Successful investment strategies need to be ahead of stock market movements. Machine learning paves the way for the development of financial theories that can forecast those movements. In this work an application of the Triple-Barrier Method and Meta-Labeling techniques is explored with XGBoost for the creation of a sentiment-based trading signal on the S&P 500 stock market index. The results confirm that sentiment data have predictive power, but a lot of work is to be carried out prior to implementing a strategy.
In this repository you will find a Jupyter Notebook with all the Python code used to generate insights--please, note that for a correct loading you may have to use the NbViewer application and that embedded widgets will not render in this preview--, a folder with some snippets of the graphs created and the final thesis.
This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.