Identication problems in ESTAR models anda new model

In ESTAR models it is usually dfficult to determine parameter estimates, as it can be observedin the literature. We show that the phenomena of getting strongly biased estimators is aconsequence of the so-called identication problem, the problem of properly distinguishing thetransition function in relation to extreme parameter combinations. This happens in particularfor either very small or very large values of the error term variance.[...]

MoreLess

Year of publication:	2010-03-29
Authors:	Donauer, Stefanie ; Heinen, Florian ; Sibbertsen, Philipp
Institutions:	Universität <Hannover> / Wirtschaftswissenschaftliche Fakultät
Subject:	Wechselkurs \| Linearität \| Nonlinearities \| Identifikation \| Exponential smoothing \| Autoregressives Modell

1 Introduction
2 Exponential Smooth Transition Autoregressive Models
2.1 The ESTAR-Model
2.2 The Identication Problem
3 The T-STAR Model
3.1 The Model and Estimators
3.2 Linearity testing
3.3 Unit Root Testing
3.4 Monte Carlo Simulations
4 Empirical Illustration
5 Conclusions
A Appendix: Proofs and Technical LemmasA Appendix: Proofs and Technical Lemmas

Persistent link: https://ebvufind01.dmz1.zbw.eu/10005870744