A. Introduction

1. Recent macroeconomic developments in Tonga underscore the importance of conducting monetary policy effectively. During the past two years, economic growth has been sluggish but credit has risen rapidly, contributing to double digit inflation and a marked decline in foreign reserves. The National Reserve Bank of Tonga (NRBT) has had relatively limited options to anchor money growth effectively so as to arrest the rapid credit expansion. It has relied on a combination of direct and some indirect instruments for monetary control, but will little success. As a result, gross reserves declined from 3¾ months of import cover at the end of fiscal year 1998/99 to 1¾ months in September 2002 and the value of the pa’anga has declined steadily against the currencies in the basket.²

2. This paper reviews the past conduct of monetary policy, considers the problems in the current system, and discusses the policy options available to the authorities. The main points are:

• Excessive credit growth has threatened Tonga’s external stability in the past and recent monetary developments confirm that direct instruments for monetary control have been ineffective.

• The NRBT’s use of indirect instruments has been impaired by its inadequate capitalization—a problem which has worsened in recent years—and by low interest rates on government securities.

• In the face of current trends in credit growth, significant changes to the monetary policy framework are needed to preserve macroeconomic stability.

B. Evolution of the Monetary Policy Framework

3. Tonga’s financial system is small but has grown with the entry of foreign-owned banks. At present the financial system comprises the central bank, three commercial banks, and the Tonga Development Bank. Despite the few number of banks, the economy is considerably monetized—broad money constituted a little less than half of nominal GDP in 2001/02.³ The NRBT was established in My 1989, prior to which the functions of the monetary authorities were carried out by the Treasury, together with other departments in the government. In the early 1990s, only one commercial bank, the Bank of Tonga, was in operation which was 40 percent owned by the government and 60 percent by foreign banks. The Tonga Development Bank, majority-owned by the government and financed mainly by concessional external borrowing, was created in 1977 primarily to promote the agriculture sector and rural development. With the enactment of the Financial Institutions Act in 1992, two commercial banks—a branch of Australia and New Zealand Banking Group Limited (ANZ) and the MBf Bank, a locally incorporated bank, whose major shareholder is the MBf Finance Company of Malaysia—started operations in late 1993.

4. After its establishment, the NRBT initially relied on direct controls for monetary policy implementation but progressively introduced indirect instruments in the early 1990s. Prior to the NRBT’s establishment, the conduct of monetary policy was explicitly tied to the level of foreign reserves—the Bank of Tonga was required to limit its lending when reserves fell below four months of imports and to completely cease lending when reserves fell below 3 months. The NRBT superseded this by introducing a maximum lending-to-deposit ratio⁴ for the Bank of Tonga—initially set at 60 percent and raised to 65 in September 1990— and interest rate ceilings. The interest rate ceilings were phased out by mid-1991 and the maximum lending-to-deposit ratio was eliminated in early 1993. The NRBT replaced these instruments with two indirect instruments:

• Reserve Requirements: Commercial banks were required to hold a proportion of their outstanding deposit liabilities—initially 5 percent—as non-interest-bearing statutory reserves at the NRBT. The reserve requirement was designed to serve both prudential and monetary policy purposes. At the time it was introduced, the Bank of Tonga was holding considerable excess liquidity and the monetary impact of the reserve requirement was not significant.

• NRBT Notes: In April 1993, the NRBT began to issue notes on a weekly basis as the primary means of influencing liquidity in the banking system. Securities, with maturities ranging from 28 days to 5 years, were issued in early 1994 with interest rates ranging from 1¼ percent on the 28-day issue to 4½ percent on the 5-year notes. The interest rate on the 3-month note was slightly above prevailing inflation rates.

Tongo: Credit Growth and Net Foreign Assets of the Banking System

(3 month centered moving average)

Citation: IMF Staff Country Reports 2003, 037; 10.5089/9781451837452.002.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Tongo: Credit Growth and Net Foreign Assets of the Banking System

(3 month centered moving average)

Citation: IMF Staff Country Reports 2003, 037; 10.5089/9781451837452.002.A001

• Download Figure
• Download figure as PowerPoint slide

Tongo: Credit Growth and Net Foreign Assets of the Banking System

(3 month centered moving average)

