COLUMBIA LAW SCHOOL

FALL 2012 TUESDAY FACULTY WORKSHOP SERIES

Tuesday, December 4, 2012

Robert Scott & Mitu Gulati

Present

“THE THREE AND A HALF MINUTE TRANSACTION: BOILERPLATE AND THE LIMITS OF CONTRACT DESIGN”

Case Lounge, Jerome Greene Hall Catered Lunch to begin at 12:00 p.m. Presentation to begin at 12:15 p.m. Discussion to follow until 1:10 p.m.; informal conversation until 1:30 p.m.

Electronic copies of the papers are available online at http://www.law.columbia.edu/faculty/fac_resources/faculty_lunch/fall12 Please contact Char Smullyan for special dietary requests: ext.4-3259, [email protected] For extra copies or other questions, please contact Rachel Jones: 4-7594, [email protected]

Date Presenter Paper

September 4 Michael Gerrard Expedited Approval of Energy Projects: Assessing the Forms of Procedural Relief 11 Carol Sanger Father and Fetuses: What Would Men Do?

18 Charles Fried The Ambitions of Contract as Promise Thirty Years On 25 Franz Mayer The Law and the Euro Crisis October 2 Legal Theory Workshop TBA 9 Curtis Milhaupt “We are Happy": A Glimpse of North Korea under Kim III

16 Daniel Bethlehem Principles Relevant to the Scope of a State's Right of Self - Defense Against an Imminent or Actual Armed Attack by Nonstate Actors 23 Sudhir Krishnaswamy Implementing the Constitution: Directive Principles as Underenforced Norms 30 Hanoch Dagan Private Law Pluralism and the Rule of Law November 6 Academic/ Curriculum N/A Lunch 13 Scott Hemphill and Tim Parallel Exclusion Wu (Thursday Workshop) 20 Fred Schauer Legal Realism Untamed 27 Legal Theory Workshop - TBA Theodore Ruger December 4 Panel Celebrating THE THREE AND A HALF MINUTE TRANSACTION : Robert Scott & Mitu BOILERPLATE AND THE LIMITS OF CONTRACT DESIGN Gulati's new book

THE THREE AND A HALF MINUTE TRANSACTION:

BOILERPLATE AND THE LIMITS OF CONTRACT DESIGN

Mitu Gulati & Robert E. Scott

Abstract

Theory tells us that the lawyers who draft standard-form contracts will respond to an erroneous court interpretation of a boilerplate contract term by revising the standard formulation of the clause or otherwise clarifying the ambiguity. Lawyers, it is assumed, have the incentive to protect clients from the risk that other courts may adopt the disfavored interpretation or that the apparent ambiguity will, in any event, produce future costly litigation. Reality is often different. Boilerplate terms in standard contracts are often sticky. Many of the theories explaining why standard contract terms are resistant to innovation focus on the efficiency benefits of market-wide standardization. Others look to economic or psychological factors that deter the production of innovative terms. In this book, we use both qualitative and quantitative data to explore the reasons for the phenomenon of stickiness in sovereign bond contracts. We focus on a novel judicial interpretation of an obscure clause in cross-border financial contracts--the pari passu clause—that rattled the chandeliers of international finance. One might have expected the practicing bar to quickly clarify their forms before the heresy could spread and gain traction. But that didn’t happen. In over 90% of the contracts subsequently issued, no attempt was made to clarify the imprecise language of the clause. None of the extant theories of stickiness explain what we find. Rather, the story that emerges is about the modern big law firm: the financial pressure on big firms to maintain a high volume of transactions contributes to an array of conflicts that that deter innovation and are largely hidden from the individual lawyers charged with drafting responsibility—conflicts that are hidden in part by the myths that the members of this legal community collectively tell themselves about the origins of boilerplate terms whose meaning has been lost in time.

Draft – January 6, 2012 Contents

TABLE OF FIGURES ...... ERROR ! BOOKMARK NOT DEFINED .

INTRODUCTION ...... 5

CHAPTER I: A STORY OF STICKY BOILERPLATE BEGINS IN BRUSSELS ...... 13

Elliott v. Peru : Brussels, September 26, 2000 ...... 16

CHAPTER II: THE SOVEREIGN BOND CONTRACT ...... 23

The Sovereign Debt Contracting Problem ...... 23

Two Different Perspectives on the Problem ...... 26

The Provisions of the Contract ...... 28

Negotiating Over the Contract Terms ...... 32

Optimal Changes to Pari Passu after Elliott ...... 34

CHAPTER III: THEORIES EXPLAINING THE STICKINESS OF CONTRACT BOILERPLATE ...... ERROR ! BOOKMARK NOT DEFINED .

Learning Externalities ...... Error! Bookmark not defined.

Network Externalities ...... Error! Bookmark not defined.

Negative Signaling ...... Error! Bookmark not defined.

A Penalty for Remedial Measures ...... Error! Bookmark not defined.

Contract Routines ...... Error! Bookmark not defined.

Nervous Nellies and Herd Behavior ...... Error! Bookmark not defined.

Free Riders ...... Error! Bookmark not defined.

Endowment Effects (and Related Cognitive Biases) ...... Error! Bookmark not defined.

One Can’t Fix What One Doesn’t Understand ...... Error! Bookmark not defined.

Stickiness as Myth ...... Error! Bookmark not defined.

CHAPTER IV: THE IMPORTANCE OF ELLIOTT : THE OFFICIAL STORY ...... ERROR ! BOOKMARK NOT DEFINED . The New Litigation Risk ...... Error! Bookmark not defined.

CHAPTER V: MARKET RESPONSES ...... ERROR ! BOOKMARK NOT DEFINED .

The Sovereign Market ...... Error! Bookmark not defined.

The Issuers ...... Error! Bookmark not defined.

The Lawyers ...... Error! Bookmark not defined.

The Bankers ...... Error! Bookmark not defined.

CHAPTER VI: THE FAITHFUL LAWYER ...... ERROR ! BOOKMARK NOT DEFINED .

Learning Externalities and Free Riders ...... Error! Bookmark not defined.

Network Externalities: Vertical not Horizontal ...... Error! Bookmark not defined.

Negative Signals: Empirical Realities and Shoulder Shrugs ...... Error! Bookmark not defined.

Hindsight Bias: Bemusement and Annoyance ...... Error! Bookmark not defined.

CHAPTER VII: THE IMPERFECT AGENT ...... ERROR ! BOOKMARK NOT DEFINED .

Mechanization, Commoditization, and the Golden Age of Lawyering ...... Error! Bookmark not defined.

Anti Innovation Bias and Herd Behavior ...... Error! Bookmark not defined.

The Barriers Caused by Legal Uncertainty (Can you revise a clause if you don’t know what it means?) ...... Error! Bookmark not defined.

Big Brother Will Fix the Problem ...... Error! Bookmark not defined.

Contract Terms as as Symbols (Herein of the Endowment Effect) ...... Error! Bookmark not defined.

Stickiness is a Myth (And, in any event, the problem was solved by DOJ briefs in 2004) ...... Error! Bookmark not defined.

CHAPTER VIII: RITUALS AND MYTHS ...... ERROR ! BOOKMARK NOT DEFINED .

Finding Meaning in Stories ...... Error! Bookmark not defined.

Tales of Pari Passu ...... Error! Bookmark not defined.

The First Story: Pari Passu as Anti-Earmarking Device ...... Error! Bookmark not defined.

The Second Story: It was Copied from Corporate Deals . . . Inadvertently ...... Error! Bookmark not defined. The Third Story: Gunboat Diplomacy and the Promise of Customs Revenues ...... Error! Bookmark not defined.

The Fourth Story: No More Haircuts ...... Error! Bookmark not defined.

The Fifth Story: From Secured Loans to Involuntary Subordination ...... Error! Bookmark not defined.

CHAPTER IX: THE HUNT FOR PARI PASSU ...... ERROR ! BOOKMARK NOT DEFINED .

The Inadvertent Copying Story Revisited ...... Error! Bookmark not defined.

No More Haircuts… Smaller Banks as the Source of Pari Passu ...... Error! Bookmark not defined.

The Anti-Earmarking Story ...... Error! Bookmark not defined.

The Buchheit-Pam Story ...... Error! Bookmark not defined.

The Gunboat Diplomacy Story ...... Error! Bookmark not defined.

CHAPTER X: THE AGENCY COSTS OF BIG LAW ...... ERROR ! BOOKMARK NOT DEFINED .

Failure of the “Faithful Lawyer” Theories ...... 36

“Imperfect Agent” Theories ...... 39

Complications, Biases and Caveats ...... 47

Tentative Generalizations ...... 59

CHAPTER XI: EPILOGUE ...... 61

Ecuador 2008-09 ...... 62

The Eurozone Crisis: 2010-11 ...... 64

Argentina; September 28, 2011 ...... 65

APPENDIX ...... 72

INTRODUCTION Many of us who purport to study and explain human behavior secretly dread becoming the punch line of the story about the jumping frog. In that tale, a scientist sets out to measure how far a bullfrog can jump. On the first day of the experiment, the scientist prods the creature’s posterior while commanding “jump frog, jump”. The frog jumps nine feet. The scientist records in his notebook: “Day One -- frog jumps 9 feet.”

On the second day, the scientist cuts off one of the frog’s legs. When prodded with the instruction “jump frog, jump”, the frog jumps five feet. The notebook entry for that day reads: “Day Two: frog with three legs, jumps five feet.”

Day Three is a repeat of the experiment, but the poor frog now has two of its legs surgically removed. “Jump frog, jump.” Notebook entry: “Day Three: with two legs, frog jumps 18 inches.”

On Day Four, the frog loses yet another leg in the interests of science. “Jump frog, jump.” Notebook entry: “Day Four: with one leg, frog jumps 1.5 inches.”

The scientist removes the last of the poor frog’s legs on Day Five and issues the by-now familiar command, “jump frog, jump.” No response. The scientist says in a louder and more imperious voice, “jump frog, jump!” And again, “jump frog, jump!!” The animal still does not stir.

Notebook entry: “Day Five: frog went deaf.”

Attempting to deduce the motivation or rationale for human behavior based solely on the record of that behavior is not so different from the study of jumping frogs. The observer can get it badly wrong. Law professors can study the practices of lawyers in preparing the tens of thousands of commercial contracts that are signed each day. From the textual evidence of those contracts, many of them in utterly standard boilerplate form, academics might ascribe rational motivations and objectives to the drafters. Fair enough, you might say. After all, these documents are prepared by seasoned, highly paid, members of what still styles itself a learned profession. The contracts themselves are intended to embody legally binding commitments of their clients, enforceable if necessary with all the majesty of the law. It is hardly the place where one would expect to encounter the casual, the unpremeditated or the indecipherable text.

And then, every once in a while, the frog goes deaf.

In the story we tell in this book, a novel (some might say singular) judicial interpretation of an obscure clause in cross-border financial contracts--the pari passu clause--briefly rattled the chandeliers of international finance. One might have expected the elite practicing bar to have quickly clarified their forms so as to discredit what they universally believed was an heretical interpretation of this boilerplate provision before the heresy could spread and gain traction. But that didn’t happen. In over 90% of the contracts subsequently issued, no attempt was made to clarify the imprecise language of the clause. To be sure, over time a small number of contracts did incorporate new provisions that had the effect, if not the clear purpose, of reversing the aberrant interpretation. But the fact that these changes were limited to a small minority of the outstanding contracts only served to deepen the puzzle. In this book, we try to unpack the puzzle of why the response that standard theory would predict did not occur. The story that emerges is partly about why these financial contracts did not change despite the on-going risk of other courts or adjudicative bodies adopting the same destabilizing interpretation.

At bottom though, this is a story of forgetfulness. It is a story of how a remarkably unconfiding contractual provision was introduced into international financial contracts over a hundred years ago (presumably to deal with a risk or contingency that arose at that time). This contractual provision was promptly absorbed into the lumpish boilerplate of such contracts and then came to be replicated, thousands upon thousands of times, even while the knowledge of its origin and purpose insensibly faded from the minds of its remote drafters. If anything, the increase in the popularity of this clause in international financial contracts seems to have been inversely related to market understandings of its meaning. As the clause became more widely used over the past century, shared understanding of its intended meaning actually diminished. So much so that this clause appears today in every one of the cross-border documents that form the subject of our study – frequently displayed prominently on the front page -- and almost no one seems to understand what the clause means. This is also a story about the organic life form known as a standard commercial contract and about how such documents pass relatively untouched through the hands of generations of lawyers much like a seed can pass unharmed through the intestinal tract of a bird. The story can be told from the standpoint of basic human psychology; novelty sparks curiosity, repetition stupefies it. Or it can be told from the perspective of a legal profession in which new lawyers are expected to learn the lore of their craft from their elders in a tutorial, master/apprentice system that no longer exists in most major law firms. Or the tale can be brought down to the individual lawyer working on a financial document at 3:00 a.m., and who briefly scratches her head over the significance of a pari passu representation in her agreement, only to yawn and pass on, comforted by the thought that someone at the firm must know why it is there; the document is, after all, the firm’s standard form for this type of deal.

Finally there is the obvious question: If the pari passu clause could have lain dormant, unchallenged for over a century in cross-border financial contracts, how many other boilerplate clauses might similarly have outlived the memory of their origins and purpose, making them prime candidates for creative interpretations by highly motivated litigants?

The origins of this project lie in an “incubator lunch” at the University of Virginia Law School roughly six years ago. Those lunches, the brainchild of our friend, George Triantis, were supposed to be informal settings where faculty who were thinking about possible research projects could present their ideas to a group of colleagues who would be able to shed light on whether the idea was worth pursuing. One of us, Gulati, who had already spent a number of years studying sovereign debt contracts pitched the idea that the standard models of contract interpretation might not work well in situations where the parties to the contract, because they were using historical boilerplate, had no idea what the provisions meant and nor did anyone else. The example used to illustrate the problem to the group was the pari passu clause in sovereign bonds, which no one seemed to understand, but was used everywhere. The lunch group was skeptical about the efficacy of the project—it seemed at once too ambitious and too obvious. Developing a new theory of contract interpretation for boilerplate contracts did not strike them as particularly interesting.

Scott, however, was interested in the example itself—and specifically, the question of why it was that these sophisticated and highly paid lawyers, working at the most elite firms, failed to alter a contract term that not only posed a litigation risk to their clients, but that no one understood. This was a phenomenon that Scott had also noticed in his experience working as an expert on a variety of contract issues in different settings involving sophisticated lawyers and large-scale commercial disputes. The phenomenon flew in the face of conventional wisdom, at least that of the economic variety, about standard form contract terms. To quote Smith and Warner’s classic 1979 article on financial contracting:

[Boilerplate contract terms] take their current form and have survived because they represent a contractual solution which is efficient from the standpoint of the firm . . . Harmful heuristics, like harmful mutations, will die out.” 1

At odds with the quote above, the pari passu example seemed to indicate that inefficient and harmful contract provisions could persist for long periods of time, even in the most sophisticated of financial markets.

The two of us had also long been intrigued by the contrast between the standard justification given to law students as to why they are asked to read hundreds of cases in their first year contracts classes and the reality of law practice. At the schools we were teaching at then, the majority of our students were hoping to go into transactional practices, where they would be drafting and negotiating contracts. But only a miniscule fraction of contracts are ever litigated, and transactional lawyers are almost never involved in the litigations. Nevertheless, contract law is taught to all students almost exclusively through case law, whether they plan to be transactional lawyers or litigators. If asked why this is so, the explanation we generally give, and that we suspect others also give, is that “students study past disputes in order to draft contractual provisions that will avoid similar disputes in the future”. 2

Yet, neither one of us had seen much evidence of this model of the ideal transactional lawyer engaged in a dynamic process of regularly reading cases and incorporating that learning into novel innovations in subsequent contracts. Some of the transactional lawyers we knew didn’t appear to have looked at a case in years. The task of reading cases seemed to be the province of the litigators, while the thinking about contract drafting remained with the transactional lawyers. In theory, the two groups might be specializing and transferring information across the artificial boundary that separated them. However, we had seen little evidence of interaction among transactional lawyers and litigators, let alone a process by which they collaborated in R&D on contract design.

Several hours later, long after the others had left the lunch table, the two of us realized that we could not suggest a plausible answer to why the clause had neither been improved nor, better yet,

1 Clifford W. Smith & Jerald B. Warner, On Financial Contracting , 7 J. FIN . ECON . 117, 123 (1979).

2 th ROBERT E. SCOTT & JODY KRAUS , CONTRACT LAW AND THEORY vii (4 ed. 2007). just deleted. The failure to revise a contract term that, owing to an aberrant interpretation, now carried a non-trivial litigation risk was completely inconsistent both with the theoretical models of how sophisticated contract drafters behaved and with the dynamic model of case law serving as the basis for contract drafting and innovation. We assumed there had to be a rational explanation for the fact that the frog didn’t jump. Our speculation was that we would find some form of an “agency problem” driving the phenomenon: lawyers were failing to represent their clients’ interests adequately owing to recognizable conflicts of interest. Perhaps, for example, lawyers were reluctant to admit that they had failed on past deals to exert appropriate efforts on behalf of the clients to remove the litigation risk that ultimately materialized. Agency problems such as this are familiar topics in the legal and economics literature. They illustrate the divergence of interests between the principals (the sovereigns who issue the bonds and the underwriters who market them) who want to maximize the joint value of the contract and their agents (the lawyers) who draft the contract terms that ideally are designed to achieve the principals’ objectives. Whether owing to this or other causes, we believed that we would be able to solve the puzzle quickly. Surely, it would only take a few months to find the answers to our questions and to publish the results.

We began by gathering information along two different dimensions. First, we collected data on the contracts themselves—to see whether what we had perceived by casual observation (that the contract provisions had not been revised to fix the offending provision) was actually the case for a large dataset. Second, we asked a sample of the senior New York lawyers who worked on sovereign debt contracts whether we could speak to them about our puzzle. In our original research plan, we proposed to interview 25-30 lawyers in New York and to examine 50-75 sovereign debt contracts over the period 2000-2005.

Our early optimism turned out to be misplaced. No coherent answers could be gleaned from either the first set of contracts or the interviews. Instead of a straight-forward agency problem or other market failure explanation, these hard-nosed Wall Street lawyers told us stories about rituals, talismans, alchemy, the search for the Holy Grail, and Zeus. Frustrated, we assumed that we simply had not talked to enough people or the right people or looked at enough sovereign debt contracts. As we write this Introduction, more than six years after we began, we have examined over 1,500 sovereign debt contracts, covering the period 1820-2010 and conducted more than 200 interviews. As we kept unpacking the story, it became more fascinating even as a straightforward conflict of interest hypothesis proved ever more elusive. No single agency problem emerged from the data; at least not in a fashion that we could assert with confidence. To be sure, we recognized that the lawyers we talked to would be unselfconscious about the array of possible conflicts that might explain the failure to amend or eliminate a troublesome clause, and also would be quite ignorant of any theoretical explanations for the faithlessness of agents. Nevertheless, the explanations we were given for why a troublesome clause was allowed to remain in subsequent contracts were both diverse and conflicting. Moreover, we determined from our research that these explanations often rested on myths that were based on quite unsupportable factual premises.

Over time, a messy but more consistent hypothesis began to emerge: there are many overlapping sources of agency conflicts in contemporary big firm law practice—at least law practice of the sort represented by the firms that draft these contracts and thus have had to grapple with the pari passu issue. The myths that we were told can be best understood as ways in which the lawyers were able to deflect what would otherwise be obvious failures to correct errors in the formulation of historic boilerplate. “Three and a half minutes” is one explanation that was candidly offered to us by a lawyer who sought to explain the trade off between the time it took to “draft a new contract” and the effort costs of redesigning boilerplate that was widely used and had been part of the standard form contract for many years. But “three and a half minutes” is also a metaphor for a business model that relies on herd behavior, fails to provide incentives for innovation and thus rises and falls on volume-based, cookie-cutter transactions. To be sure, we find that in a few instances individual lawyers, who appreciated the litigation risk, did adapt by redesigning sovereign debt contracts (often by adding new terms rather than correcting perceived errors in existing terms). But our evidence suggests that in the great majority of firms, lawyers rely on the herd and on their myths: the returns to the firm in terms of volume transactions outweigh the present value of the risk. 3 This is despite the fact that a social planner seeking to maximize the joint interests of lawyers and their clients would likely choose a different business model. In short, we conclude that social welfare is less than it would be under a different regime even though the private benefits of volume transactions over careful design may explain the firm behavior that we see.

The book divides informally into three parts. In the first seven Chapters we focus on the question with which we began; the puzzle over the legal market’s failure to respond to a litigation event by clarifying the ambiguous provision that was the source of the litigation. Chapter I describes the litigation that began the story: Elliott Associates, a vulture fund, attached payments from the government of Peru intended for those bondholders who had agreed to a restructuring of their debt, arguing that the pari passu clause entitled them to a pro rata share of the payments. In Chapter II we lay the groundwork for what follows by describing the archetypical sovereign bond transaction: We focus specifically on the purpose and meaning of the standard form contract terms at issue in the Elliott litigation and the

3 There is support in economic theory for the proposition that the time and effort involved in drafting, revising and negotiating contract terms will lead the contract writer to choose “satisficing” rather than “optimal” or “efficient” contract terms. See Patrick Bolton & Antoine Faure-Grimaud, Satisficing Contracts 77 REV . ECON . STUD . 937 (2010) respective roles of the sovereign (who issues the bonds) and the underwriter (who markets the bonds) in the drafting of those contract terms. Chapter III sets out the various theories from the academic literature that might explain why the pari passu clause would remain unchanged following the Elliott litigation. Chapters IV and V describe the nature and extent of the litigation risk generated by the Elliott litigation and set out the basics of the qualitative and quantitative datasets that we use to analyze the aftermath of the case. We then turn in Chapters VI and VII to the explanations offered by our respondents for why the clause was not revised. In these chapters our emphasis is on the qualitative—on how our respondents explained the failure to revise a contract term that virtually no one seemed to understand. We juxtapose their explanations with the theoretical explanations for the “stickiness” of standard-form terms in commercial contracts and with the data on the contracts themselves, including the evidence of price effects owing to differences in litigation risk in particular bond issues.

Our respondents resisted our framing of the problem as a search for explanations for the stickiness of contractual boilerplate. In response to our questions, they frequently attempted to reframe the discussion by answering questions that we weren’t asking. They appeared to be suggesting that we were asking the wrong question. The right inquiry (the “quest,” to quote one of them) was to unearth the historical origins of the offending provision. Once we discovered the true origins of the clause all would become clear. We took the bait. Chapters VIII and IX report both on the stories of historical origins of pari passu that respondents told us and on our attempt to find those historical origins using what we believe is the largest existing academic database of sovereign debt contracts. We think we managed to come close to determining where the clause originated. Quite possibly, the clause originated in a bond issued by the Republic of Bolivia in 1870, which was issued to finance the attempt of an American adventurer, Colonel George E. Church, to connect Bolivia to the Atlantic Ocean. Colonel Church’s expedition was a failure (among other things, as one commentator reported, he ran into “savage Indians, some of them even cannibals”, along the way4).

If we had expected to be celebrated when we returned with our tales of Colonel Church and the origins of pari passu , we were to be disappointed (admittedly, the story of Colonel Church notwithstanding, we had found ourselves no closer to discerning the original meaning of the clause). Our respondents exhibited a singular lack of interest in learning about the origins of the clause, despite their prior insistence that this was where the answers lay.

