Emory International Law Review

Settling Sovereign Debt’s “Trial of the Century”
Juan J. Cruces,
Tim R Samples Juan J. Cruces is Professor of Finance and Economics at the Universidad Torcuato Di Tella Business School.Tim R Samples is Assistant Professor of Legal Studies, Terry College of Business, University of Georgia.We are grateful for the outstanding research assistance of Federico Varni, especially on the history of NML litigation. Juan Martín Cappellini, Corey Goerdt, Patricio Goldstein, Victoria Lacasa, Martín Przybylski, and Jonathan Tonge also provided valuable research assistance. Anna Gelpern, Martin Guzman, Matías Isasa, Stephen Park, and Andrea Rotnitzky provided valuable comments and/or conversations. The usual caveat applies. A prior version of this Article was presented at the Interdisciplinary Sovereign Debt Research and Management Conference at Georgetown University Law Center in January 2016. This research was supported in part by the Latin American and Caribbean Studies Institute (LACSI) at the University of Georgia through a LACSI Ambassador Faculty Travel Grant.

Abstract

NML v. Argentina, the “trial of the century” in sovereign debt, is finally poised for settlement negotiations. International experience, incentives for the parties themselves, and even statements by the presiding federal judge, all suggest that it is high time for a settlement between the parties. However, major challenges remain. In this Article, we analyze a subset of the key economic and legal factors underlying this litigation, with a particular emphasis on issues relevant to a potential settlement. We document the wide heterogeneity of holdout rates across Argentina’s 150 defaulted bonds (of which seventy-four still have holdout rates greater than five percent) and focus the subsequent analysis on the seven most held-out bonds—which have holdout rates between twenty and eighty-two percent and account for about thirty percent of total holdout principal. We show that New York’s statutory real rate of interest on overdue interest has been 6.6% on average during the years affecting this suit compared to 3.1% during the previous forty years. As such, this rate has become more punitive than compensatory. We also illustrate the growth of the value of holdout claims for the seven bonds from their initial $1.7 billion in principal up to $4.3 to $7 billion in current value, depending on when holdouts obtained judgments. We analyze the sensitivity of holdout claims to different approaches to overdue interest—an issue that has become increasingly controversial in New York state law in recent years. We next assess the returns that investors would have obtained by purchasing the seven-bond basket at different times since 2002. We find that investors would have multiplied their money an average of eight times if they obtained judgments in 2008 or thirteen times in 2015. Finally, we compute the current value of Argentina’s 2005 exchange offer and find that is worth about one-half of the litigants’ claims for judgments obtained in 2008. Our analysis offers a framework for potential settlement negotiations. However, with so many holdouts unaccounted for, a settlement with the NML litigants exposes Argentina to the tyranny of the next litigant as long as the current injunctions remain in place. We close by underscoring the benefit of modifying or lifting these injunctions as Argentina begins negotiating in good faith to reach a reasonable settlement with its holdout creditors.

Introduction

All eyes are once again on NML v. Argentina (NML) as sovereign debt’s “trial of the century” has entered the endgame phase. 1Joseph Cotterill, Coverage of the Argentina Sovereign Debt Litigation, Fin. Times: FT Alphaville, http://ftalphaville.ft.com/tag/pari-passu-saga/ (last visited Oct. 8, 2016). Given its extraordinary nature and importance, the Financial Times coined NML the “trial of the century” of sovereign debt restructuring. Id. See NML Capital v. Republic of Argentina, 699 F.3d 246 (2d Cir. 2012). The litigation collectively involves plaintiffs, led by NML Capital, Ltd., who have claims against Argentina for New York law bonds in default. Id. at 487. Talks between Argentina and holdout creditors were even a prominent storyline at the 46th World Economic Forum in Davos, Switzerland. 2See, e.g., Chris Giles, Gillian Tett, Elaine Moore & Benedict Mander, Argentina Pledges to Honour Debts Owed to Holdout Creditors, Fin. Times (Jan. 22, 2016, 12:02 PM), http://www.ft.com/intl/cms/s/0/356cbbae-c0f1-11e5-9fdb-87b8d15baec2.html#axzz3yISe8tio. After a hostile standoff with U.S. courts under the Kirchner administration, the recently elected Macri government pledged to negotiate a settlement and quickly put forth an offer. 3Id. The outcome of Argentina’s debt dispute has critical implications for sovereign debt markets, which are a systemically important component of the global economy. 4See Hold-outs Upheld, Economist (Nov. 3, 2012), http://www.economist.com/news/finance-and-economics/21565635-court-ruling-against-argentina-has-implications-other-governments-hold-outs. Recent events—including crises in Greece, Puerto Rico, and Ukraine—underscore the implications of sovereign debt markets for policymakers, financial systems, and ordinary citizens alike. 5See Michael Davies & Tim Ng, The Rise of Sovereign Risk: Implications for Financial Stability, BIS Q. Rev., Sept. 2011, http://www.bis.org/publ/qtrpdf/r_qt1109g.pdf.

Argentina’s debt saga began with an $81.3 billion default in 2001, the largest-ever sovereign default at that time. 6See infra Part I.A (reviewing Argentina’s default and restructurings in detail). A number of “holdout” creditors—with bonds worth $6 billion at face value—sat out of Argentina’s 2005 and 2010 debt restructurings, with many opting instead for litigation. 7In this context, “holdout” creditors are those who decide not to participate in a debt restructuring whereas “exchange” creditors do. A typical restructuring involves an exchange of defaulted or distressed debt for new debt. See infra Part I.C (reviewing holdout rates in the 2005 and 2010 restructurings in detail). After years in the Second Circuit, NML finally hit a boiling point in 2014. 8Dan Rosenheck, Argentina’s Rational Default, New Yorker (Aug. 7, 2014), http://www.newyorker.com/business/currency/argentinas-rational-default (characterizing Argentina’s default as “perhaps the strangest in history”). See infra Part I.B (reviewing Argentina’s holdout litigation in detail). Following a string of losses in federal courts and a failed petition to the U.S. Supreme Court, Argentina again slipped into default as payments to exchange creditors were blocked by a court injunction. 9See Rosenheck, supra note 8. Argentina’s debt imbroglio offers valuable examples and lessons at various stages in sovereign finance—issuance, default, restructuring, litigation, and post-litigation. This Article focuses on the post-litigation stage.

NML ignited widespread commentary in academic, policy, and industry circles. Building on a substantial body of existing literature involving sovereign debt restructuring and litigation, 10See, e.g., William W. Bratton, Pari Passu and a Distressed Sovereign’s Rational Choices, 53 Emory L.J. 823 (2004); Lee C. Buchheit & G. Mitu Gulati, Exit Consents in Sovereign Bond Exchanges, 48 UCLA L. Rev. 59 (2000); Lee C. Buchheit & Ralph Reisner, The Effect of the Sovereign Debt Restructuring Process on Inter-Creditor Relationships, 1988 U. Ill. L. Rev. 493 (1988); Ugo Panizza, Federico Sturzenegger & Jeromin Zettelmeyer, The Economics and Law of Sovereign Debt and Default, 47 J. Econ. Literature 651 (2009); W. Mark C. Weidemaier, Sovereign Immunity and Sovereign Debt, 2014 U. Ill. L. Rev. 67 (2014). scholars explored the significant implications of Argentina’s default and the NML fallout. 11See, e.g., Karen Halverson Cross, The Extraterritorial Reach of Sovereign Debt Enforcement, 12 Berkeley Bus. L.J. 111 (2015); Arturo C. Porzecanski, From Rogue Creditors to Rogue Debtors: Implications of Argentina’s Default, 6 Chi. J. Int’l L. 311 (2005); John A. E. Pottow, Mitigating the Problem of Vulture Holdout: International Certification Boards for Sovereign-Debt Restructurings, 49 Tex. Int’l L.J. 219 (2014); Tim R Samples, Rogue Trends in Sovereign Debt: Argentina, Vulture Funds, and Pari Passu Under New York Law, 35 Nw. J. Int’l L. & Bus. 49 (2014); Julian Schumacher, Sovereign Debt Litigation in Argentina: Implications of the Pari Passu Default, 1 J. Fin. Reg. 143 (2015); S.I. Strong, Rogue Debtors and Unanticipated Risk, 35 U. Pa. J. Int’l L. 1139 (2014); W. Mark C. Weidemaier & Anna Gelpern, Injunctions in Sovereign Debt Litigation, 31 Yale J. on Reg. 189 (2014); Mark L. J. Wright, Sovereign Debt Restructuring: Problems and Prospects, 2 Harv. Bus. L. Rev. 153 (2012). Reactions followed from institutions such as the United Nations (U.N.) and the International Monetary Fund, among others. 12Int’l Monetary Fund, Sovereign Debt Restructuring—Recent Developments and Implications for the Fund’s Legal and Policy Framework 31 (2013), http://www.imf.org/external/np/pp/eng/2013/042613.pdf (expressing concerns about NML precedent); U.N. Conference on Trade and Dev., Argentina’s “Vulture Fund” Crisis Threatens Profound Consequences for International Financial System (2014), http://unctad.org/en/pages/newsdetails.aspx?OriginalVersionID=783.NML has also prompted feedback from industry organizations representing various constituencies in international financial markets. 13See, e.g., Brief for the American Bankers Association as Amicus Curiae Supporting Nonparty the Bank of New York Mellon at 14–15, NML Capital, Ltd. v. Republic of Argentina, 727 F.2d 230 (2d Cir. 2013) (Jan. 4, 2013) (No. 12-105-cv(L)); Brief for the Clearing House Association L.L.C. as Amicus Curiae in Support of Reversal at 25–27, NML Capital, Ltd. v. Republic of Argentina, 727 F.2d 230 (2d Cir. 2013) (Jan. 4, 2013) (No. 12-105-cv(L)); Int’l Capital Mkt. Ass’n, Standard Aggregated Collective Action Clauses (“CACs”) for the Terms and Conditions of Sovereign Notes Governed by English Law (May 2015), http://www.icmagroup.org/assets/documents/Resources/ICMA-Standard-CACs-Pari-Passu-and-Creditor-Engagement-Provisions—-May-2015.pdf.

Between scarce currency reserves and political sensitivities surrounding the holdout situation, Argentina’s government faces a delicate balancing act. 14See Charlie Devereux, Argentina’s Fix-It Man Is Slowing Down and Investors Are Worried, Bloomberg (Jan. 24, 2016, 9:00 PM), http://www.bloomberg.com/news/articles/2016-01-25/argentina-s-fix-it-man-is-slowing-down-and-investors-are-worried. Finding a reasonable value for settling the claims that the holdouts will accept and Argentina can afford—financially and politically—will not be easy. 15See Juan J. Cruces & Tim Samples, Time to Settle Sovereign Debt’s “Trial of the Century”?, Col. L. Sch.: CLS Blue Sky Blog (Jan. 26, 2016), http://clsbluesky.law.columbia.edu/2016/01/26/time-to-settle-sovereign-debts-trial-of-the-century/ (underscoring challenges for settlement negotiations); see also Charles Blitzer, Guest Post: Best Practices to Resolve Argentina’s Debt Dispute, Fin. Times: FT Alphaville (Jan. 22, 2016), http://ftalphaville.ft.com/2016/01/22/2150862/guest-post-best-practices-to-resolve-argentinas-debt-dispute/ (recommending strategies for the negotiations). Serious challenges remain, including structural questions about Argentina’s settlement offer and issues of inter-creditor equity among the holdouts. 16See Matt Levine, Argentina’s Bond Fight Comes Down to Its Worst Bonds, Bloomberg (Feb. 8, 2016), http://www.bloombergview.com/articles/2016-02-08/argentina-s-bond-fight-comes-down-to-its-worst-bonds [hereinafter Levine, Argentina’s Bond Fight]. Initial reactions to Argentina’s preliminary offer illustrate the complexity and gravity of inter-creditor issues for a potential settlement. 17Id. The role of injunctions in the NML litigation is still critical—even at the settlement stage. 18See Anna Gelpern, Love and Exhaustion in Argentina, Credit Slips (Feb. 11, 2016, 10:23 PM), http://www.creditslips.org/creditslips/2016/02/love-and-exhaustion-in-argentina.html [hereinafter Gelpern, Love and Exhaustion]; see also infra Part IV.C.

In this Article, we address economic and legal factors underlying Argentina’s holdout litigation with an emphasis on key issues for settlement negotiations. We contribute original financial data and legal analysis to the NML debate, which has critical implications for the broader world of sovereign finance. Specifically, our quantitative analysis illustrates holdout rates by bond, outstanding defaulted bonds by currency and applicable law, the current value of holdout claims compared to Argentina’s 2005 exchange offer, and returns for hypothetical holdout creditors under various investment scenarios. We close by evaluating legal and policy factors related to a potential NML settlement.

This Article is organized as follows. Part I reviews Argentina’s holdout litigation stemming from the 2001 default. Part II addresses the valuation of holdout claims, including an analysis of interest liabilities and an illustration of investment performance for hypothetical bondholders under different scenarios. Part III illustrates the current value of the 2005 offer depending on different allocations of the cash flows paid on Argentina’s exchange bonds and GDP-linked warrants between 2005 and 2015. Part IV considers legal and public interest factors at play in a potential NML settlement, including the role of ratable payment injunctions in the settlement phase of the NML litigation. We then offer concluding observations.

I. Argentina’s Holdout Litigation

This Part breaks down Argentina’s debt saga in detail across three key stages: default, restructuring, and litigation. This Part will first explain the dimensions of Argentina’s remarkable default and restructuring process, which set the stage for the extraordinary amount of holdout litigation facing Argentina. This Part then traces the evolution of the NML litigation in U.S. courts and provides a detailed bond-by-bond look at holdout claims.

A. Default and Restructuring

Following a devastating economic crisis, Argentina’s 2001 default was epic in both proportion and complexity. 19See A Victory by Default?, Economist (Mar. 3, 2005), http://www.economist.com/node/3715779. In a short but traumatic period, Argentina’s economy contracted dramatically as unemployment topped twenty percent and half of the population fell under the poverty line. 20See Republic of Argentina, Prospectus Supplement (To Prospectus Dated Dec. 27, 2004) at 11 (2005), http://www.sec.gov/Archives/edgar/data/914021/000095012305000302/y04567e424b5.htm [hereinafter Prospectus Supplement]. At that time, Argentina’s $135 billion default was the largest sovereign debt default to date. 21Sophie Arie & Andrew Cave, Argentina Makes Biggest Debt Default in History, Telegraph (Dec. 24, 2001), http://www.telegraph.co.uk/news/worldnews/southamerica/argentina/1366218/Argentina-makes-biggest-debt-default-in-history.html. The complexity of the default was also staggering with 150 different bonds, denominated in six currencies under the laws of eight different jurisdictions, and a highly fractured creditor base of over half a million bondholders. 22See A Victory by Default?, supra note 19.

But a record default was only the beginning. These factors paved the way for an extraordinarily difficult restructuring process. Sovereign debt markets exist in a legal and regulatory void. 23William W. Bratton & G. Mitu Gulati, Sovereign Debt Reform and the Best Interest of Creditors, 57 Vand. L. Rev. 1, 10–13 (2004) (underscoring gaps in sovereign debt system). Without a formal bankruptcy system, sovereign insolvency is resolved through voluntary restructuring. 24Jonathan I. Blackman & Rahul Mukhi, The Evolution of Sovereign Debt Litigation: Vultures, Alter Egos, and Other Legal Fauna, 73 L. & Contemp. Probs. 47, 48 (2010) (contrasting the voluntary nature of sovereign debt restructuring with the corporate bankruptcy process). Typically, defaulted bonds are swapped for new debt, or “exchange” bonds. 25Debt rescheduling and debt reduction are common methods for restructuring distressed or defaulted debt. See Ross P. Buckley, 8 Emerging Market Debt: An Analysis of the Secondary Market ch. 2 (1999). Growth-linked securities have also been featured in recent sovereign restructurings: Argentina (2005 and 2010), Greece (2012), and Ukraine (2015). See James Benford, et al., Bank of England, Sovereign GDP-linked bonds, Financial Stability Paper No. 39 (Sept. 21, 2016). For a discussion of Argentina’s GDP-linked warrants, see infra Part III.A. Creditor losses—the “haircut”—have a bearing on restructuring outcomes, such as creditor participation and holdout litigation. 26See Julian Schumacher et al., Sovereign Defaults in Court: The Rise in Creditor Litigation 1976-2010 22 (May 6, 2014) (unpublished manuscript) (on file at http://ssrn.com/abstract=2189997) (articulating factors that make sovereign debt litigation more likely) [hereinafter Schumacher et al., Sovereign Defaults in Court]. In Argentina’s case, a harsh haircut led to a protracted and remarkably combative restructuring process. 27Anna Gelpern, What Bond Markets Can Learn From Argentina, Int’l Fin. L. Rev., Apr. 2005, at 19 (describing the “unusually contentious” tone of the negotiations). At seventy-three percent, Argentina’s haircut was considerably higher than the average of thirty-seven percent for all sovereign restructurings from 1978 to 2010. 28See Juan J. Cruces & Christoph Trebesch, Sovereign Defaults: The Price of Haircuts, 5 Am. Econ. J.: Macroeconomics 85, 86 (2013), https://www.aeaweb.org/articles.php?doi=10.1257/mac.5.3.85 (finding that the average haircut in sovereign restructurings is thirty-seven percent); see also Juan José Cruces, ¿Cómo Resolver el Problema de los Holdouts y Bajar el Costo de Capital de la Economía Argentina?, 59 Fondo de Cultura Económica (forthcoming 2016) [hereinafter Cruces, Problema de los Holdouts] (refining the seventy-six percent haircut calculation of previous work by incorporating the present value of Argentina’s GDP-linked warrants and arriving at 73.4%).

