Exploring the limits of contract design in debt financing.
INTRODUCTION I. INCOMPLETE CONTRACTS AND MANDATORY LEGAL STANDARDS II. CONTRACT REMEDIES, THIRD PARTIES, AND PROPERTY RIGHTS III. COVENANT-LITE DEBT CONCLUSION
Two Articles in this issue, one by Professor Rock (1) and the other by Professors Adler and Kahan, (2) draw renewed attention to the contracting challenges raised by debt financing. The occasion for revisiting the agency costs of debt, according to Professor Rock, is the successful alignment of the interests of shareholders and corporate managers over the past thirty years. (3) This alignment has been achieved by a combination of contract, market, and legal measures (such as changes in compensation structure, shareholder concentration and activism, and board ideology). (4) As a result of these developments, Professor Rock observes that "managers and directors today largely 'think like shareholders.'" (5)
All investors, including creditors, benefit from the correction of inefficient incentives that lead managers to entrench themselves, build conglomerate empires, and shirk or consume perquisites. Other stakeholders such as employees, suppliers, and customers also benefit from such correction. However, the convergence of managerial and equity interests threatens to increase the agency costs of debt because debtholder and shareholder interests diverge in other respects. Most notably, as faithful agents of their shareholders, managers are more likely to (a) forego lower-risk, profitable projects ("underinvestment"); (b) invest in higher-risk, unprofitable alternatives ("overinvestment" or "risk alteration"); (c) incur additional debt to further leverage the equity in the firm; and (d) distribute firm value to shareholders in the form of dividends or share repurchases.
Professor Rock suggests that the negative externality of shareholder-centrism has been aggravated by the significant and contemporaneous increase in corporate leveraging. (6) The proposition that leveraging increased while the agency problems of debt became more severe is puzzling. To act in their shareholders' best interests, however, managers should borrow up to the point at which the marginal cost of further debt financing equals that of equity financing. The benefits of debt financing include the discipline imposed on managers by regular mandatory payments of free cash flow and the tax deduction from interest payments, while the costs stem from the resulting increase in the likelihood of bankruptcy and the agency costs of debt. The debt investors bear these costs and, to the extent they are informed, will compel the shareholders to internalize those costs. If the agency costs of debt truly have risen during the past three decades, what then explains the contemporaneous increase in corporate leveraging over that period?
While it is possible that the bankruptcy costs of debt have decreased or that the tax benefits have increased over this period, another possibility more germane to this Response is that the quality of debtholder contracts has improved as borrowing has increased. To the extent that debt investors price agency costs, a firm can lower its cost of capital by reducing the inefficiencies of debtor-creditor conflict. Empirical studies show that contractual covenants are indeed priced by debt investors, giving borrowers incentives to agree to them. (7)
The mechanisms of debtholder governance have been subject to substantial examination in both law and finance scholarship. The literature reveals that covenants play important roles in mitigating the agency costs of debt and adverse selection. (8) Professor Rock's and Professors Adler and Kahan's Articles in this issue contribute to this body of literature by focusing on the limitations of the existing technology of contractual protections and proposing reforms of the governing legal rules to address these limitations. (9)
Although debt contracts offer significant protection to debtholders, the extent to which debt contracts mitigate agency problems might be limited by three features raised by Professors Rock and Professors Adler and Kahan. First, specifying and enforcing optimal protective covenants is costly, and the costs of some conceivable protections exceed the benefits. Second, contract law limits the parties' ability to provide for effective remedies for breaches of those covenants, particularly against third parties who either have control of, or benefit from, those breaches. Third, the parties may sometimes omit even covenants whose benefits outweigh their costs because of imbalances in market conditions or bargaining power; one example might be the period of covenant-lite bonds preceding the financial crisis of 2007. Professor Rock and Professors Adler and Kahan propose legal reforms to address these limitations, including (a) extending mandatory, legally imposed standards such as fiduciary duties and the duty of good faith; and (b) expanding contract remedies to allow enforcement of contract covenants against third parties. As discussed below, the available technology for debt contracting is more potent than it may appear and the incremental gains from the authors' proposals are probably not worth the costs. Part I explains that the typical remedy contracted for by lenders is the right to terminate and assume greater control of the debtor's decisionmaking. This is a distinctive and effective remedy that contrasts with the usual contract remedy of expectation damages; indeed, debt contracts do not provide for such damages. Part II suggests that the existing combination of debt covenants and security interests can achieve much of what Professors Adler and Kahan are seeking without the social cost of imposing greater information burdens on creditors. For example, the ability of managers of a firm to engage in either massive borrowing or a leveraged buyout will be constrained if the firm has given most of its assets as collateral for a prior loan. Part III begins to tackle the more puzzling phenomenon raised by the authors' examples: debt contracts sometimes do not incorporate the available technology and are both unsecured and light on covenants. Although the omissions might be the result of market failure, they may alternatively be efficient responses to changing market conditions.
I. INCOMPLETE CONTRACTS AND MANDATORY LEGAL STANDARDS
The academic discipline of contract theory is founded on the observation that many contracts are incomplete because transaction costs prevent parties from providing for the efficient set of obligations in each possible future state of the world. Consequently, parties are relegated to choosing second-best tools of contract design. To illustrate in the context of a debt contract, suppose that Project A is the value-maximizing alternative in state i, while Project B is the value-maximizing alternative in state j. The optimal complete contingent contract would oblige the firm to invest in the efficient project in each respective state of the world. However, providing in this respect for each possible future state of the world is costly--both at the front end of contract design and at the back end of enforcement. At the front end, parties must contemplate and define each relevant state of the world, and they must agree to and specify the contingently optimal investment for each state. At the back end, they must monitor and observe the debtor's actions and, if necessary, verify performance or breach during the enforcement proceedings in court. The front- and back-end cost of providing for different obligations in states i and j may outweigh the incremental gains. Therefore, it may be more efficient to lump the two states together, by requiring, for example, Project A (or prohibiting project B) in both states. In a debt contract, the parties might prohibit sales of assets in bulk, mergers, or future borrowing, rather than incur the contracting costs of specifying the conditions under which such sales, mergers, or borrowing are permitted. In practice, then, most covenants are either under- or overinclusive in proscribing undesirable behavior, and this incompleteness leaves behind residual agency costs.
