Has ECB communication been helpful in predicting interest rate decisions? An evaluation of the early years of the Economic and Monetary Union

Abstract

We examine the usefulness of communication by the European Central Bank for predicting its policy decisions during the early years of the European Economic and Monetary Union. Using ordered probit models based on the Taylor rule, we find that statements on the main refinancing rate and future inflation are significantly related to interest rate decisions. At the same time, an out-of-sample evaluation shows that communication-based models do not outperform models based on macroeconomic data in predicting decisions. Both types of models have difficulty in predicting changes in the main refinancing rate.

Acknowledgements

We thank participants in the 21st EEA Congress, the CESifo area conference and the annual conference at the University of Crete, as well as participants in seminars at de Nederlandsche Bank and Radboud University Nijmegen for useful comments. We particularly thank Roel Beetsma, Marcel Fratzscher, Michael Funke and Elmer Sterken. Comments by two anonymous referees of this Journal greatly improved this article. Any errors are naturally our own responsibility. Views expressed in this article do not necessarily coincide with those of De Nederlandsche Bank.

Notes

¹ Jansen and De Haan (2006) show that communication by the ECB during the early years of the EMU has often been contradictory. Jansen and De Haan (2007) analyse the ECB's use of the keyword ‘vigilance.’

² In the remainder of this article, we often use the term ‘ECB communication’ for sake of brevity.

³ Also, Sager and Taylor (2004) find that announcement by the ECB Governing Council contain significant news content.

⁴ Recent studies that estimate Taylor rule models for ECB monetary policy are Belke and Polleit (2007), Garcia-Iglesias (2007), Sauer and Sturm (2007), Gorter et al. (2008) and Moons and Van Poeck (2008).

⁵ Svensson (2003) has shown that, even if the ultimate objective of monetary policy is to stabilize inflation and output, a simple Taylor rule will not be optimal in a reasonable macroeconomic model. Interest rate changes affect inflation and output with a sizable lag. Therefore, monetary policy has to be forward-looking, i.e. it should be based on expected inflation and output. Realized outcomes for inflation and output enter the optimal decision rule if they help to predict future inflation and output.

⁶ Results are reported in the Appendix. To assess stationarity, we use both augmented Dickey–Fuller tests and the Kwiatkowski et al. (1992) test. The latter test is useful as it has a higher power in small samples. See also Hu and Phillips (2004) for a discussion on stationarity in the context of Federal Reserve policy.

⁷ We use the term ‘economic growth’ rather than ‘output gap’ here. The reason is that, in our dataset, central bankers communicated in terms of economic growth (i.e. growth of gross domestic product) rather than the output gap (i.e. the difference between actual and potential output).

⁸ In the remainder of the article, we no longer report results for the model including changes in money growth. Results including M3 are similar to those reported.

⁹ The decision to lower rates by 50 basis points on 17 September 2001 was unscheduled. It came in the aftermath of the 9/11 terrorist attacks in the United States. In this individual case, the results may be biased in favour of the communication-based models. After such an event, communication will adjust more quickly and be more readily available than forward-looking variables.

¹⁰ We only describe the results of these extensions in broad terms here. Detailed results are available on request from the corresponding author.