II. Considerations for Adopting a Currency Board

Charles Enoch, and Tomás Baliño
Published Date:
September 1997
  • ShareShare
Show Summary Details

CBAs may appear attractive to small open economies with limited central banking expertise and incipient financial markets, or to countries that wish to preserve the benefits of belonging to a broader trade or currency area, or envisage joining a currency union. In addition, they may be attractive to countries where lack of credibility severely constrains the effectiveness of monetary policy or exposes the economy to recurrent currency crises and high risk premiums (Box 3). However, a CBA cannot of itself create credibility unless accompanied by firm supporting policies. Without such policies, credibility will remain low, which will undermine the sustainability of the CBA itself.

Strengths of a Currency Board Arrangement

A CBA’s strength accrues from its simplicity and the limited discretion of its operating rules. These rules eliminate or sharply limit the scope for discretion in monetary and foreign exchange rate policies, and thereby enhance the credibility of conventional fixed pegs and simplify central bank operations. They can be particularly helpful in cases where despite a strong commitment to financial discipline the political process is unlikely to give the central bank a high degree of independence or the central bank has not as yet been unable to establish a solid track record.

Administrative and Operational Simplicity

The extreme simplicity and transparency of a CBA is appealing. Its operating rules are easily understood and monitored by the general public, if appropriate information is provided. Pegging the exchange rate simplifies the operation and monitoring of the foreign exchange market. Streamlining monetary operations and delegating other central bank functions, such as payments system and fiscal agent functions, can sharply reduce the need for staff and bookkeeping.

For such reasons, a CBA is particularly attractive for post-chaos countries—such as Bosnia and Herzegovina—or for small countries that have recently attained independence and that wish to create their own currency for motives of seigniorage as well as national identity.14

Credibility of Sound Monetary and Fiscal Policies

CBAs have been used in recent years to strengthen the credibility of exchange-rate-based disinflation policies in Argentina, Estonia, and Lithuania. By eliminating or strictly limiting the monetary authorities’ ability to monetize fiscal deficits, or lend to banks, CBAs can rapidly strengthen credibility. In principle, they would be expected to be more effective than conventional fixed pegs when the monetary authorities’ reputation has been weakened by a history of lax fiscal policy, accommodative monetary policy, and failed stabilization attempts (as in Argentina), or when the authorities lack an established track record (as in Estonia and Lithuania). By providing clear signals about the policy intentions of the authorities, and ensuring monetary independence from the government, CBAs facilitate an adjustment of expectations and promote wage and price discipline, thereby lessening potential inflation biases.

CBAs appear to have made an important contribution to stabilization programs in Argentina, Estonia, and Lithuania. In all cases, they clearly played a central role in securing monetary discipline.15 They ensured that changes in the demand for base money, after the exchange rate was pegged, were met by changes in foreign assets rather than domestic assets, thereby strengthening the credibility and sustainability of stabilization efforts. For instance, in Argentina net domestic assets of the central bank stabilized immediately after the CBA was introduced. Thus, reserve money growth decelerated sharply, notwithstanding a rapid buildup of international reserves caused by capital inflows (Figure 1). It should also be noted that the CBA contributed in an important way to enforcing fiscal discipline (at the federal level). The case is less clear in the Baltics due to the fact that fiscal discipline was less of a problem there.

Figure 1.Argentina: Reserve Money and Its Components1

(In millions of pesos)

Source: IMF, International Financial Statistics.

1 The increase in net foreign assets and decline in net domestic assets in September 1992 reflect primarily a transfer of balance of payments support loans to capital of the central bank.

Box 3.Motivation for Establishing Currency Board Arrangements

Although all existing CBAs were established with the objective of maintaining the long-term stability of the currency, they can be divided into two groups, depending on the rate of inflation that prevailed at the time they were established. The high-inflation group includes Argentina, Estonia, and Lithuania, and the low-inflation group includes Brunei Darussalam, Djibouti, members of the Eastern Caribbean Central Bank (ECCB), and Hong Kong.

High-Inflation Group

In Argentina, since 1985 the authorities had implemented a series of stabilization programs that included wage and price freezes, measures to strengthen monetary and fiscal discipline, and exchange rate pegs. Nevertheless, continued lax fiscal discipline had resulted in a persistently high rate of monetary expansion, as the public sector primarily relied on the central bank for deficit financing. As a result, inflation had accelerated and dollarization had become entrenched. Following an extensive flight from assets denominated in local currency (austral) during 1988–89, a national election in 1989, and an assets freeze in January 1990, domestic liquidity became very tight and the exchange rate stabilized during the second and third quarters of 1990, resulting in a rapid fall of inflation.1 Reserve money growth expanded, however, in the last quarter of 1990 as the central bank extended sizable credits to the social security system and provided rediscounts to provincial banks that were facing runs on their deposits. The fiscal position weakened further and the austral depreciated by almost 70 percent. The CBA was introduced in April 1991 through the Convertibility Law as part of a comprehensive stabilization package to strengthen the credibility of a nominal exchange rate anchor.

