[*PG195]SUBORDINATED DEBT: A CAPITAL MARKETS APPROACH TO BANK REGULATION
Abstract: Banking organizations in the United States are growing larger, more complex and more diversified in their operations. As a result, bank regulators are becoming less able to understand and supervise their regulatory charges. The recently enacted Gramm-Leach-Bliley Act contributed to this trend by expanding the activities in which banks and their affiliates may engage. The authors argue that increasing the amount of market discipline to which banks are subject promises to remedy many of the shortcomings of government supervision and regulation. This Article proposes that large banks should be required to issue a minimum amount of long-term subordinated debt to third-party investors and sets forth a comprehensive subordinated debt program as a complement to government regulation of banks. Actual and prospective holders of bank subordinated debt will constrain bank risk taking roughly in accordance with the interests of the federal government and without the bureaucratic and other inefficiencies entailed in government regulation. Holders of bank subordinated debt, as they buy and sell bank debt securities in the secondary market and negotiate purchases in the primary market, will also signal to federal regulators the private sector’s view as to the value of a bank’s enterprise.
On November 12, 1999, President Clinton signed into law the Gramm-Leach-Bliley Act, the most significant U.S. banking legislation in over sixty-five years.1 The Gramm-Leach-Bliley Act eliminates many long-standing federal and state law barriers to affiliations between banks and securities firms, insurance companies, mutual funds, and [*PG196]other financial service providers. In so doing, the Act paves the way for a significant restructuring of the U.S. financial services industry and a modernization of the way financial services are offered in the United States. Tucked comfortably in the middle of Title I of the Gramm-Leach-Bliley Act is a provision that instructs the Federal Reserve Board and the Treasury Department to study the feasibility of requiring large banks and their holding companies to maintain a portion of their capital in the form of subordinated debt.2 This requirement is a way to bring market forces to bear on the operation of banking institutions and to reduce the risks that these institutions and their expanded activities may pose to the federal deposit insurance fund.3
The Gramm-Leach-Bliley Act’s commission of a federal study of subordinated debt is the latest in a long series of calls from academics, regulators, and public officials for investigation into the potential benefits that market discipline—that is, using the private sector to monitor and regulate bank risk taking—may provide to the U.S. system of bank regulation.4 Since the 1980s, the academic literature has contained various proposals for using different forms of market discipline to supervise and regulate banking activities, and academic writers have suggested alternatively that depositors, nondeposit creditors and shareholders are the market participants best situated to monitor and regulate bank behavior.5 Several commentators have argued that [*PG197]subordinated debtholders are the preferred source of market discipline.6
The use of market discipline, however, has not been without critics. Indeed, some commentators have contended that proponents of market discipline have not presented workable models that can achieve the goal of monitoring and curbing bank risk-taking activities.7 Other commentators have argued that, among potential market disciplinarians, subordinated debtholders are ill-suited to serve as monitors of bank behavior.8
In light of the Gramm-Leach-Bliley Act’s demand for a study of subordinated debt and in anticipation of the new financial world to be ushered in by the Act, this Article takes a fresh look at the market discipline debate. In this new era, in which banks and bank holding companies will be permitted to engage in a wide array of financial activities, the regulatory system will need to utilize market forces to help monitor the activities of banking firms. In short, banking organizations are becoming more complicated, and the government will need help in supervising these increasingly complex and geographically diverse institutions.
This Article concludes that a properly structured subordinated debt program can serve as an important source for constraining bank behavior and can complement federal and state supervision and examination efforts.9 This Article formulates a properly structured sub[*PG198]ordinated debt program, which program we believe to be the most detailed and workable proposal to-date for the issuance of subordinated debt by banks. In reaching this conclusion, this Article also evaluates the arguments made by both opponents of market discipline and proponents of forms of market discipline other than subordinated debt.
The first Part of this Article examines a predicate question: What is it about banks that causes the government to regulate and supervise them and makes us carefully consider alternative methods of supervision and regulation? Part II looks at the current approaches to banking regulation and supervision and highlights the significant limitations of these approaches. Part III examines the ways in which market discipline may complement existing bank examination and supervision, and it assesses alternative sources of market discipline. Part IV contains our detailed proposal for using subordinated debt as the source of market discipline. Part V lays out some of the benefits of using subordinated debt, especially in light of the financial reforms brought about by the Gramm-Leach-Bliley Act. Finally, Part VI responds to the concerns and questions raised by commentators about market discipline in general and subordinated debt in particular.
Despite two decades of deregulation, banks remain subject to a uniquely complex and intrusive regulatory scheme, one that makes banking one of the most comprehensively regulated industries in the American economy.10 For example, federal law restricts the ownership of banking organizations,11 places limits on the activities and invest[*PG199]ments of banking organizations12 and requires banking organizations to maintain certain prescribed levels of capital.13 Before examining the potential benefits of increasing the amount of market discipline to which banks are subject, it is necessary to answer a predicate question: Why do we regulate and supervise banks? Or, to be more precise, why do we employ a regulatory strategy in banking that is so different from the approach used in supervising other firms and even other financial intermediaries?14
Ultimately, the answer is that we subject banks to unique levels of regulatory scrutiny and supervision in order to prevent—or, more accurately, to minimize the potential adverse effects of—bank failures.15 Firm failure is, of course, in no respect unique to banking. In fact, in the context of other industries and the economy as a whole, [*PG200]the failure of firms generally is regarded as healthy and beneficial. Failure serves the useful cleansing function of eliminating mismanaged, noncompetitive and obsolete institutions, and diverting their resources to enterprises that can more efficiently utilize them.16 Bank failures, however, have been and currently are viewed differently because: (1) banks are uniquely susceptible to failure, and even banks that are well managed face unpredictable risks of sudden failure; (2) individual bank failures may pose significant systemic risks to other banks, other segments of the economy, and to the economy as a whole; and (3) under the current system of federal government guarantees, which were designed to mitigate the individual and systemic risks of bank failure, the direct cost of bank failures may fall not merely on the institutions that fail but also on all federal taxpayers.
Banks have capital structures that make them uniquely susceptible to failure. Banks are distinctive in that they tend to have relatively little equity when compared to other firms—banks tend to receive ninety percent of their capital funding from debt.17 Moreover, a significant portion of that debt is in the form of transaction accounts, which are payable on demand at par and are readily transferable by the accountholder to third parties.18 A high percentage of bank assets, however, are in the form of relatively illiquid commercial loans or [*PG201]mortgages.19 No other financial intermediary combines such a potentially dangerous mix of highly liquid liabilities and illiquid assets.20
The asymmetry of the maturity structure of bank assets and liabilities makes banks especially vulnerable to liquidity crises. If a substantial portion of a bank’s depositors wish to withdraw their money at the same time, commonly referred to as a bank “run,” the bank would face an immediate and severe liquidity crisis that could be fatal to the institution. The bank might be forced to liquidate assets, even profitable ones, quickly and at discounted prices. If that liquidation effort and other efforts to raise cash (such as interbank borrowing) proved unsuccessful in stopping the run, the bank would have to close.
Bank runs represent a classic prisoner’s dilemma. Depositors, as a whole, are better off if they do not simultaneously seek to withdraw their money from a bank. Individual accountholders, however, are better off adopting a “me-first” attitude. Being among the first to withdraw funds from the bank increases the chances that the depositor will get his or her money before the bank’s ability to obtain cash is exhausted and the institution must close its doors.21 Accordingly, once they begin, bank runs tend to take on a life of their own.22
A bank run may occur because the public receives news that the institution is facing a significant loss or that bank management has engaged in fraudulent activities. Bank runs also can occur for a variety of reasons that have nothing to do with whether the institution is solvent or well-managed.23 Bank runs may even occur when the public receives misinformation about the financial or managerial condition of a bank.24 In fact, inaccurate information may crowd out the truth; [*PG202]informed depositors, who know that the information is false, cannot ignore the information merely because they know it to be erroneous. Rather, informed depositors must take into account the reaction of ill-informed accountholders in deciding whether to withdraw their money, as the actions of ill-informed accountholders may precipitate the fatal run.25 Accordingly, banks—even well-managed banks—are susceptible to failure due to liquidity runs in ways in which other firms and financial institutions are not.
Bank runs have been viewed as potentially troubling not only because of their capacity to harm otherwise healthy and well-managed banking institutions on an individual basis, but also because individual bank failures pose a systemic risk to the banking system and the broader economy. The classic concern is that a run at one bank may cause depositors at other banks to panic.26 Through contagion, a liquidity problem at one institution may spread to other depository institutions in the same town or region of the country and, in extreme crises, extend throughout the national banking system.27 Nationwide [*PG203]bank runs of this sort occurred periodically in the eighteenth century and the first half of the nineteenth century, prior to the advent of federal government deposit guarantees.28 The problem of contagion, although not confined to banking, is more severe in banking than in other industries.29
Another systemic concern is that the failure of banks could have effects outside the banking industry. When banks face a significant demand for cash that cannot be satisfied through such means as interbank borrowing, banks must recall loans. The recall of good loans disrupts the productive financing of investments and use of assets by bank borrowers.30 In addition, banks facing a run on their deposits will be unwilling to extend additional credit to borrowers or to roll over existing lines of credit.
Bank failures also could potentially have significant macroeconomic effects because of the central position that banks, especially certain very large banks, occupy in our monetary and financial system. Although the significance of banks in the financial system has declined over the past few decades,31 the banking system still plays a key role in the money supply process. The stability of the electronic, large-dollar payment system, through which billions of dollars move daily, and the liquidity of the securities, financial derivatives and interbank funding markets are all dependent on the soundness and [*PG204]proper functioning of the banking system.32 Accordingly, the widespread failure of banks, or the failure of certain very large banks, could cause an unexpected disruption in the money supply, which certain empirical evidence suggests could lead to dislocations in the economy that may worsen, or even cause, an economic downturn.
