Abstract: As underlying of financial derivatives in China financial market, the jumps and their dynamic play import role in asset pricing and risk management. In this paper, the author builds up a new estimation of spot variance based on threshold bipower variation and establishes an asymptotically more efficent jump test statistics. With this new test jump statistics, jump test and dynamic analysis based on five-minutes data of CSI300 are conducted. The empirical evidence shows that jumps in CSI 300 are ubiquitous and driven by a compound Poisson process with time-varing density and time-varing jump size,which implies that jump size is not i.i.d.. The empirical results in this paper can be foundamental for relevent studies.

Key words: jump, threshold bipower variation, spot variance, jump dynamics