As the future movements of financial time series like the European Central Bank’s benchmark rate are exposed to uncertainty, financial market participants regularly have to rely on professional analysts’ forecasts. Not surprisingly—and for decades already—the quality of survey forecasts has been evaluated, with heterogeneous results. In addition, forecasters’ performance can change through the course of time. This may happen not only due to wrong or inadequate underlying models. Especially in times of financial turmoil or monetary crisis—like the European debt crisis—the interest rate moves made by central bankers may become even harder to predict (at least the direct reaction to the crisis). Because of this, we evaluate the performance of survey forecasts for the three months rate in the Euro zone performed by financial professionals and test for structural breaks to evidence for crisis related changes and the corresponding forecast errors.

Purpose: This paper aims to investigate the long-term relationships of long-term European Monetary Union (EMU) government bond yields. From an asset managers’ or risk managers’ perspective during the euro crisis, the relevance of sovereign credit and redenomination risk became a major issue. Furthermore, it has to be differentiated between core and non-core EMU member countries.Design/methodology/approach: Methods of applied time series analysis are used to investigate EMU government bond yields and EMU government bond yield spreads for Spain, Italy, The Netherlands, Austria and Germany. Both standard unit root testing procedures and breakpoint unit root tests are used to examine cointegrating relationships and structural changes in these relationships.Findings: The empirical results deliver clear evidence for structural shifts in the long-term relationship between German and the two non-core EMU countries (Italy and Spain). The timing of the breaks coincides with the timing of the euro crisis. On the contrary, the results for Austria and The Netherlands are different from the findings for the two non-core countries.Research limitations/implications: One major limitation of the study is the limited availability of data regarding to the reaction of asset managers or risk managers to the euro crisis. Especially in the context of the discussion with regard to the relevant risk-free rate for investors, this strand of research is relatively new.Practical implications: A deeper understanding of changes in the long-term relationship between government bond yields and the re-emergence of redenomination risk is important for asset managers and risk managers in the financial services industry. This is especially true for German life insurers.Originality/value: The study provides various empirical contributions to the literature on the euro crisis and sovereign credit risk. First, previous results with regard to the structural changes in the long-term relationship between German and Spanish, German and Italian, German and Austrian as well as Germany and Dutch government bond yields are confirmed using unit root breakpoint tests. Second, investigating the autoregressive coefficient and the timing of the breaks delivers evidence that non-core countries have been more exposed to the fear of redenomination risk. Third, we raise the question which risk free interest rate is relevant for the affected countries.