This paper studies how the Greek sovereign credit event in March 2012 impacted the credit default swap (CDS) market from market-wide and investor behaviour perspectives, using both network tools to a dataset of snapshots of the global bilateral CDS exposures and a panel analysis on CDS spreads. Regarding the CDS spreads, we find very little discernible direct impact of the Greek credit event on CDS spreads overall. This finding provides some further evidence that the Greek credit event was well anticipated by most market participants. However, we find several significant changes in the Greek CDS network structure following the credit event: the number of connections via exposures declined significantly, the directionality of the positions (net long vs net short) of the main groups of market participants reversed, while none of the non-banks returned to trade Greek CDSs until the last observation of dataset (October 2014). Regarding indirect effects to other CDS markets, we find evidence of temporary spill-over effects on CDS reference entities with credit risk associated with the risk of the Greek sovereign. In particular, the market and counterparty structures changed temporarily with all types of traders decreasing their exposures to the EU periphery sovereign reference entities and also changing their trading counterparties, while after some time, the structure of the market returned to a similar one observed before the credit event. Finally, we find some support for the bank-sovereign nexus, as there was a consistent retreat from the CDS exposures on banks in the EU periphery countries, contrary to banks residing in the other EU countries.

Building on the literature on systemic risk and financial contagion, the paper introduces estimated network linkages into an early-warning model to predict bank distress among European banks. We use multivariate extreme value theory to estimate equity-based tail-dependence networks, whose links proxy for the markets’ view of bank interconnectedness in case of elevated financial stress. The paper finds that early warning models including estimated tail dependencies consistently outperform bank-specific benchmark models without networks. The results are robust to variation in model specification and also hold in relation to simpler benchmarks of contagion. Generally, this paper gives direct support for measures of interconnectedness in early-warning models, and moves toward a unified representation of cyclical and cross-sectional dimensions of systemic risk.