Most of the fears about bank runs and systemic contagion were alleviated with the advent of the federal “safety net,” the key elements of which were put in place as a result of the banking panics of 1907 and 1929 and the Great Depression. The safety net, which places the full faith and credit of the United States behind the banking system, has three primary elements: (1) federal insurance of bank deposits up to $100,000; (2) access for banks to emergency cash through the Federal Reserve’s discount window; and (3) access for banks to the Federal Reserve’s payment system.33 With deposit insurance, most accountholders no longer face a prisoner’s dilemma in a bank run, as they know that the federal government will back the deposits that they hold in a bank (up to $100,000) regardless of the health of the institution.34 Access to the discount window gives banks facing temporary liquidity crises—due to unexpected withdrawals, operational problems or various factors outside the bank’s control—a means of meeting depositors’ cash demands without liquidating assets.35 The Federal [*PG205]Reserve’s payment system ensures riskless settlement of financial transactions.36
The safety net, which is a strong prophylactic for bank runs, has in itself given rise to another reason—and, in the minds of many commentators, the only valid reason—for bank regulation and supervision.37 The federal government, and thereby the federal taxpayer, is now ultimately responsible for bank deposits and the orderly functioning of the banking system and has a direct financial stake in keeping bank risks under control. The government’s interest is akin to that of a private insurance company, which has a stake in controlling the risks of its policyholders.38
As described in the previous Part, the federal government has created a safety net to reduce the incidence of bank failures and the systemic shocks that may accompany such failures. That safety net, however, creates incentives for excessive risk taking by banking organizations. The safety net also eliminates many of the incentives for private-sector monitoring of banks’ risk-taking activities. The government relies principally on regulation and supervisory oversight as a surrogate for private-sector monitoring of bank activities and disciplining of excessive bank risk taking. This public-sector apparatus, however, is both less efficient and less effective than a system that utilizes market-based disciplinary forces.
The federal safety net, in general, and deposit insurance, in particular, have created a twin set of problems. First, the safety net gives rise to a “moral hazard” problem, one that is common to all types of insurance.39 The safety net creates incentives for banks to take larger [*PG206]risks than they otherwise would because the rewards of any gamble accrue principally to the bank’s management and owners but the risks fall largely on the safety net.40 This same problem of excessive risk taking exists in the private insurance context. There, insurers use contractual devices, including risk-adjusted premiums, deductibles and co-insurance, to limit the risk taking of insured entities. The federal government has tried to implement some of these devices, but with only limited success.41 Instead, the government has relied largely on supervision and regulation.
Second, at the same time as it leads to increased risk taking by banks, deposit insurance leads to reduced monitoring and sensitivity to risk on the part of creditors, shareholders and managers—the three corporate constituencies that, in the normal firm setting, monitor and manage the risk profile of an enterprise. In the usual corporate setting, creditors have a powerful incentive to monitor firm risks. Creditors are the most risk averse of the three corporate constituencies because they are entitled to a fixed rate of return. Accordingly, debtholders oversee the activities of an enterprise to ensure that the firm’s ventures produce a reliable stream of income to permit payment of interest and principal on the debt but do not involve a [*PG207]significant risk of loss that may interfere with the firm’s debt servicing obligations.42
Equityholders and managers (who are often equityholders themselves) tend to be more accepting of risk than debtholders, because equityholders and managers stand to gain—either through capital appreciation, increased dividends or increased compensation—if risky ventures succeed. Equityholders and managers, however, cannot afford to ignore the risk preferences of creditors because doing so could cut the firm off from reasonably priced debt funding. In fact, the willingness of equityholders and managers to enter into agreements with debtholders that limit risk taking, such as restrictive debt covenants, may serve as a signal to other (debt and equity) investors of the firm’s prudent risk posture.43 By contrast, an enterprise with a reputation for excessive risk taking or poor risk management will likely need to pay a premium for debt financing, which will reduce the resources that the firm has available to pay dividends to equityholders and bonuses to management. In more extreme situations, excessive risk-taking activities may make it difficult or impossible for the firm to attract and retain business creditors, and the exit of creditors from a firm may precipitate the firm’s collapse.44
Deposit insurance shelters banks from the disciplinary forces that exist in the ordinary corporate setting. Depositors—the largest group of bank creditors—have little incentive not only to guard against excessive risk taking, but even to police management self-dealing, defalcation or fraud because the funds that most depositors lend to banks are guaranteed by the Federal Deposit Insurance Corporation (“FDIC”)45; the FDIC provides guaranteed coverage at least up to [*PG208]$100,000.46 Large uninsured depositors also have little incentive to monitor banks because the risk of loss to even uninsured depositors has been slight. In the past, the FDIC has paid large depositors an average of 99.5% of their funds regardless of the amount of their deposits.47
The absence of creditor monitoring allows shareholders and management to be less concerned with risk. Shareholders and managers know that depositors, secure in the knowledge that their deposits are guaranteed by the FDIC, will not demand a premium for placing their funds in a risky bank. Moreover, if a bank approaches insolvency, debtholders will not function as a check on managers’ and equityholders’ proclivity to taking excessive gambles.48
Theoretically, at least, the absence of debtholder monitoring need not be problematic. The regulatory scheme places government in the position that debtors typically occupy in the system of corporate governance. Government employs supervisory, enforcement, and other tools to regulate and monitor bank activities.49 Indeed, govern[*PG209]ment regulators and supervisors would appear to have the tools and resources to be better monitors than depositors and other private-sector creditors.50 Government regulators, who specialize in the trade of banking supervision and regulation, could be more adept at risk-monitoring than the average bank creditor and have a panoply of statutory and regulatory remedies to curb excessive risk taking and punish malfeasance that are not available to bank creditors.
Notwithstanding these apparent superiorities, the substitution of the public sector for the private sector creates a less efficient and less effective monitoring system. First, private-sector entities monitor firms at an efficient level; they monitor firms until the marginal costs of such monitoring are equal to its marginal benefits.51 Government regulation and supervision, on the other hand, is not subject to this calculus and is less likely to be as efficient. In some instances there may be excessive supervision and regulation,52 while in other situations there may be inadequate monitoring.53 Second, private-sector monitors have more reliable monitoring incentives. Private-sector monitors have their own money at risk and, therefore, have powerful [*PG210]incentives to act with greater alacrity than public-sector bureaucrats. In addition, private-sector monitors have at their disposal contractual devices designed specifically for the purpose of causing borrowers to internalize the costs of risky activities. Such contractual devices are not available to government regulators, who instead need to rely on less precise and generally applicable regulatory requirements.54
The relative effectiveness of private-sector monitoring may be further enhanced by the singleness of purpose of private oversight. Private-sector investors have a single goal—protecting their investment—and their success or failure is measured solely by their ability to achieve this financial goal. Government regulators, on the other hand, may have political and other goals that temper their monitoring efforts. For example, political capture of a regulator by a regulated institution or industry may compromise the effectiveness of the supervisory and regulatory process.55 Alternatively, competition among the various bank regulators may lead to regulatory laxity, in the classic “race to the bottom” fashion.56 Political compromises also may give rise to statutory or regulatory restrictions that are unrelated to banking safety—such as the long-standing limitations on banks’ insurance and securities underwriting and dealing activities.57 These limitations, many of which were in place for over sixty years despite widespread acknowledgment that they served little or no safety and soundness purposes, needlessly barred banks from engaging in profitable activities and diversifying their risks.58
Finally, even regulators who are not distracted by competing goals and pressures may find effective oversight to be a difficult task, given the broad range of activities in which banks and their parent [*PG211]holding companies engage.59 The size and complexity of banks severely tax the ability of bank regulators to engage in effective monitoring.60 Banks now routinely conduct a wide array of complicated derivative, securitization and other types of transactions to generate fee income and to manage risks. Moreover, the last few decades have seen the advent of truly global financial institutions that integrate a variety of banking and nonbanking businesses into the same multinational enterprise.61
The trend of financial conglomerization and product diversification will, if anything, accelerate under the framework established in the Gramm-Leach-Bliley Act, which permits banks to affiliate with securities firms, mutual funds, insurance companies, futures commission merchants and merchant banking firms. Broad, prophylactic rules that restricted the activities of banks and their affiliates—and which made monitoring bank activities more straightforward—have been swept away by the Gramm-Leach-Bliley Act. Under these increasingly complicated circumstances, using private-sector monitors to complement the efforts of government regulators is likely to result in better risk monitoring and control.62
The previous Part discussed some of the significant limitations of the current approach to bank regulation and supervision, which relies [*PG212]almost exclusively on the government to monitor bank risk taking. Many of these shortcomings in bank regulation can be ameliorated by an increased use of market discipline—that is, by heightening the ability of private providers of bank capital (stockholders, depositors or other debtholders) to influence bank behavior and to restrict excessive bank risk taking.
There are three ways in which market discipline promises to benefit the existing bank regulatory and supervisory structure. First, private suppliers of funds to a bank will constrain bank behavior through enforcement of contractual restrictions by debt investors and exercise of voting rights by stockholders. Regulators will be able to observe how the private sector responds to bank activities—such as what conditions are being imposed in debt contracts—and, thereby, determine what risks the market believes are material. Second, private actors that have already invested in a bank will punish inappropriate bank behavior by selling their shares or bonds in the secondary market. A declining share or bond price in the secondary market will provide a signal to bank management and to third parties, such as bank regulators, that an institution is moving in an unhealthy direction.63 Third, investors in the primary market that are contemplating an investment in a particular bank will exact a premium from a bank if they believe that the bank is engaging in risky activities.64
This Article proposes a program to enhance the market’s ability to regulate bank behavior and thereby take advantage of the private sector as a watchdog. Use of private-sector monitors is necessitated by the enactment of the Gramm-Leach-Bliley Act, which permits banking organizations to engage for the first time in a broad array of new activities.65
[*PG213] The first step in designing a program that deploys private-sector resources to assist bank regulators is to determine which participant in the financial markets is in the best position to monitor bank activities and discipline high-risk banks. The first criterion for selecting this market participant is the participant’s ability to monitor and discipline banks effectively. The second criterion is the alignment or commonality of that participant’s interests with the interests of the federal government as bank regulator. The greater the commonality of interests between the market participant and the federal government, the greater the likelihood that the market participant will take the same posture as the government with respect to bank risk taking. Similarly, the closer the alignment of interests, the more information the government will be able to glean by observing the market participant’s behavior.
There are three potential private-sector contenders for this regulatory duty: equityholders, both common and preferred; depositors; and other creditors. We examine each in turn.
Holders of common stock have some natural advantages as disciplinarians vis-à-vis other market participants. First, common equityholders have a unique ability to monitor bank activities. They possess legal rights to inspect the books and records of a corporation, and equityholders in a publicly held company are entitled to receive periodic financial and other information about the corporation and its operations.66 More importantly, common stockholders have tools to enforce their will directly on a firm and its management. Common stockholders elect the directors of a company and, through the directors, appoint the company’s management and supervise its operations. If common stockholders do not approve of the conduct of their company, they can exercise their voting power to change the company’s directors, management and policies. Moreover, state corporate law generally confers upon the directors and officers of a corporation [*PG214]a fiduciary duty to act in the interests of common stockholders and not in the interest of creditors or other corporate constituencies.67
In addition, if equityholders are dissatisfied with the firm’s performance, they generally have the right and ability to sell their stock in a publicly held corporation. An exodus of stockholders will create sell-side pressure on the corporation’s stock price. Corporate managers, who themselves are typically stockholders, will act in the interests of stockholders to prevent such sales and the resulting decline in the company’s stock price.
