Main Content

The Resurgence of Debt as Capital

Debt, scorned as a part of the capital structure of a banking organization by the Collins Amendment, has the potential to be cost efficient, absorb losses, and foster market discipline. It is staging a comeback both in the United States and abroad.

By Oliver I. Ireland, Anna Pinedo, and Jeremy Jennings-Mares

The Clearing HouseBefore the financial crisis, the Federal Reserve Board (FRB) had explored the use of hybrid capital instruments at the bank holding company (BHC) level that combined the cost efficiency (including favorable tax treatment) of debt with loss-absorbing capacity. These instruments inspired little investor confidence as an indicator of a bank’s condition or “going concern capital” during the crisis and were precluded from use in meeting existing capital requirements by the Collins Amendment in section 171 of the Dodd-Frank Act.

Nevertheless, in 2015, the Financial Stability Board (FSB) adopted its Final Principles on Loss-absorbing and Recapitalization of G-SIBs in Resolution, which established a total loss-absorbing capacity, or TLAC, requirement in excess of otherwise applicable capital requirements. The additional loss-absorbing capacity requirement, over and above otherwise applicable capital requirements, can be met with debt instruments, excluding deposits, secured debt, derivatives, and debt instruments with derivative-like features. In implementing these principles in the United States, the FRB went a step further by not only allowing debt instruments that meet specified terms to count toward TLAC, but also requiring that a specified portion of TLAC be met by eligible debt securities. Under the FSB Principles and the FRB’s implementation of those principles, TLAC may absorb losses directly or may be converted into equity in a new banking organization created to replace a failed global systemically important bank (G-SIB).

The idea of requiring banking organizations to issue debt instruments did not originate with TLAC. The Feasibility and Desirability of Mandatory Subordinated Debt was the subject of a report to Congress in December 2000, required pursuant to Section 108 of the Gramm-Leach-Bliley Act (Note: Under the FRB’s implementation of TLAC, TLAC debt may or may not be subordinated to other unsecured liabilities). One purpose of this report was to assess whether such a requirement would bring market forces and market discipline to bear on the operation of, and the assessment of the viability of, large banking organizations, the failure of which could have adverse effects on economic conditions or financial stability. Although the report assessed a number of potential objectives for issuing subordinated debt, the first objective was to determine whether the subordinated debt issuance costs would provide an incentive for the issuing entity to refrain from taking excessive risks because market perceptions of risk would be reflected in the price of the issuer’s debt securities. Another objective was to increase the size of the issuer’s financial cushion, or loss-absorbing capacity, to facilitate resolution.

Debt issued to meet the FSB’s requirements implemented into national law by foreign jurisdictions for G-SIBs and the FRB’s TLAC requirements for U.S. domestic banking organizations that are G-SIBs and intermediate holding companies of foreign G-SIBs will provide loss-absorbing capacity – that is, a cushion – for resolution purposes. However, the debt will be subject to differences in the triggers that would result in the debt being used to absorb losses. These differences do not impair the debt’s ability to absorb losses under the respective resolution schemes or to provide that loss-absorbing capacity more economically than might be the case for other capital instruments. However, the differences do limit the utility of using market rates for TLAC debt as a tool to assess market perceptions of the condition of individual banking organizations issuing the debt.

EU Resolution Scheme and the Role of Loss Absorption
Under the EU’s Bank Recovery and Resolution Directive (BRRD), which applies to all EU banks, not just G-SIBs, all liabilities of the bank in resolution can be written down or converted to capital (“bailed in”), unless they are excluded from bail-in. Liabilities can be excluded either by virtue of being a type of liability that is expressly excluded under BRRD, or by virtue of being excluded as a result of the exercise of the resolution authority’s discretion. Expressly excluded liabilities include but are not limited to deposits subject to a guarantee scheme, secured liabilities, liabilities arising from a fiduciary relationship, liabilities to EU credit institutions or investment firms with an original maturity of less than seven days, and short-term liabilities owed to payment or securities settlement systems or their participants.

Resolution authorities are provided with discretion in exceptional circumstances to exclude, in whole or in part, certain liabilities from bail-in, in circumstances under which, for example, it is not possible to bail in that liability within a reasonable time, the exclusion is strictly necessary and proportionate to achieve continuity of critical functions and core business lines, or the exclusion is strictly necessary and proportionate to avoid causing widespread contagion that would severely disrupt the financial markets. To the extent the exercise of a discretionary exclusion results in the need to increase the extent of the bail-in applied to liabilities, such an increase is permitted, so long as it does not breach the principle that no creditor will be worse off than the creditor would have been in liquidation (referred to as the NCWOL Principle). If such a discretionary exclusion would require a contribution from the relevant member state’s resolution financing arrangement, the European Commission must be notified and the commission has the power to prohibit or require amendments to the proposed exclusion.

