STANDARDIZED TIME-SERIES L-P-NORM VARIANCE ESTIMATORS FOR SIMULATIONS

This paper studies a class of estimators for the variance parameter of a stationary stochastic process. The estimators are based on L-p norm s of standardized time series, and they generalize previously studied estimators due to Schruben. We show that the new estimators have some desirable properties: they are asymptotically unbiased and have low as ymptotic variance. We also illustrate empirically the performance of t he L-p-norm estimators on various stochastic processes.