| With the technology development,institutions can obtain a large number of transaction data from the financial market instantly,which arouses people’s interest in analyzing the information transmitted in the data collected under different frequencies.Due to the development of computer technology,the use of higher frequency data to achieve intra day transactions based on minute frequency data can be met.High frequency data is an important tool to adapt to the current market development.The high frequency data contains more market information,which can increase the validity and accuracy of the estimation compared with the daily data.Therefore,this paper is committed to using the high-frequency financial data generated by intra day trading to study the pairs trading strategy,to achieve the minute level timing strategy under the trading rules of T+1 in China’s stock market,and to make better use of the instantaneous and rapid information provided by the market on the impact of price changes.In the observation and description of 1-minute high-frequency data,it is found that there is a large amount of microstructure noise in the higher frequency financial data.Therefore,this paper uses 1-minute high-frequency data,after wavelet denoising,to apply it to the Cointegration-OU model and design a pairs trading strategy.Through the online transaction setup,this paper realizes the pairs trading time selection strategy by daily high-frequency,which is applicable to the current Chinese stock market,and applies the strategy to the Internet industry components,which have been widely valued in recent years and have strong demand for enterprises development.This paper carries out a one-year online rolling back test through the JuKuan platform from 2018.11.1 to 2019.11.1,realizing a strategic return of 56.4% and the annual return of 58.13%.In order to test the effectiveness of the trading strategy established in this paper,this paper evaluates it from two aspects: model comparison and strategy applicability test.In the aspect of model comparison,the main strategy,Cointegration-OU model using after denoise 1 minute data established in this paper is compared with the model using 1 minute raw data,5 minutes data and daily data by the trade back test results,proving that using high-frequency data can make better use of data information and accurately estimate the model.At the same time,empirical test shows that the performance of Cointegration-OU model is better than that of single use of OU model,which proves the superiority of the model in this paper.In the aspect of strategy applicability test,it applies the main trading strategy to 8 other industries for back testing,and selects two major bull and bear markets in China’s stock market in recent years for back testing.All in all,this paper proves the robustness of the trading strategy and its applicability in Chinese market from many aspects. |
