EGARCH-RR: realized ranges explaining EGARCH volatilities

Abstract

The purpose of this paper is to investigate whether the inclusion of a realized measure of volatility as external regressor on the GARCH and EGARCH variance equation would result in more accurate ts. The estimation of the model is performed by maximum likelihood with fteen daily volatility series incorporated in the variance equation one at a time. The results show that the realized volatility measures add information to the EGARCH process; particularly, the realized range estimators that seem to outperform the realized volatility one. The scaling approach appears to be superior to the others that include squared overnight returns. GARCH is the most-adopted volatility model, which in itself justi es any improvement attempting. Besides, to the best of our knowledge, this is the rst work to include range estimators to the EGARCH variance equation.