The bot processes real-time market data from Meta Trader 5 desktop application to make predictions using multiple ML model for different symbols as a single model fails to work with every symbol.
** Note: Only a reference model is provided here which is working for all symbols, So dont trade with real $. **
Using the predictions, the bot execute trades and constantly monitors the active trades. If there are active trades which contradicts the prediction, those are closed and new trades according to predictions are taken. Also it has a trailing SL, which books partials and tries to ensure profits and minimize losses.
Clone the project
git clone https://github.com/Rohan-SSS/forex_trading_bot
cd forex_trading_botCreate virtual environment and install dependencies
python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txtNote: It only works in a windows machine as Meta trader 5 is not available for Linux.
Run the streamlit app to see the predictions
Open 2 terminals and run the following in one terminal
python take_trades.pyand in another run the following after a min or so
python monitor_trades.py