The Control of Wealth in Bankruptcy

I. Introduction

This Article is about the control of many billions of dollars of wealth that are seized in bankruptcy like the spoils of a fallen fortress. It introduces a model that unifies at a theoretical level the twin pillars of financial default: secured-credit and bankruptcy. That model also provides a unifying per¬spective on the academic literature of default. The model is based in substantial part on the insights available from the study of secured-credit and bankruptcy systems around the world.

In the years 2000–2002, more than $730 billion of assets held by large public companies came under bankruptcy administration in the United States

1. THE BANKRUPTCY YEARBOOK AND ALMANAC 2003, at 38 (Christopher M. McHugh et al.

eds., 13th ed. 2003).

2. This Article discusses only business bankruptcy cases involving corporations or other legal entities. Consumer bankruptcy presents a very different, although equally fascinating, set of financial issues.

3. This conclusion is the precise opposite of the theory offered by Professor Adler. Barry E.

Adler, Bankruptcy Primitives, 12 AM. BANK. INST. L. REV. (forthcoming May 2004) [hereinafter

The purpose of the recovery process is to maximize the value of the assets of the defaulting business and to distribute that value to designated beneficiaries. Control of the debtor’s assets in the recovery process following a general default has an important impact on both maximization and distribution. This Article provides a theoretical model whose core is the struggle between bankruptcy law and the law of secured credit for control of the recovery process.4 The model can be introduced by the oversimplified statement that secured-credit law and bankruptcy law represent the struggle between a private and a public ordering of the recovery process. By placing control at the center of the model, the model reveals secured-credit law as the necessary and singular stronghold of the movement for the privatization of the recovery process because a security interest provides the institutional mechanism for control of that process. The model offered here would re¬place the outdated and superficial notion that the struggle in bankruptcy is merely between creditors and owners.

The last decade has seen a furious debate over privatization of the recovery process.5 The debate has centered on a group of academic proposals that may be assembled under the rubric “contractualism.” Their proponents argue that the recovery process should be governed by contracts between the debtor-business and its creditors, with bankruptcy law serving as a default option for those who do not enter into bankruptcy contracts. However, these articles have proposed mechanisms for establishing priorities in distribution without explaining just how or by whom the recovery process would be managed.

Adler, Primitives]

4. The present author made a start on this project three years ago. See Jay Lawrence Westbrook, A Global Solution to Multinational Default, 98 MICH. L. REV. 2276, 2304–07 (2000) [hereinafter Westbrook, Solution]

United Kingdom study. See JULIAN FRANKS & OREN SUSSMAN, THE CYCLE OF CORPORATE DISTRESS, RESCUE, AND DISSOLUTION: A STUDY OF SMALL AND MEDIUM SIZE UK COMPANIES

34–35 (London Bus. School Inst. of Fin. & Accounting, Working Paper No. 306, 2000) (providing a detailed analysis of the rescue process for small to medium size companies in the United Kingdom, and discussing highly collateralized banks’ control over certain aspects of the recovery process), at http://www.facultyresearch.london.edu/docs/306.pdf.

5. See infra subpart V(A).

Recently, some scholars have acknowledged for the first time that a central issue in the debate is control of the debtor’s assets after default.6 Yet these recent articles persist in failing to explain how control of the debtor’s assets would be achieved, before or after default. In particular, none of these analyses identified any connection between privatization of the recovery process and a creditor’s obtaining a security interest. This Article demon¬strates that none of the contractualist proposals can succeed without a security interest. Indeed, any such proposal requires the creditor to obtain a dominant security interest, which is a security interest that covers virtually all the assets of the debtor.7 This Article then shows that widespread adoption of a privatized system depending upon dominant security interests is as undesirable as it is unlikely.

The link between contractualism and a dominant security interest would be revelatory even if it were merely heuristic: contractualism creates the same sorts of difficulties, no matter how the necessary contractual control is exercised. The converse is also true: where we find a dominant security interest today, contractualism already exists

Because the vast legal literature about secured-credit and bankruptcy law has largely ignored the struggle for control, the approach advanced here creates new perspectives on a host of issues that dominate the current debates in that literature. These new perspectives include the role of control in the arguments about the economic efficiency of secured credit,8 the proper

6. See Douglas G. Baird & Robert K. Rasmussen, The End of Bankruptcy, 55 STAN. L. REV. 751, 754–55 (2002) [hereinafter Baird & Rasmussen, End] (discussing the value of keeping different assets together in the same firm during corporate reorganizations)

7. Control of a debtor may, of course, be achieved in other ways, as by getting control of its management. See infra notes 284–85 and accompanying text.

8. See generally Thomas H. Jackson & Anthony T. Kronman, Secured Financing and Priorities Among Creditors, 88 YALE L.J. 1143 (1979)

analytical approach to distinguish “true-sale” securitizations from security interests,9 and the uses and abuses of “bankruptcy-remote vehicles.”10 It is the author’s larger project to apply the model to a number of these areas. However, this Article is introductory and limited to addressing the functional disabilities of the contractualist project.

Part II of this Article provides an overview of the theory and the model. The law governing a debtor’s financial default has two central elements: priority rights and control. While priority determines the order in which value will be distributed to claimants, control concerns the management of the debtor’s assets during the recovery process following default.11 The two laws that together govern the recovery process—Article 9 of the Uniform Commercial Code and Title 11 of the United States Code (Bankruptcy)— incorporate rules about both priority and control, but the conflict between the two bodies of law—and consequently the nexus that unites them at the theoretical level—is control. Part II explains why the elaboration of more and better methods of recovery has made control of the recovery process the principal battleground between private and public governance of that process and, therefore, between their respective strongholds: secured credit and bankruptcy.

Transactions, 41 RUTGERS L. REV. 1067 (1989) [hereinafter Shupack, Solving]

9. See, e.g., Jonathan C. Lipson, Enron, Asset Securitization and Bankruptcy Reform: Dead or Dormant?, 11 J. BANKR. L. & PRAC. 101, 103–09 (2002) (analyzing the treatment of asset securitization as a true sale or, alternatively, as a transfer for security)

Rules for Securitizations, 7 FORDHAM J. CORP. & FIN. L. 301, 302 (2002) (arguing that an approach to securitizations must include consideration of “muddy rules”)

Use and Abuse of Special Purpose Entities in Corporate Structures, 70 U. CIN. L. REV. 1309,

1315–18 (2002) (noting the difference between Enron’s special purpose entity transactions and legitimate securitization transactions).

10. See, e.g., Michael D. Fielding, Preventing Voluntary and Involuntary Bankruptcy Petitions by Limited Liability Companies, 18 BANKR. DEV. J. 51, 66–69 (2001) (discussing the structure and risks of remote vehicles).

11. Control has another function more directly related to priority, which is the structuring of distribution. Given the widespread use of Chapter 11 for liquidation of companies, rather than Chapter 7 with its specific distribution rules, the controller of the recovery process may be able to advantage or disadvantage different groups of beneficiaries by the structuring of the securities, contract rights, or other property received by each. This control of the distribution process through a Chapter 11 plan may permit the controller to have a substantial effect on de facto priorities. See infra text accompanying note 17.

Part III outlines a model of secured credit.12 The model establishes the roles of priority and control in the law of secured credit. Although they are closely related, priority and control have distinct economic and legal functions. Because these distinctions have been blurred in the vast literature about secured credit, this Article must develop some new terminology to describe the functioning of its model. The result is to identify clearly the place of control in the secured-credit model and to permit its juxtaposition with the control function in bankruptcy law. Although this Article is not about the efficiency vel non of secured credit, the model does show the relevance of the control function to the literature on that subject and thus connects the model to the prior debates in the secured-credit field.

Part III distinguishes ordinary secured parties from dominant secured parties, the latter being those with all-encompassing security interests. It ex¬plains why priority is central to ordinary secured parties, while control is merely an additional value, albeit an important one. By contrast, control has a much greater value for a dominant secured party. Part III also explains how the nature and effect of control is quite different, although equally important, prior to default and following default. A further sharp distinction is drawn between control in the context of enforcement of a single debt and control following a general default.

Part III goes on to introduce the interaction of the secured-credit model with the bankruptcy model by identifying the elements of control—including the “bankruptcy veto”—as an additional value for a dominant secured party. Finally, it describes the British model of secured-creditor management of a general default, an approach that has dominated British commercial life for over a century as a standard substitute for bankruptcy management. Although the British model represents complete privatization of the recovery process, giving dominant secured parties both priority and control, it has been ignored in the American theoretical literature.13

Part IV introduces the bankruptcy model. It explains the role of control and priority in bankruptcy and explains why control is the more basic concept. In principle, bankruptcy law can accommodate any form of priority or no priority, but control is its sine qua non. The discussion also questions the ancient chestnut that “bankruptcy is equality,” drawing on comparative law to show that most bankruptcy regimes may be viewed as existing for the

12. See infra subpart III(A).

13. One recent article has discussed British corporate experience as “manager-displacing,” indicating the role of the floating charge and the receivership system as an element of that corporate tradition. See John Armour, Brian R. Cheffins & David A. Skeel, Jr., Corporate Ownership Structure and the Evolution of Bankruptcy Law: Lessons from the United Kingdom, 55 VAND. L. REV. 1699 (2002). The authors do not, however, draw broader conclusions about the relationship between the secured-credit and bankruptcy systems or its implications for bankruptcy theory. A recent working paper that focuses on concentrated ownership compares control of the insolvency process in the United States and United Kingdom systems. It raises some of the same issues raised in this Article. See Hahn, supra note 6.

very purpose of enforcing inequalities in the form of priorities. One of the most prominent of those inequalities, found in most bankruptcy laws around the world, is priority for secured creditors. On the other hand, without control of substantially all of the assets of the debtor, bankruptcy law cannot perform its principal functions in the management of the recovery process. As the British experience shows, the only serious rival to bankruptcy as a manager of that process is the dominant secured party.

Subpart V(A) describes “contractualism” and summarizes the various proposals that fall within that rubric, including “automated bankruptcy,” “complete-system bankruptcy,” and the waiver approach. Subpart V(B) explains why contractualism is a dysfunctional theory unless it is wed to a regime based on a dominant security interest. This is because contractual bankruptcy creates a host of difficulties that cannot be resolved except through linking contractual bankruptcy schemes to a dominant security interest. The difficulties faced by contractualism are largely the same problems that arise from giving a priority to a secured creditor and protecting that priority in a default. Because these difficulties have been largely re¬solved by the secured-credit regime in the United States and some other countries, the creation of a new body of law to solve the same problems would be unnecessary and inefficient. Yet, if contractualism must be linked to secured credit, it must be assessed as no more than an aspect of the case for management of the recovery process by a secured creditor.

Part VI explores two of the major theoretical difficulties with the secured-credit model of default control. The first difficulty is the problem of transactional efficiency: The case for the efficiency of secured credit is unresolved, incomplete, and problematic—especially as concerns a dominant security interest. The second difficulty is the “incentive problem,” a term that describes the secured creditor’s disincentive to maximize value for the benefit of other claimants.14

Part VII describes the evidence available from the debate in the British literature over the effects of secured-creditor management. In Britain, the system based on secured-creditor management has dominated commercial lending for over a century but has been under increasing attack in recent years for reasons similar to the objections to secured-creditor control discussed in Part VI. That critique recently resulted in a remarkable development: the abandonment of secured-creditor management in favor of a system that steps significantly in the direction of the Chapter 11 regime found in the United States. Subject to caveats about the unknown effects of a

14. See Vannessa Finch, Re-Invigorating Corporate Rescue, 2003 J. BUS. L. 527, 536–39 [hereinafter Finch, Re-Invigorating] (discussing the effects of various creditor incentives on creditors’ future behavior). There is a third problem: the policing of state secured-credit law, a job which history has left to bankruptcy law. That point is related in turn to the problem of a “carve¬out” to serve the interests of stakeholders other than secured parties. Carve-out will be the subject of another article. A carve-out has recently been adopted in the United Kingdom. Id. at 553.

major legal reform, Part VII argues that the British experience offers sub¬stantial empirical support for the existence of serious difficulties in secured¬creditor management of a general default. That conclusion is strongly supported by the fact that the private receivership system is also being eroded or abandoned in other Commonwealth countries, although it had been dominant in those countries for most of the twentieth century. Canada is the leading example. Part VII concludes with an outline of the empirical work that is needed to test the analysis offered here.

Part VIII of this Article summarizes the elements of the model and introduces some of its implications. In particular, it does a preliminary analysis of the position presented in the recent articles from Professors Baird and Rasmussen,15 using this model to identify the strengths and weaknesses of their approach.

Because of the serious objections to contractualism and secured-creditor control, Part IX concludes that the system of public, judicial control of the recovery process should continue to be the preferred alternative. It suggests that the considerable movement in the direction of United States-type secured-credit law and United States-type bankruptcy law in other countries, both developed and developing, reflects the strength of the existing American system, although some of the variations in other countries may be worth serious consideration for adoption in the United States. Part IX suggests that further progress in the field will require theories that unify secured-credit and bankruptcy scholarship. It closes with the hope that more scholars will drink at the well of comparative scholarship in the commercial field and elsewhere.

As the foregoing discussion suggests, much of the argument in this Article rests upon lessons drawn from the study of secured-credit and bankruptcy regimes in other countries. Nearly every market economy in the world has such regimes, although they vary enormously. Study of these foreign laws rewards the student richly with an appreciation of the underlying policies that shape these laws, their relationship to national cultures, and the influence now exerted by globalization. While commercial laws in the United States are among the best in the world, there remains a great deal for Americans to learn from others.

II. Overview: Control is the Functioning

Understanding the reconceptualization of the theory of secured-credit and bankruptcy law around the concept of control requires some new perspectives and new terminology concerning phenomena that are familiar and often discussed in both fields. The result is a bit like looking at the

15. See Baird & Rasmussen, End, supra note 6

famous ink blot and seeing faces instead of the vase. It takes some retooling to achieve that result. For this reason, it may be helpful to begin with a brief summary statement of the overall theory.

After a business enters general default,16 the traditional picture of the recovery process comprises three conceptually simple steps: seizure of the debtor’s assets, sale of the assets (generally piecemeal “on the courthouse steps”), and distribution of the proceeds. The process of seizure and sale requires the taking and exercising of control. The distribution of proceeds rests upon a system of priorities: who stands where in the line in front of the distribution table. Thus recovery raises issues of control and issues of priority. With this simple model in mind, we find the vast and sophisticated literature of secured-credit and bankruptcy law has focused almost exclusively on questions of priority, rather than on the problem of control of the process by which seizure, sale, and distribution are achieved. In the context of that model, control of the process of sale has been seen as relatively simple and uncontroversial, while questions of priority in distribution have been at the center of most disputes.

The contemporary process of recovery is likely to be very different. Although the redeployment of assets by piecemeal liquidation in a Chapter 7 bankruptcy proceeding is still frequent, in current practice the recovery process often means sale of a business as a going concern or a financial restructuring of the business (reduction of debt and extension of time for payment) in a reorganization. In the United States, both of these methods of recovery are most often accomplished in a Chapter 11 bankruptcy proceeding.17 The resulting proceeds are often obtained over a period of years and distributed in the form of equity in the debtor, new bonds, and various other securities of considerable complexity depending in turn on the outcome of complex sales and other restructuring transactions. In these circumstances, control of the process of recovery has become at least as important as rules of priority. But, for the most part, the theoretical literature has not reflected these developments and has kept priority at center stage.

The reason that control of the process of recovery has become so important is that the methods mentioned above often require more time and more complex management, both operational and financial, than a simple piecemeal liquidation. At the same time, the range of possible values, from a low value in a simple liquidation to a high value obtained from a creative merger, has become much greater as well. Closely related is the fact that key decisions in this more complex environment turn importantly upon

16. “General default” means the debtor has defaulted on all or most of its obligations, in contrast with failure to pay or perform a particular debt or obligation.

17. They may also be done outside of bankruptcy, but even then the process of achieving an agreement often turns on the outcomes projected in a Chapter 11 proceeding.

evaluation of risk and a willingness to accept risk. As explained below,18 the consequence is that parties with varying priorities will have quite different interests in the management of the recovery process and, therefore, strong incentives to try to control it for their benefit.

Traditionally, the control of the recovery process rested in the trustee in bankruptcy. Outside of bankruptcy, however, a secured creditor had its own unilateral method of control and sale.19 Even when a bankruptcy proceeding began, a straight-forward liquidation might not justify preemption of secured-creditor control by the bankruptcy trustee, so the collateral might be “abandoned” to the secured party.20 However, as reorganization bankruptcy became more prominent in the United States—sooner and more extensively than anywhere else in the world —bankruptcy law changed to impose bankruptcy control on secured creditors as well. The secured creditor retained its priority rights as before, but control of the recovery process extended to the secured creditor’s collateral along with all the rest of the debtor’s assets. Secured creditor autonomy, free of bankruptcy-law control, remains the norm in many countries that still have piecemeal liquidation as the centerpiece of their recovery paradigm. However, the worldwide trend toward reorganization regimes in bankruptcy has lead in many countries, as it did in the United States, to depriving secured creditors of their control of collateral in favor of bankruptcy control.21

Because United States bankruptcy law recognizes and enforces the priority of secured credit with little deviation, there is no conflict and no particular theoretical connection between secured-credit and bankruptcy law as to priority. There is, however, a profound conflict between them as to control of a debtor’s business. In that very conflict lies the fundamental connection between them. The model of the recovery process based on that struggle lies at the heart of the analysis presented here.

18. See infra subpart VI(B).

19. See infra section III(A)(1).

20. 11 U.S.C. § 554 (2000).

21. THE WORLD BANK, PRINCIPLES AND GUIDELINES FOR EFFECTIVE INSOLVENCY AND

CREDITOR RIGHTS SYSTEMS (2001), http://www.worldbank.org/ifa/ipg_eng.pdf [hereinafter WORLD BANK, PRINCIPLES]

AM. L. INST., INTERNATIONAL STATEMENT OF CANADIAN BANKRUPTCY LAW 17 (2003)

[hereinafter ALI, CANADIAN STATEMENT]. Even there, however, the court can issue a stay if it appears that excess value might be lost by secured creditor control. Id.

III. The Secured Credit Model

A. Priority and Control

Little of the literature concerning secured credit under Article 9 of the Uniform Commercial Code has carefully distinguished its two central elements: priority and control. Conversely, many articles and books discuss the doctrine and policy of Article 9 and its relationship to the Bankruptcy Code.22 A great deal has been written about the transactional (ex ante) impact of secured credit on the efficiency of the credit system.23 But the theoretical building blocks of secured credit and the relationship between the goals and methods within its structure have been largely ignored. This Article does not join the debate as to the efficiency of secured credit, and, in any case, a full discussion of that subject would be a large project for an entire article. What follows is an outline of the priority and control elements in secured-credit law sufficient to serve present purposes. The key requirement is to understand control of the recovery process as a valuable element of a security interest distinct from priority in distribution, a point virtually unexamined in the secured-credit literature.

In this model, both priority and control are shown to be essential to secured-credit law outside of bankruptcy. As the later discussion of secured¬credit efficiency theory will demonstrate, however, priority has been the centerpiece of that literature, and control has been largely ignored .24 Only two scholars in that long debate have fully recognized the importance of some of the aspects of control discussed in this subpart. Professor Scott almost twenty years ago presented a model of the relationship between the secured creditor and debtor that emphasized the benefits of one type of control.25 More recently, in a series of articles, Professor Mann has offered powerful empirical and analytical arguments that control is a distinct, and distinctly valuable, attribute of security outside of bankruptcy.26

As we will see, once bankruptcy is filed, it largely negates the control aspect of secured credit, while leaving intact the priority function. That may be one of the reasons scholars have failed to see the importance of secured

22. The classic work remains GRANT GILMORE, SECURITY INTERESTS IN PERSONAL PROPERTY

(1965). A search of the Westlaw “JLR” database uncovers over 1,000 articles discussing Article 9 and bankruptcy from 1990–2003 (“‘Article 9’ w/p/ bankruptcy and da(aft 1989)”).

23. See infra note 8 and supra notes 168–96 and accompanying text.

24. For the article that started it all, see Jackson & Kronman, supra note 8. See infra subpart VI(A). A recent newspaper article reports the dramatic effects of the separation of control and priority in corporate law in Germany, where preference shares with a priority have lost much of their premium value because investors have been forced to realize their vulnerability to those who own the shares that provide control of a corporation. See Floyd Norris, Learning the Hard Way Just How Much Voting Rights Can Be Worth, N.Y. TIMES, Feb. 20, 2004, at C 1.

25. Scott, Relational, supra note 4. It was the type of control called “asset constraint” in this Article. See infra section III(A)(1).

