Skip to content
Using graph algorithms to find arbitrage opportunities
Branch: master
Clone or download
Latest commit aff75f4 Jul 2, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
binance_data first commit Jun 26, 2019
LICENSE Create LICENSE Jun 26, 2019 minor cleanup Jul 2, 2019 minor cleanup Jul 2, 2019
pairs_list.json first commit Jun 26, 2019 changed image Jul 2, 2019
requirements.txt minor cleanup Jul 2, 2019
snapshot.csv first commit Jun 26, 2019


This is the supporting code for my post on using graph theory to discover arbitrage opportunities in a cryptocurrency market.

Getting started

To run it, first sign up to CryptoCompare to get a free API key. Then, you can either replace it after the equals sign at the top of, or create a new text file named API_KEY.txt and paste it there directly.

Then, install dependencies with:

pip install -r requirements.txt

Lastly, run the code:



  • downloads the raw data, creating pairs_list.json, binance_data/ and snapshot.csv.
  • processes this data and puts it into a graph, before running Bellman-Ford to find arbitrage opportunities.

Your turn

Here's a brief list of a few ways that you could extend this project. Some are trivial, some are not!

  • Model transaction fees. This is literally one line of code, multiplying the arbitrage value by e.g 0.999 for each element in the path.
  • Download data from more exchanges. operates completely independently of the data collection, it just needs an adjacency matrix.
  • Automatically run the code at fixed intervals to continuously look for arbitrage.
  • Modify the Bellman-Ford so that it doesn't have to recompute everything when some weights change.
You can’t perform that action at this time.