Contract design is fundamentally an exercise in minimizing these residual costs. The technology of contract design is quite rich, allowing parties significant flexibility to tailor their contracts to their circumstances. (10) Two important mechanisms in the toolkit for reducing agency costs are (1) the use of standards (as opposed to rules) and (2) contingent control rights and renegotiation.
The first mechanism, the use of standards, invites the court to determine ex post whether the firm's decisions met the expectations of the parties--for example, whether the decision was reasonable given the realized state of the world. Performance standards, such as obligations of good faith, best efforts, and due diligence, convey this discretion to the court. These standards can be effective tools if the reason for contract incompleteness is the cost of specifying the optimal project in each state of the world. While performance standards save front-end negotiating and drafting costs, they also raise back-end enforcement costs, especially the costs of litigation and judicial error in applying the standard. In some cases, this trade is advantageous, especially if the probability of going to court is low. However, if the obstacle to contract completeness is the high cost of litigation and judicial error in verifying the existing state or the efficient project in that state, then the case for a standard is more controversial and complicated, (11) Indeed, a key reason that many commentators oppose the broadening of fiduciary duties or weakening of the business judgment or good faith defense is that such reforms would exacerbate the cost of judicial error and litigation. (12) Moreover, if the ex post determination of the court cannot be readily predicted by the firm's directors, the beneficial effect on managerial incentives is limited. Facing the risk of a surprise adverse judgment, potential directors may be reluctant to serve without indemnification or expensive professional counsel, and will exercise excessive caution in managing the firm (just as the expansion of medical malpractice liability may lead to defensive medicine). (13)
Despite the cost of the standards describe above, Professor Rock proposes that directors be subject to broader standards by regulation. (14) He suggests, for example, that the fiduciary duties of directors be recast to oblige them to maximize the value of the firm as a whole. (15) Lenders themselves could include contractual standards that mimic elements of fiduciary duties, requiring the firm to use reasonable efforts to maximize value and to retain only directors who assume this obligation. In practice, however, debt contracts typically do not use such broad obligation standards to correct the incentives of directors, even though the front-end cost of doing so is low. (16) This practice raises some doubt as to whether additional mandatory performance standards, such as broader fiduciary duties, would yield a contracting benefit net of their costs. (17) Moreover, I describe below the distinctive features of debt contracting (contingent control transfers and security interests) that are sufficiently potent such that, for most debtors, mandatory standards are likely to undermine rather than help debt financing.
The second mechanism, contingent control rights, transfers control over corporate assets from shareholders to debtholders upon covenant violation. (18) In this respect, the operation of debt covenants differs fundamentally from the conventional use of contract promises, particularly in one-shot transactions. If a seller, for example, promises to deliver a good in exchange for payment, the seller is liable for damages if she fails to deliver. The buyer generally cannot assume control of the seller's premises and manufacture or seize the good itself. The prospect of liability deters the seller from breaching, and if breach occurs, she must compensate the buyer for the loss of her expectancy interest. The breach of a debt covenant, by contrast, does not give rise to liability for expectation damages, but instead triggers a shift in control. For example, if the debtor fails to make a scheduled payment and the lender exercises its right to call the loan when market interest rates are lower than the contract rate, the lender is not entitled to collect the difference between the contract rate and the lending rate at the time of breach. Rather, the lender has the right to accelerate the maturity of the indebtedness and demand prompt payment of the principal and accrued interest.
The acceleration of maturity and calling of a loan give the lender the right to repayment--or to the seizure of assets to enforce repayment--but more often, the threat of acceleration leads to a shift in decisionmaking control to the lender and to renegotiation. (19) For example, in addition to a higher interest rate, the lender may ask for shorter maturity in a renegotiated agreement as well as management changes. Thus, in addition to deterring opportunism, debt covenants provide the triggers that enable the lender to intervene in governance. (20) Institutional lenders routinely set covenants tightly from the time of contracting, so that they are frequently tripped and renegotiated, even in the absence of impending financial distress. (21) Tight covenants protect lenders by ensuring greater bargaining power in renegotiation. (22) Provisions in debt contracts and debtor-creditor law--such as avoidable preferences, equitable subordination, and lender liability--provide incentives for the lenders to monitor compliance and act upon information of a violation without abusing their control or power. (23)
These control-shifting provisions in debt contracts are critical tools of corporate governance, but they effectively replace one imperfect agency relationship with another, less imperfect, one. The provisions address the worsening problem of shareholder governance during a firm's decline through the less severe problem of debt governance. There is substantial evidence that lender intervention following a covenant violation yields positive externalities. (24) Even when a lender exits rather than intervenes, the lender's exit is valuable because it communicates information to another stakeholder who then takes corrective action. (25) Actions such as extinguishing the liability by payment or granting new collateral might be observed by other creditors, shareholders, or employees, who may themselves intervene to address the underlying problem with the firm. Recent work provides empirical support for the value of this type of interactive corporate governance. (26)
As in the case of shareholder control, however, lender control produces negative externalities in addition to the positive ones described above. For example, creditors prefer risk-averse decisions and are more likely to urge management to liquidate assets and abandon growth options. There is also conflict between the enforcing lender and other creditors who are vying to improve their payoffs in the event the firm becomes insolvent. A number of legal doctrines in corporate and bankruptcy law police the more blatant actions that redistribute value to the enforcing creditor from the passive ones. (27)
In sum, the case for the regulatory imposition of significant performance standards, such as a fiduciary duty to maximize firm value, is complicated. The existing technology of debt contract design--particularly debt covenants-provides a sophisticated set of tools, of which contract standards and contingent control and renegotiation rights are but two. Professor Rock's proposal to broaden mandatory standards by regulation faces the objection that debtholders can already provide for these standards contractually, but do not. Instead, debtholders rely significantly on contingent control rights that are conditioned on covenant violations. These provisions are priced in the market and create value by addressing agency problems. Under Professor Rock's regulatory proposal, however, the parties may not be able to avoid costly judicial determinations of whether directors made value-maximizing decisions or whether a distribution to shareholders was made in good faith.
