项目名称: 高维高频数据下金融资产积分波动率矩阵的统计分析
项目编号: No.11501348
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 数理科学和化学
项目作者: 夏宁宁
作者单位: 上海财经大学
项目金额: 18万元
中文摘要: 本项目主要探索基于微观结构噪声干扰下的高维高频数据的积分波动率矩阵的估计问题。我们利用随机矩阵理论,建立了积分波动率矩阵与其经典估计矩阵的极限谱分布的关系表达式,阐明了积分波动率矩阵的统计机制,从而提出积分波动率矩阵的新的估计量。其中的关键步骤在于我们首次提出逆转MP方程的思想,阐明如何利用样本协方差阵来描述总体协方差阵的问题。这为高维协方差矩阵的估计问题提供了新的解决方案,为建立高维框架下的极限理论奠定基础。
中文关键词: 高维数据;积分波动率矩阵;已实现波动率矩阵;高频数据;极限谱分布
英文摘要: This project considers estimation of the integrated covariance matrices of high-dimensional diffusion processes based on high-frequency data in the presence of microstructure noise. We adopt the random matrix theory approach to establish the connection between the underlying integrated covariance matrix and itsestimator in terms of their limiting spectral distributions. We illustrate the statistical properties of integrated covariance matrices and further propose an alternative estimator. A key element of the argument is a result describing how the limiting spectral distribution of sample covariance matrices depends on that of population covariance matrices. We call this result the inversion of MP equation. It will put forward a method to estimate high-dimensional population covariance matrices and establish limiting theory in high-dimensional frameworks.
英文关键词: high-dimensional data;integrated covariance matrix;realized covariance matrix;high-frequency data;limiting spectral distribution