A bias-reduced log-periodogram regression estimator for the long-memory parameter

Donald W.K. Andrews, Patrik Guggenberger

Research output: Contribution to journalArticle

81 Citations (Scopus)

Abstract

In this paper, we propose a simple bias-reduced log-periodogram regression estimator, d̂r, of the long-memory parameter, d, that eliminates the first- and higher-order biases of the Geweke and Porter-Hudak (1983) (GPH) estimator. The bias-reduced estimator is the same as the GPH estimator except that one includes frequencies to the power 2k for k = 1, r, for some positive integer r, as additional regressors in the pseudo-regression model that yields the GPH estimator. The reduction in bias is obtained using assumptions on the spectrum only in a neighborhood of the zero frequency. Following the work of Robinson (1995b) and Hurvich, Deo, and Brodsky (1998), we establish the asymptotic bias, variance, and mean-squared error (MSE) of d̂r, determine the asymptotic MSE optimal choice of the number of frequencies, m, to include in the regression, and establish the asymptotic normality of d̂r. These results show that the bias of d̂r goes to zero at a faster rate than that of the GPH estimator when the normalized spectrum at zero is sufficiently smooth, but that its variance only is increased by a multiplicative constant. We show that the bias-reduced estimator d̂r attains the optimal rate of convergence for a class of spectral densities that includes those that are smooth of order s ≥ 1 at zero when r ≥ (s - 2)/2 and m is chosen appropriately. For s > 2, the GPH estimator does not attain this rate. The proof uses results of Giraitis, Robinson, and Samarov (1997). We specify a data-dependent plug-in method for selecting the number of frequencies m to minimize asymptotic MSE for a given value of r. Some Monte Carlo simulation results for stationary Gaussian ARFIMA(1, d, 1) and (2, d, 0) model show that the bias-reduced estimators perform well relative to the standard log-periodogram regression estimator.

Original languageEnglish (US)
Pages (from-to)675-712
Number of pages38
JournalEconometrica
Volume71
Issue number2
DOIs
StatePublished - Jan 1 2003

Fingerprint

Log-periodogram regression
Long memory
Estimator
Mean squared error
Integer
Spectral density
Asymptotic normality
Regression model
Asymptotic bias
Monte Carlo simulation
Rate of convergence

All Science Journal Classification (ASJC) codes

  • Economics and Econometrics

Cite this

@article{e430ed95354b4a4a81204ea39f2adc52,
title = "A bias-reduced log-periodogram regression estimator for the long-memory parameter",
abstract = "In this paper, we propose a simple bias-reduced log-periodogram regression estimator, d̂r, of the long-memory parameter, d, that eliminates the first- and higher-order biases of the Geweke and Porter-Hudak (1983) (GPH) estimator. The bias-reduced estimator is the same as the GPH estimator except that one includes frequencies to the power 2k for k = 1, r, for some positive integer r, as additional regressors in the pseudo-regression model that yields the GPH estimator. The reduction in bias is obtained using assumptions on the spectrum only in a neighborhood of the zero frequency. Following the work of Robinson (1995b) and Hurvich, Deo, and Brodsky (1998), we establish the asymptotic bias, variance, and mean-squared error (MSE) of d̂r, determine the asymptotic MSE optimal choice of the number of frequencies, m, to include in the regression, and establish the asymptotic normality of d̂r. These results show that the bias of d̂r goes to zero at a faster rate than that of the GPH estimator when the normalized spectrum at zero is sufficiently smooth, but that its variance only is increased by a multiplicative constant. We show that the bias-reduced estimator d̂r attains the optimal rate of convergence for a class of spectral densities that includes those that are smooth of order s ≥ 1 at zero when r ≥ (s - 2)/2 and m is chosen appropriately. For s > 2, the GPH estimator does not attain this rate. The proof uses results of Giraitis, Robinson, and Samarov (1997). We specify a data-dependent plug-in method for selecting the number of frequencies m to minimize asymptotic MSE for a given value of r. Some Monte Carlo simulation results for stationary Gaussian ARFIMA(1, d, 1) and (2, d, 0) model show that the bias-reduced estimators perform well relative to the standard log-periodogram regression estimator.",
author = "Andrews, {Donald W.K.} and Patrik Guggenberger",
year = "2003",
month = "1",
day = "1",
doi = "10.1111/1468-0262.00420",
language = "English (US)",
volume = "71",
pages = "675--712",
journal = "Econometrica",
issn = "0012-9682",
publisher = "Wiley-Blackwell",
number = "2",

}

A bias-reduced log-periodogram regression estimator for the long-memory parameter. / Andrews, Donald W.K.; Guggenberger, Patrik.

