Crypto Backtesting with Python: A Step-by-step Guide

In this guide, we'll build a crypto backtesting tool leveraging the CoinGecko API, that makes it easy to test various trading strategies

This guide will help you build a crypto backtesting tool using the CoinGecko API. With this tool, you can easily test various trading strategies, ranging from simple price-based approaches like buying the dip to more complex strategies utilizing technical analysis and indicators.

As always, you'll find a link to the GitHub repository at the end of the article, allowing you to dive right in and start experimenting.

What is Crypto Backtesting?

In the world of trading, crypto backtesting refers to the process of evaluating a trading strategy using historical market data to assess how it would have performed in the past.

This enables traders to gauge the profitability, risk, and overall effectiveness of a strategy before deploying it in live trading. By simulating trades based on past data, traders can refine their approach, identify potential weaknesses, and gain confidence in their strategy without risking real money.

Pre-requisites

Before we start building our crypto backtesting tool, we'll need the following:

To obtain a CoinGecko API key, head over to the Developer’s Dashboard and click on +Add New Key in the top right corner. For detailed instructions on generating and setting up your key, refer to this guide.

We'll be using the OHLC Chart within Time Range endpoint to fetch historical prices, which is available on the Analyst plan and above. For a free alternative, you may use this endpoint instead. The only difference is that, on the Demo endpoint, you cannot specify a time range.

Step 1. Set Up Your Environment

To get started, create an empty directory, which will serve as the root of your project. Within the root directory, let's create a new virtual environment that will allow us to install our requirements locally, without making any changes to the global Python environment.

Let’s now configure our Python application. Run the following commands to create and activate your environment:

If you're using VS Code, your IDE may also ask you if you’d like to use the local Python compiler – choose yes.

Installing Requirements

We’re now ready to install our project’s requirements. The easiest way to do this is by copying the file below to your root directory in a file called requirements.txt and then running pip install -r requirements.txt.

Installing Ta-Lib (Optional)

There is one more requirement that we need to install: ta-lib. This is a fantastic Python library for calculating indicator values from raw data. Unlike the requirements above, ta-lib requires us to use the release files and build the package ourselves, otherwise, it will error during installation.

Head over to the project’s release page and select a version that matches your OS, CPU architecture, and Python version. For instance, I’m running 64-bit Windows 11 with Python 3.11 and an x86 CPU architecture. The correct release for me was ta_lib-0.6.0-cp311-cp311-win_amd64.whl.

To run this on a Macbook with Python 3.11 and an M1 chip or higher, you may use the following release: ta_lib-0.6.0-cp311-cp311-win_arm64.whl. Once you have downloaded the correct version for your machine, drop the file inside your project root. From your project root, install the package using the file you just downloaded.

For instance: pip install ta_lib-0.6.0-cp311-cp311-win_amd64.whl. This should take care of all the project requirements.

Create project scaffold

Inside your root directory, create the services and utils directories, alongside an empty .env file and an empty main.py file. It should look like so:

Inside your .env file, define a variable called CG_API_KEY and assign your CoinGecko API key as its value. We’ll use this to securely load the key into our app, without hardcoding it in the project files.

Step 2. Defining Utilities

Inside the utils directory that we defined, create a file called load_env.py. This will help us load our API key and define any additional configuration options that we may have.

Note that in addition to our API Key, which we’re storing in cg_api_key, we’ve also defined some basic strategy settings such as a take_profit, a stop_loss, an order size, and a total_amount.

Feel free to adjust these settings to suit your needs, and play around with different settings during backtesting to find the best combination of stop loss and take profit for your strategy.

The backtesting library can behave unpredictably if the input amount is much smaller than the asset's price. To avoid this, we've set the amount high enough to prevent issues.