Printer Friendly

A new governance structure for corporate bonds.

This article proposes a new governance structure for publicly issued corporate bonds. Ownership of public bonds is both fluid and dispersed. Fluidity and dispersion generate benefits: They increase the liquidity of public bonds and make their risk more easily diversifiable. By the same token, however, fluidity generates a bonding problem and dispersion generates a collective action problem. In the context of public bonds, these problems increase the agency cost of debt and thus lower the overall value of the company. Private debt, the ownership of which is neither fluid nor dispersed, lacks easy diversification and liquidity, but also entails lower agency cost of debt. The new governance structure Professors Yakov Amihud, Kenneth Garbade, and Marcel Kahan propose is designed to overcome the collective action and bonding problems while retaining the easy diversifiability and liquidity associated with public bonds. The centerpiece of this structure is a new type of bondholder representative, the supertrustee. In contrast to the conventional indenture trustee, the supertrustee will be given the authority and the incentives to monitor the company actively, as well as to enforce and, where appropriate, to renegotiate a bond's covenants on behalf of public bondholders. Because the supertrustee is a single entity, and because the company will have some power over who is appointed as supertrustee, this structure resolves the collective action and bonding problems presently associated with dispersed and fluid ownership of public debt. At the same time, because public bond indentures will contain covenants similar to those presently contained in private debt, agency costs are reduced.


Over the last two decades, "corporate governance" has become an important focus of academics and policymakers. Newspapers editorialize on corporate governance;(1) executives, investors, and judges deliver speeches on corporate governance;(2) international conferences are devoted to corporate governance;(3) the American Law Institute drafted "Principles of Corporate Governance;(4) and countless articles analyze, compare, and criticize the corporate governance system(5)--in short, a "corporate governance movement"(6) has been sweeping through boardrooms, the halls of academia, and the popular press.

The term "corporate governance" conventionally refers to the legal and economic structures that mediate the relationship between a firm's shareholders and its managers. Companies, however, raise substantially more capital from selling debt than from issuing stock,(7) and the interests of creditors and stockholders often conflict. There is, therefore, a second dimension to corporate governance: the legal rules and economic arrangements that mediate the relationship between the firm and its creditors. This raises the question, largely ignored in the present debate,(8) of how one should design the governance structure of debt.

In discussing corporate debt, one needs to distinguish between publicly issued corporate bonds on the one hand and "private" debt--bank loans and privately placed bonds--on the other. Ownership of public corporate bonds is both dispersed, in that many different investors hold bonds of a single issue, and fluid, in that ownership of bonds changes over time and the company has little control over the changing identities of its creditors. This dispersion and fluidity in ownership creates both advantages and disadvantages. On the plus side, dispersion and fluidity make it easier for investors to diversify their holdings and increase the liquidity of their securities. On the minus side, dispersion results in a collective action problem (familiar from the shareholder context)(9) and fluidity results in a "bonding" problem (which has no shareholder analog).(10) Both of these problems raise the agency costs of debt. Private debt, which is neither dispersed nor fluid, lacks easy diversifiability and liquidity, but also entails lower agency costs.

This article proposes a substantial reform--in fact, an innovation--in the governance structure of public corporate bonds which overcomes the collective action problem and the bonding problem, but which retains the easy diversifiability and liquidity of public corporate bonds. We argue that this new structure is, for many companies, more attractive than either of the two existing schemes and that those companies can increase their aggregate market value by adopting the proposed governance scheme.

The proposed structure includes the following features:(11)

* A strengthening of the substantive rights of public bondholders by the inclusion of more and tighter covenants, thereby reducing agency costs.

* The appointment of a newly structured entity--the "supertrustee"--with the duty and power to actively monitor, renegotiate, and enforce bond covenants, and with access to confidential company information, thereby overcoming the collective action problem.

* A weakening of the procedural rights of bondholders by vesting in the supertrustee the exclusive authority to renegotiate and enforce covenants, stripping bondholders of the power to give directions to the supertrustee, and curtailing their power to replace the supertrustee, thereby overcoming the bonding problem.

* An incentive-based compensation scheme for the supertrustee, coupled with business-judgment-rule type protection for renegotiation and enforcement decisions and with capped liability for failure to monitor compliance with covenants, thereby addressing agency problems created by the supertrusteeship.

We also explain why having bondholders monitor a supertrustee, which in turn monitors the company on behalf of the bondholders, is superior to having bondholders monitor the company directly.

Two recent developments may have greatly increased the benefits of supertrustee bonds. First, the market for publicly issued "junk" bonds--bonds issued by companies with high credit risk--has grown substantially. Junk bonds entail much larger agency costs than investment-grade bonds. Nevertheless, the same governance structure for investment-grade bonds, which is not well designed to control agency costs, has been haphazardly applied to junk bonds.(12) The supertrustee governance structure, which is designed to reduce agency costs, would be especially beneficial for such junk bonds. Second, the growing emphasis on encouraging managers to maximize shareholder wealth, although on the whole beneficial, has aggravated the agency cost of debt problem.(13) In the past, creditors were to some degree protected by managers' self-interested risk aversion and their desire to reinvest (rather than distribute) cash. As this protection has eroded, reducing agency costs directly, as our proposal does, has become increasingly important.

We stress that the supertrustee proposal is market oriented rather than regulation oriented. We do not advocate a change in the laws to force companies to adopt our proposed governance scheme. While our scheme would benefit many companies, there are doubtless other companies that would prefer one of the existing schemes. Ultimately, the choice of governance structure should be made by the market participants whose money is at stake--by companies and bondholders--rather than by academics and legislators. We do, however, advocate removing regulatory barriers created by the Trust Indenture Act that impede the implementation of the supertrustee proposal.

Part I of the article examines the existing governance structures for private debt and for publicly issued corporate bonds and assesses the advantages and disadvantages of each structure. Part II provides the details of our proposed new governance scheme, explains their rationale, and analyzes how the supertrustee differs from the indenture trustee, which presently serves, in some limited respects, as a bondholder representative. Part III discusses issues related to the implementation of our proposal.


This Part examines the existing governance structures for corporate debt. The first section describes why debt generally introduces economic inefficiencies into the management of a corporation and explains how corporations and their creditors mutually benefit from loan agreements and bond indentures designed to mitigate such inefficiencies. The second section analyzes and contrasts the institutional characteristics, covenant structures, and behavior of creditors in private and public debt markets. In the third section, we argue that the differences between private and public debt give rise to two separate governance schemes, each with its own advantages and disadvantages. The proposal presented in Part II seeks to establish a new governance structure for public debt that grafts the advantages which we associate with private debt onto the governance structure for public bonds.

A. Agency Costs of Debt and Debt Covenants

Credit constitutes a large fraction of the capital companies use to finance their operations. About $3 trillion in corporate debt was outstanding in 1996. That debt constituted 31% of the capital structure of U.S. companies.(14)

But once a company has taken on debt, conflicts of interest commonly arise between the creditors and the company's shareholders and managers.(15) Managers serve the interests of shareholders by acting to maximize the value of the company's common stock. But some actions that enhance the value of equity reduce the value of the company's debt. Such actions include cash distributions to shareholders (through dividend payments(16) or stock repurchases), spin-offs to shareholders of the common stock of lightly leveraged subsidiaries owning valuable corporate assets,(17) dealings between shareholders and the company on terms preferential to shareholders, investments in projects with greater risk than originally anticipated by creditors, and financing new projects with debt rather than equity.(18) Shareholders will want the company to take such actions even if they reduce the value of the company's debt by more than they increase the value of equity--that is, even if they lower the overall value of the company. Losses in aggregate corporate value attributable to actions by shareholders seeking to advance their parochial interests while in control of a leveraged enterprise are commonly referred to as agency costs of debt.(19)

We illustrate this with an example where, for simplicity of exposition, valuation is determined solely by the expected value of cash flows (ignoring investors' risk preference and time preference). Consider a company with $100 in cash (and no other assets) and $80 in short-term debt, so the debt is currently worth about $80 and the stock is worth $20. If the company invested all of its cash in a project with a 50% chance of yielding $10 and a 50% chance of yielding $150, shareholders would benefit--the value of the equity would increase from $20 to $35(20)--and creditors would be hurt--the value of the debt would decline from $80 to $45.(21) More significantly, the value of the company as a whole would decline from $100 to $80.(22)

From the perspective of economic efficiency, a company should be managed to maximize the aggregate value of its debt and equity.(23) Minimizing agency costs is thus an important objective in designing a governance structure for corporate debt.

To reduce the agency costs of debt, loan agreements contain covenants(24) that limit the ability of shareholders to run a company in a parochial manner.(25) For example, such covenants can limit the dividends a company may pay(26) or the amount of new debt it may incur.(27) Covenants that reduce the agency costs of debt are ex ante in the interest of shareholders: creditors will be less concerned that the company will take actions that reduce the value of their debt and will lend to the company at a lower rate of interest.

But covenants also entail costs: the costs of monitoring compliance with the covenants, the costs of enforcing the covenants, and, most importantly, the costs stemming from the limitations on the company's actions imposed by the covenants. The latter costs arise because covenants are not (and cannot be) fine-tuned to restrain only actions that reduce the aggregate value of the firm. Most covenants turn out to restrict some actions that increase the value of equity by more than they reduce the value of debt--and which therefore increase the value of the firm as a whole. In such cases, the company foregoes an increase in value because of the existence of the covenants.(28)

Consider, for example, a covenant restricting dividend payments. Dividend payments to shareholders harm creditors because they reduce the assets of the company available to repay debt. Shareholders benefit, at the expense of creditors, from the payment of dividends even if such payments reduce the aggregate value of the firm. But if a company does not have profitable investment opportunities, distributing cash may increase the value of the company because money retained in the company may be invested in value-decreasing projects--such as a conglomerate acquisition that reduces managerial focus and increases managerial overhead.(29) A covenant restricting dividends may thus be either overly lax (allowing shareholders to distribute cash even when it reduces debt values by more than it increases equity values) or it may be overly tight (prohibiting distributions of cash even when available investment projects are not value-increasing).(30) And although covenants can be waived or changed, renegotiation is itself costly--sometimes prohibitively so.(31)

More and tighter covenants are therefore not necessarily preferable to fewer and looser covenants. The optimal pattern of covenants will depend on balancing the benefits of a covenant--restricting actions that would reduce firm value--with the costs of a covenant--monitoring costs, enforcement costs, renegotiation costs, and the costs of restricting actions that increase company value. As we shall explain, these costs vary significantly across different categories of debt. The characteristics of the variation affect the nature, frequency, and tightness of the covenants that appear in different debt contracts.

B. Private Debt and Public Bonds

There are three main categories of corporate debt that rely on extensive written agreements between borrowers and lenders: bank debt, bonds placed privately with large institutional investors (primarily insurance companies),(32) and corporate bonds sold to the public in registered public offerings.(33) For our purposes, we will combine bank debt and privately placed debt and refer to such debt as "private" debt.(34)

In this section we discuss first some differences in the institutional characteristics of the private and public debt markets. We then examine the differences between covenants typically included in private loan agreements and in public bond indentures and the related differences between the way private lenders and public bondholders behave after credit has been extended.

1. Institutional characteristics.

The markets for private debt and public bonds differ in the types of borrowers, average loan size, and, most significantly for our purposes, the structure of the lending relationship.

Borrower characteristics. Companies that borrow in the public corporate bond market are larger than those that borrow in the private debt market. They have lower research and development (R&D) expenditures (relative to total revenues) and more fixed assets (relative to total assets) than borrowers in the private market. These characteristics suggest that companies that borrow in the public bond market have less information asymmetry problems: more and more reliable information exists for large companies with relatively low R&D expenditures and relatively more fixed assets. In addition, companies borrowing in the public market are less leveraged and have greater interest coverage ratios, suggesting that they are more credit-worthy than those borrowing in the private market.(35)

Loan size. There is a major difference in the average size of loans in the public and private markets. In 1989, for example, bank loans were mostly below $1 million, with a median of $50,000 and a mean of $1 million. Privately placed debt was mostly in the range of $10 million to $100 million (80% of the debt issues), with a median of $32 million and a mean of $76 million. The size of public debt issues was mostly greater than $100 million (85% of the debt issues), with a median of $150 million and a mean of $181 million.(36)

Structure of the lending relationship. Finally, there are important differences in lender concentration and continuity of the lending relationship. Creditors in the private debt market are primarily banks and insurance companies. Usually, only one or a few of these lenders hold all of the private debt of any particular borrower. Most bank loans involve a lending relationship between a borrower and a single bank; privately placed bonds of a particular borrower are usually held by fewer than five lenders;(37) and even large, syndicated bank loans are often made by fewer than ten lenders.(38)

Financial institutions--insurance companies, banks, mutual funds, and pension funds--also hold a predominant share of public bonds.(39) Public bonds of a single issue, however, are held by a much larger number of investors. Exact data are difficult to obtain, but we estimate that bonds of a single issue are generally held by 40 to 100 owners.(40) Thus, ownership of public bonds is substantially more dispersed(41) than ownership of private debt (though it is substantially less dispersed than ownership of stock in publicly held corporations).(42)

Public bonds also entail a more fluid lending relationship than private debt. While private debt trades relatively rarely(43) due to explicit and implicit restrictions,(44) public bonds trade actively in over the counter and negotiated transactions.(45) The identities of the holders of a particular public issue thus change over time, sometimes quite dramatically.(46)

Implications of the dispersion and fluidity of ownership of public bonds. The difference between public bonds and private debt in terms of dispersion and fluidity of ownership has important implications. On one hand, dispersion and fluidity create two problems affecting the relationship between the company and public bondholders, both of which increase the agency costs of debt associated with public bonds. The dispersion of public bondholders creates a collective action problem, similar to (though less severe than) the collective action problem affecting stockholders in a public corporation. Because each public bondholder owns only a relatively small fraction of the company's bonds, the incentive to obtain information about the company, to monitor compliance with the bond indenture, and to assess whether and what kind of enforcement action should be taken if a covenant is breached is reduced. Correspondingly, the company faces greater costs in obtaining the consent of dispersed bondholders to an amendment of a covenant or waiver of a breach.

The fluidity of ownership (as well as the fact that public bonds are initially offered in public capital markets) means that a borrower has virtually no control over who owns its public debt.(47) As we shall argue below, in the private debt market, the reputation of lenders is an important factor for the covenant structure of private debt agreements as well as for renegotiation and post-breach behavior. Public bondholders (who cannot reputationally bond themselves) and companies in the public debt market therefore do not obtain the advantages that flow from lender reputation.

On the other hand, the fact that public bonds trade freely and in small units in secondary market transactions confers advantages on those bonds relative to private debt. Most importantly, corresponding to the greater fluidity in the creditor relationship, public bonds enjoy greater liquidity; and, corresponding to the greater dispersion in ownership, public bonds enable easy diversification. In addition, as a result of the availability of market prices for public debt, the access to information concerning market appraisals of corporate decisions is enhanced.