I now turn to the Article by Professors Adler and Kahan, who are unabashedly contractarian; they reject mandatory provisions and instead seek legal reform to broaden the parties' toolkit. Indeed, they advocate turning existing mandatory regulation--such as laws against fraudulent transfers, preferences, stock repurchases, and dividend payments by undercapitalized firms--into defaults and allowing debtors to customize their own provisions. They envisage debtors using this freedom of contract also to either increase or decrease the liability on third parties who participate in the violation of covenants. It is hard to argue, of course, against the contractual freedom to customize in settings with sophisticated and informed parties unless it inflicts external costs not borne by the parties to the contract. However, Professors Adler and Kahan's proposal may well produce such externalities because future creditors, for example, do not have an easy way of assuring themselves that their prospective debtor has not agreed to broad and severe third-party sanctions. The information costs imposed on all third parties dealing with the debtor may not be worth the incremental benefit. Although the existing legislative restrictions may seem to take the form of one-size-fits-all, significant customization is available through the technology of security interests, which the authors undervalue.
II. CONTRACT REMEDIES, THIRD PARTIES, AND PROPERTY RIGHTS
The foregoing discussion of the role of covenants in transferring control illustrates that contract remedies are critical features of contract design and are not limited to deterring breach or compensating a promisee for its loss from breach. (28) A similar approach can be found in other types of commercial contracts, such as franchise agreements, in which franchisees promise to refrain from actions that might harm the value of the franchise trademark. (29) The most important remedy for breach of this promise is that the franchisor may then terminate the franchise. The cost of litigation, the risk of judicial error, and the franchisee's insolvency make the franchisor's right to sue for damages of secondary importance. Given the franchisee's specific investment in the business, the loss of the franchise alone is a sufficient deterrent against breach.
Although armed with their respective contingent control rights, the franchisor and lender nevertheless share two concerns. First, the deterrence created by these provisions dissipates as the value of the borrower or franchisee falls and the entity approaches insolvency. Contract design presents a solution to this problem of gradual decline: trip wires that transfer control on signs of financial distress. Financial ratio covenants, in particular, play key roles in this regard, allowing the lender to intervene or exit at earlier warning signs of distress. These provisions depend on costly monitoring of the borrower or franchisee, but this cost can be reduced by coordinated monitoring among creditors and other stakeholders, which I have discussed at some length elsewhere. (30) The second, more challenging problem is that of sudden expropriation: an action that inflicts damages quickly before the franchisor or lender can either observe--given a feasible level of monitoring--or assert control.
Sudden expropriation might occur if the debtor distributes firm value to shareholders in the form of dividends or share repurchases, even if it is in violation of a restrictive covenant, thereby leaving the firm in or on the verge of insolvency. Although notice is often given in advance of such actions, the lender's control rights are ineffective unless the lender can accelerate and collect before the distribution. In contrast, firms typically do not give notice of borrowing from individual entities, so creditors would have even less timely information before their claims are diluted by the new debt. However, as long as the lender reacts before the borrowed value is dissipated, a covenant prohibiting new debt may be effective.
Professor Rock and Professors Adler and Kahan use the leveraged buyout (LBO) as an example of shareholder opportunism and describe the protection offered by fraudulent transfer laws. (31) However, because an LBO typically entails considerable lead time and notice, debt covenants that specify distinctive features of LBOs would trigger a timely shift of control to the lender. Bankruptcy courts apply fraudulent transfer laws--under which a fraudulent transfer is a transfer for less than reasonably equivalent value while the debtor is insolvent--to undo LBOs that render firms insolvent. (32) Indeed, as Professors Adler and Kahan suggest, one might imagine a set of possible variations in fraudulent transfer standards (such as the requirements of insolvency and the absence of reasonably equivalent value) that could be tailored to each debtor's circumstances. (33) The parties might also tailor the remedies to allow them to enjoin a transaction that violates the contractual standard. It is noteworthy that, in Professor Rock's LBO example, the bondholders lacked any such covenant, and the puzzle is therefore why it was omitted. Part III addresses this question.
Where there is sudden expropriation, the lender does not have the opportunity to act before value is taken and the debtor becomes insolvent. A common legal technique for dealing with similar situations in which a defendant has disappeared or is judgment-proof is to place the loss on an available party who was in the best position to prevent the loss. Consider, for example, a rogue who sells a good in which he has no title to an innocent party. If the rogue were not judgment-proof, the buyer could collect damages for breach of the warranty of title, and the goods could be returned to the owner. If the rogue were judgment-proof, however, the loss must be borne by one of the innocent parties: in this case, either the owner or the buyer. The law generally requires the buyer to return the good in order to give each buyer the incentive to check the seller's title, particularly for big-ticket items. One exception is when the buyer purchases a good from a merchant who sells that good in the ordinary course of business. To encourage owners to be careful as to whom they deliver possession and to facilitate merchant sales, the law gives title of the good to the buyer, rather than to the owner who entrusted possession of the good to a merchant of such goods. (34) For similar policy reasons, federal environmental and securities laws impose third-party liability for some harms on persons other than the corporate defendant who were in a position to prevent them.
LBOs that render firms insolvent transfer wealth from creditors to shareholders, and those payments to shareholders are difficult to recover. In cases where managers execute an LBO on behalf of rogue shareholders, Professor Rock and Professors Adler and Kahan suggest transferring the loss to third parties (unless they have acted in good faith) in order to discourage them from participating in an LBO by an undercapitalized entity. Professor Rock suggests that directors and officers who knowingly approve such a transaction should be personally liable if their firm violates fraudulent transfer laws. (35)
In their discussion, Professors Adler and Kahan see merit in imposing liability on the lenders who enable constructively fraudulent LBOs or on the shareholders who benefit from them. They propose that the allocation of loss be determined contractually at the time of the earlier debt financing. (36) Indeed, the authors envisage that the debtor would have the ability to impose liability on a variety of third parties who would control or benefit from a range of debtor misbehavior. Such third parties include future creditors, corporate affiliates, shareholders, or directors. The misbehaviors are those actions identified as the product of the shareholder-creditor conflict, particularly the distribution of firm value to shareholders or excessive future borrowing. (37)
As Professors Adler and Kahan recognize, the obstacle facing their suggestion is that contractual obligations are binding only on parties to the contract. The contract between the lender and debtor might include other parties--such as the current directors and officers, other creditors, or shareholders--as sureties or guarantors. However, this raises practical difficulties when these parties are numerous and dispersed. Moreover, directors, shareholders, and creditors change over time, and the new ones will not be bound by the previous agreement. In contrast, individual consent is not required to effect statutory or judicial provisions, like those governing fraudulent transfers, avoidable preferences, equitable subordination, lender liability, and veil piercing. The advantage of regulation is that it binds future parties that fit the class of defendants defined in the statute or common law. Given that Professors Adler and Kahan prefer contracting over regulation and wish to bind third parties, they propose changing the contract law governing debt contracts to allow such contracts to bind future third parties. (38) They contemplate a range of possible remedies, including depriving the new creditor of its security interest, subordinating its claim against the debtor's assets, or imposing liability for the payment of compensatory (but not punitive) damages.