26. See, e.g., Mann, Strategy, supra note 4. His focus was also on the period prior to a general default, although less exclusively. Id. at 213–14.

creditor control, despite the important clue presented by Professor Scott and the evidence provided by Professor Mann. Yet it is the tension over bankruptcy’s effect on the control aspect of security that creates the theoretical link between the two bodies of law at the most fundamental level.27

The first purpose of a secured-credit regime is a workable system to permit a debtor to sell a post-default priority in certain collateral to a creditor, thus achieving what are thought to be socially useful purposes.28 Priority for this purpose simply means the right to be paid first, in full, from the proceeds of disposition of the designated collateral. To achieve that goal, a secured¬credit regime has three requirements. The first is to assure the creditor priority in the proceeds of sale of the collateral in case of a default because certainty in that regard is crucial to generation of the economic benefits. Yet priority in, say, yesterday’s newspapers would not be satisfactory. So the second requirement is to maximize the value of the collateral. No matter how great its value, however, the collateral will not serve the first goal unless it is available following default, so the third requirement is to assure the availability of the collateral at the time of default.

These three requirements must be viewed in the context of three dichotomies of central importance in secured credit. The first divides the period prior to a default from the period following a default. The second di¬vide is between an “ordinary” security interest and a “dominant” security interest. The third dichotomy is between a single default and a general default. These categories are important because the nature and effect of the secured party’s control of the debtor’s assets is different on each side of these paired divides.

1. Pre-Default/Post-Default.—A secured party’s control of a debtor’s assets is crucial both before and after a default, but the character and effect of that control change radically from the pre-default to the post-default situation. Prior to default, the central function of secured credit is to satisfy its third requirement, preserving the availability of the collateral in case of a future default. There is no legal difficulty vis-à-vis the debtor. Article 9 swept away some ancient formalisms, so only a written agreement is required.29 However, promises by the debtor to preserve the collateral are of little value standing alone. There is no point in suing the debtor who ignores the promises and conveys the assets to a buyer for value in good faith or

27. It is instructive that the two lender-control articles recently published do not emphasize the role of security. See infra notes 276–81 and accompanying text.

28. Traditionally, the benefit was thought to be a reduction in borrowing costs and an increase in the availability of credit, although the efficiency of secured credit in delivering these results from the perspective of the credit system as a whole is hotly disputed. See infra subpart VI(A).

29. U.C.C. § 9-203 (1999). The lender must give value, of course, and the debtor must have rights in the collateral in which the interest is granted. See 1 GILMORE, supra note 22, § 10.2, at 297–301.

gives a security interest to a new secured party. The debtor often has no assets left, so a judgment against it is only suitable for framing. To preserve the collateral, there must be a right to recover the property from a transferee. That is, there must be a right traditionally called “in rem,” which means that the right is good as against third parties, even those acting in good faith and giving value.30 It must also extend throughout the relevant market to be good against the bankruptcy trustee.31 That result—ensuring the collateral remains with the debtor or is readily recoverable from transferees—we may call “asset constraint.”

Article 9 has succeeded in creating a workable system of asset constraint by the use of a highly effective system of notice. Giving of notice by one of the methods prescribed under Article 9 is called “perfection,” and it enables the secured party to recover property from a transferee, subject only to certain carefully defined exceptions. This is efficacious because the disclosure of the interest to the world permits an inference of acceptance by subsequent commercial actors and, therefore, justifies subjecting them to the debtor’s sale of priority and enforcement rights to the secured creditor. Article 9 provides a remarkably simple, inexpensive, and readily accessible system of notice by registration in a public office.32 That notice permits the law to provide assurance of asset constraint to the secured party. This system sounds easy but took centuries to achieve.33 Indeed, many countries have yet to achieve it.34 Any secured-credit system is of limited value if it cannot pro¬vide asset constraint

30. Although a right good against all or most third parties is conventionally described as a property right, the point is not undisputed. See Lynn M. LoPucki, The Unsecured Creditor’s Bargain, 80 VA. L. REV. 1887, 1952–54 (1994) [hereinafter LoPucki, Unsecured Bargain].

31. Westbrook, Solution, supra note 4, at 2283–84.

32. It also may have some serious imperfections, but that is a subject for another discussion.

33. The trick, of course, is to bind third parties without destroying free transferability in the marketplace because, without free transferability, inefficiencies would increase costs dramatically. Article 9 has largely achieved that result, but the many important details are not the subject of this Article. For more on this point, see generally 1 GILMORE, supra note 22, at Part IV.

34. WORLD BANK, PRINCIPLES, supra note 21.

35. A system that derives its legitimacy from the autonomous values of private bargaining must create unacceptable externalities if it affects the rights of third parties without their consent. See,

e.g., MICHAEL J. TREBILCOCK, THE LIMITS OF FREEDOM OF CONTRACT 243 (1993). Notice is the

device that attempts to solve this difficulty—or at least to ameliorate it—by a system that permits a claim that the third party with notice has implicitly assented. But see Lucian Arye Bebchuk & Jesse

M. Fried, The Uneasy Case for the Priority of Secured Claims in Bankruptcy, 105 YALE L.J. 857,

882–91 (1996) (questioning whether notice is properly held binding in certain cases)

Bebchuk & Jesse M. Fried, The Uneasy Case for the Priority of Secured Claims in Bankruptcy: Further Thoughts and a Reply to Critics, 82 CORNELL L. REV. 1279, 1286–88 (1997).

Asset constraint is a notably powerful and important function to have gone so long without a name.36 Article 9 provides asset constraint by making unauthorized transfers of collateral recoverable from the transferee, but that simple statement belies the complexity of a system of constraint that has been carefully fine-tuned over a century of experience to produce the right balance between protection of the secured creditor and the maintenance of free transferability in the marketplace.37 It ensures, to a very large extent, that the collateral will still be available if there is a general default.

We will use the term “asset constraint,” rather than “asset control,” when referring to the period prior to default38 because, during that period, the debtor has control of the assets. The security interest merely confers constraint on transfer to third parties. After a debtor’s general default, a secured party acquires the right to direct control of its collateral because it may seize it—by self-help or by judicial action—and proceed to realize upon the collateral with a broad discretion in the method of sale or other realization.39 This valuable right to seize assets of the debtor and control their disposition we will call “collateral control.”40 Outside of bankruptcy, it leaves the secured party in complete control of the recovery process, with only a broad and flexible legal duty to act in a “commercially reasonable” way . 41 This distinction between pre-default and post-default control is significant to understanding the relationship between secured credit and contractualism. Contractualism and secured credit require different kinds of control before and after a general default, while bankruptcy as a system for managing the recovery process exercises control only after default.

In the next section, we will address the fact that both asset constraint and collateral control become qualitatively different in the hands of a dominant secured party. Prior to default, constraint over all assets gives a dominant secured party a pre-default check upon any substantial changes in

36. The closest term in the literature is “hostage value,” used by Professors Scott and Hill. Claire A. Hill, Is Secured Debt Eficient? 80 TEXAS L. REV. 1117, 1134 (2002) [hereinafter Hill, Eficient]

37. Obviously, overbroad restrictions on transfer—or, worse still, undisclosed restrictions— would seriously hamper the functioning of a market. Among other things, overbroad restrictions would cause potential buyers and other transferees to refuse or undervalue trades because of the risk of losing the asset to a secured party.

38. Professors Scott and Mann have used the term “control” when referring to the legal protection granted to a secured party prior to default. Scott, Relational, supra note 4, at 927–28

39. U.C.C. art. 9, pt. 6.

40. See infra notes 58–65 and accompanying text.

41. U.C.C. §§ 9-610(b), 9-627.

the business activities of its debtor.42 The post-default collateral control given to a dominant secured party gives the secured creditor control of an entire enterprise and makes it possible for the creditor to realize going¬concern value.43

2. An Ordinary Secured Party versus a Dominant Secured Party: Enterprise Control and Going-Concern Value.—The second dichotomy in secured finance is between an “ordinary” secured party and a “dominant” secured party. In a basic and simple model of a secured transaction, the ordinary secured party is one that receives a security interest in a single item of collateral, while a dominant secured party receives a security interest in all of the assets of its debtor.44 Their priority rights are the same, but their rights to constrain and control produce importantly different effects .45

Suppose a debtor, General Kompute,46 that might give either a security interest in a large welding machine to an ordinary secured party or a security interest in all of its assets to a dominant secured party. In case of default, the ordinary secured party would have the right to take control of the single

42. That extended form of asset constraint we may call “business plan constraint.” See infra section III(A)(3). It is the sort of control or quasi-control that Professor Scott had in mind in an important article about the possible efficiencies of secured credit, although he did not establish the distinction as such. Scott, Relational, supra note 4, at 927–28.

43. We will call the dominant secured party’s all-embracing collateral control “enterprise control” because it gives effective control of the business. See infra sections III(A)(2), III(B).

44. There are many intermediate positions where the secured party has a substantial—but less than complete—security interest, but that is another issue that must await a fuller development of the secured-credit model. It is obvious, however, that a security interest in the most valuable and operationally crucial of the debtor’s assets might well amount to a dominant security interest. An example would be a typical single-asset real estate case, where the holder of a mortgage on an apartment complex may not have a security interest in the office furniture, the bank account, maintenance equipment, and so forth, but its collateral nonetheless represents so much of the total wealth of the debtor that the effect is much the same as if it did. An example that may be especially relevant to the recent “control” articles would be control of the debtor’s bank account by way of setoff and cash collateral orders.

45. Professor Scott is one of those who saw the importance of distinguishing among security interests with varying scope. See Scott, Relational, supra note 4, at 904–33 (examining the reasons that security interests differ in structure and scope). But he did not seize upon the heuristic value, as he might have put it, of the dichotomy between the secured party with an interest in a single asset and one with a blanket lien, nor did he focus on the enforcement rights of secured parties. (His work may be subject to critical comment in this Article to a greater extent than the work of others, I fear, precisely because he has published so many useful insights that overlap with the analysis offered here.) In any event, his focus was on the quasi-control right he called “hostage value,” which this Article would characterize as an element of leverage for the secured party that arises from “asset constraint” or, in the case of a dominant secured party, “business-plan constraint.” Asset constraint is the overall pre-default effect of the Article 9 rules, while hostage value is one of the benefits it generates for a secured party. See supra note 36. Professor Hill has developed the hostage-value idea in more sophisticated and useful detail. See Hill, Eficient, supra note 36, at 1133–35

46. I trust there is no such company, but, if there is one, it is not the company I have in mind.

welding machine .47 The dominant secured party would have the right to assume control of the entire enterprise because of its right to seize and sell all of the enterprise’s assets. While the dominant secured party might not have technical control of the debtor company as such,48 it would effectively control the entire business from the office lease to the drill presses to the accounts receivable and inventory. One highly important consequence would be that the secured creditor could sell the business as a “going concern.”49

When the ordinary secured party sold its welding machine, it would probably get only “liquidation value,” the value typically obtained at a liquidation sale. It is established that liquidation value is routinely well below market value.50 Because of the sophisticated procedures available to a secured creditor under Article 9,51 if the ordinary secured party tried hard (for example, by advertising widely and waiting for the right buyer), it might ob¬tain a price closer to market value for that single asset, but no more. By contrast, the dominant secured party’s control of all of the assets of the enterprise would mean that it would have the opportunity to obtain “going¬concern” value at an Article 9 sale. Going-concern value may be much greater than market value (and therefore much, much greater than liquidation value) because a living business—with established customers, knowledgeable employees, and so forth—may well bring a higher price as a unit than would the sale of each asset separately, even in the unlikely event that those separate sales would obtain market value for each asset. It is generally conceded that going-concern value is the highest possible value for assets and much to be desired.52 The highly flexible Article 9 sale provisions

47. The classic ordinary secured party is either a vendor with a purchase money security interest in an item it has sold to the debtor, or a lender who has financed such a sale. Such a creditor has no interest in control of the debtor enterprise. Indeed, if there is a default, its interest will lie in quickly separating its collateral from the rest and realizing as much as possible from the sale of its collateral. Only in unusual circumstances would it have an interest in the overall fate of the debtor enterprise.

48. The right to sell assets, and conserve them in the meantime, is not the same as ownership of a company, but it does confer considerable control, absent bankruptcy.

49. Referring to the British receivership system, Rizwaan Mokal cites as key the ability of a dominant secured party (a “main creditor”) to keep the assets of the debtor together, contrasted with the inability of an ordinary secured party (a “fixed” chargeholder) to do so. The reason is that this power protects “synergies,” by which he means something like what this Article calls going-concern

value. See Rizwaan J. Mokal, The Floating Charge—An Elegy, in COMMERCIAL LAW AND COMMERCIAL PRACTICE 479, 485–93 (Sarah Worthington ed., 2003) [hereinafter Mokal, The Floating Charge].

50. See, e.g., United States v. Whiting Pools, Inc., 462 U.S. 198, 203 (1983) (“Congress presumed that the assets of the debtor would be more valuable if used in a rehabilitated business than if ‘sold for scrap.'”).

51. U.C.C. art. 9, pt. 6.

52. James S. Rogers, The Impairment of Secured Creditors’ Rights in Reorganization: A Study of the Relationship Between the Fifth Amendment and the Bankruptcy Clause, 96 HARV. L. REV.

973, 975 (1983) (stating that a central assumption of the reorganization sections of the Bankruptcy Code is that a business will be worth more as going concern than the liquidation value of its assets)

make it legally possible for the dominant secured party to make a going¬concern sale.

As a consequence, if the ordinary secured party sells the welding machine at auction after giving such notice as it thinks satisfies the Article 9 requirement,53 it may get, say, $10,000 for it, versus a market price for such a used machine of, say, $18,000.54 If it is willing to advertise extensively, make phone calls to other companies who use such machines, and wait until the best offer is obtained, it might get the full market value of $18,000. On the other hand, the dominant secured party may be able to sell General Kompute (that is, all of its assets as a package) to Cross-Town Kompute, which has been anxious for a second location for years and would be de¬lighted to retain the employees who know the eccentricities of each machine and greet customers by name. The dominant secured party may obtain a full going-concern value for the business, much more than the sum of the amounts it could get by selling each asset.55 Thus the spread of values for the assets might be, say, $100,000 total from a liquidation sale of each asset, $180,000 market value after much time and effort in selling each asset, or $300,000 for sale as a going concern. There is no guarantee that a going¬concern sale can be made, but it is possible in a substantial number of cases, and increased payoffs like those in this example would not be surprising. That going-concern sales by secured parties are quite plausible is empirically demonstrated by the English experience described below.

Lucian Arye Bebchuk, A New Approach to Corporate Reorganizations, 101 HARV. L. REV. 775,

776 (1988) [hereinafter Bebchuk, Approach] (noting that “liquidation might well leave the participants with less than the going-concern value of the company’s assets”)

Lynn M. LoPucki, The Nature of the Bankrupt Firm: A Response to Baird and Rasmussen’s The End of Bankruptcy, 56 STAN. L. REV. 645, 653 (2003) [hereinafter LoPucki, Nature]

916 (1993). Actually, there may occasionally be higher values. For example, there may be an additional value for a buyer-competitor in being able to eliminate a competitor by an asset purchase without the risk of successor liability and some of the other risks of a corporate acquisition. In a recent article, Professors Baird and Rasmussen argue that few American companies have any going¬concern value by the time of bankruptcy, but they do not deny that any such value is likely to be the best obtainable. See Baird & Rasmussen, End, supra note 6, at 758. They offer no empirical support for their claim about the availability of going-concern value in fact.

53. U.C.C. §§ 9-611 to 9-614.

54. Unfortunately, we have little empirical data on these points and must rely on the experience of people in the field.

55. The statement in the text must be qualified because in some cases (perhaps in many cases) the managers or owners of a business may be crucial to obtaining going-concern value for the business. For example, a manager may be careless about financial discipline but may have a level of credibility and technical respect among customers that makes him or her essential to the business. Because personal services may not be legally commandeered in our society, even a dominant secured party cannot realize going-concern value in such a business without obtaining the cooperation of key personnel. See, e.g., Douglas G. Baird & Robert K. Rasmussen, Chapter 11 at Twilight, 56 STAN. L. REV. 673, 687–89 (2003) [hereinafter Baird & Rasmussen, Twilight].

Because the distinction between ordinary secured parties and dominant secured parties has been so generally ignored in the literature,56 we need a term to distinguish the effects of the two types of collateral control. The post-default collateral control given to a dominant secured party we may call “enterprise control,” because it gives the secured creditor control of an entire enterprise and makes it possible for the creditor to realize going-concern value.57 The distinction between the two types of secured parties is crucial because there is such a sharp difference between the nature and value of an ordinary secured party’s mere collateral control and the enterprise control exercised by a dominant secured party.

3. Single versus General Default.—Following default, the law’s central goal is maximizing collateral value. There is a key distinction here between a single default—only the secured party’s debt is not being paid—and a general default in which the debtor is not paying its debts generally. In the first situation, the secured-credit regime is simply a highly privileged mechanism for enforcing a particular debt, much superior to the slow and uncertain business of getting a judgment and sending out the sheriff to seize such property as can be discovered. Most systems around the world give a secured creditor some advantage in debt enforcement, but the United States’ system is among the most helpful to secured creditors.58 Part 6 of Article 9 permits the secured creditor to use self-help, without the need to go to court. The secured party may seize the collateral and sell it under a very flexible regime of reasonableness, paying itself from the proceeds.59 With or without self-help, the special bundle of enforcement rights given to a secured creditor in most legal systems is very valuable, although its independent value has

56. A recent well-documented article discusses the allocation of going-concern value between secured and unsecured creditors and makes a number of useful points analytically but puts no importance on the distinction between ordinary and dominant secured parties. See Omer Tene,

Revisiting the Creditors’ Bargain: The Entitlement to the Going-Concern Surplus in Corporate Bankruptcy Reorganizations, 19 BANKR. DEV. J. 287 (2003). Professor Shupack identified the importance of the emergence of blanket liens under Article 9 but did not develop the point. Shupack, supra note 8, at 1084. Professor Scott was nearly unique in emphasizing the importance of the “blanket lien” but focused almost exclusively on pre-default control of the debtor’s business. See supra note 45. Professor Hill has usefully discussed the emerging notion in financial circles, especially in the United Kingdom and the Commonwealth, of “whole-business securitization,” but that approach depends on the British receivership system—that is, control by the dominant secured party. Id.

57. Enterprise control is, of course, the cognate of the pre-default business constraint. See supra notes 42—43 and accompanying text.

58. The receivership system is even more generous, but the North American system, found in the United States under Article 9 and in Canada under the Personal Property Securities Acts, is the next best for the secured creditor. For an expert discussion of the latter, see generally JACOB S.

ZIEGEL & DAVID L. DENOMME, THE ONTARIO PERSONAL PROPERTY SECURITY ACT,

COMMENTARY AND ANALYSIS (2d. ed. 2000).

59. U.C.C. § 9-612. For certain collateral, like accounts receivable, the secured party may simply collect the accounts from the debtor’s customers without any court intervention. U.C.C. § 9-607.

been largely ignored in the literature. In jurisdictions like the United States that permit self-help repossession and sale, collateral control is even more valuable.60

The shift from the enforcement of a single debt to the general-default context may radically change the rights of the secured creditor because bankruptcy negates collateral control.61 As explained below in Part IV, control of the debtor’s business lies at the heart of bankruptcy law and is its sine qua non. Therefore, in principle, bankruptcy must negate a secured party’s collateral control.62 The distinction between priority rights and collateral control is crucial for the secured creditor in a debtor’s general default because bankruptcy law honors one and negates the other.

In a general default, a security interest is simply a privately bargained priority in bankruptcy that was sold by a debtor to a creditor. The sale was part of a larger transaction in which the priority was presumably exchanged for a lower interest rate and better terms in connection with an extension of credit to the debtor. It is the only privately bargained priority permitted under the United States Bankruptcy Code.63 Although our priority for a security interest is more absolute than that found in many other countries, the security-interest priority is recognized to varying degrees in the general¬default laws of many jurisdictions.64 Most of the issues in secured-credit law in the United States, including perfection under Article 9 and adequate pro¬tection under the Bankruptcy Code, relate primarily to the institutional apparatus necessary to protect the priority of the holder of a security interest in the circumstance of a general default. But its collateral control is generally lost.65

60. The text refers to the value of control independent of priority alone.

61. 11 U.S.C. § 362(a).

62. This assertion is correct despite the fact that, in a number of countries around the world, secured creditors are not restrained by the bankruptcy stay. See infra note 110.

63. That statement should perhaps be qualified, if one concedes that subordination is a sort of reverse priority. Section 510(a) makes contractual subordination clauses enforceable in bankruptcy, so they may be said to be a form of privately bargained priority. 11 U.S.C. § 510(a). On the other hand, such clauses do not constitute a general subordination of a claim but only subordination to certain other creditors, so they do not constitute a priority in the full sense of the word. Setoff also produces a priority in bankruptcy

64. WORLD BANK, PRINCIPLES, supra note 21, at P3 and accompanying text

ROLF SERICK ET AL., SECURITIES IN MOVABLES IN GERMAN LAW 108 (1990) (primarily by a

retention of title theory for both suppliers and lenders)

(discussing the ranking of “singularly preferred creditors” ahead of a secured creditor).

