[Pg. 213]

Three Faces of Creditor-on-Creditor Aggression

Douglas G. Baird †

Creditor-on-creditor aggression has become the central pathology of modern Chapter 11 reorganizations. Distressed debt investors operate under few external constraints and are most unlikely to be either Good Samaritans or Boy Scouts. Instead of confining themselves to vindicating their rights against their debtor, distressed debt investors position themselves to extract hundreds of millions of dollars from their fellow creditors.[1] Beyond the constraints of the loan contract itself, the law does little to prevent these investors from doing this. Once they find a lacuna in the language of the loan documents or the Bankruptcy Code, those bent on capturing value from their fellow creditors can insulate themselves from liability as long as they muster some evidence that they are acting in good faith.

It is possible to confront this problem by imposing more stringent good faith duties on creditors.[2] Creditor advantage-taking, however, manifests itself in modern corporate restructurings in various ways. Different strategic forces are at work, and each presents its own challenges.[3] Moreover, one [Pg. 214]can contest the amount of legal reform that is necessary. Even though the sums involved are large and the conduct is unsavory, the social costs may be small, and intervention may make things worse. In this paper, I stop short of any general solution and offer instead some landmarks to help navigate legal reform.

Part I of the paper establishes the relevant background. Much of the legal backdrop was put in place by New Deal reforms. These focused on protecting the rights of public bondholders, and as has long been recognized, they provide an awkward starting place. Many of these legal reforms were misguided and addressed an altogether different problem.[4] They focused on the use of divide-and-conquer strategies against diverse creditors who were unable to organize themselves. The inability of creditors to organize themselves is not at the center of the current storm. Indeed, the problem is the opposite. Creditors, far from being unable to organize themselves, form factions, and each faction presses for advantage at the expense of others. The creditors’ ability to organize rather than their inability to organize creates the central difficulty.

In Part II, the paper focuses on the aggression that takes place between creditors outside of bankruptcy. It is tempting to think one creditor’s appropriation of wealth from another is itself the principal problem. This is a mistake, however. The risk of being subject to such a transfer is one that sophisticated professionals can anticipate and fully price ex ante. The costs of such behavior are to be found elsewhere. Most obviously, the prospect of capturing such large sums itself invites rent-seeking. The risk of others engaging in rent-seeking in turn inspires others to take measures to guard against it. Each hedge fund has an incentive to spend to ensure it comes out ahead or at least does not come out behind.

Reducing such costs is desirable, but doing so may come at a cost. The transactions themselves may value-enhancing as well as value-reducing. To be sure, if the wealth transfer takes place in a transaction in which the debtor is merely playing for time and putting off a day of reckoning, allowing the parties to profit from such transactions encourages social waste. Limiting creditor-on-creditor aggression increases the value of the firm at the same time it eliminates rent-seeking. On the other hand, however, creditor-on-creditor aggression may be associated with transactions that preserve the value of the firm. Activist creditors may capture wealth from passive creditors at the same time they establish new sources of finance that enable the debtor to avoid a costly liquidation.

[Pg. 215]Academics are poorly positioned to distinguish between transactions that are socially wasteful and those that bring net benefits. By contrast, parties draft their contracts in the shadow of these problems and empirical uncertainties. It is a world in which contracts are dynamically updated as conditions change. Parties can anticipate the battles that take place when the debtor becomes distressed. Moreover, those that lack either the skill or the appetite for such contests can sell their positions to those that do.

Legal interventions are needed most to protect the structural integrity of the contractual mechanisms the parties put in place. Rather than provide creditors with substantive legal protections, the law should ensure that they can enforce their contracts. If, for example, they bargain for the right to vote before any of their rights against the debtor are altered, the law should ensure that every vote is counted and that no one is stuffing the ballot box.[5]

Part III shows that an altogether different problem presents itself in bankruptcy. When a bankruptcy petition is filed, all the rights of creditors are collapsed into a “claim.”[6] The specific attributes of the debt under the contract disappear, including rights creditors secure to protect themselves against other creditors. Creditors must rely on the legal protections built into the Bankruptcy Code. These protections consist of straightforward and sensible rules, but clear rules by their nature introduce an opportunity for gamesmanship. This requires an altogether different approach to creditor-on-creditor aggression.

But bankruptcy and nonbankruptcy worlds cannot be neatly separated from one another. In the run-up to bankruptcy, creditors and the debtor are well advised to meet and anticipate the direction the reorganization is likely to take. Free-fall bankruptcies, bankruptcies that arise unexpectedly and for which no one has done advance planning, usually end badly. But when major players plan together before the Chapter 11 begins, they can reshape the bankruptcy process in a way that circumvents the bankruptcy rules that protect nonparticipating creditors. Those left out of the winning coalition [Pg. 216]awake in bankruptcy to find their contractual rights are gone and the protections of the Bankruptcy Code defused. The greatest challenge that creditor-on-creditor aggression presents lives in this netherworld.

I.  Coercive Exchange Offers and Strategic Behavior

Firms that encounter hard times often cannot keep their promises. In the simplest case, a debtor is unable to make an interest payment. More often, debtors break some other promise first. Debtors operate under many covenants, and even mild headwinds can leave them in breach of one or more of them. Modern loan agreements are long and complicated. They can run dozens of pages, and each contains many promises. When even just one is broken, the creditor can declare a default and demand immediate payment in full. The creditor uses such a breach to renegotiate its deal. The debtor is excused from keeping the promise and in return the creditor acquires better terms or additional control over the direction of the business or both.

Debt, however, can be dispersed.[7] Bonds are publicly traded, and private loans are syndicated. One-on-one negotiations are sometimes not possible. The simplest alternative to relying on negotiations in the face of changed circumstances is the exchange offer. The debtor retains financial consultants and investment bankers. These professionals survey the landscape, persuade the debtor to alter its course, and recommend a new debt structure that leaves everyone better off. Creditors are then asked to trade their existing bonds for new ones. The new bonds have different covenants. These give the debtor breathing room. The debtor is better managed and has a simpler capital structure that is more consistent with the condition in which it finds itself. Each of the new bonds is worth more, as each individual creditor enjoys a stream of payments and a set of promises that takes account of the realities the debtor faces. It is in the individual interest of each creditor to accept the new bond.

Such restructurings have been around for a long time. The entirely voluntary restructuring of the Atchison, Topeka, and Santa Fe Railway in 1890 is a good example.[8] The old managers of the railway had expanded too [Pg. 217]much, and the expansion had led a chaotic structure of many different types of bonds secured by discrete assets held in multiple subsidiaries. The railway’s earnings could not both pay the bonds and keep the railroad running on a sound basis. The investment bankers brought about a management change and persuaded the diverse bondholders to exchange forty-two different bonds for two bonds backed by all the assets of the railroad.[9] This more sensible capital structure reduced the railroad’s debt burden and at the same time promised each bondholder better security and more predictable cashflows over the long term.[10]

When a debtor is sufficiently distressed, however, the investment bankers often cannot put together a package attractive enough to induce each individual investor to part with its bond. There is the familiar collective action problem. Each creditor makes the group as a whole better off when it waives its rights, but each creditor bears the entire cost of making such a waiver. The creditors as a group might benefit if all agreed to the new structure, but it does not make sense for any individual creditor to give up its rights.

A hypothetical captures the dynamics. Firm is run by managers who are honest agents. They hold no ownership stake. They aspire to maximize the value of the business.[11] If the debtor remains on its current course, the firm will be worth $120 or $40 with equal probability in a year’s time. There are a hundred equityholders, each of whom holds one share of stock. A group of one hundred dispersed creditors is owed $1 each. The loan is due in five-years’ time. No interest payments are due before then, and the loan is not in default.

With debts of $100 and assets with an expected value of only $80, the debtor is insolvent. Nevertheless, as long as the creditors have no ability to call a default and exercise their right to liquidate the firm, the firm’s equity still has option value. Half the time the firm succeeds, and there will be $20 left over after the creditors are paid. Each of the 100 shares trades for ten [Pg. 218]cents, reflecting this 50 percent chance that things will turn out well. The debt trades for $70 as there is an equal chance either the firm succeeds (and creditors will be paid the $100 they are owed in full) or the firm fails and the creditors will be paid only $40.

A new project presents itself. The project can be undertaken only if most of the firm’s debt can be eliminated from the capital structure. If undertaken, the firm has an equal chance of being worth $140 or $40. A sole owner would undertake the new project. It increases the value of the firm from $80 to $90. To do the project, however, the firm must convert all the debt into equity. The managers of the debtor attempt to persuade the creditors to trade each dollar of debt for nine shares of stock. The managers’ plan benefits the creditors.[12] With the new plan in place, the all-equity firm has 1,000 shares, and an expected value of $90. By holding 900 shares, or 90 percent of the total, the former creditors receive in expectation $81. The value of their collective stake in the firm increases by $11 (from $70 to $81).

