Abstract
In april 1972, the European Communities (EC) introduced a system of narrower margins of fluctuation for their cross exchange rates known as the “snake.” The maximum possible fluctuation between member currency spot exchange rates was halved from ±4½ per cent to ±2¼ per cent. Wth further reductions and the eventual elimination of margins of fluctuation envisaged, the snake was seen as a stepping stone toward monetary union in the EC. However, actual developments have fallen short of original aspirations. Despite the protection of exchange controls, the snake has been beset on a number of occasions by waves of speculation. Large-scale short-term capital movements forced the withdrawal of the United Kingdom in June 1972, Italy in February 1973, and France in January 1974 and March 1976. The movements of short-term capital have been seen by member governments as being occasionally disruptive in the sense that they have not always been seen as being warranted by underlying economic conditions.
In april 1972, the European Communities (EC) introduced a system of narrower margins of fluctuation for their cross exchange rates known as the “snake.” The maximum possible fluctuation between member currency spot exchange rates was halved from ±4½ per cent to ±2¼ per cent. Wth further reductions and the eventual elimination of margins of fluctuation envisaged, the snake was seen as a stepping stone toward monetary union in the EC. However, actual developments have fallen short of original aspirations. Despite the protection of exchange controls, the snake has been beset on a number of occasions by waves of speculation. Large-scale short-term capital movements forced the withdrawal of the United Kingdom in June 1972, Italy in February 1973, and France in January 1974 and March 1976. The movements of short-term capital have been seen by member governments as being occasionally disruptive in the sense that they have not always been seen as being warranted by underlying economic conditions.
No intervention arrangements for maintaining internally fixed exchange rates can be expected to survive permanently divergent rates of monetary expansion in member countries. However, the achievement of monetary union would be facilitated by intervention arrangements that successfully combat disruptive capital flows without the need for controls on intra-EC capital movements. This paper examines a reform of the snake that would fulfill this requirement. After describing the current mechanism of the snake, the paper describes and compares an alternative arrangement—a “forward snake.” Under this alternative intervention mechanism, official support would be switched from the spot to the forward market; spot exchange rates would be free; and partner currencies purchased (sold) through forward support operations would, on maturity, be sold (purchased) in the free spot market.1
Pegged forward exchange rates will be shown to be advantageous in a number of ways:
(1) Unlike spot market speculation, forward market speculation involves no movement of funds; exchange controls to restrain speculation in the officially supported exchange markets would not, therefore, interfere with the free flow of capital within the EC.
(2) During the period before maturity of official forward contracts, both strong and weak countries would be encouraged to bring interest rates into line with those in other member countries in order to avoid losses associated with exchange transactions; that is, a harmonization of monetary policies would be encouraged.
(3) The interests of traders are, in general, better served by stability of forward rather than of spot exchange rates.
Free spot exchange rates will be shown to add additional advantages:
(1) The freedom for the authorities to undertake discretionary purchases and sales of partner currencies in the spot markets would remove the need for financing and settlements between member countries.
(2) Reserves and money stocks would be unaffected by the whims of transactors in short-term capital.
(3) Speculation would be discouraged by greater risk associated with the stochastic element in the profitability equation of speculators—the expected spot rate.
The structure of this paper is as follows. Section I describes the current operation of the snake; Section II describes the operation of a forward snake. Section III compares the current snake and the forward snake according to their impact on harmonization of monetary policy, speculation and the prospects for removal of margins of exchange rate fluctuation, the need for exchange controls that impede the free flow of capital within the EC, and the well-being of traders. The conclusion presents a possible scenario for monetary union within the EC following a transition from the current snake to the forward snake.
I. The Current Intervention Arrangements
The main features of the current intervention arrangements are as follows: (1) Each participating member has an unlimited obligation to intervene in Community currencies to maintain its spot exchange rates within ±2¼ per cent of parity. (2) Debts and claims in partner currencies arising from intervention are financed and settled in a prescribed way. (3) Although reserves are temporarily unaffected by intervention, money stocks are affected immediately.
