Skip to content

Picking stocks through various screening methods. Focus on Northern Europe.

Notifications You must be signed in to change notification settings

lseffer/stock_screener

Repository files navigation

Running Python tests

Stock screener

For stocks listed on Nasdaq OMX Nordic.

This is a hobby project that helpes me make better investment decisions, while helping me learn new things. I haven't followed any investment strategy to the letter, but the screener helps me find good stocks and weed out the crap.

I'm working on this project very sporadically and the code could be more beautiful but who has time for that?

Loosely based on:

Piotroski F-Score - https://en.wikipedia.org/wiki/Piotroski_F-Score

Magic Formula - https://en.wikipedia.org/wiki/Magic_formula_investing

NCAV - Net Current Asset Value - https://www.oldschoolvalue.com/blog/investing-strategy/backtest-graham-nnwc-ncav-screen/

Architecture

This stock screener consists of 3 services

  • Postgres database
  • Worker service scheduling and executing jobs that fetch the data about the stocks and stores that in the database
  • Web server for serving a front end with a login portal
    • Flask app served by gunicorn
    • Configured for nginx

Usage:

  • Install Docker and Docker-compose
  • run the make run target
  • Go to localhost:5000 and login with credentials found in dev-vars.env
    • For a production setup use the make run_prod target with your own secret prod-vars.env

TO DO:

  • Add more screening methods
  • Add more tests
  • Possibly refactor the ETL job scripts
  • Add a job for sending same data to Google Sheets (like in previous version)
  • Ask for feedback