估计量和预测模型的选择对高频协方差阵的预测及组合收益的影响
发布时间:2018-07-04 16:40
本文选题:高频协方差阵 + 动态预测模型 ; 参考:《系统工程》2015年08期
【摘要】:围绕复杂的估计方法是否有助于提高协方差阵的预测效果和组合收益进行研究。在预测模型一定的情况下,从统计精度和经济价值的角度对不同协方差阵估计量的预测效果进行了比较;同时在协方差阵一定的情况下,也对不同的预测模型进行了比较。研究发现同时考虑了噪声和跳跃影响的双频已实现协方差阵(RTSCOV)在所有比较标准下均具有最好的表现,并且较低频协方差阵而言,高频协方差阵应用在投资组合中会获得更高的收益。
[Abstract]:Whether the complex estimation method helps to improve the prediction effect and the combination income of the covariance matrix. In the case of the prediction model, the prediction effect of the estimators of the covariance matrix is compared from the point of view of the statistical accuracy and the economic value, and the different prediction is also given under the case of the covariance matrix. The model is compared. It is found that the dual frequency realized covariance matrix (RTSCOV), which considers both the noise and the jump effects, has the best performance under all the comparative criteria, and the high frequency covariance matrix will gain a higher profit in the investment portfolio than the low frequency covariance matrix.
【作者单位】: 贵州财经大学数学与统计学院;
【基金】:2014年贵州省教育厅高校人文社科研究项目(14ZC307) 2014年贵州省哲学社会科学规划项目(14GZYB17) 贵州省科技基金资助项目(黔科合J字【2009】2062号)
【分类号】:F830.59;F224
,
本文编号:2096691
本文链接:https://www.wllwen.com/jingjilunwen/jingjiguanlilunwen/2096691.html