(1) Between all pairs of participating currencies, there exists a central or parity level for the spot exchange rate. For example, between the deutsche mark and the Netherlands guilder it is DM 1 = f. 1.04202 or, equivalently, f. 1 = DM 0.959675.2 When an actual exchange rate differs from parity by 2¼ per cent, simultaneously the strong currency is at the ceiling of the snake and the weak currency is on the floor. For example, the ceiling for the deutsche mark in terms of guilders is DM 1 = f. 1.0657; this is identical to the floor for the Netherlands guilder in terms of deutsche mark of f. 1 = DM 0.93835. To maintain the limits of the snake, the authorities of both the strong and weak currencies are required (a) to post the ceiling/floor exchange rate as the rate at which they will buy the weak currency with the strong currency; and (b) to accept all orders forthcoming. The resulting magnitude of intervention that is called forth from each country will be equal only by chance. Although it would be possible to maintain a joint float against the U. S. dollar while intervening in dollars, intervention is confined to the Community currencies that are at the limits of the snake.3 Since both the weak and strong countries intervene in the same direction at the limits of the snake, there is no danger that such intervention will be counteractive; no concertation is therefore required prior to intervention at the limits of the snake. However, because of the danger that simultaneous intervention within the limits could be counteractive, concertation is required prior to such intervention.
According to the Basle agreement of April 1972, member countries are restricted in their holdings of partner currencies to the level of working balances. As a result, a system of financing and settlement exists. Since the weak country will normally not hold the amount of the strong currency required for intervention, it has the right to borrow the requisite amount from the strong country. There is no limit to the amount of such financing that must be made available. The strong country will not require any financing, since it will be selling its domestic currency. Following a support operation, the weak country will hold liabilities denominated in the strong currency while the strong country will hold claims denominated in the weak currency. These liabilities and claims are cleared through the European Monetary Cooperation Fund. To limit exchange risk, they are denominated in European monetary units of account—a basket of participating member currencies. They bear an interest rate, rounded down to the nearest ¼ per cent, equal to the average discount rate in participating members of the snake. Settlement normally takes place at the end of the month following intervention. However, postponement of settlement is permitted for three months for amounts up to agreed quotas. Settlement is made with reserve assets; the composition of the reserve assets used in settlement is subject to the constraint that the ratio of foreign exchange to the sum of special drawing rights (SDRs) and the reserve position in the International Monetary Fund (IMF) equals the ratio in the debtor’s reserves.4
(3) During the period prior to settlement, the reserves are unaffected by intervention.5 As well as this temporary insulation of the reserves, there is an economy in the use of reserves to the extent that currencies move from one extreme to another in the snake, and debts and liabilities are cleared multilaterally within the European Monetary Cooperation Fund. When actual reserve settlements take place, a tendency toward increasing similarity in the composition of member country reserves is brought about by the afore-mentioned constraint on reserve settlement. Although reserves are temporarily insulated, money stocks are immediately affected by intervention. In the absence of sterilization, base money in the weak country will decrease by the total of all official purchases of the weak currency in the foreign exchange market, that is, by the combined intervention by the weak and strong countries. Similarly, base money in the strong country will increase by a corresponding amount.
II. The Intervention Arrangements of a Forward Snake
The operation of a forward snake would have the following characteristics: (1) Each participating country would have an unlimited obligation to maintain its forward exchange rates of a particular maturity within prescribed limits. (2) Partner currencies purchased (sold) in the forward market would, on maturity, be sold (purchased) in a free spot exchange market. (3) Reserves would be fully insulated; money stocks would be insulated except for any gains or losses on exchange transactions.
(1) Under a forward snake, forward exchange rates of a particular maturity would be maintained within prescribed limits. Judging by the current thicknesses of forward markets, the greatest benefit to traders and capital transactors would probably result if the maturity period of the supported forward markets were three months.6 The limits of the forward snake would be maintained by the requirement that each country post the ceiling forward rates for its currency against partner currencies as the prices at which it will sell its domestic currency forward, and the floor rates as the prices at which it will buy its domestic currency forward. Official forward contracts would be made in the weak and strong currencies only, and intervention would be without limit. The authorities of both the strong and weak currencies would be intervening simultaneously to maintain the limits of the snake, although the magnitude of intervention called forth from each would not necessarily be equal.
