The Cayman Islands is among very few countries in the world - fifteen (15) as of 2003 - with exchange rate regimes that are classified as currency board arrangements (CBA) by the International Monetary Fund (IMF). A currency board is defined as a special case of a rules-based monetary regime characterized by "an explicit legislative commitment to exchange domestic currency for a specific foreign currency at a fixed rate, combined with restrictions on the issuing authorities - the currency board - to ensure fulfilment of its legal obligation" (IMF, 1997).
The advantages of CBAs are well known. They are relatively simple to operate compared to a traditional central bank, assure currency convertibility, minimize exchange rate volatility, and enhance the credibility of monetary policy. CBAs theoretically help insulate the economy from monetary shocks, and are empirically associated with lower inflation and higher output growth (see Ghosh, Gulde and Wolf, 1998). Studies reviewed in this paper and elsewhere also show support for the disciplining effect of CBAs on fiscal policy, although the 2002 exit of Argentina from the CBA and recent theories have raised some doubt on the generality of this effect.
On the other hand, the paucity of CBA adherents may be traced to its strict entry and maintenance conditions. First, an adequate level of foreign exchange reserves is needed to maintain a CBA, although this requirement is less of a challenge for countries with ready access to private capital markets. Second, CBA is a legislated monetary regime and is therefore largely non-discretionary. It rules out the conduct of independent monetary policy in smoothing fluctuations in economic growth, including the use of a flexible exchange rate, and precludes the board from acting as lender of last resort when financial institutions face liquidity problems. Third, this results in the use of fiscal policy being the only available adjustment tool for government, but this, too is required to be disciplined in order to enhance the credibility of the CBA.1
This paper studies the link between fiscal policy and the currency board system in the Cayman Islands. The broader academic motivation for this topic arises from the current literature on the fiscal discipline - exchange rate nexus which offers conflicting theories on the relationship. Empirical tests using cross-country data sets have also yielded mixed results. Their policy implications in turn continue to fuel the debate on whether exchange-rate based stabilization (e.g., CBA) or monetary-based stabilization is more effective in inducing deficit reduction in the public sector.
Current literature suggests the use of fiscal policy as a credibility device and as a stabilizing tool in a CBA (see Fatas and Rose, 2001). The first implies fiscal conservatism while the second implies the opposite behaviour during periods of economic downturns. There are no known empirical studies of this fiscal-CBA relationship in the Cayman Islands. A qualified exception is the study by Fatas and Rose which included the Cayman Islands as one of the 206 sample countries in an empirical test of whether three types of extreme exchange rate regimes yield restrictive fiscal policies. Their results support the hypothesis that CBA countries display more fiscal discipline compared to all other countries in the sample.
Most of the studies on the CBA-fiscal relationship are static in the sense that they imply an automatic disciplining relationship. They do not explain periods when fiscal discipline is tightened or relaxed in CBA countries. In this study, we attempt a different approach and investigate the direction of fiscal policy and how it relates to a measure of the fundamental requirement for maintaining the credibility of the CBA - the availability of foreign reserves for CBA operations.
In addition, two issues specific to the Cayman Islands motivate this research. First, the Cayman Islands has an open capital account and recently went through a period of slow economic growth in 1999-2002 due to the after-effects of the 1997-1998 global financial crisis, and later the 9/11 tragedy. Following traditional analysis, these shocks would have presented pressures for fiscal policy to weaken in order to stabilize the economy, thus compromising the credibility requirements of the country's CBA. We hypothesize, however, that the credibility support role of fiscal policy in periods of low foreign reserves dominates the pressure to stabilize aggregate demand. This is consistent with the law which guarantees the use of government revenues for stabilizing the reserve asset requirements of the currency board.
Second, notwithstanding the aforecited rule, there have been a number of changes in the Cayman Islands legislation underlying the CBA which may have implications on the role of fiscal policy as a credibility device. Specifically, a rise in the allowable amount of external reserves for the purchase of securities of government agencies was legislated in 2002. This change provides potential incentive for the government to borrow from the CBA.
This study finds that fiscal discipline in the Cayman Islands is effectively constrained by expectations on the level of reserves needed to maintain the credibility of the CBA: lower expected reserves triggers higher fiscal surplus. Fiscal discipline is, however, relaxed when growth of reserves is higher. Fiscal discipline is also affected by economic growth but this ultimately depends on the expectations about reserves accumulation. This qualifies the results of traditional analysis suggesting the weakening of fiscal discipline during periods of destabilization as the automatic fiscal stabilizers operate. These results are, however, best taken as tentative in view of the limited number of observations from the available data set for the Cayman Islands.
