Exchange rates targets.

Obstfeld, Maurice

When massive speculative attacks shattered the system of fixed exchange rates among industrial countries in early 1973, policymakers viewed the shift to generalized floating rates as a tactical retreat. Nearly 19 years later, however, most industrial countries' exchange rates continue to float against the U.S. dollar. Nonetheless, in the past decade the world economy has evolved in the direction of more limited exchange rate flexibility.

The group of countries participating in the exchange rate mechanism (ERM) of the European Monetary System has expanded since the system's inception in 1979, and the original adjustable-peg regime has given way to one in which realignments increasingly are avoided. Developing countries, as well as the transforming economies of the former Soviet bloc, have used stable external exchange rates both as anti-inflationary anchors and as a conduit for importing informative signals from world markets about relative prices. Finally, during the late 1980s the main industrial countries intervened heavily against the dollar in foreign exchange markets, and in some periods targeted the dollar exchange rates of major currencies.

Academic economic research has mirrored and, it is fair to say, encouraged the movement toward more deliberate exchange rate management. In the late 1960s and early 1970s, a significant school of academic opinion favored abandoning the Bretton Woods adjustable peg system in favor of floating rates determined by the market; the system's final collapse in February 1973 initially appeared to vindicate that view. The subsequent performance of the world economy revealed, however, that floating exchange rates were not the panacea that their early advocates had promised. It remains widely accepted that some economic disturbances require exchange rate adjustment. But it is also recognized that floating rates can move sharply and in ways that bear no close short-term relationship to international costs; they do not prevent persistent departures from current account balance; their movements redistribute income domestically, creating political pressures inimical to free trade; and they may present governments with harmful macroeconomic incentives.

My research has focused on several aspects of exchange rate targeting. One set of studies has addressed the problem of imperfect government credibility under managed exchange rates. A second examines the theory of exchange rate target zones and regime switches under uncertainty. A final area of research concerns the role of official foreign exchange intervention in pursuit of exchange rate targets. I now summarize some of this work, along with some related studies by other NBER researchers.

Credibility and Exchange Rate Management: The Case of Europe

When the eight original members of the European ERM began to peg their mutual exchange rates in March 1979, relatively few observers would have bet heavily on the system's survival. Annual inflation rates at that time ranged from 2.7 percent in Germany to 12.1 percent in Italy, and the looming prospects of recession and higher oil prices augured poorly for the future of ERMs. Indeed, the first phase of ERM history was rather turbulent, punctuated by periodic speculative crises. To date, there have been 11 substantive realignments of intra-ERM exchange rates.(1)

The last ERM alignment--that of the French franc in January 1987--apparently inaugurated a new phase marked by startling successes. Since then, exchange rates have remained stable, even though France and Italy have dismantled their remaining controls on foreign exchange transactions and capital movements; Spain and the United Kingdom have joined the mechanism; Italy has narrowed the band of fluctuation permitted to the lira under the ERM agreement; swap networks among the participating central banks have been enhanced; and several nonparticipating countries, including Norway and Sweden, have started to orient their exchange rate policies toward maintaining the ERM peg. Some observers argue that the discipline of the ERM enabled certain countries to ride in the wake of the German Bundesbank's anti-inflationary reputation.

This impressive track record, together with the momentum toward further economic and political integration within Europe, has enhanced the credibility of ERM exchange rate parities. Yet the system's ability to weather severe shocks that may lie ahead remains doubtful--whether the shocks emanate from domestis policies, German reunification, or some other source impossible to foresee today.

Data from financial markets reflect these doubts. In 1990, for example, Italy's long-term government bond interest rate exceeded Germany's by 310 basis points.(2) Such discrepancies would imply large unexploited profit opportunities under the hypothesis that the Deutsche mark/lira rate is irrevocably confined to its current fluctuation limits.(3) More plausibly, markets may perceive implicit (and explicit) "escape clauses"--contingencies in which national authorities may realign exchange rates or suspend free capital mobility despite paper commitments to eventual monetary union.

My own analysis of exchange rate escape clauses reveals some macroeconomic implications that may be problematic.(4) Escape clauses have the potential for raising welfare by giving authorities additional freedom to respond to exceptional economic shocks. Yet it is difficult to ensure that, once this freedom is granted, it will not be exercised more often than is socially optimal. I consider a policymaker who cares about employment and exchange rate variability but faces a fixed political cost of realigning the national currency. Even if this cost is consistent with an optimal frequency of realignment, there may be additional equilibriums in which the policymaker accommodates high inflationary expectations by devaluing. Because there is no way for authorities to forswear accommodation credibly, the actual outcome of the escape clause option may be inferior to a rigidly fixed exchange rate.

