Comments: The Case for Benign Neglect
- Alexander Swoboda, and Peter Kenen
- Published Date:
- December 2000
Cœuré and Pisani-Ferry’s paper could be retitled, “In search of common international economic policy principles.” Their plea is not for an explicit commitment to coordination, which is, according to them, “too demanding.” They are looking for a softer approach to coordination. I question whether constant exchange rates can be established in this way. I begin by asking what would have to be done if we want stable exchange rates. Next, I show that the necessary conditions will not be achieved. Thus, my argument is a contrario.
Of course, everyone would like to have relatively stable exchange rates between the major currencies of the world and between the other currencies as well, but the issue is whether stable exchange rates can be obtained without jeopardizing other important policy targets. Stability of exchange rates means three different things: first, the absence of excessively high short-run volality, that is, deviations relative to a long-term trend of the exchange rate on a month-tomonth, semester-to-semester, or even year-to-year basis; second, not to have massive jumps in the exchange rate with variations of 40-50 percent or more in a very short period, representing an overshooting relative to a two- to three-year horizon; and third, not to have trend reversals in the exchange rate over longer time periods. Slumps and long-term trend reversals involve higher adjustment costs in the real sphere of the economy than short-term volality. Note that we ultimately are interested in preventing abrupt changes in the real exchange rate (that do not correspond to the long-run fundamentals) to keep real adjustment costs low, and not in the stability of the nominal exchange rate alone.
Let me review some basics. A currency loses its value if market participants move out of it. With the demand and the supply of currencies being derived from capital movements and trade in goods, an interplay between interest rate parity and PPP explains exchange rate variations. On the one hand, if interest rate parity, i=i* + êe (where êe is the expected rate of change in the exchange rate and i and i* are the domestic and foreign interest rates), is violated, portfolio capital will start to move. On the other hand, PPP determines the exchange rate expectations,
If the economic policy of a country errs away from stability and if market participants expect a stronger rise in the price level of that country than abroad, market participants will expect a devaluation of the currency of that country. As soon as interest rate parity sets in motion capital flows, markets require a risk premium on the national interest rate and eventually, if the rise in the interest rate is not sufficient to defend the exchange rate,1 markets induce a devaluation of the national currency.2
In the short and medium term, exchange rates can deviate from a longerterm trend. The exchange rate will overshoot if the national money supplies are not in line with real conditions. In a simple monetary model, the rate of depreciation of the domestic currency will be equal to the sum of (1) the excess of domestic over foreign monetary growth; (2) the excess of foreign over domestic income growth; and (3) the rate of depreciation of the real exchange rates (determined by real factors exogenous to the model). In a richer model and in the real world, other elements such as fiscal imbalances in the public budget, public debt, excessive wage policy, the financial situation of the social security systems, the credibility of government, and the competitive position of firms have to be taken into consideration as well. Thus, if the relative stability between countries gets out of line, the exchange rate has to give.
Pegged rates may add to instability of the nominal and real exchange rate if they postpone adjustment, which eventually has to be done abruptly, and then leads to instability in the real sphere of the economy. This holds for pegs to a single currency, for basket pegs, and for crawling pegs3 if the crawl does not correspond to the inflation differential. When the peg gets out of line with the fundamentals, arbitrage becomes possible at relatively low cost promising high returns. You also may call this speculation. If a financial bubble accompanies the asynchronic behavior of the pegged rates, the fundamental nature of the business cycle changes, implying deeper ups and downs (Siebert, 1999). The financial crises in the Czech Republic, in Asia, and in Brazil are evidence of the aggravating properties of pegged rates.
How can instability of the exchange rate be prevented? The target zone approach (Williamson, 1983) can be used to find the necessary conditions for the exchange rate to remain stable. This approach seeks to prevent exchange rate instability by letting the exchange rate fluctuate within a band. The exchange rate is supposed to deviate from the (real) equilibrium exchange rate only within a limited range.
