A computationally efficient method for obtaining smoothed volatilities in long-memory stochastic volatility models

Abstract

We provide a computationally e±cient method, based on Harvey (1998) proposal, to estimate the underlying volatility of asset returns using the Long-Memory Stochastic Volatility (LMSV ) model. The performance of our procedure is illustrated with an application to three series of daily exhange rates returns. A comparison of long memory GARCH-type volatilities with our smoothed ones is also presented.