Download

Work in progress


Abstract

This paper uses measures developed in the finance literature to analyse sovereign risk networks of emerging and advanced economies. Using mixed-frequency data, I decompose sovereign bond yields into observed common and country-specific factors and unobserved idiosyncratic factors, and apply elastic net regularization (LASSO and ridge) to estimate high-dimensional network adjacency matrices in the presence of sparsity and multicollinearity. Network analysis of these factors reveals important linkages and dependencies between country groups that can be used for financial stability monitoring.