With two bond exchanges—one in 2005 and another in 2010—Argentina’s restructuring process was also extraordinarily long. 29See Elena Duggar, Special Comment: The Role of Holdout Creditors and CACs in Sovereign Debt Restructurings, Moody’s Inv. Servs., Apr. 10, 2013, at 4. Even without taking the 2010 exchange into account, Argentina had an exceptionally long restructuring process. See id. The length of time before Argentina’s 2005 restructuring is over twice the average for sovereign debt restructurings. Id. Creditor participation was also unusually low. The 2005 exchange saw only seventy-six percent participation, but the second exchange in 2010 brought overall participation to about ninety-three percent with bonds worth about $6.03 billion holding out. 30See id. at 1. Even then, participation in Argentina’s restructuring remained relatively low. By comparison, between 1997 and 2013, the average participation rate in a sample of thirty-four sovereign debt restructurings was ninety-five percent. 31See id. at 8. High holdout rates spawned an extraordinary amount of litigation against Argentina, led by distressed debt hedge funds. 32Schumacher et al., Sovereign Defaults in Court, supra note 26, at 11 (explaining that post-2001 Argentina accounts for nearly one-third of all sovereign debt cases between 1976 and 2010). Eventually, after a string of major legal setbacks, Argentina’s holdout litigation led to a contested default in 2014—declared by major credit agencies but denied by the Argentine government—as payments sent by Argentina to exchange creditors were blocked by court injunctions. 33See Benedict Mander et al., Argentina: Unresolved Debts, Fin. Times (Aug. 6, 2014), http://www.ft.com/intl/cms/s/0/96b56394-1d68-11e4-b927-00144feabdc0.html#axzz3wfRLZpKM (describing “semantic disagreement” over whether or not a default had occurred).

B. The Evolution of NML Litigation, 2002-2014

Litigated for over a decade through different phases of Argentina’s debt crisis, the NML litigation has been exceptional across the board—in duration, volume, implications, and controversy. 34See Samples, supra note 11, at 63–75 (explaining the exceptional nature of NML litigation). Creditor claims were filed as early as 2002—long before NML finally reached a boiling point as injunctions led to Argentina’s 2014 default. 35Julian Schumacher, Sovereign Debt Litigation in Argentina: Implications of the Pari Passu Default, 1 J. Fin. Reg. 143, 144 (2015) [hereinafter Schumacher, Argentina Implications] (illustrating the trajectory of claims filed against Argentina since 2002). In the early stages of NML, courts were sympathetic to Argentina’s legitimate interest in restructuring, even supporting efforts to that end. 36Marcus Miller & Dania Thomas, Sovereign Debt Restructuring: The Judge, the Vultures and Creditor Rights, 30 World Econ. 1491, 1500 (2007) (describing the district court’s use of judicial discretion to promote debt restructuring in NML). In doing so, the Second Circuit prevented holdout claims from derailing restructuring efforts, citing the importance of debt restructuring for Argentina’s economy. 37EM Ltd. v. Republic of Argentina, 131 F. App’x 745, 747 (2d Cir. 2005). The Second Circuit cited concerns for “the economic health of a nation” in upholding the district court’s refusal to allow NML plaintiffs to block Argentina’s 2005 exchange. Id.

Following the 2005 exchange, the volume of claims filed against Argentina increased dramatically from eight hundred million dollars in 2004 to $3.5 billion by 2009. 38Our calculation of these amounts uses source data from Julian Schumacher, Enforcement in Sovereign Debt Markets (July 14, 2015) (unpublished Ph.D. dissertation, Humboldt-Universität zu Berlin) http://nbn-resolving.de/urn:nbn:de:kobv:11-100234666 [hereinafter Schumacher, Enforcement in Sovereign Debt Markets]. Focus turned to Argentine assets as plaintiffs began invoking alter ego arguments, but sovereign immunity thwarted these attempts. 39Alison Frankel, How Argentina Lost Game of Chicken with Renegade Bondholders, Reuters (Nov. 26, 2012), http://blogs.reuters.com/alison-frankel/2012/11/26/how-argentina-lost-game-of-chicken-with-renegade-bondholders/ (documenting the shift in strategy from asset attachment to equal footing claims). After years of litigation, holdout plaintiffs remained empty handed. But after years of defiance by Argentina, the court’s patience waned. 40See Anna Gelpern, Contract Hope and Sovereign Redemption, 8 Cap. Mkts. L.J. 132, 139–40 (2013) (noting the court’s exasperation with Argentina) [hereinafter Gelpern, Contract Hope]. Exasperation with Argentina’s defiance was clear at both the trial court 41EM Ltd. v. Republic of Argentina, 720 F. Supp. 2d 273, 304 (S.D.N.Y. 2010), vacated sub nom. NML Capital, Ltd. v. Banco Cent. de la Republica Argentina, 652 F.3d 172 (2d Cir. 2011) (“What is going on between the Republic of Argentina and the federal court system is an exercise of sheer willful defiance of the obligations of the Republic to honor the judgments of a federal court.”). and the Court of Appeals for the Second Circuit. 42NML Capital, Ltd. v. Banco Cent. de la Republica Arg., 652 F.3d 172, 196 (2d. Cir. 2011) (“We share the District Court’s understandable irritation at the Republic’s ‘willful defiance of [its] obligations to honor the judgments of a federal court.’”). Ultimately, frustration with Argentina’s non-compliance led the court to take drastic measures through injunctive relief. 43See infra notes 50–53.

Even for a plaintiff with a money judgment in hand, collecting against an unwilling sovereign is no easy task. 44Attaching valuable, non-immune sovereign assets is notoriously difficult. See Bratton, supra note 10, at 824 (“Sovereigns in default rarely leave valuables lying around subject to attachment in creditor-friendly jurisdictions.”); see also George K. Foster, Collecting from Sovereigns: The Current Legal Framework for Enforcing Arbitral Awards and Court Judgments Against States and Their Instrumentalities and Some Proposals for its Reform, 25 Ariz. J. Int’l & Comp. L. 665 (2008). As efforts to seize Argentina’s assets failed, holdout plaintiffs began invoking pari passu in claims against Argentina. 45See Schumacher, Argentina Implications, supra note 35, at 144 (illustrating the filing of NML claims invoking pari passu); see also Natalie A. Turchi, Note, Restructuring a Sovereign Bond Pari Passu Work-Around: Can Holdout Creditors Ever Have Equal Treatment?, 83 Fordham L. Rev. 2171, 2196–202 (2015) (documenting the evolution of pari passu under New York law). Often found in cross-border debt instruments, the meaning of the enigmatic pari passu or “equal step” clause is uncertain and highly contested in the sovereign debt context. 46Many prominent voices in sovereign debt have questioned—or have even rejected outright—the ratable payment interpretation of the pari passu clause. See, e.g., Lee C. Buchheit & Jeremiah Pam, The Pari Passu Clause in Sovereign Debt Instruments, 53 Emory L.J. 869, 877–91 (2004) (rejecting the ratable payment interpretation of pari passu); G. Mitu Gulati & Kenneth N. Klee, Sovereign Piracy, 56 Bus. Law. 635, 650 (2001) (making the case against the ratable payment interpretation of pari passu); Analysis of the Role, Use and Meaning of Pari Passu Clauses in Sovereign Debt Obligations as a Matter of English Law 79 Fin. Mkts. L. Comm., 17 (2005), http://www.fmlc.org/uploads/2/6/5/8/26584807/79.pdf. For arguments in favor of ratable payment interpretation, see Brief for Kenneth W. Dam as Amicus Curiae Supporting Respondents, NML Capital Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. 2012) (No. 12-105-cv(L)) (arguing in favor of third party injunctions also); Brief for Washington Legal Foundation as Amicus Curiae Supporting Respondents, NML Capital Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Oct. 26, 2012) (No. 12-105-cv(L)). Generally, the clause obligates the debtor to maintain the securities on equal footing or equal priority with other specified securities. Argentina’s pari passu clause reads:

The securities will constitute . . . direct, unconditional, unsecured and unsubordinated obligations of the Republic and shall at all times rank pari passu without any preference among themselves. The payment obligations of the Republic under the Securities shall at all times rank at least equally with all its other present and future unsecured and unsubordinated External Indebtedness. 47NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246, 251 (2d Cir. Oct. 26, 2012) (quoting the pari passu clause from Argentina’s Fiscal Agency Agreement).

The district court found that Argentina had violated its pari passu clause in (a) continuing payments to exchange bondholders without paying the holdouts and (b) enacting legislation that prohibited payments to holdouts. 48Order at *2–3, NML Capital, Ltd. v. Republic of Argentina, No. 08-6978, 2011 WL 9522565 (Dec. 7, 2011) [hereinafter 2011 NML Order]. Most importantly, the court’s interpretation of Argentina’s pari passu laid the foundation for broadly applicable ratable payment injunctions. 49See NML Capital, Ltd v. Republic of Argentina, No. 08-6978, 2012 WL 5895786, at *3 (S.D.N.Y. Nov. 21, 2012) (finding that the holdouts were entitled to one hundred percent of amounts owed by Argentina every time that Argentina pays one hundred percent of amounts owed to exchange bondholders). According to the NML ratable payment injunctions, before continuing to pay exchange creditors amounts due (coupon payments on the exchange bonds), Argentina had to make ratable payments to holdout plaintiffs (the full value of their claims). 50See 2011 NML Order, supra note 48 at *2. So the injunction forced Argentina to decide between paying the holdouts in full and defaulting on payment obligations to exchange creditors. 51Weidemaier & Gelpern, supra note 11, at 191 (“Put differently, the injunction allows Argentina to keep stiffing NML, but only if it also stiffs the exchange bondholders.”).

Even further, the injunctions were broadly applicable to third parties—including financial intermediaries—not just Argentina. 52Order, NML Capital, Ltd. v. Republic of Argentina, No. 1:09-CV-01707 (S.D.N.Y. filed Feb. 23, 2012). The scope of the injunctions included “all parties involved, directly or indirectly, in advising upon, preparing, processing, or facilitating any payment of the Exchange Bonds.” 53Id. Anticipating continued defiance by Argentina, the court aimed enforcement at innocent third parties who were more likely to comply with judicial orders. 54The court was acutely aware that Argentina would likely continue to defy its orders. See, e.g., Transcript of Hearing at 15, NML v. Argentina, 144 F. Supp. 3d 513 (Nos. 08-CV-6978 and 09-CV-1708) (S.D.N.Y. 2012). Though startling and controversial, this interpretation of pari passu was not completely unprecedented. 55In 2000, Elliott v. Peru broke new ground in pari passu litigation with ratable payment injunctions applicable to third parties. See Brief for the United States of America as Amicus Curiae in Support of Reversal, NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Apr. 4, 2012) (No. 12-105-cv(L)), 2012 WL 1150791. For criticisms of the “ratable payment” interpretation of pari passu abounded after the Brussels decision, see, e.g., Gulati & Klee, supra note 46; Rodrigo Olivares-Caminal, The Pari Passu Interpretation in the Elliott Case: a Brilliant Strategy but an Awful (Mid-Long Term) Outcome?, 40 Hofstra L. Rev. 39 (2011). Despite urging from the U.S. government, 56Brief for the United States of America as Amicus Curiae Supporting Argentina’s Petition for Panel Rehearing and Rehearing En Banc, NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Dec. 28, 2012) (No. 12-105-cv(L)) 2012 WL 5275014. the Second Circuit upheld the district court’s pari passu injunctions on appeal. 57See NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246, 250 (2d Cir. 2012). Finally, after the Supreme Court denied Argentina’s pari passu petition for review, the ratable payment injunctions came into effect, blocking Argentina’s scheduled payments to exchange bondholders and leading Argentina into another default in 2014. 58Republic of Argentina v. NML Capital, Ltd., 727 F.3d 230 (2d. Cir. Aug. 23, 2013), cert. denied, 134 S. Ct. 2819 (June 16, 2014) (No. 13-990); Exch. Bondholder Grp. v. NML Capital, Ltd., 727 F.3d 239 (2d Cir. Aug. 23, 2013), cert denied, 134 S. Ct. 2819 (June 16, 2014).

C. A Bond-by-Bond Look at NML Claims

Exhibit 1 below illustrates holdout rates after the 2005 and 2010 exchanges for each of the 150 defaulted bonds. The vertical bars illustrate the holdout rates on a bond-by-bond basis: the vertical gray bars show holdout rates in the 2005 exchange and vertical black bars show holdout rates after the 2010 exchange. The horizontal lines depict the weighted average holdout rate after each exchange: twenty-three percent and seven percent, respectively.

cruces-samples-fig1

Exhibit 1: 2005 and Post-2010 Holdout Rates by Bond

This figure shows the holdout rate for each defaulted bond in the 2005 exchange (vertical gray bars) and that remaning after the 2010 exchange (vertical black bars). The horizontal lines report the eligible-debt weighted average of holdout rates after the 2005 and the 2010 exchanges, twenty-three percent and seven percent, respectively. Seventy-one bonds had holdout rates greater than twenty percent in the 2005 exchange but only seven bonds surpass that mark after the 2010 exchange. These seven bonds also have the highest litigation rates of all bonds in the sample. Our analysis of returns from holding out, claim value, and current value of the 2005 offer focuses on these seven bonds.

The first takeaway from Exhibit 1 is that holdout rates vary significantly across bonds. The second takeaway is that after strong resistance to the 2005 exchange for a wide range of bonds, holdout rates tapered off dramatically in the second exchange. Post-2010 holdout rates are significant only in a handful of bonds. For example, seventy-one bonds had holdout rates greater than twenty percent in the 2005 exchange, but only seven bonds surpass that mark after the 2010 exchange. Holdout rates fall off steeply after those seven bonds, but non-trivial holdout rates are found in sixty-seven other bonds (with holdout rates of three percent or greater and an average holdout rate of 8.4%). 59The seven bonds with particularly high (above twenty percent) holdout rates plus the sixty-seven bonds with non-trivial holdout rates (above three percent) comprise the total of seventy-four bonds with meaningful holdout rates. The atomization of Argentina’s holdouts across so many debt instruments underscores the potential difficulty of creditor management and coordination in reaching a comprehensive final settlement. 60See infra Part IV.C (addressing the practical difficulties of coordinating a settlement across numerous debt instruments).

To keep our analysis tractable, we focus on the seven bonds with holdout rates higher than twenty percent after the 2010 exchange. 61The seven bonds, including their currency, coupon rate, ISIN codes, maturity dates and total principal still outstanding are (the number preceding each bond is the order in which each bond appears in Republic of Argentina 2004, Annex C): #41: Global bond, ARP ten percent 2001-2004 and twelve percent 2004-2008 due 2008, XS0130278467, Jun-2001, $595; #9: Global bond, USD 11.375% due 2017, US040114AR16, Jan-1997, $419; #17: Bond, USD variable rate due 2005 (FRAN), US040114AX83, Apr-1998, $298; #14: Global bond, USD 10.25% due 2030, US040114GB00, Jul-2000, $122; #7: Global bond, USD 12.375% due 2012, US040114GD65, Feb-2001, $113; #11: Global bond, USD twelve percent due 2020, US040114FB19, Feb-2000, $66; #48: Brady Par bond, EUR 5.87% due 2023, DE0004103007, Mar-1993, $58. The outstanding principal at the time of default of these seven bonds was $1.67 billion, which amounts to twenty-eight of the $6.03 billion of total remaining holdout bonds after the 2010 exchange. As explained below, claims on these holdout bonds have grown significantly from their initial face value. These seven bonds are also heavily litigated, making them an interesting sample for this paper.

Exhibit 2 below illustrates outstanding holdout principal by governing law. At the time of default, eighty-three percent of Argentina’s debt was under New York, German, or English law. But that figure has risen to ninety-five percent now. Moreover, the only substantial increase in concentration occurs for New York law. For the 2005 exchange, such bonds amounted to forty-five percent of the total, but now they amount to fifty-nine percent. 62For the sake of completeness, the bottom line of Exhibit 2 shows the principal amount of defaulted bonds outstanding at each point in time, in billions of dollars. So, arguably, the “run to the courthouse” could be considered the “run to the Southern District of New York” with regard to Argentina’s holdout litigation. 63See Schumacher et al., Sovereign Defaults in Court, supra note 26, at 11 (observing the runs to the courthouse that occurred after the debt crises of Argentina and Peru). In terms of currency, sixty percent of holdout debt is denominated in dollars and thirty-nine percent is denominated in euros. Those ratios have been quite stable since the 2001 default.

Exhibit 2: Holdout Bonds by Governing Law

cruces-samples-fig2

This table shows the breakdown by law of the bonds outstanding at different points in time. After the 2010 exchange, there is a significant concentration of bonds under New York law. Bonds under New York, German, and English law amount to about ninety-five percent of the outstanding capital. The bottom line reports the total principal outstanding in each year in billions of dollars.