4 See NEVILLE CRAIG , RECOLLECTIONS OF AN ILL FATED EXPEDITION 35, 53 (1907). In Chapter X, therefore, we turn to an analysis and interpretation of what we have learned. We conclude that the evidence supports the hypothesis that the financial pressure on big firms to maintain a high volume of transactions contributes to an array of conflicts that are largely hidden from the individual lawyers charged with drafting responsibility—conflicts that are hidden in part by the myths that the members of this legal community collectively tell themselves. Chapter XI provides an epilogue to the story: the pari passu drama has begun a second act in 2011 with the sovereign debt crisis in the Eurozone and in litigation against Argentina in federal court in New York, where a district court judge gave Elliott Associates a second victory ten years after their first one.

*****

There is a long list of those we should thank for their advice, insights and patience in bearing with us as we struggled to gain a foothold on this problem. Our biggest debt is to the many lawyers, bankers and regulators who gave up valuable portions of their days to talk to us at length about what undoubtedly seemed an utterly obscure project to many of them. For reasons of confidentiality, we cannot thank them individually. Many academic colleagues read portions of this manuscript and offered helpful comments and criticisms. Our apologies for not listing all of them; the list was simply too long. And then there are those who not only suffered through multiple iterations of various portions of this book, but have effectively been our collaborators on this project. To them we owe an incalculable debt. They include, Omri Ben-Shahar, Douglas Baird, Steve Bainbridge, Bobby Bartlett, Michael Bradley, Curt Bradley, Patrick Bolton, Ross Buckley, Devon Carbado, Stephen Choi, John Conley, Jim Cox, Victor Fleischer, Tracey George, Chris Guthrie, Anna Gelpern, Ronald Gilson, Victor Goldberg, Melissa Jacoby, Jerry Kang, Sung Hui Kim, Marcel Kahan, Dan Klerman, William Klein, Jack Knight, Kim Krawiec, Lynn Mather, Stewart Macaulay, Karl Okamoto, Rodrigo Olivares-Caminal, Randal Picker, Eric Posner, Richard Posner, Mark Ramseyer, Robert Rasmussen, Barak Richman, Larry Ribstein, Patrick Shin, Elizabeth Scott, Alan Schwartz, David Skeel, George Triantis, Tom Ulen, David Wilkins, Mark Weidemaier and Mark Wright. We are also grateful to workshop participants at UCLA, USC, the University of Colorado, Brooklyn Law School, the University of Chicago, Duke University, Suffolk University, Haifa University, Hofstra University, UC Irvine, and the University of Georgia. The librarians at the Duke University Library were tireless in their efforts to help us build our datasets and we owe thanks to Jennifer Behrens, Lauren Collins, Dick Danner and Marguerite. Finally, thanks to Briana Brake, Keegan Drake, Kim Lott, De Lira Salvatierra, Tori Simmons, and Guangya Liu for their excellent assistance in both gathering and analyzing the data.

Our two perspectives, relative to the events we describe and the people we interviewed, are significantly different. One of us specializes in the area of sovereign debt finance, does both academic and policy work in the area, worked as an expert on one of the early iterations of the pari passu litigation and has worked closely with some of the characters in this book. The other is a specialist in contracts and commercial transactions and has spent much of the recent past focusing on contract design and the role of transactional lawyers in drafting complex contracts. Before this project he had little prior experience with the world of sovereign debt finance, let alone its lawyers. We would like to think that the combination of our two different perspectives and biases has helped make this a better book.

CHAPTER I: A STORY OF STICKY BOILERPLATE BEGINS IN BRUSSELS

You have to understand the system. No one pays that much attention to the minute details like this. One cannot afford to, if one wants to stay competitive. The firm has a computer program. You know . . . one that a junior associate can go to and plug the relevant parameters into–you know, type of issuance, type of issuer, which side we are representing, etc.,–and the computer generates a standard contract. The firm spent [a large amount] on putting together this system. Associates can now produce a contract for one of these deals in three and a half minutes . This is the future of contracting in these markets.

Interview; Senior Law Firm Partner; June 18, 2007

Had it been your fate to practice law sometime during the relatively uneventful period that elapsed between the end of the Peloponnesian Wars and the first O.J. trial, you would view contract drafting very differently. Every word of every contract you prepared would have called for physical labor on the part of both the author and the typist: hands clutching pens, fingers hitting keys, eyes proofreading text. Not anymore. You have joined the legal profession at a time when the sentence “I drafted the Agreement” is universally understood to mean “I sucked 99.7% of the Agreement off some electronic blob on a word processor.” …. The munificence of these machines is therefore a mixed blessing. On the positive side, they help ensure uniformity throughout the Firm in the drafting of boilerplate provisions, they are the medium through which some carefully considered judgments about contract wording are communicated to succeeding generations of lawyers in the Firm, and they often allow us to meet a client’s expectations about delivery schedules and cost. On the darker side, the ready availability of prefabricated contracts means that new lawyers have less of an opportunity to practice the craft of contract drafting. Far too much of the revealed text is preserved in each new incarnation of the document, mostly because an inexperienced drafter will not be sure why it is there or whether taking it out would help or hurt the document.

Contract Drafting Section; Legal Writing Materials; U.S. law firm, 2008. 5

Long before the days of computers, lawyers were justifying their fondness for standardized contract terms – the paradigmatic “boilerplate” found in virtually all commercial contracts (even those that are carefully negotiated). Paul Cravath, of the Cravath firm, in a speech given in 1916, exhorted his listeners to place their implicit confidence in models and precedents. “The provisions of the modern reorganization agreement and the modern corporate mortgage”, he warned, “are the result of the experience and prophetic vision of a great many able lawyers…. It would indeed be a courageous man who would say that any of the provisions which some of these lawyers have conceived to be wise should be rejected simply because he cannot for the moment think when or how it will become useful. But we suspect that if Mr. Cravath had been asked how lawyers at his firm should respond to a court decision interpreting a contract provision in a manner different from the intent of its drafters, he would have said that it was unacceptable not to immediately clarify the offending provision. In sophisticated markets, theory tells us that the response from the Cravaths of the world to a court decision misinterpreting the meaning of a widely-used standard contract provision should be rapid, if not immediate. To be sure, there may in some cases be legitimate debate about whether or not the newly minted judicial interpretation is erroneous. In such a case, new formulations may evolve more slowly as parties grapple with the language that best represents the shared understanding of the risks the clause allocates. But sometimes the meaning of a standard, widely-used clause is universally accepted in the relevant community. Here an interpretation that differs from that common wisdom is sure to receive

5 LEE C. BUCHHEIT , TOP PEN , Cleary Gottlieb Steen & Hamilton 46-47 (2008) quick attention. Subsequently drafted contracts should amend or clarify the meaning of the clause so that there is no risk that the court’s erroneous interpretation will apply in the future.

Reality is different, particularly when it comes to contracts with boilerplate clauses. 6 Boilerplate clauses—standardized clauses that have been used by rote over long periods of time—often remain unchanged even though a court decision has created uncertainty regarding the clauses’ meaning. In short, boilerplate clauses are sticky: they seem resistant to amendment even when amendment seems desirable. Multiple theories for contract stickiness have been advanced. 7 But the academic understanding of the stickiness phenomenon sits at the point of post hoc rationalizations–understandable in light of the difficulty of empirically testing for the factors that produce resistance to change. Testing for the factors that might induce a revision in boilerplate language is straightforward; one looks to theory for factors that might cause change and then examines the data to see which of the possible causal factors moved in line with the change. 8 But with factors that impede revision, it is less clear how one might empirically test the theoretical conjectures; nothing changes, meaning there are no correlations to observe. 9

6 Many legal scholars have observed, and sometimes sought to explain, the persistence of seemingly inefficient terms. E.g., William W. Bratton, Jr., The Economics and Jurisprudence of Convertible Bonds , 1984 WIS . L. REV . 667, 689; Russell Korobkin, Inertia and Preference in Contract Negotiation: The Psychological Power of Default Rules and Form Terms, 51 VAND . L. REV . 1583 (1998); Claire A. Hill, Why Contracts Are Written in “Legalese, ” 77 CHI .-KENT L. REV . 59 (2001) ; Michael A. Woronoff & Jonathan A. Rosen, Understanding Anti-dilution Provisions in Convertible Securities , 74 FORDHAM L. REV . 129 (2005) ; D. Gordon Smith, The Exit Structure of Venture Capital , 53 UCLA L. REV . 315 (2005); Marcel Kahan, Anti-Dilution Provisions in Convertible Securities , 2 STAN . J.L. BUS . & FIN . 147, 155, 158-59 (1995) ; Lee C. Buchheit, Majority Action Clauses May Help Resolve Debt Crises , INT 'L FIN . L. REV ., Aug. 1998, at 13; Michelle Boardman, Contra Proferentem: The Allure of Ambiguous Boilerplate , 104 MICH . L. REV . 1105 (2005); Karen Halversen Cross, Arbitration as a Means of Resolving Sovereign Debt Disputes , 17 AM. REV . INT ’L ARB . 335 (2006); Steven M. Davidoff, The Failure of Private Equity , 82 S. CAL . L. REV . 491 (2008); Doron Teichman, Old Habits are Hard to Change: A Case Study of Israeli Real Estate Contracts , 44 L. & SOC . REV . 299 (2010); Christopher R. Drahozal & Quentin R. Wittrock, Is There a Flight From Arbitration ? 37 HOFSTRA L. REV . 71, 104 (2008); Claire A. Hill, Bargaining in the Shadow of the Lawsuit: A Norms Theory of Incomplete Contracts, 34 DEL . J. CORP . L. 191 (2009); Deborah Demott, Investing in Work 33 W. NEW ENGLAND L. REV . 497 (2011). Along more skeptical lines, see Bruce H. Kobayashi & Larry E. Ribstein, Choice of Form and Network Externalities, 43 WILLIAM & MARY L. REV . 79 (2001). 7 Marcel Kahan & Michael Klausner, Standardization and Innovation in Corporate Contracting (Or, the “Economics of Boilerplate”), 83 VA. L. REV . 713 (1997); Charles J. Goetz & Robert Scott, The Limits of Expanded Choice: An Analysis of the Interactions Between Express and Implied Contract Terms , 73 CAL . L. REV . 261 (1985); Omri Ben Shahar & John Pottow, On the Stickiness of Default Rules , 33 FLA . ST. U. L. REV . 651 (2006); Stephen J. Choi & G. Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds , 53 EMORY L.J. 929, 947 (2004) ; D. Gordon Smith & Brayden G. King, Contracts as Organizations 51 ARIZ . L. REV . 1 (2009)). The stickiness phenomenon in contracts is sufficiently widespread that lawyers tell jokes about it. For example, one story concerns the senior bond lawyer who had one of two responses to any “why” question regarding a contract provision. Either "We have always done it that way" or "We have never done it that way before." See Blog Posting, March 27, 2007, (available at http://www.concurringopinions.com/archives/2007/03/getting_smacked.html ). 8 E.g., Choi & Gulati, Innovation in Boilerplate Contracts, supra note 3. 9 In theory, if one can determine the full set of conditions under which change occurs, one can say that the absence of those conditions resulted in no change. In the sovereign debt market, however, because the instances of significant changes in contractual boilerplate terms have been few, it is difficult to specify with confidence the range of conditions under which change occurs. One of us has attempted in prior work, nevertheless, to approach the problem from this direction as well. See Choi & Gulati, Innovation in Boilerplate , supra note 4; Anna Gelpern & Mitu Gulati, Public Symbol in Private Contract: A Case Study, 84 WASH . U. L. REV . 1627 (2006); Anna Gelpern & Mitu Gulati, Innovation After the Revolution: Foreign Sovereign Bond Contracts since 2003 , 4 CAP . MKTS L. J. 85 (2009); Stephen J. Choi, Mitu Gulati & Eric A. Posner, Political Risk and Sovereign Debt Contracts , 2011 draft (available at http://ssrn.com/abstract=1962788 ). We attempt to shed light on the question of sticky boilerplate using what will undoubtedly strike some readers as a naïve technique. 10 We went into the field and asked the lawyers drafting these contracts why they had not altered the standard language in light of what they universally claimed was an aberrant interpretation of its meaning. These were sophisticated market actors, many of whom had clear views about why they behaved in certain ways. Our aim was not only to gain traction on the reasons why boilerplate was resistant to revision, but also to obtain insight into the world of elite law firm practice. The production (and maintenance) of boilerplate contracts is, in many respects, the lifeblood of transactional practice in today’s law firms. To the extent we are able to understand the assembly-line process that produces these contracts, we can better appreciate both the strengths and the deficiencies of the business model employed by the modern big law firm.

To put our findings in context, we put the views of the respondents together with both the regnant theories accounting for sticky boilerplate and a quantitative analysis of the contracts themselves. Viewing the stories told by the respondents against the backdrop of theory and the available empirical evidence provides a test of a consistency hypothesis: If the stories are neither consistent with theory or the evidence, then what explains what we have been told? Alternatively, if the stories are consistent with the theory but not the evidence, what explains the myths that lawyers tell themselves? Hopefully, this approach reveals a richer picture both of the stickiness phenomenon and the nature of modern law practice.

The form of narrative in this book follows the preceding structure in that explanations from the respondents are juxtaposed against the theories that purport to explain the same phenomenon. The quantitative data on sovereign bond contracts and the general legal structure governing these contracts provide the context for this juxtaposition of theory and the explanations of the market actors themselves. For example, a number of respondents report that they did not revise their contract clauses because of the view that courts would penalize them by adopting the heretical interpretation for all previously negotiated contracts that contained the same language. We then ask whether this explanation is consistent with theory and also whether the case law suggests that a court would, in fact, penalize those parties who made such a change. Alternatively, some respondents explained that they did not modify the contested contractual language because the bond market penalizes changes in contractual boilerplate. Here again, we look for parallel theoretical explanations and test whether it is the case that bonds with revised contract clauses suffer a pricing penalty on the market.

10 In setting up the project, we drew in part from Ben-Shahar and White’s study of automotive procurement contracts, which drew insights from both interviews with industry participants and the contracts themselves. Omri Ben-Shahar & James J. White, Boilerplate and Economic Power in Auto Manufacturing Contracts , 104 MICH . L. REV . 953 (2006). We begin with a description of the case that led to the drama over the meaning of a key standard clause in sovereign debt instruments.

Elliott v. Peru : Brussels, September 26, 2000

September 26, 2000, in a commercial court in Brussels, Belgium, a judge issued a ruling on what she surely thought was a routine preliminary injunction. 11 That ruling became the catalyst for some of the most radical and far-reaching proposals for reform of the International Financial System. Prior to the ruling, sovereigns were seen as largely enforcement-proof. Unpaid creditors could obtain judgments, but collecting on them was nearly impossible. The preliminary injunction granted by the Brussels judge turned out to have real bite, however, in part because it allowed for the possibility of a creditor going beyond the sovereign debtor and attacking other creditors who might have received payments from the sovereign. This was one of the first occasions in the centuries-long history of the sovereign debt market that creditors had succeeded in using a court decision, rather than political pressure or gunboats, to force a defaulting sovereign to pay its debts. 12

The ruling was ex parte -– meaning that only the plaintiffs were heard on their side of the case– and it granted an injunction against the Brussels-based financial clearinghouse, Euroclear. The injunction barred Euroclear from crediting funds given to it by the Peruvian government to holders of Peru’s restructured Brady bonds. 13 Under the terms of the restructuring agreement, Peru undertook to pay only those bondholders who had agreed to a modification of Peru’s obligation on the bonds. Elliott Associates, a hedge fund holding some Peruvian debt, had refused to enter the restructuring agreement. Now as Peru was preparing to pay the holders of the restructured debt through Euroclear, Elliot sought the preliminary injunction to bar the payments from being made. Elliott argued to the court that since its

11 Elliott Associates No. 2000QR92 (Court of Appeals of Brussels, 8 th Chamber, September 26, 2000).

12 See Alan Beattie, State of Bankruptcy , FINANCIAL TIMES , August 2, 2010, at 5 (discussing the history of enforcement attempts against sovereign debtors and noting this case as one of the only successful attempts to use litigation against a non paying sovereign). For additional background, see Laura Alfaro & Ingrid Vogel, Creditor Activism in Sovereign Debt: “Vulture” Tactics or Market Backbone , HARVARD BUSINESS SCHOOL CASE STUDY , No. 706057 (2007); Rodrigo Olivares-Caminal, Sovereign Default: More Business Opportunities , THE HEDGE FUND JOURNAL (March 2010) (available at http://www.thehedgefundjournal.com/magazine/201003/commentary/sovereign-default-more-business-opportunities.php ). 13 Brady bonds were an innovation of the late 1980s, where the syndicated loans for many Latin American countries were restructured into bonds. These bonds typically had their principal amounts collateralized by U.S. Treasury bonds. The issuance of these Brady bonds coincided (and likely influenced) the emergence of a new bond market for emerging market sovereigns. See Ross P. Buckley, Turning Loans into Bonds , 1 J. RESTRUCTURING FIN . 185 (2004). debt contracts included a term commonly designated as a pari passu clause, Peru was barred from making payments to the holders of the restructured bonds without also making proportional payments to Elliott.

The pari passu clause has been a standard provision in sovereign debt instruments since at least the late nineteenth century. Our data suggest that, apart from the payment terms, it is one of the two oldest covenants in the modern sovereign debt contract; the other being the tax gross-up clause. 14 The Latin translation for pari passu is “in equal step.” But what does it mean for debt to be in equal step? According to almost all commentators at the time of the Brussels litigation, the answer was--not much.15 In the domestic bankruptcy context, the notion of being in equal step has meaning because creditors form a queue during liquidation. 16 Those in equal step recover from the liquidation proceeds pro rata. But sovereigns cannot be liquidated. And, if there is no liquidation event in which the assets of the sovereign are aggregated and paid out, it is unclear what, if anything, the pari passu clause means. One possibility, of course, is that pari passu means that the sovereign cannot offer assets to other creditors as security for their loans and thereby impair the chances that the earlier bondholders will not be paid. But this risk -- that a sovereign’s assets will be parceled out to other creditors–-is the province of the “negative pledge clause” another standard clause in all bond contracts. 17 So the question remains: what, if anything, does pari passu add to the restrictions on the subsequent actions of the sovereign that are not already covered by other clauses whose meaning is well understood? For purposes of this study, however, whether the clause makes sense or not in the sovereign context is not crucial (although it does add to the drama). Much ink has been spilt over that argument already. What is relevant is that the vast majority of the respondents we interviewed believed that the clause made little or no sense in the sovereign context. Yet, even though the clause presented a risk of litigation, they were unwilling to change it.

At the time of the Brussels litigation, some participants in the sovereign debt markets thought that the clause was an historical relic, harking back to a time where foreign sovereigns would grant earmarks of their revenues; that is, they would promise an interest in tax or customs revenues to subsets of creditors. Because these promises arguably did not fit within the definition of traditional security interests, they perhaps would not have been barred by the standard negative pledge clauses that were a

14 Cf. Choi et al., Political Risk, supra note 5. 15 E.g., Qamar S. Siddiqui, Some Critical Issues in Negotiations and Legal Drafting , in SOVEREIGN BORROWERS : GUIDELINES ON LEGAL NEGOTIATIONS WITH COMMERCIAL LENDERS (Lars Kalderén and Qamar S. Siddiqi eds., 1984) 44, 57 (“[the pari passu clause] is likely to have little practical significance in the case of a sovereign borrower, where there may not be an occasion for a forced distribution of the assets to unsecured claimants following the bankruptcy, insolvency or liquidation of the borrower.”). 16 See, e.g., JON YARD ARNASON & IAN M. FLETCHER , PRACTITIONER ’S GUIDE TO CROSS -BORDER INSOLVENCIES , ch. On England at ENG-3 (July 2001) (“The principal of pari passu distribution applies only in liquidation.”). For discussion of that queue and the priorities it enplanes, see Thomas H. Jackson & Robert E. Scott, On the Nature of Bankruptcy: An Essay on Bankruptcy Sharing and the Creditor’s Bargain, 75 VA. L. REV . 155 (1989). 17 The negative pledge clause is a covenant in a bond indenture stating that the debtor will not pledge any of its assets to subsequent creditors if doing so impairs either the priority rank or rights to payment of existing creditors. See Robert E. Scott, A Relational Theory of Secured Financing, 86 COLUM . L. REV . 901, 922-23 (1986). common feature of sovereign bond instruments. Hence, commentators speculated that the pari passu clause might have been aimed at constraining the grants of these quasi-security interests. 18 Others suggested that the clause had migrated from cross-border corporate documents, copied by the lawyers working on drafting the initial sovereign bond contracts of the modern era (in the 1980s), who had not realized that such a clause was meaningless in the sovereign context. 19

Where some saw the pari passu clause as a relic or mere surplusage resulting from earlier drafting errors, Elliott saw opportunity. It argued that the clause, in all its Latin finery, was no relic but rather embodied critically important rights for certain creditors that could be asserted against other creditors as well as against the sovereign debtor. Given that the clause could not logically be referring to liquidation in the sovereign context, it had to be given a meaning that made sense. Elliott was implicitly taking advantage of a long-standing canon of contract construction that all clauses, especially terms in contracts among sophisticated parties, are presumed by courts to have substantive meaning. 20 Sophisticated parties, after all, would only have provisions in their contracts that had a function and, therefore, an understood meaning. That meaning, Elliott and Andreas Lowenfeld, its eminent law professor expert, argued, was that when the sovereign was in default on a payment, it could not make preferential payments to one creditor over another; not if those creditors were linked by a pari passu clause. That, in turn, meant that the clause was an inter-creditor agreement: if some creditors accepted payments—say, under the terms of

18 See, e.g., PHILIP R. WOOD , LAW AND PRACTICE OF INTERNATIONAL FINANCE 156 (1980) (“In the case of a sovereign state, . . . [t]he clause is primarily intended to prevent the earmarking of revenues of the government or the allocation of its foreign currency reserves to a single creditor and generally is directed against legal measures which have the effect of preferring one set of creditors over the other or discriminating between creditors.”); William Tudor John, Sovereign Risk And Immunity Under English Law And Practice , in INTERNATIONAL FINANCIAL LAW , Vol. I, 79, 96 (2d ed. R. Rendell ed., 1983) (“[T]he pari passu clause . . . is primarily intended to prevent the earmarking of revenues of the government toward a single creditor . . . .”); K. Venkatachari, The Eurocurrency Loan: Role and Content of the Contract, in Sovereign Borrowers , in SOVEREIGN BORROWERS : GUIDELINES ON LEGAL NEGOTIATIONS WITH COMMERCIAL LENDERS (Lars Kalderén and Qamar S. Siddiqi eds., 1984) 73, 92 (“In the case of a sovereign borrower the [ pari passu ] clause is intended to prevent the borrower giving preference to certain creditors by, say, giving them first bite at its foreign currency reserves or its revenues . . . . This kind of clause catches arrangements which merely give a right of priority of payment; it is not concerned with arrangements as to creation of security over the assets of the borrower (or others) – that will be provided for in the negative pledge clause.”); ENCYCLOPEDIA OF BANKING LAW , F1204 (Sir Peter Cresswell et al. eds., 2002) (“[A] pari passu clause in state credit is primarily intended to prevent the legislative earmarking of revenues of the government or the legislative allocation of inadequate foreign currency reserves to a single creditor and is generally directed against legal measures which have the effect of preferring one set of creditors over the others or discriminating between creditors.”); TERRENCE PRIME , INTERNATIONAL BONDS AND CERTIFICATES OF DEPOSIT (1990) (recognizing that the meaning attributed to the pari passu clause in corporate context does not apply in the sovereign context and suggesting the earmarking explanation); Ousmene Mandeng, Intercreditor Distribution in Sovereign Debt Restructuring 13 IMF WORKING PAPER (September 2004) (“in the sovereign context the prevailing interpretation of [the pari passu clause has been that] the clause should prevent sovereign borrowers from passing legislation that would increase the risk of subordination of certain lenders”). 19 See Lee C. Buchheit, Negative Pledge Clauses : The Games People Play , INT ’L FIN . L. REV ., July 1990, at 10. Only months prior to the Brussels decision, ISDA, the organization that issues the standard forms used in credit derivative transactions, had deleted the pari passu provisions in its sovereign forms because, as lawyers for one leading investment house put it, the provisions made little sense in the sovereign context. See infra note __ (citing First Boston memorandum). 20 E.g., Delta & Pine Land Co. v. Monsanto Co., C.A. No. 1970-N, 2006 WL 1510417, at *4 (Del. Ch. May 4, 2006) (“It is, of course, a familiar principle that contracts must be interpreted in a manner that does not render any provision ‘illusory or meaningless.’”); see also Goetz & Scott, The Limits of Expanded Choice, supra note 3.. a restructuring agreement—when pro rata shares had not been paid to all the other creditors, then—and this is the important part—those creditors were vulnerable to suit by the unpaid creditors.