Despite these relative advantages, holders of common stock would not be the best purveyors of market discipline to banks, in part because their interests are not well-aligned with the paramount interest of the government in averting (or at least cabining the systemic and other risks of) bank failure. Common stockholders prefer higher-yield strategies than other corporate constituencies and the federal government.68 A bank’s income must cover debt charges before it can be shared with equityholders and, while shareholders’ risk of loss cannot exceed the amount they originally invested in the corporation, their potential gain is limited only by the size of the future earnings stream of the company.69 Consequently, the primary incentive of common stockholders, especially well-diversified stockholders, is to maximize the expected profitability of the companies in which they have invested—and not necessarily to limit risk taking.70
[*PG215] Moreover, as a firm moves closer to insolvency, common shareholders may have additional incentives to encourage risky activities. In insolvency, shareholders stand last in line for repayment of their initial investment, and shareholders may not receive any portion of their investment on liquidation of the firm. Therefore, common shareholders of a distressed firm may prefer that the firm take greater risks to gamble its way back to solvency.71
These risk preferences are not consistent with the primary goal of bank regulation. Thus, although shareholders have powerful incentives to monitor and influence bank behavior and the legal authority to supervise and control effectively such behavior, shareholders’ motives are not sufficiently aligned with those of bank regulators.72
Having determined that common shareholders are not a good source of market discipline, we next look to preferred shareholders as another possible class of equityholders that may exert discipline on banks. Preferred stockholders are more risk averse than their common stockholding brethren. Preferred stock generally is entitled to receive a fixed dividend and has a fixed preference upon liquidation of the company.73 For this reason, the incentive of a preferred stockholder is not simply to maximize the risk-adjusted profitability of the company but, rather, to ensure a steady and perpetual stream of preferred dividends. The incentives of preferred stockholders, therefore, [*PG216]are more closely aligned with those of bank regulators. The incentives are not, however, aligned with the government as well as the interest of debtholders.74 The claims of a preferred stockholder are junior to the claims of all the creditors of a company,75 including, in the case of a bank, depositors and the FDIC. This ranking again gives preferred shareholders a greater incentive to allow bank risk taking because, in a liquidation, they will only be paid after all the creditors’ claims are fully satisfied.
Preferred stock has three further disadvantages as an indicator of a bank’s risk position that, along with the incentive misalignment, lead us to search elsewhere for the most efficient source of private discipline. First, preferred stockholders generally do not have voting rights; the ability of preferred stockholders to control the behavior of a company is limited to negotiated voting rights in the certificate of designations or other instrument creating the preferred stock. Second, holders of preferred stock are typically unsophisticated retail investors.76 Third, preferred stock is a relatively uncommon capital instrument that is rarely traded in a deep, liquid secondary market.77
The next class of private actors that is available to provide discipline is a bank’s depositors. In many respects, depositors are a natural choice as a disciplining agent. First, they are a bank’s most common stakeholder.78 In addition, depositors generally stand in a creditor-debtor relationship with a bank; they have loaned a sum of money to [*PG217]a bank and have a right to withdraw that money on short notice, perhaps earning interest during the period of the deposit. For this reason, depositors do not gain if a bank engages in risky behavior. Depositors benefit only if the bank is able to repay their deposits to them.
One significant problem with depositor-based discipline is that, under the current scheme of federal deposit insurance, most depositors—except for those uninsured depositors who have placed over $100,000 with a bank—have no incentive to monitor or influence a bank’s activities. If a bank fails, the federal government will fulfill the bank’s contractual obligation to return the depositors’ principal.79
Even if deposit insurance were abolished or significantly revised to lower the maximum amount insured or to require co-insurance or deductibles, depositors would not be the ideal agents of market discipline. Small depositors do not have the financial sophistication and acumen necessary to monitor bank activities, especially the types of complex financial transactions most likely to put an institution at risk, and collective action problems would inhibit joint action. Furthermore, each depositor would have so little at risk that it would make no financial sense for him or her to spend much time analyzing the riskiness of the bank. Co-insurance, a scheme in which depositors would have some money at risk, would give depositors greater incentive to discipline banks than in the curent system but would still leave depositors with something less than an adequate motivation for watching their bank. In addition, many depositors are not investors; for such depositors, the convenience of a bank’s location or service menu is paramount.80 Finally, and perhaps most importantly, the monitoring benefits of depositors generally would be outweighed by [*PG218]the increased risk of bank runs and other adverse effects in a world without deposit insurance.81
The typical uninsured depositor has more funds in the bank than the typical insured depositor and, one might expect, has a greater ability to monitor bank behavior. In addition, uninsured depositors have incentives analogous to those of the bank regulators—they want a bank to behave conservatively so as to ensure the return of their principal with interest. And, if a bank were to fail, the uninsured depositors stand immediately below the FDIC in the capital structure and insolvency waterfall of a bank.82
Notwithstanding this close incentive alignment between uninsured depositors and government regulators, uninsured depositors would not be good disciplinarians of bank behavior. First, despite the size of their deposit accounts, uninsured depositors are not necessarily savvy investors; many of them are simply large manufacturing or other nonfinancial corporations parking liquid assets. These companies may not have the time or financial sophistication to monitor closely the riskiness of the bank in which their excess cash resides. They may choose a particular bank because of its location or range of business services, not because of the bank’s risk posture. For example, many businesses establish accounts at banks that are willing to extend credit to them. For most of these depositors, the transaction costs of changing banks are enough to prevent any shifting of deposits to take advantage of a risk differential. Second, uninsured depositors have no de jure control over the activities of a bank and probably have insufficient bargaining power—due both to the small size of their investments relative to the total assets of the bank and to the short-term nature of their investment—to obtain contractual controls over the [*PG219]activities of a bank.83 Third, uninsured depositors would be the first possible beneficiaries of a regulatory extension of the safety net in the event of bank insolvency. Even if regulators were able to commit not to bail out a bank’s nondeposit creditors in the event of insolvency, uninsured depositors may yet believe that they would receive protection from the FDIC if their bank failed.84 Finally, uninsured depositors generally have short-term or demand deposits; they generally have the contractual ability to get their money back from the bank quickly and will be most inclined to do so when the bank most needs the deposits to stay put. This leads yet again to the specter of bank runs.85
Having concluded that none of the other private-sector actors is suitable, nondeposit creditors are the remaining potential source of market discipline. There are two classes of non-deposit creditors that may serve as market disciplinarians. The first class is comprised of short-term creditors, such as interbank and corporate creditors that supply funds, often on an overnight basis, to a bank. The second class is comprised of long-term creditors, such as holders of bank-issued debt securities.
Interbank creditors are an intriguing, but ultimately unsatisfactory, source of creditor discipline.86 To be certain, interbank credit arrangements do provide banks with a reason to remain well-capitalized. A bank transacts with a host of counterparties that may not hold debt securities in the bank but, nevertheless, enter into contracts with the bank. The value of these contracts is contingent on the continued solvency of the bank.87 A counterparty entering into, for example, an interest rate swap with a bank becomes exposed to the credit risk of the bank and hence, one would expect, will charge a higher price to a bank that is in weak financial condition.
Interbank credit arrangements, however, are not good sources of market discipline. To begin, most interbank transactions are conducted pursuant to standardized contracts,88 and counterparties have little flexibility to impose special terms. In addition, because many interbank contracts have special bankruptcy priority89 and are short-[*PG221]term,90 bank counterparties have a reduced incentive to price contracts according to the bank’s credit risk or to engage in significant monitoring of the bank’s financial condition. For this reason, the price one bank charges another for a swap or other interbank credit arrangement is determined only in small part by the credit risk posed by the borrower.91 Price signals, consequently, will be ambiguous and will not provide regulators with much information about the bank’s risk posture.
Leaving interbank and other short-term creditors aside, we turn to the last potential source for market discipline: the bondholder or long-term creditor of a bank. In contrast to the other potential candidates, these debtholders are well suited to act as disciplinarians. Bondholders have interests that are closely aligned with those of government regulators, and they possess tools that enable them to monitor bank conduct and cabin bank risk taking.
Although long-term debtholders, unlike holders of common stock, are not entitled to elect their borrower’s directors and do not have voting power over all of a bank’s significant policy decisions, these debtholders routinely contract with their borrowers for at least negative control, or veto power, over the significant corporate decisions that might adversely affect the interests of the debtholders. For example, bond indentures and loan agreements typically restrict a borrower’s ability to increase its leverage, enter new businesses, sell significant assets or merge with other companies without the approval of some proportion of the debtholders. Debtholders also frequently negotiate informational covenants pursuant to which borrowers agree to furnish detailed annual, quarterly and even monthly financial and other data to the debtholders. In addition, debtholders routinely sub[*PG222]ject borrowers to minimum financial ratios.92 These types of covenants are precisely the kinds of provisions that are necessary to discipline bank behavior.
The typical investment of a long-term debtholder is also much larger than the typical investment of a depositor and, accordingly, we should expect that the typical long-term creditor has a much greater incentive to monitor a bank. Moreover, these nondeposit creditors are almost uniformly institutional investors that possess the financial sophistication and business acumen necessary to monitor their borrowers. Finally, because the typical debt investment is long term, most debtholders cannot put their investment back to the bank in the same manner that depositors can withdraw their funds.93 Sudden, devastating bank runs will not be the market disciplinary mechanism employed by long-term debtholders.94
In the previous Part, we determined that the market participant that can provide the most effective market discipline of banks is the non-depositor creditor that holds a long-term debt instrument issued by the bank. This Part seeks to determine the debt instrument that best enables such creditors to discipline banks. The task is to design a capital-enhancing debt instrument, the holders of which would have incentives most closely aligned with federal bank regulators. This Part sets forth proposed terms for the instrument and analyzes which banks ought to be required to issue such debt and how much and to whom.
The first question to address is the priority of the debt security in the capital structure of the bank. As noted in the previous Part, the goal is to find debtholders with incentives as similar as possible to those of the federal government, as bank regulator, and to the FDIC, the agency responsible for guaranteeing deposits and resolving failed banks. Because the FDIC guarantees depositors, the debt instrument should stand on an equal footing with the bank’s insured depositors. The relevant security, however, should also provide a capital cushion to the bank that will absorb losses before the federal deposit insurance fund incurs a loss. Hence, the debt must, at a maximum, have a priority one stage below that of the bank’s insured depositors. Balancing these two competing interests, it seems that the best place for the debt security in the capital structure of the bank is in a tier of debt immediately subordinated to the bank’s depositors and general creditors (including trade and interbank creditors) and immediately prior to the remainder of the bank’s debt. Holders of this debt would be paid in a bank’s insolvency only after depositors’ and general creditors’ claims are fully satisfied, but before other creditors and equityholders receive money.
