Vitor Gaspar and Vitor Bento *
I. THE EXCHANGE RATE STABILITY COMMITMENT
Exchange rate stability is usually regarded as desirable by European Union member states. The pro-stability arguments can be grouped into three categories. First, the risk arising from exchange rate uncertainty creates a barrier to exports, imports, borrowing, and lending ; therefore the Single Market would require a Single Currency: "One Market, One Money". Second, exchange rate uncertainty creates problems for the management of the Community and of the national budgets, and for the working of the Community's com-mon policies themselves. Third, fixing the exchange rate provides a nominal anchor for monetary policy. Given that the (nominal) exchange rate is particularly easy to monitor the commitment to exchange rate stability may enhance the credibility of price stability.
The arguments in favour of exchange rate stability in Europe have not vanished in the face of the recent exchange market crisis. Nevertheless recent experience provides clear evidence on the limits on policy-makers' ability to impose their will on markets despite unprecedented volumes of exchange market inter-vention. Since September 1992 the ERM has experienced an unprecedented period of turbulence : on September 13, the lira devalues; on September 16, the UK withdraws the sterling from the ERM while the Italian authorities suspend their intervention obligations under the ERM rules and the peseta is devalued by 5 %; on November 22, the peseta and the escudo are devalued by 6 %; on January 30, 1993 the Irish authorities are granted a 10 % devaluation of the punt ; on May 13, 1993, the peseta is devalued by 8 % and the escudo by 6,5 %; on August 2, 1993 currencies move to fluctuation bands of 15 % around unchanged central rates. This impressive sequence of events understates the magnitude of the turmoil since it ignores episodes of unsuccessful attacks on ERM currencies (including the french franc, the belgian franc and the danish krone) and the fate of non-ERM currencies, linked to the ECU, that are now floating : the Finnish markka, and the Swedish krone and the Norwegian krone.
The adverse impact of these developments on the credibility
of the process leading to Economic and Monetary Union cannot be overemphasized.
It calls into question the idea of gradual and continuing convergence amongst
Member States's economic policies and results. There is the risk that the
ERM turmoil be interpreted by the public and the politicians alike as a
signal that the process of monetary unification might not work. Such perception
could undermine the political will to move to full EMU.
II. THE IMPACT OF FINANCIAL INTEGRATION
The Exchange Rate Mechanism (ERM) of the European Monetary System (EMS), like the Bretton Woods System, is a regime of fixed (albeit adjustable) exchange rates. One major difference in the environment in which policies were to be defined, between the latter period of the ERM and earlier experiences, was given by the much larger degree of capital mobility. In fact by mid-1980 most European currencies were fully convertible in a world of global financial markets. The emergence of global financial markets and the trend towards financial liberalization in Europe and elsewhere increased competition reduced profit margins and economic rents thereby improving economic efficiency, that is ensuring better terms for borrowers and lenders, savers and investors. As a matter of fact the continuing trend of integration of financial markets had all but eliminated short-term covered interest rate differentials for all the major industrialized countries by the late eighties. Of course the 1992 internal market programme goes much further than the simple lifting of capital controls. It envisages the full integration of the various national financial systems.
The emphasis on the impact of financial integration on
resource allocation efficiency should make clear that there is no real-financial
dichotomy in any meaningful sense. Paul Krugman made this point in a particularly
clear way : "In a deep, underlying sense, integration of national economies
through capital movements is not very different in its causes and effects
from trade in goods and services. As in the case of trade, capital movements
serve to allow countries to benefit from their differences through long
run transfer of resources to countries where they are more productive.
Also like trade, capital mobility can be beneficial through its effects
on financial market efficiency even where little net transfer of resources
takes place". Nevertheless integration of capital markets has different
implications on policies than the removal of barriers to trade.
