Banking and public policy: too big to fail.

Kaufman, George G.

I. INTRODUCTION

I have devoted much of my professional career to the study of banking (in the broadest sense to include all channels of financial intermediation operating either in the sunlight or in the shadows), its management, and its interaction with public policy. On the whole, it has been a wonderful and exciting journey. Banking and finance are at the very heart of most industrial economies. As has been vividly reinforced in recent years, what happens in the financial sector may affect not only the remainder of the financial sector but also the real economy, particularly on the downside, and vice versa. What happens in the real sector feeds back on the financial sector.

But, at times, I feel that economists know only little, if any, more about banking and its relationship with the rest of the economy than we did some 50 years earlier. Financial crises have neither been reduced in either magnitude or frequency, no less disappeared (Reinhart and Rogoff 2009). Indeed, in some countries, they have become more frequent and deeper. The recent financial crisis from 2007 to 2010 in the United States and many, but not all, industrial countries, commonly now referred to as the great financial crisis (GFC), may well be the most serious and damaging in U.S. history. It has been a rough and costly ride. This should have humbled most of us.

In this article, I describe what I believe to be one of the, if not the, major public policy issues in banking today--"too big to fail" (TBTF). Although most ideas are taken from my earlier papers, all are not mine. I have borrowed freely and shamelessly from my colleagues. If nothing else, correctly conducted plagiarism can be efficient.

As noted, banking and finance are important in nearly all economies. For the United States, income created by the financial sector accounts for about 8% of GDR Not only are banking and finance important in most countries, but, at least until the GFC, the sector was growing rapidly (Artill, Hou, and Sarkar 2014). This had important collateral benefits and costs. A number of studies have reported that the broader, deeper, and more liquid the financial system is, the greater is the growth rate in real aggregate income in that country (Barth, Caprio, and Levine 2012; Caprio 2013). Some more recent studies, however, found that these results may be sensitive to a number of factors, including the time period examined and the state of a country's economic development (Gambacorta, Yang, and Tsatsaronis 2014; Levine, Loayza, and Beck 2000). Losses in aggregate income from financial problems and crises in any country may be greater than, equal to, or less than the gains in income from greater and more efficient financial activities both before and after the crises. Which it is determines the societal contribution of the financial sector and the success or failure of our policies. Moreover, as observed in recent years, abrupt changes in the health of the financial sector have important societal (income and wealth distribution) and macroeconomic implications. That is, not only aggregate wealth but also the distribution of wealth is importantly affected by banking and financial behavior.

TBTF goes under many different names, such as

* Too big to liquidate;

* Too complex to liquidate;

* Too big to manage;

* Too complex to manage;

* Too big to prosecute; and

* Too big to jail.

Thus, the term means different things to different people. But, by any name, it suggests that the big are getting special treatment that the small are not. Other than those who benefit directly, for reasons discussed later, few like the concept and what it represents and there have been many attempts to kill it. But, like vampires and Moby Dick, TBTF refuses to stay dead for very long and returns to haunt our existence. As a result, TBTF has become one of the most important public policy issues affecting banking and financial markets today. Indeed, the issue has migrated beyond economics and finance to religion. The Financial Times of June 18, 2014, reported that the Archbishop of Canterbury has warned that banks are still "too big to fail" (Arnold 2014). Apparently, large banks should repent for the sin of largeness!

In theory, a bank fails and is legally closed (its charter revoked) when the value of its liabilities exceeds that of its assets, so that its net worth is negative. The financial stake of the bank's equity shareholders is wiped out and the unsecured creditors, who have not managed to run, share in the remaining losses according to their legal priority. But, in practice, some troubled big banks or other large financial firms may not be failed and closed when they become insolvent as would other banks or other financial firms. They might be permitted to continue to operate by their regulators or the government, their creditors would not suffer losses and not flee the institution, and their shareholders' stake possibly not wiped out totally. Alternatively, select insolvent large firms, mostly, but not exclusively (think General Motors and Chrysler) large financials may be failed but subject to a resolution process that differs from the resolution process usually applied to other insolvent firms in the same industry (e.g., federal bankruptcy code [FBC]) with respect to loss allocation.