Citation: IMF Staff Country Reports 2003, 037; 10.5089/9781451837452.002.A001

• Download Figure
• Download figure as PowerPoint slide

5. Credit growth accelerated with the entry of foreign-owned commercial banks and increased competition in bank-lending. During 1990–93, credit grew sluggishly as the Bank of Tonga kept lending restrained in response to mounting problem loans from the late 1980s and instead invested its excess liquidity in NRBT notes. With the entry of two new commercial banks, competition in the sector increased. Spreads between deposit and lending rates declined, falling by 50 basis points to around 4½ percent in 1994. Increased competition also stimulated bank lending, especially for housing. BOT reversed its strategy of investing in NRBT notes, catering to rising credit growth, and sharply grew its loan portfolio. As a result, credit to the private sector in the economy grew at unprecedented levels beginning around 1994.

6. Faced with rapid credit growth, the NRBT was unable to use its commercial paper effectively to regulate liquidity because of poor profitability and balance sheet problems. The acceleration in credit growth presented the NRBT with the first opportunity to use its commercial paper for monetary policy purposes. However, several weaknesses in the NRBT’s balance sheet prevented it from adequately absorbing the large volume of liquidity needed to contain credit growth. These weaknesses were mainly related to an outstanding claim on the government dating to the creation of the NRBT when the government was to transfer to the NRBT the domestic currency counterpart of the currency in circulation; a weak starting balance sheet position with paid-up capital of only T$1 million and capital and reserves amounting to around 5 percent of total assets; the construction of a new headquarters building in 1993/94 at a substantial expense which absorbed 20 percent of assets and amounted to 3 percent of GDP; and problems associated with foreign reserves management.⁵ Operating profits declined from T$2.17 million in fiscal year 1990/91 to T$0.25 million by 1994/95 and a loss of T$0.5 million in 1995/96. With low capital and net operating losses, the NRBT could not afford to absorb liquidity.

7. Instead, the NRBT increasingly relied on reserve requirements and eventually on direct controls to contain credit growth. With rapid credit growth pressing reserves in 1995, the NRBT finally resorted to raising minimum reserve requirements in December 1995 from 5 percent to 10 percent and persuaded banks to limit credit expansion. These measures helped to bring credit growth down under 10 percent by end-1996. However, credit growth accelerated again beginning 1997 and by mid-1998 net foreign assets of the banking system had fallen to an all time low of close to US$10 million.⁶ With continuing operating losses, the NRBT was unable again to use its notes to contain the credit growth and raised reserve requirements to 12 percent in September 1998. In addition, the NRBT imposed a 30 percent deposit cash margin on all new loans with the exception of those for export financing and tourism. Subsequently, the reserve requirement was increased to 15 percent in September 2000, and in July, the Privy Council approved the use of formal overall credit ceilings marking the reversion to the use of direct controls of monetary. The use of NRBT notes was abandoned in early 2001 and the NRBT resorted to overall credit ceilings and moral suasion to conduct monetary policy. Since June 2002, the NRBT has started to use bank-by-bank credit ceilings in order to better accommodate differences in lending policies across banks.

C. Problems in the Current Monetary Policy Framework

8. Credit ceilings and reserve requirements are the main instruments currently available to the NRBT for conducting monetary policy. Other instruments exist, but are less relevant for regulating liquidity in the banking system. These include a Minimum Lending Rate (MLR) and a Short-Term Liquidity Facility (STLF). The MLF, currently at 12 percent, indicates the minimum rate at which the NRBT would operate with banks but has little operational significance since it is only a minimum rate. The STLF allows the banks to borrow against government bonds and other acceptable securities for short-term liquidity purposes (normally less than a month) and carries, at present, an annual interest rate of 17 percent for the first ten days, and 19 percent thereafter. This facility tends to effectively act as a penalty rate and only affects market rates when banks actively use it. However, most banks have usually maintained excess reserves with the NRBT and, hence, their recourse to this facility has been limited. Therefore, the NRBT has continued to rely primarily on credit ceilings applied to each bank and reserve requirements, but these instruments are also subject to several problems.

9. Credit ceilings tend to be distortionary. International experience shows that credit ceilings are only temporarily effective and are an inefficient means to conduct monetary policy.⁷ Establishing an appropriate credit ceiling for each bank is a challenging task, given the difficulties in assessing credit plans and the liquidity available in the system. In practice, ceilings are often based on historical credit shares of banks and can therefore perpetuate banks’ market shares. In this case, their prolonged use can act to protect inefficient banks by preventing growth in market share of banks with more sophisticated and sound credit risk management. Finally, politically important borrowers tend to be protected from borrowing constraints and the use of ceilings can distort the allocation of credit across customers within and across banks.

