Issue Brief: Volume1, Number 4

“Too Big to Fail”

Authors: Professor Kent Smetters and Christopher Pericak

In July 2012, HSBC Holdings, Europe’s largest bank, was accused in a U.S. Senate report of laundering monies for Saudi Arabian terrorists, Mexican drug cartels and rogue regimes in North Korea and Cuba.


  • The Dodd-Frank Act does not provide sufficient protection against another major financial crisis.
  • A better regulatory system would promote financial stability by correcting the key market failures that lead to excessive risk taking by Strategically Important Financial Institutions (SIFIs).
  • Regulatory policies centered on contingent capital would offer a clearer and purer market signal when a SIFI is performing poorly and trigger steps to mitigate the financial risks.

In December 2012, the firm agreed to a record $1.9 billion fine but escaped criminal prosecution from U.S. Attorney General Eric Holder. In withholding criminal charges, prosecutors, in part, cited fear of “collateral consequences” of disrupting an international bank that is tightly tied to almost seven dozen economies throughout the world where HSBC operates. In sharp response, Oregon Senator Merkley claimed that the Justice Department “firmly set the precedent that no bank, bank employee, or bank executive can be prosecuted even for serious criminal actions if that bank is a large, systemically important institution.”[2] Indeed, HSBC’s record fine was equal to only about one month of earnings.

Despite the passage of the 2010 Dodd-Frank Act, U.S. regulators remain petrified of Systemically Important Financial Institutions (SIFIs). These are entities that are widely viewed as “too big to fail” because of their possible contagion effects on the rest of the financial markets.[3] Hundreds of billions of taxpayers’ dollars have been spent bailing out many of these SIFIs. And, while taxpayers might largely get paid back for their help in some of these cases, they will be under-paid for the risk they have borne. 

The federal government added to the disorder by establishing its own track record of supporting big bailouts. The counterparties of AIG were made whole for their losses—100 cents on the dollar—despite being sophisticated institutions that are better positioned to underwrite their own counterparty risks, unlike the disinterested taxpayers who ultimately bore those risks. Bondholders for Fannie Mae and Freddie Mac were also made whole with at least $291 billion of taxpayers’ money.[4] The obligations for just these two entities totaled more than $5 trillion, equal to half of all publicly-held debt when the federal government agreed to inherit their liabilities in the fall of 2008. 

The Dodd-Frank Act tries to address some of these issues by forming a foundation that empowers the nation’s key financial regulators to implement new rules—many of which still need to be enacted. Nonetheless, SIFIs still can safely rely upon taxpayers to bear some of the downside risk of their financial decisions while private stakeholders continue to enjoy the potential upside. As Federal Reserve Governor Jeremy Stein recently argued, “[W]e are quite a way from having fully solved the policy problems associated with SIFIs.” In particular, “the market still appears to attach some probability to the government bailing out the creditors of a SIFI; this can be seen in the ratings uplift granted to large banks based on the ratings agencies’ assessment of the probability of government support.” [5]

One of the main problems with the Dodd-Frank Act is that it focuses on symptoms, like excessive risk taking, rather than on the root causes related to the underlying incentives. The real question, in examining the recent financial crisis, is why the private actors were not properly incented by their own self-interested pursuit of profit to eliminate the inefficiencies that led to their excessive risk taking. Without a solid framework for understanding the root causes of their behavior, regulation quickly becomes policy Whac-A-Mole, a continuous losing battle that focuses on responding defensively to ever-changing symptoms. This reactionary approach to regulation cannot lay the real groundwork for mitigating future disasters and actually can cause economic harm by raising costs and reducing productive activity.

This brief summarizes “an economist way of thinking” about regulating SIFIs, founded on the concept of market failures. If there were no market failures, regulation would not be needed and likely would be misguided. But market failures do happen. Without an understanding of the core market failures that actually exist, regulation will be missing or slapdash. A framework built on market failures provides a foundation for thinking critically and finding the right policy solutions.

We then use this framework to analyze the current regulatory environment. While Dodd-Frank does make a couple of important key changes consistent with our framework, the world remains vulnerable.

Finally, we show how a fairly simple and actionable regulatory model, where reverse convertible debt generates market signals to help realign incentives, can be implemented to address most of these market failures. This illustrative regulatory model transfers most of the risk efficiently back to the private parties in proper order of their access to information and their roles and responsibilities: executives, shareholders, bondholders, and then counterparties. The risk to taxpayers is substantially reduced. 

