|
Abstract:
|
Since the publication of Hamilton's (1989) seminal work on Markov switching model, a large number of applications have been found in economics and finance. Since classical Markov switching models characterize regimes in finite states, it is restrictive in empirical studies. In this paper, we consider a Markov switching model with stochastic regimes, in which the regimes and model parameters are represented both categorically and continuously. Assuming conjugate priors, we develop closed-form recursive Bayes estimates of the regression parameters, an approximation scheme that has much lower computational complexity and yet is comparable to the Bayes estimates in statistical efficiency, and an expectation-maximization procedure to estimate the unknown hyperparameters. We conduct intensive simulation studies to evaluate the performance of our model. We also used our model to analyze several US economic data, such as the industrial production, manufacturing and trade inventory and unemployment rate, to show our model is more suitable than classical regime switching models.
|
