Bayesian estimation of generalized long-memory stochastic volatility

Abstract

We propose a Bayesian approach to estimating the parameters of a Generalized Long-Memory Stochastic Volatility (GLMSV) model, a versatile framework designed to address both persistent (long-memory) and seasonal (cyclic) behaviors across various frequencies. This provides an alternative method incorporating prior information about the model parameters, and allows for relatively computationally efficient sampling from the posterior distribution by a reparametrization of the model parameters. The practical applicability of this methodology is demonstrated through the analysis of intraday volatility in Microsoft stock prices.