A layman's guide to ECB watching: ten simple rules for analyzing the European Central Bank.

"ECB watchers" have had a much less successful record at forecasting the immediate outcome of ECB meetings than "Fed watchers" do on the FOMC. Using data from regular Bloomberg surveys of analysts, taken just a few days in advance of policy meetings, we found that ECB watchers correctly forecast ECB rate changes just 36 percent of the time, compared with an accuracy rate of 86 percent for "Fed-watchers" and 64 percent for UK-based "MPC-watchers." Undoubtedly, part of this poor performance by "ECB watchers" has resulted from the much higher frequency of ECB policy meetings, which during most of its first three years were running at a rate of twenty-four or twenty-five a year, compared with twelve for the Bank of England's MPC and just eight for the Fed's FOMC. In this sense, the ECB's decision last November to consider monetary policy actions at only its first meeting of the month is very welcome and should make future policy actions considerably more predictable. Meanwhile, here are ten "rules of thumb" to help investors better understand the ECB.

1. THE ECB'S MANDATE DIFFERS FROM THAT OF THE FED.

Much of the explanation for the ECB's different behavior to that of the Fed appears to result from a different mandate. The ECB's primary objective is "to maintain price stability"; only without prejudice to this may it then contemplate its secondary objective, "to support the general economic policies of the Community." In contrast, the Fed's objective is "to promote effectively the goals of maximum employment, stable prices and moderate long-term interest rates." The ECB's interprets price stability as harmonized CPI inflation over the medium term of "below 2 percent"; more recently officials have begun to imply that this represents a 1-2 percent range. Recall also that the ECB is a new institution, responsible for an economic region larger by population and geography, with a more inflation-prone history than the US, as well as being more diverse economically, politically, and culturally, than is the Fed. As such the ECB is still highly preoccupied with reputation building--in particular, to emphasize to economic agents in the euro area that it will not tolerate inflation above 2 percent in the medium term.

2. A DIFFERENT NOTION OF PRIMUS INTER PARES ("FIRST AMONG EQUALS")

The phrase, primus inter pares ("first among equals") might also be used to describe the authority that the Chairman of the Federal Open Markets Committee (FOMC) exerts over his fellow members. In practice, what the Fed Chairman wants, the Fed Chairman appears to get. In contrast, this may not always be the case for the President of the ECB, who strives for consensus amongst the eighteen-person Governing Council (GC). This collegiate approach reflects several factors--the newness of the institution; cultural inclination; the personalities involved. The formation of the ECB has involved a tremendous concentration of power--from twelve national central banks to effectively one organization. Many senior central-banking officials, especially governors of national central banks, had been accustomed to a highly visible public profile, and inevitably still want their voices heard in public and in private.

Another difference with the Fed is that the ECB's economics and research directorate works in principle for the GC as a whole, but its research tends to be aggregated by Chief Economist Issing, who is said to open the meeting with an economic exposition that concludes with a recommendation on interest rates. In contrast, Fed Chairman Greenspan has de facto become the Fed's "Chief Economist" and therefore has exerted a degree of influence over the research conducted by the Fed's staff and its presentation to the FOMC. That said, the ECB President does usually represent the ECB before the European Parliament and takes responsibility for monetary policy at the ECB press conferences, and therefore can still exert a high degree of suasion over the committee.

3. THE "STEADY HAND" TRADITION CONTINUES.

The ECB's perception of the effectiveness of monetary policy appears to differ from that of the Fed. It would argue that above all monetary policy should be "stability-oriented"--a legacy of Bundesbank thinking. It regards Europe's growth prospects as best achieved through achieving a high rate of investment, which in turn requires low and stable long-term rates. It sees these as particularly important for Europe, given a higher dependency of corporate and household financing using long-term rates. The ECB focuses less than the Fed on stock market movements, partly because it does not believe that equities play much role in affecting the euro area economy. It does, however, pay close attention to (and takes considerable pride in) inflation expectations derived from index-linked bonds. As well, the volatility of the euro and its potential impact on the formation of inflation has meant that the currency's external value has assumed proportionately more importance in ECB considerations and actions than it has for the Fed (which has traditionally paid scant attention to the dollar's value).