65. In a routine liquidation, the secured creditor may regain its collateral control. If the trustee in bankruptcy is selling off assets piecemeal, it will often be efficient to release bankruptcy control of collateral back to the secured party to permit that party to realize upon its sale. 11 U.S.C. § 554. On the other hand, where rescue or sale of the business as a going concern is a possibility (for

Thus the distinction between a single default and a general default is important because of the difference in the secured party’s collateral-control rights. In enforcing a single debt outside of bankruptcy, the secured party has complete control of the recovery process as to the assets constituting its collateral. If it is a dominant secured party, it has control as to the entire enterprise. However, in a general default the debtor and other creditors hold an option to remove the secured party’s control by filing a bankruptcy proceeding, an option often exercised in the United States. Once bankruptcy is filed, the secured party loses its collateral control.

In summary, priority represents the right to be first in line for payment from the disposition of collateral after default. Control is the right to control the collateral in some sense both before and after default. The importance of control is that it protects priority—that is, it protects the distribution scheme.

Asset constraint is the secured party’s power to recover its collateral from a transferee. In the hands of a dominant secured party, asset constraint amounts to “business-plan constraint,” a negative power to prevent unapproved major changes in the debtor’s business. More direct control of the collateral arises after default. Collateral control is the right outside of bankruptcy to seize the collateral and control the recovery process, including the sale or other disposition of the collateral. For a dominant secured party, collateral control becomes enterprise control because the dominant secured party can sell the entire business and obtain going-concern value. For both types of secured parties, however, a general default may sharply reduce collateral control because another party may exercise the bankruptcy option, which negates collateral control while preserving priority.

Asset constraint prior to bankruptcy is a valuable right that has received little attention in the economic analysis of secured credit. Collateral control is a distinct, additional value that has also been largely ignored. Because bankruptcy operates only after a general default, it is in that context that control becomes the center of conflict between secured-credit and bankruptcy law and thus constitutes the basic nexus between them at a theoretical, as well as an operational, level.

B. Benefits of Enterprise Control to a Dominant Secured Party

The two most important benefits a dominant secured party gains from enterprise control are the opportunity to realize going-concern value after default and a possible veto over a bankruptcy proceeding.

example, in a Chapter 11 reorganization), the debtor in possession or trustee in bankruptcy is unlikely to relinquish control. Whether the court will return control to the secured party under § 362 depends largely on the court’s view of the prospects for reorganization and the required protection of the secured party’s priority. See infra subpart VI(B).

A dominant security interest offers debtor and creditor a substantial potential payoff, absent bankruptcy. The ordinary secured party must lend based on its estimate of the future value of its one item of collateral.66 If the item’s estimated depreciating value at sale will at all times be in excess of the declining balance on the loan, the creditor’s analysis is complete.67 It will determine the amount of its loan and price it as a function of being fully secured up to the sale value.68 By contrast, the dominant secured party is in a position to lend on the basis of being secured by the going-concern value of the debtor’s enterprise, over which it will take full control in case of default.69 That higher value may permit the lender to lend more money at a lower secured rate of interest. That is, it may lead to an increase in capital availability and a reduction in costs. That result is arguably desirable for both debtor and creditor as well as for society generally, other things being equal. Given the large gap between going-concern value and other values, the enterprise control given to the dominant secured party has the potential to be substantially more valuable to the parties than the mere collateral control granted to the ordinary secured party. The institutional implications of enterprise control are also important because the realization of going-concern value is a principal justification for Chapter 11.70

66. The text is not inconsistent with the elementary proposition that all lenders are interested in the debtor’s capacity to pay the loan without any need for enforcement. See Paul M. Shupack,

Preferred Capital Structures and the Question of Filing, 79 MINN. L. REV. 787, 793 (1995)

[hereinafter Shupack, Filing]. The text refers to the secured party’s analysis of its lending position qua secured party, that is, with sole reference to the value of its collateral as part of its lending package.

67. Note that this estimate would be higher if there were no prospect of bankruptcy because the creditor could assume that, in case of default, it would be able to enjoy its collateral-control right to manage the sale of the collateral. In reality, the possibility of bankruptcy means the estimate of value by the ordinary secured party will depend in material part on the likelihood of bankruptcy and the consequent loss of its collateral-control right.

68. That is, it will apply its secured-loan pricing and risk standards to the loan to the extent of the value of the collateral. If it is willing to lend more, it will apply its unsecured price and risk standards to the remainder above the projected collateral value ignoring for the moment the question of collateral-control rights.

69. The description in the text is shorthand. All lending is done—at least in theory—on the basis of discounted probabilities. The actual value of a dominant secured party’s security will be the going-concern value discounted by the improbability of obtaining the going-concern value. More fully, where the going-concern value is X, the piecemeal-sale market value of the assets is Y, and the liquidation value of the assets is Z, there will be a composite value equal to the discounted probability of obtaining X plus the discounted probability of obtaining Y plus the discounted probability of obtaining Z. Assuming a fairly good probability of obtaining going-concern value, the composite value will be higher than either Y or Z, permitting the lender to lend more or to lend at a lower interest rate.

70. H.R. REP. NO. 95-595, at 220 (1978), reprinted in 1978 U.S.C.C.A.N. 5963, 6179 (“The premise of a business reorganization is that assets that are used for production in the industry for which they were designed are more valuable than those same assets sold for scrap.”)

As we have seen, however, bankruptcy negates collateral control. To that extent, the value of enterprise control must be discounted by the risk of a bankruptcy filing following default.71 Yet that risk may be reduced, and the value of enterprise control enhanced, by a bankruptcy veto. A possible veto over the maintenance of a bankruptcy proceeding is a second added value arising from enterprise control. The bankruptcy veto arises from the facts that enterprise control may prevent the funding of a bankruptcy proceeding and may also make the likely results of bankruptcy unattractive to the debtor and other potential filers of a bankruptcy petition.72 The ability to sidestep bankruptcy control and enjoy the benefits of enterprise control over the recovery process may represent a substantial additional value to the secured creditor, materially adding to the benefit associated with potential sales at going-concern value.73 Neoclassical theory would suggest that this increase

sufficient resources to acquire it”)

71. Of course, the risk of bankruptcy also reduces the value of collateral control for an ordinary secured party. See supra note 67.

72. In a fair number of reported cases, a secured party has agreed to a “carve-out” of value from its collateral. See, e.g., In re Debbie Reynolds Hotel & Casino, Inc., 255 F.3d 1061 (9th Cir. 2001)

see also LAWRENCE P. KING, 4 COLLIER ON BANKRUPTCY ¶ 506.05[6], at 506-134 (15th ed. rev. 2000) [hereinafter COLLIER ON BANKRUPTCY]. Because § 9-315 of the U.C.C. and § 522(b) of the

Bankruptcy Code give the secured party a broad right to the proceeds of its collateral, a dominant secured party will ordinarily have a security interest in all the cash flow of the business, leaving nothing to pay attorney and trustee fees and other expenses of the bankruptcy unless the secured party agrees to the use of its proceeds for that purpose. See, e.g., In re Hotel Syracuse, Inc., 275 B.R. 679 (Bankr. N.D.N.Y. 2002)

(1994) (discussing cases and suggesting reforms). Nickles and Adams criticize the cases for having been unwilling to interpret this provision more broadly in favor of paying bankruptcy costs. Id. at 1174. The arguments that Nickles and Adams, two knowledgeable scholars, make in favor of requiring payment of bankruptcy costs from collateral are closely related to those that can be made for a bankruptcy trump of the bankruptcy veto.

73. Of course, the bankruptcy veto would have substantial value even where the debtor does not have a going-concern value because elimination of bankruptcy control would leave the dominant secured party in complete control of the recovery process.

The benefits of control are not limited to the bankruptcy situation. See, for example, Gau Shan Co. v. Bankers Trust Co., 956 F.2d 1349 (6th Cir. 1992), in which the debtor tried to block a suit by the

in value would be shared between debtor and lender in the same way as other enhancements obtained by the secured party by virtue of a dominant security interest.74 The benefit comes with certain difficulties discussed below.75

C. British System

Partly for historical reasons, this way of looking at the position of a dominant secured party has been obscured in the United States.76 By contrast, it is the classic and typical picture of secured lending in Britain. Britain provides an empirical example of the consequences of a system in which a dominant security interest is typical and a dominant secured party retains control even in a general default.77 Since the late nineteenth century, Britain has had a security system known as the “floating charge.” The word “charge” means security interest. The scope of the floating charge is roughly equivalent to a “blanket lien” in American terminology

creditor in Hong Kong. Its key argument was that the creditor had an English-style floating charge and therefore could block the debtor from access to the courts by exercising its rights, making it impossible to respond properly to the Hong Kong suit. Id. at 1352. A similar result might be obtained in the United States if a dominant secured party argues that its all-embracing security interest includes any cause of action the debtor might have against the secured party.

74. See infra subpart VI(A).

75. See infra subpart VI(B).

76. A striking characteristic of Article 9 is that it makes “blanket” liens readily obtainable and enforceable without establishing any different rules for the holders of such liens. The literature neither suggests such differentiation nor explains why it is unnecessary. Few things are more obvious than the dramatic difference in the positions of ordinary and dominant secured parties, but the difference is ignored like the purloined letter. EDGAR ALLAN POE, The Purloined Letter, in THE PURLOINED POE, at 6 (John P. Muller & William J. Richardson eds., Johns Hopkins Univ. Press 1988) (1844). The reason for this oversight is the failure to take account of the centrality of the concept of control. Also contributing may have been the existence of many middle cases, secured creditors with multiple security interests but not blanket liens—a point not addressed in this Article.

77. My British friends will find the following description painfully simplistic to the point of error, but I eschew all nuance in an effort to be clear to an American audience. For far fuller and

better discussions of the floating charge, see, for example, INSOLVENCY LAW AND PRACTICE, REPORT OF THE REVIEW COMMITTEE (1982) Cmnd. 8558, at 31–34, 344–58 [hereinafter Cork Report]

INSOLVENCY SYSTEMS 1, 2–3 (1999), at http://www4.worldbank.org/legal/insolvency_ini/WG10

paper.htm

Empirical Evidence from England, in CURRENT DEVELOPMENTS IN INTERNATIONAL AND COMPARATIVE CORPORATE INSOLVENCY LAW 191 (Jacob Ziegel ed. 1994).

intervention or supervision by a court. A legal fiction makes the receiver, who is now called an “administrative receiver,”78 the agent of the company, not of the secured party, so the secured creditor is not liable for the receiver’s acts. A key element in the structure is that the receiver’s primary responsibility is to the secured party

The creditor-appointed receiver, who is invariably an accountant, then proceeds to liquidate the debtor.80 In theory, the receiver should make every effort to sell the business as a going concern and thus to realize going¬concern value. If no buyer is found for the business as a whole, the receiver proceeds to an asset liquidation with broad powers to choose the methods of sale. As a matter of course, there is no court involvement—or supervision— of any kind. Often no bankruptcy proceeding81 is filed at all. If there is a need to resolve issues beyond the sale of the assets, then a voluntary or involuntary liquidation might follow the receivership.82 An example would be the happy, if perhaps rare, instance in which there is a surplus after the secured party is paid in full from the sale of the assets and there is need for a bankruptcy proceeding to distribute the remaining proceeds according to the statutory scheme.83

The creditor who holds the floating charge in this system almost always has a “fixed” charge as well, often over most, if not all, of the debtor’s assets. It is the fixed charge that provides priority in distribution because the floating charge is subordinate to a number of “preferences” (priorities), including tax claims.84 Thus the British bankers’ saying that introduces this Article: the

78. Perhaps to keep foreigners confused, the private receiver, appointed out of court by the secured party, is called an “administrative receiver,” while the court-appointed manager of an attempted reorganization is called an “administrator.” (Now if we can only learn how to pronounce “Leiscester.”)

79. See Rajak, supra note 77, at 211.

80. Cork Report, supra note 77, at 104.

81. I will use the term “bankruptcy” per American usage, although in Britain (and most of the English-speaking world) the term for a corporate bankruptcy is an “insolvency proceeding.” The word “bankruptcy” is used in Britain and elsewhere to mean the bankruptcy of a natural person.

AM. L. INST., TRANSNATIONAL INSOLVENCY PROJECT, PRINCIPLES OF COOPERATION IN TRANSNATIONAL INSOLVENCY CASES AMONG MEMBERS OF THE NORTH AMERICAN FREE TRADE

AGREEMENT 1 n.2 (2003). Although the analysis in this Article is limited to business debtors that are legal entities, I will use “bankruptcy” in the service of clarity for an American audience.

82. See Gabriel Moss, Comparative Bankruptcy Cultures: Rescue or Liquidation? Comparison of Trends in National Law—Britain, 23 BROOK. J. INT’L L. 115, 121 (1997) (noting that “[i]n England, most rescues of businesses have involved the sale of the business to other entities, while the insolvent rump of the corporation has, generally speaking, gone into liquidation”).

83. As the quotation that leads this Article suggests, the lender often takes “fixed charges” on specific assets. A fixed charge is similar to a security interest of the sort found in North America. It has several advantages, including the fact that the charge is not subordinated to certain other priority creditors, as a floating charge may be. See GOODE, SECURITY, supra note 77, at 52.

84. See id.

fixed charge for priority, the floating charge for control.85 For our purposes, the most important point is the separation of the two functions of a security interest, one that is rarely so obvious in American practice and therefore often ignored.86

This sort of lending arrangement is typical in Britain, while the “blanket lien” is only one of a number of lending patterns in the United States and not necessarily the most frequent.87 In Britain, banks, especially the four dominant “clearing” banks, employ the floating charge routinely,88 although apparently it is not so often found where the debtor is a major company.89 Thus Britain presents a “natural experiment” for observation of the effects of a lending system that depends primarily on dominant security interests.

English accountants report that they are fairly frequently successful in selling businesses intact and thus obtaining going-concern values.90 They do not quantify these reports, and one may wonder about how often going¬concern values are actually achieved. Nonetheless, there is no reason to doubt the honesty of their reports that such sales are made. The English ex¬perience certainly demonstrates that such sales can be achieved by dominant secured parties exercising a fully realized enterprise control. Although the receiver system no doubt facilitates this process, there is no reason in principle that dominant secured parties cannot achieve the same results under the very flexible provisions of Article 9 and no reason to doubt that they have done so.91

IV. The Bankruptcy Model

Much less development is needed to describe a basic model of bankruptcy for the purpose of establishing the centrality of control and its relationship to priority because there is a considerable consensus on some

85. The explanation for the combination of fixed and floating charges is complex, and its detail

would take us too far out of our way. See generally GOODE, SECURITY, supra note 77

86. It must be said, however, that the distinction has often been ignored in the British literature as well. See, e.g., Mokal, The Floating Charge, supra note 49.

87. See Shupack, Filing, supra note 66, at 792–807 (providing some empirical data on the uses of different types of capital structures). There is little empirical data to help us on this point, but the statement in the text would command considerable agreement in the field.

88. Cork Report, supra note 77, at 32

RESCUE AND BUSINESS RECONSTRUCTION MECHANISM 9 (2000), available at

http://www.insolvency.gov.uk/introduction/condoc.condoc/review.pdf.

89. See Cork Report, supra note 77, at 32 (examining the disadvantages of floating charges, especially their capability for working great injustice in the context of major enterprises).

90. A recent study confirms that conclusion. See FRANKS & SUSSMAN, supra note 4, at 15–16.

91. Cf. Bezanson v. Fleet Bank-NH, 29 F.3d 16, 18–20 (1st Cir. 1994) (involving an ultimately unsuccessful deal wherein a secured creditor refused to sell a defunct company for a price yielding excess value in order to pursue an opportunity to invest in the company and attempt a joint venture, which could have resulted in potentially higher returns to the secured party). There may be reason to wonder, however, if secured creditors have often received going-concern value at the point of the “official” Article 9 sale. Id. See also infra subpart VI(B).

central points. It has long been understood that bankruptcy is a collective proceeding to be used primarily, if not exclusively,92 when a debtor has entered general default and that bankruptcy’s central purpose is the maximization of the value to be distributed to its designated beneficiaries. It can be argued that other purposes are also important, or even central, to the bankruptcy function,93 but the maximization of distributions to beneficiaries is a consensus goal.94 On the other hand, we find a considerable debate about the choice of beneficiaries—especially in the business context, which pro¬vides the only sorts of cases considered in this Article. Some argue that only creditors should be regarded as beneficiaries of the process,95 while others would find a place for employees, entrepreneurs, and even communities.96

92. A bankruptcy filing by a debtor not in general default is likely to be dismissed. The courts have made it clear that bankruptcy is not ordinarily to be used in single-creditor disputes as opposed to in general defaults. See, e.g., In re Better Care, Ltd., 97 B.R. 405, 419 (Bankr. N.D. Ill. 1989) (dismissing a creditor’s involuntary petition against a debtor because “[n]ot a scintilla of evidence was offered as to a general default by the alleged debtor in the payment of its obligations”). Bankruptcy is a blunt instrument for debt enforcement, and legal systems should not rely on it for

that purpose. See WORLD BANK, PRINCIPLES, supra note 21, § 2.1, para. 45 (stating that a seizure

and sale of assets is a more efficient method of recovering debt for unsecured creditors than is an involuntary insolvency proceeding). Furthermore, a filing by a debtor who is not in general default is unlikely because it will often lack the incentive to risk control of its business to resolve a single¬creditor dispute.

93. Jay L. Westbrook, The Globalisation of Bankruptcy Reform, 1999 N.Z. L. REV. 401, 404– 10 [hereinafter Westbrook, Globalisation] (discussing six purposes for bankruptcy, beginning with social control).

94. See, e.g., THOMAS H. JACKSON, THE LOGIC AND LIMITS OF BANKRUPTCY LAW 7–19, 22–

27 (1986) [hereinafter JACKSON, LOGIC] (arguing that bankruptcy law and nonbankruptcy law should work as complementary systems for the maximization of the value of a certain pool of assets and the distribution of those assets, respectively)

Financial and Political Theories of American Corporate Bankruptcy, 45 STAN. L. REV. 311, 319–

23 (1993) [hereinafter Adler, Financial] (discussing market-based pre- and post-insolvency proposals as potentially improved methods of maximizing value for creditors)

Loss Distribution, Forum Shopping, and Bankruptcy: A Reply to Warren, 54 U. CHI. L. REV. 815,

822 –24 (1987) [hereinafter Baird, Loss Distribution] (discussing the need for parity between bankruptcy and nonbankruptcy priority rules)

95. See JACKSON, LOGIC, supra note 94, at 24–27 (arguing that bankruptcy law cannot simultaneously solve the common pool problem and protect non-creditors, such as workers and society).

96. The long-ago debate between Professors Warren and Baird remains fresh in this regard. See Elizabeth Warren, Bankruptcy Policy, 54 U. CHI. L. REV. 775, 777–78 (1987) (arguing that bankruptcy is an attempt to deal with a debtor’s multiple defaults and to distribute the consequences among a variety of different actors)

Elizabeth Warren, Bankruptcy Policymaking in an Imperfect World, 92 MICH. L. REV. 336, 336–38 (1993) [hereinafter Warren, Policymaking] (providing a brief discussion of the alternative theoretical approaches taken by bankruptcy scholars).

But each disputant would want to maximize value for its favored beneficiaries.

After the identification of beneficiaries comes the establishment of priorities. Every bankruptcy system establishes priorities in distribution.97 These priority schemes vary greatly from one country to another. There is a pattern at the top, where three groups are favored in most systems: employees, secured creditors, and governments.98 Even as to those privileged creditors,99 however, the national schemes vary greatly in their generosity to each group. For example, employees are given only a modest priority in the United States100 but have a dominant position in Mexico, per its constitution.101 By contrast, secured creditors have a nearly absolute priority in their collateral in the United States but are subordinated to other interests in a number of countries.102

The only real constant as to priorities in bankruptcy regimes around the world is the omnipresence of priorities, which is the same as saying there is a universal absence of equality of distribution. Equality of distribution (in the

97. WORLD BANK, PRINCIPLES, supra note 21, § 2.2, para. 58

Bankruptcies, in MAKING COMMERCIAL LAW: ESSAYS IN HONOUR OF ROY GOODE 419 (Ross

Cranston ed., 1997) (discussing priorities for distribution in transnational bankruptcies under the EU system)

98. Id. at 30

1.06 (1986) (stating that, in many countries, there are general preferences in bankruptcy for secured creditors, employee claims, and governments)

K. Morgan, Should the Sovereign Be Paid First? A Comparative International Analysis for the Priority of Tax Claims in Bankruptcy, 74 AM. BANKR. L.J. 461, 462 (2000) (noting that government claims are among the highest in bankruptcy priority internationally).

99. The favorable distribution rights that are called “priorities” in the United States are often called (in English) “privileges” in civil-law systems. See, e.g., ALI, MEXICAN STATEMENT, supra note 64, § IV.F.1.c–d (referring to singularly privileged creditors)

100. 11 U.S.C. § 507(a)(3)–(4) (granting employees of bankrupt entities wages earned within 90 days of bankruptcy up to $4,000 per employee, plus limited amounts of employee benefits).