Liquidity of an asset reflects the ease of converting the asset into cash and converting cash into the asset.(48) Liquidity is a valuable attribute of a financial instrument and investors are willing to pay for it.(49) A study by Silber, for example, found that the price of illiquid "restricted stock" that could not be freely sold in the secondary market was about one third lower than the price of equivalent unrestricted stocks.(50) While the liquidity differential between private and public issues is likely to be lower for private bonds and bank debt than for restricted stock,(51) yields on private debt do include some premium to compensate for illiquidity.(52)

Moreover, relative to private debt, public bonds facilitate risk diversification because they are sold in smaller units (typically $1,000 principal value) and because their greater liquidity enables a holder to adjust her portfolio in response to variations in risk following purchase. An investor can thus diversify away the firm-specific, or idiosyncratic, risk associated with the bonds of any single company. Since investors are risk averse, the ability to reduce risk by diversification enhances the value of public bonds, thus lowering their yield.(53) In contrast, private debt comes in a unit size--often millions of dollars--that makes it difficult for all but the largest institutions to hold diversified portfolios of such debt.(54) The small unit size of public bonds also enhances the bonds' liquidity because smaller units may be less costly to trade and the investor base may be broader.(55)

An additional, less prominent advantage of public debt is the benefit to issuers of a market assessment of managerial policies.(56) The secondary market prices of a public bond reflect the views and opinions of a wide range of investors.(57) Having access to these views, as mediated through the market mechanism, is valuable to managers.(58) Although managers usually have better information than outside investors about internal corporate affairs, the ultimate effect of many corporate acts depends on their interaction with the economy at large. Outside investors--especially professional analysts with substantial industry expertise--may be better able than managers to gauge this interaction. Corporate managers can thus obtain informed assessments of their decisions by observing the secondary market prices of their bonds.(59) This information may be valuable for the managers and could lead them to revise their policies directed towards maximizing shareholder value.(60)

2. Covenant structures.

In addition to differences in institutional characteristics, there are also significant differences in the covenant structures of private debt and public bonds. Private debt contains more extensive covenants and carries more extensive monitoring rights than public bonds.(61)

Covenant protection. Covenant protection is substantially greater in private debt than in public corporate bonds.(62) Three elements contribute to the greater protection: greater covenant frequency, greater covenant tightness, and the presence in private debt of certain types of covenants that are rarely found in public bonds.

First, covenants that are found in both private loan agreements and public bond indentures are found more frequently in the former agreements. For example, "restricted payment" covenants--which limit the amount a company can distribute through dividends and stock repurchases--are found in over 90% of private debt agreements but in only 20% to 50% of public bond indentures.(63)

Second, if a public bond indenture contains a particular covenant, the covenant is usually looser than the same covenant in a private loan agreement. Consider again the restricted payment covenant, which usually limits the cumulative amount of dividends a company can pay to the sum of (i) a specified fraction of the company's cumulative profits and (ii) a specified initial reservoir.(64) Dividend restrictions in public bond indentures permit companies to pay out a larger fraction of cumulative earnings and establish a larger initial reservoir than dividend restrictions in private debt.(65)

Third, private loan agreements contain "state-of-the-firm" covenants of a type that are rarely found in public bonds. These include minimum financial ratio covenants--for example, current ratio and interest coverage ratio covenants, and minimum net worth covenants.(66) Unlike the "discretion-limiting" covenants found in both bond indentures and loan agreements (such as restricted payment covenants), state-of-the-firm covenants concern events (such as business losses) that are often not subject to direct managerial control.(67) As a result, violations of state-of-the-firm covenants are much more common than violations of discretion-limiting covenants.(68)

The purpose of a state-of-the-firm covenant is substantially different from the purpose of a discretion-limiting covenant. Some corporate actions that harm creditors--for example, actions that increase the business risk associated with the company's operations--are difficult to specify contractually: It is difficult to specify ex ante what actions actually increase the volatility of the value of an enterprise, and even harder to determine whether such actions harm creditors (if at all) more than they benefit shareholders.(69) State-of-the-firm covenants give creditors a qualified de facto opportunity to prevent such actions in cases where they are most likely to occur and most likely to be harmful to creditors (or to react to them before they result in more serious harm).(70)

Consider, for example, a minimum interest coverage ratio. As the interest coverage ratio of a firm falls, shareholders have an increasing economic incentive to increase the riskiness of the firm's operations, and creditors become progressively more exposed to harm from such risk shifting. But once the company breaches the specified minimum threshold for its interest coverage ratio, creditors obtain a de facto veto over any corporate action. If management wants to engage in an action that increases the company's business risk, even one not specifically proscribed by the loan agreement, creditors can invoke the breach of the interest coverage ratio covenant to forestall the action.

Thus, state-of-the-finn covenants are not meant to be strictly enforced--in the sense that a creditor will call a loan as soon as a state-of-the-firm covenant is violated.(71) Rather, they give creditors a greater voice in the management of the finn in circumstances where the creditors are particularly exposed to exploitation by the stockholders.

Monitoring Rights. Compared to holders of public corporate bonds, holders of private debt also have access to substantially more information to determine whether a covenant has been breached.(72) Public bondholders must generally rely on public documents--for example, annual and quarterly reports to the SEC--and on annual "compliance certificates" in which the company self-reports any covenant violations. In contrast, holders of privately placed bonds have access to such public documents as well as to confidential company information, such as reports submitted by the company's accountants, financial statements for subsidiaries and operating divisions, and any other financial information that they may reasonably request. They also have the right to inspect the company's properties, to examine its books and records, and to discuss its affairs with its officers and accountants.(73) In addition, companies must generally self-report any covenant violation promptly after it occurs and must provide creditors with quarterly compliance certificates that (unlike those for public bonds) contain detailed calculations showing whether or not a covenant has been breached.

3. Creditor behavior.

Differences between private debt and public bonds in terms of institutional characteristics and covenant structures lead to differences in the way creditors act after a credit has been extended. Compared to holders of public bonds, lenders in private transactions engage in more monitoring; they agree more frequently to covenant amendments sought by the borrower; and, following a covenant violation, they take more graduated enforcement actions.

Intensity of monitoring. Creditors who hold private debt engage in more monitoring, and thus are better informed about the company than creditors who hold public bonds.(74) This higher level of monitoring is due to the factors discussed above: more extensive covenant protection, which requires more monitoring;(75) more extensive monitoring rights; and greater concentration and stability of debt holding, which increases incentives to exercise monitoring rights.

Frequency of amendments. Contractual amendments are much more common for private debt than for public bonds. As a result of the differences between private and public debt, amendments of private loan agreements generate greater benefits and entail lower costs than amendments of public bond covenants.(76) Specifically, since private debt contains more numerous and more rigidly constrained covenants, private debt covenants are more likely to restrict an action that increases company value and for which the company will seek creditor consent. In addition, the state-of-the-firm covenants commonly found in private loan agreements can give rise to unavoidable, or non-discretionary, violations, thus necessitating more frequent amendments.(77) Finally, the fact that holders of private debt are more concentrated and better informed about the company facilitates the renegotiation of the covenants and makes it cheaper to obtain creditor consent to an amendment of a private loan agreement.(78)

Renegotiation and post-breach behavior. The third difference between the behavior of holders of private versus public debt relates to the way creditors act when a covenant is breached, or when an amendment is sought prior to taking an action which (absent the amendment) would precipitate a breach.

Creditors have substantial bargaining power in these circumstances. If a covenant has been breached, they can threaten to accelerate their loan or to take other actions that could impose substantial costs on the company. If an amendment is sought, they can threaten to withhold consent. Creditors can use this bargaining power opportunistically. They can, for example, use a covenant breach as a pretext to accelerate a fixed rate loan if interest rates have increased since the loan was made. Alternatively, they can demand a consent payment in excess of any losses they would suffer from an otherwise prohibited action. If creditors regularly use their bargaining power in such an opportunistic fashion, tight covenants could entail substantial costs to the issuer. As a result, companies would be reluctant to take actions that could result in a covenant breach, or that could put them in a position of having to seek an amendment, even if such actions increase the overall value of the firm.

Lenders in the private debt market generally do not use their bargaining power to extract excessive concessions for amendments, nor do they opportunistically accelerate loans after interest rates have increased. To the contrary, evidence shows that they often consent to amendments and grant temporary or permanent waivers gratuitously--without demanding any concessions or payments by the borrower. Or, they may merely step up their monitoring of the borrower's business activities, without taking any immediate enforcement action.(79) These lenders act in this nonopportunistic fashion because they want to attract new borrowers in the future. Since holders of private debt, i.e., banks and insurance companies, are professional lenders and since borrowers control who they borrow from, lenders that do not act in this fashion would incur costs from the loss of future lending opportunities that may exceed the benefits to be derived from any immediate payment or concession from the borrower.(80)

In contrast, holders of public corporate bonds do not have incentives to act nonopportunistically. Public bonds are initially sold and subsequently traded in public capital markets. Issuers thus have no control over who owns their public bonds. Accordingly, unlike in the case of private bonds, companies virtually always have to make a payment to get holders of public bonds to consent to a covenant amendment.(81) Moreover, consistent with public bondholder opportunism, evidence shows that this consent payment over-compensates public bondholders for the loss in bond value due to the amendment. That is, public bondholders do not just break even, but actually gain when the company seeks an amendment to their bond indenture.(82)

C. Summary of the Governance Structures of Private Debt and Public Bonds

The differences in institutional characteristics, covenant structure, and creditor behavior create separate governance structures for private and public debt. In the case of a conventional private loan, one or more professional lenders lends money to a borrower and holds the loan until it matures. The lender will expend resources in gathering information about the borrower and monitoring its compliance with the terms of the loan. If the borrower has breached or desires to breach a covenant, the lender is able to assess whether the breach is significant, whether an enforcement action should be taken, and whether renegotiation is in its interest. Whatever it decides to do, it does not require the consent of numerous co-creditors. The fact that the lender is a professional lender gives the borrower confidence that the lender will not exploit its bargaining position in an opportunistic fashion. Since monitoring and amending covenants is easy and the lender has incentives to act nonopportunistically, the loan agreement will contain extensive covenants and monitoring rights. Monitoring and covenants, coupled with a high ability to renegotiate, in turn reduce agency costs and prevent losses in value.

In the case of conventional public bonds, many independent bondholders own and trade bonds in an open market. This results in liquidity and easy diversifiability of firm-specific risks, but also creates a collective action problem and a bonding problem. Due to the collective action problem, bondholders have little incentive to monitor, to become informed about the company, or to decide what enforcement action to take if a covenant has been breached. Because of the bonding problem, they have little incentive to act nonopportunistically should they be asked to consent to a covenant amendment. Since monitoring, amending and enforcing covenants is difficult, the bond indenture will contain few covenants; and those that appear entail little monitoring, are rarely breached, and, if breached, can be properly enforced without intimate knowledge of the issuer. Such covenants, of course, lead to a smaller reduction in the agency costs of debt.

The choice between private and public debt depends on the relative advantages and disadvantages of participating in the respective markets. For some borrowers--for example, those whose debt is more likely to give stockholders more opportunities to exploit creditors, and those who are subject to greater information asymmetries--the benefits of greater monitoring, easier renegotiation, and graduated enforcement are relatively more important. Such companies generally borrow in the private debt market.

For larger, more credit-worthy companies with greater borrowing needs, the benefits of borrowing on the basis of an instrument which is liquid and easily diversified are relatively more important. Liquidity will be higher for a large bond issue than for a small one,(83) and larger loans are more amenable to diversification than smaller ones. These companies tend to borrow in the public debt market.(84)


The preceding discussion presented an analysis of the governance structures for private and public debt. The governance structure for private debt is designed to reduce agency costs, but in the process limits the liquidity and diversifiability of that debt. The governance structure of public debt facilitates liquidity and diversifiability of risk, but fails to reduce agency costs to the extent of private debt.

In this Part, we propose a new governance structure for public corporate bonds that is designed to replicate the lower agency costs of private debt while retaining the liquidity and easy diversifiability of public bonds. The keystone of this new structure is the supertrustee.

The supertrustee would serve as an agent of public bondholders, thereby substantially ameliorating the collective action problem. Specifically, the supertrustee would:

* actively acquire information about the borrowing company;

* monitor compliance with the bond indenture;

* renegotiate covenants when the company seeks an amendment; and

* decide whether and what kind of enforcement action to take when a covenant is breached.

Since the power to renegotiate and enforce covenants will rest with the supertrustee rather than the bondholders, and since companies will have some control over who can serve as supertrustee for their bonds, the supertrusteeship will also overcome the bonding problem. The covenant structure for a public bond with a supertrustee will, in turn, more closely resemble contracts currently used in private lending: The indenture will contain relatively strict covenants, including state-of-the-firm covenants that may be breached unintentionally; and the supertrustee will have extensive monitoring rights. The supertrusteeship will thus reduce agency costs for public debt. At the same time, because the debt is publicly issued and traded, it will retain the benefits of liquidity, diversifiability, and availability of information. In this way the supertrustee would enable an issuing company to enjoy many of the benefits of private debt as well as the benefits of public bonds.

The supertrustee that we propose is different from the indenture trustee presently associated with public bonds. While one purpose of the indenture trustee is to assist dispersed public bondholders to overcome the collective action problems described above, the indenture trustee lacks the authority and incentives to execute effectively the functions of active monitoring, enforcement, and renegotiation that we assign to the supertrustee. Specifically, as we discuss below, the indenture trustee is not obliged to engage in active monitoring, its compensation creates no incentive to expend effort, and it does not have the authority to renegotiate covenants. Finally, the indenture trusteeship is neither intended nor designed to reduce the bonding problem.

There are costs associated with executing the functions of the supertrustee. The decision to use the supertrustee governance structure therefore depends on a balancing of these costs against the benefits flowing from the reduction in agency costs. Each issuing company should voluntarily decide whether to use the supertrustee structure. The issuing company should also determine the specific features of the supertrusteeship, and these features ought to be fully disclosed in the prospectus so that a prospective bondholder can assess the governance structure before making an investment decision. Our discussion should be seen as commentary on how the supertrusteeship might be designed to accomplish the functions that we have identified, and not as a prescription for how it must be designed.

In the sections that follow, we describe in detail the appointment and replacement of the supertrustee, its duty to monitor the company, its power to renegotiate and enforce indenture provisions, its legal liability for failure to make proper renegotiation and enforcement decisions, its compensation, and its reputational incentives.

A. The Appointment and Replacement of the Supertrustee

The company should choose the initial supertrustee before the bonds are issued. Once the bonds are issued, the company should have no right, and bondholders should only have a limited right, to replace the supertrustee.

Having the company appoint, and making it difficult for bondholders to replace, the supertrustee might seem odd. Why should a corporate borrower appoint a supertrustee that will take its tasks of monitoring, renegotiating, and enforcing covenants on behalf of bondholders seriously?(85) The reason is, of course, the same as the reason why a company agrees to be bound by covenants in the first place: It is in the company's self-interest to do so. Bondholders will be aware of the identity of the supertrustee--just as they are aware of the extent of covenant protection--when a company issues its bonds. They will pay less for the bonds if the supertrustee is unreliable--just as they will pay less for the bonds if they are dissatisfied with the covenant protection. The company thus has an incentive to pick a supertrustee that represents the interests of bondholders effectively.