The government has enacted special TBTF resolution regimes for select firms, for example, a systemic risk exemption (SRE) or override to least cost resolution (LCR), the usual failure resolution process for chartered banks in the Federal Deposit Insurance Corporation Improvement Act (FDICIA) of 1991, and the Orderly Liquidation Authority (OLA) in place of bankruptcy for large bank holding companies and large non-bank financial companies in the recently enacted Dodd-Frank Act Wall Street Reform and Consumer Protection Act (DFA) of 2010. In TBTF resolutions, the firm is failed but some or all creditors and other in-the-money stakeholders may receive more than they would under the usual (no-TBTF) resolution process, which is likely to impose losses. Loss allocation is not determined by the usual rules. The difference between the amount paid to creditors at resolution and the corresponding lesser pro-rata recovery amount received is the loss--"protection" or "bail-out"--funded by a third party, for example, the government.

Why is there a perceived need for TBTF resolutions? In the resolution of most insolvent firms, losses are spread pro-rata over all creditor claims of the same legal priority. But there is fear in some quarters that losses to some or all uninsured and unsecured creditors at large financial institutions will generate unacceptable large adverse externalities (collateral damage) to financial and economic stability. These firms, and banks in particular, are structured in ways that make them provide critical/essential services to the economy that likely would be discontinued or continued on a smaller scale if the firm became insolvent and was resolved the usual way with losses to depositors/creditors. This is because

* Bank deposits collectively comprise the largest share of the country's money supply and are the primary medium of exchange;

* Bank deposits represent a significant portion of the public's most liquid assets;

* Bank deposits are held in a wide range of amounts, including very small amounts by a large proportion of the population, including households of limited financial means and expertise;

* Banks are major providers of credit to households, business firms, and governments;

* Banks have a large proportion of their liabilities in very short-term debt that is shorter in maturity on average than their assets and that can easily be withdrawn or not renewed (run);

* Banks operate much of the payments system;

* Some individual banks are large relative to GDP, and individual bank asset size and market concentration have increased greatly through time;

* Banks are closely interconnected with each other through interbank deposits, loans, and derivative transactions;

* Dealer banks are particularly closely interconnected with large financial institutions;

* Bank assets are widely perceived to be less transparent than assets of most nonbank firms;

* Large banks tend to have complex organizational structures and operate cross-border both nationally and internationally and are thus subject to different legal and regulatory jurisdictions; and

* Ownership and location of bank assets can be transferred quickly and at low costs (Bliss and Kaufman 2011).

The view that banks are "special" was clearly summarized recently by William Dudley, the President of the Federal Reserve Bank of New York, as follows:

... the root cause of "too big to fail" is the fact that in our financial system as it exists today, the failure of large complex financial firms generate large, undesirable externalities. These include disruption of the stability of the financial system and its ability to provide credit and other essential financial services to households and businesses. When this happens, not only is the financial sector disrupted, but its troubles cascade over into the real economy. (Dudley 2012, 1)

II. TREATMENT OF TROUBLED LARGE BANKS

If large banks experience difficulties and threaten to become insolvent, regulators (the government) have two options to avoid or limit the feared collateral damage. First, they could prevent the failure and the accompanying losses to depositors/creditors and possibly also on a more limited scale to shareholders. The firms are rescued and continue to live. Second, they could permit the firms to fail, but protect the creditors either partially or totally. That is, the firm and shareholders die, but all creditors need not.

In a rescue, firms are rescued from insolvency through the injection of sufficient explicit or implicit capital by the government to maintain or restore positive net worth. Injection of explicit capital generally takes the form of preferred stock and of implicit capital in the form of guarantees of depositor and creditor claims. No losses accrue to any party but shareholders. Senior management may or may not be changed, and essential services are maintained. Old shareholders' interests are partially or totally wiped out, for example, the TARP (Troubled Asset Relief Program) in the United States in 2008 and the RFC (Reconstruction Finance Corporation) in the United States in 1933. A more descriptive and accurate name for firms that are likely to be rescued than TBTF is too big to not rescue (TBTNR).

