2002
DOI: 10.1017/s0266466602183083
|View full text |Cite
|
Sign up to set email alerts
|

Testing for Zero Autocorrelation in the Presence of Statistical Dependence

Abstract: The problem addressed in this paper is to test the null hypothesis that a time series process is uncorrelated up to lag K in the presence of statistical dependence. We propose an extension of the Box–Pierce Q-test that is asymptotically distributed as chi-square when the null is true for a very general class of dependent processes that includes non-martingale difference sequences. The test is based on a consistent estimator of the asymptotic covariance matrix of the sample autocorrelations under the nu… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

2
76
0

Year Published

2009
2009
2023
2023

Publication Types

Select...
5
2
1

Relationship

0
8

Authors

Journals

citations
Cited by 80 publications
(78 citation statements)
references
References 28 publications
2
76
0
Order By: Relevance
“…Assuming also that {ε θ 0 t } t∈Z follows a Gaussian GARCH process, then a (i,j) θ 0 = 0, i = j, which makes the estimation easier; see Lobato, Nankervis, and Savin (2002).…”
Section: Introductionmentioning
confidence: 99%
“…Assuming also that {ε θ 0 t } t∈Z follows a Gaussian GARCH process, then a (i,j) θ 0 = 0, i = j, which makes the estimation easier; see Lobato, Nankervis, and Savin (2002).…”
Section: Introductionmentioning
confidence: 99%
“…The …rst one is to modify the Q p statistic by introducing a consistent estimator of the asymptotic null covariance matrix of the sample autocorrelations, b T ; so that the modi…ed Q p statistic retains the 2 p asymptotic null distribution. Lobato, Nankervis and Savin (2002) name this statistic…”
Section: Tests Based On a …Nite-dimensional Conditioning Setmentioning
confidence: 99%
“…In other words, the standard portmanteau goodness-of-fit tests are not reliable when the model is ARMA and the DGP is nonlinear. Is is however possible to modified these tests to take into account conditional heteroscedasticity or any other dependence in the linear innovations (see [89] and [59]). …”
Section: Detection Of Autocorrelationsmentioning
confidence: 99%
“…In concrete applications the true significance limits are unknown, but can be estimated (see [105], [89] and [61]). Table 1 reports the results of the standard and modified portmanteau tests, for a simulation of length n = 5000 of the GARCH(1,1) model (13).…”
Section: Examplementioning
confidence: 99%