Influencer Detection meets Network AutoRegression – Influential Regions in the Bitcoin Blockchain

Abstract

Known as an active global virtual money network, Bitcoin blockchain with millions of accounts has played an ever-growing important role in fund transition, digital payment and hedging. We propose a method to Detect Influencers in Network AutoRegressive models (DINAR) via sparse-group regularization to detect regions influencing others cross-border. For a granular analysis we analyze if the transaction record size plays a role for the dynamics of the cross-border transactions in the network. With two-layer sparsity, DINAR enables discovering 1) the active regions with influential impact on the global digital money network and 2) if changes in the transaction record size impact the dynamic evolution of Bitcoin transactions. We illustrate the finite sample performance of DINAR along with intensive simulation studies and investigate its asymptotic properties. In the real data analysis on Bitcoin blockchain from Feb 2012 to December 2021, we found that in the earlier years (2012-2016) network effects came surprisingly from Africa and South America. In 2017 Asia and Europe dominate whereas from 2018 effects majorly originate from North America. The effects are robust in regard to different groupings, evaluation periods and choice of regularization parameters.

Keywords: Bitcoin Blockchain, Network Dynamics, Two-Layer sparsity

JEL Classification: C55, C58, C60, G17