As the authors note, their proposal to allow for remedies for breach against third parties lies at the interface between contract and property law. Some scholars distinguish between contract and property by defining contractual rights as those enforceable only against other contracting parties and property rights as enforceable against all the world. (39) From this perspective, Professors Adler and Kahan wish to import a distinctive feature of property into contract, and such blurring of the line between the two types of rights has some precedent. As described earlier, various areas of law (like avoidable fraudulent transfers) bind third parties in order to incentivize them to take precautions or to exercise influence over circumstances in which the primary actor might become insolvent or otherwise judgment-proof. Even within contract law, injunctive remedies in contract law play a similar role in effectively binding third parties who have not consented to the contract. (40) Consider, for example, a covenant not to compete in an employment contract. In many instances, the covenant is enforceable by injunction, which prevents another employer from hiring the worker. This remedy against a third party is valuable because of the risk that the employee is judgment-proof and unable to compensate her employer for breach of the covenant. The threat of an injunction gives the former employer the leverage to extract compensation from the new employer in exchange for a release from the covenant. In this sense, the future employer is bound by the initial contract to which it is not a party because the covenant restricts the future employer's ability to hire the worker, irrespective of the worker's solvency.
A core legal principle requires that property interests be observable by third parties at relatively low cost in order for them to be enforceable. Commerce would be impeded if the contours of a seller's title were opaque to buyers. The notice requirement is in large part attributable to enforcement costs and the risk of insolvency. After all, if buyers could be fully compensated by their sellers for damages due to breach of warranty of title, they would be more willing to transact despite incomplete information. As Professors Merrill and Smith observe, the concern with information costs has led to legal restrictions on the ability of owners to create new forms of property and, at least in some contexts, the law appropriately limits property to a finite number of possible interests. (41) For example, an owner cannot create and transfer a time-share property interest in a watch that is limited to Mondays, Wednesdays, and Fridays. (42) The property interest must be enforceable either on all days or on none.
The principle of a closed number of property forms limits the contracting parties' flexibility to maximize the surplus of the transaction by tailoring the property interest. However, as Professors Merrill and Smith argue, the restriction yields countervailing information cost savings. (43) For example, each time a prospective buyer looks to buy a watch, she does not need to investigate whether the owner's title is limited to only some of the days of the week. Nor will the buyer be called upon in court to rebut the existence of a prior partial interest (for example, ownership on Tuesdays, Thursdays, and Saturdays). The social cost of such activity depends, of course, on how numerous and definite the affected group of parties is. (44) As a normative matter, the optimal number of interests is determined by weighing the information cost savings against the cost of foregone flexibility. Yet, as Professors Smith and Merrill point out, a lot of flexibility can be attained by combining the available property forms as building blocks. (45)
A similar concern would arise if the tailored and varied fraudulent transfer provision and other covenants were enforceable against third parties. Professors Adler and Kahan are sensitive to the concern about information costs and propose ways to lower investors'--particularly creditors'--cost of investigating firms. For example, they would restrict the enforceability of covenants, such as a negative pledge clause, to relatively sophisticated lenders who later make material investments and who would likely screen the debtor for other risk factors such as cash flow, credit record, and asset ownership. (46) The authors also contemplate that third parties would be liable only if they had actual or constructive notice of the contractual provision and the remedy. They propose a central depository for contracts filed against the legal name of the debtor that might operate alongside the current registry for personal property security interests. (47) These suggestions for reducing information costs make the extension of quasi-property rights more feasible, especially if the records themselves can be standardized and digitized to dramatically lower search costs. The authors do not indicate that they would require registration, presumably to reduce the cost burden on the debtor. However, allowing for other forms of constructive notice would increase uncertainty and information costs on creditors dealing with all debtors. It could also lead to other complications, including undermining the elegance and simplicity of the priority system of Article 9 of the UCC. For example, consider a negative pledge clause that explicitly subordinates future creditors who violate by lending on a secured basis. Suppose the clause is not filed in a public registry so as to constitute constructive notice to all third parties. Two secured creditors, C1 and C2, subsequently deal with the debtor. C1 has either actual or constructive notice of the provision and files to perfect its security interest in order to enjoy priority over existing and future creditors. C1 lends on a secured basis without knowledge and files. The debtor (or its bankruptcy trustee) can subordinate C1 but not C2. However, because it files first, C1 has priority over C2.
Professors Adler and Kahan argue that debtors will internalize the social costs of incorporating third-party remedies because creditors, shareholders, and directors will charge more for their capital and services to compensate for the costs of search and uncertainty. Debtors would therefore only adopt third-party liability for breach of covenants if the benefits would outweigh the information costs incurred by those parties. However, the core justification for the closed number of property interests, advanced by Professors Merrill and Smith, stems from the external costs imposed on debtors who choose not to incorporate such provisions (or who impose mild remedies) and nevertheless bear the information costs incurred by third parties who deal with them.
It is helpful, however, to look at the other side of the ledger, to the expected incremental benefits from the greater flexibility that debtors would have to impose sanctions on third parties, including liability for compensatory damages. The existing contract doctrine already enforces some promises--such as noncompete covenants--by injunction without requiring prior notice to affected third parties. Tortious interference with contract also might be invoked to impose liability on a future creditor who intentionally lends in violation of the covenant. (48) Professors Adler and Kahan's proposal would give debtors the flexibility either to relax or to opt out of this rule. Debtors could impose liability on parties who had constructive knowledge of the breached covenant but who did not intend to induce the debtor to breach, thereby reducing the burden of proof on the debtor or its trustee. Or the debtor could take the opposite tack by contractually limiting the sanction either completely or to the avoidance of a security interest or subordination of a claim. The mere possibility of third-party liability raises costs for all lenders who would be inclined to search or bear the risk, and therefore also all debtors.
