Chapter

Chapter 16: The Optimal Exchange Rate Regime in the OECS/ECCU: A Medium-Term View

Author(s):
Alfred Schipke, Aliona Cebotari, and Nita Thacker
Published Date:
April 2013
Share
  • ShareShare
Show Summary Details
Author(s)
Sebastian Acevedo and Chris Walker 

For more than 35 years, the countries comprising the Eastern Caribbean Economic and Currency Union (OECS/ECCU) have maintained a fixed exchange rate at 2.7 Eastern Caribbean dollars per U.S. dollar. The fixed rate has been administered through a quasi–currency board arrangement, managed first by the Eastern Caribbean Currency Authority, and following its establishment in 1983, by the Eastern Caribbean Central Bank.1 The exchange rate regime has been a stabilizing factor for the economies of the region (see, e.g., Pineda, Cashin, and Sun, 2009), and contributed to the development of the financial sector.

For small or micro states such as those in the OECS/ECCU, the prevailing view is that a fixed exchange rate is the most practical basis for monetary and exchange rate policy. Imam (2010), for example, points out that the economic structure of microstates2 (a category that includes all of the OECS/ECCU countries) tends to be best suited to hard peg regimes such as dollarization, currency boards, or fixed exchange rates. Flexible exchange rate regimes are seen to have major disadvantages for microstates, among them the need for a costly central bank and high exchange rate volatility. Also, for small states the theoretical advantages of a flexible regime, including the possibility of maintaining an independent monetary policy, may offer no more than limited benefits. Some recent experiences, such as that of Seychelles (see below), however, suggest that maintaining an effective flexible exchange rate regime may not be as difficult for a small state as previously believed.

Given that the structure of the OECS/ECCU economies changes over time, it is useful to reassess, on occasion, which exchange rate regime would be the most appropriate in the medium term. This assessment can be made on the basis of empirical comparisons between the OECS/ECCU and neighboring Caribbean countries with respect to inflation, interest rates, bank balance sheets, and other factors, as well as economic integration with the United States and trade relations. The focus is on the long term, and short-term considerations, such as the level of the exchange rate, are not addressed.3

This chapter seeks to determine the optimal exchange rate regime for the OECS/ECCU countries, given that they are small, open, island economies highly dependent on tourism. The chapter reviews the theoretical advantages and disadvantages of different exchange rate regimes for small countries, looking to extract lessons for the OECS/ECCU. The next section briefly discusses the structure of the OECS/ECCU, and is followed by a section that provides background on relevant international experiences with exchange rate regimes. The different regime options (dollarization, currency board, and free float) and their implications are then explored, followed by a section that focuses on the optimal choice for the OECS/ECCU, based on a standard methodology. The final section discusses the policy implications and concludes.

The Eastern Caribbean Economic and Currency Union

The eight island states that make up the OECS/ECCU are all former British colonies (two are still territories), with populations ranging from 56,000 (St. Kitts and Nevis) to 166,000 (Saint Lucia). The total OECS/ECCU population is only 600,000. Per capita income levels vary considerably, from US$6,400 per year (St. Vincent and the Grenadines) to US$13,500 (Antigua and Barbuda). The member countries are small open economies, vulnerable to external shocks and natural disasters. They have a long history of boom-and-bust cycles, beginning with sugar and other agricultural products in the colonial period, and carrying through to tourism and tourism-related investments since independence in the 1980s. Most of the union’s members have high ratios of public debt to GDP, and several have experienced significant fiscal or banking problems. The Eastern Caribbean Central Bank (ECCB) conducts monetary policy for the union, overseeing a common pool of members’ foreign exchange reserves and maintaining the full convertibility of the currency.

Relevant International Experiences

Imam (2010) highlights that the economic structure of microstates favors the adoption of hard pegs (dollarization, currency board arrangements [CBAs], or other fixed exchange rate arrangements). Within the class of hard peg regimes, he concludes that the relative advantage of one particular regime over another depends on country-specific factors. Estimating a multinomial probit model to determine the factors that affect the decision to adopt a particular exchange rate regime, Imam (2010) finds that countries with CBAs usually are former British colonies that gained independence more recently and that have larger governments compared with countries with fixed exchange rates but no currency board. Dollarized countries are likely to have gained independence earlier, are smaller, less likely to be open, and more likely to have balanced budgets.

Although it is too soon to draw general conclusions, the recent experience of Seychelles offers an interesting case showing that a free floating exchange rate regime might be compatible with the characteristics of a small island economy. After having relied on a fixed exchange rate pegged to a basket of currencies from 1967 through 2008, Seychelles, a country with a population of 82,000, was faced with a balance of payments crisis, caused in part by the 2007–08 petroleum and food price spike. The pressure on the exchange rate led to abandonment of the peg. After an initial large depreciation against the U.S. dollar, the exchange rate stabilized and has been free floating ever since (IMF, 2010).

Exchange Rate Regimes: Advantages and Disadvantages

The following discussion focuses on three types of regimes: outright dollarization, quasi–currency board (i.e., the present regime in the OECS/ECCU), and full float.4

Dollarization

Official dollarization generally means the use of another country’s currency—most commonly, but not necessarily, the U.S. dollar—in lieu of a national currency.5 For a small state, dollarization offers the opportunity to import the credibility of a larger trading partner’s currency. A state without its own currency cannot be subjected to a speculative attack. The use of the U.S. dollar or other hard currency is a highly effective bulwark against inflation because it denies the fiscal authority the possibility of monetizing a deficit. Dollarization also eliminates the need for a central bank, thereby helping to minimize administrative costs that may otherwise represent a sizable share of national income for a small country. Official dollarization may also provide conditions conducive to a higher degree of economic integration than is otherwise possible.

