Mixed Causal–Noncausal Autoregressions: Bimodality Issues in Estimation and Unit Root Testing1

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Standard

Mixed Causal–Noncausal Autoregressions : Bimodality Issues in Estimation and Unit Root Testing1. / Bec, Frédérique; Nielsen, Heino Bohn; Saïdi, Sarra.

I: Oxford Bulletin of Economics and Statistics, Bind 82, Nr. 6, 12.2021, s. 1413-1428.

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Harvard

Bec, F, Nielsen, HB & Saïdi, S 2021, 'Mixed Causal–Noncausal Autoregressions: Bimodality Issues in Estimation and Unit Root Testing1', Oxford Bulletin of Economics and Statistics, bind 82, nr. 6, s. 1413-1428. https://doi.org/10.1111/obes.12372

APA

Bec, F., Nielsen, H. B., & Saïdi, S. (2021). Mixed Causal–Noncausal Autoregressions: Bimodality Issues in Estimation and Unit Root Testing1. Oxford Bulletin of Economics and Statistics, 82(6), 1413-1428. https://doi.org/10.1111/obes.12372

Vancouver

Bec F, Nielsen HB, Saïdi S. Mixed Causal–Noncausal Autoregressions: Bimodality Issues in Estimation and Unit Root Testing1. Oxford Bulletin of Economics and Statistics. 2021 dec.;82(6):1413-1428. https://doi.org/10.1111/obes.12372

Author

Bec, Frédérique ; Nielsen, Heino Bohn ; Saïdi, Sarra. / Mixed Causal–Noncausal Autoregressions : Bimodality Issues in Estimation and Unit Root Testing1. I: Oxford Bulletin of Economics and Statistics. 2021 ; Bind 82, Nr. 6. s. 1413-1428.

Bibtex

@article{5c462604bd344f1abd697f2abac5ccc2,
title = "Mixed Causal–Noncausal Autoregressions: Bimodality Issues in Estimation and Unit Root Testing1",
abstract = "This paper stresses the bimodality of the likelihood function of the Mixed causal–noncausal AutoRegressions (MAR), and it is shown that the bimodality issue becomes more salient as the causal root approaches unity from below. The consequences are important as the roots of the local maxima are typically interchanged, attributing the noncausal component to the causal one and vice-versa. This severely changes the interpretation of the results, and the properties of unit root tests of the backward root are adversely affected. To circumvent the bimodality issue, this paper proposes an estimation strategy which (i) increases noticeably the probability of attaining the global MLE; and (ii) selects carefully the maximum used for the unit root test against a MAR stationary alternative.",
author = "Fr{\'e}d{\'e}rique Bec and Nielsen, {Heino Bohn} and Sarra Sa{\"i}di",
year = "2021",
month = dec,
doi = "10.1111/obes.12372",
language = "English",
volume = "82",
pages = "1413--1428",
journal = "Oxford Bulletin of Economics and Statistics",
issn = "0305-9049",
publisher = "Wiley-Blackwell",
number = "6",

}

RIS

TY - JOUR

T1 - Mixed Causal–Noncausal Autoregressions

T2 - Bimodality Issues in Estimation and Unit Root Testing1

AU - Bec, Frédérique

AU - Nielsen, Heino Bohn

AU - Saïdi, Sarra

PY - 2021/12

Y1 - 2021/12

N2 - This paper stresses the bimodality of the likelihood function of the Mixed causal–noncausal AutoRegressions (MAR), and it is shown that the bimodality issue becomes more salient as the causal root approaches unity from below. The consequences are important as the roots of the local maxima are typically interchanged, attributing the noncausal component to the causal one and vice-versa. This severely changes the interpretation of the results, and the properties of unit root tests of the backward root are adversely affected. To circumvent the bimodality issue, this paper proposes an estimation strategy which (i) increases noticeably the probability of attaining the global MLE; and (ii) selects carefully the maximum used for the unit root test against a MAR stationary alternative.

AB - This paper stresses the bimodality of the likelihood function of the Mixed causal–noncausal AutoRegressions (MAR), and it is shown that the bimodality issue becomes more salient as the causal root approaches unity from below. The consequences are important as the roots of the local maxima are typically interchanged, attributing the noncausal component to the causal one and vice-versa. This severely changes the interpretation of the results, and the properties of unit root tests of the backward root are adversely affected. To circumvent the bimodality issue, this paper proposes an estimation strategy which (i) increases noticeably the probability of attaining the global MLE; and (ii) selects carefully the maximum used for the unit root test against a MAR stationary alternative.

U2 - 10.1111/obes.12372

DO - 10.1111/obes.12372

M3 - Journal article

AN - SCOPUS:85087726684

VL - 82

SP - 1413

EP - 1428

JO - Oxford Bulletin of Economics and Statistics

JF - Oxford Bulletin of Economics and Statistics

SN - 0305-9049

IS - 6

ER -

ID: 255049744