Learning Financial Networks using Quantile Granger Causality

In the post-crisis era, financial regulators and policymakers require data-driven tools to quantify systemic risk and to identify systemically important firms. We propose a statistical method that measures connectivity in the financial sector using time series of firms' stock returns. Our method is based on system-wide lower-tail analysis, whereby we estimate linkages between firms that occur when those firms are distressed and that exist conditional on the financial information of all other firms in the sample. This is achieved using Lasso-penalized quantile vector autoregression. By considering centrality measures of the estimated networks, we can assess the build-up of systemic risk and identify risk propagation channels. We apply our method to monthly returns of large U.S. firms, demonstrating that we are able to detect many of the most recent systemic events, in addition to identifying key players in the 2007-2009 U.S. financial crisis. Importantly, these players are not identified using standard Granger causality, which estimates connectivity by averaging across good, bad, and normal days of the market.