Structure and influence in a global capital–ownership network

Abstract Relationships between legal entities can be represented as a large weighted directed graph. In this work, we model the global capital ownership network across hundred of millions such with goal establishing methodology for extracting and analysing meaningful patterns capitalistic influence from graph structure. To do so, adapt employ metrics analytics algorithms area maximization. We characterize relationships extracted show that our analysis aligns information macro-economic studies; example it recovers presence known tax heavens, which appear in dense subgraphs countries. also identify quantify cases where is principally owned by others, corresponding to influence. Beyond confirming justifying novel application maximization area, outcome offers new insight domain, highlighting existence strong communities property. leverage methods means evaluate impact these contexts. Finally formulate results study into recommendations future analyses kind.