摘要
构建了FIEGARCH-EVT-ES风险测度模型,并将其应用于期货市场风险度量。FIEGARCH能捕捉资产收益的条件异方差、波动的长记忆性以及对正负信息反应的非对称性;极值理论(EVT)能处理资产收益的厚尾性;期望损失ES是一致性风险测度,可有效弥补VaR无法考察超过分位数信息的缺陷。以上海期货交易所铜期货为例验证了模型的有效性,并与GARCH-N-VaR模型进行比较分析,结果表明,GARCH-N-VaR低估了风险,FIEGARCH-EVT-ES预测效果较好,同时也不会使预测结果过于保守,从而过高估计风险。
FIEGARCH-EVT-ES model, a risk measure model, is built and applied in futures market. FIE- GARCH can capture conditional heteroscedasticity of asset return series, and long memory of volatility; Extreme Value Theory （EVT） is able to deal with fat tail of asset return series; ES is a coherent risk measure and can effectively remedy the defect of VaR that does not reflect the information more than quantile. Taking SHFE Copper futures as an example, effectiveness of model is tested and comparative analysis between GARCH-N-VaR and FIEGARCH-EVT-ES is conducted. The results show that GARCH-N-VaR underestimates risk; FIEGARCH-EVT-ES has a better forecasting result, and simultaneously it is not conservative so that overestimate risks.
出处
《南京航空航天大学学报（社会科学版）》
2011年第2期28-33,共6页
Journal of Nanjing University of Aeronautics & Astronautics（Social Sciences）
基金
国家自然科学基金项目(71071034)