An additional factor straining ERM credibility is the sharp variation in public indebtedness among member countries. In 1990, ratios of net general government debt to GDP ranged from a low of 23 percent in Germany to a high of 121 percent in Belgium, with Italy's ratio of 98 percent near the upper end.(5) Observers worry that such high-debt governments as Italy face a powerful incentive to inflate away their normal obligations, and suggest that this incentive may build an expected inflation premium into long-term nominal interest rates.(6) Several studies have examined whether skillful management of the currency composition and maturity structure of government liabilities may reduce the temptation to depreciate the public debt via inflationary surprises. A related literature considers how the maturity structure of debt may limit a government's vulnerability to exchange rate crises and related lapses of confidence.(7) One serious concern is whether the ERM central banks' commitment to mutual intervention support could make credibility crises contagious.

Related work by NBER Research Associates Kenneth A. Froot and Kenneth Rogoff looks at the stability of ERM exchange rates during the transition to a single European currency.(8) They describe an "end-game" scenario in which national central banks, restrained until the moment of transition by the need to preserve their anti-inflation reputations, devalue national debts just as the European currency is born. Their general point is that credibility problems may become more evident as major steps toward economic union proceed.

Target Zones and Regime Switches

ERM exchange rates fluctuate within narrow limits, as did Bretton Woods exchange rates and even exchange rates under a gold standard. In recent literature, NBER Research Associate Paul R. Krugman finds that a perfectly credible band will have a stabilizing effect on the exchange rate. The exchange rate will remain within its limits for a wider range of its fundamental economic determinants, and its response to changes in these determinants will be muted.(9)

Froot and I have extended Krugman's analysis of target zones and have shown how to apply its principles to a wide range of anticipated shifts in the exchange rate regime: for example, a switch from floating to fixed rates that is triggered when a random variable such as the money supply reaches a pronounced threshold. One important implication of this work is that relationships between exchange rates and their economic determinants may well be nonlinear when drastic regime changes are possible. But these nonlinearities may be hard to detect without long historical data series.(10)

Other empirical work has focused on testing the target-zone model against ERM data. Robert P. Flood, Andrew K. Rose, and Donald J. Mathieson find little support for the model. However, Giuseppe Bertola and Lars E. O. Svensson, and Svensson and Rose, find that the introduction of time-varying realignment risk improves the model's ability to match the data. Kathryn M. Dominguez and Peter B. Kenen argue that ERM interention practices have not conformed to the simple target zone model. They provide empirical evidence to support that view.(11)

Foreign Exchange Intervention

On September 22, 1985, finance ministers and central bank governors of the Group of Five industrial countries dramatically announced their determination to depreciate the U.S. dollar. This Plaza Accord--named after the New York hotel that hosted the G-5 meeting--marked the start of a period of highly publicized management of the dollar exchange rates of major industrial currencies. The subsequent Louvre Accord of February 22, 1987 set up informal target zones for the dollar against the yen and the Deutsche mark, zones that were definitively breached after the October 1987 stock market crash.

Although the Louvre currency zones have not been abandoned formally--they were never announced formally--recent ministerial communiques have shifted attention away from exchange rate issues. Nonetheless, the experience of the late 1980s offered a fresh chance to rethink the economics of foreign exchange intervention and to evaluate its effects anew.

Naturally, a number of questions arise. First, does "pure" or sterilized intervention--intervention that is not allowed to affect national money supplies--have significant effects on the exchange rate? (Much post-Plaza intervention was sterilized in some way.) In other words, can intervention be viewed as a tool of macroeconomic policy that somehow can be exercised independent of monetary and fiscal policy?

Second, if intervention has effects on exchange rates, do they arise exclusively because intervention is a signal of official intentions about future monetary and fiscal policies? Or, does intervention have distinct effects by altering the currency composition of the bond portfolio that private markets must hold? If there are only signaling effects, then intervention is not an independent policy tool, since it must be followed by concrete policy shifts for signals to remain credible. But it is still crucial to ask how intervention signals can acquire credibility, and to determine what information (if any) policymakers communicate through intervention that they could not communicate through some other medium.(12)

In a review of the 1985-8 period of coordinated intervention, I argue that the broad realignment of dollar exchange rates is explained adequately by shifts in fiscal, and especially monetary, policies.(13) The signaling effect of intervention appears to have been important, particularly on occasions when central banks intervened on a concerted basis. But there are significant instances, both from 1985-8 and after, when intervention was shrugged off by the foreign exchange market pending more substantive shifts in policy.

Not all investigators concur with these conclusions about the effect of intervention. In a study based on data provided by the Bundesbank, the Federal Reserve, and the National Bank of Switzerland, Jeffrey A. Frankel and Kathryn M. Dominguez find that pure intervention has a statistically detectable portfolio effect, but that the effect is not large when the intervention is not announced.(14) This suggests some independent role for intervention, but is consistent with the assertion that much of the apparent effectiveness of intervention comes from its effect on expectations of future monetary and fiscal policies.

