1
The Convergence Criteria – How Tight a Constraint under
Inflation Targeting?
Jan Frait – The Czech National Bank
*
Paper for the 10
th
Dubrovnik Economic Conference
This version: June 2004
ABSTRACT
The paper focuses on how tight a constraint are the convergence criteria for the conduct of
inflation targeting. This particular question is relevant to those EU member countries staying
outside the euro area which are currently operating under flexible exchange rate
arrangements. For these countries, it will be difficult to avoid a double shift in monetary
policy. To meet the convergence criteria, they will be pressured to switch to exchange rate
targeting and adopt a regime similar to inflation targeting after joining the euro area again.
For those that would like to avoid the double shift, the conditions for using inflation targeting
as an independent anti-cyclical monetary policy after joining ERM2 are defined. The
approach applied is rather informal and builds on potential financial market reactions to the
existing policy constraints. The experience of the current euro area members during the
ERM2 period is analysed and the implications of the term structure of interest rates and the
uncovered interest rate parity condition are discussed.
JEL
: E31, F31, O11, P17
Keywords
: ERM2, convergence criteria, uncovered interest rate parity, inflation targeting,
exchange rate regime, Balassa–Samuelson effect
*
Jan Frait – Member of the Board and Chief Executive Director of the Czech National Bank (e-mail:
jan.frait@cnb.cz). I am grateful to Juraj Antal, Viktor Kotlán and Luboš Komárek for extremely helpful
contributions and to Jacek Rostowski and Velimir Šonje for his comments. The opinions expressed in the paper
are those of the author and, unless explicitly stipulated, do not necessarily represent the official views of the
Czech National Bank.
2
Introduction
The discussion of the ten new Member States (NMSs) on euro adoption and ERM2
participation has recently shifted from the shaping of positions to questions of a more
practical nature. Among those, suitable policies for meeting the convergence criteria before
adopting the euro play a prominent role. Countries are searching for ways of making the
transition to ERM2 and later the euro as smooth as possible. This is a question relevant not
just to the new EU entrants, but to all the non-euro EU members, which either are expected to
enter the euro area (Sweden) or might want to (UK, Denmark).
Among them, there are five inflation targeters (UK, Sweden, the Czech Republic, Poland,
Hungary) plus two non- inflation targeting floaters (Slovakia, Slovenia), which all may face
the same question. Namely, how tight a constraint is the exchange rate criterion (and its
combinations with other criteria) for the conduct of inflation targeting (IT) or a policy similar
to IT? This is the main question I try to answer in this paper. It is natural that the question is
relevant to only some of the EU member countries now staying outside the euro area. These
are the countries currently operating under flexible exchange rate arrangements. On the
contrary, the countries operating currency boards will face rather different challenges.
The issue has at least two levels. First, one can think about it in terms of the right policy
regime to use in the pre-euro phase. Second, one can understand the question as concerning
the specific way monetary policy is conducted in this period, irrespective of the declared
regime. The discussion on the regime design question may seem relatively simple. One of the
most important and easily observable characteristics of inflation targeting as a policy regime
is that policy is formulated as an explicit numerical target for inflation. And this is what the
inflation convergence criterion represents – a numerical inflation target. This is true
notwithstanding the fact that it is a moving target – depending on how low the inflation rate in
the three EU countries with the lowest inflation is. Thus, even countries that have not applied
IT in the past will, at least from the outside view, operate some form of IT prior to euro area
entry. There is another reason why having an explicit goal for inflation may be the right
strategy prior to euro area entry. The ECB has been practising a policy based on a (more or
less) specific numerical target of “below but close” to 2 percent inflation. Thus, monetary
policy communication with the public formed around the logic of attaining a specific rate of
inflation may manifest important continuity if IT is practised before euro area entry. This is,
of course, even more relevant to the countries that have already practised this strategy for
some time in the past.