Nevertheless, it is not an equilibrium for each individual creditor to take the managers’ offer. It is not a best strategy for a creditor to exchange debt for equity if every other creditor does so. By turning down the offer when everyone else accepts it, a creditor will be repaid in full even in bad states of the world.[13]

[Pg. 219]Investment bankers have long found ways to limit this holdout problem and prevent lone dissenters from scuttling desirable reorganizations. In the nineteenth century, they used the equity receivership. The managers would persuade a well-disposed general creditor to ask an agreeable federal judge to order a foreclosure of the entire firm to the highest bidder. The foreclosure sale was orchestrated in such a way that it attracted only one, exceedingly low bid from a group working at the behest of the investment bankers.[14] Once they acquired control of the assets, the investment bankers would then give each cooperating creditor what it would have received if it had accepted an exchange offer. If the creditor refused to participate, it would be left with nothing except its pro rata share of the pitifully small amount the investment bankers paid for the firm when they went through the formalities of a foreclosure sale.

Everyone understood that the “sale” at the heart of the equity receivership was a sham. The defenders of the equity receivership thought the fictitious nature of the sale was beside the point. The foreclosure sale was a legal fiction that solved a holdout problem that would otherwise leave everyone worse off. There was no reason to feel sorry for the creditors who refused to participate. They had the same chance as everyone else to participate. Those who orchestrated the equity receivership never thought the sale that concluded it had much substance. In the course of solving a holdout problem, it happened to follow the rituals associated with a true sale, but such mimicking was characteristic of every legal fiction.

From the perspective of legal reformers in the 1930s, however, the equity receivership was not merely an ordinary exchange offer coupled with a benign mechanism that solved a holdout problem.[15] These high-minded academics, many of whom were affiliated with the Yale Law School, [Pg. 220]believed that equity receiverships and other practices deprived individual investors of their rights. These academics thought the equity receivership did little more than give investment bankers and reorganization lawyers the ability to line their own pockets and those of corporate insiders. They believed legal intervention was necessary to protect unsophisticated individual investors. As Jerome Frank explained, “Courts of equity have a tradition of aiding the helpless, such as infants, idiots, and drunkards. The average security holder in a corporate reorganization is of like kind.”[16]

There was, however, some substance to the concerns of these reformers. Public bondholders of this era were successful small entrepreneurs and professionals who could not park their wealth in real estate, nor could they store it safely in banks. They bought bonds to save for retirement. They lived in small cities or towns removed from financial centers, and they had little ability to engage with each other, the debtor, or investment bankers.[17] Making matters worse, bonds were available only in relatively large denominations. Hence, individual investors could not easily diversify their holdings.

The New Deal reformers put in place laws they thought would protect passive and unsophisticated investors. One of the ideas that animated these reforms was that each individual creditor’s right to principal and interest was sacrosanct. Alterations of the right to principal and interest outside of a corporate reorganization were forbidden, and a new reorganization regime, Chapter X, was put in place. In contrast to its predecessors, Chapter X subjected any modifications to strong oversight by a trustee, the SEC, and a federal judge.

Chapter X, however, failed spectacularly. It was unwieldy, expensive, and time-consuming. As a result, bankruptcy restructurings of large firms became impossible. More to the point, these reforms also failed to prevent coercive exchanges outside of a collective proceeding. The new laws prohibited only those modifications outside of bankruptcy that altered the obligation to pay principal and interest.[18] The reformers did not appreciate [Pg. 221]that it was possible to alter the rights of nontendering bondholders without altering rights to interest or principal and still render them worthless. By threatening to make such a change, bondholders could be induced to exchange their bonds. Bondholders could insist upon their right to principal and interest, but they would be foolish to do so.[19]

In the decades that followed, these coercive exchange offers were the focal point of much academic criticism.[20] Many thought the debtor would take advantage of this regulatory gap and put forward changes that, although contrary to the interests of the creditors as a group, were in the self-interest of each creditor to accept. They feared the debtor would deploy the familiar divide-and-conquer strategy.[21]

Consider several examples. Assume the debtor faces financial and economic distress and owes creditors $100. Without a change of course, the debtor must liquidate. The debtor is worth only $95 if it liquidates today. The debtor puts forward a new plan that allows it to stay open. If the plan succeeds, the debtor will be worth $110, but if the debtor fails, which it will with fifty-fifty probability, it will be worth only $50. The debtor can adopt this new plan only with the permission of a majority of its creditors.

This plan is in the interest of the equityholders. They receive nothing if the firm continues its present course, but with the plan they have a fifty-fifty chance the firm will be worth $110. To gain consent for the plan, the debtor offers each of the creditors five cents for each dollar it is owed provided: (1) the creditor accepts the plan; and (2) a majority of the other creditors do as well. The creditors are dispersed and cannot coordinate with one another.

Given the lack of coordination, it is in the interest of each creditor to take the offer and vote in favor of the plan.[22] To be sure, each creditor is worse off accepting the debtor’s plan than it would be if the class turned the plan down. As it stands, the creditors will be paid 95 cents on the dollar. [Pg. 222]By contrast, the new plan would give each consenting creditor only 80 cents on the dollar (a 5 cents on-the-dollar cash payment and an equal chance of either payment in full or 50 cents on the dollar). Moreover, the plan is not in the interest of the firm. The plan reduces the firm’s expected value from $95 to $80.

The parties offered the deal, however, do not focus on the welfare loss the group suffers. Instead, each creditor focuses on its own interests. If everyone else turns down the offer, a creditor loses nothing by accepting it. The firm will be liquidated, and the creditor will still receive 95 cents on the dollar. But if others accept, the creditor that refuses loses the extra five cents on the dollar. Instead of having an equal chance of 105 or 55, it will have only an equal chance of 100 or 50.

Consider another variation, one that captures the dynamic of the exchange offer more closely. The debtor owes creditors $100, and the terms of the loan forbid the debtor from taking on any senior debt without the consent of a majority of the creditors. The debtor is again worth $95 if it liquidates today, and unless the firm restructures, it must liquidate. The debtor puts forward a new plan. The benefits of the restructuring are not clear, but there is a fifty-fifty chance the firm will fail. In this event, the liquidation value of the business will then be only $50. Once again, none of the creditors can coordinate its decision with another.

The debtor approaches each creditor in sequence. It offers each creditor a package deal. The creditor must agree to vote to permit the issuance of the new tranche of secured debt. In return, each creditor can trade every dollar of unsecured debt for 90 cents of the new secured debt and some equity. The plan becomes effective only if a majority accept.

A creditor must weigh the possibilities. If the creditor accepts the offer and others do not, it will still be paid 95 cents on the dollar. If the creditor accepts the offer and the debtor does put together a majority, the creditor will be paid at least 90 cents on the dollar if the project succeeds and no less than 50 cents if it fails.

Each creditor must weigh the value of accepting the debtor’s proposal against its fate if it votes against the plan and a majority consents. In that event, the dissenting creditor’s claim would be worth only 50 cents on the dollar. It will be paid in full the half the time the project succeeds, but nothing half the time, as everything the debtor has will go to the majority that accept and would therefore acquire senior debt. The expected value of being paid in full or nothing is worse than receiving at least 50 cents and the possibility of receiving more. Each individual creditor will accept this offer regardless of how much the offered equity is worth. If the equity is of small value, the new bundle may be worth less than the 95 cents it would be [Pg. 223]worth if everyone voted against the proposal, but this is not the relevant choice.

In both hypotheticals, those running the debtor are buying votes.[23] The consenting creditors receive something of value (cash in the first case; equity and seniority in the second) in return for their willingness to support the plan. Vote buying here is problematic. To be sure, there is a large literature that defends vote buying in the corporate context.[24] In these cases, however, the interests of all the players are aligned. As long as those buying the votes cannot loot the corporation, what benefits the corporate raider and the majority benefits the minority as well. For example, if two corporate raiders try to take control of a firm by buying votes, the amount each is willing to pay turns on the value each can bring to the business. Allowing vote buying allows the assets to be put to their highest valued use. But defenses of vote buying depend upon those buying votes using their own money. This is not at work when a distressed debtor buys votes. The value the debtor is using to gain consent for its plan is coming from the minority creditors, not from it.

The engine at work in these examples is a divide-and-conquer strategy.[25] Central to a successful divide-and-conquer strategy is treating players differently depending upon how they vote. Divide-and-conquer strategies turn on the debtor’s offer containing both a carrot and a stick for the individual creditor—good things that happen if it accepts (e.g., a senior position in bad states of the world) and bad things if it refuses (e.g., a junior position in bad states). As a result, approval by a majority of the creditors no longer necessarily reflects the interests of the creditors as a group. Each creditor includes the value of the carrot and the cost of the stick in accessing the merits of the plan. Exchange offers are coercive in this sense.[26]

One should hesitate, however, before concluding too quickly that these exchange offers necessarily lead to creditors voting in favor of plans that are contrary to their self-interest. Any divide-and-conquer strategy depends critically on the inability of its targets to organize themselves. Even if it is [Pg. 224]not possible for the intended victims to speak with a single voice, it takes far fewer of them to be able to join together and prevent mischief from being done to them. In the context of debt exchange offers, the ability of creditors to organize, at least to some extent, likely prevents divide-and-conquer strategies from working.

The facts of Marblegate Asset Management v. Education Management Corporation illustrate how creditors can organize themselves sufficiently to protect themselves from a debtor intent on pitting them against one another. EDMC was a for-profit provider of college and graduate education. It enjoyed an enrollment of more than 100,000 students,[27] and held more than $1 billion in secured debt and $200 million in unsecured debt. The parent corporation of EDMC guaranteed the debt.