This paper is organized as follows. Section II provides a brief survey of the theoretical discussions on the topic and some empirical findings. Section III presents key features of the CBA in the Cayman Islands. Section IV discusses a framework for analyzing the direct role of fiscal policy in the CBA, and present the results of a simple test using Cayman data. Section V provides a summary of findings and areas for further research.
Although the establishment of a CBA is legislated, there is no strict legal requirement as to how fiscal authorities should behave in the regime. Theoretically, there are two roles for fiscal policy as summarized in Fatas and Rose (2001): as a credibility device and as a stabilizing tool.
The first role follows from the canonical model of currency crisis (Krugman, 1979) which points to a fundamental inconsistency between domestic policies - typically the persistence of money-financed budget deficits - and the attempt to maintain a fixed exchange rate as the key sources of a crisis. The inconsistency leads to a collapse of the exchange rate when foreign reserves become inadequate. The implication is that, a government cannot peg its exchange rate simply by establishing a CBA without a policy commitment to a stricter fiscal regime. Thus, the IMF, for instance, considers fiscal discipline as an entry condition for countries wishing to establish CBA (IMF, 1997). Similarly, a strong fiscal position is considered crucial in keeping the CBA: this leads to the accumulation of foreign reserves that enhances the position of the CBA and reduce the likelihood of speculative attacks.
There are intrinsic features of a CBA that reinforces fiscal discipline. For one, a CBA prevents the government from availing of seiniorage or monetary financing of a deficit inasmuch as local currency can be issued only against foreign exchange reserves. Alberola and Molina (2000) shows that it is the absence of "fiscal seigniorage" or credit from a central bank, and not a pegged exchange rate per se, which acts as a key instrument for enforcing fiscal discipline in a CBA.
The preceding constraint, however, need not be binding if the government has access to the debt market. In this case, the impulse for irresponsible behaviour can be mitigated by the threat of sizeable adjustments in the fixed exchange rate should fiscal debt becomes unsustainable.
Tornell and Velasco (1994) argue, however, that such a threat does not automatically deter governments because the impact of unsound fiscal balances do not show up immediately, rather they manifest themselves in unsustainable debt and falling reserves after sometime. Their delayed consequences would require adjustments in the economy that fall more heavily on the part of future governments. (In contrast, under a flexible exchange rate system, current authorities will be immediately punished by weakening exchange rates and rising inflation discouraging them from incurring deficits). By implication, a CBA will not necessarily lead to strict fiscal policy. Only in cases where governments are "sufficiently patient" or are aiming for long-term authority will future adjustments have effective deterrent power. Yan Sun (2003) provides a more general extension of Tornell and Velasco's model but contends that there is no inherent correlation between any exchange rate regime and fiscal discipline; rather, the latter depends on structural reforms that will reduce fragmented policy-making.
The second theoretical role of fiscal policy in a CBA relates to its use in stabilizing business cycles. Discussions in standard macroeconomic textbooks hold that under a fixed exchange rate regime, fiscal expansion is the only effective tool for restoring equilibrium in the economy. In an orthodox CBA, the option of devaluation is ruled out by law. The use of fiscal policy for stability implies looser fiscal position in times of economic disturbance relative to a regime with flexible exchange rates.
The two roles present a possible trade-off between the use of fiscal policy as a credibility device and as an adjustment tool. In times of stress in economic growth, the need for stability will tend to raise government spending which weakens fiscal discipline and impact negatively on the credibility of the CBA.
Fatas and Rose (2001) attempted to address the issue empirically by regressing five measures of fiscal policy (total expenditure, total current revenue, overall budget balance, general government consumption and tax revenue - all expressed as percentage of GDP) against dummy variables for three extreme exchange rate regimes (unilateral currency union, multilateral currency union and CBA). Among their key results are as follows: (a) CBAs have relatively conservative fiscal policies which imply that the goal of adding credibility to the monetary regime dominates the stabilization goal. Compared to all other countries in the sample, CBA countries have the smallest total and current expenditures, lowest revenues, and highest budget balance (surplus); (b) CBAs tend to carry out the stabilization role by ensuring the operation of automatic stabilizers such as lower income taxes, and by tilting the composition of their budgets towards components that can provide more social insurance (transfers, subsidies).