II. The Value of Holdout Claims

This Part addresses the valuation of Argentina’s holdout claims for the seven bonds in the sample, beginning with the dramatic growth of Argentina’s pre-judgment interest liabilities. Interest alone—at somewhere between $2.6 and $5.3 billion depending on the judgment year—represents a significant portion of Argentina’s liabilities, about 1.6 to 3.2 times the initial value of the litigated debt. Next, this Part turns to the current value of holdout claims. In doing so, this Part illustrates returns on holdout investments by comparing the purchase price of the seven-bond basket under different hypothetical investment scenarios.

A. Pre-judgment Interest Under New York Law

Although NML was litigated in federal courts, New York’s statutory interest may apply when a federal court is deciding a matter of New York law. 64Federal rules are silent on pre-judgment interest but provide a floating post-judgment interest rate. 28 U.S.C. § 1961 (2012). Generally, pre-judgment interest applies to the award of a breach of contract from the time of the breach until a judgment is obtained. 65N.Y. C.P.L.R. 5001(b)–(c) (McKinney 1992). Courts understand the policy behind pre-judgment interest as making a plaintiff whole by recognizing the time value of money pending litigation. 66NML Capital v. Republic of Argentina, 621 F.3d 230, 240 (2d Cir. 2010). Courts have broad discretion in applying pre-judgment interest. 67See Michael S. Knoll & Jeffrey M. Colon, The Calculation of Prejudgment Interest 3, (U. of Pa. L. Sch., Public Law Working Paper No. 06-21, 2005) (citing Kansas v. Colorado, 533 U.S. 1 (2001) (“[L]awyers and their experts have wide latitude in persuading the court on just how much pre-judgment interest the defendant should pay to make the plaintiff whole.”)). In federal court cases, once a judgment is meaningfully ascertained, pre-judgment interest ceases to accrue and post-judgment interest begins to accrue at the substantially lower Treasury bill rate. 6828 U.S.C. § 1961 (2012).

Pre-judgment interest has two components: contract interest and interest on overdue interest. The latter is sometimes referred to as statutory interest because under New York law this rate is set by statute when a financial contract is silent on default rates of interest. 69NML Capital v. Republic of Argentina, 621 F.3d at 239 certified question accepted, 15 N.Y.3d 859 (2010), certified question answered, NML Capital v. Republic of Argentina, 952 N.E.2d 482 (2011) (“If the parties failed to include a provision in the contract addressing the interest rate that governs after principal is due or in the event of a breach, New York’s statutory rate will be applied as the default rate.”). Contract interest, at the rate provided in the bond contract, applies to principal whereas statutory interest applies to missed interest payments. 70NML Capital v. Republic of Argentina, 952 N.E.2d 482, 489–490 (“[I]t is undisputed that Argentina must pay interest on principal at the contract rate” while the “statutory interest on the unpaid interest payments compensates the bondholders for a different loss”). Another key determination stemming from the NML litigation is related to the phrase “until the principal hereof is paid or made available for payment” in Argentina’s bond documents. 71See id. This language renders maturity and acceleration irrelevant for the purposes of interest liabilities. As a result, Argentina’s pre-judgment interest liabilities continue accumulating until the court enters into a final judgment or a settlement occurs. 72See id. at 488–89. Given the extraordinarily long duration of Argentina’s debt litigation, this language makes a massive impact on interest rate liabilities. 73See infra Part II.B (quantifying Argentina’s interest liabilities).

New York’s statutory pre-judgment interest rate has been fixed at nine percent since 1981. 74David D. Siegel, Supplementary Practice Commentaries, in N.Y. C.P.L.R. § 5004 (McKinney 1992 & Supp. 2014). Previously, the rate was set at six percent from 1962 to 1967 and from 1972 to 1980. However, during the 1968-1971 interval, New York’s Banking Board was tasked with setting the rate between five percent and 7.5%. 75N.Y. C.P.L.R. § 5004. For a full discussion of New York’s prejudgment interest rates, see Laila Abou-Rahme & Stephen Scotch-Marmo, Is It Time for N.Y.’s Prejudgment Interest Rate to Float?, Special Rep. (N.Y. L.J.) Sept. 9, 2013, at S4. In 1981, the fixed rate was adjusted to nine percent to approximate the historically high inflation environment of the time, which was 8.9% during that year. Although contracts normally specify nominal interest rates, the true return obtained by an investor is the real (or inflation-adjusted) interest rate—as would be the case in a U.S. Treasury Inflation-Protected security, or TIP.

Since inflation varied greatly after 1961, realized ex-post real rates of return have been much lower and more volatile than the nominal rate. The average real rate from 1962 until 2001 was 3.1%, whereas it was 6.6% from 2002 until 2015—more than twice as high. Exhibit 3 below shows the frequency distribution of the real rate of interest since 1962. The horizontal axis shows different bin ranges for the real rate, while the vertical axis reports the number of years during which the real rate fell within the range indicated by each bin. The gray bars correspond to the forty years from 1962 until 2001, while the black bars correspond to the period from 2002 to 2015, which is the default range. Simple inspection reveals that the black distribution bar sits to the right of the gray bar, which means that real rates since Argentina’s default have been higher than the historical ones. 76We conduct a bilateral Wilcoxon-Mann-Whitney test of the null hypothesis that real statutory rates were equally distributed during the periods of 1962–2001 and 2002–2015 against the alternative that real rates after 2002 were different. See Paul Newbold, Statistics for Business and Economics 394 (4th ed. 2005). The p-value for the test statistic is .00007. We repeat the test, narrowing the first period to 1981–2001, and thus focusing on two subperiods that had a constant nominal nine percent statutory rate. The p-value in this case is .017. In both cases, the interpretation is that it is extremely unlikely that the two samples that we observe (the gray bars and the black bars) come from the same original distribution, or in plain English, that they had the same mean real rate. The statistical test thus supports the view that real statutory rate since Argentina’s default has been significantly higher than during previous periods.

Exhibit 3: Frequency Distribution of New York State Real Statutory Rate, 1962–2015

cruces-samples-fig3

This figure shows the frequency distribution of New York’s real statutory interest rate from 1962 to 2015. The gray bars correspond to the period until 2001 and the black bars correspond to the period that starts in 2002. The black bar distribution is clearly shifted to the right, a fact that is confirmed by a test of statistical significance. The average real rate for the first period was 3.1% per annum whereas it was 6.7% during the second period.

As the fixed nine percent rate has grown apart from the market, it has become more controversial. 77See Abou-Rahme & Scotch-Marmo, supra note 75, at 3. A real rate as high as we have observed since 2002 is more punitive than compensatory. 78See Advisory Comm. Civil Practice, Report to the Chief Administrative Judge of the Courts of the State of New York 120, (2013), [hereinafter Advisory Report], https://www.nycourts.gov/ip/judiciaryslegislative/pdfs/2013-CivilPractice-ADV-Report.pdf. Numerous bills have been introduced and advisory recommendations made for a floating rate, but the New York legislature has yet to respond to calls for change. 79See Abou-Rahme & Scotch-Marmo, supra note 75; see also Advisory Report, supra note 78, at 120 (finding the fixed nine percent rate “both illogical and unfair” and proposing a rate equivalent to a one-year Treasury bill plus three percent). Of the fifty states, thirty-seven have fixed pre-judgment rates and thirteen have floating rates. 80See Abou-Rahme & Scotch-Marmo, supra note 75. Of the states with fixed rates, eighteen have pre-judgment rates equal to or higher than New York’s nine percent rate. 81Id.

B. Argentina’s Extraordinary Interest Liabilities

Pre-judgment interests represent a substantial part of Argentina’s holdout liabilities. The high contract interest rates in Argentina’s bonds, in addition to New York’s nine percent prejudgment interest rate and the extremely lengthy period of the holdout disputes, led to an extraordinary accumulation of interest liabilities. Interest alone—at somewhere between $2.6 and $5.3 billion—represents about 1.6 to 3.2 times the initial value of the litigated debt. As a result, plaintiffs who obtained judgments earlier have lower claim values than those who obtained them later. 82The timing of judgments is so critical because, once the court enters into a final judgment, the claim is thereafter subject to interest at the federal post-judgment rate, which matches yields on U.S. Treasury bills. This post-judgment rate is drastically lower than contract interest plus the New York nine percent pre-judgment interest. See supra Part II.A (explaining the application of pre-judgment and post-judgment rates). Holdout claims for the seven bonds in our sample would total approximately $4.3 billion with judgments obtained in 2008, but could add up to $7 billion with judgments obtained in 2015.

To help assess the importance of the New York statutory rate in the case at hand, we compute the claim value of defaulted bonds at the end of 2015 under three alternative interest rates on overdue interest: the statutory nine percent rate, a nominal rate equal to the real New York statutory rate that prevailed from 1962 until 2001 plus 2015 inflation, 83This comes out to be 3.22% per annum. We use the realized inflation for the year ending in October 2015. Data were taken from the Federal Reserve web site. Selected Interest Rates (Daily), Bd. Governors Fed. Reserve Sys., http://www.federalreserve.gov/releases/h15/data.htm (last updated Oct. 11, 2016). and the one-year Treasury bill rate, which is used for post-judgment interest and, commonly, as a pre-judgment rate in federal courts deciding questions of federal law. 84Here we took the rate prevailing on the last week of the year in which the contractual interest was originally due. The average such rate comes out to be 3.03% per annum if judgment was handed down in 2008, and 1.54% per annum if it was handed down in 2015. Id.

We compute interest on overdue interest as follows: (a) it begins to accumulate at the end of the calendar year in which the contract interest was originally due, and (b) it applies directly, that is with no compounding. So if a bond had an annual coupon of $12, and the holder obtained a judgment in 2008, the interest on overdue interest for the coupon from year 2002 that was missed is $12 x 6 x interest rate. 85For bonds that had more than one coupon per year, we neglect this subtlety and assume that all coupons were due on the last day of the year. For the coupons missed in 2003, it would be $12 x 5 x interest rate on overdue interest, and so forth.

As noted above, when a final judgment is handed down, all amounts due are merged into a court decision and the judgment amount accrues the post-judgment interest rate until it is paid or a settlement is reached. This rate is the weekly average one-year constant maturity U.S. Treasury yield prevailing on the week before judgment is entered. 86Unlike pre-judgment interest, which is calculated on a simple basis, post-judgment interest is compounded. So, if it took two years from judgment to settlement, the judgment value is grossed up by (1+T bill rate) squared. Given the gap between the pre- and post-judgment interest rates, the current claim value of defaulted bonds critically depends on the time when judgment was entered. We compute two scenarios thereof: (i) claimants who obtained judgments in December 2008, and (ii) claimants who obtained their judgments in December 2015. Scenario (i) approximately corresponds to the weighted average filing date of claims against Argentina in New York, which is August 2006. 87See Schumacher, Enforcement in Sovereign Debt Markets, supra note 38, at 146–47. Scenario (ii) would correspond to “me-too” litigants who filed their claims around late 2013.

For the seven bonds at hand, since most of the held-out bonds have already been litigated, scenario (i) more closely approximates the claim values. However, we also compute scenario (ii) because it may better approximate the claim value of the other 119 bonds that still have holdouts—and many of those bonds have not been litigated yet. We compute this current claim value for each of the seven bonds and then aggregate it in the value of the overall portfolio.

Exhibit 4 below presents the results. The first row just below the column headings reports the principal outstanding of the seven-bond portfolio, which is $1.67 billion. The next row reports the overdue contract interest, which is $1.9 billion in scenario (i) and $3.21 billion in scenario (ii). The following row shows the interest on overdue interest under each of the three rates discussed above. The shaded columns report the benchmark scenario using the nine percent New York statutory rate. The next row of the table reports the judgment amount, which would be the total pre-judgment interest plus the accelerated capital. Since neither payment nor settlement has occurred, the subsequent row reports the post-judgment interest accrued until the end of 2015. The first shaded row reports the total claim value at the end of 2015.

Exhibit 4: Value in 2015 of Holdout Portfolio Claims and of the 2005 Exchange Offer in Different Scenarios

cruces-samples-fig4

Thus, in the benchmark scenario, if all holdouts had obtained their judgments in 2008, the original $1.67 billion would have grown to $4.27 billion, while if they all had obtained their judgments in 2015, their claims would be worth seven billion dollars. The two alternative assumptions about interest on overdue interest give similar results with a reduction in total claim value of around four hundred million dollars for the 2008 judgment scenario, and a reduction of almost the full amount of the original outstanding capital for the 2015 judgment scenario. With the alternative approach to overdue interest, reductions in total claim value range from ten percent (2008 judgment) to about twenty-one percent (2015 judgment). The two lines of Exhibit 4 in between the shaded horizontal lines depict these savings. The bottom line is that the claim value of the portfolio is between 2.56 and 4.19 times the principal owed initially—as shown in the bottom shaded row. 88The bottom four lines in Exhibit 4 compare the current value of the 2005 offer with the claim value in each scenario and are discussed fully in the text of Part III. The sheer amount of the multiplication for late litigants reinforces the importance of dealing with bondholders who have not litigated yet—an issue whose full treatment exceeds the scope of this paper. As discussed in Part III.A, given the extraordinary length of Argentina’s holdout litigation, the historically high statutory pre-judgment rate is critically important. 89Argentina’s bond documents did not specify a default rate. See NML Capital v. Republic of Argentina, 952 N.E.2d 482 (2011). Going forward, sovereigns issuing debt under New York law may consider setting a reasonable default rate in their bond contracts to avoid excessive pre-judgment interest liabilities. See Lee C. Buchheit, How to Negotiate Eurocurrency Loan Agreements 36–40 (Richard Forster et. al. eds., 2d ed. 2000) (discussing default interest clauses in loan agreements). Argentina is now paying the price—a costly instance of boilerplate contracting. 90See generally Mitu Gulati & Robert E. Scott, The 3 1⁄2 Minute Transaction: Boilerplate and the Limits of Contract Design (2012).

C. Hypothetical Bondholder Returns

Though perhaps lacking direct legal consequence for a breach of contract dispute between sophisticated parties, the overall fairness or legitimacy will figure prominently in potential settlement negotiations between Argentina and the holdouts. Likewise, specific components of investor returns such as pre-judgment interest will likely be the subject of scrutiny. Additionally, the outcome of the NML negotiations carries broader implications for sovereign finance. 91For examples of scholarship discussing these implications see supra notes 11–13 and accompanying text. Against this backdrop, we analyze the returns that holdout litigants could have obtained by purchasing bonds at different points in time, now standing to recover the claim value documented in Exhibit 4.

We then present the results of this exercise for a portfolio of the seven bonds in the sample weighted by outstanding principal of each bond at the end of 2001. Specifically, these calculations illustrate returns from purchasing bonds in a given year, holding out from the 2005 and 2010 exchanges, and litigating.

Exhibit 5: Returns from Purchasing Defaulted Bonds and Holding Out

cruces-samples-fig5

This table reports the returns for holdouts who purchased our seven-bond basket in the secondary market after Argentina’s default and collected the claim value at the end of 2015. We compute such returns under three hypothetical rates of interest on overdue interest, and for two judgment dates. Purchase prices and claim values are expressed per $100 of principal outstanding of each bond. We report two measures of return: compound annual average returns, and cumulative wealth from investing one dollar. In the benchmark case (shaded columns), investors in defaulted bonds multiplied their wealth an average of between eight and thirteen times depending on when judgment was handed down. See Part II.C for details.

The first column in Exhibit 5 shows the year during which the bond portfolio was purchased, ranging from 2002 until 2013. The second column reports the purchase price of the basket of bonds. 92For details on the sources of the price construction, see Appendix. One difference between these figures and those in Exhibit 4 is that here we report bond prices and claim values per one hundred dollars of outstanding principal, whereas Exhibit 4 uses the aggregate outstanding amount of the seven bonds and their claim value. So, as noted in the first two columns of the table, the bond basket cost $25.63 in 2002, its price hovered around the thirty dollar range for most of the decade, rose to about forty dollars in 2011, went back to twenty-eight dollars in 2012, and then up to $41.71 in 2013.

The first row in the table reports the claim value at the end of 2015 under the three different assumptions about interest on overdue interest discussed in the previous section. These figures are consistent with the data in Exhibit 4: for example, the claim value under New York’s nine percent statutory interest for litigants who obtained a judgment in 2008 is $4.27 billion for a portfolio that had an initial outstanding amount of $1.67 billion at the end of 2001. Hence, for each one hundred dollars of initial principal, the claim value is $100 x 4.27 / 1.67 = $255.63, which is the figure appearing in the corresponding cell of Exhibit 5.

Finally, for each combination of purchase year and interest on overdue interest, the table shows two measures of return on holding out: average compounded annual return (in percentage points) and total accumulated wealth from having originally invested one dollar. For example, someone who purchased the bond basket in 2012 for $28.41 and litigated in 2014 would be entitled to collect $419.19 under New York’s nine percent statutory rate. This implies an average return of 115.8% per annum. Exhibit 5 shows that this investor would have multiplied her money by almost fifteen times in these three and a half years. 93Since purchase prices are annual averages, we measure the time elapsed from purchase to final claim value, starting from the middle of the purchase year until December 2015. Hence, the basket of bonds bought in mid-2012 was held during 3.5 years: $1 x (1+1.158)3.5 = $14.76. In fact, the last column of Exhibit 5 shows that, from 2002 until 2013, all investors in Argentina’s seven most held-out bonds multiplied their wealth at least tenfold, with an average gross return surpassing thirteen times the initial investment.