Professors Adler and Kahan's proposal should be assessed on the basis of the incremental improvement it would yield over existing mechanisms in both contract and property law. Security interests are very significant in this respect. A security interest gives the secured creditor priority in the collateral as well as rights to repossess the collateral on default without invoking the judicial process. There is a sophisticated regime in place to give notice of security interests. (49) Security interests have many possible roles in corporate finance and have been shown to restrict overinvestment, cure underinvestment, and constrain distributions to shareholders and LBOs. (50)
Professors Adler and Kahan, however, say that their goal is "to broaden the scope of creditor protection that secured credit now provides in a needlessly narrow fashion." (51) They may understate the potency of security interests. As I have discussed at length elsewhere, security interests restrict the ability of debtors to engage in the various types of misbehavior that fall under the agency costs of debt. (52) In order to pay out firm value to shareholders, for example, debtors must have liquid assets. Security interests constrain the ability to convert nonliquid assets to liquid assets through sales or leases. As another example, the debtor will have difficulty borrowing large amounts if its assets have been pledged as collateral, especially to finance an LBO. (53)
These restrictions are somewhat rigid in that they allow the secured party to block even value-increasing activity. It is more difficult for a debtor to finance a profitable opportunity if it lacks liquidity and substantially all its assets are covered by security interests. However, there are other valuable exceptions that ameliorate this rigidity, such as setting up a separate entity, borrowing with purchase money security interests, or debtor-in-possession financing in bankruptcy. (54)
The wisdom behind security interests is that, despite their simplicity, they offer flexibility along two dimensions: the collateral assets covered (current and after-acquired) and the obligations secured (current and future indebtedness). Professors Adler and Kahan's concern is that the security interest operates only to alter priority of claims while their proposal would permit more significant sanction in the form of damages liability. (55) It is not clear, however, why such liability would be necessary to deter a future creditor, for instance, from financing an undesirable LBO. After all, the LBO lender would lose some or all of the capital it contributes to the insolvent entity because it would be subordinated to the earlier secured interest.
In sum, Professors Adler and Kahan argue persuasively that creditors would value enforcement rights against subsequent parties who enable violations of covenants. Before legal rules are altered to allow such rights, however, it is worth considering whether the existing technology of contract rights and property interests, particularly security interests, creates sufficient flexibility and potency in addressing agency conflicts. The incremental improvements from additional quasiproperty rights may not be worth the added social costs of information imposed even on the limited group of debtholders they prescribe.
III. COVENANT-LITE DEBT
A distinct puzzle that Professor Rock alludes to in passing is that debt contracts are sometimes designed inside the frontier of available contracting technology. In particular, he notes the ebb and flow of "covenant-lite" bond indentures over the past several decades. Indeed, in the two principal case studies of creditor exploitation that he presents, the bondholders could have protected themselves by incorporating in the indenture fairly common covenants that restrict change-of-control and corporate-restructuring transactions. (56) To justify the broader mandatory legal standards that Professor Rock proposes on the basis of these gaps, we first need to investigate the explanation for these omissions.
Empirical observations of variations in covenant patterns raise two sets of puzzles. First, bond indentures tend to have less intense and less extensive covenants than the contracts of institutional lenders, particularly banks, (57) Some commentators suggest that this is because it is easier for the borrower to renegotiate violated covenants with a bank than with dispersed bondholders. Therefore, covenant default in bond indentures might lead to the collapse of the firm. However, indenture trustees are lax in enforcing nonpayment covenants, so renegotiation is not essential. A better explanation stems from the observation that banks are better positioned to monitor and exert positive influence over firm governance. They are given broad powers to exercise this influence by broad covenants. Indeed, in many cases, banks have the right to call loans at any time either because the contract provides that payment is due on demand or because the covenants are so tight as to be violated even at the time of the agreement. Banks use this threat to exert control over the management of the borrower. As noted earlier, the governance activity of a bank generally benefits bondholders, and a number of protections in debtor-creditor law prevent banks from using their control to improve their position relative to other creditors, (58) As a result, the incremental value to bondholders of stronger covenant protection in their indentures is limited, and the bondholders are better off relying on the bank's governance activity. (59)
Second, covenants in bond indentures appear to ebb and flow with cycles in capital markets. In a borrower's market flush with investors, for instance, covenant-lite instruments and unsecured debt are more likely to appear. Industry observers attribute this phenomenon to the fact that borrowers have greater bargaining power in such markets. In theory, however, changes in market conditions--as in bargaining power--seem more likely at first blush to affect the price of capital and less likely to affect nonprice terms such as covenants or collateral. Put simply, if a protective covenant or security interest is efficient in reducing agency costs when capital is scarce, it ought to be efficient when capital is abundant. The parties should maximize the size of their contracting surplus by minimizing agency costs and then use the interest rate to divide the resulting surplus.
In recent work, Professor Albert Choi and I have explored the theoretical explanation for why the ebb and flow of covenants and security interests might correlate with supply and demand. (60) One set of explanations relates to the effect of price changes on the severity of moral hazard and adverse selection. All else equal, an exogenous increase in the interest rate due, for example, to shrinking supply of capital, tends to lower the quality of the pool of prospective borrowers because the increase in the expected interest payments is lower to high-risk borrowers that are more likely to fail. An increase in the interest rate also increases the incentive for opportunistic behavior because the shareholders have less of an interest in the value of the firm (since a greater share will go to the lender). Therefore, the optimal covenant may be tighter and more expansive in times of limited capital supply. The risk of opportunism, in contrast, is lower when capital is cheap; at the margin, it is therefore more efficient and rational for the parties to loosen covenants and security interests ex ante to avoid the risks of over-inclusive covenants and the attendant costs of litigation. Imposing a stricter standard by mandatory provision of law would undermine this efficiency calculus and increase the cost of capital.
