WPS3780 An Analysis of the 2002 Uruguayan Banking Crisis Luis de la Plaza Sophie Sirtaine World Bank Policy Research Working Paper 3780, December 2005 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. The authors are inmensely grateful to Aquiles Almansi, Mariluz Cortes, Augusto de la Torre, Juan Gaviria, Daniel Oks, John Pollner and Axel van Trotsenburg, for the very helpful comments given to previous versions of this paper. ABSTRACT This paper reviews the series of events that led to the 2002 Uruguayan banking crisis, assesses the current status of the Uruguayan banking sector, and analyzes the policy responses undertaken by the Uruguayan authorities to counteract the crisis. The main conclusion from this analysis is that although the immediate trigger for the crisis was caused by contagion resulting from Argentina's financial crisis, the spread and magnification of the crisis that engulfed the Uruguayan economy was amplified by certain weaknesses of the Uruguayan economy in general, and the domestic banking sector in particular. The authors also believe that the policy responses adopted by the Uruguayan authorities were mostly adequate, allowing Uruguay to successfully counteract simultaneous banking and public debt crises. Most importantly, the Uruguayan authorities were able to overcome a severe crisis while preserving the necessary trust in banking contracts, achieving a high level of social stability and political cohesion, and maintaining a fluid dialogue with multilateral financial institutions and all affected parties. The cooperative and consensual approach taken by the authorities created the necessary conditions to overcome some of the important obstacles to the recovery of the domestic banking sector. The paper is organized in three sections: (i) Section 1 follows the chronology of events before, during and after the crisis, and reviews the policy responses undertaken by the Uruguayan authorities ­ assisted by multilateral institutions­ to counteract and stem the crisis; (ii) Section 2 assesses the current status (as of June 2004) of the Uruguayan banking sector, its immediate prospects and the remaining areas of weakness; and (iii) Section 3 attempts to draw useful policy lessons derived from the actions undertaken by the Uruguayan government. 2 Table of Contents SECTION I: THE URUGUAYAN BANKING CRISIS ............................................................ 4 Situation prior to the crisis.................................................................................................. 4 Description of the Crisis...................................................................................................... 8 First Wave of Policy Responses: Crisis Management .................................................... 12 Second Wave of Policy Responses: Recovery from the Crisis....................................... 15 Restructuring of the Banking System .............................................................................. 15 Strengthening the Regulatory and Supervisory Frameworks .......................................... 17 Debt Rescheduling........................................................................................................... 18 SECTION II: THE URUGUAYAN BANKING SECTOR: CURRENT STATUS............... 20 Current Situation............................................................................................................... 20 Challenges........................................................................................................................... 23 SECTION III: POLICY CONCLUSIONS AND LESSONS LEARNED .............................. 25 Levels of and Banking Exposure to Government Debt.................................................. 25 Dollarization and Contagion ............................................................................................. 25 Government's Debt Restructuring................................................................................... 26 Institutional Stability......................................................................................................... 27 Relations with International Financial Institutions........................................................ 27 Exchange Rate Dynamics.................................................................................................. 27 Costs of the Crisis............................................................................................................... 28 Specific Banking Measures ............................................................................................... 29 Preservation of Banking Contracts and the Payment System.......................................... 29 Internalization of currency mismatch. ............................................................................. 30 Early Bank Intervention................................................................................................... 30 Targeted Policies Toward the Private Banks................................................................... 31 Conclusions......................................................................................................................... 31 Bibliography....................................................................................................................... 32 3 SECTION I: THE URUGUAYAN BANKING CRISIS Situation prior to the crisis At the end of 2001, the Uruguayan banking sector was considered by most to be reasonably healthy. Although a protracted economic recession that started in 1999 had weakened the profitability of Uruguayan banks ­particularly the public banks, the system was perceived to be, with some exceptions, properly capitalized, to have adequate liquidity and ­contrary to many other Latin American countries­ to lack a large exposure to the public sector. The system was highly segmented into two large public banks (Banco de la República Oriental del Uruguay, or BROU and Banco Hipotecario del Uruguay, or BHU, together accounting for 40 percent of the system's assets) and a group of approximately 30 private ­mostly foreign­ banks, that also included some local investment banks and savings & loans cooperatives. Among these, Banco Galicia Uruguay (BGU) and Banco Comercial (BC) clearly dominated the segment, with combined assets representing approximately 20 percent of the overall banking system. Uruguay: Selected Banking Indicators Public Private As of December 31, 2001 (%) Total Banks Banks Asset Quality NPLs/Total Loans 17.9 39.1 5.6 Provisions/NPLs 49.7 39.2 91.7 Capital Adequacy Assets/Capital 16.7 12.2 22.3 Capital/Risk Adjusted Assets 11.8 17.5 7.6* Profitability R.O.A. (after tax) -2.3 -4.5 -0.9 R.O.E. (after tax) -28.1 35.4 -16.2 Liquidity Loans/Deposits 93.2 89.5 96.4 Liquid Assets/Deposits 15.9 20.9 13.6 Memorandum Total Assets (share) 100.0 40.9 59.1 FX Deposits/Total Deposits 89.1 84.2 92.1 FX Loans/Total Loans 80.6 56.0 93.0 * This number includes ratios for two intervened bankrupt banks, Banco de Crédito and Caja Obrera. Source: Banco Central del Uruguay and IMF's estimates Notwithstanding the overall picture of relative soundness, the system was, however, intrinsically vulnerable to external shocks due to its deposit structure, which was highly dollarized and with a large presence of non-resident depositors. In fact, as of December 2001, total deposits in the system amounted to US$15.4 billion (representing 83 percent of Uruguay's 2001 GDP), of which 90 percent were foreign currency deposits, with 47 percent of these deposits being held by non-residents. Despite this skewed deposit structure, however, the Uruguayan Central Bank (Banco Central del Uruguay, or BCU) lacked specific prudential regulations regarding foreign currency denominated deposits 4 held by non-residents, there were no specific liquidity requirements or limitations on such deposits, and supervision of the state-owned banks was generally weak and untimely1. Total Deposits by Currency US Dollar Deposits by Nationality of Depositor Local Non- Currency Residents 10% Foreign 47% Residents Currency 53% 90% Source: BCU and IMF, as of December 31st, 2001 Dollarization was not limited to the banks' liabilities, but extended also to their asset side. As of the end of 2001, total loan book for the system amounted to US$11.5 billion, out of which US$8.6 billion (or 75 percent) was denominated in foreign currency, mainly in US dollars. Paradoxically, US$6.1 billion (or 71 percent) of these loans had been extended to residents, even though the vast majority of them did not earn revenues in any foreign currency. Within the system, the financial condition of the two largest public banks (BROU and BHU) was very weak. These two banks had questionable lending practices and weak corporate governance and, as a consequence, their level of non-performing loans was greater than that of the rest of the system, their costs higher and their profits lower. In fact, as of December 2001, the ratio of non-performing loans to total loans for the public banks was 39.1 percent, versus 5.6 percent for the private banks, while after tax return on equity was minus 4.5 percent for the public banks, versus minus 0.9 percent for the private banks. Particularly vulnerable was the case of the BHU, which as of the end of 2001 represented approximately 10 percent of total deposits within the system. This institution, the country's almost exclusive provider of mortgage lending, was especially susceptible to external shocks, given the substantial currency and maturity mismatch in its balance sheet. In fact, the majority of the BHU's deposits (77 percent) was denominated in US dollars and had relatively short maturities, while the greater part of its loans (94 percent) were peso- denominated, long-maturity loans extended to peso-earners. 