In this system, once the limit of the band is reached, the central banks have to intervene. For instance, if the supply of a currency, say the dollar, increases because of an excessively rising money supply, a depreciation of the dollar can be avoided only if the central banks intervene and purchase the excess supply of dollars in exchange for other currencies.
The target zone approach is not promising. It is likely to fail for three reasons, which are specific to this approach:
(1) The first problem arises from having to determine the reference (real) rate. A political agreement of sovereign states is needed to determine the ex ante “equilibrium exchange rate” and it will therefore likely become a political variable and not an equilibrium exchange rate determined by economics. The latter is difficult to determine anyhow because all the elements that could change the expectations of market participants would have to be identified.
(2) The second problem is who has to intervene at the limits of the band. Take the euro and the U.S. dollar. If the U.S. Federal Reserve has to buy euros in order to support the euro and, in doing so, supplies U.S. dollars, this means that the dollar money supply is, de facto, steered by a foreign central bank. The way out of this dilemma is for the intervention to be undertaken by the central bank whose currency is under pressure. It is highly unlikely, however, that political agreement can be reached on this point because countries would have to be prepared to raise their interest rate and go into recession in order to defend the exchange rate.
(3) The system has a destabilizing effect as soon as the limits of the band are reached and markets doubt the credibility of the band limit. This is because the costs of speculation are relatively low in the neighborhood of the band limits. The story that Paul Krugman (1991) told about the smoothing properties of target zones depends on the credibility of the band and is quite unrealistic; witness the partial collapse of the European Exchange Rate System in 1992.
Another approach that sheds light on the necessary conditions for stable exchange rates is the McKinnon proposal (1982). In this approach, the exchange rates are used as a nominal anchor of monetary policy and are stabilized by a coordination of national monetary policies. The world quantity of money should increase according to some rule, let us say according to expansion of production potential. If a currency comes under pressure to depreciate, then monetary expansion of this country has to be reduced, whereas a currency with a tendency for a revaluation would need a more generous monetary expansion. It has to be kept in mind that such a system would attempt to stabilize the exchange rates but there would be the risk of fluctuating national price levels. This means instability at home.4 Moreover, the sovereign states would need to agree on a definition of a reasonable monetary expansion. But it would be insufficient to coordinate monetary policies exclusively because expectations on interest rates and exchange rates play a role. Consequently, stabilization policy, including fiscal policy, is of great importance as well, if monetary stability is to be established.
It is a different matter if, in contrast to a multilateral system of stable exchange rates, individual countries pursue an exchange-rate-oriented monetary policy. They then choose the exchange rate as a nominal anchor. Their domestic currency is tied to a foreign currency and they import price level stability from abroad. That implies that monetary policy has to influence the domestic price level in such a way that the exchange rate remains stable. An exchange rate change, ê=0, is targeted and, following PPP,
In practice, attempts at coordination have been disappointing. Thus, the Plaza Agreement (1985), the Louvre Accord (1987), and the ensuing Japanese bubble would suggest that attempts to stabilize the exchange rate include a high risk of failure. It can be argued that the attempts at coordination were a root cause of the Japanese bubble. If I were a Japanese economist looking at the (misleading) advice that was given to Japan in the late 1980s, under the name of coordination, by internationally renowned economists, I would sing the song “What have you done to my yen (and to my financial system)?” (Siebert, 1999). In addition, economic summits have not been successful in coordinating interest rate policy (Schweickert, 1999) and the experience with the financial crises in Mexico (1994), the Czech Republic (1997), and Brazil (1999) showed that a severe divergence between a defended, constant, nominal exchange rate and the real exchange rate is at the heart of currency crises.
There are several reasons why international coordination to make currencies stable has failed in the past:
(1) Economic policy philosophies differ among countries, implying different policy stances. Witness the Anglo-Saxon and the Continental approach to the unemployment problem, and the reluctance of European countries to undertake structural reforms of their labor markets and social security systems.
(2) Countries have different economic situations and cycles, so if they followed the same paradigm they would still tend to choose different policy measures.
