Do Aluminum Future Returns Have Long Memory?

This study examines the long memory behavior in the aluminum future returns using the rescaled range technique. The rescaled range technique is robust to detect the long-term dependence and non-period cycle in aluminum data. Our empirical results suggest that the long memory behavior in aluminum future returns is very stable. It means that the aluminum future returns follow a biased random walk. That is, they tend to trend in one direction until some exogenous event occurs to change their bias. The result of long memory and non-period cycle analysis provides new approaches for assessing the behavior of aluminum future prices.