The ECB appears to have a stronger preference than the Fed to minimize volatility in its official policy rate (the "repo rate"). It therefore has a very "gradualist" approach that, it believes, helps to foster lower long-term rates (by lowering the risk premium). In general, it would prefer not to reverse interest rate decisions within a considerable time-span--indeed, some officials would doubtless have preferred, with the benefit of hindsight, not to have lowered rates 50bp in April 1999 only to reverse this decision seven months later. Overall, the ECB eschews "fine tuning"--it does not believe that it has sufficient historical and forward-looking information about the business cycle to be able to deliver a counter-cyclical policy, especially given its belief that there are considerable lags in the interest rate transmission process--of "a number of quarters or even years." That said, it is not unresponsive to the economic cycle--for example, it lowered interest rates 150bp last year, 100bp of which came in two 50bp installments, on September 17 and November 8.

4. IF YOU THINK THE ECB WILL LOWER RATES, THINK AGAIN.

On account of its collegiate nature, the Governing Council therefore seeks a high level of consensus prior to changing policy. In an interview President Duisenberg reportedly commented last year that he would rather that the ECB deferred a contentious decision to change interest rates to such a time when the decision commanded more widespread support on the GC. This is not to say that all ECB decisions are unanimous--although the ECB's minutes are not published, there is evidence to suggest that it will take decisions even if agreement is not reached with all eighteen GC members. But there has undeniably been a tendency for the ECB to lag, rather than lead, money market expectations, both on the upside and downside. This arises partly from the collegiate nature of decision taking and partly from the ECB's self evidently gradualist approach.

5. THE SIZE OF RATE CHANGES DEPENDS ON THE TASK IN HAND.

During the first year of the euro (1999) the ECB altered interest rates just twice--it lowered rates 50bp to 2.5 percent in April and raised them back to 3 percent in November. This led to the view at the time that the ECB would prefer to change rates infrequently, entailing perhaps some inclination for 50bp rather than 25bp. Since the start of 2000, however, there have been only three rate changes of 50bp--in June 2000 and in September and November 2001. The remaining seven rate changes were of 25bp only. It would appear that the magnitude of rate changes does depend on what particular circumstances, and interpretation thereof, were behind the action. For example, the June 15, 2000 rate increase of 50bp was justified at the time as resulting from the need to signal a clear horizon since it coincided with a major change to the procedure in the ECB's money market operations. The implication then was that were the ECB not to have switched to a variable rate repo, it would have perhaps been more inclined to move by 25bp.

The September 17, 2001 50bp rate cut resulted from exceptional economic and market uncertainty--and a desire to act in concert with the Fed. One can argue that many of the rate changes during 2000 resulted from concerns about the depreciating euro, and that the rate cuts in May and August 2001 were of a more tentative nature that could have reflected compromises (given that they were made at a time when inflation was closer to 3 percent than being below 2 percent). Meanwhile, the 50bp rate cut in November 2001 was apparently motivated by concern about a sharp deterioration in economic prospects. Overall, there is perhaps a slight predisposition in favor of altering rates in 25bp increments, especially if the ECB wishes to influence the external value of the euro (which it regards as an important transmission route to euro area inflation). The 25bp incremental moves are more likely to reflect compromises borne out of the consensus-driven nature of decision taking. That said, the ECB has demonstrated that if the time is considered right, it will move in 50bp increments--that is, if business confidence is shifting rapidly enough.

6. ECB ACTION IS IN INVERSE PROPORTION TO EXTERNAL JAWBONING.

The ECB jealously protects its high level of independence, and a consequence of this is that the more vocal the pressure from politicians on the ECB to act, the less likely the ECB is to do so. Perhaps the best example of this was in March 1999, when German Finance Minister Lafontaine publicly urged the ECB to cut rates, and attended its early March meeting as chairman of the euro group of finance ministers. The ECB desisted, but lowered rates 50bp at their April 8 meeting, which was shortly after his resignation. This was a crucial defining period whereby the ECB demonstrated that it was in charge of monetary policy.

Last year also, heavy vocal pressure during late March and early April from some Finance Ministers, especially from the then "euro group" chairman, Reynders, seems to help explain why the ECB did not lower rates on April 11 (as many in the markets had expected). But at the EU Finance Ministers meeting (Ecofin) in early May, Mr. Reynders resisted media temptation to call again for lower rates. The ECB duly obliged with a 25bp rate cut on May 10.

One should not assume, however, that politicians exert no influence on the ECB. There is dialogue, both informally and at formal meetings, and the President of Ecofin has the right to attend and speak at ECB Governing Council meetings (though not to vote, and attendance has not been especially frequent). As well, governments are responsible for appointing members of the Governing Council when their terms expire, and in this way could indirectly influence policy. Moreover, the scope of politicians to discredit ECB officials and/or the institution could be considerable were the ECB to fail to meet its objectives. Politicians and the ECB realize this, which is why ultimately the ECB does require a mutually satisfactory working agreement with politicians.

