|Institution:||University of Otago|
|Keywords:||Systemic Risk, Systematic Risk, Idiosyncratic Risk, Decomposition, Risk Components, Sovereign Credit Risk, Credit Default Swaps, Granger Causality in Value at Risk, Time Varying, Rolling Window Analysis.|
|Full text PDF:||http://hdl.handle.net/10523/5316|
This thesis examines to what extent systemic, systematic and idiosyncratic risk drive sovereign credit risk and how these risk components vary over time during the pre-Global Financial Crisis (GFC), GFC and post-GFC periods. Using a sample of 77 countries over the period 1 January 2005 – 3 September 2013, I investigate the time-varying dimension of systemic, systematic and idiosyncratic risk components of sovereign credit risk. This study uses daily returns on sovereign Credit Default Swap (CDS) spreads to measure sovereign credit risk. First, I review the existing systemic risk literature and derive a definition for systemic risk as the contagion spillover of extreme risk caused by a trigger event. Second, I perform the Granger causality in Value at Risk (VaR) test by Hong, Liu and Wang (2009) on a rolling window basis to identify the contagion paths of extreme risk spillover to measure the systemic component of sovereign credit risk. The systematic risk component is estimated using common global risk factors from a principal component analysis (PCA) applied on a rolling window to the sovereign CDS returns. Third, using a multi-stage regression model I decompose sovereign credit risk into its respective risk components. Idiosyncratic risk is the largest component of sovereign credit risk but is smaller during the crisis period. In contrast, the systemic risk component constitutes a larger proportion of sovereign credit risk during the crisis period than non-crisis periods. The systematic risk component increases over the time frame of this study. Consistent with higher sovereign credit risk exposure the systematic risk component moves in the opposite direction to global indices. Furthermore, there is a trade-off between the systemic and systematic risk components throughout the crisis period. My analysis shows that when a crisis strikes, systemic risk increases significantly due to a sudden shock triggering contagion spillover of extreme risk from one entity to another. As the extreme risk spillover impacts more and more sovereign entities, it becomes a common global risk that affects most countries simultaneously, that is, systematic risk.