As long as the exchange rates between the currencies of the major industrial countries continue to fluctuate widely, either for reasons that are not clearly linked to changes in economic fundamentals or on account of inappropriate policies, suggestions for a fundamental reform of the system aimed at diminishing these exchange rate fluctuations are likely to continue to be proposed. These proposals often call for the establishment of target zones for the bilateral exchange rates of the three largest industrial countries, with the aim of reducing exchange rate volatility and the probability of exchange rate misalignments while retaining a significant level of flexibility for domestic policies.51 More generally, advocates of target zones seek to move the present international monetary system in the direction of greater centralization and more explicit cooperation and coordination of economic policies, in order to prevent the exchange rates of the three largest industrial countries from fluctuating to the extent that has been recorded since the end of the Bretton Woods system. Underlying these proposals is the view that the revealed preference of Germany, Japan, and the United States for a system of managed floating is not conducive to the proper functioning of the international monetary system. With target zones, it is argued, an explicit structure of specific exchange rate arrangements will lead to both better exchange rate behavior and more disciplined and appropriate economic policies on the part of the three largest industrial economies.
The distinction between a system of target zones and the present system of managed floating between the three most important currencies is a matter of degree. A “quiet” or “soft” system of target zones—with wide and unannounced exchange rate ranges, limited commitments to intervene to defend these zones, and no firm understandings and little willingness to adjust monetary policies for exchange rate purposes—could be little different from the present system of policy coordination among the major countries. It would be a mistake to characterize the formalization of such a weak system of target zones as a “fundamental reform” of the international monetary system or to expect that it would accomplish something significantly different from the present system. Accordingly, to provide a meaningful contrast between the proposals for fundamental reform and evolutionary improvement of the international monetary system, it is useful to define three characteristics of a target zone system that would clearly distinguish it from the present system. First, the permitted ranges for exchange rates, together with the mechanisms for adjusting these ranges, should imply notably smaller fluctuations between the exchange rates of the major currencies than has been observed in recent years. Second, relatively firm commitments should be made to intervene in substantial amounts to defend the agreed ranges for exchange rates. Third, although domestic monetary policies would retain considerable independence and the commitments to defend exchange rate zones would not be absolute, it should clearly be understood that monetary policies would sometimes need to be adjusted for exchange rate purposes, even when this might be inconvenient from the perspective of the key domestic objectives of monetary policy.
Benefits and Costs of Target Zones
Supporters of target zones make a number of arguments as to why the establishment of these zones would improve the functioning of the exchange rate system. First, target zones would impose more discipline on macroeconomic policies in two ways: to maintain exchange rates within the zones, monetary policy—and sometimes also fiscal policy—would need to be adjusted; and, if the authorities opted to alter the zone rather than their policies, they would need to explain why a new zone was appropriate and convince other participants accordingly. The latter requirement would be likely to strengthen the element of peer pressure in the policy formulation process.
Second, target zones are said to improve the international consistency of macroeconomic policies. These zones would have to be negotiated and would therefore require a mutual consistency among exchange rates. It is argued that the exchange rate implications of alternative stances and mixes of macroeconomic policies would thus have to be dealt with directly. In a related vein, supporters argue that the negotiation and revision of target zones could act as a convenient organizing framework for multilateral surveillance.
Third, by promoting greater exchange rate stability, target zones are viewed as providing an anchor for medium-term exchange rate expectations. The peer pressure to keep the exchange rate within the zone would give market participants useful information about the future course of monetary policy, thereby lessening the danger that short-term deviations of policy would be erroneously extrapolated into the future.
Fourth, because the members of a target zone system would be the key-currency industrial countries, it is claimed that target zones would reduce the asymmetry in adjustment in the present exchange rate system. In particular, it would subject the countries whose policies have the greatest spillover effects on the world economy to the same scrutiny and pressure experienced by smaller countries with external and internal imbalances.