7. THE ECB DOES NOT ORDINARILY WISH TO SURPRISE THE MARKETS.

The ECB takes comfort from independent assessments suggesting that, on the whole, it has conducted monetary policy in an appropriate way, but has acknowledged that it has not always well communicated its process. In particular, it appears in retrospect to acknowledge that its 25bp rate cut on May 10, 2001 surprised the markets. Shortly after that decision, Chief Economist Issing commented that its policy was not ordinarily to surprise the markets: "... the evidence from the financial markets indicates that ECB monetary policy has always been credible and also--for the most part--has been well anticipated by the markets. There is also some evidence that volatility in financial markets at the time of interest rate decisions has been declining over time. The ECB has, on occasions, been forced to surprise markets with policy decisions. However `surprise' is never the goal or deliberate intention of monetary policy."

Note that this is a different attitude from that often thought to have been held by the Bundesbank, which almost seemed to delight in surprising the markets. Indeed, the outcome of meetings since May 2001--with the exception of the emergency meeting on September 17--have been a little more discernible from previous comments by ECB officials. Moreover, an analysis of the change in the Euribor three-month futures contract on the days of rate changes suggests that the ECB suggests that the ECB has surprised the market on a three-month horizon with only three of its twelve rate changes (see Figure 2). But surprises do occasionally happen, such as the 50bp rate cut on September 17th, a decision that resulted from exceptional economic and financial circumstances and so need not be considered inconsistent with the doctrine of not ordinarily surprising the markets.

However, another way of seeking to establish whether the ECB does surprise the markets is to look at polls of "ECB-watchers" undertaken immediately ahead of ECB meetings and to compare the results with polls of "Fed-watchers." The difference is striking: of the eight previous occasions when the ECB has changed interest rates at a pre-scheduled GC meeting, the accuracy of forecasts conducted a few days previously was just 36 percent (1). In contrast, of the past ten Fed changes at pre-scheduled FOMC meetings, the forecast accuracy of "Fed watchers" was 86 percent, while for "MPC-watchers" it was 64 percent (2). If one considers forecasting accuracy ahead of all pre-scheduled meetings, then the accuracy of"ECB-watchers" increases sharply (as the "safe" option is to forecast inaction, given so many meetings), to 77 percent, coincidentally the same as for "MPC-watchers," whereas that for "Fed-watchers" improves to 91 percent.

Probably the most important explanation for this gulf is simply that until last November, the ECB was considering interest rates on an almost fortnightly schedule, that is, three times as frequently as the FOMC, and compared with the MPC's monthly schedule. This meant that "ECB watchers" had much less chance of forecasting the exact occasion of a rate change. For an institution that eschews "fine-tuning" the ECB has met very frequently to consider interest rates. Under the Maastricht Treaty, the ECB Governing Council is required to meet at least ten times a year, but in practice, until last November, was able to consider altering interest rates virtually every two weeks. This high frequency of meetings was driven partly by the requirements to consider many technical issues surrounding the ECB's first years of operation, but it meant that the ECB has been subject to far greater and more continuous speculation about impending policy changes than the Fed of Bank of England.

Last November's decision to consider monetary policy actions "normally" on only the first meeting of the month is therefore welcome, and already appears to have reduced the tide of speculation that used to precede many of the GC's fortnightly meetings. Lucas Papademos, Governor of the Bank of Greece, has commented, "the high frequency of meetings in the past probably caused increased expectations for changes in interest rates." As well, the reduced frequency of monetary policy meetings to only twelve a year could help the GC to become more forward-looking.

Nonetheless, there are some other reasons that come to mind to explain the gap between the accuracy of "Fed" and "ECB" watchers:

(i) Fed policy has tended to be dominated by its Chairman in recent years, who in turn has considerable experience at communicating with markets, the government and the media, whereas the ECB's more collegiate structure has led to occasional "policy dissonance," such as appeared apparent during the spring of 2001.

(iii) "Fed-watchers" are more experienced -- they have had more time and information to acquire an accurate template for Fed behaviour.

(iv) "Fed-watchers" observe a more homogenous economy and financial system, with a more readily identifiable fiscal policy, and with much more comprehensive and timely statistical information.

