This is a school project. We do not recommend the use of the information output by the algorithm as a financial advice.
- We assigned ESG scores to each of the CAC40 companies and selected the top x performers.
- For the selected companies, we conducted a monthly Twitter sentiment analysis, utilizing Natural Language Processing (NLP) on tweets related to each company.
- Additionally, we calculated scores based on their half-yearly reports.
- Leveraging machine learning and these scores, we allocated weights to each company within its respective industry sector.
- In parallel, we conducted return forecasts using time series analysis (ARIMA model) and employed Markowitz methodology to determine sector weights.
- Finally, we multiplied the two sets of weights to construct our monthly portfolio.
-
Python 3.9 (pandas does not yet support python 3.10) : link
-
Python packages :
You can also use the windows command prompt and pip, for example pip3 install -r requirements.txt