The foregoing hypotheses as to why (a) covenants are lighter in bond indentures than bank loans and (b) covenant-lite indentures and unsecured loans are more likely to appear under conditions of apparently excess supply of capital are far from dispositive. I describe them here selectively to illustrate the possibility that the weaknesses of covenant protection may be efficient, rather than as evidence of contracting failure. This may give further pause before adjusting legal rules to address the apparent gap in covenant protection.
We have not reached a steady state in the technology of contract design, and we ought to encourage steps to promote valuable contract innovation. This objective raises two fundamental questions: First, are there legal obstacles to innovation? Second, do contracting parties (and their lawyers) have the appropriate incentives to innovate and to incorporate innovations in their contracts? These essential questions are receiving increased attention in legal scholarship, and the Articles in this issue by Professor Rock and Professors Adler and Kahan contribute to the research agenda addressing these questions in the context of debt financing. Their respective proposals for law reform are provocative and worthy of further investigation. However, for reasons outlined in this Response, I have doubts in each case that the incremental value of the prescribed change outweighs the respective costs.
(1) Edward B. Rock, Adapting to the New Shareholder-Centric Reality, 161 U. PA. L. REV. 1907 (2013).
(2) Barry E. Adler & Marcel Kahan, The Technology of Creditor Protection, 161 U. PA. L. REV. 1773 (2013).
(3) Rock, supra note l, at 191:2-13.
(4) Id. at 1910.
(6) Id. at 1919.
(7) See Michael Bradley & Michael R. Roberts, The Structure and Pricing of Corporate Debt Covenants 30 (May 13, 2004) (unpublished manuscript), available at http://papers.ssrn.com/ id=466240 (finding a negative correlation between the inclusion of a covenant and the resulting loan yield). On the equity side, Lucian Bebchuk, Alma Cohen, and Charles Wang provide evidence suggesting that equity markets have learned to internalize the significant factors of quality governance. Lucian A. Bebchuk, Alma Cohen & Charles C.Y. Wang, Learning and the Disappearing Association Between Governance and Returns, 108 J. FIN. ECON. 323, 238-45 (2013).
(8) See, e.g., Matthew T. Billett et al., Grewth Opportunities and the Choice of Leverage, Debt Maturity, and Covenants, 62 J. FIN. 697, 699 (2007) (finding that covenants mitigate the agency costs of debt for high-growth firms); Clifford W. Smith, Jr. & Jerold B. Warner, On Financial Contracting: An Analysis of Bond Covenants, 7 J. FIN. ECON. 117 (1979) (describing how covenants lower the various types of agency costs of debt).
(9) The Articles discuss, in particular, the gaps left behind by (a) the current legislative and common law rules governing fraudulent transfer laws, and (b) the preferences and restrictions on corporate dividends and share repurchases. Adler & Kahan, supra note 2; Rock, supra note 1.
(10) The parties have further flexibility to tailor their governance provisions to asset- or business-types within the borrower's enterprise by splitting assets between affiliated legal entities. For example, a conglomerate with a manufacturing and a hotel division might choose instead to put each division in a separate subsidiary in order to tailor the financing and governance structure to each line of business. This is an important motivation for the parent-subsidiary structure in corporate groups. See Edward M. Iacobucci & George G. Triantis, Economic and Legal Boundaries of Firms, 93 VA. L. REV. 515, 560-65 (2007) (explaining the complex tradeoff between economic integration and capital-structure tailoring considerations in the decision to either integrate assets within a single firm or establish parent-subsidiary relationships between distinct legal firms).
(11) See Albert Choi & George Triantis, Completing Contracts in the Shadow of Costly Verification, 37 J. LEGAL STUD. 503, 526 (2008) (showing that contracts can be designed to leave the costly verification of standards off the equilibrium path so that verification costs are not incurred); Albert Choi & George Triantis, Strategic Vagueness in Contract Design: The Case of Corporate Acquisitions, 119 YALE L.J. 848, 881-96 (2010) (describing the efficiency of contract vagueness).
(12) See, e.g., Stephen M. Bainbridge, The Business Judgment Rule as Abstention Doctrine, 57 VAND. L. REV. 83, 117-24 (2004) (arguing that judges' lack of business expertise would lead to judicial error if the business judgment rule were relaxed).
(13) See Adler & Kahan, supra note 2, at 1787 ("Delaware judges and lawmakers.., realize that the prospect of personal liability--whether to shareholders or creditors--would make directors excessively cautious....").
(14) Rock, supra note 1, at 1977-86.
(15) See id. at 1955 ("[D]irectors' understanding of their role should return from the contemporary exhortation to maximize equity value to the traditional goal of maximizing firm value."). A related important question is whether this reform would change the deference that courts give to directors under the business judgment rule. In a similar vein, the sanction imposed on directors of corporations for improper share repurchases or dividends--where corporations are insolvent in either the balance sheet or equity sense--would be limited by the full defense available for directors who rely in good faith and reasonably on corporate records, officers, employees, or experts. DEL. CODE ANN. tit. 8, [section] 172 (2011).
(16) See Adler & Kahan, supra note 2, at 1787 ("[L]oan agreements give no indication that companies and creditors would want to impose liability on directors."). Contract scholars have speculated that novel contract terms may be discouraged by the risk of adverse judicial interpretation and by resistance from capital markets. E.g., Michael Klausner, Corporations, Corporate Law, and Networks of Contracts, 81 VA. L. REV. 757, 774 (1995).
(17) Alan Schwartz and Robert Scott raise a similar doubt as to the value of state-imposed standards in commercial law. Alan Schwartz & Robert E. Scott, Contract Theory and the Limits of Contract Law, 113 YALE L.J. 541, 594 (2003).
(18) See Philippe Aghion & Patrick Bolton, An Incomplete Contracts Approach to Financial Contracting, 59 REV. ECON. STUD. 473, 486-90 (1992); Mathias Dewatripont & Jean Tirole, A Theory of Debt and Equity: Diversity of Securities and Manager-Shareholder Congruence, 109 Q.J. ECON. 1027, 1049-50 (1994).
(19) See, e.g., Michael R. Roberts & Amir Sufi, Control Rights and Capital Structure: An Empirical Investigation, 64 J. FIN. 1657, 1660 (2009) (finding in their sample that only four percent of covenant violations led to a termination of the relationship within two quarters after the violation, but the breaching borrowers were subject to disciplinary actions by their creditors).