At times, it is necessary to protect fully only some but not all creditors or all creditors only partially to restrict the collateral damage to amounts the economy may be expected to absorb without serious damage. In this case, the troubled firm may not be rescued but failed by the regulators, particularly if the resolution regime gives the regulators the authority to select the creditors to be protected against loss and by how much. At times, the FDIC failed a bank, but fully protected all depositors, and other unsecured creditors explicitly insured or not. If some depositors or creditors will not be protected, some market discipline is maintained and a failure resolution may be preferred to a rescue. But this process requires an appropriate resolution regime. As the firm is failed, it was never strictly speaking TBTF. If the bank is failed and, because of fear of excessive collateral damage, all depositors and creditors are fully protected, the bank may be TBTNR but not TBTF. It is ironic that the term TBTF was first applied to the Continental Illinois National Bank in Chicago, the seventh largest bank in the country at the time, which was failed by the regulators in 1984. The shareholders were effectively wiped out.

There appears to be considerable confusion between rescues and failure resolution. For example, the multinational Financial Stability Board recently stated that

The "too-big-to-fail" (TBTF) problem arises when the threatened failure of a SIFI leaves public authorities with no option but to bail it out using public funds to avoid financial instability and economic damage.... The SIFI framework sets out recommendations for improving the authorities' ability to resolve such institutions in an orderly manner, without exposing tea-payers to loss, while maintaining continuity of their vital economic functions. This may require changes to resolution regimes. (Financial Stability Board, September 2, 2013, 27, emphasis added)

Likewise, a recent academic paper argues that

The too-big-to-fail (TBTF) doctrine postulates that the government will not allow large financial institutions to fail if their failure would cause significant disruption to the financial system and economic activity. It is commonly claimed that, because of the TBTF doctrine, large financial institutions and their investors expect the government to back the debts of these institutions should they encounter financial difficulty. (Acharya, Auginer, and Warburton 2014, 1, emphasis added)

In both statements, the first sentence describes a rescue and the second sentence a failure resolution, but the two are not related. Changing the resolution regime, as many reform proposals do, does not affect the rescue process, only the resolution process. The remainder of the article discusses only the resolution of failed or dead firms. For banks, their death is officially recognized when their charter is revoked. Resolution is frequently confused with rescue in which the firm is not failed but survives, though under modified ownership and management. In the recent crisis years, rescues greatly outnumbered failure resolutions for comparable troubled firms.

III. COST OF TBTF POLICIES

As noted, TBTF resolution regimes are designed to protect some or all explicitly uninsured and unsecured depositors/creditors. Unless offset by ex post correctly computed premiums, the direct cost of the protection is shifted from the depositors/creditors to third parties that are likely to be more broadly based and dispersed and therefore often less able to lobby against this shift, for example, government (taxpayers) and/or private parties (other banks). In addition, there are indirect costs that do not appear until later. Underpriced insurance encourages depositors/creditors to reduce their due diligence and monitoring of their banks and accept a lower deposit rate than if there was no protection. As is well recognized, banks respond to the reduced monitoring and lower cost of funds by increasing their risk exposures. The resulting bank losses will only become visible later. Likewise, the lower cost of funds gives TBTF banks a competitive advantage over non-TBTF (smaller) banks, whose deposits are not expected to be protected and must offer a higher unsubsidized deposit rate. (1) The resulting allocation of both financial and real resources is likely to be inefficient. Because these effects do not occur immediately but through time, the costs of TBTF promise to also increase through time.

Lastly, because all banks do not serve the same markets, they are likely to have different legislators "guarding" their well-being. These guardians may be expected to insure that troubled banks in their "protection domains" receive the same benefits as other banks in other domains, even if they are smaller in size than the other banks. As a result, ceteris paribus, the number of assumed TBTF banks may be expected to increase through time with an accompanying increase in societal costs, as occurred between 1984 and 1991. But the higher the visible costs, the greater is the likely opposition to a continuation of TBTF resolutions.

