Argentina - Debt and Debt Service Reduction Loan Project

Document of The World Bank FOR OMCIL USE ONLY Repot No. 13885 PROGRAM COMPLETION REPORT ARGENTINA DEBT AND DEBT SERVICE REDUCTION LOAN (LOAN 3555-AR) JANUARY 20, 1995 Country Operations Division Country Department I Latin America and the Caribbean Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY UNIT - PESO (A$) The peso was fixed at A$1 = US$1 as of April 1, 1991. ABBREVIATIONS DDSR - Debt and Debt Service Reduction DM - Deutsche Mark DRE - Debt Reduction Equivalent EFF - Extended Fund Facility FSAL - Financial Sector Adjustment Loan GRA - Guaranteed Refinarncing Agreement IFIs - International Financial Institutions IRR - Internal Rate of Return LIBOR - London Inter-Bank Offered Rate MYRA - Multi-Year Rescheduling Agreement PDI - Past Due Interest PERAL II - Second Public Enterprise Reform Adjustment Loan PSRL - Public Sector Reform Loan RRP - Report and Recommendation of the President on DDSR Loan TDRE - Total Debt Reduction Equivalent FISCAL YEAR January 1 - December 31  FOR OFFICMIL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation January 20, 1995 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Program Completion Report on Argentina - Debt and Debt Service Reduction Loan (Loan 3555-AR) Attached is the Program Completion Report on Argentina - Debt and Debt Service Reduction Loan (DDSR-Loan 3555-AR). Parts I and III were prepared by the Latin America and the Caribbean Regional Office. Part II has not been received from the Borrower. The DDSR loan was part of a package of external assistance to finance collateral required for a major debt and debt service reduction Agreement under the Brady Plan. Included in this package were set-asides from two other Bank loans, a Second Public Enterprise Reform Adjustment Loan (PERAL II) and a Financial Sector Adjustment Loan (FSAL), as well as support from the IMF, the Inter-American Development Bank and the Eximbank of Japan. The reduction in debt and debt service, linked to ongoing economic reforms, was expected to generate increased confidence, both domestically and abroad, in the government's economic management. The DDSR Agreement was concluded successfully and Argentina's debt and debt service were reduced considerably. The PCR discusses in some detail the direct benefits derived from the Agreement. Estimates of direct benefits, however, are not particularly reliable, given the difficulty of establishing realistic counterfactuals. Furthermore, of much more importance than direct benefits are the potential indirect benefits which come from combining a DDSR program with other structural reforms. Viewed from this perspective, the reduced uncertainty and increased investor confidence generated by the successful conclusion of the DDSR Agreement reinforced the positive impact of other structural changes in Argentina. In particular, the country's external creditworthiness increased significantly as a result. Thus, the outcome of the DDSR loan is rated as satisfactory. Sustainability is rated as likely and institutional impact as modest: in the process a small group of Government specialists increased their expertise in financial engineering. Given the complementarity between the DDSR loan and other structural reform loans, OED will audit this operation as part of a cluster audit including this loan, PERAL II and the FSAL. Attachment ( This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   FOR OFFICIAL USE ONLY PROGRAM COMPLETION REPORT ARGENTINA DEBT AND DEBT SERVICE REDUCTION LOAN (LOAN 3555-AR) TABLE OF CONTENTS PW PREFACE ......................................... EVALUATION SUMMARY . ............................. PART I. PROGRAM REVIEW FROM THE BANK'S PERSPECIIV .... 1 A. Program Identity .............................. 1 B. Background ............................... 1 C. Description of the Debt Agreement ....... ............. 4 D. Results .................................. 8 Debt Reduction ............................... 8 Cost and Benefits .............................. 10 Debt Management Issues .......................... 15 Sustainability ............................... 17 E. The Bank's Role and Performance ....... ............. 17 F. Lessons from Experience ........................... 18 PART II. PROGRAM REVIEW FROM BORROWER'S PERSPECTIVE . . 20 PART III: BASIC DATA . ............................... 21 Annex Argentina - Macroeconomic Accounts .................. 22 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   - i - PROGRAM COMPLETION REPORT ARGENTINA DEBT AND DEBT SERVICE REDUCTION LOAN (LOAN 3555-AR) PREFACE This is the Program Completion Report (PCR) for the Debt and Debt Service Reduction Loan (DDSR Loan) to the Argentine Republic in an amount of US$450 million equivalent. The DDSR Loan was approved on January 5, 1993; fully disbursed on March 26, 1993; and closed on July 31, 1993. The PCR was prepared by the Country Operations Division of Country Department I of the Latin American and Caribbean Regional Office (Preface, Evaluation Summary, Parts I and III). The Borrower's comments on Parts I and III, and contribution to Part II, have not yet been received. No specific comments have been received from the Co-financiers--the IMF, IDB and the Eximbank of Japan. Preparation of this PCR started in February 1994 and is based on the Report and Recommendation of the President, on the loan and implementation agreements, on internal Bank memoranda, and on the Memorandum to the Board 'Analytic Aspects of Debt and Debt Service Reduction Operations" (January 1992).   - iii - PROGRAM COMPLETION REPORT ARGENTINA DEBT AND DEBT SERVICE REDUCTION LOAN (LOAN 3555-AR) EVALUATION SUlMMARY The Debt and Debt Service Reduction Loan (3555-AR) of US$450 million equivalent was fully disbursed in March 1993 to finance collateral required for debt and debt service reduction (DDSR) instruments issued under the Republic of Argentina 1992 Financing Plan ("the Agreement"). The DDSR Loan, together with a total of US$175 million in set-aside funds from the Second Public Enterprise Reform Adjustment Loan (3556-AR) and the Financial Sector Adjustment Loan (3558-AR), was an integral part of support for the implementation of the Agreement from the IMF, the IDB, the Eximbank of Japan, and the World Bank which made possible the successful completion of the Agreement. Under the Agreement, all of eligible commercial debt (US$28.2 billion) was restructured- -about one half of Argentina's total public external debt. Argentina exchanged US$19.3 billion in principal debt for two new instruments, respectively involving debt reduction (discount bonds) and debt service reduction (par bonds). Both instruments are fully collateralized as to principal, and two semi-annual interest payments are secured through collateral instruments with a matching maturity held by the collateral agent. In addition, Argentina regularized past due interest (PDI) of US$8.9 billion by making a downpayment of US$700 million and exchanging the remainder at par against non-collateralized floating rate PDI bonds. The Agreement's implementation achieved a commercial debt reduction equivalent of 38 percent of eligible debt (including eligible PDI). Additional official lending of about US$2.1 billion was provided. The total debt reduction equivalent was 15 percent of total public external debt. The Agreement achieved debt reduction at a substantially lower financial cost to Argentina than an alternative market-based debt reduction operation would have entailed. These savings are estimated at about US$3.7 billion, or 1.4 percent of GDP. Implementation of the Agreement required Argentina to provide up-front enhancements at a cost of about US$3.8 billion. Evaluating costs and benefits against a counterfactual involving debt rescheduling and a PDI arrangement as included in the Agreement leads to the following results: the net present value of financial savings amounted to US$5.3 billion, or about 2 percent of GDP. The internal rate of return over the stream of financial savings is about 18 percent, which is well above the Bank's critical level of the social discount rate used for project lending. This indicates significant direct benefits to Argentina from the Agreement. In addition, the Agreement generated substantial indirect benefits as a low and permanent ceiling on future payments to the banks reduced uncertainty about future fiscal and exchange rate developments.  - iv - Against a market-consistent counterfactual involving partial default, the evaluation of the Agreement recognizes financial costs of about 1.2 percent of GDP and associated negative direct benefits. However, much higher indirect benefits must be ascribed to the Agreement under this counterfactual (compared to the rescheduling counterfactual), because of the link between DDSR and economic reforms in this case. A positive net result is established. The Agreement reduced the overall burden which the possibility of a future shortfall in public external debt service imposes on creditors as a group. Among the reduced debt burden, the share of commercial creditors declined and the share of official creditors increased, primarily that of the bilaterals. Debt servicing flexibility declined due to the exchange of loans against bonds which the Agreement explicitly exempts from future restructuring. The implications for the Bank's exposure risk appear manageable due to greatly improved prospects for the sustainability of Argentina's economic program. The Agreement could not have been concluded without the lending support of the Bank and the other donors. The Bank's performance in coordinating support, evaluating the project, and processing the DDSR Loan and the set-asides was outstanding. With the Agreement, the Government successfully completed a well-designed strategy for ending the debt crisis. The Government fielded a highly competent team for negotiating and implementing the Agreement. The analysis of the Argentine DDSR operation confirms lessons from previous DDSR operations. Specifically: Brady operations should be undertaken only if a reform program is in place; the concerted and menu-based approach to debt reduction leads to significant cost savings to the debtor country compared to market-based debt reduction; due to the voluntary nature of Brady deals, creditor banks improve their financial position; and there is a need for more flexible pricing of the par bond option, given the objective of a balance between debt reduction and debt service reduction. The experience of the Argentine operation also indicates that inclusion of a new money option in the menu of options is not essential for obtaining both full participation of creditors and substantial cost reduction compared to market-based debt reduction.  PROGRAM COMPLETION REPORT ARGENTINA DEBT AND DEBT SERVICE REDUCTION LOAN (LOAN 3555-AR) PART I. PROGRAM REVIEW FROM THE BANK'S PERSPECTIVE A. Progrm Identity Program Name Debt and Debt Service Reduction Loan Loan Number 3555-AR RVP Unit Latin American and the Caribbean Region Country Argentina Type of Loan Support for debt and debt service reduction (DDSR) B. Background 1. In the late 1970s, Argentina joined other middle-income countries in borrowing heavily from international markets at low interest rates. External indebtedness increased six-fold during 1976-83 to about US$45 billion--approximately 35 percent of GDP and 470 percent of exports. Public external debt issues covered persistent fiscal deficits resulting from a weak tax system, losses of public enterprises, generous industrial promotion and a military build-up. Private debt accumulation to a large part financed capital flight as residents sensed the unsustainability of a macro-economic program combining fiscal expansion with exchange rate-based disinflation. In early 1981, devaluation caused private debtors to default. Over the following five years, the Government progressively assumed nearly all private long-term external debt. 2. Debt Crisis. Rising international interest rates in the early 1980s lowered the ceiling on country debt commercial banks were willing to hold. The intemational debt crisis, triggered by the Mexican debt moratorium in 1982, quickly spread to Argentina. Voluntary credit was cut off and extemal financing became limited to official lending and negotiated, involuntary commercial debt reschedulings combined with modest amounts of new money. Serving interest on the existing, mostly public debt henceforth required Argentina to absorb less than it produced, and to transfer commensurate resources from the private to the public sector. 3. Adverse direct and indirect effects of the debt crisis on investment and nolicies worsened Argentina's economic decline during the 1980s, which basically resulted from delays in overdue economic reforms. Lack of external financing directly reduced domestic investment. Investors also were discouraged by uncertainty about future fiscal and foreign exchange developments generated by often contentious debt rescheduling negotiations, and their failure to generate sustainable debt service schedules. Partial term credit agreements were achieved in 1983 and 1985, but the Government repeatedly fell into interest arrears. A comprehensive Guaranteed  Financing Agreement (GRA)--with 20 years maturity, 4 years grace and a reduced spread of 13/16 percent over LIBOR--was reached in August 1987. However, it became soon apparent that the monthly interest service of about US$250 million was fiscally unsustainable; the Government suspended payments to the banks in April 1988. 4. Lack of fiscal adjustment produced further adverse indirect effects because the required external adjustment and internal resource transfer had to be brought about by other, highly distorting means. The military government used foreign exchange restrictions to generate the necessary shift in the resource balance--about 5 percent of GDP between 1980/81 and 1982/83. In 1984, the constitutional government of President Alfonsin established a scheme of quantitative import restrictions (QRs) covering virtually all competing imports. Reopening of the economy subsequently met with the strong resistance of vested interests. A new import liberalization process started in 1987, but proceeded only gradually, and was reversed in the turmoil of the final months of the Alfonsin Administration in 1989. 5. Most damaging proved the Government's inability to generate the necessary internal transfer of resources in a non-distortive manner. Starting from a high primary fiscal deficit, this required the authorities to rapidly implement deep public sector reforms. Preoccupied with the South Atlantic war and its aftermath, the military authorities made no attempt at fiscal adjustment in 1982/83, instead relying on Central Bank financing. To avoid the inflation tax, residents responded by withdrawing from the domestic financial system and reducing currency holdings. This rendered the macroeconomy inherently more unstable. The Alfonsin Administration conducted four stabilization programs between 1984 and 1988, but each time inflation surged to higher levels. Significant deficit reductions were achieved, but were not sufficient relative to available financing. In the absence of public sector reforms, furthermore, deficit reductions relied on unsustainable or inefficient means such as non-payment of pensions and supplier bills and taxes on trade and financial intermediation. High and unpredictable inflation became the main impediment to a recovery of savings and investment. In March 1989, the Alfonsin Government was forced to abandon its fourth stabilization program. Having lost all policy credibility, it was subsequently unable to prevent a run on the currency followed by hyperinflation. Monthly inflation peaked at 200 percent in July; real annual output declined about 6 percent in 1989. 6. Economic Adjustment. The shock of hyperinflation and recession provided the newly elected President Menem, who assumed office ahead of time in July 1989, with the necessary broad political support for addressing the root cause of Argentina's economic decline. This also would generate the conditions for ending the debt crisis. The new administration--with intellectual and lending support from the Bank--enacted profound structural reforms to eliminate the fiscal deficits that were fueling inflation.' Over the following 2-3 years, the Government 1. Studies prepared for the new Government included Tax Policy for Stabilization (1989), Reforms for Price Stability and Growth (1989), and Provincial Government Finance (1990). Lending included the Public Sector Reform Loan (PSRL, US$325 million, July 1991), which was cofinanced by the Inter-American Development Bank, and the complementary Tax Administration Technical Assistance Loan (US$6 million, January 1989) and Public Sector Reform Technical Assistance Loan (US$18 million, July 1989). The privatization program was supported through two Public Enterprise Reform Adiustment Loans (PERAL I in February 1991 and PERAL II in January 1993, each in an amount of US$300 million). In addition, the Bank made a Financial Sector Adiustment Loan (FSAL, US$400 million, January 1993).  - 3 - extended the value added tax while removing inefficient taxes on trade and financial intermediation, modernized the tax administration and public financial management, downsized the public administration, suspended costly subsidies, and privatized nearly all public enterprises. To strengthen incentives for private investment, the Government accelerated trade liberalization and deregulated markets for goods and services. Fiscal strengthening facilitated the April 1991 Convertibility Law which fixed the exchange rate and disciplined monetary policy by requiring the monetary base to be fully backed by international reserves. Consumer price inflation fell from 1355 percent in 1990 to 84 percent in 1991, 18 percent in 1992 and 8 percent in 1993. Real output recovered 18 percent in 1991-92 after a 6 percent decline in 1989 and zero growth in 1990. 7. Debt Strategy. Linked to the Menem reform program was a strategy to end the debt crisis by obtaining a low, permanent ceiling on debt service to the banks. To that end, the existing commercial debt stock (about US$31 billion plus US$5.5 billion in accumulated interest arrears at end-1989) needed to be significantly reduced, a settlement for past due interest be negotiated with the banks, and a substantial part of the remaining debt be transformed into fixed rate instruments. The divestiture of public assets in the context of privatization offered a possibility for retiring debt at a discount. However, the Government was aware that the scope for market-based debt reduction would be limited. Progress in the reform program would increase the market valuation of Argentine debt by raising expectations for debt repayment. In addition, more-than-marginal debt reduction operations would increase market prices as the remaining debt would be perceived as more likely to be serviced. As debt sellers would be forward-looking, such operations would take place at the higher post-operation price. (As described below, the Government retired about US$11 billion of commercial debt in 1990-92. The secondary market price of Argentine debt increased from 13 cents (per dollar of face value) at end-1989 over 20 cents at end-1990 and 38 cents at end-1991 to 48 cents at end-1992.) 8. The Government set its sight on the Brady plan which had been successfully demonstrated in the 1989 debt agreement between the Mexican Government and its commercial creditors. Central to the Brady approach was the notion that a severely indebted country and its creditor banks as a group have a common interest in major debt and debt service reduction (DDSR) linked to economic reforms. This would lead to more and better domestic investments, the gains of which can be shared. However, agreements on market-based debt reduction cannot achieve the common objective--because of coordination problems between banks (free riding cannot be prevented); and because the sovereign debtor country cannot credibly commit to the reforms which, in turn, leads the banks to demand heavy front-loading. The Brady plan promised to overcome these obstacles with a concerted and menu-based approach. Official support was offered for comprehensive, menu-based DDSR agreements between the debtor country and the banking community. Such agreements would minimize free riding by: (i) tailoring the menu of options to the needs of the various banks: and (ii) requiring nearly full participation, reinforced by the exchange of debt holdings for new instruments considered senior to loans. Official support would mitigate the up-front liquidity crunch and would be conditional on economic reforms.  9. The Menem Government could not immediately access the Brady plan for two main reasons. The cash flow impact of an agreement would have been highly negative compared to actual zero interest payments; with the primary balance still in deficit, payments would have been fiscally unsustainable. Relations with commercial banks, furthermore, needed to be restored. While the fiscal position gradually strengthened in 1990-91, the Government used the privatization program for debt-public equity conversions retiring about US$6.5 billion of commercial debt in 1990-91 (and a further US$4.5 billion in 1992). This repaired relations with the banks which proved eager to participate in the conversion operations. Relations improved further when the Government resumed partial interest payments in June 1990, at a monthly rate of US$40 million, raised to US$60 million in January 1991. By end-1991, fiscal improvement and debt reduction appeared sufficiently advanced to seek negotiations with the banking community about a comprehensive Brady-deal. C. Description of the Debt Agreement 10. Negotiations between the Government and the creditors' Working Committee on the restructuring of principal debt and the regularization of past due interest (PDI) commenced on January 30, 1992. The Government unilaterally increased monthly interest payments from US$60 million to US$70 million, effective April. An "Agreement in Principle" was announced on April 8, 1992; the final term sheet--called the Republic of Argentina 1992 Financing Plan (the Agreement)--was distributed on June 23, 1992; and was signed on December 6, 1992. The principal debt restructuring component of the Agreement was concluded on April 7, 1993, and the PDI component on October 29, 1993. However, about 5 percent of the eligible principal debt and about 13 percent of the eligible PDI had not been reconciled at the respective closing dates; these were settled on September 27, 1993 and December 29, 1993, respectively. 11. Terms of the Agreement. The Agreement covers all eligible principal debt and PDI held by commercial banks--about half of Argentina's total public external debt. Principal debt holdings could be exchanged for either of two exit instruments: collateralized floating rate discount bonds; or collateralized fixed rate par bonds. Regarding eligible PDI, which includes interest on unpaid interest, participating banks had to accept on a pro-rata basis: a downpayment of US$700 million; and a par exchange of the remainder against non-collateralized floating rate PDI bonds having a backloaded amortization schedule. The terms of the Agreement are described in more detail in Table 1.  -5- Table 1: ARGENTINA - TERMS OF DEBT AGREEblENT | lMaturity Interest ?/ | l Instruments 1/ I (years) I (Dercent) C Collateral Comnents Menu of Options 1. Par Bonds (with 30 yr. bullet Yr. 1: 4.00% Principal secured by zero Registered form. DM below-market Yr. 2: 4.25% coupon U.S. Treasury option issued in bearer interest rates) Yr. 3: 5.00% securities. TweLve-month form with 5.87% Yr. 4: 5.25% rolling interest interest; equivalent Yr. 5: 5.50% guarantee at assumed rate cost of collateral. Yr. 6: 5.75% of 6%, secured by cash or Eligible for debt-equity Yr. 7: to permitted investments. conversion. Can be maturity: 6% repurchased at any price if not in default. 2. Discount Bonds (35% 30 yr. bullet LIBOR + 13/16% Same as for Par Bonds, Same as for Par Bonds, discount on face but assumed rate of 8% but interest at LIBOR + vaLue) for interest guarantee. 13/16% for DM option. Past Due Interest (PDI) 1. Cash payment of For USS400 million face US$400 million value of PDI. 2. USS300 million in None; new instruments to For USS300 million face cash or new be repurchased on the value of PDI. instruments S Closing Date. 3. PDI Bonds 12-yr. LIBOR + 13/16% None. Bearer form. (par exchange) maturity, 3- Payments 1-7 = 1%; yr. grace, 19 payment 8 = 5%; semiannual payments 9 - 19 = 8%. payments 1/ Denominated in US dollars; Dl options described under "Comments.M 2v Payable semiannually. 3/ Subsequent to the issuance of the Term Sheet, Argentina agreed to replace collateralized Argentine zero-coupon bonds having a market value of USS300 million with USS300 million in cash or new instruments at the option of the holders of eligible PDI. New instruments would be repurchased on the Closing Date. Source: Republic of Argentina, 1992 Financing Plan. 12. With a menu of only two exit options, Argentina's creditors were given less choice than participants in other DDSR agreements. Notably absent was a new money option without DDSR which in other agreements served the dual purpose of helping to finance the enhancements and accommodating banks wishing to remain exposed because of high exit costs. The Government was confident that financing could be obtained at lower cost. The Agreement accommodated banks not wishing to mark their portfolio to market--the main source of high exit costs--by requiring the Government to issue on request instruments that would never be listed on any stock exchange, but would be freely assignable. Outside the Agreement, the ongoing debt-equity conversion program provided debt holders with an additional exit option. Debt holdings committed under the Agreement could be reassigned to such conversions until 75 days prior to the closing date.  - 6 - 13. Under the terms of the Agreement, the par bonds, discount bonds and PDI bonds are exempt from future requests for new money or reschedulings. The Government, furthermore, is entitled to repurchase these bonds at any time and price provided it is not in payment default on any of them. Also, these instruments are eligible for debt-equity conversion. 14. Implementation. The Agreement achieved virtually full participation resulting in the restructuring of US$19.3 billion in principal debt and the regularization of US$8.9 billion in PDI (Table 2). About 66 percent of principal debt was exchanged for par bonds and 33 percent for discount bonds. However, these shares were obtained only after the Government and the Working Committee had impressed on the banks the need for modifying their first selections, which had resulted in a split of 85/15 between par and discount exchanges. As discussed below, this indicates a mispricing of options. About 2 percent of principal debt was exchanged for bonds denominated in Deutsche Mark, which was the only currency option other than US dollar. TABLE 2: DEBT EXCHANGE UNDER THE ARGENTINE DEBT AGREEMENT Eligible Debt New Debt US$ Million Percent US$ Million DM-denominated' Principal Debt 19,266 100.0 16,951 1.9 Par Bond 12,653 65.7 12,653 1.3 Discount Bond 6,613 34.3 4,298 3.9 Past Due Interest 8,913 100.0 8,213 -- Cash 700 7.9 - PDI Bond 8,213 92.1 8,213 -- Total 28,179 25,164 1.3 1) in percent of new debt. 15. The Government provided the required enhancements--principal and interest collateral for the par bonds and discount bonds plus the downpayment on PDI--at a total cost of about US$3.8 billion. All interest collateral and about 60 percent of principal collateral was purchased in the market, with the remainder directly acquired from the US Treasury or Kreditanstalt fuer Wiederaufbau. Total costs were US$27 million higher than estimated in the Report and Recommendation of the President (RRP). This was the net result of an increase in collateral costs per unit of debt exchanged due to falling interest rates, and a reduction in the amount of eligible principal debt due to ongoing debt-equity conversions. Table 3 provides a breakdown of costs of enhancements and sources of financing.  TABLE 3: ARGENTIA - USES AND SOURCES OF ENHANCEMENT FUNDS' PDI Par Bond Collateral Discount Bond Collateral Total Expected Downpayment Principal Interest Principal Interest End-1993 Total 2 IMF 986 986 Augmentation 460 460 460 Set-asides 522 4 526 526 IBRD 625 625 Additional Lending 450 450 450 Set-asides 175 175 175 IDB 475 475 Additional Lending 124 276 400 400 Set-asides 75 75 75 Japan Eximbank 473 473 800 Government 700 419 3 78 1,200 873 Total 700 1,476 729 522 332 3,759 3,759 Additional Lending 1,057 726 1,783 2,110 Set-asides 522 254 776 776 Government 700 419 3 78 1,200 873 1. Situation at end- 1993. 2. After full disbursement of funds committed by Japan's Eximbank. 16. The Government was the sole source of funding for the US$700 million downpayment. Direct financing for the purchase of collateral was provided by the IFIs and the Government. In addition, Japan's Eximbank supported the Agreement indirectly by committing a total of US$800 million in parallel financing with the IMF and cofinancing with the Bank and the IDB, which disburses against imports. Ahead of the settlement (April 7, 1993), the IMF disbursed US$986 million, the Bank US$710 million, the IDB US$475 million and Japan's Eximbank US$443 million; the balance (US$445 million) was funded by the Government. As described below, the Bank's contribution was subsequently reduced by US$85 million. On the other side, Eximbank disbursed an additional US$30 million in parallel financing between the settlement date and year's end. At end-1993, the balance funded by the Government was therefore US$500 million. (This is reflected in Table 3.) Eximbank disbursed a further US$125 million in March 1994; together with the expected disbursement of the balance of Eximbank's commitments, this would reduce the Government's contribution to the funding of collateral purchases to US$173 million, which would be in addition to the US$700 million downpayment. 17. The Bank's contribution ahead of the April 7 closing date consisted of the DDSR Loan (US$450 million) and set-asides of US$100 million from the PERAL II and US$160 million from the FSAL--all disbursed March 25-26, 1993. Regarding the FSAL, the loan agreement set aside US$200 million for the purchase of discount bond collateral, with the proviso that amounts deemed by the Bank not to be necessary for this purpose could be reallocated by the Bank to the financing of imported goods. US$40 million were reallocated and disbursed against imports on March 22, and US$160 million were disbursed as set-asides three days later. As  - 8 - described above, the April 7 settlement of principal claims was incomplete because some claims were still being reconciliated. Ahead of the September 27 settlement date, it became apparent that some US$85 million in FSAL set-asides were not needed. Accordingly, a further US$85 million were reallocated on September 22, 1993, which lowered the Bank's total contribution to US$625 million. 18. Funds from the Bank were used consistent with Bank guidelines.2 Particularly, all proceeds of the DDSR Loan financed par bond interest collateral, and all set-asides from the PERAL II and the FSAL financed discount bond collateral. The Govemment did not see a need to make use of a modification in the guidelines which would have allowed the use of additional lending for the purchase of principal collateral for par bonds.3 D. Result 19. The analysis of likely results in the RRP indicated the Argentine Brady operation would: achieve debt reduction and cost savings (against an open market operation) on the scale of the Mexican debt accord; generate significant financial savings and associated direct benefits as well as indirect benefits against a counterfactual involving multi-year debt rescheduling; and greatly reduce the burden on extemal creditors as a group, but also render debt service less flexible increasing the risk to the Bank in case of adverse developments. As discussed in more detail below, the analysis of actual results confirms the extent of debt reduction and cost savings of concerted and menu-based DDSR relative to market-based debt reduction. Financial savings and direct benefits tumed out lower than estimated in the RRP, but indirect benefits appear to be higher. Substantial net benefits of the Agreement to Argentina also are confirmed by an evaluation using an altemative debt service scenario consistent with market expectations of partial default in the absence of the Agreement. Argentina's good macroeconomic performance in the year following the conclusion of the Agreement may have mitigated the increase in risk to the Bank resulting from the commercial debt exchange, by improving the outlook for the sustainability of Argentina's economic program. Debt Reduction 20. The Agreement resembles the successful Mexican DDSR operation in the extent of debt reduction achieved (Table 4) and costs saved compared to market-based debt reduction of the same amount (Table 5). 2. 'Operational Guidelines and Procedures for Use of IBRD Resources to Support Debt and Debt Service Reductions'. R89-104, May 22, 1989. 3. 'Review of the Program to Support Debt and Debt Service Reduction". R92-34, March 9, 1992.  -9- Concluding the Agreement reduced Argentina's commercial bank debt by the equivalent of about US$10.8 billion, or 38 percent of eligible debt including PDI4. Net of additional official borrowing, the total debt reduction equivalent is US$8.8 billion, or 15 percent of total public external debt. The operation occurred at an average market price of 50 TABLE 4: ARGENTINA - DEBT REDUCION EQUIVALENT Commercial Total TDRE in Present Prepayment Debt Debt percent of Eligible Face Value Value of Equivalent Reduction DRE in Additional Reduction public Debt of Debt Interest of of Equivalent percent Official Equivalent extend (ED) Reduction Reduction Collateral (DRE) of ED Lending (TDRE) debt Argentina 28,179 3,015 4,732 3,059 10,806 38 2,110 8,696 15 Discount Bonds 6,613 2,315 -- 1,061 3,376 51 Per Bonds 12,653 -- 4,732 1,998 6,730 53 PDI Arrangement 8,913 700 -- -- 700 8 Mexico 47,170 6,034' 7,090 7,166 20,290 43 3,732 16,558 17 Venezuela 19,011 755' 2,491 1,729 4,975 26 687 4,288 13 1. Nct of new money. 