Classical and Bayesian aspects of robust unit root inference

Henk Hoek, André Lucas, Herman K. van Dijk

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

This paper has two themes. First, we classify some effects which outliers in the data have on unit root inference. We show that, both in a classical and a Bayesian framework, the presence of additive outliers moves 'standard' inference towards stationarity. Second, we base inference on an independent Student-t instead of a Gaussian likelihood. This yields results that are less sensitive to the presence of outliers. Application to several time series with outliers reveals a negative correlation between the unit root and degrees of freedom parameter of the Student-t distribution. Therefore, imposing normality may incorrectly provide evidence against the unit root.

Original languageEnglish
Pages (from-to)27-59
Number of pages33
JournalJournal of Econometrics
Volume69
Issue number1
DOIs
Publication statusPublished - 1995

Fingerprint

Unit Root
Outlier
Students
Additive Outliers
Time series
t-distribution
Stationarity
Normality
Likelihood
Degree of freedom
Classify
Unit root
Outliers
Inference

Keywords

  • Bayesian analysis
  • Outliers
  • Robustness
  • Student-t distribution
  • Unit root inference

Cite this

Hoek, Henk ; Lucas, André ; van Dijk, Herman K. / Classical and Bayesian aspects of robust unit root inference. In: Journal of Econometrics. 1995 ; Vol. 69, No. 1. pp. 27-59.
@article{fcb01c0018924620bd645792df4bffaa,
title = "Classical and Bayesian aspects of robust unit root inference",
abstract = "This paper has two themes. First, we classify some effects which outliers in the data have on unit root inference. We show that, both in a classical and a Bayesian framework, the presence of additive outliers moves 'standard' inference towards stationarity. Second, we base inference on an independent Student-t instead of a Gaussian likelihood. This yields results that are less sensitive to the presence of outliers. Application to several time series with outliers reveals a negative correlation between the unit root and degrees of freedom parameter of the Student-t distribution. Therefore, imposing normality may incorrectly provide evidence against the unit root.",
keywords = "Bayesian analysis, Outliers, Robustness, Student-t distribution, Unit root inference",
author = "Henk Hoek and Andr{\'e} Lucas and {van Dijk}, {Herman K.}",
year = "1995",
doi = "10.1016/0304-4076(94)01661-I",
language = "English",
volume = "69",
pages = "27--59",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier BV",
number = "1",

}

Classical and Bayesian aspects of robust unit root inference. / Hoek, Henk; Lucas, André; van Dijk, Herman K.

In: Journal of Econometrics, Vol. 69, No. 1, 1995, p. 27-59.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - Classical and Bayesian aspects of robust unit root inference

AU - Hoek, Henk

AU - Lucas, André

AU - van Dijk, Herman K.

PY - 1995

Y1 - 1995

N2 - This paper has two themes. First, we classify some effects which outliers in the data have on unit root inference. We show that, both in a classical and a Bayesian framework, the presence of additive outliers moves 'standard' inference towards stationarity. Second, we base inference on an independent Student-t instead of a Gaussian likelihood. This yields results that are less sensitive to the presence of outliers. Application to several time series with outliers reveals a negative correlation between the unit root and degrees of freedom parameter of the Student-t distribution. Therefore, imposing normality may incorrectly provide evidence against the unit root.

AB - This paper has two themes. First, we classify some effects which outliers in the data have on unit root inference. We show that, both in a classical and a Bayesian framework, the presence of additive outliers moves 'standard' inference towards stationarity. Second, we base inference on an independent Student-t instead of a Gaussian likelihood. This yields results that are less sensitive to the presence of outliers. Application to several time series with outliers reveals a negative correlation between the unit root and degrees of freedom parameter of the Student-t distribution. Therefore, imposing normality may incorrectly provide evidence against the unit root.

KW - Bayesian analysis

KW - Outliers

KW - Robustness

KW - Student-t distribution

KW - Unit root inference

UR - http://www.scopus.com/inward/record.url?scp=0039175067&partnerID=8YFLogxK

U2 - 10.1016/0304-4076(94)01661-I

DO - 10.1016/0304-4076(94)01661-I

M3 - Article

VL - 69

SP - 27

EP - 59

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 1

ER -