We construct a sovereign default network by employing high-dimensional vector autoregressions obtained by analyzing connectedness in sovereign credit default swap markets. We develop four measures of centrality, namely, degree, betweenness, closeness, and eigenvector centralities, to detect...