There are, however, a number of disadvantages to formal dollarization. Without a national currency, it is not possible to have an independent monetary or exchange rate policy. This may not be a major handicap if the economic cycle of the dollarized economy is closely linked to that of the anchor currency economy, but even economies with strong ties may be subject to idiosyncratic shocks. In the face of a negative shock, an exchange rate depreciation or interest rate reduction may be desirable. The lack of a central bank with the power to create money also precludes the lender-of-last-resort function, which could be important during times of financial stress if local banks encounter a liquidity squeeze. A government without its own currency also gains no seigniorage, which can be an important source of income, even with a currency board (see below). Finally, there is the one-off cost of acquiring dollars to be used in place of the national currency, although this may not be an issue if the monetary authority holds sufficient dollar reserves at the time of dollarization.

Currency Board

Currency board arrangements were introduced in some British colonies in the nineteenth century, and are currently in use in Hong Kong Special Administrative Region (SAR), some Eastern European countries (where they are linked to the euro), and in the OECS/ECCU, among others.6 Under a currency board, the national monetary authority restricts its issuance of base money so that it does not exceed its reserves of the anchor currency.7 With a fixed exchange rate and an open capital account, net outflows of capital lead to reductions in the stock of reserves, while inflows cause an increase. One important consequence of this arrangement is that monetary policy becomes largely passive—as capital flows out, domestic interest rates rise, eventually attracting some capital back in and bringing the economy back into equilibrium.

The currency board offers most of the benefits of dollarization, but with some flexibility. Currency boards are generally viewed as more credible than simple pegs, because base money is covered by foreign reserves and outflows associated with a decline in confidence automatically lead to an increase in domestic interest rates and resulting stabilization. Moreover, the system rules out the possibility of monetizing a large fiscal deficit because that would require foreign financing to provide the necessary increase in the stock of foreign reserves. Also, in contrast to dollarization, the currency board yields some seigniorage. Some limited space for the monetary authority to function as a lender of last resort may also be available, depending on whether it holds any reserves in excess of the monetary base, or for a quasi–currency board such as in the OECS/ECCU, if foreign reserves are in excess of mandatory requirements.8

On the negative side, the currency board may encounter credibility issues that do not arise under dollarization. To the extent that domestic interest rates diverge from those prevailing in the anchor currency’s country, the difference is likely to be undesirably procyclical—interest rates may need to rise further under the currency board than in the anchor currency country in the event of a common shock. Depending on the perceived strength of the authorities’ commitment to the regime, the market may demand a foreign exchange risk premium that could not exist under dollarization. The market may also test the authorities’ commitment to the currency board, as happened in Hong Kong SAR in 1998, or in Argentina in 2002.

Free Float

A freely floating currency offers many advantages and is one of the hallmarks of a financially mature economy, but it may cause difficulties for a smaller state. Under a free float, the central bank is, in principle, able to set the monetary policy most appropriate for the cyclical position of the economy. With a floating currency, seigniorage revenue may be substantial, and the central bank is relatively free to act as the lender of last resort to domestic banks. As distinguished from the currency board or other hard peg, a floating currency is likely to be more resilient to a negative change in sentiment. This resilience occurs because this regime does not allow the accumulation of one-way bets—negative sentiment on the currency will be immediately reflected in exchange rate movements. Moreover, economic actors in a flexible exchange rate regime are generally, although not necessarily, aware of the potential for rapid exchange rate movements, and protect themselves by limiting or hedging dollar liabilities. The possibility of responding to a negative real shock with a depreciation confers a further advantage on a flexible exchange rate in an economy with nominal rigidities.

Among the potential disadvantages to a small state of a freely floating currency, the greatest may be the difficulty of establishing an anti-inflationary policy anchor. In recent years, larger developing countries with floating currencies have generally opted for some form of inflation targeting, although the degree of commitment to the target can vary. For a small state, however, inflation targeting can be difficult, requiring a large central bank staff and greater statistical resources than it may have. A floating exchange rate regime without a credible inflation or monetary target is also susceptible to pressure to monetize fiscal deficits. Taking these factors into account, market traders (both foreign and domestic) may demand higher interest rates or an especially depreciated exchange rate to hold local assets. A further consideration is that the volatility of the exchange rate itself may carry additional costs, particularly for businesses that are not able to hedge their foreign exchange risks effectively. The availability of appropriate hedging instruments in small countries is likely to be negligible.

As with the currency board, the free float is also compatible with the concept of a currency union. The leading example of an external free float combined with a currency union is the euro area. In addition to the OECS/ECCU, the West African Monetary Union provides an example of a currency union that has a pegged exchange rate (pegged to the euro).

See Table 16.1 for a comparison of the various regimes.

Table 16.1Comparison of Exchange Rate Regimes
IndicatorDollarizationCurrency boardFree floatOptimum
CredibilityPerfectVery highLowPerfect
Seigniorage gainNoYes (low)Yes (high)Yes
Interest premiumLowMediumHighLow
Resilience to external shocksHighMediumHighHigh
One-off reserves cost of acquiring currencyYesYesNoNo
Reserve coverage100% or moreVariable foreign reservesNo
Lender of last resortNoNo (or very limited)YesYes
Staffing requirementsNo staffSmall staffLarge staffNo staff
Transaction costsNoneLowHighNone
Budget disciplineYesYesNoYes
Monetary policyNoneSmall (if capital markets imperfect)YesYes
Source: Adapted from Imam, 2010.Note: — = Not applicable.

