This paper examines the information content of trading volume in terms of forecasting the conditional\nvolatility and market risk of international stock markets. The performance of parametric\nValue at Risk (VaR) models including the traditional RiskMetrics model and a heavy-tailed\nEGARCH model with and without trading volume is investigated during crisis and post-crisis periods.\nOur empirical results provide compelling evidence that volatility forecasts based on volume-\naugmented models cannot be outperformed by their competitors. Furthermore, our findings\nindicate that including trading volume into the volatility specification greatly enhances the performance\nof the proposed VaR models, especially during the crisis period. However, the volume effect\nis fairly overshadowed by the sufficient accuracy of the heavy-tailed EGARCH model, during the\npost-crisis period.
Loading....