This project deals with Quantative Trading, Generating & Combining alpha for better trading strategies .

This project is divided into 3 sections :

1. NLP on 10k-Financial Statements:
   • Scraped and Pre-processed 10K-Financial Statements that are the annual reports that publicly traded companies are required to file with the SEC within 60 days of the fiscal year end.
   • Natural Language Processing Analysis on 10-k financial statements to generate an alpha factor.
2. Analyze Stock Sentiments from StockTwits using Deep Learning:
   • Built a deep learning model to classify the sentiment of messages from StockTwits (a social network for investors and traders).
   • Model predicts if any particular message is positive or negative. From this, a signal of the public sentiment for various ticker symbols is generated.
3. Combining Alpha using Random Forest:
   • Combined signals on a random forest for enhanced alpha.
   • While implementing this, solved the problem of overlapping samples.

Since the project was under Udacity's AI for Trading Nanodegree the dataset was provided by thier partners Quotemedia and Loughran-McDonald sentiment word lists. Moreover, StockTwits and SEC sites were used to fetch data for Stock Sentiment and NLP repectively.

pip install -r requirements.txt

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

MIT