Abstract:
In this paper, weexplore the linear estimator (LE) for the parameters of ARCH models. The accuracy of LE in terms of estimation and volatility forecasting is compared with the well-known quasimaximum likelihood estimator (QMLE). Monte Carlo simulations are used to check the relative performance of both estimators. Our study indicates that the linear estimator provides accurate estimates for the parameters of ARCH models and also predicts volatility better than the QMLE in most of the cases. Application to real data sets is also presented.
Page(s):
113-118
DOI:
DOI not available
Published:
Journal: Science International, Volume: 22, Issue: 1, Year: 2010