Portal de Periódicos da CAPES

What life after default? Time horizons and the outcome of the Argentine debt restructuring deal

2009; Routledge; Volume: 16; Issue: 3 Linguagem: Inglês

10.1080/09692290802454653

1466-4526

Giselle Datz,

Fiscal Policies and Political Economy

ABSTRACT ABSTRACT Although a recurrent phenomenon, sovereign debt restructurings following defaults on debt obligations do not engender today the exact same dynamics they did in the past. On the one hand, prevalent reputational models have overestimated the costs of default in terms of access to new credit. On the other hand, debtors' restructuring strategies have been underestimated in the face of anticipated constraints on developing countries' policy autonomy believed to flow from financial globalization. By focusing on the latest Argentine debt restructuring (the largest and most complex in history to date), I find that the outcome of sovereign debt restructurings in terms of credit (re)access are much less contingent on market punishments based on reputations than on the short-term nature of financial incentives in the bond industry as well as on international lending cycles. Time horizons play a crucial role in how debtors take advantage of different creditors' incentives to participate in a restructuring. A critical implication is that governments have more room to maneuver in taking contentious policy stances (such as stringent debt restructuring deals) than the literature on economic globalization accounts for. Also, far from a zero-sum game, for both debtors and some of their creditors, debt restructurings can prove to be attractive deals for their specific purposes. KEYWORDS: Sovereign debtrestructuringcreditorstime horizonsreputationsArgentina ACKNOWLEDGEMENTS I am grateful to Phil Cerny, Yale Ferguson, Anna Gelpern, Robert Kaufman, Louis Pauly and three anonymous reviewers for their comments and insight. Notes 1 Default is defined here as the failure to make required debt payments on a timely basis or to comply with other conditions of an obligation or agreement. This is the standard definition used by Standard and Poor's, one of the most influential credit rating agencies. Since my focus is on defaults in the 1990s, where credit to developing countries takes predominantly the form of international bond purchases of sovereign debt, the definition of default needs to be in agreement with the market's benchmark. 2 Credit access has no universal definition. Here I use the term as a synonymous to 'market access', defined by the IMF (2001) International Monetary Fund. 2001. 'Assessing Determinants and Prospects for the Pace of Market Access of Countries Emerging from Crises', Washington, DC: IMF. Policy Development and Review Department [Google Scholar] 'as the resumption of capital inflows mainly in the form of bonds and to a lesser extent bank loans, equity and short-term credit' (IMF, 2001 International Monetary Fund. 2001. 'Assessing Determinants and Prospects for the Pace of Market Access of Countries Emerging from Crises', Washington, DC: IMF. Policy Development and Review Department [Google Scholar]: 11). 3 See Datz (2007a) Datz, G. 2007a. "'Global-National Interactions and Sovereign Debt Restructuring Outcomes'". In Deciphering the Global: Its Scales, Spaces, and Subjects, Edited by: Sassen, S. New York: Routledge. [Google Scholar] for an account that includes other cases. 4 This index tracks on a daily basis the prices and yields of bonds issued by various emerging market countries, which rise and fall based on news, developments and interest rate trends. Each country in the index has a percentage 'weighting' – for example, 10% for Russia, 15% for Argentina, 4% for Venezuela, etc. – that depends largely on the amount of bonds it has issued compared with the total issued by all the countries combined. 5 Indeed, 19% of all sovereigns have defaulted on their debts since the beginning of the 1990s. This represents one of the highest records in the last 180 years. Higher average default rates were only registered in the 1830s where 31 sovereign countries defaulted on their obligations, in the 1980s, when the rate was 22%, and in the depression of the 1930s when the rate was 21% (Beers and Chambers, 2002 Beers, D. T. and Chambers, J. 2002. 'Sovereign Defaults: Moving Higher Again 2003?'. Standard & Poor's, Reprinted from RatingsDirect, 24 September [Google Scholar]). 6 Rarely do authors provide a definition for reputation in economic studies. When that is the case, implicit in the definition is the idea of learning over time from observed past behavior. 