ثبات و اثربخشی ارتباطات بانک مرکزی : شواهد از ECB

In this paper we analyse the introductory statement of the ECB President in his monthly press conference held on Governing Council meeting days. We provide a glossary that translates the qualitative information of the press conferences into an ordered scale. We find that the predictive ability of these statements is similar to that implied by market-based measures of monetary policy expectations. Moreover, ECB words provide complementary, rather than substitutable, information with respect to macroeconomic variables. Finally, we show that market expectations react to the unexpected component of the information released by the ECB, after controlling for the monetary policy shock.

“If you understood what I just said, you must not have heard me correctly” Alan Greenspan, Testimony to a Congressional committee “To me, that is the hallmark of credibility: matching deeds to words. (…) Credibility means that your pronouncements are believed — even though you are bound by no rule and may even have a short-run incentive to renege. In the real world, such credibility is not normally created by incentive compatible compensation schemes nor by rigid precommitment. Rather, it is painstakingly built up by a history of matching deeds to words. A central bank that consistently does what it says will acquire credibility by this definition almost regardless of the institutional structure.” Alan Blinder, Central Bank in Theory and Practice (1998, page 64) What is European Central Bank (ECB) communication about? More generally, is the ECB transparent about its future monetary policy stance? What will the level of its policy rate be in three months from now? Do financial intermediaries understand the messages sent by the ECB about the future path of European monetary policy? Are these messages credible and, thus, promptly incorporated in market expectations? These are some of the questions this paper attempts to answer. To put it differently, our goal is to analyse central bank communication, focusing in particular on ECB practices, its consistency (are words subsequently matched by actions?) and effectiveness (how effective are ECB's words in moving financial markets?). Blinder et al. (2001, page 9) first ask: “Does [central bank] communication really matter?” They then answer: “To date there is no research to report on, so we can only call upon casual experience to back our claim that it does, and quite a lot”. This paper tries to support their claim econometrically.1 In particular, this paper contributes to the empirical literature on central bank transparency in three important ways.2 First, from a methodological point of view, we analyse the semantic content of ECB communication in an explicit way: we provide a glossary to rank the information contained in the ECB President's monthly press conference into an ordered scale about the danger to European price stability and sustained economic growth. Second, we test the ECB's consistency between words and deeds. Third, we verify empirically whether, and to what extent, the public not only understand but also believe the signals sent by the European monetary authority. Our main findings can be summarized as follows. First, by looking at the semantic content of the ECB President's introductory statement to the monthly press conference, we can predict the ECB near future monetary policy moves. Therefore, ECB words are usually followed by facts (i.e., a monetary policy move consistent with what the ECB has previously announced). For instance, when the introductory statement states that “the overall prospects remain in line [or “are appropriate”] with price stability over the medium term”, the Governing Council will not normally change the official rate at its next meeting. Second, the informational content of ECB rhetoric is substantial. On the one hand, the forecasting ability of European monetary authority words is similar to the one implied by market-based measures of monetary policy expectations. On the other, we find that words and data on macroeconomic variables are essentially complementary, rather than substitutable, pieces of information to correctly predict the future Repo rate (which is the key ECB policy rate). Third, we show that innovations (due to ECB's announcements) in market expectations about future monetary policy can be explained by the unexpected component of the information released by the ECB, after controlling for the monetary policy shock. Hence, we infer that even if the ECB is a relatively young multinational financial institution, it has already acquired a reputation for telling the truth. Finally, a by-product of our empirical analysis is that non-standard econometric techniques, such as fuzzy regression, do not perform better than standard OLS and ordinal regressions. The rest of the paper is organized as follows. In the next section, we explicitly measure ECB communication. In other words, we construct a glossary that translates the qualitative information of the ECB President's speeches during press conferences into an ordered scale. In Section 3, we investigate the consistency of ECB communication (i.e., between its words and deeds). In Section 4, we further analyse the informational content of the wording of the ECB. First, we compare the forecasting ability between ECB words and Euribor rates (i.e., interest rates in force in the interbank market). Second, we determine whether words are complements or substitutes to macroeconomic variables in empirical reaction functions. Then in Section 5, we investigate the significance of news surprises after controlling for monetary policy shocks. In Section 6, we suggest some important issues left for future research and conclude. The Appendix provides both the description and the sources of the time series used in the estimations.

In this paper, we have shown that in recent times central bankers are not boring at all (cf. Lambert, 2004): since the ECB has so far been consistent between its words and future deeds, its communication conveys useful information about the short-run dynamics of the Repo rate. In fact, by looking at the verbatim content of the ECB President's monthly press conferences, it is possible to predict the European monetary authority's interest rate setting behaviour fairly well. Furthermore, we show that the informational content of the wording of the ECB is substantial. On the one hand, the forecasting ability of ECB words is similar to the one implied by the Euribor term structure. On the other, the information of ECB rhetoric does not disappear once we control for Taylor-type macroeconomic variables. Finally, we verify empirically that financial markets not only understand but also believe the signals sent by the European monetary authority. Therefore, we conclude that the ECB is effective in its job of communication to the public, and is able to influence market expectations on the short-term interest rate path using just words. To put it another way, not only can the ECB reinforce the effects of a policy action by adding a statement in the same direction (Kohn and Sack, 2003), but it can also affect the short-end of the money market term structure just by the words contained in the President's press conference. Of course, some important issues are not considered in this paper and deserve further study. In Section 4.1 we have made the implicit assumption that the risk premium on Euribor contracts (proxied by − α) is constant over time. Piazzesi and Swanson (2004) show that for the period 1988–2003 (sixteen years) excess returns on U.S. federal funds futures have been time-varying and, more specifically, strongly countercyclical. Moreover, they found that excess returns could be well predicted by macroeconomic indicators. Therefore, they conclude that futures-based measures of the future monetary policy path should be adjusted to account for these variable excess returns. Is this also true for the European case? We have tentatively tried to replicate their exercise by regressing (OLS with White standard errors) the risk premium for various maturities (defined as R¯t + k,t + k + 1 - rt,t + k,1 in our notation) on a constant and a vector of variables known to financial markets at time t (regression results not reported). The constant turns out to be always positive and significantly different from zero, while the regressor coefficients on employment growth and financial business cycle indicators such as corporate bond spreads were systematically insignificantly different from zero. Perhaps, our sample of six years was too short to investigate this issue or we have not identified the right set of explanatory variables. The validity of the horse race proposed at the end of Section 4.1 crucially depends on the correct measurement of ECB opinions. We have already noted that the interpretation of verbal documents cannot but be imprecise and subjective, in spite of all our efforts, and these ambiguities will never be fully cancelled out. The solution suggested in this paper about turning verbal expressions into an ordinal variable is just one possibility. Others may include either the use of textual analysis software based on the relative frequency of words or running an experiment. Finally, and more generally, since in this paper we have shown empirically that central bank unexpected communication is very effective (at least as effective as monetary policy deeds) in moving financial markets, in our opinion it would be worthwhile to investigate the economic consequences of monetary authority announcements in a New-Keynesian framework of the monetary policy transmission mechanism (cf. Woodford, 2003). 23 Woodford (2005) represents a first step in this direction. Unfortunately, his analysis deals only with a very special case, i.e. the role of central bank communication when the policy rate instrument cannot be used because it has hit the zero lower bound (i.e., Japan liquidity trap situation).