III. The Current Value of the 2005 Restructuring Offer

As previously noted, Argentina’s present value haircut was high compared to the international historical record. 94See supra Part I.A. However, that present value haircut reflects the market’s valuation of the exchange bond and GDP-linked warrant basket as of June 2005. That measure is useful to understand the high holdout rates for the 2005 exchange and the wave of litigation that occurred thereafter. However, the exchange bonds have performed very well thus far. Argentina’s GDP-linked warrants, in particular, have provided their holders with phenomenal returns. This Part illustrates returns on participation in Argentina’s 2005 restructuring and calculates the current value of Argentina’s 2005 exchange offer under two investment scenarios. This Part then closes with a discussion of the haircut that would be taken by holdouts in a hypothetical settlement offer that pays the current value of Argentina’s 2005 exchange offer.

A. Returns on Participation

Compared to the historical record of sovereign debt restructurings, the creditor haircut in Argentina’s 2005 exchange was high. 95See supra note 28 and accompanying text. The harsh present value haircut explains why holdout rates in the 2005 exchange were so high and why litigation mushroomed. 96See Schumacher et al., Sovereign Defaults in Court, supra note 26, at 22 (articulating conditions that make sovereign debt litigation more likely). In Argentina’s 2005 and 2010 exchanges, a GDP-linked warrant was offered as a “sweetener” to entice creditor participation. 97Stephen Park & Tim Samples, Ukraine’s Quietly Revolutionary Debt Restructuring, Fin. Times; beyondbrics, 12 (Sept. 17, 2015), http://blogs.ft.com/beyond-brics/2015/09/17/ukraines-quietly-revolutionary-debt-restructuring/ (analyzing the role of GDP-linked “value recovery instruments” in Ukraine’s 2015 restructuring). These securities, which are detachable and tradable independently from the exchange bonds, provide payments linked to GDP growth. 98For a detailed analysis of the terms of Argentina’s GDP-linked warrants, see Stephen Kim Park & Tim R Samples, Towards Sovereign Equity, 21 Stan. J.L. Bus. & Fin. (forthcoming 2016), http://ssrn.com/abstract=2630772. These GDP-linked warrants have performed phenomenally since the 2005 exchange. 99See infra notes 102–04.

Comparing the ex-post realized returns of different investments with Argentina’s 2005 exchange bonds and GDP-linked warrants provides some perspective. Exhibit 6 below illustrates returns on an investment of one U.S. dollar in various securities when Argentina’s first exchange settled on June 2, 2015. Our calculations assume that all dividends and coupon payments were reinvested in the original security that paid them. It is important to assess the return on Argentine exchange bonds against the broader canvass of other well-known assets. The table shows that the accumulated wealth from investing one dollar in 2005 in U.S. Treasuries, the S&P 500, Argentina’s stock market indices as well as Chilean, Uruguayan, and Brazilian government bonds would range between $1.68 and $2.03. 100These figures are comparable to the Wealth columns of Exhibit 5. As a reference, for investors who bought defaulted bonds in 2005, the accumulated wealth from collecting the claim value under NY’s nine percent rate would range between $7.63 and $12.51 depending on when they obtained judgment.

Exhibit 6: Total Return on Various Securities Since Argentina’s 2005 Debt Exchange

cruces-samples-fig6

This table shows the wealth that an investor would have at the end of 2015 if she had invested one dollar in different assets at the time of Argentina’s 2005 exchange and had reinvested all interim cash-flows (dividends or coupons) paid by each holding in that same security. All figures are in U.S. dollars. The purchase date for all assets is June 2, 2005, (or first observed price thereafter) except for the Apple stock and the GDP warrants. The warrants began trading by themselves on November 24, 2005, so we use that purchase date for the securities marked with an asterisk. Argentine exchange bonds are a simple average of pars and discounts, and bonds in hard currency are a simple average of those in dollars and those in euros. Returns on government bonds from Brazil, Chile, and Uruguay are from JP Morgan’s EMBI Global index. The returns on U.S. Treasuries are measured via the IEF iShares exchange-traded fund. The table vividly shows that Argentine exchange instruments have outperformed similar assets. The lavish return on the GDP warrants in hard currency even surpasses that of the Apple stock.

At $3.38 per dollar invested, the 2005 exchange bonds denominated in dollars and euros have performed very well. 101To circumvent the problem that exchange bonds under New York law have been in default since 2014, we use the prices of their Argentine-law U.S. dollar-denominated counterparts. Only the jurisdiction differed among these bonds, not the promised payments. Argentina has been making payments on the local law bonds despite the injunctions that came into effect in 2014. But at $16.17 per dollar invested, Argentina’s GDP-linked warrants have provided astronomical returns. 102For an early assessment of the high realized ex-post returns on the exchange bonds and GDP-linked warrants, see Brief for Alfonso Prat-Gay as Amicus Curiae in Support of the Republic of Argentina’s Petition for Panel Rehearing, NML Capital, Ltd. v. Republic of Argentina, No. 12-105-cv(L) (2d Cir. Jan. 4, 2013). Therefore, for participating bondholders who held on to their exchange bonds, the overall ex-post performance has been much less painful than initially expected. For example, the GDP-linked warrants provided windfall gains but were given virtually no value in the 2005 exchange. 103Stephany Griffith-Jones & Krishnan Sharma, GDP-Indexed Bonds: Making It Happen, in Innovative Financing for Development 85 (Suhas Ketkar & Dilip Ratha eds., 2009); Argentine GDP Warrants, Euromoney (Jan. 25, 2006), http://www.euromoney.com/Article/1014876/Argentine-GDP-warrants.html. As a result, the exchange has actually been very costly for the Argentine government. 104Guido Sandleris & Mark L. J. Wright, GDP-Indexed Bonds: A Tool to Reduce Macroeconomic Risk?, in The Future of Sovereign Borrowing in Europe 108 (Morten Balling, et al. eds., 2013). A harsh present value haircut hindered creditor participation and spawned high rates of litigation. 105See supra Part I. Unfortunately, after imposing a drastic haircut, Argentina actually ended up paying out a great deal on the exchange instruments. In a way, Argentina neither has its cake, nor ate it.

As will be discussed in Part IV, settling the holdout claims will involve a haircut on the legal claims. 106See Cruces & Trebesch, supra note 28, at 90. If Argentina chooses to pay with a new bond, the market value of that bond compared to the claim value will determine the settlement haircut. Id. Analyzing the bonds potentially offered in an NML settlement is beyond the scope of this Article. However, given the history of declining interest rate spreads as countries emerge from default, issuing bonds with call rights might merit consideration. Many of the Brady bonds, in restructurings sponsored by the U.S. government, provided debtors with call rights. Juan J. Cruces & Nicolás Merener, Holdouts: pagar con bonos precancelables nos puedeahorrar u$s 2000 millones, casi la mitad de lo que salió YPF, El Cronista (July 21, 2014), http://www.cronista.com/columnistas/Holdouts-pagar-con-bonos-precancelables-nos-puedeahorrar-us-2000-millones-casi-la-mitad-de-lo-que-salio-YPF-20140721-0029.html. When thinking about a haircut, the current value of the 2005 exchange offer is compelling and interesting for two reasons. First, it reflects an interest in respecting a principle of inter-creditor equity vis-á-vis exchange bondholders who accepted the 2005 offer. Second, it allows a simple benchmarking of whatever settlement offer is ultimately made compared to the 2005 restructuring, which had a high degree of support in Argentine society.

B. The Holdout Trust

The subsequent analysis assumes that Argentina’s government issued exchange bonds and GDP-linked warrants in 2005 for our seven-bond portfolio on the same terms as the average bondholder participating in the exchange. Furthermore, we assume that these exchange bonds were put in a trust account held on behalf of bondholders and kept for them until the end of 2015. 107This move would have violated the Padlock Law (or Ley Cerrojo 26,017), which prohibited the Argentine government from making an exchange offer to holdout creditors after the 2005 exchange, including of course giving bonds in their favor to a trustee. We abstract from this fact here. See Law No. 26017, art. 2, Feb. 11, 2005 [CXIII] B.O. 30590 (Arg.). We assume that each time that the Argentine government made payments to the exchange bondholders, the holdout trust received ratable payments as well.

In order to carry out the 2005 restructuring, Argentina defined an “eligible” amount for each defaulted bond. This amount equaled the principal outstanding at the time of default plus the accrued and unpaid interest up to and including December 31, 2001. 108Including these unpaid interests was done at the request of—among others—Argentina’s Bond Restructuring Agency, a bondholder group. Our seven-bond portfolio, which had an outstanding principal of $1.67 billion as mentioned above, would have had an exchange-eligible amount of $1.71 billion.

Exhibit 7 below shows the face value amounts of bonds and warrants in each currency that would have been given to the trustee. These values are computed in proportion to the total amount of new bonds issued in the 2005 exchange relative to the total eligible value of old bonds tendered in it. In other words, the trust would receive the same basket of bonds that the average participating bondholder obtained in the 2005 exchange for each dollar of eligible old debt. As the table shows, the trust would have GDP-linked warrants for $1.71 billion (823 million denominated in pesos, 473 million denominated in dollars, and 416 million denominated in euros). 109All figures in our calculations are in U.S. dollars using the official exchange rates noted in Argentina’s Prospectus Supplement. See Prospectus Supplement, supra note 20. Moreover, the trust would have received a total $968 million face value of new bonds: 431 million dollars of bonds denominated in pesos, 278 million in dollars, and 259 million dollars of bonds denominated in euros. The table also shows the breakdown of discount, par, and quasi-par by currency of denomination of the new bonds. 110For tractability, and given the extremely low liquidity of Japanese yen exchange bonds, we assumed that the 0.7% corresponding to new bonds issued in that currency in 2005 were actually denominated in euros.

Exhibit 7: Face Value of the New Bonds Issued for the Benefit of the Seven Held-Out Bond Portfolio

cruces-samples-fig7

This table reports the face value amount of each new security that the holders of our seven-bond portfolio would have obtained in the 2005 exchange if they were to get the same basket of new securities than the other tendering bondholders. All values are in millions of U.S. dollars, using the official exchange rates for the 2005 exchange.

C. Reinvestment of Intermediate Cash Flows

One critical question that the trustee would have had to address is how to deal with the cash paid by the Republic to the trust over the years. There are two sources of cash that the trust would have received. First is the initial cash that was given at the time of the exchange to pay bondholders for the interest on the new bonds that accrued from December 31, 2003, which was the issue date of the new bonds, until June 2, 2005, which was the exchange settlement date. 111Arturo C. Porzecanski, From Rogue Creditord to Rouge Debtors: Implications of Argentina's Default, 6 Chi. J. Int'l L. 311, 324 (2005). Argentina never recognized the interest on the old bonds that accrued during 2002 and 2003. Id. Second, and more important, are the coupons that were paid over time on the new bonds and on the GDP-linked warrants.

We make two assumptions as to the allocation of these interim cash flows. In the first scenario, we assume that they were used to purchase fractional units of the same mother security that paid those cash flows. To this end, we used the closing price of the new securities on the ex-coupon date, at each point in time from 2005 until 2015. Thus, every time that Argentina paid a service on the new securities, we are assuming that the trustee went to the market and bought more units of the same security at the market price prevailing at that time. 112The secondary market prices were taken from Bloomberg and Thomson Reuters-Datastream. In the second scenario, we assume that the trustee invested all interim cash flows at the six-month U.S. Treasury bill rate and rolled over the accumulated cash position every semester. 113For bonds issued in pesos and euros, we convert all interim cash flows to U.S. dollars at the free market rate at the time that they were paid and maintain that position in dollars throughout the sample.

D. Results

Exhibit 8 below shows the results. To facilitate comparison with other work in the literature and with the common market practice, this figure shows the claim value and the current value of the exchange offer per hundred dollars of original debt outstanding. The two top lines (both gray) show the value of holdout claims for the seven-bond portfolio at the end of each year from 2001 until 2015. The dashed gray line reflects the value for the holdouts that obtained judgments in 2008 and have been accruing the (low) post-judgment rate since then. The dotted gray line reports the value of the claims for the holdouts that litigated in 2013 and obtained judgments at the end of 2015, and so continued to accrue the (high) pre-judgment rate until the end of the sample. In accord with the figures in Exhibit 5, for each hundred dollars of original principal outstanding, the total claim value of the seven-bond portfolio at the end of 2015 is $256 for claimants whose judgments are dated 2008, and $419 for those whose judgments are dated 2015. 114These figures are also consistent with Exhibit 4. For example, $100 x 7.01 / 1.67 = $419.76. The difference in values shows the very steep growth of Argentina’s liabilities due to pre-judgment interests discussed above.

Exhibit 8: Value of the 2005 Exchange Offer and of Holdout Claims

cruces-samples-fig8

This figure shows the evolution over time of the claim values (top lines) and of the restructuring offer made by Argentina in 2005 (bottom lines). All figures are expressed per $100 of principal outstanding at the end of 2001, which was the time of Argentina’s default, and refer to the seven-bond portfolio of the most held-out bonds analyzed in this paper. The dashed gray line reports the claim value of a portfolio of the seven most held-out bonds assuming that litigants obtained a judgment at the end of 2008. The dotted gray line reports the same value but assumes that judgment was handed down in 2015, which is a typical situation for the “me-too” plaintiffs. The solid black line reports the value of the 2005 offer assuming that interim cash flows were reinvested in the same mother security that paid those cash flows. The dashed black line reflects the same value but assumes that interim cash flows were invested at the six-month U.S. Treasury bill rate and rolled over until the end of the sample. At the end of the sample, the value of the 2005 offer with reinvestment in the same mother security strategy amounts to fifty-two percent of the claim value for holdouts that obtained judgments in 2008.

As noted above, the majority of the holdouts of the seven bonds in our sample have litigated and obtained money judgments for their claims. Hence, the gray dashed line better represents the value of their claims for the specific case of these seven bonds. However, we provide the dotted gray line because it reflects the value of the “me-toos” who have litigated only recently or have not litigated at all—a situation that may be more representative of the other 119 bonds that still have holdouts. The material difference between these two lines again underscores the difficulty of settling with a broad array of holdouts with significant variation in their claims. 115See infra note 152 and accompanying text.

The two bottom curves show the value of the trust just described at the end of each quarter since the 2005 exchange. The solid black line reports the investment-in-same security strategy while the dashed black line depicts the investment-in-U.S. Treasury bills strategy. Since values are expressed per one hundred dollars of original outstanding principal of old bonds, the trust for the holders of the seven-bond portfolio would have obtained a basket of new bonds that had a market value of thirty-seven dollars at the time of the exchange. 116A cursory reading of these figures would suggest that the haircut was sixty-three percent, which differs from the 73.4% figure mentioned for Argentina’s 2005 exchange. See Cruces & Trebesch, supra note 28. Note that the former figure uses the face value of the old debt, while the latter one uses the present value of the old debt. See id. at 88–89 (explaining the difference between these two haircut concepts). It may come as a surprise that the value of the trust sometimes rises but other times falls. This is because such value uses the market price of the trust’s securities at the end of each quarter. Thus, in times like the 2008 crisis, when the prices of risky assets fell worldwide, so too did the value of the trust, regardless of how interim cash flows were allocated. 117To sum the value of the holdings in different currencies, we convert the value of the securities denominated in pesos and euros to dollars using the free market exchange rates prevailing at the end of each quarter.

In spite of these cyclicalities, the value of the trust displays a secular rise in value. By the end of 2015, the holdings of the trust fund are worth $133 under the reinvest-in-same security strategy (reinvestment strategy) and ninety-nine dollars under the invest-in-Treasury bill strategy (T-bill strategy). 118The end of sample prices correspond to November 24, 2015. The thirty-five percent gap among the results of the two strategies is notable, in part because Argentina paid the same amount of money under both options at each point in time. The difference poses an intriguing question, which, because of its sheer magnitude, is an important one: as of December 2015, how much did the 2005 exchange really cost Argentina? We now address this question.

The first option has a higher value because the interim cash flows were reinvested on very favorable terms. This is because the price of Argentine bonds and warrants over time has been low relative to its current value and also relative to the services that they have paid. It was therefore much more profitable to reinvest interim cash flows in the same security that paid them than to park those cash flows at the Treasury bill rate. In a way, the reinvestment strategy reflects the joint effect of what Argentina has been paying combined with unduly pessimistic ex-ante expectations that the market has had about Argentina since the 2005 offer compared to the ex-post reality. Nevertheless, the solid black line does reflect the current value of the 2005 offer in the following way: if, at the time that each cash-flow service came due after issuance and until the end of the sample, Argentina did not pay it in cash (to the trust fund caring for the holdouts) but rather paid it by issuing new quantities of the instrument whose coupon came due, then the solid black line exactly reflects the value that holder would now have, and the value that the 2005 offer really cost Argentina expressed in money at the end of 2015. We conclude that the reinvestment strategy better approximates the cost of the 2005 offer to Argentina expressed at the end of the sample.

E. Haircut of a Hypothetical Offer

Here we explore the implications of a hypothetical settlement offer resembling the value of the 2005 exchange. We then compare that hypothetical offer with the current value of holdout claims for our seven-bond portfolio. In doing so, we show below that the final haircut in this hypothetical settlement could be as low as forty-eight percent, even relative to the full benchmark claim value.