Estonia and Lithuania introduced CBAs to deal with high inflation, which resulted, inter alia, from the liberalization of administered prices, terms of trade shocks, and the depreciation of the Soviet ruble. After they introduced economic reforms and liberalized prices in 1991, their annual inflation rates accelerated from around 200 percent at the end of 1991 to over 1,000 percent at the end of 1992.

In Estonia, the use of the rapidly depreciating Soviet ruble led to high inflation and mounting political pressures to introduce a national currency. Lacking a track record in monetary management and believing that policy credibility was important to the success of stabilization, the authorities adopted a CBA in June 1992.

In Lithuania, which also experienced high imported inflation during the early stage of transition to a market economy, the authorities introduced in October 1992 talonas, or coupons (at par with the Russian ruble). They initially pursued a relatively relaxed monetary policy, and stabilization was delayed. In June 1993, a new currency, the litas, was issued to replace the talonas, but without sufficiently tight monetary policy, inflation continued. To maintain the litas, credibility and speed up stabilization, and encouraged by the success of Estonia’s CBA, the authorities introduced a CBA through the Litas Stability Law on April 1, 1994.

Low-Inflation Group

In Hong Kong, the decision to adopt a CBA was driven by the need to end the 1983 crisis of confidence in the Hong Kong dollar. During the crisis, the Hong Kong dollar depreciated against the U.S. dollar by more than 50 percent; property and stock markets collapsed; and financial institutions experienced widespread deposit runs. As Hong Kong’s economy was highly dependent on international trade and financial activities, a credible fixed exchange rate was deemed crucial for economic recovery.

Brunei Darussalam, Djibouti, and member governments of the ECCB adopted a CBA partly due to their limited central banking expertise. In the case of the ECCB, the CBA’s simplicity and transparency also facilitate policy coordination between member countries. Similarly, owing to a lack of central banking infrastructure and the need to maintain a transparent mechanism with minimal discretion, Bosnia and Herzegovina agreed to establish a CBA as their main monetary arrangement in the Dayton peace agreement.2

1 On January 1, 1990, the government converted all of its austral-denominated public debt maturing in 1990, and the bulk of the austral-denominated time deposits of the banking system, to long-term U.S. dollar-denominated bonds; see Ize and Mackenzie (1992). Between February and December 1990, the austral stabilized at an exchange rate equivalent to around Arg$0.5–0.6 per US$1. But it depreciated to around Arg$0.95 per US$1 in January 1991 and to around Arg$0.98 peso per US$1 by the time the CBA was introduced and the peso replaced the austral.2 For a comparison between the ECCB and the Central African Monetary Union, see Nascimento (1994).

In Argentina, fiscal policy aimed at achieving an annual primary surplus (excluding privatization proceeds) sufficient to meet the interest obligations of the nonfinancial public sector deficit and the quasi-fiscal deficit (net foreign interest outlays) of the central bank. As a result chiefly of revenue-enhancing measures and the lower interest bill following the conversion of domestic debt into dollar-denominated debt, the overall deficit of the combined public sector fell from 7.3 percent of GDP to 0.4 percent of GDP during 1989–92.

The CBA countries mentioned above experienced strong economic recoveries after the stabilization periods. Lower interest costs contributed to these recoveries, as interest rates converged quickly to international levels. Furthermore, CBAs’ credibility helped attract large capital inflows (as in Argentina, Estonia, and Hong Kong); however, the inflation record has been mixed. In Argentina, inflation was brought down to low levels, and has remained low. Since mid-1993, it has remained lower than that of Chile and Mexico (Figure 2). In the Baltic countries and Hong Kong, on the other hand, inflation has remained moderate but higher than in their trading partners owing to the CBAs’ nominal exchange rate rigidity (Figure 3). In the ECCB countries, Marston (1995) shows that they had lower inflation than other Caribbean countries. During the period 1987–92, annual inflation fluctuated between 2 percent and 5 percent, while that of Barbados, Guyana, Jamaica, and Trinidad and Tobago ranged from 12 percent to 42 percent.16

Figure 2.Argentina, Chile, and Mexico: Inflation1

(In percent)

Source: IMF, International Financial Statistics.