The Optimal Exchange Rate Regime for the OECS/ECCU

Metrics that bear on the suitability of the various exchange rate regimes for the OECS/ECCU countries are examined in this section. The first subsection looks at specific indicators such as banks’ exposures to foreign currency debt, seigniorage, interest rates, inflation, and trade and other economic linkages. The second subsection applies an established empirical test for optimal currency areas to the OECS/ECCU, using some of the preceding indicators to assess the suitability of a hard currency link between OECS/ECCU members and the U.S. dollar and other potential anchor currencies.

Empirical Indicators of the Suitability of Various Regimes for the OECS/ECCU

The following assesses some specific characteristics of the OECS/ECCU economies, with a view to determining their implications for different exchange rate regimes.

Bank Balance Sheets

The foreign currency positions of banks in the OECS/ECCU are well matched in the aggregate (Table 16.2). As of December 2010, the banking system’s foreign currency liabilities totaled US$841 million, whereas foreign currency assets exceeded liabilities at US$991 million (equivalent to 18.8 percent of total assets). However, US$766 million of those assets consisted of loans in foreign currency to private businesses. Such assets often carry a degree of credit risk that can increase in a downturn or if the currency is devalued. Should the latter occur, debtors without access to a steady stream of foreign currency income could find it difficult to service their loans, raising the possibility of default. This type of hidden credit risk, which can be difficult to assess in the absence of a shock, arose in both Mexico and Brazil during sharp depreciations in 2008. In light of the liquidity-constrained condition of some OECS/ECCU banks, the potential for a devaluation to cause damage to bank balance sheets appears to be moderate. Bank balance sheet exposure is one of the factors to consider before country authorities consider moving from a fixed exchange rate system or a currency board to a free float.

Table 16.2Foreign Currency Deposits and Loans of Residents in OECS/ECCU Banks(EC$ millions)
December

2009
December

2010
Public sector
Foreign currency deposits376.2332.5
Foreign currency loans84.193.1
Private business
Foreign currency deposits908.2899.8
Foreign currency loans1,967.82,068.1
Households
Foreign currency deposits945.2916.3
Foreign currency loans533.8509.6
Nonbank financial institutions
Foreign currency deposits34.585.3
Foreign currency loans1.81.6
Subsidiaries and affiliates
Foreign currency deposits42.235.7
Foreign currency loans4.23.4
Total
Foreign currency deposits2,306.42,269.6
Foreign currency loans2,591.72,675.7
Memorandum
Foreign currency deposits/total deposits (percent)16.015.4
Foreign currency loans/total loans (percent)18.118.8
Sources: Eastern Caribbean Central Bank; and IMF staff estimates.Note: For the whole banking system.

Foreign Exchange Reserves

Given the long-functioning quasi–currency board system and a reserve coverage ratio of almost 100 percent, members of the OECS/ECCU already have at their disposal an adequate stock of foreign reserves (Table 16.3). No additional one-off cost would be incurred in moving from a currency board to outright dollarization, other than the loss of seigniorage (see below). Conversely, a move to a free float could free up some reserves (note the Dominican Republic’s ratio of reserves to GDP), but potentially at the cost of some credibility. Jamaica and the Dominican Republic serve as regional referents with freer exchange rate arrangements than the OECS/ECCU. Both countries’ de jure exchange rate arrangements are classified as “floating.” However, Jamaica’s de facto regime is a “stabilized arrangement,” and the Dominican Republic’s de facto regime is a “crawl-like arrangement.”9

Table 16.3Foreign Reserves
CountryAverage

(US$ millions)
Percent of GDP
2000–042005–092000–042005–09
Antigua and Barbuda931364.64.8
Dominica39615.46.7
Grenada831086.86.6
St. Kitts and Nevis621016.47.2
Saint Lucia1001445.15.6
St. Vincent and the Grenadines59815.95.7
OECS/ECCU Total4376315.55.8
Dominican Republic6502,3330.10.2
Jamaica1,5282,0440.30.2
Sources: IMF, International Financial Statistics; and IMF staff calculations.

Seigniorage Gains and Losses

The total seigniorage accrued by the central bank is equal to the monetary base (adjusted for price changes over time). To shift from a currency board to dollarization, the central bank would have to purchase back the monetary base outstanding, which would constitute a reversal of the accrued seigniorage gains. Conversely, a shift to a free float could allow for some additional seigniorage, although potentially at the cost of some loss in anti-inflationary credibility. Table 16.4 details the monetary base, and thus the potential loss in seigniorage that would be associated with dollarization, in the OECS/ECCU.

Table 16.4Monetary Base
CountryAverage

(US$ millions)
Percent of GDP
2001–042005–092001–042005–09
Antigua and Barbuda11014614.713.8
Dominica435316.315.7
Grenada8910219.516.9
St. Kitts and Nevis679718.318.8
Saint Lucia10313214.013.9
St. Vincent and the Grenadines617516.014.2
OECS/ECCU Total51866317.516.6
Dominican Republic2,4234,90810.812.8
Jamaica1,0901,26911.810.6
Sources: IMF, International Financial Statistics; and IMF staff calculations.