Professor Lowenfeld’s affidavit, after noting that the leading practitioners had not articulated a clear meaning for the pari passu clause in the sovereign context, explained:

I have no difficulty in understanding what the pari passu clause means: it means what it says – a given debt will rank equally with other debt of the borrower, whether that borrower is an individual, a company, or a sovereign state. A borrower from Tom, Dick, and Harry can’t say ‘I will pay Tom and Dick in full, and if there is anything left over I’ll pay Harry.’ If there is not enough money to go around, the borrower faced with a pari passu provision must pay all three of them on the same basis.

Suppose, for example, the total debt is $50,000 and the borrower has only $30,000 available. Tom lent $20,000 and Dick and Harry lent $15,000 each. The borrower must pay three fifths of the amount owed to each one – i.e. , $12,000 to Tom, and $9,000 each to Dick and Harry. Of course the remaining sums would remain as obligations of the borrower. But if the borrower proposed to pay Tom $20,000 in full satisfaction, Dick $10,000 and Harry nothing, a court could and should issue an injunction at the behest of Harry. The injunction would run in the first instance against the borrower, but I believe (putting jurisdictional considerations aside) to Tom and Dick as well. 21

The battle, then, was joined over the meaning of this otherwise obscure clause. Under ordinary circumstances, the battle would have been fought and resolved in the commercial court in Brussels. Peru’s lawyers were sovereign debt specialists, the New York-based firm of Cleary Gottlieb Steen & Hamilton. 22 They would have challenged the injunction, pointing out that Elliott was advancing an interpretation of the pari passu clause that virtually no one else in the sovereign market would accept: whatever pari passu meant, if anything, no one believed it meant that a debtholder who had not accepted the terms of the restructuring could attack the payments that were subsequently made to consenting creditors. If such an interpretation were allowed to stand it would undermine any efforts to restructure sovereign debt in the future—a prospect that threatened to destabilize the international financial system. Odds are that the Brussels court, even if sympathetic to Elliott’s position, would have recognized that Elliott was seeking to establish a completely novel understanding of the relationship among the bondholders of sovereign debt instruments. Under those circumstances, the court would likely have been

21 See Lee C. Buchheit & Jeremiah Pam, The Pari Passu Clause in Sovereign Debt Instruments, 53 EMORY L. J. 869, 878 (2004) (quoting the Lowenfeld affidavit). 22 Prior to the Brussels litigation, Peru’s counsel for a number of years had been the Washington D.C. firm of Baker & Hosteller. Cleary Gottlieb appears to have taken over sometime during the litigation in Belgium in late 2000. unwilling to offer a definitive judgment on the meaning of a clause in a contract that was governed by New York law. But the circumstances in this case were not ordinary.

Alberto Fujimori, the Peruvian strongman, was fighting for political survival. On September 14 th , a few weeks prior to the case, a Peruvian television station had broadcast scenes of one of Fujimori’s henchmen bribing an opposition congressman to secure his support. 23 The resulting furor forced Fujimori to announce that elections in which he would not be running would be held in the next year. Fujimori faced a real threat of immediate overthrow and prosecution. 24 As a result, the financial markets became jittery about Peru’s prospects. 25 One story about the events that transpired in Brussels, therefore, is that in light of his precarious position Fujimori was unwilling to risk the political costs of a default on the Brady bonds, however temporary the default might have been. One veteran litigator opined:

Fujimori was scheduled to meet with a senior U.S. official, I have heard it could have been [Madeline] Albright [the then U.S. Secretary of State] . . . He wanted this settled before he met with her. Maybe he wanted to work out a deal for himself . . . he was unstable politically . . . didn’t want a default to deal with as well.

To avoid default, the injunction on Euroclear had to be lifted and the only way to do that immediately was to settle the case, which Fujimori did. 26 And so, Elliott realized a very nice return on its long-shot litigation and walked away with over $58 million on bonds that it had purchased on the secondary market for around $11 million. Some sovereign-side lawyers were displeased with the quick settlement. One frustrated lawyer said:

The [defense] lawyers were literally on their way to the courthouse. They would have beaten this crazy theory. [I hear that] . . . they got a call while they were en route that said that the case had been settled. It was frustrating for us all . . . but the end result was that people began talking about this ex parte decision as precedent. It was not precedent. It was ex parte.

A more colorful account comes from the “vulture hunter”, Greg Palast. Our interviews revealed nothing as exciting as what Palast reports. Palast’s account, however, is illustrative of the near mythical proportions that the Elliott case has taken on over time. He writes:

23 Soenke Haseler, Individual Enforcement Rights in International Sovereign Bonds (2008 draft) (available at www.bepress.com/cgi/viewcontent.cgi?article=1253&context ) (citing Euromoney, October 2000, p. 20) 24 Id.

25 Credit Firms Lower Rating on Peru Debt , Houston Chronicle, September 20, 2000.

26 For background on the pressure Peru was facing, see US Fund Takes Legal Steps in Peru Brady Bond Row , Reuters, September 30, 2000 (available at http://www.financialexpress.com/old/fe/daily/20000930/fns30073.html ). One of my favorite Singer scores was his successful scheme to legally loot the Treasury of Peru. The nation's US lawyer told me, aghast, how Singer let Peru's rogue President, Alberto Fujimori, flee his nation to avoid murder charges. Singer had seized Fujimori's getaway plane. The Vulture named his price: one of Fujimori's last acts as president before he fled was to order his dirt-poor nation to pay Singer $58 million. 27

Back to reality, the important point is that the Brussels court’s ruling–that there was a sufficient possibility that Elliott might prevail on its suggested interpretation to justify the preliminary injunction— was left standing. While the major players in the sovereign debt field were excoriating the Brussels ruling and discussing the increased risks to the international financial system resulting from further unruly litigation, the language of pari passu clauses in subsequently issued debt contracts did not appear to be changing in response to Elliott to at least clarify what it did not mean.

To be sure, over time there were efforts to respond to the Elliott decision in a number of different ways. Prior to the Elliott litigation, bonds issued in New York typically required the unanimous consent of all creditors, while English bonds set the consent requirement at 75%. In the turmoil that followed Elliott, the 75% requirement became the standard for New York bonds as well, thereby reducing the barrier to a successful restructuring agreement. Pressure from the G-20 finance ministries, Euroclear and its own Central Bank, induced Belgium to enact legislation in January 2005 that barred suits against third parties such as Euroclear. 28 Moreover, around the same time as the Belgian legislation was passed, the U.S. Department of Justice, for only the third time in the history of the sovereign debt market, filed an amicus brief in litigation in New York against Argentina expressly disapproving of the Elliott interpretation. Finally, in a small subset of deals, we do see innovation engineered by Lee Buchheit, a senior lawyer at Cleary Gottlieb, who is the undisputed guru of the field of sovereign debt. These innovations include the use of a trustee structure and the removal of the individual right to sue that took the bite out of any future holdout threat. 29 But importantly for our purposes, these collateral responses, extending well over a decade, did not remove the threat posed by the Elliott interpretation of pari passu , a threat that continues to this day, culminating in a second victory for Elliott in federal court in New York in September, 2011.

27 See Greg Palast, Uber-Vultures: The Billionaires Who Would Pick Our President, Truthout, October 6, 2011 (available at http://www.truth-out.org/print/7125 ). Palast has written more extensively about his adventures pursuing vultures. GREG PALAST , VULTURES ' PICNIC : IN PURSUIT OF PETROLEUM PIGS, POWER PIRATES AND HIGH -FINANCE CARNIVORES (2011). 28 For discussion, see National Bank of Belgium, Financial Stability Review 2005 at 162-63 (available at http://www.bnb.be/doc/ts/Publications/FSR/FSR_2005_EN.pdf) 29 For a discussion of these innovations and their impact on the level of holdout creditor threat, see Michael Bradley & Mitu Gulati, Collective Action Clauses for the Eurozone (October 2011 draft) (available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1948534). In prior work, each of us had remarked on the phenomenon of the stickiness of standardized contract terms and had theorized as to the causes of the phenomenon of “rote use” and “encrustation” in contractual boilerplate. 30 Elliott and its aftermath presented the possibility of testing the extant theories explaining the stickiness of boilerplate. To triangulate the stickiness question, we use three sources of insight: the regnant theories for why standard contract terms are resistant to modification; the explanations of market actors for the stickiness of the pari passu clause in the Elliott context; and the empirical data on the effects of the exogenous shock caused by Elliott on the relevant contract language. The different perspectives yield answers that are sometimes at odds with each other. Nevertheless, integrating these different perspectives not only sheds new light on the unresolved analytical puzzle of sticky boilerplate, but also on the workings of that mysterious institution that is the modern big law firm. We begin with a description of the unique contracting problems facing the parties to a sovereign bond deal and the standard terms that lawyers have devised to solve them.

CHAPTER II: THE SOVEREIGN BOND CONTRACT

The sovereign debt market is among the oldest securities markets in existence. For reasons ranging from a monarch’s desire to go to war or a dictator’s need to erect monuments to his achievements to the routine funding of daily expenditures in many modern industrialized economies, sovereign states have been borrowing from time immemorial. Frequently, that borrowing has been from foreign creditors. The fact that a sovereign is the debtor and the creditors are often in a distant jurisdiction creates a unique set of contracting problems. The most obvious one, of course, is that it is difficult to enforce the obligation if a sovereign state decides it doesn’t wish to repay its debts.

In this chapter, we describe those contracting problems and outline the types of contract provisions that debtors and creditors have devised to help solve them. The basic issues with sovereign debt have largely remained the same over many centuries and this allows scholars to examine data over a longer period than for almost any other contracting problem. In turn, this broader temporal perspective not only permits one to determine how the basic contracting problems are solved, but also how those solutions are modified in response to changes in the global economy from exogenous events such as the formation of the Bretton Woods institutions (the IMF and the World Bank), the World Wars or, as in our case, a unexpected court decision in a commercial court in Brussels.

30 Charles J. Goetz & Robert Scott, The Limits of Expanded Choice: An Analysis of the Interactions Between Express and Implied Contract Terms , 73 CAL . L. REV . 261 (1985) (defining rote use as a loss of meaning owing to constant repetition and encrustation as an overlaying of legal jargon that destroys intelligibility); Stephen J. Choi & Mitu Gulati, Contract as Statute , 104 MICH . L. REV . 1149 (2006); Choi & Gulati, Innovation in Boilerplate , supra note 3. The Sovereign Debt Contracting Problem Many are familiar with corporate borrowing and the contracting problems it raises in terms of issues such as limited liability and the agency problems arising from managers’ incentives to misbehave. We use corporate borrowing as the backdrop against which to contrast the sovereign debt contracting problem. Sovereign lending is different from corporate borrowing in at least five important ways, each of which is likely to cause the contractual terms in sovereign bonds to be different from those in corporate bonds. 31

Unlimited Liability : Perhaps the most defining characteristic of the modern corporation is limited liability. Shareholders, who own the equity interest in the firm, share in all of the upside gains from firm projects, but are limited in their exposure on the downside to the extent of their investment in the firm. Unlike a corporate debtor, a sovereign has unlimited liability. It cannot decide to cease operations or go out of business. The deadbeat sovereign cannot be liquidated and have all its assets sold and the proceeds distributed to creditors when it is unable to repay its debts in full. The implications for creditors of this unlimited liability is that the debts of the sovereign never get extinguished. No matter how long ago that credit was extended, and assuming the creditors have not allowed the statute of limitations to run, they can continue to make claims for repayment. For example, there are still bondholders today who are trying to recover on bonds that were issued by the Imperial government in China in the early 1900s and that have long been disclaimed by China’s communist government. 32 French holders of Czarist bonds issued in the early 1900s, subsequently disclaimed by the Soviet government, eventually obtained recoveries during the Boris Yeltsin era. 33 In effect, the unlimited liability aspect of sovereign borrowing means that sovereigns can issue debt with far longer maturities than their corporate cousins.

The Government-Creditor Agency Problem : The standard agency problem in the corporate context is the conflict between shareholders and debtholders. Shareholders control the firm (they hire and fire the managers) and have a residual claim on the earnings of the firm. Debtholders, in turn, are entitled only to a fixed return on the firm projects that they finance. Because shareholders capture all of the upside gains from firm projects in excess of the fixed interest payment to the debtholders and also enjoy limited liability, they have an incentive to increase the riskiness of the firm’s investments so as to transfer wealth from the debtholders to themselves. In the sovereign context, there are no shareholders. In a

31 For further discussion of the contrast between corporate and sovereign borrowing, see Lee C. Buchheit & Mitu Gulati, Responsible Sovereign Lending and Borrowing , 73 L. & CONTEMP . PROB . 63 (2010); George Triantis & Mitu Gulati, Contracts Without Law: Corporate versus Sovereign Debt , 77 U. CINN . L. REV . 977 (2007). 32 See, e.g., Jackson v. People’s Republic of China, 794 F. 2d 1490 (11 th Cir. 1986).

33 See Russia Redeeming Czar’s Bonds , N.Y. Times, Nov. 19, 2010. sense, the citizens of the sovereign state are the equity holders, but their control and earnings rights are significantly different from those of shareholders in a corporation.

There are, however, significant agency problems in the sovereign debt context. Governments are supposed to invest the proceeds of sovereign loans wisely, and use the fruits of those investments to repay the creditors. But because politicians’ primary interest is staying in office, governments have an incentive to use those funds to do things that please voters, and the things that please voters are not necessarily the same things that ensure that creditors are repaid. Moreover, since the government has control over the local laws, if it anticipates difficulty in repaying the debt it can seek legislation to either tax bondholders or restrain their ability to collect payments for other reasons (for example, on the ground that the bondholders are citizens of a country in conflict with the debtor country).

The Intergenerational Agency Problem : A second and different type of agency problem that applies in the sovereign context stems from the incentives of governments to cater to the interests of the current generation of voters. The managers of corporations answer directly to the market in that their employment is a function of the will of the shareholders, who in turn care about the price of their shares. Governments, however, not only have a different constituency, but it is a narrow one – the set of citizens eligible to vote. Since governments primarily care about getting reelected, they will have an incentive to borrow in a fashion that keeps the current generation of voters happy while penalizing later (perhaps as yet unborn) generations with a high debt burden. After all, the later generations do not get to vote in current elections. The problem for creditors arises when these later generations decide that they do not wish to pay what they consider unfairly imposed debt burdens.

The Enforcement Problem : Sovereigns may have unlimited liability, but what is limited is the ability of creditors to enforce legal judgments against a sovereign. To the extent the assets of the sovereign are largely located within the boundaries of the nation, it will be difficult for a creditor to collect by seizing those assets. Local courts, particularly in the context of a populist default, may be reluctant to allow foreign creditors to remove scarce assets from a distressed sovereign debtor. In theory, sovereigns typically do have some assets that are located in foreign jurisdictions – embassies, airlines, ships, gold reserves, central bank assets, and so on – but the sovereign’s primary asset is almost always its tax revenues that are the product of the local citizens’ willingness to pay taxes. And it will be difficult for a foreign creditor to enforce its claim against unwilling taxpayers. In addition, the governments of the foreign jurisdictions where sovereign assets such as the embassy property or state airlines are located are often reluctant to permit creditors to enforce their claims against these assets. Asymmetry of Information : For many corporate debtors, a wealth of information reaches the market through active trading in various types of securities and from analysts who carefully monitor corporate decisions. On the contrary, it is often difficult to obtain good information about the finances of sovereigns. In particular, major items of sovereign expenditures, such as defense, are likely to be kept secret. This, in turn, means that creditors, and especially foreign creditors, will be concerned about the risk of not discovering problems in time to recoup their losses.

The Absence of a Bankruptcy Workout Mechanism : Even if sovereigns cannot be liquidated, there remains the theoretical possibility that there could be judicial oversight (perhaps under the auspices of an international court) of a resolution process to assist with the restructuring of the obligations of a distressed sovereign debtor. Over the years, there have been numerous proposals to create such a mechanism to assist sovereign debtors and their creditors. 34 But these attempts at instituting a global statutory scheme have always met with failure (the most ambitious of these failed proposals was made by the IMF in 2002 35 ).

Two Different Perspectives on the Problem Economists have long been fascinated by sovereign debt. In particular, they have been intrigued by one basic puzzle. Sovereigns are extremely difficult to sue, especially if their assets are largely located on home soil. That, in turn, means that sovereigns can choose when and whether they wish to repay their creditors. If the sovereign chooses not to repay, the creditors are helpless. In other words, the central issue is “willingness to pay, not ability to pay.” 36 The question for an economist then is: Why do sovereigns ever repay?

Broadly speaking, this literature has produced three sets of answers. The first reason that sovereigns are thought to repay voluntarily is the discipline imposed by the prospect of future relations with foreign creditors. Sovereigns repay their debts because they need to borrow again in the future and thus it is important to maintain a creditworthy reputation. The second explanation focuses on other sanctions than can be imposed by the outside world, such as disruptions to trade and impairment of important aspects of international relations. 37 Third, and most recently, scholars have suggested that local

34 See Kenneth Rogoff & Jeromin Zettlemeyer, Bankruptcy Procedures for Sovereigns: A History of Ideas 1976-2001 , IMF Working Paper (2002). 35 See Sean Hagan, Designing a Legal Framework to Restructure Sovereign Debt , 39 GEO J. INT ’L L. 299 (2005); Sean Hagan & Anne O. Krueger, Sovereign Workouts: An IMF Perspective , 6 CHI . J. INT ’L L 203 (2005). __ 36 E.g., John Kay, Loans to a King Do Not Always Pay , FINANCIAL TIMES , August 10, 2011. 37 For an overview of the literature and the sovereign debt borrowing problem, see Ugo Panizza et al., The Economics and Law of Sovereign Debt and Default , 47 J. ECON . LIT . (2009). Among the articles in this field are: Jeremy Bulow & Kenneth Rogoff, politicians in power tend to be myopic. 38 These politicians rationally recognize that the immediate costs of default are likely to be high for them individually. And that gives them an incentive to delay defaults and restructurings, even when doing so would be a rational long-term strategy for the nation.

For our purposes, what is interesting about the economists’ perspective is that the debt contracts themselves are treated as largely irrelevant. 39 The implicit claim is that the design of these contracts is irrelevant to the question of whether sovereigns repay. On its face, this claim is somewhat at odds with the fact that resources are spent in every deal to draft contracts. Elite law firms are hired to document the deal, negotiations take place, and the resulting contracts are then advertised to investors as providing them with important rights and protections. More importantly, the contract provisions frequently impose costly restrictions on both sovereigns and creditors should a debt crisis subsequently arise. For example, the litigation provisions in most bonds specify that the sovereign has waived its sovereign immunity, agreed to specify an agent who will take service of process, and has consented to jurisdiction in a foreign location such as New York or England. Litigation provisions that subject the sovereign to foreign courts and foreign law rather than to the sovereign’s domestic law impose additional financial and reputational costs on a defaulting sovereign. If the bonds are only subject to the sovereign’s domestic law, then the sovereign can forestall creditors’ actions by the simple expedient of changing the law. Such an action would have reputational costs, but those costs would be less severe than if the sovereign were to default on a judgment entered by a court in London or New York.

Sovereigns presumably would not agree to these expected costs unless they believed that it would result in a lower cost of capital when the bonds were issued. To be sure, creditors may sometimes overestimate the value of the provisions in the sovereign bond contract that purport to protect them against default and enforcement risks. But that is a far cry from treating the detailed set of contract provisions that have evolved over two centuries as essentially irrelevant. In any case, as contracts scholars, we begin with the

Sovereign Debt: Is to Forgive to Forget ? 79 AMER . ECON . REV . (1989); Jeremy Bulow & Kenneth Rogoff, A Constant Recontracting Model of Sovereign Debt , 97 J. Pol. Econ. (1989); Jonathan Eaton, & Mark Gersovitz, Debt with Potential Repudiation: Theory and Empirical Analysis , 48 REV . ECON . STUD . (1981); Hershel Grossman & John B. Van Huyck, Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation, and Reputation , 78 AMER . ECON . REV . (1988) 38 See Viral Acharya & Raghuram Rajan, Sovereign Debt, Government Myopia and the Financial Sector , NBER Working Paper Number w17542 (October 2011 draft); Jeromin Zettelmeyer, How to Do a Sovereign Debt Restructuring in the Eurozone: Lessons From Emerging Market Debt Crises , EBRD Working Paper (December 2011 draft). In the context of the current Eurozone sovereign debt crisis, scholars have suggested that one of the structural constraints that may cause countries to work hard to avoid defaults or restructurings are high levels of local holdings of sovereign bonds and the difficulty of distinguishing between local holders and foreign holders of the bonds. For example the holdings could be by local financial institutions, whose failure might trigger a domestic economic crisis. See Acharya & Rajan, supra ; Nicola Gennaioli, Alberto Martin & Stefano Rossi, Sovereign Default, Domestic Banks and Financial Institutions (April 2011 Draft) (available at http://www.econ.upf.edu/~martin/GMR.pdf ). 39 There are exceptions. See Ashoka Mody, What is an Emerging Market , 35 GEO . J. INT ’L L. 601 (2004). assumption that the contracts matter, that they help ameliorate the distinctive set of contracting problems presented by sovereign debt. 40

Contract provisions can operate to ameliorate the basic problems in sovereign lending in a variety of ways. Sovereigns who want to lower their borrowing costs can offer contract terms that make it easier for creditors to seize their assets or sue them. Sovereigns can also offer to constrain their behavior in credible ways that provide their creditors with greater assurance that they will repay. Creditors, in turn, can negotiate for mechanisms that provide early warnings of potential sovereign misbehavior: These provisions enable a creditor to quickly terminate its loan and demand repayment before the risk of default is apparent to others. Contract terms can also be devised that rely on intermediaries who have close and long-term relationships with the sovereign (banks, law firms, international organizations) to enhance the likelihood that the sovereign will repay. Finally, if the lack of a bankruptcy or liquidation mechanism is seen to be a problem, contract provisions can be designed to simulate the effects of a typical corporate reorganization.

Many of the foregoing contractual mechanisms will be costly to the sovereign because they impose constraints on its behavior. It is important to recognize, therefore, that at the time of contract both parties have a shared interest in agreeing to contract terms that will efficiently reduce the risk of nonpayment by the sovereign. To the extent that the sovereign can reassure the creditor that misbehavior in the future has been constrained, its borrowing costs will decline. To the extent that the creditor can propose constraints that are effective in reducing the risk of default but are less costly to the sovereign, the returns to both parties will increase. Both parties, therefore, will work to find contract terms that best address the contracting problems they confront. Consider, for example, restrictions on the types of spending the sovereign can engage in or limitations on the ability of the sovereign to go to war. Sovereigns would presumably not enter into contracts that impose these costs unless they produced expected benefits in terms of an even greater reduction in their borrowing costs. Moreover, we should observe that sovereigns whose reputation for creditworthiness is strong, and thus have less need to provide assurances to investors, are correspondingly less likely to enter into contracts containing such costly provisions.

