The Markov Modeling In High Frequency Data

With the completion of the financial market, we are required the analysis on mi-crostructure of market of market more and more, thus explain the relevant financial phenomena and offer the reference for formulation of the financial policy. The high-frequency data can reflect the change of the market in time because of its special property;it is very helpful for us to study the micro- structure. This text utilize Markov chain to describe the high-frequency trade state of data, define the observed income and obtained the negative first order autocorrelation of the income, it is identical with others results of study. At the same time, we defined the time interval of transaction;find no linear correlation between the intervals. Finally we carried on the analysis of the real example to IBM data, found that supports our model.