Information Networks in the Financial Sector and Systemic Risk

Abstract

We create and test two novel network-based measures of interconnectedness in the financial industry during 1996 to 2013. A network based on informed trading in financial firms predicts firm-specific risk and performance, while one formed on financial firm returns predicts future macroeconomic risk. The measure of informed trading is robust to variable order arrival rates more common in modern algorithmic trading. A trading strategy based on informed trading network centrality in the financial sector delivers an annualized risk-adjusted return of 7.73%. This risk-adjusted return shows that the network centrality has an economic impact that is relevant beyond the statistical results of the paper.