The implementation of this priority scheme may require a transition period. Some banks have debt outstanding, and requiring these banks to issue a senior subordinated debt instrument may compel them to violate certain covenants in their existing debt indentures or loan agreements. Those banks able to redeem the existing problematic debt without paying a substantial premium should be required to do so; other banks may be permitted to issue junior subordinated debt until such time as their existing problematic debt instruments mature.
The maturity of the debt instrument should be determined by balancing two competing considerations. On the one hand, a debt security with a long maturity is preferable. From a capital adequacy perspective, only long-term debt contributes to a bank’s capacity to incur losses over a period of time and remain solvent.95 Short-term [*PG224]obligations do not provide much support to a bank in distress because short-term creditors can withdraw their funds from the bank precisely when the bank most needs them.96 Moreover, holders of long-term debt securities would have incentives more closely aligned with those of the government—ensuring the perpetual health of the bank—than would holders of short-term debt securities. Accordingly, long-term debt securities would provide more relevant pricing information to regulators than would short-term securities.97
There is, however, a countervailing consideration. Requiring a firm repeatedly to issue short-term obligations has the potential benefit of forcing the bank to access the capital markets more frequently. As discussed above, a firm that knows it will have to raise capital frequently in the primary market will be more likely to behave appropriately—in a fashion that will make prospective investors consider the firm to be desirable.98
In balancing these competing interests, we propose that Subject Banks (as defined below) should be required to issue subordinated debt securities with a minimum maturity of six years, and should be required to roll over a proportionate amount of their subordinated debt at least once every two years.99 Six-year debt should provide a [*PG225]sufficiently long-term investment to prevent runs on the bank’s capital and to align investors’ interests with those of the government and the long-term solvency of the bank.100
Although most of the currently outstanding subordinated debt issued by U.S. banking organizations has a ten-year maturity,101 banks should be accorded as much flexibility as possible in determining the tenor of their subordinated debt. Some banks will find that they can issue six-year debt much more cheaply than ten-year debt; other banks will find that ten-year debt, or debt of even a longer term, is the most cost-effective approach for them. Under this proposal, banks would be able to issue ten-year debt, the current market standard, or debt of a shorter or longer term to meet their individual corporate finance needs. This flexibility also will permit banks to adapt their capital structures quickly to keep up with evolving market trends without having to wait for regulators to recognize the evolution and react with revised regulations.
Requiring banks to issue subordinated debt in staggered tranches, at least once every two years, will provide substantial capital markets discipline on bank behavior. A requirement to tap the capital markets at least once every two years should not be too burdensome—large bank holding companies typically issue subordinated debt into the capital markets once or twice per year.102 Providing banks flexibility as to the frequency of their primary offerings would allow banks to avoid having to issue debt in times of market stress and to design a debt program that best suits their particular funding needs.
[*PG226] One commentator has proposed that certain banks be required to issue puttable subordinated debt.103 Under that scheme, large banks would be required to have a minimum amount of subordinated debt outstanding at all times, and banks would have to stand ready to redeem at par value within ninety days any debt “put” to them by investors. If investor puts reduced a bank’s outstanding subordinated debt below the prescribed minimum, the bank would have to issue additional subordinated debt, sell risky assets to reduce its minimum debt requirements or face closure by regulators.104
This approach certainly would increase the amount of market discipline to which banks are subject, but requiring banks to issue redeemable debt is not a workable proposal for many reasons. Foremost among them is the threat of bank runs. Puttable subordinated debt too closely resembles a demand deposit and, while holders of bank subordinated debt will by and large be more sophisticated investors than depositors, the creation of a tier of uninsured quasi-demand obligations places a bank’s capital base at risk of rapid erosion.105 Second, it seems unlikely that a bank experiencing significant puts of its subordinated debt would be able to replace potentially one-third to one-half of its capital base on ninety days’ notice. Issuing securities, especially public securities to new investors, is typically a lengthy process. Third, the puttable debt would need to be in the form of floating-rate debt securities,106 and demandable floating-rate debt securities are a rare financial instrument. When introducing a relatively new type of security into the capital markets, such as bank subordinated debt, it would be best to keep the novel elements to a minimum. Fourth, a right to put a security back to the issuer at par encourages investors to exercise the put at the slightest sign of the [*PG227]issuer’s eroding financial condition. Why sell into the market at 95% of par value, if the issuer stands ready to redeem at 100% of par value at any time?107 A ninety-day put waiting period will somewhat mitigate the tendency to put at the first sign of trouble, but not enough. The put mechanism will substantially eliminate the debt securities’ ability to convey pricing information to bank managers, the market and regulators. Regulators, for example, will have little advance warning of a bank’s financial difficulties; rather, at the first sign of weakness, all of the investors would rush to put their securities back to the bank.
We propose that each Subject Bank should be required to have outstanding at all times an amount of subordinated debt equal to two percent of the bank’s risk-adjusted assets, as calculated under the Capital Guidelines applicable to such bank.108 The Capital Guidelines implicitly require banks to hold tier 2 capital against four percent of their risk-adjusted assets.109 The subordinated debt would count as tier 2 capital for a bank only to the extent currently permitted by the Capital Guidelines.110
We expect that most Subject Banks will use the new subordinated debt to replace other outstanding securities counting as tier 2 capital. Because the Capital Guidelines require banks to reduce proportionally the amount of tier 2 capital credit received for subordinated debt securities with less than five years remaining until maturity, our proposal will require banks to operate with a slightly higher aggregate [*PG228]outstanding amount of capital securities.111 This implicit raising of the minimum capital requirements will increase a bank’s cost of financing and may reduce its return on equity. Of course, if bank subordinated debt has the desirable effects laid out in this Article, regulators should consider at a later date reducing the levels of required tier 2 capital.112
Our proposal requires that banks, as opposed to their parent holding companies, issue the minimum amount of subordinated debt.113 This aspect of the proposal will help ensure that debtholders are concerned solely with the safety and soundness of the debt-issuing banks, rather than the well-being of banks’ parent holding companies or affiliates, and that debtholders’ interests will be closely allied with the interests of bank regulators.114
Because the activities of banks would be the principal, if not exclusive, source of funds for bank debt servicing, debtholders would monitor the activities of banks and would discipline banks that seek to engage in high-risk activities. The activities of a bank’s affiliates or its parent holding company would affect debtholders only to the extent [*PG229]that those activities alter the bank’s ability to service its subordinated debt. Any inter-corporate transactions that are beneficial to the bank should meet with debtholder approval. On the other hand, debtholders, like government regulators, would guard vigilantly against any attempts by a parent or an affiliate of a bank to divert resources from the bank or to engage in transactions that are detrimental to the bank.115 This vigilance against inter-corporate transactions that are disadvantageous to banks—yet beneficial to non-bank affiliates within the same holding company—is especially vital in the post-financial modernization world, where banks may freely affiliate with a wide variety of other financial firms.
It is important that banks issue the required subordinated debt because, as detailed in the first Part of this Article, it is banks—and not their parent companies—that are special. After all, the goal of our proposal is to give banks a private-sector constituency that is interested in banks—not the corporate owners of banks—and that helps federal regulators safeguard banks.116
Requiring the corporate owners of banks to issue subordinated debt would not serve the goals of risk monitoring and disciplining at the bank level. Indeed, the holders of parent-company debt may be significantly insulated from risk taking at the subsidiary bank level by the other varied businesses in which the parent company may engage. This will be especially true in the wake of the Gramm-Leach-Bliley Act, which authorizes the creation of broadly diversified financial enterprises. The banking operations of the new financial holding companies may represent only a small part of the overall operations of the enterprise, and subordinated debtholders of such a financial holding company may have no need to be vigilant as to the activities of the [*PG230]firm’s subsidiary banks so long as the other parts of the business are managed prudently.117
Admittedly, requiring banks that are subsidiaries of parent holding companies to issue subordinated debt may create accounting and disclosure issues for parent firms. For example, parent holding companies may need to issue detailed public financial reports and break-out operating results at the subsidiary bank level, practices with which they may not be familiar and which may involve added expense. Moreover, financial analysts and other market participants may, for the first time, raise questions regarding a subsidiary bank’s performance, rather than just the overall financial holding company’s results.
Subsidiary banks also may need to be managed differently than under current practice, with greater attention given to the legal entity of the bank and how transactions affect the bank’s subordinated debtholders. Currently, most holding companies are operated as integrated organizations that make use of centralized risk-management systems.118 That management system may need to be changed because subordinated debtholders will place greater emphasis on the performance and risks undertaken at the bank level. That change may involve greater inefficiencies than a system that permits business-line management across the entire financial holding company, but that change also may result in better firewalls between the bank and its affiliates, in better corporate policies to safeguard against inter-company transactions that adversely affect the bank, and in less risk that a court might pierce the corporate veil and hold the bank liable for an affiliate’s debts or misconduct.
Although requiring banking organizations to issue debt at the bank level may create the inefficiencies described in the preceding paragraphs, those inefficiencies may be offset by the reduced costs of bank-level issuance. Empirical evidence has shown that the market requires a higher return on bank holding company subordinated debt than bank subordinated debt.119 Accordingly, banking organiza[*PG231]tions should be able to issue subordinated debt at the bank level more cheaply than at the holding company level.