There are contract provisions that impose expected costs on creditors as well. Consider, for example, the standard modification provision that restricts creditors from renegotiating the terms of the debt unless those holding 75% of the principal amount of the outstanding loan agree to the modification.

40 See also Stephen Choi, Mitu Gulati & Eric A. Posner, Political Risk and Sovereign Debt , University of Chicago Law & Economics Working Paper Number 583 (2011 draft). It can be very costly to obtain the agreement of such a large percentage of the bondholders and the greater the fraction of creditors who must agree to any change the greater is the risk of holdouts. Nevertheless, creditors who wish to deter sovereigns from too easily asking for restructurings might rationally enter into such contracts. In short, if sovereigns simply made repayment decisions without regard for the terms of the contracts they had entered into, we would probably not see the complex and detailed contracting that characterizes the standard sovereign debt contract.

The Provisions of the Contract Broadly speaking, the various provisions of the sovereign bond contract fall into six categories that roughly correspond to (and respond to) the different sovereign debt problems we have identified. 41

Payment Terms : Most obviously and importantly, bond contracts have payment terms. These are provisions such as the interest rate, the currency of payment, and the maturity. And our economist friends do study variations in the use of these terms. Even a casual examination of the sovereign market reveals different borrowing patterns for the strongest borrowers and everyone else. Issuers like the U.S., Germany and Japan are able to issue much of their sovereign debt in domestic currencies, which of course gives them the flexibility to inflate their way out of a debt crisis. By contrast, weaker issuers tend to have to borrow in foreign currencies. Scholars have also observed differences in the maturities of the debts of different issuers, with weaker issuers having to issue debt with shorter durations. 42 Our focus, however, is on the secondary contract terms—the ones that form the bread and butter of the lawyer’s trade.

Financing Constraints : Creditors can and do impose contractual constraints on the actions of sovereign debtors in at least five different ways. One obvious approach is little used in modern bonds: creditors could, in theory, negotiate for provisions that constrain the use of funds by specifying particular projects for which the funds are to be used or limit the amount of future debt the sovereign is permitted to issue until it has repaid its prior debts to the bondholders. These types of severe contractual restrictions, although occasionally present in bonds issued in the 1800s, are almost never seen in modern sovereign bonds. A similarly severe constraint -- the Sinking Fund or amortization provision requiring the debtor to make periodic payments of the principal amount of the debt over the term of the bond--was common in bonds in the 1800s and early 1900s, but has almost completely disappeared from modern bonds.

41 For further detail, see Choi et al., supra note 10.

42 E.g., Olivier Jeanne, Debt Maturity and the International Financial Architecture . 99 AMER . ECON . REV . 2135 (2009); Cristina Arellama & Ananth Ramanarayanon, Default and the Maturity Structure in Sovereign Bonds, Minneapolis Fed Working Paper Number 410 (2010 draft) (available at http://www.econ.umn.edu/~arellano/arellrama.pdf ). Modern bonds do have two primary covenants that restrict the sovereign’s ability to borrow additional funds. The first is the negative pledge clause: This term provides that if the sovereign issues additional debt and grants those new creditors a security interest in particular assets owned by the sovereign, then the debtor has to grant the prior bonds an equivalent security interest. In other words, the sovereign is allowed to borrow as much as it wants, but is restricted in its ability to grant collateral to future creditors without similarly treating existing creditors.

A typical companion provision to the negative pledge clause (occasionally even in the same paragraph) is the pari passu clause. The pari passu clause typically states that the bond will “rank equally” or will “rank equally in right of payment” with all other external indebtedness of the sovereign. And here is where the puzzle arises. As we noted in Chapter I, equal ranking is a concept that applies in domestic bankruptcy proceedings when an entity is liquidated (its assets are sold to generate cash) and creditors are paid out of the liquidation proceeds. Those who rank equally will receive equal shares of the cash proceeds. But if a sovereign is never “liquidated,” and if its liability is unlimited, what is the meaning of a clause that appears to be of value only in a context that will never occur?

Prior to the court decision in Elliott, the leading commentators had suggested two meanings for the pari passu clause in the sovereign context; neither one particularly satisfying. One view held that the clause was a meaningless historical relic, an example of boilerplate that was repeated by rote from contract to contract whose meaning, if any, was lost in the past. The second view was that the clause did have a narrow and specific function: it was intended to protect against the earmarking of streams of revenues to particular creditors. The term “earmark” refers to a type of security interest that sovereigns frequently gave to creditors in the 1800s and early 1900s; a promise to pay those creditors out of a particular stream of assets such as the customs revenues from a particular port or the taxes on tobacco and liquor in a particular locality. The problem with the earmarking explanation, however, is that the negative pledge clause in the modern era has evolved to clearly cover “quasi” security interests such as an earmark. That means that the earmarking explanation also boils down to an admission that the pari passu clause is an historical relic. Critical to our analysis of what follows, however, is the following point: Regardless of what the clause means, if anything, virtually all commentators and market participants took pains to explain that it did not mean that a sovereign in distress was restricted from discriminating among creditors either in a restructuring or in other ways. 43 For example, everyone agreed that a decision to pay the World Bank on a preferred creditor basis was acceptable under the clause. 44

Our principal focus in seeking to understand what occurred after the Elliott case remains the pari passu clause and its relationship to the negative pledge clause and the other constraints on financing found in the standard sovereign debt contract. Nevertheless, to provide a sense of the overall structure of the bond contract, we briefly summarize the other key provisions in the standard sovereign bond that respond to related contracting problems.

Early Warning Provisions : Sovereign creditors, and particularly foreign investors, fear that they may learn too late of events that are likely to trigger a sovereign debt crisis. For this reason, they negotiate for provisions that will serve as early warning signals of a potential default. For example, a cross-default clause permits a creditor to accelerate all of the future payments that are owed on a particular bond if there is a default on any other bond or debt instrument that the sovereign has issued. A similar function is served by a provision requiring that the sovereign debtor continue to be a member of the IMF and be entitled to participate in its programs. The type of early warning trigger that creditors demand will vary, however, depending on the sovereign’s reputation for creditworthiness. Thus, most bonds provide for grace periods–a period of time within which the sovereign is allowed to cure whatever contract breach has occurred. These grace periods can range from 0 days, on the low end, to 90 days, at the high end. There is also variety in the terms of acceleration provisions: Some bonds allow individual investors to accelerate their bonds while others require there to be a vote of over 25% of the outstanding bonds before an acceleration can occur.

Third-Party Assurances : Given the difficulties involved in enforcing a claim for payment against a sovereign, creditors have long relied on third parties to provide reputational assurances. In the 1800s, investment banks like the Rothschilds, Barings, and Paribas acted as reputational intermediaries. Bonds issued through the Rothschilds, for example, were considered sacrosanct–the Rothschilds themselves would stand behind them. 45 Until the mid 1900s, sovereign issuers maintained long-standing relationships with particular banks that were able to supply the sovereign with valuable reputational capital. In the modern era, however, the reputational value brought to a deal by the investment bank is considerably diminished: The lead bank on a particular sovereign’s issuances literally changes from deal

43 E.g., Rodrigo Olivares-Caminal, The Pari Passu Interpretation in the Elliott Case: A Brilliant Strategy but an Awful Outcome , HOFSTRA L. REV . (forthcoming 2011). 44 E.g., Thomas A Duvall III, Legal Aspects of Sovereign Lending, in THOMAS MARTIN KLEIN , EXTERNAL DEBT MANAGEMENT : AN INTRODUCTION , WORLD BANK TECHNICAL PAPER # 245, 41-43 (1994); PHILIP R. WOOD , PRINCIPLES OF INTERNATIONAL INSOLVENCY , 25-062 (2007); LEE C. BUCHHEIT , HOW TO NEGOTIATE EUROCURRENCY LOAN AGREEMENTS 83 (2d ed. 2000). 45 E.g., Marc Flandreau & Juan Flores, Bonds and Brands: Foundations of Sovereign Debt Markets 1820-1830 (2008 draft) (available at http://graduateinstitute.ch/webdav/site/international_history_politics/shared/bondsandbrands%5B1%5D.pdf ). to deal. Big law firms (and their lawyers) have replaced banks as the primary institutions that retain long- term relationships with the sovereign, and presumably they provide some reputational capital when they are working on a sovereign issuance. Other forms of contractual third-party assurances are the promise that the sovereign will continue as a member of the IMF and the promise to list the bonds on a particular exchange. 46

Constraints on Opportunism : Creditors inevitably are concerned with the risk that states will use their sovereign power to reduce the amounts that they owe. Sovereigns, in turn, are willing to offer reassurances against such behavior in order to increase the value of the contract and thereby lower their borrowing costs. As a consequence, the standard contract contains a variety of promises that the sovereign will not misbehave in particular ways. Foremost among these constraints is the currency specified for payment. Foreign creditors, in particular, are concerned about the sovereign attempting to inflate its way out of a debt crisis. As a result, they are willing to charge less if they are promised payment in some relatively stable external currency such as the U.S. dollar or the Euro. Alternatively, the sovereign might be tempted to reduce its obligations by imposing taxes on the payments due to the bondholders. This risk is addressed by a standard provision requiring the sovereign to make the creditors whole for any new taxes that are imposed (this is called the “tax gross-up” clause). Other constraints that reduce the risk of the sovereign behaving opportunistically include specifying the governing law and jurisdiction and a waiver of sovereign immunity for purposes of the bond. Typically, the governing law is specified to be either that of New York or England and the jurisdiction is specified to be the courts of those locations. These are the two jurisdictions that have legal systems that have developed an expertise in adjudicating cross-border contractual disputes and can be relied upon to issue consistent rulings.

Bankruptcy and Collective Action : As discussed, there is no bankruptcy or equivalent mechanism by which a sovereign’s creditors can engineer a restructuring of a distressed sovereign’s debt. As the market for sovereign bonds has expanded over the past two decades, the potential problems in coordinating an agreement among thousands of dispersed bondholders have become more salient. Modern bonds, therefore, contain detailed provisions that attempt to simulate a bankruptcy system. These contract terms are typically referred to as Collective Action Clauses (CACs). They specify matters such as the vote required to alter the sovereign’s payment obligations. Prior to the Elliott litigation, bonds issued in New York typically required the unanimous consent of all creditors, while English bonds set the consent requirement at 75%. As we noted in Chapter II, in the turmoil that followed Elliott, the 75% requirement became the standard for New York bonds as well, thereby reducing the barrier to a successful

46 For a discussion of the various strategies to bring in third-party enforcers in the sovereign debt context, see W. Mark C. Weidemaier, Contracting for State Intervention: The Origins of Sovereign Debt Arbitration , 35 L. & CONTEMP . PROBLEMS 335 (2010). restructuring agreement. In some cases, these clauses require a physical meeting of the bondholders before any agreement can be made modifying the terms of the bond. In such a case, contract terms typically specify the appropriate quorum for holding the meeting and how the quorum is adjusted should the first meeting be adjourned without a quorum. Since creditors facing a restructuring are particularly concerned about debtor opportunism, they often specify disenfranchisement provisions, where bonds “owned or controlled” by the sovereign are excluded from any vote affecting the terms of the bond. 47

Negotiating Over the Contract Terms The negotiation over the typical sovereign bond contract is a decidedly unglamorous event; for the most part not much happens. Both sides in the transaction, the issuer and the underwriter, want the negotiation over the deal documents to take as little time as possible. The sales process-- persuading the underwriter’s stable of clients to buy this particular offering—is where the effort and time is likely to be invested.

Assuming the sovereign has issued bonds in the recent past, those documents are used as the template and marked up to take into consideration recent events. Typically, the standard contract provisions—the contractual “boilerplate”--remain the same. The bulk of the revisions are done by the lawyers who are assigned the task of ensuring that the disclosures being made have been updated since the last offering. In the U.S., this process is driven by concerns about complying with the federal securities laws, rather than with any concerns about the contract terms themselves.

There are two situations in which a different dynamic might operate. The first is in the context of a sovereign doing its first deal. Here, the level of attention paid to the contract provisions is likely to be higher, if only because this deal will form the template that will be followed in deal after deal. However, new issuers are unlikely to experiment much with innovative contract terms. A sovereign that is new to the markets has an incentive to use the standard provisions, as far as possible, so that the likely quite thin market for its bonds is made as liquid as possible. And that may mean using the deal documents that the lawyers have used on a regular basis for their other sovereign clients. The second situation where the negotiations are more complex occurs when a sovereign is going through a debt restructuring and is attempting to do a bond exchange. Here, the issuer is often making a unilateral offer to the existing creditors and the underwriters do not play a big role. Given the prior default and the current restructuring effort, the creditors are going to be wary of the issuer. Investors in this situation are going to expect terms that protect them against the kinds of financial risks that concern investors in the aftermath of a large loss.

47 For additional detail on these provisions , see Michael Bradley & Mitu Gulati, Collective Action Clauses for the Eurozone (2011 draft) (available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1948534). As a formal matter, the parties at the negotiating table are the issuer and the lead underwriter (although, in a routine issuance, there may be no negotiating at all). The lead underwriter usually represents a group of other underwriters who will each commit to sell some portion of the issue to its stable of clients. Our impression from conversations with lawyers is that responsibility for the deal documents tends to lie with the lawyers for the underwriter (the restructuring scenario being the exception). There are several reasons why the underwriter’s counsel is likely the primary contract drafter. The underwriters in a sovereign deal potentially face a greater risk of liability than does the sovereign, since they are inevitably an easier target for disgruntled local creditors to sue. Thus, underwriters will often insist that their standard forms be used. Moreover, the underwriters are the ones who have to sell these bonds to potential investors. They would prefer not to have to argue about legalities in marketing the “product.” Rather, it is much easier to close the deal if they can simply tell their clients that they are using the standard contract document that they have used many times before. A final reason for underwriter control of the contract terms is that the historical practice for many years in the sovereign bond world was for there to be only one set of external lawyers on a sovereign deal--the counsel for the underwriters. In all the deals done between 1945 and 1980 in our dataset, both in New York and England, there were no more than a handful (fewer than 5%) of the transactions where issuers were represented by a New York or London law firm.

Optimal Changes to Pari Passu after Elliott Thus far, we have described a) the unique contracting problems facing the parties to a sovereign bond contract, b) the ways particular contract terms attempt to address those problems, c) how these terms evolve and are standardized and d) the responsibility of transactional lawyers (primarily those representing the underwriter) to update and revise this boilerplate as new conditions warrant. All of this is a prelude to the central question that animated our project: If the interpretation of pari passu that was left standing in Elliott was as wrong as the various participants in the sovereign debt world believed, why wasn’t the standard contract revised to make clear in subsequent bond issues that this was not the intended meaning of the clause?

Before we attempt to answer that question, however, we must scrutinize two of our unstated assumptions. First, note that our question presupposes that the Elliott outcome reduced rather than enhanced the value of the contract to the parties. So we should first consider the possibility that Elliott serendipitously hit on an interpretation that actually reduced contract risks efficiently. If so, one would predict that there would be no revision: the new and presumably better understanding of pari passu would become part of the standard boilerplate. In short, perhaps the industry participants simply failed to see that this was a benign innovation. There are three problems with this argument. First, no one can offer a convincing explanation for what contract risks were ameliorated by the Elliott interpretation that were not otherwise addressed by other provisions, most notably the negative pledge clause. To be sure, investors who subsequently purchase distressed debt would prefer a sharing clause as a lever to demand payment in full should the debtor default. But at the time the contract is drafted and the risks are allocated, prospective creditors must weigh the risks of holding out for full recovery against the certain returns from a restructuring agreement to which most other creditors have subscribed. Second, the “benign innovation” story is challenged by the quantitative and qualitative data we have assembled. We will see that there is some evidence that borrowing costs increased for sovereigns following the Elliott decision, and, meanwhile, the Official Sector expended substantial efforts to add new terms to the contract that would reduce the barriers to a restructuring that Elliott had created. Finally, our goal is not to prove that the Elliott decision reduced the joint value of the contract to the parties but rather to ask why those who had the responsibility for revising the standard terms, and who said, repeatedly and loudly, that the decision was counterproductive to the interests of both parties, nevertheless undertook no action to solve the problem.

A second assumption underlying our project is that, following Elliott , the parties to subsequent sovereign debt contracts would be able to reach agreement on an ameliorating change to pari passu . It is tempting to argue, to the contrary, that this negotiation would be a zero sum game—any change that was good for sovereigns would be bad for creditors and vice versa. Thus, the failure to change may simply reflect the inability to secure an agreement on an appropriate revision. But this argument fails as well. It ignores the point made earlier that when the contract is being drafted both parties have a shared interest in finding the best contract terms that will ameliorate the risks of nonpayment. 48 To be sure, once bondholders have purchased the sovereign’s bonds, they might prefer a contract provision that made restructuring more difficult, thus enhancing the prospect of repayment in full. But the question is whether a provision allowing holdout creditors to attack a restructuring would enhance the value of the contract before bonds were purchased. If the Elliott interpretation (making a future restructuring of the debt more difficult) would reduce the total value of the contract as of the time of issuance, then theory predicts that parties would eliminate this inefficiency in future contracts. In short, it may be more difficult to reach agreement on modified terms once bonds have been purchased or default has occurred, but it should be possible to agree on the optimal terms for modification at the time the loan is first negotiated.

In sum, theory would predict that the following, least intrusive, revision to pari passu would improve contract value following Elliott. In the future, the clause could specify that “the bonds will rank

48 For discussion of this important point, see Alan Schwartz & Robert E. Scott, Contract Theory and the Limits of Contract Law, 113 Yale L.J. 541, 550-55 (2003). pari passu, which ranking does not mean that, following a qualifying restructuring agreement, non- consenting creditors are entitled to recover their pro rata share in the distribution of funds to consenting creditors. ” Such a clarification could preserve whatever meaning, if any, was embodied in the pari passu clause, even though that meaning might have been lost in time and, at the same time, clarify that the Elliott interpretation was disapproved.

CHAPTER X: THE AGENCY COSTS OF BIG LAW The stories told by our respondents fall into two categories. First are those that describe either institutional limitations or exogenous factors that effectively bar transactional lawyers from attempting to revise standard boilerplate even in the face of a non-trivial risk of an adverse interpretation. The second category is more difficult to decipher: Here we find stories that deny the existence of the problem or repeat myths about the origin and original meaning of the pari passu clause. Both the denials and the origin stories reinforce the futility of attempting to revise the language of historic boilerplate such as pari passu. In each of these cases, however, the facts show that the claims stand on shaky ground.

At first blush, none of the extant theories that purport to explain the phenomenon of sticky contract terms fit comfortably with either the stories told by our respondents or the facts as we have come to see them. In that sense, there is no single dominant theory that best explains the observed reluctance to revise boilerplate clauses. In cases where the stories matched one of the efficiency theories explaining stickiness (for example, the negative signal or satisficing contracts stories), there remained either logical flaws in the story or inconsistencies with the data. But one theme does emerge from the aggregation of stories we were told: transactional lawyers who practice in the sovereign debt field are subject to a host of institutional pressures that raise barriers to innovative contract design. These institutional routines work in tandem with the psychological effects of herd behavior and the conflicting loyalties between old and new clients to deter the individual lawyer from revising contractual boilerplate to accommodate the increased risk of a “formulation error.” Such errors are revealed when the occurrence of a contingency— in this case the Elliott litigation—produces surprising (and unintended) consequences, suggesting the instructions embedded in the now-suspect clause are defective.

The failure to revise a clause that appears to contain a formulation error exposes a gap between the interest of the client as principal and that of the lawyer as agent. In combination, these various agency problems appeared to motivate many, if not most, lawyers in our sample to eschew revisions to standard form boilerplate even when inaction increased the risk of adverse consequences for their clients. In sum, our data suggests that agency problems operate both at the level of the firm and also at the level of the individual lawyer and collectively provide the best explanation for the failure of the legal market to implement welfare enhancing revisions to the pari passu clause following the Elliott litigation. Before we attempt to support this claim, however, we first review the ways in which extant theories of stickiness fail to explain the pari passu case.

Failure of the “Faithful Lawyer” Theories i. Risky Clauses that Lack a Standard Meaning will be Deleted Since the pari passu clause frequently appears on the front page of sales documents and is one of the oldest and most widely used covenants in sovereign debt instruments, it is perplexing that so few lawyers have much of an understanding of what it purports to mean. Moreover, despite the litigation risk, our respondents expressed little interest in deleting the clause altogether. The learning externality story posits that the stickiness of a suboptimal boilerplate contract term is a function of a common understanding among market participants as to the meaning of the term. Thus, in the case of bonds traded on the market, the term becomes easy to price. To be sure, there is also an equilibrium if all market participants understand the term to mean nothing (in that case the clause has a zero price). However, if that is so, then deleting the clause seems preferable to retaining a term that everyone understands is largely meaningless. Since all parties would have a common understanding that mere surplusage had been deleted, there would be no risk of litigation.

Any explanation for the stickiness of pari passu has to account, therefore, for the apparent preference for retaining boilerplate terms even when they have lost their understood meaning and in spite of the risk of litigation that could harm either side in the transaction. One possibility is that the clause has value, not because it has a current meaning, but because it is a placeholder for future contingencies; a vague clause promising equal treatment of creditors that can take on new meaning to tackle as yet unforeseen circumstances. In short, perhaps the clause is an empty vessel that can be filled with meaning by courts when conditions warrant. If new contract provisions are costly to formulate and it is difficult for parties to coordinate a move to new terms, there may be a value to retaining vague formulations for courts to fill with content when circumstances demand. This imagined preference for retaining boilerplate as a placeholder for the uncertain future might explain why parties appear to prefer continued litigation over the meaning of the pari passu clause rather than deletion. Litigation is a means by which, in a public setting, a clause can be tested and given a new, and well understood, meaning. If the meaning established in an initial dispute is unsatisfactory, parties can continue to contest that interpretation of the clause until, in a quasi-Darwinian process, an eventual meaning that the market accepts is achieved. There is, however, a problem with this story: this strategy leads quite explicitly to the Elliott outcome where a new meaning emerges that carries significant risks and thus imposes high short-term litigation costs on all market players. Surely, more explanation than this is required.

ii. Stickiness will be Overcome in Time Existing theories of stickiness posit implicitly that, over time, offending boilerplate terms will be revised; that is, that stickiness is a cost barrier but not a complete impediment to intelligent contract design. In our case, however, the resistance to modification of the pari passu clause appears to increase over time. We see this first in the absence of observed changes in the clause following Elliott. In addition, the same parties who decried Elliott also agreed, starting in 2003 (a little over two years after Elliott ), to raise the voting threshold required to alter or remove the clause after the bonds had been issued. 49 Indeed, as the plot of the pari passu data from Table 2 illustrates, the frequency with which parties have used the standard pari passu clause has increased over time even as understanding of the meaning of the clause has decreased.

The data reported in Chapter IX show that, over the years, lawyers have frequently modified the formulation of the pari passu clause in small ways that have arguably altered its meaning. These mutations, however, have occurred quietly, under the radar. The heightened scrutiny that Elliott brought to the clause, however, has apparently made it more difficult to effect revisions. Prior to Elliott , and particularly in the period 1960-2000, a number of mutations of pari passu occurred. Subsequent to Elliott and until the time of this writing in 2011, we find no new formulations. In other words, the pari passu language appears largely to have been frozen by Elliott .

