Topic and Objectives
The government support programs for the financial sector of the 11 leading Western economies amount to 5 trillion Euro or 18.8 percent of their combined gross domestic product (Panetta et al. 2009). These immense costs cannot readily be absorbed by major economies. Archarya et al. (2013) provide evidence that the "financial sector bailouts were an integral factor in igniting the rise of sovereign credit risk in the developed economies". Even worse, their results point towards a "loop effect" between the credit risk of sovereigns and banks. However, their empirical analysis uses standard OLS regressions.
This project applies the PVAR (panel vector autoregression) approach used in Kraft and Schmidt (2013) using CDS data of European banks and sovereigns. Several rounds of panel vector autoregressions are run in order to estimate the spillover effects between European financial sectors and the corresponding sovereigns. Compared to OLS regressions, cleaner causal relations can be identified; compared to standard VAR approaches, the PVAR allows for a more detailed analysis of the data, as individual characteristics (e.g. accounting and market valuation variables of individual banks) can also be taken into account. It will be possible to link systemic risk of a particular country to the characteristics of the local banking system. The authors can thus address questions like whether a separate banking system ("Trennbanken¬system") or a system with full service banks ("Universalbankensystem") is more prone to systemic risk.
The work builds and extends upon the work of Stanga (2011) analyzing the default swap markets. Using a VAR approach, she studies spillover effects between the banking sector and the sovereign by using a sign restriction approach to identify certain shocks and to calculate impulse responses. Her approach is different as it does not analyze the cross-sectional dimension and for instance does not allow to address crucial issues like the importance of the form of the banking system for the financial stability of a country. In particular, the team can study the spillover effects of single banks to the sovereign sector, while conditioning for accounting and economic variables such as size of the banking sector in relation to GDP, debt-to-GDP-ratio and balance sheet composition.
- We used Bankscope, SNL. Bloomberg, Thomson Reuters and Markit to build a dataset of bank CDS and balance sheet data of 53 banks in 7 major Eurozone countries, accompanied by sovereign CDS and macro data of their respective home countries covering the years 2008-2013.
- After careful review of our findings with all available data we could not confirm our initial findings to be significant. Our analysis of the impact of banking sector and macro variables on the spillover effects of banks to the sovereign sector is inconclusive.