Sovereign Debt Litigation in Argentina: Implications of the Pari Passu Default

On 31 July 2014, Argentina defaulted on its sovereign bonds for the second time in the 21st century. It was also its eighth default since independence ¹ ; at such frequency, this was perhaps not an especially noteworthy event. What made it so extraordinary was not that another domestic financial crisis triggered the payment default, but rather an injunction handed down by a federal district court in New York. However, despite public outrage, the wider impact of this decision is likely to be limited. That is even more so if reforms that have already started continue to be implemented.

1. BACKGROUND

The case dates back to Christmas Eve 2001, when the interim Saá administration declared a payment suspension on foreign debt. ² The first scheduled payments affected by this decision were due in January 2002. Only weeks later, in March, the first lawsuits by American investors against the Argentine government were filed in US courts (see Figure 1 ). But a real rush to the courthouse took place from September 2003 onwards, when the first details of the intended debt restructuring were published. With a present value loss (‘haircut’) for investors of approximately 90 per cent, they were considerably harsher than most bond exchange offers since the 1990s, which saw average haircuts of 35 per cent. ³ A debt exchange finally took place in June 2005 with a haircut of approximately 77 per cent. About 76 per cent of creditors accepted the offer, later increased to slightly more than 90 per cent after a reopening at identical terms in 2010. Afterwards, Argentina resumed payments to participating investors on the new ‘exchange’ bonds, but refused to make any payments to the ‘holdout’ creditors who kept the original bonds.

Figure 1.

Lawsuits filed since 2002. Cumulative nominal claims in litigation against the Republic of Argentina, Provinces of Buenos Aires or Mendoza in the aftermath of the 2001–02 default, by date of filing. Data based on Schumacher ((n 4).

During the subsequent 12 years, more than 180 separate actions demanding repayment of the original bonds were filed in New York alone. ⁴ All cases brought to trial seeking judgment were decided in favour of the plaintiffs (with a single exception against the Province of Mendoza). However, none of the cases succeeded in seizing assets for enforcement.

This lack of success led a group of hedge funds, most notably NML Capital, to invoke a controversial interpretation of the pari passu , or equal step, clause. ⁵ In a previous case against Peru, ⁶ courts had interpreted the clause as mandating ratable equal payment on all obligations subject to such a provision. Many observers have rejected this interpretation as unfounded. But in the fiscal agency agreement governing the defaulted bonds, the contracting parties had gone one step further: besides promising the pari passu ranking of the securities , the agreement additionally stated that ‘the payment obligations  … shall at all times rank at least equally with all its other … External Indebtedness’ (emphasis added). ⁷ Formulations enhancing pari passu clauses with explicit reference to equal payment obligations are considered particularly vulnerable to a ratable payment interpretation. ⁸

The application of the argument to the Argentine default had long been coming. In 2003, in one of the first cases filed, ⁹ Argentina had unsuccessfully applied for declaratory relief that the pari passu clause could not be the basis for ‘interfering with payments to other creditors’. Since the creditors had not invoked this argument, the court rejected the motion. Yet seven years later, in October 2010, NML (later joined by Aurelius and a number of smaller investors) did just that. In December 2011, judge Thomas P Griesa followed the plaintiffs’ argument and ordered that by repaying the exchange bonds, Argentina was violating the terms of the original bonds. ¹⁰

At first glance, all this meant was another judicial confirmation to the holdouts that Argentina owed them full repayment according to the original bonds’ terms. But this was something they had already achieved in virtually all cases before, too. What made the development special was Griesa’s decision in February 2012 to grant an injunction prohibiting Argentina from making 100 per cent of the reduced payments due on the exchange bonds unless the government would also transfer 100 per cent of the full payments due on the original bonds to the plaintiffs.

This meant not only that Argentina was breaching its obligations towards the holdouts, but that the banks assigned with processing the payments were assisting in that breach of contract. The Argentine government, as a foreign sovereign entity, had by and large been able to ignore judgments entered against it. But Bank of New York Mellon, trustee for the bondholders, is subject to the full enforcement powers of the US court system, and as such can be held directly accountable. The district court’s decision was confirmed by the Court of Appeals for the Second District, ¹¹ and an application for certiorari at the Supreme Court was rejected in June 2014. ¹²

2. CREDITOR COORDINATION BECOMES MORE DIFFICULT

Due to the lack of enforcement possibilities against foreign governments, the economic literature has looked for alternative default penalties inducing governments to repay their debt. A set of papers have argued that the disturbances caused by holdout litigation, such as disruptions to market access or trade, can be an additional default penalty—giving governments incentives not to default in the first place. ¹³ Under this assumption, strengthening creditor rights through sovereign debt litigation may provide the positive externality of deterring government default and excessively harsh treatment of creditors, thereby stabilizing sovereign debt markets. ¹⁴