The Market Failures

For starters, we must be clear about what we mean by a “market failure.” Economists distinguish between reasonable risk-taking and risk-taking that is inefficient ex-ante. For example, did the CEO take a risk that her shareholders would have approved had they been fully informed? Or, did the CEO simply gamble because she profited disproportionately from the upside in the form of stock options? The former is simply bad luck, the latter a market failure.

Our analysis has identified four key market failures facing the financial system.

I. Bounded Rationality (in particular, a form known as “disaster myopia”) occurs when low probability events or events with little history are ignored in the face of considerable complexity costs. The possibility of a 9/11 terrorist attack, for example, became more obvious in hindsight. Ignoring low probability events is rational if the associated severity is low, since there are costs associated with planning for every possible event. It becomes a problem, however, when those events also have a large severity of loss: that’s the reason why people buy house insurance even though the chance of a significant loss is small. There is ample evidence that prior to the financial crisis, CEO’s, CIO’s, CFO’s and other key employees of investment firms, insurers and other financial intermediaries took large risks they didn’t understand. Of course, today, those private actors better appreciate these risks. However, they can continue to ignore these risks if they face little cost of being wrong because the government has their backs. Indeed, bounded rationality typically amplifies another market failure, adverse selection.

II. Adverse Selection (in this context) occurs when securities sellers have better information about the risk of their products than buyers do. During the financial crisis, many contracts were greatly mispriced relative to their default risk, and not just in hindsight. “Lemons” (higher risk securities) were being successfully sold as “cherries” (lower risk). The evidence suggests that adverse selection didn’t act alone but exploited the bounded rationality of key decision makers who misjudged the odds and severities. Think of bounded rationality as the needle and adverse selection as the toxic drug. 

III. A Samaritan’s Dilemma, an expression coined by Nobel laureate James Buchanan, is a type of moral hazard that occurs when a party takes advantage of the protection offered by (usually) the government, because the government, like the Good Samaritan, can’t credibly commit to not bailing out after a loss. Fannie Mae and Freddie Mac, although technically private entities, benefitted from the implicit Congressional protection provided to their bondholders. In exchange, Fannie and Freddie responded to federal pressure to increase homeownership by buying “Alternative A” mortgages and substantial amounts of securitized loans. To finance these operations, both entities were able to borrow at low rates and carry substantial amounts of leverage, often 40 - 70 times their equity, and much higher if you count the face value of the guarantees that they extended. The bailout of the bondholders of both institutions only enhanced the Samaritan’s Dilemma. The government also bailed out many financial institutions that were not previously viewed as having this implicit guarantee, only adding to the Dilemma.

IV. A Principal-Agent Problem exists when an agent who is employed by a principal does not act in the principal’s best interest. Executives don’t always maximize shareholder value if the executives’ actions are hard to monitor. Likewise, salespeople working on commission might increase sales regardless of costs. Large firms are especially difficult to risk manage. A single small unit in AIG collapsed the entire company; similar events occurred previously at Barings and other firms. Such diseconomies of scale are challenging for non-key personnel and shareholders to ascertain before actual losses emerge. In the meantime, a larger firm affords generous compensation to those who run it.

Figure 1 summarizes the role that each market failure played in the 2008 financial crisis. While each market failure remains important today, the Samaritan’s Dilemma is the most challenging; indeed, all major G20 countries have all but given up on the pretense that they won’t bail out SIFIs after ruin.

Figure 1: Housing Bubble and Financial Crisis Caused by Four Market Failures

The Dodd-Frank Act Leaves America and the World Exposed

The Dodd-Frank Act may appear to address some of the market failures illustrated above. However, by focusing on symptoms instead of root causes, Dodd-Frank leaves much work to be done.

Bounded Rationality and Adverse Selection

Some of Dodd-Frank’s most successful policies come from enhancing information and lowering the cost of obtaining it, thereby reducing bounded rationality and adverse selection.[6] First, the Act attempts to formalize the derivatives trading process by encouraging that more trades be executed through clearinghouses and exchanges, where a third party provides more due diligence and engages in more risk management. In practice, though, loopholes and clever industry workarounds are emerging.[7] Second, the Act also reduces the role of the credit rating agencies by creating more competition and reducing the SEC’s and Federal Reserve’s reliance on ratings for determining capital buffers and collateral. There is little evidence that the standard Moody’s-S&P-Fitch oligopoly, though, is facing any serious competition. Moreover, coordination with Basel III’s reliance on ratings remains unresolved. Third, the Act mandates using stress tests to evaluate risk across financial institutions. But, so far, those tests emphasize worst-case scenarios, and their methodology is not transparent to outside parties, reducing their effectiveness.

