Using Company-Specific Headlines and Convolutional Neural Networks to Predict Stock Fluctuations

Readshaw, Jonathan; Giani, Stefano

Abstract

This work presents a Convolutional Neural Network (CNN) for the prediction of next-day stock fluctuations using company-specific news headlines. Experiments to evaluate model performance using various configurations of word-embeddings and convolutional filter widths are reported. The total number of convolutional filters used is far fewer than is common, reducing the dimensionality of the task without loss of accuracy. Furthermore, multiple hidden layers with decreasing dimensionality are employed. A classification accuracy of 61.7% is achieved using pre-learned embeddings, that are finetuned during training to represent the specific context of this task. Multiple filter widths are also implemented to detect different length phrases that are key for classification. Trading simulations are conducted using the presented classification results. Initial investments are more than tripled over an 838 day testing period using the optimal classification configuration and a simple trading strategy. Two novel methods are presented to reduce the risk of the trading simulations. Adjustment of the sigmoid class threshold and re-labelling headlines using multiple classes form the basis of these methods. A combination of these approaches is found to more than double the Average Trade Profit (ATP) achieved during baseline simulations.

Citation

Readshaw, J., & Giani, S. (2021). Using Company-Specific Headlines and Convolutional Neural Networks to Predict Stock Fluctuations. Neural Computing and Applications, 33(24), 17353-17367. https://doi.org/10.1007/s00521-021-06324-9

• Journal Article Type: Article
• Acceptance Date: Jul 8, 2021
• Online Publication Date: Jul 29, 2021
• Publication Date: 2021-12
• Deposit Date: Jul 13, 2021
• Publicly Available Date: Nov 23, 2021
• Journal: Neural Computing and Applications
• Print ISSN: 0941-0643
• Electronic ISSN: 1433-3058
• Publisher: Springer
• Peer Reviewed: Peer Reviewed
• Volume: 33
• Issue: 24
• Pages: 17353-17367
• DOI: https://doi.org/10.1007/s00521-021-06324-9
• Public URL: https://durham-repository.worktribe.com/output/1245268