(2) Since forward contracts do not immediately involve movements of funds, the weak country would not require financing prior to intervention. Immediately before maturity, the weak country would purchase in the spot market the amount of the strong currency required to meet its maturing forward contracts. Similarly, the strong country would sell in the spot market the weak currency acquired through its forward contracts. Spot operations would be made in the weak and strong currencies only. Under these arrangements, no reserve settlement would be required. Rather, the official spot sales of weak currency and purchases of strong currency would induce a private capital flow from the strong to the weak currency.7 Since the reserves would be fully insulated, the induced capital flow would necessarily be such as to ensure overall payments balance in both the weak country and the strong country. The actual mechanism whereby these stabilizing capital flows would be induced is as follows. The official spot operations would tend to depreciate the spot rate of the weak currency in terms of the strong currency. This would increase the forward premium on the weak currency. The forward premium is an implicit interest rate that must be added to the actual interest rate to obtain the total return from investing in the weak currency on a covered basis. Whenever the forward premium on the weak currency was greater than the interest rate differential in favor of the strong currency, a profit could be earned by the covered interest arbitrage operation of borrowing the strong currency and investing on a covered basis in the weak currency. Provided that there was considered to be no risk of default, the profit would be riskless; in the absence of exchange controls and banking regulations that prohibited or penalized resident lending to or borrowing from nonresidents, only a minimal depreciation of the spot rate of the weak currency would be required to induce stabilizing covered interest arbitrage flows from the strong to the weak currency. Good illustrations of how interest parity could be expected to be maintained in the absence of such exchange controls and banking regulations are provided, first, by the deutsche mark/Euro-dollar covered interest differential, which has remained in the vicinity of zero since the relaxation of controls on capital inflows in the Federal Republic of Germany in early 1974, and, second, by the normal experience of the Euro-currency market. In this latter context, a statement by Prissert is particularly significant: “… there is no example—this is a regular observation of foreign exchange dealers—of any single bank announcing, at the same time on the market, swap rates and Euro-rates which are not exactly and precisely in line with each other.” 8
(3) Reserves would be fully insulated, since all partner currencies purchased (sold) in the forward market would, on maturity, be sold (purchased) in the spot market. However, both the strong and weak countries would make gains or losses on exchange transactions whenever the forward rate was not equal to the spot rate prevailing on maturity. If the weak currency were to be dearer in the spot market than it was in the forward market, the weak country would need to sell less of its domestic currency in the spot market than it had purchased in the forward market in order that its purchases and sales of the strong partner currency should be equal; the strong country would buy more of its strong domestic currency in the spot market than it had sold in the forward market in order that its sales and purchases of the weak partner currency should be equal. The stock of base money in each country would decrease by the amount of the gain made by each country.9 Similarly, if the weak currency were to be cheaper in the spot market than it was in the forward market, losses would be made on exchange transactions by both the weak and strong countries, resulting in increases in base money. Assuming that there was no risk of default on forward contracts and that forward exchange rates remained unchanged, the difference between the forward rate and the spot rate prevailing on maturity would equal the interest rate differential prevailing on maturity. The gains on exchange transactions would occur whenever the interest rate on the weak currency prevailing on maturity was higher than the interest rate on the strong currency; losses would occur whenever the interest rate on the weak currency was lower than the interest rate on the strong currency.
III. The Current Snake Versus the Forward Snake
incentive toward harmonization of monetary policy
Under the current snake, it could be hoped that divergent rates of domestic credit expansion would result in reserve movements that would, first, bring conformity in rates of monetary expansion and, second, provide an incentive for the authorities to harmonize their rates of domestic credit expansion. Under the forward snake, reserve movements are eliminated, and therefore do not in themselves provide an incentive for harmonization of monetary policy. However, an incentive does occur through the possibility of losses on exchange transactions. Whenever the rate of interest on the weak currency is lower than the rate of interest on the strong currency, both countries are likely to incur losses on exchange transactions.
Expansionary monetary policy may be associated with either a relatively low rate of interest or a relatively high rate of interest. The incentives to harmonize monetary policy under the current and forward snakes will be compared under each of these alternatives.
Expansionary monetary policy associated with a relatively low rate of interest
If member countries began with similar rates of inflation and unemployment, a faster (slower) rate of growth in domestic credit in one country would normally result, at least initially, in a lower (higher) rate of interest than in partner countries. Under both the current and forward snakes, the currency of the expansionary country would be likely to weaken in the supported exchange market while the currency of the contractionary country would be likely to strengthen. Under the current snake, a flow of reserves from the expansionary to the contractionary country would occur; under the forward snake, losses on exchange transactions would occur. The resulting incentive for harmonization of monetary policy could be greater under the forward snake for the following reasons:
(1) Under the current snake, the need to protect the reserves is an incentive for the weak country to curtail its rate of domestic credit expansion. However, there is no incentive for the strong country to increase its rate of domestic credit expansion. On the contrary, the strong country may respond by further reducing its rate of domestic credit expansion in order to sterilize the impact of the reserve inflow. In contrast, there would be an incentive for both countries to adjust under the forward snake. For as long as interest rates in the two countries remained out of line, both countries would incur losses on exchange transactions. To avoid such losses, the weak country would be encouraged to reduce its rate of domestic credit expansion while the strong country would be encouraged to increase its rate of domestic credit expansion.10
(2) Although the losses on exchange transactions that would occur under the forward snake would be small in relation to the changes in reserves that would occur under the current snake, it is likely that they would be more embarrassing politically.11 They would probably be explicitly identified in the balance of payments accounts as “losses.” Many would consider them to be real losses that should be financed out of taxation. In contrast, it is unlikely that reserve losses are seen as real losses, since they are associated with the net importation of goods or financial assets.