Interest Rates

The gap between interest rates in the OECS/ECCU and those in the United States and elsewhere may be useful in highlighting potential gains from dollarization. In many fixed exchange rate regimes, such gaps signal the degree of credibility the market attributes to the current regime. For market participants, all else being equal, the greater the likelihood of depegging, the higher the interest rate demanded for holding local currency assets. As shown in Table 16.5, interest rates in the OECS/ECCU have remained within 1 or 2 percent of those in the United States, even as neighboring countries have endured much higher nominal interest rates. It would appear then that the currency board has provided a substantial degree of credibility to the OECS/ECCU, to the extent that a shift to dollarization might not offer a very large reduction in rates.

Table 16.5Money Market Rates(Percent)
Region or countryAverageAverage
1990–941995–992000–042005–09
OECS/ECCU average5.35.35.65.0
Canada7.64.83.52.9
Dominican Republicn.a.14.921.510.3
Jamaican.a.23.018.59.8
United States4.95.42.93.1
Sources: IMF, International Financial Statistics; and IMF staff calculations.Note: n.a. = not available.

Inflation

The rate of inflation is another key indicator of the effectiveness of an exchange rate regime. As shown in Table 16.6, the currency board regime appears to have been highly effective, keeping inflation at approximately the same rate as experienced in the United States, although with somewhat higher variability. As is the case with nominal interest rates, the much higher rates of inflation in Jamaica and the Dominican Republic, for example, point to some of the potential costs of a move away from a hard peg.

Table 16.6Inflation
CountryAverage

(percent)
Coefficient

of variation
1990–992000–091990–992000–09
Antigua and Barbuda2.11.91.40.8
Dominica1.92.20.80.7
Grenada2.62.50.51.2
St. Kitts and Nevis3.33.70.90.7
Saint Lucia3.62.60.81.0
St. Vincent and the Grenadines2.93.31.01.0
OECS/ECCU average2.72.70.90.9
Dominican Republic14.512.71.61.0
Jamaica27.211.20.80.4
United States2.92.50.40.4
Canada2.12.10.70.4
Sources: IMF, World Economic Outlook database; and IMF staff calculations.Note: Inflation is measured using the end-of-period consumer price index.

Comovements in Output

The discussion above suggests that, all else equal, the closer the business cycles of any two economies, the more suited they are to be linked by a fixed exchange rate. For the OECS/ECCU economies, the key issue is the extent to which domestic business cycles correspond to those of its anchor currency, that is, the United States. The long-term success of the currency union also depends on the degree to which business cycles within the OECS/ECCU are synchronized. The matrix in Table 16.7 measures comovement as stdevyi – Δyj), in which yi and yj are natural logarithms of the output of economies i and j. This measure is inversely related to the covariance in output between i and j.

Table 16.7Variation in Output, 2000–09
Antigua

and

Barbuda
DominicaGrenadaSt. Kitts

and Nevis
Saint

Lucia
St. Vincent

and the

Grenadines
CanadaDominican

Republic
JamaicaUnited

States
Antigua and Barbuda0.000
Dominica0.0480.000
Grenada0.0650.0640.000
St. Kitts and Nevis0.0420.0360.0590.000
Saint Lucia0.0320.0300.0540.0350.000
St. Vincent and the Grenadines0.0360.0350.0670.0390.0260.000
Canada0.0420.0360.0530.0280.0280.0280.000
Dominican Republic0.0490.0360.0600.0440.0400.0370.0350.000
Jamaica0.0390.0360.0560.0370.0250.0250.0150.0380.000
United States0.0400.0340.0530.0280.0240.0260.0060.0370.0130.000
Sources: IMF, World Economic Outlook database; and IMF staff calculations.Note: The comovements in output are measured as the standard deviation of the difference in the logarithm of the real output of two countries. Lower values indicate closer business cycles.

The analysis reveals the surprising result that the comovement in output between the individual OECS/ECCU economies and the United States is not significantly greater than that between the OECS/ECCU countries and Canada. Nor is the comovement in output between any given OECS/ECCU economy (e.g., Antigua and Barbuda) and the other economies within the OECS/ECCU notably greater than that with other Caribbean countries not in the currency union. Although not shown in the matrix, but discussed below, the degree of comovement between OECS/ECCU members and the United States deteriorated somewhat between 1990–99 and 2000–09. Overall, the evidence presented in the matrix does not indicate any special justification for a hard currency link between the OECS/ECCU and the United States, or even among the OECS/ECCU economies.

Terms of Trade Variation

Industrial structure is another factor bearing on the decision to fix the exchange rate between two economies. The more similar the industrial structures of two separate economies, the more likely it is that a common external shock (e.g., a rise in commodity prices) will affect the economies in the same way, and thus the more likely it is that the same monetary and exchange rate policy response will be appropriate for both. However, because data on industrial structures may be difficult to interpret and are often not widely available, comovements in terms of trade are used here as a proxy.10 The intuition is that the ex post response to a shock, that is, a shift in the terms of trade, reveals the underlying export and (to some extent) industrial structure (Table 16.8). Somewhat surprisingly, as applied to the OECS/ECCU economies and the United States and Canada, this measure does not appear to signal a wide divergence in structure, despite differences in the relative levels of economic development.