The bottom of Exhibit 4 shows these values but, instead of expressing them per one hundred dollars of initial outstanding principal, it expresses them for the full outstanding principal of the basket, which is $1.67 as noted before. As shown therein, the current value of the 2005 offer, expressed in these latter units, is $2.23 billion in the reinvestment strategy and $1.65 billion in the T-bill strategy. Compared to the current value of the claim for litigants who obtained judgments in 2008, they would amount to forty-eight percent and sixty-one percent haircuts under New York’s nine percent statutory rate. If we used a more lenient interest on overdue interest, like the Treasury bill rate that is used in federal courts, the haircut would be forty-two percent and fifty-seven percent, respectively.

These haircut figures fall within the range of the historical record. In fact, the average haircut in the twenty sovereign debt restructurings involving non-HIPC countries after the Brady plan (and excluding Argentina in 2005) is exactly forty-two percent. 119See Cruces, Problema de los Holdouts, supra note 28, at 69. In a nutshell, the haircut that would stem from an exchange offer in which Argentina gave holdouts the current value of the 2005 offer would be within the ballpark of sovereign debt restructuring experiences since the end of the Brady Plan. The vast majority of these restructurings did not lead to significant holdout litigation. 120See infra note 155 and accompanying text.

In analyzing these haircut calculations, it is important to bear in mind that they underestimate the true haircut that claimants will take upon the settlement of the suit if they hold on to new Argentine bonds. This is due to the fact that, by reducing uncertainty, a settlement will create value. Specifically, a settlement will mean that—all else equal—the yield on Argentine government securities will fall, propping up their market price, and hence reducing the haircut. This will happen both to the basket of 2005 exchange securities, if Argentina offered holdouts the “actual” 2005 exchange bonds and GDP-linked warrants that would be in the hypothetical trust fund, or to any new bonds that Argentina might issue to pay a settlement. In this sense, it is worth noting that by agreeing to a given haircut, the holdouts will automatically reduce their own pain, which strengthens settlement incentives among holdouts.

IV. Towards a Reasonable Settlement

Settling a large number of holdout claims at the post-trial stage, the situation facing Argentina in some ways resembles a typical sovereign restructuring. Argentina’s holdout settlement will likely involve a creditor haircut and another debt issuance. Less certain is how Argentina will achieve—in strategic and practical terms—the goal of settling claims with a fractured and diverse group of creditors. 121See supra Part I.C. Indeed, concerns surrounding inter-creditor equity and holdout participation flared up almost immediately after Argentina’s initial settlement offer. 122See Levine, Argentina’s Bond Fight, supra note 16; see also Charles Blitzer, Guest Post: Argentina’s Debt Offer—Don’t Pop the Champagne Just Yet, Fin. Times: FT Alphaville (Feb. 12, 2016), http://ftalphaville.ft.com/2016/02/12/2153239/guest-post-argentinas-debt-offer-dont-pop-the-champagne-just-yet/ (describing dramatic differences in haircuts among creditors in Argentina’s initial settlement offer). Many of Argentina’s holdout creditors—the “me toos”—have not even filed claims yet. This Part will first address challenges facing courts tasked with adjudicating sovereign debt disputes and policy factors at play in Argentina’s holdout situation. Then, this Part considers the role of ratable payment injunctions in potential NML settlement negotiations, ultimately concluding that the injunctions should be modified or lifted to facilitate the settlement process.

A. Adjudicating Sovereign Debt Disputes

Adjudicating sovereign debt disputes is a complicated task. The legal system—or lack thereof, rather—for sovereign debt is patchy and awkward. 123See Bratton & Gulati, supra note 23, at 10–13 (contrasting corporate and sovereign debt). For one, sovereign debt obligations are simultaneously “unenforceable-yet-nondischargeable,” which creates complex pressures for courts and disputing parties alike. 124See Anna Gelpern, A Skeptic’s Case for Sovereign Bankruptcy, 50 Hous. L. Rev. 1095, 1098 (2013). Furthermore, sovereign debt markets exist in a legal void, lacking a direct regulatory or institutional authority. 125See id.; see also supra notes 23–24 and accompanying text. As a result, in sovereign debt litigation, courts of general jurisdiction are called upon to adjudicate contractual disputes involving complicated insolvency situations better suited for a bankruptcy system. 126See Pottow, supra note 11, at 227 (“Judges are nevertheless asked to make important policy decisions in one-off interventions that occur every few years, a task to which they are poorly suited.”); see also Gelpern, Contract Hope, supra note 40, at 133 (describing sovereign debt litigation as a source of “hard cases prone to make bad law”). Without the bankruptcy toolkit, courts are left with blunt mechanisms for intricate situations. 127Sovereign debt lacks the crucial restructuring tools of a formal insolvency system: automatic stays, cram-downs, debtor-in-possession financing, clear priority rules, etc. See Lee C. Buchheit & G. Mitu Gulati, Sovereign Bonds and the Collective Will, 51 Emory L.J. 1317, 1347 (2002) (explaining collective action dilemmas due to the lack of cram-down mechanisms in sovereign debt workouts). Together, these legal vacuums in the status quo sovereign debt system make for unpredictable and dysfunctional results. 128See Gelpern, Contract Hope, supra note 40, at 134 (predicting dysfunctional results in NML litigation); see also Park & Samples, supra note 98 (explaining dysfunctional outcomes and “rogue” trends in sovereign debt generally). Exacerbating these problems, creditor fragmentation in sovereign debt markets has complicated coordination and collective action problems. 129See W. Mark C. Weidemaier & Mitu Gulati, A People’s History of Collective Action Clauses, 54 Va. J. Int’l L. 52, 56 (2013). Sovereign finance transitioned from syndicated loans to bonds with the Brady Plan. See id. Increasingly numerous and divergent creditors have created a vastly more complicated landscape for sovereign debt restructurings. See id. “Bondholders, by contrast, are widely dispersed, may have divergent interests, and are less subject to regulatory pressure.” Id. at 56. For a contrasting view, see Ran Bi, Marcos Chamon & Jeromin Zettelmeyer, The Problem That Wasn’t: Coordination Failures in Sovereign Debt Restructurings (Int’l Monetary Fund, Working Paper No. 11/265, 2011), https://www.imf.org/external/pubs/ft/wp/2011/wp11265.pdf. These problems have been most visible in the restructuring phase of sovereign debt, but NML demonstrates that creditor coordination problems can exist even in the settlement phase. 130See supra notes 16–18.

Without a formal bankruptcy system, sovereigns have relied on the limited enforceability of debt contracts to encourage participation in restructurings. 131See Gelpern, Contract Hope, supra note 40, at 133. Enforcing judgments against an unwilling sovereign remains a highly uncertain and expensive venture. 132Weidemaier & Gelpern, supra note 11, at 190 (“Courts can inconvenience sovereigns; they cannot make them pay.”); Foster, supra note 44, at 670. These risks incentivize creditor participation in restructurings, even when sovereigns lack the threat of bankruptcy. Even though participation is voluntary, holdout rates in sovereign debt restructurings are generally quite low. 133Between 1997 and 2013, the average of creditor participation in sovereign restructurings was approximately ninety-five percent. See Duggar, supra note 29, at 1. Out of thirty-four debt exchanges, all but two—Dominica in 2004 and Argentina in 2005—exceeded ninety percent participation during that time. See id. at 8. Although retail investors or pensioners are sometimes among the holdouts, the business of holding out on a sovereign restructuring to litigate for a profit is largely limited to highly specialized distressed debt hedge funds. 134See Jill E. Fisch & Caroline M. Gentile, Vultures or Vanguards?: The Role of Litigation in Sovereign Debt Restructuring, 53 Emory L.J. 1047, 1101–16 (2004) (addressing the role of holdout litigation in sovereign debt markets). The business model requires a hearty appetite for risk and an ample war chest to fund—potentially—several years of litigation and asset hunting. 135See Robin Wigglesworth, Vulture Funds Come Under Sovereign Fire, Fin. Times (Apr. 24, 2013, 10:09 AM), http://www.ft.com/intl/cms/s/0/41a633ae-ab3d-11e2-8c63-00144feabdc0.html#axzz2rQrOZWe2 (referring to the “risky and difficult” nature of the business model).

Sovereign debt contracts have evolved in response to legal voids, collective action, and coordination problems—albeit in an incomplete, piecemeal fashion. 136See W. Mark C. Weidemaier, Remarks Made Before the United Nations General Assembly Ad-Hoc Committee on Sovereign Debt Restructuring (Feb. 5, 2015) in Univ. N.C. Legal Studies Research Paper Series, Feb. 10, 2015, at 1. While contractual responses to problems in sovereign debt are relatively easy to implement, they are often limited in scope. 137Park & Samples, supra note 98, at 43 (highlighting the shortcomings of contractual and institutional responses to problems in sovereign debt). Collective action clauses (CACs) are designed to alleviate coordination problems among creditors by enabling a qualified majority of creditors (usually seventy-five percent) to change critical bond payment terms. 138A majority modification clause (a specific type of CAC) enables a percentage of bondholders—often seventy-five percent—to make restructuring decisions that bind one hundred percent of the entire bond issuance. For a full discussion of CACs and their evolution in sovereign debt, see Weidemaier & Gulati, supra note 129, at 6. CACs have been celebrated as a cure for holdout problems in sovereign debt. 139See, e.g., NML Capital, Ltd. v. Republic of Argentina, 727 F.3d 230, 247–48 (2d Cir. 2013), cert. denied, 134 S. Ct. 2819 (2014). But that view is probably too optimistic. Many outstanding sovereign bonds simply do not have CACs. 140See Declaration of Stephen Choi at 8, NML Capital, Ltd. v. Republic of Argentina, No. 08-CV-6978, 2012 WL 7656066 (S.D.N.Y. Nov. 16, 2012) (noting that “25.3% of New York law governed bond issuances with a maturity date of 2013 or later employ [unanimous action clauses] for changes to payment related terms.”). When they do, CACs are often limited in scope. 141See Yan Liu, et al., Int’l Monetary Fund, Strengthening the Contractual Framework to Address Collective Action Problems in Sovereign Debt Restructuring 19 (2014), https://www.imf.org/external/np/pp/eng/2014/090214.pdf (comparing types of CACs). Recent improvements, responding in part to the NML decisions, are especially promising. 142Key improvements include guidance for drafting superior CACs as well as a pari passu clause that expressly preempts the ratable payment interpretation. See Int’l Capital Mkt. Ass’n, supra note 13; see also Anna Gelpern, ICMA CACs, New York Edition—Vietnam!—and More Un-Boilerplate, Credit Slips (Nov. 18, 2014, 11:13 AM), http://www.creditslips.org/creditslips/2014/11/icma-cacs-new-york-edition-vietnam-and-more-un-boilerplate.html (discussing implementation of ICMA-proposed revisions in recent issuances by Kazakhstan, Vietnam, and Mexico). But contractual solutions to problems in sovereign debt remain incomplete, despite the persistence of longstanding problems. 143See Weidemaier, supra note 136, at 1 (“Still, it is fair to say that, despite two centuries of attempted reform, many believe the contract template for sovereign lending remains flawed.”); see also Martin Guzman & Joseph E. Stiglitz, Creating a Framework for Sovereign Debt Restructuring that Works, in Too Little, Too Late: The Quest for Resolving Sovereign Debt Crises 15–20 (Martin Guzman, José Antonio Ocampo & Joseph E. Stiglitz eds., 2016) (underscoring the limitations of contractual solutions).

The NML court responded to Argentina’s unwillingness to pay with injunctive relief broadly applicable to third parties. 144See supra notes 52–54 and accompanying text. Enforcement through ratable payment injunctions solves certain problems posed by a recalcitrant sovereign defendant. 145See Sung Hui Kim, Pari Passu: The Nazi Gambit, 9 Cap. Mkts. L.J. 242, 243 (2014) (observing that broad injunctions gave the pari passu clause teeth, “a concrete remedy that could be used by the holdout creditor to induce sovereign debtors to pay”). In doing so, however, this approach creates a number of new problems. 146See Weidemaier & Gelpern, supra note 11, at 213–17 (explaining collateral costs for third parties stemming from NML injunctions). First, ratable payment injunctions endanger restructuring incentives for sovereign creditors by aggravating the classical prisoner’s dilemma problem that affects them. 147See Schumacher, Argentina Implications, supra note 35, at 146. Second, enforcing sovereign debt through injunctive remedies shifts costs and burdens to innocent third parties. 148See Cross, supra note 11, at 136–37 (citing external costs for third parties caused by Argentina’s 2014 default); Weidemaier & Gelpern, supra note 11, at 210–18 (articulating fundamental problems associated with enforcing sovereign debt through injunctions); An Illusory Haven, Economist (Apr. 20, 2013), http://www.economist.com/news/finance-and-economics/21576391-what-lessons-should-investors-learn-argentine-and-greek-restructurings (“But the battle has inflicted collateral damage on a host of third parties, from Ghanaian ports to American custodian banks.”). Exchange bondholders from the 2005 and 2010 swaps have gone unpaid since the NML injunctions came into effect. 149Vivianne Rodrigues & John Paul Rathbone, Argentina Bond Investors Challenge Long Arm of US Law, Fin. Times (July 3, 2014), www.ft.com/intl/cms/s/0/471b5be2-02c7-11e4-a68d-00144feab7de.html; Matt Levine, Argentina’s Bond Mess Gets Slightly More Complicated, Bloomberg (Feb. 13, 2015), http://www.bloombergview.com/articles/2015-02-13/argentina-s-bond-mess-gets-slightly-more-complicated [hereinafter Levine, Argentina’s Bond Mess]. A wave of secondary litigation also followed, dragging financial intermediaries into the fray. 150See Levine, Argentina’s Bond Mess, supra note 149, at 3–4. Finally—and most pressing for the current situation—ratable payment injunctions could also paralyze a sovereign defendant’s incentives to settle with holdouts. 151See infra Part IV.C. This is especially true when a uniform settlement offer, like Argentina’s, creates inter-creditor inequity in terms of disparity in returns. 152See Levine, Argentina’s Bond Fight, supra note 16; see also Benedict Mander & Robin Wigglesworth, ‘Holdout’ Slams Argentine Debt Offer to End Financial Blockade, Fin. Times (Feb. 12, 2016, 5:39 PM), http://www.ft.com/intl/cms/s/0/d17fb6b4-d1a3-11e5-986a-62c79fcbcead.html#axzz3zzJjx1DC (reporting on resistance amongst certain holdout creditors to Argentina’s initial settlement offer). Absent inter-creditor coordination to distribute returns in an equitable manner, Argentina’s offer would result in dramatically different returns (or haircuts) on the various creditor claims. 153See Blitzer, supra note 122. Exhibit 8, for example, illustrates the dramatic differences in the value of various holdout claims just based on when a judgment is obtained. Reactions to Argentina’s initial settlement proposal showed the potential for holdout problems to undermine settlement negotiations as well. 154Id.

B. Settling Sovereign Debt Disputes

The historical record indicates that, in the vast majority of sovereign debt disputes, the litigating parties arrive at a negotiated settlement. 155From 1976 to 2010, 120 holdout suits were filed against sovereigns in the United States, England, and the World Bank’s International Centre for the Settlement of Investment Disputes. See Schumacher et. al., Sovereign Defaults in Court, supra note 26, at 10–12. Of the sixty-five cases that reached a final outcome, holdout claims were paid in full in eleven of those cases whereas negotiated settlements were reached in forty-eight. See id. Moreover, the presiding judge has repeatedly encouraged the parties to reach a negotiated settlement, but to no avail. 156Transcript of Hearing at 23–24, NML Capital, Ltd. v. Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. May 29, 2015) (No. 14-CV-8601 (TPG)) [hereinafter May Hearing] (“We are dealing with large amounts of money. But negotiations and settlements have occurred before where large amounts of money are involved, and that can occur again. The Court, of course, cannot order a settlement. But I want to say in this courtroom before this entire group that the way to ultimately resolve this litigation must come through settlement. Unless the Republic wants to pay 100 percent of the judgments, which the Republic doesn’t indicate that it will do, then there has to be a settlement.”). A settlement between Argentina and the NML holdouts would likely resemble a typical sovereign debt restructuring in fundamental ways. For one, sovereign insolvency situations are often resolved through a voluntary exchange of existing (distressed or defaulted) debt obligations for new debt obligations. 157See Buckley, supra note 25, at 225, 229; Schumacher et al., Sovereign Defaults in Court, supra note 26. A settlement with the NML holdouts would likely involve the issuance of new bonds by Argentina. 158Argentina settled an expropriation dispute with Repsol with a bond issuance. Stanley Reed & Raphael Minder, Repsol in $5 Billion Settlement with Argentina, N.Y. Times (Feb. 25, 2014), http://www.nytimes.com/2014/02/26/business/international/repsol-said-to-reach-settlement-with-argentina.html?_r=0. Also, Argentina’s settlement with NML holdouts would likely include a haircut on the full value of the claim. 159See supra notes 155–156. Structuring a settlement with numerous creditors is challenging enough. But settling a large number of claims with vastly different valuations and highly divergent creditors—from individual pensioners to distressed debt hedge funds—promises to be even more challenging. Below we explain how the NML injunctions make an exceedingly challenging task a deeply irrational one.

