Design, Backtest and Run your Binance Trading Bot on GCP

Description

Description

In this course you will learn an example of how to combine three technical indicators (RSI, Bollinger Bands and Engulfing Pattern) to define a trading strategy for Bitcoin (BTCUSD) using Python, Ta-Lib and Binance API. You will perform a backtesting of this strategy to see if it is a successful strategy or not. The execution of the backtesting will take several hours, so you will learn how to containerize your Python script using Docker and how to push it to Google Cloud Platform, specifically how to push the container to Artifact Registry and then run the container on Google Compute Engine. Finally, you will see how to run your trading bot 24/7 on Google Cloud again by using Docker, Artifact Registry and Google Compute Engine.

Overview of the contents

Section 1: Basics

• Create local environment for our experiments using Docker (a Jupyter notebook with specific libraries).

• Usage of Binance API: getting credentials, extraction of historical prices, checking filters for BTCUSDT, executing market and OCO orders, getting the id and status of an order. Also canceling an order.

• Reviewing RSI, Bollinger Bands and Engulfing Pattern.

• Programming them using Ta-Lib and building some basic plots.

Section 2: Defining and Visualizing the Trading Strategy

• Describe the trading strategy.

• Programming the buy signal.

• Plotting simultaneously the Bollinger Bands, RSI and Engulfing Pattern.

• Based on previous plot, define the stoploss and takeprofit.

• Check a couple examples (a winning trade and a losing trade).

Section 3: Backtesting the Trading Strategy on Google Compute Engine

• Creation of Dockerfile, requirements.txt and the main Python script (bot_backtesting).

• The bot_backtesting script Includes a connection to BigQuery to save logs during the execution and to save the final result at the end of the execution. Also it will handle Binance fees and any existing open trades.

• In bot_backtesting script, construct the main function (compute_sl_tp) steps:

      1 Browsing for buy signals.

      2 Setting entry, takeprofit and stoploss prices.

      3 Determine when we reach each one of them.

      4 Which one happened first.

      5 Retrieve additional information.

• Building container with Docker, push it to Artifact Registry and run it on Google Compute Engine.

• Analyze backtesting results.

Section 4: Building and Running the Trading Bot on Google Compute Engine.

• Creation of Dockerfile, requirements.txt and the main script (main py) which includes a class called TradingBot.

• The class will handle buy signals. Also it will define the entry, takeprofit and stoploss prices, estimated fees and returns, and whether we have an existing open trade or not.

• The class will define the required conditions to enter a trade.

• It will create market and OCO orders. Also it will check and update the status of both market and OCO orders.

• Running the class TradingBot every minute.

• Building container with Docker, push it to Artifact Registry and run it on Google Compute Engine. Also see how to run it on local machine.

• Results after running trading bot.

Section 5: Appendix

• Installing basic tools: Notepad++, Google Cloud SDK, Docker.

• Enable Google Cloud components: Compute Engine, Artifact Registry and BigQuery

• Create Json Service Account for connection from Python to BigQuery.

• Provide additional permissions to SDK: Gmail, Docker, Artifact Registry and repositories in Artifact Registry.

Who this course is for:

• Python developers interested in trading bots and cryptocurrencies.
• Traders with programming skills.