Finally, issuance of subordinated debt at the bank level will help avoid corporate control issues that may arise if the debt were required to be issued at the parent company level. Because, under this proposal, the subordinated debt will be typically issued by a bank that is a wholly owned subsidiary of a parent holding company, there is little risk that the subordinated debt will be accumulated by a potential corporate suitor as a means for acquiring the parent holding company. By contrast, there would be a greater risk that acquirers of debt that is issued at the parent company level would seek to combine that debt holding with other direct or indirect stake holdings in the parent company in an attempt to exercise control or controlling influence over that company.120
Our proposal divides banks into three tiers and imposes different requirements on banks in each tier. The smallest banks (“Tier III banks”)—those with total consolidated assets less than $1 billion—would not be compelled to issue subordinated debt but would be allowed to opt-in to the system. Intermediate-size banks (“Tier II banks”)—those with total consolidated assets between $1 billion and $10 billion—would be required to issue subordinated debt but would be permitted to do so in a variety of manners. The largest banks (“Tier I banks”)—those with total consolidated assets in excess of $10 billion—would have to issue subordinated debt in the form of securities registered with the Securities and Exchange Commission (“SEC”).121
We believe that thresholds based on asset size alone would be the most practical. Our subordinated debt program is designed to mini[*PG232]mize losses to the deposit insurance funds and to assist regulators in managing systemic risk in the financial infrastructure of the United States. Although certain non-size criteria, such as the geographic scope of an institution or the nature and complexity of an institution’s activities, are relevant to systemic risk assessment, we believe that it is more important to provide banking organizations with clarity as to the scope of application of the subordinated debt rules than to pick up a handful of small organizations that pose an outsized risk to the financial system.122 Requiring such small banks to issue debt securities would impose on them unacceptably high financing costs.123
Holding companies with multiple bank subsidiaries would be required to aggregate the assets of all their banking affiliates to determine which tier their banks fall under. Additionally, the subsidiary banks of a holding company would be required to issue, in the aggregate, an amount of subordinated debt as if they were a single bank in that tier. If a holding company determines that direct issuance of subordinated debt by multiple banks would be inefficient, the holding company could merge its banks or the banks could collectively issue a single pool of subordinated debt securities under the rules applicable to that tier.124 The aggregation requirement would ensure that two holding companies that control the same amount of banking assets do not receive disparate treatment under our proposed rule. The requirement also would ensure that holding companies do not maintain separate banks merely to escape the subordinated debt requirement.
The proposal envisions excluding Tier III banks from the mandatory elements of the subordinated debt plan.125 The legal, administra[*PG233]tive, disclosure and other fixed costs of issuing subordinated debt would be excessive for the small banks in Tier III. In addition, if small banks were made subject to this requirement, they would be required to issue relatively small amounts of debt.126 The amount of the debt issued may not be sufficient to draw adequate investor interest, and small banks might have to pay a substantial interest-rate premium to attract debtholders. Investors also might demand a rate premium from Tier III banks to compensate for the fixed costs of monitoring, which would be relatively high compared to the small amount of debt issued by such banks.127
The exclusion of Tier III institutions is not problematic for our proposal. Tier III banks, by and large, tend to engage in traditional banking activities with which regulators have the greatest familiarity and expertise. As a result, the marginal benefit of the additional monitoring and discipline provided by private-sector creditors would be negligible in the case of small, traditional banks. Moreover, and perhaps more importantly, the failure of a Tier III bank would not pose the same systemic risk to the economy or to the federal deposit insurance fund as would the failure of a larger institution. Finally, the proposal would permit Tier III banks to opt-in to the subordinated debt scheme at their own discretion.128
Banks with sufficient assets to place them in Tier II would be required to issue subordinated debt. Tier II institutions are large enough to be able to absorb the costs of periodic debt issuance and to [*PG234]issue debt without having to incur a sizable interest-rate premium. Furthermore, the amount of debt issued by a Tier II bank would be of sufficient quantity to attract investor interest. Tier II banks also are far more likely than Tier III institutions to engage in non-traditional financial activities, such that these institutions and their federal regulators may benefit from the discipline of debtholders. Perhaps most importantly, Tier II banks are of sufficient size that their failure may have adverse systemic consequences.129
Under our proposal, Tier II banks would be permitted to issue unregistered debt securities through private placements.130 Although Tier II banks are large enough to be required to issue subordinated debt, the size of a Tier II bank’s debt offering may not be sufficient to require that the debt be issued in a registered public offering.131 Moreover, a Tier II bank may not issue enough debt securities to facilitate the creation of a public market in the security.
Because of the lack of a public market for the Tier II banks’ subordinated debt securities, the debt securities of Tier II banks likely would be held principally by small numbers of sophisticated institutional investors, such as pension and mutual funds and insurance companies. These investors would possess the financial and business acumen necessary to evaluate the risk posture of the banks. The debt-holders also would have sufficient resources invested so that they will have incentives to discipline bank managers and limit their risk-taking activities. The investors may demand risk premiums to protect themselves from excessive risk taking on the part of a particular institution, or the debtholders may negotiate ex ante restrictive covenants that impose limits on bank activities. The existence of such premiums or covenants should serve as important signals to federal regulators both as to the private sector’s evaluation of a particular bank’s risk position and a particular bank’s willingness to limit its activities in response to investor concerns.
Tier I banks are the largest banks in the country.132 These institutions and their parent holding companies are typically at the forefront of financial modernization and change. Banks in Tier I engage in a broad array of financial activities beyond the traditional banking functions of taking deposits and making loans. As a result, these institutions and their federal regulators would benefit most from private-sector discipline. In addition, by virtue of being the nation’s largest banks, their failure poses the greatest risk to the financial system.
Under our proposal, Tier I institutions would be called upon to issue subordinated debt to the public in offerings registered with the SEC.133 These Tier I banks and their bank holding companies already issue a substantial amount of subordinated debt; consequently, transition costs for these banks should be low.134
A bank’s asset size is subject to a fair amount of volatility. In order to avoid logistical inefficiencies for a bank with an asset size that meanders around $1 billion or $10 billion, a bank with assets that cross the Tier I or Tier II thresholds from below should be permitted to defer compliance with the subordinated debt requirements of the new tier for a period of two years. Only banks with assets that exceeded $1 billion for eight straight quarters would be automatically subject to the privately issued subordinated debt program; and only banks with assets that exceeded $10 billion for eight consecutive quar[*PG236]ters would be automatically subject to the publicly issued subordinated debt program. This transition rule would save banks from the regulatory burden of moving in and out of the subordinated debt program and would provide growing banks with a substantial period of time to prepare for compliance with the program. Similarly, we propose that a bank should only be allowed to escape the subordinated debt requirements for a particular tier if its assets remain below the tier’s minimum thresholds for eight consecutive quarters.135
The question arises as to how foreign banks with U.S. branches and agencies should be treated and whether these institutions also should be required to issue subordinated debt in order to engage in banking and non-banking businesses in this country. On the one hand, to the extent that the subordinated debt requirement is viewed as an additional burden on U.S. banks, foreign banks that operate in the United States may gain a competitive advantage over their U.S. competitors if these foreign institutions do not need to comply with the requirements of this proposal.
At the same time, however, imposing the subordinated debt program on foreign banks raises several difficulties. To begin with, requiring foreign banks operating in the United States to issue subordinated debt would result in a U.S. capital requirement being imposed extra-territorially on foreign-based institutions and would be contrary to the internationally accepted policy of consolidated “home country” supervision, in which a bank’s home country takes the lead in setting the capital and other standards that should apply to that institution. The United States’ imposition of a subordinated debt requirement on foreign banks could effectively supersede capital benchmarks established by the home country supervisors of those foreign banks and would effectively require those foreign banks to manage their international operations to conform to the U.S. subordinated debt requirements. Were each host country to follow the United States’ precedent and to adopt similar unique capital requirements, internationally active banks could be subject to an array of conflicting and possibly irreconcilable standards. In addition, imposing the subordinated debt requirements on foreign banks may be impractical given differences [*PG237]in bank accounting practices, balance sheet structures and bank and capital market regulation in different countries.136 Perhaps most importantly, few foreign banks with U.S. branches and agencies hold deposits that are insured by the U.S. government.137 For this reason, the failure of a foreign bank does not expose a U.S. federal taxpayer to the same potential liability as the failure of a U.S. bank, and the government does not have the same interest in constraining a foreign bank’s risk as the government has in keeping a domestic institution’s risk profile in check.138
In any event, the imposition of a subordinated debt requirement on U.S. banks, and not on foreign banks operating in the United States, should not give rise to significant competitive inequities. As discussed in greater detail in the next Part, U.S. banks that issue subordinated debt under our proposal will benefit from fewer examinations and less general regulatory oversight. Foreign banks, by contrast, would not be eligible to take advantage of these benefits. In addition, to the extent that a subordinated debt program is successful in the [*PG238]United States, it may also be implemented internationally as part of the Basle Capital Accord, in which case all banks—both U.S. and foreign—will be subject to similar subordinated debt requirements.139
Our proposal mandates that the subordinated debt be issued to investors that are different from and independent of the managers and equityholders of the bank, its affiliates or its parent holding company.140 In addition, our proposal requires that debtholders not have significant business relationships with the debt-issuing bank or its affiliates.141 Subordinated debt held by such insiders of a bank or its affiliates would not be counted towards the amount of subordinated debt required to be issued by the bank.
This aspect of the proposal seeks to ensure that subordinated debtholders do not have relationships with the bank that would compromise their willingness and ability to monitor and, where necessary, impose limits on the activities of debt-issuing banks. Our concern is that if, for example, an insurance company not only held the bank’s subordinated debt but also sold a significant volume of its insurance policies to the bank’s customers and obtained credit through the bank, that insurance company may have priorities and goals that conflict with its role as private-sector monitor of the bank’s activities. The insurer may be willing to purchase and hold the bank’s subordinated debt not because the bank’s subordinated debt securities are an attractive investment for the insurer but because ownership of the securities may be a means to protect the insurer’s business relationships with the bank.
[*PG239] This restriction will necessitate regulatory monitoring in order to be successful. Bank examiners may need to review bank subordinated debt periodically to ensure that outstanding debt is actually held by independent investors. Regulators also may need to craft special rules or give heightened supervisory scrutiny to banks that have issued the required amount of subordinated debt to third parties but have also issued a substantial amount of such debt to insiders and affiliates. Negotiation of the terms of the debt and decisions about enforcement of these terms may be affected by the presence of bank insiders and affiliates in the class of investors. The quality of pricing information conveyed by the secondary market for the debt may also be weakened if a substantial proportion of the debtholders are related to the bank in ways other than as debt investors.
A single federal financial regulatory agency could be entrusted to administer and ensure uniform compliance with the subordinated debt program. A natural choice for this duty would be the FDIC, which is the only federal banking agency that has a role in supervising all banks, whether state or federally chartered.142 The FDIC’s principal mission is to maintain public confidence in the banking system, and the agency is specifically entrusted with preserving the safety and soundness of the federal deposit insurance fund—the very goal of our proposal.143 For these reasons, the FDIC has a direct stake in keeping bank risk under control, and the FDIC’s interests and those of subordinated debtholders are closely matched.
If the FDIC were chosen as the administrator of the subordinated debt program, banks required to—or, in the case of Tier III institutions, that choose to—issue subordinated debt would be required to report annually to the FDIC and to certify that they have issued the [*PG240]required amount of debt to non-insiders. Banks also would be required to reveal the pertinent terms of their subordinated debt contracts. The FDIC’s principal obligations would be to review these submissions to ensure compliance with the requirements of the program and to watch for signals from the market as to the risk posture of the issuing banks. In the ordinary course, the FDIC’s role should not result in an added level of day-to-day oversight of banks or give rise to significant added compliance burdens to subordinated debt issuers.144 The FDIC, however, would have the residual authority—such as the authority it currently possesses by virtue of its administration of the federal deposit insurance fund—to inquire of and to conduct on-site examinations for compliance in situations in which bank submissions prove insufficient, unclear or otherwise inadequate. The FDIC also would have the power to sanction or levy fines against banks that fail to meet their subordinated debt obligations.145 In addition, the FDIC would be empowered to conduct examinations if market signals or subordinated debt contract terms indicate that the issuing bank was engaged in risky activities.
