Let's End "Too Big to Fail"

Tuesday, October 20, 2015
Image credit: 
Third Way Think Tank

Editor’s note: This essay is an excerpt of the Hoover Press book, Making Failure Feasible: How Bankruptcy Reform Can End Too Big to Fail.

Any process for resolving the affairs of failed financial institutions other than banks, whether under Title II of the Dodd-Frank Act of 2010 or the Resolution Project’s proposed new version of a Chapter 14 of the Bankruptcy Code, takes as its starting point a firm whose organizational form and financial structure have been determined by a complex set of statutory and regulatory requirements. At this writing, many of those requirements are still being developed, important aspects are uncertain, and terminology is not set.

A note on terminology: the phrase “systemically important financial institution” or SIFI is nowhere defined (or even used) in the Dodd-Frank Act, though it has come into common parlance. I will use it here to refer to those financial companies whose distress or failure could qualify for seizure under Title II and Federal Deposit Insurance Corp. (FDIC) receivership, as threatening serious adverse effects on US financial stability. Presumably they come from bank holding companies with more than $50 billion in consolidated assets and nonbank financial companies that have been designated for supervision by the Federal Reserve Board.

Revised Chapter 14 2.0, at places, makes assumptions about pending requirements’ final form, and may have to be modified in the light of what is settled on. It also contains recommended changes in the application of stays to QFCs (qualified financial contracts).

The Resolution Project’s original proposal (Chapter 14 1.0) contemplated resolving a troubled financial institution through reorganization of the firm in a manner similar to a familiar Chapter 11 proceeding, with a number of specialized adjustments. Subsequently, the FDIC has proposed that the failure of those large US financial institutions (mostly bank holding company groups) that are thought to be systemically important (SIFIs) and not satisfactorily resolvable under current bankruptcy law will be handled by (1) placing the parent holding company under the control of the FDIC as a Title II receiver and (2) transferring to a new “bridge” financial company most of its assets and secured liabilities (and some vendor claims)—but not most of its unsecured debt. Exactly what is to be left behind is not yet defined, but will be here referred to as bail-in debt (BID) or capital debt. (Any convertible debt instruments—CoCos—that the firm may have issued are required to have been already converted to ­equity.) The losses that created a fear of insolvency might have occurred anywhere in the debtor’s corporate structure, but the takeover would be of the parent company—a tactic described as a “single point of entry” (SPOE).

The desired result would be a new financial company that was strongly capitalized (having shed a large amount of its prior debt), would have the capacity to recapitalize (where necessary) operating subsidiaries, and would have the confidence of other market participants, and therefore be able to immediately continue its critical operations in the financial system without any systemic spillover effects or problems. But all of that depends on a number of preconditions and assumptions about matters such as: the size and locus of the losses, the amount and terms of capital debt and where it is held, the availability of short-term (liquidity) debt to manage the daily flow of transactions, and agreement on priorities and dependable cooperation among regulators in different countries where the firm and its subsidiaries operate—to name some of the most salient.

If the failed financial institution is not deemed to present a threat to US financial stability, even though large, it is not covered by Title II but would come under the Bankruptcy Code. Chapter 14 2.0 is our proposal for a bankruptcy proceeding that is especially designed for financial institutions and includes provisions for the use of SPOE bridge transfers where desired, and it too will be affected by the regulatory regime in force—especially as it relates to BID.

Not all of these matters are, or can be, determined by Dodd-Frank or in the Bankruptcy Code. But they can be affected for better or worse by regulations still being proposed or adopted. This paper represents my attempt, for readers not unfamiliar with these topics, to highlight some of the problems and Chapter 14’s responses, and to recommend some other measures that would facilitate successful resolutions.

Capital Debt

1) In FDIC’s proposal, the debt that is not to be transferred (and thus fully paid) is not precisely specified. It is suggested that accounts payable to “essential” vendors would go over, and “likely” secured claims as well (at least as deemed necessary to avoid systemic risk), but not (all?) unsecured debt for borrowed funds. Unless ultimately much better specified, this would leave a high degree of uncertainty for creditors of financial institutions, with corresponding costs.

