Crypto Currency Trading Bot in Python

Windows desktop app for crypto portfolio management - using REST API and live data streams. User can place, cancel order, overview status and balance of the their account and see live market data. It is a multithreaded application

Tools

• Python 3.7.9
• numpy-1.21.2
• pandas-1.3.3
• python-dateutil-2.8.2
• pytz-2021.1 six-1.16
• numpy
• python-dateutil
• python-dotenv
• requests
• websocket-client v0.58.0

Installation & Usage

Installation

• pipenv --python 3.7.9 - to setup project with required python version
• pipenv shell - enter virtual env
• pipenv install pandas numpy python-dateutil pytz requests --dev - install required dependencies
• pipenv requirements --dev > requirements.txt - generate the requirement list

Usage

• pipenv shell - enter VE
• python main.py - to run application

App Planning

Requirements

1. Interface

• Display market data (decision helper)
• Display messages to the user (loggin)
• Set up a strategy and start/stop it
• Follow the orders/trades (PNL, order status...)
  (4 components)

Interface design version 1.0

2. Connectivity (APIs)

• Binance futures API
• Bitmex API
• Authentication to both APIs
  Using 2 types of connections REST API and web socket API (2 connectors)

3. Trading Automation

• Parse parameters set by the user in the interface
• Parse market data received by connectors
• Define a logic for each strategy (entry and exit)
• Manage orders / positions (Strategy class)

4. Other

• Multi-threading: to enable live data updates in paraller to user interactions
• Logger to log actions in the dedicated file

Desktop vs Web App

This application is developed for the desktop use to avoid number of security issues related to the web applications. As well as avoid the need of high compute power when deployed, thus reduce the cost. The main goal of this app is to explore the principles and strategies of trading applications and related algorithms, as well as utilise the existing APIs for crypto trading.

Desktop	Web App
Pros
Easier to handle security	Strict separation of front and back end
Easier to develop	Interface design is superior
Uses local computing power	Easier to make it available to multiple users
Can run on the server as well	Accessible through any device
Cons
Interface design is limited	Needs to handle authentication
Needs to be installed on every machine	Expensive and consumes a lot of compute power
Needs a separate development to handle different operational OS