In the same way that losses on exchange transactions would be considered undesirable under the forward snake, gains on exchange transactions would be welcomed. To reap such gains would require a higher rate of interest in the weak country than in the strong country—a desirable situation for the continued maintenance of fixed exchange rates. To the extent that an intrinsic discount (premium) existed on the weak (strong) currency owing to a conjectured risk of default on forward contracts, avoiding losses and reaping gains would require an even higher rate of interest in the weak country relative to the strong country.
(3) Under the forward snake, losses on exchange transactions would occur only when interest rates were out of line (in the absence of default risk on forward contracts). After the inception of a forward snake, there would be an initial period before maturity of official forward contracts during which, if interest rates remained out of line, the approximate magnitude of the forthcoming loss would be known. The fact that a loss would be impending for, say, three months would have a salutary effect. Once interest rates had been brought into line, there would be an incentive to so maintain them.
Under the current snake, reserve changes occur owing to random and cyclical factors as well as interest rate differentials. It is occasionally difficult to tell whether reserve changes are caused by divergent monetary policy or by temporary phenomena that should not induce monetary policy changes. Owing to this uncertainty, the policy response to reserve losses may be delayed. In fact, one reason why delays in settlement are permitted under the current arrangement is the prospect that the direction of intervention may be reversed and reserve liabilities and claims may be cleared within the European Monetary Cooperation Fund. To the extent that small reserve changes are seen as normal occurrences, they may fail to provide an incentive toward policy harmonization. To the extent that large reserve changes can undermine confidence in the currency and may trigger a currency run, monetary policy changes alone may be insufficient to correct them; exchange controls may be the preferred alternative.
If monetary policy harmonization were achieved under the forward snake, that is, if interest rates in the strong and weak countries were equal, official gains and losses on exchange transactions would be eliminated. However, under similar circumstances under the current snake, reserve movements would not necessarily be eliminated. Identical interest rates in the strong and the weak countries would mean that there would be no interest rate cost associated with speculation. Disruptive capital flows could still affect the reserves and could necessitate policy responses that were inimical to the pursuit of monetary union.
(4) Under the forward snake, the loss on exchange transactions would steadily increase if divergent monetary policy resulted in overall payments balance being maintained only through a continuing flow of interest arbitrage funds from the strong to the weak country. A continuing flow would give rise to an increasing stock of arbitrage funds in the weak country. Since this increasing stock would be regularly rolled over, the magnitude of official intervention in the forward market and the resulting loss on exchange transactions would steadily increase. Owing to the need to take account of the possibility of an ever-increasing loss on exchange transactions, a decision to maintain a divergent monetary policy would be less likely.
Expansionary monetary policy associated with a relatively high rate of interest
A relatively high rate of domestic credit expansion need not be associated, even initially, with a low rate of interest. Expansionary monetary policy may simply be accommodative. For example, it may be designed to curb the increase in the rate of interest that would otherwise accompany an expansionary fiscal policy. Whatever is the initial impact, expansionary monetary policy that is not fully corrected by either policy action or an outflow of reserves is likely to lead eventually to higher inflation and lower unemployment, and to be reflected in a higher rate of interest.
Under the current snake, a high rate of interest accompanying an expansionary monetary policy is likely to lead to an improvement in the capital account for as long as there is considered to be no risk of a parity change. The improvement in the capital account may be sufficient to prevent any reserve loss. Any attempt to curb the rate of monetary expansion will, other things being equal, initially increase the rate of interest even more and lead to further capital inflows that may frustrate the attempt to curb the rate of monetary expansion. Thus, for as long as parity changes are not expected, reserve movements may fail to bring rates of monetary expansion into line and may even frustrate appropriate monetary policy. As soon as parity changes are expected, large-scale capital outflows from the high-inflation country are likely. Rather than performing the role of bringing rates of monetary expansion into line and enabling fixed exchange rates to be maintained, such capital flows may force withdrawals from the snake or parity changes within the snake.