IV. BENEFITS OF TBTF POLICIES

Deposits in a failed bank to be protected under TBTF resolution may be expected to be those for which the societal benefits estimated by the regulators exceed the estimated societal costs. The largest benefit is likely to be the estimated reduction in collateral damage from what would have occurred if the protection had not been provided. A review of past TBTF resolutions, however, shows that these benefits have not been quantified very rigorously. Rather, they are sized by subjective broad generalizations. For example, the decision to protect all depositors and creditors of the Continental Illinois National Bank in Chicago when it failed in 1984 was justified by the regulators as follows. C.T. Conover, who was the Comptroller of the Currency at the time of the failure, testified before Congress at the time that

Had the Continental failed and been treated in a way in which depositors and creditors were not made whole, we could very well have seen a national, if not an international, financial crisis, the dimensions of which were difficult to imagine. None of us wanted to find out. (Conover 1984, 288)

Maybe they did not want to experience the actual crisis, but they should at least have tried to estimate the damage. John La Ware, a Governor of the Federal Reserve System, later testified before Congress that

The only analogy that I can think of for the failure of a major international institution of great size is a meltdown of a nuclear generating plant like Chernobyl. (LaWare 1991, 34)

It would have been highly unlikely that any estimated costs of providing the protection would have exceeded the estimated benefits of "saving the world" and protection not provided. As Federal Reserve Governor Daniel Tarullo observed

... No matter what its general economic policy principles, a government faced with the possibility of a cascading financial crisis that can bring down its national economy tends to err on the side of intervention. (Tarullo 2009, 2)

Moreover, the benefits are recognized immediately and decline through time, while the recognized immediate costs are small but increase through time. Today triumphs over tomorrow! This is a classic time inconsistency problem in economics. Thus, TBTF resolutions were applied to large banks far more often than not. (2)

But, through time, the regulators' estimates of such large benefits raised doubts, were increasingly questioned, lost considerable credibility, and were implicitly downsized. At the same time, the increasingly large delayed costs became more visible. The effective benefit-cost ratio declined through time after the TBTF resolution was applied. At the time of a failure, benefits tend to be overestimated and costs underestimated and the greater the crisis, the larger the perceived collateral damage and the greater the overestimate of the benefits of protection.

Another benefit of TBTF resolutions accrues to the regulators. Recognizing a bank insolvency and failing the bank is politically unpopular and depicts regulators in a bad light. They are charged to keep banks safe and open and are evaluated by doing so. By hiding losses and keeping insolvent banks open and operating, TBTF provides political cover for regulators from public criticism, at least in the short run. Likewise, as failing a bank reduces the popularity of regulators among bankers, not failing a bank or delaying the decision does not necessarily reduce the likelihood of the regulator being able to use the revolving door in the future to gain employment with a bank at considerably higher compensation.

V. ALTERNATIVE TBTF RESOLUTION REGIMES

TBTF resolution regimes differ according to the type and amount of depositor/creditor claims protected. On the one extreme, no accounts are protected. This may be termed a no-TBTF resolution. On the other extreme, all depositor/creditor claims are protected. This may be termed a 100%-TBTF resolution. But, depending on the benefit-cost trade off, regulators may wish to protect some but not all depositor/creditor claims, for example, demand deposits but not time deposits. Demand deposits may be viewed as a more important part of the payments system than are time deposits and losses to them may do more economic damage than do equivalent losses to time deposits. Protecting some but not all claims in a failure resolution may be viewed as a hybrid TBTF resolution. Thus, TBTF resolutions are not only black and white but come in different shades of gray.

Hybrid resolutions are frequently misclassified as no-TBTF resolutions. For example, ex-Congressman Barney Frank, the former Chairman of the House Banking Committee and the coauthor of the Dodd-Frank Wall Street Reform and Consumer Protection Act that, among other things, was intended "to end too big to fail," has argued that in a no-TBTF (no protection) resolution

... shareholders are wiped out, the CEO is fired, and the institution no longer exists, and the regulators may, at that point, the FDIC, pay off some debts if it is necessary to prevent a downward spiral, but any penny paid out must be recouped from the large financial institutions. (Frank 2011, emphasis added).