Michael W. Klein and Karen K. Lewis empirically model the foreign exchange market's beliefs about the authorities' dollar/yen and dollar/Deutsche mark exchange rate targets.(15) After analyzing daily data for the period of the Louvre Accord preceding the stock market crash, Klein and Lewis conclude that market perceptions of the implicit Louvre exchange rate bands evolved substantially in response to observed interventions. Their result underscores the importance of the signaling channel, and highlights the volatility of expectations even during a period of supposedly successful policy coordination.

(1)F. Giavazzi and A. Giovannini, Limiting Exchange Rate Flexibility: The European Monetary System, Cambridge, MA: MIT Press, 1989, provides a comprehensive review. (2)See OECD Economic Outlook 49 (July 1991), Table 4, p. 11. (3)For a detailed analysis, see A. Giovannini, "European Monetary Reform: Progress and Prospects," Brookings Papers on Economic Activity 2 (1990). (4)M. Obstfeld. "Destabilizing Effects of Exchange Rate Escape Clauses," NBER Working Paper No. 3603, January 1991. NBER researchers R. P. Flood, V. U. Grilli, T. Persson, and G. Tabellini have done related work. (5)OECD Economic Outlook 49 (July 1991), Table 32, p. 113. (6)I explore some dynamic implications of these incentives in "A Model of Currency Depreciation and the Debt-Inflation Spiral," NBER Reprint No. 1498, January 1991, and Journal of Economic Dynamics and Control 15 (January 1991); and in "Dynamic Seigniorage Theory: An Exploration," Centre for Economic Policy Research Discussion Paper No. 519, March 1991. (Original version issued as NBER Working Paper No. 2869, February 1989.) (7)See, for example, T. Persson and G. Tabellini, Macroeconomic Policy, Credibility, and Politics, Chur, Switzerland: Harwood Academic Publishers, 1990; G. A. Calvo and M. Obstfeld, "Time Consistency of Monetary and Fiscal Policy: A Comment," Econometrica 58 (September 1990); and the papers by A. Alesina, A. Prati, and G. Tabellini; by F. Giavazzi and M. Pagano; and by G. A. Calvo and P. Guidotti in Public Debt Management: Theory and History, R. Dornbusch and M. Draghi, eds., Cambridge, U.K.: Cambridge University Press, 1990. For a general discussion of links between monetary union and fiscal harmonization, see B. J. Eichengreen, "One Money for Europe? Lessons from the U.S. Currency Union," Economic Policy 10 (April 1990). (8)K. A. Froot and K. Rogoff, "The EMS, the EMU, and the Transition to a Common Currency," NBER Working Paper No. 3684, April 1991. (9)P. R. Krugman, "Trigger Strategies and Price Dynamics in Equity and Foreign Exchange Markets," NBER Working Paper No. 2459, December 1987, and "Target Zones and Exchange Rate Dynamics," NBER Working Paper No. 2481, January 1988, and forthcoming in Quarterly Journal of Economics. Further references can be found in Exchange Rate Targets and Currency Bands, P. R. Krugman and M. H. Miller, eds., Cambridge, U.K.: Cambridge University Press, 1991. (10)See K. A. Froot and M. Obstfeld, "Exchange Rate Dynamics under Stochastic Regime Shifts: A Unified Approach," NBER Working Paper No. 2835, February 1989, and forthcoming in Journal of International Economics; and "Stochastic Process Switching: Some Simple Solutions," NBER Reprint No. 1585, August 1991, and Econometrica 59 (January 1991). (11)R. P. Flood, A. K. Rose, and D. J. Mathieson, "An Empirical Exploration of Exchange Rate Targe Zones," NBER Working Paper No. 3543, December 1990; G. Bertola and L. E. O. Svensson, "Stochastic Devaluation Risk and the Empirical Fit of Target Zone Models," NBER Working Paper No. 3576, January 1991; A. K. Rose and L. E. O. Svensson, "Expected and Predicted Realignments: The FF/DM Exchange Rates During the EMS," NBER Working Paper No. 3685, April 1991; and K. M. Dominguez and P. B. Kenen, "On the Need to Allow for the Possibility That Governments Mean What They Say: Interpreting the Target-Zone Model of Exchange Rate Behavior in the Light of EMS Experience," NBER Working Paper No. 3670, April 1991. (12)For a more extensive discussion of the general issues, see M. Obstfeld, "Exchange Rates, Intervention, and Sterilization," NBER Reporter, Fall 1982. (13)M. Obstfeld, "The Effectiveness of Foreign Exchange Intervention: Recent Experience, 1985-8," in NBER Reprint No. 1525, February 1991, and in International Policy Coordination and Exchange Rate Fluctuations, W. H. Branson, J. A. Frenkel, and M. Goldstein, eds., Chicago: University of Chicago Press, 1990. (14)J. A. Frankel and K. M. Dominguez, "Does Foreign Exchange Intervention Matter? Disentangling the Portfolio and Expectations Effects for the Mark," NBER Working Paper No. 3299, March 1990. (15)M. W. Klein and K. K. Lewis, "Learning About Intervention Target Zones," NBER Working Paper No. 3674, April 1991.