Under the forward snake, a higher rate of interest on the weak currency than on the strong currency would be likely to give rise to official gains on exchange transactions for both the weak and the strong country, provided that forward parities remained unchanged. Any attempt to reduce the rate of monetary expansion in the weak country or to increase the rate in the strong country would, other things being equal, initially increase the official gains on exchange transactions. In contrast to the current snake, there would be no danger that appropriate changes in monetary policy would be frustrated by reserve movements.
If parity changes took place under the forward snake, both the weak and the strong countries would be likely to incur large losses on exchange transactions.12 As a result, parity changes would be resisted. Such resistance would be a further encouragement to harmonization of economic policies. Countries would be less prone to think that they could, if necessary, use a future exchange rate change to accommodate a currently divergent economic policy. Parity changes are not conducive to the achievement or maintenance of uniform rates of inflation. A high-inflation country would be likely to find greater difficulty in curbing its rate of inflation following a devaluation; inflationary expectations are likely to be worsened by a devaluation. Although the supply of real money balances would be reduced by a devaluation, such a reduction could be achieved by appropriate monetary policy in the absence of devaluation. For countries that are trying to move toward a monetary union, exchange rate adjustment should not be seen as a policy alternative.
A transition from the current to the forward snake should not be criticized on the grounds that countries would inevitably suffer losses; parity changes and interest rate differentials give rise to similar gains and losses under the current snake, although this is seldom recognized.13 Rather, it is an advantage of the forward snake that the official loss would be explicitly identified and that this would provide a powerful stimulus toward policy harmonization within the Communities and a deterrent to parity changes.
speculation and the prospects for removal of margins of fluctuation
It is likely that speculation would be less of a problem under the forward snake than under the spot snake for the following reasons: (1) Reserves and money supplies would not be affected. (2) Speculation would be riskier. (3) There would be no constraint on the reduction and elimination of the permitted margin of fluctuation of the supported exchange rate.
(1) When a country is at the limits of the current snake, the reserves and money supply are affected by speculation occurring in both the spot and forward exchange markets. Speculation occurring in the spot market directly affects the money supply and the net official reserves; speculation occurring in the forward market is transmitted through interest arbitrage to the spot market and indirectly affects the reserves and the money supply. Under the forward snake, neither speculation occurring in the forward market nor speculation occurring in the spot market would affect the reserves or money supply. The official support of the forward rate would ensure that speculation in the forward market would be maintained in the forward market and would induce no movement of spot funds. The free spot rate would ensure that speculation in the spot market would induce offsetting interest arbitrage flows; a speculative capital inflow (outflow) would simply tend to appreciate (depreciate) the free spot rate and, with the forward rate and interest rates unchanged, induce an offsetting interest arbitrage outflow (inflow). Since interest arbitrage is essentially risk- less, only minimal changes in the free spot rate would be induced.
Given that speculation would not affect the reserves or money supply under the forward snake, there would be no danger that disruptive capital flows would compel unwanted devaluations to protect reserves or unwanted revaluations to protect money supplies.
Under the current snake, official action can obviously be undertaken in an attempt to offset the impact of speculation. In particular, changes in domestic interest rates could be so used. However, small changes in interest rates would not necessarily combat speculative capital flows under the current snake in the same way that small changes in the free spot exchange rate would combat speculative capital flows under the forward snake. When only spot exchange rates are supported, interest rate changes influence uncovered, not covered, capital flows. In the absence of official forward intervention, capital transactors in aggregate cannot move funds on a covered basis; individual capital transactors are able to obtain cover only to the extent that they are able to induce other individuals to bear the exchange risk. Because uncovered capital flows are risky whereas covered capital flows are essentially riskless, significant changes in interest rates might be necessary under the current snake to combat speculative capital flows, whereas only small changes in the free spot rate are required under the forward snake.
An alternative means of offsetting the impact of speculation under the current snake would be discretionary intervention in the forward market. The traditional policy for a currency under attack is to purchase the domestic currency forward in the hope of inducing stabilizing arbitrage inflows. However, since this would tend to appreciate the forward rate of the weak currency, speculation against the currency may be encouraged and the trade balance may worsen. In contrast, discretionary intervention in the spot market by a weak country under the forward snake would tend to depreciate the spot rate. As well as inducing stabilizing arbitrage inflows, spot market speculation in favor of the currency may be encouraged and the trade balance may improve.
There is another difference between the means for combating speculative capital flows under the current and forward snakes. Under the forward snake, stabilizing capital flows occur automatically; under the current snake, they require policy action. Inevitably, there is a delay before action is taken. In a world in which crises sometimes seem to spring overnight from nowhere, small delays in policy action could mean that a “currency run” may have already gained momentum before action was taken.
