Skip to content

Automatic Cryptocurrency Trading Bot using Triangular or Exchange Arbitrages

License

Notifications You must be signed in to change notification settings

alimogh/crypto-arbitrage

 
 

Repository files navigation

Crypto Arbitrage

Introduction

This is an automatic trading bot using Triangular or Exchange Arbitrages. It reguarly checks and detects arbitrage opportunities, and place orders when a profit can be made. This works on any cryptocurrency pairs with minor configuration.
Started with $1000 in October 2017, there were some times that this made about $40/day for a few weeks with Triangular Arbitrage on Bittrex, but as the market is getting very unstable, the profit is hard to outrun the high price fluctuation.

Exchanges

Bittrex, Bitfinex, Bitstamp, Kraken, Gatecoin (Not maintained anymore due to its extremely low volume)

Setup

  1. Install Python 2 at here https://www.python.org/downloads/ if necessary. This bot will not work with Python 3.
  2. pip install grequests
  3. Rename or copy .key_sample files under keys to .key file, and add the APIs. E.g. Create a file bittrex.key. Make sure not to push any API keys or this .key file.
  4. Modify arbitrage_config.json for any other ticker pairs or exchanges

Usage

Triangular: python main.py -m triangular
Exchange: python main.py -m exchange
Mock mode is enabled by default, which does not place any order and just check and show any arbitrage opportunities. To turn off mock mode and run in production, add the argument -p.

Difficulties

  1. The trading fee is the largest obstacle. Most of the exchanges have a 0.25% fee. The profit will be larger if the fee can be lower.
  2. Sometimes not all the placed orders are executed, so there will be some manual work to rebalance. The bot should be able to deal with this situation, such as placing a market order, instead of just cancelling the open orders.

Further Improvement

  1. Implement exchange rebalancing
  2. Handle open orders strategically
  3. Refactoring

I will put this project on hold now as the price goes up and down so much. Hope this can help anyone with similar interests.

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%