Alberola and Molina's (2000) study clarifies the peculiar feature of CBAs which makes them more conducive to fiscal discipline compared to other fixed exchange rate regimes. They show that the latter generally reduced money creation but have no positive effect on fiscal discipline because fiscal seigniorage or the monetary financing of the deficit are not automatically curtailed even when money growth is reduced. On the other hand, legislation and the reserve-backing requirement for monetary creation in CBAs rules out discretionary fiscal seiniorage. The study's empirical test consisted of regressions of the overall fiscal balance, primary balance, monetary seigniorage and fiscal seigniorage against a dummy variable for CBA. Their results show negative effects of CBA for all fiscal variables, although these were not found to be significant for the overall fiscal balance and primary balance.
Similar results of a significant dampening effect of CBA on government spending and positive effect on fiscal balance were found among transition European economies who are recent CBA adopters (Lithuania, Estonia and Bulgaria) by Grigonyte (2003). The effects remain significant after controlling for the independent impact of institutional budget reforms on fiscal discipline. The latter result raises a question on the recent model of Yan Sun (2003) that relies on structural reforms, rather than exchange rate regime, as a sufficient condition for fiscal discipline.
Tornell and Velasco (1994) present evidence where fixed exchange rate regimes had looser fiscal policies compared to flexible regimes based on the stabilization experiences of Latin American and Sub-Saharan African countries. In view, however, of Alberola and Molina's exposition that there are significant differences in the fiscal constraints facing CBAs and other fixed regimes, Tornell and Velasco's results cannot be easily generalized and applied in the case of CBAs.
In summary, empirical studies provide evidence to the hypothesis that a currency board enhances discipline in fiscal policy. The results in turn imply the active use of fiscal policy to support the credibility of the CBA. That is, the role of fiscal policy dominates its stabilization role. However, as Fatas and Rose points out, the results are at best viewed as correlations rather than causal statements.
Williamson (1995) identifies three cases where a CBA is superior to other alternative exchange rate arrangements: (a) where the collapse of confidence in the local monetary authority has been so complete that only the renunciation of monetary sovereignty will serve to restore it; (b) where the economy is small and very open to world trade and finance, as in most historical cases of currency boards, so that the cost of not being able to use the exchange rate as an instrument of adjustment is unimportant; and (c) where a country is determined to use a fixed exchange rate as a nominal anchor in stabilizing inflation, whatever the cost.
The last two conditions aptly describe the conducive factors behind the adoption of the CBA in the Cayman Islands. The CBA in the country became effective in May 1972 based on the statutory authority from The Cayman Islands Currency Law of 1971. The new CBA system succeeds a long history of what is presently called "dollarization" or the local use of foreign currencies as units of exchange. Prior to 1972, the Spanish currency, the British sterling system, and later the Jamaican pound were successively used; the latter became the sole currency legal tender in Cayman Islands starting in 1969 until 1972 (see Johnson, 2001).
The local dollar was initially pegged to the pound sterling in 1972, and allowed to float against the United Sates dollar. However, this was reversed in 1974 following the pegging of the Jamaican dollar to the US dollar amidst the floating of the British currency. The parity with the US dollar was also influenced by the dominance of the US in the Island's external trade.
The announced dominant macroeconomic objective for the CBA was price stabilization, as presented in the government's statement preceding the actual fixing of the parity with the US dollar in 1974 (see Johnson, 2001): "In an endeavour to ease the current high level of inflation, which is largely imported, it has been decided to remove one element of uncertainty by fixing the Cayman dollar exchange rate to the US dollar..."
In 1997, the Cayman Islands Currency Board was merged with the Financial Services Supervision Department under the 1997 Monetary Authority Law into the present Cayman Islands Monetary Authority (CIMA). Despite the merger, the independence of the currency board function was preserved as the Law mandated separation of its assets from the rest of the Authority.
A CBA's credibility vis-à-vis other forms of pegged exchange rate arrangements arises mainly from legislated restrictions on the level of exchange rate and sources of reserve money creation (IMF, 1997). In the case of the Cayman Islands CBA, the Monetary Authority Law does not specifically fix the rate of exchange between the Cayman dollar and the US dollar; instead, it authorizes the Governor in Council to fix the rate in accordance with the Authority's advice. Spot rates quoted by the Authority are restricted to be within a margin from the fixed rate prescribed by the Governor after consultation with the Authority. Notwithstanding the lack of legislation, the fixed rate first established in 1974 (CI$1.00 = US$1.20) remains the official exchange rate to date.