C. Ratable Payment Injunctions Versus Negotiated Settlement

For years, the district court has correctly recognized that the only realistic way out of the NML litigation is through settlement. In doing so, the court has recognized that a settlement for less than the full claim amount is the most likely outcome. 160May Hearing, supra note 156, at 23–24; Transcript of Hearing at 9, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Aug. 8, 2014) (No. 08-CV-6978 (TPG)). More recently, the court has continued to reiterate the view that a negotiated settlement is the only answer. 161Transcript of Hearing at 11, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Aug. 21, 2014) (No. 08-CV-6978 (TPG)) [hereinafter August 21, 2014 Hearing]; Transcript of Hearing at 29, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Oct. 28, 2015) (No. 14-CV-8601 (TPG)). However, the court faces a dilemma between enforcing holdout claims and encouraging settlement. As currently drafted, the district court’s very own ratable payment injunctions present a serious impediment to settlement. This dilemma exposes an additional problem associated with using injunctions to enforce sovereign debt judgments: settlement complications. 162Previous work by scholars has articulated more fundamental issues related to injunctions as an enforcement tool in sovereign debt. See Weidemaier & Gelpern, supra note 11, at 213–17 (explaining public interest considerations and costs for innocent third parties); see also Cross, supra note 11, at 135–40 (critiquing the NML injunctions from a Foreign Sovereign Immunities Act perspective).

The NML injunctions prohibit Argentina from paying exchange creditors from previous restructurings until holdout claims have been paid in their entirety. Since the outstanding holdout creditors are not bound by collective action mechanisms, certain holdouts within the holdouts may remain even after a settlement exchange offer. 163Argentina’s holdout bonds lack CACs. Mario Blejer, Argentina’s Deal with the Holdouts is a Mixed Blessing, Fin. Times (Mar. 31, 2016), https://www.ft.com/content/db6779d6-f729-11e5-96db-fc683b5e52db. In sovereign bonds issued under New York law prior to 2003, unanimous action clauses were standard. See Stephen J. Choi & G. Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds, 53 Emory L.J. 929, 932–33 (2004). As a result, under the current injunctions, even just one settlement holdout could hijack payments to existing exchange creditors and future settlement creditors. Put differently, NML-style injunctions mean that anything less than one hundred percent participation could derail the entire settlement effort. These issues raise doubts about the viability and participation incentives for a potential settlement of NML. Making matters worse, on a practical level, locating and coordinating all the holders in Argentina’s 126 outstanding bonds with holdouts is a difficult task. 164See supra Part I.B (detailing the allocation of holdouts across Argentina’s defaulted bonds).

Unlike more straightforward commercial cases between sophisticated parties, the enforcement of sovereign debt litigation can involve weighty social and policy questions, including serious collateral costs for third parties. 165See Pottow, supra note 11, at 225, 229; Gelpern, Contract Hope, supra note 40, at 132–33. As an equitable remedy, public interests and collateral costs for third parties are especially relevant in considering the use of injunctive relief. 166See Weidemaier & Gelpern, supra note 11, at 199–200 (discussing injunctive relief as a remedy in sovereign debt litigation). With a new administration at the helm, Argentina demonstrated good faith in quickly putting forth a reasonable settlement offer. 167See Gelpern, Love and Exhaustion, supra note 18. If Argentina’s publicly stated defiance of court orders was a primary driving force behind the injunctions, perhaps Argentina’s good faith efforts towards a negotiated settlement would be cause to reconsider the NML injunctions. 168See id. at 1; see also Mark Weidemaier, Argentina’s Settlement Negotiations and Lifting the Injunction, Credit Slips (Feb. 9, 2016, 8:56 AM), http://www.creditslips.org/creditslips/2016/02/argentinas-settlement-negotiations-and-lifting-the-injunction.html. As a matter of policy, a sovereign’s legitimate interest in restructuring unsustainable debt is just the beginning. Collateral costs for innocent third parties loom large as well. 169The NML court has pointed to collateral damage to “very innocent third parties” while advocating for a settlement between the parties. August 21, 2014 Hearing, supra note 161, at 11.

Conclusions

In earlier phases of the NML litigation, the Southern District of New York recognized Argentina’s economic realities and interest in restructuring a substantial debt burden. 170See Miller & Thomas, supra note 36, at 1492–93, 1499–501 and accompanying text; see also EM Ltd. v. Republic of Argentina, 131 F. App’x 746, 747 (2d Cir. 2005); supra text accompanying note 37. But adjudicating sovereign debt disputes is no easy task. Institutional voids and limited enforceability only make matters more difficult for courts. 171See supra notes 123–129 and accompanying text. As a result, striking a balance between the legitimate restructuring needs of a sovereign debtor, the interests of innocent third parties, and the legitimate rights of creditors is a complicated goal. Though injunctive remedies may serve to force an unwilling sovereign debtor to the negotiating table, collateral costs can be significant. NML also demonstrates that enforcing sovereign debt through injunctions can potentially obstruct a settlement process by exacerbating creditor coordination problems. Additionally, our analysis offers a framework for what can be considered a baseline for comparisons in holdout negotiations and Argentina’s settlement offer.

Appendix

Secondary market price sources and average annual price computations:

We use what we consider the best source for the closing price of each defaulted bond on each day in the sample and then take the simple average of observed prices by year (or by semester for 2001). The data come from various Bloomberg sources [i.e. Bloomberg Valuation Service (BVAL), Deutsche Bank (DAB), Frankfurt Exchange (FRNK), German Exchange (GERM), and Stuttgart Exchange (STGT)], and they are improved upon with data obtained from a leading investment bank in New York (LIB). According to Bloomberg, “BVAL draws on market data from a wealth of sources. . . . [and] combines these market observations with sophisticated analytics and asset class-specific relative value models to produce credible, defensible and independent valuations. . . . BVAL’s prices are highly reactive and most closely reflect current market conditions.” 172BVAL Provides Objective, Independent & Defensible Evaluated Pricing, Bloomberg for Enterprise, https://www.bbhub.io/solutions/sites/8/2015/10/BVAL-Evaluated-Pricing-fact-sheet.pdf (last visited Nov. 13, 2016). Following are the sources used for each bond and time period using the same bond codes as supra note 26. #7: Average of LIB and DAB until 2011. Average of LIB and BVAL starting in 2013. As for 2012, we average DAB and LIB until 02/20/2012, we take LIB from 02/21/2012 until 04/02/2012, and then average BVAL and LIB starting 04/03/2012. Finally, we take the simple average of daily prices to obtain the 2012 average, just like we do for all other years and bonds #9, #11, and #14: LIB until 2011. As for 2012, we use LIB until 04/02/2012 and an average of BVAL and LIB starting on 04/03/2012. The latter sources carry on after 2012. #17: LIB until 2012. As for 2013, we use LIB until 01/31/2013 and BVAL starting 02/01/2013. BVAL for 2014. #41: LIB until 2007 and BVAL from 2012 onwards. Since none of the sources had market data for this bond from 2008 until 2011, we impute its price as follows. We start with the annual prices of an equally-weighted portfolio of like bonds 7, 9, 11, and 14 for which we do have market prices (we exclude bonds 17 (FRAN) and 48 (Brady) given their differing characteristics with 41). We next fit a linear trend from the portfolio’s price in 2007 to its price in 2012. We next compute, for each year from 2008 until 2011, the ratio between the actual portfolio price and the imputed trend price for that year. We next fit a trend from bond 41’s price in 2007 to that in 2012. Finally, for each year from 2008 until 2011, we multiply this trend price by the ratio just computed for the portfolio. The final result is an imputed price that has the observed prices in 2007 and 2012, but which for each year in the interim has a variation that is proportional to that in the portfolio of like bonds. #48: Average of DAB, STGT, GERM and FRNK.

Regarding Exhibit 6:

Exchange bonds were handed to participating exchange bondholders on June 2, 2005. The GDP-linked warrants originally attached to those exchange bonds started trading independently of the bonds when they detached on November 24, 2005. We assumed that the holder of the “mother” exchange bond sold those warrants in the market at the ongoing price for the warrants alone and reinvested that money in the post-detachment mother bond.

The end date of all the security prices in Exhibits 6 and 8 is November 24, 2015.

Footnotes

Juan J. Cruces is Professor of Finance and Economics at the Universidad Torcuato Di Tella Business School.Tim R Samples is Assistant Professor of Legal Studies, Terry College of Business, University of Georgia.We are grateful for the outstanding research assistance of Federico Varni, especially on the history of NML litigation. Juan Martín Cappellini, Corey Goerdt, Patricio Goldstein, Victoria Lacasa, Martín Przybylski, and Jonathan Tonge also provided valuable research assistance. Anna Gelpern, Martin Guzman, Matías Isasa, Stephen Park, and Andrea Rotnitzky provided valuable comments and/or conversations. The usual caveat applies. A prior version of this Article was presented at the Interdisciplinary Sovereign Debt Research and Management Conference at Georgetown University Law Center in January 2016. This research was supported in part by the Latin American and Caribbean Studies Institute (LACSI) at the University of Georgia through a LACSI Ambassador Faculty Travel Grant.

1Joseph Cotterill, Coverage of the Argentina Sovereign Debt Litigation, Fin. Times: FT Alphaville, http://ftalphaville.ft.com/tag/pari-passu-saga/ (last visited Oct. 8, 2016). Given its extraordinary nature and importance, the Financial Times coined NML the “trial of the century” of sovereign debt restructuring. Id. See NML Capital v. Republic of Argentina, 699 F.3d 246 (2d Cir. 2012). The litigation collectively involves plaintiffs, led by NML Capital, Ltd., who have claims against Argentina for New York law bonds in default. Id. at 487.

2See, e.g., Chris Giles, Gillian Tett, Elaine Moore & Benedict Mander, Argentina Pledges to Honour Debts Owed to Holdout Creditors, Fin. Times (Jan. 22, 2016, 12:02 PM), http://www.ft.com/intl/cms/s/0/356cbbae-c0f1-11e5-9fdb-87b8d15baec2.html#axzz3yISe8tio.

3Id.

4See Hold-outs Upheld, Economist (Nov. 3, 2012), http://www.economist.com/news/finance-and-economics/21565635-court-ruling-against-argentina-has-implications-other-governments-hold-outs.

5See Michael Davies & Tim Ng, The Rise of Sovereign Risk: Implications for Financial Stability, BIS Q. Rev., Sept. 2011, http://www.bis.org/publ/qtrpdf/r_qt1109g.pdf.

6See infra Part I.A (reviewing Argentina’s default and restructurings in detail).

7In this context, “holdout” creditors are those who decide not to participate in a debt restructuring whereas “exchange” creditors do. A typical restructuring involves an exchange of defaulted or distressed debt for new debt. See infra Part I.C (reviewing holdout rates in the 2005 and 2010 restructurings in detail).

8Dan Rosenheck, Argentina’s Rational Default, New Yorker (Aug. 7, 2014), http://www.newyorker.com/business/currency/argentinas-rational-default (characterizing Argentina’s default as “perhaps the strangest in history”). See infra Part I.B (reviewing Argentina’s holdout litigation in detail).

9See Rosenheck, supra note 8.

10See, e.g., William W. Bratton, Pari Passu and a Distressed Sovereign’s Rational Choices, 53 Emory L.J. 823 (2004); Lee C. Buchheit & G. Mitu Gulati, Exit Consents in Sovereign Bond Exchanges, 48 UCLA L. Rev. 59 (2000); Lee C. Buchheit & Ralph Reisner, The Effect of the Sovereign Debt Restructuring Process on Inter-Creditor Relationships, 1988 U. Ill. L. Rev. 493 (1988); Ugo Panizza, Federico Sturzenegger & Jeromin Zettelmeyer, The Economics and Law of Sovereign Debt and Default, 47 J. Econ. Literature 651 (2009); W. Mark C. Weidemaier, Sovereign Immunity and Sovereign Debt, 2014 U. Ill. L. Rev. 67 (2014).

11See, e.g., Karen Halverson Cross, The Extraterritorial Reach of Sovereign Debt Enforcement, 12 Berkeley Bus. L.J. 111 (2015); Arturo C. Porzecanski, From Rogue Creditors to Rogue Debtors: Implications of Argentina’s Default, 6 Chi. J. Int’l L. 311 (2005); John A. E. Pottow, Mitigating the Problem of Vulture Holdout: International Certification Boards for Sovereign-Debt Restructurings, 49 Tex. Int’l L.J. 219 (2014); Tim R Samples, Rogue Trends in Sovereign Debt: Argentina, Vulture Funds, and Pari Passu Under New York Law, 35 Nw. J. Int’l L. & Bus. 49 (2014); Julian Schumacher, Sovereign Debt Litigation in Argentina: Implications of the Pari Passu Default, 1 J. Fin. Reg. 143 (2015); S.I. Strong, Rogue Debtors and Unanticipated Risk, 35 U. Pa. J. Int’l L. 1139 (2014); W. Mark C. Weidemaier & Anna Gelpern, Injunctions in Sovereign Debt Litigation, 31 Yale J. on Reg. 189 (2014); Mark L. J. Wright, Sovereign Debt Restructuring: Problems and Prospects, 2 Harv. Bus. L. Rev. 153 (2012).

12Int’l Monetary Fund, Sovereign Debt Restructuring—Recent Developments and Implications for the Fund’s Legal and Policy Framework 31 (2013), http://www.imf.org/external/np/pp/eng/2013/042613.pdf (expressing concerns about NML precedent); U.N. Conference on Trade and Dev., Argentina’s “Vulture Fund” Crisis Threatens Profound Consequences for International Financial System (2014), http://unctad.org/en/pages/newsdetails.aspx?OriginalVersionID=783.

13See, e.g., Brief for the American Bankers Association as Amicus Curiae Supporting Nonparty the Bank of New York Mellon at 14–15, NML Capital, Ltd. v. Republic of Argentina, 727 F.2d 230 (2d Cir. 2013) (Jan. 4, 2013) (No. 12-105-cv(L)); Brief for the Clearing House Association L.L.C. as Amicus Curiae in Support of Reversal at 25–27, NML Capital, Ltd. v. Republic of Argentina, 727 F.2d 230 (2d Cir. 2013) (Jan. 4, 2013) (No. 12-105-cv(L)); Int’l Capital Mkt. Ass’n, Standard Aggregated Collective Action Clauses (“CACs”) for the Terms and Conditions of Sovereign Notes Governed by English Law (May 2015), http://www.icmagroup.org/assets/documents/Resources/ICMA-Standard-CACs-Pari-Passu-and-Creditor-Engagement-Provisions—-May-2015.pdf.

14See Charlie Devereux, Argentina’s Fix-It Man Is Slowing Down and Investors Are Worried, Bloomberg (Jan. 24, 2016, 9:00 PM), http://www.bloomberg.com/news/articles/2016-01-25/argentina-s-fix-it-man-is-slowing-down-and-investors-are-worried.

15See Juan J. Cruces & Tim Samples, Time to Settle Sovereign Debt’s “Trial of the Century”?, Col. L. Sch.: CLS Blue Sky Blog (Jan. 26, 2016), http://clsbluesky.law.columbia.edu/2016/01/26/time-to-settle-sovereign-debts-trial-of-the-century/ (underscoring challenges for settlement negotiations); see also Charles Blitzer, Guest Post: Best Practices to Resolve Argentina’s Debt Dispute, Fin. Times: FT Alphaville (Jan. 22, 2016), http://ftalphaville.ft.com/2016/01/22/2150862/guest-post-best-practices-to-resolve-argentinas-debt-dispute/ (recommending strategies for the negotiations).

16See Matt Levine, Argentina’s Bond Fight Comes Down to Its Worst Bonds, Bloomberg (Feb. 8, 2016), http://www.bloombergview.com/articles/2016-02-08/argentina-s-bond-fight-comes-down-to-its-worst-bonds [hereinafter Levine, Argentina’s Bond Fight].

17Id.

18See Anna Gelpern, Love and Exhaustion in Argentina, Credit Slips (Feb. 11, 2016, 10:23 PM), http://www.creditslips.org/creditslips/2016/02/love-and-exhaustion-in-argentina.html [hereinafter Gelpern, Love and Exhaustion]; see also infra Part IV.C.

19See A Victory by Default?, Economist (Mar. 3, 2005), http://www.economist.com/node/3715779.

20See Republic of Argentina, Prospectus Supplement (To Prospectus Dated Dec. 27, 2004) at 11 (2005), http://www.sec.gov/Archives/edgar/data/914021/000095012305000302/y04567e424b5.htm [hereinafter Prospectus Supplement].

21Sophie Arie & Andrew Cave, Argentina Makes Biggest Debt Default in History, Telegraph (Dec. 24, 2001), http://www.telegraph.co.uk/news/worldnews/southamerica/argentina/1366218/Argentina-makes-biggest-debt-default-in-history.html.

22See A Victory by Default?, supra note 19.

23William W. Bratton & G. Mitu Gulati, Sovereign Debt Reform and the Best Interest of Creditors, 57 Vand. L. Rev. 1, 10–13 (2004) (underscoring gaps in sovereign debt system).

24Jonathan I. Blackman & Rahul Mukhi, The Evolution of Sovereign Debt Litigation: Vultures, Alter Egos, and Other Legal Fauna, 73 L. & Contemp. Probs. 47, 48 (2010) (contrasting the voluntary nature of sovereign debt restructuring with the corporate bankruptcy process).

25Debt rescheduling and debt reduction are common methods for restructuring distressed or defaulted debt. See Ross P. Buckley, 8 Emerging Market Debt: An Analysis of the Secondary Market ch. 2 (1999). Growth-linked securities have also been featured in recent sovereign restructurings: Argentina (2005 and 2010), Greece (2012), and Ukraine (2015). See James Benford, et al., Bank of England, Sovereign GDP-linked bonds, Financial Stability Paper No. 39 (Sept. 21, 2016). For a discussion of Argentina’s GDP-linked warrants, see infra Part III.A.