7 Recently, Borensztein and Panizza (2006) Borensztein, E. and Panizza, U. 2006. 'Do Sovereign Defaults Hurt Exporters?', Washington, DC: Inter-American Development Bank. unpublished manuscript[Crossref] , [Google Scholar] find a 'statistically significant and economically sizable effect of default on trade', even though this effect is short lived. Despite working with an impressive sample of 9700 observations for 24 countries, the study focuses on data from 1980 to 2000, leaving aside the Argentine default discussed here. 8 An example of such a study is found in Rose (2001) Rose, A. 2001. 'One Reason Countries Repay Their Debts: Renegotiation and International Trade' NBER Working Paper 8853 [Google Scholar]. In order to check for default's detrimental spillover effects on trade, the author looks at data from 1948 to 1977. Despite being a relatively recent study, it is surprising that the data set used to make a contemporary argument – to be applicable to debt theory today – deals with data that does not encompass any one of the three key phases of sovereign indebtedness and defaults: the 1930s, the 1980s, and the 1990s/early 2000s (Rose, 2001 Rose, A. 2001. 'One Reason Countries Repay Their Debts: Renegotiation and International Trade' NBER Working Paper 8853 [Google Scholar]). For another example focusing on a possible collusion of creditor banks, see Wright (2002) Wright, M. L.J. 2002. 'Reputations and Sovereign Debt', Palo Alto, CA: Stanford University. unpublished manuscript [Google Scholar]. 9 This periodization includes the severe financial crises of the 1990s and only leaves out the Argentine default of December 2001. 10 When reputations were not enough to explain repayment, economists tried to relate it to the threat of direct sanctions – usually trade related. However, as Dooley (2000) Dooley, M. 2000. 'International Financial Architecture and Strategic Default: Can Financial Crises be Less Painful?', Santa Cruz: University of California. manuscript [Google Scholar] pointed out, the trouble with these enforcement mechanisms is that they are not observed. Instead, Dooley suggested that repayment occurs because sovereigns want to prevent output loss. Indeed, some severe financial crises provoked output contraction, but not all of them led to default. The Russian default, for example, was not matched by severe output loss (Roubini and Setser, 2004 Roubini, N. and Setser, B. 2004. Bailouts or Bail-ins? Responding to Financial Crises in Emerging Economies, Washington, DC: Institute of International Economics. [Google Scholar]). Alternatively, crises in Thailand, Korea, Indonesia and Malaysia were followed by large output loss and no default. 11 It is worth clarifying that in this analysis the focus is on credit (re)access post-default as a key component of what I call 'policy leeway'. To be sure, other elements also define leeway to follow an alternative course of action than that favored by private creditors, the G-7, and international financial institutions. In the case of study here, Argentina, the retreat of the IMF – as determined to a large extent by (in)action from the Bush administration – vis à vis the renegotiation of Argentine debt may have played a role in allowing the country more room to move towards a more radical stance than any debt negotiation in the past (Helleiner, 2005 Helleiner, E. 2005. 'The Strange Story of Bush and the Argentine Debt Crisis'. Third World Quarterly, 26(6): 951–69. [Taylor & Francis Online], [Web of Science ®] , [Google Scholar]). Yet, it should be noted that this apparent inaction does not exclude the fact that the US Treasury was keen on following the events closely. A particular concern of the US was not to make sure the IMF's seniority regarding Argentine debt would not be undermined (Interview with former Argentine official, Buenos Aires, November 2005). In addition, the inclusion of collective action clauses in recent bond contracts has been associated primarily with 'moral suasion' on the part of the US government on sovereigns issuing new bonds in the early 2000s (Gelpern and Gullati, 2006 Gelpern, A. and Gullati, M. 2006. 'Public Symbol in Private Contract: A Case Study' Duke Law School Faculty Scholarship Series, Paper 50 [Google Scholar]). 12 In their model for determinants of bank lending to EMs, the dependent variable used is the change in cross-border claims on EMs listed by banks from the USA, the UK, France, Germany, and Spain (all BIS-reporting countries). In their model for the determinants of bond spreads, the dependent variable used is the log spread on JP Morgan's EMBI Global index 13 When it comes to long-term US interest rates, however, Ferrucci et al. (2004) Ferrucci, G., Herzberg, V., Soussa, F. and Taylor, A. 2004. 