But governments’ costs from litigation are only one side of the game. Creditors’ individual decision-making depends on the relative expected value from either participating in an exchange offer and receiving the new, reduced claim with relative certainty, or rejecting the offer and trying to recover the full claims in court. The latter option has previously been a risky choice, since success was far from certain. This implied a rising, but still relatively small share creditors filing lawsuits. Most of these were led by highly specialized creditors contesting especially those public debt restructurings with high investor losses. ¹⁵

Pari passu , however, could obscure the positive default penalty-effect of litigation by giving too much power to holdouts. Enforcing ratable payment of original and exchange bonds by binding domestic financial institutions makes the repayment probability of holding out exactly as high as of accepting the offer. This could induce a much larger share of creditors to reject future restructuring scenarios. Applying the ratable payment provision would then inflate the default penalty of litigation to such prohibitive levels that orderly debt restructuring would become impossible. Negotiated agreements, which have been a successful way of overcoming public debt crises, would be ineffective.

3. CONTRACTUAL REFORM IS NECESSARY—AND POSSIBLE

The Argentine case thus causes uncertainty about whether orderly restructurings will still be possible. Observers have reacted with numerous proposals for establishing a full-scale international bankruptcy court for governments, as last contemplated by the International Monetary Fund (IMF) more than a decade ago. ¹⁶ The 68th UN General Assembly adopted a resolution suggesting such a measure in September 2014, calling for a multilateral solution instead of the previous market-based approach. ¹⁷ While a large majority of 124 countries voted in favour, the list of countries abstaining or voting against the resolution is more revealing. Of the 20 permanent members of the Paris Club, representing global bilateral debt claims of more than USD 330bn, only one supported the resolution—Russia, incidentally also the only member country that defaulted in the past 60 years. With such a clear divide in political support between creditor and debtor countries, the implementation of a statutory framework seems simply not feasible.

Furthermore, the Argentine case provided an exceptional, if not unique, combination of factors. A high haircut, combined with a very tough negotiation pattern on behalf of the government, ¹⁸ and a pari passu clause explicitly promising equal payment meant that a strong default penalty by market participants was likely. For each of these reasons, a less radical reform effort seems more practical. In fact, first steps have already been undertaken. ¹⁹

Two crucial contractual features need to be amended. First, stronger collective action clauses (CACs) ought to force holdouts into accepting an offer, provided an aggregate supermajority across all eligible instruments is achieved. A proposal by the International Capital Markets Association (ICMA), recently implemented by Mexico in a New York bond prospectus, ²⁰ contains such a possibility. According to this new CAC, a bond will be completely restructured if either (1) a security-specific majority of 75 per cent accepts; (2) an aggregate majority of 66.67 per cent and a security-specific majority of 50 per cent are reached, or (3) an aggregate majority of 75 per cent accepts.

CACs have been criticized as not solving a situation like Argentina’s crisis with many separate instruments. ²¹ But the criticism usually refers to the first option: ie. reaching a security-specific 75 per cent majority. Indeed, as Figure 2 and Table 1 show, security-specific CACs would have significantly reduced the share of holdouts in Argentina, but would still have failed in 63 bond issues. A ‘two-limb’ approach under option 2, given the aggregate participation rate of 76 per cent, would have reduced that number to only nine bonds. And applying option 3 would have completely wiped out all holdouts—despite the fact that the participation rate was the lowest of all recent sovereign bond restructurings. ²²

Figure 2.

Holdout rates in Argentine bonds. The graph shows a histogram of holdout rates in Argentine sovereign bonds in the 2005 restructuring. The bonds to the left of the vertical lines would have been completely restructured had the bonds contained the respective collective action clauses. Data are based on Schumacher (n 4).

Table 1.

Participation rates under different CACs

CAC Majority Method	Number of Bonds Fully Restructured	Percentage of All
None	16	11
(#1) Bond-specific 75%	82	57
(#2) Bond-specific 50% + aggregate 66.67%	136	94
(#3) Aggregate 75%	145	100

The table shows in how many Argentine bond holdouts would have been forced into accepting the restructuring under different collective action clauses, given the observed participation rates. Data based on Schumacher (n 4).

Secondly, an adjustment of the pari passu clause should be explicit about only implying equal rank in terms of treatment (such as extending exchange offers), but not with respect to payment. The ICMA proposal and Mexico’s example lead the way here, too: ‘this shall not be construed so as to require Mexico to make payments … ratably’. Compared to the explicit ‘equal payment’ promise of Argentina’s bonds it is hard to see how an interpretation similar to NML could arise.