26See Julian Schumacher et al., Sovereign Defaults in Court: The Rise in Creditor Litigation 1976-2010 22 (May 6, 2014) (unpublished manuscript) (on file at http://ssrn.com/abstract=2189997) (articulating factors that make sovereign debt litigation more likely) [hereinafter Schumacher et al., Sovereign Defaults in Court].

27Anna Gelpern, What Bond Markets Can Learn From Argentina, Int’l Fin. L. Rev., Apr. 2005, at 19 (describing the “unusually contentious” tone of the negotiations).

28See Juan J. Cruces & Christoph Trebesch, Sovereign Defaults: The Price of Haircuts, 5 Am. Econ. J.: Macroeconomics 85, 86 (2013), https://www.aeaweb.org/articles.php?doi=10.1257/mac.5.3.85 (finding that the average haircut in sovereign restructurings is thirty-seven percent); see also Juan José Cruces, ¿Cómo Resolver el Problema de los Holdouts y Bajar el Costo de Capital de la Economía Argentina?, 59 Fondo de Cultura Económica (forthcoming 2016) [hereinafter Cruces, Problema de los Holdouts] (refining the seventy-six percent haircut calculation of previous work by incorporating the present value of Argentina’s GDP-linked warrants and arriving at 73.4%).

29See Elena Duggar, Special Comment: The Role of Holdout Creditors and CACs in Sovereign Debt Restructurings, Moody’s Inv. Servs., Apr. 10, 2013, at 4. Even without taking the 2010 exchange into account, Argentina had an exceptionally long restructuring process. See id. The length of time before Argentina’s 2005 restructuring is over twice the average for sovereign debt restructurings. Id.

30See id. at 1.

31See id. at 8.

32Schumacher et al., Sovereign Defaults in Court, supra note 26, at 11 (explaining that post-2001 Argentina accounts for nearly one-third of all sovereign debt cases between 1976 and 2010).

33See Benedict Mander et al., Argentina: Unresolved Debts, Fin. Times (Aug. 6, 2014), http://www.ft.com/intl/cms/s/0/96b56394-1d68-11e4-b927-00144feabdc0.html#axzz3wfRLZpKM (describing “semantic disagreement” over whether or not a default had occurred).

34See Samples, supra note 11, at 63–75 (explaining the exceptional nature of NML litigation).

35Julian Schumacher, Sovereign Debt Litigation in Argentina: Implications of the Pari Passu Default, 1 J. Fin. Reg. 143, 144 (2015) [hereinafter Schumacher, Argentina Implications] (illustrating the trajectory of claims filed against Argentina since 2002).

36Marcus Miller & Dania Thomas, Sovereign Debt Restructuring: The Judge, the Vultures and Creditor Rights, 30 World Econ. 1491, 1500 (2007) (describing the district court’s use of judicial discretion to promote debt restructuring in NML).

37EM Ltd. v. Republic of Argentina, 131 F. App’x 745, 747 (2d Cir. 2005). The Second Circuit cited concerns for “the economic health of a nation” in upholding the district court’s refusal to allow NML plaintiffs to block Argentina’s 2005 exchange. Id.

38Our calculation of these amounts uses source data from Julian Schumacher, Enforcement in Sovereign Debt Markets (July 14, 2015) (unpublished Ph.D. dissertation, Humboldt-Universität zu Berlin) http://nbn-resolving.de/urn:nbn:de:kobv:11-100234666 [hereinafter Schumacher, Enforcement in Sovereign Debt Markets].

39Alison Frankel, How Argentina Lost Game of Chicken with Renegade Bondholders, Reuters (Nov. 26, 2012), http://blogs.reuters.com/alison-frankel/2012/11/26/how-argentina-lost-game-of-chicken-with-renegade-bondholders/ (documenting the shift in strategy from asset attachment to equal footing claims).

40See Anna Gelpern, Contract Hope and Sovereign Redemption, 8 Cap. Mkts. L.J. 132, 139–40 (2013) (noting the court’s exasperation with Argentina) [hereinafter Gelpern, Contract Hope].

41EM Ltd. v. Republic of Argentina, 720 F. Supp. 2d 273, 304 (S.D.N.Y. 2010), vacated sub nom. NML Capital, Ltd. v. Banco Cent. de la Republica Argentina, 652 F.3d 172 (2d Cir. 2011) (“What is going on between the Republic of Argentina and the federal court system is an exercise of sheer willful defiance of the obligations of the Republic to honor the judgments of a federal court.”).

42NML Capital, Ltd. v. Banco Cent. de la Republica Arg., 652 F.3d 172, 196 (2d. Cir. 2011) (“We share the District Court’s understandable irritation at the Republic’s ‘willful defiance of [its] obligations to honor the judgments of a federal court.’”).

43See infra notes 50–53.

44Attaching valuable, non-immune sovereign assets is notoriously difficult. See Bratton, supra note 10, at 824 (“Sovereigns in default rarely leave valuables lying around subject to attachment in creditor-friendly jurisdictions.”); see also George K. Foster, Collecting from Sovereigns: The Current Legal Framework for Enforcing Arbitral Awards and Court Judgments Against States and Their Instrumentalities and Some Proposals for its Reform, 25 Ariz. J. Int’l & Comp. L. 665 (2008).

45See Schumacher, Argentina Implications, supra note 35, at 144 (illustrating the filing of NML claims invoking pari passu); see also Natalie A. Turchi, Note, Restructuring a Sovereign Bond Pari Passu Work-Around: Can Holdout Creditors Ever Have Equal Treatment?, 83 Fordham L. Rev. 2171, 2196–202 (2015) (documenting the evolution of pari passu under New York law).

46Many prominent voices in sovereign debt have questioned—or have even rejected outright—the ratable payment interpretation of the pari passu clause. See, e.g., Lee C. Buchheit & Jeremiah Pam, The Pari Passu Clause in Sovereign Debt Instruments, 53 Emory L.J. 869, 877–91 (2004) (rejecting the ratable payment interpretation of pari passu); G. Mitu Gulati & Kenneth N. Klee, Sovereign Piracy, 56 Bus. Law. 635, 650 (2001) (making the case against the ratable payment interpretation of pari passu); Analysis of the Role, Use and Meaning of Pari Passu Clauses in Sovereign Debt Obligations as a Matter of English Law 79 Fin. Mkts. L. Comm., 17 (2005), http://www.fmlc.org/uploads/2/6/5/8/26584807/79.pdf. For arguments in favor of ratable payment interpretation, see Brief for Kenneth W. Dam as Amicus Curiae Supporting Respondents, NML Capital Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. 2012) (No. 12-105-cv(L)) (arguing in favor of third party injunctions also); Brief for Washington Legal Foundation as Amicus Curiae Supporting Respondents, NML Capital Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Oct. 26, 2012) (No. 12-105-cv(L)).

47NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246, 251 (2d Cir. Oct. 26, 2012) (quoting the pari passu clause from Argentina’s Fiscal Agency Agreement).

48Order at *2–3, NML Capital, Ltd. v. Republic of Argentina, No. 08-6978, 2011 WL 9522565 (Dec. 7, 2011) [hereinafter 2011 NML Order].

49See NML Capital, Ltd v. Republic of Argentina, No. 08-6978, 2012 WL 5895786, at *3 (S.D.N.Y. Nov. 21, 2012) (finding that the holdouts were entitled to one hundred percent of amounts owed by Argentina every time that Argentina pays one hundred percent of amounts owed to exchange bondholders).

50See 2011 NML Order, supra note 48 at *2.

51Weidemaier & Gelpern, supra note 11, at 191 (“Put differently, the injunction allows Argentina to keep stiffing NML, but only if it also stiffs the exchange bondholders.”).

52Order, NML Capital, Ltd. v. Republic of Argentina, No. 1:09-CV-01707 (S.D.N.Y. filed Feb. 23, 2012).

53Id.

54The court was acutely aware that Argentina would likely continue to defy its orders. See, e.g., Transcript of Hearing at 15, NML v. Argentina, 144 F. Supp. 3d 513 (Nos. 08-CV-6978 and 09-CV-1708) (S.D.N.Y. 2012).

55In 2000, Elliott v. Peru broke new ground in pari passu litigation with ratable payment injunctions applicable to third parties. See Brief for the United States of America as Amicus Curiae in Support of Reversal, NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Apr. 4, 2012) (No. 12-105-cv(L)), 2012 WL 1150791. For criticisms of the “ratable payment” interpretation of pari passu abounded after the Brussels decision, see, e.g., Gulati & Klee, supra note 46; Rodrigo Olivares-Caminal, The Pari Passu Interpretation in the Elliott Case: a Brilliant Strategy but an Awful (Mid-Long Term) Outcome?, 40 Hofstra L. Rev. 39 (2011).

56Brief for the United States of America as Amicus Curiae Supporting Argentina’s Petition for Panel Rehearing and Rehearing En Banc, NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246 (2d Cir. Dec. 28, 2012) (No. 12-105-cv(L)) 2012 WL 5275014.

57See NML Capital, Ltd. v. Republic of Argentina, 699 F.3d 246, 250 (2d Cir. 2012).

58Republic of Argentina v. NML Capital, Ltd., 727 F.3d 230 (2d. Cir. Aug. 23, 2013), cert. denied, 134 S. Ct. 2819 (June 16, 2014) (No. 13-990); Exch. Bondholder Grp. v. NML Capital, Ltd., 727 F.3d 239 (2d Cir. Aug. 23, 2013), cert denied, 134 S. Ct. 2819 (June 16, 2014).

59The seven bonds with particularly high (above twenty percent) holdout rates plus the sixty-seven bonds with non-trivial holdout rates (above three percent) comprise the total of seventy-four bonds with meaningful holdout rates.

60See infra Part IV.C (addressing the practical difficulties of coordinating a settlement across numerous debt instruments).

61The seven bonds, including their currency, coupon rate, ISIN codes, maturity dates and total principal still outstanding are (the number preceding each bond is the order in which each bond appears in Republic of Argentina 2004, Annex C): #41: Global bond, ARP ten percent 2001-2004 and twelve percent 2004-2008 due 2008, XS0130278467, Jun-2001, $595; #9: Global bond, USD 11.375% due 2017, US040114AR16, Jan-1997, $419; #17: Bond, USD variable rate due 2005 (FRAN), US040114AX83, Apr-1998, $298; #14: Global bond, USD 10.25% due 2030, US040114GB00, Jul-2000, $122; #7: Global bond, USD 12.375% due 2012, US040114GD65, Feb-2001, $113; #11: Global bond, USD twelve percent due 2020, US040114FB19, Feb-2000, $66; #48: Brady Par bond, EUR 5.87% due 2023, DE0004103007, Mar-1993, $58.

62For the sake of completeness, the bottom line of Exhibit 2 shows the principal amount of defaulted bonds outstanding at each point in time, in billions of dollars.

63See Schumacher et al., Sovereign Defaults in Court, supra note 26, at 11 (observing the runs to the courthouse that occurred after the debt crises of Argentina and Peru).

64Federal rules are silent on pre-judgment interest but provide a floating post-judgment interest rate. 28 U.S.C. § 1961 (2012).

65N.Y. C.P.L.R. 5001(b)–(c) (McKinney 1992).

66NML Capital v. Republic of Argentina, 621 F.3d 230, 240 (2d Cir. 2010).

67See Michael S. Knoll & Jeffrey M. Colon, The Calculation of Prejudgment Interest 3, (U. of Pa. L. Sch., Public Law Working Paper No. 06-21, 2005) (citing Kansas v. Colorado, 533 U.S. 1 (2001) (“[L]awyers and their experts have wide latitude in persuading the court on just how much pre-judgment interest the defendant should pay to make the plaintiff whole.”)).

6828 U.S.C. § 1961 (2012).

69NML Capital v. Republic of Argentina, 621 F.3d at 239 certified question accepted, 15 N.Y.3d 859 (2010), certified question answered, NML Capital v. Republic of Argentina, 952 N.E.2d 482 (2011) (“If the parties failed to include a provision in the contract addressing the interest rate that governs after principal is due or in the event of a breach, New York’s statutory rate will be applied as the default rate.”).

70NML Capital v. Republic of Argentina, 952 N.E.2d 482, 489–490 (“[I]t is undisputed that Argentina must pay interest on principal at the contract rate” while the “statutory interest on the unpaid interest payments compensates the bondholders for a different loss”).

71See id.

72See id. at 488–89.

73See infra Part II.B (quantifying Argentina’s interest liabilities).

74David D. Siegel, Supplementary Practice Commentaries, in N.Y. C.P.L.R. § 5004 (McKinney 1992 & Supp. 2014).

75N.Y. C.P.L.R. § 5004. For a full discussion of New York’s prejudgment interest rates, see Laila Abou-Rahme & Stephen Scotch-Marmo, Is It Time for N.Y.’s Prejudgment Interest Rate to Float?, Special Rep. (N.Y. L.J.) Sept. 9, 2013, at S4.

76We conduct a bilateral Wilcoxon-Mann-Whitney test of the null hypothesis that real statutory rates were equally distributed during the periods of 1962–2001 and 2002–2015 against the alternative that real rates after 2002 were different. See Paul Newbold, Statistics for Business and Economics 394 (4th ed. 2005). The p-value for the test statistic is .00007. We repeat the test, narrowing the first period to 1981–2001, and thus focusing on two subperiods that had a constant nominal nine percent statutory rate. The p-value in this case is .017. In both cases, the interpretation is that it is extremely unlikely that the two samples that we observe (the gray bars and the black bars) come from the same original distribution, or in plain English, that they had the same mean real rate. The statistical test thus supports the view that real statutory rate since Argentina’s default has been significantly higher than during previous periods.

77See Abou-Rahme & Scotch-Marmo, supra note 75, at 3.

78See Advisory Comm. Civil Practice, Report to the Chief Administrative Judge of the Courts of the State of New York 120, (2013), [hereinafter Advisory Report], https://www.nycourts.gov/ip/judiciaryslegislative/pdfs/2013-CivilPractice-ADV-Report.pdf.

79See Abou-Rahme & Scotch-Marmo, supra note 75; see also Advisory Report, supra note 78, at 120 (finding the fixed nine percent rate “both illogical and unfair” and proposing a rate equivalent to a one-year Treasury bill plus three percent).

80See Abou-Rahme & Scotch-Marmo, supra note 75.

81Id.

82The timing of judgments is so critical because, once the court enters into a final judgment, the claim is thereafter subject to interest at the federal post-judgment rate, which matches yields on U.S. Treasury bills. This post-judgment rate is drastically lower than contract interest plus the New York nine percent pre-judgment interest. See supra Part II.A (explaining the application of pre-judgment and post-judgment rates).

83This comes out to be 3.22% per annum. We use the realized inflation for the year ending in October 2015. Data were taken from the Federal Reserve web site. Selected Interest Rates (Daily), Bd. Governors Fed. Reserve Sys., http://www.federalreserve.gov/releases/h15/data.htm (last updated Oct. 11, 2016).

84Here we took the rate prevailing on the last week of the year in which the contractual interest was originally due. The average such rate comes out to be 3.03% per annum if judgment was handed down in 2008, and 1.54% per annum if it was handed down in 2015. Id.

85For bonds that had more than one coupon per year, we neglect this subtlety and assume that all coupons were due on the last day of the year.

86Unlike pre-judgment interest, which is calculated on a simple basis, post-judgment interest is compounded. So, if it took two years from judgment to settlement, the judgment value is grossed up by (1+T bill rate) squared.

87See Schumacher, Enforcement in Sovereign Debt Markets, supra note 38, at 146–47.

88The bottom four lines in Exhibit 4 compare the current value of the 2005 offer with the claim value in each scenario and are discussed fully in the text of Part III.

89Argentina’s bond documents did not specify a default rate. See NML Capital v. Republic of Argentina, 952 N.E.2d 482 (2011). Going forward, sovereigns issuing debt under New York law may consider setting a reasonable default rate in their bond contracts to avoid excessive pre-judgment interest liabilities. See Lee C. Buchheit, How to Negotiate Eurocurrency Loan Agreements 36–40 (Richard Forster et. al. eds., 2d ed. 2000) (discussing default interest clauses in loan agreements).

90See generally Mitu Gulati & Robert E. Scott, The 3 1⁄2 Minute Transaction: Boilerplate and the Limits of Contract Design (2012).

91For examples of scholarship discussing these implications see supra notes 11–13 and accompanying text.

92For details on the sources of the price construction, see Appendix.

93Since purchase prices are annual averages, we measure the time elapsed from purchase to final claim value, starting from the middle of the purchase year until December 2015. Hence, the basket of bonds bought in mid-2012 was held during 3.5 years: $1 x (1+1.158)3.5 = $14.76.

94See supra Part I.A.

95See supra note 28 and accompanying text.

96See Schumacher et al., Sovereign Defaults in Court, supra note 26, at 22 (articulating conditions that make sovereign debt litigation more likely).

97Stephen Park & Tim Samples, Ukraine’s Quietly Revolutionary Debt Restructuring, Fin. Times; beyondbrics, 12 (Sept. 17, 2015), http://blogs.ft.com/beyond-brics/2015/09/17/ukraines-quietly-revolutionary-debt-restructuring/ (analyzing the role of GDP-linked “value recovery instruments” in Ukraine’s 2015 restructuring).