The second feature that anchors the credibility of an orthodox CBA is a clear, legislated rule on the foreign currency reserves needed to back up the issuance of the local currency. This backing rule for the Cayman CBA is legislated but has two characteristics that are not of the orthodox type:
Laws relating to the above features have been subjected to a number of changes over the years. As of 1995, there was no clear-cut foreign currency component of the then-called currency fund which had two components:
Thus, at most thirty (30) percent of the currency reserve backing of the board's demand liabilities may be held in the form of local assets, i.e. investments in government securities or as balances with local banks and the Treasury. In such cases, the board needed to raise only seventy (70) percent of foreign assets to back up local currency issuance.
There have been subsequent changes since then. As of the January 2003 amendment of the Law, the currency backing rule is clearer but gives greater discretionary authority for government financing:
This study quantitatively measures the credibility of the currency board in terms of the reserve backing of its total demand liabilities. Based on data available from the Cayman Islands Currency Board Annual Reports for 1992-1996 and the Cayman Islands Monetary Authority Annual Report for 1997-2002, we examine four indicators of currency backing: (a) total currency reserves comprising external and local assets; (b) total currency reserves net of government security purchases; (c) foreign currency component of total reserves; and (d) foreign currency reserves net of foreign-denominated government securities.
Recorded total currency reserves of the Cayman CBA have historically exceeded one hundred percent of demand liabilities, or an average of 127 percent during 1992-2002 (see Table 1). Recorded backing with external assets comprise 117 percent of demand liabilities on average and have annually exceeded the 90 percent required in the latest amendment of the law (January 2003). With regard to the local asset component of reserves, the recorded percentages were relatively high in 1992 (23% of demand liabilities), and 1993 (32%) but these have since been brought down to single digits in recent years at rates much lower than required by law.
| Table 1: Currency Reserve Backing of the Cayman Islands (% of Demand Liabilities) | |||
|---|---|---|---|
| Year | Total Currency Assets | Total External Assets | Total Local Assets |
| 1992 | 127 | 104 | 23 |
| 1993 | 128 | 96 | 32 |
| 1994 | 116 | 107 | 9 |
| 1995 | 124 | 113 | 10 |
| 1996 | 120 | 110 | 10 |
| 1997 | 133 | 126 | 7 |
| 1998 | 132 | 128 | 4 |
| 1999 | 124 | 123 | 1 |
| 2000 | 135 | 129 | 6 |
| 2001 | 132 | 125 | 7 |
| 2002 | 130 | 122 | 8 |
Source of basic data: Cayman Islands Currency Board Annual Report 1992-1996, Cayman Islands Monetary Authority Annual Reports 1997-2002
To remove the impact of discretionary fiscal policy from total currency reserves, we need estimates of government securities purchased by the currency board. Exact figures for these variables are not available. Hence, we base our estimates on the following special note in the 2002 CIMA Annual Report which indicates the composition of external assets in 2002 and previous years. "The investment managers were given added investment flexibility following Executive Council's approval in May 2001 to invest in mortgage-backed securities issued by US Government agencies and AAA-rated corporate bonds. Executive Council also advised that the Law be amended to allow marketable securities issued by approved government agencies as separate category under permitted external assets. As at the Balance Sheet data, such investments totalled 55 percent of external assets, which exceeded the 20 percent previously allowed under the Law. The relevant amendment to the Law authorizing these investments was passed prior to the Balance Sheet date but was not gazetted until January 2003." (our emphasis).
Considering the above-cited percentages, estimated total currency reserves net of government securities averaged 100 percent of demand liabilities during the period, although it dropped precariously to 63 percent in 2002 as shown in Figure 1. (A crucial estimation assumption is that the 20% ceiling for government security purchases was applied throughout 1992-2001). The drop in 2002 is directly related to, and explained in, the preceding CIMA note.
It is noted that external reserve assets net of government securities for 2002 is estimated at fifty-five (55) percent of total demand liabilities. Taking this into account, and assuming that twenty percent of external assets in previous years were in the form of foreign-denominated government securities, the overall ratio of external assets (net of government securities) to total demand liabilities in 1992-2002 would average 90 percent. This suggests that investment holdings in government securities did not on the whole undermine the required backing with external assets (90% based on the latest version of the law) except in 2002.