There are some specifics that have been suggested—for example, that capital debt be limited to unsecured debt for borrowed money with an original (or perhaps remaining) maturity of over a year. That would imply a regulatory requirement that a SIFI hold at all times a prescribed minimum amount of such debt—at a level yet to be determined but perhaps equal to its applicable regulatory capital requirements and buffers, giving a total loss absorbing capacity (TLAC) of as much as 20 percent to 25 percent of risk-weighted assets

Would that total amount be sufficient to cover all losses the firm might encounter, and enough more to leave it still well capitalized? That depends on the magnitude of the losses it has incurred. In effect, the debt requirement becomes a new ingredient of required total capital (beyond ­equity), and impaired total capital could trigger resolution (but not necessarily continuance of operations, unless a grace period of a year or more for restoration of the mandated TLAC were included). The operative constraint is the mandated total amount of regulatory capital plus BID; the exact split between the two is less significant, and could be a matter for management judgment. Until such requirements are actually specified and instituted, however, their effectiveness is hard to analyze.

The definition of bail-in debt continues to be controverted. Is it a species of unsecured bonds for borrowed money, with specified staggered maturities? Is it all unsecured liabilities, with an extensive list of exceptions? Whatever the category, does it apply retroactively to existing liabilities? Will investors realize their risk status? Should disclosure requirements be spelled out? (It is hard to see why it is not defined simply as newly issued subordinated debt, without any cumbersome apparatus for conversions or ­write-downs or loss of a priority rank.)

2) A capital debt requirement would function the same way in Chapter 14, but without discretionary uncertainty. Section 1405 provides for the transfer to a bridge company of all the debtor’s assets (which should include NOL [net operating loss] ­carry-forwards) and liabilities (except for the capital debt and any subordinated debt); in exchange, the debtor estate receives all of the stock in the new entity. And the external capital debt is given a clear definition: it must be designated unsecured debt for borrowed money with an original maturity of one year or more. To be effective, minimum capital debt requirements (an issue outside of bankruptcy law) would again need to be specified.

It should be noted that Chapter 14 applies to all financial companies, not just SIFIs that pose systemic risk and not just to resolution through a bridge. The firm may go through a familiar Chapter 11 type of reorganization, following on a filing by either management or supervisor after losses have impaired compliance with whatever are the total capital plus BID (TLAC) requirements then in force. In that case, the BID is not “left behind” but should all automatically (under the provisions of its indenture) either be written down or converted to a new class of senior common stock, or to preferred stock or subordinated debt with similar terms. (If conversion were to a security on a parity with outstanding common stock, there would be immediate time-consuming and disputable issues about how to determine asset valuations and losses and the possible value of existing common shares. These are avoided by simply converting instead to a new class with a priority above outstanding common and below ordinary liabilities.)

3) What is the locus of the capital debt? The question is central to whether subsidiaries necessarily continue in operation. The FDIC proposal seems to contemplate that it is issued by a parent holding company (or, in the case of a foreign parent, its intermediate US holding company), and thus removed from the capital structure of the new bridge company, which is thereby rendered solvent.

But what if the large losses precipitating failure of the US parent were incurred at a foreign subsidiary? There have been suggestions that the new bridge parent would be so strongly capitalized that it could recapitalize the failed subsidiary—but who makes that decision, and on what basis? The supervisory authorities of foreign host countries have understandably shown a keen interest in the answer, and it is high on the agendas of various international talks.

A core attribute of separate legal entities is their separation of risk and liability. Under corporation law, the decision to pay off a subsidiary’s creditors would be a business judgment for the parent board, taking into account financial cost, reputational cost, future prospects, and the like—and the decision could be negative. In a Title II proceeding, perhaps the FDIC, through its control of the board, would override (or dictate) that decision—and perhaps not.

The clearest legal ways to try to ensure payment of subsidiary creditors would be (1) to require parents to guarantee all subsidiary debt (which amounts to a de facto consolidation) or (2) to have separate and hopefully adequate “internal” capital debt (presumably to the parent) requirements for all material subsidiaries. Again, at time of writing it is an issue still to be resolved, and perhaps better left to the host regulators and the firm’s business judgment in the specific circumstances.


