The Determinants of Sovereign Bond Yield Spreads in the EMU: A GVAR Model

Abstract

The literature about the sovereign yield spreads grew substantially with the euro debt crisis, mainly because of the rising concerns about the Eurozone future and the skepticism regarding the radical shifts in few EMU countries’ interest rates which cannot be explained by standard models. With unsustainable borrowing cost levels, Greece, Ireland and Portugal lost access to the market and part of the public opinion condemned what they called a market irrational behavior. Academics try to bring new lights to these market reactions.Our paper aims to contribute to the growing body of literature that studies the determinants of bond yield differentials in the Eurozone and differs in many ways from the previous works. First of all, we use longer data sets, which cover the recent Eurozone crisis events. Then, we use a new risk aversion proxy based on European data, which can be described as the difference between the euro corporative bond indexes Bbb and Aaa. We also perform an event study of the European Commission fiscal forecasts publication for European countries to help specify our model with the fiscal fundamentals that significantly influence the spreads. Finally, this paper proposes a new application of a GVAR model for the sovereign bond spreads analysis in the Eurozone that outperforms standard models thanks to the inclusion of fiscal fundamental and contagion variables. Indeed we extend the Global Vector Autoregressive (GVAR) model introduced by Favero (2013) in whom we add variables capturing the credit risk (real GDP growth), the contagion effect and a dummy variable for the crisis.The results estimated with our GVAR model suggest a lack of consensus among euro countries and significant differences between the GIIPS and the Core countries in terms of spreads determinants. Moreover by taking the model to country by country data, the R-squared drastically increase which shows the importance of idiosyncratic behavior in the spreads movements. Overall, we find that the sovereign bond yield spreads were mainly influenced by the expected debt to GDP, the risk aversion and the contagion factors, which is in line with the related literature findings.