For purposes of reckoning foreign currency backing of demand liabilities, we add to external assets the foreign currency (US dollar and British pound) deposits in local banks. In 1992-2000, these averaged 97 percent of local assets. In 2001-2002, the proportion dropped to 53 percent and 50 percent, respectively.
In summary, the foreign-currency denominated reserves of the Cayman currency board during 1992-2002 was approximately 126 percent of total demand liabilities. Deducting the amount of foreign-denominated government securities purchases will bring down the ratio to 99 percent of demand liabilities, although the 2002 ratio fell drastically to 59 percent.
The stance of fiscal policy in this study is simply measured by the traditional overall balance of revenues over total expenditures (excluding principal debt payments).4 As shown below, fiscal policy over the period 1992-2002 yielded surpluses in most years.
It is noted that the average surplus of 0.9 percent of GDP during the period is above the obtained historical average for CBA countries in the Fatas and Rose sample set, as well as those of other studies summarized in Table 2. Historically, fiscal discipline in the Cayman Islands was achieved not so much by a higher-than-average revenue or tax effort in the absence of income taxation, but by lower spending, specifically lower capital expenditures.
| Table 2: Comparative Fiscal Indicators (% of GDP) | ||||||
|---|---|---|---|---|---|---|
| Source | Sample countries; years | Revenue | Expenditure | Surplus | Taxes | Govt Consumption |
| This study | The Cayman Islands; 1992-2002 | 22.0 | 21.1 | 0.9 | 17.2 | 17.1 |
| Fatas and Rose (2001) | 206 countries; 1960-1998 | |||||
| Mean for all sample countries | 23.9 | 28.2 | -3.7 | 19.2 | 15.3 | |
| Mean for CBA countries | 25.5 | 28.1 | -1.9 | 21.8 | 15.0 | |
| Alberola and Molina (2000) | 15 Latin American, 9 European countries; 1972-1998 | |||||
| Mean for CBA countries | -0.71 | |||||
| Mean for all fixed regimes | -1.79 | |||||
| Ghosh, Gulde and Wolfe (1998) | All IMF member countries; 1970-1996 | |||||
| Full sample | -4.2 | |||||
| Pegged | -4.1 | |||||
| Currency Board | -2.8 | |||||
| Gulde-Wolfe and Keller (2000) | EU Accession Countries; 1995-2000 | |||||
| Hard-peg/CBA countries | 36.4 | 38.5 | -2.4 | 32.8 | ||
| Other countries | 42.4 | 45.0 | -2.6 | 36.8 |
Source of basic data for this study: 2002 Cayman Islands Compendium of Statistics and The Cayman Islands Government Accountant General's Annual Reports (1992-2002)
Current literature as surveyed above analyze the impact of fiscal policy on the credibility of CBA by relating the budget constraint of the government to the availability of financing from the board. A cut in the budget deficit necessarily reduces the amount of local government securities (whether foreign or local currency denominated) that can be purchased by the board, and increases the amount of foreign currency and foreign assets held as reserve assets. This analytical framework, however, is limited in explaining the dynamic behaviour of fiscal policy in maintaining the CBA.
Alternatively, a broader framework of analysis is suggested from the balance of payments (BOP) identity, as follows:
ΔR = zBOP
where ΔR denotes change in the foreign reserves of the currency board, 1 ≥ z > 0 is the proportion of the country's balance of payments that is remitted to the board for exchange with the local currency. Total reserves assets for backing the board's demand liabilities rises as the country generates balance of payments surpluses.
Noting that the current account can be segregated into the savings-investment (S-I) gap of the private sector (denoted by subscript "p") and public sector (denoted by subscript "g") and assuming for simplicity that z is constant for all sectors in the economy, we have:
ΔR = z(S - I)p + z(S - I)g + zK
This simple identity summarizes the three sources of changes in reserves for the board: (a) the current account balance generated by the surplus of domestic private savings over private investments; (b) the current account balance due to fiscal policy; and (c) net capital flows. Positive balances from these three sources increase the availability of foreign reserves that the board can use for defending the fixed parity. It is straightforward to see that fiscal discipline which yields S > I in the public sector is expected to have a positive impact on reserves.