101. ALI, MEXICAN STATEMENT, supra note 64, § IV.F.1.

102. Interestingly, the United States priority is nowhere explicitly stated in the Code. Germany is only one of a number of countries to provide for surcharges on collateral, reducing the secured party’s return from the collateral. See Kenneth N. Klee, Barbarians at the Trough: Riposte in Defense of the Warren Carve-Out Proposal, 82 CORNELL L. REV. 1466, 1477–78 (1997)

(describing the rule recently adopted in Germany). Other countries subordinate secured parties to other creditors. See, e.g., GOODE, SECURITY, supra note 77, at 95 (discussing the subordination of certain types of security interests under British law)

and many other debts come ahead of a security interest in the distribution of the sale proceeds of property subject to a security interest, with the result that the benefits of secured credit are unavailable.”).

common law tradition, “equity is equality”) is often said to be a central tenet of bankruptcy law in every country.103 However, the exceptions do not prove that rule but swallow it.104 A tenet that lacks a single observation cannot be taken seriously. It is true that bankruptcy has as a major purpose equality of distribution to all those within a legal “class” of creditors. That is, those with equal rights should receive equal treatment. As a statement of priority, that point is nearly trivial. Presumably, all law aims to treat similarly situated persons similarly. But as a statement of control, it has content. To under¬stand its content, it is necessary to understand the role of control in the collective process that is bankruptcy. In brief, control is necessary to enforce any system of priority, including a system of equality (an absence of priority).

Every bankruptcy system includes a moratorium (stay) that can effectively block creditor access to some or all of the debtor’s assets.105 The moratorium often restrains lawsuits and other actions to collect debt as well.106 Along with the moratorium, bankruptcy systems routinely give control of the debtor’s assets to the designated controller of the bankruptcy process,107 although with great variation in the legal doctrines that produce

103. JACKSON, LOGIC, supra note 94, at 29–30 (stating the principle is “perhaps the most common—and uncontroversial—of bankruptcy’s policies”)

James W. Bowers, Groping and Coping in the Shadow of Murphy’s Law: Bankruptcy Theory and the Elementary Economics of Failure, 88 MICH. L. REV. 2097, 2101–02 (1990) (describing the equality principle as a “ritual incantation,” questioning its utility, and citing numerous exceptions).

104. This point has recently become recognized in the British literature. See Rizwaan J. Mokal, Priority as Pathology: The Pari Passu Myth, 60 CAMBRIDGE L.J. 581, 582 (2001) (arguing that the equality principle “does not fulfil[l] any of the functions often attributed to it” and “does not constitute an accurate description of how the assets of insolvent companies are in fact distributed”)

see also Look Chan Ho, On Pari Passu, Equality and Hotchpot in Cross-Border Insolvency, LLOYD’S MAR. & COM. L. Q. 95 (2003).

105. WORLD BANK, PRINCIPLES, supra note 21, at 31. In many countries, however, the stay is

not automatic as it is in the United States. Instead, the responsible court is given the power to issue orders restraining seizure of assets and other conduct. See, e.g., The Summary of the Bill for the

Law on Recognition and Assistance of a Foreign Insolvency Proceeding, in 43 JAPANESE ANNUAL OF INTERNATIONAL LAW 331, 334 (Junichi Matsushita trans., 2000)

supra note 64, § II.F.1.

106. See WORLD BANK, PRINCIPLES, supra note 21, at 30 (“[The moratorium] should enjoin

actions by creditors to enforce claims against the enterprise’s assets through collection efforts, adjudication, execution or otherwise.”).

107. For example, in the United States, the controller is the trustee in bankruptcy or the debtor¬in-possession. 11 U.S.C. § 1104. In several other countries, the controller is an administrator or the old management subject to the supervision of a court-appointed official. See, e.g., ALI, MEXICAN STATEMENT, supra note 64, §§ II.C, IV.B (conciliador

supra note 21, §§ I.C.7, I.E.3.c (2003) (trustee

that result.108 On the basis of these two legal preliminaries, the controller can marshal assets, manage an ongoing business, seek bids for assets, bring avoiding actions, and, ultimately, distribute proceeds or otherwise benefit the chosen beneficiaries of the process.

The key point is control. If individual creditors were allowed to seize the debtor’s assets, an orderly liquidation or reorganization would obviously be impossible. That is, employees, secured creditors, or the state— favored in most systems—might well lose value to unsecured commercial creditors in the rush for assets. The control provided by the moratorium or stay ensures that these competitors are restrained, while other doctrines109 ensure that the bankruptcy regime will control the liquidation or recapitalization of the debtor’s assets and the distribution of resulting value to the preferred beneficiaries.110 It has long been recognized that the collective proceeding that is bankruptcy is required to maximize value,111 but it is equally true that the control imposed by bankruptcy law is essential to enforcing the inequality of distribution—that is, priorities—mandated by each legislature. A recent study by a Spanish scholar makes a persuasive historical case for understanding the role of bankruptcy as a system for enforcing priorities and defeating a general or overall equality of distribution.112

108. Some systems contemplate the creation of an estate to which the debtor’s property is conveyed by the law. See, e.g., ALI, MEXICAN STATEMENT, supra note 64, § II.G.1 (“masa” or estate assumes all of the debtor’s property). Other systems conceive that the debtor’s assets become the property of the trustee. See, e.g., ALI, CANADIAN STATEMENT, supra note 22, § I.C.3.e. Virtually all limit the debtor’s power to convey or otherwise control those assets. WORLD BANK, PRINCIPLES, supra note 21, princ. 11, at 32–33.

109. The most important additional rules for enforcing priority systems are as follows: (i) the vesting of control of the debtor’s assets in someone equivalent to our trustee in bankruptcy

110. For this reason, the very absence of control identifies some of the beneficiaries chosen by a particular regime. For example, in some countries, the opening moratorium does not apply to labor (employee or union) claims. See, e.g., ALI, MEXICAN STATEMENT, supra note 64, § II.F.1.a. In others, secured creditors and their collateral are exempt from its reach. See, e.g., ALI, CANADIAN STATEMENT, supra note 21, § I.C.3.d (in liquidation cases). The practical effect of their exemption may be to maximize the value of their recoveries, while lowering the overall recovery for beneficiaries generally. It is not coincidental that the same creditors exempted from the stay in a particular system are invariably favored within a bankruptcy proceeding as well. See, e.g., ALI,

MEXICAN STATEMENT, supra note 64, §§ II.F.1.a, IV.F.1 (labor claims)

111. JACKSON, LOGIC, supra note 94, at 5.

112. See generally JOSE MARIA GARRIDO, TRATADO DE LAS PREFERENCIAS DEL CRÉDITO

(2000) (English translation by Jonathan Pratter and Gloria E. Avila-Villalha, on file with author).

In a model of bankruptcy based on the goal of enforcement of priorities, equality of distribution still has a role, but only to enforce equality (generally understood as pro rata distribution) within a given priority class. Just as bankruptcy control is necessary to ensure the priority of employees as a group, for example, that control is also necessary to ensure that each employee will share pro rata in the enjoyment of that priority, rather than having favored employees, lucky employees, or more aggressive employees do better than the rest. It is in that sense that equality of distribution is an important goal of bankruptcy.

The necessary conclusion is that bankruptcy exists to enforce a set of priorities, but it can be used to enforce any set of priorities that might be chosen by policymakers. The only thing that is essential to bankruptcy is that the bankruptcy regime gets control of the debtor’s assets and be able to dispose definitively of its liabilities, so the assets can be safely sold or re¬capitalized and the value distributed to the chosen beneficiaries.113 Bankruptcy could, in theory, be used to enforce a priority system of nonpriority —that is, complete equality—although it does not appear that even one such system is extant. Thus, priority is the usual goal of bankruptcy, but no particular choice of priority is necessary. On the other hand, control is the sine qua non of any bankruptcy system.114

At this point, it is necessary to introduce the idea of neutrality, a concept never identified in the literature about bankruptcy, but almost as central in many bankruptcy systems as control. This concept deserves an article of its own, but can only be briefly introduced here and modestly elaborated in Part VI. If equality is not the necessary centerpiece of a bankruptcy priority regime, it can be argued that neutrality occupies that position. The core idea is this: If a bankruptcy involves competing interests, then control may be exercised either impartially or with partiality to one interest or another. If the policymaker recognizes both interests as worthy of protection, then its policy requires neutrality in the default manager.115

Professor Garrido describes a tension as bankruptcy law was developing in the seventeenth century between the pro rata rule of the ius mercantile and the priority-heavy “concurso” schemes of government systems in Italy, Spain, and elsewhere in Europe. He suggests that the purpose of changes in the Spanish insolvency laws in that period was to protect the property of impecunious nobles. Necessary to that end was the creation of priorities and control of the entire process by a publicly appointed administrator. Id. at 232–33.

113. In a reorganization, debts may be restructured by being reduced and extended in payment, or the company or its assets may be sold as a unit to pay some or all of the debts. In a restructuring, value is created by effectively recapitalizing the assets at the restructuring “price” because the new capital structure is thought to be sufficiently reduced that the assets can provide a sufficient return on the restructured investment. Often that process requires new loans or new equity investments as well. This explains the usage “recapitalization of assets.”

114. As noted earlier, this statement includes the notion that pervasive control from which favored creditors are excepted is perfectly consistent with bankruptcy as a system of priority enforcement. See supra note 110.

115. It may be that the importance of neutrality has been ignored because it was too obvious, like speaking in prose.

To get an initial idea how this point applies in the context of secured credit, suppose that Congress has decided to accept without reservation the claim that secured credit is the most efficient form of credit in every instance. On that basis, it enacts a bankruptcy system that has only dominant secured parties as beneficiaries.116 In such a system, the manager of the default has only to ensure that the dominant secured party gets the maximum possible recovery. Then its work is done. Although the British Parliament and courts never explicitly adopted such a rule, it has been argued that the receiver system frequently operated as if that were the rule.117 And that system made sense once a decision was made to protect a dominant secured party (there, the floating-charge holder) at all hazards. If there is only one beneficiary, the management of the general default should be in the hands of that beneficiary or its agent, which will then act in the beneficiary’s best interests.118

On the other hand, Congress might have decided instead to have a system that encourages the market to decide on the appropriate combination of secured credit, unsecured credit, and equity financing for each business. That would be a fair description of the present system in the United States. In such a system, the congressional objective of maximizing value for each class of beneficiaries would require a neutral manager charged not merely with the distribution of proceeds according to a set of priorities, but first and foremost with managing the default and arranging the deployment of assets so as to produce the maximum return for each class.119 Such a task is formidable and probably cannot be achieved fully, but it would be the ideal objective of a neutral manager in response to the congressional policy.120

116. If Congress included ordinary secured parties holding interests in a variety of assets of the debtor, then those interests would be in potential conflict, and the result would not be a single¬beneficiary class.

117. See infra notes 262–63 and accompanying text.

118. Cf. Alan Schwartz, A Theory of Loan Priorities, 18 J. LEGAL STUD. 209, 216–17 (1989) [hereinafter Schwartz, Priorities]. In such a system, a necessary premise would be that virtually all the working capital for the business would be supplied by the entrepreneur and the secured party, a point of considerable interest we will not explore further here.

119. The Code explicitly requires neutrality in a trustee’s lawyer and other professional advisors to the estate. 11 U.S.C. § 327(a) (2002) (requiring “disintrestedness” in the trustee’s attorney). Although the Code does not require neutrality of the trustee in so many words, the case law has long imposed on a trustee in bankruptcy fiduciary duties to all those interested in the estate.

See generally 3 COLLIER ON BANKRUPTCY, supra note 72, ¶ 327.04, at 327-59.

120. The task would be difficult because the beneficiary classes would have conflicting interests. For example, the first priority class might be paid in full with a very low risk, low return management, and deployment of assets, while a lower priority class might be better served with a little more risk on the hope of a little more return. See infra notes 231–33 and accompanying text. This point deserves an article of its own, which it will not get here. It should be noted, however, that a congressional decision to give top priority to class A is not the same as a decision that the default manager should maximize returns for that class while ignoring the rest of Congress’s beneficiaries. Cf. Jonathan C. Lipson, Director ‘s Duties to Creditors: Power Imbalance and the Financially Distressed Corporation, 50 UCLA L. REV. 1189, 1192–93 (2003) (arguing that is a mistake to assume a strong connection between duty and priority in right of payment). It is the distinction between those two ideas—priority versus management—that explains the apparent anomaly that Congress is obviously very concerned with both debtors and unsecured creditors yet

Even under the present system, an exception can be posited, in principle, to the proposition that bankruptcy neutrality is essential to the management of a general default. The exception is the case in which there is a valid dominant security interest that is undersecured. That is—even assuming a near-absolute, secured-credit priority as in the United States, so that the secured party has a top priority in all the debtor’s assets—it may be that the maximized value of those assets will still be less than the amount owed to the secured party.121 In that case, it can be argued that there can be no competing priority and no need for bankruptcy. Among other things, this exception explains why almost all instances in which bankruptcy waivers are enforced are single-asset real estate cases.122 In that sort of case, the control exercised by the dominant secured party is arguably all that is necessary or, perhaps, appropriate.

V. Contractualism and Secured Credit

A. Contractualism

A number of scholars have proposed replacing bankruptcy law with bankruptcy bargains established by contracts between debtors and creditors.123 They do not make much attempt to relate their work, each to the other, but it is fair to group these theories as “contractualism,” understood to mean any approach that would permit important bankruptcy rules to be modified or abrogated by contract between the debtor and one or more creditors.124 They are vague about the role of the courts or other institutions in enforcing the hypothetical bargains they imagine being struck, but, presumably, these bargains would establish a set of legal rules governing the recovery process or, at least, contractual waivers and exceptions to the existing bankruptcy regime. Nearly all of the proposals that have been advanced envision a bankruptcy bargain struck ex ante, at the time credit is granted.125

gives no explicit priority to equity. It wants to protect the priority of certain creditors, while hoping to maximize value for lower-priority classes at the same time. As usual, it leaves the hard-to¬reconcile details to the courts.

121. Note that this statement assumes the case in which there is not a going-concern value in excess of the secured debt.

122. See David S. Kupetz, The Bankruptcy Code Is Part of Every Contract: Minimizing the Impact of Chapter 11 on the Non-Debtor’s Bargain, 54 BUS. LAW. 55, 69 (1998) (noting that most

of the decisions enforcing waivers of the automatic stay involve single-asset real estate cases).

123. See, e.g., Barry E. Adler, Finance’s Theoretical Divide and the Proper Role of Insolvency Rules, 67 S. CAL. L. REV. 1107 (1994)

124. Professor Block-Lieb calls them “neoliberterian” theorists. Susan Block-Lieb, The Logic

and Limits of Contract Bankruptcy, 2001 U. ILL. L. REV. 503, 504.

125. See, e.g., infra notes 131–36 and accompanying text. The exceptions include Steven L.

Schwarcz, Rethinking Freedom of Contract: A Bankruptcy Paradigm, 77 TEXAS L. REV. 515

(1999) [hereinafter Schwarcz, Rethinking] (proposing the enforcement of certain waivers in bankruptcy) and Edward S. Adams & James L. Baillie, A Privatization Solution to the Legitimacy of

The term “proposals” with respect to contractualist writings is accurate because none of these approaches devotes a great deal of attention to theory as such.126 They assume for the most part that bankruptcy should have the functions that private parties, especially lenders, want it to have. Following Professor Jackson,127 they assume it has no other functions. On the basis of a general notion of market efficiency, they further assume that a private regime bargained by the parties would produce a more efficient method of achieving those functions than would be provided by any legal regime.128

On the other hand, the contractualists understand that their key difficulty is to govern by contract an event—general default—that is collective and multiparty by definition. They are keenly aware that parties may not ordinarily use a contract between them to bind third parties. As a result, they devote much of their efforts to advancing proposals by which they believe the problems of third-party rights and multiparty relationships can be avoided in a system that turns on a private bargain.

One group of contractualist proposals can be called “automated bankruptcy.”129 One version of automated bankruptcy assumes a tiered system of debt priority. Upon a general default, the lowest priority tier can either invest sufficient funds to pay off all higher tiers or forfeit its interest entirely. The next tier up the ladder then has the same chance and faces the same forfeiture. If the process reaches the top tier without payment having been made, the top tier simply takes over ownership of the debtor and disposes of it as the holders of top-tier debt think most advantageous.130

Another approach is potentially more elaborate than automated bankruptcy. It puts in place an entire bankruptcy system by contract. One variation on this “complete-system” approach supplies a menu of such

Prepetition Waivers of the Automatic Stay, 38 ARIZ. L. REV. 1, 25 (1996) (urging the acceptance of waivers of the automatic stay even while acknowledging that such waivers amount to waivers of bankruptcy protection).

126. The description that follows draws heavily upon that found in ELIZABETH WARREN & JAY

L. WESTBROOK, THE LAW OF DEBTORS AND CREDITORS 1029–42 (4th ed. 2001).

127. JACKSON, LOGIC, supra note 94, at 7–9 (emphasizing the maximization of distributions to creditors as the primary role of bankruptcy law). Professor Jackson’s book, with its “creditors’ bargain,” was the most important book about bankruptcy theory in its generation. Among other effects, it influenced the new German law profoundly. Control was implicit in the creditors’ bargain, but that aspect of Jackson’s approach was unrecognized and abandoned by the contractualists.

128. “Of Coase.” Yet it is interesting to note that a leading contractualist, Professor Alan Schwartz, once proposed the imposition of priority for the first lender by law. See Schwartz, Priorities, supra note 118, at 249–54 (arguing that U.C.C. Article 9 should be amended to give absolute priority to the initial creditor, whether secured or unsecured, except in the case of purchase money secured creditors).

129. Professors Adler and Bebchuk provide the leading examples. See Adler, Puzzle, supra note 8

130. See, e.g., Adler, Financial, supra note 94, at 324–26 (describing the tiered system as a “Chameleon Equity firm” and noting its benefits over the traditional bankruptcy system).

systems, one of which is selected by a debtor at its corporate birth.131 Creditors can then decide whether to extend credit to the debtor given its selection of a particular bankruptcy system from the menu.132 Some other advocates of a complete system by contract propose that each new contract adopting a bankruptcy system would automatically apply to the debtor’s prior creditors, replacing the systems that were in prior contracts.133 In this way, what might be called the “ever-green” complete-system approach seeks to avoid the objection that the menu system is too rigid. Finally, a third variation is more cautious. It would leave the bankruptcy system largely in place, at least in theory, but permit bargains for waiver and replacement of certain provisions—subject to limited post-hoc review by the courts under stated tests of fairness and efficiency.134 A central focus of this last approach is an advance, irrevocable waiver of the automatic stay if bankruptcy should ensue, generally given in exchange for a secured loan after the debtor is already in financial distress.

Recently, contractualist scholars have gone farther, to claim that contractualism is already in place.135 It is claimed that lenders obtain control of debtors by contract from the start of the lending relationship and maintain it through the debtor’s default and bankruptcy.136 There is support for the assertion that lenders have achieved control in some very prominent recent cases, but they do not offer evidence to support the claim that this control results from contracts that pre-date the debtor’s financial distress.137

This short summary does not do justice to the extent to which these contractualist proposals have dominated debate about bankruptcy theories over the last decade. Article after article has proposed the privatization of the management of financial distress through ever more-ingenious contractual schemes. On the other hand, so preoccupied have these scholars

131. See Robert K. Rasmussen, Debtor’s Choice: A Menu Approach to Corporate Bankruptcy, 71 TEXAS L. REV. 51, 100–07, 116 (1992) [hereinafter Rasmussen, Menu] (providing a list of five basic bankruptcy options and arguing that firms should make their choice at the time of capital formation).

132. This proposal allows for change in the agreed bankruptcy regime at a later time by implicit agreement of all creditors. Id. at 118 (noting that “change in the corporate charter should be allowed only with the consent of all of the creditors”).

133. See, e.g., Schwartz, Contract Theory, supra note 123, at 1834–36.

134. See, e.g., Schwarcz, Rethinking, supra note 125, at 585 (arguing for enforcing pre¬bankruptcy contracts that waive bankruptcy rights or change bankruptcy procedures only where the debtor receives value reasonably equivalent to the value of the contract).

135. See, e.g., Adler, Financial, supra note 94, at 334 (arguing that courts may dismiss bankruptcy claims brought after private insolvency arrangements are in place)

136. See Baird & Rasmussen, End, supra note 6, at 778–81 (discussing contracts giving investors control rights).

137. See infra Part VIII.

been with the problem of third-party rights that they have devoted little attention to the institutional management of a general default or to explaining why a private system would be more efficient.138 They have made a useful contribution to the literature because they have forced onto the table the question of privatizing bankruptcy law, one inevitable in the academic epoch of Law and Economics. In the process, their work has identified the central problems with private management of a general default. Professor Rasmussen’s ingenious “menu” idea, for example, put the question of public notice at the center of things, where it belongs.139 Professor Schwartz has seen for some time the central importance to a private system of giving an overriding priority to a single creditor.140 Professor Schwarcz, by focusing on the phenomenon of security interests given after financial distress has arisen, has begun an exploration of the centrality of the choice of a manager of a general default.141

B. Secured Contractualism

The contractualists say almost nothing about secured credit. Their imagined systems of bargained bankruptcies ignore the central pillar of credit bargaining, the security interest, which is also the only privately bargained priority recognized in bankruptcy laws in the United States and around the world. The critics of the contractualists offer a number of sound arguments against these theories142 but stay largely within the same frame of reference, ignoring the relevance of the secured-credit system. This subpart will show that the existence of a highly sophisticated system of secured credit permits achievement of all the goals of the contractualists and solves the otherwise intractable problems presented by their proposals. The only conceivable form of contractualism is secured contractualism.

