Continuous time regime-switching model applied to foreign exchange rate

Abstract

The continuous time modified Cox-Ingersoll-Ross (1985) stochastic model is employed, combining with the Hamilton (1989) type Markov regime-switching framework, to study daily foreign exchange rates where all parameter values depend on the value of a continuous time Markov chain. The generalized Expectation-Maximization algorithm is applied to a more general class of regime switching models and used to study some exchange rate data. We compare the obtained results with non regime switching models and notice that the regime switching outcomes match much better the reality than the others without Markov switching; and two regimes in most of the cases are better than more regimes.