When EDMC’s earnings declined nearly $400 million over the course of a single year, it became clear that EDMC would soon no longer be able to pay its creditors as promised. EDMC proposed a deal in which the secured debt would receive $400 million in new bonds and 77 percent of EDMC’s common stock, a haircut of about 50 percent off the face amount of their debt. The unsecured debt would receive equity, estimated to be worth about a third of the face amount of the debt.

A stand-alone creditor was left with little choice other than to accept this offer. Under the plan, a majority, consistent with its powers under the indenture, would release the parent corporation from its guarantee and enable the debtor to transfer all its assets to a new entity that was liable only on the new debt and not on any of the old. As a result, those who did not tender would find their debtor without assets. Their right to insist on being paid principal and interest would be untouched, but it would also be worthless. As a matter of substance, individual creditors of EDMC were forced to accept a dramatic scaling back of the debtor’s obligations to them or to be left with rights against an entity that had been stripped of its assets.[28] 

It would be rational for individual creditors to tender their debt instruments and vote in favor of the exit consents even if they would be [Pg. 225]collectively better off holding tight and doing nothing. And when given the choice, over 90 percent of EDMC’s unsecured creditors and 99 percent of its secured creditors opted for the deal. This consent standing alone tells us nothing about whether the exchange offer was in the interests of the creditors as a group. Nevertheless, it is possible to infer that it was in the collective interests of the creditors, given their ability to organize themselves.

When a coercive exchange offer is open to all the creditors and when a blocking coalition can act in concert, divide-and-conquer strategies that leave each creditor worse off are not possible. When a blocking coalition can organize itself, it will approve only those offers that leave them better off, and without their support the plan will fail.

Consider the simple case in which there are three creditors. Two control 35 percent and the third controls 30 percent, and the debtor’s plan requires consent from two-thirds of the creditors.[29] In such a situation, no divide-and-conquer strategy is possible. Each of the two creditors with a 35 percent stake has a veto right. Matters become only slightly more complicated when multiple creditors hold large positions. A coalition of a small number of them may be all that is necessary to put a stop to any mischief.

These dynamics are plain when there are four creditors, each with an equal stake, and again a two-thirds vote is necessary to approve any plan. The debtor proposes a risky plan in which it offers a haircut, but those creditors who consent to the plan would be given a senior position. Assume the firm without the plan is worth $120 or $40 with equal probability and the creditors are owed $100. (In this case, the expected value of each creditor’s claim before the plan is 70 cents on the dollar.) Under the plan, each creditor must agree to a 10 percent haircut. If the plan is approved and succeeds, the debtor will be worth $140.

Whether the plan makes sense turns on how much the debtor will be worth if the plan fails, but the payout under the plans does not affect the way creditors vote when they cannot coordinate. Assume the debtor would be worth $60 in the bad state. In this case by voting in favor, the creditors will receive in expectation at least 75 cents on the dollar if others accept and will fare no worse if they do not as the status quo remains in place.[30] If the same creditor refused to accept the plan, however, it would be paid in full if the plan succeeded, but it would receive nothing if the plan failed, as the [Pg. 226]other creditors would have a senior position. This would leave the creditor who rejected the plan with a claim with an expected value of only 50 cents on the dollar.[31] Receiving 75 cents is better than receiving 50 cents.

Accepting this plan is a good deal for the creditors. It leaves them better off than they would be if the status quo remained. (With claims worth 75 cents on the dollar rather than 70 cents). But the voting dynamic would be the same even if the firm would be worth only $20 if the plan failed. In this case, the plan would not change the value of the firm. (When this plan is put in place, the firm’s expected value remains at $80.) The plan merely shifts value from creditors to the equityholders. Nevertheless, when the creditors cannot organize, each creditor will still accept the plan. Each creditor will receive 90 cents half the time and 20 cents the other or an expected value of 55 cents. This is better than the 50 cents each would receive if it voted against the plan and others accepted (with payment in full half the time and nothing the other). If the creditors cannot coordinate with one another, each will again end up voting in favor of the plan even though it leaves each creditor with less than the 70 cents on the dollar they would have received if each voted against the plan.

But if two of the four creditors coordinate their actions, the two plans no longer share the same voting dynamics. The two creditors who can coordinate will vote against any plan that is not in their interest. Because the first plan leaves each creditor better off and the second does not, they will both agree to vote in favor of the former, but against the latter. To be sure, the other creditors who cannot coordinate will still vote in favor of both plans, but their support of the second plan, even though contrary to the self-interest of the creditors as a group, is harmless. The coordination of two creditors is enough to ensure only favorable plans are approved.[32]

Both plans are coercive in the sense that a rational individual creditor will consent to them, but as long as some coordination is possible among the creditors, coercion of the minority is not a problem. Indeed, such coercion is affirmatively desirable in many environments. The pressure put on solitary creditors neutralizes the ability of any individual creditor to hold out. It prevents an obstructionist creditor from shaking down the others. Ideally, the debtor should have enough power to overcome the holdout problem, but not too much.

[Pg. 227]As it happened, the creditors in Marblegate were able to organize themselves. A group of creditors that possessed a majority of the vote formed an ad hoc committee. This committee was well-positioned to assess the merits of any plan the debtor put forward. It was led by seasoned investors with skilled advisors by their sides. The committee could assess the debtor’s plan and reject it or insist on a different one if it chose.

Dissenting creditors might not be able to participate in any of the negotiations with the debtor, but these dissenters shared the same interests as the creditors who organized themselves. A creditor might have a different view of the merits of the plan, but one should not think that this creditor was more likely to be right than the committee.

When public bonds are being restructured, the debtor has little interest in setting creditors against each other. The debtor wants to clean up its balance sheet. The debtor’s first best option is to make everyone part of the new capital structure. The more people who consent to its plan, the better. The holdout problem is front and center, and the debtor is eager for all the bondholders to participate on equal terms.

Creditors who hold publicly traded debt engage in comparatively little monitoring. The interactions between the creditors and the debtor are comparatively few. In this environment, it is also comparatively simple to ensure at the outset that creditors cannot take advantage of each other. There is no downside to a provision in the bond that all creditors receive share “pari passu” or receive the “same treatment.” Pari passu clauses are common in publicly traded debt.[33]

In a case like Marblegate, dissenters enjoy the same terms as the majority coalition as long as they voted in favor of the plan.[34] Voting against the plan is therefore foolish and the right to vote for them is essentially meaningless. [Pg. 228]Courts reviewing publicly traded debt even suggested that allowing everyone to participate on equal terms was an essential feature of such debt:

Had [the debtor] not made its offer to all bondholders on the same terms, but had it privately paid money to sufficient holders to carry the election, one would, without more, feel some confidence in concluding, provisionally at least, that such conduct was so inconsistent with the concept of voting implied by the amendment provision that it constituted a violation of what must have been the reasonable expectation of the contracting parties.[35]

To be sure, the individual creditor is forced to accept the majority’s view of the merits. It is being denied a voice, but it is not obvious that we should care. A meaningful right to vote might matter in political debate, but democratic values are not at stake when the governance mechanism is in service of professional investors trying to maximize returns. The vote of the majority when all participate in the plan on the same terms is trustworthy. It reflects the wisdom of crowds.[36] When a majority can speak with a single voice and accedes to a plan in which every creditor is treated the same, it is a fair inference that the plan leaves each creditor better off than the status quo. As long as the minority creditors receive the same treatment as the majority when they approve the plan, allowing the majority to leave them worse off if they withhold their consent is a straightforward mechanism that overcomes the holdout problem.

One can fault the reformers who brought about the Trust Indenture Act for failing to understand that their reforms did not limit the ability of debtors to engage in coercive exchange offers as they had intended. The harm from coercive exchange offers is easy to overstate, however.[37] The apparent cost of coercion in exchange offers—the ability of the debtor to engage in a divide-and-conquer strategy—is small in a world in which creditors can organize themselves.

Coalition formation, however, has a dark side. A coalition of creditors can join forces with the debtor and take advantage of the other creditors. Return to the hypothetical in which there are four creditors and any three can bind the fourth. Two creditors can defeat a divide-and-conquer strategy, and this inures to everyone’s benefit. But trouble starts when three can coordinate. As with a coalition of two creditors, no deal can happen that leaves the three creditors worse off. But the three have the power to [Pg. 229]approve a deal over the objection of the fourth. This allows them to transfer value from this creditor to themselves.

This danger of coalition formation plays a small role in the case of public debt. Syndicated debt, however, has an altogether different dynamic, and the dark side of coalition formation manifests itself here. It might seem that creditors could prevent these problems by putting a pari passu clause into their agreements. Granting similarly situated stakeholders the ability to participate in any offer on equal terms, however, is not necessarily always in their collective interest.

We see analogous provisions—known as tag-along rights—included in some corporate by-laws. They work in analogous fashion by protecting minority equityholders when the majority decides to sell. But it is well understood that tag-along rights come with trade-offs. Giving everyone an option to participate takes more time and is more costly. Moreover, when the debtor becomes distressed and needs reorganizing, designing the new capital structure requires someone to invest money and muster expertise. Those positioned to take action may have too little incentive to do so if they have to share the benefits of their work equally with everyone else.[38] The debate here may replicate the familiar one from battles between controlling and noncontrolling shareholders.[39] Mandatory rules designed to limit creditor-on-creditor aggression might result in creditors having too little incentive to monitor the debtor and keep it on track.