1 Inflation is calculated on a year-on-year basis.

Figure 3.Estonia, Latvia, and Lithuania: Inflation1

(In percent)

Source: IMF, International Financial Statistics.

1 Calculated on a year-on-year basis.

Currency Stability, Interest Rate Convergence, and Financial Intermediation

CBAs are less prone to policy reversals than conventional fixed pegs. In particular, in an era of high and growing capital mobility, they are perceived to be less vulnerable to destabilizing capital outflows and self-fulfilling currency crises.17 Due to the higher credibility, interest rates should converge rapidly to levels in the reserve currency country and remain close to international levels.18 In addition to lowering the risk premium in interest rates, containing exchange rate uncertainty and maintaining orderly monetary conditions should also help promote, in the longer run, international trade and facilitate access to international capital markets.19 For instance, in Hong Kong, during the currency crisis before the CBA was established, international trade and investment were disrupted by the sharp depreciation of the Hong Kong dollar caused by capital outflows, and real economic growth declined from 9.2 percent in 1981 to 2.7 percent in 1982. After the CBA was established, economic and exchange rate stability led to a return of capital, and a recovery of international trade, international investment, and economic growth.20

Experience with CBAs generally substantiates these expected benefits. Indeed, their resilience contrasts sharply with that of conventional fixed pegs, possibly because the introduction of a CBA provides greater impetus for a strengthening of policies overall.21 With the possible exception of earlier Argentine experiences with CBAs in the early part of this century, there are no other examples of CBAs being abandoned under stress, although as noted earlier, in some cases they were modified in response to pressures. Although the experience of colonial CBAs is only partially relevant to current ones, it is still important to note that they functioned adequately during long periods, notwithstanding large fluctuations in the terms of trade. Since the introduction of its currency board in 1983, Hong Kong has been exposed to only two short-lived attacks on its currency, notwithstanding the uncertainties related to the process of reunification with mainland China. Similarly, Argentina’s current CBA has been put to the test twice since it was introduced, and it survived successfully the aftermath of the Mexican crisis. Those in the Baltics have also withstood speculative attacks.22

In Argentina and Estonia, the performance of CBAs in lowering interest rates was indeed superior to that of countries such as Mexico and Latvia, which faced somewhat similar economic environments but had different currency and exchange rate arrangements.23 Similar evidence exists for Ireland, which operated a CBA during the period 1928–71.24 Also Argentina’s experience in 1978–81 with a pre-announced devaluation path produced a much less rapid and significant interest rate convergence than had been the case under the CBA (Box 4). During that period, money market and deposit rates remained high and only started to decline along with inflation in the last quarter of 1979. Interest rates increased sharply after a banking crisis started in the second quarter of 1980.

Box 4.Backing Rules Tablita System and the Currency Board Arrangement

An important feature of CBAs has been the convergence of domestic interest rates to international levels. In this regard, it is helpful to compare such convergence under a CBA and under an alternative exchange rate arrangement that also offers potential investors an exchange rate horizon. Argentina offers an interesting case for such a comparison.

Beginning in December 1978, the Central Bank of Argentina began to publish a schedule of future daily devaluations of the peso (informally known as the “tablita”) for a given period of time. This became a centerpiece of the stabilization strategy, which intended to make the prices of tradable goods follow international prices and assumed that nontraded goods prices would follow a similar path, perhaps after some initial changes in relative prices. Also, it was expected that by drastically reducing uncertainty about the future exchange rate, the arrangement would make domestic interest rates converge to international levels.

Both the tablita and the CBA depend on the nominal exchange rate anchor to achieve some common objectives. However, besides differences in economic circumstances and policies, there are two major differences between the two regimes. First, while the tablita was a key element of the economic program, it did not have the legal basis of the current CBA, under which both the exchange rate and the backing rule are determined by the Convertibility Law and the central bank’s charter. Second, while under the tablita there was a perception that domestic credit expansion had to be consistent with the maintenance of the exchange rate regime, it was left to the discretion of the central bank to set that expansion. Thus, the workings of the tablita system were less rigid and less transparent than those of a CBA. Moreover, this difference made a crucial difference in the way that the central bank could react to banking problems. When a banking crisis developed in 1980, the tablita allowed the central bank to provide substantial lender-of-last-resort support: its loans to the banking system jumped from less than 2 percent of reserve money just before the start of the crisis to almost 31 percent two months later (Baliño, 1991). This led to a loss of credibility and contributed to the eventual demise of the tablita in February 1981.

The differences between the two regimes are consistent with the differences in the degree of interest rate convergence. The figures below compare the convergence of deposit and money market interest rates (in the case of the tablita, the interest rates were adjusted for the announced devaluation path) to the international levels under both regimes. They indicate a lower and slower degree of convergence during the tablita period than under the CBA.

