Abstract:
Standard long memory models are in abundance in the literature
today. Selecting the best such a model in terms of capturing key requisite
features and trends in data becomes a challenge. This paper
addresses the issue through a sequence of Monte Carlo experiments
on simulated data and introduces an interval estimate on the asymptotic
variance for the long-range dependence parameter of the entire
family of standard long memory time series considered within the
scope of the study.