II. Private Debt and the Rise of Creditor Control

Syndicated loans are often used in leveraged acquisitions. The buyer of a business borrows against its assets to raise part of the purchase price. For example, when Revlon acquired Elizabeth Arden in 2016 in a leveraged transaction, it needed to borrow $1.8 billion dollars, both to pay $870 million to the old equity and to retire the existing debt. Syndicated loans are the avenue of choice for such transactions. More than a trillion dollars of such loans are issued each year in the United States.[40]

[Pg. 230]In a syndicated loan, a large bank lends some of the money and finds other banks or financial institutions to fund the balance. For arranging such transactions, the bank receives a fee that ranges between 1 percent and 5 percent, depending upon the complexity of the transaction and the risks it is assuming.[41] Because these loans carry a large measure of risk, lenders will both demand a high rate of interest and insist upon many restrictions on the debtor’s freedom of action. In particular, the debtor will enjoy only a limited ability to sell or transfer its assets without obtaining permission. Similarly, the debtor is sharply limited in its ability to take on additional debt.

The parties need to reach agreement at the time of the loan about which actions require permission from the creditors and, if they do, how exactly that permission is obtained.[42] Putting together the right mix of covenants is hard. The purpose of loan covenants is not to prevent the debtor from making major decisions, but rather to allow creditors to exercise control when conditions warrant. In bad states of the world, creditors bear the costs if things turn out badly, and they will want to be able to take control. Among other things, they will want to have the power to replace the CEO and other managers when the business falters.

Spelling out exactly what the debtor can and cannot do without the creditors’ permission requires contracts that are long and complex. The more complicated the enterprise, the harder the job becomes. Such contracts are necessarily imperfect. Creditors need to give the debtor some freedom of action to prevent good opportunities from being missed. Doing this, however, allows room for advantage-taking under different circumstances. It is not possible to write a contract that perfectly distinguishes between the two in advance.

As important as the covenants themselves, however, is the governance mechanism the creditors in a syndicate create among themselves. Seeking waivers is an ordinary part of the debtor-creditor relationship in these loans. Hence, an issue as critical as the covenants themselves is the mechanism used to grant a waiver. The creditors commonly agree to empower an agent, such as the bank that arranged the loan, to issue some sorts of waivers on their behalf. For routine decisions, such delegations make sense.

The creditors can also agree that, with respect to certain covenants, only a majority or a supermajority of the creditors can agree to waive their rights. Other rights are even more important. Each creditor enjoys these unless and until it explicitly waives them. These are known as “sacred rights.” A [Pg. 231]prohibition on any change in principal or interest in private loans is almost always included among the sacred rights set out by contract.[43] But sacred rights do not stop there. Common are clauses prohibiting the sale of assets free and clear of security interests as well as limitations on the debtor’s ability to give other creditors security interests. In stark contrast to the straightforward prohibitions of the Trust Indenture Act, however, sacred rights in private loan agreements are qualified and subject to exceptions. For example, notwithstanding each creditor’s right to pari passu treatment, the debtor might have the ability to buy back debt in an arms’ length transaction without being obliged to buy back debt pro rata from everyone.

Designing the best mechanism for creditors to govern themselves is akin to constitution making. The group needs to be able to act as one, but there is always the danger one group can hijack the process and force through a decision that benefits itself at the expense of others. The ability of factions to organize themselves is double-edged. On the one hand, the ability to create alliances disarms the ability of outside forces to engage in divide-and-conquer strategies. On the other hand, factions are themselves a source of internecine conflict. Governance mechanisms cannot eliminate these problems. They can only minimize them. And the appropriate sort of mechanism turns itself on time and place.

It is often hard to determine whether a particular course advances everyone’s interest or is mere advantage-taking. Contrast the following two hypotheticals. In the first, a firm owes its creditors $100 and is worth $100 with certainty. Debtor finds a project that will make the firm worth $270 or $50 with equal likelihood, but it needs permission from half of the creditors to take it on. Debtor approaches a bare majority of the creditors and offers, in return for their permission to take on the project, to exchange their debt for half of the equity in the firm. The remaining creditors are not consulted and are not given the chance to participate.

The majority coalition will accept the project. The equity they receive in the deal has an expected value of $110 half the time and is worthless the other half. When it is successful, $270 will be available. After $50 is paid to the remaining creditors, success will be yield $220. Half this goes to the majority coalition. This package has an expected value of 110 cents on the dollar. (The majority under the status quo is owed $50 and under the plan will be paid $110 half the time.) For a risk-neutral investor, this is better than being paid 100 cents on the dollar with certainty. And the equity holders are better off. With the plan in place, their equity has an expected value of $55 instead of $0. Most significantly for present purposes, the minority creditors are no worse off. Even if the bad state of the world [Pg. 232]manifests itself, there will be enough to pay them in full. (In the bad state, the assets are worth half as much as before, but the minority creditors are the only creditors, so there is enough to pay them everything they are owed.)

This transaction is one in which the debtor is giving value to the majority in return for their permission to go forward with the plan, but none of the value is coming from the minority creditors. The minority creditors are being paid everything they are owed. It might therefore seem unobjectionable. The problem, however, is that such plans are hard to distinguish from those in which a plan is advanced at the expense of the minority.

Consider a second hypothetical. The debtor is worth $130 or $90 with equal likelihood, and it owes its creditors $100. The expected value of the debt is $95. The debtor develops a plan that again can be implemented only with permission of a majority of the creditors. With the plan in place, the debtor will be worth $150 or $50 with equal likelihood. The debtor negotiates with a majority coalition. The debtor agrees to give the majority coalition priority over the other creditors and in return, the coalition permits the debtor to go forward with its plan.

With the plan in place, the value of the debt of the coalition members increases from 95 cents to 100 cents on the dollar. They will be paid in full even in the worst-case scenario. The equityholders are also better off. Their equity stake rises in value from $15 (a fifty-fifty chance of $30) to $25 (a fifty-fifty chance of $50). But the minority is worse off. The value of their stake falls from 95 cents on the dollar to 50 cents. They will be paid in full half the time and will receive nothing half the time.

It is a mistake, however, to think that it is in the interest of the parties to prohibit such a transaction in advance merely because it effects a wealth transfer from some creditors in a class to other creditors in the same class. If the creditors are homogenous, the creditors in the majority when one debtor is restructured will be in the minority when another debtor attempts to restructure. When creditors are indistinguishable, any given creditor is as likely to be a victim or participant in creditor-on-creditor aggression. When the creditors do not know in advance whether they will be part of the winning coalition, the risk of the wealth transfer is itself diversifiable. The expected value of its investment in any debtor is constant notwithstanding the risk of expropriation in any given case.

If the creditors in a class are homogenous and hold diversified portfolios, transfers within a particular class offset each other. Investors before the fact would not know whether they will be on the winning or losing end of such a transfer, and the possibility that a creditor will be in the majority and capture wealth from the minority offsets the risk that a creditor will find [Pg. 233]itself in the minority. The transfer of value from the minority to the majority is a zero-sum game.

It might seem that the debtor’s ability to transfer wealth in a restructuring will affect ex ante investment. Investors might be reluctant to lend when faced with the prospect that the debtor will later transfer value away from them. This effect on the ability to assemble capital is one of the principal reasons for ensuring that priority among creditors is respected.[44] But the principal effect when majority coalitions take advantage of a minority is a transfer of wealth inside a class. Such transfers do not themselves change the amount the class as a whole receives. Absolute priority across classes is respected.

Creditors as a group, however, would still want to prevent this second transaction from going forward if they could. The plan produces a deadweight welfare loss of $10 in addition to transferring wealth from the minority to the majority coalition and the equityholders. The problem with the plan is not the majority captures wealth from the minority, but that it destroys value in the process. This problem, however, is easy to exaggerate.

Hypotheticals in which the debtor puts in place value-reducing plans are simple enough to invent (such as taking all the firm’s assets to a casino and betting on black), but such opportunities are hard to find in practice.[45] There are only so many ways in which the debtor can change its operations over a relatively short time horizon. There are few opportunities for a debtor to reconfigure its assets in a way that is better for the debtor and a majority of its creditors, but so much worse for the minority of creditors that the transaction as a whole produces a net social loss.

The plan a distressed debtor typically puts forward is usually not a radical change in asset deployment, but rather a way to ride out a business cycle. All that is happening is the debtor is continuing for too long on the same course, a course the creditors approved in the first instance. Such a path may be bad, but it is unlikely to be catastrophically bad. The debtor continues to lose money and postpones what will be an inevitable, but expensive restructuring. But pulling the trigger on an established business is costly, and other options may be nearly as bad.