10. In addition, the use of credit ceilings has not been effective in Tonga. Since 1995, the NRBT has used credit ceilings, either explicitly or through moral suasion, but only with temporary impact on credit on each occasion. After the reserve requirement was last raised in September 2000, and with the cessation of NRBT notes issuance, direct credit controls have been the sole monetary policy instrument. However, the lack of significant penalties for breaching credit ceilings has rendered these ceilings ineffective—by end-September 2002, three of the four banks were in excess of the ceilings.

11. Relying on reserve requirements is also problematic. A key shortcoming with reserve requirements is that they cannot be used to fine-tune management of liquidity in the banking system. Therefore, monetary policy is constrained from reacting promptly to shocks. Even a small change in the reserve requirement can impose considerable costs on banks in terms of the needed adjustments in bank portfolios. In addition, frequent changes in reserve requirements, aside from imposing considerable adjustment costs on the banks, can also prompt banks to maintain excess reserves at the central bank, thereby reducing the sensitivity of their portfolios to changes in the reserve requirement and undermining the very use of the reserve requirements as an instrument of monetary policy. Another detrimental consequence of long-term reliance on reserve requirements is that, unless they are remunerated, they act as a tax on bank intermediation and can adversely affect the development of the financial system. Therefore, the choice between increasing reserve requirements and using open market operations amounts to who bears the direct cost of monetary policy implementation: the financial system or the government. International evidence shows that development of the financial system is associated with faster economic growth in countries. As such, monetary policy should rely less on indirectly taxing the commercial banks through the reserve requirement and more on open market operations.

12. International experience supports the view that increasing reserve requirements can be useful in one-off sterilization of excess liquidity, but costly if used frequently. As a reflection of the costs associated with frequently changing reserve requirements, their active use for monetary policy purposes has declined in advanced countries. Central banks in these countries seldom change the reserve ratio and over time have lowered or eliminated them. Many developing countries continue to use reserve requirements but primarily to supplement other instruments of monetary policy and not as the primary instrument for monetary policy.

13. In Tonga, raising the reserve requirement has generally provided temporary relief. In both 1995 and 1998, the increase in the reserve requirement was followed by a slowdown in credit growth in the subsequent few months. Credit growth bounced back eventually as deposits grew. By comparison, the recent increase in the reserve requirement in September 2000 was less effective as subsequent credit growth was concentrated in a few specific, large corporate projects in key sectors of the economy.

D. Improving the Conduct of Monetary Policy

14. Given Tonga’s past experience with rapid credit growth, a framework based on open market operations is needed soon to regain monetary control. Restoration of the use of open market operations is key to regulating liquidity more effectively. Key advantages of the use of open market type operations are that they allow for greater flexibility in managing liquidity in the financial system, they allow the NRBT to manage liquidity pro-actively (as opposed to passively as would be the case if the NRBT relied on the short-term liquidity facility), and their use can help to develop a secondary market for securities. The NRBT has issued its notes in the past for this purpose and could re-initiate weekly issues of varying maturities at positive yields for this purpose.

15. Without adequately capitalizing the NRBT, use of open market operations will be fraught with problems, as Tonga’s experience has already demonstrated. The NRBT needs to be recapitalized so that it can bear the financial costs associated with absorbing liquidity from the system. Such recapitalization would require budgetary assistance. In addition, the NRBT has financed budgetary needs increasingly in recent years, as evidenced by the increased holdings of government securities. The NRBT needs to be made more autonomous from budgetary needs to effectively conduct monetary policy.

16. In the transition period to a strengthened monetary policy framework, several measures could be considered to tighten monetary conditions. Despite the imposition of bank-specific credit ceilings, private credit has continued to expand rapidly, growing at an annualized rate of over 25 percent in the first three months (July-September) of 2002/03. Several options could be considered:

• Penalize banks in breach of credit ceilings. The NRBT could impose penalties; however, it is effectively constrained by the NRBT Act from imposing meaningful penalties since approval of the Privy Council would be required for these measures. Alternatively, the government can penalize those banks that breach ceilings by withdrawing its deposits from those banks and shifting them to the NRBT. This would pose a modest cost to the government in terms of the forgone interest from keeping deposits with commercial banks. However, persistent use of government deposits for this purpose is generally not considered appropriate as it raises concerns of transparency.

