【摘要】 高频数据分析是理解市场微观结构极为有效的手段.文章在GARCH模型对高频数据低效的情况下,从非参数的角度给出风险价值一个相合估计量,这个估计量不依赖于总体分布.实证分析表明深圳综合指数5min对数收益率偏离低频数据常具有的正态分布,且本文的非参数方法可以比较精确地度量风险价值.更多还原

【Abstract】 The analysis of high frequency data in Financial Markets is a very effective approach to understand the microcosmic structure of market.Considering the low effective of GARCH model for high frequency data,the nonparametric method is adopted to estimate the value at risk of high frequency data.The demonstration analyses showes that such an estimate is good and stable.更多还原