Even if the debtor can embark on a strategy that is excessively risky, the majority creditors will resist. In practice, restructuring transactions have two possible outcomes. In one, the new cash infusion works, the firm turns around, prosperity returns, and the minority creditors are paid their [Pg. 234]principal and interest in full. There is no expropriation. Alternatively, things go badly and there is expropriation. But in this event, the senior creditors end up owning the equity of the firm. Creditors in a majority coalition anticipate this possibility, and this gives them the same incentives as any residual owner. They will tend to resist any transactions that reduce the total value of the firm given that they ultimately bear the consequences of excessive risk-taking. This will dampen their eagerness to enter transactions that capture wealth from junior stakeholders and reduce firm value at the same time.

Further complicating the picture are the negotiating dynamics among the creditors when the debtor puts forward plans that would allow one coalition of creditors to take advantage of another. Instead of agreeing to a deal with the debtor, the creditors may make common cause with each other. The creditors themselves might form a coalition and work in concert to prevent a transaction reduces the value of the firm.[46] For all these reasons, the risk that creditor-on-creditor aggression outside of bankruptcy generates value-reducing transactions might be overstated.

It still, however, is in the interest of creditors at the outset to create governance mechanisms that minimize the incentive of creditors to try to capture wealth from one another. The simplest mechanism is one that mandates equal treatment for everyone. In such an environment, much can be done with a simple rule that empowers a majority (or a supermajority) to bind dissenters while at the same time ensuring the same treatment among all members of the group. Many other mechanisms are available as well. The way in which contracts require supermajorities for some decisions and not for others is one example of how contract design can dampen the ability of parties to engage in rent-seeking.

There is one more complication. Members of loan syndicates are not all cut from the same cloth. Some are sophisticated hedge funds that hold a large piece of a particular loan. They might have invested in the debtor from the start, but as likely they specialize in distressed debt and assembled their positions after the firm encountered trouble. These hedge funds believe they have a comparative advantage in navigating the challenges presented when a debtor is in trouble, and active oversight of their debtor is an integral part of their business model. Their returns depend not only on making good investments, but on influencing how business is done.

[Pg. 235]These activist investors bring in financial advisors. Among other things, they scour all the loan documents to understand what kinds of opportunities exist. Many of these opportunities involve both restructuring existing debt and making new capital investments in the business. They are well-situated to oversee the debtor and assess what needs to be done. They can observe whether the debtor needs a new injection of capital or whether it is time to cut their losses. Similarly, they can assess whether the current managers are doing the best that can be done in hard times or are themselves the problem.

By granting or withholding waivers, activist investors can do much to ensure that the debtor is put on the right course. It goes without saying, of course, that the activist investors are pursuing their own interests, not the greater good. They are relatively indifferent to whether the changes they bring about expand the size of the pie, increase the size of their slice, or both. They do not care whether the new opportunities they discover are ones in which other investors will share equally or at all. As a doctrinal matter, individual creditors, in contrast to controlling equityholders, do not owe fiduciary duties to other creditors. Their only obligation is to act in good faith and adhere to reasonable commercial standards of fair dealing.

Activist creditors as a group can protect their own interests. They know the covenants they will enjoy when they buy their positions, and they can contend with other activist investors on something like equal terms. But members of loan syndicates do not consist entirely of activist investors competing with one another. For example, managers of pools of collateralized loan obligations (“CLOs”) have increasingly participated in loan syndicates.[47] To be sure, the players in the private debt space are small in relative terms only. They are still managing many millions of dollars. They do not include any of the small investors the New Deal reformers worried about when they put forward the Trust Indenture Act. Nevertheless, they complicate the picture as they differ from activist creditors in distinctive ways.

Consider the position of a CLO manager caught up in a contest between competing groups of creditors.[48] A CLO manager assembles a pool of syndicated loans into a legal entity. That entity in turn issues securities backed by this pool of debt to other investors, such as pension funds, university endowments, and high-wealth individuals. Because each pool consists of many loans, these investments offer returns that are remarkably [Pg. 236]stable. CLO debt have performed consistently well over time, even during the Great Recession.[49]

Much of the wealth activist investors capture as a result of creditor-on-creditor aggression is taken from such passive investors. Of course, these CLO managers still should receive a market return on their capital. Nevertheless, these do not have the ability to take action when the debtor is distressed. They are at the mercy of whatever loan covenants are in place, and they do not necessarily account for the particular covenants in the loan document when they decide which loans to buy.[50]

When CLO managers decide whether to add a loan to their portfolios, the terms of the loan are far from their primary consideration. They care in the first instance about the quality of the firm and then about the interest rate the loan will carry. They are generally aware of the loan covenants, their sacred rights, and the voting rules in place, but these are at best third-order concerns.

To assemble their pools, CLO managers rely on their connections with these private equity sponsors. The sponsors are repeat players, and they can choose to do business with some CLO managers and not with others. This in turn gives CLO managers only a limited ability to pick and choose among the deals they are offered. They stand in much the same position as wholesale diamond buyers. In the wholesale diamond market, at least as it existed in the 1980s, favored buyers were given a box (or “sight”) of diamonds of a particular grade with a take-it-or-leave-it price. Buyers who refused to take the sight they were offered would not be invited to return. Because the wholesaler enjoyed a cartel, it had the ability to exclude potential buyers who spend time and energy scrutinizing the diamonds in each sight, taking only good sights and refusing to accept inferior ones. Conditional upon the existence of the cartel, this arrangement may have been mutually beneficial. It prevented cherry-picking. No one could buy only sights in which the diamonds were better on average than other sights within the same grade.[51] Hence, no one wasted resources taking advantage of the inevitable differences across boxes.

Private equity sponsors may possess a power similar to that of the diamond cartel. The market makers for these high-yield loans consists of a [Pg. 237]few major players, and they have control over deal flow. The CLO manager may be better off taking the bundle of deals a large private equity fund offers, rather risking the loss of future deals by kicking the tires of individual deals too hard. The arrangement saves the CLO manager from investing resources distinguishing among deals the private equity firm has assembled. The CLO manager is better off trusting that the private equity fund’s book of deals as a whole will perform well.

Even if the CLO managers have some choice over which deals they include in their pools, it makes little sense to hire lawyers to focus on the terms of the contract. Even if legal risk were all that mattered, curing imperfections in contract language is not the low-hanging fruit. Before the fact, the improvements one can make in the terms of any individual loan are small relative to the other legal risks associated with the transaction. There is, for example, in every leveraged loan transaction a risk of a fraudulent conveyance attack. A lawyer’s time is likely better invested in assessing this sort of risk rather than tweaking covenants.

Merely understanding what the debtor is and is not allowed to do requires delving into documents many tens of thousands of words long. Relatively few (perhaps only 3 percent) of these high-yield loans default on their payment obligations. Moreover, because they are backed by collateral, the payout in the event of default is high (perhaps as much as 80 percent). If the CLO manager were to invest $10 million in a particular loan and if she were able to eliminate all default risk associated with it, she would pocket only $60,000. The expected losses as a result of expropriation by activist investors is only one of the costs associated with default, and it would take many billable hours to identify all the terms activist investors could exploit.

Even if CLO managers completely understood the ways in which activist members of the loan syndicate can capture wealth from them with contractual terms, it is not necessarily in their interest to insist on changes that restrain their freedom of action. The pricing of the loan should, in expectation, compensate them for the risks that they are running.[52] More to the point, the inefficiencies associated with such transfers must be balanced against the ways in which the activist investors can use these powers to enhance the value of the debtor. The CLO managers want to free-ride on the efforts of the large activist investors to keep the debtor in line and on track. It may not be possible to do this and still prevent expropriation.

It is against this background that the creditors must fashion the mechanism that must govern their relationship among one another. In this environment, simple solutions (such as pari passu clauses) may be inapt. For [Pg. 238]example, the debtor might find itself in trouble and need additional funds. In such a case, the debtor might seek permission from its existing creditors to engage in new borrowing that takes priority over the existing debt. At the same time those most likely to be in a position to extend new capital may be the activist creditors. The new borrowing could leave the firm as a whole better off, but leave passive creditors who do not participate in the refinancing worse off.

The loan agreement could, of course, require all the existing creditors be given the same opportunity to participate in the refinancing, but CLO managers themselves are not in a position to double-down on their investment in a distressed business. Moreover, forcing activist investors to run deals past them might be costly. It is conceivable that it would be in the interests of all creditors before the fact to allow activist creditors to act unilaterally. Hence, even passive investors might agree to limit the scope of a pari passu clause to make such lending possible.

There is much to be said for relying on the contract the parties themselves create. Outside of bankruptcy, the loan agreement reconciles the tension between ensuring that the debtor can get permission to engage in desirable transactions and the imposition of costs on minority creditors. The judge’s job is to discover this agreement.

