Characterising Cryptocurrency Project Networks Using Graph-Based Analysis

Abstract

With the introduction of smart contracts by the Ethereum blockchain in 2015, cryptocurrencies can now function as decentralized applications (dApps). Over the years, the proliferation of dApps has grown exponential and in 2020, the cryptocurrency market has onboarded and grown from a marketcap of 200 billion at the start of 2020 to almost 3 trillion at its peak valuation by the end of 2021. Users’ interactions with these dApps via the blockchain results in immutable and publicly available records which stores rich information and complete traces of financial activities. Such interactions are often studied from a network perspective but lack a unified approach and tend to focus on the blockchain rather than the dApps. This thesis aims to use networks as a general language to describe dApps as an interacting system in the real world. This thesis proposes a new framework titled Graph Analysis of Cryptocurrency Project Network (GAPNET) to analyze dApps using a collection of graph-based methods. The framework brings together different methods based on Graph Properties and Token Price Correlation. In the thesis, I use this framework to analyze the growth of selected dApp networks in the Ethereum and Binance Chain, investigate relations to real-world networks, and correlate graph properties to token price. Lastly, I close by formulating conjectures from the experiments and provide future research directions. In the long-term, this framework could be used for providing insights to users interested in investing and participating in dApps.