Spuriousness of information criteria when selecting the number of breaks in stationary AR(p) process

This note proves analytically and shows by a Monte Carlo analysis the spuriousness that arises by some model selection criteria when selecting the number of breaks in stationary AR(p) process without changes for a regression with mean-shifts. This brings a theoretical support to the Perron's (1997) simulation results which indicate that this phenomenon occurs for an AR(1) process.

MoreLess

Year of publication:	2007-08-31
Authors:	JOUINI, Jamel ; BOUTAHAR, Mohamed
Published in:	Economics Bulletin. - AccessEcon, ISSN 1545-2921. - Vol. 3.2007, 38, p. 1-11
Publisher:	AccessEcon

More details

Extent:	application/pdf
Type of publication:	Article
Classification:	C2 - Econometric Methods: Single Equation Models
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10010629410