Moreover, giving the company substantial control over the identity of the supertrustee emulates the situation in private debt, where borrowers decide to borrow from a specific lender and where lenders rarely sell their loans. From the company's control flow incentives for lenders in the private debt market not to use their bargaining power to extract excessive concessions when a company seeks to amend a loan agreement or has breached a covenant. In addition, such control makes borrowers more willing to grant lenders access to confidential information. Such incentives and monitoring rights are important factors in reducing agency costs in the private debt market.

In the public market, of course, companies cannot control who owns their bonds. But the identity of the bondholders becomes less relevant when the supertrustee exercises power in their stead. Thus, the company's control of who acts as supertrustee is, for public bonds, the functional equivalent of the company's control over whom it borrows from privately.(86) Seen from this perspective, an extensive right of bondholders to replace the supertrustee would undermine the benefits that flow from the company's control.

One extreme way to limit the ability of bondholders to remove the supertrustee is to permit removal of the supertrustee only for cause--for example, if the supertrustee has grossly neglected its duties, acted in bad faith, or become subject to a disabling conflict of interest.(87) A less extreme way is to let bondholders periodically elect the supertrustee, but to place limits on who can be elected.(88) For example, bondholders could be required to select a supertrustee from a list of candidate institutions specified by the company when the bonds were issued.(89) The choice between these two (and any intermediate) structures depends on whether it is more important to provide incentives to the supertrustee to act in a nonopportunistic manner vis-a-vis the company or to provide incentives to the supertrustee to represent effectively the interests of bondholders.

B. The Duty to Monitor

The supertrustee will have a duty to monitor, actively and independently, the compliance of the borrower with the covenants in the bond indenture. In fulfilling this duty, the supertrustee will not be permitted to rely on the company's own possibly self-serving "compliance certificates" stating, without more, that the company has not breached any covenants. Rather, the supertrustee will be given access to, and will use, the same type of information as lenders in the private market, such as reports submitted by accountants, non-public financial data, inspection rights, and compliance certificates containing detailed calculations showing whether or not a covenant has been breached.(90)

1. Comparison with monitoring by an indenture trustee.

The monitoring role of the supertrustee differs markedly from that of the indenture trustee. The indenture trustee has no legal or contractual obligation to monitor actively the company's finances in order to detect a breach of a covenant.(91) The indenture trustee also has no access to non-public information, and is permitted to rely on conclusory compliance certificates provided by the company.(92) Moreover, the indenture trustee has no financial incentive to monitor. To the contrary, since the indenture trustee has limited obligations as long as it is not aware of a covenant breach, but has to perform additional tasks (without receiving additional compensation) if it becomes aware of a breach, the indenture trustee has a financial disincentive to monitor intensively and to detect breaches.

2. Scope of monitoring and liability for failure to monitor.

The affirmative duty of a supertrustee to monitor raises the issue of the scope of the required monitoring and the liability of the supertrustee for failure to monitor.

As to scope, the supertrustee should be required to engage in the type and intensity of monitoring which a reasonable lender in the private debt market would engage in under similar circumstances. This standard is commercially practicable and offers a readily available benchmark for evaluating whether a supertrustee has fulfilled its duties.

As to liability of the supertrustee, there is a trade-off. On the one hand, liability for failure to satisfy monitoring obligations will strengthen the supertrustee's incentive to monitor. On the other hand, a low threshold for bringing a claim against a supertrustee will expose the supertrustee to nonmeritorious lawsuits and second-guessing by courts, thereby raising the implicit costs of the supertrusteeship. The liability regime needs to strike a balance between these considerations.

We recommend that issuers and bondholders include several structural safeguards in the supertrustee's liability regime. First, only bondholders who, in the aggregate, own a material fraction of the bonds, say 10%, should be permitted to sue the supertrustee. Second, the supertrustee's monetary liability for ordinary neglect of its duty to monitor should be limited to, for example, some multiple of the annual fees earned by the supertrustee from the bond issue. That is, the supertrustee should not ordinarily be liable for the full amount of the losses bondholders may suffer as a result of the supertrustee's failure to monitor. The supertrustee's liability should be more extensive only if bondholders can show that the supertrustee's failure to fulfill its duties was reckless, willful, or in bad faith--for example, if the company bribed the supertrustee not to monitor.

C. The Power to Renegotiate the Indenture

The supertrustee should be authorized to renegotiate bond covenants on behalf of bondholders and make independent decisions concerning covenant amendments. In this regard as well, the supertrustee differs from the indenture trustee, which lacks authority to consent on behalf of bondholders to substantive changes in the bond indenture.(93)

Empowering the supertrustee to renegotiate covenants independently is crucial to our proposal because it eliminates both the collective action problem and the bonding problem resulting from the dispersion and fluidity in ownership of public bonds. Renegotiation of public bonds issued under a supertrusteeship will thus more closely resemble renegotiation of private debt than renegotiation of public bonds issued under the existing indenture trustee governance structure. Precisely because it will be relatively easy to renegotiate bond covenants, public bonds issued with a supertrustee will include more extensive and tighter covenants, thereby reducing agency costs and bondholder risk.

From the bondholders' viewpoint, authorizing the supertrustee to decide independently on covenant changes may sometimes be preferred to having such matters decided by a majority vote. A priori, bondholders may be less reluctant to cede control to a supertrustee--a known entity with a reputational interest in representing the interests of bondholders qua bondholders--than to an unknown group of majority bondholders who may have interests that conflict with those of the minority.(94)

1. Renegotiation of core financial terms.

A question arises as to whether the power to renegotiate should be extended to changes in the core financial terms that are adverse to bondholders, such as interest rate reductions and principal forgiveness. Unlike the power to renegotiate bond covenants, giving the supertrustee the power to renegotiate core financial terms is not crucial to our proposal. While the more extensive covenant structure of supertrustee bonds will necessitate more monitoring of covenant compliance, more frequent covenant renegotiation, and more complex enforcement decisions if a covenant is breached, the structure does not increase the likelihood that a company will seek changes in core financial terms. To the contrary, other things being equal, more effective monitoring and stricter covenants should result in less frequent renegotiations of core financial terms.(95)

On the other hand, there may be other benefits in giving the supertrustee some power to renegotiate core financial terms, possibly within some prescribed limits.(96) Renegotiation of core financial terms, like other renegotiations, involves a collective action problem.(97) Moreover, the present requirement of unanimous approval of changes in the core financial terms gives a disproportionate power to any group of bondholders whose interests may be inconsistent with those of the great majority of bondholders or who want to hold out for better terms.(98) The supertrustee, as agent for all bondholders, could ameliorate both the collective action problem and the holdout problem affecting restructurings of core financial terms. Therefore, we propose that companies should be permitted (but not required) to issue supertrustee bonds in which the supertrustee's power to renegotiate extends to core financial terms, with such limits as prescribed in the indenture.(99) As we discuss below, the Trust Indenture Act presently does not permit the issuance of such bonds.(100) Thus, although it is not crucial to the implementation of our proposal, we favor amending the Act to permit a supertrustee to renegotiate core financial terms on behalf of the bondholders.

Even if the supertrustee lacks authority to approve changes in core financial terms independently, it can play an important role in the restructuring process. Through its past monitoring, the supertrustee will have acquired substantial knowledge of the company and will be well positioned to evaluate the merits of a proposal. The supertrustee is thus well positioned to act as advisor to and representative for bondholders--even if ultimately the bondholders themselves must consent to changes in core financial terms.(101)

D. The Power to Enforce the Indenture

The supertrustee should have the power to take enforcement actions against the company if the company violates provisions of the bond indenture. While the indenture trustee also has such power, there are a number of important differences between our proposal and the present situation.

First, due to its greater monitoring activities and closer association with the company, the supertrustee will be in a better position than a generally passive indenture trustee to assess what enforcement steps should be taken. Moreover, the supertrustee will have incentives to enforce the indenture reasonably in two respects: in relation to the bondholders, the supertrustee will have incentives to protect their interests effectively; and in relation to the company, the supertrustee will have incentives not to act opportunistically. The enforcement of the indenture of supertrustee bonds should thus resemble the enforcement of private debt agreements.

Second, absent a payment default, the power to enforce should be vested exclusively with the supertrustee. Bondholders should not be entitled to direct the supertrustee to take any particular enforcement action, and, as discussed below, the right of bondholders to sue the supertrustee if they disagree with its enforcement decisions should be limited. These limitations on the powers of bondholders are designed to assure the company that the indenture provisions will be enforced nonopportunistically. In contrast, the indenture trustee holds enforcement power concurrently with bondholders; it must follow directions given by holders of a majority of the bonds; and bondholders have broader rights to sue the indenture trustee with regard to its enforcement decisions.(102)

We understand that vesting enforcement power exclusively with the supertrustee exposes bondholders to the possibility that the supertrustee will abuse that power. However, we do not believe that this possibility is a major problem. For one, the practical potential for abuse is limited both by the fact that bondholders regain enforcement power if the company defaults on an interest or principal payment and by the supertrustee's incentive structure, which we discuss below.(103) Moreover, but for the supertrustee scheme, the bond indenture would not contain many of the covenants with regard to which the supertrustee is to have exclusive enforcement power. Thus, it is hard to see how bondholders could be worse off under our scheme, where a supertrustee has discretion in enforcing covenants, than under the present regime where, due to the scarcity of covenants, few limits are imposed on management's discretion. Finally, shareholders are limited in their ability to control and to sue corporate directors in ways that are similar to the limits we propose for bondholders and the supertrustee,(104) and holders of bond mutual funds have even less authority over enforcement decisions of the fund managers than bondholders in a supertrusteeship. Even closer to the point, bondholder trustees in the United Kingdom exercise discretion(105) to enforce only those breaches in trust deeds that they deem to be materially prejudicial to bondholders, and may, without obtaining bondholder consent, agree to amendments to the terms of a bond deed that are not materially prejudicial and that do not result in substantial permanent changes. (In contrast to our proposed scheme, however, U.K. bonds do not contain covenants comparable to those in private debt; U.K. trustees do not engage in significant independent monitoring; and they lack direct financial incentives to maximize the value of the bonds.)(106) Though the grant of such authority may well create problems in these contexts, the fact that such authority is commonly granted indicates that the potential for abuse can be sufficiently limited to ensure that such grants are, on balance, beneficial.

E. The Supertrustee's Liability for Renegotiation and Enforcement Decisions

Many renegotiation and enforcement decisions made by the supertrustee will involve difficult judgment calls. To reduce the potential for nonmeritorious lawsuits and to avoid having courts second-guess these judgment calls, the supertrustee's decisions with regard to enforcement and renegotiation should be evaluated under a liability standard analogous to the "business judgment rule" applicable to decisions taken by a board of directors. Under this standard, bondholders would not be entitled to sue a supertrustee if its enforcement or renegotiation judgments are merely wrong or substantively unreasonable. They could, however, bring a claim if the supertrustee does not act in good faith or acts in the presence of a disabling conflict of interest--the supertrusteeship-equivalent of self-dealing by corporate directors. (Moreover, as discussed earlier, bondholders can sue the supertrustee if it fails to monitor adequately and as a result bases its enforcement and renegotiation decisions on insufficient information.)(107) Importantly, for purposes of this standard, the interest of the supertrustee in acting reasonably toward issuer companies (in order to increase the likelihood of being appointed as a supertrustee for other bond issues) does not constitute a conflict of interest. To the contrary, it is integral to our scheme that the supertrustee will take its interest in future appointments into account in dealing with the company.

The liability standard that we propose differs from the standard applicable to enforcement decisions by the indenture trustee. The indenture trustee must use "the same degree of care and skill [to enforce the bond indenture] as a prudent man would exercise or use under the circumstances in the conduct of his own affairs"(108) as judged by a court after the fact. If a "prudent man" for purposes of this standard is equated with a prudent public bondholder (rather than, say, a prudent private creditor), this standard would not permit the supertrustee to take into account the effect of its enforcement decisions on future business. Moreover, the "prudent man" standard permits courts to engage in too much after-the-fact second-guessing of the supertrustee's enforcement decisions.(109)

F. Compensation

The compensation of a supertrustee should be designed to achieve two goals. First, the level of compensation must provide sufficient income to the supertrustee to pay for expenses incurred in fulfilling its duties and to earn a reasonable profit.(110) Second, the compensation should be structured to enhance the supertrustee's incentive to represent bondholders effectively.

The issuing company will set the details of the compensation scheme for the supertrustee before the bonds are issued. Once set, the scheme cannot be changed without the consent of the bondholders. Otherwise, the issuing company could alter the compensation scheme in a manner that reduces (or even reverses) the supertrustee's incentive to represent bondholders effectively.

As discussed earlier, the supertrustee will have significant responsibilities with respect to, and will need to expend substantial efforts in, monitoring, renegotiating, and enforcing covenants. These responsibilities and efforts greatly exceed the present responsibilities and efforts of an indenture trustee. Thus, the fees paid to the supertrustee should be materially higher than those paid to the indenture trustee.(111) The higher fees should be recoverable through the company's ability to issue supertrustee bonds at a higher price (or lower yield) than it can issue indenture trustee bonds or, more fundamentally, by the reduction in the agency costs of public debt resulting from the supertrusteeship.(112)

The supertrustee's compensation should include incentives for the supertrustee to fulfill its duties effectively.(113) Unlike the indenture trustee's fixed annual fee, the amount of the supertrustee's compensation should be positively tied to changes in the market value of the bonds. More precisely, the compensation should be tied to changes in bond value that result from variations in the bonds' credit risk, but not to changes that result from fluctuations in interest rates on U.S. Treasury debt--a matter over which the supertrustee has no control.

Such a tie can be achieved in several ways. For example, the supertrustee could be compensated by a set of derivative securities the value of which declines when the spread between the yield on the bonds and the yield on Treasury debt widens. This will give the supertrustee an economic incentive to act to prevent reductions in the value of the bond, without also rewarding or penalizing the supertrustee for changes in the prices and yields of U.S. Treasury debt. Though technically somewhat more complex, these derivative securities do not differ fundamentally from executive stock options commonly granted to provide incentives to managers. Other incentive-based compensation regimes that could align the interests of the supertrustee with those of the bondholders include bonuses for each payment that the company makes in full and on time, and higher fixed payments to the supertrustee combined with a partial guarantee by the supertrustee of the company's debt. (Insurance companies commonly give analogous guarantees for municipal bonds.)

The supertrustee could also commit to holding a stake in the same bond issue for which it serves as a supertrustee. This will emulate the incentive structure of private loans, where a creditor owns the debt and suffers a direct loss if it shirks in the performance of its monitoring, renegotiation, and enforcement duties. A supertrustee can thus signal its self-interest in fulfilling its duties by purchasing, and pledging to hold during the term of its supertrusteeship, a substantial fraction, (e.g., 10% or 20%) of the bond issue. Such ownership would provide the supertrustee with a direct benefit from diligent performance of its tasks, as well as a direct loss from shirking. However, it has the disadvantage of exposing the supertrustee to the risk of changes in yields on Treasury debt(114) and of making the supertrustee's holding illiquid.(115)

Finally, once supertrustees have developed a positive reputation for the quality of their services, there may no longer be any compelling need for special incentive compensation. Bond rating agencies such as Moody's and Standard and Poor's perform their tasks diligently without such compensation. The economic interest of the rating agencies in preserving their reputations suffices as an incentive for diligent service.