'Understanding Capital Flows to Emerging Market Economies'. Financial Stability Review, : 89–97. [Google Scholar] agree with Eichengreen and Mody (1998) Eichengreen, B. and Mody, A. 1998. 'What Explains Changing Spreads on Emerging-Market Debt: Fundamentals or Market Sentiment?' NBER Working Paper 6408[Crossref] , [Google Scholar] and McGuire and Schrijvers (2003) McGuire, P. and Schrijvers, M. 2003. 'Common Factors in Emerging Market Spreads'. BIS Quarterly Review, : 65–78. December [Google Scholar] in that these rates have a strong negative effect on the spreads of EM bonds. 14 Eichengreen and Mody study a sample of 1300 developing country bonds launched in the years 1991–97. When it comes to their finding for Latin American bond in particular, the authors conclude that correlation coefficients between the Latin American countries in their sample for the period between December 1997 and September 2000 indicate that although markets have some capacity to differentiate amongst Latin American countries on the basis of policies and fundamentals, they also perceive Latin American countries as a group when assigning risk. This practice is conducive to market volatility and indicates that market sentiment is indeed a key component in the determination of spreads. 15 The study compared sovereign bond yields in emerging markets in 1870–1913 – a golden era for international capital flows – with the yields of the period from 1992 to 2000, using a sample of 16 historical 'emerging markets', including Brazil, Argentina, Chile, and Mexico. 16 It is interesting to note that, in 2001, the biggest country to move in the index was Argentina, the weighting of the country dropping from more than 20% at the beginning of the year to less than 3% by the end of the year (Santiso, 2003 Santiso, J. 2003. The Political Economy of Emerging Markets: Actors, Institutions, and Financial Crises in Latin America, New York: Palgrave. [Crossref] , [Google Scholar]). 17 Several scholars and market players are aware that index following and benchmark imperatives contribute to herd behavior in financial markets, playing a crucial role also in explaining contagion effects at times of financial crises (Calvo and Mendoza, 2000 Calvo, G. and Mendoza, E. 2000. 'Rational Herding and the Globalization of Financial Markets'. Journal of International Economics, 51: 79–114. [Crossref], [Web of Science ®] , [Google Scholar]). 18 Funds with incentive fees usually take on higher risk, often pursuing non benchmark strategies (Santiso, 2003 Santiso, J. 2003. The Political Economy of Emerging Markets: Actors, Institutions, and Financial Crises in Latin America, New York: Palgrave. [Crossref] , [Google Scholar]: 111). An incentive fee is a reward structure that makes management compensation a function of investment performance relative to some benchmark. For example, hedge funds typically charge investors a fixed fee plus an incentive fee equal to between 5% and 25% of the funds' annual return (Elton et al., 2001 Elton, E., Gruber, M. and Blake, C. 2001. 'Incentive Fees and Mutual Funds', NY: NYU and Fordham University. unpublished manuscript[Crossref] , [Google Scholar]). 19 Outside of a benign context, however, even non defaulters may 'pay' for simply being more risky options of investment than liquid instruments such as US Treasury bonds. 20 Local banks' loans to the public sector constituted about 27% of the banks' assets in October 2001 (Blustein, 2005 Blustein, P. 2005. And The Money Kept Rolling in (And Out), New York: Public Affairs. [Google Scholar]). 21 However, the deal did not change negative perceptions of Argentina in the market. Country risk for Argentina, a few days later, continued to follow a downward spiral (Machinea, 2002 Machinea, J. L. 2002. 'La Crisis de La Deuda, El Financiamiento Internacional y la Participación del Sector Privado' [The debt crisis, international funding and private sector participation], Santiago, CA: CEPAL. Serie Financiamiento del Desarollo, No. 117 [Google Scholar]: 66). 