Some New Results for Threshold AR(1) Models
John Knight and
Satchell Stephen
Additional contact information
Satchell Stephen: Trinity College, University of Cambridge and University of Sydney
Journal of Time Series Econometrics, 2011, vol. 3, issue 2, 42
Abstract:
The purpose of this paper is to derive some new results for threshold models. We consider AR(1) threshold models, with either self-exciting or exogenous triggers. In the latter case, we derive necessary and sufficient conditions for the existence of a stationary distribution, which are wider than the sufficient conditions that are the consequence of theorems provided in the literature by previous authors. We note that there appear to be no results in the literature for closed-form expressions for steady-state distributions for threshold models under econometrically relevant conditions. We provide such a result for the case of a threshold AR(1) with an exogenous trigger variable and normal innovations. It turns out to be a known distribution, namely, a compound geometric sum of normals.
Keywords: nonlinear time series; TAR models; autoregressive models; Markov chains (search for similar items in EconPapers)
Date: 2011
References: Add references at CitEc
Citations: View citations in EconPapers (7)
Downloads: (external link)
https://doi.org/10.2202/1941-1928.1085 (text/html)
For access to full text, subscription to the journal or payment for the individual article is required.
Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
HTML/Text
Persistent link: https://EconPapers.repec.org/RePEc:bpj:jtsmet:v:3:y:2011:i:2:n:1
Ordering information: This journal article can be ordered from
https://www.degruyter.com/journal/key/jtse/html
DOI: 10.2202/1941-1928.1085
Access Statistics for this article
Journal of Time Series Econometrics is currently edited by Javier Hidalgo
More articles in Journal of Time Series Econometrics from De Gruyter
Bibliographic data for series maintained by Peter Golla (peter.golla@degruyter.com).