Table 16.8Variation in the Terms of Trade
Antigua

and

Barbuda
DominicaGrenadaSt. Kitts

and

Nevis
Saint

Lucia
St. Vincent

and the

Grenadines
CanadaDominican

Republic
United

States
Jamaica
Antigua and Barbuda0.00
Dominica0.120.00
Grenada0.140.070.00
St. Kitts and Nevis0.120.050.080.00
Saint Lucia0.130.090.130.110.00
St. Vincent and the Grenadines0.150.130.170.130.120.00
Canada0.140.050.080.070.110.160.00
Dominican Republic0.130.040.090.040.090.120.080.00
Jamaica0.110.040.080.020.100.120.080.030.00
United States0.110.030.080.040.090.110.080.020.020.00
Sources: IMF, World Economic Outlook database; and IMF staff calculations.Note: The comovements in terms of trade are measured as the standard deviation of the difference in the logarithm of the terms of trade of two countries. Lower values indicate less variation and more similar industrial structures.

Bilateral Trade

Bilateral trade relations constitute another criterion affecting the suitability of an exchange rate peg. The stronger the trade ties between any two economies, the more likely they are to gain from reduced bilateral exchange rate volatility. A further consideration is that countries that are already linked in a currency union or through a fixed exchange rate may develop stronger trade ties due in part to bilateral exchange rate stability (see Micco, Stein, and Ordoñez, 2003; and Rose, 2000). As shown in the trade matrix (Table 16.9), most of the OECS/ECCU countries send a substantial proportion of their exports to the United States and Canada. An interesting exception is Antigua and Barbuda, which sends only 2 percent of its exports to the United States. However, although it does not appear in this table, the United States is a substantial market for Antigua’s tourism services, which account for a larger share of economic activity than do exports.

Table 16.9Bilateral Export Shares, 2000–09(Percent)
CountriesAntigua

and

Barbuda
CanadaDominicaDominican

Republic
GrenadaJamaicaSt. Kitts

and

Nevis
Saint

Lucia
United

States
St. Vincent

and the

Grenadines
Antigua and0.550.290.030.190.050.720.402.020.05
Barbuda
Canada0.010.000.030.000.040.000.0082.510.00
Dominica8.000.340.330.6910.583.263.633.470.68
Dominican0.060.860.030.010.400.040.0574.120.01
Republic
Grenada7.121.356.410.380.916.3910.4819.641.10
Jamaica0.2614.390.150.360.110.150.2433.620.10
St. Kitts and Nevis0.507.191.180.020.040.142.1163.670.28
Saint Lucia3.620.353.460.032.400.530.8817.711.39
United States0.0221.870.010.540.010.190.010.010.01
St. Vincent and the Grenadines2.680.182.390.002.300.791.404.341.73
Sources: IMF, Direction of Trade Statistics; and IMF staff calculations.Note: Entry aij denotes the percentage of country i’s total exports that are shipped to country j.

Testing for an Optimal Currency Area Index for the OECS/ECCU

Following earlier work by Bayoumi (1994), Bayoumi and Eichengreen (1997) devised an index to evaluate the suitability of different countries to be part of an optimum currency area (OCA), applying it to the countries taking part in the European Monetary Union—the predecessor to the euro area. The idea behind the index is simple: countries with greater comovements in output, higher bilateral trade, more similar industrial structures, and smaller economies are better suited to participate in an OCA because the bilateral exchange rates should be more stable in such cases. For countries that are well suited to membership in the same currency area, the change in the standard deviation of the logarithm of the bilateral exchange rate between countries i and j should be small. This value—the “fitted” standard deviation of the bilateral exchange rate—is the index measure that Bayoumi and Eichengreen employ in their paper.

In the construction of the index, output disturbances are measured as the standard deviation of the difference in the change in the log of real output in the two countries SDyi – Δyj) so countries with synchronized business cycles will have a lower value of this measure. Bayoumi and Eichengreen (1997) also include a measure of the dissimilarity of the commodity composition of exports as a proxy for the effects of industry-specific shocks. However, because of the lack of data on the commodity composition of bilateral trade for some of the developing countries included in the sample, an alternative measure proposed by Kim and Papi (2005), as described earlier, is used here.

Trade is a key component of the OCA index because countries with stronger trade relations will benefit more from belonging to a currency union. Higher trade volumes are also expected to yield a more stable exchange rate. To assess trade links, bilateral trade data are used to compute the mean of the exports-to-GDP ratios of the two countries, Tradeij. The potential benefits of a common currency are expected to be higher in small economies because the costs of forgoing an independent monetary policy are lower.11 To measure the effect of size, the mean of the logarithms of the GDP of countries i and j in U.S. dollars is computed (Sizeij).

For the purpose of the estimation, the sample is restricted to the countries of the western hemisphere. These total 34,12 yielding 561 distinct country pairs. However, as a result of data availability, the sample is reduced to a maximum of 553 pairs depending on the time period used. The sample period covered is between 1980 and 2009, but to test the stability of the results regressions were run for five subperiods (five successive moving averages of 10 years). The equation estimated by OLS is

where the dependent variable is the standard deviation over the sample period of the change in the log of the bilateral exchange rate between each country pair ij. The output and terms of trade variables are measured as the standard deviation over the sample period, while the trade and size variables are measured as averages over the period.

The estimation results are presented in Table 16.10. The coefficients have the expected signs with the exception of the terms of trade, which in two subperiods has a negative sign. However, the results for the whole sample period (1980–2009) have all the correct signs and are statistically significant at the 5 percent level. The values of the coefficients over the different sub-periods do not seem to be particularly stable, but unlike Bayoumi and Eichengreen (1997) and Kim and Papi (2005), the focus here is not on measuring the speed of convergence toward a potential currency union, but on evaluating the existing currency union in the Eastern Caribbean.

