1229-2893",
publisher = "Korean Econometric Society",
number = "2",

}

TY - JOUR

T1 - Standardization and estimation of factor numbers for panel data

AU - Greenaway-McGrevy, Ryan

AU - Han, Chirok

AU - Sul, Donggyu

PY - 2012/6/1

Y1 - 2012/6/1

N2 - Practitioners often standardize panel data before estimating a factor model. In this paper we show an example that the standardization leads to inconsistent estimation of the factor number. When the common component exhibits strong heteroskedasticity, the conventional eigenvalue-based decompositions are consistent but standardization does not necessarily result in consistent estimation. To overcome this issue, we recommend using a "minimum-rule" whereby the minimum factor-number estimated from both the conventional and standardized panel is used. Monte Carlo studies and an empirical application are provided.

AB - Practitioners often standardize panel data before estimating a factor model. In this paper we show an example that the standardization leads to inconsistent estimation of the factor number. When the common component exhibits strong heteroskedasticity, the conventional eigenvalue-based decompositions are consistent but standardization does not necessarily result in consistent estimation. To overcome this issue, we recommend using a "minimum-rule" whereby the minimum factor-number estimated from both the conventional and standardized panel is used. Monte Carlo studies and an empirical application are provided.

KW - Bai-Ng criteria

KW - Factor model

KW - Panel data

KW - Principal components estimator

KW - Selection criteria

KW - Standarization

UR - http://www.scopus.com/inward/record.url?scp=84863463392&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863463392&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84863463392

VL - 23

SP - 79

EP - 88

JO - Journal of Economic Theory and Econometrics

JF - Journal of Economic Theory and Econometrics

SN - 1229-2893

IS - 2

ER -