In a bank failure, either before or in conjunction with any resolution action (including the use of the bail-in tool), the resolution authority may write down or convert to equity the bank’s regulatory capital instruments. Such action is permitted if the bank is failing or likely to fail and there is no reasonable prospect that any action other than a write-down or conversion of regulatory capital would prevent failure. In this circumstance, Common Equity Tier 1 capital (CET1) is written down and in the case of shares, canceled or transferred or, if the resolution valuation demonstrates that the bank has a net positive value, diluted by Additional Tier 1 capital (AT1) and/or Tier 2 (T2) capital being converted into common shares. Once bail-in action is taken, either after or in conjunction with the write-down or conversion of capital instruments, any CET1 capital (including common shares issued pursuant to a write-down or conversion) will first be written down, followed by any remaining AT1 capital by the extent necessary, and then followed by any remaining T2 capital as needed. If further bail-in is required, this must be satisfied first by writing down non-T2 subordinated debt as needed, and thereafter, other liabilities eligible for bail-in, provided that any contingent convertible bonds or other instruments containing contractual loss-absorption language have absorbed losses. In bailing in unsecured, unsubordinated eligible liabilities, the creditor insolvency hierarchy that would apply on a normal winding-up must be respected.

As discussed further below, a bail-in action can only insulate the public sector from the cost of a bank failure to the extent that the bank has sufficient total loss-absorbing capacity. Generally, under the FSB Principles, which are implemented into national legislation in the EU, G-SIBs are subject to a TLAC requirement that equals the sum of a G-SIB’s resolution entity’s (1) T1 and T2 regulatory capital instruments that are in the form of debt, plus (2) other eligible TLAC that is not regulatory capital, and is equal to or greater than 33% of the G-SIB’s minimum TLAC requirement. Regulatory capital instruments may count toward the minimum TLAC requirement, subject to certain conditions. Eligible TLAC must contain a contractual trigger or be subject to a statutory mechanism that permits the resolution authority to write down or convert it to equity. There is no separate eligible long-term debt requirement as part of the FSB Principles – although, as we mention below, under the BRRD, each EU bank (not just G-SIBs) must hold a certain amount of MREL (minimum requirement for own funds and eligible liabilities), as prescribed on a bank-by-bank basis by its resolution authority.

U.S. Approach to Resolution and the Role of Loss Absorption
In December 2013, the FDIC requested comment on the single point of entry (SPOE) strategy that is based, at least in part, on its experience in dealing with failed banks. The proposed U.S. SPOE approach would require that only the BHC of the failed G-SIB be placed in receivership. Title II of the Dodd-Frank Act requires that the BHC be liquidated with losses imposed on the stockholders and creditors of the BHC. The stockholders of the BHC would bear the first losses, and the claims of holders of the BHC’s long-term debt obligations would be converted into equity that would be used to capitalize the successor entity.

This approach assumes that the BHC truly functions, in the FRB’s terms, as “a clean holding company” in which business is conducted by the entity through its operating subsidiaries. The Federal Deposit Insurance Corporation generally is expected to treat creditors of the failed entity within the same class and priority of claim similarly. However, the FDIC can favor creditors, such as vendors, that provide critical services. Title II of the Dodd-Frank Act sets out the following priority for claims: administrative costs of the receiver, amounts owed to the government; employee salaries and benefits; general or senior unsecured liabilities, subordinated debt obligations, salaries for directors and officers, and shareholder claims. To date, a statutory mechanism for bail-in or conversion to equity has not been adopted, so there are no detailed mechanisms or principles like those set out in the BRRD.

The SPOE strategy contemplates that the assets of the failed holding company, including ownership interests in and intercompany loans to the holding company’s operating subsidiaries, and these operating subsidiaries would be transferred to a bridge financial company established by the FDIC. The bridge financial company would be capitalized first by the bail-in of outstanding long-term debt of the failed BHC, which presumes that sufficient long-term unsecured debt would be outstanding at the holding company level. As noted above, in order to address this, the FRB’s final rules create a total of loss-absorbing debt and equity instruments at the top-tier holding company. This is the total TLAC number, which is the sum of T1 capital issued directly by the G-SIB BHC and the amount of eligible external long-term debt (LTD) together with the G-SIB surcharge and applicable regulatory capital buffers.

This distinct eligible LTD requirement is intended to facilitate the SPOE resolution approach. Given the need for simplicity and efficiency in a resolution, the eligible LTD requirement must be met by plain-vanilla debt, which is paid-in, issued directly by the covered BHC, is unsecured, is governed by U.S. law, and has a maturity of greater than one year from the issuance date. The bridge financial company, which would be the new holding company for the failed G-SIBs operating subsidiaries, would be capitalized by the bail-in of the eligible external LTD of the failed BHC.