Table 16.10Estimations of the Optimum Currency Area Index
Variables1980–89

SD(eij)
1985–1994

SD(eij)
1990–99

SD(eij)
1995–2004

SD(eij)
2000–09

SD(eij)
1980–2009

SD(eij)
SDyi – Δyj)0.5327.661***6.070***3.130***2.568***5.776***
(0.850)(1.048)(1.596)(0.279)(0.221)(1.259)
SDtoti – Δtotj)2.296***1.602***–0.4400.003–0.159**0.723**
(0.261)(0.374)(0.352)(0.069)(0.073)(0.331)
Tradeij–0.748–4.532***–3.701**–0.375–0.142–3.310***
(0.495)(1.478)(1.724)(0.307)(0.158)(0.922)
Sizeij0.067***0.098***0.063***0.008***0.007***0.090***
(0.004)(0.007)(0.010)(0.002)(0.002)(0.006)
Constant–0.257***–0.388***0.073–0.0120.010–0.220***
(0.059)(0.079)(0.078)(0.012)(0.010)(0.073)
Observations523534549553553553
R20.500.360.090.250.220.34
Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.Note: The total number of countries in the sample is 34. Standard errors in parentheses. *** significant at the 1 percent level. ** significant at the 5 percent level. * significant at the 10 percent level.

The entire period 1980–2009 is used to construct the OCA index. The index is the predicted value of the dependent variable using the coefficients from the 1980–2009 regression with the values of the exogenous variables in the 1990s (1990–99) and the 2000s (2000–09). Because the OECS/ECCU has a quasi–currency board with a fixed exchange rate to the U.S. dollar, all the countries are evaluated vis-à-vis the United States. In addition, because the presence of Canadian banks in the OECS/ECCU region is large and there are important economic ties, the index is also calculated with respect to Canada.

As shown in Figure 16.1, the OECS/ECCU countries (round dots) are among those better suited in the sample to have a fixed exchange rate pegged either to the U.S. or the Canadian dollar. However, the location of most of the OECS/ECCU countries above the 45 degree line indicates that they were better placed to benefit from a fixed exchange rate in the 1990s than in the past decade. A further point is that, compared with the indices calculated by Bayoumi and Eichengreen (1997) for the European countries in 198713 that lie in the range of 0.003 and 0.099, the indices for the OECS/ECCU are all above 0.1 in the 2000s. These values suggest that the case for a fixed exchange rate with either Canada or the United States may not be particularly strong, at least on the basis of OCA index criteria.

Figure 16.1Optimum Currency Area Indices with Canada and the United States, 1990–2009

Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.

Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate. Points below the 45-degree line denote increased suitability for a fixed exchange rate arrangement in 2000–09 compared with 1990–99. Antigua and Barbuda (ATG), Argentina (ARG), The Bahamas (BHS), Barbados (BRB), Belize (BLZ), Bolivia (BOL), Brazil (BRA), Canada (CAN), Chile (CHL), Colombia (COL), Costa Rica (CRI), Dominica (DMA), Dominican Republic (DOM), Ecuador (ECU), El Salvador (SLV), Grenada (GRD), Guatemala (GTM), Guyana (GUY), Haiti (HTI), Honduras (HND), Jamaica (JAM), Mexico (MEX), Nicaragua (NIC), Panama (PAN), Paraguay (PRY), Peru (PER), St. Kitts and Nevis (KNA), Saint Lucia (LCA), St. Vincent and the Grenadines (VCT), Suriname (SUR), Trinidad and Tobago (TTO), United States (USA), Uruguay (URY), and Venezuela (VEN).

An assessment of which currency would make the best anchor for the OECS/ECCU—assuming that a currency board is preferred—shows that in the last two decades the Canadian dollar had a better fit. Figure 16.2 compares the OCA indices of the OECS/ECCU countries with respect to Canada and the United States in the periods 1990–99 and 2000–09. The points that lie below the 45 degree line indicate that a peg to the U.S. dollar would be preferred to a link to the Canadian dollar. So the fact that most of the OECS/ECCU countries lie above the 45 degree line in both decades indicates that they might have benefited more from pegging to the Canadian dollar than to the U.S. dollar. The exceptions are St. Kitts and Nevis and St. Vincent and the Grenadines, which show a greater economic affinity with the United States than with Canada.

Figure 16.2Optimum Currency Area Indices: United States versus Canada, 1990–2009

Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.

Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate. Points below the 45-degree line denote increased suitability for a fixed exchange rate arrangement with the United States compared with Canada. Antigua and Barbuda (ATG), Dominica (DMA), Grenada (GRD), St. Kitts and Nevis (KNA), Saint Lucia (LCA), and St. Vincent and the Grenadines (VCT).

This result is also supported by the overall ranking of the OCA index presented in Appendix 16A. The results seem to suggest that the OECS/ECCU countries would be better off with a fixed exchange rate with Canada rather than with the United States. The OECS/ECCU has four countries in the top ten places in the Canadian rankings, whereas it has only two in the U.S. rankings. However, these results do not take into consideration that most of the foreign currency loans and deposits are denominated in U.S. dollars; hence, a peg to the dollar limits exchange rate and credit risk for banks, companies, and households. To get an idea of how the OECS/ECCU compares with other country groupings in the sample, the analysis computes the average of each group’s index with respect to the anchor countries. Table 16.11 shows the averages for different groups in the region and for the OECS/ECCU, including a group of countries in the western hemisphere that have fixed exchange rates with the United States. The lowest indices are presented in boldface type. Notably, in the last decade, the OECS/ECCU slipped to third among the different country groupings (in the western hemisphere) with respect to suitability for a fixed currency link to the United States.