(v) Fed action may be conditioned by market expectations, perhaps because the Fed may have a greater aversion towards precipitating market volatility than the ECB. Related to this, some observers believe that Fed officials engage in greater discussion with market participants.

(vi) The surveys we have used above are of the FOMC primary dealers, whereas no such group exists within the euro area, so the survey of "ECB watchers" focuses upon a less clearly defined group of major financial institutions, which could increase the potential for "rogue" results.

(vii) Our sample excludes un-scheduled rate changes (i.e., inter-meeting moves). There was much less chance of these happening at the ECB because of the three-times higher meeting frequency. However, if one were to include such results, then it would lower only slightly the forecasting accuracy of Fed watchers, who would still have a significant lead over the performance of ECB watchers.

8. THE MONTHLY BULLETIN IS THE PUREST STATEMENT OF ECB THINKING.

In contrast to the monthly bulletin of the Fed, the ECB's Monthly Bulletin is the single most important statement of the ECB's thoughts. It strives, particularly in its Editorial, to truly represent the collective thoughts of the ECB. The Opening Statements at the ECB's press conferences also merit close attention, but the Bulletin is more extensive and sometimes gives a framework for evaluating the future directional risks for monetary policy. Since the ECB does not believe in expressing biases on monetary policy, one needs to read between the lines. For example, in its August 2001 Bulletin, the ECB stopped talking about expected growth being in line with potential. This was a material omission that signaled that the ECB had altered its growth assessment, which in turn helps to account for its 25bp rate cut on August 30.

The ECB does not publish minutes of its Governing Council meetings primarily because it does not wish to personalize the internal debates it conducts. This seems fair enough--a public airing of the inevitable disagreements that will arise internally would undermine collective responsibility and lead to a less productive discussion, in at least two respects. First, a public reporting of their discussion might make central bank officials--who anyhow are a watchword for caution--excessively conservative in expressing ideas at Council sessions, thereby potentially limiting free ranging and creative debate. Second, ECB officials are proud to observe that the ECB's decisions are based on what policy is believed to be appropriate for the euro area as a whole--to the extent that discussion of specific regional developments is discouraged at Governing Council meetings. In mm, this means that the twelve national central bank governors who sit on the Council meetings act as representatives of the euro area, not as national delegates. ECB officials observe that if minutes were published then Council members would risk coming under heavy pressure from local politicians and media to vote in accordance with national interests, which would run counter to the European spirit of decision making.

9. M3 IS USED FOR EX POST RATIONALISATION.

The ECB's "twin pillar" monetary strategy has perhaps aroused the most criticism from external observers. Basically, the first pillar is "analysis according a prominent role to money" while the second is "analysis focusing on a wide range of other economic and financial indicators." The ECB believes that "monetary developments contain information about future price developments and can therefore help in the overall assessment of risks to price stability." Increasingly it points to a focus on monetary and credit aggregates as a way to avoid asset price instability. Nonetheless, the ECB is always careful to observe that, "Monetary policy does not respond in a mechanical way to deviations of M3 growth from the reference value," since M3 velocity may not be stable in the short run.

In practice, the ECB use the information contained in the monetary and credit aggregates for ex post rationalization of interest rate decisions. M3 growth is routinely the first aggregate mentioned in the Bulletin's Editorial and so the ECB has deliberately put itself in a position where interest rate decisions need to be justified first with reference to M3 as well as other variables.

On account of various distortions that have prevailed in the monetary aggregates since EMU's inception, the ECB has persistently sought to account for policy actions that might be inconsistent with M3 growth. Hence, for example, it lowered rates in April 1999 despite M3 growth being above the 4.5 percent reference rate, by commenting that there were "special factors at the start" of EMU. It lowered rates in August (by 25bp) and in September and November last year (each by 50bp) despite reported M3 growth being above the reference rate, by citing a distortion in M3 measurement that was boosting the aggregate by 0.75pp, as well as, increasingly, shifts from equity funds into money markets.

Overall, the ECB has shown itself willing to adopt a very pragmatic attitude towards the M3 reference rate, provided that it can show there are special factors accounting for distortion in M3 growth, and when other economic and financial variables under its "second pillar" argue in favor of easing. It will be interesting to see if the prominence accorded to the money and credit aggregates remains so strong as the Governing Council's composition and size changes in coming years.

