A DOUBLY MARKOV SWITCHING AR MODEL: SOME PROBABILISTIC PROPERTIES AND STRONG CONSISTENCY

In this work, we consider doubly Markov switching AR models, where analytic tractability and flexibility are quite simply a competitive advantage, which becomes an attractive tool for modeling economic and financial time series. In these models, the parameters are allowed to depend on an unobservable time-homogeneous Markov chain with finite state space. So, we discuss some basic probabilistic properties of [Formula omitted] model such as conditions ensuring the existence of strict, second-order stationarity solution, causal and ergodic solution and its moments properties. The quasi-maximum likelihood estimator of the parameters in the model is shown to be strongly consistent.
Author(s): Ahmed Ghezal [sup.1] Author Affiliations: (1) Department of Mathematics and Computer Sciences, University Center of Mila, , Mila, Algeria Introduction An autoregressive model with Markov switching has been introduced [...]