论文部分内容阅读
本文从方差分析、结构相关和回归分析三个层面研究ETF基金市场效应的统计检验方法。首先,基于高频交易数据构造条件异方差波动、实例波动、相对价差、价格水平和交易量的均值比及中值比;第二步,基于t和Wilcoxon符号秩统计量,以方差分析方法检验ETF基金上市前后标的成份股的量价变化;第三步,基于Pearson线性相关、Spearman Rho秩相关、Kendall Tau秩相关,以结构相关方法检验ETF基金上市前后标的成份股的结构变化;第四步,基于实例波动、价格水平、交易量和相对价差的均值比,以回归方法检验ETF基金上市前后标的成份股的弹性变化。最后,实证结果表明这样三层次统计方法可以到检验上证50ETF基金的市场效应。
This article from the variance analysis, structural correlation and regression analysis of three levels of ETF fund market effect statistical test method. First, based on the high-frequency trading data, we constructed the conditional heteroskedasticity volatility, the instance volatility, the relative spread, the average price ratio and the median ratio of the price level and the transaction volume. Secondly, based on t and Wilcoxon signed rank statistic, ETF fund listed before and after the listing of the underlying stock price changes; the third step, based on Pearson linear correlation, Spearman Rho rank correlation, Kendall Tau rank correlation, the structural correlation method to test the ETF fund listing constituent stocks before and after the change; fourth step , Based on the volatility of the instance, the price level, the trading volume and the average ratio of the relative spread, the return method was used to test the elastic changes of the constituent stocks before and after listing of ETF funds. Finally, the empirical results show that such a three-level statistical method can test the market effect of the SSE50 ETF.