The Samaritan’s Dilemma

Dodd-Frank is fairly weak when it comes to dealing with the Samaritan’s Dilemma. First, the Act requires SIFIs to design and regularly update their “living wills”—their own funeral instructions—in order to expedite their liquidation upon failure. In theory, an orderly liquidation process quickly unlocks remaining capital to reduce the cascading nature of defaults across financial institutions. A living will also clearly identifies subordinated stakeholders who are most at risk, giving them more incentive to closely monitor the SIFI’s risks. Still, much of the Samaritan’s Dilemma problem remains: counterparties, bondholders and maybe even some equity owners can depend, in full or in part, on the government backdrop and therefore are less likely to engage in due diligence and risk management. Moreover, initial drafts of living wills were due last summer and fell short of the Federal Reserve’s and FDIC’s expectations.[8] They are also hard to coordinate across international legal standards.[9]

Second, the emergency lending powers of the Federal Reserve have been restricted to assisting a bank’s liquidity rather than supporting an insolvent institution.[10] Instead, a bailout requires the backing of the new Financial Stability Oversight Council (FSOC), representing a more diverse set of opinions of the nation’s largest federal regulators. In fact, many actions require supermajority support.[11] In theory, therefore, a SIFI can’t count on being bailed out, thereby being allowed to fail like Lehman (although presumably not in the same manner). Such a random or “mixed strategy” is indeed often optimal in the presence of moral hazard. In practice, however, a supermajority of FSOC members would likely support a bailout of an institution that has already been designated as a SIFI by the G20. Most of the voting members are presidential appointments anyway, and so they would likely pay attention to the wishes of the Administration.

Third, the so-called “Volcker Rule,” which garnered significant media attention, attempts to prohibit a bank from trading for its own account (“proprietary trading”), rather than trading on a client’s behalf. The thrust of the Rule is that “prop desks” don’t serve a regular business purpose and so taxpayers should not be exposed to seemingly personal bets tucked underneath the too-big-to-fail implicit guarantee. However, a well-functioning bank dealer doesn’t simply pass through a client’s risk, like a pure broker, to an outside counterparty. It pools clients’ idiosyncratic risks, bringing to the market the remainder that it can sell at a price less than its internal risk cost. It’s nearly impossible to determine at the margin when that remainder constitutes trading for the bank’s own interests or a client’s. Even government regulators can’t apparently agree on how to detect a violation, delaying the Rule’s actual implementation. Even though obvious violations could be detected, the Rule, including its watered down implementation in Senate conference, will mainly only impact a few large banks yet increase compliance costs for all.[12]

Fourth, the Act requires the study of additional firm capital to augment a firm’s ability to suffer a loss, including contingent capital (discussed more below) and GAAP-based capital requirements, which follow quarterly accounting statements. But the Act stops short of requiring capital requirements to reflect the market’s valuation of on- and off-balance sheet assets and liabilities and contingencies. 

The Principal - Agent Problem 

Dodd-Frank also does very little to address principal-agent problems. First, to deal with the conflicts of interest that are inherent in contract sellers paying for their own credit ratings, the Act now makes credit rating agencies just as liable as any other “expert,” such as the firm’s auditor. Moreover, it requires that credit agencies include an attestation in any credit rating that they were not influenced “by any other business activities” and also requires agencies to create informational barriers between their marketing departments and their expert raters. But the Act does not ban unrelated consulting. Moreover, sellers still pay for their own ratings, which is an almost unavoidable conflict of interest in the modern era where information on ratings can be easily shared.

Second, to deal with principal-agent problems between firm employees and shareholders, the Act enhances the executive “clawback” provisions already found in the Sarbanes-Oxley Act of 2002, requiring that firms seek repayment from any current or former executive of any incentive-based compensation paid during the three-year period preceding a “material noncompliance.” However, executives are still incented to take big bets that can earn them nice bonuses, provided that such actions are revealed in accounting statements, which are often challenging for the market to understand. In fact, executives would only have to return their bonuses if their bets fail, rather than suffer a large loss like their shareholders—and, even then, only if “material noncompliance” can be proven, potentially litigated.