G. Monitoring the Supertrustee and the Significance of Reputation

An important element of the supertrusteeship is that a supertrustee would develop a reputation for responsible conduct in the eyes of two distinct clienteles. First, investors must be willing to pay more for bonds with a supertrustee governance structure even though some of their individual powers are curtailed. Second, companies that issue bonds must be persuaded to confer on the supertrustee substantially greater power than that held by an indenture trustee or by bondholders under the present governance structure. The reputation of a supertrustee is an important factor in giving both bondholders and the company comfort that the supertrustee will not abuse its power.

For a supertrustee to develop its reputation, it is necessary that investors and issuers have the ability to monitor the supertrustee to determine whether the supertrustee represents bondholders faithfully and acts nonopportunistically towards issuer companies. As we will argue in this subsection, we believe that investors and issuers have a sufficient ability to monitor the supertrustee and that the supertrustee will have incentives to develop a reputation for responsible conduct vis-a-vis both bondholders and companies. In making this case, we will address two concerns. First, how can investors monitor the behavior of the supertrustee if they are apparently unable to monitor the company's compliance with indenture covenants? Second, how can a supertrustee develop a reputation if indenture trustees have failed to develop significant reputations regarding the effectiveness of their representation of bondholders?

1. The ability to monitor.

For a supertrustee to develop a reputation, investors and issuers must be able to monitor the behavior of the supertrustee. At first glance, it may appear that issuers have such an ability, while investors do not. Issuers have high stakes in assuring that the supertrustee on their bonds will act responsibly. Issuers can obtain information about supertrustees from lawyers, investment bankers, and other executives. In addition, issuers who have issued supertrustee bonds in the past will have had direct experience with at least one supertrustee.

Investors, on the other hand, may seem hampered by the collective action problem, which partly accounts for the need to appoint a supertrustee in the first place. Moreover, given their passive role envisioned by our proposal, investors gain little direct experience in dealing with a supertrustee. If the foregoing were to hold true, then investors' relative inability to judge supertrustee performance would expose bondholders to two forms of exploitation: first, the supertrustee may, in order to attract supertrusteeship business from issuers, fail to represent bondholders faithfully; second, the supertrustee may, in order to reduce its own expenses, fail to expend efforts in monitoring the company or fail to take enforcement action if a breach occurs.

Upon closer inspection, however, the same collective action problem that renders it difficult for bondholders to monitor, enforce, and renegotiate covenants does not substantially inhibit their ability to assess supertrustee reputations. First, such reputational monitoring is substantially easier than the "preventive" monitoring required to ensure covenant compliance. Unlike reputational monitoring, which is ex post and can be sporadic, preventive monitoring requires continuous, real-time supervision. Assume, for example, that the company has defaulted on its bonds. If, at this point, bondholders learn that the company had breached a covenant two years before, it would not help them: They would have had to detect the breach earlier to take steps to prevent further damage. However, it would be useful for bondholders to learn that, two years before, the supertrustee had acted improperly, because that knowledge will make them more reluctant to buy bonds with the same supertrustee in the future.(116)

Second, bondholders have greater incentives to monitor supertrustee reputations than they have to monitor, enforce, and renegotiate covenants. The benefits of monitoring a supertrustee's reputation accrue to all supertrustee bonds that an investor purchases. For example, if an investor considers purchasing ten different bond issues, for $10 million each, with the same entity acting as supertrustee, assessing its reputation will be relevant for an $100 million investment. By contrast, the benefits of monitoring covenant compliance accrue only to the specific bond issue (or, at most, to all bonds issued by the same company). In this respect, it is important to recall that most holders of public bonds are financial institutions--insurance companies, pension funds, mutual funds--which hold large portfolios of bonds.(117) Such investors obtain economies of scale from reputational monitoring that counterbalance, and probably exceed, the diseconomies resulting from the collective action problem. Moreover, over time, such investors will have many direct dealings with supertrustees even if they are largely passive in any particular bond issue.

Third, investors have greater incentives to monitor reputation than to monitor covenants because they can more easily use the information gained. The information gained by reputational monitoring is used by each investor individually to determine whether and at what price to buy supertrustee bonds. Conversely, the information gained by covenant monitoring must be used by all bondholders collectively to determine how to deal with a breach.(118) Covenant monitoring thus entails a more severe collective action problem than reputational monitoring.(119)

Finally, we do not think that the possibility that a supertrustee will fail to represent borrowers faithfully in order to attract supertrusteeship business from issuers is of serious concern. Most bond purchasers are sophisticated financial institutions, hold large bond portfolios, and enjoy access to law firms, investment bankers, and other market participants comparable to that of bond issuers.(120) It is therefore highly unlikely that the fact that a particular supertrustee acts in this manner could be well known among issuers, but not equally well known among investors.

2. Will supertrustees develop a reputation?

As we have shown, bondholders can more easily monitor the supertrustees' reputations than the company's covenant compliance. Yet it is also easy for bondholders to monitor the reputations of the current indenture trustees. Yet, indenture trustees have not developed significant reputations regarding the effectiveness of their representation of bondholders. Why, then, should supertrustees?

We note at the outset that reputation is an important aspect of many business enterprises. In the financial services industry, lenders in the private debt market have developed reputations of substantial value regarding the way they renegotiate and enforce covenants,(121) underwriters have developed similarly valuable reputations for the initial public offerings which they underwrite,(122) and bond rating agencies have comparable reputational concerns. Thus, companies in general, and financial companies in particular, are clearly able to develop reputations in certain circumstances.

In our view, indenture trustees have failed to develop a reputation for the quality of their bondholder representation because such a reputation is not important for an indenture trustee given how its tasks are currently defined. Indenture trustees have a narrow scope of authority until default occurs.(123) Since covenants in public bonds are lax, covenant defaults are rare. Payment defaults, of course, do occur, but are often accompanied by bankruptcy filings, where a creditors' committee takes over the task of representing bondholders.(124) Hence, in practice, the trustee does not engage frequently in important post-default functions either.

In general, an agent's reputation enables those who seek its services to predict its behavior when this behavior cannot be monitored directly and inexpensively. This is particularly valuable when the agent's authority contemplates a substantial range of discretionary behavior. Such predictive ability is not of much use in the case of the indenture trustee, who rarely exercises significant discretion in representing bondholders. Consequently, indenture trustees compete on the basis of price and the ability to perform routine administrative tasks (such as mailing redemption notices to bondholders), rather than on the quality of their bondholder representation.(125)

In contrast, the supertrustee will have substantially greater responsibilities and powers than the indenture trustee and can take a greater range of discretionary action. The supertrustee will be called on more frequently to exercise its judgment and will receive greater fees. The combination of higher remuneration and a greater measure of responsibility and discretion will provide incentives for supertrustees to develop, and to compete on the basis of, reputation for representing bondholders effectively, as well as treating issuers fairly, in renegotiating and enforcing covenants.


The previous Part outlined the potential advantages of the supertrustee proposal and some of its specific elements. This Part considers some issues that may arise in the implementation of the proposal. We first discuss legal barriers to the proposal. While most elements of the proposal can be implemented without any change in the law, some elements would conflict with the Trust Indenture Act. We suggest revisions to the Act to eliminate these conflicts. Second, we examine several economic barriers to the adoption of the proposal. Third, we consider which companies may gain the most from issuing supertrustee bonds.

A. Legal Barriers

The governance structure for public corporate bonds is, to a large extent, set out by the Trust Indenture Act of 1939.(126) The Act requires that an indenture trustee be appointed to represent corporate bondholders and sets up a framework for the indenture trustee's rights and duties.(127) It creates the presumption that the indenture trustee has no independent duty to monitor covenant compliance,(128) that the indenture trustee can rely conclusively on compliance certificates provided by the company stating that the company has not violated any indenture provisions,(129) and that a majority of bondholders may direct the indenture trustee to take certain enforcement steps.(130)

For corporate bonds with a supertrustee, having a separate entity act as indenture trustee would lead, at best, to a costly duplication of effort and, at worst, to conflicts between the trustees. Preferably, therefore, the supertrustee proposal should be implemented with a trustee that fulfills the duties of a supertrustee and also conforms to the requirements of the Trust Indenture Act.

Fortunately, most of the provisions of the Trust Indenture Act that are inconsistent with the supertrustee proposal are "default" rules and can thus be overridden by explicit provision in the bond indenture. For example, even though the Trust Indenture Act creates a presumption that the indenture trustee can rely conclusively on compliance certificates provided by the company,(131) the bond indenture can provide that the indenture trustee may not rely on such certificates. Thus, the supertrustee proposal can largely be implemented within the present boundaries of the Trust Indenture Act.

Several provisions in the Trust Indenture Act are, however, mandatory and cannot be overridden by the bond indenture. In particular, the Act mandates a liability standard for the indenture trustee different from the liability standard we consider appropriate for the supertrustee,(132) and requires consent by all affected bondholders to changes in the timing and amount of interest and principal payments on the bonds.(133)

1. Liability standard for a trustee.

With regard to the trustee's duties, once a covenant violation is detected and a "default" occurs, section 315(c) of the Trust Indenture Act prescribes that the indenture trustee must "use the same degree of care and skill [in exercising its enforcement powers] as a prudent man would exercise or use under the circumstances in the conduct of his own affairs."(134) As discussed above, this liability standard is stricter than the "business judgment rule" standard we propose for the supertrustee.(135) It can lead to excessive second-guessing of enforcement decisions by courts and may preclude the supertrustee from taking account of its self-interest in dealing fairly with the company.

Similarly, section 315(d) of the Act specifies that the indenture trustee may not be relieved from liability for its own negligent actions or its own negligent failure to act.(136) This negligence standard applies prior to default to the extent that the trustee is required to perform any pre-default duties.(137)

Unlike the indenture trustee, the supertrustee will have an express duty to monitor the company, actively and independently, prior to default. While we suggest that the supertrustee be liable for neglect of its pre-default duty to monitor, we additionally recommend the inclusion of some important safeguards in the specification of the supertrustee's liability regime.(138) Although section 315(d) can be construed to mandate only a liability standard (negligence) and to be silent on other features of the liability regime, the wording of the section--the indenture may contain no provision "relieving the trustee from liability"(139)--suggests that it may in fact extend to at least some of our recommended safeguards, such as the quantitative limit on a trustee's liability. Thus, section 315(d) is likely to be inconsistent with our suggested liability regime for a supertrustee's neglect of its monitoring duties.

2. Renegotiation of core financial terms.

Another provision of the Trust Indenture Act that may constrain implementation of the supertrustee proposal is the "unanimous consent rule" in section 316(b).(140) Section 316(b) provides that "the right of any holder of any indenture security to receive payment of the principal of and interest on such indenture security ... shall not be impaired or affected without the consent of such holder ..."(141) In effect, section 316(b) requires the unanimous consent of all bondholders to indenture amendments that adversely affect the right to interest and principal payments.(142)

While it is not crucial to our proposal, we suggest that companies and bondholders should have the option to empower the supertrustee to consent, on behalf of the bondholders, to some changes in the core financial terms of a supertrustee bond issue.(143) At present, section 316(b) prevents the implementation of the supertrustee proposal in this expansive fashion.

3. Implementation of supertrusteeship.

We favor exempting bonds with the proposed supertrustee governance structure from the mandatory liability provisions of section 315 and, if the parties so desire, from the unanimous consent provisions of section 316(b). Creating such exemptions does not necessarily require a legislative amendment to the Act. The Securities and Exchange Commission enjoys the statutory power to "exempt ... any ... security or transaction ... from any one or more provisions of [the Act], if and to the extent that such exemption is necessary or appropriate in the public interest and consistent with the protection of investors and the purposes fairly intended by" the Act.(144) The SEC should use this power to exempt supertrustee bonds from the liability provisions in section 315 and the unanimous consent rule in section 316(b) of the Trust Indenture Act.

Absent a statutory amendment or action by the SEC, the supertrustee proposal could be implemented in any of three alternative ways. First, the supertrustee could be subjected to the liability provisions in the Act. Second, corporate bonds could be issued with two bondholder representatives: an indenture trustee subject to the provisions of the Act as well as a supertrustee as outlined in our proposal. However, having two representatives entails higher cost and makes implementing the proposal more complicated by creating a potential for conflict between the trustees. Third, one could try to evade the grasp of the mandatory provisions by aggressive drafting. For example, "default" could be defined, for purposes of the Act's liability provisions, not to include a covenant breach, thus assuring that the heightened liability standard for the indenture trustee does not apply.(145) The legal validity and effect of such clauses would be subject to uncertainty which, until resolved, would detract from the benefits of the supertrustee proposal.

B. Economic Issues

Successful implementation of our proposal will also require informing and educating investors about the benefits of the proposal, as well as developing institutions to act as supertrustees that can be relied upon by both investors and issuers. The difficulty is that externalities in the development of the supertrusteeship may inhibit the genesis of these institutions, even if the institutions would ultimately enhance efficiency.

Given the cost of a supertrusteeship, a company will issue a supertrustee bond only if its provisions are "priced" by the market, that is, only if bondholders accept a lower yield on a supertrustee bond than on otherwise equivalent bonds issued under the present governance structure. This requires that a broad class of investors recognize the benefits emanating from the supertrusteeship, because the initial bondholders may wish to resell the bonds in the secondary market.(146) The supertrustee proposal, like other innovations in securities design, thus involves "network externalities:"(147) Even if some market participants find the proposal attractive, none may be willing to be the first to adopt it and none may want to bear the burden of educating the market about its benefits.

Underwriters can ameliorate this externality problem because they can expect to be engaged repeatedly in bringing supertrustee bonds to market and to benefit from the activity. There is still a problem of externalities here, because no single underwriter will reap all of the benefits from the advent of supertrustee bonds, while the initiating underwriter will bear much of the costs. On the other hand, if an underwriter believes that its competitive position can be enhanced by introducing a bond with the novel feature of supertrustee, it may be induced to pioneer the adoption of our proposal.(148)

The implementation of the proposal also requires the development of institutions with a reputation for acting as supertrustees in a manner that is fair and acceptable to both companies and investors. The institutions that presently serve as indenture trustees do not possess the requisite reputation, and it will take some time before such reputation is developed. This problem may also obstruct the implementation of the supertrustee proposal. Again, at the beginning, a single institution may be unwilling to invest in developing a reputation for providing a service whose economic viability is uncertain.

