Picking Stocks based on Fundamental Factor Model

Inspiration

The idea was from the process of building a valuation model(e.g.DCF DDM) as analysts would need to evaluate the financial data of a company

What it does

Creat factors based on company financial performance and select the top 20 stocks to build a portfolio for backtesting.

How we built it

Program written in Python on Jupyter notebook Data collected through Thomson Reuters Eikon

Challenges we ran into

The quality of the raw data was poor and therefore we need to spend time cleaning the data.

Accomplishments that we're proud of

The result of the backtest was good

What we learned

A tree-based model could create better factors than that of the linear models

What's next for Picking Stocks based on Fundamental Factor Model

1. Apply the model to mid-cap and small-cap companies
2. Introduce deep learning model in factor generation

Built With

• python