Asymptotic and bootstrap tests for linearity in a TAR-GARCH(1,1) model with a unit root

A-Tier
Journal: Journal of Econometrics
Year: 2008
Volume: 146
Issue: 1
Pages: 146-161

Score contribution per author:

4.022 = (α=2.01 / 1 authors) × 2.0x A-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

This paper derives the limiting distribution of the Lagrange Multiplier (LM) test for threshold nonlinearity in a TAR model with GARCH errors when one of the regimes contains a unit root. It is shown that the asymptotic distribution is nonstandard and depends on nuisance parameters that capture the degree of conditional heteroskedasticity and non-Gaussian nature of the process. We propose a bootstrap procedure for approximating the exact finite-sample distribution of the test for linearity and establish its asymptotic validity.

Technical Details

RePEc Handle
repec:eee:econom:v:146:y:2008:i:1:p:146-161
Journal Field
Econometrics
Author Count
1
Added to Database
2026-01-25