1"Third Review under the Stand-By Arrangement and Request for Modification and Waiver of Applicability of Performance Criteria", Supplementary Information, IMF, June 2003, page 2. 5 On the fiscal front, the protracted economic recession that started in 1999 had led the government to accumulate increasingly large deficits (from 38 percent of GDP in 1998 to 58 percent by 2001) that were financed through correspondingly sizeable levels of public debt, the vast majority of which was financed via external issuance of foreign currency denominated debt. Thus, just before the crisis erupted, total public sector debt amounted to US$10.7 billion, out of which 83 percent was denominated in foreign currencies. Cyclicality of Uruguay's Government Debt Total Government Debt as % of GDP (Left) GDP Growth (annual %) (Right) 140% Crisis 10% Starts 120% 5% 100% 80% 0% 60% -5% 40% -10% 20% 0% -15% 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 Source: Global Development Statistics This high level of government debt as a percentage of GDP represented the highest recorded level since the mid 1980's ­when Uruguay also experienced severe GDP contractions­ and, unfortunately, preceded the eruption of the Argentinean crisis2. This reality, combined with the fact that most of the government debt was denominated in foreign currencies, made Uruguay highly vulnerable to an external shock of the nature and magnitude of the Argentine crisis. The external debt situation would, as we will see, only worsen with the crisis, with total external debt reaching US$11.3 billion by December 2002 (or 93 percent of that year's GDP) and US$12.1 billion by December 2003 (or 108% percent of that year's GDP). 2Uruguay's level of public debt as a percentage of GDP in 2001 (i.e. 58 percent) compares unfavorably versus, for instance, 51 percent for Argentina that same year and 44 percent for Brazil in 1999 6 Partially responsible for the weak performance of the Uruguayan economy since 1999 was the one-two punch inflicted by the successive devaluations of the Brazilian Real and the Argentinean Peso in 1999 and 2001 respectively. In fact, the loss of competitiveness caused by the considerable appreciation that the Uruguayan Peso's real exchange rate experienced as a consequence of Brazil's crisis in 1999 was only to be made worse by Argentina's devaluation in late 2001. Thus, by January 2002, and as can be seen in the chart below, the Uruguayan Peso was seriously out of equilibrium not only against Uruguay's two major trading partners, but also vis-à-vis the rest of the world. Uruguay: Real Exchange Rate (1995=100) vs. Argentina vs. Brazil vs. Rest of the World 120 110 100 90 80 70 60 50 40 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Source: BCU In summary, at the end of 2001, the Uruguayan economy was characterized by the weakness of the public banks, its high level of foreign currency indebtedness ­both private and public­ and the sluggishness of the economy derived from an appreciated real exchange rate against major trading partners. Under these circumstances, the Uruguayan banking sector was highly vulnerable to an external shock and potential currency devaluation. 7 Description of the Crisis The crisis erupted in December 2001 when the Argentine government imposed capital controls and deposit freezes to Argentina's nationals (popularly known as the "corralito"). At that time, the two largest private banks in Uruguay (Banco Galicia Uruguay, or BGU; and Banco Comercial, or BC) began facing liquidity problems as a result of their high level of exposure to Argentina. In fact, these two banks ­which combined represented approximately 20 percent of total deposits within the system­ were both owned by Argentinean financial groups and were, therefore, highly susceptible to Argentina's economic turmoil. The case of BGU is particularly worth noting since its demise has been considered by many as the event that truly triggered the wider crisis. BGU was a subsidiary3 of Banco de Galicia ­Argentina's largest private bank­ and was, as of December 2001, the second largest Uruguayan bank in terms of assets. BGU's vulnerability stemmed from its business model, which was concentrated almost exclusively in taking deposits from Argentine nationals and corporations and lending those same funds also to Argentine nationals and corporations. Therefore, when the Argentine government imposed capital controls and deposit freezes, BGU was severely affected as access to its assets in Argentina was drastically curtailed and as cash-strapped Argentineans begun withdrawing their deposits from Uruguay. As a consequence, during the month of January 2002 alone, BGU lost an estimated 15 percent of total deposits, completely exhausting its available liquidity. Faced with this situation, the Uruguayan central bank had no choice but, on February 13, 2002, to temporarily halt BGU's operations and intervene the bank and, subsequently, to permanently suspend it. Similarly, BC, which as of December 2001 was the largest private bank in Uruguay, was also highly exposed to the Argentine economy due to its large holdings of Argentine government debt as well as its extensive lending to the Argentina conglomerate Grupo Banco General de Negocios, of which BC was part. Thus, as in the case of BGU, BC was confronted in early 2002 with severe liquidity constraints, with the situation being exacerbated by accusations of management fraud. After several attempts to re-capitalize the bank, the BC was also intervened by the BCU and subsequently restructured (please see below). Over the following weeks, and as the economic crisis deepened in Argentina, deposit withdrawals gradually continued and, by March 2002, 12 percent of total bank deposits ­mostly from non-residents­ had left the country. Although the authorities were expedient in providing liquidity support to the affected banks and a financial support program from the IMF was announced, negative public perception of the unfolding crisis continued exerting pressure on the Uruguayan currency, which the government attempted to counteract by introducing greater exchange rate flexibility and widening the existing crawling exchange rate band from 6 to 12 percent. 3Although the BGU is often referred to as a branch of Banco de Galicia, in reality it had obtained a full banking license and should, therefore, had been considered as an Uruguayan banking institution (Authors' Note) 8 Chronology of the Uruguayan Banking Crisis 3-Dec-01 Argentina establishes the "Corralito" 12-Dec-01 IMF suspends loan disbursements to Argentina 23-Dec-01 Argentina defaults 15-Jan-01 Uruguay widens its Crawling Exchange Rate Band 3-Feb-02 Argentina establishes the "Corralón" 13-Feb-02 Banco Central del Uruguay suspends operations of Banco Galicia-Uruguay 15-Feb-02 Uruguay losses investment grade rating 28-Mar-02 First attempts to re-capitalize Banco Comercial 31-Mar-02 Cumulative 12.2% of total deposits have been withdrawn since January 1st, 2002 21-Jun-02 Banco Montevideo-Caja Obrera is intervened 30-Jul-02 Bank holiday is declared 31-Jul-02 Cumulative 37.6% of total deposits have been withdrawn since January 1st, 2002 31-Jul-02 Banco Central del Uruguay has lost 79% of international reserves since January 1st, 2002 5-Aug-02 Bank holiday is lifted after a substantial bailout package provided by multilaterals As the financial crisis in Argentina worsened and deposit freezes were tightened (the "corralón"), a second wave of deposit withdrawals ensued in April 2002, following Uruguay's downgrade from investment grade status and amidst mounting concerns that Uruguay would be forced to follow in Argentina's footsteps and impose itself deposit freezes. Thus, by May 2002, an additional 18 percent of deposits had been withdrawn from the system. Most importantly, by that date, deposit withdrawals were no longer confined to non-resident deposits and a few specific institutions, but had spread to include also resident deposits withdrawing their funds from the public banks4. Evolution of Deposits (US$ millions) Dec-01 Mar-02 Jun-02 Sep-02 Dec-02 Total Deposits 15,403 13,475 10,744 8,007 8,374 Foreign Currency 13,970 12,261 9,824 7,458 7,747 Residents 7,413 6,953 5,958 5,250 5,431 Non-residents 6,557 5,308 3,866 2,208 2,315 Local Currency 1,433 1,214 920 549 627 Source: Banco Central del Uruguay As deposit withdrawals accelerated during May and June 2002, the authorities continued to actively provide widespread liquidity support to both private and public banks. Despite this continued support, on June 21, 2002, the third largest private bank, the Banco de Montevideo-Caja Obrera, run into severe liquidity shortages and had to be intervened by the authorities, which took over its operations and replaced its management. 4The Uruguayan banking system prior to the crisis did not distinguish between resident and non-resident deposits. Although special entities (Cajas Bancarias) had been created specifically to service the non-resident segment of the banking system, the initiative had failed to catch on due to adverse fiscal consequences vis-à- vis local banking institutions. Therefore, resident and non-resident deposits were mixed up within the system and, in some cases (e.g. Banco Galicia Uruguay) clearly dominated the deposit base of some institutions. 