1) The FDIC’s SPOE bridge proposal seemingly applies only to domestic financial companies posing systemic risk (currently, eight bank and three or four non-bank holding companies are so regarded, although more may be added, even at the last minute), not to the next hundred or so bank holding companies with more than $10 billion in consolidated assets, or to all the (potentially over one thousand) “financial companies” covered by Dodd-Frank’s Title I definition (at least 85 percent of assets or revenues from financial activities). Will the capital debt requirement be limited to those dozen SIFIs, or will it be extended to all bank holding companies with more than $250 billion or even $50 billion in consolidated assets (though posing no threat to US financial stability)? That will determine how failure resolutions may be conducted under the Bankruptcy Code, as they must be for all but that small number of SIFIs that Title II covers.

2) Resolution under Chapter 14 (in its original version) can take the form essentially of a familiar Chapter 11 reorganization of the debtor firm (often at an operating entity level). Where systemic risk or other considerations dictate no interruptions of business operations, it may (in its current version 2.0) take the form of transfers to a new bridge company (usually at the holding company level—thus leaving operating subsidiaries out of bankruptcy). Therefore, any capital debt requirement should apply explicitly to both situations, and Chapter 14 would accommodate both options.

3) What triggers the operation of the capital debt mechanism? A filing of a petition under Chapter 14, for which there are two possibilities. The management of a firm facing significant deterioration in its financial position can choose to make a voluntary filing, to preserve operations (and perhaps their jobs) and hopefully some shareholder value, as often occurs in ordinary Chapter 11 proceedings. Depending on circumstances, this could take the form of a ­single-firm reorganization or a transfer of assets and other liabilities to a new bridge company in exchange for its stock.

The second possibility is a filing by the institution’s supervisor, which could be predicated on a determination (1) that it is necessary to avoid serious adverse effects on US financial stability (as our proposal now specifies) or (2), more broadly, that there has been a substantial impairment of required regulatory capital or TLAC. There can be differing views on how much regulatory discretion is advisable, so this too is to some extent an open issue. But the ability of the supervisor to force a recapitalization short of insolvency might alleviate concern that institutions that are “too big to fail” must be broken up or they will inevitably receive government bailouts.


Banks perform vital roles in intermediating transactions between investors and businesses, buying and selling risk, and operating the payments system. They have to manage fluctuating flows of cash in and out, by ­short-term borrowing and lending to each other and with financial firms. Bank failures often occur when creditors and counterparties have lost confidence and demand full (or more) and readily marketable collateral before supplying any funds. Even if over time a bank’s assets could cover its liabilities, it has to have sufficient immediate cash or it cannot continue in business. For that reason, the Basel Committee and others have adopted, and are in the process of implementing, regulations governing “buffer” liquidity coverage ratios that global systemically important banks (G-SIBs) would be required to maintain.

FDIC’s SPOE Proposal

The new bridge company is intended to be so well-capitalized, in the sense of book net worth, that it will have no difficulty in raising any needed funds from other institutions in the private market. But this is an institution that, despite all the Title I regulations, has just failed. There may be limited cash on hand and substantial uncertainty (or controversy) about the value of its loans and investments. So if liquidity is not forthcoming in the private market, Dodd-Frank creates an Orderly Liquidation Fund (OLF) in the Treasury, which the FDIC as receiver can tap for loans or guarantees (to be repaid later by the bridge company or industry assessments) to assure the necessary cash. Critics fear that this will open a door for selected creditor bailouts or ultimate taxpayer costs.

Chapter 14

As with the FDIC proposal, under favorable conditions there may be no problem. But what if cash is low or collateral value uncertain, and there is a problem? It depends on which type of resolution is being pursued.

In a standard Chapter 11 type of reorganization, the debtor firm can typically obtain new (“debtor in possession” or DIP) financing because the lenders are given top (“administrative expense”) priority in payment; those provisions remain in effect under Chapter 14. In a bridge resolution, the new company is not in bankruptcy, so the existing Bankruptcy Code priority provision would not apply. Therefore, Chapter 14 2.0 provides that new lenders to the bridge would receive similar priority if it were to fail within a year after the transfer.