Aligning Incentives with Reverse Convertible Debt

The most effective regulatory policy aligns the incentives of each stakeholder according to his or her access to information and responsibilities. Most current regulatory discussions thus far have focused on building a SIFI’s capital base as a buffer against losses and contagion. Of the different types of capital, contingent capital is likely to be the most effective at aligning incentives because its price can be highly responsive to the specific actions taken by a SIFI. But it must be implemented correctly to be effective.

Contingent Capital 

Contingent capital is debt that can convert to equity when a firm is in trouble. One particular type is a reverse convertible (debenture). The more familiar standard convertible gives the debt holder the option to convert into a predetermined number of company shares, an action normally taken as the share price increases. A reverse convertible, by contrast, gives the company the right to force this conversion, an action usually taken as the share price declines. A reverse convertible, therefore, effectively reduces the debt load during challenging financial times. While used by some European banks in the past, reverse convertibles are uncommon in the U.S., where the tax treatment is less clear.

The policy goal of contingent capital is to protect taxpayers by allowing SIFIs to recapitalize during times of economic distress, known as a “bail in.” Contingent capital also gives the broad base of shareholders strong incentives to push for corrective actions before the trigger point to avoid dilution.[13]

Among its 2319 pages, the Dodd-Frank Act requires that regulators study “contingent capital” in more detail, but it doesn’t go much further. Federal Reserve Governor Daniel K. Tarullo, who serves as the Fed’s point person for non-consumer financial regulation, recently stated that “for all the attention paid to [contingent capital] in the last few years, it is even now not clear as a practical matter that an instrument can be developed which would be cheaper than common equity but still structured so as to convert in a timely, reliable fashion.”[14] Under Dodd-Frank, the Federal Reserve has until July 22, 2014, to suggest a firm position. Until then, U.S. law remains fairly silent on contingent capital.

The G20’s Financial Stability Board, though, is studying contingent capital in more detail. One idea being discussed is that a top holding company of the SIFI would amass a large amount of long-term unsecured debt. A “single entry” regulator in each country would have the power to force the conversion to equity as part of liquidation, thereby providing some buffer to taxpayers. In the U.S., this role would likely fall to the FDIC under the orderly liquidation powers granted to it by Dodd-Frank.[15] As currently being discussed, though, this contingent capital does not address Governor Tarullo’s concern about conversion occurring in a “timely, reliable fashion.”[16] Instead, its objective is to mainly protect taxpayers from losses from a “gone concern” rather than to support a “going concern.” Furthermore, its price also would not serve as a quality early warning signal to policymakers or be the basis for pegging various pre-failure incentives.

Characteristics of Optimal Contingent Capital

The optimal implementation of contingent capital would: (I) allow market prices to help predict a failure early enough to help avoid a liquidity crisis, (II) produce a pure signal that is not distorted by endogenous government actions themselves (including the bail-in itself), (III) be specific to a SIFI’s own actions and (IV) promote consistency across borders. Let’s briefly consider each criterion in turn.

(I) Without harnessing the power of market prices, the various regulatory bodies would have to be well informed, coordinated and proactive. The competitive market, in contrast, is typically faster and more accurate. Regulators, for example, entered Bear Sterns only days before it would have defaulted; in contrast, its share price actually fell long before. The credit rating agencies don’t do much better. Their ratings, for example, often substantially lag credit default spreads for sovereign debt, which explains why credit downgrades often have little impact. 

(II) Any pure signal also should not be contaminated by the regulators’ decisions or the bail-in action. A SIFI’s standard bond yield, or credit default swap risk premium trading on those bonds, therefore is not a good pick for the trigger that decides when to convert the contingent capital into equity.[17] At a minimum, those prices are noisy, since they would capitalize expected government action.[18] When government action is highly likely, focusing on those instruments could even lead to an indeterminacy of government action. The firm’s share price, while a potential contender, is also at least somewhat distorted by government action and the potential of the bail-in itself.[19] Instead, the price signal should track an instrument that is also subordinate to government help and even the contingent capital.