If regime choice is not the issue, the question on constraint then really boils down to a
question on the mutual consistency of the inflation, exchange rate and, potentially, other
criteria. This can be a tricky issue since it depends not only on how one understands the
definition of the criteria, but also on several country-specific factors. Methodologically, there
are several ways of examining the consistency between the inflation and exchange rate
convergence criteria. Égert and Kierzenkowski (2003), Szapary (2001) and Buiter and Grafe
(2002) use a rather informal way of presenting the topic. Grauwe and Schnabl (2004) employ
a more formal partial equilibrium model. Mihaljek (2002) and Flek, Marková and Podpiera
(2002) perform empirical analysis, while Natalucci and Ravenna (2003) investigate the issue
by means of a dynamic stochastic general equilibrium model.
The formal approaches have significant drawbacks. Econometric treatment is useful, but
as the transfer to ERM2 will represent a regime change for most of the countries concerned,
the results from the old situation cannot easily be transferred to the new one, since the Lucas
critique applies
1
. Moreover, there is the usual lack-of-data problem with the NMSs. The
1
Apart from the Lucas critique, Natalucci and Ravenna (2003) disregard the impact of nominal exchange rate
appreciation on firms’ profits, which may transform to investment and, later, potential GDP growth. This
3
general equilibrium approach covers the full scope of the interrelationships and is thus
particularly well-suited to exploring the interactions among the criteria. However, unless fully
micro-based, it may also be subject to the Lucas critique typical of the econometric models. In
this respect, partial equilibrium models and informal description may serve the purpose well,
too. They can accommodate appropriate and relevant assumptions, constraints and scenarios,
although they are not well-suited to exploring the general equilibrium consequences.
As indicated above, the goal of this paper is to provide an answer to the question posed in
the title, namely, “How tight a constraint is the exchange rate criterion for the conduct of
IT?”. In addition, I define conditions for using IT as an independent anti-cyclical monetary
policy at least to some extent even after joining ERM2. The approach applied is informal
because, among other reasons, it is rather difficult to apply formal economics to the process
that the NMSs are burdened with, given that this process has been designed without having
economics in mind. As politics was a major force behind the design, it is more appropriate to
analyse potential financial market reactions to the policy constraints created. I will do this by
analysing the experience of the current euro area members during the ERM2 period and
evaluating the constraints stemming from the term structure of interest rates and the
uncovered interest rate parity (UIP) condition. The use of UIP instead of formal analysis is
justified by the very specific situation of the new entrants. Once they decide to participate in
ERM2, they will be exposed to an environment in which the risk premium is approaching
zero, exchange rate expectations are influenced by the explicit central parity and interest rates
are converging to the euro area levels.
The remaining discussion is structured as follows. In Section 1, the debate on the
definition and extent of the exchange rate constraint is reviewed. In Section 2, the current
state of nominal convergence is described and the Czech Republic’s position on euro adoption
and ERM2 participation is presented. In Section 3, the tightness of the constraints is assessed,
and the subsequent Section 4 deals with monetary policy autonomy in ERM2. Section 5 then
focuses on potential strategies for entering the ERM2 and the implications of financial market
mechanisms. Section 6 then concludes with a debate on how to pursue IT in ERM2. The final
section concludes with several policy reflections.
1. The Constraints and their Relevance
1.1 The Exchange Rate Convergence Criterion
The exchange rate convergence criterion is defined in the Treaty on European Union (the
Treaty). as follows:
“The criterion on participation in the Exchange Rate Mechanism of the
European Monetary System … shall mean that a Member State has respected the normal
fluctuation margins provided for by the Exchange Rate Mechanism of the European Monetary
System without severe tensions for at least the last two years before the examination. In
particular, the Member State shall not have devalued its currency’s bilateral central rate
against any other Member State’s currency on its own initiative for the same period”.