(20) See George G. Triantis, Debt Financing, Corporate Decision Making, and Security Design, 26 CAN. BUS. L.J. 93, 94 (1996) ("Debt covenants therefore constrain the firm's decision space and, if they are violated, may trigger the more active intervention of the lender in the firm's decisions."); George G. Triantis & Ronald J. Daniels, The Role of Debt in Interactive Corporate Governance, 83 CALIF. L. REV. 1073, 1093-94 (1995) (discussing the role debt covenants play in triggering investor activism).
(21) See Ilia D. Dichev & Douglas J. Skinner, Large-Sample Evidence on the Debt Covenant Hypothesis, 40 J. ACCT. RES. 1091, 1093 (finding that covenant violations occur frequently and often in the absence of financial distress); Nicolae Garleanu & Jeffrey Zwiebel, Design and Renegotiation of Debt Covenants, 22 REV. FIN. STUD. 749, 749 (2009) (asserting that tight covenants protect lenders from information asymmetry that favors the borrower); Michael R. Roberts, The Role of Dynamic Renegotiation and Asymmetric Information in Financial Contracting 1 (Dec. 29, 2010) (unpublished manuscript), available at http://ssrn.com/id=1732364 (finding that the typical loan in the sample is renegotiated four times). However, Roberts's other interesting finding is that renegotiation in his sample is "initiated by borrowers primarily in response to changing conditions, as opposed to lender interventions due to default". Id.
(22) Michael R. Roberts & Amir Sufi, Renegotiation of Financial Contracts: Evidence from Private Credit Agreements, 93 J. FIN. ECON. 159, 166 (2009) (noting that tight covenants allocate bargaining power between the borrower and the lender in future renegotiations).
(23) See Triantis & Daniels, supra note 20, at 1094-95 (noting that one of the effects of avoidable preference provisions in bankruptcy law is to incentivize lenders to intervene upon early signs of distress instead of waiting for insolvency, at which point additional firm value will have been lost); see also Raghuram Rajan & Andrew Winton, Covenants and Collateral as Incentives to Monitor, 50 J. FIN. 1113, 1114 (1995) (finding that long-term debt with covenants increases the bank's incentive to monitor by decreasing the bank's payoff it if does not monitor); Joshua D. Rauh & Amir Sufi, Capital Structure and Debt Structure, 23 REV. FIN. STUD. 4242, 4243-44 (2010) (concluding that low credit-quality firms are more likely than higher credit-quality firms to spread their debt issues across multiple tiers that include secured bank debt with tight covenants and subordinated debt with loose covenants).
(24) For example, Professors Greg Nini, David Smith, and Amir Sufi show that covenant violations are followed immediately by an increase in CEO turnover, an increase in the hiring of turnaround specialists, a decline in acquisitions and expenditures, and a sharp reduction in leverage and shareholder payouts. See Greg Nini et al., Creditor Control Rights and Firm Investment Policy, 92 J. FIN. ECON. 400, 401 (2009) ("Capital expenditure restrictions are 20% more likely to be observed in a renegotiated agreement following a covenant violation."). They also show that firm operation and stock price performance improve following a violation. Id. at 415-16; see also Sudheer Chava & Michael R. Roberts, How Does Financing Impact Investment? The Role of Debt Covenants, 63 J. FIN. 2085, 2106 (2008) ("[C]apital expenditures decline significantly in response to covenant violations. We observe a quarterly decline in investment of approximately 1% of capital, a 13% decline relative to the level of investment outside of violation states."); Greg Nini et al., Creditor Control Rights, Corporate Governance, and Firm Value, 25 REV. FIN. STUD. 1713, 1748, 1752 (2012) (providing data showing an increase in operating performance and stock price performance following covenant violations).
(25) See Triantis & Daniels, supra note 20, at 1108-12 ("[T]he sale of stock by a shareholder, the termination of deliveries by a supplier or of orders by a customer, and the acceleration of debt obligations by a lender often send useful signals to other stakeholders through a variety of channels.").
(26) See Liangliang Jiang & Hui Zhou, Do Auditors Play a Positive Role in the Resolution of Debt Covenant Violations? 13-14 (Oct. 15, 2012) (unpublished manuscript), available at http://papers. ssrn.com/id=2166174 (finding that covenant violations lead to closer monitoring by external auditors).
(27) See, e.g., Triantis & Daniels, supra note 20, at 1094 (discussing the voidable preference rule under bankruptcy law, which "encourages timely monitoring and pre-insolvency action by threatening to reverse any attempt [by an enforcing creditor] to exit after the debtor has become insolvent").
(28) George Triantis, The Evolution of Contract Remedies (And Why Do Contracts Professors Teach Remedies First?), 60 U. TORONTO L.J. 643, 653 (2010) (asserting that contractual remedies are expected to achieve efficiency across a broad range of incentive and risk-bearing objectives, of which efficient breach is of relatively little consequence); George Triantis, Promissory Autonomy, Imperfect Courts, and the Immorality of the Expectation Damages Default, 45 SUFFOLK U. L. REV. 827, 833 (2012) (describing the significance of contingent remedies, including termination).
(29) See, e.g., Sample Business Contracts: Form of Krispy Kreme Doughnut Corporation Franchise Agreement, ONECLE, http://contracts.onecle.com/krispy-kreme/franchise.shtml (last visited May 6, 2013).
(30) See generally Triantis & Daniels, supra note 20.
(31) Adler & Kahan, supra note 2, at 1792-93; Rock, supra note 1, at 1939-44.
(32) 11 U.S.C. [section] 548(a)(1)(B) (2006).
(33) See Adler & Kahan, supra note 2, at 1794-1804 (suggesting that creditors could create contract rights with results mimicking those of bankruptcy law).
(34) U.C.C. [section] 2-403 (2012).
(35) See Rock, supra note l, at 1946 ("Given the complexities of clawing back payments made to dispersed shareholders [as fraudulent transfers], corporate law's strategy of providing directors with incentives not to make such payments in the first place makes sense.").