In: Econometrica, Vol. 71, No. 2, 01.01.2003, p. 675-712.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A bias-reduced log-periodogram regression estimator for the long-memory parameter

AU - Andrews, Donald W.K.

AU - Guggenberger, Patrik

PY - 2003/1/1

Y1 - 2003/1/1

N2 - In this paper, we propose a simple bias-reduced log-periodogram regression estimator, d̂r, of the long-memory parameter, d, that eliminates the first- and higher-order biases of the Geweke and Porter-Hudak (1983) (GPH) estimator. The bias-reduced estimator is the same as the GPH estimator except that one includes frequencies to the power 2k for k = 1, r, for some positive integer r, as additional regressors in the pseudo-regression model that yields the GPH estimator. The reduction in bias is obtained using assumptions on the spectrum only in a neighborhood of the zero frequency. Following the work of Robinson (1995b) and Hurvich, Deo, and Brodsky (1998), we establish the asymptotic bias, variance, and mean-squared error (MSE) of d̂r, determine the asymptotic MSE optimal choice of the number of frequencies, m, to include in the regression, and establish the asymptotic normality of d̂r. These results show that the bias of d̂r goes to zero at a faster rate than that of the GPH estimator when the normalized spectrum at zero is sufficiently smooth, but that its variance only is increased by a multiplicative constant. We show that the bias-reduced estimator d̂r attains the optimal rate of convergence for a class of spectral densities that includes those that are smooth of order s ≥ 1 at zero when r ≥ (s - 2)/2 and m is chosen appropriately. For s > 2, the GPH estimator does not attain this rate. The proof uses results of Giraitis, Robinson, and Samarov (1997). We specify a data-dependent plug-in method for selecting the number of frequencies m to minimize asymptotic MSE for a given value of r. Some Monte Carlo simulation results for stationary Gaussian ARFIMA(1, d, 1) and (2, d, 0) model show that the bias-reduced estimators perform well relative to the standard log-periodogram regression estimator.

AB - In this paper, we propose a simple bias-reduced log-periodogram regression estimator, d̂r, of the long-memory parameter, d, that eliminates the first- and higher-order biases of the Geweke and Porter-Hudak (1983) (GPH) estimator. The bias-reduced estimator is the same as the GPH estimator except that one includes frequencies to the power 2k for k = 1, r, for some positive integer r, as additional regressors in the pseudo-regression model that yields the GPH estimator. The reduction in bias is obtained using assumptions on the spectrum only in a neighborhood of the zero frequency. Following the work of Robinson (1995b) and Hurvich, Deo, and Brodsky (1998), we establish the asymptotic bias, variance, and mean-squared error (MSE) of d̂r, determine the asymptotic MSE optimal choice of the number of frequencies, m, to include in the regression, and establish the asymptotic normality of d̂r. These results show that the bias of d̂r goes to zero at a faster rate than that of the GPH estimator when the normalized spectrum at zero is sufficiently smooth, but that its variance only is increased by a multiplicative constant. We show that the bias-reduced estimator d̂r attains the optimal rate of convergence for a class of spectral densities that includes those that are smooth of order s ≥ 1 at zero when r ≥ (s - 2)/2 and m is chosen appropriately. For s > 2, the GPH estimator does not attain this rate. The proof uses results of Giraitis, Robinson, and Samarov (1997). We specify a data-dependent plug-in method for selecting the number of frequencies m to minimize asymptotic MSE for a given value of r. Some Monte Carlo simulation results for stationary Gaussian ARFIMA(1, d, 1) and (2, d, 0) model show that the bias-reduced estimators perform well relative to the standard log-periodogram regression estimator.

UR - http://www.scopus.com/inward/record.url?scp=0037244640&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0037244640&partnerID=8YFLogxK

U2 - 10.1111/1468-0262.00420

DO - 10.1111/1468-0262.00420

M3 - Article

AN - SCOPUS:0037244640

VL - 71

SP - 675

EP - 712

JO - Econometrica

JF - Econometrica

SN - 0012-9682

IS - 2

ER -