Abstract:

The main aim of this paper is to study the dynamic relationships between return volatility-volume of the Malaysian stock exchange using a trivariate Vector AutoRegression (VAR) model. The latent volatility is obtained from the estimated component generalized autoregressive conditional variance (CGARCH) model. For trading volume, the linear and non-linear trends are adjusted using multiple regression analysis. The Granger causality tests show only unidirectional causality from volume to volatility. Hence, this result is not enough to support the presence of sequential arrival of information hypothesis. It is also worth noting that the return series has a much stronger role than volume in explaining volatility.