138. This myopia is consistent with the economic focus of much of this literature. Economists are notoriously deficient in describing the organization of the markets about which they theorize.

See JOHN MCMILLAN, REINVENTING THE BAZAAR: A NATURAL HISTORY OF MARKETS 8–9

(2002).

139. See Rasmussen, Menu, supra note 131, at 66–68.

140. Schwartz, Priorities, supra note 118, at 235–49.

141. See Schwarcz, Rethinking, supra note 125, at 587–89 (arguing that directors will have incentives to choose Chapter 11 reorganization over pre-bankruptcy contracts). His focus is the mid-distress workout, after a company is in financial trouble. The special analysis that is required for mid-distress security interests is another element in the author’s larger project that is beyond the scope of the present Article.

142. See, e.g., Block-Lieb, supra note 124, at 528 –29 (arguing that minimizing the costs of commercial credit through contract does not maximize social welfare)

On the other hand, that conclusion makes the contractualist project a relatively small aspect of the debate that matters most: the costs and benefits of secured credit. The case for secured credit at the transaction (credit¬extending) stage has been the subject of a long and inconclusive debate, leaving the claim of benefits from a secured-credit system problematic at best. Beyond that fundamental difficulty lies a second difficult obstacle that must be overcome by proponents of secured contractualism. As we have seen, an ordinary security interest is insufficient for management of a general default. To be successful, contractualism requires a dominant security interest to support its objective of privatizing management of a general default. We will see that there are substantial concerns about giving over management of a general default to a dominant secured party and that these concerns must extend to any contractualist proposal.

The contractualist theorists envision a world in which creditors can bargain in advance for an assured position in case of the debtor’s general default. A security interest provides just that result. A dominant secured creditor that exercises reasonable care in its arrangements will have a highly predictable legal environment following default—an environment in which it will control the debtor enterprise, the disposal of its assets, and the distribution of the proceeds. Even under present law, a dominant secured party will often be able to veto a bankruptcy. There is little more that any contractualist could ask of a legal regime.

Without the support of a security interest, the contractualist proposals are crippled by intractable problems. The three most important are notice, serial contracting, and control. Bankruptcy law provides standing notice as to the rules governing a general default, permitting credit grantors and others (for example, employees and venture capitalists) to make market decisions on that firm basis. As Professor LoPucki has explained, any contractualist approach that permits bankruptcy laws to be waived or modified must pro¬vide a method by which creditors will know that the debtor has entered into a binding agreement concerning events following its general default.143 The other creditors must also know the contents of that agreement. Otherwise, market actors would have to price their contracts on worse-case assumptions about the agreements entered into by their counterparties (borrowers or traders), which might affect the actors’ risks in the transaction. For the most part, contractualist theories have failed even to address this problem.144 Yet

143. Id. at 2244–45.

144. See, e.g., Bebchuk, Approach, supra note 52

the terms of an agreement for automated bankruptcy, for example, would be of pressing importance to a subsequent creditor considering extending unsecured credit to a debtor. If that party later discovered that the debtor had contracted for an automated bankruptcy regime in case of general default, and if that regime were binding upon the subsequent creditor, that party would recover nothing unless it could organize (and contribute capital to) a “class” effort to buy out higher tiers of debt. Absent notice, any creditor would have to price its credit on the risk that such an agreement existed.145

Similar difficulties arise with respect to the complete-system approaches to contract bankruptcy. At the heart of such a system must be an agreement binding on third parties that a certain creditor will have the right to appoint a new manager of the debtor’s business following default. The provisions of that agreement would be important to various parties, including those considering entering into joint ventures or distribution agreements with a debtor, because of the uncertainties about the competence and conflicting interests of the new manager. To avoid inefficient overpricing in reaction to these risks, a contractualist system must provide for the giving of notice.146

To his credit, Professor Rasmussen sees the centrality of the notice problem and provides for notice to the market through his “menu” approach.147 But his approach does not solve the second problem: serial contracting, which is addressed by Professor Schwartz.148 The difficulty arising from serial contracting is that Professor Rasmussen’s proposal would lock the debtor into an unalterable choice of a bankruptcy regime at an early stage in its business life—a scheme that might be wholly unsatisfactory to its subsequent creditors and to itself at a later stage.149 However, when Professor Schwartz attempts to solve that problem with his last-to-contract or

major banks), available at http://www.enron.com/corp/por/pdfs/examiner3/ExaminersReport3.pdf

145. Another very serious problem is that the subsequent creditor would have to learn and understand the detailed terms of the automated bankruptcy agreement rather than being able to rely upon the rules of a general-default system like the current bankruptcy law that serves, in effect, as a standard, boiler-plate agreement in every case. That difficulty is not resolved by linking the system to a security interest but will not be further addressed in this Article. See generally Elizabeth Warren & Jay L. Westbrook, Contracting Out of Bankruptcy: An Empirical Intervention (unpublished manuscript, on file with author) [hereinafter Warren & Westbrook, Contracting Out] (reporting empirical data that demonstrates the potential for serious informational difficulties arising from contractualism).

146. These are, of course, the same kinds of problems created by a system of secured credit with priority and control assurances for the secured party, although the contractualist problems are more complex and difficult to solve.

147. See Rasmussen, Menu, supra note 131, at 114–16 (explaining that the selection from the “menu” will provide notice as to the actor’s risk-taking behavior and the risks that might be involved).

148. Schwartz, Contract Theory, supra note 123, at 1833–36.

149. See LoPucki, Cooperative Territoriality, supra note 142, at 2243–51 (analyzing the weaknesses of Rasmussen’s contractualism).

ever-green approach,150 he is left with a somewhat bizarre system in which a bankruptcy scheme in the contract of a supplier of copy paper to the debtor might be held binding on J.P. Morgan Chase, the debtor’s principal lender. Any proposal that creates a hard-to-change contractual scheme (menu) or that separates the bargaining process from a major supplier of the credit that will be at risk (ever-green) is unworkable on its face.151

Furthermore, in an ever-green contractualist system with successive amending contracts, there would inevitably be ambiguities as to which contract was the controlling one. How would disputes of this sort be resolved and who would control the enterprise while they were being resolved? These issues would involve a thousand points of important detail. The details themselves need not detain us. The point here is that, if there is to be legal enforcement of these contracts, especially in the context of third¬party rights and multiparty proceedings, all sorts of provisions would have to be developed, adopted into law, found wanting in various respects, and revised and tried again—all before such a system could hope to succeed. Such a project—fat with expense, dislocation, and risk—would never be undertaken if the end could be achieved in another fashion. It would be academic in the worst sense of the word. As explained below, a far simpler and more elegant solution is available by linking contractualism to a dominant security interest, thereby creating secured contractualism.

The central difficulty for contractualists is the problem of control of the debtor’s assets. There are two control problems: pre-default and post¬default. If the contractualist creditor is unsecured, the most intractable problem is the absence of asset constraint152 prior to default. Even the debtor who has locked itself into one of the menu choices proposed by Professor Rasmussen is left free to manipulate its assets, especially through the use of multiple corporations in multiple jurisdictions and conveyances to good-faith transferees. Certain transfers that are fraudulent or without real benefit to the debtor might be set aside under bankruptcy or other law, but many transfers to good-faith transferees will not be recoverable for the benefit of unsecured

150. Schwartz, Contract Theory, supra note 123, at 1833–36

151. Indeed, that scheme seems flatly inconsistent with Professor Schwartz’s interesting earlier proposal for absolute priority to the first lender. See supra note 128.

152. See supra section II(A)(1).

creditors.153 As noted earlier, a lawsuit against the debtor is generally an exercise in futility.154

Furthermore, the debtor may not be making the transfers in bad faith. Its manipulations may be motivated by tax or financial-reporting considerations (as may have been true on some occasions with Enron) rather than an intent to defraud creditors. But, whatever the motive, the effect of the manipulations may be to leave creditors with a marvelous bankruptcy scheme vis-á-vis an empty shell of a debtor. An unsecured contractualist system leaves the counterparty utterly exposed to a complete restructuring of the business and conveyance of assets. While such transfers may be breaches of contract or even ultra vires, they will often be effective as property transfers, leaving the contractualist with only an unsecured claim for damages against a debtor already in general default who will pay little or nothing to unsecured creditors.155

The easiest example of the risk of such transfers is the debtor’s grant of a dominant security interest in violation of its bargain with the contractualist creditor. The debtor signs a contract with the creditor containing elaborate bankruptcy provisions. A year later, under financial pressure, the debtor grants a dominant security interest in violation of that contract. Despite any covenants against such a grant, the grant is fully effective under Article 9.156 Upon the debtor’s general default, the contractualist creditor is left with an unsecured contract scheme while the secured party is selling all the assets of the company and distributing the proceeds to itself.157

That example invokes the wisdom of Professor Gilmore. In his famous treatise, he gave the back of his hand to “negative covenants,” mere

153. 11 U.S.C. § 548. Indeed, a transfer of assets in breach of contract is not viewed as necessarily dishonest or evidence of bad faith, as “efficient breach” theory demonstrates. See

generally Peter Linzer, On the Amorality of Contract Remedies—Eficiency, Equity, and the Second Restatement, 81 COLUM. L. REV. 111 (1981)

154. See supra section II(A)(1).

155. One who thought that such things could not happen with public companies because of disclosure requirements would be mistaken. See supra note 144.

156. U.C.C. § 9-401(b). Professor Mann found that a major benefit to security is the power to

enforce covenants. Ronald J. Mann, Explaining the Pattern of Secured Credit, 110 HARV. L. REV. 625, 669–74 (1997) [hereinafter Mann, Explaining].

157. A legal reform putting the unsecured contractualist ahead of the secured party would change that result, but that seems an obviously bad idea—exchanging for no good reason a perfectly workable and well-understood priority and control system for a new and untested one. Rather different is the proposal to give the beneficiary of a negative-pledge clause—a sort of reverse security interest—priority over subsequently perfecting secured parties. Carl S. Bjerre, Secured

Transactions Inside Out: Negative Pledge Covenants, Property And Perfection, 84 CORNELL L.

REV. 305 (1999). Nonetheless, such a system would constitute fruit-basket-turnover of existing systems, no doubt with unforeseen consequences. Thus only a powerful case for its benefits could support its adoption—yet another interesting discussion we must omit here. But it should be noted that the Bjerre proposal is really a modification of the secured-credit system itself through notice and a sort of priority, so it is conceptually consistent with the model presented here—depending upon the details.

contractual promises made to unsecured lenders to the effect that the debtor would not grant security to anyone else.158 Professor Gilmore viewed them as nearly valueless for the reason just given.159 The elaborate contracts contemplated by the contractualists are simply expanded versions of those illusionary bargains.160

Given that the contractualists fail to address collateral protection before default, it is not surprising that they also ignore the problems of collateral control and enterprise control after a general default, although control is essential to their schemes. To enforce their covenants, they would have to procure legislative enactment of a new code with extensive provisions enforcing the rights necessary to their proposed regime. It is not possible to manage a general default without enterprise control, which is given to the trustee in bankruptcy by the automatic stay. In a contractualist regime, a bristle of court orders would be required following a general default. Even if the procedures are to be governed by contract, the new code would have to provide default rules for all the points not covered in a given contractual bargain and regulate the inevitable abuses that arise in reaction to all legal innovations. For example, one element required under a system of auto¬mated bankruptcy, although not discussed by its proponents, would be a method for determining interim control of the defaulting enterprise while the parties to the various automated debt instruments worked through the schemes’ contractual option process. Who would appoint the controller pending the completion of the process? Would it be the highest tier creditors or the lowest or a neutral party? Would the controller’s incentive be to protect the minimum asset value, potentially leaving the lower tiers with no remaining value? Or would its incentive be to maximize asset value for all tiers although some risk would be necessary? The power to appoint the controller would undoubtedly affect the likelihood that the appointee would choose one path or the other. This problem is discussed further in Part VI.161

158. 2 GILMORE, supra note 22, § 38.4, at 1017.

159. Id. For a contemporary confirmation of that view, see Jonathan R.C. Arkins, “OK—So

You’ve Promised, Right?”: The Negative Pledge Clause and the Security It Provides, 15 INT’L

BANKING L. 198, 204 (2000) (concluding that a negative pledge clause has little value). Another recent author agrees with that conclusion, but would change it by making negative-pledge clauses registrable and therefore enforceable. Bjerre, supra note 157, at 306–07. However, Professor LoPucki argues that negative covenants do work in the case of large, public companies. LoPucki, Unsecured Bargain, supra note 30, at 1926–27. Short of a full rebuttal of that interesting idea in this Article, it must suffice to say here that it applies only to large companies and assumes that security interests are always obvious as such. The kind of off-book deals revealed to the public in Enron and other cases suggest that even public companies may enter into what amount to secured transactions without calling them by that name and that lenders, one is sorry to say, may be willing to ignore the legal risks involved. See supra note 144.

160. One article has claimed that mere covenants give priority in distribution and make the “ownership rights” given by a security interest “unimportant” in a commercial context. These assertions are unsupported legally or empirically. See Schwartz, Priorities, supra note 118, at 212– 13, 243.

161. See infra subpart VI(B).

These three problems—notice, serial contracting, and asset control—are impossible to solve in the contractualist system but are easily solved once a security interest is granted to a dominant secured creditor.162 As discussed above, we have a highly sophisticated system of secured credit, long refined, that has largely overcome these very difficulties.163 If the contractual powers desired by the contractualists are given to a dominant secured creditor, a simple Article 9 filing would give effective notice to the world, directly and through private credit agencies, that would solve instantly the otherwise hopeless problems of notice. By virtue of that notice (that is, perfection),164 the secured party would have a high level of assurance of asset constraint and thus priority.165 The debtor could not dispose of its assets, or interests in them, because good-faith transferees would be deterred by notice and transfers in the teeth of notice would be recoverable. As against a dominant

162. The critique of contractualism in this Article applies a bit differently with regard to waiver proposals like the one offered by Professor Schwarcz. See Schwarcz, Rethinking, supra note 125, 584–90 (proposing enforcement of certain bankruptcy waivers). This Article focuses primarily on the automated and complete-system approaches, although the fundamental difficulties of a waiver proposal are closely related. The critique of waiver proposals cannot be fully realized in this Article, but five points are key. The first is notice. Permitting enforcement of waivers will effectively make many debtors reorganization ineligible in the sense that bankruptcy relief will be ineffective to permit reorganization because of the waiver. It would be very important for creditors, both those existing at the time of waiver and future creditors, to have notice of the debtor’s dramatic change in status from a reorganization-eligible to a reorganization-ineligible debtor. An Article 9 filing would provide this notice, albeit inadequately

163. See supra section III(A)(1).

164. Naturally, once the role and effect of a dominant secured creditor became fully appreciated, questions would arise about the adequacy of the present Article 9 safeguards. Those safeguards have been debated for the most part as if most secured creditors were ordinary secured parties. Much more elaborate disclosures might be appropriate for secured creditors claiming the power to manage a general default. Indeed, one may ask if more disclosure should be required now for those with dominant security interests.

165. The Article 9 provisions operate through a negative predicate. Secured parties trump all others with interests in the collateral—unless there is a stated exception in favor of a stated claimant with a stated sort of interest under stated circumstances. U.C.C. § 9-201(a) (“Except as otherwise provided in the [U.C.C.], a security agreement is effective according to its terms between the parties, against purchasers of the collateral, and against creditors.”). An exemplary exception is that in which a buyer of consumer goods from another consumer may defeat a secured party’s interest in the goods if the security interest was perfected by a method other than filing. Id. § 9-320(b).

secured creditor, the debtor would be well aware that it would violate covenants at the risk of almost instant loss of control of its entire enterprise. Finally, in the case of general default, the dominant secured creditor would simply take enterprise control and manage affairs in its own best interests, within broad limits. Virtually everything the contractualists desire is available under present law to a dominant secured creditor.

This simple move would solve most of the problems to which so many law review pages have been devoted by providing Article 9 notice, Article 9 collateral protection, and Article 9 enterprise control—all under well¬established, routinely-enforced existing law. A dominant security interest is the only extant method for achieving the asset constraint and enterprise control necessary to contractualism. It is the necessary and sufficient condition for the privatization of bankruptcy law, that is, privatization of the management of a general default.

Yet there is a serpent in this contractualist fruit, to be discussed in the next Part.

VI. The Problem with Secured Contractualism

On the analysis just completed, the whole contractualist project collapses into secured-credit theory, which gives rise to two sorts of serious problems for its proponents. The first difficulty —the transactional efficiency problem—is that, after twenty-five years of debate, the efficiency of secured credit remains problematic. Its proponents in that debate have gotten no farther than the Scottish verdict, “not proven.” Furthermore, the value of control to the secured party and its cost to the debtor have received little attention in the efficiency debate, leaving the debate incomplete, as well as unresolved. Given that contractualism would require a widespread shift to financing by means of dominant security interests, the inconclusive and incomplete evidence for the efficiency of such a system would make a move to a contractualist system quite risky.

The second difficulty is that the secured-contractualist proposal requires a certain kind of security interest, a dominant one, to permit management of a general default in substitute for a public bankruptcy regime. But management by a dominant secured party raises serious and unresolved questions in any system that encourages the extension of credit by anyone other than a dominant secured party because a system under the secured party’s control may often fail to realize full value for the debtor’s assets. This point makes patent what has been only implicit in the contractualist proposals: that there will be a conflict of interest between the party appointed by contract to manage the debtor’s general default and the remaining creditors and other beneficiaries. When the contractualist creditor is revealed as a dominant secured party, the connection becomes clear.

These two points, which represent severe obstacles to the contractualist project, are discussed in this Part.

A. The Eficiency of Secured Credit

For more than two decades, academics have engaged in a wide-ranging debate about the efficiency vel non of secured credit. No other subject has dominated commercial-law scholarship to such an extent for so long with such inconclusive results. Yet the fact that contractualism must rest upon a secured-credit foundation means that a positive finding for the efficiency of secured credit is essential to contractualism.

The long debate began with a classic article by Professors Jackson and Kronman in the Yale Law Journal in 1979.166 They asked if secured credit is socially beneficial.167 This article was seminal indeed, as a vast law review literature grew from it.168 Traditionally, secured credit was thought to be useful because it reduced the risk to the secured party, permitting a lower interest rate, as well as enabling loans that might otherwise not be made at all.169 In effect, the debtor and creditor could split a reduction in costs that followed a reduction in risk. But Jackson and Kronman identified a difficulty with that argument: Any reduction in the secured party’s risk was necessarily accompanied by an increase in risk for unsecured parties, who would therefore increase their charges and ration their credit, leaving the overall costs at best the same.170

These authors, like many who followed them, were sure that secured credit must be efficient.171 The task then became, in Professor LoPucki’s felicitous phrase, like that of Cinderella’s sisters, trying “to fit the glass slipper” of efficiency to a theory of secured credit.172 All the resulting theories began with debtor misbehavior, labeled “moral hazard,” which was the common enemy of all creditors. The “moral hazard” was that the debtor would borrow on the basis of business plan A and then operate under business plan B.173 Plan B would have a greater upside, all of which would

166. Jackson & Kronman, supra note 8.

167. It is interesting that their central concern in this article was to demonstrate the efficiency of the priority choices made in Article 9. Id. They raised the “fundamental question[]” of the efficiency of security interests generally only as a necessary predicate. Id. at 1146.

168. See supra note 8.

169. Carlson, Eficiency, supra note 8, at 2180 (noting that this justification has been dismissed as “hackneyed”).

170. Jackson & Kronman, supra note 8, at 1153–54. Given that—even under the Article 9 system—secured credit has some cost, the result would actually be a loss of social wealth because of the net increase in costs.