However, lack of fiscal discipline (S < I) need not necessarily decrease the amount of reserves as long as the domestic capital market is well developed and able to generate sufficiently high private savings to finance both private investments and a fiscal deficit. Since this is only possible if the interest rate of government securities is higher than the equilibrium interest rate needed to balance private savings and investments, the classic crowding-out effect ensues. However, in an open economy like the Cayman Islands where private investors have access to external financing, this crowding-out effect can be completely mitigated, except perhaps in cases where capital market imperfections such as costly information for project evaluation keeps external capital from financing small domestic investors.
The above simple framework suggests the following:
The preceding safety-net characterization of the fiscal balance in the Cayman CBA is facilitated by the following provision in the Monetary Law:5 "If, at any time, the total assets of the Currency Reserve shall be less than one hundred percent of the demand liabilities, such deficiency shall, so far as possible, be met by transfer from the General Reserves6 and, to the extent such deficiency is not hereby removed, it shall be a liability of the Government, and the Government shall appropriate such funds from the general revenue of the Islands as are required to extinguish such deficiency." (our emphasis).
The above law represents a reversal of the traditional "fiscal seiniorage" associated with traditional non-CBA regimes. In the latter, the monetary authorities provide credit to the government as a form of deficit financing. On the other hand, the Cayman legislation compels government revenue financing of the currency board's liquidity and solvency needs.
The above discussions suggest that the need for achieving S > I in the public sector is influenced by expectations of the board's foreign reserves. Expected increases in reserves relaxes the need for discipline on the part of the fiscal authorities. That is, fiscal balance may decline with increases in expected foreign reserves of the board. However, during times of instability where reserves from private capital flows are expected to decline, fiscal balance is strengthened as a contingency measure for the board.
Evidence for the above proposition is obtained via ordinary least squares regression of the Cayman Islands central government's savings-investment gap or overall fiscal balance against indicators of expected changes in reserves.
In addition, we examine whether the safety-net role of fiscal policy is also affected by economic growth. We would expect that higher growth allows the government to generate more revenues, and counters the tendency for the fiscal balance to weaken in periods of expected increases in reserves.7 However, the relationship becomes ambiguous when economic growth and expected reserves are both falling. The relationship between economic growth and fiscal discipline is not therefore straightforward as it depends on whether expected reserves are rising or falling at the same time as growth occurs.
Accordingly, we have a simple empirical equation:
Yt = a + bXt-1 + cXt-1*Gt-1 + ut
where Yt is the overall fiscal balance alternately measured in levels and as ratio to GDP. Xt-1 is a proxy indicator for the expected changes in reserves, specifically measured by the lagged (previous year's) value of the four indicators of currency backing described above (changes in foreign currency reserves, foreign currency reserves net of government securities, total currency reserves, and total currency reserves net of government securities). Gt-1 is lagged GDP growth. This paper's hypothesis is supported if b < 0. We also expect c > 0, i.e., higher economic growth generates greater fiscal balance and tempers the negative impact of higher expected additional reserves.
Results of the various regressions with the level of fiscal balance as the dependent variable are shown in Table 3; regressions with fiscal balance expressed as a ratio to GDP were also conducted but yielded similar results, hence they are not presented here. Results of regressions without the interaction variable are given in Table 3.
| Table 3: Results of OLS regression runs* Dependent Variable: Central government's overall fiscal surplus |
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| Equations | |||||||||
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | ||
| Constant | 17.57395 | 9.471377 | 18.19659 | 8.65411 | 18.26437 | 13.83068 | 19.11541 | 12.93863 | |
| (4.145256) | (2.139635) | (4.283176) | (1.706245) | (3.683937) | (3.298603) | (3.937752 | (2.764794) | ||
| Lag of Change in Foreign Currency Reserves Net of Govt Securities | -1.777619 | -8.024806 | |||||||
| (-1.908412) | (-3.207722) | ||||||||
| Lag of Change in Foreign Currency Reserves Net of Govt Securities*Lag of GDP Growth Rate | 1.786424 | ||||||||
| (2.597989) | |||||||||
| Lag of Change in Foreign Currency Reserves | -1.550399 | -6.553872 | |||||||
| (-2.052037) | (-3.086172) | ||||||||
| Lag of Change in Foreign Currency Reserves*Lag of GDP Growth Rate | 1.483139 | ||||||||
| (2.448267) | |||||||||
| Lag of Change in Currency Reserves Net of Govt Securities | -1.9094 | -9.13992 | |||||||
| (-1.61743) | (-2.965949) | ||||||||
| Lag of Change in Currency Reserves Net of Govt Securities*Lag of GDP Growth Rate | 1.817606 | ||||||||
| (2.453527) | |||||||||
| Lag of Change in Currency Reserves | -1.723037 | -7.018022 | |||||||
| (-1.862167) | (-2.898783) | ||||||||
| Lag of Change in Currency Reserves*Lag of GDP Growth Rate | 1.41447 | ||||||||
| (2.293693) | |||||||||
| Adjusted R2 | 0.248264 | 0.587268 | 0.286406 | 0.583529 | 0.16806 | 0.515501 | 0.235742 | 0.524927 | |
| DW | 2.105283 | 2.473377 | 2.055481 | 2.426618 | 2.097694 | 1.943595 | 2.050628 | 2.072403 |
* numbers in parentheses are t-stats. Data are annual data for the period 1992-2002. Data on reserves are from Cayman Islands Monetary Authority; fiscal data are from the Cayman Islands Government Accountant-General's Annual Reports 1992-2002.; GDP data are from the Cayman Islands Compendium of Statistics 2002.