9 Peso/US$ Exchange Rate and Evolution of US$ Reserves (January to December 2002) Official International Reserves (Left Axis) Exchange Rate (Right Axis) 3,000 30 2,500 . m 2,000 m 20 D 1,500 US Peso/USD 1,000 500 10 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Source: BCU The tilting point came in early July, when ­following a further deterioration of market sentiment that translated into widening government's bonds spreads­ the run on dollar deposits extended also to affect local currency deposits, causing, by late July, the position of the BCU to become untenable. After months of widespread deposits withdrawals and substantial liquidity support to the banking system, the level of available international reserves reached the critically low level of US$650 million (versus USD3.1 billion in December 2001, or an 80 percent decline). This low level of reserves was clearly insufficient to service the mounting external debt and to continue backing the large proportion of foreign currency denominated deposits still present within the system (i.e. US$8.7 billion as of July 2002). Confronted with this situation, the authorities opted to freely float the peso ­which immediately depreciated by 27 percent­ compelling the government to declare a five-day bank holiday on July 30, 2002. Evolution of Official International Reserves and Deposits Official International Reserves (Left Axis) "Encajes" at the BCU on USD Deposits and Local Currency Deposits (Left Axis) US$ Deposits (Right Axis) 3,500 16,000 3,000 14,000 2,500 12,000 .mm 10,000 2,000 8,000 .mm USD1,500 6,000 USD 1,000 4,000 500 2,000 - - Dec-00 Jun-01 Dec-01 Jun-02 Dec-02 Jun-03 Dec-03 Jun-04 Source: BCU 10 As the crisis was unfolding during 2002, the drastic and sustained deposit withdrawals translated into a system-wide credit crunch, as banks ­both private and public­ scrambled to find any available liquidity by suspending new loans as well as by requesting early repayment of existing loans. Thus, credit to the non-financial sector shrunk by 37 percent during 2002 alone, greatly contributing to a GDP contraction of 10.7 percent for that same year. Evolution of Deposits andCredit to the Non-Financial Sector (December 2001 to December 2002) Credit to Non-Financial Sector Total Deposits 15,000 14,000 13,000 12,000 mm. 11,000 US$ 10,000 9,000 8,000 7,000 Dec-01 Mar-02 Jun-02 Sep-02 Dec-02 Source: Banco Central del Uruguay By the end of July 2002, a cumulative and staggering 38 percent of total deposits had been withdrawn from the system. In addition, with the peso having also depreciated by 57 percent over the same period, the majority of banks had become technically insolvent. In just seven months, the Uruguayan banking sector had changed beyond recognition, with 67 percent of the system's total deposits in the hands of the government5 (either as public or as "intervened" banks), and with 51 percent of deposits held by non-residents having left the country. As far as the public banks were concerned, and despite massive liquidity support amounting to US$1,211 million6 in total, the financial condition of BROU was extremely weak, while the situation of the BHU ­given the inherent currency mismatch of its balance sheet­ had become critical. Conversely, of the private banks ­whose liquidity support had also amounted to US$1,208 million7­, Banco Galicia Uruguay had been suspended and Banco Comercial and Banco de Montevideo-Caja Obrera had been intervened and were under government control8. In addition, a large number of foreign-owned banks had had to receive liquidity and equity support from their headquarters abroad. 5As opposed to 37 percent of total deposits as of December 31, 2001. 6As of August 2002 (Source: IMF) 7Ibid. 8A third bank, Banco de Crédito, had also been intervened by the authorities prior to the crisis, with the government owning 51 percent of its capital and the remaining 49 percent in the hands of the Moon sect. At the time of the crisis, the Banco de Crédito was the country's fourth largest bank in terms of assets. 11 First Wave of Policy Responses: Crisis Management The Uruguayan banking crisis was characterized ­and, arguably, mitigated­ by the swift and targeted intervention of the authorities, which, assisted by multilateral institutions, stood ready during 2002 and 2003 to both provide liquidity to the system and promptly intervene to restructure and/or liquidate troubled institutions. As the initial phases of the crisis appeared to be instances of limited deposit runs at specific banks strictly caused by the troubles in neighboring Argentina, the actions of the Uruguayan authorities (through the Corporación Nacional para el Desarrollo, or CND9 and the CBU) were limited to providing liquidity support to the affected banks. However, as the crisis deepened through June and July 2002, and liquidity needs became increasingly large and systemic, the authorities had to discriminate in between, core banks with a critical participation in the payment system and a nation-wide branch network and client base, and, non-core banks, most of which were the Uruguayan branches of foreign banks. Accordingly, the CBU's liquidity support was to benefit exclusively the core banks (i.e. the BROU, the BHU, the three intervened banks10 and some small domestic cooperatives, all of which accounted for 55 percent of the system's deposits), while non-core banks were to rely on their own liquidity support. However, and most importantly, the discriminating nature of the authorities' liquidity support was predicated upon the tacit understanding that no additional restrictions were to be imposed on the foreign banks, which were able throughout the unfolding of the crisis to act without any specific limitation. As the crisis grew and intensified, a US$2.5 billion facility (the Fondo para la Fortificación del Sistema Bancario, or FFSB) was created in June 2002 to supplement the fast decreasing lender-of-last-resort facilities of the central bank, and to provide additional liquidity and equity support to struggling core banks. The resources of the FFSB came from the first augmentation of the Stand-By Agreement with the IMF, with additional resources to be provided by other multilateral institutions and the government in the form of US dollar-denominated bonds. Despite the FFSB's comparatively large resources, however, deposit withdrawals continued unabated in June and July 2002, which combined with the authorities' support for the then prevailing exchange rate, critically drained the available reserves at the central bank and exhausted the government's ability to continue to provide further liquidity to the system11. Ultimately, the FFSB proved insufficient and, by July 30, 2002, its activities were suspended and the bank holiday was declared. 9The CND is a government-owned investment company created in 1985 (Ley N° 15,785) to foster economic growth through selective investment in a variety Uruguayan economic sectors. 10The three intervened banks were Banco Comercial, Banco de Montevideo-Caja Obrera and Banco de Crédito. 11International reserves at the BCU fell from US$3,300 mm. at the end of 2001 to approximately US$650 mm. by the end of July 2002. 12 Following the lifting of the bank holiday on August 5, 2002, the Uruguayan authorities created a new legislative framework (Ley 17.523, known as the Ley de Fortalecimiento del Sistema Financiero) that included a series of resolute measures in order to finally contain the crisis: · First, the law created the Fondo de Estabilización del Sistema Bancario ("FESB"), a US$1.4 billion stabilization fund financed by multilateral institutions12. The FESB's resources ­which were in addition to the resources already used by the suspended FFSB­ were sufficient to fully back the entire book of US dollar sight and savings deposits at public and intervened banks. · Second, the US dollar time deposits of the public banks (i.e. BROU and BHU) were re- programmed and their maturities stretched over a three-year period13. · Third, no restrictions were imposed on foreign banks' operations, as long they were to rely on their own resources to provide liquidity support. · Fourth, the BROU absorbed all foreign currency and timed deposits of the BHU14, which, although it remained in operation, was no longer allowed to receive deposits. · Fifth, the operations of the three intervened (i.e. Banco Comercial, Banco de Montevideo-Caja Obrera and Banco de Crédito) were permanently suspended and actions were initiated towards their eventual restructuring and/or liquidation. Evolution of Total US$ Deposits (January to December 2002) 15,000 14,000 13,000 12,000 mm. 11,000 10,000 US$ 9,000 8,000 7,000 6,000 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Source: BCU The creation of the FESB and the measures above mentioned were finally able to stop the crisis. Although deposit withdrawals picked up substantially after the lifting of the bank holiday, they gradually diminished and, by October 2002, deposits started to return to the system and ­as far as resident deposits are concerned­ have today reached their July 2002 levels, although are still, as of June 2004, only 78 percent of their December 2001 level. 12Funding of the US$1.4 billion facility was split as follows: IMF US$788 mm., IADB US$385 mm., and World Bank US$200 mm. 13Total deposits reprogrammed amounted to US$2.2 bn. 14Transfer was executed by the BHU issuing US$776 million bond ­which was guaranteed by the government­ and placed with the BROU in return for absorbing such deposits. 