We investigate the volume impact on volatility for 14 Chinese ADRs and their underlying H-shares. We decompose volume into expected and unanticipated components and include those as determinants of conditional volatility in a bivariate GARCH model for each ADR and its underlying H-share. Expected volume denotes liquidity, while unanticipated volume implies information content in volume. The GARCH model fits the data well. In addition to the conventional GARCH parameters, for ADRs and their underlying H-shares, expected and unanticipated volumes significantly but asymmetrically affect both the variance and the covariance functions. Further, volume components asymmetrically impact volatility of ADRs and H-shares in high- versus low-liquidity and high- versus low-liquidity-risk buckets denoted by volume and standard deviation of volume, respectively.
- Volume decomposition
ASJC Scopus subject areas
- Business and International Management
- Strategy and Management
- Information Systems and Management