(2) The profitability of both spot and forward market speculation depends on only one unknown variable—the future spot rate. The official support given to the spot rate under the current snake reduces the risk involved in speculation. When a currency is being maintained at the floor by official intervention, speculators know that a large depreciation would occur if the official support limits were removed or changed and that a maximum appreciation of 4½ per cent could occur if the limits were maintained. A loss of 4½ per cent would obviously be significant. However, the fact that official intervention is being needed to prevent the currency slipping through the floor means that a given strengthening of the currency would be required before the currency could even begin to move off the floor. Hence, the risk of an appreciation in the rate is reduced. The possibility of a large depreciation or, at most, a small appreciation of the currency gives rise to the well-known “one-way options” enjoyed by speculators when spot rates are pegged.
Under the forward snake, the spot rate is not officially supported. Besides having no official boundary to the possible future spot rate, there would be no stickiness to the spot rate caused by official intervention; spot rates would respond immediately and fully to interest rate changes.
(3) If the margins of fluctuation of the current snake were to be reduced, currencies would become overvalued sooner than otherwise because of the reduced scope of equilibrating exchange rate movements, and the possible appreciation of weak currencies would be lowered. At the extreme when no fluctuation in the spot rate was allowed, there would be no risk associated with speculation (except the unlikely possibility that weak currencies would be revalued rather than devalued). Speculation could be expected to occur whenever there was the slightest possibility of a devaluation, and an immediate snowballing effect would be likely. Speculation would feed on itself in the anticipation that the very magnitude of speculation would force a parity change.
Under the forward snake, the margins of fluctuation of the forward rate could be completely eliminated without undue problems from speculation. Although riskless speculation against the future forward rate could occur in longer nonsupported forward markets, this would involve no movement of funds and would simply place a constraint on the term structure of interest rates.14 Speculation against the future spot rate occurring in the supported forward market would remain risky. Assuming that the forward parity was maintained, the future spot rate would be determined by the future interest rate differential. After speculative positions against the currency had been taken out, the authorities would have the power to penalize speculators by changing interest rates. An increase in the rate of interest on the weak currency would appreciate the spot rate of the weak currency by a corresponding amount; a decrease in the rate of interest on the strong currency would depreciate the spot rate of the strong currency. A pronouncement or an occasional policy action to this effect would be likely to have a salutary effect. It is unlikely that there would be any snowballing effect. The greater the magnitude of speculation, the greater would be the official loss that would be incurred if parities were changed. Large-scale speculation would be expected to induce penalizing interest rate changes rather than profit-giving parity changes.
the need for exchange controls that impede the free flow of capital within the EC
In the first year of the snake’s operation, a barrage of exchange controls was introduced to defend the snake from speculative capital movements. One reason why the controls were needed was an inappropriate alignment of some currencies stemming from the December 1971 Smithsonian agreement. However, another reason was simply that the narrower margins of fluctuation in the snake encouraged speculation. The current snake appears to be beset by a dilemma: Two characteristics of a monetary union are rigidly pegged exchange rates and freedom for capital movements; however, the more rigidly spot exchange rates are pegged, the greater is likely to be the problem of speculative capital movements and the greater will be the need for exchange controls that interfere with the free flow of capital. Alternatively, the greater the freedom of capital movements within the Communities, the greater is likely to be the need for increased exchange rate flexibility.
Under the forward snake, there would be no need for exchange controls that interfered with the free flow of capital within the EC. The flexibility in the spot exchange rate would achieve the desired result. This flexibility, however, would not prevent investment in any country by any individuals wishing to invest. Such investment would be accommodated by offsetting interest arbitrage flows by banks. An individual buying a currency for investment would tend to appreciate the spot rate of the currency; the appreciation of the spot rate would reduce the forward premium on the currency and induce an offsetting interest arbitrage outflow. Equivalently, the individual would buy the currency spot from a bank, which would simply borrow the currency and then cover its shortened position in the currency by buying the currency forward.