Table 16.11Optimum Currency Area, Country Groups with the United States and Canada, 2000–09
United StatesCanada
Country groups1990–992000–091990–992000–09
OECS/ECCU10.2380.3250.2700.299
Caribbean20.2880.2680.3370.301
Caribbean (excluding OECS/ECCU)0.3250.2260.3870.303
South America30.6410.5890.6020.541
Central America40.3830.2730.4230.368
Countries with fixed exchange rates50.3240.3190.3410.325
Countries with fixed exchange rates (excluding OECS/ECCU)0.4110.3130.4120.352
Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.

The OCA index can also be used to assess the appropriateness of fixed exchange rate links among the different OECS/ECCU members. Figure 16.3 presents the OECS/ECCU OCA indices with respect to each other. Here as well the indices have deteriorated during the past decade. Economic coordination in the OECS/ECCU seems to have been more conducive to a currency union in the 1990s than in the 2000s. Nonetheless, the still-low values of the indices do suggest that the OECS/ECCU is well suited to a currency union, but further study of the causes of the divergence in economic coordination since 1990 would be advisable.

Figure 16.3Optimum Currency Area Indices within the OECS/ECCU, 1990–2009

Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.

Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate. Points below the 45-degree line denote increased suitability for an Optimum Currency Area in 2000–09 compared with 1990–99. Antigua and Barbuda (ATG), Dominica (DMA), Grenada (GRD), St. Kitts and Nevis (KNA), Saint Lucia (LCA), and St. Vincent and the Grenadines (VCT).

Factors Underlying the OCA Results

To gain some further insight into the reasons for the deterioration in the OCA indices for some OECS/ECCU countries, the contributions of the separate independent variables to the OCA index computation for each country are analyzed separately. See Figure 16.4 and Table 16.12. Notably, the two OECS/ECCU countries that appear least suited for a currency link to the U.S. dollar in the 2000s—Antigua and Barbuda and Grenada—both show greater divergence in the business cycle, and greater differences in industrial structure, as revealed by comovements in the terms of trade.

Figure 16.4Contribution to the Optimum Currency Area Indices versus the United States, 1980–2009

Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.

Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate.

Table 16.12Contribution to the Optimum Currency Area Indices versus the United States, 1980–2009
CountriesSDyi – Δyj)SDtoti – Δtotj)TradeSizeIndex
2000–09
Antigua and Barbuda0.2330.082–0.0100.3980.483
Dominica0.1940.023–0.0190.3440.323
Grenada0.3060.057–0.0390.3690.473
St. Kitts and Nevis0.1620.026–0.1750.3580.152
Saint Lucia0.1410.063–0.0420.3970.340
St. Vincent and the Grenadines0.1490.083–0.1990.3640.178
1990–99
Antigua and Barbuda0.1790.181–0.0180.3670.490
Dominica0.1150.033–0.0260.3260.229
Grenada0.2060.093–0.0510.3390.366
St. Kitts and Nevis0.1080.016–0.1620.3280.071
Saint Lucia0.1860.054–0.0660.3760.331
St. Vincent and the Grenadines0.2150.033–0.4260.336–0.061
Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate.

Conclusion

The existing exchange rate regime remains appropriate for the OECS/ECCU in the medium term, and the currency board and peg to the U.S. dollar continue to offer substantial benefits to the economies in the region. These benefits are reflected in low inflation and relatively low nominal interest rates enjoyed within the OECS/ECCU, both of which are less than average rates elsewhere in the Caribbean and in other developing economies. On the basis of the optimal currency area framework of Bayoumi and Eichengreen (1997), the analysis determined that, although the currency link between the OECS/ECCU and the United States remains appropriate, the basis for it has deteriorated slightly in the past decade. An interesting result of this chapter is that for most of the OECS/ECCU economies, fitted bilateral exchange rates against the Canadian dollar have lower variability than those against the U.S. dollar. This suggests—although far from conclusively—that the Canadian dollar could be an alternative anchor for the currency board.

Full dollarization would provide the credibility and stability benefits that accrue to the currency board, but dollarization would carry additional costs that would likely outweigh any further stability gains. First among these would be seigniorage losses, the loss of any potential monetary flexibility, and the possibility of functioning as a lender of last resort. By contrast, a free float would offer large potential gains to the OECS/ECCU economies, including policy flexibility, some additional seigniorage, and the possibility of acting as a lender of last resort. But the loss in stability and credibility would likely be substantial, particularly if the move to a float were not accompanied by a credible and sustainable fiscal consolidation program.

Thus, there is no compelling case for a shift away from the combination of a currency union and currency board now employed by the OECS/ECCU. Nevertheless, this chapter does point to divergence between the member economies and the economy of the United States, the anchor currency country. Accordingly, it may be useful to reassess whether the U.S. dollar remains the appropriate anchor currency.