As hypothesized, the tests consistently yield b > 0 across all regressions: expected increases to the board's reserves have a downward pressure on fiscal balance. Conversely, fiscal policy tightens up with an expected decline in reserves. These results hold up regardless of how "currency reserves" is measured. For the succeeding discussion, however, we use the results of the best-fit equation, equation 2 where reserves is measured by foreign-denominated reserve assets excluding government securities.
That b < 0 is consistent with findings in other studies reviewed earlier showing the positive impact of CBA as a monetary regime on fiscal discipline (eg. Fatas and Rose) if we assume that fiscal authorities in CBA countries are risk averse, i.e. their expectations are dominated by declines rather than additions to foreign reserves. In the case of the Cayman Islands, the decline in 2001 and 2002 of the expected foreign currency-denominated reserves (excluding foreign-denominated government securities) triggered an increase in fiscal surplus of up to CI$15.3 million (or 77% of total surplus) in 2001, and CI$9.2 million (or 30% of total surplus) in 2002.8
With regard to the influence of economic growth, the coefficient of the interaction variable "c" between reserves and GDP growth is positive across all tests. This suggests that fiscal discipline is strongest during periods when expected changes in both reserves and GDP are negative. In the Cayman Islands, 2001 and 2002 come closest to this characterization, albeit the economy suffered from slower positive GDP growth rather than actual contractions. It is during these two years that the level of fiscal surpluses were highest during the ten year-period.9
On the other hand, simulations using equation 2 explain the high fiscal deficit in 2000 (when GDP grew at the average rate) mainly as a result of an expected increase of foreign-currency denominated assets which reached the highest level during that period. That is, when economic growth was positive and expected changes in reserves was also positive, fiscal policy tended to relax.
This paper supports the findings of other studies showing the disciplining effect of CBAs on fiscal policy. We show, however, that this is not "automatic" as it depends on the authorities' assessment of changes in reserves accumulation. In the case of the Cayman Islands, this is facilitated by the safety-net character of the CBA legislation which guarantees financing of the currency board's deficit by central government revenues. It is noted that this mechanism is a reversal of the monetary credit financing of government deficits in traditional non-CBA regimes. Moreover, it is shown that the influence of GDP growth on fiscal policy is also dependent on the authorities' expectations about reserves accumulation. The need for stabilization is subordinated to the maintenance of the monetary regime. (Despite this subordination, it is worth pointing out that GDP growth in the country is much more stable than the flow of reserves).10
The results of this paper are however suggestive rather than conclusive in view of the limited number of observations used for the empirical test. Further research with a longer time series in the case of the Cayman Islands, as well as its extension to other countries is warranted to make the results more robust. Another reasonable extension is the use of the consolidated public sector position as a measure of total public savings-investment gap, especially if this has significant deviation from the central government's fiscal balance. Extending the research further on how fiscal policy behaves in a cross-section of CBA countries with open vs. controlled capital accounts during periods of stable and volatile financial flows will enrich policy-making in this area.
1 The other requirements for a CBA are a sound banking system which should be resilient to fluctuations in demand for deposit withdrawals, and interest rates; and flexible labour and goods markets that can help avoid exchange rate misalignments that could lead to severe economic adjustments and pressure the abandonment of the CBA (see IMF, 1997).
