School of Banking and Finance , University of New South Wales , Australia

Graduate School of Economics, Kobe University , Japan

This article examines the pattern of information flow between the percentage price change and the trading volume in gold futures contracts using daily data over a ten-year period. We employ the robust two-step procedure proposed by Cheung and Ng (1996) to detect the causality in variance. We find evidence of strong contemporaneous causality that is indicative of the mixture of distribution hypothesis of information flow. We also detect, although not as strong, lagged causality running from percentage price change to trading volume. This indicates mild support for sequential information flow as well directed from price change to trading volume. This is contrary to the documented behavior in agricultural futures and crude oil futures, where bi-directional causality has been reported. We hypothesize that this is probably due to the special nature of gold as a commodity and the fact that the gold market takes on added importance when the equity market underperforms.

GARCH model

Back