22 The logic the government offered to justify this initiative was that the funds would go into the Treasury's account at the Banco de la Nación (the state bank) so the bank could manage to make regular payments and so other banks could make all pension and salary payments. The guarantee for the loans would be the revenues collected through a new financial transactions tax. 23 Two incentives were attached to this swap deal. First, the new instruments would be valued at par rather than marked-to-market (i.e., on a daily basis), which was characterized as an accounting benefit. Second, the threat that in the near future the government would face a default and thus force an 'involuntary' (tougher) restructuring on its creditors was a clear incentive for creditors to participate in the swap of bonds at that time rather that at a later date. 24 Phase 2 of this swap operation, which aimed at restructuring debt held by external creditors (almost $58 billion in bonds), was not completed until the beginning of 2005. 25 This initiative raised the total investments in government paper in pension funds' portfolios to 70% in that year, and would increase further in 2002 (to 76.7%), as government reliance on pension fund investment increased (Kay, 2003 Kay, S. J. . 'Pension Reform and Political Risk'. paper presented at the LASA XXIV International Congress. March27–9, Dallas, TX. [Google Scholar]). 26 Pension fund managers complained that the government was forcing them to act against the best interests of their clients since they would be forced to accept lower overall returns. The only fund that ended up accepting this move was the publicly managed the Nacíon fund, administered by the public bank, the Banco de la Nación. The other funds, by not accepting the bond exchange, were then swapped back to the old defaulted bonds. 27 Roberto Lavagna replaced Minister Cavallo in 2002, during the interim rule of President Eduardo Duhalde. Lavagna remained in the government after Kirchner became the newly elected President in 2003 and left office on 28 November 2005. 28 The Argentine executive branch counted on the support of the majority of the country's senators who interrupted vacations to approve the bill. The House also voted in favor of this piece of legislation on 10 February 2005, soon enough to still impact favorable the outcome of the swap. The 'cerrojo' bill made it illegal for the government to enter into any 'judicial, extra-judicial or private agreement' with non participating bondholders (Dow Jones Newswires, 2006). Though it is hard to access how much the most favored creditor clause impacted the swap outcome, it is clear that it was a powerful domestic tool, mounting substantial support to the government's handling of the debt problem (Interview with former official at the Ministry of the Economy, Buenos Aires, November, 2005). 29 Holders of Argentina's dollar-denominated and euro-denominated bonds could exchange them for Argentine peso-denominated bonds under the swap deal. Some hedge funds also bought Argentine debt in the hope that a better offer would materialize. 30 The bonds would be upgraded to a scale of 'B–', which was reported as 'a potentially good sign for the government' that angered bondholder activist who still wanted to pressure for a better offer (Dow Jones Newswires, 2005). It is worth noting that it took Ecuador 5 years after its default to attain the same rating. 31 There was indeed no negative reaction from these institutions post-default, except for the IMF's recurrent mentioning that the Argentine government needed to address the case of those bondholders who did not join the swap. 32 Author's translation. 33 It is also important to note that Brazil received large loans from the IMF, especially in the 2002 debacle preceding the Presidential election of Lula da Silva. 34 A government official involved in the restructuring resisted the term 'compensated' but confirmed that the funds were offered a 'very good deal'. Interviews with pension funds' CEOs and CIOs backed up this view from an industry perspective. These interviews were conducted by the author in Buenos Aires in November of 2005. 35 The sample included 1200 households in different regions of Argentina. 36 Kirchner has taken on a role played by former President Carlos Menem, who implemented the reforms of the 1990s, within the Peronist party. 37 That amount included early repayments of debt due in 2006, 2007 and 2008 (La Nación, 2005d). 38 The announcement of early repayment took Wall Street by surprise but was only delivered after the Kirchner team had communicated its intentions to both the IMF's director, Rodrigo de Rato, and the undersecretary of the US Treasury, Timothy Adams (La Nación, 2005d). The initiative to break ties with the Fund came after minister of the economy Roberto Lavagna was replaced by former president of the state bank (Banco Nación), Felisa Miceli in earlier December 2005. Lavagna had mentioned that he intended to enter talks with the IMF over a new agreement before he left office (Financial Times, 2005b). The announcement was supposed to only be made by late December, but the Brazilian initiative to announce their own early repayment and the favorable reception that had in financial markets made the Kirchner team anticipate its own parallel move which had – till then – remained a 'well kept secret' (La Nación, 2005d; author's translation). However, the market was less thrilled by Argentina's repayment than it was by Brazil's initiative. Analysts saw Brazil's decision as one deriving from its strong position in credit markets, a reality not shared by Argentina, which was seen in a more delicate position then (La Nación, 2005e, 2005f). Yet, the loss of reserves due to the early repayment did not lead to substantial vulnerability. 39 Author's translation. Argentina had some leeway prior to the early repayments. The country suspended its agreement with the Fund in August 2004 and had since repaid its debts to the organism as they fell due (The Economist, 2005). 40 Since 2003 the bolivar has been pegged to the dollar. 41 The figure announced by Venezuela's finance team was $309 million. Some sources, like The Economist, reported on the deal but suggested skepticism in trusting the government's numbers. Indeed, according to the Financial Times, the profit was around $200 million (Financial Times, 2006). 42 This notion holds also for European retail investors who flocked the market for emerging bonds after the Brady Plan in 1989 while commercial banks who had suffered from the 1980s defaults were more timid in joining in. I thank one of the careful anonymous reviewers of this article for suggesting the analogy. 43 Author's translation. 44 In Argentina a key vulnerability concerns a possible return of higher rates of inflation. 45 Backing up the government in every stage of the process were lawyers Roger Thomas, Carmen Corrales and others at Cleary Gottlieb, the law firm working with Argentina. According to Nielsen's recollections they played a 'key role in the drafting of the prospectus that was also been over seen by the lawyers of the banks [UBS, Barclays and Merrill Lynch]. Cleary Gottlieb was also crucial in dealing with the regulators in the many markets we had to exchange'. In addition: a hidden part of the transaction was in the hands of the exchange agent (the Bank of New York), the liability management deposits of the banks and all the clearing systems and we had to deal with all of them, in most cases with the able assistance of the people at the Caja de Valores, the Argentine clearing system. The banks had all the radars pointed to the markets, fine tuned with the transaction. (Euromoney, 2006) 46 Interview with former debt negotiator for the Argentine government. Some Italian investors who bought Argentine bonds following the unwise advice of their banks sued those local intermediaries for pushing assets that were not to be taken on by risk-averse, savings-oriented investors. Italians held about 16.5% of Argentine debt, others in the European Union held 9.6%. 47 Interview with a private macroeconomic consultant operating in Buenos Aires, November 2005. 48 Author's translation. 49 In early studies of the Argentine default, analysts noted that Nación (the pension fund that accepted the 'pesification' of their assets after the default, when other pension fund firms did not) had profited and attracted more contributors than its competitors (Kay, 2003 Kay, S. J. . 'Pension Reform and Political Risk'. paper presented at the LASA XXIV International Congress. March27–9, Dallas, TX. [Google Scholar]). However, in November 2006, Nación was being outperformed exactly by those pension funds that, by not accepting the pesification, were stuck with defaulted bonds. Those pension funds are now profiting from the high returns derived from their swapped, GDP-indexed, bonds. For further details, see Datz (2007b) Datz, G. . 'Pension Funds and Sovereign Debt in Latin America'. paper presented at the Annual Convention of the International Studies Association. 28 February–3 March, Chicago, IL. [Google Scholar]. 50 Most of these gains were enjoyed in the last 4 months of the 10-month period. 51 Though it could be argue that political costs need to include an analysis of the banking sector, which certainly incurred in some losses, those were never computed and a direct association is less empirically feasible than when one studies the local pension funds. In addition, it is important to note that pension funds were mostly owed by international banks.