In addition, a new financial institution could be given the same access to the Fed’s discount window as its competitors have. In a time of general financial crisis it could be eligible to participate in programs established by the Fed under its section13(3) authority. If all that is not enough assurance of liquidity in case of need, skeptics might support allowing (as a last resort) the supervisor of the failed institution (as either the petitioner or a party in the bankruptcy proceeding) the same access to the OLF as under Title II.

Qualified Financial Contracts

Even with a prompt “resolution weekend” ­equity recapitalization and measures to bolster liquidity, the first instinct of derivatives counterparties could well be to take advantage of their current exemption from bankruptcy’s automatic stay and exercise their contractual termination rights—which could have an abrupt and heavy impact on the firm’s ability to continue to conduct business.

Therefore, to simplify a bit, the proposed Chapter 14 amends the Bankruptcy Code to treat a counterparty’s derivatives as executory contracts and make them subject to a two-day stay, for the debtor to choose to accept or reject them as a group—provided the debtor continues to fulfill all its obligations. If they are accepted, they remain as part of the firm’s book of continuing business.

This would enact into governing US law some of what the International Swaps and Derivatives Association (ISDA) has sought to achieve in its 2014 Resolution Stay Protocol, to stay or override certain ­cross-default and ­close-out rights, through amending the master agreements of adhering parties (initially the eighteen largest dealer banks).

Due Process

Title II of Dodd-Frank Act

Section 202 of the Act prescribes a procedure to take over a SIFI posing systemic risks that the Secretary of the Treasury has determined to be in danger of default, with FDIC as receiver instructed to immediately proceed to liquidate it. The secretary’s determination, if not consented to, is filed in a petition in the District of Columbia federal district court to appoint the receiver. Unless in ­twenty-four hours the district court judge has held a hearing, received and considered any conflicting evidence on the financial condition of a huge firm, and either (1) made findings of fact and law, concluded that the determination was arbitrary and capricious, and written an opinion giving all the reasons for that conclusion, or (2) granted the petition, then (3) the petition is deemed granted by operation of law.

Obviously, the pre-seizure judicial hearing is an empty formality, and it is quite possible that most judges would prefer to simply let the ­twenty-four-hour clock run out. The company can appeal the outcome as arbitrary and capricious (although the record may be rather one-sided), but the court cannot stay the receiver’s actions to dismantle the firm (or transfer operations to a bridge), pending appeal. So in the unlikely event that there is a successful appeal, an adequate remedy would be hard to design. The whole procedure invites constitutional due process challenge.

Chapter 14

Most debtors are likely to go through a straightforward, one-firm reorganization, which entails claimant participation, public hearings, and well-defined rules, all presided over by an Article III (life tenure) judge. Criteria of due process and fundamental fairness are observed in a procedure developed over many years.

In the case of a SIFI going through the bridge route in order to promote continuity of essential services, the transfer motion is subjected to a somewhat more substantial hearing, in terms of both time and content. If the Fed is filing the motion, it has to certify (and make a statement of the reasons) that it has found (1) that a default by the firm would have serious adverse effects on US financial stability and (2) that the new bridge company can meet the transferred obligations. If the Treasury Secretary decides to assert authority to put the proceeding into Title II, he would be required in addition to certify and make a statement of the reasons for having found that those adverse effects could not adequately be addressed under the Bankruptcy Code (as amended by Chapter 14).

Nonetheless, the court would not be in a position, given the time constraints, to conduct a genuine adversary hearing and make an independent judgment. To overcome the serious due process shortcomings attached to the Title II section, Chapter 14 provides for an ex-post remedy under section 106 of the Bankruptcy Code: an explicit damage cause of action against the United States. And rather than the very narrow judicial oversight possible under the “arbitrary and capricious” standard of review (as in Title II), there is the standard of whether the relevant certifications are supported by “substantial evidence on the record as a whole.”

International Coordination

Most SIFIs are global firms (G-SIFIs), with branches and subsidiaries in many countries. To resolve them efficiently and equitably would require cooperation and similar approaches by regulators in both home and host nations. Optimally, that would mean a multilateral treaty among all the countries affected—a daunting undertaking that would take years at best. The Financial Stability Board, in its Key Attributes paper, has outlined a framework for procedures and cooperation agreements among resolution authorities, but they are in general not legally binding or enforceable in judicial proceedings.

