A1636
Title: Generalized fractional stochastic volatility models
Authors: Shelton Peiris - University of Sydney (Australia) [presenting]
Manabu Asai - Soka University (Japan)
Abstract: In recent years fractionally differenced processes have received a great deal of attention due its flexibility in financial applications with long memory. The aim is to consider a generalized class of fractionally differenced processes generated by Gegenbauer polynomials, known as Gegenbauer Autoregressive Moving Average (GARMA) models. We extend this class to incorporate both the long memory and stochastic volatility (SV) components. The existence and uniqueness of second order solutions will be established. Under certain regularity conditions, it is shown that the estimation of parameters, identification and forecasting through the underlying process are not affected. Furthermore, various new results associated with the moments, kurtosis and minimum mean square error forecasts will be reported. A simulation study has been added and a potential application will be discussed to justify the usefulness of this new class of models in financial econometrics.