Quite apart from the parties themselves being able to identify their own interests, these contracts are subject to evolutionary pressures in ways the rules a statute puts in place are not. Members of loan syndicates are repeat players. They enter into many loan transactions with the same counterparties. If a loophole is discovered that allows costly advantage-taking, parties can fix it in their subsequent deals. Unlike mandatory legal rules, the governance mechanism that parties create is self-correcting. This feature itself pushes against strong intervention.[53]

It is possible, of course, to have a legal rule that requires activist investors to act in the interests of all the creditors, but such a legal rule makes strong assumptions about the ability of the judge to distinguish between transactions that advance the joint interests of the parties and those that do not. Merely showing that the restructuring effects a transfer of wealth after the fact is not itself sufficient. More to the point, parties could, if they wanted, write agreements in which those in positions of power agreed to act as fiduciaries on behalf of the creditors as a whole. We do not see such [Pg. 239]undertakings, however, and it is possible to draw an inference from silence.[54] If parties do not bargain for such protection, it is not obvious why it should be forced upon them.[55]

There are comparatively few cases that examine the ability of minority creditors to prevent a majority coalition from extracting value from them. The state of the law is unclear. A number of recent efforts of creditor coalitions to dismiss such actions outright have failed.[56] When they reach the merits, however, courts tend to enforce its various clauses as best they can. They give relatively little scope to general duties to act in good faith.[57] In a recent case, for example, the debtor sought the consent of the lenders to take on senior debt. The debtor offered a majority of the existing creditors, but only a bare majority, the new debt on favorable terms.[58]

We can infer that the offer left the majority creditors better off. When considering this offer, creditors in the majority likely weighed the benefit they would enjoy from making the senior loan on favorable terms against the cost to them of having new debt senior to their existing debt. By contrast, the minority had no chance to participate in the loan, and after the fact it left them worse off. None of the benefits of enjoying a high return on a new loan offset the cost of being subordinated. Any net benefit would come indirectly only if the new loan improved the overall fortunes of the firm by more than it reduced their likelihood of being paid in bad states of the world.

Because the plan impacted the majority and the minority differently, one cannot not infer from the willingness of the majority to approve the new loan that the new loan was in the interests of the creditors as a group. At the same time, however, the debtor’s plan might have been consistent with a coherent ex ante bargain. The loan could improve the overall value of the firm even if it left the minority worse off after the fact. And it could even leave the stranded creditors better off after the fact. Existing creditors are sometimes better off when the debtor acquires new debt with higher priority. Nor is it surprising that some existing creditors were the source of new debt or that the terms offered these creditors were especially attractive to them.

[Pg. 240]Against this background, the court did not look beyond the language of the loan agreement itself. The arguments the parties made allowed the court to reduce the dispute to the narrow question of whether the transaction tripped up the sacred rights of the minority.[59] The court examined the sacred rights in the loan. The court confronted the following language:

[W]ithout the consent of each Holder affected thereby, an amendment, supplement or waiver under this Section 9.02 may not (with respect to any [Senior Secured] Notes held by a nonconsenting Holder) . . . make any change in the provisions in the [2019 Intercreditor Agreement] or this Indenture dealing with the application of proceeds of Collateral that would adversely affect the Holders.

The minority creditors argued that by making a new loan senior to existing loan, the amendment changed the way proceeds of the collateral would be ultimately distributed. Hence, the debtor’s plan amended the loan agreement in a way that violated a sacred right.

The court rejected this argument. The minority’s argument was not itself implausible, but neither was the argument that the minority did not possess a right to be included in every deal. If those drafting the contract wanted to ensure that a minority could veto any issuance of senior debt to the other creditors, there was a much simpler way for them to write the contract. They could have added a provision to the loan document that allowed each creditor to veto any new loan that enjoyed a higher priority in the same collateral. Clauses that bar subordination of one lien to another are easy to write. The availability of this simpler route provides a reason against interpreting the “application of proceeds” language expansively. Moreover, another provision of the loan contract allowed a supermajority to release the collateral entirely. The greater usually includes the lesser. It seemed odd to give each creditor a veto over priming liens, but to allow only a supermajority to release the same collateral entirely.

The court’s reasoning here is not ironclad. Contracts that are this complicated are inherently ambiguous. Allowing a priming lien that affects some creditors in a group differently from others is not the same as releasing collateral, an act that affects all equally. Nevertheless, the court’s approach to interpreting lending agreements in this narrow fashion is easy to defend. Everyone can be better off if courts interpret syndicated loan documents narrowly. To be sure, such interpretations may lead to an outcome that the parties would not have wanted had they confronted this outcome in advance, but no interpretative approach is perfect. Narrow interpretations [Pg. 241]offer predictability, and because the parties to syndicated loans are repeat players that adapt their contracts in the wake of judicial opinions that are too literal, being excessively literal comes at low cost. It is nothing like an excessively mindless interpretation of a statute that takes both houses of Congress and the president to correct.

Interpreting contracts narrowly, however, still leaves much unsettled. There are limits to how much a contract can spell out, and there are always gaps to fill. Consider the challenge facing the parties in negotiating their loan agreement when they need to decide whether the debtor should be able to repurchase its own debt. It is easy enough to write a contract that requires a debtor to approach every member of the syndicate and offer to repurchase the debt of everyone on the same terms, but it might be cumbersome, and parties might rationally choose not to force a debtor to take this extra step.

To be sure, some parties might want to prevent their debtor from picking and choosing among creditors in bad times. But an outright ban on repurchasing debt has costs. A debtor that is doing well may have excess cash, and it may want to retire some of its debt rather than issue a dividend. It may benefit everyone if the debtor can approach a creditor that wants to deploy its capital elsewhere and strike a deal with it. Such a deal might leave everyone better off. A creditor that desires to exit has a chance to do so. The creditors who want to stand pat with their investment in the company benefit from a debtor with less leverage. They are better off than they would be if the company had merely issued a dividend to shareholders.

Parties solve this sort of problem by having a clause in the loan agreement that forbids the debtor from paying off debt as a general matter, but still allows the debtor to make “open-market” purchases of its debt. Debtors cannot pick and choose which creditor to pay, but an arms’ length repurchase of debt is permitted. The debtor can buy debt, and creditors can sell it, but the debtor cannot cut special deals on the side.

The basic idea behind a provision giving the debtor the right to make open-market purchases is clear enough, but establishing the exact contours of this right is not easy. Syndicated debt is not traded in the same fashion as a stock listed on an exchange. What constitutes an “open-market” purchase is not self-evident. One can argue that only offers made to everyone can be open-market transactions. Deals made between the debtor and only some creditors in private without the knowledge of others are not open-market purchases, at least not when it is part of an elaborate transaction in which the repurchase is made in conjunction with a waiver of loan covenants and an issuance of senior debt against the same collateral. On the other hand, it is possible to argue that an open-market purchase is any transaction that reflects a deal between a willing seller and a willing buyer. If the debtor is [Pg. 242]buying the debt on the same terms as others in the marketplace, it should be unobjectionable.

The contract, of course, could limit what constitutes an “open-market purchase.” Given all the parties are sophisticated and that in any event the contours of the definition are not self-evident, there is every reason to defer to the written contract. Such a contract could either prevent or allow the debtor and a majority coalition strike a new deal. But when the contract itself does not offer a definition, the court must supply one.

In addition to focusing on inferences that can be drawn from the language of the contract and alternative language that could have been used in its place, but was not, a court enjoys a license to examine the transaction as a whole and ask whether the parties have carried out their contractual undertakings in good faith.

It has long been the law that creditors owe some duties to one another. One creditor cannot, for example, orchestrate a foreclosure sale that allows it to buy a firm at a bargain price and at the same time take active steps to ensure that the other creditors do not know the sale is taking place.[60] In addition to exploring the meaning of the words in the contract, the judge must decide how expansively to interpret the duty of good faith. Every party to a contract must act in good faith, and the duty of good faith extends beyond mere honesty in fact. “Good faith” is not a general mandate that parties be nice to each other, but rather a principle that allows the court to intervene to prevent parties from exploiting the gaps that exist in any agreement. It requires each party to adhere to “reasonable commercial standards of fair dealing.”[61]

Minority creditors can argue that a majority coalition fails to act in good faith when it engages in an elaborate transaction that leaves them worse off. Such good faith arguments are often tied to arguments interpreting specific contractual language. In Serta Simmons, for example, the debtor and a bare majority of creditors agreed to a restructuring that left the remaining creditors in the same class subordinated and stripped of guarantees. The debtor argued the exchange was an “open-market” purchase within the meaning of the loan agreement. The minority creditors insisted that the plan was merely a scheme to transfer wealth away from them into the pockets of the majority coalition. As such, it violated the duty of both the debtor and [Pg. 243]the coalition to act in good faith according to reasonable standards of fair dealing.[62]

 Arguments grounded on good faith, however, have met with only limited success so far.[63] The rise of private debt and recent cases of advantage-taking by majority coalitions, however, may change matters.[64] Empowering courts to invoke the duty of good faith is not the same as asking the judge to play tennis without a net. Outside of bankruptcy, good faith serves as a marker for the general ability of courts to flesh out necessarily incomplete contracts. As Arthur Corbin explained, “[P]arties occasionally have understandings or expectations that were so fundamental that they did not need to negotiate about those expectations.”[65]

Parties are not able to imagine in advance every sort of maneuver that constitutes advantage-taking, so it makes sense that they do not have to write contracts that spell out in chapter and verse all the ways in which their contracting opposites are forbidden from playing gotcha games.[66] In every contractual relationship, some things are so much a part of the warp and woof of the contractual relationship that they go without saying. At bottom, discovering good faith duties is the same thing as identifying implicit contract terms. As Richard Posner explained:

[W]hether we say that a contract shall be deemed to contain such implied conditions as are necessary to make sense of the contract, or that a contract obligates the parties to cooperate in its performance in “good faith” to the extent necessary to carry out the purposes of the contract, comes to much the same thing. They are different ways of formulating the overriding purpose of contract law, which is to give the parties what they would have stipulated for expressly if at the time of making the contract they had had complete knowledge of the future and the costs of [Pg. 244]negotiating and adding provisions to the contract had been zero.[67]

When interpreting a bond indenture, the Delaware Chancery Court put forward a test for “good faith” that captured the same idea:

[Courts need to ask whether it is] clear from what was expressly agreed upon that the parties who negotiated the express terms of the contract would have agreed to proscribe the act later complained of as a breach of the implied covenant of good faith—had they thought to negotiate with respect to that matter. If the answer to this question is yes, then . . . a court is justified in concluding that such an act constitutes a breach of the implied covenant of good faith.[68]

Answering such a counterfactual question, of course, is not easy, but this is the task at hand.