As some creditor claims are protected, I consider this is a hybrid-TBTF resolution, not a no-TBTF resolution. However, Frank appears to assume that because protecting these claims avoids a damaging downward spiral in the economy, the protection is a necessity, not arbitrary.

Likewise, Fed Governor Daniel Tarullo, who is in charge of bank supervision and regulation at the Board, has argued that

A fourth reform, intended to ensure that no firm is too big to fail, was the creation by Dodd-Frank of orderly liquidation authority. Under this authority, the FDIC can impose losses on a failed institution's shareholders and creditors ... while avoiding runs by short-term counterparties.... [F]or the resolution mechanism to be credible ex-ante and effective ex-post, the capital and liability structure of major firms must be able to absorb losses without either threatening short-term funding liabilities or necessitating injections of capital from the government. (Tarullo 2012, 2, emphasis added)

But the only way short-term counterparties will not run when the solvency of a bank is in doubt is if they were expected to be protected. Again, this represents a hybrid TBTF resolution, not a no-TBTF. Tarullo is also confusing a no-TBTF resolution with a hybrid resolution, in which stability in short-term claims is given great weight and viewed as a necessity. Thus, no-TBTF resolutions are not defined equally by everyone.

Moreover, differences in definition of TBTF resolutions also arise depending on who finances any protection provided. Some argue that protection is TBTF only if it is financed by the government. But Peter Wallison, a frequent commentator on financial issues, argues that

... the source of funds for a bailout is not the real issue. The possibility of a creditor bailout creates moral hazard, no matter where the bailout funds originate. and it is moral hazard that provides the largest batiks or other large financial firms with competitive advantages. (Wallison 2012, emphasis added)

Thus, as long as any funds are shifted from third parties to depositors/creditors to compensate them for losses, even only if private sources and not the government is the provider, the resolutions are effectively considered TBTF resolutions.

Differences in definition generate confusion as to who is being protected by TBTF, by whom, and for what reason. As different observers may see different parties as being protected for the same named TBTF resolution, the task of reaching agreement on how to modify or end TBTF altogether becomes more difficult and awaits greater clarity and agreement in definitions.

VI. WILL DFA KILL TBTF?

A primary objective of the DFA is to kill TBTF. But past experience suggests that like vampires, even if killed, TBTF may not stay dead for very long unless killed with a wooden stake or equivalent (depending on the story) stabbed through the heart. Does DFA give the regulators the right cross and can the regulators find TBTF's heart to succeed? That would depend. Arguments can be made on either side.

Two arguments suggest that TBTF may be killed for good. In its Title 2, DFA establishes the OLA to resolve the failure of select troubled large bank holding companies and nonbank financial firms, whose resolution by the FBC would threaten financial stability. DFA gives the FDIC, as receiver, the ability to protect some uninsured and unsecured claimants, but only if doing so would reduce the resolution cost to the FDIC. However, the earlier enacted FDICIA permits the FDIC to also protect claims at chartered banks if doing so reduces the adverse effects on financial stability. OLA does not permit this. Thus, OLA provides one less rationale for providing protection to large nonbank financial firms and limits the potential loss to the FDIC. More importantly, OLA permits regulators to close (place in receivership) covered firms sooner than under the FBC. The regulators can do so when they perceive a threat of default by the firm rather than waiting for an actual default. The faster a closure, the smaller are any losses likely to be and the less reluctant would regulators be to close the firm.

On the other hand, it should not be very difficult for regulators to argue that by reducing potential financial instability through providing additional protection, the cost to the FDIC of resolving a firm is lowered. Thus, they are likely to do so and effectively apply a TBTF resolution. Likewise, the DFA does not increase the incentives of regulators to not only "talk-the-talk" but also to "walk-the-walk" in implementing sanctions and denying protection and TBTF resolution to covered institutions. Thus, there is little to lessen the severe principal-agent problem between taxpayers and the regulators that contributed significantly to the recent and also earlier banking crises. Regulators did not have the political will to take strong unpopular actions that could reduce the cost of resolutions. It is the lack of political will by regulators, not the lack of appropriate regulations, that makes killing TBTF difficult and permits its resurrection when killed (Kaufman and Malliaris 2010). Prompt corrective action and LCR provisions in FDICIA look better on paper than in action. Lastly, in crises, regardless of the letter or spirit of the law, policy-makers are likely to act to minimize their own short-term costs. Short-term survival is key. Ad-hocery is likely to triumph! TBTF by any name will survive.