(III) Ultimately, getting an early and pure signal should be part of a structure that incentivizes each SIFI to be prudent. This is an area where Basel III, while improving the quality of capital, falls short. Most of the required capital in Basel is a constant fraction of equity or tied to a measure of risk-weighted assets. But a significant amount of variation can occur within each risk bucket, thereby allowing for substantial amounts of moral hazard in the presence of the government guarantee.[20] Additional forms of capital, including countercyclical capital and a “SIFI surcharge” are discussed below; while they can likely play a role in reform, they also fail to align marginal incentives.

(IV) Having a consistent cross-border standard is also important in order to not create competitive advantages for SIFIs who “game” the market failures. As it currently stands, the existing and simpler Basel regulations have not been consistently applied across countries. In Congressional testimony, Governor Tarullo argued, “Despite extensive sharing of information on supervisory practices, the Basel Committee has, over the years, found it difficult to achieve … rigorous and consistent application of those rules by supervisors and firms across countries.”[21] Achieving consistent application will become only harder, even with peer review, as regulators’ rules move toward even more complex bank-specific capital standards. In contrast, markets are highly incented to make such judgments.

Illustrative Use of Reverse Convertible Debt as Integrated Part of Regulatory Reform

One such regulatory structure using contingent capital is illustrated in Figure 2. During normal economic times, SIFIs would sell unsecured contingent capital to the open market that is subordinate to the firm’s bond holders and most other claims. To be concrete, we’ll think of these contracts as reverse convertibles, although the exact nomenclature is unimportant. 

Figure 2: Illustrative Use of Reverse Convertible Debt as Early Warning Signals

SIFIs would be required to sell these contracts in addition to the other potential capital requirements of Basel III, including risk-weighted capital, a countercyclical buffer, and a capital “SIFI surcharge.” The exact amount of reverse convertibles required to be sold would, at a minimum, be set to equalize the average cost of capital between the SIFI and non-SIFIs during normal times, as SIFIs will likely continue to enjoy a lower cost of capital in the presence of the Samaritan’s Dilemma, even post-Basel III. Of course, since the larger interest paid on the reverse convertible is in exchange for the firm’s option to recapitalize, the equalization of the cost of capital is a lower bound. In practice, a large share of the firm debt would be required to be held in the form of reverse convertibles to reduce any SIFI advantage. 

SIFIs then would be required to run a battery of standardized stress tests. These testing suites would be updated regularly by the appropriate regulator in the United States, likely the Federal Reserve and SEC working together. Internally, SIFIs could automate many of these tests, much like how eBay or Facebook routinely runs a large suite of automated “unit,” “regression,” and “penetration” tests on its code base. 

Unlike the current bank stress tests conducted in the United States, however, we envision that stress tests would be standardized, extensive, and public. Arguments made by SIFIs for keeping stress test results secret seem to have more to do with protecting executives and rent seeking than safeguarding intellectual secrets. Indeed, stress tests in Europe have become more extensive over time. But if U.S. regulators were truly concerned about these issues or the costs associated with stress tests, the tests could be broken up into multiple and conditional testing suites, including “Level 1” (L1) tests during normal times and more detailed “Level 2” (L2) tests performed only after negative market signals, as shown in Figure 2.

Getting the right market signal is critical for measuring a SIFI’s risk of default. As noted earlier, a reverse convertible note must offer a higher yield than the bank’s senior standard debt because the note will be converted to equity when the SIFI is performing poorly. The difference between these two yields is a relative risk premium that indicates the value for this added risk on top of the firm’s normal risk. This risk premium, however, is not pure enough of a signal for the conversion trigger, since a severe default could impact the yield paid by the senior debt as well, unless it were fully protected by government action. 

Instead, a “conversion swap” market that tracks the conversion of these reverse convertibles into equity would need to arise. After a conversion of the contingent debt to equity, the seller of the conversion swap would take delivery of the underlying equity in exchange for making the originally-promised bond payments. Unlike credit default swaps trading on senior debt, the volume in this new conversion swap market should be robust for two reasons. First, even equity investors would be attracted to the potential of buying equity “on the cheap” if the risk premium of a conversion swap rose to an inefficient large level. To ensure that the market has sufficient time to respond to even a potential “bear run” that artificially inflates the risk premium, we assume that the conversion of the contingent debt to equity only happens after the risk premium exceeds a certain level (6% in our example) for a given time threshold (30 days). Second, the ability to insure the reverse convertibles with swaps should let large institutional investors such as insurance companies hold the reverse convertibles as reserves. The combination of reverse convertibles and the swaps trading on them then should support a large demand for each.

