Research On Statistical Methods Of High Frequency Data Of Shenzhen Stock Exchange Market

The stock value is mainly determined by the fundamentals of the company, and the price of the stock mainly focus on fluctuations in the value, because the value of the stock is in a dynamic process of change and also by market factors etc. there are many factors influence the fluctuation of stock price is difficult to estimate. Grasp of ordinary investors in the market it is difficult to the company’s fundamentals, market factors and investors psychological factors, but according to the technical analysis of stock market basic assumption, all of these factors are the stock price to reflect. So the study of stock price has very important significance. Each pen market transactions form a transaction data, before due to technical reasons, unable to each transaction data are used to analyze, only the use of assumed several important data, such as K-map opening price, closing price, the highest price and lowest to represent the whole day basic situation of stock trading.With the development of computer and network technology, the daily transaction data of stock can also be obtained. According to the need, we can provide the transaction data of different time interval, which is called high frequency data. On high-frequency trading data reflect the day of stock trading, the company fundamentals, market factors and investors’ psychological factors such as, how to use statistical methods to study the high frequency data has important value. Firstly in this paper, by using the method of descriptive statistics, distribution of the study of the high frequency data using graphical bar charts, box shaped, mean, kurtosis, skewness, statistical inference and hypothesis testing and parameter method to study the distribution of stock price high frequency data. The results show that the distribution of high frequency data is generally not well distributed and the log normal distribution, and the distribution of the data is very difficult to describe the distribution of the data. By the bootstrap method), on the high frequency data of sampling simulation, obtain the high prices mean and confidence intervals, so as to obtain the K line graph representation of stock transaction possible errors. Finally, using logistic regression model to study the dynamic factors based on the high frequency data on stock price volatility quantitative model, using this model can short-term fluctuations of stock prices provides forecasts for stock investment provides a new way of thinking.