10. THE ECB PRESIDENT IS DE FACTO "MR. EURO."

Despite responsibility for foreign exchange policy in the Maastricht Treaty being a somewhat grey area, in practice this ECB has claimed for itself, in the absence of guidance from the Finance Ministers, the prerogative over the external value of the euro. Under the Treaty, the Finance Ministers have the authority to issue a "general orientation" to the ECB concerning the external value of the euro, but this must not prejudice the ECB's primary objective of price stability (3). As such, the ECB could legitimately oppose such an orientation if it considered that doing so would compromise its price stability objective. As the ECB is charged with interpreting as well as maintaining this price stability objective, and is granted formidable independence, in practice it would appear implicitly to support the ECB as the final arbiter of what constituted a risk to price stability, in the event of a dispute with the Ecofin concerning exchange rate policy.

In practice, the Finance Ministers have so far refrained from giving the ECB a general orientation. Doing so would probably entail a major wrangle--the ministers would be required first to consult with the ECB, and if there were such a disagreement that required an orientation, then the ECB could dig in its heels and claim that to obeying it would prejudice its primary objective. By opening up this particular hornets' nest, the ministers might achieve a weaker euro, but they could also find a higher risk premium in the form of higher interest rates on euro assets, thereby negating any benefit from a weaker currency. Hence it would appear hard to dispute the claim by the ECB President that he is, de facto, "Mr. Euro."

That said, being "Mr. Euro" de facto is not the same as de jure. Some officials suggest that the finance ministers are not happy at the ECB President assuming this moniker, and that the euro group might at some stage try to reclaim some influence over the euro. This issue, ambiguous in the Treaty, has not yet gone away.

Figure 1

Key statistics on policy actions by the ECB, FOMC and BoE
MPC, January 1999-November 2001

                                             ECB   Fed   MPC

Total number of rate changes                  12    16    15
Average absolute size of rate change (bp)     35    39    28
Historic probability of rate change at
  pre-scheduled meeting (per cent)           17%   70%   43%
Number of quarter point rate change            7     7    13
Number of half point rate changes              5     9     2
Historic probability that a rate change is
  25bp rather than 50bp (per cent)           58%   44%   87%

Source: CSFB, MPC, Federal Reserve
Figure 2

ECB rate changes since the start of EMU

                                                          M3 growth
                                                          3mma (as
                                                          published
                  Repo       Press          Market         before
Date of change   action   conference?   anticipation? *   meeting)

08-Apr-99        -50bp         Y              yes         5.1%
04-Nov-99        +50bp         Y              yes         5.9%
03-Feb-00        +25bp         Y              yes         5.7%
16-Mar-00        +25bp         N              yes         5.8%
27-Apr-00        +25bp         N              yes         6.0%
08-Jun-00        +50bp         Y              no          6.3%
31-Aug-00        +25bp         N              yes         5.5%
05-Oct-00        +25bp         Y              yes         5.3%
10-May-01        -25bp         Y              no          4.6% **
30-Aug-01        -25bp         Y              yes         5.9% ***
17-Sep-01        -50bp         N              no          6.4% ***
08-Nov-01        -50bp         Y              yes         6.9% ***

* The change is defined as anticipated if the Euribor 3m contract
moved by 5bp or less on the day of the change

** Reported to have a distortion adding around 0.5pp to growth

*** Reported to have a distortion adding around 0.75pp to growth

Source: CSFB estimates, Bloomberg, ECB
Figure 3

Analysis of economists' expectations of rate changes
ahead of impending policy meetings

Data from the past two years             ECB   Fed   MPC

Overall accuracy rate, for predictions
  ahead of all meetings                  77%   91%   77%
Accuracy rate for predictions ahead of
  meetings where policy was changed      36%   86%   64%

Source: CSFB, Bloomberg

(1) Accuracy is measured as the proportion of analysts who correctly forecast the rate outcome of the upcoming meeting.

(2) The surveys of central bank "watchers" were taken from polls by Bloomberg of leading financial market institutions' expectations of policy action at the next policy meeting, usually conducted up to one week ahead of pre-scheduled FOMC/GC meetings.

(3) Technically the Treaty states that "the [EU] Council, acting by qualified majority either on a recommendation from the Commission, and after consulting the ECB or on a recommendation from the ECB, may formulate general orientations for exchange rate policy ... These general orientations shall be without prejudice to the primary objective of the ESCB to maintain price stability." Elsewhere the Treaty assigns to the ECB the power "to conduct foreign exchange operations" and to "hold and manage official foreign reserves of the Member States" as well as "to define and implement the monetary policy of the Community".

Julian Callow is Chief Continental Economist with Credit Suisse First Boston (Europe) LTD.