Sovereign credit risk spillover

This thesis examines cross-market correlations between means and variances in sovereign credit markets and captures the presence of any contagion effect by focusing on parallel movements between markets in the wake of the recent crisis. Furthermore, it focuses on the effect of policy interventions on the dynamics of these correlations.\ud \ud First, to look at the correlation between markets, we investigate the interaction between sovereign spreads and creditworthiness. Our results suggest that there are stable long-term cointegration relationships and significant short-term reactions between government CDS spreads to rating and outlook changes, with rating and outlook leading CDS spreads. After confirming the leading role of credit ratings, we further investigate the spillover effect from ratings to CDS spreads across markets and countries. We are concerned with the spillover effect of a change in the sovereign credit rating and outlook of one country on the sovereign CDS spreads of other countries. We find that rating and outlook announcements originating from different countries have a strong spillover effect across countries but not across regions, while countries’ initial credit status has limited effect on such spillover. Moreover, the US market is a strong source of global spillover to all the countries. After controlling for US factors, the international spillover effects are found to be stronger during crisis periods than in tranquil periods. In addition, credit outlook changes have a greater impact on sovereign CDS spread responses than rating change announcements, suggesting that outlook changes carry more new information.\ud \ud Furthermore, we are also concerned with the influences of rescue plans by the European Union (EU) and the International Monetary Fund (IMF) on the interdependence of sovereign credit risk, measured by CDS spreads, in the Eurozone. The study focuses on the interaction between two groups of nations, ‘cores’ (Austria, Belgium, France, Germany and the UK) and ‘PIIGS’ (Portugal, Ireland, Italy, Greece and Spain), before and after these bailouts. We are able to control for the rating and other external influences affecting sovereign CDS spreads. There are three principal findings. (1) Before the EU interventions, the spreads of the rescued countries – Greece, Ireland, Portugal and Spain (PIGS) – had a strong influence on rating changes in Austria, Belgium, France, Germany and the UK (core European countries). (2) After bailout, our results underline increased interdependencies between sovereign credit risk in the EU area, especially between the rescued country and the core countries. This suggests that these bailout plans not only increase the influence of the rescued country on the development of the core nations, but also amplify the sensitivity of PIIGS to changes in the cores. (3) Different countries will vary in their financial stability and their fundamentals will differ, so they will be expected to respond differently to a bailout. Indeed, distinctive interaction behaviours across countries, related to country-specific characteristics (fiscal outlook), is found for each of the financial policy interventions.\ud \ud Second, to look at the correlation between variances, this study investigated correlation between 9 major EMU countries’ CDS markets during the sovereign debt crisis, and hence examined the impacts of policy interventions on these markets, using the DCC-GARCH model. The main purpose was to assess the extent to which the policy interventions influenced the dynamics of correlations in sovereign CDS markets, after controlling for international influence (US VIX), and both domestic and foreign sovereign credit rating and outlook. Our results suggest that correlations are time-varying for all the sample countries. Most of the policy interventions led to a significant increase in the pairwise correlations. Our interpretation is that the “two-way feedback” between the healthy country and the bailed-out country causes the public-to-public risk transfer. The increased debt and deficit partly result from assisting other troubled nations. Through policy interventions, any deterioration in the sovereign creditworthiness of the healthy countries could transmit back to the bailed-out countries. Moreover, the estimation result suggests that policy interventions, rather than VIX and credit rating/outlook, play the most direct and significant role in shaping the structure of dynamic correlation in the EMU markets