171. See id. at 1149–57.

172. LoPucki, Unsecured Bargain, supra note 30, at 1894 n.23.

173. Somehow in these analyses, it was generally only the debtor who faced moral hazards, rarely the creditors. See, e.g., Levmore, supra note 8, at 51–52 (discussing the two potential categories of debtor misbehavior: (1) conversion misbehavior where “an individual or group that is involved in the management of the firm takes company assets and uses the proceeds for its benefit”

be captured by the debtor, at the cost of greater risk, most of which would be borne by the creditor. The debtor would propose to the lender, let’s say, a video shop. Then, with the lender’s money in hand, the debtor would purchase a race car because all of the upside of the risky business would fall to the equity owners, while the risk of failure would be borne primarily by the lender (and the driver).174 To protect against this moral hazard, the lender takes a security interest,175 which makes it difficult for the debtor to transfer assets and therefore difficult for it to change businesses.176

Jackson and Kronman postulated that the cost of “monitoring” the debtor was the key.177 If taking the security interest reduced the secured creditor’s risk and monitoring costs enough, it would outbalance the increase in the unsecured creditor’s added risk and monitoring costs whenever the unsecured creditor had lower monitoring costs than the creditor who wished to be secured.178 Professor Schwartz, in two classic articles, destroyed this and other monitoring arguments, and they have been largely abandoned.179 Yet many other defenders of secured credit proffered other feet to fit the efficiency slipper, only to face withering criticism from Schwartz and others.180

ventures,” or “fail to exercise the effort needed to develop profitable opportunities fully”)

Book Publishing, Venture Capital Financing, and Secured Debt, 8 J.L. ECON. & ORG. 628, 649

(1992) (noting that lenders may be irrationally loathe to allow debtors to enter into new ventures because significant profits will not accrue to the lender)

174. See, e.g., Shupack, Filing, supra note 67, at 790 (exploring how debtor and creditor motivations diverge and conflict when debtors contemplate pursuing business activities different from the original purpose for which credit was advanced).

175. These articles failed to make it clear that a security interest would broadly constrain the debtor’s misconduct only if it was a “dominant security interest,” a concept described supra section III(A)(2).

176. Carlson, Eficiency, supra note 8, at 2188–89

177. See Jackson & Kronman, supra note 8, at 1149–57 (discussing the costs and benefits of a monitoring policy).

178. Id.

179. Alan Schwartz, Security Interests and Bankruptcy Priorities: A Review of Current Theories, 10 J. LEGAL STUD. 1, 9–14 (1981)

180. Compare White, supra note 8, at 479–502 (making the case for the economic efficiency of granting personal property security and discussing Professor Schwartz’s criticisms), and Kripke, supra note 8, at 971–72 (arguing that the “judgment of the market” provides strong support for the economic efficiency of security arrangements and that Professor Schwartz “ought to carry the burden of showing why the judgment of the market is not to be accepted”), with Thomas H. Jackson & Alan Schwartz, Vacuum of Fact or Vacuous Theory: A Reply to Professor Kripke, 133 U. PA. L. REV. 987, 1001 (1985) (concluding that “[Professor] Kripke’s methodology is utterly bereft of serious intellectual support”).

In general, the task was to find a benefit to other creditors that offset the increased risk to those creditors. Schwartz argued that a classic economic model would always return the overall costs to equilibrium, rising on the unsecured side as they fell on the secured side.181 He also argued that every theory that showed the efficiency of secured credit necessarily proved too much.182 If secured credit were efficient on the theory presented, all debt would be issued secured, a pattern far different from the reality observed in the financial markets.183 A number of articles written since have attempted to show that there are a series of reasons that security would be used in one circumstance or industry and not in another, so that security might be efficient despite its failure to be ubiquitous.184 However, these articles have not succeeded in establishing a persuasive response to Schwartz’s first point. That is, they have not been able to demonstrate a benefit to unsecured creditors that would reduce the increase in their costs arising from the loss of access to the collateral in case of default. Absent that reduction, no mechanism has been persuasively proposed by which the increase in unsecured credit costs would not equal or exceed the reduction of cost to the secured party, producing a net loss of wealth.

Given the failure to achieve that goal, there remained the question: Why does secured credit flourish if it is inefficient? Professor LoPucki’s conclusion was that the parties who bore the primary effects of its inefficiency were involuntary creditors, especially tort creditors, who could not raise their “prices” to offset the costs imposed by security interests.185 Thus the benefits of security interests obtained by Citigroup are built on the backs of the maimed and killed, an incendiary proposition that has drawn strong responses.186 Subsequently, Professors Bebchuk and Fried have further developed that point, arguing that not only tort creditors but many other unsecured creditors are “weakly adjusting” or nonadjusting.187 To the extent they cannot adjust their costs or ration their extensions of credit, there is a surplus created from which the debtor and the secured party can split the benefits of security.188 On this theory, of course, there is no efficiency gain from secured credit, but only a redistribution of a distinctly troubling sort.

181. Schwartz, Puzzle, supra note 181, at 1054. But see David G. Carlson, Secured Lending as a Zero-Sum Game, 19 CARDOZO L. REV. 1635 (1998) [hereinafter Carlson, Zero-Sum] (arguing that

security is not a zero-sum game because the secured party makes available capital that the debtor would not otherwise be able to obtain).

182. See Schwartz, Puzzle, supra note 181, at 1060 (explaining that such theories are unsatisfactory because their predictions diverge from reality).

183. Id. at 1062.

184. See, e.g., Hill, Eficient, supra note 36, at 1148–56 (discussing different circumstances affecting the use of secured credit and citing a number of articles addressing the same).

185. LoPucki, Unsecured Bargain, supra note 30, at 1896–97.

186. See, e.g., Hill, Eficient, supra note 36, at 117 (suggesting that the practice of securing debt is driven more by the economic efficiencies it generates than by its tendency to externalize costs).

187. Bebchuk & Fried, supra note 35, at 864–66.

188. Id.

Until recently, the only writer to propose a solution related to the concept of control was Professor Scott.189 Although he did not elaborate on the idea of a dominant security interest, he clearly had in mind a dominant secured party.190 The social benefit that would arise from a security interest would often be part of a long-term relationship between lender and debtor.191 The security interest would give the lender a tether that would do more than prevent “misbehavior”

presented. 196

From the perspective of the defenders of security, the result of these years of debate was at best inconclusive.197 No one was able to make the case that there is some benign general effect of secured credit that outweighs the zero-sum relationship between its reduction of the secured party’s costs and its increase in the costs of unsecured creditors, adjusting or not. Given

189. See Scott, Relational, supra note 4, at 926–28.

190. See id. at 904 (explaining that “the leverage obtained by holding the debtor’s assets hostage empowers the secured creditor to influence the debtor’s business decisions”).

191. Id. at 917–19.

192. Id. at 926–27.

193. Id. at 927 –28. For examples of other scholars who have discussed the concept of leverage, see Mann, Verification, supra note 36, at 2244 (noting that some scholars believe that repossession and liquidation costs “asymmetrically” disadvantage the borrower in relation to the lender and provide the lender with leverage)

194. See supra note 42 and accompanying text

195. Cf. Robert E. Scott & Thomas H. Jackson, On the Nature of Bankruptcy: An Essay on Bankruptcy Sharing and the Creditor’s Bargain, 75 VA. L. REV. 155, 171–74 (1989) (analogizing

conflicts among creditors on the “eve” of debtor insolvency to the concept of “general average contribution” in admiralty law—under which doctrine a captain may jettison whatever cargo necessary to prevent a ship from sinking, and all cargo owners share the expense of the lost cargo according to their percentage of ownership of the total cargo traveling on the ship).

196. It may be that one reason for its neglect is that its central argument relates to what this Article calls a dominant security interest. Many scholars have wanted to make arguments that apply to all secured credit, ignoring the crucial distinction between the ordinary and the dominant security interest. As this Article demonstrates, there is much to be done to understand the functioning of secured-credit law and its interaction with bankruptcy in the pure case of a dominant security interest, before undertaking the much more complex application of these concepts to the enormous variety of security interests found in the market.

197. See, e.g., Note, Switching Priorities: Elevating the Status of Tort Claims in Bankruptcy in Pursuit of Optimal Deterrence, 116 HARV. L. REV. 2541, 2552 (2003) (stating that the view that secured credit is efficient is “dominant” but “most definitely not unanimous”).

that, Professor Mann decided to ask a new question: What do secured creditors believe they get from security interests?198 What do they value about security interests? He undertook rigorously structured empirical research, a combination of interviews with lenders and an examination of a large number of lending files.199

Professor Mann marshaled evidence from his research to show that control was the central benefit to secured parties, with priority upon sale a distant second.200 This finding went to the heart of the prior debate because the problem of equilibrium in the allocation of priority was the central difficulty raised by Professor Schwartz and addressed by a generation of scholars. If, as Professor Mann demonstrated, control is the key benefit of secured credit to the secured party, then control must be essential to secured¬credit theory—including any claims about efficiency. But except for Professor Scott, the debate had little to say about control. While Professor Mann’s work was enormously valuable, it did not attempt to resolve the efficiency issue,201 and little has been done by others to build on his work.202

This brief summary of hundreds of law review pages demonstrates that the result of the debate concerning the efficiency of secured credit from the perspective of the debt-capital system as a whole is inconclusive and incomplete. It is inconclusive because of the failure to demonstrate that the Schwartz, LoPucki, and Bebchuk/Fried critiques are substantially unfounded. If those critiques are cogent, secured credit may be redistributional and inefficient in many circumstances. The debate is also incomplete because the benefits and costs of control in its various aspects have been almost entirely ignored. No one has attempted to include the positive value of this material factor in any of the equations written regarding the efficiency of secured credit. Equally important, no one, not even Professor Scott, has attempted to include in the analysis the costs that may be associated with the diminution of debtor control, especially in the case of a dominant security interest.203

198. Mann, Strategy, supra note 4, at 160.

199. Id. at 163.

200. However, he does not make the distinction between “asset constraint” and post-default “collateral control” that is made in this Article. Id. at 160–61.

201. His work did make another very important contribution to that debate. It demonstrated that the impact of default rules ex ante (that is, transactionally) may be marginal, meaning that any efficiency gains may have only minor effects on the cost and availability of credit. Id. at 234–35. That point is of great and ongoing importance to theoretical debates in the field of financial distress because of the claim that ex ante (transactional) effects, as opposed to procedural fairness and similar values, are the ones that really matter. See Douglas G. Baird, Bankruptcy’s Uncontested Axioms, 108 YALE L.J. 573, 589–93 (1998) (highlighting the debate between “traditionalists” and “proceduralists” over whether ex ante effects should be a key consideration in formulating bankruptcy law).

202. It should also be noted that his work deliberately focused on problematic loans, not loans in bankruptcy or necessarily in general default. Mann, Strategy, supra note 4, at 163.

203. It has been observed that dominant security interests are less likely to be observed as the size and creditworthiness of a company increases. Indeed, that observation is one of the telling

Indeed, the special benefits and costs associated with dominant security interests have barely been mentioned, much less analyzed or measured.

As things stand, it is quite possible that dominant security interests produce net inefficiency rather than an efficiency gain when viewed across a wide range of transactions. Yet that conclusion would doom contractualism. Adoption of contractualism would require wholesale adoption of dominant security interests, a state quite foreign (literally as well as figuratively) to the United States market.204 If secured credit may be inefficient, always or often, then a wholesale shift in the United States debt market to one in which dominant security interests are universal or even typical would seem a very risky experiment and one unlikely to be undertaken. If that is true, contractualism is not a serious alternative to the existing system.

Efficiency aside, there is an irony in the fact that contractualists are primarily interested in large public companies.205 These are the very companies least likely to grant dominant security interests, absent financial distress.206 Thus they find themselves required to assert that the large companies that disdain dominant security interests now will be eager to adopt them when they are linked to a contractualist system. That proposition is highly implausible.

B. The Incentive Problem

Beyond the general arguments about the efficiency of secured credit, contractualism assumes management of a general default by a person or persons designated by contract. As we have seen, contractualism necessarily rests on a dominant security interest

arguments against the efficiency of secured credit: Would large companies shun financial efficiency? See Hill, Eficient, supra note 36, at 1136–37, 1141 (stating that a “key factor in an account of secured debt is the classification of firms” and observing that higher-quality firms “secure a smaller proportion of their assets than do lower-quality firms”)

204. Professor Hill has developed some evidence that small businesses always grant blanket (that is, dominant) security interests, but the evidence is mixed. See Hill, Eficient, supra note 36, at 1139 n.103.

205. See, e.g., Baird & Rasmussen, End, supra note 6 (passim).

206. See Hill, Eficient, supra note 36, at 1141 (conceding that large companies do not ordinarily grant sweeping security interests, but asserting that they do grant security interests in particular assets).

a general default by a dominant secured party is problematic. One reason is the incentive problem.

The incentive problem applies to both ordinary secured parties and dominant secured parties. It arises in ordinary debt enforcement as well as in a general default. Any creditor-controlled sale—the sale aspect of what we have called “collateral control”207—gives rise to a serious risk of socially undesirable results.208 Specifically, creditor control risks realization of substantially less than the full value of the asset being sold. These risks are well-recognized. Indeed, even the drafters of the Revised Article 9, despite their profound commitment to vindicating the rights of secured parties, have attempted to address some of these concerns in the debt-enforcement process. 209 A full discussion of the details of this problem is for another article, but the main points are described below. The incentive problem is important in a number of contexts, but it is of the highest order of importance when the issue is management of a general default by a dominant secured party.

There are two aspects to the problem. The first is a classic “free-rider” problem: the secured party lacks incentives to achieve efficient and socially desirable results. The second is the risk of self-interested behavior leading to socially suboptimal results. Exemplary of the risks created is the fact that both types of incentive difficulties may cause secured lenders to want to liquidate a debtor quickly to maximize the value of their security interests, even if delayed liquidation or reorganization might be in the best interests of other stakeholders.210

207. The power of self-help, nonjudicial seizure, is also very valuable. It also creates serious social dangers, which is why Article 9 requires as a precondition that self-help not threaten “a breach of the peace.” U.C.C. § 9-609(b)(2). The social danger implicates a major purpose of bankruptcy law, in my conception, although one little noted in the literature. That purpose is social control of circumstances likely to lead to social disruption and community conflicts. Westbrook, Globalisation, supra note 93, at 405. I do not mean to ignore its importance by putting it to one side for the purposes of this Article.

208. See U.C.C. § 9-615 cmt. 6 (recognizing that a “secured party sometimes lacks the incentive to maximize the proceeds of disposition”)

(positing situations in which a secured party might lack the incentive to maximize disposition proceeds)

209. U.C.C. § 9-615 cmt. 6. See generally Andrea Coles-Bjerre, Trusting the Process and Mistrusting the Results: A Structural Perspective on Article 9’s Low-Price Foreclosure Rule, 9 AM. BANKR. INST. L. REV. 351 (2001)

210. See, e.g., Riva D. Atlas & Jonathan D. Glater, WorldCom’s Collapse: Market Place

The free-rider problem arises from the obvious fact that the secured party managing collateral sales has no incentive to realize more than the amount of its debt. Anything above that amount must be distributed to other secured parties, the bankruptcy trustee, or the debtor—none of whom bear the costs and risks of the sales.211 Although Article 9 imposes significant procedural requirements for secured-party sales,212 these provisions leave a broad zone of reasonableness within which quite different sale values might be obtained. As we have seen, great gaps separate liquidation value, market value, and going-concern value, but the secured party has no incentive to realize more than the value that will pay the secured debt in full.

Imagine the asset-realization department at a finance company.213 The manager estimates that the assets of General Kompute can be quickly sold to realize sufficient proceeds to pay the company’s secured debt in full. Is she likely to do the job herself or give it to the best and most creative marketer in the department? Surely not, if she is the rational maximizer we usually suppose people to be. Because satisfaction of the lender’s interest will be simple, she surely will give the job to the dumbest asset-disposal person in the office, the one hired solely because he was qualified as someone’s brother-in-law. She will also give the brother-in-law a list of Article 9 procedural requirements—requirements designed to be idiot-proof in most circumstances.214 As a result, if substantially higher values could have been realized by a highly competent seller investing time and resources, they will be lost in this commonplace circumstance—to the injury of other stakeholders. Within a broad zone of reasonableness,215 the secured creditor will not be liable for that loss of value. In the context of that branch of secured-credit law devoted to enforcement of particular debts in a single default, this balance between efficiency and protection of other possible claimants may be reasonable.

The second part of the incentive problem has a somewhat darker moral hue. Indeed, it fills that near void in the literature concerning the “moral

insolvency law has also run into difficulties when a few large banks with conflicting interests have control of the insolvency proceeding. See Christoph G. Paulus, Germany: Lessons from the Implementation of a New Insolvency Code, 17 CONN. J. INT’L L. 89, 92–93 (2001)

supra note 6, at 937 (discussing the DIP financier’s conflict of interest analysis)

211. U.C.C. §§ 9-608(a)(4), 9-615(d).

212. These include notices to interested parties and a commercially reasonable approach to advertising and sale. U.C.C. §§ 9-610 to 9-614.

213. This discussion suggests an attractive empirical project. Professor Mann has done some important work in this regard, but much remains to be done. See supra subparts III(A), VI(A).

214. U.C.C. § 9-627. It should be noted that Article 9 protects the selling party by permitting deduction of all reasonable costs of sale off the top. U.C.C. § 9-615(a)(1). The same rule applies in bankruptcy. 11 U.S.C. § 506(c).

215. See Jack F. Williams, Debunking the Myth Regarding Article 9 Collateral Dispositions, 9 AM. BANKR. INST. L. REV. 703, 707 (2001) (noting that Article 9’s requirement of commercial reasonableness allows room for considerable flexibility).

hazards” of creditors as a complement to the oft-discussed moral hazards of debtors.216 The problem is the creditor’s positive incentive to acquire the collateral at its own sale at a very low price and then resell the collateral at a much higher price for its own account. The hazard is created by two legal rules. The first is the flexible-sale rule already discussed, which grants a broad zone of reasonableness in the advertising and conduct of a sale. The second rule permits a secured party to “bid in” at its own sale. That is, the secured party is allowed to use the debt owed to it in lieu of an actual cash payment for any bid it may make at its own sale.217 Thus, if the creditor is owed $100,000, it can bid any amount up to $100,000 and defeat any lesser bid without having to produce any cash.

In both judicial and private auction sales, there are often strict requirements for a bidder other than the secured party.218 In particular, the bidder may have to bring sufficient cash to cover its bid or to provide cash payment very shortly after the bidding.219 For this and other reasons, it is often the case that few other bidders appear at foreclosure and repossession sales.220 This fact combines with the bidding-in rules to make it possible for secured parties to buy at their own sales at a price well below market value while avoiding sanctions for violating Article 9’s notice and sale procedures.

The classic recent instance was the sale of the infamous Brentwood home of O.J. Simpson .221 Although the sale was of real estate rather than personal property, the economic factors were typical. At the well-attended foreclosure sale, the bank holding the mortgage bought the property for the amount of the mortgage ($2.6 million) “bid in” plus $31,000. It thus defeated the only other active bidder, who began the bidding at the amount of the mortgage plus one dollar. Less than a month later, the bank listed the property for sale at $3.95 million, the approximate price for which it sold about six months later .222

216. See supra note 173 and accompanying text.

217. U.C.C. § 9-610(c). “The secured creditor has peculiar advantages over other possible purchasers. Not only does he determine the time and place of the sale but to the extent of the secured obligation, he does not have to put up money since he pays by offering a credit against the obligation.” 2 GILMORE, supra note 22, at 1241–42.

218. Id.

219. LYNN M. LOPUCKI & ELIZABETH WARREN, SECURED CREDIT: A SYSTEMS APPROACH

69 (3d ed. 2000) (stating that “[t]ypically, that bidder must immediately make a deposit of a portion of the purchase price in cash or by cashier’s check” and generally the balance of the purchase price must be paid within a few hours or days).

220. See, e.g., id. at 76–84 (discussing the problems with judicial sale procedures that lead to prices far below market value and few bidders other than the secured creditor).

221. Carla Hall, More Hoopla at Simpson Home Auction, L.A. TIMES, July 15, 1997, at A1

222. See, e.g., In re Cearley, 295 B.R. 892 (Bankr. W.D. Ark. 2003) (refusing to reduce a deficiency claim although secured party after “a few months” obtained in a subsequent sale $125,000 over the bid-in price). More formal discussions of this problem can be found in the

This sort of circumstance creates a temptation for the secured party to underbid at its own sale and then to resell for a substantial profit. Whether this result is called abusive or merely self-interested, the consequence may be that far less value will be received at an Article 9 collateral sale than would be achieved by a seller anxious to maximize that value. We may call the value that might have been obtained over and above the amount of the secured debt “excess value.”

In the context of debt enforcement, absent a general default, Article 9’s resolution of the competing values and risks is plausible, although still controversial .223 It can be argued that commercial debtors are able to protect themselves against the loss of excess value. Furthermore, other claimants against a debtor are not generally identified outside of a general default. When the discounted risks of creditor-controlled sales are balanced against the real gains from the flexible Article 9 sale provisions, it can be argued that the drafters have struck the right balance in the context of debt enforcement.224

Once a general default has occurred, however, the proper balance may be very different. In the recovery process following a general default, we find a debtor in general financial distress and a number of other claimants to any excess value. These considerations are central to the explanation for trumping the collateral-control rights in bankruptcy even as to a dominant secured party that could potentially obtain going-concern value .225 Protection of the other claimants requires that a neutral bankruptcy controller conduct the management of the assets and their redeployment or recapitalization. Where excess value can be obtained, the secured party would still receive its full priority payment, but there would be something left for the other claimants. By seizing control, the bankruptcy regime protects the secured party’s priority right, while also achieving the bankruptcy purpose of maximizing distribution for all the chosen beneficiaries.226 Among other things, this point clearly distinguishes the branch of financial-distress law concerned with the secured party’s enforcement of a particular debt227 from its enforcement in the context of

literature. One of the common explanations is the qualifications buyers must bring to liquidation auctions. In the case of the Simpson house, a buyer was required to have cash or a cashier’s check in an amount not less than the mortgage—about $2.6 million. Hall, supra note 221.