13 In summary, by the end of 2002, the Uruguayan banking system had lost 46 percent of total deposits and the level of non-resident deposits had decreased by 65 percent. As a result of this massive deposit run, one bank had to be closed and three additional banks had to be intervened and restructured by the government, which, as of the end of 2002, controlled approximately 70 percent of the total deposits in the system. In total, liquidity support provided by the government during 2002 amounted to US$2.4 billion, or approximately 20 percent of that year's GDP, while the ensuing GDP contraction amounted to approximately 11 percent. The government's actions in providing liquidity support to the ailing banking system ­and the need to finance this support­ has translated not only into a significant increase in the absolute level of public sector debt, but has also caused a significant shift in the composition of such debt, with official creditors becoming much more important than prior to the crisis. Thus, total public sector debt has grown 26 percent since the beginning of the crisis, with official creditors representing 45 percent of such debt, versus only 25 percent in December 2001. Evolution and Composition of Uruguay's Public Sector Debt Official Creditors Private Creditors 8,000 7,000 6,000 mm. 5,000 US$ 4,000 3,000 2,000 Dec-01 Mar-02 Jun-02 Sep-02 Dec-02 Mar-03 Jun-03 Sep-03 Dec-03 Mar-04 14 Second Wave of Policy Responses: Recovery from the Crisis Once deposit withdrawals finally receded and the banking system resumed a more normal operating environment, the authorities undertook during the second half of 2002 and 2003 a series of additional legal and regulatory initiatives aimed at, first, restructuring the still ailing banking system, second, strengthening the financial sector's regulatory and supervisory framework, and, third, resolving the impending crisis of the government finances and the increasingly large foreign debt. Restructuring of the Banking System On the issue of bank restructuring it is important to distinguish between the situation of the public banks (BROU and BHU) and that of the private intervened banks (Banco Comercial, Banco Montevideo-Caja Obrera and Banco de Crédito). Public Banks As a consequence of the crisis, the BROU lost approximately 66 percent of total deposits15 during 2002, very seriously affecting its liquidity and highlighting the weaknesses of its balance sheet, not to mention its questionable lending practices and corporate governance. Consequently, the BROU ­assisted by the IMF­ has undertaken a drastic restructuring program along the following lines: · A government-guaranteed Asset Management Company was created in late 2003 to absorb all category 4 and 5 loans16 and attempt to recover these loans over a five-year period17. The BROU has committed to complete the transfer of these loans by December 2004. · The government has covered the arrears on the BHU bond18 and the BHU has committed to keep the bond current in the future. · BROU's commercial focus will be re-directed towards peso-based products. · BROU's cost structure will be rationalized with the objective of reducing operating costs by 15 percent over the next two years. · BROU's credit risk management and loan monitoring will be improved to reduce non- performing loans in the future. 15This figure does not include the re-programmed deposits from BROU and BHU, which represented in excess of 62 percent BROU's deposit base at the end of 2002 16Non-performing loans in Uruguay are defined as category 4 and 5 loans within the loan book. These loans represent the two riskiest categories of loans (out of a total of five categories) and with the least likelihood of repayment. 17The trust, which will be controlled and managed by the BROU, will issue notes to purchase the non- performing loans from the BROU at their net book value. 18This was the US$776 million government-guaranteed bond that the BHU issued to the BROU when this institution absorbed the BHU's dollar-denominated time deposits in July 2002. The BHU has since been unable of fully service this bond. 15 On the other hand, the BHU was by far the Uruguayan financial institution hardest hit by the 2002 crisis as a consequence of the inherent vulnerability of its balance sheet. In fact, at the end of 2001, most of BHU's deposits (77 percent) were denominated in dollars, while the vast majority of its loans (94 percent) were peso-denominated, indexed-to-wages loans extended to peso-earners. Naturally, when the crisis erupted and the peso was sharply devalued, the resulting deterioration of the BHU's balance sheet made it to become highly insolvent. Despite the severity of its crisis, the BHU ­with the assistance of the World Bank­ is the institution whose restructuring plan is more defined and advanced. Today, the BHU has become19 a non-banking housing institution that is no longer allowed to take on new deposits (with the exception of Ahorro Previo or housing savings plans) and can only issue mortgages up to a total amount of US$50 million. In addition, the BHU is in the midst of implementing a radical streamlining plan aimed at strengthening its balance sheet, reducing costs and improving its management and information technologies. Private Intervened Banks In January 2003, the authorities placed Banco Comercial and Banco Montevideo- Caja Obrera under liquidation and created a new bank, the Nuevo Banco Comercial, or NBC. Thus, in late February 2003, the NBC issued CDs to finance the purchase of the assets of the two liquidated banks20, with the proceeds from that sale used in turn to partially compensate the original depositors of the liquidated banks. The NBC opened for business in late March 2003 and was, as of the end of 2003, the third largest bank in the country, representing 9.5 percent of total deposits within the system. On the other hand, the Banco de Crédito was finally placed under liquidation in February 2003, after several failed attempts to re-capitalize it. Other Private Banks This segment of the Uruguayan banking, composed of a small network of mostly foreign-owned banks, has proven to be the most resilient of all. As mentioned before, these banks were purposely left out of the liquidity support provided by the authorities and had to rely on their own resources to weather the crisis. Despite this fact, the private banks have been the great beneficiaries of the "flight-to-quality" during the 2002 crisis and, to date, no major consolidation or downsizing has taken place among these banks. 19The restructuring of the BHU is delineated by the Ley 17.596 de Fortalecimiento del Banco Hipotecario del Uruguay, approved in December, 2002 20A "Put Option" existed over category 4 & 5 loans, by which the NBC could return such assets back to the government until December 2003. 16 Strengthening the Regulatory and Supervisory Frameworks Lax regulatory and supervisory frameworks greatly contributed to creating some of the conditions which made the Uruguayan banking sector highly vulnerable to external shocks. To eliminate this weakness, the Uruguayan authorities have made a substantial effort in overhauling the existing regulatory framework and strengthening the supervisory institutions with the double purpose of, first, re-establishing credibility in the system as a whole and, second, creating the institutional and supervisory architecture to avoid and/or mitigate future financial crises. With these objectives in mind, Congress passed in December 2002 a new banking law (Ley 17.613, Ley de Bancos), which clearly delineated the basic operating procedures for bank restructuring and/or liquidation as well as re-defined and expanded the role of the BCU in strengthening the regulatory and supervisory frameworks. As far as bank restructuring is concerned, the new law strengthened the role of the BCU in the liquidation of failed banks and established the creation of trust funds (Fondos de Recuperación de Patrimonio Bancario) for asset recovery. It is precisely under the umbrella of this new piece of legislation that the Nuevo Banco Comercial was created and the Banco de Crédito was ultimately liquidated. On the issue of the regulatory framework, the new banking law expanded the supervisory scope of the BCU ­specifically of the unit within the BCU in charge of bank supervision, the Superintendencia de Instituciones de Intermediación Financiera­ and took the first steps towards the strengthening of the much deficient prudential regulatory framework with the aim of (i) promoting the overall quality of the banks' balance sheets, (ii) fostering the early detection of troubled institutions to avoid system-wide crisis and (iii) insulating the system itself from potential external shocks. Specifically, the new banking supervisory framework is centered on the following parameters: · Risk Limitation: Limits have been introduced for certain types of borrowers and risks (e.g. country limits, government limits, risks to associated entities, etc.). In addition, consolidated supervision has been introduced to account for international exposures. · Liquidity Requirements: Higher reserve requirements have been introduced for non- resident deposits as well as to cover maturity mismatches in the event of internal and/or external shocks. · Internalization of Foreign Exchange Risk: Regulations have been introduced to encourage banks to subjectively account and adequately provision for foreign exchange risk ­even if such risk is indirect­ in an effort to incorporate currency risk in the lending decision-making and monitoring processes. · Transparency: There is increased public dissemination of relevant financial information, including system-wide ­through the BCU­ sharing of borrower's credit history among all financial institutions. · Expansion and upgrade of the supervisory function of the BCU: Both in terms of staffing as well as of its technical proficiency. 