(3) Their political situations are different. For instance, countries are in different stages of the election cycle or are characterized by different constitutional safeguards for stability and have different constraints on policy.
(4) Time inconsistency of economic policy comes into play.
(5) Countries would have to give up national sovereignty if they want stable exchange rates.
I am skeptical that rules on how to respond to economic shocks can meaningfully be agreed upon as suggested by Cœuré and Pisani-Ferry. Of course, analysis should be done on how countries may react to external shocks and how to be prepared when shocks arise, but countries are likely to have different economic policy approaches. Moreover, there will be debates about the type of crisis, most prominently whether a shock is truly external or whether it is homemade.7 Thus, the relatively weak GDP growth rate in 1999, in two major Continental countries, Germany and Italy, may be viewed as an external negative demand shock by some, but as an expression of weak investment in the 1990s by others. For Cœuré and Pisani-Ferry to suggest that it is the role of monetary policy to react to shocks is highly questionable and, for me, even scary. Imagine the reaction function of a central bank responding to external shocks, instead of attempting to steer the money supply according to some long-term target, for instance, in line with the growth of the production potential. Monetary policy would tend to become unsteady and thus contribute to monetary and financial volatility.
I am also skeptical about systematically reducing portfolio capital flows in order to have more stable exchange rates, for instance, by “throwing sand into the wheels” of the international capital markets with a Tobin tax. Care must be taken that such interventions do not impede real capital flows because this would imply welfare losses. An exit constraint for portfolio capital will deter inflows, and uncertainty about the right to take portfolio capital out of a country may make a country less attractive for direct foreign capital as well. Moreover, an entry constraint will not prevent a sudden outflow once enough foreign capital is in the country. Finally, entry controls should not prevent necessary structural reforms. The strategy for individual countries should therefore be to make themselves more attractive for foreign direct investment. At another (systemic) level, there is also the issue of the instability generated by excessive lending and putting into place institutional mechanisms that make creditors bear a higher proportion of the risks of credit failure (bail in).
What is the conclusion? A multilateral approach to stable exchange rates among the major currencies would require a commitment on the part of the relevant countries to establish a credible currency system and to pursue a stability-oriented policy. The major countries would have to submit to a stability-guaranteeing rule system. This implies that all relevant policies in the three major regions of the world—monetary, fiscal, and income policies—would have to be more or less harmonized. This also requires that countries be willing to give up their sovereignty in important domains of economic policy. For instance, they have to be prepared to defend an exchange rate with a higher interest rate, that is, by going into recession. They should also accept high opportunity costs to defend the exchange rate. This would seem to be both unrealistic and undesirable. In addition, instability may be aggravated.
It is a realistic assessment that, too often, the triggers for exchange rate volatility are political, reflecting unstable economic policy conditions—above all, failed stabilization policy, fiscal disorder, and misguided monetary policy. Exchange rate movements thus represent a barometer of fundamental disturbances. If we want to stabilize exchange rates, we have to emphasize national responsibility; each country targets monetary and fiscal stability at home. For instance, a country expands its money supply according to the growth of the national production potential if the money demand function exhibits stability. Or it aims for a stable price level with other monetary strategies. Then, the price level remains stable in each country and, as a consequence, exchange rates do not change. A country cannot have two different nominal anchors for monetary policy. A stability-oriented monetary policy alone is insufficient to keep exchange rates stable; fiscal policy and the whole domain of economic policy, including wage setting, also have to be orientated toward stability for exchange rate expectations to be stable. Soft coordination will not do it. Finally, if rules are not binding, and only intentions exist, they will not help to coordinate national policies.
In the implementation of economic policy, the issue of coordination is, first of all, an issue of assignment, that is, the question of which agent or which institution is responsible for which policy target and which policy tool is assigned to which agent (Siebert, 1996). My answer to the assignment problem among countries is for each individual country to keep its own house in order and to maintain a stable and sustainable economic situation at home. This approach should be complemented by some minimum agreement on stability standards, including prudential rules for the financial sector, and mechanisms to internalize externalities caused by unstable behavior of countries via market forces, to shield the overall system against instability and contagion.