In addition to the benefits deriving from discipline and cooperation, proponents have argued that the existence of a target zone could reduce the probability that markets would test the boundaries of the exchange rate range.52
If a target zone for exchange rates could be established without significant cost, it would be desirable to do so. In practice, however, a number of considerations make it likely that proposals to limit exchange rates within well-defined ranges would be ineffective in achieving the stated objective or would entail costs that would outweigh any potential benefits.
In particular, there are reasons to doubt whether an effective target zone could be imposed without incurring a significant loss in domestic monetary independence. The argument that a target zone could be self-stabilizing follows from the initial theoretical model of a target zone.53 A fully credible target zone would change the relationship between the exchange rate and the economic fundamentals in such a manner that the impact of unfavorable movements in the fundamentals on the exchange rate would weaken as the rate moves toward the edge of the band; hence, market participants would help to maintain the exchange rate within the band. The implication is that the loss of monetary independence caused by the imposition of a target zone could be substantially smaller than what might be expected from past experience. Subsequent work has made clear, however, that this result depends upon the assumption that the target zone is fully credible. If it is not, the relationship between the target zone and the fundamentals becomes more complex. The empirical work on target zones surveyed in Svensson (1992) suggests that, while the relationship between the exchange rate and the fundamentals appears to be somewhat different from what would be predicted under a free float, a target zone “appears very similar to a ‘managed float’ with a target central parity but without an explicit band.” (Svensson (1992), p. 140).
The operation of a target zone would therefore involve some loss of domestic monetary independence, and there is little evidence that the major industrial countries would be prepared to compromise their domestic objectives in this manner. Even if they were willing to do so, many analysts doubt whether a target zone would be beneficial. For example, the belief that a target zone provides a useful anchor for exchange rate expectations can be questioned. As most target zone proposals involve relatively wide bands and frequent reassessments of these ranges, it is not clear how this system would provide such an anchor. Similarly, the argument that a target zone would provide more symmetry in adjustment is open to doubt. The evidence does not indicate that international adjustment under the Bretton Woods exchange rate regime was faster or more symmetric.54 Given this experience of regimes with fixed exchange rates, it is not clear that the operation of a target zone would generate the hoped-for benefits.
More fundamentally, many would argue that establishment of target zones for the three key currencies could actually harm the world economy. By focusing on exchange rates rather than on underlying macroeconomic policies, such a system runs the risk of directing attention to symptoms rather than diseases and could actually lessen the pressures for corrective action. In addition, while movements in the exchange rate are often a useful signal to policymakers, the macroeconomic discipline provided by a policy of stabilizing the exchange rate need not always be beneficial.
A significant amount of work has been done on the impact of different exchange rate regimes.55
Much of the recent empirical work has used international macroeconomic models to assess the benefits of different international regimes, including target zones.56 The results from comparing a target zone to a regime based on domestic indicators, such as the money supply, have been mixed; they appear to depend upon the nature of the underlying disturbances and the way in which the different regimes are specified. While it is always possible to criticize such work, as it is very difficult to analyze the impact of a fundamental change in a monetary regime, the evidence that a successful target zone would improve macroeconomic stability does not appear to be strong.
Also, a number of practical considerations with regard to target zone agreements would have to be resolved. It would be necessary to reach agreement about the appropriate range of fluctuations in the chosen exchange rate (bilateral or effective) and, most important, the responsibilities of participants for keeping the exchange rate within that range. The size of the range would depend on whether the range was aimed primarily at limiting exchange rate misalignments, or whether it should also aim to reduce volatility. The former objective would imply relatively wide bands, with exchange rates allowed to vary significantly over the short term. The latter combination of objectives would imply a much narrower band, such as under the original ERM in Europe, but, as the example of the ERM shows, such a band would be less flexible in the face of perceived policy dilemmas.
In either case, it would also be necessary to agree on the underlying parities and on how these parities might be moved in the light of changes in economic fundamentals. Agreement on underlying parities could well prove to be difficult. Indicators of appropriate exchange rates, such as alternative measures of international cost competitiveness, often provide imprecise and somewhat conflicting answers.57 Moreover, different country authorities may not always share a common view or interest in defining appropriate ranges for exchange rates. In the first half of 1994, for example, the U.S. and Japanese authorities were not in complete accord regarding the range of the bilateral exchange rate between the yen and the dollar that would both contribute to a gradual reduction of payments imbalances and avoid undermining the recovery of the Japanese economy.