17 Finally, the new banking law of December 2002 also established a Deposit Guarantee Program (Fondo de Garantía de Depósito Bancario), although final implementation of such plan is still pending. Debt Rescheduling Simultaneously with the above-mentioned policy responses, the Uruguayan government successfully completed in May 2003 a US$5.4 billion re-scheduling of foreign-currency denominated debt. This operation ­which restructured nearly half of Uruguay's long-term sovereign debt­ provided the authorities with the necessary room to maneuver to continue the task of reforming and strengthening the banking sector. In fact, after the sharp devaluation of the peso experienced in mid 2002 and the sharp GDP contraction that same year, the ratio of total debt to GDP had grown by early 2003 to an overwhelming 100 percent, hinting towards the possibility that Uruguay may have to follow in Argentina's footsteps and declare a sovereign default. Instead, by successfully restructuring the debt, the authorities eliminated any residual financial need until 2005 and, most importantly, enhanced the government's ability to continue servicing its debt obligations. To improve confidence on the government's broad restructuring efforts, the authorities designed the debt exchange plan to include the following innovative features that ensured the success of the exchange: · Consultation with Investors: A formal process of consultation with affected investors was established two months before the debt exchange took place, allowing for a negotiated solution and ultimately arriving at a debt exchange proposal that was agreeable to all the parties involved. · Maturity versus Liquidity Option: Investors were offered the choice of exchanging their old bonds for either a longer maturity bond or a more liquid, benchmark bond. This feature was designed to appeal to both buy-and-hold retail investors as well as to large institutional investors and index-tracking funds. · Collective Action Clauses (CACs): The new bonds, issued under New York law, allowed for aggregation clauses21 by which a potential future restructuring of these bonds will be facilitated in the future, should that become necessary. This was the first time that a sovereign issuer included such clauses, which are increasingly becoming standard for emerging market issuers. · Regulatory Incentives: A series of regulatory incentives were introduced to encourage domestic and foreign22 participation in the exchange. On the domestic front, the authorities announced preferential reserve requirements for the new bonds as well as the implicit subordination of the old bonds in case of a shortage in debt servicing funds in the future. These measures greatly contributed to ensuring the participation of domestic financial institutions, particularly domestic pension funds and insurance companies. 21An aggregation clause permits the terms of the bonds to be changed with the consent of investors representing 75 percent of outstanding principal. 22The Chilean regulatory authorities facilitated the participation of Chilean pension funds 18 As a consequence of these features, participation in the debt exchange was unusually high. In total, US$5 billion worth of principal amount was rendered for exchange, representing approximately 93 percent of eligible bonds23. Uruguay: Evolution of Sovereign Bond Spreads (Uruguayan Bond Index) 2,500 2,000 stn 1,500 Pois Basi1,000 500 0 May-01 Nov-01 May-02 Nov-02 May-03 Nov-03 May -04 Source: República AFAP The success of the debt exchange was validated by the return of deposits to the system, with total level of deposits in the system growing by 3% during the month of May 2003 alone, reversing a declining trend that had started in late 2002. In addition, the debt exchange caused a sharp decline in perceived country risk, with country risk premia declining from 2,337 basis points in April 2003 to 735 basis points by June 2003. This trend has continued until today and, as of the end of June 2004, Uruguayan sovereign bond spreads stood at 697 basis points. 23Domestic participation was particularly high, with 100 percent participation by domestic financial institutions and 98 percent by domestic retail investors. 19 SECTION II: THE URUGUAYAN BANKING SECTOR: CURRENT STATUS Current Situation As a consequence of the 2002 crisis, the Uruguayan banking sector is today highly segmented into large, public banks with national-wide branch networks, and a small number of mostly foreign-owned banks. As of June 2004, total assets of the Uruguayan banking system amounted to US$11.7 billion. The Uruguayan Banking Sector (US$ millions) Assets % Liabilities % Equity % BROU 4,941 42% 4,651 42% 290 40% BHU 1,202 10% 1,231 11% -29 -4% NBC 1,057 9% 917 8% 140 19% Total Public Banks 7,200 61% 6,799 62% 401 55% Private Banks 4,568 39% 4,246 38% 322 45% TOTAL 11,768 100% 11,045 100% 723 100% Source: BCU, as of 30 of June, 2004, except for BHU as of May 31st, 2004 The public banks represented, as of the end of June 2004, approximately 61 percent of total assets within the Uruguayan banking system24, with the BROU remaining as the flagship of the government's presence in the banking sector, representing 45 percent of all deposits and 50 percent of all loans within the system. The situation with bank deposits' level has continued to improve since the time of the crisis. Although total deposits within the system in June 2004 still represent only 61 percent of their level in December 2001, both Peso deposits and resident dollar-denominated deposits have surpassed their levels as of July 2002, when the bank holiday was imposed. As far as non-resident deposits are concerned ­a client base that may be difficult to fully recover­ they represent today 38 percent of their pre-crisis level. 24Versus approximately 36 percent as of the end of 2001. 20 Evolution of USD and Peso Deposits (2000-2004YTD) Resident USD Deposits Non-Resident USD Deposits Peso Deposits 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 As far as deposit structure is concerned, and in marked difference to the situation before the crisis, non-resident deposits represented, as of the end of June 2003, only 17 percent of total deposits within the system, instead of 43 percent as of the end of 2001, immediately before the crisis. However, foreign currency denominated deposits continues to be the norm, with 92 percent of total deposits within the system denominated in foreign currency as of the end of June 2004. Deposit Structure Resident Resident Non Resident Non Resident (US$ millions) Deposits (US$ mm) Deposits (%) Deposits (US$ mm) Deposits (%) BROU 3,860 50% 326 21% BHU 529 7% - 0% NBC 743 10% 110 7% Total Public Banks 5,132 66% 436 28% Private Banks 2,599 34% 1,111 72% TOTAL 7,732 83% 1,546 17% Source: BCU, as of 30 of June 2004, except for BHU as of May 31st, 2004 Of the public banks, the BROU (representing 42 percent of the system's total assets as of June 2004) is clearly the cornerstone of the Uruguayan banking sector. Thus, the financial condition of the BROU has had and will continue to have a significant impact over the health of the rest of the Uruguayan banking sector. As of the end of 2003, the financial condition of the BROU remained very weak, with non-performing assets to total assets of approximately 33 percent (60 percent for private sector loans). The BROU has unfrozen the first and second tranches of the deposits that were reprogrammed in July 2002, and has done so ahead of schedule and with very high rates of retention. The BROU is still in the midst of a far-reaching restructuring plan aimed at increasing efficiency, reducing costs and improving corporate governance. Although results have so far been promising, the government should endeavor to continue along this path, particularly as it relates to final resolution of Non Performing Loans. 21 The BHU (representing 10 percent of the system's assets as of June 2004) is currently also under restructuring after having become a non-banking institution. After cutting its staff by half and adopting other cost-cutting measures, the BHU continued to face liquidity problems during 2003 as has had to receive financial assistance from the government in order to service its bond with the BROU. As part of the measures to increase its liquidity and refocus its activities to become a more market-driven house financing institution, the BHU is in the process of transferring its Cartera Social (i.e. a book of loss- making housing loans granted to low income families) to the Housing Ministry and is also planning to issue mortgage-backed bonds securities in the near future. The NBC (representing 9 percent of total assets within the system as of the end of 2003) has also gone through a restructuring period that lasted most of 2003. The NBC is today probably the best-capitalized domestic bank in Uruguay (approximately 14 percent capital-to-assets ratio as of the end of 2003) and has been, so far, relatively successful in restructuring its portfolio of non-performing loans and in attracting new deposits. However, the NBC, whose branch network extends throughout the country, competes head to head with the BROU, making uncertain its very viability in the future as an independent entity. As part of a broader business plan, the NBC is currently negotiating the potential entry in its share capital of a foreign strategic investor, which may help guarantee its future independence. The private banks (representing the remaining 39 percent of total assets within the system as of the end of 2003) are, as mentioned before, mostly foreign-owned banks almost exclusively dedicated to the Montevideo area. The private banks offer today almost exclusively dollar-based products, mostly concentrated also in short maturities. The reputation and competitive position of some of the largest international banks has arguably improved as a consequence of the crisis, having honored their deposit withdrawals in full during the crisis and as they may choose to take advantage of the vacuum left by some of the liquidated domestic banks25. 