A theory that would explain stickiness here, therefore, would have to also explain increasing stickiness in the face of a negative litigation outcome. One possible explanation for this counter-intuitive outcome is salience : The heightened focus of market participants on a particular boilerplate term can

49 In addition, in one case, a sovereign that had its pari passu language embedded in its negative pledge clause, subsequently (after Elliott ) separated it out and put it into its own separate provision. freeze a clause that might not be difficult to alter when scrutiny of changes in boilerplate is low. Negotiators, the story goes, focus on a limited set of key terms in doing a deal. They make sure that the deal looks very much like the prior deal and similar deals on the market and they have a checklist of terms that then must be put in the contract. The pari passu clause may not have been on the checklist prior to Elliott . But following the hue and cry over the potential damage Elliot would cause to the international financial system, all market participants learned of the clause and thus it migrated to the checklist. Once on the checklist, clauses become difficult to change. More obscure clauses, the prediction would be, are easier to change.

This story suggests that once the Elliott litigation created such a stir in the markets, the problem would be solved by other revisions to the boilerplate rather than changes to pari passu. There is some support in the data for this conjecture. As we noted in Chapter I, Elliott has stimulated some changes in sovereign bond terms. Some bonds now incorporate new terms that reduce the likelihood of holdout litigation. These include prohibitions on the right of individual bondholders to sue as well as the creation of a trustee mechanism for collecting payments. 50 But this “revision by other means” explanation is seriously incomplete. These innovations have occurred in just a small percentage of new issuances. For the large majority of bonds, the litigation risk has been unresolved. Thus, just as with the claim that “litigation will solve the problem eventually,” inaction in revising a vague clause carries substantial costs in the short term to the contracting parties.

“Imperfect Agent” Theories i. Asymmetries in Information and Incentives None of the theories that explain stickiness in terms of the “faithful lawyer” pursuing the interests of the contracting parties is supported by our data. We turn, therefore, to examine the case for an agency cost explanation. Here we begin with the assumption that there is asymmetric information between the contracting parties—the sovereign debtors and their creditors—on the one hand and, on the other hand, between the principals in the transaction and their agents, the transactional lawyers who draft the contract clauses. The parties to sovereign debt contracts—the sovereigns who issue the debt and the creditors who lend the funds and buy the bonds -- may understand the substantive risks of the loan that arise because of the probability that certain contingencies may materialize in the future. But what they do not observe are the costs of formulation error in the contract terms that purport to allocate that risk. These costs are a function both of the risk of error and the efforts necessary to correct that error. The efforts required to

50 The trustee mechanism on its face does not seem to preclude dissenting creditors from pursuing holdout litigation. However, the financial burden of indemnifying the trustee may make it functionally impossible for holdout creditors to sustain the litigation. avoid formulation error are an example of what economists call hidden action—the inability of the principal to observe and thus monitor the actions necessary to perform a task on behalf of the principal. Only the lawyers know these actions.

To understand how things can go wrong in formulating the terms of the contract, it is helpful to view contracts as essentially collections of conditional instructions. Contractual terms first communicate to the parties what consequences are to occur if certain conditions are met and then, when those conditions materialize, define for the adjudicators of any dispute (courts or arbitrators) the appropriate consequences. If the instructions that make up the contract were ideally formulated, then the only uncertainty associated with contractual outcomes would be the probability of the contingencies themselves being realized. By contrast, when the occurrence of a contingency produces surprising and unintended consequences—such as occurred in the Elliott litigation—it suggests that the instructions embedded in the contract term were defective. An agency problem arises, therefore, if the interests of the parties would motivate them to repair a formulation error, but the interests of the lawyers impedes the desired change.

It is tempting to begin with a fairly straightforward agency cost story. One possible story is that lawyers are reluctant to undertake the efforts needed to revise language used in prior deals because, by doing so, they thereby acknowledge that they had erred in using such imprecise language in earlier deals. This would be a costly admission that inadequate efforts had been taken to avoid this formulation error in the large number of the debt contracts containing the earlier version of the clause that were still on the market. In fact, some respondents made this precise admission (while sometimes also stating that the pari passu clause was not the cause of the big problem that needed to be solved). In the words of one lawyer:

When you amend a clause, you acknowledge a problem. No one wants to acknowledge that they had messed up in a prior deal. . . . When you move away from loans to bonds, you have to have a method to constrain the debtor from misbehaving. Pari passu clauses are one way to constrain. Pari passu is not the problem. Pari passu is just a technicality. The real problem is that the holdout creditor is attaching assets . . . What one needs as a solution is a way of binding dissenting creditors.

We can find other versions of this agency cost story in our data. For example, several respondents acknowledged a reluctance to revise the pari passu clause in new bond issues with new clients because of the possible negative signaling effect such changes might have on the risk of an Elliott- style attack on the bonds of existing clients whose contracts contained the vulnerable language. Since the negative signaling story by itself is unpersuasive, a plausible interpretation of this explanation for inaction is reluctance by lawyers to confront old clients with the fact of the revisions in new contracts.

Despite some evidence to support this simple agency cost story, this account is weakened by several facts. First, the clients here are sophisticated, both on the investor and the issuer side. They inevitably have in-house lawyers who previously had worked at the big law firms and had drafted prior iterations of these documents. Thus, it is implausible to assume that any agency problems are attributable solely to a failure by in-house counsel to monitor the effort that lawyers put in to drafting contractual boilerplate. Second, because everyone in the market has some version of the pari passu clause, it is difficult to imagine a client penalizing its lawyer for lack of effort in failing to revise pari passu in a past deal. We talked to a number of client-side lawyers; not only did all of them recognize the problem, but not one of them blamed their outside counsel for including pari passu clauses in prior deals.

In the case of one large international bank that had been the subject of a recent pari passu attack, the bank explicitly asked its lawyers to protect it against future holdout litigation with respect to the bonds of a particular country. Instead of deleting or revising the pari passu clause, the solution was to add a new term in the contract restricting the individual bondholder’s right to sue. One of the lawyers explained:

There was concern among our clients, particularly some of the banks, about lawsuits by aggressive creditors seeking to hold up a restructuring. One large bank in particular had been sued using the pari passu clause. It did not want to have to deal with such problems again if there was a need for a restructuring. That was big reason why we took away the individual suit possibility.

This episode confirms a drafting strategy for correcting problematic clauses that we discussed earlier: Rather than revising the clause to solve the problem directly, the lawyer proposes new clauses (such as a prohibition on individual creditor’s suing the sovereign) that purport to solve the problem indirectly. There are two problems with this approach. First, as noted earlier, this strategy of “innovation by indirection” has only been adopted in a small minority of bonds issued after Elliott : It has not led to a market-wide revision of the contract terms. Second, because the innovation occurs around the periphery of the standard boilerplate, it leaves difficult interpretation questions unresolved. To be sure, once litigation has focused attention on the problem, a lawyer may well spend more time and effort explaining to clients and others why it is necessary to delete or revise troublesome boilerplate than it would take to attempt to solve the problem with a new clause. But rather than drafting a new term that appears to conflict with pari passu , the parties would be better served by a straightforward revision that corrects the “erroneous” interpretation. Yet this action appears to be constrained by factors other than the mere physical task of revising standard language. In the following section, we consider the nature of those additional barriers that make the effort needed to revise pari passu greater than it would otherwise appear.

ii. Barriers to Innovation As noted above, our respondents had little concrete information as to the distribution of various types of clauses in the market. Nor did the majority of respondents appear to have given any consideration to the implications of having any one version of pari passu. At best, there was a sense that the litigation risk was adversely affected if a clause used the word “payment” to modify the concept of “rank.” While some of the senior lawyers among the respondents mentioned that someone might be able to unearth the true meaning of the clause by doing a historical study of its evolution, no one thought that their firms would have any interest in undertaking this research. Moreover, no one appears to have reexamined the bonds issued in the 1960s and 1970s—bonds that were worked on by many of the existing U.S. and U.K. law firms—for clues as to why these clauses were originally used in unsecured sovereign debt instruments. This indifference might not be surprising in the case of an obscure clause that had been routinely used without controversy for generations. But it demands explanation given that the pari passu clause has been the subject of multiple litigations over the past decade (and, as of this writing, it is at the center of the most recent round of creditor litigation against Argentina).

We pushed on this question, and the respondents pushed back. This was not about lawyers, they said. Lawyers are eager to bill more hours for research, especially when it might help in training junior associates. During a conversation with two lawyers, a senior partner and a junior associate, we heard:

[Senior Lawyer]: Part of the story is the competition for business. Clients want to pay as little as possible on these deals. And the firms who computerize the production of documents can compete better on price. These documents are no longer treated like traditional contracts–where provisions are negotiated and the lawyers invest in understanding the provisions. Lawyers are beginning to treat these contracts as commodities. [And in that world of commoditization], clients will say–Why should we pay you for investments into understanding [contracts]? The victims are the younger lawyers.

[Junior associate interjects]: “[It is] too easy not to ask the tough questions. Very difficult to ask a question . . . it causes delays to ask questions . . . and everyone knows that . . .

[Senior Lawyer]: Theory is that law firms are instructing juniors in the history. But that does not happen. This incident is deeply embarrassing. What happened to the old rule about nothing there in a contract that is not understood? History has shown us how dangerous it is to express agnosticism regarding a clause or provision we do not understand . . . because an [Elliott] will always be out there . . . waiting.

The information deficit does not end with knowledge about the meaning of the clause and the distribution of the types of clauses in the market. It extends as well to materials that would have been readily available to all of our respondents. Consider, for example, the article by Lee Buchheit and Jeremiah Pam, two Cleary lawyers, reporting their research on the origins of the pari passu clause. This article was circulated widely among sovereign debt practitioners. But only a few of our respondents appeared to have read it carefully. Fewer still had looked at the article on pari passu by the legendary Allen & Overy lawyer, Philip Wood, or the Bank of England’s Financial Markets Law Committee report.

Then there is the case law on corporate bonds. It is not voluminous; there are roughly a half dozen cases from which predictions might be made as to what federal courts in New York might do with an interpretation question of the pari passu variety, particularly when a foreign jurisdiction had made an earlier interpretation. Finally, there are the cases involving pari passu and the various amicus briefs that were filed. Again, but for a handful of respondents, most had no more than a vague sense of what these materials said. 51

It wasn’t that these lawyers were uninterested in the history of these clauses or in the data about the variation in terminology or in the empirics of whether there was a pricing penalty for a certain form of the clause. A number of respondents helped us to answer these questions. Moreover, respondents were often fascinated by the antique bonds from the late nineteenth century when we would show them in the course of conducting interviews. 52 One senior partner of a U.S. firm, whom we met in London after a day at the Rothschild archives, brought a magnifying glass to dinner to be able to read more easily the small cursive print on the old bonds. Another senior partner found for us a database of old newspaper articles from the early 1800s that provided clues as to the provisions in early bonds. Yet another sent us references to articles on the practice of earmarking in sovereign debt from the 15 th century. 53 Finally, a handful of respondents searched old files at their firms for copies of the handful of bonds issued in the 1960s, 70s and early 1980s that are not available on the public databases. 54 But again, these lawyers made it clear that they were motivated by curiosity and that they couldn’t ask one of their junior associates or paralegals to assist them because the time could not be billed to a client. One of the last

51 Among the handful were the litigators. Despite, for the most part, being generalists who worked on cases from a variety of areas, they were more likely to have read the background research materials than their transactional colleagues. 52 Both in London and New York we would frequently spend the mornings in the archives, before turning to the interviews in the afternoons. 53 See Christophe Chamley, Default More than 400 Years Ago Leaves Scars , BLOOMBERG NEWS , July 31, 2011. 54 Multiple senior lawyers flew to Durham to talk to the students about their research projects in a class that one of us taught in 2008 about the evolution of clauses in Latin American sovereign debt instruments between 1825 and 1940. interviews we conducted was with a senior U.S. lawyer, who was prompted to respond after being showed copies of a bond for Brazil from the late 1800s (that looked to have one of the earliest pari passu clauses) and a Mexican bond from the early 1970s (that had a version of a pari passu clause different from the contemporary version). At the end of dinner that evening, he volunteered the following about our project:

You are not seeing the big picture. The answer to your question is embedded in the larger question of why law firms do not have R&D departments. We just don’t do basic research. With contract provisions, for example, we might make modifications at the margins. Small changes . . . maybe. When we make those small changes, we do them quietly. No fanfare.

And why don’t we have R&D departments doing research on the meaning of existing clauses and devising new ones? It is hard to see the economic justification for that. It would be easy for others to copy any new clause we developed . . . the reputations that matter are for getting the deals done . . . not [for] knowing the historical origins of some obscure contract provision.

The senior lawyer had phrased his answer in structural terms: the lack of R&D departments. But the question isn’t really about R&D departments, but why, independent of departmental structures, law firms do not invest in doing more basic research into the tasks that they routinely perform? This is not to say that law firms and their lawyers do not or cannot ever devise innovative products. There are exceptions: Wachtell Lipton famously developed the poison pill in the 1980s. 55 Elliott’s strategy of using the pari passu clause to sue sovereigns was itself an innovation that we presume originated with its law firm. 56 And these firms devised those strategies even though they could (and were) copied by others; the first-mover and reputational advantages must have been sufficient to justify the initial expenditure. That said, the broader point is that law firms, and particularly the transactional departments of big law firms, seem to underproduce innovation.

There are several reasons for the apparent barriers to innovation in Big Law. On the demand side, corporate clients, facing competition in their respective industries, have cut costs by bringing routine transactions in-house. They assign less value to one-stop shopping and relationship lawyering, and are more willing to bid out transactional work in pieces. On the supply side, new information technologies

55 See Michael J. Powell, Professional Innovation: Corporate Lawyers and Private Lawmaking , 18 L. & SOC . INQUIRY 423 (1993); William H. Starbuck, Keeping a Butterfly and an Elephant in a House of Cards: The Elements of Exceptional Success , 371, in THE STRATEGIC MANAGEMENT OF INTELLECTUAL CAPITAL AND ORGANIZATIONAL KNOWLEDGE (Chun Wei Choo & Nick Bontis eds. 2002). 56 Ribstein and Kobayashi explore this question in the context of class action litigation, making the point that the current system might not provide adequate incentives for innovation in law making for lawyers competing to represent a class. See Larry Ribstein & Bruce Kobayashi, Class Action Lawyers as Lawmakers , 46 ARIZ . L. REV . 733 (2004). have reduced the cost of storing, cataloguing and retrieving documents. This change is a double-edged sword for law firms. On the one hand, it reduces the cost of maintaining and using contract precedents and thereby generates economies of scale within categories of transactions. So, while clients have brought in-house the standard contracts that they enter into regularly, they continue to turn to a law firm with scale economies across a large number of clients for help in extraordinary transactions such as sovereign debt issuances. But, on the other hand, these scale economies have led to deficiencies in the organizational structure of large law firms, including the secondary role given to R&D and the manner in which legal services are marketed. Innovation in contract design requires investment in an interactive collaborative process, a process that is largely absent from the modern firm.

iii. Institutional Constraints Reinforce Herd Behavior The institutional constraints that deter innovation in the modern law firm reinforce cultural patterns of behavior among the individual lawyers working in the firm. As noted in Chapter VII, a number of our respondents explained their inaction in terms of the psychology of herd behavior. What makes herd behavior intractable is that it is often a rational strategy for an individual lawyer working in a large firm to pursue. While herd behavior has been attributed to informational asymmetries, 57 in the law firm setting it is largely a function of the individual lawyer’s concern with her personal success, which is tied to her reputation. 58 Here the key is to avoid distinguishing oneself as a ‘bad’ decision maker. By staying with the herd there is the possibility of ‘sharing-the-blame’ even if the decision—say, not to revise the risky pari passu clause—ends up being a poor one. Since it is difficult to know what is a good innovation in contract design, a fortiori the safest bet is to mimic others’ decisions. Under these conditions, only those lawyers who have either very good or very bad reputations innovate. The former do so because they are accustomed to a high level of success in all that they do. The latter do only because they under-perform in relation to the herd and may have better success by plotting a new course. 59

While it may seem that herd behavior is inefficient, there are inherent benefits that account for its existence. First, there is a strong draw to retain standardized terms not only for the benefit of

57 Under an informational asymmetry rationale, herd behavior occurs when a decision maker ignores her information and follows the actions of another, based upon an underlying assumption that the initial decision maker has information that might be beneficial to her. See Abhijit V. Banerjee, A Simple Model of Herd Behavior , 107 Q. J. ECON . 797 (1992). This blind ‘follow- the-leader’ behavior is inefficient because it encourages decision makers to join the herd at the expense of ignoring their own information. This behavioral bias is similar to Kahan and Klausner’s discussion of anchoring bias, where a fixed point drags the responses (or decisions) in that direction. See Marcel Kahan & Michael Klausner, Path Dependence in Corporate Contracting: Increasing Returns, Herd Behavior and Cognitive Biases , 74 WASH U.L.Q. 347 (1996). 58 E.g., David S. & Jeremy C. Stein. Herd Behavior and Investment , 80 AM. ECON . REV . 465 (1990); Jeffrey Zwiebel, Corporate Conservatism, and Relative Compensation , 103 J. POL . ECON . 1 (1995). 59 See Zwiebel, supra note 10. contemporary use but also for the benefit of later use with a familiar term. 60 Moreover, once a level of standardization has been reached there is pressure to maintain this status quo because of the costs of switching to a new standard form. A benign way of putting the same point is that elite lawyers may have a culture of sharing; that is, of refusing credit for innovations. And that culture may translate into a disincentive to do the research needed to innovate. Both in our research on the pari passu clause and in a related study examining a different set of clauses, we found that the lawyers who had made modifications or innovations in contract language were often reluctant to take credit for the innovation. 61 If anything, it appeared to cause them discomfort that we had identified changes in the documentation that they, specifically, had authored.

This herd behavior story provides for us the most compelling explanation for the origin myths that we were told. The broad theme that emerges from these stories is that efforts to revise historic boilerplate are costly and counterproductive. In particular, the persistence of the "inadvertent copying" story in the face of the evidence of systematic changes through the years reinforces the view that there are institutional norms that have developed specifically to support a) the futility (and, indeed, the riskiness) of attempting to change the boilerplate, and b) the myth that, therefore, changes never occur. This origin story feeds into the overly-candid admission recounted in Chapter II that the contract only takes “three and a half minutes" to draft. After all, why should any lawyer waste time (and her client’s money) worrying about contractual language that is fixed in stone?

iv. Structural Divisions: Litigation versus Transactional Practice A further institutional explanation for the absence of R&D and the lack of innovation in the modern law firm may be the separation between litigation and transactional work. The modern law firm tends to be organized around practice groups. Two of these groups—litigation practice and corporate transactions—often fail to mesh at the interface of particular transactions because the firm that litigates a transaction is often not the firm that did the deal in the first instance. By connecting these two services, rather than treating them as distinct tasks or modules, law firms could internalize some of the lost benefits from providing contract design services. But even within a single firm, the two sets of lawyers do not seem to talk to each other. Few of the litigators who have worked on the pari passu matter over the years (and, at least for purposes of litigation, have learned an enormous amount about the sovereign markets) appeared to have been asked by their transactional colleagues either about the legal risks inherent in using one type of a clause over another. A firm learns of legal risk—such as the risk of not altering a problematic clause or the risk of using one form of the clause over another— from the information known

60 See Kahan & Klausner, supra note 9. 61 See Anna Gelpern & Mitu Gulati, Innovation After the Revolution: Foreign Sovereign Bond Contracts After 2003 , 4 CAP . MKTS . L. J. 85 (2009). to the litigators working on actual cases. But this knowledge is not systematically transferred to the transactional lawyers who are drafting new contracts and preparing disclosure statements about matters such as litigation risks.

This Chinese wall between litigators and transactional lawyer has costly effects. A lawyer’s advantage lies in her knowledge of the adjudicative—typically adversarial—process used to enforce contract terms. Understanding the litigation risk embedded in contractual boilerplate is thus the core competency from which lawyers derive comparative advantage in designing transactions for their clients. Yet, this advantage seems underappreciated in contemporary big firm practice, other than in a handful of specialized fields.

To summarize, our data suggest that a combination of explanations for the stickiness of the pari passu clause may be at work. Many complementary factors point in a single direction: the lack of investment in R&D, herd behavior and the culture of not giving (or taking) credit for innovations or background research if not immediately billable to a client, the separation between litigation and transactional practice, and the residual embarrassment of having to explain contract revisions in new contracts to clients who are bound by the older, risky, language all impel lawyers to resist revising standardized language. These factors are reinforced by a powerful mythology about the origins of historic boilerplate that in tandem motivate lawyers to choose to retain old formulations despite an increase in the risk of an adverse interpretation.

Complications, Biases and Caveats The theme that we have identified runs through most of the stories told by our respondents, is consistent with the data we have gathered and fits much of the theoretical literature on the institutional barriers to innovation . But it is not a slam dunk. There are a number of complicating factors that cause us to pause before we confidently declare “Big Law” and the lawyers who work there the culprit in the story.

i. Lee Buchheit and the “Wise Man” Model Lee Buchheit, one of the senior most lawyers at Cleary Gottlieb, complicates our story. Tall, thin and sporting a moustache that suggests he is from a different era, he speaks both softly and slowly, in perfectly formed paragraphs. A distinctive accent and an understated sense of humor that brings peals of laughter from his audience even when he is sketching out the darkest of scenarios, complete this picture.

Buchheit is frequently referred to as the “guru” of the sovereign debt world.62 He advises

62 E.g ., Julie Triedman, Front Row Seats at a Financial Disaster , AM. LAWYER , Dec. 1, 2011. governments around the world, publishes widely, teaches periodically at Yale, Harvard, NYU and Columbia, advises graduate students, collects his own statistics on the market, and has an encyclopedic grasp of the history of the international bond markets. 63 He rarely does routine transactions. He is retained when a country has an unsustainable debt load and his first task is generally to break the news that there is no option but to do a difficult and painful restructuring. He has engineered some of the most successful sovereign restructurings of the past two decades, including those involving Iraq, Ecuador, Uruguay, and Belize. After helping the Icelandic government negotiate out of its financial crisis in 2010, he was selected “man of the year” in business by Frettabladid, an Icelandic newspaper. 64 Although the deal he helped negotiate for Iceland was ultimately rejected by the populace in early 2011, Buchheit earned the admiration of many Icelanders. In late 2011, the Greek Government engaged Buchheit for what may prove to be his biggest challenge yet. 65 As John Dizard of the Financial Times wrote:

Last week, it was announced that Cleary Gottlieb, the New York headquartered international law firm, had been engaged by the Hellenic Republic. I believe this tells us a lot about the direction euro area finance will take… the Cleary team working on Greece will be headed by Lee Buchheit, the partner who has previously represented Iceland, Argentina, etc, etc. Trusted by distressed governments, cursed by former optimists with long positions, his is the mobile number of choice for sovereigns that believe they have unsustainable debt burdens. 66

Buchheit complicates our story for multiple reasons. He innovates on a regular basis, often by devising new transactional structures as well as novel contract terms. He is also willing to take credit for his innovations, in that he does not try to hide them; he writes about them in the most prominent trade journals, gives presentations at industry meetings, and is a fixture on every major international drafting group. He structured the use of Exit Consents by Ecuador in 2000, designed some of the earliest collective action clauses and advocated for their adoption, designed the first aggregation provisions for Uruguay in 2004, designed the trust structures for Iraq in 2005 and was key in removing the individual right to sue in the restructured bonds for Grenada and Belize in 2005 and 2007. 67 Even before he was

63 One of us (Gulati) has worked closely with Buchheit on a number of sovereign debt articles over the past decade. Buchheit has a number of former associates who are now academics or policy makers, all of whom tend to be fiercely loyal to him. 64 See http://uti.is/2010/12/lee-buchheit-man-of-the-year-in-iceland/ 65 See Get Him to the Greeks: Greece Taps Cleary’s Buchheit , AMLAW DAILY , August 2, 2011 (available at http://amlawdaily.typepad.com/amlawdaily/2011/08/cleary-gottlieb-greece.html ). 66 Joseph Cotterill, Cleary Gottlieb’s New Client , August 1, 2011, FINANCIAL TIMES ALPHAVILLE (available at http://ftalphaville.ft.com/blog/2011/08/01/639096/cleary-gottliebs-new-client/ ). 67 See Lee C. Buchheit, How Ecuador Escaped the Brady Bond Trap , 19 INT ’L FIN . L. REV . 17 (2000); Lee C. Buchheit & Elizabeth Karpinski, Grenada’s Innovations , 21 J. INT ’L BANKING & REG . 227 (2006); Lee C. Bucbheit & Jeremiah S. Pam, Uruguay's Innovations 19 J. INT ’L BANKING & REG . 28 (2004); Lee C. Buchheit & Elizabeth Karpinski, Belize’s Innovations , BUTTERWORTHS J. INT ’L BANKING & FIN . 278 (May 2007); Lee C. Buchheit, Supermajority Control Wins Out , INT ’L FIN . L. REV . 2 (April 2007). engaged by the Greek government to represent them, his views on how to tackle the Greek crisis were the ones to which the financial press frequently turned.68 On the pari passu matter alone, Buchheit and his colleague, Pam, did an historical excavation of the meaning of the clause, examining contracts from back in the mid-1800s to the modern era. 69 For that project, they did informal surveys of market practices and also talked to retired lawyers who might remember stories about the origins of the clause. In other words, they did everything that we have been saying that lawyers do not do.

