摘要
文章利用中国证券市场的日内交易数据实证了非参数ACD模型。非参数ACD模型不依赖条件均值的函数形式和误差项的分布形式,更具有一般意义。文章从多个方面进行实证分析。利用非参数方法进行分析的结果表明:数据不能用线性ACD模型来刻画,根据非参数拟合曲面的形状可以把此ACD模型的函数形式设定为某种非线性形式。
In this paper, we have demonstrated the nonparametric ACD model on intraday transaction data of the stock market of China. Because the nonpararnetric ACD model does not rely on the functional form of conditional mean value and the error distribution form, it has more typical implication. The article conducts empirical analysis on several aspects. We draw the conclusion that the result of the nonpararnetric method analysis has indicated that the data in the paper can not be depicted by the linear ACD model, and we can set the functional form of the ACD model as the nonlinear form by the nonparametric fitting.
出处
《统计与信息论坛》
2006年第3期96-99,共4页
Journal of Statistics and Information
关键词
非参数ACD模型
非参数估计
模型形式设定
nonparametric ACD model
nonparametric estimation
the model form specification
