This project is an automated trading system that trades cryptocurrencies on the Binance exchange. By default it trades BTC/USDT, but any other pairs on the exchange can be traded.
The project contains a number of pre-built trading strategies, located inside strategies.py. The default strategy uses an LSTM neural network, which has been trained on all available 1-minute BTC/USDT data from Binance (around 3 years worth in total).
Prices and executed trades are visualised in an automatically updating chart, allowing for easy monitoring of positions. Minute by minute candle data is continually appended to a CSV for analysis, alongside a log of trades and account balances to track profitability.
Please read the disclaimer before trading using this bot.
Ensure that you have installed all packages in requirements.txt, either manually or by using:
pip install -r requirements.txtBy default, the script looks for Binance API keys saved in a folder called config in the parent directory of the project. The keys are expected to be in a file called algo_config.json which takes the form:
{"api_key": "INSERT API KEY HERE", "api_secret": "INSERT API SECRET HERE"}This location can of course be modified in trading_bot_main.py. Ensure that these keys remain secure as they permit access to all funds on the account.
Run trading_bot_main.py to begin trading. Ensure that TRADE_QUANTITY is modified to be the amount you would like to trade: this must be at least 0.001 when trading BTC/USDT.
Running visualise.py will create an automatically updating chart that displays the price movements since the bot began running, as well green dots for points at which a buy order was executed and red dots to indicate executed sell orders. Note that this uses matplotlib.animation.FuncAnimation which is a little temperamental about the IDEs it works well with - I found issues running this in VSCode, but no problems using IDLE.
The strategies used for trading are defined as classes in strategies.py, and can be easily extended when new ideas are developed. Once a new strategy class is defined, simply modify the initialisation of the Strategy object in trading_bot_main.py.
Please review all code in this project before allowing access to your funds. This project is presented without license or warranty, and no responsibility can be taken for money lost through trading or otherwise. No promise is made here about the profitability of the strategies used to trade, and users should be aware that investments may decrease in value or become worthless.