Volume 17 | Issue 1 | Article 4

<< Previous Article | Next Article >>
Back To: Table of Contents

Value-at-Risk Aanlysis for KOSPI 200 Index Futures: Evidence from Long Memory Volatility Models with a Skewed Student-t Distribution
Sang Hoon Kang and Seong-Min Yoon (C32, C52, G11)

We computed daily Value-at-Risk (VaR) for KOSPI 200 Index futures returns using two long memory volatility models (FIGARCH and FIAPARCH) based on the normal, Student-t, and skewed Student-t innovation distributions. We considered asymmetric long memory in the conditional variance and asymmetric fat-tailed distributions. Both the in-sample and out-of-sample VaR analyses indicated that the FIGARCH and FIAPARCH models with the skewed Student-t distribution innovation provide more accurate volatility forecasting for KOSPI 200 Index futures than do the models with the normal and Student-t distribution innovations. These VaR analyses provide an optimal margin level of risk in the KOSPI 200 Index futures market.