• Raising reserve requirements. Reserve requirements are already at reasonably high levels and further increases would pose considerable costs on the banks. However, this measure could be used as a last resort option to temporarily freeze credit at its current levels and reversed in the future, once a new framework for monetary policy is established. As a result of increased competition between major commercial banks to gain market share, liquidity has tightened considerably recently, so that an increase in the reserve requirement would likely be effective in inducing banks to shrink their balance sheets. However, the impact across borrowers would likely be quite asymmetrical since recent credit expansion at one major bank is tied to specific corporate projects which may be quite inelastic to higher interest rates. The NRBT could also use temporarily a marginal reserve requirement defined as a reserve requirement at a higher rate on banks deposit liabilities in excess of the amount at a given cut-off date (for instance, the end of the most recent quarter). This would discourage further growth in monetary aggregates without changing the needed reserves on the existing stock.

• Using government bonds to conduct open market operations. Until the NRBT is recapitalized, the government could issue treasury securities for the NRBT to use, on behalf of the government, for the sole purpose of conducting open market operations. However, to use such government bonds for liquidity management, two conditions are necessary: first that the proceeds from such issues are not used for financing the deficit (by, for instance, placing these bonds in an escrow account) and, second, that the securities are issued at positive real yields. A key problem with the current issue of government bonds is their very low yields in real terms has rendered them unattractive for banks and effectively precluded them from being used for open market operations.

17. The experience of selected other developing countries illustrates the transition to the use of indirect instruments of monetary policy. Based on a sample of 19 developing countries, Alexander et. al. (1995), identify the following three stages which countries typically followed in the transition to adopting indirect instruments for monetary control:⁸

• In the initial stage, central banks typically needed to absorb liquidity at their own initiative and relied on the use of reserve requirement to ensure the required sterilization. The use of securities sales at this stage was found less reliable since the lack of development of markets may make banks unwilling to hold such securities. In addition, since reserve requirements affect all banks uniformly and cannot take into account asymmetrical distribution of liquidity across banks, central banks also provided an overdraft or Lombard facility to provide temporary accommodation to individual banks having difficulties in meeting their interbank settlement obligations. Central banks also used credit auctions to provide credit to the banks at their own discretion and eliminated most interest rate controls.

• In the second stage, central banks introduced open market operations often using their own commercial paper and liberalized interest rates further. This helps to develop a secondary market for government securities and can underpin other transactions such as repos as well as provide collateral for other borrowing. At this stage central banks began to reduce reserve requirements as securities operations were able to sterilize excess liquidity. Sales of securities were combined with other instruments—such as short-term credit auctions, a Lombard window, or bill rediscounting—to facilitate both monetary control and money market developments.

• The third stage can take considerable time and refers to the development of efficient secondary markets for central bank and treasury bills. This allows the central banks to operate flexibly and continuously in the secondary markets for securities.

18. Compared to this stylized pattern, Tonga’s experience shows that the transition to indirect instruments was abandoned in the second stage. However, experience also shows that Tonga is not unique in experiencing problems in the transition to adoption of indirect instruments. About half of the countries in the sample considered above experienced some reversal of reforms. In many such cases, however, the causes of the reversals in reforms were externally driven, such as the onset of a financial crisis, which required direct monetary controls to return stability to financial markets (as was the case in Argentina, Chile, and the Philippines). In contrast, the reasons for the reversal of reforms in Tonga are mainly home grown due to the unwillingness of the government to support the use of open market operations by either recapitalizing the NRBT or using treasury bills and bringing on-budget the fiscal costs of monetary policy.

E. Conclusion

19. A new framework for conducting monetary policy effectively is needed in Tonga. Ensuring price stability and protecting foreign reserves will require an effective conduct of monetary policy. This has become especially important in the last two years as inflation has risen to double digit levels and official reserves have been almost depleted. Establishing the necessary conditions for conducting open market operations will help provide a nominal anchor to the economy and allow Tonga to enjoy price stability and assist in economic growth.