III. THE CREDIBILITY OF A PRICE STABILITY POLICY
As already indicated, one argument for exchange rate stability is derived from a (assumed) link between the exchange rate commitment and the credibility of policies for inflation prone countries. This argument comes from a relatively recent literature allowing the behaviour of policy-makers to be endogenous. Policy is determined (in part) by political considerations and certainly reacts to the state of the economy. In this context a credible commitment to exchange rate stability makes inflation less likely (or disinflation more likely) therefore moderating wage increases, and contributing to improve the unemployment-inflation trade-off. One may recall that the title of one of the more often-quoted papers making this argument: " The advantage of tying ones' hands" .
The problem of course, is that there is no such a thing as irrevocably tying one's hands. Specifically short of a single currency it is very hard to envisage credible commitment mechanisms ensuring that devaluations are no longer an option. In other words it is extremely dangerous for market participants but much more for policy-makers to take credibility for granted. Furthermore since realignments have always been possible under ERM rules the mechanism provided a partial commitment devise at best. Therefore the fact that countries going through a realignment incur a political cost works like a double-edged sword. On one hand the existence of significant costs enhances the credibility of the commitment to exchange rate stability but, on the other hand, makes it likely that even with divergent fundamentals, private economic agents can unfold their long positions in weak currencies in the event of an expected devaluation since they may rely on central bank intervention providing sufficient time and liquidity.
It is generally accepted that there can be no sustained inflation in the absence of accommodation mechanisms and, in particular, of steady increases in the money supply. Further it is accepted that there is no longer run unemployment-inflation trade-off . Nevertheless most countries are far from the price stability benchmark. The best available explanation of this paradox is provided by the litterature on time-(in)consistency . Dynamic inconsistency arises when the best ex ante plan is no longer optimal when the moment of implementation comes. Typically inflation results from the attempt to explore short-run trade-offs in order to force higher employment, the financing of the government budget deficit, and the avoidance of large balance of payments deficits. Furthermore the concern of the authorities with financial stability may motivate interest rate smoothing that leads likewise to an inflationary (devaluation) bias . The point is that rational economic agents anticipate the short-run dilemmas of the authorities and react accordingly so that the final outcome is worst than what would result if the authorities were able to pre-commit.
It is interesting that in a report from the Committee of Central Bank Governors 6 it is said that: "The crucial element for the short run credibility of a given central parity comes from the commitment of monetary policy in its defence. It is therefore important that there are no inhibitions to the use of interest rates preventing an effective defence of the currency against speculative attacks. The existence of constraints on nominal interest rates seems to be related to a number of factors: the size of public debt (in particular the share of floating rate debt) and deficits; the size of private debt; the extent to which GDP deviates from potential, the level of unemployment, the degree of fragility of the financial sector (...)". It is certainly no coincidence that most of the factors quoted are exactly the ones that would endogenously contribute to a time consistency problem for the authorities.
These problems were exacerbated by the business cycle
slow down in most European countries, the increase of unemployment to unprecedented
levels, the increase in German short term interest rates and the strength
of the DM 7 given its anchor
role within the ERM. The conflicts between the internal and external stability
requirements were clear for market participants from the reluctance in
raising their interest rates to match the German's
IV. THE COMMON CREDIBILITY OF THE EMS
It is, however, clear that the previous argument is only part of the story. One striking feature of the period of exchange market turmoil was the sequential transmission of pressures from one currency to the next. One may take this fact as an illustration of the collective good character of the mechanism. Each participant has credibility derived from its own past record and other idiosyncratic factors and from the credibility of the system. Given that each member's credibility depends on the credibility of the system it is not unlikely that it is adversely affected by the perception of exchange market turmoil and that realignments are (in general) likely. Given that the defence of a given central parity is the more costly the least credible its defence, the decline in credibility may become a self-fulfilling prophecy. Specifically, the fact that credibility is undermined makes exchange rate stability more costly, triggering a realignment or the floating of the currency.
The dependence of the authority's policy response on market
expectations based, in turn, on the policy rules themselves makes the existence
of multiple equilibria likely, some of which may be unstable. If these
were the main causes for the ERM crisis then the decision to widen the
fluctuation margins to 15 % in August 1993 may (paradoxically) contribute
to exchange rate stability.