25Such was recently the case of Credit Agricole, which recently acquired the cooperative ACAC. 22 Challenges Despite the significant progress achieved so far in resolving the 2002 crisis, the Uruguayan financial sector still faces considerable challenges. Although some of these challenges are structural matters (e.g. final resolution of the non-performing assets at BROU, NBC and BHU) that should be eventually resolved overtime, other more deeply rooted issues are likely to remain challenging in the medium and long terms. Today, the banking sector ­both public and private­ remains extremely liquid. Although under a different set of circumstances this may have appeared as a sign of strength, in the current Uruguayan reality it masks the fact that a crisis mentality still remains deeply embedded across the financial sector. To be sure, on the banks' liability side, the vast majority of deposits is sight deposits and could, therefore, flee the system at the slightest sign of trouble. Conversely ­and partially as a consequence of it­ the banks' lending policies are broadly restrictive, tending to concentrate on the export sector and in short maturities. In fact, as of the end of June 2004, bank's lending to the non-financial sector, as a percentage of GDP, was only 27 percent, versus 57 percent at the end 2001. In terms of flows, risk aversion and liquidity considerations are causing lending to occur predominantly within the financial sector, with 61 percent of the total loan book being among financial institutions. Banking Activity Loans to Loans to Deposits from the Deposits from the (US$ millions) Financial Sector Non Financial Sector Financial Sector Non Financial Sector BROU 2,717 1,235 372 4,185 BHU 12 288 652 529 NBC 285 317 9 853 Total Public Banks 3,014 1,840 1,033 5,567 Private Banks 1,693 1,219 292 3,711 TOTAL 4,707 3,059 1,325 9,278 Source: BCU, as of 30 of June, 2004, except for BHU as of May 31st, 2004 In total, and since December 2001, loans to the non-financial sector have decreased by almost 60 percent, despite the fact that deposit re-flow into the system has been steady since early 2003 and resident deposits have, as of June 2004, reached 78 percent of their level prior to the crisis. In summary, the banking system is simply not fulfilling its mission of transforming short term deposits into long term lending and channeling capital towards to the most productive investments within the real sector of the economy. 23 Evolution of Deposits andCredit to the Non-Financial Sector Credit to Non-Financial Sector Total Level of Deposits 16,000 14,000 12,000 mm. D 10,000 US 8,000 6,000 4,000 Dec-01 Jun-02 Dec-02 Jun-03 Dec-03 Jun-04 Partially responsible for this state of affairs is the high degree of dollarization still present in the economy, which the crisis has only managed to exacerbate. Indeed, supply and demand of financial products remains today predominantly dollar-based (e.g. 92 percent of total deposits), despite comparatively higher rates in peso deposits and broad market consensus of the likely appreciation of the peso in the short and medium terms. Dollarization has constituted a mixed blessing towards the expansion of the Uruguayan financial sector. Although it has encouraged comparatively high levels of financial intermediation and has enabled Uruguay to attract substantial amounts of deposits by foreigners ­particularly from Argentina, it has also increased exponentially the costs generated by past financial crisis. This fact still constitutes ­as it did at the end of 2001­ a significant vulnerability factor for the Uruguayan economy. It also thwarts any further deepening of the financial sector and prevents the development of non-banking capital markets. The recent introduction of deposits denominated in inflation-indexed units ("Unidades Indexadas", or UI) is undoubtedly a step in the right direction, but the initiative has still to catch on within the financial sector and the public at large. It may also be argued that the new entity created out of the liquidation of three banks, a new public bank, might have reduced competition and may be difficult to privatize (although the new institution is a clearly stronger institution and may act as an incentive for BROU to reform). 24 SECTION III: POLICY CONCLUSIONS AND LESSONS LEARNED With the invaluable benefit of hindsight, helpful lessons can be learnt by analyzing not only some of the immediate causes of the crisis, but, most importantly, the policy choices undertaken by the Uruguayan authorities and the results that such measures have achieved since they were implemented. Although some of the policies adopted were and have remained highly controversial, the results obtained so far have mostly validated the actions undertaken by the authorities, particularly when considering the very difficult operating environment under which such measures had to be implemented. Overall, and although the full impact of some of these measures is still to be achieved, the government's actions during and after the crisis seemed adequate responses to the challenge and highlight the importance of resolute and expedient measures when confronted with such a volatile predicament. Levels of and Banking Exposure to Government Debt As we have seen, Uruguay's overall levels of government debt had been on the rise in the years prior to the crisis. Although it is very difficult to quantify what level of debt constitutes too much debt26, it is nonetheless obvious that historically high levels of government debt ­particularly foreign currency denominated debt­ make for less than an ideal platform from which to attempt to counteract a growing financial crisis. If, in addition, one's economy and financial sector are highly geared to macroeconomically volatile partners, then the conclusion must surely be to err on the side of caution and attempt to always remain fiscally cautious. The case of Uruguay clearly illustrates such an assertion. On the positive side, however, and in marked contrast to some of its Latin American neighbors, the Uruguayan banks had very low levels of exposure to government debt. In fact, as of the end of 2001, the Uruguayan banking sector's total exposure to government debt amounted to only 3 percent27 of total banking assets and 18 percent of total bank's capital28. Thus, when the crisis erupted, the Uruguayan authorities were capable of partially compartmentalizing the effects of simultaneous deposits runs and deteriorating government finances, therefore being able to adopt resolute measures to counteract both unfolding crises effectively. Dollarization and Contagion Uruguay was ­and remains­ a highly dollarized economy. Although this fact has historically been considered beneficial to the development and growth of the Uruguayan 26Please see Carmen Reinhart, et al, "Debt Intolerance", National Bureau of Economic Research, August 2003. 27Source: IMF and BCU 28Conversely, and for instance, the Argentine banking system's exposure to public debt was much more substantial, amounting at the end of 2000 to 21 percent of total banking assets and 160 percent of total bank's capital (Source: Joaquin Gutierrez and Fernando Montes-Negret, "Argentina's Banking System: Restoring Financial Viability", Working Paper 2/2004, page 4, The World Bank, 2004). 25 financial sector, high levels of dollarization have also made the Uruguayan economy exceedingly exposed to sudden reversals of capital flows ("sudden stops"29). In fact, in a context of high domestic liability dollarization, floating exchange rates and economic openness, Uruguay was highly vulnerable to an external shock such as the Argentine crisis and, should, therefore, have either remained more fiscally cautious to combat such a shock, or, have adopted early on policies towards reducing domestic liability dollarization. These policies towards des-dollarization should have encouraged the development of peso-based financial services through a combination of economic incentives and regulatory initiatives, such as, for instance, preferential reserve requirements for Peso (and UI) deposits, differential risk weighting for Peso (and UI) loans, mandatory minimums for Peso (and UI) investments by domestic AFAPs, increased supply of Peso-denominated (or UI- denominated) government debt products, etc. Government's Debt Restructuring As mentioned before, Uruguay successfully re-scheduled in May 2003 a large proportion of its foreign currency denominated debt and has, consequently, gained a window of opportunity within which to complete the final restructuring of both the banking sector and the government's finances. It is obvious that much work still needs to be done in resolving the crisis and that the re-scheduling of the debt is only a transitory