A. Three Distinct Phases in Exchange Rate Arrangements

3. Tonga’s exchange rate arrangements have evolved significantly in the past 20 years. Different arrangements have been established to reflect changes in the composition of external trade and trading partners, the structure of production and exports, a gradual but persistent deterioration in macroeconomic fundamentals, and adverse external shocks. Three distinct phases can be distinguished.

4. The first phase, during the 1980s, was characterized by a fixed exchange rate. Tonga’s currency, the Pa’anga, was pegged at par to the Australian dollar. Given the structure of the economy and the exchange regime, domestic inflation was primarily determined by external factors. In particular, the depreciation of the Australian dollar against the New Zealand dollar¹⁰ in the mid-1980s and higher oil prices in the late 1980s played a major role (Table 1). Nonetheless, other domestic factors, such as droughts and cyclones, increases in public sector wages, and government borrowing from commercial banks may have played some role as well. During this phase, although inflation was not only high but also volatile, the economy was able to grow at a rapid pace.

Table 1.

Tonga: Economic Performance

Sources: Official statistics; analysts reports; IFS database; and staff calculations

^1/Figures correspond to averages over the period 1981–1991

^2/The new basket includes the US dollar, the Australian dollar, the New Zealand dollar, and the Japanese Yen. Figures correspond to averages over the period 2000–2002.

Table 1.

Tonga: Economic Performance

		Hard Peg to Australian Dollar, 1980s 1/	Basket Peg (US, Australian and New Zealand Dollars) 1991–2000	New Basket (US, Australian and New Zealand Dollars, and Japanese Yen) Post-2000 2/
Real GDP Growth (percent)		2.4	2.0	1.0
CPI Inflation (percent)		10	3.6	8.6
Average Trading
Partners’ year-on-year CPI Inflation, (percent)		7.0	2.3	2.4
of which:
	Australia	6.0	1.5	3.3
	New Zealand	7.9	1.3	2.2
	United States	4.7	2.8	2.6

Sources: Official statistics; analysts reports; IFS database; and staff calculations

^1/Figures correspond to averages over the period 1981–1991

^2/The new basket includes the US dollar, the Australian dollar, the New Zealand dollar, and the Japanese Yen. Figures correspond to averages over the period 2000–2002.

View Table

Table 1.

Tonga: Economic Performance

		Hard Peg to Australian Dollar, 1980s 1/	Basket Peg (US, Australian and New Zealand Dollars) 1991–2000	New Basket (US, Australian and New Zealand Dollars, and Japanese Yen) Post-2000 2/
Real GDP Growth (percent)		2.4	2.0	1.0
CPI Inflation (percent)		10	3.6	8.6
Average Trading
Partners’ year-on-year CPI Inflation, (percent)		7.0	2.3	2.4
of which:
	Australia	6.0	1.5	3.3
	New Zealand	7.9	1.3	2.2
	United States	4.7	2.8	2.6

Sources: Official statistics; analysts reports; IFS database; and staff calculations

^1/Figures correspond to averages over the period 1981–1991

^2/The new basket includes the US dollar, the Australian dollar, the New Zealand dollar, and the Japanese Yen. Figures correspond to averages over the period 2000–2002.

5. The second phase, the 1991–2000 period, was characterized by greater exchange rate flexibility. In February 1991, the authorities decided to peg the pa’anga to a basket of currencies that included the U.S., Australian, and New Zealand dollars. In addition, the value of the Pa’anga could be adjusted by the NRBT up to 2 percent per month if needed. Despite the increased flexibility, domestic inflation declined markedly, owing to lower inflation rates in major trading partners, at relatively low output costs.

6. Changes in Tonga’s structure of production and exports prompted further adjustments, marking the beginning of the third phase. In July 2000, the currency basket peg was extended to include the Japanese Yen, as squash became more predominant among exports, which were destined to Japanese markets.¹¹ After the collapse of the copra industry in the 1980s, and restrictions on exports of watermelons to New Zealand because of poor disease control, agricultural production shifted to squash and vanilla. Given the nature of external shocks facing the economy, a gradual deterioration in macroeconomic fundamentals, and a decline in the level of official foreign exchange reserves, the monthly limit of exchange rate adjustment by the NRBT was increased from 2 percent to 5 percent in order to preserve competitiveness, boost growth, and ensure external viability.