Although it is unlikely that speculation in the supported forward market would be a serious problem, exchange controls could be used to restrain it without interfering with the free flow of capital. Exchange controls can deter speculation occurring in the officially supported market without constraining the free flow of capital only if it is the forward rate that is supported. Already, under the current snake, controls exist to deter purely speculative forward operations in order to prevent flows of funds through the supported spot markets from being induced. At the transition from the current snake to the forward snake, the controls on forward contracts could be maintained while the controls on spot contracts could be removed. This would bring a significant saving in administrative costs owing to the large number of small-denomination spot contracts that would no longer need to be monitored.
the effect on trade
Before a trade payment is actually made, two types of lag normally exist: first, the search and negotiation period before exchange of contracts and, second, the period between exchange of contracts and actual payment (which itself incorporates the production and transportation lag before delivery, and the period of credit following delivery). In comparison with the spot snake, the forward snake would, in general, reduce the exchange rate risk associated with each of these two lags. First, stability of forward exchange rates would reduce the risk that forward exchange rates would change during the search and negotiation period, and would therefore reduce the risk that plans would be made on the basis of incorrect price signals. Second, it is likely that official support of a forward rate with a particular maturity rather than of the spot rate would lengthen the availability of forward facilities and so increase the likelihood that long lags between exchange of contracts and eventual payment could be covered in the forward market; whichever forward rate was supported, it is likely that banks would be willing to utilize their ability to cover in the supported forward market to offer longer-period forward facilities to their customers.
Many current account transactions currently involve payments through the spot market. Some such transactions would obviously not benefit from a switch from the current snake to a forward snake. However, it is unlikely that the well-being of traders would be significantly hurt by the removal of official support for the spot rate. Many transactions through the spot market are either known about in advance and would benefit from having access to forward facilities—for example, household purchases of foreign exchange for tourism—or occur without notice and would therefore not be affected by the future or past stability of the spot exchange rate.
IV. Conclusion: A Scenario for Monetary Union
This paper has examined a reform of the European currency snake in which, first, official support is shifted from the spot market to the forward market and, second, the procedure of financing and settlement is replaced by official purchases and sales of partner currencies in a free spot market. The following is a possible scenario for the contribution of such a forward snake to monetary union in the EC.
At the transition from the current snake to the forward snake, previously nonparticipating member countries would be encouraged to join by the knowledge that their spot exchange rates would remain officially unsupported, thus preventing disruptive capital flows from having adverse effects on reserves and money stocks. Following the introduction of the forward snake and the freeing of spot exchange rates, exchange controls on the flow of funds between member countries would be unnecessary and would be removed; exchange controls would still be applied to prevent purely speculative forward operations. Spurred on by the movement toward monetary union resulting from the freeing of intra-Community capital flows, the movement would be continued by the elimination of any permitted margin of fluctuation in forward exchange rates; the freedom of the spot exchange rates would continue to deter or offset speculation and random disturbances.
Throughout this time, countries would be encouraged to follow appropriate monetary policies by the threat of embarrassing losses on exchange transactions and the reward of official profits. As monetary policy became more harmonized within the Communities, interest rate differentials would diminish and spot exchange rates, although still free, would become increasingly tied to the pegged forward rates. When economic policy was finally conducted on a Community, rather than national, level and the free spot exchange rates departed only minimally from the fixed forward rates, the Communities could offer fixed and equal buying and selling rates for both spot and forward contracts. It would then be only a small step to the introduction of a European currency.
Mr. Day, an economist in the Central European Division of the European Department when this paper was prepared, is currently in the Financial Studies Division of the Research Department. He is an external graduate of the University of London and received his doctorate from the University of Birmingham. Before joining the Fund staff, he taught at the University of Manchester.
Outside the context of the snake, a theoretical foundation for the policy proposal is provided in William H. L. Day, “The Advantages of Exclusive Forward Exchange Rate Support,” Staff Papers, Vol. 23 (March 1976), pp. 137–63; “Dual Exchange Markets Versus Exclusive Forward Exchange Rate Support,” Staff Papers, Vol. 23 (July 1976), pp. 349–74; “Domestic Interest Rates Under a Pegged Forward Exchange Rate” (unpublished, International Monetary Fund, June 10, 1975); and “Destabilizing Capital Flows: An Exchange Market Policy Proposal” (doctoral thesis, University of Birmingham, September 1973).
On October 18, 1976 [after this paper was written], the deutsche mark was revalued by 2 per cent against the Netherlands guilder.
It is not strictly correct to state that “… a Community level for the dollar is an essential prerequisite of any system of coordinated intervention in dollars” (“The European System of Narrower Exchange Rate Margins,” Deutsche Bundesbank, Monthly Report, Vol. 28 (January 1976), p. 23). Maintaining a joint float against the dollar requires only that the combined sum of member official spot dollar purchases equals zero. The limits of the snake could be maintained while floating against the dollar by the authorities of the weak currency selling dollars and purchasing its weak domestic currency, and the authorities of the strong currency buying an equal amount of dollars with its strong domestic currency. Concertation would be required concerning the magnitude of intervention, but there would be no settlements problem. See Joanne Salop, “Dollar Intervention Within the Snake” (unpublished, International Monetary Fund, August 13, 1976).
