An empirical analysis on long memory of Shanghai copper and fuel oil in the supply chain

• 1、Economics and Trade Schoool,Hunan University ,Hunan ,410079

Abstract：Empirical research increasingly suggests that long-term memory exists in futures return rate in the sequence. According to the historical information of financial derivative asset\' prices, the majority of investors can make roughly estimate of the trend.This article extracts the data of the copper and fuel oil\' closing prices of each trading day, with acquisition time from Dec 16, 2005 to Mar 27, 2017. For a sequence of long-term memory characteristics, simple analysis to be done to observe some basic features of their return series, including normal distribution test and stability test. In addition, this article attempts to make a concrete analysis through the R/S test and modified R/S method. Finally, based on the long- term memory model- ARFIMA model, the basically forecasts are made from1 to10 periods. The research shows that the prediction\' effectiveness of the established ARFIMA model is good. The results of this study show that the two futures\' return rate and volatilityseries exist significant long-term memory, but long-term memories in the copper slightly weaker in the fuel oil. Therefore, the study of the raw material price fluctuations , such as copper, oil and other household electrical appliances, it has played an importantly theoretical and practical significance in reducing the cost of home appliance supply chain and promoting the rational pricing and further development of products in home appliance supply chain.

Keywords： financial market；long-term memory Hurst R/S analysis Modified R/S analysis ARFIMA model