Entrusting all of the administrative and monitoring duties of the subordinated debt program to the FDIC, however, could be problematic. For example, the primary federal regulator of each issuing bank currently has responsibility for the capital adequacy of that bank, and it would be natural for that regulator to play a leading role in determining whether the bank’s subordinated debt issuance was adequate. In addition, a bank’s primary federal regulator is the agency most familiar with the issuing bank and would be in the best position to understand the various market signals provided by subordinated debtholders regarding the risk profile of that bank.
Forcing financial institutions to alter their capital structure will have secondary effects. For example, requiring certain banks to issue debt (in replacement of equity or as a supplement to existing equity) may reduce the value of their enterprises and will certainly raise the [*PG241]financing costs of those banks.146 In addition, compelling a bank to issue debt and to increase its leverage may increase the bank’s expected costs of financial distress by an amount greater than the expected benefits of the debt. Banks are by their nature highly leveraged companies—deposits are a form of leverage—and forcing them to obtain still more leverage may have high costs. These costs may outweigh the advantages conferred by increased discipline. We cannot say outright, however, that our proposal would impose inefficiencies in capital financing decisions simply because banks are not currently issuing the debt that the proposal would require. The Capital Guidelines already distort bank capital structures in an anti-debt direction; it is possible that our rule would restore banks’ capital structures to optimal levels.
At least one commentator has indicated a preference for standardized terms if banks are to be required to issue subordinated debt.147 Regulators could require all banks to issue subordinated debt of the same maturity, with the same frequency, and pursuant to debt contracts that contain the same set of covenants and conditions. We believe, however, that banks should be granted the maximum amount of flexibility possible in designing and issuing their subordinated debt consistent with achieving the regulatory goals of increasing market discipline and market signaling to regulators. The most obvious benefit of standardization is that it would provide regulators with signals that are more comparable across different banking organizations. This benefit is outweighed by numerous costs, many of which have been mentioned above. Investors in bank subordinated debt will value highly an ability to tailor covenants to the particular bank, and regulators will be able to obtain a great deal of relevant information about a bank’s risks by evaluating these covenants. Furthermore, banks will value an ability to negotiate covenants in, and determine maturities of, their debt instruments.
Some commentators have proposed establishing a maximum interest rate on a bank’s subordinated debt securities. Imposing such a cap on bank subordinated debt would be counterproductive. Determining the appropriate cap would be an insurmountable task; the cap would need to be raised or lowered continually to reflect the existing [*PG242]interest rate environment; and a uniform cap would treat different banking organizations with different capital structures unevenly. Similarly, banks should be free to issue fixed or floating-rate securities. Most U.S. banks and bank holding companies that currently issue subordinated debt issue fixed-rate securities.148 The capital markets, however, are sophisticated enough to price floating-rate debt as accurately as fixed-rate debt, and banks’ flotation of both fixed- and floating-rate debt should not hinder greatly regulators from making peer group comparisons.149
The one requirement that we would place on a bank’s subordinated debt is that it not be puttable at the option of the holder, for all the reasons set forth above. Banks should be free, on the other hand, to make the debt callable at the option of the bank at any time. Although, to be consistent with the Capital Guidelines, the bank should be required to consult with its appropriate federal banking agency prior to redeeming any such debt.150
Requiring banks to issue subordinated debt securities would improve the stability and soundness of the banking system by complementing federal regulators’ efforts to place limits on banks’ risk-taking activities. As this Article has detailed, subordinated debt-based market discipline would be an important complement to the existing scheme of bank regulation and supervision. Our proposal also would have other important collateral benefits. In addition to contributing meaningfully to risk monitoring, our proposal should lead to reduced frequency and cost of bank examination while simultaneously enhancing examination efficiency; increased disclosure of bank activities; an efficient set of restrictions regarding the proper scope of bank product diversification; and other collateral benefits.
A significant benefit of this proposal is that it would permit less frequent examinations of banks by federal regulators, while facilitat[*PG243]ing better assessments of the risk posture of institutions.151 By relying on signals from subordinated debtholders, regulators would be able to monitor the well-being of an institution without engaging in a direct examination.152 For example, regulators would have an indication of private-sector concerns about an institution if holders of subordinated debt began demanding a premium from a Tier I bank that was seeking to roll over its maturing debt securities or if a variance emerged between the secondary market price of subordinated debt securities issued by an institution and its peer group. Similarly, the imposition by subordinated debtholders of certain contractual covenants on a bank might indicate a particular set of risks that ought to concern regulators. The unwillingness of bank management to enter into certain covenants agreed to by other banking peers also may be an important indicator of a bank’s risk policies. By use of such market signals, regulators would not need to conduct full-scale on-site examinations with their current degree of frequency; rather, examinations could be more precisely and effectively targeted. Regulatory scrutiny could be directed at the particular institutions that the private sector signaled as raising concerns. When covenants indicate that particular activities are of greatest concern, regulators may direct attention to those bank activities.153
Reducing the frequency of bank examinations would reduce the monitoring and administrative costs of federal regulators, ease the regulatory burden on banks, and result in lower examination fees.154 These lower examination burdens and fees may be sufficient to entice Tier III banks—which, as noted in Part IV above, would not be required to issue subordinated debt securities—to opt-in to the subordinated debt program. These small banks may find that the cost of [*PG244]issuing subordinated debt is less than the cost of the current stricter examination procedures.
This proposal also would enable the FDIC to establish more quantitatively precise risk premiums for federal deposit insurance. Since the passage of the Federal Deposit Insurance Corporation Improvement Act (“FDICIA”) in 1991, the FDIC has been required to operate a risk-based assessment system for insured depository institutions.155 The FDIC’s current system determines a bank’s insurance premium based on the bank’s reported capital ratios and a subjective assessment of the bank’s financial soundness.156 The FDIC could use the pricing mechanism created by bank subordinated debt to assist it in its efforts to price banks’ deposit insurance premiums. Rather than rely on publicly reported capital ratios (which are increasingly subject to regulatory capital arbitrage) and supervisory assessments of financial soundness (which are highly subjective), insurance premiums could be calculated based in part on debt ratings or the market prices of the subordinated debt issued by banks under this proposal. Increasing the objectivity of the data used by the FDIC to determine insurance assessments will enable the FDIC to be both aggressive and fair in imposing insurance fees on banks.
In addition, this proposal would provide federal bank regulators with credible market signals of the deteriorating financial condition of a bank. As a bank deteriorates, the price of the bank’s subordinated debt securities will fall, and the implicit yield on the bank’s debt will rise. Bank regulators will easily be able to monitor these prices and could impose restrictions on banks with falling debt prices similar to those dictated by the prompt corrective action rules—restrictions on payment of dividends, acquisition of assets and engagement in new lines of business.157 Because different banks have different capital structures and will have subordinated debt securities with different terms, regulators monitoring bank debt prices will not be able to use uniform thresholds, like those used in prompt corrective action, to [*PG245]trigger sets of restrictions. Rather, regulators will be able to make rough interbank comparisons (between banks) and fairly precise intrabank, intertemporal comparisons (a single bank over time).
Another significant collateral benefit of our proposed system of subordinated debt would be to increase the amount and quality of financial information provided by banks and bank holding companies. At the present time, banks, especially banks that are wholly owned subsidiaries of parent holding companies, lack incentives to disclose financial information to the public. Under a system of market discipline, however, banks will have strong incentives to report timely, relevant financial information or face the prospect of being punished by a market that does not adequately comprehend the institution’s risk profile. Increased public disclosure about bank operations will improve the ability of market participants to distinguish strong banks from weak banks158 and also will provide regulators with another important source of information for their own assessment of bank operations and risks.159 In addition, market participants likely will be most interested in the types of transactions that expose banks to the greatest risks, including bank derivative activities.160 These also are the precise transactions in which regulators have the greatest interest.
A further benefit of our proposal may be to decrease opportunities for banks to engage in regulatory arbitrage and to prevent competition among various bank regulatory agencies from becoming a “race to the bottom.” The U.S. bank regulatory system does not vest supervisory authority over banks and banking organizations with a single regulator; rather, supervisory authority over banking firms is divided among several competing regulatory agencies. Banks may choose to be state or federally chartered. State-chartered institutions [*PG246]are principally regulated and supervised by state banking commissioners; federally chartered banks are regulated and supervised by the OCC, a division of the Treasury Department.161 Meanwhile, the Federal Reserve regulates bank holding companies and the nonbanking affiliates of banks.162
This division of authority allows banking firms to choose to structure their operations so that they fall within the purview of the regulator that the organization believes may be the most accommodating.163 For example, a bank holding company that owns both state-chartered and OCC-chartered banks can determine to engage in an activity (e.g., municipal bond underwriting) in its state bank or through a subsidiary of the state bank, both of which would be regulated principally by its state chartering authority; in its national bank or a subsidiary of the national bank, which are regulated by the OCC; or at the holding company level, regulated by the Federal Reserve. Giving regulated entities such choices leads to competition among the regulators, as each seeks to enhance the scope of its authority and the number of institutions and activities that it oversees.164
[*PG247] Such agency turf battles, however, can also create undesirable regulatory risks.165 For example, regulators may ignore the risks associated with an activity and may fail to penalize excessive risk taking for fear that such action will lead a bank to switch charters or to house that activity in an entity that is supervised by a competing agency. Subordinated debtholders will serve as an important check on this form of regulatory arbitrage. Debtholders will have a vital interest in preventing excessively risky activities from being conducted in a bank in which they have invested, and they will penalize banks, through the various mechanisms outlined above, that seek to take advantage of excessively permissive regulatory authority.166 Unlike bank regulators, subordinated debtholders will have no competing pressures that will dampen their ability and interest in safekeeping the bank’s assets.167
Finally, our subordinated debt proposal may lay the groundwork for further financial modernization—namely, authorization of the mixture of banking and commerce through the holding company structure.168 As a general matter, U.S. law has both barred banks from affiliating with firms that are engaged in commercial or non-financial businesses and prevented banks from engaging directly in commercial activities.169 The twin separations of banking and commerce have [*PG248]been maintained for several reasons. Banks have been prevented from affiliating with commercial firms because such banks may face conflicts of interest when they engage in lending activities. Banks may experience irresistible pressures from their commercial affiliates to extend credit, and otherwise lend support, to those enterprises or to withhold credit from their competitors. Banks have been prevented from engaging directly in commercial activities because such activities may pose risks to the safety of the bank. Commercial activities are viewed as inherently more volatile than traditional lending and deposit-taking activities, and are seen as likely to tax the ability of financial regulators to comprehend their supervisory charges.170
To the extent that these concerns are warranted, and there is significant debate as to whether they are, our subordinated debt proposal may represent a vehicle that can be used to permit the mixture of banking and commerce, either through affiliation or within a universal bank. For example, once a bank has issued the required amount of subordinated debt (or perhaps some additional amount of subordinated debt to be determined by federal regulators), it could be free to affiliate with commercial enterprises. Whatever risks to the bank arise from its commercial affiliations would be counter-balanced by the monitoring role of the subordinated debtholders. In short, the bank would have a private-sector constituency that would help federal regulators to police transactions between the bank and its commercial affiliates. If the private sector deemed a particular affiliation between a bank and a commercial firm to expose the bank to undue risks or exploitation, the bank’s subordinated debtholders would demand a risk premium from the bank or would negotiate covenants that would restrict the nature and extent of the transactions between the bank and that affiliate. Similarly, to the extent that a bank engages directly in risky commercial activities, subordinated debtholders would require a higher interest rate, insist upon restrictive covenants, deny credit or sell their debt into the market.
