estimation of value-at-risk with time varying skewness and kurtosis

this paper studies the effect of considering time varying skewness and kurtosis on the estimation of value at risk (var) for both long and short positions using the hyaparch model and daily data for tehran stock exchange price index (tepix). our results show that applying conditional distributions with time varying or constant skewness and degrees of freedom is able to capture the asymmetry appropriately compared to the normal distribution. however, the var estimations of these models are conservative, and they are appropriate for risk averse investors.