Argentina: Deposit Rate Minus U.S. Treasury Bill Rate

(In percent)

Source: IMF, International Financial Statistics.

Argentina: Money Market Rate Minus U.S. Federal Funds Rate

(In percent)

Source: IMF, International Financial Statistics.

The credibility of CBAs appears to have stimulated remonetization and financial reintermediation. In Argentina, after having declined for more than two years, the ratio of broad money to GDP increased steadily when the CBA was introduced and prices stabilized. Between 1992 and 1994, the real growth rate of credit to the private sector became significantly positive, and the ratio of deposits to broad money increased (Figure 4).25 Similarly, in Hong Kong, the credibility of the CBA’s fixed exchange rate, together with the presence of active international banks, has facilitated the development of offshore banking. In Estonia and Lithuania, despite successful stabilizations, severe banking crises and large structural reforms slowed down the growth of financial intermediation.26 However, in Estonia persistent capital inflows, in part encouraged by sound macroeconomic policies within the CBA framework, led to continued real credit growth (Figure 5).

Figure 4.Argentina: Real Growth in Broad Money and Domestic Credit1

(In percent)

Source: IMF, International Financial Statistics.

1 Adjusted by consumer price index.

Figure 5.Estonia: Real Growth Rate of Claims on the Private Sector

(In percent)

Source: IMF, International Financial Statistics.

Weaknesses of Currency Board Arrangements

The weaknesses of CBAs are the flip side of their strengths. While the commitment to preserve the parity is an asset in times of currency instability, it can become a liability in the presence of large exchange rate misalignments. This risk can be particularly serious whenever it is difficult to have any firm idea as to what the appropriate exchange rate should be (such as in a post-chaos economy). While the market may be no more successful than the authorities at initially setting the exchange rate at an appropriate level, a more flexible exchange rate regime would enable early errors to be corrected without incurring the costs of loss of credibility. Similarly, a rigid backing rule can increase the financial sector’s vulnerability to crises. Operational simplicity also comes at a cost, in that it may rule out the exercise of what would be viewed in other countries as important central bank functions. In addition, by removing the possibility of monetary financing of the fiscal deficit CBAs require a level of immediate fiscal consolidation that may be difficult to attain.

Box 5.Policy Content and Policy Framework: The Baltic Countries

While both Estonia and Latvia adopted stabilization programs at an early stage, Estonia adopted a CBA while Latvia had followed a conventional money-based stabilization program before adopting a fixed exchange rate regime in February 1994.1 Notwithstanding these differences, the stabilization performance of the two countries was quite comparable. As argued by Saavalainen (1995), sound fundamentals, rather than the exchange rate regime and currency arrangements per se, appear to have been the key factor underlying price stabilization in both countries. At the same time, the CBA’s fixed exchange rate appears to have hampered the reduction of inflation to levels comparable with those prevailing in industrial countries. Thus, Latvia’s inflation outperformed Estonia’s after the periods of rapid stabilization.

In Lithuania, the introduction of the CBA does not appear to have directly accelerated the decline of inflation. However, it was instrumental in achieving adequate control of the central bank’s net domestic assets. Although inflation started to decline after most administered prices were liberalized in mid-1993, broad money growth did not stabilize until the CBA was introduced in April 1994. As in Estonia, Lithuania’s inflation performance during the CBA period was below that of Latvia.

As Estonia and Lithuania started their economic reforms with an undervalued real exchange rate, their real exchange rates adjusted toward their equilibrium levels through inflation. However, it is difficult to quantify the degree of undervaluation, as various structural reform measures were under way. In Latvia, by contrast, the effect on inflation was mitigated by an appreciation of the nominal exchange rate, as the lats/U.S. dollar exchange rate appreciated by almost 35 percent between the end of 1992 and the end of 1994.

The real value of the Estonian kroon rose sharply during the six-month period after the CBA was established and continued to appreciate afterwards, albeit at a much lower rate (see figure at right). Estonia’s export growth remained strong, however, suggesting that its competitiveness may not have been overly affected. In U.S. dollar terms, its exports of goods and services grew by 75.9 percent and 56.2 percent in 1993 and 1994, respectively. The trade balance, however, was slightly in deficit during the period, mainly because of a rapid expansion of imports financed by foreign direct investment. In Lithuania, the talonas appreciated strongly in real terms before the CBA was introduced, owing to high inflation. After the CBA was introduced, the real effective exchange rate of the litas remained stable.2 The stability of the real effective exchange rate reflects, in part, Lithuania’s slower shift in the pattern of trade toward the West, as compared with Estonia and Latvia. Nevertheless, export growth declined from around 54 percent in 1993 to 10 percent in 1994.

