Estimation of Generalized Long-Memory Stochastic Volatility for High-Frequency Data

We consider the generalized long-memory stochastic volatility (GLMSV) model, a relatively general model of stochastic volatility that accounts for persistent (or longmemory) and seasonal (or cyclic) behavior at several frequencies. We employ the decorrelating properties of discrete wavelet packet transform (DWPT) to provide a wavelet-based approximate maximum likelihood estimator that allows for analysis of high-frequency data by simplifying the variance-covariance matrix into a diagonalized matrix, whose diagonal elements have the least distinct variances to compute using a computationally efficient quadrature. We apply the proposed method to the estimation of highfrequency simulated data.