Some agreement on the relationship between the agreed target ranges and the economic fundamentals would also be desirable, so that it would be understood that changes in the fundamentals would prompt discussion of the need for changes in the ranges. Otherwise, these ranges would be in danger of becoming too rigid. While clearly not inevitable, arrangements to limit exchange rate variability have a tendency to become less flexible over time. The problems of the ERM in 1992–93, for example, are generally believed to have stemmed, at least in part, from a reluctance to change parities following the upward pressure on the deutsche mark as a result of the macroeconomic consequences of German unification. Similarly, the collapse of the Bretton Woods system was marked by problems in agreeing on changes in parities.
Policies to Maintain Target Zones
Exchange Market Intervention
The essential ingredient required to make an exchange rate range credible and effective is a clear agreement on both the policies required to keep the exchange rate within this range and the responsibilities of the participating governments in implementing these policies. One approach to maintaining such a range is sterilized exchange market intervention. However, the consensus view, as embodied in the Jurgensen Report (Jurgensen (1983)), is that the influence of sterilized official intervention is generally short term and cannot be relied upon to have either a substantial or a durable influence on exchange rates. Subsequent to the Jurgensen Report, some studies have suggested that coordinated intervention might be more useful than was previously appreciated, particularly as a signaling device for the implementation of other policies.58 However, the experience with quite massive interventions during the European exchange rate crises of 1992–93, as well as with other recent interventions, appear to confirm the thrust of the Jurgensen Report, namely, that there are relatively narrow limits to what intervention alone may be expected to achieve on a sustained basis. This does not imply, of course, that intervention is irrelevant for target zones; indeed, such zones would be superficial if the authorities were not prepared to intervene when exchange rates threatened to breach their limits. Also, if there were a commitment to use other, more basic policies to defend the zone, intervention could carry more weight in the market. Nevertheless, a wide range of experience clearly demonstrates that exchange market intervention by itself is not adequate to significantly affect exchange rates over an extended period of time.
One reason for the limited effectiveness of official intervention is the expansion in private capital markets over time. As the magnitude of private foreign exchange market transactions has increased, the ability of governments to influence these markets through intervention has declined. This has led some observers, most notably James Tobin, to propose taxing international capital transactions or requiring the opening of interest-free deposits at the central bank by those undertaking capital transactions, with the aim of putting “sand in the wheels” of the capital markets.59 Such a tax, it is argued, would increase the cost of speculation and, hence, the efficacy of official intervention. However, as with all capital controls, such a tax would also limit useful capital market transactions. There is no clear-cut way to separate inappropriate speculation and socially unproductive capital flows from those that are desirable. If restrictions based on taxes are not successful in distinguishing between productive and unproductive flows, many of the benefits of liberalization will be sacrificed, including increased returns to savers, lower costs of capital to firms, and better hedging instruments. Moreover, given the fungibility of capital and the ease with which taxed transactions can move off-shore, it is not even clear that such a tax would be effective in limiting exchange rate volatility. While international capital markets may not always work perfectly, it is clear that they provide significant efficiency gains. The objective of reforms should therefore be to enhance their operations, not to limit them.
More generally, it may be noted that transactions costs, especially in stock markets, have declined precipitously over the course of this century as technology has improved. This decline in transactions costs has been associated with a massive increase in trading volumes in virtually all forms of marketable assets. Studies of the behavior of asset price volatility, however, do not reveal any upward trend, although the degree of volatility has varied to some extent from one decade to the next. This evidence does not suggest that a tax on transactions would be an effective method of reducing asset price volatility; indeed, it is difficult to see how taxes on transactions could lower asset price volatility, unless the taxes were set sufficiently high as to virtually eliminate these markets.