The assessment of fulfilment of the criterion is based on the relevant provisions of the
Treaty. The European Commission (EC) expressed its standpoint on the fulfilment of the
exchange rate criterion in ERM2 in its 2000 Convergence Report as follows: 1) Participation
in the ERM2 at the time of assessment is mandatory and expected for at least two years. Some
exchange rate stability during a period of non-participation before entering ERM2 can be
taken into account, too. 2) No downward realignment (devaluation) of the central parity
within the two year examination period. 3) Exchange rate to have been maintained within a
omission may be one of the reasons why they conclude that “allowing for a sustained appreciation of the
nominal exchange rate would deliver a lower volatility of both the output gap and inflation”.
4
fluctuation band of ±2.25% around the currency’s central parity against the euro. An
assessment of any deviation from the ±2.25% fluctuation band would have to take account of
the reasons for that deviation. A distinction is to be made between exchange rate movements
above the ±2.25% upper margin and movements below the ±2.25% lower margin.
Article 121 of the Treaty stipulates that both the EC and the European Central Bank
(ECB) are to examine the state of convergence of the Member States. The Convergence
Reports are then to be submitted to the Council of the EU, which, based on the
recommendation of the EC, judges whether a given country fulfils the necessary conditions
for the adoption of the single currency
2
.
As Égert and Kierzenkowski (2003) notice, the assessment practice of the EC in general
has been different from – and much more explicit than – that of the ECB (EMI). While the EC
often refers to the ±2.25% band, the ECB refrains from mentioning it. In any case, the EC
emphasises that a deviation exceeding ±2.25% will not automatically imply a violation of the
criterion on exchange rate stability. Whether or not a larger deviation corresponds to a
violation of the normal fluctuation margins or to severe tension hinges mainly on how long
the deviation lasts, how large it is, and, most importantly, whether it occurs on the weaker or
the stronger side of the fluctuation band. Also, the ECB has always understood that wider
bands should have been taken into account in a way.
The ECB in its latest policy position (December 2003) points out that the assessment of
exchange rate stability against the euro will focus on the exchange rate being close to the
central rate while also taking into account factors that may have led to an appreciation, which
is in line with what was done in the past. The ECB further stresses that the width of the
fluctuation band within ERM2 shall not prejudice the assessment of the exchange rate
stability criterion. Moreover, the issue of absence of “severe tensions” is, according to the
ECB, addressed by examining the degree of deviation of exchange rates from the ERM2
central rates against the euro, by using indicators such as short-term interest rate differentials
vis-à-vis the euro area and their evolution, and by considering the role played by foreign
exchange interventions.
All this means that the exchange rate criterion should basically be understood as 2.25% on
the weaker side and 15% on the stronger side. In addition, going beyond the 2.25% limit on
the weaker side does not automatically mean a violation of the criterion, and at the same time,
the possibility of revaluation of the central parity questions the existence of any limit on the
stronger side. It is thus possible to conclude that to meet the criterion, it is necessary to avoid
devaluation of the central parity and to ensure that the exchange rate is not frequently well
beyond the 2.25% limit on the weaker side despite interventions via interest rate hikes and
exchange reserves sales.
There is thus some room for manoeuvre concerning the mix of interest and exchange rate
policy; uncertainty remains as to the size of it. What is open is how the EC would assess
occasional breaches of the 2.25% limit that are not of a fundamental nature and that are not
accompanied by any interventions. In other words, the crucial question is how to assess the
underlying tensions and tell fundamental breaches from non-fundamental ones. We know that
the “severe tensions” conditions will be assessed by taking into account indicators like the
side, timing, size and duration of the deviations during the assessment period. In addition,
factors like exchange rate volatility, short-term interest rate movements or size of foreign
exchange interventions will be taken into account. Such a framework offers some scope for
desirable exchange rate flexibility
3
.
2
There is a substantial difference between the EC’s and the ECB’s responsibility. Although both the EC and the
ECB are required to prepare a convergence report when a given country is analysed to see whether it complies
fully with the convergence criteria, it is the EC that makes a direct recommendation to the European Council.
