This study provides a comprehensive analysis of the possible influences of jump dynamics, heavy-tails, and skewness with regard to VaR estimates through the assessment of both accuracy and efficiency. To this end, the ARJI model, and its degenerative GARCH model with normal, GED, and skewed normal (SN) distributions were adopted to capture the properties of time-varying volatility, time-varying jump intensity, heavy-tails and skewness, for a range of stock indices across international stock markets during the period of the U.S. subprime mortgage crisis. Empirical results show that, with regard to the evaluation of accuracy, the role of jump dynamics is more substantial than heavy-tails or skewness as it pertains to VaR accuracy at the 90% and 95% levels, while heavy-tails become more important at the 99% level for a long position. However, the influence of the abovementioned properties on VaR estimation does not appear substantial for a short position. In addition, the properties of jump dynamics and skewness appear to be beneficial for the improvement of efficiency.