Multiple Instruments, Differing Terms
The FSB has stated that it expects at least 33% of a G-SIB’s minimum TLAC requirement to be met in the form of debt regulatory capital instruments (because, for EU banks, certain instruments that are debt securities for accounting purposes still qualify as regulatory capital) and other TLAC-eligible liabilities that are not regulatory capital. For example, for European G-SIBs, contingent capital instruments that qualify as additional Tier 1 and Tier 2 capital can satisfy this category of TLAC if the contingent capital instruments are subordinated to all those liabilities that the FSB expressly excludes from TLAC eligibility. Generally, in the case of a contingent capital instrument, the conversion to equity or the write-down of principal occurs before the failing financial institution enters resolution, rather than in a bail-in on failure of the institution. EU banks, not just G-SIBs, are also mandated by BRRD to hold a certain amount of MREL. The amount of MREL will consist of a loss-absorption amount (which as a baseline default will equal its prudential capital requirements, including Pillar 2 and the combined buffers) plus an amount required to recapitalize the bank in order to ensure both that it meets the conditions for authorization (its licensing requirement), with a minimum of 8% of its total risk exposure amount, and that it can ensure market confidence following resolution.

Like the FSB’s TLAC requirement, the MREL obligations can be satisfied with contingent capital instruments. The eligibility requirements for a liability to count toward MREL are similar to the eligibility requirements for TLAC. However, MREL requirements are set on a bank-by-bank basis, and are set as a percentage of the bank’s non-risk-weighted assets. Currently, there are a number of differences in the features of liabilities that are excluded from MREL eligibility, compared to those excluded from TLAC eligibility, as well as differences relating to the subordination of MREL (to liabilities that are excluded from MREL eligibility). However, it is expected that once certain harmonizing legislative amendments are made, there should be no substantive differences between TLAC and MREL for EU G-SIBs.

Where subordination of MREL is required, subordination may be achieved by contractual language, by statutory subordination, or by structural subordination. As bailing in some classes of unsubordinated senior liabilities but not others increases the risk of breaching the NCWOL Principle, it became clear that there is a need for banks to satisfy a significant portion of their MREL requirements with liabilities that rank junior in insolvency to ordinary senior unsecured liabilities but senior in insolvency to other subordinated debt liabilities (so-called “senior non-preferred” or “Tier 3” liabilities).

Several individual member states (such as France and Germany) have already amended their national insolvency laws to allow for just such a new class of liabilities, albeit with different features and with different results for TLAC eligibility purposes. The United Kingdom had no need to introduce legislation, because U.K. banks rely on structural subordination rather than statutory subordination. In an effort to ensure that banks in all member states would have at least one TLAC-eligible option for senior nonpreferred debt, the European Commission introduced a new directive, the Insolvency Hierarchy Directive, which should come into force in the EU in early 2018. This new directive will require member states to amend the creditor hierarchy in their national insolvency laws to allow for a new class of senior nonpreferred debt, which must have an initial maturity of at least one year, must have no derivative features, and whose contractual documentation must specifically refer to its ranking in insolvency.

Additional loss-absorbing capacity can be achieved through a variety of instruments with varying terms, and cost efficiency can likewise be achieved through different instruments, including instruments having different triggers that may result in that debt being used to absorb losses ahead of other liabilities or triggers that result in conversion into equity. However, the use of such instruments as tools to compare the market’s assessment of the issuer’s credit risk prior to a resolution event is limited.

As set forth above, the post-crisis regulatory framework now includes, in the case of European banks, debt instruments that may count for regulatory capital purposes. These may include contingent capital instruments, the principal of which may be written down or which may convert to equity before the bank issuer’s failure or resolution. Likewise, these also include debt securities that convert into equity at the point of nonviability, and certain of these instruments, depending on the jurisdiction, may allow for regulatory discretion with regard to the non-viability assessment. Even in the case of a “plain-vanilla” debt security that for U.S. purposes would constitute eligible LTD, a holder might be subject to a different creditor hierarchy and different outcome in a resolution. The differences in approaches to resolution across EU jurisdictions, when added to the differences in the features of TLAC instruments, are likely to preclude the development of risk-sensitive pricing differences that would assist supervisors in assessing the credit quality of the issuers in anything but the most extreme circumstances. The lack of comparability in EU jurisdictions is compounded by differences between the EU requirements and the U.S. requirements, including that the U.S. bail-in process, which is articulated in policy, is not built into law as is bail-in under the BRRD.

We believe that for the foreseeable future this result is inevitable given that each jurisdiction will adopt its own regulations to implement Basel and other internationally agreed-upon standards. It is also probably inevitable to the extent that supervisors retain discretion in triggering the resolution process. Certainly, it is unlikely that pricing will accurately reflect risk when investors cannot be expected to understand the differences among jurisdictions, differences among eligibility criteria and requirements, etc. Nevertheless, given the increases in T1 capital, which is viewed by investors as an indicator of the viability of a banking organization as a going concern, the differences in the terms and triggers for debt that provides additional loss-absorbing capacity in the event of a failure and the lack of comparability should not impair the role of T1 capital and should help supervisors to limit the market disruption resulting from the failure of a banking organization.