As stated in the Introduction, this Article is not written upon a clean slate. Numerous commentators have registered opposition to market discipline approaches to bank regulation on various grounds. This Part examines some of the most common and powerful objections to market-based proposals and attempts to explain how these objections can be overcome. Although none of the criticisms are fatal to the market disciplinary project, a discussion of their strengths and weaknesses provides valuable insight into the limits of market discipline.
The objection: Market discipline can only be an effective complement or supplement to bank regulation if market participants have timely access to comprehensive and credible information about bank assets and activities. Investors simply do not have a sufficient quantity or quality of information about banks.
The response: Some theory and evidence suggest that market participants have less (and less reliable) information about banks than they do about other kinds of firms. The theoretical argument generally runs as follows: banks hold few fixed and easily valued assets and the risks to banks’ mostly financial assets are hard to observe and easy for banks to change.171 The primary assets of most banks are loans to private-sector borrowers. The value of each loan in a bank’s portfolio is hard to observe and is contingent on the terms of the loan contract and the health and riskiness of the borrower.172 The loan portfolio of [*PG250]most banks is also volatile, as banks package and sell loans frequently into the growing secondary market for bank loans.173 The biggest banks, which have been shifting their activities over the past few years from more conventional lending into exotic securities and derivatives trading, are even more opaque than the smaller banks. The biggest banks are rapidly growing more dynamic, and their assets are quickly becoming more complex.174
A substantial body of economic literature has examined the question of bank opacity, and the results have been mixed. Some studies have concluded that banking organizations are more opaque than non-banking companies. Studies by economists Donald Morgan and Richard Cantor and Frank Packer have shown that the Moody’s and Standard & Poors rating agencies have more ratings disagreements over bank holding companies than over other firms of comparable size and risk.175 Similarly, a study by Robert Avery, Terrence Belton and Michael Goldberg showed that risk premiums on bank-related long-term debt are virtually unrelated to traditional accounting measures of bank performance and that the risk premiums are only weakly related to ratings conferred by private-sector rating agencies.176 Finally, Gary Gorton and Anthony Santomero concluded that accounting measures of risk only marginally predict the market-determined volatility of bank assets.177
Although these studies have shown that banks are somewhat opaque, many recent studies suggest that the market does exact an [*PG251]institution-specific risk premium from banking organizations.178 Julapa Jagtiani, George Kaufman and Catharine Lemieux concluded that, from 1992 to 1997, the market-priced credit risk for the debt of banking organizations was determined by accounting risk measures, private agency ratings and the regulatory rating of the borrower.179 Professors Mark Flannery and Sorin Sorescu found that, between 1983 and 1991, subordinated debt yields were sensitive to bank risk and that the sensitivity increased as the government withdrew de facto insurance coverage from uninsured liabilities.180 Moreover, a study by Thomas Cargill found that, between 1981 and 1987, bank certificate of deposit (“CD”) rates reflected “CAMELS” scores, the system used by federal banking regulators to rate the condition of banks.181 Herbert Baer and Elijah Brewer also found that rates of return to holders of uninsured CDs were correlated with changes in the banks’ market-to-asset ratios and the volatility of bank returns on equity.182
The empirical studies indicating the existence of some bank opacity do not amount to much of an objection to the market discipline hypothesis. First, the empirical studies that have shown a weak correlation between market-exacted risk premiums and accounting measures of bank risk only establish that the market uses more than accounting data to assess the riskiness of banks. Unless research can show that accounting measures of bank risk are the best predictors of bank failures, these studies do not prove enough. Second, the empirical studies showing that rating agencies disagree over bank securities more often than over non-bank securities only show that banks are somewhat more opaque than non-banks. Slight relative opacity, however, does not imply that the market is unable to evaluate the riskiness of banks; rather, it indicates only that the market considers assessing bank value and risk somewhat more difficult than assessing non-bank value and risk. Banks will pay for this relative opacity and consequent [*PG252]uncertainty since bond investors will demand a premium for it.183 Third, most of the instruments analyzed in the first set of empirical studies discussed above were short-term, and short-term investors do not care so much about borrower risk profile.184
Finally, and most importantly, the empirical studies conducted to date have suffered contamination from a bank regulatory environment that provides an implicit federal government guarantee to bank investors. If bank debtholders believe that the federal government will protect them from loss if their bank fails, and issuing banks understand this, debtholders will be unable to exact a risk premium from volatile banks. Although this implicit government guarantee is thought to protect uninsured creditors of only the biggest banks, it is precisely those banks that issue most of the debt that the empirical studies have analyzed. While the implicit guarantee has weakened over the past decade,185 most of the studies were performed when the guarantee was in full effect.
One method for improving the capital markets’ ability to assess bank risk levels is to improve the quality and increase the amount of public disclosures that banks are required to make. Banks are already subject to a host of disclosure requirements. Publicly held banks and their holding companies are required to make annual, quarterly and special event reports to the SEC.186 Banks also are required to submit year-end consolidated reports of income and quarterly consolidated reports of condition,187 and make special disclosures in areas of heightened supervisory concern.188 The Basle Committee on Bank Supervision has repeatedly issued guidance to bank supervisors relat[*PG253]ing to increasing the transparency of banking organizations.189 Federal bank regulators have made similar proposals.190
Banks already have some natural incentives to disclose information. For example, banks with a reputation for providing timely and accurate information to the public can access the capital markets more cheaply.191 Increased public disclosure about bank assets and operations will have the collateral benefit of reducing the potential for systemic disruptions of the financial landscape by improving the ability of market participants to distinguish strong banks from weak banks in troubled times.192
The objection: Market discipline can only be effective if market participants do not expect the government to compensate them for losses they may accrue in connection with their investments in banks. Federal bank regulators historically have been reluctant to close large failed banks and impose losses on creditors and uninsured depositors because of a fear that such substantial failures might destabilize the banking system.193
[*PG254] The response: The implicit federal guarantee of the investments of creditors and uninsured depositors of large banking organizations—the “too-big-to-fail” doctrine—has historically placed a significant obstacle in the way of effective market discipline. FDICIA, however, requires the FDIC to resolve failed banks in the manner that imposes the lowest cost on the deposit insurance fund.194 Moreover, since January 1, 1995, the FDIC has been forbidden to take any action “that would have the effect of increasing losses to any insurance fund by protecting . . . depositors for more than the insured portion of deposits [or] creditors other than depositors.”195 While these provisions of FDICIA may theoretically restrict the scope of the too-big-to-fail doctrine, it is not clear how restrictive they will be in practice.196 In order to implement a successful market discipline approach to bank regulation, the federal government must credibly commit not to insure the losses of the relevant market participants.197 The recent empirical [*PG255]studies cited in the preceding section suggest the government has made a reasonably credible commitment.198 We believe that federal regulators should not take any actions that would weaken the plausibility of this pledge.
The objection: For market discipline to be successful, bank managers must alter their investment and activity strategies as a result of market influence. Bank managers, however, are not sensitive to declines in the market value of the bank’s securities because their primary source of funding is deposits (mostly insured deposits). Moreover, to the extent that a bank can finance its operations by insured deposits and internally generated cash flow, it will not need to access the capital markets and will not pay a penalty for declining stock or bond prices.199
The response: As discussed above, market participants can discipline a bank in various ways.200 First, in addition to selling their investments in the bank and thereby signaling to other investors and bank managers that the bank is behaving badly, subordinated debtholders also will be able to control bank manager behavior through negotiated covenants in the debt indenture. Second, our proposal will require Subject Banks to access the capital markets once every two years, notwithstanding the amount of deposits or internally generated cash flow available to the bank. Third, banks, especially Tier I banks, are relying less on insured deposits to fund their operations and more on capital market instruments.201 Fourth, even in the absence of a [*PG256]need or desire to access the capital markets, bank managers will strive to avoid falling share prices or bond prices because such falling prices may reduce the manager’s compensation, cause shareholders to replace managers for permitting the share prices to fall, or trigger a takeover.202
The objection: Bank managers are naturally risk averse. Failure of their bank will result in a significant diminishment of the value of their human capital; consequently, bank managers will make conservative investments and engage only in low-risk activities. At the same time, investors generally diversify their investment portfolios to reduce firm-specific risk and hence maximize the risk-adjusted rate of return of their investments.203 Diversified investors have a greater tolerance for risk than does the federal government, since losses in one investment in the portfolio likely will be offset by gains in other investments in the portfolio. Hence, making bank managers more sensitive to the desires of their investors may lead to an increase in risk taking.204
The response: As a preliminary matter, bank managers typically are substantial equityholders or stock option holders in their employer. As such, bank managers are not especially risk adverse.
Investors, especially debt investors, should be expected to reduce bank risk profiles. While there can be no doubt that investors increasingly are becoming more diversified, and that diversified investors care less about firm-specific risk than non-diversified investors, investor diversification does not pose a problem. First, and most importantly, even diversified investors care about the risk profile of their individual investments. They care about the nature of the risks because, in constructing investment portfolios, investors need to know [*PG257]the amount of negative correlation between packages of investments205; they care about the quantity of risks because eliminating losses on any investment maximizes the value of the entire portfolio. Second, many investors are not diversified; the non-diversified investors, as well as the diversified investors with a large percentage of their portfolio invested in the subordinated debt of one issuer, will provide appropriate risk monitoring. Third, debt investors are different from stockholders. Because their upside gains are limited, debt investors have a direct risk aversion for each company in which they invest, despite diversification. Even diversified debtholders make efforts to ensure that their portfolio companies limit their risks.