We expect, however, that institutions that already possess considerable reputation in related activities will provide a base for recruiting supertrustees. These institutions include banks, insurance companies, underwriters, accounting firms, and bond rating agencies. For such institutions, the cost of developing a reputation as a supertrustee may not be high. The institutions may thus be more inclined to enter the market for supertrustee services even when there is substantial uncertainty about the depth and persistence of demand for those services. Again, as in the development of any new product or service, an institution may wish to establish a reputation early, in order to seize a larger share of the prospective market. Moreover, until a firm reputation is developed, these institutions can commit to holding a certain percentage of the company's bonds.(149)

C. Borrowers That May Use a Supertrustee

A supertrustee may be appropriate for some but not all companies. Because the costs of fulfilling the duties of a supertrustee are largely fixed or increase less than proportionally with the size of a bond issue, there will be economies of scale and, therefore, a propensity to establish a supertrusteeship for larger issues. This prediction corresponds with the evidence that larger bond issues are more likely to be in the form of public bonds,(150) for which the function of the supertrustee is intended. Also, if a company has issued several bonds with similar covenants and seniority, it may prefer to appoint the same supertrustee for all these issues: A group of such bonds requires no more information on the company and no greater monitoring effort than what is needed for a single bond. This design will reduce the cost of the supertrusteeship per dollar value of bonds and enhance the economic feasibility of the proposal.

Similarly, the supertrustee proposal is more attractive for bonds with longer maturities. Liquidity is relatively more important for such bonds, because there is a greater likelihood that an original buyer will want to sell before maturity. In addition, a supertrusteeship involves temporal economies of scale. Monitoring costs increase less than proportionally with time because a supertrustee benefits from information acquired in past monitoring of the company.

It is also expected that a supertrustee is more likely to be appointed by companies issuing debt with more than de minimis credit risk. In such cases there is a more significant conflict of interest between bondholders and stockholders and a greater incentive for stockholders to take actions that transfer wealth from bondholders to themselves. A supertrusteeship would thus be more valuable in reducing agency costs of debt.

Another source of gain attributable to a supertrustee may appear in the context of companies sliding into financial distress. In that case, the supertrustee may facilitate prebankruptcy negotiations, and possibly a workout,(151) that avoids a bankruptcy(152) that would inflict dead weight costs on both stockholders and bondholders.(153)

Finally, the ownership and control structure of the issuing company may affect the incentives of stockholders to extract wealth from bondholders. For example, owner-controlled companies may have a greater propensity to increase the operating risk of the firm (compared to manager-controlled companies) and thus act in a way that is contrary to the interests of bondholders.(154) In this case, bondholders' interests are more likely to be violated and there is a greater need for a supertrustee.


In this article, we propose a new governance structure for publicly issued corporate bonds. For many companies, this new structure will be preferable to the existing governance structures for public and private debt.

Compared to private debt, ownership of public bonds is both fluid and dispersed. While the fluidity and dispersion of ownership generate benefits--public bonds have greater liquidity and their risk is more easily diversifiable--they also entail two problems: the "collective action" problem, which results from the dispersion of ownership; and the "bonding" problem, which results from fluidity in ownership. The collective action and bonding problems, in turn, make it difficult to reduce agency costs of debt in public bonds.

Our new governance structure is designed to overcome the collective action problem and the bonding problem while retaining the greater liquidity and easy diversifiability that stern from the dispersion and fluidity in ownership of public bonds. The key element of the governance structure is a new type of bondholder representative--the supertrustee--which will have the power and the duty to monitor, renegotiate, and enforce extensive bond covenants. Thus, while bond ownership will remain dispersed and fluid, the power to act for bondholders will be concentrated and remain stable in the supertrustee. We show how the supertrusteeship may be designed to create these benefits and specifically discuss the incentives of the supertrustee to act as faithful representative vis-a-vis bondholders and as reasonable lender vis-a-vis issuing companies.

Presently, the adoption of our proposal is impeded by several provisions of the Trust Indenture Act. We therefore propose that the SEC use its regulatory authority to exempt supertrustee bonds from these provisions to enable companies to issue supertrustee bonds if they choose to do so.

We would like to thank Bill Allen, Lucian Bebchuk, Bernie Black, Robert Brokaw, John Coates, Tamar Frankel, Alon Harel, Louis Kaplow, Ehud Kamar, David Norris, Eric Otts, Ed Rock, George Triantis, Philipp v. Randow, and the participants at the Harvard Law School Law and Economics Workshop, the University of Pennsylvania Law School Law and Economics Workshop, the University of Toronto Law and Economics Workshop, and the Columbia Law School Conference on Alternative Corporate Governance for their helpful comments. We would also like to thank the Filomen D'Agostino and Max E. Greenberg Research Fund for its financial assistance.

(1.) See, e.g., John E. Berthoud, Is There a Future for Corporate Governance?, WASH. TIMES, Jan. 22, 1997, at A15 (op-ed discussing merits of corporate governance proposals); Reforming the Firm, ECONOMIST, Aug. 9, 1997, at 16 (editorial arguing that directors should try to make shareholders rich, and not merely to make managers accountable).

(2.) See, e.g., Kayla J. Gillan, CalPERS--Corporate Governance: Planning for the 21st Century, Address to the Institutional Investor Conferences, Corporate Governance Presentation, May 15, 1997, available in LEXIS, CURNWS File; Maureen Reilly, California Public Employees' Retirement System: Why Corporate Governance Today? A Policy Statement, in HANDLING MERGERS & ACQUISITIONS IN A HIGH-TECH AND EMERGING GROWTH ENVIRONMENT, at 979 (PLI Corp. Law Practice Course Handbook Series No. 985, 1997); Chancellor William T. Allen, Corporate Directors in the Dawning Age of Post-Materialism: New Problems and New Solutions, Address at Stanford University Center for Economic Policy Research Conference on Corporate Governance (May 1, 1992).

(3.) See, e.g., Columbia School of Law Conference on Corporate Governance Today (May 21-22, 1998); Stanford University Center for Economic Policy Research Conference on Corporate Governance (May 1, 1992); ALI-ABA Conference on Current Issues in Corporate Governance (Nov. 30, 1995); World Bank Conference on Corporate Governance in Eastern Europe and Russia (Dec. 1994); Conference on the Influence of Corporate Governance and Financing Structures on Economic Performance, sponsored by the French Ministry of Industry, Postal Services and Telecommunications and External Trade, OECD.


(5.) A LEXIS search in the ALLREV file retrieved over 1,500 articles containing the term "corporate governance." A partial list of law review articles whose titles contain "corporate governance" includes Jayne W. Barnard, Institutional Investors and the New Corporate Governance, 69 N.C. L. REV. 1135 (1991); Michael Bradley & Cindy A. Schipani, The Relevance of the Duty of Care Standard in Corporate Governance, 75 IOWA L. REV. 1 (1989); Victor Brudney, Corporate Governance, Agency Costs, and the Rhetoric of Contract, 85 COLUM. L. REV. 1403 (1985); William J. Carney, The ALI's Corporate Governance Project: The Death of Property Rights?, 61 GEO. WASH. L. REV. 898 (1993); Daniel R. Fischel, The Corporate Governance Movement, 35 VAND. L. REV. 1259 (1992); Carol Goforth, Proxy Reform as a Means of Increasing Shareholder Participation in Corporate Governance: Too Little, But Not Too Late, 43 AM. U. L. REV. 379 (1994); Martin Lipton, Corporate Governance in the Age of Finance Corporatism, 136 U. PA. L. REV. 1 (1987); Mark J. Loewenstein, The SEC and the Future of Corporate Governance, 45 ALA. L. REV. 783 (1994); Jonathan R. Macey & Geoffrey P. Miller, Corporate Governance and Commercial Banking: A Comparative Examination of Germany, Japan, and the United States, 48 STAN. L. REV. 73 (1995); John H. Matheson & Brent A. Olson, Corporate Law and the Longterm Shareholder Model of Corporate Governance, 76 MINN. L. REV. 1313 (1992); Lawrence E. Mitchell, A Critical Look at Corporate Governance, 45 VAND. L. REV. 1263 (1992); Lawrence E. Mitchell, Private Law, Public Interest?: The ALI Principles of Corporate Governance, 61 GEO. WASH. L. REV. 871 (1993); Donald E. Schwartz, Federalism and Corporate Governance, 45 OHIO ST. L.J. 545 (1984); G. Richard Shell, Arbitration and Corporate Governance, 67 N.C.L. REV. 517 (1989); D. Gordon Smith, Corporate Governance and Managerial Incompetence: Lessons from Kmart, 74 N.C.L. REV. 1037 (1996); Elliott J. Weiss, Economic Analysis, Corporate Law, and the ALI Corporate Governance Project, 70 CORNELL L. REV. 1 (1984).

(6.) See, e.g., Reforming the Firm, supra note 1, at 16 (noting successes of the "corporate governance movement"); see also Daniel R. Fischel, The Corporate Governance Movement, 35 VAND. L. REV. 1259 (1982); Gillan, supra note 2, (discussing impact of "corporate governance movement"); Judith H. Dobrzynski, Big Investors Train Their Fire on Nonperforming Directors, N.Y. TIMES, Mar. 7, 1997, at D6 (noting increasing sophistication of the "corporate governance movement"); Profiles in Power, NAT'L L.J., Apr. 28, 1997, at C10 (describing Joseph Grundfest and Martin Lipton as, respectively, leader and key conceptualizer of the "corporate governance movement"); Richard Waters, GM Unites Top Jobs Under Smith, FIN. TIMES, Dec. 5, 1995, at 21 (noting that GM has become a rallying point of the "corporate governance movement"); World's Third Largest Pension Fund Sees Problems With an Around-the-Clock, 24-Hour Global Market, INT'L SEC. REG. REP., May 22, 1997 (noting CalPERS' leadership role in the "corporate governance movement").

(7.) See How Sweet It Was!, INVESTMENT DEALERS' DIG., Jan. 11, 1993, at 14 (stating that companies sold $534 billion of bonds, $72.4 billion of common stock, and $49 billion of preferred stock during 1992). In terms of market values of outstanding securities, corporate debt also accounts for a significant portion of a company's capital structure. See text accompanying note 14 infra.

(8.) At present, debt, if accorded any function at all, is generally seen as ancillary to equity and as a means to amplify managerial efforts to maximize the value of the company's shares. See, e.g., Michael C. Jensen, Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers, 76 AM. ECON. REV. 323, 324 (1986) (suggesting that greater levels of debt inhibit wasteful investment decisions). Articles dealing with the governance of debt in its own right have typically focused on a more narrow aspect of the governance structure. See, e.g., William W. Bratton, Jr., The Economics and durisprudence of Convertible Bonds, 1984 WIS. L. REV. 667, 735-39 (opposing fiduciary duty to bondholders); John C. Coffee, Jr. & William A. Klein, Bondholder Coercion: The Problem of Constrained Choice in Debt Tender Offers and Recapitalizations, 58 U. CHI. L. REV. 1207, 1271-73 (1991) (favoring rules against bondholder coercion); Marcel Kahan & Bruce Tuckman, Do Bondholders Lose from Junk Bond Covenant Changes?, 66 J. BUS. 499, 513 (1993) (arguing that bondholders can protect themselves against coercion); Marcel Kahan, The Qualified Case Against Mandatory Terms in Bonds, 89 NW. U. L. REV. 565, 621-22 (1995) (arguing that governance structure should be determined by contract, not by mandatory legal rules); Morey W. McDaniel, Bondholders and Corporate Governance, 41 BUS. LAW., 413, 449 (1986) (arguing for fiduciary duties to bondholders); Lawrence E. Mitchell, The Fairness Rights of Corporate Bondholders, 65 N.Y.U.L. REV. 1165, 1222-25 (1990) (arguing for fairness rights for corporate bondholders).

(9.) See, e.g., ROBERT C. CLARK, CORPORATE LAW 389-400 (1986) (discussing collective action problem in shareholder context).

(10.) See text accompanying note 47 infra.

(11.) Prior commentators have made proposals that are similar in spirit to single features of our scheme. See Mitchell, supra note 8 (arguing for stronger protections for bondholders); Stewart M. Robertson, Debenture Holders and the Indenture Trustee: Controlling Managerial Discretion in the Solvent Enterprise, 11 HARV. J.L. & PUB. POL'Y 461, 485 (1987) (suggesting that trustee should have duty to monitor compliance with indenture). Moreover, trustees of corporate bonds issued under U.K. law already have greater powers to renegotiate and enforce bond covenants than U.S. trustees do. However, nobody has proposed, and U.K. practice does not resemble, the integrated scheme of stronger covenants and greater trustee powers and duties that we advance in this article.

(12.) The first junk bond issued to finance an LBO contained covenants comparable to those in bank and private debt. Obtaining bondholder consent to amend the covenants, however, was more bothersome than obtaining consents to bank and private debt amendments. After several rounds of amendments, the issuer decided to seek bondholder consent to relax the financial covenants to reduce the need for future consents. Letter from Professor Bernard Black, Stanford Law School to Professor Marcel Kahan, New York University School of Law (June 4, 1998). Subsequent junk bond issues included the more lax set of covenants. See text accompanying notes 61 & 73 infra.

(13.) See John C. Coffee, Jr., Unstable Coalitions: Corporate Governance as a Multi-Player Game, 78 GEO. L. J. 1495, 1509-15 (1990).

(14.) See BOARD OF GOVERNORS OF THE FED. RESERVE SYS. OF THE UNITED STATES, FLOW OF FUNDS ACCOUNTS: FLOWS AND OUTSTANDINGS OF THE FOURTH QUARTER 1996, at 63 tbl.L 102 [hereinafter FED. RESERVE FUNDS FLOWS] (showing total debt of $2.966 trillion and total market value of equities of $6713 billion).

(15.) See generally Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305 (1976) (discussing conflicts of interest between shareholders and creditors).

(16.) See, e.g., Geren v. Quantum Chemical Corp., 832 F. Supp. 728,730 (S.D.N.Y. 1993) (discussing bondholder lawsuit complaining that special dividend caused market value of bonds to decline by fifty percent. See generally Upinder S. Dhillon & Herb Johnson, The Effect of Dividend Changes on Stock and Bond Prices, 49 J. FIN. 281 (1994) (presenting empirical evidence that large dividend increases transfer wealth from bondholders to shareholders).

(17.) The best-known example is the spinoff of Marriott International by Marriott Corporation as originally proposed in 1992. See, e.g., Robert Parrino, Spinoffs and Wealth Transfers: The Marriott Case, 43 J. FIN. ECON. 241, 263 (1997) (finding that Marriott spinoff resulted in $114 million decrease in combined value of company's stock and bonds); Eben Shapiro, Marriott Bonds Reflecting Doubt, N.Y. TIMES, OCt. 7, 1992, at D6.

(18.) See, e.g., MOODY'S SPECIAL REPORT, SPECIAL EVENT RISK IN THE U.S. AND EUROBOND MARKETS 1984-88, at 9, 10, 21 (1989) (reporting that bonds of 183 companies lost value between 1984 and 1988 due to leveraged acquisitions); Paul Asquith & Thierry A. Wizman, Event Risk, Covenants, and Bondholder Returns in Leveraged Buyouts, 27 J. FIN. ECON. 195, 200-10 (1990) (presenting empirical evidence for bondholder losses in LBOs).

(19.) See Jensen & Meckling, supra note 15, at 308-10.

(20.) If the company makes the investment and the project yields $10, shareholders will not get paid anything; if the project yields $150, shareholders will receive $70. Since there is a 50% probability that the project will yield $10 (and the stock will be worthless) and a 50% probability that the project will yield $150 (and the stock will be worth $70), the expected value of the stock is $35.