B. Greater Rigidity or Greater Flexibility?

7. Like many other Pacific Island economies, Tonga is a small open economy which is distant from major markets, with a narrow and undiversified production and export base, and prone to cyclones, droughts, and terms of trade shocks. In addition, the public sector plays a large role in the economy and monetary and financial instruments are undeveloped.

8. On balance, most of the features characterizing the economy would indicate that the current exchange rate arrangement serves Tonga well. The flexibility in the basket peg, given by the diverse currency composition and by the possibility of adjusting the value of the pa’anga relative to the basket peg, provides a good balance between the pros and cons associated with a fixed and a floating exchange rate regime. It could be argued that, as a small and highly open economy, a fixed exchange rate would be preferable for Tonga. However, with production and exports relatively undiversified, a more flexible regime may help the economy better ensure macroeconomic and external stability. More importantly, Tonga has faced largely asymmetrical shocks in recent years, many policy related and others largely exogenous (such as cyclones), which would also support a more flexible regime. The basket peg has provided some nominal anchoring and has also contributed to restrain the rapid growth in domestic demand associated with expansionary macroeconomic policies. In particular:

• With a productive structure highly dependent on agriculture, fisheries and forestries, the exchange rate regime helps the economy insulate from the high vulnerability to climatic conditions. Agriculture, fisheries, and forestries accounts for around 23 percent of real GDP; while tourism is expanding rapidly, and now accounts for around 12 percent of real GDP. Agricultural production is highly concentrated on a few commodities, namely: squash, fish, vanilla, kava and root crops. Most of these commodities are destined for exports.

• While the economy is heavily dependent on remittances from Tongans living abroad, the implications for the exchange rate regime are less clear cut. Empirical evidence suggests that export volume is highly sensitive to both terms of trade and foreign demand shocks, while the volume of imports depend on gross national income and relative prices to an extent that is consistent with estimates for other countries (Box 1). Therefore, preserving adequate competitiveness will significantly impact export volumes and earnings, and assist in maintaining an adequate level of foreign exchange reserves. However, the narrow export base and the low capacity of its earnings to finance imports, implies a persistent trade deficit, which, in Tonga, is mostly financed by private remittances from Tongans living abroad. Private remittances, which averaged around 60 percent of total imports over the 1994–2002 period, have financed imports. Their role in the economy is crucial, and any exchange rate policy must account for it. Private remittances may constitute an alternative or complementary source of wage income that limits, to some extent, the effectiveness of a nominal exchange rate depreciation in achieving a real depreciation through lower wages in real terms.¹² Also, expenditure switching may be somehow limited because domestically produced goods are not perfect substitutes for imports.

Table 2.

Tonga: The Structure of the Economy Percent Change

Source: Tongan authorities; and Fund staff estimates.

Table 2.

Tonga: The Structure of the Economy Percent Change

		2001/02 Preliminary
(In percent of total GDP)
Primary sector		23.1
	Agriculture, forestry, and fisheries	23.1
Secondary sector		12.8
	Manufacturing	4.5
Tertiary sector		46.9
	Commerce, restaurants, and hotels	11.4
Total GDP at factor cost		82.8
Net indirect taxes		17.2
GDP by Sector at 1995/96 Prices		100.0

Source: Tongan authorities; and Fund staff estimates.

View Table

Table 2.

Tonga: The Structure of the Economy Percent Change

		2001/02 Preliminary
(In percent of total GDP)
Primary sector		23.1
	Agriculture, forestry, and fisheries	23.1
Secondary sector		12.8
	Manufacturing	4.5
Tertiary sector		46.9
	Commerce, restaurants, and hotels	11.4
Total GDP at factor cost		82.8
Net indirect taxes		17.2
GDP by Sector at 1995/96 Prices		100.0

Source: Tongan authorities; and Fund staff estimates.

• Policy commitments to preserve a fixed exchange rate regime have weakened. Over the past two years, fiscal policy has been expansionary owing to an increase in wages and higher public spending, bringing the public wage bill to around 15 percent of GDP in fiscal year 2002. At the same time, the ability of monetary policy to focus on price stability has been very limited. The lack of effective instruments for monetary control have allowed for the rapid expansion of credit to continue, putting pressure on the external current account and depleting reserves. In addition, the loss of the Tonga Trust Fund, and the limited access to capital markets, have heightened the need for a more flexible exchange rate regime.

