本文旨在探究貨櫃輪運價指數波動率的長期記憶現象。運用GPH、GSP、R/S檢定及FIGARCH、HYGARCH、FIAPARCH長期記憶GARCH模型來檢視。研究結果顯示，採用t分配與偏態t分配的長期記憶GARCH模型可能對於貨櫃輪運價指數波動率較能精確估計，並且提升長期預測與定價的精確性。因此，對於貨櫃輪運價指數波動率的風險估計，應將其長期記憶現象納入考量，同時所採用的GARCH模型應能一併考量波動的叢聚現象、不對稱性、厚尾 及長期記憶等因素。這些結果可以應用在實務界從事貨櫃輪運費市場之風險管理。

This study aims to investigate the features of the container freight indices when there is a long memory effect. We employed GPH test, GSP test, the Rescaled Range Tests of Mandelbrot (1972) and Lo (1991), FIGARCH, HYGARCH and FIAPARCH models for the long memory test and estimation. Our results suggest that precise estimates of container freight indices may be acquired from a long memory in volatility models with Student-t and skewed Student-t distribution. Such models might improve the long-term volatility forecast and more precise pricing of container freight contracts. We could extend these findings to the risk management in the container freight markets. Moreover, for appropriate risk evaluation of container freight indices, the degree of persistence should be examined and appropriate modelling that includes volatility clustering, asymmetry, leptokurtosis and long range dependence should be take into consideration. We could extend this implication to the connection of the container freight market management.