(36) Under Professors Adler and Kahan's framework, placing liability on future creditors--whether under fraudulent transfer rules or under debt contracts--would induce them to be more cautious in learning about the transaction and to make sure, for example, that the debtor is not left undercapitalized. Adler & Kahan, supra note 2, at 1800-03. In contrast, Professor Rock says that "LBO lenders ... are neither the initiating parties, nor the actors with fiduciary duties to the corporation or the actors with direct access to the relevant information.... The real justification for imposing obligations on them, backed by the threat of losing priority ... seems to be to recruit them to force the LBO sponsors and the Target firm to adopt a sound financial structure.... This is in tension with the general principle that creditors do not have duties to look out for the interests of other creditors." Rock, supra note 1, at 1945.
(37) Adler & Kahan, supra note 2, at 1798.
(38) Adler & Kahan, supra note 2, at 1794-95.
(39) See, e.g., Henry E. Smith, Standardization in Property Law, in RESEARCH HANDBOOK ON THE ECONOMICS OF PROPERTY LAW 148, 157 (Kenneth Ayotte & Henry E. Smith, eds., 2011).
(40) Professor Henry Smith discusses another borderline case of servitudes in real, tangible, and intellectual property. Id. at 167-68.
(41) Thomas W. Merrill & Henry E. Smith, Optimal Standardization in the Law of Property: The Numerus Clausus Principle, 110 YALE L.J. 1, 4 (2000); Smith, supra note 39, at 159.
(42) Merrill & Smith, supra note 41, at 27.
(43) "In rem rights are directed at a wide and indefinite audience of dutyholders and other affected parties, who would incur high information costs in dealing with idiosyncratic property rights and would have to process more types of information than in the absence of [a closed number of property forms]." Smith, supra note 39, at 149; see also Merrill & Smith, supra note 41, at 30.
(44) Smith, supra note 39, at 154. But see, e.g., Henry Hansmann & Reinier Kraakman, Property, Contract, and Verification: The Numerus Clausus Problem and the Divisibility of Rights, 31 J. LEGAL STUD. 373, 379-82 (2002) (finding Professor Merrill and Smith's marginal cost analysis "unconvincing").
(45) Smith, supra note 39, at 152-53; Merrill & Smith, supra note 41, at 39-40.
(46) Adler & Kahan, supra note 2, at 1801-03. The materiality threshold, they suggest, would depend on the extensiveness of the remedy, that is, whether it subordinates or imposes damages liability on the third party. Id.
(47) Id. at 1801
(48) RESTATEMENT (SECOND) OF TORTS [section] 766A (1977). This tort has been subject to considerable criticism and has not yet been applied in the context of breached covenants. Adler & Kahan, supra note 2, at 1810-1811 & 1809 nn.90-93.
(49) It should be noted that the registration regime for security interests still imposes significant search responsibility on the third party. The registration against the debtor's name simply discloses that the debtor may have granted a security interest in some or all of the identified assets. U.C.C. [section] 9-502 cmt. 2 (2000) ("The notice itself indicates merely that a person may have a security interest in the collateral indicated. Further inquiry from the parties concerned will be necessary to disclose the complete state of affairs."). The notice requirement reflects a balancing of burdens on secured creditors, on the one hand, and creditors and other third parties who contemplate dealing with the debtor, on the other.
(50) See, e.g., Robert E. Scott, d Relational Theory of Secured Financing, 86 COLUM. L. REV. 901, 904 (1986) (describing how security interests ameliorate agency conflicts in debt financing).
(51) Adler & Kahan, supra note 2, at 1798.
(52) See George G. Triantis, Financial Slack Policy and the Laws of Secured Transactions, 29 J. LEGAL STUD. 3.5 (2000); George G. Triantis, A Free-Cash-Flow Theory of Secured Debt and Creditor Priorities, 80 VA. L. REV. 2155 (1994).
(53) Indeed, much of the value in Professors Adler and Kahan's proposal to enforce negative pledge clauses against future creditors can be attained by granting the early creditor a security interest. See Barry E. Adler, Accelerated Resolution of Financial Distress, 76 WASH. U. L.Q. 1169, 1201 (1998) (advocating for the use of contingent security interests for early creditors).
(54) See Triantis, Financial Slack Policy and the Laws of Secured Transactions, supra note 52; Triantis, A Free-Cash-Flow Theory of Secured Debt and Creditor Priorities, supra note 52.
(55) "[O]ur proposal would allow a breach victim to collect directly against a third party associated with a breach, including one who is not a creditor, while a security interest in a debtor's assets operates only to alter the priority of claims against the debtor. The value of priority enhancement for a covenant breach is limited to the value of the debtor's assets.... [B]ecause a reduction in a claim's priority does not eliminate the claim, a potential third-party creditor may have a relatively low incentive to avoid participation in the debtor's covenant breach." Adler & Kahan, supra note 2, at 1813.
(56) In the case of the hypothetical LBO, Professor Rock states that "change-in-control covenants would have protected them, but during some periods of the business cycle, bonds are issued with minimal protection." Rock, supra note a, at 1940.
(57) See Triantis & Daniels, supra note 20.
(58) See supra notes 20-23 and accompanying text.
(59) See Triantis & Daniels, supra note 20, at 1090 (explaining that the combination of extensive bank covenants and light bond covenants solve the dangers that either both sets of creditors will freeride or duplicate monitoring activity, as can security interests); see also Rajah & Winton, supra note 23; Rauh & Sufi, supra note 23.
(60) See Albert Choi & George Triantis, The Effect of Bargaining Power on Contract Design, 98 VA. L. REV. 1665 (2012); Albert Choi & George Triantis, Market Conditions and Contract Design: Variations in Debt Contracting, 88 N.Y.U.L. REV. 51 (2013).
GEORGE TRIANTIS, James and Patricia Kowal Professor of Law, Stanford Law School. I thank Mike Klausner for helpful comments on an earlier draft and am much indebted to Trevor Lynch, Stanford Law School Class of 2014, for his valuable assistance in preparing this Response.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||response to articles in this issue, p. 1773 and 1907|
|Author:||Triantis, George G.|
|Publication:||University of Pennsylvania Law Review|
|Date:||Jun 1, 2013|
|Previous Article:||Poor pitiful or potently powerful preferred?|
|Next Article:||Protecting search terms as opinion work product: applying the work product doctrine to electronic discovery.|