风险管理视角下的高频波动率测度比较

发布时间：2017-12-31 22:01

本文关键词：风险管理视角下的高频波动率测度比较　出处：《中国管理科学》2014年S1期 　论文类型：期刊论文

【摘要】：鉴于波动率研究的一个重要应用是金融风险管理,提出在风险管理视角下比较各种高频波动率测度。具体考虑已实现波动、双幂次变差、中位数已实现波动、双尺度已实现波动、已实现极差和最优线性组合已实现波动,借助偏学生t分布假设下的Realized GARCH模型,预测未来一日,五日、十日和二十日的在险价值(VaR),并从统计精度、监管精度、资本运作效率和巴塞尔Ⅱ规定的市场风险资本需求四个角度,对六种不同高频波动率测度的VaR预测效果进行比较。使用沪深300指数1分钟数据的实证表明,最优线性组合已实现波动产生的VaR预测具有明显最低的市场风险资本需求,较高的监管精度和较高的资本运作效率,以及最高的统计精度,是风险管理视角下比较可靠的高频波动率测度。
[Abstract]:As one of the important applications of volatility research is financial risk management, it is proposed to compare various high-frequency volatility measures from the perspective of risk management. The two-scale fluctuation has been realized, the range difference and the optimal linear combination have been realized, and the Realized GARCH model under the assumption of partial student t distribution is used to predict the future on 1st and 5th. In 10th and 20th, the value at risk is VaR, and from the statistical accuracy, regulatory accuracy, efficiency of capital operation and Basel II requirements of market risk capital. The VaR prediction results of six different high-frequency volatility measures are compared. The empirical results of Shanghai and Shenzhen 300 index 1 minute data show that. The VaR forecast produced by the optimal linear combination has obvious minimum market risk capital demand, higher supervision accuracy and higher capital operation efficiency, and the highest statistical accuracy. It is a reliable measure of high frequency volatility from the perspective of risk management.
【作者单位】：南京大学工程管理学院;
【基金】：国家自然科学基金资助项目(71201075) 江苏省自然科学基金资助项目(BK2011561) 高等学校博士学科点专项科研基金资助项目(20120091120003) 教育部留学回国人员科研启动基金资助项目
【分类号】：F832.51;F224
【正文快照】： 1 在险价值(Value at Risk,VaR)是目前业界最为广泛使用的风险管理工具,它衡量一段时期内金融资产在给定置信水平下可能出现的最大损失,因此对其预测的关键在于对收益及其波动率的建模刻画。早期的波动率模型研究主要采用低频数据,通过对收益条件方差建模(ARCH/GARCH模型或者S

【共引文献】