(21.) If the company makes the investment and the project yields $10, debtholders get $10; if the project yields $150, debtholders get $80. Since there is a 50% probability that the project will yield $10 (and the debt will be worth $10) and a 50% probability that the project will yield $150 (and the debt will be worth $80), the expected value of the debt is $45.

(22.) The value of the company as a whole following the investment is the sum of the value of the debt ($45) and equity ($35). This is also the expected value of receiving $10 or $150 with a probability of 50% each.

(23.) See, e.g., Kahan, supra note 8, at 570.

(24.) Covenants--which for purposes of this article, include both traditional financial covenants and security interests--are probably the most important device constraining agency costs for larger companies in the United States. Other devices used for other types of companies and in other countries include (i) personal guarantees by shareholders, (ii) short debt maturity, (iii) creditor representation on company boards, and (iv) generally applicable provisions of creditor protection laws.

(25.) See Jensen & Meckling, supra note 15, at 337-39 (discussing the role of covenants). Empirical evidence shows that such covenants are effective in protecting bondholders against some forms of wealth expropriations. See, e.g., Asquith & Wizman, supra note 18, at 201-03 (finding that publicly traded bonds with strong covenant protection gained value in leverage buyouts, while those without such covenants lost value); Mark Carey, Credit Risk in Private Debt Portfolios, 53 J. FIN. 1363 (1998) (finding that private debt, which contains more extensive covenants and is more closely monitored, has lower losses from default than public debt with ex ante lower risk).

(26.) See Marcel Kahan & Bruce Tuckman, Private Versus Public Lending: Evidence from Covenants, in THE YEARBOOK OF FIXED INCOME INVESTING 253, 259 (John D. Finnerty & Martin S. Fridson eds., 1996) (study of covenants in privately placed bonds); Kenneth Lehn & Annette Poulsen, Contractual Resolution of Bondholder-Stockholder Conflicts in Leveraged Buyouts, 34 J. L. & ECON. 645, 653 (1991) (discussion of covenants in public bonds); Clifford W. Smith, Jr. & Jerold B. Warner, On Financial Contracting: An Analysis of Bond Covenants, 7 J. FIN. ECON. 117 (1979) (noting frequent use of dividend-limiting covenants in publicly traded bonds); Avner Kalay, Stockholder-Bondholder Conflict and Dividend Constraints, 10 J. FIN. ECON. 211 (1982) (same).

(27.) See Kahan & Tuckman, supra note 26, at 260 (study of covenants in privately placed bonds); Lehn & Poulsen, supra note 26, at 653 (noting use of covenants to limit additional debt in public bonds); Smith & Warner, supra note 26, at 136-37 (same).

(28.) See Elazar Berkovitch & E. Han Kim, Financial Contracting and Leverage Induced Over- and Under-Investment Incentives, 45 J. FIN. 765, 767 (1990) (arguing that covenants can increase or decrease finn value).

(29.) See Larry H.P. Lang & Robert H. Litzenberger, Dividend Announcements: Cash Flow Signalling vs. Free Cash Flow Hypothesis?, 24 J. FIN. ECON. 181, 182-89 (1989) (showing that increase in the dividend payment by a company facing bad investment opportunities, measured by Tobin's Q ratio, significantly increases the company value, whereas an increase in dividend payment by a company with good investment opportunities does not have a beneficial effect on value).

(30.) See Kalay, supra note 26, at 211-12 (discussing the conflicts of interest between shareholders and bondholders regarding dividend constraints). Even if a covenant seems appropriate at the time debt is incurred, it may later come to be viewed as harmful if conditions change. For example, a company's investment opportunities may dry up--thereby increasing the optimal level of dividends--after a bond has been issued.

(31.) For example, the dispute between Burlington Northern Railroad Company and its bondholders about the release of certain collateral resulted in a two-year litigation. See Burlington Northern Railroad Co. Bond Settlement Approved, BUS. WIRE, Nov. 30, 1987 (noting that litigation commenced in 1985); see also Rievman v. Burlington N. R.R. Co., 618 F. Supp. 592 (S.D.N.Y., 1985).

(32.) Privately placed bonds, as the name indicates, are sold directly to institutions, not in a public offering, and are exempt from registration with the SEC. Since April 1990, when the SEC adopted Rule 144A, privately placed securities issued under this Rule can be traded among sophisticated investors designated as qualified institutional buyers. Consequently, privately placed debt can be divided into two types: the traditional privately-placed bonds and the Rule 144A bonds. The latter are now underwritten in a similar way that public bonds are, and their covenants are less restrictive than those on traditional privately placed bonds, making them closer in characteristics to public bonds. See text accompanying notes 37-73 infra. Also, the size of underwritten 144A bond issues is similar to that of public bonds issues, which is much larger than that of privately placed debt. The borrowers are also closer in nature to those of public debt: They are noninformation sensitive. Also, there is a considerable trading activity in bonds issued under SEC Rule 144A. This enhances the liquidity of this market, although it is not quite as liquid as the market for public bonds, where ownership is unrestricted. Finally, there is considerable fluidity of ownership in Rule 144A bonds, that is, they are quite likely to change ownership after issuance, unlike the case of private debt. See FRANK J. FABOZZI, BOND MARKETS, ANALYSIS AND STRATEGIES 148-49 (3d ed. 1996); Mark Carey, Stephen Prowse, John Rea & Gregory Udell, The Economics of Private Placements: A New Look, 2 FIN. MARKETS, INSTITUTIONS & INSTRUMENTS, No. 3, 1993, at 51-57. Economically, therefore, bonds issued under Rule 144A resemble public bonds, although they are legally privately placed bonds.

(33.) Commercial paper and trade credit are important additional sources of corporate debt financing.

(34.) This dichotomy is convenient for our primary objectives of contrasting the dispersion of ownership of a public bond with the concentration of ownership of a private loan and exploring the implications of that dispersion for liquidity and covenant structure. One can also distinguish debt by maturity (short term vs. long term) or according to whether the interest rate on the debt is fixed or floating. Given the characteristics of bonds issued under Rule 144A, see note 32 supra, we will not generally classify these bonds in either category, although their characteristics make them closer to public corporate bonds.

(35.) See Carey et al., supra note 32, at 13-22.

(36.) See id. at 7-8.

(37.) See Kahan & Tuckman, supra note 26, at 257 (finding in sample of privately placed bonds by publicly traded companies, mean number of lenders was 3.9 and median was 2).

(38.) For example, seven banks provided a $885 million loan to finance the 1989 leveraged buyout of GAF Corp. See Credit Agreement among G Industries Corp. et al. and Chase Manhattan Bank et al. (Mar. 28, 1989).

(39.) See Kahan, supra note 8, at 584 (estimating that households hold only 11-13% of corporate bonds).

(40.) We are not aware of any study that reports beneficial ownership figures for public corporate bonds. The best source of data on bond ownership is BEST'S MARKET GUIDE CORPORATE BONDS (1995) (the last year the guide was published) which lists insurance company holdings of each outstanding bond. To derive an estimate of overall bond holdings, we randomly selected twelve public bond issues with a principal amount of $250 million, calculated various ownership figures for insurance companies, and determined overall ownership figures assuming that other owners held the bonds in the same concentration as insurance companies do. Consistent with Federal Reserve figures, the insurance companies held about 44% of the corporate bonds in our sample. See FED. RESERVE FUNDS FLOWS, supra note 14, at 87 tbl.L212 (showing 1994 holdings of corporate and foreign bonds by life and other insurance companies was $889 billion, or 36% of total corporate and foreign bonds outstanding).

(41.) To represent these dispersed bondholders, the Trust Indenture Act requires the appointment of an indenture trustee for each corporate bond. However, as we shall discuss in detail below, the indenture trustee plays no substantial role in monitoring or renegotiating a loan. The indenture trustee plays some role in enforcing indenture provisions (once a breach has been detected), but even in this regard the indenture trustee functions poorly. See text accompanying notes 90-104 infra.

(42.) The number of beneficial holders of stock of public companies ranges from 100 to over 1,000,000. Telephone Interview with John Wilcox, Georgenson & Co. (June 2, 1998). The New York Stock Exchange generally lists only companies with at least 2,000 shareholders. See Listing and Delisting of Securities, 2 N.Y.S.E. Guide (CCH) [paragraph] 2501 (1996).

(43.) See Elliott Asarnow, Pricing Corporate Loans, in THE YEARBOOK OF FIXED INCOME INVESTING, supra note 26, at 349 (noting that banks have traditionally been "buy-and-hold" investors). Although ownership of bank debt is sometimes transferred when one bank sells its loan portfolio to another bank, when it securitizes its loans (such as car loans or mortgages), or when it assigns a loan, bank loans are rarely traded. See id. (noting thin volume of secondary market bank loan trades). Private placement debt held by institutions such as insurance companies and pension funds can be more easily traded than bank debt, but is still far less tradable than corporate bonds. See Carey et al., supra note 32, at 51-57. The trading of private placement debt is regulated by the Securities Act of 1933, 15 U.S.C. [subsections] 77a-77aa (1994 & Supp. 1996), and the rules promulgated thereunder, 17 C.F.R. pt. 230 (1998), which limit the liquidity of privately placed debt. Because private debt is less liquid than public debt, it is allocated in equilibrium to long-term investors who can depreciate the higher cost of liquidity over a longer horizon. See Yakov Amihud & Haim Mendelson, Asset Pricing and the Bid-Ask Spread, 17 J. FIN. ECON. 223, 228 (1986) [hereinafter Amihud & Mendelson, Asset Pricing]; Yakov Amihud & Haim Mendelson, Liquidity, Asset Prices and Financial Policy, FIN. ANALYSTS J., Nov.-Dec. 1991, at 56-58 [hereinafter Amihud & Mendelson, Liquidity].

(44.) See Carey et al., supra note 32, at 44-46.

(45.) Some corporate bonds trade on organized exchanges, such as the New York Stock Exchange, but most corporate bonds trade over-the-counter and in negotiated transactions between investors and dealers. See FABOZZI, supra note 32, at 147. To get a sense of bond trading, it may be illustrative to consider the sample of corporate bonds that we used to derive ownership data. See note 40 supra. In that sample, during 1994, 6.5% of the insurance companies sold all their holdings of a bond issue, 32.3% bought (and did not sell) bonds of an issue which they previously did not hold, and 12.7% changed the amount of bonds of a single issue held in other ways. (Percentages are of the number of insurance company owners at the beginning of 1994. For this calculation, we excluded two bonds issued in 1994.)

(46.) In companies under financial distress, there are investors (often called "vulture investors") who buy large blocks of public bonds in order to gain a strong bargaining position when negotiating with management. See generally Edith S. Hotchkiss & Robert M. Mooradian, Vulture Investors and the Market for Control of Distressed Firms, 43 J. FIN. ECON. 401 (1997).

(47.) Fluidity in ownership also reduces the incentives of bondholders to become informed and to engage in monitoring to the extent that the value of the information obtained accrues over time.

(48.) See Amihud & Mendelson, Liquidity, supra note 43, at 56.

(49.) See id.; Yakov Amihud & Haim Mendelson, Liquidity and Asset Prices: Financial Management Implications, FIN. MGMT., Spring 1988, at 5.

(50.) See William L. Silber, Discounts on Restricted Stock: The Impact of Illiquidity on Stock Prices, FIN. ANALYSTS J., July-Aug. 1991, at 60, 60-64 (comparing prices of restricted stock with prices of publicly traded stock from same company to demonstrate relationship between illiquidity and value of stock).

(51.) The value of asset liquidity is greater when the asset has higher volatility and when there is greater asymmetry of information between investors in the assets. These problems are more severe for stocks.

(52.) Empirical evidence shows that privately placed corporate bonds have a higher yield to maturity, compared to similar-risk public bonds. See Burton Zwick, Yields on Privately Placed Corporate Bonds, 35 J. FIN. 23, 23-29 (1980). Evidence also shows that less liquid U.S. Treasury notes have higher yields than more liquid U.S. Treasury bills. See Kenneth D. Garbade, Bankers Trust Co., Analyzing the Structure of Treasury Yields: Duration, Coupon and Liquidity Effects, in TOPICS IN MONEY AND SECURITIES MARKETS 3-4 (1984); Yakov Amihud & Haim Mendelson, Liquidity, Maturity and the Yields on U.S. Treasury Securities, 46 J. FIN. 1411 (1991); Avraham Kamara, Liquidity, Taxes, and Short-Term Treasury Yields, 29 J. FIN. & QUANTITATIVE ANALYSIS 403, 405-09 (1994).

(53.) See William F. Sharpe, Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk, 19 J. FIN. 425, 426 (1964).

(54.) To be sure, noninstitutional investors can indirectly access the private debt market by purchasing shares of mutual funds or of institutions that in turn invest in private debt. But this is costly too. There are agency costs when there is an intermediary that manages the debt portfolio and the composition of that portfolio may not suit some investors' objectives. Indeed, closed-end funds usually trade at a discount. In addition, open-end mutual funds are generally limited to hold no more than 15% of their net assets in illiquid securities. See Revision of Guidelines to Form N-1A, Securities Act Release No. 6927, SEC Docket (CCH), at 1659 (Mar. 12, 1992) (increasing permitted percentage from 10% to 15%).

(55.) See Robert C. Merton, A Simple Model of Capital Market Equilibrium with Incomplete Information, 42 J. FIN. 483, 500-04 (1987) (proving that an increase in a security's investor base increases its value); Yakov Amihud, Haim Mendelson & Jun Uno, Number of Shareholders and Stock Prices: Evidence from Japan, 54 J. FIN. (forthcoming June 1999) (showing that when investors are able to trade a security in smaller units, its value rises); Sandip Mukherji, Yong H. Kim & Michael C. Walker, The Effect of Stock Splits on the Ownership Structure of Firms, 3 J. CORP. FIN. 167 (1997) (showing that stock splits increase the numbers of both individual and institutional shareholders, and positively affect stock values); Chitru S. Fernando, Srinivasan Krishnamurthy & Paul A. Spindt, Does Share Price Affect Marketability? Evidence from Mutual Fund Share Splits (May 1997) (Working Paper, Freeman School of Business, Tulane University, on file with Stanford Law Review) (showing that splits of mutual funds shares, while not affecting their values, increase the number of their shareholder accounts because of the greater convenience in trading their shares, thus enhancing their marketability and increasing the demand for the units).

(56.) When private debt is incurred, the yield on the debt provides the issuer with information on the lender's evaluation. But through the term of the private debt, the company obtains no further information. Thus, by issuing a private rather than a pubic bond, the company forgoes the valuable information that is continuously provided by the public market on the company's debt value.

(57.) See generally Ronald J. Gilson & Reinier Kraakman, The Mechanisms of Market Efficiency, 70 VA. L. REV. 549 (1984) (discussing how public information is translated into security prices). Not all transactions are informative. Some investors do not trade on information but instead buy and sell bonds (and other financial instruments) because of their liquidity needs. They buy bonds when they have excess cash that they wish to convert into assets to earn higher yields, and sell bonds when they need cash. Such liquidity-motivated transactions convey no information as to the value of the bond. However, there are sufficient other investors who are informed, and who trade on their information. Their transactions will reflect the information they have and will affect the market price.