3
The Czech experience with 2001–2002 appreciation pressures provides evidence that properly addressing
5
1.2 ERM2 as a Constraint
There was a very lively debate on the role and sense of ERM2 between the Eurosystem
and the acceding countries. The Eurosystem position regarding ERM2 is set out in the “Policy
position of the Governing Council of the ECB on exchange rate issues relating to the acceding
countries” (ECB, 2003). This document builds on the Position Paper “The Eurosystem and the
Accession Process” endorsed by the Governing Council on 21 November 2002. The Position
Paper puts forward the view that “
ERM2 offers a meaningful framework for combining
nominal and real convergence and should therefore not be seen as a mere “ante-chamber”
before the adoption of the euro … ERM2 should be seen as a useful regime on its own right,
as a number of policy challenges can be tackled within that framework in the run-up to the
adoption of the euro … ERM2 will provide for a certain degree of exchange rate stability and
macroeconomic policy discipline, while still leaving room for adjustments to shocks and
market developments. Moreover, its multilateral nature – with both the accession country’s
central bank and the ECB being involved – will enhance its credibility. For these reasons,
ERM2 is likely to be beneficial for the accession countries in their pursuit of real and nominal
convergence.
”
One of the most striking features of this particular argumentation is the apparent lack of
emphasis on the target zones literature that was so popular in the 1990s (Krugman, 1991,
Bertola and Caballero, 1992 or Svensson, 1994). It is not easy to find a reference to the
smooth pasting, honeymoon effects and other desirable features that were discussed during
the glorious days of ERM. Financial crises and the subsequent literature on multiple
equilibria, self-fulfilling speculative attacks and reversals of capital flows (e.g. Eichengreen
and Wyplosz, 1993) made us more aware of the complicated dynamics of modern financial
markets. This paved the way for an understanding of the benefits of corner solutions (e.g.
Fischer, 2001) in the form of either hard pegs or more flexible regimes. That is why most of
the NMSs want to minimise the length of ERM2 participation now that they know it will not
be possible to avoid it entirely. Unfortunately, the proponents of European monetary
integration in the EU institutions still pretend as if nothing has happened and as if we are not
in 2004 but in 1994. That is where we stand now. Clearly, at the beginning of the 1990s IT
could hardly have been viewed as a potential substitute for ERM2, owing to the lack of
experience with it. Today, thanks to the experience with intermediate regimes, one can hardly
think of ERM2 as a reasonable substitute for IT.
The position of the NMSs’ central banks towards ERM2 differs from that of the
Eurosystem and the EC. They generally perceive ERM2 as a mere “waiting room”. Some
even argue that it is at best of zero or negative value-added and may even entail significant
risks. Of course, for some small NMSs, ERM2 may not represent a
de facto
exchange rate
regime shift. ERM2 participation may deliver negative effects for countries with a free or
lightly managed float and a credible IT framework in place.
Personally, I cannot see any value added in using the ERM2 regime. The Czech Republic
successfully stabilises inflation by means of IT, which is now a flexible and pragmatic policy
based on occasionally managed floating and a tailor-made communication strategy. ERM2
cannot be superior to it. In addition, ERM2 can hardly constitute a tool for stabilising the
nominal exchange rate. The width of the band, ±15%, cannot stabilise anything, intramarginal
interventions of the ECB are unlikely, and large-scale marginal intervention of the
ECB cannot be guaranteed. Under such circumstances an agreed central parity in the narrow
pressures in the foreign exchange market (especially non-fundamental and expectations-driven ones) can involve
allowing the exchange rate to change sharply and then adjust back with the support of accommodating and
offsetting interest rate policy.