The objection: Holders of heavily subordinated debt (like the debtholders in our proposal) will have equity-like risk preferences, especially as a bank approaches insolvency and the value of the bank’s equity approaches zero.206
The response: Admittedly, subordinated debtholders will prefer riskier bank activities and projects as a bank approaches insolvency. For two reasons, however, this fact does not threaten the viability of the proposal presented in this Article. First, as discussed above, subordinated debtholders are appropriately risk averse in contexts prior to impending insolvency,207 and most of the decisions that impel banks into insolvency are taken prior to the appearance of insolvency on the horizon. Moreover, even in the face of an impending insolvency, subordinated debtholders are more risk averse than equityholders: equityholders will enjoy all of the benefits of the success of a risky project, but debtholders will enjoy benefits only up to the return of their principal and accrued interest. Debtholders also are more likely than equityholders to receive their invested principal back in an insolvency proceeding.
The objection: Capital markets investors are fickle and too short-term oriented for the good of banks.208
[*PG258] The response: While this argument with respect to shareholders is off the mark,209 it is true that subordinated debtholders under our proposal will have their principal returned as soon as six years after the initial investment. While six years is something less than an eternity (the time horizon of bank regulators), it is a substantial period of time. The slippage resulting from the difference in time horizons should be more than offset by the gains accruing to the banking system from increasing the amount of bank monitoring by sophisticated fixed claimants and forcing banks to go into the capital markets every two years.
The objection: Holders of securities of a bank with an escalating risk profile, especially holders of debt securities whose disciplining tools are crude and generally negative in nature, will find it cheaper to sell the securities and sever their connection to the bank than to maintain their investment and attempt to alter the bank’s behavior.210
The response: This may be true with respect to holders of subordinated debt securities in Tier I banks. These debt securities will be bought and sold in a liquid, public market. For most of these investors, the cheapest method of expressing their thoughts about the prospects of a bank may be to sell the debt. Such selling pressure, however, will reduce the market price of the bank’s debt and signal to the market, to bank management and to bank regulators that the market frowns upon the bank’s increasing risk profile. Because our proposal requires banks to enter the debt capital markets every two years, such selling pressure and the resultant declining market price also will increase the future cost of funding to the bank.
Holders of subordinated debt in Tier II banks, however, may not enjoy a liquid market in which to sell their securities. Moreover, the federal securities laws will prevent them from selling any privately issued debt for several years.211 For these investors, it is unlikely that exit will be seen as a primary investment management strategy.
The objection: Monitoring by market participants is costly. Holders of bank securities will be too diffuse to exercise control over the bank; collective action problems will prevent market participants from providing effective discipline over bank behavior.212
The response: To the extent that a set of market participants determines that collective action problems—such as high information costs and free riders—pose a threat to the quality of their investments, nothing would prevent them from appointing one of their own as a compensated monitoring agent. Institutional investors routinely appoint an indenture trustee to represent their interests with respect to a bond issuance.213 Bank lenders also routinely appoint an agent bank to represent their interests with respect to syndicated loans.214 In addition, to the extent that market participants believe that their investments will suffer because collective action problems will interfere with their monitoring ability, market participants will pass along those costs to borrower banks. As a consequence, banks will likely assist market participants in establishing some sort of agent to represent their interests.215 Moreover, disciplining through direct monitoring is only one method of disciplining. The requirement that a bank return to the capital markets periodically and the ability of market participants to withdraw their investment and reduce the share or bond price of the bank are also effective mechanisms. No collective action problem will interfere with these disciplinary mechanisms.
Holders of subordinated debt, who can make intelligent investment decisions ex ante, can solve the collective action problem related to ex post monitoring by investing in firms that must continually return to the capital markets. A bank under our proposal will not attempt to increase its riskiness at current debtholders’ expense because it knows it will have to return to the market to issue additional debt in the near future.
The objection: Market discipline will not be effective because market participants are unable to take actions to align a bank’s incentives [*PG260]with their own. Bondholders are unable to constrain company behavior through their contractual covenants.216 The typical covenants negotiated by debtholders (asset maintenance and prohibitions on incurring senior debt) are unduly restrictive for a bank and would be ineffective in controlling bank risk.217 Banks, because of the very nature of the business of banking, would insist on preserving their flexibility to substitute assets without substantial restriction. Banks can increase their risk profile very easily by substituting assets.
The response: Admittedly, bank debtholders would probably not negotiate strict asset sale prohibitions or strict prohibitions on the incurrence of senior debt. Instead, they would likely negotiate minimum financial ratios, activity restrictions, and limits on the bank’s ability to incur additional senior debt. Bank debtholders would likely insist on a covenant that prevents the bank from entering “critically undercapitalized” status, since FDICIA prevents such banks from making any payments on their subordinated debt.218 Furthermore, bank regulators will be able to learn a great deal about preventing the insolvency of financial institutions by observing the kinds of covenants that bank debtholders employ in their bond indentures and reviewing over time the efficacy of the various covenants in controlling bank risk.
Public stockholders and bondholders also can punish bad bank behavior by selling their shares or their debt securities. This selling pressure will lower the price of the bank’s stock or debt, as the case may be, and will signal to the market that bank managers are taking actions inconsistent with the best interests of bank shareholders or debtholders. Bank managers will have trouble retaining their jobs as the value of the bank’s public securities falls. Moreover, if the bank must raise funds in the capital markets on a consistent basis, it will find that excessively risky behavior and the resultant drop in securities prices will make such financing substantially more expensive.
The objection: Banks find raising debt and equity in the capital markets to be very expensive. Banks do not issue much subordinated debt today because it would be an expensive form of financing and because there is not sufficient market appetite for the subordinated debt of banks.219 Moreover, the market for existing bank subordinated debt is a dealer market220; regulators will have a difficult time acquiring reliable price information on the debt.
The response: The market for the subordinated debt of banking organizations with greater than $50 billion in assets is actually highly liquid.221 The institutional investors that dominate the demand side of the market and the bank subordinated debt dealers agree that the secondary market prices for these securities are very efficient. Although some market participants have admitted that publicly available bank debt prices are hard to locate, expectations are that such price information will become more public in the near future.222
A handful of Web sites have recently sprung up to provide bond price information on a wide variety of municipal and corporate debt issues.223 Moreover, dozens of new internet companies are scrambling to establish on-line bond exchanges. While their efforts have not displaced the bond trading desks of the established investment banks, which make sizable commissions from bond trading, the SEC has made it a priority to illuminate debt security prices.224 The liquid market for large banking organization debt securities should only grow more transparent in the coming years.
Admittedly, the market for smaller bank subordinated debt is currently thin. A significant cause of the lack of volume in this market, however, is the Capital Guidelines adopted by the bank regulatory agencies. Banks do not issue much subordinated debt because such debt does not count as tier 1 capital and does not fully count as tier 2 capital under the Capital Guidelines.225 Our proposal would simply [*PG262]counteract this regulatory obstacle. Some banks may also resist issuing nondeposit debt because they do not want to be bound by the covenants a nondeposit creditor would impose. It is precisely these sorts of covenants that can impose additional discipline on a bank.
The objection: Market discipline will have perverse systemic effects—the relevant private investors will exit a bank quickly as adversity approaches, hastening a depletion of the bank’s capital and accelerating the bank’s approaching insolvency.226
The response: Exit by subordinated debtholders can only be accomplished through sale of the securities to another purchaser. The debt instrument proposed in this Article is not puttable. The debt securities will remain outstanding until maturity, at least six years after issuance. Hence, exit by subordinated debtholders does not create the type of systemic risk of a run on the bank that exit by depositors creates. While our proposal does permit one-third of the subordinated debt capital to leave the bank every two years, this sort of a staggered long-term departure of capital should not pose anything like the systemic threat of a deposit run. Indeed, the inability of a bank to roll over the required amount of subordinated debt would provide a valuable signal to regulators that the bank needs attention.
Admittedly, however, there is some risk that the price and other signals transmitted by subordinated debtholders will serve not only as risk indicators for bank regulators but also as a signal to bank depositors and other creditors. Those depositors—especially large, uninsured depositors—may then “run” when subordinated debtholders indicate concerns about the risk profile of a particular institution. Although the likelihood of such runs cannot be completely dismissed, we think that the risk of uninsured depositor runs precipitated by subordinated debt signals will not be severe. As noted in Part III.B. above, many uninsured depositors are not sophisticated investors, and they will not pay careful attention to the informational conveyances of subordinated debtholders. Accordingly, we expect that federal regulators will react with greater alacrity to subordinated debt signals than depositors and that regulators (and subordinated debtholders) will take measures to counteract risky bank activities before the situation becomes so severe as to cause a fatal depositor run.
The objection: Market discipline may interfere with a bank’s capacity to assist federal regulators in implementing macroeconomic policy.227 Federal exhortations to lend or provide liquidity to the economy may run counter to the interests of debtholders.
The response: The Federal Reserve’s most powerful weapons for implementing macroeconomic policies—buying and selling government securities, altering reserve requirements for banks, and changing the discount and federal funds borrowing rates228—will not be affected by the existence of additional subordinated debt in the capital structure of banks; these methods merely make quantitative adjustments to the interest rate environment in which banks operate. When the Federal Reserve and other federal regulators attempt to effect macroeconomic policies through more informal means (for example, urging banks to raise or lower their underwriting criteria), a bank’s profit motives may run counter to the federal regulatory nudge. This conflict of interest between banks and regulators is not made worse, however, by the presence of subordinated debt investors. Indeed, where the informal nudge from regulators is a recommendation of stricter loan underwriting standards,229 debtholders will be more likely to support this goal than the equityholders whom such debtholders partially replace in this proposal.
Requiring large banks to issue subordinated debt promises to remedy many of the shortcomings of government supervision and regulation. Actual and prospective holders of bank subordinated debt will constrain bank risk taking roughly in accordance with the wishes of the federal government and without the bureaucratic and other inefficiencies entailed in governmental regulation. Holders of bank subordinated debt, as they buy and sell bank debt securities in the secondary market and negotiate purchases in the primary market, will also signal to federal regulators the private sector’s view as to the value of a bank’s enterprise. While such a market discipline approach [*PG264]should not supplant government regulation, supervision and examination of banks, it can and should serve as an effective complement to government oversight of financial institutions.