Does the yield spread have predictive power to signal future economic activity? : an analysis for selected European countries

Abstract

The government bond yield curve’s ability to flatten or invert months or years before a recession is a very prevalent phenomenon and it has managed to signal the majority of recessions in the United States since 1952. Motivated by the fact that currently, the government bond yield spread for diverse major European countries, such as the United Kingdom and Germany, is at a record low point since the 2008-2009 global financial crisis, if the hypothesis holds, this is signalling a recessionary period in such countries in the next 2 years. Thus, this study makes use of historical data spanning from January 2000 till December 2019 to analyse how the relationship of the yield spread against economic indicators, such as real Gross Domestic Product (GDP) and Industrial Production Index (IPI), changed over this time span. In recent years, there has been a renewed interest in the yield curve’s ability to signal future economic activity. In line with the methodology commonly employed in literature, this study makes use of a multiple regression model employing the Ordinary Least Square method. This is further accompanied by a graphical analysis to illustrate how the relationship between the yield spread and the economic indicators changed over time. Conclusions from the empirical analysis, which are also substantiated by the graphical analysis, suggest that this relationship differs from one country to another, even for countries in the same monetary union and with a homogenous monetary policy. For Italy, the yield spread resulted to have the predictive power to signal future economic activity. The same cannot be said in the case of Greece. Secondly, both approaches conclude that there was a structural break in this relationship during the 2008 global financial crisis, where the theoretical negative relationship between the yield spread and both GDP and IPI, changed during this period. This analysis is hindered by a number of limitations mainly regarding data availability for Malta and Greece, which will be discussed furthermore in the study. A number of recommendations were suggested for future research, mainly to improve data availability for certain countries that would increase the timespan of the analysis.