Infinite-state Markov-switching for dynamic volatility and correlation models

Dynamic volatility and correlation models with fixed parameters are restrictive for time series subject to breaks. GARCH and DCC models with changing parameters are specified using the sticky infinite hidden Markov-chain framework. Estimation by Bayesian inference determines the adequate number of regimes as well as the optimal specification (Markov-switching or change-point). The new estimation algorithm is studied in terms of mixing properties and computational time. Applications highlight the flexibility of the model.