solution, but the conditions do exist ­even if only temporarily­ to attempt such reforms in an orderly fashion. It is also particularly important to note not only the completion of the debt exchange itself, but also the manner in which the Uruguayan authorities completed such an exchange. Indeed, Uruguay completed its debt exchange program by involving both the IMF and the affected investors in the design and implementation of the exchange. This fact is, as we have seen, credited with partial responsibility for the ultimate success of the exchange. Despite the positive outcome to the debt-rescheduling program, the sustainability of Uruguay's total level of external debt level (that increases substantially as a consequence of both the sizeable liquidity support provided at the time of the crisis and Uruguay's decision to float the currency in July 2002, please see chart on page 6) continues to pose a significant challenge in the medium and long terms. Indeed, with a total public debt level of 108 percent of GDP as of December 2003, the very measure of success of Uruguay's crisis management approach may ultimate prove to be the sustainability of its level of external debt. Several studies30 have attempted to assess this sustainability and forecast public debt to decrease to 65 percent of GDP by 2012. However, the Uruguayan public debt situation continues to be highly vulnerable to macroeconomic changes such as, for instance, weak GDP growth, rising international interest rates, or possible weakness in the Uruguayan peso. Consequently, large shocks from any of these key variables may raise doubts as to the ultimate sustainability of public debt levels. 29 Please see Guillermo Calvo, et al, "On the Empirics of Sudden Stops: The Relevance of Balance Sheet Effects", National Bureau of Economic Research, May 2004 30Please see, for example, IMF's "Fifth Review Under the Stand-By Arrangement and Request for Modification and Waiver of Nonobservance of Performance Criteria", August 2004 26 Institutional Stability Uruguay's crisis management could be considered a model for political stability and continuity, with a high degree of coordination among the three branches of government and overall consensus on the actions overtaken by the government to overcome the crisis. Specifically as the Central Bank and the Superintendency of Banks are concerned, the role of both of these institutions has been further defined and strengthened, and additional reforms are currently being considered to ensure their future political independence. Most important, the manner in which the Uruguayan authorities handled the crisis ­including their overall respect for the sanctity of contracts­ seemed to have enjoyed the tacit approval of the various parties involved, due partially to the very fluid dialogue the authorities maintained with all parties affected during the resolution of the crisis. Consequently, and once political and social consensus was reached, no major legal challenge has impeded the implementation and progress of the various policy measures undertaken by the authorities. Relations with International Financial Institutions Uruguay has, since the beginning of the crisis, been able to maintain a harmonious relationship with the IFIs, as it has been more willing to engage in a proactive dialogue and to follow their policy advice. This factor greatly contributed to creating an environment conducive to considerable advances in resolving some of the most important obstacles towards the ultimate recovery of the domestic banking system. Thus, the IMF has been intimately involved in the restructuring of BROU, the World Bank in a similar exercise with the BHU, the IFC may participate in the privatization of NBC while the IADB has provided advice in the modernization of the BCU's regulatory framework. This concerted effort of public support of the IFIs during and after the crisis can, we believe, be credited with helping to restore confidence in the government's ability to resolve the crisis and has likely reinforced the public's tacit endorsement of the government's actions. Exchange Rate Dynamics One of the most controversial policy choices adopted by the Uruguayan authorities at the time of the crisis was the decision in July 2002 to abandon the then prevailing crawling peg and allow for the free float of the Uruguayan Peso. Undeniably, the Uruguayan authorities were facing at that time an extremely complex and fluid situation and were confronted with a series of difficult policy alternatives, all of which would invariably inflict an onerous cost to the economy as a whole. To be sure, a nominal devaluation of the Peso would ­in theory­ induce a similar re-alignment of the real exchange rate, providing much needed impetus to the competitive position of Uruguay. The economic reactivation that such measure would achieve was considered to be essential in the medium and long-terms, after years of economic recession and, specially, after several months of widespread liquidity support to the ailing banking sector and the increased levels of public indebtedness that such liquidity support had ensued. In addition, a nominal devaluation would ­arguably­ contribute to stop the widespread deposit runs, as the very expectation of such a potential devaluation (and of 27 measures such as the "corralito") could be credited with partial responsibility for the deposit runs. However, the consequences of a large devaluation could be devastating in the context of a highly dollarized economy such as the case of Uruguay, where the levels of foreign currency indebtedness of the government were inordinately high and where a weaker currency would very seriously compromise bank debtor's solvency. This situation was made worse, as we have seen, by the low level of available reserves at the central bank, not strictly as it relates to the backing of the still existing dollar-denominated deposits, but rather by the realization that critically low levels of international reserves would very seriously impede servicing the mounting levels of public debt which the banking crisis had only exacerbated. It is difficult to prove or disprove the true need for allowing the Peso to depreciate, and for doing so at the specific juncture at which Uruguay chose to do it. However, given the very peculiar set of circumstances already described, the Uruguayan authorities had really no choice other than to the let the Peso float. The recent evolution of the Peso exchange rate has certainly shown that although the initial currency overshooting seemed to have made the situation worse, bank deposits started to return to the system by October 2002, and the exchange rate finally reached a certain level of stability which has continued until today. Costs of the Crisis As mentioned before, the total liquidity support provided by the government during 2002 amounted to US$2.4 billion, or approximately 20 percent of that year's GDP, while the ensuing GDP contraction and exchange rate devaluation amounted to approximately 11 and 95 percent respectively. Although the true and final costs of the crisis are still to be fully determined and absorbed (according to some IMF estimates, the government is expected to recover only about half of the liquidity extended to the banking system in 200231), the Uruguayan crisis seems to have been relatively benign when compared to similar systemic banking crises (please see table below for an illustration of other banking crises). However, it is nonetheless worth noting the devastating effects that such crises can inflict in the economic fabric of a nation. Thus, the lesson to be learnt is clear: banking crises are, by definition, very expensive, and avoiding them altogether should be any government's first priority. 31"Third Review under the Stand-By Arrangement and Request for Modification and Waiver of Applicability of Performance Criteria", Supplementary Information, IMF, June 2003, page 11. 28 Costs of Systemic Banking Crises Fiscal Cost Real Change Change in Year (Share of GDP) in GDP Exchange Rate Indonesia 1998 50% -15% -58% South Korea 1998 37% -11% -29% Thailand 1998 33% -5% -14% Mexico 1995 19% -6% -40% Malaysia 1998 16% -13% -14% Finland 1992 11% -5% -6% Sweden 1991 4% -3% 1% Philippines 1998 1% 1% -13% Source: Gerard Caprio and Daniela Klingebiel, "Episodes of Systemic and Borderline Banking Crises", in Klingbiel, Daniela and Luc Laeven, editors, "Managing the Real and Fiscal Effects of Banking Crises", World Bank Discussion Paper Num. 428, The World Bank, January 2002. Specific Banking Measures Although highly controversial, the measures specifically adopted to resolve the banking crisis may have proven to be the most relevant in achieving an orderly resolution of the crisis and the most enduring in ensuring the ultimate recovery of the confidence in the country's banking system. As we have seen, the Uruguayan authorities first attempted to counteract the fast developing deposit runs by pumping large amounts of liquidity into the ailing banks in an attempt to restore confidence in the system. However, as these efforts ultimately proved insufficient ­i.e. late July 2002­ the policy choices had to necessarily become more drastic. Preservation of Banking Contracts and the Payment System Given the severity of the crisis (by late July 2002, the Uruguayan banking sector had lost an estimated US$6 billion worth of deposits, or approximately 40 percent of total deposits within the system at the beginning of the year), the authorities chose, as we have seen, first, to declare a five-day bank holiday, second, to reprogram all time deposits at the state banks32, third, to guarantee 100 percent of sight and time deposits held at public and intervened banks, and, fourth, not to impose any limitations on the functioning of the foreign banks. Thus, by ­mostly­ respecting the nature of banking contracts33 and preserving intact the payment system, Uruguay managed not only to stop the crisis from further enveloping the economy and from causing even deeper economic hardship, but also to avoid the judicial quandary that breaching those contracts would have surely ensued. As a consequence, Uruguay has been able of maintaining a high degree of social stability and cohesion throughout the crisis, thus facilitating the eventual reconstruction of its banking sector. 