223. Because it is controversial, aspects of the problem as it affects consumer debtors have been left open for continued common law control. See, e.g., U.C.C. § 9-615.

224. We need not decide in this discussion if the balance struck is correct, only that it is arguable.

225. See supra section III(A)(2).

226. This discussion ignores any argument that maintenance of the secured party’s collateral control rights would produce transactional savings in excess of the combined transactional and situational losses resulting from the incentive problem. Given that the efficiency debate over secured credit has ignored the valuation of the control right, we have no estimate of savings against which to balance costs.

227. See supra subpart III(A).

general default. It is when a general default has occurred and the recovery process has begun that other claimants can be identified, implicating the collective purposes of bankruptcy.228

The incentive problem just described is merely one aspect of the larger problem of managing the recovery process in a way that protects the various interests chosen by Congress as beneficiaries under the Bankruptcy Code. Suppose, for example, that Congress adopted the automated bankruptcy scheme proposed by Professor Bebchuk.229 Although his proposal does not address the necessary interlude between general default and exercise of his various options up the ladder of priority, it is obvious that process would take some time. This is especially true in a large public company in the typical case where there is a desperate shortage of necessary information about the financial condition of the company, information crucial to the investment decisions about to be made at each tier.230

During this interim period, someone has to manage the company and its business. That management may profoundly affect the situation of the interested parties. A manager appointed by the first, or lowest, tier of investors may be likely to take some significant risks with the company’s assets,231 seeking an increase in asset value sufficient to cover the position of that lowest tier.232 Any manager who failed to take such risks would be very unlikely to receive another appointment from a bottom-tier creditor class. If the top tier appoints the manager, however, we may fairly predict that the

228. See supra Part IV.

229. See supra text accompanying notes 129–30.

230. The central importance of information and the difficulty of obtaining it, even from a public company, is captured in the phrase “due diligence” in the merger and acquisition practice. See, e.g.,

Bryan Burrough & John Helyar, BARBARIANS AT THE GATE: THE FALL OF RJR NABISCO 301–02,

368–69 (1990).

231. For example, assume a value in the enterprise of 100 units with five tiers of debt each owed 20 units. One form of management of the general default pending the option process would yield a certain value of 20 units but make it unlikely that more than 40 units of value would be preserved, while a second approach to management would give a reasonable chance of preserving 80 units of value but require a 10% chance of a complete loss. The top-tier creditors would prefer the first safe, low-value approach, while tiers three and four would prefer the second choice. Can such matters be efficiently resolved by contract, given problems of notice and the lessons of behavioral economics? These questions need serious discussion, which the model in this Article may provoke. Professors Baird and Rasmussen believe that the default should be managed by a creditor who is the agreed manager for all creditors. Douglas G. Baird & Robert K. Rasmussen, Four (or Five) Easy Lessons from Enron, 55 VAND. L. REV. 1787, 1808 (2002). Professor Lubben points out that many parties will have some right and some power to influence the direction of the firm at a given moment. Stephen J. Lubben, The Ambiguity of Control: Why the Third Lesson of Enron is Wrong (unpublished manuscript, on file with author).

232. This point was assumed at first to be the position with the DIP concept, where management would be protecting shareholders, the bottom tier of claimants in a corporation. However, the hoary, intractable agency problem so familiar in the corporate literature rose again here soon after the Code was adopted. Shareholders have been disappointed ever since. See, e.g., Baird & Rasmussen, End, supra note 6, at 780–85

the Reorganization of Large, Publicly Held Corporations, 78 CORNELL L. REV. 597, 610–11

(1993)

business will be managed into a coma of risk avoidance if that is necessary to protect the minimum remaining value that will satisfy the top tier of investors. Indeed, managers seeking top-tier appointments may actually aim for short-term value reduction which, after bankruptcy, might leave the top¬tier investors with an appreciating asset as the business cycle continues. Thus the incentive problem exists in any system in which there are competing claimants with different priorities or otherwise differing interests.233

The consequence of the incentive problem for secured creditors is that bankruptcy must trump collateral-control rights, whether or not going¬concern value is obtainable, because of the risk that a creditor-controlled sale will sacrifice excess value. Bankruptcy serves as a mechanism by which private parties can determine whether that trump will be invoked. That is, in the circumstance of general default, the debtor and other claimants hold an option to file bankruptcy and, thereby, to invoke its protection for all beneficiaries. It is plausible that the debtor and unsecured parties have the right set of incentives to act as bankruptcy gatekeepers. If they do not believe there is a general default or a reasonable chance to obtain excess value, over and above the secured party’s debt, they are unlikely to accept the costs and risks necessary to generate the collective trump represented by a bankruptcy petition.234 In those situations, they may conclude that the secured party might as well have the collateral control and conduct the sale.

On the other hand, where there is a general default and other claimants believe that the business may realize excess value—whether by sale or restructuring—they can invoke bankruptcy and ensure negation of the secured party’s collateral control. By giving the bankruptcy option to those most likely to suffer from the effects of the secured-creditor incentive problem, the law provides private parties with a choice between private, secured-creditor control and public, bankruptcy control.

Subject to the possible effects of a de facto bankruptcy veto,235 current United States bankruptcy law clearly follows the model proposed here— providing the option of the bankruptcy trump for the protection of beneficiaries other than a secured party. American bankruptcy law provides that secured creditors should not be allowed to sell collateral after bankruptcy is commenced, if the realizable value of the collateral is likely to exceed the

233. The differing interests may be economic, rather than resting on legal priorities. A bondholder and a supplier owed equal amounts may differ sharply. The bondholder may simply want the best tradable instrument available, looking to get as far away from this company as possible, while the supplier wants cash but also is looking for future business from the reorganized debtor. See Schwartz, Contract Theory, supra note 123, at 1838.

234. It is at this point in the analysis that the argument arises that debtors have too few disincentives to file bankruptcy, and, therefore, the choices are unbalanced—an argument that will have to be addressed in another article.

235. See supra subpart III(B).

secured debt, either in liquidation or reorganization. If there is such value, then, as the analysis would predict, the automatic stay cannot be lifted at all under § 362(d)(2)236 and will rarely be lifted for lack of adequate protection under § 362(d)(1).237 Unless the stay is lifted, the secured creditor’s collateral control remains negated. The point is driven home by the fact that all of the burdens in adequate protection litigation are on the trustee (DIP), except the burden to prove there is equity in the property.238 The presence of value in excess of the secured debt requires bankruptcy control to maximize returns for all beneficiaries. If the bankruptcy controller is prepared to exer¬cise that control, the secured party is required to show there is no excess value or to yield to the bankruptcy trump. The existence of the “bankruptcy veto” may defeat this statutory scheme as a practical matter because of a dominant secured party’s control of all of the resources of the estate,239 but the statutory scheme demonstrates the intent and purpose of existing law.

While the incentive problem is logical and intuitive, it has not been demonstrated empirically. Some of the empirical work done by Professor Mann may cast doubt on its importance as a practical matter. In a sample of seventy-two commercial loan files identified by the lender as “problem loans,” he found no evidence of secured creditors sacrificing excess value in the sale of collateral.240 Specifically, he found that most often debtors paid the loans from refinancing, sales of collateral, or continued business operations, thanks to the lender’s forbearance from enforcement.241 In only three cases did the lender sell the collateral, and it lost money in each of

236. That section requires that the collateral be surrendered to the secured party if there is no value above the secured debt (equity) and the collateral is not necessary to a reorganization.

237. The presence of equity works in a different way under § 362(d)(1). There, the excess value over the secured debt serves to provide the adequate protection necessary under that provision to block the lifting of the stay sought by the secured party. By its very existence, it protects against a decline in value below the level of the secured debt. See, e.g., In re Rogers Dev. Corp., 2 B.R. 679 (Bankr. E.D. Va. 1980) (refusing to grant relief from the automatic stay because the plaintiffs were adequately protected by an equity cushion that existed between the amount of the debtor’s obligation and the value of the collateral). There might be an exception if the equity is so small as to make loss of adequate protection likely over the relevant period.

238. 11 U.S.C. § 362(g) (allocating the burden of proof on the issue of the debtor’s equity in the property to the party requesting relief and allocating the burden of proof on all other issues to the party opposing such relief).

239. See supra subpart III(B). A reader might think this interpretation of the statutory scheme is inconsistent with the assertion that a dominant secured party may have a bankruptcy veto. It is not, for two reasons. First, the statute applies to both ordinary secured parties and dominant secured parties. As to ordinary secured parties, who do not have a bankruptcy veto, the rule operates invariably to protect possible equity in the collateral. Second, as to dominant secured parties, the bankruptcy veto may evade the application of the rule because a dominant secured party can prevent the funding of a bankruptcy and therefore the application of the statute. See supra text accompanying note 72. However, if a principal of the debtor or a new investor can fund the expenses of a reorganization, then the dominant secured party will also be subject to the equity¬protecting effect of § 362.

240. Mann, Strategy, supra note 4, at 163–64.

241. Id. at 164.

those cases,242 suggesting there was no excess value to forfeit in those instances. Mann’s interviews suggested that the reason for so few enforcement actions was the loan officers’ conviction that seizing collateral meant a nearly certain loss.243

Mann’s work is very helpful, but it is well short of showing that the incentive problem does not exist. As he concedes, the size and nature of his sample makes the data suggestive at most .244 He was deliberately sampling the “middle” case of loans—problematic but not in bankruptcy or foreclosure.245 Although many of the debtors refinanced their loans with “his” lenders, his methodology did not provide for following those debtors after the refinance—so we do not know their ultimate fate or the fate of their assets. His work strongly supports the conclusion that most problem loans work out and that formal legal actions, whether under Part 6 of Article 9 or in bankruptcy, are relatively rare. It does not tell us to what extent the incentive problem should be a serious concern in the cases that do go to legal action. His conclusion rests strongly on the belief of loan officers that liquidations mean loss.246 That conclusion is consistent with the literature as to the effect of forced sales,247 but it does not tell us what happens in the universe of cases that go into bankruptcy, especially Chapter 11 reorganization

What results obtain in a bankruptcy, especially a reorganization bankruptcy, remains for future empirical report. It may be, for example, that Mann’s refinancing debtors were the ones that had excess value over and above their existing secured loans and that is why they were able to refinance. On the other hand, it would generally be conceded that bankruptcies are fairly often filed where the lenders are oversecured—that is, where there is value in the collateral in excess of the secured debt

242. Id. at 221 –22 (stating that the results from the lender’s sale of collateral were “disastrous” in that the “lender’s proceeds on resale of the properties left [it] losing more than seventy-five percent of the original loan amounts”).

243. Id. at 230 (reporting that “[t]he substantial losses my lenders see coming upon repossession pose a powerful deterrent to repossession as a common strategy”).

244. Id. at 226 (conceding that his “case studies present only a limited amount of direct

evidence of the results of foreclosures, and that evidence is limited to real estate foreclosures”).

245. Indeed, many of the files were identified by the lenders as “problem loans,” but they were

not necessarily in default at all. Id. at 172, 199 –201.

246. Id. at 179.

247. See, e.g., Mann, Strategy, supra note 4, at 221.

248. Id. There are a number of other variables. For example, to what extent is there a going¬concern market for a particular kind of business? Absent such a market, the likelihood of a loss upon liquidation will be quite high.

joined by substantial anecdotal evidence249 to suggest that the incentive problem is a serious one in a significant number of cases that require legal intervention. The fact that the incentive problem was perceived as operative and harmful in Britain250 offers some empirical proof as well.

As explained earlier,251 neutrality is a necessary concept in any system for managing a general default in which the policymaker provides for multiple beneficiaries and charges the manager with maximizing value for all of them. A dominant secured party cannot be a neutral manager, and its management creates a serious potential of loss for other beneficiaries. It is just that result that provided the impetus for a major restructuring of the insolvency system in Britain.252

VII. Empirical Evidence

A. Abandonment of Administrative Receivership in Britain

A further blow for contractualists is the fact that a long-established, functioning system very near to secured contractualism has been abolished in one of the most successful commercial societies in the world.253 The British

249. See supra note 210 and accompanying text.

250. See infra subpart VII(A).

251. See supra text accompanying note 115.

252. For this sort of criticism of the floating-charge regime in the United Kingdom, see Mokal, The Floating Charge, supra note 49, 496 n.70, 497 n.80–82 and accompanying text. Mokal also criticizes the system for excessive costs. Id. at 497 n.75. On the other hand, he believes the great power given to the secured creditor to oust management even on a merely technical default is a benign feature retained by the new British system, a viewpoint that many would not share. See, e.g., Finch, Re-Invigorating, supra note 14, at n.31.

253. Enterprise Act, 2002, §§ 72A–72H, c. 40, (Eng.). Andrew McKnight, The Reform of

Corporate Insolvency Law in Great Britain, 17 J. INT’L BANKING L. 324, 324, 332–33 (2002)

(explaining that § 72A of the 2002 Enterprise Act terminated administrative receivership). The statement in the text must be qualified to the extent that there are exclusions from the new regime that permit administrative receivership to continue for certain types of transactions and that grandfather existing loans. Enterprise Act §§ 72B–72G

The effect of those provisions is not yet clear, although it may be to defeat the government’s hopes. See Finch, Re-Invigorating, supra note 14, at 536 (suggesting that “the regime creates a number of legal uncertainties and areas of potential court challenge,” and arguing that these uncertainties could work against the goals of the new regime). On the other hand, the concept of administrator responsibility to all creditors represents a fundamental change in attitude. Id. at 534 (noting that the Enterprise Act “produces a significant alteration in the substantive rights” of creditors since “the administrator has a duty to act in the interests of the creditors as a whole”).

Furthermore, British abolition of administrative receivership (i.e., private receivers) is consistent with a Commonwealth trend. Our friends in Canada relegated their English-style receivership system to a backwater twenty years ago, even though their commercial law remains very protective

of secured creditors. ALI, CANADIAN STATEMENT, supra note 21, § I.B.3–5. Australian law seems

to be moving away from secured-creditor control as well, although it has not yet gone as far as the new Enterprise Act in Britain. See G. Dal Pont & L. Griggs, The Resuscitation of the Corporate Cadaver: An Autopsy of Business Rescue Laws, 4 AUSTL. J. CORP. L. 309, 330–31 (1994)

system described earlier254 is very similar to a contractualist system. In effect, it permits liquidation or going-concern sale through an entirely private system of managing the recovery process. It is in almost every respect what the contractualist theorists propose—with the addition of the dominant security interest this Article has shown is necessary for the success of their project. Yet Britain has adopted legislation designed to abolish this system or radically move it in the direction of the sort of reorganization procedures found in Chapter 11 and the other modern reorganization regimes emerging around the world. This change has been called “a seismic shift for the country that invented . . . privately appointed receivers.”255 From the perspective of English lenders, proposals to abolish motherhood and the flag would pale in comparison.256 The reforms may have the effect of profoundly altering the commercial lending market in Britain. They overturn over a century of well-established British law. Obviously, the perceived need for reform was great.

The reforms, which came into effect during 2003,257 prohibit the appointment of an “administrative receiver” out of court under a debenture— the power that was central to the position of a floating-charge holder throughout the twentieth century.258 Instead, an “administrator” will be

Anderson, The Australian Corporate Rescue Regime: Bold Experiment or Sensible Policy?, 10 INT’L INSOLVENCY REV. 81, 84–85 (2001) (both discussing Australian bankruptcy procedures).

254. See supra subpart III(C).

255. E. Bruce Leonard & Eric Kupka, The Coming Revolution in U.K. Insolvency Law: Part I, AM. BANKR. INST. J., June 2003, at 26.

256. The abolition came less than twenty years after a very prestigious and influential report

had flatly stated, “we accept that the floating charge has become so fundamental a part of the

financial structure . . . of the United Kingdom . . . that its abolition can no longer be contemplated.”

Cork Report, supra note 77, at 34. The centrality of the floating charge in English lending practice

was also acknowledged recently by the Privy Council. Agnew v. Comm’r of Inland Revenue,

[2001] 2 A.C. 710, 717–18 (P.C. 2001) (appeal taken from N.Z.). There, Lord Millett said: The floating charge is capable of affording the creditor, by a single instrument, an effective and comprehensive security upon the entire undertaking of the debtor company and its assets from time to time, while at the same time leaving the company free to deal with its assets and pay its trade creditors in the ordinary course of business without reference to the holder of the charge. Such a form of security is particularly attractive to banks, and it rapidly acquired an importance in English commercial life which . . . should not be underestimated.

Id. Yet the central provision of the new legislation generally prohibits debentureholders (holders of floating charges) from appointing a receiver (an “administrative receiver”)—subject to certain exceptions. See, e.g., Finch, Re-Invigorating, supra note 14, at 531 n.22 and accompanying text (discussing the Act’s prohibition on the use of administrative receiverships by holders of floating charges and listing six exceptions). As noted earlier, the power to appoint a private administrative receiver has long been seen as the essence of the floating charge. See supra subpart II(C).

257. Leonard & Kupka, supra note 255, at 26.

258. See McKnight, supra note 253 (noting that a prohibition on the appointment of administrative receivers “will be a far cry from the favourable regime hitherto enjoyed by those secured creditors who have been able to appoint administrative receivers and thereby prevent the making of an administration order”).

appointed.259 The appointment may be by the court, by the company, or by the debentureholder, but the administrator’s responsibility will be to the court.260 The administrator is explicitly charged with attending to the interests of all creditors and with seeking a restructuring of the company, rather than a liquidation .261 These provisions work a conceptual revolution in British financial law.

The reasons for this radical change included a conviction that secured creditors acting in their own interests were failing to permit or encourage reorganization (in Britain, “rescue”) of viable businesses—an important aspect of the incentive problem.262 Specifically, a conviction arose that the balance of financial law in Britain was skewed too far in favor of the lender and against the entrepreneur, discouraging the development of an entrepreneurial culture like that in the United States.263

The abandonment of an essentially contractualist regime by an advanced and sophisticated commercial society, especially on the stated grounds, is a devastating commentary on the social value of the contractualist project—especially when the system seems to be moving toward abandonment or serious decline in other developed countries in the Commonwealth. Secured credit may have its uses, but in Britain it has come to be perceived as materially suboptimal when used to support a system that, in the midst of the economic crisis following a general financial default, hands over complete control of the recovery process to a secured creditor.

B. Empirical Work

The establishment of the fact that privatization of the management of general defaults must rest upon a dominant security interest generates a considerable empirical agenda. The top item must be a further exploration of the British experience, which has been a “natural experiment” for the ideas of the contractualists. The contractualists must start with the hypothesis that the British were wrong to think that the receivership system had serious flaws.

259. Enterprise Act 2002 (Eng.)

260. Enterprise Act 2002 (Eng.)

261. Enterprise Act 2002, sched. B 1, para. 3 (Eng.)

262. THE INSOLVENCY SERV., DEP’T OF TRADE AND INDUS., INSOLVENCY—A SECOND

CHANCE §§ 2.1–2.3, C.16 (2001) [hereinafter DTI WHITE PAPER]

263. DTI, WHITE PAPER, supra note 262, at 1.

In the United States, an important object of inquiry would be actual experience with management of general defaults by secured creditors. That would require an opportunity to examine creditors’ files systematically, with all the attendant difficulties, but would follow a path already blazed by Professor Mann. It might also be revealing to study bankruptcy files to compare recoveries by general creditors in cases with or without secured parties, both dominant and ordinary. Yet another significant question is the frequency and extent of the exercise of the bankruptcy veto and the results of dominant secured party management.

Because of the difficulties with these types of inquiries, however, these issues are among those that might be illuminated by simulations or “gaming.” Both game theory and actual game experiments, using players given an appropriate set of rules and rewards, might suggest the existence and strength of various incentives, both positive and perverse, that are difficult to study in the wild.

VIII. Summary and Implications of the Control Model

By establishing the fundamental elements of a model that links secured credit and bankruptcy and identifies control as the central issue in the law governing the recovery process for distressed businesses, this Article has explained why control is the governing issue and has begun the process of identifying the key implications of that finding. It has also introduced the concept of neutrality and its relationship to the case for public management of general defaults.

