Estimation in threshold autoregressive models with correlated innovations
Preprint
Abstract
Large sample statistical analysis of threshold autoregressive (TAR) models is usually based on the assumption that the underlying driving noise is uncorrelated. In this paper, we consider a model, driven by Gaussian noise with geometric correlation tail and derive a complete characterization of the asymptotic distribution for the Bayes estimator of the threshold parameter.Keywords
All Related Versions
This publication has 11 references indexed in Scilit:
- On non-stationary threshold autoregressive modelsBernoulli, 2011
- Threshold models in time series analysis — 30 years onStatistics and Its Interface, 2011
- STRONG CONSISTENCY OF THE LEAST SQUARES ESTIMATOR FOR A NON-ERGODIC THRESHOLD AUTOREGRESSIVE MODEL2009
- Ergodicity and invertibility of threshold moving-average modelsBernoulli, 2007
- Testing for a linear MA model against threshold MA modelsThe Annals of Statistics, 2005
- Threshold Autoregression with a Unit RootEconometrica, 2001
- A new weak dependence condition and applications to moment inequalitiesStochastic Processes and their Applications, 1999
- Markov Chains and Stochastic StabilityJournal of the American Statistical Association, 1997
- Consistency and Limiting Distribution of the Least Squares Estimator of a Threshold Autoregressive ModelThe Annals of Statistics, 1993
- A Poisson-Type Limit Theorem for Mixing Sequences of Dependent 'Rare' EventsThe Annals of Probability, 1973