Novel modelling strategies for high-frequency stock trading data

Financial Innovation

Research

Ke Xu¹  Li Xing²  Xuekui Zhang³  Yuying Huang⁴ 

[1] Economics Department at University of Victoria, Victoria, Canada;Mathematics and Statistics Department at University of Saskatchewan, Saskatchewan, Canada;Mathematics and Statistics Department at University of Victoria, Victoria, Canada;Mathematics and Statistics Department at University of Victoria, Victoria, Canada;Statistics and Actuarial Science at University of Waterloo, Waterloo, Canada;

关键词: High-frequency trading;    Machine learning;    Mid-price prediction strategy;    Raw data processing;    Multi-class prediction;    Ensemble learning;   

DOI  :  10.1186/s40854-022-00431-9

 received in 2022-05-19, accepted in 2022-11-24,  发布年份 2022

来源: Springer

【 摘 要 】

Full electronic automation in stock exchanges has recently become popular, generating high-frequency intraday data and motivating the development of near real-time price forecasting methods. Machine learning algorithms are widely applied to mid-price stock predictions. Processing raw data as inputs for prediction models (e.g., data thinning and feature engineering) can primarily affect the performance of the prediction methods. However, researchers rarely discuss this topic. This motivated us to propose three novel modelling strategies for processing raw data. We illustrate how our novel modelling strategies improve forecasting performance by analyzing high-frequency data of the Dow Jones 30 component stocks. In these experiments, our strategies often lead to statistically significant improvement in predictions. The three strategies improve the F1 scores of the SVM models by 0.056, 0.087, and 0.016, respectively.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

40249_2022_1049_Article_IEq56.gif

Fig. 3

40249_2022_1049_Article_IEq20.gif

Fig. 2

Fig. 1

Fig. 11

Fig. 7

Fig. 1

42004_2022_803_Article_IEq55.gif

12936_2022_4438_Article_IEq9.gif

41116_2022_35_Article_IEq27.gif

Fig. 6

41116_2022_35_Article_IEq25.gif

41116_2022_35_Article_IEq24.gif

41116_2022_35_Article_IEq23.gif

41116_2022_35_Article_IEq22.gif

41116_2022_35_Article_IEq20.gif

Fig. 2

41116_2022_35_Article_IEq18.gif

41116_2022_35_Article_IEq17.gif

Fig. 1

41116_2022_35_Article_IEq15.gif

41116_2022_35_Article_IEq14.gif

41116_2022_35_Article_IEq13.gif

41116_2022_35_Article_IEq12.gif

41116_2022_35_Article_IEq10.gif

41116_2022_35_Article_IEq8.gif

Fig. 2

41116_2022_35_Article_IEq4.gif

41116_2022_35_Article_IEq3.gif

41116_2022_35_Article_IEq2.gif

Fig. 4

Fig. 7

Fig. 3

Fig. 1

Fig. 3

42004_2022_800_Article_IEq108.gif

Fig. 2

Fig. 2

Fig. 4

Fig. 1

Fig. 1

41116_2022_35_Article_IEq1.gif

Fig. 2

Fig. 6

Fig. 1

Fig. 2

40249_2022_1049_Article_IEq55.gif

Fig. 4

42004_2022_800_Article_IEq92.gif

42004_2022_803_Article_IEq36.gif

Fig. 66

Fig. 6

Fig. 6

Fig. 1

42004_2022_800_Article_IEq86.gif

Fig. 4

42004_2022_800_Article_IEq84.gif

42004_2022_800_Article_IEq83.gif

42004_2022_800_Article_IEq82.gif

Fig. 5

Fig. 65

Fig. 2

Fig. 64

42004_2022_800_Article_IEq75.gif

Fig. 2

42004_2022_800_Article_IEq73.gif

Fig. 1

13690_2022_1010_Article_IEq4.gif

Fig. 3

Fig. 4

40798_2022_490_Article_IEq60.gif

Fig. 4

Fig. 59

Fig. 3

42004_2022_800_Article_IEq39.gif

Fig. 3

Fig. 1

Fig. 5

Fig. 1

Fig. 1

Fig. 5

42004_2022_800_Article_IEq32.gif

Fig. 4

Fig. 4

Fig. 2

40798_2022_490_Article_IEq58.gif

40798_2022_490_Article_IEq57.gif

40798_2022_490_Article_IEq56.gif

40798_2022_490_Article_IEq55.gif

40798_2022_490_Article_IEq54.gif

40798_2022_490_Article_IEq53.gif

Fig. 1

40798_2022_490_Article_IEq46.gif

40798_2022_490_Article_IEq44.gif

40798_2022_490_Article_IEq42.gif