A bankruptcy regime exists to enforce a set of priorities chosen by legislators. A study of bankruptcy laws around the world reveals that no particular set of priority choices, including a choice of general equality, is essential to bankruptcy, but that the control necessary to enforce the chosen priorities is essential. Another way to state that conclusion is that most priority decisions are exogenous to bankruptcy law,264 arising from efficiency and distributional goals and ideas of fairness that are external to bankruptcy

264. The major exceptions to the statement in the text are administration, marginality, and entrepreneurship. Administration is the necessary priority for the operation of the bankruptcy process itself. See 11 U.S.C. §§ 503(b), 507(a)(1) (prioritizing administrative expenses above unsecured claims). Marginality is the idea that the bankruptcy-regime benefits of a policy decision may be so marginal when compared to its bankruptcy-regime costs that the policy would be better vindicated in a larger context. Thus a demonstration of marginality might be persuasive in defeating a proposed vindication of some social or economic policy through the bankruptcy process. See, e.g., Manfred Baltz, Market Conformity of Insolvency Proceedings: Policy Issues of the German Insolvency Law, 23 BROOK. J. INT’L L. 167, 174 & n.34 (1997). The recent German decision to abolish employee priorities in bankruptcy was driven by this sort of argument. The workers are protected by an unemployment fund instead. Id. The third exception, encouragement of entrepreneurship, is an independent economic policy, but it is so closely related to risk allocation and other factors affected by bankruptcy law that it represents a combination of exogenous and endogenous factors.

policy as such.265 Control decisions, by contrast, are central to bankruptcy policy.266

The struggle for control of the recovery process is ultimately the struggle between a public and a private ordering. The primary approach to privatization is contractualism. This discussion has shown that the only plausible form of contractualism is secured contractualism. This is because contractualism requires control of a debtor’s assets, not merely priority in the proceeds of their sale, and only secured-credit law provides both control and priority outside of bankruptcy. We have looked more deeply to see that con¬tractualism requires both pre-default constraint and post-default control. Further, it requires complete constraint and control which can be provided only by a dominant security interest. Thus, the constraints and controls provided by a dominant security interest are essential to a privatized, contractual bankruptcy regime.

Control is the intersection and the link between the major components of the recovery process: secured-credit law and bankruptcy law. Although the elaboration of that proposition must await another day, the centrality of control to the question of contractualism—the most discussed reform program in the field in recent years—is the key step. Because control is central to realization of maximum value for the beneficiaries chosen by Congress, and because there will be conflicts of interest among beneficiaries as to the best management for maximization of each beneficiary’s interest, management by any one creditor or creditor interest will often be inconsistent with the congressional scheme. If the law establishes multiple beneficiaries, only a neutral manager will maximize value appropriately. Necessarily, any attempt to modify priorities or control by contract must therefore be incon¬sistent with those policy choices. On the other hand, where policy choices and specific circumstances result in only one class of beneficiaries—as with an undersecured dominant secured party—then control by that class may be appropriate.267

265. This point, along with marginality, has contributed to the conventional, but mistaken, idea that nonbankruptcy law should determine all of the policy content of bankruptcy law. See, e.g., Thomas H. Jackson, Bankruptcy, Nonbankruptcy Entitlements, and the Creditors’ Bargain, 91 YALE L.J. 857, 858 (1982) (stating that bankruptcy law simply prioritizes creditors who are entitled to superior rights originating in nonbankruptcy law)

266. The decision about the role of management and equity in a reorganizing company lies at the crossroads between control and priority which is no doubt one reason it has been so difficult and controversial. That decision is related to the function and importance assigned to venture capital and entrepreneurs. See DTI, WHITE PAPER, supra note 262, § 1.1 (advocating a system that encourages entrepreneurs to take reasonable risks by reducing the stigma of failure).

267. Another important issue that arises from the analysis discussed here is the propriety of the use of a collective, publicly supported proceeding—bankruptcy—for the benefit of a single secured creditor who arguably should support the burdens of the recovery process by itself when no other

Although secured contractualism has not been discussed by the contractualists, it is clear that two central issues that must be addressed: (1) the transactional efficiency of secured credit

The control model carries a number of implications for the theory of the law governing financial defaults. Their development will have to await other articles, but it may be useful to discuss the relationship of one concept to the control model, albeit in a preliminary way. The concept is “lender control.”

In two recent articles,269 Professors Baird and Rasmussen have advanced our understanding by putting forward control as a central issue in bankruptcy law. Although they have not named the phenomenon they have identified, to call it “lender control” will be close to the mark.270 They assert that lenders have recently come to dominate the Chapter 11 cases of large public companies and that this development should be applauded.271

person or social interest will benefit from a bankruptcy proceeding. Elizabeth Warren & Jay L.

Westbrook, Secured Parties in Possession, 22 AM. BANKR. INST. J., Sept. 2003, at 12.

268. Secured contractualism also raises important questions about bankruptcy policy as to entrepreneurial activity and venture capital investments in an economic system that depends upon them, but we put those points to one side for now.

269. Baird & Rasmussen, End, supra note 6

270. They are quite vague about just who will be in control and how that determination will be made. Baird & Rasmussen, End, supra note 6, at 777–85. Aside from an interesting discussion of management control when the manager is uniquely essential to the business, they generally refer to “creditors” in control, but without much specificity. Id. Nonetheless, they suggest that an “institutional lender” or investor may somehow take charge. Baird & Rasmussen, Control Rights, supra note 6, at 957. The End of Bankruptcy is almost entirely about lenders taking control, so it seems correct to take that as their paradigm. See Baird & Rasmussen, End, supra note 6.

271. See id. at 752, 784–85 (noting that the authors “are not troubled by such a shift in bankruptcy practice” because senior lenders are better situated to exercise control over the bankrupt company than either a manager or a bankruptcy judge). They also suggest that bankruptcy reorganization is no longer very useful for smaller, nonpublic companies and that therefore the only nonpublic companies using Chapter 11 are largely marginal, oddball outfits with some specific, perhaps shady, use for bankruptcy—like dodging the IRS. See id. at 752–53. Their only published support for this claim is that the number of Chapter 11 filings in 2000, at the end of the 1990s bubble, was only half as great as the number in 1991, in the midst of a recession. Id. at 752. They fail to note that the total number of business filings in 2000 was also about half as large as in 1991,

not just the number of Chapter 11 filings. AM. BANKR. INST., U.S. BANKRUPTCY FILINGS 1980–

2002, at http://www.abiworld.orgstats/1980annual.html (last visited Nov. 23, 2003). Yet they do not suggest bankruptcy in general has become obsolete. See, e.g., Baird & Rasmussen, End, supra note 6, at 788. This Article is not the place to respond to their larger claims about fundamental

Although their approach is stimulating and helpful, their description of the phenomenon presents some serious difficulties. Its empirical deficiencies have been well-addressed by Professor LoPucki,272 and it presents some important conceptual problems as well. The discussion here will assume without conceding that their factual assertion is correct.273 It will be limited to two points: (1) comparing and contrasting control by a dominant secured party with the type of lender control they discuss

In one important respect, the Baird and Rasmussen’s analysis is supportive of the model presented here. They see security interests as an important aspect of lender control in many cases.274 They do not distinguish, however, between ordinary and dominant security interests or between pre¬default and post-default control. As to the first, one key question implied by the analysis in this Article is whether control by an ordinary secured party is legitimate. For example, is it appropriate—economically or legally—for a secured party with a security interest only in accounts receivable or only in equipment to attempt to hold reorganization hostage to its demands—even though it is not a dominant secured party and therefore is incapable of realizing going-concern value itself? This power might be called “hostage value,” but it seems very different from the concept put forward under that name by Professor Scott.275 “Coercion” might be closer. This Article stops short of addressing this interesting question which is answered firmly in the negative by Chapter 11 of the Bankruptcy Code.276 Given that Baird and Rasmussen do not make the ordinary-dominant distinction, it is not surprising that they do not address this question either.

Indeed, these two authors seem unsure whether the lenders whose control they celebrate are secured or not. They have difficulty identifying the “creditors” that will take charge. They flirt with the idea of a secured creditor playing that role277 but never settle on it—although many of their exemplary cases involve security interests.278 In short, their argument for a system in which “investors” contract for “control rights” comes to the edge

changes in the nature and structure of American business that make Chapter 11 bankruptcy irrelevant.

272. LoPucki, Nature, supra note 52

273. The present author is inclined to think they are correct in this particular regard as to a portion of the major bankruptcy cases—although perhaps a smaller percentage than they assert.

274. Baird & Rasmussen, End, supra note 6, at 784–85.

275. See supra note 193 and accompanying text.

276. A secured party may be “crammed down” by a Chapter 11 debtor’s plan as long as it receives the value of its collateral plus interest over time. 11 U.S.C. § 1129(b)(2)(A). (There is a twist in the “1111(b) election,” but it is not relevant to this point.) Thus, an ordinary secured party cannot hold a reorganization hostage by refusing to agree to a plan. A dominant secured party might be able to do so where it can exercise a bankruptcy veto. See supra subpart III(B).

277. Baird & Rasmussen, Control Rights, supra note 6, at 975.

278. Baird & Rasmussen, End, supra note 6, at 784–85.

of presenting a model based on a dominant security interest, but it stops short of making any connection between bankruptcy and secured credit.

Baird and Rasmussen do not ignore the pre-default period, but they offer evidence as to control only during the recovery process.279 As to the period beginning with the initiation of the lending relationship, they merely assert that “investors” have established methods of pre-default control, without offering any evidence other than a citation to two articles that are purely theoretical.280 It does not seem that any real company has yet adopted the financial structures suggested by either of the cited articles. Of course, a dominant security interest would provide the desired control, but they do not propose it as their model. Indeed, they do not propose any method for solving the problems of contractualist lack of control discussed in Part V of this Article because they do not recognize those problems.

The only pre-bankruptcy control that is discussed in the The End of Bankruptcy seems to refer to post-distress control —that is, control that lenders obtain prior to bankruptcy but after the debtor has fallen into serious financial trouble.281 The authors may well have in mind that the lenders obtain dominant security interests in that context, although they do not say so. To the extent they are not relying on a dominant security interest, the problem presented shades into that presented by post-bankruptcy control, to which we now turn.

According to The End of Bankruptcy, one important way lenders have seized post-default control is through the post-bankruptcy lending process.282 This suggestion raises both empirical and normative issues. Why and how has this happened? Is it a desirable development? Baird and Rasmussen devote little attention to either issue.

Absent a dominant security interest, it is not clear why post-bankruptcy lenders (usually called “DIP lenders”) should have great leverage in a Chapter 11 bankruptcy—at least at the outset of the case. Odd as it seems to many non-Americans, DIP lending is highly sought after and competitive in the United States.283 There is no need, ordinarily, to go hat in hand to existing lenders to beg for more money. So why would the debtor-in¬possession (DIP) give control to the DIP lenders, as Baird and Rasmussen say they do? If the authors are correct in their factual claims, it is clear that a serious empirical inquiry —by questionnaire, data compilation, or

279. Id.

280. Id. at 778 n.125 (citing Adler, A Theory of Corporate Insolvency, supra note 144, at 345, 367–75

281. Id. at 784–85.

282. Id.

283. See James G. Connolly, DIP Financing: New Life for Ailing Companies, FIN. EXECUTIVE,

May 2003, at 31 (observing that the issuance of DIP loans tripled from 2001 to 2002)

Triantis, A Theory of the Regulation of Debtor-in-Possession Financing, 46 VAND. L. REV. 901,

901–03 (1993) (noting the rapid expansion of DIP lending in the United States and the emergence of a competitive debt market in this area).

otherwise—should be undertaken to answer this question. Could it be that management, which controls the DIP, sometimes abandons the equity owners and other company constituencies in favor of lenders who tacitly agree to maintain management in control? That result would be somewhat consistent with the positions of those who claim that management should have an ex¬clusive duty to creditors once the business is “in the vicinity” of insolvency284 but would seem highly problematic to those who disagree.285 One can imagine a number of motives and pathways by which lender control through management might be reached, but hard evidence would be much better than speculation. In any event, Baird and Rasmussen do not explain.

The normative issue raised by the claims made in The End of Bankruptcy is whether it is appropriate for the DIP, through management, to be controlled by the “lenders.” The authors exhibit the tendency found in much of the commercial-law literature to confuse “lenders” with creditors— when the creditor body in fact consists of many classes of creditors, often including classes of lenders conflicted inter se.286 By assuming some lender or lenders represents all creditors, they can assume that control by lenders is the same as control by creditors—which they applaud. In fact, control by a group of lenders is highly unlikely to be neutral as among creditors—much less as among a broader range of company constituencies287—and, therefore, it is unlikely to serve the congressional purposes for the reasons explained in Part VI. On the analysis presented here, a takeover of the Chapter 11 process by one group of creditors would seem to be the occasion for concern, not celebration.288

284. See, e.g., Credit Lyonnais Bank Nederland, N.V. v. Pathe Communications Corp., No. 12150, 1991 WL 277613, at *34 (Del. Ch. Dec. 30, 1991) (stating that “[a]t least where a corporation is operating in the vicinity of insolvency, a board of directors is not merely the agent of the residue risk bearers, but owes its duty to the corporate enterprise”)

Duty of Directors to Take Account of Creditors’ Interests: Has It Any Role to Play?, J. BUS. L., July

2002, at 385. It would be only somewhat consistent because the process would not necessarily be limited to companies “in the vicinity” of insolvency, putting to one side the very difficult problem of defining and identifying that condition. Lipson, supra note 9. Furthermore, Credit Lyonnais commended a duty to all creditors, not just lenders. See Credit Lyonnais, 1991 WL 277613, at *34 n.55 (stating that directors of corporations in the vicinity of bankruptcy, in choosing the most efficient and fair course for the company, have to consider the “community of interests the corporation represents”—stockholders, creditors, employees, or “any single group interested in the corporation”).

285. See, e.g., Lipson, supra note 9

286. See generally Stephen J. Lubben, The Direct Costs of Corporate Reorganization: An Empirical Examination of Professional Fees in Large Chapter 11 Cases, 74 AM. BANKR. L.J. 509

(2000) [hereinafter Lubben, Costs]. A recent working paper by an Israeli scholar suggests a “co¬determination” governance scheme for countries that have concentrated ownership structures in public corporations generally. See Hahn, supra note 6.

287. See Warren, Policymaking, supra note 96, at 353–54 (discussing how a bankruptcy action must address the competing interests of diverse parties).

288. There is a blunt suggestion in Professor Schwartz’s work that management should be bribed to be faithless to shareholders. Schwartz, Contract Theory, supra note 123, at 1827 & n.58.

IX. Conclusion

Although the literature in the field of commercial finance and financial default has long been preoccupied with questions of priority, control is the central concept in any persuasive model of the field. A lack of understanding of the role of control explains the failure to recognize and analyze the crucial distinction between an ordinary secured party and a dominant secured party and to see that the latter offers a possible alternative to the bankruptcy trustee. The blurring of the types of control generated by these two different types of security interest and confusion between pre-default and post-default control has prevented development of the important insights provided by Professors Scott and Mann with regard to control as a central value¬generating element in pricing secured transactions and in measuring their economic efficiency. A focus on control as the fundamental concept in the law of default, linking secured-credit law and bankruptcy law at their roots, will—among other things—force recognition of priority issues as largely exogenous to bankruptcy law, while causing neutrality issues to rise to the top. Most immediately, as shown above, an analysis based on control shows that a dominant security interest is the sine qua non of contractualism. This structural relationship has been ignored by both contractualists and their critics, but its exposure has serious consequences for the contractualist enterprise.

With regard to bankruptcy reform, the analysis presented here suggests that contractualism or some other entirely private ordering of the recovery process may be a dead end, while the alleged trend towards lender control of that process may require new legislation to ensure neutrality.289 Those conclusions are supported by the fact that there is considerable movement around the world in the general direction of existing United States secured¬credit law and bankruptcy law. Conversely, the Canadian and British experiences demonstrate, other developed nations are experimenting with significantly different versions of bankruptcy law that may well offer guidance to American policymakers—but that is another article for another day.

Yet if management accepts a bribe from lenders, there is no reason to believe it will not be faithless to non-lender creditors as well as shareholders. That point can be ignored by assuming that all creditors are lenders, that there are no conflicts among creditors, or that the interests of other creditors do not matter. The first two assumptions are demonstrably false. See Lubben, Costs, supra note 286, at 514–22. The last assumption is not only inconsistent with the congressional scheme, but is unsupported by any coherent argument that our economy would function better if it were adopted. Therefore, there is every reason to believe that faithless management would betray creditors whose interests conflicted with those of its paymasters. For an insightful discussion of the evolving role of management in Chapter 11, see Ethan Bernstein, All’s Fair in Love, War & Bankruptcy (unpublished manuscript, on file with author).

289. Yet another aspect of this emerging trend is the tendency to use Chapter 11 as a liquidation chapter in preference to Chapter 7, with no rules or principles to determine when avoidance of the statutory liquidation scheme is appropriate.

Two forward-looking thoughts conclude the discussion. First, one motive for the work leading to this Article was an attempt to understand the strange separation between secured-credit law and theory and bankruptcy law and theory. The decade-long debate over contractualism without a serious discussion of any connection between it and secured credit is the defining example of that separation. A second myopia, the absence of serious reference to the English system of secured credit and receivership in the great debates over secured credit and over contractualism, is another remarkable example of the effect of falsely dichotomous thinking about secured credit and bankruptcy. As a result, important insights into the issues of control before and after default have been left unconnected, like random pages from the blueprints for an intricate machine.

The disregard of the obviously relevant English system of secured credit is also a striking instance of the lack of comparative law study in this and many other legal fields.290 If there are any useful thoughts in this Article, they arise for the most part from the perspectives provided by some years of comparative study of secured-credit and bankruptcy law. During the last decade, many scholars from the United States have traveled abroad and talked with foreign lawyers and scholars, yet the influence of comparative study on American legal scholarship remains relatively slight. It is wonderful that we can go elsewhere to teach, but it is even more wonderful to learn. And we have much to learn.

290. The void is not merely an absence of comparative articles and books, valuable as they are, but a lack of study of other laws in relation to our own, a study that would greatly affect scholars’ understanding of our legal institutions.

Articles

The Control of Wealth in Bankruptcy

Jay Lawrence Westbrook*

The fixed charge for priority

Nearly every country in the world has a legal regime devoted to the process of recovery and distribution of value following a general default by a debtor in business. In the United States, we have both a public and a private regime that may govern the recovery process. The public regime is bankruptcy and the private regime is secured credit. Any regime governing the recovery process must include two components: (1) control of the process that maximizes value and allocates that value to chosen beneficiaries

* Benno C. Schmidt Chair of Business Law, The University of Texas School of Law. Many of the ideas in this Article arose from my work as a member of the World Bank’s Insolvency Initiative Task Force to develop international principles for benchmarking the effectiveness of insolvency and creditor rights systems and as a coordinating author of a related academic study on comparative insolvency practices under preparation with the World Bank. The insolvency principles project is under the direction of Gordon Johnson, World Bank Lead Counsel, with whom I have had many enlightening discussions touching on the issues analyzed here. I have learned a vast amount from my co-authors on that project, Professors Charles Booth, Christoph Paulus, and Harry Rajak. I am grateful to Edward Janger, Douglas Laycock, Ronald Mann, Lynn LoPucki, and Elizabeth Warren for comments on versions of this Article. I am also in debt to those attending the Fall 2001 Moller Workshop at The University of Texas School of Law for their comments on the initial concept and to suggestions made at a presentation of an earlier version of this Article at Brooklyn Law School in February of 2003. Finally, I much appreciate the assistance I received on this Article from Ludmila Yamalova, Texas ’01, Maravillas Oviedo, Texas LL.M. ’02, Derrick Talerico and Mitchell Mills, Texas ’03, and Tonya Shotwell and Benjamin Ellison, Texas ’04. None of these talented people is responsible for its remaining faults.

linkage between them. A number of consequences follow from these new models. Development of these consequences is the author ‘s larger project, but only one of them is fully explored in this Article.

“Contractualism ” is a term describing a variety of proposals for privatizing the recovery process through contracts between a debtor and its creditors. This Article argues that these proposals cannot be legitimate or eficient unless linked to a dominant security interest encumbering substantially all of the debtor’s assets, even though none of the scholars proposing contractual regimes has acknowledged that the contractual schemes require security to be plausible. Because secured-credit law in the United States is in place and highly functional, it represents the only method for resolving the otherwise intractable dificulties presented by these privatization proposals. That conclusion is confirmed by the United Kingdom ‘s experience with a contractualist system based upon a dominant security interest, a system that has served as the core structure of commercial finance in that country for over a century.

Recognition that secured credit is the sina qua non for contractualism presents some serious dificulties for its proponents, including underdeveloped claims of eficiency and substantial conflicts of interest in secured -party management of the recovery process. The abolition of the British system in 2003 demonstrates empirically the serious weaknesses of secured contractualism. This Article concludes that the case for secured contractualism cannot be made.

The control models give rise to a number of consequences mentioned here but left for subsequent development, including: (1) the need to introduce valuation of control rights, pre- and post-default, to the long-standing debate about the eficiency of secured credit—a point closely tied to the specific debate about securitization of assets

I. Introduction 797

II. Overview: Control is the Functioning

III. The Secured Credit Model 806