Graphical identification of TAR models

This paper proposes an automatic procedure to identify Threshold Autoregressive models and specify the threshold values. The proposed procedure is based on recursive estimation of arranged autoregression. The main advantage of the proposed procedure over its competitors is that the threshold values are automatically detected. The performance of the proposed procedure is evaluated using simulations and real data.

MoreLess

Year of publication:	2009-12
Authors:	Bermejo, Miguel Ángel ; Peña, Daniel ; Sánchez, Ismael
Institutions:	Departamento de Estadistica, Universidad Carlos III de Madrid
Subject:	Nonlinear time series \| Recursive estimation \| Arranged autoregression \| TAR models \| Nonlinearity test

More details

Extent:	application/pdf
Series:	Statistics and Econometrics Working Papers.
Type of publication:	Book / Working Paper
Language:	English
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10008543185