School of Mathematical Sciences

External Seminar: Jie Cheng (Keele)

Location
Pharmacy A05
Date(s)
Thursday 28th February 2019 (14:00-15:00)
Contact
Gilles Stupfler
Description

Statistics & Probability Seminar

A Latent Factor-Based Endogenous Regime-Switching Model for Non-Gaussian Processes

We propose a new approach for modelling endogenous regime-switching (RS) non-Gaussian processes. Similar to Chang et al. (2015), we assume that the change of regime is determined by whether an unobservable latent factor goes above or below an unknown threshold level. However, instead of assuming that the innovations of the state-dependent process and the latent process are bivariate normal, we allow the state-dependent process to be general, i.e. non-linear and non-Gaussian. This is achieved by introducing a copula-like conditional distribution-based transformation of the state-dependent transitional innovations. We show that our approach nests the Chang et al. (2015) approach as a special case and that our approach has wider and more general applications, such as RS processes with non-Gaussian error distributions and RS continuous-time processes.

School of Mathematical Sciences

The University of Nottingham
University Park
Nottingham, NG7 2RD

For all enquiries please visit:
www.nottingham.ac.uk/enquire