But courts must first gain their bearings before they ask whether anyone is acting in bad faith. Courts should not strike down a plan merely because it leaves some parties worse off after the fact. Ex post transfers between the parties are not necessarily something parties would prohibit in a fully dickered agreement. Sophisticated parties can account for these in their original bargain. A sensible understanding of good faith requires asking what sorts of plans parties would want to prohibit before the fact.

The contours of good faith duties are complicated in a world in which CLO managers find themselves in the same class as activist hedge funds. As noted, CLO managers do less monitoring, and they are most likely to be the losers when a majority coalition engages in creditor-on-creditor aggression. This cost, however, should not be exaggerated. CLO managers can exit and transfer their debt into the hands of activist investors well equipped to look out for their own interests. Debt trades. The creditor that holds a relatively small portion of the debt can leave when the debtor is distressed and activist traders enter the scene. Moreover, CLO managers free-ride on the activism in the hedge funds. One can look at the risk of expropriation as the price that they pay for this service.

There is still room for good faith to operate, however. A court might not be equipped to tell whether the majority coalition sold out the minority, but it can ensure that the majority is in fact a majority. Parties should not have to imagine all the ways in which the voting process itself can be corrupted before a court should be able to step in to ensure that the votes are properly cast and properly counted. Parties should be able to provide [Pg. 245]that consent of a majority is necessary to amend a contract without having to list all the sham transactions that might be put together to create the appearance of a majority.

Revlon is an example of a case where judicial intervention may be sensible. Revlon’s debt was trading for 43 cents on the dollar, and a majority of the creditors, albeit a bare majority, had lost patience. They formed a coalition to oppose the debtor’s efforts to restructure its debt. The debtor resisted them by making common cause with a coalition of creditors who fell just short of constituting a majority. The debtor then exploited its ability under the loan document to issue additional debt. To give the minority coalition the necessary power to bind the group, the debtor issued just enough additional debt to its creditor coalition to transform it into the holder of a majority of the debt. Even though the debt was trading at a large discount, there were takers for it as the debtor promised to pay this new debt in full shortly after the plan was consummated. The creditors making this new “loan” were not exposing themselves to any credit risk.

In this fashion, the debtor was able to gain the votes it needed to acquire the refinancing it needed to put off what was likely an inevitable day of reckoning. The majority coalition that struck the deal with the debtor obtained a senior position in the capital structure, one that gave it the upper hand in Revlon’s subsequent bankruptcy. And the stranded minority coalition was left holding crumbs.[69] 

If presented with the issue, a court might have found the debtor’s obtaining of consent in this fashion was not done in good faith.[70] A court might do this even if it accepted the prevailing view that the good faith duty outside of bankruptcy is exceedingly narrow. Striking down such a transaction is not the same as second-guessing contractual language or introducing fuzzy notions of fair play into dealings between adults. The new loan Revlon needed to transform the minority coalition into a majority was no loan at all. No one extends a dollar at arm’s length in return for a promise that trades at only 43 cents on the dollar. A “loan” that no sane commercial actor would make and that disappears in a nanosecond should not enjoy the voting power that comes with an actual loan in which creditors are putting their capital at risk.

Any approach to creditor-on-creditor aggression should pay attention to the stakes. When the expropriation takes place within a class of creditors [Pg. 246]and can be anticipated and priced in advance, the welfare costs should be modest. The risk is diversifiable. The mere fact that some creditors win and some lose should not trouble anyone greatly. The largest risk in fact may be the rent-seeking problem. The amounts one creditor can capture from another are sufficiently eye-popping that they are worth fighting for. And once they are worth fighting for, creditors are willing to spend resources, both attempting to capture value from others and guarding against others taking it from them. These are deadweight losses. The problem of rent-seeking and creditor-on-creditor aggression is particularly salient in bankruptcy, and the next Part of this paper addresses it.

III. Creditor-on-Creditor Aggression in Bankruptcy

Negotiations among creditors and their common debtor also take place in bankruptcy. Bankruptcy judges must take several steps back, take stock, and ensure everyone is acting in good faith. As the Supreme Court explained long ago, bankruptcy judges cannot be “mere silent registrars” of the debtor’s plan. Plans are “absolutely subject to the independent judgment of the court,” and the court must “see to it that all equitable rights in or connected with the property are secured.”[71]

Ensuring creditors act in good faith, however, presents a strikingly different challenge in bankruptcy. In contrast to a credit agreement that provides an elaborate mechanism to limit the debtor’s freedom of action, the only explicit directive for the judge is a provision of the Bankruptcy Code that requires the judge to determine whether, in a particular case, the debtor’s plan provides the “same treatment” for each claim. The minority’s protection lies not in contractually created sacred rights, but in a simple rule that all claims receive the same treatment. Section 1123(a)(4) works in the same fashion as pari passu clause in a contract, but parties have no ability to refine it as they can with sacred rights in a loan agreement.[72] The inflexibility of § 1123(a)(4) or any other simple statutory rule puts pressure on the judge to exercise more oversight over bargaining between the parties.

LATAM Airlines illustrates how controlling creditors can extract value from minority creditors in bankruptcy.[73] The debtor’s plan required raising additional cash. To ensure the availability of the funds, the debtor approached a coalition of creditors whose support was critical to ensuring that the plan could be confirmed without an expensive cramdown hearing. [Pg. 247]The plan offered a treatment for all the claims in the class, whether those holding it belonged to the coalition or not. In addition, the debtor promised to pay the majority coalition a fee in return for their willingness to invest new cash if no one else appeared.

In the language of investment bankers, the majority coalition agreed to “backstop” the rights offering. For the backstop, the controlling creditors pocketed a fee of several hundred million dollars. The Bankruptcy Code is organized around claims, not creditors, and courts interpreting § 1123(a)(4) have focused narrowly on the treatment of the claim itself, not whether the creditor was receiving a benefit wearing some other hat.[74] Even though not everyone in the class had the chance to participate in the backstop, every claim in the class received the same payout. The controlling coalition argued that the plan therefore met the requirement that each claim receive the same treatment as § 1123(a)(4). When seen through this lens, it was not relevant that only some of the class was being paid for providing a backstop.

The minority that was not given the chance to participate in the backstop, however, argued the plan was not proposed in good faith. Although each step (identical treatment under § 1123(a)(4) and an arms’ length payment for a backstop) might appear regular, the effect of the entire transaction was to transfer wealth from the minority to the majority. The amount paid on each claim was small, and the fee for the backstop was large.[75] Looking beyond form to substance, the majority coalition was in effect taking a side payment in return for its support of the plan. The majority’s vote was being bought, and the money to pay for the vote was coming out of the pockets of the minority creditors. Such a plan, the minority creditors argued, was not made in good faith.

Exactly how the judge should police creditor coalitions in bankruptcy, however, is not easy to say. A backstop provides certainty. Like any other put option, however, the debtor must pay for it, and those most willing to sell it may be those who know the debtor best. The debtor’s best option may be to buy the backstop from a preexisting creditor.

[Pg. 248]Buying a backstop from a related party would not be a problem if backstops were easy to value, but they are not. A court cannot easily tell whether the majority coalition is being paid a sensible price for the risks it is bearing. Pricing a backstop must be done without many familiar benchmarks. Something as bespoke as a backstop for a plan of reorganization does not have a readily available market price. The value of a distressed firm fluctuates in a way the value of a healthy firm does not. This makes valuing a backstop hard. It may not be possible to know whether the debtor’s payment is an underhanded way to pay the majority coalition for its support of the plan.

Many of these difficulties would disappear if the minority were allowed to participate on equal terms. Indeed, if bargaining were frictionless, it would seem affirmatively desirable to give everyone a chance to participate on equal terms. Bargaining, however, takes place in a world in which there are many frictions. The value of a put changes continuously. The majority coalition that commits to providing a backstop at a set price when the plan in first formed is not in the same position as a late-arriving minority creditor that wants the same chance to buy the backstop at some later time.

The Bankruptcy Code and the decided cases give each judge wide latitude. It is possible that judges can craft a variety of different approaches to the problem, each of which is equally effective. The judge is a repeat player. Some judges might establish hard rules, believing that these provide the best environment for parties seeking to reach deals with one another.