Among the three Baltic countries, the output loss was most severe in Lithuania, in part because of delayed stabilization and the authorities’ relatively limited commitment to privatization and structural reforms.3 Saavalainen (1995) argues that the CBA was one of the factors underlying the better growth performance in Estonia, compared with Latvia. In Estonia, the CBA was perceived to guarantee exchange rate and price stability, thereby leading to larger net inflows of foreign direct investment.4 In 1993, foreign direct investment in Latvia was only $51 million as compared with $154 million in Estonia. During 1994-95, cumulative foreign direct investment in Estonia and Latvia was around $400 million and $300 million, respectively.5 In addition, interest costs in Estonia were lower as interest rates converged at a much faster pace to levels in the international financial market.

Estonia, Latvia, and Lithuania: Real Effective Exchange Rates

(January 1992= I)

Source: IMF staff estimates.

Before the CBAs were introduced, Estonia and Lithuania were already pursuing fiscal consolidation, resulting from budgetary reforms in 1990–1991. Their central banks were explicitly prohibited from lending to the government. Although fiscal discipline continued after the CBAs were introduced, Estonia’s and Lithuania’s fiscal performances were not particularly remarkable and not significantly better than Latvia’s.6 Thus, CBAs were not essential in ensuring fiscal discipline in the Baltics.

1 See Richards and Tersman (1995), Saavalainen (1995), and Camard (1996).2Richards and Tersman (1995) show that since 1992 Lithuania’s real exchange rate appreciated more than that of Latvia and Estonia when calculated using only five industrial countries in the SDR basket.3Saavalainen (1995, p. 19) estimates cumulative output losses between the second quarter of 1992 and the end of 1994 (in percent of their initial levels) as 6.7 percent in Estonia, 11.8 percent in Latvia, and 17.3 percent in Lithuania.4 Other factors included adverse supply shocks, a faster speed of privatization in Estonia, and closer historical and political ties between Estonia, and Finland and Sweden.5 See Ize (1996) for more information on capital inflows.6 In 1994, budget deficits in Estonia and Lithuania amounted to 2.2 percent and 1 percent of GDP, respectively. In view of the financing constraint imposed by the CBA, the Lithuanian government resorted to external borrowing at the end of the year and issued Euro-bonds in the international financial markets in 1995. In Estonia, the fiscal deficit has so far been largely foreign financed. Also, due to the limited financial support from the central government, the City of Tallinn issued Euro-bonds to finance its infrastructure development project in 1996.

Nominal Exchange Rate Rigidity

Most exchange-rate-based stabilization programs have experienced poststabilization booms and some degree of real exchange rate appreciation.27 How-ever, as devaluing in a CBA or abandoning the arrangement entails a substantial loss of credibility, exchange rate misalignments are a more serious cause for concern when CBAs are used as part of a stabilization program. Correction of a real exchange rate misalignment in a CBA could require a prolonged period of tight liquidity and high unemployment that could cast doubts on the CBA’s sustain-ability. Owing to limited price arbitrage, countries with substantial nontradable sectors are particularly vulnerable. This probably explains why most CBAs, with the notable exception of Argentina’s, have been established in small open economies.

In Argentina and Estonia, the real exchange rate appreciation following the CBA’s introduction was indeed substantial, as inflation, albeit declining, continued to be higher than in their main trading partners (Box 5 and Figure 6).28 However, it is unclear whether significant losses of competitiveness have occurred, as residual inflation may have been mainly associated with initially undervalued exchange rates or a systematic “productivity bias.”29 The latter interpretation has been used to explain recent inflation in the Baltics, as well as the persistently high inflation rate in Hong Kong.30 In any event, inflationary biases and the resulting negative real rates of interest—but not necessarily a real appreciation—might have been avoided with a more flexible exchange rate policy that would have allowed for nominal appreciation. Thus, Singapore has had a significantly better inflation record than Hong Kong since it exited its CBA, due to the substantial nominal appreciation of the Singapore dollar against the U.S. dollar (Figures 7 and 8).31

Figure 6.Argentina, Chile, and Mexico: Nominal and Real Effective Exchange Rates

(January 1985 = 100)

Source: IMF staff estimates.

Figure 7.Hong Kong and Singapore: Inflationã1

(In percent)

Source: IMF, International Financial Statistics.

1 Inflation is calculated on a year-on-year basis.

Figure 8.Hong Kong and Singapore: Nominal and Real Effective Exchange Rates

(1980 = 100)

Source: IMF staff estimates.

