Maximum likelihood estimation for score-driven models

We establish strong consistency and asymptotic normality of the maximum likelihood estimator for stochastic time-varying parameter models driven by the score of the predictive conditional likelihood function. For this purpose, we formulate primitive conditions for global identification, invertibilit...

Full description

Saved in:
Bibliographic Details
Published in:Journal of econometrics 2022-04, Vol.227 (2), p.325-346
Main Authors: Blasques, Francisco, van Brummelen, Janneke, Koopman, Siem Jan, Lucas, André
Format: Article
Language:English
Subjects:
Asymptotic methods
Asymptotic normality
Consistency
econometrics
Estimating techniques
Invertibility
Markov processes
Maximum likelihood method
Normality
Parameter estimation
probability
Specification
Stationarity
t-test
Time-varying parameters
Volatility
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We establish strong consistency and asymptotic normality of the maximum likelihood estimator for stochastic time-varying parameter models driven by the score of the predictive conditional likelihood function. For this purpose, we formulate primitive conditions for global identification, invertibility, strong consistency, and asymptotic normality both under correct specification and misspecification of the model. A detailed illustration is provided for a conditional volatility model with disturbances from the Student’s t distribution.
ISSN:0304-4076
1872-6895
DOI:10.1016/j.jeconom.2021.06.003