For example, the judge might make plain her hostility to any plan in which the debtor offers to sell backstops to a majority coalition without giving the minority a chance to participate in the sale on the same terms. If a judge can lay out such ground rules, parties will not spend resources trying to use backstops as a back channel to expropriate value. Some valuable backstops might be lost in the process, but it may come at acceptable cost if the judge is otherwise adept at motivating the parties to get a deal done.

In assessing whether it makes sense for a bankruptcy judge to impose such a rule, however, it is important to recognize that preventing backstops from enriching the majority coalition at the expense of the minority is not a goal in itself. Once a class of claims held by sophisticated creditors is in bankruptcy and the class as a whole is enjoying its nonbankruptcy entitlements, it may be more important to avoid costly rent-seeking among the players than ensure that the court scrupulously respects the individual entitlements of each creditor holding claims in the class.[76] For this reason, [Pg. 249]an outright ban on backstops may make sense only to the extent it establishes an environment in which parties can bargain with each other with a minimum of frictions and rent-seeking. To achieve this goal of settling on value-maximizing plans quickly and cheaply, a bankruptcy judge, far from banning backstops that enrich a majority coalition, might do exactly the opposite. A bankruptcy judge might be able to embrace backstops warmly, permit massive transfers, and still be every bit as effective with respect to what matters as a judge who flatly prohibited them.

 Consider an extreme case. The bankruptcy judge appoints a mediator to oversee the negotiations, and the mediator, a well-respected retired bankruptcy judge of impeccable integrity, announces she will hold a meeting in a conference room at a particular time and place. Any creditors who are interested can line up outside the conference-room door at the appointed hour. The mediator warmly greets the creditors when she arrives, and she randomly invites individual creditors into the conference room until those in the room hold two-thirds of the claims in the fulcrum class.[77] Once the creditors are inside the conference room, the mediator tells them a deal will be struck that afternoon. If any creditors do not want to strike a deal, they are free to leave, and the mediator will invite others to replace them.

The bargaining begins, and a plan emerges. Those managing the company will be given a soft landing. The plan gives each creditor a choice between a certain number of cents on the dollar in cash or a share of stock in the new company for each dollar it is owed. To make this choice possible, the debtor needs a backstop so it will have the cash to the extent some of the existing creditors opt for the cash rather than the stock. Those in the room agree to provide the backstop for a few.

This plan has attractive features for outsiders, such as small trade creditors and even CLO managers. Outsiders have no easy way to tell how much the stock in the firm is worth. The plan proponent can say it is worth a certain amount, but talk is cheap.  Moreover, an outsider may have no interest in holding stock in the reorganized company. A plan that gives the outsider a choice reduces doubt that the outsider has about value of what it is receiving at the same time it makes it easier for the creditor to liquidate its position.

But everything turns on how much the debtor is paying the creditors inside the room for the backstop. Benchmarking all transactions against the market is a reliable way to prevent mischief, even if the division of the bankruptcy estate proper is not the issue at hand. A debtor should sell or buy assets only at market. The sale of any interest in a reorganizing [Pg. 250]company, whether it is a conventional equity or debt instrument or whether it is a more exotic beast like a backstop, is suspect if it is made for less than its market price.[78] But the judge has only a limited ability to put a fair price on backstops.

The market price of a backstop turns critically on volatility. Sooner or later, the ability to rely on market prices runs out. At this point, bankruptcy judges have no choice but to oversee a bargaining game in which the players know more than they do. It may not be possible both to capture the benefits of such things as backstops and prevent value leakage. Again, those in the conference room are neither Boy Scouts nor Good Samaritans. Creditors are lining up outside the door because they think there is a good deal to be had. One can be confident that some value is being transferred. The creditors in the majority end up with more and those in the minority end up with correspondingly less.

This might seem a reason for the judge to refuse to confirm the plan. But diversion of value should not necessarily be a showstopper. Under this hypothetical, a reorganization was put together in a single afternoon in a conference room. There is no risk assets will be inefficiently deployed. The conference room creditors become the owners of the firm, and they have every incentive to ensure that the assets of the firm are put to their highest and best use. And there is no violation of the absolute priority rule. The equityholders were wiped out. Crucially, no creditors spent any money jockeying for position apart from the cost of lining up outside the conference room.

To be sure, some similarly situated creditors received more than others. Such non pro rata distributions violate deeply seated bankruptcy norms. In a perfect world, we would not tolerate them. But our world is not perfect. This hypothetical reorganization was successful by every other metric. It was quick, cheap, and successful. Moreover, the mediator chose the members of the winning coalition randomly. The creditors who were part of the majority coalition were simply lucky.

The dissenting creditors do lose out, but in this respect their position is no different from anyone who loses a lottery. The only difference is that the person drawing the balls from the urn was a court-appointed mediator. Because all the creditors are at the same priority level, there should be no effect on the willingness of creditors to invest in the first place. Before the fact, each was as likely to be invited into the conference room as any other.

[Pg. 251]The point here is not to suggest such heavy-handed mediation is the best way to reorganize a distressed firm. Instead, it is to underscore that the enterprise in bankruptcy when it comes to policing coalition formation is altogether different from the challenge outside of bankruptcy. Each bankruptcy judge should be predictable and consistent and ensure a process that leads to speedy and effective reorganizations, but it may not be essential that each bankruptcy judge does this in the same way. In the end, the most desirable outcome might be a world in which each bankruptcy judge sets her own rules within established bounds, just as each referee has her own practices about what conduct merits cards and what does not. What may be optimal is an unusual combination of standards and rules—entrusting bankruptcy judges with discretion, but with each judge using this discretion to lay down her own clear rules.

This idea, however, that bankruptcy judges have alternative paths to minimizing creditor-on-creditor aggression must be qualified in a significant way.[79] The Bankruptcy Code encourages parties to negotiate with each other in advance of bankruptcy. Before the petition is even filed, parties can agree to the shape of the reorganization. The debtor and key creditors agree on a general structure for the reorganization, the key components of the plan, and a pathway for gaining additional support. The agreement both lays out the details of a plan and sets the timetable for the reorganization. A controlling coalition and the debtor can, however, use such plans to tie the judge’s hands.

A key component of the pre-bankruptcy planning is the debtor-in-possession financing. A successful reorganization depends critically upon the debtor lining up financing for the reorganization before it begins, and this often requires the help of the controlling creditor coalition. This coalition can reserve the right to terminate the financing if the bankruptcy judge refuses to approve a generous backstop.

Once the backstop is part of a restructuring support agreement, the judge’s decision to reject the backstop has collateral consequences. If the backstop is rejected, the debtor risks losing its financing and the ability to conduct its operations going forward. The debtor might have negotiated a different financing package, but when the matter is presented to the bankruptcy judge, it is too late. A refusal to approve the restructuring [Pg. 252]support agreement and the backstop it envisions may sink the Chapter 11 before it even starts.

It is hard for bankruptcy judges to prevent parties from painting them into such a corner. Bankruptcy judges do a tolerably good job of policing misbehavior over the parties after the bankruptcy begins, but incentivizing good behavior before the fact is another matter. In theory, individual bankruptcy judges might develop strategies to keep from being put in such positions. Among other things, bankruptcy judges can develop a reputation for calling the bluff of creditor coalitions that try to force their hands. After all, it is the lender, and not the judge, who stands to lose if the reorganization fails and the debtor is liquidated.

IV. Conclusion

The new arena for gladiatorial combat in the reorganization world lies in the shadow of reorganization. Parties jockey for position and attempt to form coalitions between themselves and between themselves and their common debtor. The winners stand to benefit handsomely once the reorganization begins. No one should shed any tears for the losers in this game of thrones. The losers, like the winners, are overwhelmingly sophisticated investors who can anticipate the costs of such combat and price it accordingly. But such conflicts produce deadweight loss. The need to keep such mischief in mind is one more burden that bankruptcy judges must bear.

There are no silver bullets here, but recognizing the fundamental difference between contractual provisions agreed upon at the outset and legal rules that govern negotiations once bankruptcy begins may be a useful place to start. Contract terms respond to market forces over time. Judges are well-counseled to put on narrow visors when interpreting contracts that are themselves commodities that are bought and sold in a market. They should be reluctant to exercise much discretion when it comes to interpreting them. Policing negotiations in bankruptcy, however, is a different matter altogether. Statutory commands, while similarly subject to interpretation, do not change dynamically the way contracts do. A contractual pari passu clause can evolve, but the statutory requirement that claims be treated the same is essentially frozen. Hence, a judge should have more latitude to find treatment unequal in interpreting § 1123(a)(4) than a corresponding terms in a contract.

Similarly, assessing whether past behavior violated the contractual duty to act in good faith is altogether different from the Bankruptcy Code’s mandate that plan negotiations be undertaken in good faith. When a new form of mischief is discovered, the contract can change quickly to prevent it, but the statute changes slowly, if at all. Controlling misbehavior once [Pg. 253]contractual constraints disappear is a bigger problem and requires a greater exercise of judicial discretion.

Discretion is necessary because it is not possible to advance the competing goals of bankruptcy simultaneously. We should not expect bankruptcy judges to make the necessary trade-offs the same way. The available benchmarks—even those of the market—often fall short. And the goals often stand in tension with one another. Resolving such tensions is the Sisyphean challenge from which judges cannot escape.