Bond Return Predictability: An Investigation for the European Market

Abstract

The predictability of security returns has received considerable attention in the finance literature. Notwithstanding, the predictability of bond returns, in particular outside the US, has been far less explored. In this paper we analyse the ability of several predetermined information variables in predicting bond returns in the European market. We test if variables, commonly used for that matter in the context of other markets (such as inverse relative wealth, term spread, real bond yield and a January dummy) are also useful predictors of European bond returns. Due to some particularities of the sample period of analysis, characterised by the EMU convergence, we also make another contribution by including the yield spread in relation to German bonds. Furthermore, we analyse the return predictability across different bond maturities: 1-3, 3-5 and 5 or more years to maturity. The results indicate that variables like the term spread, IRW and a January dummy represent useful information in order to predict bond returns for different maturities. The other two variables add little in terms of explanatory power. Surprisingly, the DM yield spread does not seem to have any predictive ability for the countries expected to participate in the EMU. However, a puzzling result was obtained: this variable appears to be significant for the UK market! Additionally, we find that investors, using simple trading strategies that exploit this information, may obtain higher returns. This outperformance is observed for different maturities, being more evident for long-term Government bonds. These findings may have important implications on other related issues such as market efficiency, asset pricing and portfolio performance evaluation.