The response of ISDA in its Resolution Stay Protocol was to seek a contractual solution in the master agreements, with the expectation that it would be enforced under the laws of six major jurisdictions. But since adherence is voluntary and coverage will be partial, there are gaps best filled by a statutory approach.

To make a modest legal beginning, a binding international agreement just between the United States and the United Kingdom would cover a large fraction of total transactions. The FDIC and Bank of England in a 2010 Memorandum of Understanding agreed to consult, cooperate, and exchange information relevant to the condition and possible resolution of financial service firms with ­cross-border operations. The Memorandum specifically, however, does not create any legally binding obligations.

A treaty, or binding executive agreement, could go further to determine how a resolution would proceed between the United States and United Kingdom as home or host countries. To get that process under way, the Resolution Project would provide in Chapter 15 (added to the code in 2005 to deal with ­cross-border insolvencies) new substantive provisions dealing with US enforcement of foreign home country stay orders and barring domestic ring-fencing actions against local assets, provided that the home country has adopted similar provisions for US proceedings. Unilateral action by the United States, conditioned on such a basis of reciprocal treatment, would be desirable on its merits and might contribute to much broader multilateral efforts.

The Problem of Systemic Risk

The special concern with the failure of a systemically important financial institution is based on the fear that it may lead to a collapse of the financial system which transfers savings, loans, and payments throughout the economy and is essential to its functioning. There are several different ways in which this might occur.

Knock--On Chains

In this scenario, a giant, “interconnected” financial firm incurs very large losses (from poor investment decisions, fraud, or bad luck) and defaults on its obligations, inflicting immediate losses on its counterparties, causing some of them to fail in turn. As a wave of failures spreads, the whole financial system contracts and so does the real economy.

Some observers attribute the panic of 2008 to losses caused by the failure of Lehman Brothers. That belief powered much of the Dodd-Frank Act and in particular its Title II mechanism for taking over a SIFI and putting it into a government receivership. It is not clear how a government receivership per se of a failed firm (without any bailout) is supposed to prevent direct spillover losses, other than that the process will be more “orderly” than was the case for Lehman. The fact that Lehman had done absolutely zero planning for a bankruptcy reorganization makes that a low standard, and the Dodd-Frank section 165 “living wills” requirement for firms to have resolution plans can’t help but be an improvement, however limited their “credibility” in an actual case may turn out to be. Their best practical use might be as rough preliminary drafts for “pre-packaged” bankruptcy petition filings.

In any event, Title II and FDIC’s SPOE proposal are all focused on a new procedure for handling the impending failure of an individual SIFI, and accordingly so is the Chapter 14 proposal for bankruptcy reform.

Common Shocks

In this scenario, a very widely held class of assets or investments turns out to perform unexpectedly poorly and becomes increasingly hard to value and trade. The example in 2007 and 2008 was ­asset-backed securities, and in particular over $2 trillion in residential (and commercial) real estate ­mortgage-backed securities that had been promoted as a matter of government policy and were held by financial institutions and investors around the world.

Until December 2006, subprime mortgages had been sustained by the Fed’s drastically low interest rates and ever-increasing house prices. But then that bubble burst. Delinquencies and foreclosures started rising, adversely affecting the tranches of complex securitizations. Rating agencies downgraded hundreds of subprime mortgage bonds. Financial firms became concerned about the solvency of counterparties with large but opaque holdings, and they responded by reducing or cutting off extensions of credit.

The situation came to a head in early September 2008. The giant mortgage insurers Fannie Mae and Freddie Mac were put into conservatorships, Merrill Lynch was forced into acquisition by Bank of America, Lehman filed for bankruptcy, and the Fed made an $85 billion loan to AIG—all in a ten-day period. With such unmistakable signals of the scope and severity of the problem, the flow of funds through the financial system dried up and business firms in general were forced to contract operations. A severe recession in the real economy was under way.

This kind of common asset problem affecting a great many firms cannot be prevented or cured by the early resolution of an individual SIFI. It should be understood to be beyond the reach of Title II or Chapter 14, though they remain relevant to the extent the two categories of systemic risk overlap and some SIFIs can be resolved.