2 The law (e.g. 1995 Currency Law and the Monetary Authority Law) defines the currency board's demand liabilities to comprise currency notes in circulation plus an amount not less than ten percent of currency coins in circulation; and deposits of local banks and the government. Historically, however, demand liabilities of the board have been limited to currency notes and bills in circulation (see CIMA Annual Reports 1997-2002).
3 Other external assets consist of gold coin or bullion, foreign notes and coins; money on call and deposits in foreign countries; and treasury bills issued by foreign governments maturing within 184 days.
4 Data for 1992-2002 are taken from the 2002 Cayman Islands Compendium of Statistics and the Cayman Islands Government's General Accountant's Annual Reports.
5 Section 28 (7) Monetary Authority Law (2002 Revision). A similar provision is also found in earlier laws of the currency board.
6 The General Reserves is recorded as part of "Reserves and Capital." Under the Law, this must be maintained at 15 percent of demand liabilities to provide additional funding, if necessary, for demand liabilities and other obligations of the board.
7 It may be noted that growth of reserve assets of the currency board exhibits much greater instability compared to GDP growth. For example, growth of total currency reserve assets (net of government securities) in 1992-2002 has a mean of 3.0 % and a standard deviation of 17.1% while real GDP growth has a mean of 3.7% and standard deviation of 1.9%.
8 Estimates are based on the results of Equation 2 in Table 3. Note that the equations use adaptive expectations, hence expected decline in reserves in 2001 equals actual decline in 2000.
9 Fiscal surplus as a percentage of GDP was also highest in 2002.
Alberola, Enrique and Luis Molina (2000), "Fiscal Discipline and Exchange Rate Regimes: A Case for Currency Board?" Banco de Espana Document de Trabajo No 0006.
Aloy, M, B. Moreno and G. Nancy (2003), "Does Fiscal Policy Matter in a Currency Board Regime? The case of Argentina." Paper prepared for Eco Mod 2003: International Conference in Policy Modelling (Istanbul).
Balino, Tomas J.T, Charles Enoch, Alan Ize, Veerathai Santiprabhob and Peter Stella (1997), "Currency Board Arrangements Issues and Experiences," International Monetary Fund Occasional Paper 151.
Cayman Islands Currency Board, Annual Reports 1992-1996.
Cayman Islands Monetary Authority, Annual Reports 1997-2002.
Ghosh, Atish R., Anne-Marie Gulde and Holger C. Wolf (1998), "Currency Boards: The Ultimate Fix?" IMF Working Paper WP 98/8.
Gulde-Wolf, Anne Marie and Peter Keller (2002), "Another Look at Currency Board Arrangements and Hard Exchange Rate Pegs for Advanced EU Accession Countries," International Monetary Fund.
Fatas, Antonio and Andrew K Rose (2001), "Do Monetary Handcuffs Restrain Leviathan? Fiscal Policy in Extreme Exchange Rate Regimes," IMF Staff papers, Vol. 47, pp. 40-61.
Grigonyte, Dalia (2003). "Impact of Currency Boards on Fiscal Policy in Central and Eastern European Countries," Economics of Planning 36: 111-133, 2003 (Netherlands: Kluwer)
Johnson, Vassel (2001). As I See It. How Cayman Became A Leading Financial Centre. (Sussex, England: The Book Guild Ltd).
Krugman, Paul (1979) "A Model of Balance-of-Payments Crises," Journal of Money, Credit and Banking 11 [3]: 311-25.
Sun, Yan (2003), "Do Fixed Exchange Rates Induce More Fiscal Discipline?," IMF Working Paper WP/03/78.
Taylor, John B. "Monetary Policy Implications of Greater Fiscal Discipline" in Budget Deficits and Debt: Issues and Options, Federal Reserve Bank of Kansas City, 1996
Tornell, Aaron and Andres Velasco (1994). "Fiscal Policy and the Choice of Exchange," Inter-American Development Bank Working Paper 303.
Williamson, John (1995). Currency Boards are Not the Answer. (Washington, DC : Institute for International Economics).
This paper is prepared for the University of West Indies' Cayman Islands Country Conference, May 27-28, 2004. Paul Byles and Maria Zingapan are Director of Regulatory and Economic Consulting, and Manager of Economic Consulting of Deloitte in the Cayman Islands, respectively. The opinions in this paper, and all errors are the authors' and do not represent those of Deloitte.
© Paul Byles and Maria Socorro Zingapan, 2004.
HTML last
revised 16th June, 2004.
Return to Conference papers.