Further, Buchheit is one of the few who says there is a clear modern meaning for the clause. As we described in Chapter IX, Buchheit and Pam concluded that the meaning of the pari passu clause in the sovereign context had evolved over time, taking on different meanings to solve different problems over the decades. 70 Although this was not a meaning for the clause that we heard from other respondents, Buchheit was confident in the meaning he and Pam had unearthed.

The question is: How do our general observations about the stickiness of pari passu and the reluctance to innovate mesh with what we know about Buchheit? To complicate the matter further, we do not see a systematic removal or modification of the pari passu clause in the deals done by Buchheit’s law firm. What we do see is that, in a subset of his recent deals, including those for Grenada, Iraq, and Belize, there were other innovations such as the use of the trustee structure and the removal of the individual right to sue that took the bite out of any future holdout threat.

Our initial thought was to treat Buchheit as sui generis. He is a maverick, working outside the firm’s traditional structures, spending significant amounts of non-billable time advising governments and teaching law school classes, and often doing his own research in remote archives. Nevertheless, as one respondent pointed out, for all his idiosyncrasies Buchheit is an integral part of the traditional structure. Cleary Gottlieb presumably pays him a large income to do research, advise governments, and design innovations. Buchheit was front and center when the pari passu battle against Elliott was joined, devising arguments and orchestrating strategy, even though he is a restructuring lawyer and not a litigator. Furthermore, Buchheit’s research, when examined in hindsight, rarely looks to be purely a matter of intellectual curiosity; it inevitably provides the legal basis for a new technique or clause he is developing.

Other firms, such as Allen & Overy and Clifford Chance, have similar figures. Philip Wood and Andrew Yianni in London–both of whom frequently write articles—have an encyclopedic knowledge of the contracts, and regularly serve on international drafting and policy making committees. Sullivan & Cromwell, Linklaters, Davis Polk, White & Case, Arnold & Porter and Shearman & Sterling, have

68 E.g ., John Dizard, In an Ideal World, Kafka Would Restructure Greece , FINANCIAL TIMES , July 31, 2011. 69 See Lee C. Buchheit & Jeremiah Pam, The Pari Passu Clause in Sovereign Debt , 53 EMORY L. J. 869 (2004). 70 See discussion in Chapter VIII. analogous figures with oracular status. These individuals do not fit the model of relentless accumulators of billable hours and rarely do they have stables of clients. But they are the closest we see to R&D departments in major law firms. Moreover, these lawyers are also the sources of historical information about the firm’s documents. One associate described this as “the wise man model” of information and innovation.

It is not obvious how this wise man model relates to contract stickiness, nor does this structure appear to be an ideal business model for the modern law firm. But skepticism aside, aspects of the foregoing—in particular Buchheit’s maverick status—are consistent with our predictions about the barriers to innovation in the modern law firm. Successful firms that develop routines and standardize products often find it hard to reconfigure those standard production processes to tackle innovation, let alone do the research necessary for innovation. As we have seen, the processes that are most effective at churning out standard products often work against innovation. 71 Perhaps Buchheit and the other wise men are able to innovate precisely because they work outside the traditional channels of the firm.

Another possibility is that the aura that has attached to Buchheit (and the other wise men) is itself partly based on myth. To be sure, they are all extraordinary lawyers. But the importance and influence that is attributed to them is much larger than life. On multiple occasions in our interviews, respondents would begin their answers with “Have you talked to Lee Buchheit? He can answer your question” or “Have your read Lee Buchheit’s article on this?” or “If Lee Buchheit were to say that the clause should be changed, it would be.” That these sophisticated market actors would attach so much importance to the views of any single lawyer in the field may suggest that the legend has outgrown the man.

Buchheit himself freely acknowledges this fact. When he was asked at a public lecture about one of the many innovations that had been incorrectly attributed to him, he replied: "I find myself in the same position as Saint Joseph . . . there are persistent rumors of paternity that I persistently deny." The point here is simply that this industry not only has its myths, but it also has mythical figures. As a critic from inside the industry pointed out to us, perhaps it was only the myth of Buchheit that did not fit our story. Maybe the reality fit better. In sum, figures like Buchheit and Wood are portrayed as having power and influence that it is utterly implausible to think that they actually have, even within their own firms.

And it is not just Buchheit who has achieved a mythical status. The perceived influence of Elliott Associates following the events of the Brussels litigation against Peru has achieved this stature as well. As time has past, the stories about what went on during the Brussels litigation get imbued more and more

71 See, e.g., Jeffrey Macher & Barak Richman, Organizational Responses to Discontinuous Innovation, A Case Study Approach , 8 INT ’L J. INNOVATION MGT . 87 (2004). with intrigue and drama. Consider the following from the BBC reporter, Greg Palast, on Elliott’s Paul Singer:

One of my favorite Singer scores was his successful scheme to legally loot the Treasury of Peru. The nation's U.S. lawyer told me, aghast, how Singer let Peru's rogue President, Alberto Fujimori, flee his nation to avoid murder charges. Singer had seized Fujimori's get-away plane. The Vulture named his price: One of Fujimori's last acts as president before he fled was to order his dirt-poor nation to pay Singer $58 million. 72

The reality, at least as we have learned from our respondents, was much more mundane.

There is a general point to be made about this tendency to mythologize a few individual actors in the sovereign debt market. These myths may serve to reinforce a sense of futility: How can ordinary lawyers influence events in what they see as a cosmic struggle between mythical figures of extraordinary power and influence? Under these circumstances, no one would be so foolhardy as to attempt to revise the terms of the contract that is the focal point of this historic struggle.

ii. The Move to Systems: The Death of the Wise Man Model? Even if many of these law firms have relied on a wise man model to stimulate innovation in the past, this model may soon be extinct. Every one of the wise men we spoke to or heard about was either near retirement or had already retired. 73 We talked to a number of these gurus and some of them spoke wistfully about the heydays of the 80s when they would gather, periodically, to talk about common problems and how to solve them. Apparently, those gatherings no longer occur, at least not in the sovereign debt area.

Our initial inclination was to dismiss the talk of the “good old days” as the product of nostalgia. Surely, there was a new generation of wise men who were in the process of replacing these senior statesmen. But that might not be the case. One of our respondents, upon reading an early draft of this book, called to say that what we had observed might have been the tail end of a very different model of the modern law firm. The respondent explained:

The senior statesman stuff you guys talk about . . . I’m not sure that works anymore. It may be a thing of the past. Take Buchheit. His associates are extremely junior; they are not going to take over his mantle. . . . I noticed that yet again at the [international conference] LCB brought a very

72 Greg Palast, Uber Vultures: The Billionaires Who Would Pick Our President , October 5, 2011 for TruthOut/Buzzflash (available at http://www.gregpalast.com/uber-vultures-the-billionaires-who-would-pick-our-president/ ). 73 When the pari passu drama occurred in 2001, a number of these figures had to be tracked down in retirement to find out whether they had any recollections from their time in practice as to what the clause might have meant. junior associate. This made me think of his protégés and the only one I can think of [who is in private practice at a senior level] is [. . .] and he wasn’t really one of the disciples . . . He just does not seem to have much of an institutional base, despite having hordes of swooning admirers and loyal protégés in all kinds of interesting places. . . . Your draft doesn’t seem to recognize that. I suspect that there simply aren’t junior partners at Cleary or A&O who are in their mid forties and have the kind of stature and knowledge that Buchheit or Wood had when they were that age. The senior statesman model has been scrapped. It is a new model – things are automated. In the next generation, there won’t even be the senior statesman to turn to in order to find out what a clause means.

Assuming our respondent is correct that the wise man model is an artifact, what is going to take its place? We not see a ready substitute, but perhaps these firms are enlisting specialists in automated drafting. A different respondent, one of the wise men himself (albeit in a different department) explained:

An “expert systems model” is replacing the model you describe. It is like the Three and a Half Minutes quote you have at the outset. The English firms pioneered this model–they spent millions on building these expert systems where all you have to do is punch in parameters and out pops a contract. With these contracts, it is like playing horseshoes. You get a lot of “almost there” contracts. The pari passu issue is the tip of the iceberg. Many lawyers, including partners, don’t even understand how ordinary subordination works in the commercial context, let alone the sovereign context. I bet if you did a survey of junior partners that two out of three would not understand basic subordination. . . . The mentorship mode, which was what you call the wise man model, is gone. The place to look for the new model is the big accounting firms – they are already using systems. . . . I suspect some of the big US law firms have already moved to it . . .?

iii. The Latin Effect and the “Fairness” Factor The pari passu clause is the only clause in the standard sovereign debt contract that sports a bit of Latin. Moreover, it articulates a vague notion of fairness and equality for all creditors. Finally, it shows up at the front of the sales document, often at the top of the cover page. A fair question is whether these three characteristics make it a unique clause; one that is especially susceptible to stickiness. Surely, a clause tackling something specific, like the place or currency of payment, would not be so resistant to modification. And certainly, terms such as the interest rate change from deal to deal. A clause that has been used for over a hundred years and that sets forth some general notion of fairness is probably more difficult to change than most of the standard clauses in the debt contract. In one of the first interviews, a junior partner working on the sovereign debtor side said: Can you imagine the negotiation dynamic? It doesn’t matter that the creditors themselves disagree with Elliott . They like the notion of inter-creditor parity . . . you know, equal treatment. How do I go to the creditors and say that I want to remove a clause that says that all of you will be treated equally?

In short, the claim is that the negative signal of asking for a fairness term to be altered can be much stronger than the generic negative signal story. Perhaps certain types of clauses (vague, fairness terms, with a bit of Latin) tend to be stickier than clear terms laid out in plain English.

iv. Pariahs and Vultures The aggressors in this case, the ones pushing for the expansive interpretation of pari passu , are unusual. They are vulture funds; supposedly the bottom-feeders of the industry. In other words, they are pariahs. Both the Official Sector institutions such as the IMF and the Paris Club and the anti-debt activists, such as Jubilee International, who rarely agree on anything, condemn the vulture funds. 74 There is little ambiguity as to who are the bad guys. Almost all our respondents agreed that Elliott had acted opportunistically, pushing an interpretation of the clause that few believed to be plausible. The windfall it had received, at Peru’s expense, outraged many (and caused envy in others). 75

Academic research on bargaining suggests that perceptions of fairness play a role in determining whether parties settle disputes. 76 When one side perceives that the other side is acting unfairly, the aggrieved party is often willing to suffer costs so as to reduce its opponent’s payoff. 77 This version of the pariah/negotiation story does not fit neatly with our facts, however. By refusing to alter their clauses, the sovereigns and the big creditors have not hurt the vulture funds that supposedly feed on them. Instead, they have assisted the vultures by retaining their own vulnerability to litigation. 78 To be sure, it might be that the big players and their lawyers are refusing to modify pari passu because they feel their opponents are pariahs and thus not worthy of engagement. But their refusal to act is not because they think they are imposing costs on the other side. Instead, it may be that their view of the other side influences their assessment of the outcome of pari passu litigation in a New York court (as reflected in comments such as

74 See John Barham, Vultures of Profit , PORTFOLIO .COM , July 2008 (available at http://www.portfolio.com/news- markets/international-news/portfolio/2008/06/16/Vulture-Funds-Sue-Over-Bad-Debt). 75 There have even been articles trying to shame the big law firms for representing vulture funds. See James Lewis, The Vultures , LEGAL BUSINESS , September 2007 at 112. 76 E.g., Max H. Bazerman & Margaret A. Neale, The Role of Fairness Considerations and Relationships in a Judgmental Perspective of Negotiation , in BARRIERS TO CONFLICT RESOLUTION 90-91 (Kenneth Arrow et al. eds., 1995); Ernst Fehr & Simon Gachter, Altruistic Punishment in Humans, 415 NATURE 137 (2002); Andrew Oswald & Daniel Zizzo, Are People Willing to Pay to Reduce Others' Incomes?, ANNALES D’E CONOMIE ET DE STATISTIQUE , July/December at 39. 77 Id. 78 With the exception of the small set of cases where the sovereigns have removed moved to using trust structures and removed the individual right to sue. “a New York court would never make a decision like that in Brussels”). 79 We are unaware, however, of any direct behavioral evidence suggesting that a disregard for an adversary leads to an overconfidence bias. 80

Ultimately, we are un-persuaded that the observed resistance to innovation in the sovereign debt industry is a function of the fact that the vulture funds are pariahs. As a threshold matter, if these players truly were as bad as some suggest, they would not be able to buy sovereign debt in the over-the-counter market. Moreover, in the course of this project we talked to a number of players on the vulture fund side and they struck us as far removed from the Darth Vader image that numerous news stories have made them out to be.

The two most glamorous figures on the vulture side of the equation are Paul Singer, the principal owner of Elliott Associates and Kenneth Dart, of the Dart family (who are largest manufacturers of the Styrofoam cup). Part of their glamour likely stems from the combination of the shroud of secrecy that surrounds them and their supposedly fabulous wealth. 81 We met neither in the course of the project. However, we did talk to a number of those who had worked on sovereign matters on the vulture side. In the sovereign context, Jay Newman of Elliott Associates is the best-known figure. He is the supposed mastermind of the strategy that was used against Peru. Immaculately dressed when we have seen him, he looks like a runner–leaner and fitter than anyone else we spoke to in the context of this project. 82 Newman knows the sovereign bond world inside and out and, we suspect, takes pleasure in the fact that he and his team actually read the contracts. 83

Newman and his team are constantly innovating, looking for ways to recover on the distressed sovereign debt that they have purchased at deep discounts on the secondary market. Among the most creative strategies they have used has been the one they utilized (supposedly, with great success) against the Congo (Brazzaville). Newman’s team embarked on an investigation of the corruption of the Congo

79 This could relate to research finding that, under conditions of ambiguity, people think that referees will/should rule in their favor. E.g., Linda Babcock & George Loewenstein, Explaining Bargaining Impasse: The Role of Self Serving Biases, 11 J. ECON . PERSP . 109 (1997). 80 In theory, this does not have to be an irrational prediction. Judges who recognize that one side is acting opportunistically may be more willing to penalize that side. 81 See Elizabeth Lesly, The Darts: Fear, Loathing and the Foam Cup , BUSINESS WEEK , July 10, 1995 (http://www.businessweek.com/archives/1995/b343279.arc.htm ); Kambiz Foroohar, The Opportunist , BLOOMBERG NEWS , February 2008 ( http://elliottprofiles.com/pdf/BloombergMarkets-TheOpportunistFebruary2008Eprint.pdf ). 82 An article in Playboy describes Newman as always dressed in a black suit (and apparently his nickname is “the undertaker”). See Aram Rostom, Vulture Capitalism , PLAYBOY , December 2010. 83 Among the ironies in this story is that Newman and Buchheit both have the same nickname, “the undertaker” (they both work on distressed debt). Adding to that, Buchheit was Kenneth Dart’s lawyer in Dart’s infamous (and successful) battle against the Brazilian government over its sovereign debt. government and found a good deal of evidence. 84 The Congolese powers were not pleased; Newman’s team, however, had managed to get their findings onto the website of an anti-corruption NGO (Global Witness) and also to garner significant press attention. Apparently, the Congo’s government decided that it was not worth fighting Elliott any longer and settled the debt claims. 85

In effect, Elliott had managed to come up with yet another creative solution to the problem of being unable to enforce against a sovereign. It would have done Newman’s team little good to simply uncover dirt on the Congolese president and his son: One cannot demand payment in exchange for keeping negative information about someone secret; that’s garden variety blackmail. But Elliott didn’t trade the information for the settlement that it reputedly received. The impetus for a settlement was the implicit threat of future information disclosure that would occur as part of the actual lawsuits that had been filed. As a legal matter, it was perfectly appropriate for Elliott to ask for the payment of the obligations it was already owed, even if part of the incentive for the Congo to make those payments was the fear that more information about presidential shenanigans would be unearthed.

At bottom, the lawyers who work at, and sometimes run, these funds are, like Newman, among the most sophisticated players in the industry. The ones we spoke to often had pedigrees that rivaled their sovereign-side opponents; generally having cut their teeth at the elite sovereign-side law firms before moving to the banking side of things. They also tend to be actively involved in industry groups and legislative reform proposals. And, some of their staunchest opponents were willing to concede that these funds not only provided valuable liquidity but also were exceptionally innovative. 1

84 An article in The Australian , describing some of what Newman’s team unearthed, began: On a spring morning in Paris three years ago, a young African oil executive went on a shopping spree. He spent £1,600 at Escada on the Avenue Montaigne, and £3,700 a few doors down at Christian Lacroix. A couple of weeks later he spent £4,000 at Ermenegildo Zegna and £3,200 at Louis Vuitton. By the end of 2005, Denis Christel Sassou-Nguesso [the son of the President] had charged more than £112,000 to his personal Hong Kong-issued credit card, up from £64,000 the previous year. . . . When Sassou-Nguesso senior [the President] visited New York for a recent United Nations summit, bottles of Kristal champagne popped as his entourage ran up a hotel bill of £169,000–of which £100,000 was paid in cash. Yet he repeatedly claims his country is too poor to pay off its debts. Tony Allen-Mills, Congo Sapped of Riches as Denis Menaces Boulevard Saint-Germain , THE AUSTRALIAN , June 17, 2008 (available at http://www.theaustralian.com.au/news/lifestyle/congo-sapped-of-riches/story-e6frga06-1111116650941 ). 85 According to PLAYBOY , The country settled with most of the aggressive vulture funds at 55 cents on the dollar, but Newman and his financier at Elliott scored better than the others. Apparently by agreeing to stop providing reporters with negative information about the ruling family, Newman is said to have collected about $90 million from the Congo. He had paid less than $20 million for the old debt. His biggest cost may have been for lawyers, private eyes and lobbyists. See Rostom, supra note 33; see also Felix Salmon, Vulture Funds Exposed in Playboy , COL . JOURNALISM REV ., The Audit, February 24, 2011 (available at http://www.cjr.org/the_audit/vulture_funds_stripped_bare_in.php?page=all&print=true ).

As we have come to see it, the vulture funds play an important role in disciplining the market. They make money by exploiting arbitrage opportunities; and they find those arbitrages by reading the contracts carefully. 86 Long term, one effect of the vulture funds successes may be to improve the quality of lawyering in the industry. Indeed, the pari passu saga might suggest that there is an undersupply of vulture fund litigation.

v. Sovereigns are Different The industry we examine is unusual for a variety of reasons. Sovereigns cannot easily be sued and their affairs are intensely political. Moreover, the market is relatively small and dominated by a handful of law firms and banks around the world. Three characteristics may be particularly important in limiting the extent to which our story can be generalized.

First, the agency problems we have identified above are magnified in the sovereign debt context. Sovereign clients are unusual in that their time horizons can be short in comparison to a corporate debtor. Unlike a corporate board at a U.S. corporation, governments in many democracies face real risks of being voted out of office. That in turn may mean that the finance ministers care little about the impact of any given contract term on a hypothetical default litigation ten years from the date of issue. If there is a default, this finance minister will likely have retired long ago. The primary goal of the finance minister is to get the deal done, so as to have the capital in hand. If so, she will not be pleased to hear that the debt issuance was derailed because her lawyer was trying to negotiate over a clause that has been around for over a hundred years and that every other sovereign issuer uses. What we have, then, is a situation where there are two agents whose interests’ conflict with that of the sovereign client. One agent, the Finance Minister for country X, generally expects to have a short-term tenure as representative of the country and is unwilling to pay in time and legal fees for revisions to boilerplate language. The other agent, the outside lawyer, is more likely to have a long-term relationship with the sovereign client. However, the agent with the shorter-term horizon decides whether the longer-term agent gets paid on time or gets to keep his job. In particular, the finance minister is likely to resist any efforts by the lawyer that interfere with his ability to raise the billions of Euros or dollars that he requires. The lawyer, however, will be the one responsible if and when a restructuring has to be done in the future. Hence, the lawyer-agent is

86 One of the most recent successes of one of Elliott’s incarnations was with respect to set of defaulted Argentine bonds known as FRANs (Floating Rate Accrual Notes). These bonds had had a floating interest rate that was determined as a function of third- party estimates. The contracts failed to specify what would happen if the debtor declared a moratorium, as Argentina did in late 2001. NML Capital, the litigant in question, argued that the rate to be applied during the period of the default, as during any other period, was the market rate (which, given that Argentina was in default, was sky high, rising as high as 101%). Argentina tried to argue that surely, the contracts could not have meant to pay such an exorbitant rate and that to force it to do so was unconscionable and usurious, among other things. The court, however, found the language of the contract clear. Plus, Argentina, with its sophisticated lawyers, could hardly argue that it had not known what it was bargaining for. Argentina lost. See NML Capital v. Republic of Argentina, 621 F. 3d 230 (2d Cir. 2010); NML Capital v. Republic of Argentina, 17 N.Y. 3d 250 (2011). motivated to protect himself against bad outcomes in the future, but conditioned on keeping the current finance minister satisfied. What we should expect to see, then, are lawyers working to ensure that deals go through on time, while seeking the least cost means of producing contracts that are free of litigation risks.

Consistent with the notion that the agency relationships with a sovereign client are unique, lawyers working in areas such as M&A, tax and private equity have suggested that our findings are idiosyncratic; that their clients demand contract provisions that are carefully negotiated and that every term is well understood by both client and lawyer. 87

Second, sovereigns sometimes issue debt for reasons other than to raise capital. Some sovereigns issue debt to set a benchmark rate for loans that firms from their country can use when they go to the international debt markets. 88 By doing an issuance, the sovereign is testing the market and setting a price for the country risk. For a sovereign issuing debt for these purposes, the key is to issue a debt instrument that looks like all the other instruments in the market.

