摘要
高频数据分析是理解市场微观结构极为有效的手段,在正态性假设不成立情况下,本文采用半参数极值理论对高频数据的分布尾部进行估计,进而估计风险价值。返回检验的结果表明,此方法可以比较精确地度量风险价值。
The analyses of high frequency data are a very effective approach to understand the microcosmic structure of market.Under the condition that the normality of data is not true,extreme value theory based on method of partially parameterization is adopted to fit the tail of the return distribution for high frequency data,and further to estimate value at risk.The statistical test for the data of Shenzhen index is conducted and indicates that value at risk can be well estimated by the semi-parameterized extreme value theory.
出处
《河南科技大学学报(自然科学版)》
CAS
2008年第4期101-104,共4页
Journal of Henan University of Science And Technology:Natural Science
基金
国家自然科学基金项目(10771015)
关键词
金融学
风险价值
极值理论
高频数据
Finance
Value at risk
Extreme value theories
High frequency data