Essays on the transaction networks of digital currencies

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

In this thesis I explore the transaction networks that result from the transfer of digital currencies. First, I analyse the relationship between digital currency transaction networks and their US dollar prices. In particular evaluating the presence of a cointegrating relationship between the total number of users and the cryptocurrency price. I find that a stable relationship between network size and price is not common and only exists for Ethereum once time series econometric issues are addressed. Significant attention is devoted to obtain inference in this non-standard setting and a network density interpretation for the parameter of interest is provided. The next chapter analyses three years of Bitcoin blockchain data. I construct a network (or graph) of transactions, then condense it to take into account the presence of spurious volume, which is behind 90% of the bitcoin traded. I find that Bitcoin activity drops drastically on the weekend and that the reduction in volume is driven by large players. Looking at the concentration of the network, a small number of users are responsible for a disproportionately large amount of the volume traded. The top 2% of agents generate over 90% of total network volume. The final chapter is concerned with the formation of Bitcoin transactions. In a preliminary step, I identify clusters of activity within the graph and use a register of over four million addresses to assign activity types to the identified communities. After reducing the transaction graph to the community level I use a formation model, which depends on node characteristics and network structure, to test for the presence of homophily and strategic interactions. I find that both homophily and strategic interactions are driving forces in the creation of links between communities in the Bitcoin network. However, these forces are less prevalent when considering high volume connections.
Date of Award20 Jun 2023
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorSenay Sokullu (Supervisor) & Sami Stouli (Supervisor)

Keywords

  • Economics
  • Econometrics
  • digital currencies
  • networks

Cite this

'