Research On Daily Trading Of China Stock Market Based On High-Frequency Data

Now China's stock market is still at a relatively early stage.And because of the “T+1” trading system of China's stock market,Chinese investors usually focus more on short-term and medium-long-term investment transactions with daily time units,which are less concerned about the ultra-short-term intraday trading.The emergence of the CSI 300 stock index futures provides a good opportunity for the development of China's stock market,making the trading methods of China's stock market more diverse.This paper studies whether the strategy of hedging through the CSI 300 stock index futures and conducting intraday trading in the stock spot market is effective and is intended to provide investors an investment strategy with low risk and stable income.This paper is divided into two parts: the construction of the spot combination and the intraday transaction.The first part builds a stock spot combination to track the CSI 300 Index.Firstly,different methods of constructing the spot are studied,and some methods of optimizing replication are selected to construct the spot combination.The PCA and K-Means algorithm were used to cluster the CSI 300 stocks to obtain ten?twenty?thirty?fortyand fifty types of stocks,and the most volatility constituents are selected from each type of stocks,and the genetic algorithm is used to calculate the optimal weight of the tracking CSI 300 Index.Then the tracking errors of “cluster and optimization weight”,“cluster and equal weight”,“market value arrangement and optimization weight”,“market value arrangement and equal weight” are compared and analyzed,and the tracking error of “cluster and optimization weight” is the best.Finally,use the CSI 300 stock index futures for hedging to help follow-up intraday trading to avoid systemic risks.The second part studies the method of realizing the intraday trading by pre-configuring the corresponding stocks to establish the bottom position.The intraday trading is based on the high-frequency data in the stock market for low-buy and high-selling,mainly using the technical indicators of the stock's5-minute data,such as MACD,KDJ,BOLL and RSI.This paper uses the newer XGBoost lifting algorithm to analyze the discretized technical indicators for the first time.Then based on the prediction result of the XGBoost model,a intraday trading system is built for backtesting,which obtains a stable and low-risk result.The analysis of the backtest results shows that the intraday trading is better when the stock market is in a period of shock.In summary,under the “T+1” trading system of China's stock market,the“T+0” transaction can be carried out by pre-configuring the stock bottom position to achieve low-risk stable returns in a shock stock market.