As frequently observed in exchange-rate-based stabilizations, capital inflows initially delayed the lowering of inflation in most of the recently established CBAs, including Argentina and Estonia. While a CBA’s rule-based environment may attract more long-term capital than conventional pegs, capital inflows can nevertheless also contribute to a rapid credit expansion often associated with a decline in credit quality. The rigid exchange rate rule, which prevents an early exchange rate correction, and the perception that CBAs guarantee macroeconomic stability, which could exacerbate inflows, may increase a CBA’s vulnerability to such problems, compared with conventional pegs.

As with conventional pegs, fixing the exchange rate firmly to the reserve currency can also imply important costs when the economy is subject to shocks or when the value of the reserve currency changes in relation to the currencies of other trading partners.32 In particular, changes in monetary conditions in the reserve currency country may be inopportune to the CBA country if the business cycles of the two countries do not coincide. For example, Hong Kong’s inflation during the early 1990s was partly the result of low nominal interest rates imported from the United States at a time when a rapid economic expansion (driven by China’s business cycle) and an asset-price boom would have called for higher rates (Figure 7).33 In addition, a weakening (strengthening) of the reserve currency can impart a significant inflationary (deflationary) bias if it leads to a depreciation (appreciation) of the CBA country’s currency vis-a-vis the currencies of its other trading partners. Thus, the depreciation of the pound sterling against the U.S. dollar helps explain Singapore’s 1972 decision to abandon the pound sterling as reserve currency and to switch, shortly thereafter, from a CBA to a managed float. Countries with limited trade relations with the reserve currency country are particularly exposed to such shocks. For example, in Djibouti—which is the only CBA country in the sample that has experienced very low real economic growth and which pegs its currency to the U.S. dollar while transacting mainly with France and other EU countries—fluctuations in the value of the French franc in relation to the U.S. dollar have led to wide fluctuations in inflation and competitiveness (Figure 9).

Figure 9.Djibouti: Nominal and Real Effective Exchange Rates

(1980= 100)

Sources: IMF, International Financial Statistics;and IMF staff estimates.

Financial Fragility

CBAs are required to have sufficient foreign reserves to ensure the convertibility of their monetary obligations. Moreover, as pointed out above, CBAs help discourage systemic runs on banks motivated by expectations of an exchange rate realignment.

Nevertheless, the vulnerability of the banking system to incipient runs, when they occur, can increase in CBAs that are not lenders of last resort.34 These risks are particularly important in countries where capital mobility is high but banks have limited access to foreign funds or in countries with weak banking systems.35

In a CBA, the convertibility of the monetary base does not guarantee that of bank deposits. Banks may be hard pressed to honor requests for withdrawals of deposits if they are unable to obtain short-term financing from the central bank. While selling liquid domestic assets—such as government securities—may provide liquidity to individual banks, it cannot solve a systemic crisis as such sales do not increase the amount of reserve money, unless those instruments can be sold abroad or can be rediscounted by the monetary authorities. However, pure CBAs prohibit such rediscounts.

Interest rate increases may subject banks (and borrowers) to excessive stress and damage credibility about the sustainability of the arrangement over time. Although containing of exchange rate uncertainty in CBAs helps moderate interest rate increases, relative to conventional pegs, the recent Argentine and Lithuanian experiences suggest that some interest rates (particularly long-term bank deposit and lending rates) can reach very high levels for significant periods of time, due to expectations of a banking crisis or a collapse of the CBA. Weak bank supervision and unsound banks increase the banking system’s vulnerability to such systemic perturbations.36

As in the case of bank runs and for similar reasons, CBAs, like any fixed exchange rate arrangement, may be vulnerable to bond-led speculative attacks if there is a large outstanding stock of short-term government debt, even if the current fiscal deficit is sustainable. Thus, in addition to exposing the government to a debt crisis, failure to sustain the securities market could also contribute to banking or payments system crises when government debt forms an important component of the economy’s liquidity.37

Loss of Central Bank Functions

The constraints placed on CBAs imply that they cannot undertake certain functions routinely performed by central banks, particularly those related to monetary operations and the payments system.

Monetary Management

Experience suggests that capital flows and interest rate arbitrage may not perfectly substitute for central bank liquidity management.38 Thus, CBAs that abstain from performing monetary operations and instead rely totally on capital flows to regulate liquidity may subject the economy to unnecessary fluctuations. Interest rate arbitrage is limited in the short run, owing to transaction costs, credit risk, market imperfections, and possibly a lack of full credibility in the survival of the CBA. By increasing interest rate volatility, the failure to absorb day-to-day liquidity mismatches can increase intermediation spreads and penalize financial transactions.39 In addition, the CBA’s automatic adjustment mechanism may fail to act with sufficient speed to prevent temporary monetary imbalances from affecting the economy.

While capital account transactions were fully liberalized and domestic interest rates followed the London interbank offered rate (LIBOR) on an average basis in Argentina and Hong Kong, capital flows were unable to fully arbitrage interest rates on a daily basis. Thus, short-term liquidity imbalances were perceived to be detrimental to financial and exchange markets. As a result, both countries were led to develop (or maintain) their capacity to engage actively in open market operations and other day-today monetary operations.40

Having some scope for monetary operations has been helpful in a number of circumstances. In particular, as illustrated by recent events in Argentina, such operations can moderate increases in interest rates in the case of systemic outflows. Sterilization can also limit the adverse effects of capital inflows when these are inflationary and temporary, or mop up excess liquidity injections resulting from LOLR operations.


In the British colonies, most commercial banks operating under currency boards were branches of banks with headquarters in London; consequently, large interbank transactions were settled mainly through their home offices. Local clearing houses, generally managed by a large commercial bank, were established only in places where a substantial demand for domestic settlements existed. In contrast, where CBAs are established, settlement services are provided by the central bank, when the latter exists, or by the private sector, where no central bank exists, as in Brunei Darussalam and Hong Kong.

In Brunei Darussalam, the monetary authorities maintain accounts with three domestic banks for currency settlements. In Hong Kong, where the payments system was inherited from the colonial period, the Hongkong and Shanghai Banking Corporation (HSBC) manages the Hong Kong Bankers’ Association’s (HKBA) Clearing House. The clearing house operates on a two-tier system consisting of ten settlement banks in the first tier and other financial institutions in the second tier. The settlement banks hold accounts at the HSBC for their own purposes and on account of their secondtier customers. The HSBC does not pay interest on other banks’ clearing balances, while the operating cost of the clearing house is shared by all HKBA’s members. As the management bank of the clearing house, the HSBC benefits from having access to information on other banks’ positions.

In cases where CBAs do not provide banks the opportunity to settle in the books of the central bank, some problems can emerge. In particular, this raises the risk of settlement failures, especially if the CBA is unable to provide LOLR services. Furthermore, not settling in the books of the central bank may complicate monetary management. For instance, in Hong Kong, although bank reserves are used to settle financial transactions, the Hong Kong CBA does not guarantee the convertibility of bank reserves at the official exchange rate. Rather it guarantees the convertibility of cash at the official rate. But as cash is an ineffective medium to arbitrage rates in the exchange market, the market exchange rate can in principle deviate from the official exchange rate and monetary interventions are needed to reduce foreign exchange and interest rate volatility.

Constraints on Fiscal Policy

CBAs promote but do not guarantee fiscal discipline. Although the fiscal performance of CBA countries has generally been satisfactory, large fiscal deficits were financed in some cases through an accumulation of public debt and payment arrears. For example, in Djibouti, the government ran budget deficits in every year since 1982. As a result, government arrears amounted to almost 7 percent of GDP in 1993. Some ECCB member governments, which used up their borrowing limits at the central bank after running persistent budget deficits, also accumulated substantial domestic and external arrears. In Antigua and Dominica, new central government arrears in 1994 were equivalent to 4.0 percent and 1.5 percent of GDP, respectively. In Grenada, new arrears amounted to 2 percent of GDP in 1992. Brunei Darussalam has also experienced fiscal deficits since 1992, when government revenue started to decline following a weakening in oil and gas prices. However, budget deficits were financed in this case by a reduction in extra-budgetary funds accumulated from oil and gas revenues.

By restricting the scope for crowding out private spending, high capital mobility enhances the inflationary impact of fiscal deficits in CBAs, even when deficits are financed by domestic debt. By appreciating the real exchange rate, fiscal deficits can thus compromise the CBA’s sustainability. Moreover, a debt crisis could induce the authorities to abandon their commitment to defend the parity. In extreme cases, chaotic fiscal or monetary conditions can occur when CBAs are established in a context of lax fiscal discipline.41 In some cases, treasury IOUs have become an alternative currency that has circulated, at a discount, alongside the official currency. For example, the restrictions imposed by the Convertibility Law in Argentina on deficit financing drastically curtailed the ability of the central government to assist provincial governments. In reaction, with the slowdown in economic activity, in the aftermath of the Mexican crisis, some provincial governments issued promissory notes during 1995 that circulated locally as a means of payment at a discount.42 Of course, significant fiscal tightening could have triggered a similar reaction under a different monetary arrangement.

While CBAs share with all fixed pegs the need for sound public finances, their limited capacity to accommodate fluctuations in treasury cash flows can put an additional burden on public finances. To cope with this problem, the market for short-term treasury bills may be developed, but this takes time and could undermine the CBA’s inherent incentives for fiscal discipline, unless subject to strict limits on outstanding amounts.

In view of these risks and of the lack of discretion in exchange rate and monetary management, fiscal policy acquires particular importance in CBAs. For instance, reliance will have to be placed on fiscal policy, insofar as the authorities wish to dampen the business cycle. In addition, fiscal policy can be used to help restore confidence during an incipient debt crisis, as when Argentina reduced its structural fiscal deficit in the aftermath of the Mexican crisis. Systematic fiscal surpluses can be used to limit the country’s dependence on foreign savings and build up its international reserves position, as in Hong Kong. Thus, fiscal strength and (when needed) fiscal reforms are as much a precondition for establishing CBAs as they are an operational constraint in this environment.

Implications for Entry Conditions

The key weaknesses and strengths of CBAs discussed in the previous sections have implications for the conditions for entry into a CBA. In particular, strong macroeconomic policies are crucial for establishing and sustaining a CBA. A CBA’s key function is to enhance policy credibility, but credibility cannot be maintained—or the CBA sustained—without fiscal discipline.

In turn, a strong fiscal position can contribute to the accumulation of foreign reserves, which is also crucial to enhance the credibility of the arrangement and reduce the likelihood of attacks. The close integration with capital markets that CBAs entail make them particularly vulnerable to shocks from abroad. Excess foreign reserves coverage will let the public know that the CBA can withstand such shocks.43

Policies that encourage sound banking are also important for the success of a CBA. If banks are unable to meet demands for deposit withdrawals, this would put pressure on the authorities to provide liquidity support to banks, which may undermine the CBA. Similarly, its sustainability may well depend upon banks’ ability to withstand large fluctuations in interest rates. Moreover, knowledge of the constraints a CBA puts on the LOLR function of central banks could increase the potential for a run. Thus, a weak banking system may encourage attacks against the CBA.

Flexible labor and goods markets can help avoid an exchange rate misalignment that could bring about undesirable real effects (e.g., an increase in unemployment) and fuel demands for an abandonment of the regime. In this connection, the choice of an appropriate initial exchange rate also helps, as discussed below. Finally, other factors—over which the authorities have little control—can affect the suitability of a CBA, such as the economy’s openness to the rest of the world, its vulnerability to external shocks, and a history of disappointment with policy alternatives.

Clearly, not all the elements discussed above are likely to exist at the time a CBA is set up. However, a sound fiscal policy, a sufficient level of reserves to honor the conversion commitment, and a fairly sound banking system should be in place or be part of the policy package adopted when a CBA is established.44 Other elements, such as strong prudential supervision and flexible labor laws, are also desirable and, if absent, need to be implemented and strengthened, as rapidly as possible.

Determination of the Exchange Rate

Two important and interrelated issues for establishing a CBA are the level of the exchange rate and the size of the backing. Clearly, if the rate obtaining when contemplating the move to the CBA has been kept overvalued through administrative measures, it will need to be devalued to at least a level that reflects market forces. What is more difficult to determine is whether depreciation beyond a rough market clearing rate should be sought, either to reduce the required foreign reserve backing or to build in a cushion for any ensuing real exchange rate appreciation.

In establishing the initial exchange rate, the authorities should take into account inflation inertia and the likely price effect of devaluation and of other concomitant measures, such as price liberalization. However, to knowingly build in a margin for sustained inflation could jeopardize the credibility of the arrangement. While some allowance for inflation inertia and structural rigidities could be allowed, these problems should be addressed directly, in order to engender credibility in the permanence of the arrangement. It should not, therefore, be necessary to build in more than a small margin.45

The second motivation for overdepreciation would be simply to increase the foreign exchange backing to the currency. However, if the current exchange rate is roughly in equilibrium, a devaluation could readily provide an impulse to domestic price inflation that would reestablish the previous real exchange rate—an inauspicious way to introduce the new regime. An alternative would be to allow the CBA to build up its reserve cover gradually or to borrow the reserves through a medium- or long-term foreign loan. If underlying policies are strong, a commitment to build up full reserve cover may have sufficient credibility to establish an effective regime. Argentina followed such a course when it established its CBA. Nevertheless, the credibility resulting from the currency board, per se, as opposed to the constellation of supporting policies, is lessened to the extent that less than full coverage is provided.46

    Other Resources Citing This Publication