• The structure of the financial system would suggest the need for some form of pegged arrangement. The foreign exchange market is small and heavily dependent on the NRBT as a “market-maker of last resort”. Under a floating regime, the exchange rate could potentially be subject to significant fluctuations if, under certain circumstances, banks withdrew from making a market for the pa’anga. In addition, the absence of instruments for hedging against currency risks implies that these banks, and other private and public enterprises, could face significant losses were the currency to depreciate. Exchange rate volatility could expose the economy to high volatility in prices, given the exchange rate pass-through, as well as in output, given the economy’s narrow production and export base.

C. Conclusion

9. The review of the structure of Tonga’s economy supports maintenance of the present exchange rate arrangement, with arguments in favor of both a greater rigidity and a greater flexibility in managing the basket. However, even though the existence of a high pass-through from foreign prices to domestic prices as well as that of financial costs associated to a depreciation of the currency plead in favor of greater rigidity, the lack of fiscal discipline, the absence of instruments of monetary control and, rigidities in nominal wages plead in favor of greater flexibility to ensure macroeconomic stability. Another argument in favor of the latter is that the economy is highly subject to shocks that affect its supply side, such as terms of trade and unfavorable climatic conditions. Finally, the history of Tonga’s economic performances under alternative exchange rate arrangements has shown that flexibility in managing the pa’anga has served the country well. Therefore, it is argued that flexibility should be maintained.

A. Introduction

1. The loss of most of the assets in the Tonga Trust Fund (TTF) has brought to the fore the importance of establishing sound and transparent management guidelines for off-budget funds. TTF assets, which reached a book value of about US$37 million at end-June 2000, declined to about US$3 million at end-June 2002. The Tongan authorities began legal action in U.S. courts in June 2002 on the grounds of failure to pay or default in payment on a promissory note, fraud, and negligent misrepresentation.¹⁴ This loss has been significant for a small open economy like Tonga, as evidenced by the level of gross international reserves, which stood at about US$ 18 million in June 2002. As a result, Tonga has lost an important means of buffering external shocks and, in the event of future adverse external developments, it would have to rely more on policy measures than in the past.

2. This chapter discusses some of the key features of the framework supporting the TTF, assesses its main shortcomings in light of the best international practices, and draws some lessons for the future. In particular, the chapter aims to highlight the major deficiencies in the management of the TTF, and how these could have been addressed in order to prevent its loss. The chapter presents a comparison of how the TTF was managed against the guidelines for the management of foreign exchange reserves recommended by the IMF.

B. What was the Tonga Trust Fund and How was it Managed?

3. The TTF was established as an off-budget fund to primarily finance specific capital investments or expenditure for the Kingdom. Created on December 16, 1988, through the Tonga Trust Fund Act, the TTF was established to “accumulate moneys to be held in trust as a reserve fund for the Kingdom for use in exceptional circumstances and for future major development projects”. The TTF has been maintained separately from the budget and the official international reserves. It has been managed by a Board of Trustees chaired by the Prime Minister, and includes the Minister of Finance and one other person designated by the King in the Privy Council (the King plus his cabinet).

4. The main source of revenues for the fund was initially the proceeds from the sale of Tongan passports to foreigners, and subsequently, part of the revenue from the lease of Tongan satellite space. The Act established that the TTF should be invested in trustee securities or bank deposits, that an annual statement of receipts and payments be prepared, and that the TTF be audited annually by the Auditor General. Furthermore, it established that the annual statement and the auditor’s report be published in the Gazzette and presented to the Legislative Assembly at the first available opportunity.

5. In practice, the management of the TTF was characterized by a lack of transparency, unclear investment guidelines, and weak governance and public accountability on investments. TTF funds were mainly invested offshore and its operations and investment objectives and results, in terms of portfolio allocation and return, were not sufficiently transparent. Financial statements were usually not published and the fund accounts, while audited by the Auditor General, were not subject to external or more independent auditing. In addition, the investment strategy seems to have resulted in the investment of a significant proportion of the assets in a few high-risk investments.

6. Given the relative importance of the TTF for the economy, its management framework have been a concern for the Fund, which recommended major changes in previous Article IV consultations. These recommendations included: (i) the need to consolidate the TTF with the budget and to make its financial statements public in a timely manner; (ii) the need to establish clear rules for the investment and drawdown of TTF resources, to ensure that its capital base was maintained and to prevent misuse of its resources; and (iii) to limit drawdown of the TTF only as a last resort and for short-term lending to the NRBT, in the event of severe pressure on official reserves arising from short-term external shocks.