Third, the billing structure for sovereign deals may be different. For many law firms, the primary reason to represent a sovereign on a bond issue is reputational; it enhances the firm’s image to be able to claim, for example, that the firm represents the Government of Spain. Moreover, the law firm anticipates that corporate firms from that sovereign state will look to the firm doing the sovereign deal for representation in their private transactions. If marketing is a major motivation for doing the sovereign deal, it follows that the firm’s financial return on the issuance is modest. One respondent described a recent deal done for the Ukraine [2008] as having “essentially [been] done for free.” The billing structures for these sovereign deals, therefore, might also be playing a part in the reluctance to modify standard boilerplate.

vi. Busy Respondents Our respondents were largely high-level officials, executives and senior practitioners. Their time was scarce and they might not have had either the interest or inclination to spend time thinking about our questions. Had the stakes been higher and had they an incentive to think hard about the question (for example, had a paying client asked 89 ), their answers may have been different. On the other hand, we were

87 But see Megan Murphy, Why do Lawyers Witnesseth Such Dreadful Legalese ? FIN . TIMES , March 25, 2009. 88 In our interviews, the two examples we were given were Abu Dhabi and China, both of which issued bonds at times when they did not look to be in need of external capital. 89 In at least one case, a client did ask its lawyers why they had the type of pari passu clause that they had. The back story being that a competitor firm had suggested to the client that its lawyers might have fallen down on the job in not making sure that the pari passu clause was optimal. The result: The firm in question produced a research memorandum research showing not only that a number of other sovereigns used similar clauses, but also that some of those sovereigns were similarly solid credits, thereby not asking about an obscure issue; this was a topic familiar to every one of the respondents. More importantly, almost no one questioned the premise that there was stickiness in boilerplate contracts. Sticky boilerplate was a common phenomenon to the respondents. Multiple respondents told stories of their time as associates being chastised by a senior partner when they attempted to improve the language in a standard form contract. We heard this story from both the U.K. and U.S. lawyers: Being reprimanded for having attempted to improve some otherwise incoherent standard language was an important learning occasion for them. Being disciplined for attempting modifications of the standard documents may be one of the modern rites of passage in an elite large law firm. Perhaps the transition in question is the one from the young lawyer (who thinks that transactional lawyering and contract drafting is about protecting against all contingencies) to the experienced lawyer (who understands that real world practice is about getting the deal done).

vii. Disingenuous Respondents and the “Hairy Hand” A handful of respondents from the hedge fund world, upon reading an early draft of the book, told us that we had been taken in by the lawyers we interviewed: In their view, pari passu was not sticky. By looking primarily at the pari passu clause, they asserted, we had missed the forest for the trees. The real action in these contracts over the past ten years had been in the inclusion of Collective Action Clauses (CACs) together with restrictions on the individual right to sue and the adoption of trustee mechanisms. These provisions had essentially decimated the abilities of creditors to sue sovereigns. In other words, pari passu had been rendered irrelevant. While contracts still had pari passu clauses on their front page, what was described to us as “Buchheit’s hairy hand” had been decimating creditor rights on the back pages, rendering the bonds “functionally unenforceable” according to one of our frustrated readers.

These respondents were right to make the point that some contract terms in sovereign debt contracts have changed significantly in recent years. Most importantly, CACs authorize 75% of the holders of any issue of New York-law sovereign bonds to approve changes in payment terms that bind all

raising the possibility that credits of a certain kind might be using that particular form of the clause. So, the only occasion when (to our knowledge) a client questioned its lawyers as to their choice of pari passu clause, the lawyers produced a memo justifying their use of the existing clause as opposed to trying to determine whether the clause could be improved. But even here, the story does not readily make sense. The problematic clause had been originally put in place by a different (and prominent) firm. And the current lawyers had been instructed to essentially keep the debt issuance program running in the fashion that the prior law firm had established. Thus, the current lawyers could have easily (we think) blamed the form of the clause on the prior firm and then proceeded to revise it. bondholders. These clauses have likely diminished the potency of the holdout creditor problem. 90 But pointing to the fact that some contract provisions changed—after a decade worth of efforts by the Official Sector and particularly the U.S. Treasury—hardly shows that contract terms are not sticky. The big move towards CACs did not begin until three years after the Elliott decision and widespread adoption in the New York market took several additional years. 91 During that period, the holdout threat was real.

If anything, the experience with Collective Action Clauses shows that it can take enormous and sustained effort in order to revise key boilerplate terms. 92 Moreover, the puzzle of pari passu remains even if one thinks that Collective Action Clauses have reduced significantly the potency of the holdout risk. Vulture funds clearly face a substantial burden if they are to acquire more than 25% of the outstanding bonds on the secondary market so as to forestall collective action. But nevertheless, unless other barriers are also included, the holdout risk remains real. Additional barriers have been devised, but they appear only in a small number of new issuances. For example, fewer than 10% of issuers (as of this writing) have begun using trust structures in their bonds, where litigation powers are concentrated in the trustee, significantly reducing the incentives for an Elliott or a Dart to want to sue on its own. It is not only difficult in these new bonds for an individual creditor to usurp the trustee’s litigation authority, but whatever revenues are obtained have to be shared among the other bondholders in a non-discriminatory fashion. 93

While the solution of adding new terms to a contract that undermine the effect of other terms in the same contract may be able to solve the problem of formulation error in the short-run, the practice weakens the communicative properties of all the terms in the contract, thus reducing their reliability as signals of what the parties really intended. In short, the risk of misinterpretation (and consequent exploitation by an arbitrageur) remains. So long as there is still a risk of court misinterpretation, why not revise or delete the clause to eliminate that possibility?

Tentative Generalizations The sheer number of complicating factors that distinguish the sovereign debt case study from other areas of transactional practice suggests caution in offering any lessons for the future. Nevertheless, despite the many caveats there remains substantial evidence that the institutional structure of the modern

90 See Michael Bradley & Mitu Gulati, Collective Action Clauses for the Eurozone: An Empirical Analysis (2011 draft) (available on ssrn.com). 91 Bradley & Gulati, supra note 41. 92 See Anna Gelpern & Mitu Gulati, Public Symbol in Private Contract: A Case Study , 84 WASH . U. L. Q. 1627 (2006); Stephen Choi & Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds , 53 EMORY L. J. 929 (2004). 93 See Stephen J. Choi, Eric Posner & Mitu Gulati, Political Risk and Sovereign Bond Contracts (Nov. 2011 draft) (available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1962788 ). large law firm impedes innovation in contract design. Whether we have correctly identified all the contributing causes for inertia and whether that inertia extends to all areas of transactional practice may remain an open question. But the evidence does justify some cautionary comments about the future of transactional legal practice. Our account of the treatment of the pari passu clause following the Elliott litigation gives rise to a fundamental question about the future: What organizational changes must take place within law firms to permit lawyers to effectively engage in transactional work for their clients?

We suggest that lawyers can remain significant contributors in corporate and commercial transactions by focusing on two features: (a) the link between contract design and litigation and (b) innovative contract design. The core expertise of lawyers concerns the operation of courts and other adjudicative venues in administrative agencies, arbitration and elsewhere. These institutions have specialized rules of procedure and evidence that take considerable time, instruction and experience to master. The lawyer’s advantage in this realm is more significant even than her mastery of the substantive rules of law and of government regulation. The lawyer’s advantage is in her knowledge of the process used to enforce contracts. Understanding litigation risk, in its various forms, is thus the core competency from which lawyers can derive comparative advantage in designing transactions for their clients.

One of us has written elsewhere about the importance of this link in contract design. 94 This link seems underappreciated in practice, other than in a handful of specialized fields. As noted earlier, litigators and transactional lawyers rarely combine their specialized skills at the level of the individual transaction: the firm that litigates a transaction is often not the firm that did the deal in the first instance. If law firms can manage to connect these two functions, they can capture more of the gains from providing innovative contract design.

Law firms have ceded much of the production of innovative transactions to non-law organizations, such as investment banks, marketing consultants and commodities exchanges. Despite the absence of intellectual property rights protection for contract design, 95 there is a significant and under exploited first-mover advantage in the development and marketing of novel contractual products. 96 Wachtell Lipton’s process of development and marketing bore close resemblance to the development of new industrial or consumer products but, remarkably, has not been widely followed in transactional legal practice. This is due partly to legal education and culture, but perhaps more significantly, as noted above, to deficiencies in the organizational structure of law firms, including hourly billing and compensation, the secondary role given to R&D, and the manner in which legal services are marketed. Indeed, the big law

94 See Robert E. Scott & George G. Triantis, Anticipating Litigation in Contract Design , 85 YALE L. J. 814 (2006). 95 For a review of the barriers to contractual innovation, see Charles J. Goetz & Robert E. Scott, The Limits of Expanded Choice, 73 CAL . L. REV . 261 (1985). 96 See Peter Tufano, Financial Innovation and First Mover Advantages , 25 J. FIN . ECON . 213 (1989). firm, with its emphasis on volume production and high leveraging of legal talent, appears to actively impede the efforts of the lawyers in the organization to be entrepreneurial. One hedge fund manager said to us, “we never hire the law firm, we only hire lawyers.” The successful law firms of the future, therefore, are likely to be those that evolve governance mechanisms that facilitate and support innovation. New forms of governance that overcome the obstacles to innovative contract design can preserve a meaningful role for the modern law firm and the legal profession in the full range of corporate and commercial transactions including, but certainly not limited to, the field of sovereign debt.

CHAPTER XI: EPILOGUE Some things were lost that should not have been forgotten. History became myth. Myth became legend.

From one of our respondents: A quote from Tolkien’s, “Lord of the Rings”

As we write this in December 2011, the saga of the pari passu clause continues to follow an unpredictable course. The past two years have seen three twists in the story.

First, there was the Ecuadorian drama in 2009. Ecuador selectively defaulted on some of its prior bonds on the grounds that they were “illegitimate.” There, we saw some players on the establishment side (almost) embrace the meaning of the pari passu clause that the Brussels court implicitly adopted in Elliott. Second, the question of what it means for a sovereign to affirm in a contract that a creditor is pari passu surfaced, at least in backroom discussions, in the context of the bailout that was given by the Eurozone in 2010 to Greece. In designing the loan instrument governing their loans to Greece, members of the Eurozone wanted all of the protections they could get and, therefore, included a pari passu clause that, roughly speaking, said that the debt under the Eurozone facility was pari passu with all other unsecured debt of Greece. That then raised the question of what kind of protection they were getting by including this clause. After all, the standard story that had been told about the pari passu clause prior to this was that Official Sector lending (of which the Eurozone loan facility was clearly an example) had priority regardless of the pari passu clauses in private debt contracts. So, by insisting that they were pari passu were the Eurozone nations lowering their priority ranking vis-à-vis, say, the IMF? That made no sense.

Third, there was the litigation by holdout creditors against Argentina in federal court in New York: In late 2011, as we were putting the finishing touches on this book, a federal district court judge finally ruled on the meaning of pari passu in a sovereign debt instrument. This judge said that he had little trouble discerning that Argentina’s actions vis-à-vis its creditors clearly violated the clause. And although he didn’t quite adopt the meaning of the clause endorsed in the Brussels case a decade earlier, he didn’t reject it either.

These twists in the story get us no closer to unearthing an original meaning of the clause. But by late 2011 we were probably the only ones who cared a whit about that. The twists in the tale do, however, provide additional evidence in support of the broader agency cost story. Ecuador 2008-09 In late 2008, Ecuador defaulted on two of its bonds (due 2012 and 2030) but continued making payments on a third bond (due 2015). 97 This was the rare case where a sovereign refused to pay even though it had available resources (Ecuador, while in some financial trouble at the time, still had considerable oil revenues). 98 Rather, the decision to default was grounded in the supposed illegitimacy of the debt issued by a prior government. In this case, the leftist government of Rafael Correa had replaced a more capitalist regime. The rationale given by the Correa government for defaulting on the 2012 and 2030 bonds but not the 2015 bonds was that the 2012 and 2030 bonds were illegal and illegitimate (the Correa administration had issued the 2015 bonds itself, hence those were deemed legitimate). 99 On the basis of the findings of an audit committee, whose international members were almost all prominent anti- debt activists appointed by the Correa regime, the Ecuadorian government asserted that aspects of the 2012 and 2030 bonds were lacking in requisite formalities. 100 Among other things, the Correa government attacked certain of the standard features of sovereign bonds that were contained in the Ecuadorian bonds (such as the acceptance of jurisdiction in New York and the waiver of sovereign immunity) as unjustified and unacceptable incursions on its sovereignty. 101

The creditor community reacted to Ecuador’s claims with anger and concern. The anger was directed at the implausibility of the justifications for the default. The concern was over the possibility that other Latin American countries would follow in Ecuador’s footsteps and also assert that the debts incurred by prior governments were odious or illegitimate and therefore did not need to be repaid. 102 Surprisingly, the meaning attributable to the pari passu clause was the subject of debate in a controversy that involved an altogether different set of actors than the ones we had interviewed for this book. 103

The Ecuador dispute engaged parties who had little or no familiarity with the Elliott episode, and yet they also focused on the pari passu clause: Could it possibly be used to attack Ecuador’s

97 See Abby McKenna & Federico Kaune, Ecuador’s Dangerous Game , Jan 28, 2009, Portfolio.com (available at http://www.portfolio.com/views/columns/2009/01/28/Latin-American-Debt-Defaults). 98 See Thomas Trebat, The Curious Case of Ecuador’s Default and Why it Matters , EconoMonitor (Roubini Global Economics), January 2009 (available at http://www.economonitor.com/blog/2009/01/the-curious-case-of-ecuadors-default-and-why-it- matters/). 99 See Mckenna & Kaune, supra note 1. 100 See Arturo C. Porcezanski, When Bad Things Happen to Good Sovereign Debt Contracts: The Case of Ecuador , 73 L. & CONTEMPORARY PROBLEMS 251 (2010); Lee C. Buchheit & Mitu Gulati, The Coroner’s Inquest , 28 INT ’L FIN . L. REV . 22 (September 2009). 101 See Porcezanski, supra note 4. 102 See Adam Feibelman, Ecuador’s Sovereign Default: A Pyhrric Victory for Odious Debt , 25 J. Int’l Banking L. & Reg. 357 (2010). 103 By happenstance, one of us had done research on the question of what types of debts might fall under the category of “Odious Debts” (debts of a prior government that in some narrow circumstances arguably do not transfer to a successor government). As a result, both sides of the Ecuadorian dispute solicited our support. misbehavior? 104 Some who were involved in strategizing as to how to counter Ecuador’s default asked: “Couldn’t creditors holding defaulted 2012 and 2030 bonds use the pari passu clause to seize the payments being made to the favored creditors holding 2015 bonds?” 105 The irony was that this was the precise meaning of the clause that Elliott had asserted almost a decade ago. Now, the claim was being pressed by at least some of the lawyers who earlier had pilloried the Elliott interpretation, and those on the Ecuadorian side were worried that this argument might be effective. This was not hypocrisy: By 2010, Elliott was distant history for many of the younger lawyers in the sovereign debt community. But this newer generation of sovereign debt lawyers was vaguely aware that one of the possible meanings of the pari passu clause was the proportional payments interpretation that had won the day in Brussels in 2001. And that meaning, which seemed outrageous in the context of the Peruvian payments on its restructured debt in 2001, seemed to make intuitive sense with respect to Ecuador in 2009. Ultimately, however, some of the senior lawyers who had fought the battle with Elliott silenced the talk of using the pari passu clause to attack Ecuador’s selective default. Illustrative of this reaction is a report that Barclays issued telling its clients that the Elliott argument about the meaning of pari passu had been discredited and was not likely to be of assistance in the Ecuador case. 106

The Ecuador episode brings to the forefront a question that we have taken pains to avoid. Perhaps the reason why pari passu has remained in debt contracts despite Elliott is that the markets actually like the Elliott interpretation. As one colleague put it, “Ex post, they may not have liked what happened in Peru. But ex ante, maybe the Brussels interpretation of the clause is precisely what everyone wanted.” If indeed this were the case, we would have a neat conclusion to the story. We could announce that we had discovered that what these market insiders were saying was not in line with what they were doing. And there is something to this point. After all, there are the following two findings from our research: On the one hand, there was all the brouhaha over Elliott and the expressions of outrage from most of our 100-plus respondents. And, on the other hand, the data show that the use of the clause

104 We did not conduct systematic interviews here, for the simple reason that we had not expected to find anything of relevance to our pari passu project. 105 McKenna & Kaune, supra note 1; see also Naomi Mapstone & David Oakley, Ecuador Vows Partial Bond Buy-Back , FIN . TIMES , January 16, 2009. Robert Kozak, Ecuador’s Debt Strategy Likely Leading to Legal Battles , DOW JONES NEWSWIRE , February 1, 2009 (available at http://forexdaily.org.ru/Dow_Jones/page.htm?id=412853 ); Analytica Securities, Ecuador Weekly Report , March 23-29 (2010) (reporting on the legal strategies being contemplated to counter the Ecuadorian default in 2009) (http://equatorialis.net/reporte_economico/24%20March%202009.pdf ). 106 Joe Kogan, Pari Passu and the Ecuador 15s , BARCLAYS CAPITAL , EMERGING MARKETS RESEARCH , Jan. 27, 2009 (“Contrary to the assumptions of many investors, the pari passu clause does not ensure that a sovereign will service its bonds equally.”). Some years from now, if another Ecuador-type episode occurs, we wonder whether the Elliott interpretation of the clause will emerge more forcefully from the establishment side. In the context of the Ecuador episode, one senior sovereign-side lawyer said to us in jest: “Where is Jay when we need him most?” (referring to lawyer, Jay Newman, considered one of the masterminds behind Elliott’s success against Peru).

actually increased after Elliott , thus indicating that the respondents were saying one thing and doing another. However, there are at least two reasons why this conclusion to our story, neat though it is, does not withstand sustained analysis.

First, assuming that the Elliott interpretation is the one preferred by the sovereign debt community, the puzzle of sticky boilerplate does not disappear. In such a case, one should see lawyers modifying their new contracts to make clear that the pari passu clause would now have the meaning given by the Brussels court. After all, the vast majority of pari passu clauses being used today are vague and imprecise and thus still unclear as to their meaning; a confusion that has been exacerbated by the large volume of commentary condemning the Brussels interpretation. Perhaps hoping for precisely this kind of clarification, one of the creditor associations (where Elliott was a member) proposed such amended language in the wake of Elliott . No one adopted that formulation (although a handful of nations who were already using something close to that proposal continued to use the same language in subsequent issuances). The point here is simply that the stickiness puzzle remains whether one favors the Elliott interpretation or not.

Second, if the sovereign market really did have a preference for the Brussels interpretation, we would not expect to see the most prominent lawyers in the field loudly proclaiming that the Brussels interpretation was bunk. If there really were a preference for the Brussels interpretation, statements such as those would be contrary to interest since opponents could point to those statements in litigation in arguing for an alternate meaning.

The Eurozone Crisis: 2010-11 To emphasize the never-ending aspect of the puzzle of pari passu , the question of what exactly it means to have pari passu protection against a sovereign who cannot go bankrupt and liquidate has resurfaced in the context of the ongoing Eurozone sovereign debt crisis. The EU/IMF rescue facility that was put in place for Greece in early 2010 has a pari passu clause. 107 To be clear, this rescue facility is not a public bond issuance. It is a loan facility designed to establish the rules governing the loan granted by the Eurozone nations to Greece. Given that context, we should be able to say with confidence that the inclusion of a pari passu clause in this document was not attributable to an inadvertent copying from some prior version of the document—there have been no prior versions of the Eurozone crisis or this type

107 The fact that the loans out of the rescue facility are pari passu with other unsecured Greek debt means that the Official Sector lending does not receive its usual seniority. See Daniel Gros, Bail Out Countries, But Bail In Private, Short Term Creditors , VOX EU, n.5, December 5, 2010 (available at http://voxeu.org/index.php?q=node/5891#_ftn3 ). For the actual provisions, see Euro 80,000,000,000 Loan Facility (May 8, 2010) (available at http://www.minfin.gr/content- api/f/binaryChannel/minfin/datastore/30/2d/05/302d058d2ca156bc35b0e268f9446a71c92782b9/application/pdf/sn_kyrwtikoimf_ 2010_06_04_A.pdf ). The provision is at Section 4, Representations, Warranties and Undertakings, Subsection 1 (a) (Representations). of loan. Moreover, the document was supposedly tailored to particular needs of this cross-nation lending facility. As a conservative estimate, given the number of nations that were involved in designing this loan facility, there must have been at least two-dozen senior lawyers from a dozen different nations who evaluated this document before it was released to the public. The inclusion of the pari passu clause, therefore, must have been consciously considered. That, in turn, suggests that the clause provides the kind of protection that the creditors in this instance—the various Eurozone nations participating in the loan to Greece--desired. We suspect that the debtor in this case, Greece, was utterly powerless and had to take whatever terms were provided by the lenders. The natural question is what protection did the Eurozone lenders under this facility expect from the pari passu clause? Does this mean that no one else can get paid before them (the Elliott interpretation)? Or does it mean that they expect to get paid first, before the other creditors, and that the pari passu clause is only there for decoration? Or is it a protection against the possibility that Greece will somehow engineer an earmark of assets to some other creditor, just as it used to do in the 1800s?

One of our respondents thought the use of the pari passu clause in the Greek facility confirmed the story that he had told us earlier about how clauses migrate from one set of documents to another, even when they make no sense in the new context (a story that we have tried hard to resist). The terms for the Greek facility, he insisted, had simply been copied from documents for private bonds. No one had given much thought to the fact that this was an instrument governing Official Sector lending rather than private creditor lending and that the clause did not fit. Indeed, under the Elliott interpretation the inclusion of the clause reduced rather than enhanced the priority of the Eurozone lenders.

Argentina; September 28, 2011 The vulture funds were fighting a different battle at the time the Ecuador episode occurred. In the biggest sovereign default in history, Argentina had defaulted in 2001 on upwards of $100 billion of bonds. A significant portion of the creditors (roughly 70% or so) had accepted a restructuring offer in 2005, where their claims received a significant haircut (returning roughly 30 cents on the dollar). 108 The remaining creditors refused to participate and turned to the courts instead. Included among these holdout creditors were prominent distressed-debt investors such as Elliott, Dart and Stone Harbor.

