In this thesis we develop a new approach within the framework of asset pricing models that incorporates two key features of the latent volatility: co-movement among conditionally heteroscedastic financial returns and switching between different unobservable regimes. By combining conditionally heteroscedastic factor models with hidden Markov chain models we derive a dynamical local model for segmentation and prediction of multivariate financial time series. We concentrate, more precisely on situations where the factor variances are modelled by univariate GQARCH processes. The EM algorithm that we have developed for the maximum likelihood estimation is based on a quasi-optimal Kalman filter approach combined with a Viterbi approximation which yields inferences about the unobservable path of the common factors, their variances and the latent variable of the state process. Extensive simulation experiments and the analysis of a financial data set show promising results.