C. Some Lessons from Best International Practices

7. Most of the successful international experiences with asset fund management show the importance of establishing prudent management principles, strict and transparent rules, and that the fund is well integrated into the budgetary process. In general, prudent management has been associated with high coordination with government financing operations, offshore investment of funds to avoid large and potentially distortionary effects on the domestic economy, frequent reporting of financial statements and objectives, and independent auditing by external agencies, to facilitate an assessment of its performance and prevent mismanagement. Perhaps, the only possibly prudent principle supporting the TTF might have been that its assets were mainly invested offshore given that its assets represented about 25 percent of GDP.

8. While the TTF was not conceived as a stabilization fund or part of the official international reserves, the IMF guidelines for foreign exchange reserve management provide some useful lessons.¹⁵ Effectively, the TTF provided, in addition to the official international reserves, a cushion against external shocks and other exceptional circumstances. The IMF’s guidelines cover five areas generally regarded as essential for sound reserve management practices, including: (i) clear objectives for the management of reserves, (ii) a framework of transparency that ensures accountability and clarity of reserve management activities and results, (iii) sound institutional and governance structures, (iv) prudent management of risks, and (v) the conduct of reserve management operations in efficient and sound markets. In light of these principles, the experience with the TTF suggests that:

• Asset management objectives have not been clearly defined. The IMF guidelines emphasizes the importance of adequately assessing and controlling for liquidity, market and credit risks and, given these risks, reasonable earnings are generated over the medium to long term on the funds invested. However, the objectives to be met by the TTF were never clearly defined. Over time, its resources were used for a variety of purposes, as prioritized by the Board of Trustees. While funds were mainly invested overseas given their amount and the illiquidity of domestic markets, a desire to increase return led to investing the funds in relatively high risk investments.¹⁶

• Public disclosure tends to be usually minimal. The IMF guidelines suggest the need to disclose publicly objectives and performance. The role of independent external auditors is important in ensuring the genuineness of financial statements when these are publicly disclosed. The audit should be performed in accordance with internationally recognized auditing standards, a critical element of these auditing standards is the valuation of assets at their fair or market value. However, the relationship of TTF operations with many of its counterparts were not disclosed. Audits were performed late, and did not follow international auditing standards (in particular, assets were not valued at market prices). No principles for internal governance were ever established.

• Its institutional framework is weak. The guidelines recommend that such framework should contain a clear identification of responsibilities to help ensure good governance and accountability, as well as that reserves are managed effectively and efficiently. Sound management of internal operations and risks require appropriately qualified and well-trained staff. Effective monitoring of internal operations and related risks should be supported by reliable information and reporting systems. Staff involved in reserve management should be subject to a code of conduct and conflicts of interest guidelines regarding the management of their personal affairs. However, such structure was absent for the TTF. Information and reporting systems were weak, and the independent audit function did not perform its outside vigilance role. No code of conduct or conflict of interest guidelines were established.

• The framework failed to identify and assess risks. The IMF guidelines have stressed the importance of having a framework that identifies and assesses the risks of reserve management operations and that allows the management of risks within acceptable parameters and levels. In practice, reserve management involves a number of financial and operational risks. The experience with the TTF suggests that the framework failed to deal with credit risk and, as suggested by recent legal actions by the government, with control system failure risks, where outright fraud, money laundering, and theft of reserves assets are made possible by weak or missing control procedures, inadequate skills, poor separation of duties, and collusion among staff.

• There were serious deficiencies in choosing a reputable external manager and in establishing a mandate clearly. According to the guidelines, these arrangements should be set out in separate written contracts with each appointed manager to ensure accountability. It is also important that appointment of an external manager should not result in implicitly accepting operations and risks that would not normally be considered, or that are not fully understood. The external manager of the TTF engaged in active investment strategies without an adequate understanding of the risks involved on its part or that of the authorities.

D. Conclusion

9. The absence of clearly defined objectives, poor transparency and accountability of its management and operations, a weak institutional structure, and the absence of a risk management framework have contributed to the loss of most of the assets in the TTF. Putting the framework supporting the TTF on a sound footing would require major changes. It would be important that these changes follow the international best practice.