With respect to the assignment of the main policy areas to different national or supranational economic agents (such as the ECB), it can, of course, be shown in game theoretical models that coordination of different policy agents can improve welfare. But from a political economy or public choice viewpoint, responsibilities should not be blurred. Whoever is assigned a policy instrument should be responsible for the main target associated with that tool:
Monetary policy instruments are assigned to the central bank; consequently, the central bank should be responsible for a stable money.
Determination of the wage rate is, in most European countries, assigned to the social partners; consequently, they should be responsible for employment and unemployment.
Taxation and spending tax revenue is assigned to the government; consequently, the government should be responsible for the allocation of resources and for economic growth.
The right to define the institutional setting of a market economy—that is, to make the laws—is assigned to parliament; consequently, parliament should be responsible for creating the right institutions with the appropriate incentives and constraints—for instance, for the labor market.
KrugmanPaul1991“Target Zones and Exchange Rate Dynamics,”Quarterly Journal of EconomicsVol. 116 (August) pp. 669–82.
McKinnonRonald I.1982“Currency Substitution and Instability in the World Dollar Standard,”American Economic ReviewVol. 72 (June) pp. 320–33.
SchweickertRainer1999“Stabilizing Exchange Rates in the Triad Asia and Latin America—Chances and Risks” (unpublished; Kiel).
SiebertHorst1996Institutionelle Arrangements für die Zuweisung von Opportunitätskosten. In: Ulrich Immenga/Wernhard Möschel/Dieter Reuter (Hrsg.) Festschrift fur Ernst-Joachim Mestmäcker 309-320 Nomos Verlag Baden-Baden.
SiebertHorst1999“How to Improve the Incentive System to Prevent Currency Crises” paper presented at a symposium on “Reforming the World’s Financial Architecture—The Role of the IMF” at the Kiel Institute of World EconomicsJune.
Compare the 500 percent money market interest rate in Sweden in 1992.
Governments have to anticipate the consequences of a stability-averse policy, which are made explicit by the interplay of interest rate parity and PPP; consequently, governments have to aim at stability. An example is the economic policy situation in France in 1983, when the Mitterand government had to completely reverse its economic policy approach, which had led to budget and balance of payments deficits, a rising price level, and a devaluation of the French franc.
A crawling peg has to be the choice if a stringent stabilization policy, which would be necessary for a constant exchange rate, cannot be expected to be maintained and if the inflation rate in the domestic country is expected to be higher than in the foreign country over a longer period of time.
Compare the problem when monetary policy wants to stabilize asset prices as in the Japanese bubble, or the price of gold as in the United Kingdom in the 1920s (Siebert, 1999).
The currency board is a special form of an exchange-rate-oriented monetary policy. In such an approach, the domestic currency has to be covered completely by foreign currency reserves. The central bank binds itself in the sense that it only provides domestic money to the extent that foreign reserves are available (Argentina since 1991, Estonia since 1992). Countries choosing a currency board use the currency of another country as an anchor, because they cannot provide an anchor themselves. Thus, they import the stability of the anchor currency while renouncing, however, seigniorage from their own money. An important condition is that the country must be able to digest external shocks by price adjustments. This condition is especially relevant if the country is sensitive to external shocks. For instance, a currency board is difficult to sustain if a country is a resource exporter, resource prices are volatile, and labor markets are not flexible.
Dollarization is the extreme form of a currency board. Neglecting the issue of whether the Federal Reserve is willing to act as a lender of last resort for the dollarized country, the country has to buy each dollar by giving away export goods in exchange for dollar bills (seigniorage costs). Dollarization would not cause problems if the country traded only with the United States. If the country trades with other currency areas and if these currencies fluctuate with respect to the U.S. dollar, the dollarized country experiences disturbances in the real sphere of the economy; witness the Asian financial crises where countries were pegged to the U.S. dollar.
Another issue is whether a shock is permanent or transitory.