Appendix 16A.
Table 16A.1Optimum Currency Area, Country Rankings with the United States and Canada, 2000–09
CountriesUnited StatesCanada
RankingIndexRankingIndex
Suriname1–0.0210.02
Trinidad and Tobago20.08270.49
Canada with United States (United States with30.1220.12
Canada)
St. Kitts and Nevis40.1530.20
St. Vincent and the Grenadines50.1870.28
Costa Rica60.19160.37
El Salvador70.21140.34
Nicaragua80.2160.25
Belize90.2480.29
Mexico100.26290.54
Dominican Republic110.27230.45
Barbados120.2850.24
Haiti130.28110.31
The Bahamas140.29130.34
Honduras150.30150.35
Jamaica160.31100.31
Dominica170.3240.24
Ecuador180.33200.41
Saint Lucia190.3490.29
Guyana200.34210.42
Guatemala210.37180.40
Bolivia220.43120.33
Chile230.47240.45
Grenada240.47170.38
Antigua and Barbuda250.48190.40
Paraguay260.50220.43
Colombia270.51250.46
Panama280.52260.49
Peru290.55280.52
Uruguay300.65310.62
Brazil310.65300.56
Argentina320.87320.78
Venezuela330.93330.84
Sources: IMF, World Economic Outlook database; Direction of Trade Statistics; and IMF staff estimates.Note: Based on estimated indices from the regression for the period 1980–2009. Lower values indicate greater suitability for a fixed bilateral exchange rate.
References

    Bayoumi, Tamim,1994, “A Formal Model of Optimum Currency Areas,IMF Staff Papers, Vol. 41, No. 4, pp. 53758.

    Bayoumi, Tamim, and BarryEichengreen,1997, “Ever Closer to Heaven? An Optimum-Currency-Area Index for European Countries,European Economic Review, Vol. 41, pp. 76170.

    Imam, Patrick,2010, “Exchange Rate Choices of Microstates,IMF Working Paper 10/12 (Washington: International Monetary Fund).

    International Monetary Fund (IMF), 2010, “Seychelles: First Review Under the Extended Arrangement, Request for Modification of Performance Criteria, and Financing Assurances Review,IMF Country Report No. 10/204 (Washington).

    International Monetary Fund (IMF), 2011, Annual Report on Exchange Arrangements and Exchange Restrictions (Washington).

    Kim, Jun Il, and LauraPapi,2005, “Regional Integration and Exchange Rate Arrangements,” in Central America: Global Integration and Regional Cooperationed. byM.Rodlauer and A.Schipke,IMF Occasional Paper No. 243 (Washington: International Monetary Fund).

    Micco, Alejandro, ErnestoStein, and GuillermoOrdoñez,2003, “The Currency Union Effect on Trade: Early Evidence from EMU,Economic Policy Vol. 18, No. 37, pp. 31556.

    Pineda, Emilio, PaulCashin, and YanSun,2009, “Assessing Exchange Rate Competitiveness in the Eastern Caribbean Currency Union,IMF Working Paper 09/78 (Washington: International Monetary Fund).

    Rose, Andrew,2000, “One Money, One Market: Estimating the Effect of Common Currencies on Trade,Economic Policy, Vol. 15, pp. 745.

Before this, the region had a fixed exchange rate system with the British pound through the Caribbean Currency Board (see Chapter 15 on the Eastern Caribbean Central Bank).

Imam (2010) defines microstates as those with populations of less than 2 million.

The focus of the analysis is the exchange rate regime of the OECS/ECCU vis-à-vis the rest of the world. This chapter does not attempt to assess the inner workings of the currency union, and therefore does not address issues of labor mobility and economic convergence within the union. Neither does it consider the need for a risk-sharing mechanism and more fiscal policy and financial sector coordination, as highlighted by the European Union crisis. Some of these issues are discussed in Chapter 3 on Economic Integration and the Institutional Setup of the OECS/ECCU.

Various intermediate regimes are also possible, such as a sliding or crawling peg and a managed float.

There are currently 13 countries in the world that officially use the currency of another country (that is, they have no separate legal tender), including Panama, El Salvador, Ecuador, Palau, Marshall Islands, Micronesia, and Timor-Leste, which use the U.S. dollar; Kosovo, Montenegro, and San Marino, which use the euro; Kiribati and Tuvalu, which use the Australian dollar; and Zimbabwe, which uses various currencies.

There are 12 countries with currency board arrangements. In addition to the six OECS/ECCU countries, Bosnia, Brunei, Bulgaria, Djibouti, Hong Kong SAR, and Lithuania have currency boards.

In the OECS/ECCU, the currency board is a quasi–currency board because the ECCB is mandated to hold foreign exchange equivalent to at least 60 percent of demand liabilities. The operational target, however, is 80 percent and in practice coverage has been close to 100 percent (see also Chapter 15 on the Eastern Caribbean Central Bank).

Alternatively, the currency board arrangement may not require strict one-to-one reserve backing of the monetary base. This is the case in the OECS/ECCU, where the ECCB is required to hold reserves equivalent to 60 percent of the monetary base.

A stabilized arrangement is one in which the spot market exchange rate remains within a margin of 2 percent for six months or longer (with the exception of a specified number of outliers or step adjustments) and is not floating, and the exchange rate remains stable as a result of official action. In a crawl-like arrangement the exchange rate must remain within a narrow margin of 2 percent relative to a statistically identified trend for six months or longer (with the exception of a specified number of outliers) and the exchange rate is not considered to be floating. For a more detailed description of the different types of arrangements see IMF (2011).

A similar approach to the same issue in the Central American context is taken in Kim and Papi (2005).

Small economies have less scope to implement truly independent monetary policies, hence, the costs of forgoing them are lower.

Antigua and Barbuda, Argentina, The Bahamas, Barbados, Belize, Bolivia, Brazil, Canada, Chile, Colombia, Costa Rica, Dominica, the Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, St. Kitts and Nevis, Saint Lucia, St. Vincent and the Grenadines, Suriname, Trinidad and Tobago, the United States, Uruguay, and Venezuela.

This is the last year for which Bayoum and Eichengreen have a complete set of data.

    Other Resources Citing This Publication