32The state-owned banks have an implicit 100% government guarantee 33Although the re-programming of deposits was obviously mandatory, these actions were, nonetheless, well received by depositors, given comparatively high deposit rates. 29 Internalization of currency mismatch. As we have seen, the Uruguayan banking system was, prior to the crisis, highly dollarized. Consequently, when the crisis struck and the Uruguayan peso was sharply devalued, the inherent currency mismatch made the majority of the banking system highly insolvent. At that juncture, and in order to internalize the costs of such mismatch, the Uruguayan authorities successfully resisted the temptation to forcefully alter the currency denomination of deposits. Instead, by maintaining the currency denomination of deposits and implementing other measures to internalize the cost of the currency mismatch (i.e. write-offs, reprogrammed deposits, re-capitalization, etc.), the Uruguayan authorities managed to ­albeit temporarily and only nominally­ maintain the financial viability of the banking system as a whole. Although the full extent of the crisis is yet to be fully acknowledged and paid for34, by maintaining the majority of the banks as an on-going concern, the Uruguayan authorities stand ­arguably­ a better chance of rebuilding the country's banking sector. Early Bank Intervention The Uruguayan approach towards bank intervention was distinctly targeted. In fact, from the initial days of the crisis, the authorities were able to identify the institutions most affected by the crisis (i.e. Banco Galicia Uruguay and Banco de Comercio), and were, therefore, able to implement effective policies towards their resolution and/or liquidation. In addition, throughout the crisis, the Uruguayan authorities were very willing to admit the true extent of the problem at specific institutions and were not reluctant to initiate steps towards the eventual liquidation or resutructuring of certain institutions (i.e. Banco Montevideo-Caja Obrera and Banco de Crédito). Being able to differentiate between solvent and insolvent banks ­and doing so publicly­ had a very beneficial effect in restoring credibility in the banking system as a whole. However, although the December 2002 new banking law was an important step in facilitating the liquidation of failed banks, its applicability was specific to certain institutions and at a specific point in time. This law is geared in favor of open-bank resolution and lacks a well-defined process and mechanism for the closure, resolution and liquidation of banks. We believe, therefore, that the authorities should endeavor to develop a general framework for bank restructuring and market exit of insolvent institutions. The mechanism should go from the adoption of regularization plans (in case of capital deficiencies) to suspension of operations, exclusion of assets and liabilities of insolvent banks and bank closure. A bank resolution mechanism that contemplates the possibility of bank closure should operate in conjunction with a limited deposit insurance scheme. In summary, a general regulatory framework needs to be developed to both finalize the resolution of the banks that failed during the 2002 crisis and to facilitate the quick resolution of potential future bank failures before bank runs become systemic. 34A significant proportion of the losses incurred by this currency mismatch are expected to be eventually assumed by the government. 30 Targeted Policies Toward the Private Banks As mentioned before, the Uruguayan authorities placed no specific restrictions on the private35 banks, as long as they were able to rely on their own resources for liquidity and capital support. By so doing, the authorities effectively obtained incremental international liquidity and capital support provided by the headquarters of such institutions, while, at the same time, managed to ensure that a proportion of the fast fleeing deposits were, at least, retained within that segment of the Uruguayan banking system instead of irrevocably leaving the country. Conclusions Banking crises typically start when confidence in the system is lacking and end when confidence is finally restored. The case of Uruguay was no exception. Today, while many issues are still pending in the final resolution of this crisis, it can credibly be claimed that reasonable confidence has been restored in the country's banking system. In the ultimate analysis, it is probably true to state that the 2002 Uruguayan banking crisis would not have happened had Argentina's collapse not occurred. Nonetheless, it is also safe to affirm that Argentina's impact on the Uruguayan financial sector was magnified by some inherent weaknesses of the Uruguayan economy and banking sector. Therefore, what was a clear case of exogenous contagion became, by feeding on endogenous vulnerabilities, an altogether different and much bigger phenomenon that engulfed the entirety of the Uruguayan economy. While it is always helpful to look into the causes that contributed to the crisis, its chronology and the internal and external factors that magnified the situation, it is also worth noting that, in the case of Uruguay, the very manner in which the authorities behaved themselves during and after the crisis may have been, in and on itself, a crucial element in stopping and recovering from the crisis. A key element of the government's strategy that contributed to its effectiveness was their willingness to relatively quickly and publicly intervene troubled banks to both prevent systemic contagion and assure the worrying public about the solidity of the financial system. Furthermore, not only has Uruguay been capable of simultaneously counteracting concurrent banking and public debt crises, but it has been able to do so by preserving the necessary trust in banking contracts, achieving a high level of social stability and political cohesion, and maintaining a fluid dialogue with multilateral financial institutions and all the affected parties. In fact, it can be argued, that it is precisely the congenial and consensual approach taken by the authorities that can be credited with creating the necessary conditions to effectively overcome some of the most important obstacles towards the ultimate recovery of the domestic banking sector. * * * 35"Private" is defined in this context as non-public, not-intervened financial institutions. Thus, the vast majority of these private banks were Uruguayan subsidiaries of international banks. (Authors' Note) 31 Bibliography Banco Central del Uruguay, "Evolución Reciente y Situación Actual del Sistema Bancario", Banco Central del Uruguay, October and December 2003 Banco Central del Uruguay, "Reporte de Estabilidad Financiera", March and June 2004. Credit Suisse First Boston Research. Several Issues. Calvo, Guillermo A., Alejandro Izquierdo, and Luis-Fernando Mejía, "On the Empirics of Sudden Stops: The Relevance of Balance Sheet Effects", National Bureau of Economic Research, May 2004 de la Torre, Augusto, Eduardo Levy Yeyati and Sergio L. Schmukler, "Living and Dying with Hard Pegs: The Rise and Fall of the Argentina's Currency Board", Policy Research Working Paper Num. 2980, The World Bank, March 31st, 2003 Gutierrez, Joaquin G., and Fernando Montes-Negret, "Argentina's Banking System: Restoring Financial Viability", Working Paper Num. 2, The World Bank, January 2004. IMF, "Third Review Under the Stand-By Arrangement and Request for Modification and Waiver of Applicability of Performance Criteria", IMF, June 2003. IMF, "Fourth Review Under the Stand-By Arrangement and Request for Modification and Waiver of Nonobservance of Performance Criteria", IMF, February 2004. IMF, "Fifth Review Under the Stand-By Arrangement and Request for Modification and Waiver of Nonobservance of Performance Criteria", IMF, August 2004. IMF, "Uruguay: Letter of Intent, Memorandum of Economic and Financial Policies, and Technical Memorandum of Understanding", IMF, February 6th, 2004. IMF, "Uruguay: 2003 Article IV Consultation", IMF, August 2003. J.P. Morgan Research. Several Issues. Klingbiel, Daniela and Luc Laeven, editors, "Managing the Real and Fiscal Effects of Banking Crises", World Bank Discussion Paper Num. 428, The World Bank, January 2002. Moody's Investors Service, "Uruguay: Banking Statistical Supplement", March 2004 Perry, Guillermo, and Luis Servén, "The Anatomy of a Multiple Crisis: Why Was Argentina Special and What Can We Learn From It?", Policy Research Working Paper Num. 3081, The World Bank, June 30th, 2003 32 Reinhart, Carmen M., Kenneth S. Rogoff, and Miguel A. Savastano, "Debt Intolerance", National Bureau of Economic Research, August 2003. The World Bank, "Uruguay: Financial Sector Review", Report No. 20199-UR, The World Bank, November 15th, 2000. ___ 33