(58.) The informational content of bond prices is relatively more valuable for companies whose stock is not publicly traded, either because they are closely held or because they are units of larger publicly held companies. Companies whose stock is publicly traded can glean a substantial amount of information about how outsiders view the company from moves in the company's stock price.

(59.) Cf. Jos van Bommel, Messages from Market to Management: The Case of IPOs (Working Paper 1997) (arguing that managers of firms that go public learn information from the price of their shares that helps them make investment decisions).

(60.) For example, suppose a retail company makes a strategic decision to replace its low-price merchandise that appeals to middle-income clientele with luxury goods that appeal to an upperscale clientele that can pay higher prices. While the company insiders consider this a value-increasing decision, outsiders may consider it a bad strategy that will reduce the company's value. The evaluation of outsiders is based on vast amounts of information on the state of the economy, the marketing environment, and the reaction of competitors, that company insiders may not have. This may serve as a signal to the company insiders and lead to reassessment of their newly announced strategy.

(61.) It could be argued that public bondholders do not need covenants since they are protected by the covenants contained in the private debt issued by their company. To be sure, public bondholders benefit from covenants contained in private debt when these covenants deter debtor misbehavior. See Kahan, supra note 8, at 596-98 (explaining how third party security holders can benefit from covenants between a corporation and its other creditors). But public bondholders may also be hurt by such covenants when private debtholders obtain benefits as a result of an actual or impending violation that comes at the expense of other creditors. See id.; see also Asquith & Wizman, supra note 18, at 212 (reporting that where the same company had bonds outstanding with strong protection and no protection, bonds with no protection fared worse in a leveraged buyout than bonds with strong protection). For example, holders of private debt may receive additional collateral or higher interest in consideration for renegotiating or waiving a covenant, thereby depleting the assets available to repay the public bondholders. See, e.g., Messod D. Beneish & Eric Press, Costs of Technical Violation of Accounting-Based Debt Covenants, 68 ACCT. REV. 233, 248 (1993) (finding that increased collateral accounts for a third of new restrictions imposed as a result of violation); Amy Patricia Sweeney, Debt-Covenant Violations and Managers' Accounting Responses, 17 J. ACCT. & ECON. 281 (1994) (finding that in more than half of the cases where private debtholders did not waive a breach, they received additional collateral; in the remainder, they received a higher interest rate or stricter covenants). This latter effect is especially important for state-of-the-firm covenants which, as discussed in the text accompanying notes 66-71 infra, serve only a limited deterrence function and which are the most frequently violated covenants.

(62.) See, e.g., Kevin C.W. Chen & K.C. John Wei, Creditors' Decisions to Waive Violations of Accounting-Based Debt Covenants, 68 ACCT. REV. 218, 220-21 (1993) (finding that private debt agreements are violated before public debt agreements); Kahan & Tuckman, supra note 26, at 259 (comparing covenants in publicly issued and privately placed bonds with similar ratings); Sweeney, supra note 61, at 290 (finding that private debt agreements are violated before public debt agreements); Jeffrey Abt, Bank Loan Agreement Covenant Analysis (1993) (unpublished manuscript, on file with the Stanford Law Review) (analyzing covenants in syndicated bank loans to companies with publicly rated senior debt). For a general discussion and review of evidence, see Carey et al., supra note 32, at 28-32.

(63.) See Kahan & Tuckman, supra note 26, at 259.

(64.) Companies are also permitted to pay dividends to the extent that they raise new capital from the sale of stock.

(65.) See Kahan & Tuckman, supra note 26, at 260 (finding that private placements on average permitted companies to distribute 54% of earnings as dividends, while corporate bonds that restrict dividends on average permitted companies to distribute 95% of earnings as dividends).

(66.) See id. at 260-63. Other covenants containing broad restrictions on a company's activities, such as prohibitions of mergers and changes in the company's business nature, function similarly to financial ratio covenants. See id. at 264. They enable creditors to assess the effects of these action ex post and to block only those which are harmful.

(67.) Compare, for example, a dividend restriction--commonly found in both corporate bond indentures and private debt agreements--with a minimum net worth covenant--commonly found in private debt agreements, but not in corporate bond indentures. Both of these covenants relate to the minimum amount of equity that must remain in the company. In a dividend restriction, that minimum is established by the equity in the company when the debt is incurred less the initial reservoir, plus the amount of profits made after the debt is incurred that may not be distributed as dividends. In a net worth covenant, this minimum is established by the initial amount of net worth plus (usually) a specified part of the profits made after the debt is incurred. A dividend covenant, however, can only be violated by an affirmative decision by the company--to pay a dividend while the company's equity is below the minimum equity threshold or in an amount that would cause the equity to fall below that threshold. In the case of a net worth covenant, falling below the threshold--for example, as a result of losses--is a per se violation of the covenant. Since companies do not have full control over the amount of profits or losses they make, the net worth covenant may be violated without an affirmative decision by the company to do so. See generally Kahan & Tuckman, supra note 26.

(68.) See Beneish & Press, supra note 61, at 238 (finding that violations of tangible net worth, working capital or current ratio, and minimum leverage covenants account for 90% of covenant violations); Sweeney, supra note 61, at 290 (finding that four most commonly breached covenant categories are covenants relating to net worth; working capital and current ratio; debt-equity ratio; and interest coverage and net income).

(69.) Cf. Smith & Warner, supra note 26, at 125 (noting that it is difficult to directly constrain production policy and that covenants generally do not entail direct constraints).

(70.) See George G. Triantis & Ronald J. Daniels, The Role of Debt in Interactive Corporate Governance, 83 CAL. L. REV. 1073, 1093 (1995) (explaining that such covenants ensure that the lender is "immediately notified of the borrower's financial difficulties" and is given "the opportunity to act well before the borrower becomes insolvent").

(71.) Consistent with this analysis, studies find that creditors commonly waive covenant violations without extracting concessions. See Beneish & Press, supra note 61, at 249 (finding evidence of renegotiation in 61 of 91 firms that violated covenants); Chen & Wei, supra note 62, at 224 (finding 57 out of 128 finns received temporary or permanent waivers); Sweeney, supra note 61, at 281 (reporting that lenders required concessions in 52% of cases). These studies overstate the extent to which creditors demand concessions to relax covenants as they include only cases where a covenant has actually been breached and fail to include cases where a covenant has been relaxed prior to the time where a breach would have occurred.

(72.) See Kahan & Tuckman, supra note 26, at 265-67.

(73.) Borrowers in the private market are more willing to grant lenders access to nonpublic information because they have greater control over the identities of their lenders.

(74.) See Carey et al., supra note 32, at 40-44; James R. Booth, Contract Costs, Bank Loans, and the Cross-Monitoring Hypothesis, 31 J. FIN. ECON. 25 (1992) (providing empirical evidence for the importance of monitoring in bank lending agreements); see also Douglas W. Diamond, Monitoring and Reputation: The Choice between Bank Loans and Directly Placed Debt, 99 J. POL. ECON. 689 (1991) (constructing a model of borrower choice between issuing bonds, without monitoring, and borrowing through a bank that monitors to alleviate moral hazard).

(75.) Raghuram Rajan & Andrew Winton, Covenants and Collateral as Incentives to Monitor, 50 J. FIN. 1113 (1995) (arguing that covenants are designed to provide incentives to monitor).

(76.) See Mitchell Berlin & Loretta J. Mester, Debt Covenants and Renegotiation, 2 J. FIN. INTERMEDIATION 95 (1992) (arguing that it is easier to renegotiate private debt than public debt and that this difference accounts for tighter covenants in private debt); Carey et al., supra note 32, at 43.

(77.) See Beneish & Press, supra note 61, at 238 (reporting that most covenants that are violated are state-of-the-firm covenants); Sweeney, supra note 61, at 290 (same).

(78.) See Dianna Preece & Donald J. Mullineaux, Monitoring, Loan Renegotiability, and Firm Value: The Role of Lending Syndicates, 20 J. BANKING & FIN. 577 (1996) (providing empirical evidence that a small number of lenders makes it easier to renegotiate loans and that ease of renegotiability is a source of value to borrowers).

(79.) See note 71 supra.

(80.) See Carey et al., supra note 32, at 40-44 (noting that holders of private debt that regularly make "loans may profit from a reputation for being constructively flexible"); see also Ivo Welch, Why is Bank Debt Senior? A Theory of Priority Based on Influence Costs (Institute of Finance & Accounting, London School of Economics, Working Paper No. 217, 1995) (tying contractual seniority of bank debt over public debt to banks reputational interests). See generally Amoud W.A. Boot, Stuart I. Greenbaum & Anjan V. Thakor, Reputation and Discretion in Financial Contracting, 83 AM. ECON. REV. 1165 (1993) (arguing that parties to financial contracts have reputational incentives to exercise discretion in a reasonable manner).

(81). A consent payment is either paid explicitly or the "payment" may be embedded in the terms of a tender or exchange offer where the corporate bonds of holders who consent to the amendment are repurchased, or exchanged for a new issue of bonds, at a value in excess of the value of the bonds that are not tendered or exchanged. See Kahan & Tuckman, supra note 8, at 502.

(82.) See id. at 513. The behavior of public bondholders when a covenant has been breached is difficult to assess. Due to the frequency and type of covenants in public bonds, covenant breaches occur relatively rarely, and due to the lack of monitoring, some of the breaches that do occur may not be detected. Moreover, since covenants in public bonds are usually breached only by voluntary actions of the company--for example, the payment of a dividend when such payment is prohibited by the restricted payments covenant--public bondholders ought to take stricter enforcement actions after a covenant breach than holders of private debt should take.

(83.) But see Leland E. Crabbe & Christopher M. Turner, Does the Liquidity of a Debt Issue Increase with Its Size? Evidence from the Corporate Bond and Medium-Term Note Markets, 50 J. FIN. 1719, 1733 (1995) (finding that the total amount of outstanding bonds by the same borrower, rather than the amount of outstanding bonds of the same issue, affects liquidity).

(84.) See Carey et al., supra note 32, at 13-22 (discussing categories of issuers that choose to borrow in the private debt market). Similarly, lenders will tend to prefer to participate in the private loan market to the extent that they have a comparative advantage in monitoring and place a lower value on liquidity. Those who value liquidity and ease of diversifiability, and those who are less able to monitor, renegotiate, and enforce the terms of a loan, will participate in the public bond market. Banks and insurance companies thus tend to lend directly to corporate borrowers, while individual investors and mutual funds tend to purchase corporate bonds.

(85.) This has occasionally been a source of conflict for public accounting firms. See SEC and AICPA Announce a New Independence Standards Board, SEC News Release, May 21, 1997, available on WESTLAW, ALLNEWS file (discussing the formation of a new private sector body to establish independence standards for the auditors of public companies).

(86.) To protect the company's interest in nonpublic information, the initial supertrustee, as well as any successor trustee, may be asked to sign a confidentiality agreement similar to agreements between borrowers and lenders in the private debt market.

(87.) The indenture would specify how the determination of cause is made (e.g., by a court or arbitrator) and how a successor trustee is selected.

(88.) In this alternative as well, bondholders should be permitted to remove the supertrustee for cause. If so removed, the new trustee would be chosen in the same way as a successor trustee if the initial trustee were replaced in an election.

(89.) The present indenture trustee is initially appointed by the company and does not come up for periodic election. Holders of a majority of bonds, however, may generally remove the trustee at any time. If removed, the company appoints a successor trustee. Holders of a majority of bonds may replace the company-appointed trustee with one of their own choice within one year after the appointment. See, e.g., Model Simplified Indenture, 38 BUS. LAW. 741, 761 (1983).

(90.) See text accompanying notes 72-75 supra.

(91.) See Trust Indenture Act of 1939 [sections] 315(a), 15 U.S.C. [sections] 77ooo(a) (1994).

(92.) See Trust Indenture Act of 1939 [sections] 315(b), 15 U.S.C. [sections] 77ooo(b) (1994).

(93.) See Model Simplified Indenture, supra note 89, at 763 (granting trustee power to approve only technical amendments without consent of bondholders).

(94.) For example, the majority bondholders may also hold stock or another issue of bonds or they may receive a consent payment that was not offered to the minority bondholders, for agreeing to the covenant changes.

(95.) See Carey, supra note 25, at 1385 (finding lower incidence of default in private bonds than in public bonds with ex ante lower risk). Moreover, even in private debt agreements (which our proposal is intended to emulate) with several lenders, changes in core financial terms generally require the consent of every lender, while covenant changes merely require the consent of holders of a majority of the debt. This suggests that private debt creditors value the right to veto changes in core financial terms much more than they value the right to veto covenant changes. The benefit of permitting a supertrustee for public bonds to approve changes in core financial terms, however, exceeds the benefit of making such changes subject to majority control in private debt. Since private debt ownership is concentrated, lenders' retention of control over changes in core terms occurs in a context where negotiation over such changes and agreement about them is quite feasible. Public debt ownership, however, is dispersed, and unanimous agreement is difficult to achieve. Therefore, public bondholders may want to cede to the supertrustee some power to approve changes in the bond's core terms, even if private lenders prefer to retain individual control.

(96.) Such limits could be procedural or substantive. On the procedural side, for example, a specified fraction of bondholders--say, a third--could challenge the supertrustee's decision within a specified period, say thirty days. If challenged, the supertrustee should seek the bondholders' approval for the proposed change. This procedure naturally shifts the burden of collective action over to those bondholders who disagree with the proposed changes in the core terms, and will make opposition to such changes more costly. On the substantive side, the supertrustee's power could be limited, for example, to reduce the interest rate by a specified percentage or to postpone the maturity of principal payment for a specified time period.

(97.) Mitigating the collective action problem of bondholders with regard to changes to core terms may be less important than mitigating that problem with regard to covenant changes. The need to obtain bondholders' consent to changes in core financial terms typically arises in a workout, when the company will also seek concessions from many other creditors, including other lenders and trade creditors. Thus, even if public bondholders were represented collectively by a supertrustee, the collective action problem facing creditors at large would remain and may well impede a workout. By contrast, the need for covenant changes arises more frequently but requires a less broad consent, usually by holders of a single or a few debt issues, and thus does not entail a larger collective action problem.

(98.) See, e.g., Robert Gertner & David Scharfstein, A Theory of Workouts and the Effects of Reorganization Law, 46 J. FIN. 1189 (1991) (discussing holdout problems in workouts).

(99.) Cf. Mark J. Roe, The Voting Prohibition in Bond Workouts, 97 YALE L.J. 232 (1987) (arguing that majority bondholder consent, rather than unanimous consent, should be required for changes of core financial terms).

(100.) See text accompanying notes 142-143 infra.

(101.) In this respect, the role of the supertrustee would resemble the role of a creditors' committee in a Chapter 11 reorganization. See 11 U.S.C. [subsections] 1102-1103 (1994) (discussing the appointment, powers, and duties of the creditors' committee).