The pricing impact of these contractual reforms is likely to be limited. To date, there is no sufficient data available to estimate the costs of amending the pari passu clause. But a number of studies have shown that borrowing under contractual provisions that make restructuring easier does not require higher borrowing costs. Bardozzetti and Dottori show that CACs may actually lower the borrowing costs unless countries are close to default. ²³ Similarly, Chamon et al provide evidence that issuing ‘harder-to-restructure debt’ under foreign law only leads to marginally lower interest rates for countries experiencing severe financial distress. ²⁴

Reforming the contractual framework of an outstanding debt stock with residual maturities reaching, in some cases, decades into the future will not be quick or easy. But the earlier refinancing starts with new provisions, the faster the process can be completed. In the absence of more feasible alternatives, it is the only way to keep the Argentine experience confined to uniquely recalcitrant debtors.

1 Carmen Reinhart and Kenneth Rogoff, This Time is Different: Eight Centuries of Financial Folly (Princeton University Press 2009).

2 The data in this paragraph are from Federico Sturzenegger and Jeromin Zettelmeyer, Debt Defaults and Lessons from a Decade of Crises (MIT Press 2006).

3 For the haircut data, see Juan J Cruces and Christoph Trebesch, ‘Sovereign Defaults: The Price of Haircuts’ (2013) 5(3) Am Econ J Macroecon 85.

4 All litigation data are based on Julian Schumacher, ‘Coordination Problems in Sovereign Debt Restructurings: Holdouts and Litigation in Argentina’ mimeo 2014, Humboldt University Berlin.

5 NML Capital v Republic of Argentina , 08 Civ 6978, 09 Civ 1707, and 09 Civ 1708, SDNY.

6 Elliott Associates v Republic of Peru , 96 Civ 7917, SDNY, 25 September 2000; also 2000/QR/92, Brussels Court of Appeals, 26 September 2000.

7 Fiscal Agency Agreement between The Republic of Argentina and Bankers Trust Co, 19 October 1994; < http://www.shearman.com/~/media/Files/Services/Argentine-Sovereign-Debt/2013/FiscalAgencyAgreementOct191994.pdf >, accessed 15 December 2014.

8 Mitu Gulati and Robert E Scott, TheThree and a Half Minute Transaction: Boilerplate and the Limits of Contract Design (University of Chicago Press 2012).

9 EM Ltd v Republic of Argentina , 03 Civ 2507, SDNY.

10 See NML , order handed on 7 December 2011.

11 NML Capital, Aurelius Capital Master v Republic of Argentina , 12 Civ 105, 2nd Circ, 26 October 2012.

12 Republic of Argentina v NML Capital et al , 13-990, Supreme Court.

13 See, among others, Jeremy Bulow and Kenneth Rogoff, ‘A Constant Recontracting Model of Sovereign Debt’ (1989) 97 J Pol Econ 155; Patrick Bolton and Olivier Jeanne, ‘Structuring and Restructuring Sovereign Debt: The Role of a Bankruptcy Regime’ (2007) 115 J Pol Econ 901; Julian Schumacher, Christoph Trebesch and Henrik Enderlein, ‘Sovereign Defaults in Court’ mimeo 2014, Humboldt University Berlin/University of Munich.

14 See Michael P Dooley, ‘International Financial Architecture and Strategic Default: Can Financial Crises Be Less Painful?’ (2000) 53 Carnegie-Rochester Conference Series on Public Policy 361; Andrei Shleifer, ‘Will the Sovereign Debt Market Survive?’ (2003) 93(2) Am Econ Rev 85.

15 Julian Schumacher, Christoph Trebesch and Henrik Enderlein, ‘What Explains Sovereign Debt Litigation?’ mimeo 2014, Humboldt University Berlin/University of Munich.

16 Among many others, see Lee C Buchheit and others, Revisiting Sovereign Bankruptcy (Brookings Institution 2013).

17 UN Resolution A/RES/68/304.

18 Henrik Enderlein , Christoph Trebesch and Laura von Daniels, ‘Sovereign Debt Disputes: A database on government coerciveness during debt crises’ (2012) 31 J of Intl Money and Finance 250.

19 See International Capital Markets Association, Standard Collective Action and PariPassuClauses for the Terms and Conditions of Sovereign Notes (2014) < http://www.icmagroup.org/resources/Sovereign-Debt-Information/ >, accessed 15 December 2014.

20 SEC 2014. United Mexican States registration no 333-185462 , < http://www.sec.gov/Archives/edgar/data/101368/000119312514405539/d816959d424b2.htm >, accessed 15 December 2014.

21 Anna Gelpern and Mitu Gulati, ‘The Wonder-Clause’ (2013) 41 J Comparative Economics 367.

22 Moody’s, The Role of Holdout Creditors and CACs in Sovereign Debt Restructurings (Moody’s 2013).

23 Alfredo Bardozzetti and Davide Dottori, ‘Collective Action Clauses: How Do They Affect Sovereign Bond Yields?’ (2014) 92 J Intl Econ 286.

24 Marcos Chamon, Julian Schumacher and Christoph Trebesch, ‘Foreign Law Bonds: Can They Reduce Sovereign Borrowing Costs?’ mimeo 2014, University of Munich.