A note on integrated periodic GARCH processes

This note examines some probabilistic properties of periodic and integrated periodic generalized autoregressive conditionally heteroskedasticitic ((I)PGARCH) processes. In these models, the parameters are allowed to switch between different regimes and thus constitute an important subclass of switching GARCH process. We show that, similarly to the classical IGARCH processes, a stationary (in periodic sense) solution with infinite variance for the IPGARCH processes may exist.