Local Likelihood for non paramentric ARCH(1) models
Journal
Journal of Time Series Analysis
ISSN
0143-9782
ISSN-Digital
1467-9892
Type
journal article
Date Issued
2005-03-01
Author(s)
Abstract
We propose a non-parametric local likelihood estimator for the log-transformed autoregressive conditional heteroscedastic (ARCH) (1) model. Our non-parametric estimator is constructed within the likelihood framework for non-Gaussian observations: it is different from standard kernel regression smoothing, where the innovations are assumed to be normally distributed. We derive consistency and asymptotic normality for our estimators and show, by a simulation experiment and some real-data examples, that the local likelihood estimator has better predictive potential than classical local regression. A possible extension of the estimation procedure to more general multiplicative ARCH(p) models with p > 1 predictor variables is also described.
Language
English
HSG Classification
not classified
Refereed
Yes
Publisher
Blackwell
Publisher place
Oxford
Volume
26
Number
2
Start page
251
End page
278
Pages
28
Subject(s)
Eprints ID
32656