Consideration could be given to deploying fiscal policy as an instrument to keep exchange rates within desired ranges. Again, however, there would seem to be formidable difficulties. First, the magnitude—and possibly even the sign—of the effect of fiscal policy on the exchange rate is uncertain and may change with special circumstances. The usual presumption is that expansionary fiscal actions result in an appreciation of the domestic currency; however, this may not be the case if the fiscal expansion undermined the credibility of government policies, in particular with regard to inflation. Furthermore, situations might arise in which the confidence-building effect of fiscal consolidation might induce exchange rate appreciation when such appreciation was not desirable.
Second, and more generally, it may be asked whether a system of target zones for the major currencies would usually operate in the right direction in terms of disciplining errant fiscal policies. When a country with a large fiscal deficit sees its currency depreciate excessively, the traditional advice given is to cut the deficit, both for its own sake and to alleviate pressures in the exchange market. Such traditional advice is based on the presumption that the confidence effect of cutting the deficit will outweigh the direct effect of budgetary consolidation on the real exchange rate. Many countries in this situation—including some industrial countries—have properly interpreted exchange rate depreciation as sending a message about the need for fiscal consolidation. It is far from clear, however, that exchange rate movements will consistently send the right messages about desirable adjustments of fiscal policies, especially to the three largest industrial countries. The United States in the mid-1980s is one example of this problem; it is unclear whether a much-needed move toward fiscal consolidation at that time would have made the over-valued dollar weaker or stronger. Another example is Japan in 1993–94, when an expansionary fiscal policy was being used to combat recession, and when a sharp appreciation of the yen helped to undermine an incipient economic recovery. What message was the appreciation of the yen sending to Japanese fiscal policy in that situation? Should Japanese fiscal policy have been tightened, in line with the normal presumption about the effects of fiscal policy on the exchange rate? Would a tightening of Japanese fiscal policy have made any sense in view of the cyclical situation in Japan and in other industrial countries, and in view of Japan’s substantial current account surplus?
Moreover, even if the impact of fiscal policy on the exchange rate were known with reasonable certainty, the inflexibility of fiscal policy would remain as an important limitation on its use for exchange rate stabilization. Large exchange rate movements often occur relatively quickly with little or no advance warning, while fiscal policy is generally adjusted on a less frequent basis and often requires legislative approval. Fiscal policy also has its own important objectives, which should significantly limit its availability for exchange rate stabilization. For example, many industrial countries are currently engaged in medium-term programs to reduce their fiscal deficits and their ratios of government debt to GDP. The implementation of appropriate fiscal policies may well contribute indirectly to greater stability in exchange rates by making the domestic economic environment more predictable and by providing more flexibility for monetary authorities. However, the direct use of fiscal policy as a primary tool for exchange rate stabilization appears to be neither likely nor appropriate.
The principal benefit of fiscal consolidation in industrial countries is to strike a better balance between world saving and investment. By boosting the level of world saving, such consolidation would lead to lower real interest rates, greater productive investment, and higher growth in real output. Furthermore, these benefits would be worldwide, reaching well beyond the countries actually implementing such policies. As transmitted through the international monetary system, the implications of a country’s economic policies and performance for other countries are therefore by no means limited to exchange rates alone.
The Role of Monetary Policy
As an instrument for maintaining exchange rate stability under the proposed target zones, monetary policy is the obvious candidate. It has the short-term flexibility required to respond to day-today events, and it is generally acknowledged to be among the more important determinants of exchange rate behavior. Also, countries that have successfully pegged their exchange rates over long periods of time have generally done so by targeting their monetary policies to this objective. This approach has usually entailed keeping domestic interest rates close to those in the foreign country to which the exchange rate is pegged, although, on occasion, it has required the adjustment of domestic interest rates to maintain the exchange rate peg.