(102.) See Model Simplified Indenture, supra note 89, at 756-59 (establishing enforcement rights of bondholders and duties of trustee); see also Dresner Co. Profit Sharing Plan v. First Fidelity Bank, No. 95 Civ. 1924 (MBM), 1996 U.S. Dist. LEXIS 17913 (S.D.N.Y. Dec. 4, 1996) (refusing to dismiss a complaint against indenture trustees who allegedly violated the "prudent person" standard by failing to seek adequate protection of the interests of bondholders in bankruptcy proceedings).

(103.) See text accompanying notes 108-125 infra (discussing incentives resulting from potential liability, compensation scheme, and reputation).

(104.) See, e.g., Jeffrey N. Gordon, Shareholder Initiative: A Social Choice and Game Theoretic Approach to Corporate Law, 60 U. CIN. L. REV. 347, 347-53 (1991) (discussing limited power of shareholders to give directions to directors). The reasons for precluding bondholders from giving directions to the supertrustee, however, are stronger than the corresponding reasons in the shareholder context. We favor limits on bondholders' authority in order to overcome the bonding problem created by the fluidity in the lending relationship. No equivalent problem exists in the shareholder context. Rather, the standard justification for limiting shareholders' authority seems to be that shareholders are less well informed than directors about how the company should be run (though shareholders, realizing that they are ordinarily less well informed, may well want to retain authority but use it sparingly). See id. at 353-57 (rejecting standard justifications for shareholder authority). We also note that while it is, at least in theory, easier to replace directors than it is to replace the supertrustee, directors, especially inside directors who dominate corporate boardrooms, are subject to greater conflicts of interest than is the supertrustee.

(105.) Though greater than the powers of U.S. indenture trustees, the powers of U.K. trustees, at least in practice, fall short of the powers of a supertrustee. For example, if a company has breached a material covenant, a U.K. trustee would seek bondholder approval for an amendment that, say, waived the breach in exchange for higher interest payments or additional collateral. In our proposed scheme, a supertrustee would not seek bondholder approval for such an amendment.

(106.) We are grateful to Philipp v. Randow, University of Osnabrueck, and David Norris, Law Debenture Trust Company, for information about the U.K. practice. See also David B. Citron, The Incidence of Accounting-based Covenants in UK Public Debt Contracts: An Empirical Analysis, 25 ACCT. & BUS. RES. 139, 143 (1995) (finding that seventy percent of bonds in sample contained no accounting-based covenants).

(107.) See text accompanying notes 90-94 supra.

(108.) Trust Indenture Act of 1990 [sections] 315 (c), 15 U.S.C. [sections] 77ooo(c) (1994).

(109.) The "prudent man" standard does not pose a major problem for the indenture trustee. Breaches of covenants in public bonds under the indenture trustee governance structure tend to be rare, both because there are fewer and less tight covenants and because most covenants can be breached only by an affirmative act of the company. Some breaches of covenants also may not be detected. Moreover, the nature of the covenants makes enforcement decisions less complex. Finally, indenture trustees often consult bondholders if a serious breach occurs, thus limiting their exposure to suits.

(110.) Because the supertrustee's efforts are likely to vary with the type of company and the type of debt, so should the compensation. The compensation should be increasing with the number and complexity of the bond covenants, with the likelihood that covenants will be breached, and with the difficulty of monitoring, renegotiating, and enforcing the covenants. In general, bonds issued by companies with greater credit risk and less transparent operations will impose greater monitoring, renegotiation, and enforcement burdens, and the compensation of the supertrustee should accordingly be greater. But these issuers also have the most to gain from a reduction in the agency costs of debt, and thus the most to gain from issuing supertrusteeship bonds.

(111.) Indenture trustee fees for an unsecured bond generally range from $5,000 to $10,000 per year. Telephone Interview with Mike Campbell (May 27, 1998).

(112.) Companies that cannot recoup the higher fees in this manner are better off with the current bond governance structure of indenture trusteeship.

(113.) This may be more important when the supertrustee is not ordinarily replaceable by bondholders and need not fear removal if it shirks on its tasks.

(114.) This risk can be hedged through the capital market by buying or selling financial claims whose values are contingent on the values of Treasury securities, but such hedging is costly.

(115.) It would also limit the source of supply of supertrusteeship services to large institutional investors. This problem may be negligible if other potential supertrustees could offer their services in return for incentive compensation in the form described above.

(116.) To be sure, investors can also reputationally monitor the company and not buy bonds from those that abuse their creditors. However, a company's reputational interest is substantially lower than a supertrustee's since the supertrustee wants to act as such for many bond issues, while most companies access the public debt market only once every few years.

(117.) See Kahan, supra note 8, at 583-86.

(118.) To be sure, investors can also use the information gained from monitoring covenant compliance in order to make investment decisions. Such use of information, however, does not directly reduce the agency costs of debt.

(119.) The differences in payoff between reputational monitoring and covenant compliance monitoring thus resemble the differences in payoff between obtaining information to determine whether to buy stock in an IPO and obtaining information to determine how to vote one's shares. Due to these differences in payoffs, shareholder voting decisions are commonly viewed as less perfect than the pricing of shares in capital markets. See, e.g., Lucian Arye Bebchuk, Federalism and the Corporation: The Desirable Limits on State Competition in Corporate Law, 105 HARV. L. REV. 1435, 1473 n. 127 (1992) (noting that an individual has a greater incentive to make an informed purchase decision than an informed voting decision because a stock purchase will more directly impact her financial interests).

(120.) To be sure, investors in a particular bond issue will know less than the issuer about whether the supertrustee in that specific issue acted faithfully. However, the supertrustee has little to gain, and a lot to lose, by colluding with an issuer. The supertrustee's gain can take the form of either a direct bribe or of the promise of future business from the issuer itself. Our proposal envisions severe penalties for bribes, which, together with reputational penalties and incentive compensation, should be sufficient to deter most of such conduct. The promise of future business, on the other hand, is of limited value to the supertrustee in a context where the supertrustee is asked to tolerate actions that endanger the company's continued viability--and thus its very ability to bestow future business on the supertrustee. Though this form of collusion is harder to detect and sanction than a direct bribe, incentive compensation as well as the possibility of reputational penalties should be sufficient to deter it as well.

(121.) See text accompanying notes 79-82 supra.

(122.) See Richard Carter & Steven Manaster, Initial Public Offerings and Underwriter Reputation, 45 J. FIN. 1045 (1990) (finding that underwriter reputation is significant).

(123.) See FREDERICK D. LIPMAN, VENTURE CAPITAL AND JUNK BOND FINANCING 325 (1996) ("Before a default under an indenture occurs, the trustee under the Trust Indenture Act has very limited duties, primarily that of a reporting nature.")

(124.) See 11 U.S.C. [sections] 1102(a)(1) (1994) (ordering the U.S. trustee to appoint a creditors' committee after the order for relief under Chapter 11); see also 11 U.S.C. [sections] 503(b)(5) (1994) (permitting indenture trustee to receive additional compensation for serving on a creditors' committee, thus providing an incentive for a trustee to serve). Still, the trustee's post-default functions are limited. See 11 U.S.C. [sections] 1102(b)(1) (1994) (noting that creditors' committees consist of seven members, and the six members besides the trustee will generally be large creditors of the company); see also Telephone Interview with Mike Campbell (May 27, 1998) (recounting that the trustee often resigns if a default occurs, so that the trustee who serves on a creditors' committee is a successor trustee, rather than the original trustee in place when the bonds were issued).

(125.) See AMERICAN BAR FOUNDATION, COMMENTARIES ON INDENTURES 250 (1986) ("the functions of the Trustee under ordinary conditions are largely administrative").

(126.) 15 U.S.C. [subsections] 77aaa-bbbb (1994).

(127.) See Trust Indenture Act of 1939 [sections] 310(a), 15 U.S.C. [sections] 77jjj(a) (1994).

(128.) See Trust Indenture Act of 1939 [sections] 315(a), 15 U.S.C. [sections] 77ooo(a)(1) (1994).

(129.) See Trust Indenture Act of 1939 [sections] 315(b), 15 U.S.C. [sections] 77ooo(a)(2) (1994).

(130.) See Trust Indenture Act of 1939 [sections] 316(a)(1), 15 U.S.C. [sections] 77ppp(a)(1) (1994).

(131.) See Trust Indenture Act of 1939 [sections] 315(d), 15 U.S.C. [sections] 77ooo(d)(2) (1994).

(132.) See Trust Indenture Act of 1939 [sections] 315, 15 U.S.C. [sections] 77ooo (1994).

(133.) See Trust Indenture Act of 1939 [sections] 316(b), 15 U.S.C. [sections] 77ppp(b) (1994). Other mandatory provisions of the Act include the conflict of interest rules for the indenture trustee, which authorize a single holder to petition the court for removal of the indenture trustee ([sections] 310); rules on notifying bondholders about defaults ([sections] 315(b)); rules on the bondholders' power to sue the company for overdue principal and interest payments ([sections] 316(b)); and rules on the release of collateral securing bond obligations ([sections] 314(d)). While these mandatory provisions are not inconsistent with any principal elements of our supertrustee proposal, they may constrain the ability of companies to implement the proposal in an optimal manner.

(134.) See Trust Indenture Act of 1939 [sections] 315(c), 15 U.S.C. [sections] 77ooo(c) (1994).

(135.) See text accompanying notes 107-109 supra.

(136.) See Trust Indenture Act of 1939 [sections] 315(d), 15 U.S.C. [sections] 77ooo(d) (1994); see also AMERICAN BAR FOUNDATION, supra note 125, at 251 (noting that under [sections] 315(d) the trustee is liable for negligence).

(137.) Section 315(d) is expressly modified by sections 315(a) and (b), which provide presumptively that, prior to default, the indenture trustee must fulfill only those duties expressly set forth in the indenture and that the indenture trustee may rely conclusively on compliance certificates provided by the company. See generally Trust Indenture Act of 1939 [sections] 315(a)-(b), (d), 15 U.S.C. [subsections] 77ooo(a)-(b), (d) (1994). In the present indenture trustee context, sections 315(a), (b), and (d) taken together mean that the negligence standard does not apply pre-default, since the trustee has virtually no pre-default duties. Section 315(d) applies to the indenture trustee post-default, but in this respect essentially duplicates the post-default "prudent person" standard of section 315(c). See generally 15 U.S.C. [sections] 77ooo(d) (1994).

(138.) See Section II.E supra.

(139.) Trust Indenture Act of 1939 [sections] 315(d), 15 U.S.C. [subsections] 77ooo(d) (1994) (emphasis added).

(140.) See Trust Indenture Act of 1939 [sections] 316(b), 15 U.S.C. [subsections] 77ppp(b) (1994).

(141.) Id.

(142.) In practice, the unanimous consent requirement is overcome by structuring a workout as an exchange offer in which those bondholders who consent participate. This structure, however, creates a holdout problem that impedes out-of-bankruptcy workouts. See Gertner & Scharfstein, supra note 98, at 1200-01 (describing and providing an example of this holdout problem).

(143.) See text accompanying notes 95-101 supra.

(144.) Trust Indenture Act of 1939 [sections] 304(d), 15 U.S.C. [sections] 77ddd(d) (1994).

(145.) See AMERICAN BAR FOUNDATION, supra note 125, at 249 (noting that the Trust Indenture Act "leaves the definition of `default' to the discretion of the draftsman").

(146.) Initial purchasers who plan to hold bonds to their maturity may fully price the provisions even if the provisions are not priced in the secondary market. However, as discussed above, see text accompanying notes 47-55 supra, one of the advantageous features of corporate bonds is their liquidity.

(147.) See Marcel Kahan & Michael Klausner, Standardization and Innovation in Corporate Contracting (or "The Economics of Boilerplate"), 83 VA. L. REV. 713 (1997) (examining network externalities in event-risk bond covenants).

(148.) Innovators of a new governance scheme may benefit from introducing it, even if they can be easily copied by others. As an example, consider the introduction of shareholders' rights plans, known as "poison pills." In 1984, two law firms were involved in about 75% of all poison pill adoptions. This share quickly eroded, and in 1990 these two firms had less than 10% of the market, with the next seven firms having a similar share. See Robert Comment & G. William Schwert, Poison or Placebo? Evidence on the Deterrence and Wealth Effects of Modern Antitakeover Measures, 39 J. FIN. ECON. 3, 12-13 (1995). This result was brought about by competition and emulation, but still it did not deter the initiation of this important innovation in corporate governance.

(149.) See text accompanying notes 114-115 supra.

(150.) See Carey et al., supra note 32, at 7-9 (discussing the relative issue size for bank loans, private debt placements, and publicly traded bonds). One reason for that is the diversifiability benefits associated with the bonds being publicly traded, and the fact that liquidity benefits are increasing in scale. See Amihud & Mendelson, supra note 49, at 6-8 (discussing the costs and benefits of increased liquidity).

(151.) See Elizabeth Tashjian, Ronald C. Lease & John J. McConnell, Prepacks: An Empirical Analysis of Prepackaged Bankruptcies, 40 J. FIN. ECON. 135 (1996) (reporting that costs of prepackaged bankruptcies and out-of-court restructurings are less than costs of bankruptcies).

(152.) In public debt negotiations there is often the problem of holdout, by which a small group of holders of a bond issue can squeeze beneficial terms from the rest of the holders of the issue in return for their consent. It is sometimes the case that the company slips into Chapter 11 bankruptcy because an agreement cannot be reached; then, both shareholders and bondholders are worse off. If the supertrustee can make decisions on some changes in the terms of the bond without bringing it to bondholders for approval, it prevents this problem of holdout. See text accompanying notes 96-99 supra.

(153.) See Edward I. Altman, A Further Empirical Investigation of the Bankruptcy Cost Question, 39 J. FIN. 1067 (1984) (providing empirical evidence of both the direct and indirect costs of bankruptcy).

(154.) See Yakov Amihud & Baruch Lev, Risk Reduction as a Managerial Motive for Conglomerate Mergers, 12 BELL J. ECON. 605 (1981) (demonstrating that managers engage in conglomerate mergers to decrease their employment risk).

Yakov Amihud,(*) Kenneth Garbade,(**) and Marcel Kahan(***)

(*) Research Professor of Finance, Stern School of Business, New York University, and at the Faculty of Management, Tel Aviv University.

(**) Clinical Professor of Finance, Stern School of Business, New York University.

(***) Professor of Law, New York University School of Law.
COPYRIGHT 1999 Stanford Law School
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Amihud, Yakov; Garbade, Kenneth; Kahan, Marcel
Publication:Stanford Law Review
Geographic Code:1USA
Date:Feb 1, 1999
Previous Article:Enjoining the constitution: the use of public nuisance abatement injunctions against urban street gangs.
Next Article:Anachronism of the moral sentiments? Integrity, postmodernism, and justice.

Related Articles
Board practices: its' time to name a chief governance officer.
IFAC steps up campaign to restore investors' faith.
Corporate Governance journals from Inderscience.
Corporate Governance journals from Inderscience.
'Corporate Governance'.
NBO Wins Regional Corporate Governance Award.
New journal on corporate governance.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters