Currency Board Arrangements : Issues and Experiences

Charles Enoch, and Tomás Baliño

• Publisher: INTERNATIONAL MONETARY FUND
• Year: 1997
• ISBN:
  9781557756688

NONE

¹ The general literature on CBAs has expanded rapidly in recent years. See in particular Hanke and Schuler (1991 and 1994), Bennett (1993 and 1994), Humpage and McIntire (1995), Walters (1992), Osband and Villanueva (1993), Schwartz (1993), Guitian (1995), Williamson (1995), and Zarazaga (1995).

² Other existing CBAs not covered include those of Bermuda, Cayman Islands, Falkland Islands, and Gibraltar. Hong Kong, China is referred to in this publication as Hong Kong.

³ In addition to the British colonics, currency boards operated in the Philippines, when it was a U.S. colony, and for a brief period, the Italian colony of Somalia. In most French colonies. Instituts d' Emission operated: they backed their currencies with both French franc and domestic assets. For more information on currency boards during the colonial period, see Hanke and Schuler (1994), Schwartz (1993). and Williamson (1995).

⁴ For discussion of this subject, see Guitian (1992) and (1994).

⁵ Currency boards were introduced in the British colonies in the nineteenth century to economize on specie transport costs and to let local governments benefit from seigniorage. The first currency board was introduced in Mauritius in 1849. based on the ideas of the Currency School. In the words of Earl Grey, the British Secretary of State for the Colonies, who is cited in Gunasekera (1962, p.31), currency boards were designed to “unite die ‘advantages of cheapness and convenience which belongs to a paper currency' with the ‘steadiness and uniformity of value, of a metallic currency, … [thus, they] ought to be so regulated that the amount in circulation should vary according to the laws which govern the latter'.” Currency boards became public note-issuing institutions, following the failure of several private note-issuing banks.

⁶ In this regard it may be easier to introduce a CBA in a country where the banking system is relatively small, since this makes it easier to back the currency.

⁷ The central bank’s obligation is to sell U.S. dollars at that rate.

⁸ Obstfeld and Rogoff (1995) argue that the main reason central banks break commitments to fixed exchange rates is not that they exhaust their reserves and draw down all available credit lines but that their respective governments are unwilling to accept the consequences of continuing to defend the exchange rate. In particular, they cite the unwillingness of the Swedish and U.K. authorities to accept a prolonged period of high interest rates to maintain the peg to the European currency unit (ECU) in 1992.

⁹ However, by changing reserve requirements central banks have a limited scope to change the supply of money without changing the monetary base.

¹⁰ “Among the CBA countries covered in this paper, capital controls remain in place only in the ECCB countries. Estonia maintained capital controls up to the end of 1993.

¹¹ It should be noted, though, that capital controls in Estonia during 1992–93 were relatively unbinding and probably had considerable effect only on speculative capital inflows from Russia.

¹² This, however, excludes interest earnings on fiscal reserves deposited at the Exchange Fund, which are transferred to the government.

¹³ In part because of the need to borrow reserves for initial backing and, hence, to repay the loans over time, the Bank of Lithuania was allowed to retain part of total profits as reserves. The Bank of Estonia has the authority to retain at least 50 percent of total profits as statutory and reserve capital; the remainder must be transferred to the budget or used to establish special funds. In Argentina, central bank profits may be maintained in a reserve fund until they reach 50 percent of the bank’s capital; further profits must be transferred to the budget. The ECCB may retain profits or distribute them among member governments in proportion to their shares of currency in circulation. Similarly, the National Bank of Djibouti must distribute profits among its shareholders, which include the central government, local governments, and other government entities.

¹⁴ See Collyns (1983).

¹⁵ As discussed in Saavalainen (1995), however, there is no clear evidence that disinflation was achieved faster in Estonia and Lithuania than under a monetary role and later a conventional fixed peg in Latvia. In Argentina, the introduction of the CBA in early 1991 helped consolidate the downward path of inflation and avoid a recurrence of the hyperinflation conditions experienced in the first quarter of 1990, caused by a flight from financial assets denominated in domestic currency. It should be noted, however, that in all three cases the introduction of a CBA was preceded by significant exchange rate depreciation, the pass-through of which is likely to have delayed the deceleration of inflation.

¹⁶ Marston (1995) suggests that the ECCB countries grew faster, on average, than the other Caribbean countries during 1987–91, in part, owing to their relatively stable inflation. During the period, the average annual growth rate of the ECCB countries ranged from 4.0 percent to 9.5 percent, while that of the other Caribbean countries ranged from 0.4 percent to 3.1 percent.

¹⁷ Recent discussions of self-fulfilling currency attacks can be found in Obstfeld (1994 and 1995), Eichengreen, Rose, and Wyplosz (1994), Drazen and Masson (1994), and Davies and Vines (1995).

¹⁸ However, bank rates (particularly lending rates) could remain above comparable rates in the reserve currency country, due to credit risk or inefficient or unsound banks. See Appendix I for detailed information on interest rate convergence in Argentina, the Baltic countries, and Hong Kong.

¹⁹ Similar credibility benefits are reported for countries that adhered to strict gold standard rules during the period 1870–1914. Those that had poor adherence records were charged considerably higher rates on long-term government bonds, even when denominated in gold, than those with good records (see Bordo and Rockoff, 1995).

²⁰ Economic growth was close to 10 percent in 1984 and averaged 8.8 percent during the remainder of the 1980s.

²¹ Klein and Marion (1994) study a sample of 61 pegged exchange rates in Latin America since the 1950s and find that they had a mean duration of 32 months and a median of 10 months.

²² The Lithuanian CBA came under attack at the end of 1994 and beginning of 1995, following rumors of an impending devaluation and incipient banking system difficulties. The attack appeared to have been encouraged by a weak institutional commitment to defend the parity.

²³ In Lithuania, interest rates remained high after the introduction of the CBA, due to the persistent expectations of devaluation. See Camard (1996).

²⁴ After Ireland joined the European Monetary System (EMS) in 1979, average interest rates on assets denominated in Irish pounds substantially exceeded those in deutsche mark, owing to the possibility of realignments within the EMS. In contrast, during 1928—71, when a CBA was in operation based on the pound sterling, interest rates in Ireland closely approximated those in the United Kingdom. It should be noted that the Irish economy was much more closely linked to that of the United Kingdom than to those of other EMS members. See Honohan (1994).

²⁵ However, in the aftermath of the Mexican crisis deposits at commercial banks declined by around 17 percent, between the end of November 1994 and the end of May 1...