The spread on these conversion swaps therefore would be based on the fundamental outlook of the bank’s solvency and serve as a pure early warning signal to regulators. If the risk spread exceeded some threshold (shown as 6% in Figure 2, for the sake of concreteness) some minimal actions to preserve capital, falling on executives, are taken along with additional stress tests and voluntary actions by the firm to restore capital to regulated minimums. If the spread stays above the threshold for a period of time (say 30 days) then additional actions are taken that dilute shareholders are mandatory, including conversion of contingent capital into equity. If still required, additional equity offerings are made.[22]

Post-conversion, the role of regulators is necessarily discretionary. The Federal Reserve may make some limited investment, in exchange for preferred shares, if they believe that default of non-contingent debt could result in liquidity problems; if those shares are not paid off within a reasonable period then the Federal Reserve would sell them in the open market and buyers naturally would retain the option to convert to common. However, the size of the Federal Reserve’s actions is limited and bondholders face risk. Key employees (including previous ones) suffer additional losses, including repayment of some past compensation. Experimental work has demonstrated that people are often more highly incented to the loss of previous income rather than the potential of earning additional income. If large contagion effects are suspected, additional government action for counterparties can be taken by the Financial Stability Oversight Council. In most cases, counterparties would receive a haircut on their investments.

Contingent Capital is More Effective at Aligning Incentives Than Other Forms of Capital 

Besides the Basel III core capital requirements noted earlier, Basel III and U.S. regulators also are contemplating other forms of capital, including “countercyclical capital” and a capital “SIFI surcharge.” The SIFI surcharge, in particular, attempts to peg capital requirements to a SIFI’s own complexity and potential contagion impact upon failure. It is also apparently the form of additional capital being most closely studied by the Federal Reserve. But, while such a surcharge can certainly be part of a comprehensive regulatory reform, it still does not tie at the margin to the actual actions taken by the firm: after the surcharge is paid, the firm can still engage in moral hazard, and it is doubtful that the Fed would, or even could, increase the surcharge once it finally realizes that the SIFI is in trouble. Even the FDIC has struggled to implement its existing risk-based assessment. Taxpayers could find themselves quickly on the hook for losses without a “bail in” buffer. To compensate, the Fed would be tempted to simply overestimate the surcharge ex ante, which is also inefficient. Indeed, Governor Tarullo created a stir when he suggested that some capital surcharges could lead to a doubling of Basel III requirements—a statement from which the Federal Reserve appeared to back away subsequently. 

At a minimum, any SIFI surcharge should at least be guided by signals of risk from contingent capital rather than relying on a regulator’s internal model. Simply put, it is unlikely that regulators could ever develop models for assessing firm-level SIFI surcharges that are more accurate than market prices.

Updating the “Automatic Stay” Rule Exemption

Our illustrative regulatory framework encourages SIFI counterparties to conduct due diligence, since government action might not fully cover their losses upon bankruptcy. In practice, though, counterparties holding “qualified financial contracts” (derivatives) are exempt from the “automatic stay” rule in bankruptcy, allowing them to close out their positions and seize underlying collateral outside (and before) the normal liquidation process. Congress in 1982 believed this exemption from the automatic stay was necessary to prevent the “insolvency of one commodity or security from spreading to other firms and possibly threatening the collapse of the affected market.”[23] Of course, as we saw during the recent financial crisis, the first-come-first-served approach to recovering collateral can lead to an intense run on the bank, especially for trades not cleared through a clearinghouse or an exchange. Moreover, the derivative exemption allows for a potentially large “loophole” where non-derivative contracts are repackaged as derivatives for circumventing the automatic stay. Dodd-Frank did little to change these rules other than allowing for a one-day moratorium and banning “walkaway” clauses, where nondefaulting parties can “walk away” from their own payment obligations.

Recent thoughtful and detailed analysis has argued for modifications to the automatic stay rule for various types of transactions, especially for SIFIs and transactions not cleared through a third party.[24] These changes are likely an integral part of any serious regulatory reform that aims to encourage counterparties to conduct a greater amount of their own due diligence.


