Volatility Forecasting Using APARCH with Skewed Conditional Distributions

It is well known that distributions of financial return are fat-tailed and many models have been developed to capture fat-tail. It is not so well known that distributions are skewed. We construct skewed distributions based on symmetric distributions using Fernandez-Steel method and research the prediction of volatility together with APARCH model. Empirical results show that there are significant influences of the skewedness on simulation and prediction of volatility models. More steps ahead of prediction, more significance of influence.