基于高频交易数据的在险价值模型实证研究

An Empirical Study of Risk Value Model Based on High Frequency Trading Data

高频时序分析是近些年来金融界研究的热门课题之一,而风险价值则是衡量金融工具风险大小的重要工具之一,因此对该问题进行研究具有重要意义。我们以沪深300期货收盘价的五分钟交易数据为研究对象,利用正态分布参数法、历史模拟法和蒙特卡罗模拟法测度了相应的VaR值。研究结果表明,数据整体分布非随机,并且三种方法测度的VaR值均较小,同时,蒙特卡罗模拟法测度的结果与其他两种方法所估计的VaR值差异较大。

high-frequency time series analysis is one of the hot topics in financial circles in recent years,and risk value is one of the important tools to measure the risk of financial instruments.Taking the five-minute trading data of the closing price of Shanghai and Shenzhen 300 futures as the research object,the corresponding VaR values were measured by normal distribution parameter method,historical simulation method and Monte Carlo simulation method.The results show that the overall distribution of the data is not random,and the VaR values of the three methods are all small.At the same time,the results of the Monte Carlo simulation measure are quite different from the VaR values estimated by the other two methods.