论文信息 - HETEROSKEDASTICITY-AUTOCORRELATION ROBUST STANDARD ERRORS USING THE BARTLETT KERNEL WITHOUT TRUNCATION

HETEROSKEDASTICITY-AUTOCORRELATION ROBUST STANDARD ERRORS USING THE BARTLETT KERNEL WITHOUT TRUNCATION

In this paper we analyze heteroskedasticity-autocorrelation (HAC) robust tests constructed using the Bartlett kernel without truncation. We show that while such an HAC estimator is not consistent, asymptotically valid testing is still possible. We show that tests using the Bartlett kernel without truncation are exactly equivalent to recent HAC robust tests proposed by Kiefer, Vogelsang and Bunzel (2000, Econometrica, 68, pp 695-714).

Nicholas M. Kiefer | Timothy J. Vogelsang | N. Kiefer | T. Vogelsang

[1] W. Newey,et al. A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelationconsistent Covariance Matrix , 1986 .

[2] D. Andrews. Heteroskedasticity and Autocorrelation Consistent Covariance Matrix Estimation , 1991 .

[3] K. Abadir,et al. Two Mixed Normal Densities from Cointegration Analysis , 1997 .

[4] Nicholas M. Kiefer,et al. Simple Robust Testing of Regression Hypotheses , 2000 .

[5] Paolo Paruolo,et al. Simple Robust Testing of Regression Hypotheses: A Comment , 2001 .

[6] Nicholas M. Kiefer,et al. HETEROSKEDASTICITY-AUTOCORRELATION ROBUST TESTING USING BANDWIDTH EQUAL TO SAMPLE SIZE , 2002, Econometric Theory.

[7] T. Vogelsang. TESTING IN GMM MODELS WITHOUT TRUNCATION , 2003 .