本計畫旨在提出一新的價格跳躍與共躍的檢定方式，想法值基於現存的一些應用高頻資料的檢定方法所存在不一而足的缺失，因而利用簡單的盒鬚圖所使用的分量統計量作為日內判定跳躍與共躍的方法。其中一項重要的觀察為一旦跳躍與共躍的存在會使日內報酬(或交乘項)的分配呈現厚尾與偏態，因此可以座位檢定的來源。在計畫中我們討論了本法相關的優點與比較。The main purpose of our study is to propose amd construct a new nonparametric test based on quantiles as we visualized in a typical Box-whisker plot for price jumps and cojumps among assets in financial markets. The idea is originated from a simple observation on the jump contributed skewness and kurtosis of the high-frequency intradaily returns. One nice feature to go with the simple approach is that we no longer need the Gaussianity assumption and thus is more flexible. The other is that it does not hinge on the higher order moments approximations of various power variations or integrated quarticity for standardization of its jump/cojump statistic for jump identification. Moreover, the issues of market microstructure noise are naturally mitigated within this approach. 研究期間：10008 ~ 10107