Clearly, there are circumstances in which the conduct of monetary policy can be improved by constraining it through a commitment to stabilize a nominal exchange rate. This is essentially the “monetary discipline” argument in favor of pegged exchange rates or of target zones. When a country has difficulty in committing its monetary policy convincingly to the objective of maintaining reasonable price stability in the medium term, pegging the exchange rate of its currency to a partner country with an established reputation for price stability can be a very useful policy discipline. This was the experience of several European countries that pegged their exchange rates to the deutsche mark in successful efforts to bring down domestic inflation. It has also been the experience of a number of developing countries that have used exchange rate anchors in stabilization programs to reduce inflation dramatically.60
However, the situation in the largest industrial countries is not comparable to that of smaller countries, whose significant inflation problems could be addressed through a nominal exchange rate peg. After the upsurge of inflation in the 1970s and the costly but necessary efforts at disinflation in the 1980s, the three largest industrial countries have generally maintained relatively low inflation rates. When upsurges of inflation threatened in the late 1980s and early 1990s, all of these countries tightened monetary policies to contain the inflationary threat; moreover, inflation in these countries is falling toward the lowest rates recorded in the past three decades. In all three countries, there appears to be a firm commitment to use monetary policies to achieve reasonable price stability in the medium term. Accordingly, the idea that exchange rate discipline would be valuable in guarding against inflationary monetary policies does not appear to be particularly relevant to the situation of the largest industrial countries.
Moreover, orienting monetary policy toward exchange rate stabilization has distinct disadvantages. Because it is also the most flexible and timely instrument available for achieving domestic policy objectives, any use of monetary policy to limit exchange rate variability would inevitably reduce its flexibility to address domestic concerns. This dilemma was brought out most clearly in the European exchange rate turbulence of 1992–93. The anchor country in the system, Germany, was pursuing a policy of relatively high interest rates to resist inflationary pressures caused by the fiscal expansion following German unification. Such a policy was clearly less appropriate for countries that were experiencing economic downturns, most notably the United Kingdom and the Nordic countries. The perceived conflict between the relatively high interest rates required to maintain the exchange rate peg and the lower interest rates needed to revive domestic activity was clearly an important element in the market pressures in 1992 and 1993.
If such tensions were true in Europe, with its high level of intra-EU trade and integration, it is likely to be even more of a problem across the three largest industrial countries. To illustrate the difficulties involved, consider the situation in the summer of 1992, just before the beginning of the exchange rate turmoil in Europe, when the U.S. dollar fell to a new low against the deutsche mark. This movement, which some regarded as excessive and undesirable, was at least in part a result of the different cyclical positions of the two economies. Monetary policy in Germany was relatively tight, while the weakness of the economic recovery and the low level of inflation in the United States led the Federal Reserve to continue to ease monetary conditions through the first half of 1992. Ultimately, the federal funds rate was pushed down to 3 percent in August 1992. To keep the dollar from depreciating as much as it did against the deutsche mark would have required a change in the actual or perceived monetary policy objectives either in Germany, which would have had to relax its anti-inflationary stance, or in the United States, which would have had to focus less on the domestic recovery. In both cases, such a change in economic policy would appear to have been neither popular nor economically desirable.
The history of the past 20 years supplies other important examples where the targeting of monetary policy on the exchange rate could have interfered with appropriate domestic objectives of monetary policy. In addition, to the extent that the commitment of monetary policy to exchange rate objectives might have undermined the achievement of growth and price stability objectives in the largest industrial countries, the harmful effects would probably not have been limited to these countries alone.
Target zones have been proposed for some time by Williamson; see Williamson (1985), Williamson and Miller (1987), and Williamson and Henning (1994). For a discussion and analysis of many of the issues related to target zones, see Frenkel and Goldstein (1986).
This is one of the key implications of the literature on target zones started by Krugman (1991). Svensson (1992) provides a review of the literature.
Indeed, the collapse of the system is often attributed to the lack of these properties. See the discussion in Eichengreen (1993).
For example, Canzoneri (1982) and Flood and Marion (1982).
Much of this work is contained in the books edited by Bryant and others (1989) and Bryant, Hooper, and Mann (1993). Most of this work focuses on the implications for the participants; unfortunately, there is very little work on the effects on countries outside the target zone.
Such indicators are described in Clark and others (1994).
See in particular Dominguez and Frankel (1993).
See Tobin (1978).
For an analysis of their experience, see Bruno (1993), Calvo and Végh (1994), and Végh (1992).