Until the return of the French franc to the snake in July 1975, settlement was subject to the constraint that the ratio of gold, SDRs, and the IMF reserve position to foreign exchange equaled the ratio in the debtor’s reserves. However, gold was not actually used in settlement.
This assumes that claims and liabilities of the European Monetary Cooperation Fund are not included in the definition of the reserves. This is not true, for example, in the Federal Republic of Germany.
The advantages from supporting forward markets of alternative maturities are described in Day’s article in the March 1976 issue of Staff Papers (cited in footnote 1), pp. 155–58. Since arbitrageurs would roll over their positions at intervals equal to the maturity period of the supported forward rate, it would become increasingly costly both administratively and as far as exchange transactions losses were concerned to operate the system on forward rates of shorter maturity.
The policy proposal cannot be put into reverse by supporting spot exchange rates and selling any acquired currency in the forward market. This would involve forward intervention in the opposite direction to conventional forward intervention policy. It would tend to reduce (increase) the forward premium on the weak (strong) currency and induce destabilizing interest arbitrage capital flows. A comparison between conventional forward intervention policy under a supported spot rate and discretionary spot intervention under a supported forward rate is made in Day, ibid., pp. 158–61.
Pierre Prissert, A Critical Re-Examination of the Forward Exchange Theory, reprinted by Société Universitaire Européenne de Recherches Financières (1974), p. 16.
These base money changes would be, at most, a small percentage of the magnitude of intervention of the particular country. This is in contrast to the current arrangements where base money changes equal the combined magnitude of intervention by both countries.
Under the forward snake intervention arrangement described in this paper, the authorities of both the strong and weak currencies would incur losses simultaneously and would be induced, respectively, to loosen and to tighten monetary policy. An alternative arrangement to that described in this paper would be to maintain the limits of the forward snake by requiring intervention by the authorities of the weak currency only. Such an arrangement would be more deflationary, since there would be no inducement for looser monetary policy in the strong currency. To the extent that excessive monetary expansion in isolated member countries is likely to be more common than excessive monetary tightness in isolated countries, such an arrangement may be preferable economically. It is not proposed in this paper, however, because, first, it would represent a further departure from the current snake arrangements and, second, it would be farther from the spirit of Community policy and action than combined intervention would be.
The losses on maturing forward contracts made by the Bank of England following the 1967 devaluation of sterling are here a good example.
Obviously, a strong country would not like to see a devaluation of a weak currency in which it held a long position. This is precisely why there is an incentive for the strong country as well as the weak country to adjust through appropriate changes in demand management rather than through exchange rate changes. If the forward snake were seen as a system of adjustable par values rather than as a precursor of monetary union, agreement might be possible only if the limits of the snake were to be maintained by intervention by the weak countries only. It is the author’s view that it would be preferable to go all the way to a forward snake with combined intervention by both strong and weak countries for those countries that were determined to maintain the initial parities rather than to go halfway to a forward snake with intervention by weak countries only in which countries reserved the right to use the exchange rate as an instrument of policy.
Under both the current and forward snakes, the official open foreign currency position is shortened by trade deficits and speculation against the currency. If the same amount of speculation and the same trade deficits occurred under the current and forward snakes, the reduction in the official spot foreign currency holdings under the current snake would equal the official forward commitment to supply foreign currency under the forward snake. In the event of a given devaluation, the loss on forward commitments under the forward snake would equal the smaller reserve revaluation profit that would be incurred under the current snake.
In the absence of parity changes, losses on exchange transactions under the forward snake would have been matched by similar losses under the current snake. Under a pegged spot rate, excess demand for foreign currency in the forward market is transmitted to the supported spot market by banks covering their open positions, that is, by passive interest arbitrage. As a result, both the official reserves and the official domestic currency financial requirement decrease. Interest receipts on official foreign exchange holdings and interest payments on the national debt both decrease. An implicit official loss is incurred equal to the interest differential against the domestic currency multiplied by the excess demand in the forward market. This loss is approximately equal to the forward premium on the currency multiplied by the excess demand in the forward market, that is, approximately equal to the loss that would have been incurred if official support had been given exclusively to the forward rate with an unchanged interest differential and an unchanged magnitude of excess demand in the forward market.
See Day, “Domestic Interest Rates Under a Pegged Forward Exchange Rate” (cited in footnote 1).