Comovement between stock and bond markets and the ‘flight-to-quality’ during financial market turmoil – a case of the Eurozone countries most affected by the sovereign debt crisis of 2010–2011

Abstract

This article examines the comovement between stock market returns and the dynamics (i.e. changes) of sovereign bond yields for the Eurozone countries affected by the sovereign debt crisis and for Germany, whose sovereign bonds are normally regarded as ‘safe havens’ in the financial markets. To investigate whether the periods of the greatest financial market turmoil in the last decade coincided with a Flight-To-Quality (FTQ; the flight from the riskier stocks to safer sovereign bonds), a simple FTQ indicator is defined and calculated. We find that the comovement between stock market returns and the dynamics of sovereign bond yields is time varying. For Germany, the correlation between stock market returns and the dynamics of sovereign bond yields is mostly positive during the observed period, whereas for the countries affected by the sovereign debt crisis, the correlation is mostly negative, especially after the start of the sovereign debt crisis in the Eurozone. Financial crises normally coincide with higher correlations, whereas Greece's sovereign debt crisis in the Eurozone led to a reduced correlation in Ireland, Italy, Portugal and Spain. The financial market crises prior to 2010 led to an increased FTQ in all investigated countries, whereas after the start of the sovereign debt crisis in the Eurozone, the FTQ is only observed in Germany.

Keywords:

• Eurozone
• flight-to-quality
• financial market turmoil
• sovereign debt crisis

JEL Classification:

• G11
• E44
• H63

Notes

¹ Asset return series, entering in a DCC-GARCH model, have to be filtered, so that the expected (mean) value of the series is null. In this article, this is achieved by demeaning the return series. As Engle (2002) argues, the SEs of the DCC-GARCH model do not depend on the choice of filtration.

² The description of the DCC-GARCH models is summarized from Engle and Sheppard (2001). We use the same notation as the authors.

³ The bond yield dynamics was calculated this way in Durré and Giot (2005) and was also suggested by a referee. Alternatively, bond yield changes could also be calculated as differences in bond yields ln(y_t ) − ln(y _t−1) as suggested by Durré and Giot (2005) or by Kim and In (2007). We run the dynamical conditional correlation analysis also for the later definition of bond yield changes and the results (not presented here) were very similar to the results of the former definition of the bond yield changes.

⁴ The results are not presented here but can be obtained from the author.