Estimation of seasonal fractionally integrated processes
Loading...
Date
2006
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ELSEVIER SCIENCE BV
Abstract
This paper discusses the estimation of fractionally integrated processes with seasonal components. In order to estimate the fractional parameters, we propose several estimators obtained from the regression of the log-periodogram on different bandwidths selected around and/or between the seasonal frequencies. For comparison purposes, the semi-paramenic method introduced in Geweke and Porter-Hudak (1983) and Porter-Hudak (1990) and the maximum-likelihood estimates (ML) are also considered. As indicated by the Monte Carlo simulations, the performance of the estimators proposed is good even for small sample sizes. (c) 2004 Elsevier B.V. All rights reserved.
Description
Keywords
fractional differencing, long memory, maximum-likelihood estimates, periodogram regression, seasonality, LONG-MEMORY PROCESSES, ARFIMA PROCESSES, MODELS