Forecaster

A trading software that uses various algorithms to predict trend in regolar time spans and make transaction to Trading212 broker service using the Trading212 APIs (that I've made before). In other branches it uses XTBApi (another api built by me) and different algorithm tested with Foreanalyser. There's a lot of work to do here, if you like you can help me build this. Contact me at [email protected].

Behavior

Algorithm

One of the many algorithms I used is the Mean Reversion with this formula:

with avg as a price average, mult for a costant and dev for a deviation. In my tests I found most effective the use of a linear regression as price average and a finantial index named Average True Range (that defines volatility) as deviation.

How to install

Install just with pip:

   cd Forecaster/
   pip install -e .

Then run setup.sh to save the tokens needed by the software.

   chmod +x setup.sh
   ./setup.sh

Developing

Will be used these Design Patterns:

• creational: singleton, factory method
• structural: Proxy, Adapter, Decorator
• behavioral: Chain of responsability, Mediator, Strategy

Main Libraries

• Telegram API
• Trading212 API

The Bot uses Telegram APIs to communicate with the user news and receive commands (asyncronously) and Trading212 API to make transactions and drive predictive algorithms.

Backlog

To-Do list before the v1.0 release.

• Tidy up code
• Add thread handler
• Add more telegram commands
• Add market closure time watcher
• Add database integration