98For a detailed analysis of the terms of Argentina’s GDP-linked warrants, see Stephen Kim Park & Tim R Samples, Towards Sovereign Equity, 21 Stan. J.L. Bus. & Fin. (forthcoming 2016), http://ssrn.com/abstract=2630772.

99See infra notes 102–04.

100These figures are comparable to the Wealth columns of Exhibit 5. As a reference, for investors who bought defaulted bonds in 2005, the accumulated wealth from collecting the claim value under NY’s nine percent rate would range between $7.63 and $12.51 depending on when they obtained judgment.

101To circumvent the problem that exchange bonds under New York law have been in default since 2014, we use the prices of their Argentine-law U.S. dollar-denominated counterparts. Only the jurisdiction differed among these bonds, not the promised payments. Argentina has been making payments on the local law bonds despite the injunctions that came into effect in 2014.

102For an early assessment of the high realized ex-post returns on the exchange bonds and GDP-linked warrants, see Brief for Alfonso Prat-Gay as Amicus Curiae in Support of the Republic of Argentina’s Petition for Panel Rehearing, NML Capital, Ltd. v. Republic of Argentina, No. 12-105-cv(L) (2d Cir. Jan. 4, 2013).

103Stephany Griffith-Jones & Krishnan Sharma, GDP-Indexed Bonds: Making It Happen, in Innovative Financing for Development 85 (Suhas Ketkar & Dilip Ratha eds., 2009); Argentine GDP Warrants, Euromoney (Jan. 25, 2006), http://www.euromoney.com/Article/1014876/Argentine-GDP-warrants.html.

104Guido Sandleris & Mark L. J. Wright, GDP-Indexed Bonds: A Tool to Reduce Macroeconomic Risk?, in The Future of Sovereign Borrowing in Europe 108 (Morten Balling, et al. eds., 2013).

105See supra Part I.

106See Cruces & Trebesch, supra note 28, at 90. If Argentina chooses to pay with a new bond, the market value of that bond compared to the claim value will determine the settlement haircut. Id. Analyzing the bonds potentially offered in an NML settlement is beyond the scope of this Article. However, given the history of declining interest rate spreads as countries emerge from default, issuing bonds with call rights might merit consideration. Many of the Brady bonds, in restructurings sponsored by the U.S. government, provided debtors with call rights. Juan J. Cruces & Nicolás Merener, Holdouts: pagar con bonos precancelables nos puedeahorrar u$s 2000 millones, casi la mitad de lo que salió YPF, El Cronista (July 21, 2014), http://www.cronista.com/columnistas/Holdouts-pagar-con-bonos-precancelables-nos-puedeahorrar-us-2000-millones-casi-la-mitad-de-lo-que-salio-YPF-20140721-0029.html.

107This move would have violated the Padlock Law (or Ley Cerrojo 26,017), which prohibited the Argentine government from making an exchange offer to holdout creditors after the 2005 exchange, including of course giving bonds in their favor to a trustee. We abstract from this fact here. See Law No. 26017, art. 2, Feb. 11, 2005 [CXIII] B.O. 30590 (Arg.).

108Including these unpaid interests was done at the request of—among others—Argentina’s Bond Restructuring Agency, a bondholder group.

109All figures in our calculations are in U.S. dollars using the official exchange rates noted in Argentina’s Prospectus Supplement. See Prospectus Supplement, supra note 20.

110For tractability, and given the extremely low liquidity of Japanese yen exchange bonds, we assumed that the 0.7% corresponding to new bonds issued in that currency in 2005 were actually denominated in euros.

111Arturo C. Porzecanski, From Rogue Creditord to Rouge Debtors: Implications of Argentina's Default, 6 Chi. J. Int'l L. 311, 324 (2005). Argentina never recognized the interest on the old bonds that accrued during 2002 and 2003. Id.

112The secondary market prices were taken from Bloomberg and Thomson Reuters-Datastream.

113For bonds issued in pesos and euros, we convert all interim cash flows to U.S. dollars at the free market rate at the time that they were paid and maintain that position in dollars throughout the sample.

114These figures are also consistent with Exhibit 4. For example, $100 x 7.01 / 1.67 = $419.76.

115See infra note 152 and accompanying text.

116A cursory reading of these figures would suggest that the haircut was sixty-three percent, which differs from the 73.4% figure mentioned for Argentina’s 2005 exchange. See Cruces & Trebesch, supra note 28. Note that the former figure uses the face value of the old debt, while the latter one uses the present value of the old debt. See id. at 88–89 (explaining the difference between these two haircut concepts).

117To sum the value of the holdings in different currencies, we convert the value of the securities denominated in pesos and euros to dollars using the free market exchange rates prevailing at the end of each quarter.

118The end of sample prices correspond to November 24, 2015.

119See Cruces, Problema de los Holdouts, supra note 28, at 69.

120See infra note 155 and accompanying text.

121See supra Part I.C.

122See Levine, Argentina’s Bond Fight, supra note 16; see also Charles Blitzer, Guest Post: Argentina’s Debt Offer—Don’t Pop the Champagne Just Yet, Fin. Times: FT Alphaville (Feb. 12, 2016), http://ftalphaville.ft.com/2016/02/12/2153239/guest-post-argentinas-debt-offer-dont-pop-the-champagne-just-yet/ (describing dramatic differences in haircuts among creditors in Argentina’s initial settlement offer).

123See Bratton & Gulati, supra note 23, at 10–13 (contrasting corporate and sovereign debt).

124See Anna Gelpern, A Skeptic’s Case for Sovereign Bankruptcy, 50 Hous. L. Rev. 1095, 1098 (2013).

125See id.; see also supra notes 23–24 and accompanying text.

126See Pottow, supra note 11, at 227 (“Judges are nevertheless asked to make important policy decisions in one-off interventions that occur every few years, a task to which they are poorly suited.”); see also Gelpern, Contract Hope, supra note 40, at 133 (describing sovereign debt litigation as a source of “hard cases prone to make bad law”).

127Sovereign debt lacks the crucial restructuring tools of a formal insolvency system: automatic stays, cram-downs, debtor-in-possession financing, clear priority rules, etc. See Lee C. Buchheit & G. Mitu Gulati, Sovereign Bonds and the Collective Will, 51 Emory L.J. 1317, 1347 (2002) (explaining collective action dilemmas due to the lack of cram-down mechanisms in sovereign debt workouts).

128See Gelpern, Contract Hope, supra note 40, at 134 (predicting dysfunctional results in NML litigation); see also Park & Samples, supra note 98 (explaining dysfunctional outcomes and “rogue” trends in sovereign debt generally).

129See W. Mark C. Weidemaier & Mitu Gulati, A People’s History of Collective Action Clauses, 54 Va. J. Int’l L. 52, 56 (2013). Sovereign finance transitioned from syndicated loans to bonds with the Brady Plan. See id. Increasingly numerous and divergent creditors have created a vastly more complicated landscape for sovereign debt restructurings. See id. “Bondholders, by contrast, are widely dispersed, may have divergent interests, and are less subject to regulatory pressure.” Id. at 56. For a contrasting view, see Ran Bi, Marcos Chamon & Jeromin Zettelmeyer, The Problem That Wasn’t: Coordination Failures in Sovereign Debt Restructurings (Int’l Monetary Fund, Working Paper No. 11/265, 2011), https://www.imf.org/external/pubs/ft/wp/2011/wp11265.pdf.

130See supra notes 16–18.

131See Gelpern, Contract Hope, supra note 40, at 133.

132Weidemaier & Gelpern, supra note 11, at 190 (“Courts can inconvenience sovereigns; they cannot make them pay.”); Foster, supra note 44, at 670.

133Between 1997 and 2013, the average of creditor participation in sovereign restructurings was approximately ninety-five percent. See Duggar, supra note 29, at 1. Out of thirty-four debt exchanges, all but two—Dominica in 2004 and Argentina in 2005—exceeded ninety percent participation during that time. See id. at 8.

134See Jill E. Fisch & Caroline M. Gentile, Vultures or Vanguards?: The Role of Litigation in Sovereign Debt Restructuring, 53 Emory L.J. 1047, 1101–16 (2004) (addressing the role of holdout litigation in sovereign debt markets).

135See Robin Wigglesworth, Vulture Funds Come Under Sovereign Fire, Fin. Times (Apr. 24, 2013, 10:09 AM), http://www.ft.com/intl/cms/s/0/41a633ae-ab3d-11e2-8c63-00144feabdc0.html#axzz2rQrOZWe2 (referring to the “risky and difficult” nature of the business model).

136See W. Mark C. Weidemaier, Remarks Made Before the United Nations General Assembly Ad-Hoc Committee on Sovereign Debt Restructuring (Feb. 5, 2015) in Univ. N.C. Legal Studies Research Paper Series, Feb. 10, 2015, at 1.

137Park & Samples, supra note 98, at 43 (highlighting the shortcomings of contractual and institutional responses to problems in sovereign debt).

138A majority modification clause (a specific type of CAC) enables a percentage of bondholders—often seventy-five percent—to make restructuring decisions that bind one hundred percent of the entire bond issuance. For a full discussion of CACs and their evolution in sovereign debt, see Weidemaier & Gulati, supra note 129, at 6.

139See, e.g., NML Capital, Ltd. v. Republic of Argentina, 727 F.3d 230, 247–48 (2d Cir. 2013), cert. denied, 134 S. Ct. 2819 (2014).

140See Declaration of Stephen Choi at 8, NML Capital, Ltd. v. Republic of Argentina, No. 08-CV-6978, 2012 WL 7656066 (S.D.N.Y. Nov. 16, 2012) (noting that “25.3% of New York law governed bond issuances with a maturity date of 2013 or later employ [unanimous action clauses] for changes to payment related terms.”).

141See Yan Liu, et al., Int’l Monetary Fund, Strengthening the Contractual Framework to Address Collective Action Problems in Sovereign Debt Restructuring 19 (2014), https://www.imf.org/external/np/pp/eng/2014/090214.pdf (comparing types of CACs).

142Key improvements include guidance for drafting superior CACs as well as a pari passu clause that expressly preempts the ratable payment interpretation. See Int’l Capital Mkt. Ass’n, supra note 13; see also Anna Gelpern, ICMA CACs, New York Edition—Vietnam!—and More Un-Boilerplate, Credit Slips (Nov. 18, 2014, 11:13 AM), http://www.creditslips.org/creditslips/2014/11/icma-cacs-new-york-edition-vietnam-and-more-un-boilerplate.html (discussing implementation of ICMA-proposed revisions in recent issuances by Kazakhstan, Vietnam, and Mexico).

143See Weidemaier, supra note 136, at 1 (“Still, it is fair to say that, despite two centuries of attempted reform, many believe the contract template for sovereign lending remains flawed.”); see also Martin Guzman & Joseph E. Stiglitz, Creating a Framework for Sovereign Debt Restructuring that Works, in Too Little, Too Late: The Quest for Resolving Sovereign Debt Crises 15–20 (Martin Guzman, José Antonio Ocampo & Joseph E. Stiglitz eds., 2016) (underscoring the limitations of contractual solutions).

144See supra notes 52–54 and accompanying text.

145See Sung Hui Kim, Pari Passu: The Nazi Gambit, 9 Cap. Mkts. L.J. 242, 243 (2014) (observing that broad injunctions gave the pari passu clause teeth, “a concrete remedy that could be used by the holdout creditor to induce sovereign debtors to pay”).

146See Weidemaier & Gelpern, supra note 11, at 213–17 (explaining collateral costs for third parties stemming from NML injunctions).

147See Schumacher, Argentina Implications, supra note 35, at 146.

148See Cross, supra note 11, at 136–37 (citing external costs for third parties caused by Argentina’s 2014 default); Weidemaier & Gelpern, supra note 11, at 210–18 (articulating fundamental problems associated with enforcing sovereign debt through injunctions); An Illusory Haven, Economist (Apr. 20, 2013), http://www.economist.com/news/finance-and-economics/21576391-what-lessons-should-investors-learn-argentine-and-greek-restructurings (“But the battle has inflicted collateral damage on a host of third parties, from Ghanaian ports to American custodian banks.”).

149Vivianne Rodrigues & John Paul Rathbone, Argentina Bond Investors Challenge Long Arm of US Law, Fin. Times (July 3, 2014), www.ft.com/intl/cms/s/0/471b5be2-02c7-11e4-a68d-00144feab7de.html; Matt Levine, Argentina’s Bond Mess Gets Slightly More Complicated, Bloomberg (Feb. 13, 2015), http://www.bloombergview.com/articles/2015-02-13/argentina-s-bond-mess-gets-slightly-more-complicated [hereinafter Levine, Argentina’s Bond Mess].

150See Levine, Argentina’s Bond Mess, supra note 149, at 3–4.

151See infra Part IV.C.

152See Levine, Argentina’s Bond Fight, supra note 16; see also Benedict Mander & Robin Wigglesworth, ‘Holdout’ Slams Argentine Debt Offer to End Financial Blockade, Fin. Times (Feb. 12, 2016, 5:39 PM), http://www.ft.com/intl/cms/s/0/d17fb6b4-d1a3-11e5-986a-62c79fcbcead.html#axzz3zzJjx1DC (reporting on resistance amongst certain holdout creditors to Argentina’s initial settlement offer).

153See Blitzer, supra note 122.

154Id.

155From 1976 to 2010, 120 holdout suits were filed against sovereigns in the United States, England, and the World Bank’s International Centre for the Settlement of Investment Disputes. See Schumacher et. al., Sovereign Defaults in Court, supra note 26, at 10–12. Of the sixty-five cases that reached a final outcome, holdout claims were paid in full in eleven of those cases whereas negotiated settlements were reached in forty-eight. See id.

156Transcript of Hearing at 23–24, NML Capital, Ltd. v. Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. May 29, 2015) (No. 14-CV-8601 (TPG)) [hereinafter May Hearing] (“We are dealing with large amounts of money. But negotiations and settlements have occurred before where large amounts of money are involved, and that can occur again. The Court, of course, cannot order a settlement. But I want to say in this courtroom before this entire group that the way to ultimately resolve this litigation must come through settlement. Unless the Republic wants to pay 100 percent of the judgments, which the Republic doesn’t indicate that it will do, then there has to be a settlement.”).

157See Buckley, supra note 25, at 225, 229; Schumacher et al., Sovereign Defaults in Court, supra note 26.

158Argentina settled an expropriation dispute with Repsol with a bond issuance. Stanley Reed & Raphael Minder, Repsol in $5 Billion Settlement with Argentina, N.Y. Times (Feb. 25, 2014), http://www.nytimes.com/2014/02/26/business/international/repsol-said-to-reach-settlement-with-argentina.html?_r=0.

159See supra notes 155–156.

160May Hearing, supra note 156, at 23–24; Transcript of Hearing at 9, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Aug. 8, 2014) (No. 08-CV-6978 (TPG)).

161Transcript of Hearing at 11, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Aug. 21, 2014) (No. 08-CV-6978 (TPG)) [hereinafter August 21, 2014 Hearing]; Transcript of Hearing at 29, NML Capital, Ltd. v. The Republic of Argentina, 144 F. Supp. 3d 513 (S.D.N.Y. Oct. 28, 2015) (No. 14-CV-8601 (TPG)).

162Previous work by scholars has articulated more fundamental issues related to injunctions as an enforcement tool in sovereign debt. See Weidemaier & Gelpern, supra note 11, at 213–17 (explaining public interest considerations and costs for innocent third parties); see also Cross, supra note 11, at 135–40 (critiquing the NML injunctions from a Foreign Sovereign Immunities Act perspective).

163Argentina’s holdout bonds lack CACs. Mario Blejer, Argentina’s Deal with the Holdouts is a Mixed Blessing, Fin. Times (Mar. 31, 2016), https://www.ft.com/content/db6779d6-f729-11e5-96db-fc683b5e52db. In sovereign bonds issued under New York law prior to 2003, unanimous action clauses were standard. See Stephen J. Choi & G. Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds, 53 Emory L.J. 929, 932–33 (2004).

164See supra Part I.B (detailing the allocation of holdouts across Argentina’s defaulted bonds).

165See Pottow, supra note 11, at 225, 229; Gelpern, Contract Hope, supra note 40, at 132–33.

166See Weidemaier & Gelpern, supra note 11, at 199–200 (discussing injunctive relief as a remedy in sovereign debt litigation).

167See Gelpern, Love and Exhaustion, supra note 18.

168See id. at 1; see also Mark Weidemaier, Argentina’s Settlement Negotiations and Lifting the Injunction, Credit Slips (Feb. 9, 2016, 8:56 AM), http://www.creditslips.org/creditslips/2016/02/argentinas-settlement-negotiations-and-lifting-the-injunction.html.

169The NML court has pointed to collateral damage to “very innocent third parties” while advocating for a settlement between the parties. August 21, 2014 Hearing, supra note 161, at 11.

170See Miller & Thomas, supra note 36, at 1492–93, 1499–501 and accompanying text; see also EM Ltd. v. Republic of Argentina, 131 F. App’x 746, 747 (2d Cir. 2005); supra text accompanying note 37.

171See supra notes 123–129 and accompanying text.

172BVAL Provides Objective, Independent & Defensible Evaluated Pricing, Bloomberg for Enterprise, https://www.bbhub.io/solutions/sites/8/2015/10/BVAL-Evaluated-Pricing-fact-sheet.pdf (last visited Nov. 13, 2016).