To induce its creditors to accept the restructuring in 2005, Argentina promised those who agreed to the restructuring that the other creditors would not get paid. 109 Moreover, Argentina didn’t merely

108 Lester Pimentel, Argentina May Default Without Debt Accord , Bloomberg, October 13, 2009 (available at http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aGTGKOvk9DW0 ) 109 See Rodrigo Olivares-Caminal, Sovereign Bonds: A Critical Analysis of Argentina’s Debt Exchange Offer , 10 J. BANKING REG . 28 (2008) (describing the “padlock” law). affirm that it was not going to pay those non-tendering creditors: the legislature enacted a law—the “padlock” law—that formally restricted the state from agreeing to a deal with the non-tendering creditors. 110 The problem with Argentina’s strategy was that this law arguably gave one set of creditors (the ones that had tendered in the exchange offer) a superior rank to the others (who had refused to tender). And the meaning of the pari passu clause that had been asserted in opposition to the decision in Elliott almost a decade ago was that the clause provided protection against a sovereign enacting legislation to provide one set of creditors favored access to a particular stream of earnings in preference to others. In fact, in 2004 Argentina had submitted a brief in opposition to NML Capital, an Elliott subsidiary that was asserting the Brussels interpretation. In their brief, Argentina’s lawyers (Cleary Gottlieb) argued that the purpose of the pari passu clause was to “prohibit the Republic from creating, or permitting the creation of, senior debt that has a legal right (‘rank’) to be paid before the debt issued under the [Fiscal Agency Agreement] and allegedly held by NML.” 111

The debate over pari passu that had occurred until that point can be summed up in the following way: In order to counter the Brussels interpretation, the lawyers on the sovereign side had had to conjure up a meaning for the clause (it would hardly do for them to be arguing to a judge that they knew that Elliott was wrong but that they had no clue as to what the clause actually meant). It was also important that the meaning that was asserted was one that rendered the clause harmless and essentially meaningless in the modern context so as to protect against future episodes of that kind. The meaning that they settled on was one that Lee Buchheit, Philip Wood and other luminaries in the world of sovereign debt had suggested in their articles and books. 112 Recall that Buchheit, Wood and others had suggested that perhaps the clause had been originally aimed at protecting against legislative actions that granted preferential access to sovereign funds to one subset of creditors over others. 113 This suggestion as to a possible meaning of the clause, one irreverent observer explained, was a narrow one:

Violating the pari passu covenant in a debt instrument has at least one thing in common with the sin of bestiality; it is unlikely to be done unwittingly. In both cases, the offense requires an affirmative, deliberate action on the part of the transgressor. To violate the pari passu clause in a

110 Id. 111 See Memorandum of Law in Support of the Motion by NML Capital, Ltd,. For Partial Summary Judgment and For Injunctive Relief Pursuant to the Equal Treatment Provision, pg. 13, S.D.N.Y. 08 Civ. 6978; 08 Civ. 1707; 08 Civ. 1708 (October 20, 2010). 112 See, e.g., PHILIP R. WOOD , LAW AND PRACTICE OF INTERNATIONAL FINANCE 156 (1980) (“In the case of a sovereign state, . . . [t]he clause is primarily intended to prevent the earmarking of revenues of the government or the allocation of its foreign currency reserves to a single creditor and generally is directed against legal measures which have the effect of preferring one set of creditors over the other or discriminating between creditors.”); Lee C. Buchheit & Jeremiah S. Pam, The Pari Passu Clause in Sovereign Debt Instruments , 53 EMORY L. J. 869 n. 10 (2004); WILLIAM TUDOR JOHN , SOVEREIGN RISK AND IMMUNITY UNDER nd ENGLISH LAW AND PRACTICE IN INTERNATIONAL FINANCIAL LAW 95-96 (ROBERT S. RENDELL ED . 2 ed. 1983). 113 See Buchheit & Pam, supra note 16; PHILIP WOOD , LAW AND PRACTICE OF INTERNATIONAL FINANCE , VOLUME 2, 6-23 (1984). sovereign debt instrument, the sovereign debtor must go to the extraordinary length of attempting by legislation or decree to elevate one category of creditor over another in terms of the legal ranking of their claims. Such things just don't happen.

Until they do.

But now Argentina might have done the specific bad act that the leading establishment commentators had said the pari passu clause was meant to prevent. In other words, according to the very arguments offered by Elliott’s opponents, Argentina had probably violated the pari passu clause by enacting a law to favor the creditors who had agreed to exchange their bonds. Gleefully, Elliott (through a subsidiary, NML Capital), argued in reference to the interpretation that Argentina’s lawyers had asserted in 2004:

Argentina’s construction is unduly narrow and renders the contractual right to equal treatment almost meaningless. Under its view, Argentina may always make preferential payments, as it pleases, to favored creditors while continuously denying any payment at all to bondholders who are entitled to equal treatment; the only limitation, Argentina contends, is that it cannot enact laws or enter into contracts that purport to codify such disparate treatment as Argentina’s obligation. This interpretation would relegate a crucial contractual right of bondholders to a paper-thin restriction on Argentina’s ability to formalize its discrimination against them. Even under Argentina’s anemic view of its contractual obligations, however, NML is entitled to partial summary judgment. Through its enactment of the Lock Law, which repudiated all of NML’s claims for payment, Argentina lowered the rank of its payment obligations to NML relative to its payment obligations to other investors. The Lock Law made clear to potential participants that if they did not tender their bonds, they would receive nothing and that no other offer would or could legally be made, whereas if they acceded to the exchange offer they would receive at least some of the interest for which they had originally contracted. 114

To add salt to the wound, Elliott’s brief supported its argument by citing articles written by two disciples of Buchheit, both professors and experts on sovereign debt, Rodrigo Olivares-Caminal of the University of Warwick, and Anna Gelpern of American University. 115 Olivares-Caminal and Gelpern had both suggested in prior writings that Argentina’s law might have violated the pari passu clause. 116

114 Id. at 14 (emphasis added). 115 Id. at 16. 116 Id. As a technical matter, Argentina could (and did) argue that Elliott had misconstrued their prior statements and the general idea behind understanding the pari passu clause as a prohibition on the earmarking of assets for particular subsets of creditors. As a formal matter, under the earmarking argument the pari passu clause bars sovereigns from granting subsets of creditors access to payments from particular streams of income, such as oil revenues–hence the term “earmark.” The Padlock Law that Argentina enacted in 2005 arguably did not grant special access to a stream of earnings. Instead, it barred Argentina from entering into a separate agreement with the subset of creditors who had not entered into the first restructuring. Cleary Gottlieb’s brief explained:

The passage of the Lock Law, which merely prohibits the Executive from unilaterally settling defaulted claims or re-opening the exchange offer without Congressional approval, therefore did not create a “a legal priority to those creditors that entered the exchange offer,” Pl. Inj. Mem. at 15, in the sense prohibited by the pari passu clause or in any other sense. The Law did not create any preferred creditor class. 117

Thomas P. Griesa, the judge before whom the case came, was no stranger to sovereign debt litigation. Griesa had been the judge on the most famous holdout creditor case before Elliott some four decades earlier. That case, Allied Bank v. Banco Credito Agricola de Cartago , involved Costa Rica’s attempt to restructure its syndicated loan debt in the face of a holdout creditor’s attempt to block the deal. 118 That litigation produced what are among the most important cases in the sovereign debt pantheon. When the Argentine litigation showed up in New York after Argentina’s default in late 2001, it is no surprise that the multitude of investor cases were consolidated in front of Griesa, one of the few judges with any familiarity in tackling these matters. If any judge anywhere could be said to know the industry, and understand sovereign restructurings and litigation, it was Griesa. 119 At the time the case was filed, Griesa was a senior judge. He had taken “senior status” in 2000, which gave him the option of skipping a case as onerous as the Argentine litigation was likely to be. He took it on.

Given both the billions at stake in the litigation and Argentina’s determination not to pay the holdouts, it was inevitable that the pari passu issue would arise. It first came up in 2005, but the U.S.

117 Memorandum of Law in Opposition to Plaintiff’s Motions for Partial Summary Judgment and for Injunctive Relief Pursuant to the Pari Passu Clause, S.D.N.Y. 08 Civ. 6978; 08 Civ. 1707; 08 Civ. 1708 (December 10, 2010). 118 In Allied Bank thirty-eight of the thirty-nine banks had agreed to a restructuring. One, Fidelity Union Trust, however, did not want to collaborate and litigation commenced. Judge Griesa’s initial decision had been in favor of enabling the restructuring to go forward, but the Second Circuit reversed him. The combination of cases have now become part of the sovereign debt canon and Allied Bank is generally seen as the first major holdout creditor case. See Allied Bank Int’l v. Banco Credito Agricola de Cartago, 757 F.2d 516, 522–23 (2d Cir. 1985). For a discussion, see FEDERICO STURZENEGGER & JEROMIN ZETTELMEYER , DEBT DEFAULTS AND A DECADE OF CRISES 64-66 (2007). 119 Griesa’s attempts to manage the aftermath of the Argentine default and the various litigations is discussed in Dania Thomas & Marcus Miller, Sovereign Debt Restructuring: Judge Griesa, the Vultures, and Creditor Rights , University of Warwick Working Paper (2007) (available at http://wrap.warwick.ac.uk/26/1/WRAP_Miller_twerp_757.pdf ). Department of Justice and the Clearinghouse filed amicus briefs and the Elliott-Dart team withdrew the issue. In 2011, however, thanks to Argentina’s Padlock Law and the judge’s likely frustration with Argentina’s failure to resolve the litigation, the issue was ripe. Moreover, the U.S. government was no longer eager to intervene on Argentina’s behalf; in the intervening years, it had likely also gotten frustrated with Argentina’s treatment of its creditors. The briefs in this second incarnation of pari passu were filed in late 2010. Judge Griesa though took his time; no doubt understanding the importance of the issue. Industry insiders kept predicting that a decision was bound to come down by June 2011, before Griesa left on his customary summer vacation. June came and went, however, and no decision was released.

Then, at a hearing on a Motion to Amend on September 28, 2011, out of the blue Griesa lowered the boom. Ruling directly from the bench, as opposed to writing a lengthy opinion, he appeared to have made up his mind that the there was a straightforward answer to the interpretive question. 120 Whatever the pari passu clause meant, surely Argentina’s passage of the Padlock law constituted a violation of the pari passu clause. After all, this surely altered the rank of the debt in the holdouts’ hands.

Frustrated with the attempts of the lawyers to complicate the matter, Griesa interrupted them, saying:

I hate to cut off very fine arguments by very fine lawyers, but I have to tell you that I have studied the matter -- I say to both sides -- and it's hard for me to believe that there is not a violation of the pari passu clause accomplished by the congressional legislation in '05 and '10, simply saying that the Republic will not honor these judgments.

It is difficult to imagine anything would [more] reduce the rank, reduce the equal status or simply wipe out the equal status of these bonds under the pari passu clause.

I have to say, I don't think it is a terribly hard question. I regret not hearing more of your argument and Mr. Boccuzzi [Argentina’s lawyer from Cleary Gottlieb] but I don't see -- I think the hard question is what to do. But to me, there's simply no doubt at all that what the Republic has done is to violate the pari passu clause, no matter what interpretation. It can't be interpreted to allow the Argentine government to simply declare that these judgments will not be paid, and that's what they have done.

120 Case Number 08 CV 6978 (Republic of Argentina v. NML Capital) (transcript on file with authors). What we have then, finally, is a New York litigation—ten years after Brussels--interpreting the very same unrevised clause, with Elliott arguing that the clause had an obvious meaning and with the issuer's counsel fumbling around to offer an alternative meaning and being unconvincing because, we suspect, he didn’t know its real meaning. It is worth adding here that Griesa, unlike with the case in Brussels a decade ago, was quite cognizant of the fact that what Elliott wanted as its remedy was an kind of intercreditor obligation; effectively, an injunction that would apply to US agents of Argentina holding Argentine reserves. The court was nervous about granting a remedy of that magnitude. But, for our purposes, what this shows is that Elliott is not going away and is prepared to pursue all kinds of remedial options even after its Euroclear alternative was taken away by Belgian legislation. This takes us back to what has been Elliott’s position all along: "If pari passu doesn't mean ‘equal sharing’ then what does it mean?" That argument is compelling to a court and equally unconvincing is the response "these clauses have been in these bonds for many, many years and whatever it means it doesn't mean that." At the end of the day, it seems hard now to make the argument that the reason for not needing to fix the pari passu clause was that the “risk was really trivial”.

Regardless of which side of the argument one takes, however, the entire episode is a reflection of the costly consequences of the sticky pari passu boilerplate. In this latest litigation, hundreds of thousands of dollars are being spent on legal fees to argue whether Argentina’s passage of the 2005 Padlock law violated the earmarking interpretation of the pari passu clause. The part of this episode that is both comic and tragic is that (we suspect) no one seriously believes the earmarking interpretation. This interpretation was the product of speculation by commentators prior to the Brussels case, but as we have seen it was speculation without any real basis in fact. After the Brussels case, sovereign-side lawyers adopted this view because they needed some interpretation of the clause’s meaning to counter Elliot’s argument. But the historical evidence is at odds with the earmarking meaning of the pari passu clause. Indeed, the evidence flatly contradicts that view. Nevertheless, the historical facts are irrelevant in the current litigation, where the argument is joined over the contours of the earmarking interpretation.

Maybe the frog hadn’t gone deaf?

*****

Our pari passu journey ends here. It was more fun than we could ever have imagined and, if you are a reader who made it to this point, we hope you have enjoyed the trip as well. We realize, however, that we may have opened more questions than we have answered. Two questions in particular continue intrigue us. First, as with the Ecuador episode, the Argentina case shows how pari passu continues to take on opportunistic meaning far removed from any original meaning or any contemporary market consensus as to the rights and obligations embodied in the clause. Nonetheless, the outcome of the litigation might be to attach such a meaning to the clause going forward. In short, we see boilerplate contract language remaining the same, but its meaning evolving to suit ad hoc litigation strategy. In terms of explanations for contract stickiness, then, the persistence of the pari passu clause may be a function of its very vagueness: the clause is a chameleon, able to take on whatever meaning is advanced by the prevailing party in litigation. The problem with this story, however, is that, at least in the Ecuador and Argentine cases described above, the interpretations given to the clause hardly seem optimal for either sovereign debtors or bondholders. And so, these episodes return us yet again to the question of why sophisticated parties would allow contract provisions to remain unchanged that could potentially assume unexpected and costly meanings in future states of the world. We have offered a conjecture that focuses on the structural forces that deter individual lawyers from modifying boilerplate formulations, but opportunities exist for others to develop data that support complementary or competing views.

Second, the current litigation raises the question of whether lawyers will modify the pari passu clauses now that a New York judge has used the clause to give Elliott a victory (although, the extent of that victory will depend on the remedy that is crafted). Recall that a number of respondents justified inaction on the grounds (a) that the prior decision had come from a Brussels court, and (b) that the debate over the meaning of the clause had been put to rest in 2004-5 by the United States government filing its brief and Belgium passing legislation barring attachment of funds.

Ultimately, our story is about the modern world of big law firms and their mass- produced contracts. In a world where contracts have to be produced in three and a half minutes, there is little opportunity to invest in understanding the meaning of standard contract terms, especially those vague terms that are historical boilerplate, let alone to invest time and effort to determine how best to clarify the meaning of those terms. The lawyers producing these contracts seem, for the most part, to follow the herd. There is safety in following what everyone else does, regardless of whether it makes sense. Maybe there just isn’t enough time to figure out whether it makes sense or not.

Appendix

Table A-1

Table A-2

Table A-3

Table A-4

Table A-5

Table A-6

A-7. Issuers, by frequency of appearance (percent of total)

Argentina: 72 (3.9%) Dominican Republic: 8 (0.4%) Moldova: 2 (0.1%) Perth: 1 (<0.1%) Belgium: 69 (3.7%) Slovakia: 7 (0.4%) Milan: 2 (0.1%) Persia: 1 (<0.1%) Finland: 67 (3.6%) Qatar: 7 (0.4%) Luxembourg: 2 (0.1%) Palestine: 1 (<0.1%) Sweden: 62 (3.3%) Egypt: 7 (0.4%) Kenya: 2 (0.1%) Oxaca: 1 (<0.1%) Brazil: 61 (3.3%) Ceylon: 7 (0.4%) Iraq: 2 (0.1%) Ottoman Empire: 1 (<0.1%) Mexico: 56 (3.0%) Barbados: 6 (0.3%) Iran: 2 (0.1%) Oman: 1 (<0.1%) Greece: 56 (3.0%) Ukraine: 5 (0.3%) Honduras: 2 (0.1%) Nuremberg: 1 (<0.1%) Italy: 53 (2.9%) Trinidad & Tobago: 5 (0.3%) Helsinki: 2 (0.1%) Nova Scotia: 1 (<0.1%) Portugal 50 (2.7%) Thailand: 5 (0.3%) Hamburg: 2 (0.1%) Nizam State Railways Co.: 1 (<0.1%) Japan: 49 (2.6%) Sudan: 5 (0.3%) Grenada : 2 (0.1%) Nicaragua: 1 (<0.1%) Philippines: 46 (2.5%) India: 5 (0.3%) Georgia: 2 (0.1%) New Westminster: 1 (<0.1%) China: 45 (2.4%) Bolivia: 5 (0.3%) Fiji: 2 (0.1%) New Brunswick: 1 (<0.1%) Australia: 43 (2.2%) Berlin: 5 (0.3%) Dubai: 2 (0.1%) Netherlands: 1 (<0.1%) Turkey: 36 (1.9%) Belize: 5 (0.3%) Copenhagen: 2 (0.1%) Montevideo: 1 (<0.1%) Norway: 35 (1.9%) Slovenia: 4 (0.2%) Congo: 2 (0.1%) Mongolia: 1 (<0.1%) South Africa: 34 (1.8%) Salvador: 4 (0.2%) Cape of Good Hope: 2 (0.1%) Micronesia: 1 (<0.1%) Russia: 34 (1.8%) Nigeria: 4 (0.2%) Bosnia: 2 (0.1%) Medellin : 1 (<0.1%) Denmark: 34 (1.8%) Morocco: 4 (0.2%) Bahrain: 2 (0.1%) Mauritius: 1 (<0.1%) Colombia: 34 (1.8%) Latvia: 4 (0.2%) Bahamas: 2 (0.1%) Maisonneuve: 1 (<0.1%) Venezuela: 31 (1.7%) Eur. Coal/Steel Comm.: 4 (0.2%) Aruba: 2 (0.1%) Madras: 1 (<0.1%) Uruguay: 31 (1.7%) Dutch East Indies: 4 (0.2%) Abu Dhabi: 2 (0.1%) Macedonia: 1 (<0.1%) Iceland: 30 (1.6%) British Guiana: 4 (0.2%) Yokohama: 1 (<0.1%) Louisiana: 1 (<0.1%) Austria: 30 (1.6%) Tokyo: 3 (0.2%) Wynberg: 1 (<0.1%) Launceston: 1 (<0.1%) New Zealand: 29 (1.6%) Switzerland: 3 (0.2%) Westphalia: 1 (<0.1%) King. Serbs, Croats & Slovenes: 1 (<0.1%) Chile: 29 (1.6%) Seychelles: 3 (0.2%) Western Australia: 1 (<0.1%) Kieff: 1 (<0.1%) Poland: 25 (1.3%) Serbia: 3 (0.2%) Wellington: 1 (<0.1%) Jordan: 1 (<0.1%) Jamaica: 24 (1.3%) San Paulo: 3 (0.2%) Warsaw: 1 (<0.1%) Johannesburg: 1 (<0.1%) Panama: 23 (1.2%) Manitoba: 3 (0.2%) Vietnam: 1 (<0.1%) Haiti: 1 (<0.1%) Ireland: 23 (1.2%) London: 3 (0.2%) Vienna: 1 (<0.1%) Ghana: 1 (<0.1%) United Kingdom: 20 (1.1%) Liberia: 3 (0.2%) Valle del Cauca: 1 (<0.1%) Gabon: 1 (<0.1%) Buenos Aires: 20 (1.1%) Kazakhstan: 3 (0.2%) Trondhjem: 1 (<0.1%) Fed. Malay States: 1 (<0.1%) Malaysia: 19 (1.0%) Hong Kong: 3 (0.2%) Transvaal: 1 (<0.1%) Eur. Econ. Comm.: 1 (<0.1%) Lithuania: 17 (0.9%) Guatemala : 3 (0.2%) Tasmania: 1 (<0.1%) Edmonton: 1 (<0.1%) Croatia: 16 (0.9%) Eur. Inv. Bank: 3 (0.2%) Tanganyika: 1 (<0.1%) Dresden: 1 (<0.1%) Canada: 16 (0.9%) Estonia: 3 (0.2%) Sydney: 1 (<0.1%) Dominica: 1 (<0.1%) Spain: 15 (0.8%) Ecuador: 3 (0.2%) St. Vincent: 1 (<0.1%) Cundinamarca : 1 (<0.1%) Peru: 14 (0.8%) Danzig: 3 (0.2%) St. Petersburg: 1 (<0.1%) Cologne: 1 (<0.1%) Hungary: 14 (0.8%) Cordoba: 3 (0.2%) St. Chris. Nevis/Anguilla: 1 (<0.1%) Christiana: 1 (<0.1%) El Salvador: 14 (0.8%) Bergen: 3 (0.2%) South : 1 (<0.1%) Carlsbad: 1 (<0.1%) Czech: 14 (0.8%) Belarus: 3 (0.2%) Silesia: 1 (<0.1%) Cape Town: 1 (<0.1%) Costa Rica: 14 (0.8%) Victoria: 2 (0.1%) Sierra Leone: 1 (<0.1%) Cape of Good Hope: 1 (<0.1%) Cyprus: 13 (0.7%) Tucuman: 2 (0.1%) Senegal: 1 (<0.1%) Calgary: 1 (<0.1%) Romania: 12 (0.6%) Stockholm: 2 (0.1%) Santos: 1 (<0.1%) Caldas : 1 (<0.1%) Germany: 12 (0.6%) Sri Lanka: 2 (0.1%) Santa Fe (Argentina): 1 (<0.1%) Budapest: 1 (<0.1%) Lebanon: 11 (0.6%) Southern Rhodesia: 2 (0.1%) Salisbury: 1 (<0.1%) Bengal: 1 (<0.1%) Korea: 11 (0.6%) South Carolina: 2 (0.1%) Saarbruecken: 1 (<0.1%) Belgian Congo: 1 (<0.1%) Indonesia: 11 (0.6%) Siam: 2 (0.1%) Rotterdam (Netherlands): 1 (<0.1%) Bavaria: 1 (<0.1%) France: 11 (0.6%) Queensland: 2 (0.1%) Rosario: 1 (<0.1%) Auckland : 1 (<0.1%) Tunisia: 10 (0.5%) Paraguay: 2 (0.1%) Riga: 1 (<0.1%) Antofagasta: 1 (<0.1%) Israel: 9 (0.5%) Pakistan: 2 (0.1%) Rhodesia: 1 (<0.1%) Antioquia : 1 (<0.1%) Cuba: 9 (0.5%) Newfoundland: 2 (0.1%) Prince Albert: 1 (<0.1%) Amsterdam: 1 (<0.1%) Bulgaria: 9 (0.5%) Natal: 2 (0.1%) Pretoria: 1 (<0.1%) Albania: 1 (<0.1%) Oslo: 8 (0.4%) Montenegro: 2 (0.1%) Prague: 1 (<0.1%) Akershus : 1 (<0.1%) New South Wales: 8 (0.4%) Poyais: 1 (<0.1%)

Table A-7

A-8. Issuances By Decade Decade N Percent of Total 1820s 6 0.3% 1830s 1 0.1% 1840s 2 0.1%

1850s 8 0.4% 1860s 23 1.2% 1870s 26 1.4% 1880s 31 1.7% 1890s 35 1.9% 1900s 58 3.1% 1910s 85 4.6% 1920s 144 7.8% 1930s 66 3.6% 1940s 24 1.3% 1950s 44 2.4% 1960s 70 3.8%

1970s 77 4.1% 1980s 121 6.5% 1990s 346 18.6% 2000+ 692 37.2% Table A -8

A-9. Pari Passu usage over time, secured and unsecured bonds Secured Unsecured Pari Passu No Pari Passu Pari Passu No Pari Passu Pre-1860 0 8 0 9 1860-1879 1 22 0 26 1880-1899 3 22 1 40 1900-1919 8 54 4 77 1920-1939 22 78 5 105 1940-1959 0 7 25 36 1960-1979 0 3 123 21 1980-1999 1 2 429 35 2000-present 1 0 684 7

Table A-9

A-10. Post WW-II pari passu mutations (unsecured bonds with pari passu only)

Level of risk of Elliott interpretation (Fig. 1) Exception for "mandatory law" (Fig. 2) Date of issue/currency of payment mutation (Fig. 3) Low Medium High English-law bonds NY-law bonds English-law bonds NY-law bonds 1970s 67 0 0 0/12 0/28 1/12 15/28 1980s 96 0 6 2/39 0/18 16/39 6/18 1990s 211 101 16 18/137 3/114 41/137 6/115 2000-present 325 318 40 118/272 28/351 35/272 15/352

Table A-10