The too-big-to-fail problem remains too big of a problem. Government regulations to date have fallen short. But a more solid regulatory framework can be built. Regulations should not simply aim to tame the market by focusing on one apparent symptom and then the next. Rather, the power of the market itself must be used to discipline itself. Contingent capital is one effective way to harness market signals and encourage restraint, in conjunction with other reforms. Its use is currently being debated by the G20. Successful implementation, however, requires keen attention to details, including layering it correctly within the SIFI’s capital structure.

About the Authors

Kent Smetters, PhD

Professor of Business Economics & Public Policy The Wharton School[1]

Kent Smetters is the Boettner Chair Professor at the Wharton School of the University of Pennsylvania and a Faculty Research Fellow at the National Bureau of Economic Research. Previous policy positions include the Congressional Budget Office (1995-1998) as well as Economics Policy Coordinator (Deputy Assistant Secretary) for the United States Treasury (2001-2002). He also was the Kaiser Visiting Professor of Economics at Stanford University (2000-2001). He received his PhD in Economics from Harvard University.

Christopher Pericak

Chris Pericak conducted graduate studies at the University of Chicago in public policy and international affairs.  He is currently a Consultant at The Advisory Board Company.


1 Corresponding author. 3620 Locust Walk, Philadelphia, PA 19104. Email: This research was supported by the non-profit Insurance Studies Institute. We thank Richard Herring, Paul Siegert, and David Skeel for their valuable comments.

2 Jeffrey Merkeley, “Letter to Attorney General Holder,” December 13, 2012,

3 An exact definition of a SIFI has not yet been determined, but the G20 Financial Stability Board published a list of 29 firms that they deem to be “global SIFIs.” Financial Stability Board, “Policy Measures to Address Systemically Important Financial Institutions” (November 4, 2011), 

4 Congressional Budget Office, The Budgetary Cost of Fannie Mae and Freddie Mac and Options for the Future Federal Role in the Secondary Mortgage Market (June 2, 2011),, which also notes that “losses have increased somewhat since [the original $291B estimate] because of the continued deterioration of conditions in the housing market” (p. 3).

5 Jeremy C. Stein, “Regulating Large Financial Institutions,” Presented at IMF Conference, Washington, DC, April 17, 2013. 

6 See, for example, David A. Skeel, The New Financial Deal: Understanding the Dodd-Frank Act and its (Unintended) Consequences (John Wiley & Sons, 2011).

7 See, for example, “A $4 Trillion Dodd-Frank Loophole,” BusinessWeek, September 11, 2012.

8 William C. Dudley, “Solving the Too Big to Fail Problem,” Remarks at the Clearing House’s Second Annual Business Meeting and Conference, New York City, November 15, 2012. 

9 Richard Herring, “Wind-down Plans as an Alternative to Bailouts: The Cross Border Challenges,” in Kenneth Scott and John Taylor, eds., Ending Government Bailouts as we Know Them (Hoover Press, 2010), pp. 125-151.

10 The Dodd-Frank Act: Title XI—Federal Reserve System Provisions,

11 The Dodd-Frank Act: Title I—Financial Stability,

12 In Senate conference, the Volcker Rule was liberalized to allow banks to make outside investments in hedge funds, private equity and other instruments. Only 0.5% - 2.0% of bank revenue comes from proprietary trading, although Goldman Sachs, Morgan Stanley, Citigroup, JPMorgan and Bank of America earn substantially more from such trades.  See Standard and Poors, “For U.S. Bank Ratings, the Volcker Rule’s Impact Depends on the Final Details,” October 12, 2002.

13 See William C. Dudley, “Some Lessons from the Crisis,” Remarks at the Institute of International Bankers Membership Luncheon, New York, October 13, 2009. 

14 See Daniel K. Tarullo, “Regulating Systemically Important Financial Firms,” Remarks at the Peter G. Peterson Institute for International Economics, Washington, DC, June 3, 2011. However, Tarullo later seems to be more receptive to contingent capital. See Daniel K. Tarullo, “Industry Structure and Systemic Risk Regulation,” Remarks at the Brookings Institution Conference on Structuring the Financial Industry to Enhance Economic Growth and Stability, December 4, 2012, p. 13. New York Federal Reserve President William Dudley also argues that “the introduction of a contingent capital instrument seems likely to hold real promise.” See Dudley, “Some Lessons from the Crisis.” It appears, therefore, that the Federal Reserve has not yet arrived at a concrete position. Dodd-Frank allows, but does not require, the Federal Reserve to issue a report on contingent capital by July 21, 2014, after the FSOC issues its findings to Congress. The FSOC issued its report and “recommends that contingent capital instruments remain an area for continued private sector innovation, and encourages the Federal Reserve and other financial regulators to continue to study the advantages and disadvantages of including contingent capital and bail-in instruments in their regulatory capital frameworks.” See Financial Stability Oversight Council, “Report to Congress on Study of a Contingent Capital Requirement for Certain Nonbank Financial Companies and Bank Holding Companies,” July 2012, p. 3. 

15 Federal Deposit Insurance Corporation and the Bank of England, “Resolving Globally Active, Systemically Important, Financial Institutions: A Joint Paper,” December 10, 2012. 

16 Similarly, as Darrell Duffie writes: “The trigger that converts the debt to equity should be set as to eliminate the debt claims before a liquidity crisis is likely to begin, and hopefully with a sufficiently strong impact on the balance sheet to forestall a self-fulfilling presumption of a liquidity crisis.” See Darrell Duffie, How Big Banks Fail: And What to Do about It (Princeton University Press, 2011), p. 50.

17 A literature has examined optimal conversion triggers. See the extensive and excellent review in Charles Calomiris and Richard Herring, “Why and How to Design a Contingent Convertible Debt Requirement,” Brookings-Wharton Papers on Financial Services, 2012. The Squam Lake Working Group on Federal Regulation, “An Expedited Resolution Mechanism for Distressed Financial Firms: Regulatory Hybrid Securities,” Council on Foreign Relations Press, April 2009, recommends a dual trigger: an aggregate trigger based on the regulator declaring a systemic emergency and a firm-specific one based on a firm’s GAAP-based equity. Robert McDonald, “Contingent Capital with a Dual Price Trigger,” Working Paper (Northwestern University, 2010) also considers a dual trigger while Mark Flannery, “Stabilizing Large Financial 

Institutions with Contingent Capital Certificates,” Working Paper (University of Florida, 2009) argues for the benefit a single trigger based on firm’s equity. Duffie, How Big Banks Fail, reviews various triggers, although he does not settle on one. The literature has also discussed the potential of market manipulation (e.g., “bear runs”) to manipulate various triggers. But the literature has not examined a trigger in the presence of a likely government backstop.

18 See Drew Fudenberg and Jean Tirole, Game Theory (MIT Press, 2000), pp. 17-18. In the setting herein, the government’s actions should be contingent on a negative signal, but the market would not produce this signal if it is confident that the government will help.

19 The distortion to the share price by the anticipated bail-in will be greatest at a low share price. Suresh Sundaresan and Zhenyu Wang also argue that, even without additional government action, the conversion of the contingent capital itself could lead to a downward spiral in share price: if the market believes that the SIFI is in trouble, the share price will fall by even more than without the contingent capital because its conversion will dilute shareholders even more.  See the Federal Reserve Bank of New York, “Design of Contingent Capital with Stock Price Trigger for Conversion,” Staff Report 448, 2010. Charles Calomiris and Richard Herring, in “Why and How to Design a Contingent Convertible Debt Requirement,” argue that their approach to an equity-based trigger is less susceptible to this problem if firms have enough time to issue capital. 

20 Patrick Slovik shows that over the past two decades, SIFI’s were able to drop their risk-weighted assets from almost 70% to about 35% and even “transform high-risk subprime loans into seemingly low-risk securities on a scale that would spark a global financial crisis.” See “Systemically Important Banks and Capital Regulation Challenges,” OECD Working Paper No. 916, 2012, p. 10. See also Robert Pozen, Too Big to Save? How to Fix the U.S. Financial System (John Wiley & Sons, 2010).

21 Daniel K. Tarullo, Statement before the Committee on Financial Services of the U.S. House of Representatives, June 16, 2011. 

22 Contingent capital and mandatory equity offerings are essentially the only two methods of recapitalization possible at this point and the ones most commonly discussed by regulators. Several commentators believe that a mandatory equity offering is challenging for regulators to enforce after a shock, while others (including Duffie, How Big Banks Fail) argue that they have worked in Europe. Nonetheless, most attention has focused on contingent capital.

23 1982 U.S.C.C.A.N. 583, 583.

24 Darrell Duffie and David A. Skeel, Jr., “A Dialogue on the Costs and Benefits of Automatic Stays for Derivatives and Repurchase Agreements,” Stanford University Working Paper No. 108, 2012.