6
range could be prone to testing by the foreign exchange markets. The probability of this is
increased by the uncertainty of market participants as to the “fair” value of the currency,
owing to the multiple-equilibria phenomenon, and by the potential conflict between exchange
rate targets and the inflation criterion. Paradoxically enough, the stabilising properties
promised by the ERM2’s proponents could easily be overcome by non-fundamental
destabilising forces. In addition, a double-shift from IT, to exchange rate targeting and then to
the ECB policy which is on the euro area level close to IT and on a small country level close
to currency board
4
. One of the key rules of monetary policy makers is “do not change a
monetary policy regime that is working well”. The shift from IT to a fixed exchange rate
regime may be risky. That is the main reason why inflation targeters question the claims of
the EU institutions regarding the process towards the euro. One of the key successes of the
CNB’s policies based on flexible inflation targeting and a floating exchange rate regime was
the stabilisation of both inflation and nominal interest rates at very low levels
5
. That would be
extremely difficult to achieve with a fixed exchange rate in an environment of convergence
trends, free movement of capital and unstable conditions in the world economy.
There are numerous papers highlighting the undesirable features of the ERM2 regime (e.g.
Lipschitz, 2004 or Bubula and Otker-Robe, 2003
6
). A comprehensive description of the risk
embodied in intermediate exchange rate regimes like ERM2 can be found in Begg et al.
(2003), or Buiter (2004). I will deal briefly with two of them. Lipschitz (2004) explains that
real economic forces (not only the equilibrium rate of real appreciation, but also various
structural changes) limit monetary independence in the NMSs and make these countries
highly sensitive to conditions in external capital markets. Given that, monetary and exchange
rate policies during the interregnum between joining the EU and adopting the euro will be
particularly difficult to formulate. He proposes a flexible exchange rate with a relatively large
amplitude of exchange rate swings as the most efficient measure to contain vulnerability.
From the point of view of domestic borrowers, it will reduce the incentive for excessive
foreign exchange exposure. From the point of view of the domestic authorities, less foreign
exchange exposure militates against a fear-of-floating phenomenon with the government
trying to resist market-driven exchange rate changes. And, from the point of view of
speculators, less intervention of this sort will reduce one-way bets and opportunistic
speculation. Unfortunately, the NMSs are forced to maintain a very low degree of exchange
rate flexibility during the interregnum. Lipschitz concludes that an asymmetric band with an
ostensible guarantee against significant depreciation seems to be the most dangerous policy. It
seems likely that capital flows will produce the following pattern: an initial over-appreciation
of the nominal exchange rate, coupled with an expected depreciation and a correspondingly
higher interest rate than in the euro area.
According to Buiter (2004), a serious design weakness of the Maastricht criteria for full
EMU membership is that they specify a number of nominal convergence criteria that jointly
constrain the behaviour of real economic variables in ways that may not be desirable or,
worse, not even feasible. He argues that ERM2 is a pointless and potentially dangerous
arrangement, especially if the nominal exchange rate constraint it incorporates is combined
4
The only way how to avoid the double shift is to switch to a currency board when entering ERM2 since for a
small country having the euro is just like having a currency board. Nevertheless, currency board is not exactly
the same as membership in a monetary union, and what is more important, establishing a currency board in a
country with a history of floating exchange rate is not a risk-free option.
5
The Czech Republic used the interest rate channel in reaction to the 2002 appreciation and subsequent
slowdown of the world economy with the aim of supporting domestic demand. Despite the external constraints
already existing, monetary policy proved to be autonomous enough to achieve the desired outcomes.
6
Bubula and Otker-Robe (2003) document that during 1990–2001 the frequency of crisis under intermediate
regimes was substantially higher than under polar regimes.
7
with an inflation target and a nominal interest rate target. The simultaneous pursuit of three
nominal targets greatly enhances the likelihood that a major financial accident will happen. A
state-contingent set of criteria creates an environment for indeterminacy and multiple
equilibria if the state variables in question are expectational and non-predetermined. For most
NMSs it will bring unnecessary exposure to potentially destabilising international capital
Предмет: | Комуникационна техника и технологии, Технически науки |
Тип: | Лекции |
Брой страници: | 15 |
Брой думи: | 944 |
Брой символи: | 5875 |