Infrastructure and Techniques to Collect Data and Detect Market Manipulation on Crypto-Exchanges

Abstract

Bitcoin's approximate average daily volume, as reported by the different exchanges individually, was $2.46 billion in March 2021. During the same time, only $110.6 million worth of Apple stock (the most liquid publicly traded stock in the world) was traded given Apple's market cap is nearly 3.4 times the size of Bitcoin's. There are multiple players in the crypto-markets but the ones central to this research are: 1) the exchanges that are also the brokers unlike the public stock markets, 2) the organizations listing their coins on these exchanges, and 3) the traders that do not yet include the institutional traders trading in the public stock markets. The crypto markets have little to no regulations and/or oversight, enabling each player to take measures that are deemed illegal in the public markets but incentivize other players in the markets to interact with them to generate alpha. For example, crypto exchanges are notoriously known to use bots to create and execute fake orders on either side of the spread to drive up the volume traded on their exchanges. This creates an illusion of highly liquid markets thereby incentivizing the traders to come to these exchanges and the organizations to list their coins on these exchanges.

The goal of this research was two-fold: first to build a robust and highly optimized cloud-based infrastructure for capturing and storing multiple streams of data (like order book depth, trades placed, ticker prices, etc.) and second, to identify market manipulation by different players (exchanges, organizations, and traders) using the aforementioned data.