2. Relative to market rates for fixed interest rate loans prevailing at the time of the operation. TABLE 5: ARGENTINA - DEBT PRICES' Average Pre-deal for Deal Post-deal (P2-P1)/P2 Eligible Debt (po)2 (Pl)3 (P2)4 xlOO PI-Po (US$M) 5 Argentina 39 50 63 21 11 28,179 Mexico 36 41 52 21 5 47,170 Venezuela 37 50 61 18 13 19,011 1. Cents per dollar of face value. 2. For Argentina, P0 is the average price in the quarter before negotiations on the Brady deal were announced; for Mexico and Venezuela, P0 is the average price in the month before the Brady plan was announced. 3. Market value of new portfolio as percent of face value of eligible debt including PDI. 4. The stripped post-deal price P2 is estimated as the market value of debt after the deal minus the prepayment equivalent of collateral, plus the amount of cash used for buybacks or donor-payments, divided by the face value of eligible debt minus the DRE, plus the amount of cash used for buybacks or donor-payments. 5. Face value. cents on the dollar--estimated as the market value of the banks' new portfolio (including enhancements) divided by the face value of eligible debt and PDI. The price of the Agreement is about 21 percent lower than the post-deal market price at which an open market operation with 4. The commercial debt reduction equivalent (DRE) includes: US$3.0 billion in face value reduction through the discount bond exchange and the PDI downpayment; US$4.7 billion in net present value of interest service reduction through the par bond exchange; and US$3.1 billion in prepayment equivalent of collateral placed beyond Argentina's reach. Not included are about US$400 million in interest reduction between January 1, 1992 and the closing date resulting from the interest arrangement included in the Agreement.  - 10- a like amount of debt reduction would occur, indicating savings of about US$3.7 billion or 1.4 percent of GDP5. Cost and Benefits 21. Counterfactuals. The RRP evaluated the Agreement against an altemative scenario of contractual debt service on principal debt rescheduled through two consecutive MYRAs and PDI as arranged in the Agreement. Rescheduling counterfactuals have been used in the presentation of other Bank loans in support of DDSR agreements, but have been faulted for being inconsistent with market expectations of partial default reflected in the pre-deal prices of country debt.6 Obviously, a DDSR operation generates financial savings and associated direct benefits compared to contractual service on rescheduled debt, but financial costs and negative direct benefits if default is the alternative. Positive net benefits of DDSR therefore require sizable indirect benefits against a market-consistent counterfactual, but not against a rescheduling scenario. Depending on the counterfactual selected, the evaluation of a DDSR agreement would therefore lead to opposite net results if the operation failed to generate significant indirect benefits. 22. Indirect benefits of DDSR can be expected if economic reforms address the structural problems that led to the debt crisis in the first place. The RRP correctly pointed out that Argentina's reform program was already advanced prior to the Agreement and included difficult to reverse reforms such as privatization. The selection of the rescheduling counterfactual was based on the judgement that the Government would be unlikely to put at risk the benefits of previous and planned reforms, and would therefore seek to regularize relations with commercial banks also in the absence of the Agreement. While not consistent with market expectations, the Region could make such judgement on the basis of better insight into the Government's intentions. Consistent with this reasoning, the RRP did not ascribe benefits of the reform program to the Agreement, as would be necessary in an evaluation against a counterfactual involving partial default. 23. Financial Savings. The decline in international interest rates since the writing of the RRP drastically lowered the interest relief related to the par bond exchange. As shown in Table 6, subsequent cash flow savings against the rescheduling counterfactual remain modest in the 1993-2000 period, but increase in the long-term when rescheduled debt would have to be amortized.7 In net present value terms, direct financial savings from the Agreement amount to 5. The post-deal price is approximated by the price of the non-collateralized portion of the banks' new portfolio. This stripped price is calculated as the market value of the new instruments (including the PDI downpayment) minus the prepayment equivalent of the collaterals, divided by the face value of the new instrunents (including the PDI downpayment) minus the DRE. See: World Debt Tables 1993-94, Vol. 1, Box 3.3; 6. Claessens, S., I. Diwan and E. Fernadez-Arias (1192), 'Recent Experience with Commercial Bank Debt Reduction," World Bank, PRE Working Paper, No. 995. 7. The Bank's LIBOR projection used in the analyses declined from 7.0 percent to 5.8 percent.  - 11 - approximately US$5.3 billion--about 2 percent of GDP. The severe impact of the up-front costs on liquidity is greatly mitigated by additional official lending. The RRP estimated that this would suffice to generate a positive overall cash flow impact already in the 1993-95 period. With lower interest relief, however, Table 6 indicates that the overall cash flow impact will be slightly negative for the 1993-2000 period. Higher international interest rates, of course, would improve the cash flow impact. However, the Bank's medium-term LEBOR projections, which underly the cash flow depicted in Table 6, are consistent with the recent interest rate increase. TABLE 6: ARGENTINA - DEBT AGREEMENT: CASH FLOW AND FINANCIAL SAVINGS Cash Flow Financial Savings 1993 1994 1993-2000 2001-23 NPV IRR Debt Agreement (1) Up-front Costs 1' 3,759 0 470 0 (2) Debt Service2' 338 1,131 1,461 1,285 (3) Additional Net Transfers 31 1,728 210 39 -76 Rescheduling Counterfactual (4) Alternative Debt Service ' 1,121 1,337 1,853 1,744 (5) Subsequent Cash Flow Savings (4-2) 782 206 392 459 (6) Financial Savings (5-1) -2,977 206 -78 459 5,336 17.71 (7) Cash Flow Impact (6 +3) -1,249 416 -39 382 Market-consistent Counterfactual -3,10V 1.02 1/ Collateral purchases plus PDI downpayment. 2/ Interest payments on new mstnuments plus amortization of PDI bonds, minus interest earnings on interest collateral and release of interest collateral in 2023. 3/ Disbursement of additional lending (US$1,783 million in 1993 and US$327 million in 1994) net of interest and amortization. 4/ Two consecutive MYRAs for principal debt (with total grace period of 8 years and stepped-up amortization over 16 years) and PDI arrangement. 5/ Eligible debt (including PDI) evaluated at difference between average market price of debt reduction operation and pre-deal price of debt. 24. In a market-consistent evaluation of the Agreement, direct financial costs to the country (in net present value terms) are equivalent to commercial banks' gains from the Agreement. The gains/costs are indicated by the difference between the average market price at which the Agreement was concluded--about 50 cents on the dollar as described above--and an appropriate pre-deal price of eligible debt not reflecting expectations of the deal (Table 5). That price is difficult to establish both because market participants were aware of the Govemment's intention to access the Brady plan and because prices were moving upwards due to ongoing debt-equity conversions and good policy performance, as discussed above. Selecting the average price in the fourth quarter of 1991--before negotiations with the banks were announced--gives a price difference of 11 cents on the dollar. This puts the financial costs against the market-consistent counterfactual at US$3.1 billion--about 1.2 percent of GDP.  - 12 - 25. Direct Benefits. Financial savings due to the Agreement directly affect the liquidity of the public sector, because the debt being converted is owed by the Government. Given constraints on public consumption expenditures and a ban on Central Bank financing, the Government could react by adjusting public investment; due to complementarities with private investment, the total investment impact would be magnified. The authorities could instead adjust domestic debt, which would affect private investment through the effect on domestic interest rates. A summary of the direct benefits of the Agreement is provided by the internal rate of return (IRR) of the stream of financial savings. For direct benefits to be positive, the IRR must exceed the domestic discount rate. As shown in Table 6, the IRR is about 18 percent against the rescheduling counterfactual. While lower than the IRR of 23 percent estimated in the RRP, this is still well above the domestic discount rate. The critical level used in the Bank's project lending is 10 percent. The result of significant direct benefits of the Agreement therefore is confirmed against this counterfactual. 26. Against the market-consistent counterfactual, the estimated IRR is just 1 percent. ( Since up-front enhancements exceed the net present value of financial costs, subsequent cash flow savings against this counterfactual are positive.) The IRR is way below the critical level of the domestic discount rate, thus indicating significantly negative direct benefits. 27. Indirect Benefits. DDSR operations are expected to improve investment incentives and macroeconomic policy by lowering the "tax" on future output and the uncertainty over future payments associated with the debt overhang, and increase resource transfer by increasing country creditworthiness. As discussed below, the notion of a tax on future output appears to have little relevance for Argentina. Indirect benefits resulting from reduced uncertainty and increased creditworthiness, however, are pertinent for Argentina, particularly if the alternative scenario involves default. Lower domestic interest rates due to reduced uncertainty would also lead to indirect financial savings in countries with substantial public debt denominated in domestic currency; however, this is not the case in Argentina. 28. The debt overhang hypothesis states that increasing output of a highly indebted country enables creditors to extract proportionally higher actual debt repayments. This would give rise to a "tax" on domestic investment and reform efforts, which in turn would be reduced through DDSR. For the hypothesis to be relevant in the context of the Argentine debt deal, the net present value of expected service on risky debt would need to be large as a fraction of output, and decline as a result of the operation. Valued at the pre-deal price of 39 cents, eligible debt amounted to about US$11 billion--barely 4.5 percent of GDP in 1993. In the month following the conclusion of the Agreement, the market value of risky commercial debt--the non- collateralized portion of the banks' new portfolio described above--was also US$11 billion. Any debt overhang "tax" therefore appears small in the first place, and not affected by the Agreement. 29. Uncertainty about fiscal and exchange rate developments adversely affects credibility and sustainability of macroeconomic policy, and causes investors to hold back investments that would be irreversible. High and volatile contractual payment obligations to the banks are a major source of uncertainty, which is addressed by the Agreement. Specifically: the net present value  - 13 - of payment obligations is reduced by 38 percent as described above; the need for often contentious rescheduling negotiations is eliminated; and about 50 percent of interest payments to the banks are protected against future interest rate shocks. A positive effect on uncertainty is indicated by developments of interest rates (Figure 1 ) and exchange rate risk (Figure 2 ) during periods in which new information on the debt deal became available. Between the quarter preceding the announcement of negotiations (1991,4) and the quarter following the Agreement in Principle (1992,2) lending rates declined 10 percentage points and exchange rate risk dropped 8 percentage points. Between the quarters preceding and following the conclusion of the Agreement (1993,1 and 2), interest rates and exchange rate risk fell 7 and 2 percentage points, respectively. Figure 1: Argentina - Infladon and Lindng Rat.. Figue 2: Argentina - Exohang- Rate Fdc (90 day P.- DOCA Aw- Ce.&-RdP- depi ts ) 50 DD'o Au- 20 40 - Cw,binhd 0 Price Inflation 1S Intermst Rate 0 _ _ _ _ 0 10'i - = 2- = E = _ s E 2 _ s _ P1gw. 3: A,gentina - Inettuion i Inv-stor Creit # af h1gv 4: Arg-nt0na - Foreign Regowo. Transfers */ Rating. GDP 40 L DOS JW_n 2 _ 30 i2 35 2 20 Ratings GDP~~loiaArrss 15 10 25~~~~~~~~~~~~~~~~~~~~~~~~~ 0 -3 0 'x-3880o 000i)3 I. T X .4D ~~~~ ~~~~~~~ ~~~~~~~ ./~~~~~~~a Detinod as Reourc Deficit phi. Liquid Resarn. Accunulation.  - 14 - 30. Uncertainty had already been reduced prior to the Agreement as a result of fiscal reforms and the convertibility program, but also the Government's strategy for regularizing relations with commercial creditors, as described above. This achievement would likely be undone by continued partial default--rather than debt rescheduling--in the absence of the Agreement. Accordingly, against a market-consistent counterfactual the Agreement is associated with higher indirect benefits. 31. Since 1990, Argentina has experienced a dramatic improvement in its external creditworthiness and ability to attract foreign financing, mainly as a result of domestic developments. The country's creditworthiness as measured by the semi-annual rating of the Institutional Investor suffered a low of 18 percent in 1990(2), when it was 21 percentage points below the average for some 100 countries (Figure 3). It has since increased each semester and reached 36 percent in 1994(1), just 1 percentage point below the global average. Although the global average was affected by the entry of countries of the Former Soviet Union, the closing of the gap between the ratings still suggests that the improvement in Argentina's creditworthiness was predominantly due to domestic factors. Consistent with increasing creditworthiness, Argentina's ability to attract foreign financing improved, as indicated by the development of the adjusted resource trar,sfer from abroad, i.e., the resource balance plus reserve accumulation (Figure 4). As a share of GDP, the adjusted resource transfer increased from -5 percent in 1990 to 4 percent in 1993. While the increase is partly due to external factors such as the fall in LIBOR rates and sluggish activity in the OECD space, recent analysis strongly suggests that in the case of Argentina domestic developments have been more important.8 Domestic sources in the case of Argentina include good policy performance, a coherent strategy for regularizing relations with commercial creditors, and the achievement of a low and permanent ceiling on payments to the banks through the Agreement. What role is ascribed to the Agreement, obviously depends on the counterfactual. It would be modest against an alternative scenario assuming that relations with the banks will be regularized through rescheduling. Against a counterfactual of involving default, the contribution of the Agreement would be recognized as crucial, because the debt strategy would in that case be abandoned and macroeconomic policy would loose credibility. 32. Development Impact. Direct and indirect benefits of DDSR contribute to long-term changes in economic growth, investment and savings (development impact). The Bank's projections for 1993-2000 compared to the record for 1988-92 indicate a 2.5 percentage point increase in the growth rate, and upward changes in the GDP shares of gross investment and national savings of 3.7 percentage points and 2.2 percentage points, respectively. Given the impact of external developments and the complex link between domestic policy performance and DDSR, it is difficult to quantify the development impact of the Agreement. However, it appears 8. A decomposition analysis of the increase in net portfolio inflows into 13 middle- income countries between 1989 and mid-1993 indicates that on average only 14 percent can be ascribed to domestic factors. Dominant domestic contributions are shown for only three countries: Argentina (63 percent), Mexico (80 percent) and Korea (63 percent). E. Fernandez-Arias (1994), "The New Wave of Private Capital Inflows: Push or Pull?', Mimeo, World Bank.  - 15 - safe to say that a Brady deal for Argentina would likely not have been agreed without the previous dramatic improvement in domestic policies; and would not benefit development if it did. As argued above, the reverse may also hold in the case of the market-consistent counterfactual, because continued default would undermine the benefits of previous adjustment and the sustainability and credibility of macroeconomic policy. In this light, the Agreement can be viewed as having met a necessary, but not sufficient condition for development. Against the rescheduling counterfactual, the RRP estimated an increase of 1 percentage point in the growth rate based primarily on direct benefits associated with a positive overall cash flow impact in the sort-term and the medium-term. As described above, this was erased by the fall in international interest rates since the writing of the RRP. However, output and investment actually grew much stronger in 1992-93 than projected in the Debt Agreement scenario of the RRP, and the Bank's long-term growth rate projections have been raised by one-half of a percentage point. It is widely agreed that the Agreement and its successful conclusion have the main cause for the strengthening of investor confidence underlying this development. This suggests that the RRP took insufficient account of the indirect benefits of the Agreement. 33. Net Benefits. Positive net benefits of the Agreement are assured against the rescheduling counterfactual as financial savings of about 2 percent of GDP add to the development impact, which may be on the order of one half of a percentage point increase in the growth rate. In the case of the market-consistent counterfactual, financial costs of about 1.2 percent must be recognized. The adverse development implications of default, however, strongly suggest a positive net result of the Agreement. Consistent with the market-consistent counterfactual, net benefits may also be estimated strictly on the basis of market valuations, as has been proposed by various authors.9 In this approach, the net benefits of a Brady operation are approximated by the financial savings against an open market operation involving the same amount of debt reduction. As shown above, these savings are approximately US$3.7 billion, or 1.4 percent of GDP. Debt Management Issues 34. The Agreement's implementation and improved economic performance reduce the burden which the possibility of a future shortfall in public debt service imposes on Argentina's external creditors as a group. The share of preferred creditors in the reduced burden has increased, however, and debt service has become less flexible. With regard to the Bank's exposure risk, this may outweigh the beneficial effect of a reduced overall burden. 9. Bulow and Rogoff show that, under reasonable assumptions, in an open market buyback creditors can expect to capture all indirect benefits resulting from the reduction of the debt overhang. This suggests that the gross benefits of a Brady operation can be approximated by the increase in the market value of the banks' portfolio following an open market buyback of equal size; and net benefits of the debtor country by the financial savings of the Brady operation against the open market operation. J. Bulow and K. Rogoff (1991), "Sovereign Debt Repurchases: No Cure for Overhang", Quarterly Journal of Economics, Vol. CVI, pp. 1219- 1235.  - 16 - TABLE 7: ARGENTINA - EXTERNAL DEBT SERVICE RATIOS Average Average Percentage 1988-92 1993-2000 Change Share of GDP Total DS 5.4 3.5 -35.6 Public DS 5.1 2.2 -55.8 Multilateral Creditor DS 1.3 0.7 -50.7 IBRD DS 0.3 0.2 -40.0 Share of Exports Total DS 56.4 49.7 -11.8 Public DS 52.6 31.8 -39.6 Multilateral Creditor DS 14.0 9.5 -32.5 IBRD DS 3.5 2.8 -17.9 Share of Federal Revenue Total DS 42.5 27.4 -35.5 Public DS 39.7 17.5 -55.8 Multilateral Creditor DS 10.6 5.2 -50.6 IBRD DS 2.6 1.6 -40.0 Share of Public DS Multilateral DS 26.6 29.7 11.7 Bilateral DS 3.0 17.0 474.1 Private Creditors DS 70.4 53.3 -24.3 Inflexible Foreign DS as Share of l/ Public DS 43.9 79.4 80.9 Total DS 44.2 73.3 65.7 1/ Inflexible public DS includes DS to multilaterals and holders of Government bonds including new instruments under the Agreement. Total inflexible DS also includes private bond issues and interest payments on short-term debt. 35. The reduction in the overall burden on external creditors results from the Government's increased willingness and capacity to service the remaining obligations. The option of default has become more costly because the Government has already provided the enhancements, and substantial benefits from the Agreement have become apparent. The increased capacity to pay is reflected in the drastic reduction--between 1988-92 and 1993-2000--in public debt service as a share of federal revenues (55 percent), GDP (55 percent) and exports (38 percent) as depicted in Table 7.  - 17 - 36. Public debt service to preferred creditors also is shown to decline sharply as a share of GDP, exports and federal revenues, though less than for external creditors as a group (Table 7). The associated increase in the debt service share of preferred creditors by 2.8 percentage points appears modest, however, compared to a 14 percentage point increase in the share of bilateral creditors. Arguably more important for the relative burden on preferred creditors is the exchange of commercial bank loans, which are potentially subject to reschedulings and new money calls, against bonds which the Agreement explicitly exempts from future restructuring. This could reduce the Government's flexibility in addressing future debt servicing difficulties should downside scenarios be realized. As shown in Table 7, the public debt service share of preferred creditors and bondholders increases by 36 percentage points to nearly 80 percent 37. Reduced flexibility increases the risk to the Bank in case the economic program is not sustained and debt servicing difficulties arise, as discussed below. It is worth noting that the Government in such case would have room for reducing other expenditures to sustain payments to preferred creditors: the fiscal program includes expenditure increases that could be cut; the interest collateral could be used up to pay up to 12 months in interest on the discount bonds and par bonds, without the Government being obligated to replenish the collateral account; and, as a last resort, the Government could seek to renegotiate registered bonds in a force majeure situation. Sustainability 38. After three successful years, the Government's convertibility program is still on track. The key factor has been growth in private sector confidence brought about by a series of structural reforms, and establishment of a low and permanent ceiling on the Government's payments to commercial creditors. The main risks to the program are associated with possibly higher than expected international interest rates; failure to improve competitiveness through productivity growth; and unforeseen political developments threatening fiscal equilibrium. Strongly rising international interest rates could reverse capital inflows inducing a sharp economic contraction with adverse consequences for the financial system and public finances. Over the medium-term meeting growth expectations crucially depends on strong export growth. In the context of the legally fixed exchange rate, this requires competitiveness to be restored through productivity growth above rates achieved by international competitors. The authorities would otherwise come under pressure to devalue the exchange rate, which would carry its own inflation risk, in the context of a highly dollarized economy. A weakening of fiscal discipline for political reasons constitutes another risk, in the context of a very open capital account. In such an event, international reserves and therefore the monetary base could decline sharply, which would severely test the convertibility program. However, memories of the hyperinflation make it unlikely that the electorate will endorse in the near future populist policies that would undermine the hard-won gains in fiscal discipline. E. The Bank's Role and Performance 39. Bank support for the Agreement was crucial for two main reasons. Under the currency board arrangement, the Government could not apply reserves to the funding of enhancements.  - 18 - Eligibility for the Bank's Program to Support Debt and Debt Service Reduction and similar programs of other donors was therefore a necessary requirement for negotiating the Agreement; and timely disbursement of funds was essential for settlement. Bank staff advised the Government early in the process that fiscal sustainability of the outcome would be a key criterion for the Bank's support, as would be continued progress in the reform agenda. The Bank took the lead in coordinating donor support. In this, the Bank's early and frequently updated analysis of the merits of the deal and of funding needs proved critical, particularly with regard to timely lending development of the IDB and Japan's Eximbank. The Bank's technical expertise was also crucial for coordinating the required negative pledge waivers; and for the disbursement of funds- -other than those of the IMF--into the collateral accounts on the exact date needed for the purchase of the collateral instruments. Secondly, the Bank's seal of approval was critical for the participation of commercial creditors. Once the final term sheet was agreed, the Bank provided a comfort letter to the banking community, and participated in the presentation of the Agreement in the main financial centers. 40. The Bank kept close contact with the Government throughout the process and provided advice. Coordination of donor support was flawless in all phases. Also, intra-Bank coordination of the DDSR Loan, the PERAL II and the FSAL was a challenging task, which was performed effectively and efficiently, i.e., without in any way compromising the quality of the adjustment loans. The outstanding quality of the RRP for the DDSR Loan was recognized by the Executive Directors. The contributions of the Bank's Cofinancing and Financial Advisory Services, Legal Department, Loan Department, and Cash Management Department all were of high order. F. Lessons from Experience 41. The Agreement was the sixth in a series of debt accords between middle-income debtor countries and their commercial creditors under the Brady plan. The experience of the Agreement confirms lessons of the previous operations relating to the linkage between DDSR and economic reforms; the superiority of the concerted, menu-based approach to resolving the debt crisis; gains to creditors; and inefficiencies in the pricing of options. It also contradicts arguments to the effect that the menu must include a new money option to obtain debt reduction at below-market prices. 42. In DDSR operations the debtor country incurs a heavy up-front cost to reduce and stabilize its contractual obligations to creditor banks. To be worthwhile to the country, however, economic reforms must address the structural sources of the debt crisis. Without reforms, partial default as expected by the market would be the appropriate counterfactual. As shown for the Agreement, financial savings and direct benefits of the operation would in such case be highly negative, and so would be net benefits in the absence of indirect benefits that depend on reforms. This supports the general conclusion of previous analyses that Brady operations should not be undertaken without a reform program in place. Bank lending support would in such case fail to benefit the country while shifting a part of the debt burden from private to official creditors.  - 19 - 43. The Agreement adds to the evidence supporting the rationale of the Brady plan. Given economic reforms, a highly indebted country and its creditor banks as a group have a common interest in DDSR as it leads to efficiency gains that can be shared. Concerted action, however, is necessary to overcome problems of coordination between the banks and of credible commitment to reforms by the sovereign debtor country. The Brady approach achieves this by providing official support, conditioned on economic policies, for comprehensive and menu-based DDSR agreements between the debtor country and its creditor banks as a group. As confirmed by the Agreement, this leads to substantial debt reduction in net present value terms combined with significant savings to the country relative to market-based debt reduction. 44. The voluntary nature of Brady operations, as has been observed before, makes it likely that the banks will end up with an improved financial position. The analysis of the Agreement provides supporting evidence for creditor banks' gains. In this context, the Argentine experience suggests that accumulated interest arrears tend to weaken the debtor country's negotiating position. Argentina had to undertake significant reforms before it could access the Brady plan as post-DDSR contractual payments would otherwise not be sustainable. With reforms already made and a momentum for further reforms developed, the Government was eager to reach a settlement, allowing the banks to hold out for larger gains. 45. The banks' selection between the two options of the Agreement revealed a strong preference for par bonds. Since official support for DDSR depends on a balance between debt reduction and debt service reduction, pressure was put on the banks to accept a higher share of discount bonds. Despite the limitation of choice, however, full voluntary participation was achieved, which suggests that the par bond option was overpriced. The mispricing of options, which was observed also in the Venezuela DDSR operation, results from the fixing of the interest rate schedule of the par bond option at the date of the Agreement in Principle. At that date, preferences of the banks are not known; particularly, interest rate expectations may change before selections are made several months later. The Agreement thus lacked an efficient instrument for reconciling selections with the objective of a balance between debt reduction and debt service reduction. Such instrument could be a program of interest rate swaps. 46. The Agreement differs from other Brady deals in offering creditor banks only exit options, i.e., no option for maintaining their loan exposure by providing new money. The case for including a new money option rests on the premise that some banks have very high exit costs and would therefore only be willing to exit when the value of at least one exit instrument is very high. Buying out the high exit cost banks using an exit menu option would dramatically increase the cost of debt reduction; instead, allowing them to remain on the sideline would undermine the whole menu approach. The Agreement worked on the assumption that the relevant source of high exit costs is the regulatory requirement of some countries that banks selling part of their loans to a given country mark to market their country portfolio. The offer of bonds never to be traded on any stock exchange circumvented this requirement. The outcome of the Agreement suggests that the assumption was correct, i.e., a new money option is not required to secure both full participation and reduced costs of debt reduction.  - 20 - PART II. PROGRAM REVIEW FROM BORROWER'S PERSPECTIVE Part II has not yet been received from the Borrower.  - 21 - PART m: BASIC DATA A. Debt and Debt Service Reduction Loan (Loan 3355-AR) ORIGINAL DISBURSED CANCELLED REPAID OUTSTANDING Loan 3555-AR US$450M US$450M 0 0 US$450M ORIGINAL LOAN DATES ACTUAL Briefing Note 06/15/92 Completion of Negotiations 11/13/92 Board Approval 01105/93 Loan Agreement Signing 02/02/93 Release (US$450 million) 03/15/93 Effectiveness 03/26/93 Loan Closing 07/31/93 CUMULATIVE LOAN DISBURSEMENT (US Millions) FY93 (a) Planned 450 (b) Actual 450 STAFF INPUTS FY9 FY9 FY94 TOTAL To Negotiation 9.0 15.9 0.0 24.9 Negotiations 0.0 13.9 0.0 13.9 Supervision and Administration 0.0 5.5 0.6 6.1 Total 9.0 35.3 0.6 44.9 FOLLOW-ON SAL OPERATIONS Second Enterprise Reform Adjustment Loan (3556-AR): US$300 million. Financial Sector Adjustment Loan (3558-AR): US$400 million.  - 22 - A-t ARGEN4TINA - A Ant. (T- p-oo -6- -h-c-God) 1991 I992 1993 W1994 995 1996 I99W7 199 199 mm GDP 5-.h 1.9 t.7 6.0 4.0 4.0 3.1 4.0 4_3 4.5 47 Glti .1pkpiu pwz. -pt. 10. 9.5 3.2 1IJ l.t L4 1.9 2-3 2-3 2L3 GDP (1.u bil-) 19.7 221-. 2S7.6 276.5 2912 3161 340.5 30 359.9 42.65 2S.- I GDP' T.a.1l Lnv_^u 14.6 16.7 It.0 19.1 19.4 197 19.9 20.1 2tU 20.6 PrwL 12.4 14.M 15.9 16J 172 17.4 17.5 17.6 17.7 17. PubU- (-o-d.d) 2.3 1.9 2.2 22 2.2 2.3 2.4 2.5 2.6 2.J S.tul S 13.2 1l.0 14.1 15.9 16.1 16.6 16.9 17-3 17.7 1L2 1323 II. 13.1 14.1 14.3 14 5 14.7 14.9 15.1 15.4 PNblU (4<- oldct.d) (0.1) 13 1.7 1I L9 2.0 2.2 2.4 2.6 2. Foip. S"vr 1.4 3.7 33 3.2 3.3 3.2 3.0 2.5 2.6 2.4 ICtOR (l6gd) 1.6 1.9 3.3 5.3 5.6 6.0 5. 5.4 5.2 5.1 Fcd-I PubUc S5_ ( lt b..i. cs dcnt GDP). T.c.1 Cul t R_ IL2.5 LL5 12.2 1.4 119 1. L3 10 L3 12 12. T~l CtD Eqt.& _= 12.45LL ILS LLO LL-5 12 5 Ma LL6 L LL6 InXl_t EX-YI;IYI 2-5 L.5 LO LO Li Li L3 L3 L2 Li 5=iml Scsrity Svw 0.1 0.5 1 L1 0.7 0J 0J 0.A 0.6 LO FE N--im-d S.-k 0.2 0.2 0L3 0.0 (C-O) (0-/l ) (0