VII. HOW DOES DFA AFFECT RESCUES (TBTNR)?

In 2008, rescues were used in large volume to provide assistance to many financial institutions through both direct Treasury funding and a broadened Federal Reserve discount window (Section 13(3) lending) (Garcia, forthcoming). Title 11 of DFA restricts the Fed's use of its discount window to provide emergency only through broadly based programs rather than directing the aid to individual institutions. It must also receive prior approval from the Secretary of the Treasury for such assistance. Again, the regulators' will to accept these restrictions during a future crisis may be questioned. Lastly, the DFA requires large complex and systemically important institutions to develop a road map or "living will" that shows how the institution is internally organized and how, if it became insolvent, it could be resolved efficiently. This plan must be approved by the regulators. If accurate, this knowledge should make both rescues and failure resolutions more efficient. But the criticisms by the regulators of the first round of wills submitted by the largest banks suggest that there may be a long and slow learning process (Board of Governors 2014).

VIII. WHY REVIEW THE PAST TO DEVELOP REFORMS FOR THE FUTURE

Every crisis brings forth calls for reforms. But, it must be remembered the "reforms" are only changes that one likes. Changes that one does not like are "deforms." To design reform solutions correctly, one need first to correctly identify the problems(s). Getting it right requires knowing what went wrong.

The Spanish-American philosopher George Santayana (1863-1952) warned us that "those who cannot remember the past are condemned to repeat it." This is an optimistic and hopeful view. It suggests that while the past was not always optimum, the future could be made better. What is required to develop the right reforms is to study and learn from history. But what do those policymakers who do remember history do? Unfortunately, I believe more pessimistically that those who can remember the past are condemned to agonize first and then to repeat it.

IX. CONCLUSION

Technically, TBTF describes a resolution regime in which some or all depositor/creditor claimants of failed large firms, mostly banks and nonbank financial firms, are partially or totally protected against loss from insolvency (negative net worth) by the government in order to reduce collateral damage to other firms and the economy from increased financial instability. But TBTF can incur significant societal costs. Reducing the probability and magnitude of loss to these claimants permits large firms to obtain funding at lower cost. This both incentizes the affected banks to increase their risk exposures and gives them a competitive advantage over smaller no-TBTF banks. The benefits of reducing collateral damage tend to be perceived to be large at the time of a crisis and decrease over time thereafter. But the regulators' estimates of the collateral damage caused by not providing protections are generally highly subjective and not rigorously developed. They tend to overstate the actual reduction in collateral losses from the protection. The costs, on the other hand, are small at first, but increase over time. This results in a strong bias toward providing protection in a TBTF resolution.

But resolutions need not only protect either all claimants or no claimants. Some resolutions may protect only accounts for which the benefits are perceived to exceed the cost but not others for which the benefits are smaller and the costs higher. These may be classified as hybrid-TBTF resolutions. Hybrids are often misclassified. Hybrids in which accounts are protected, because the benefits are perceived to be much greater than the costs, are not infrequently classified incorrectly as no-TBTF institutions. Likewise, protecting some accounts that are perceived to provide critical services is frequently not considered a TBTF resolution. Thus, there is a disagreement about the definition of TBTF. One person's TBTF resolution may be another's no-TBTF resolution. It may be possible to "kill" TBTF by defining it away.

The article also differentiates between resolutions and rescues. In rescues, the firm is not killed. Equity support is provided by the government generally in exchange for an ownership stake and managerial control. As the firm is kept solvent, all depositors/creditors are fully protected. In failure resolutions, net worth is negative. The firm is killed. But some or all depositors/creditors are not. They may be partially or fully protected by the government (FDIC) in order to limit collateral damage. If perceived to be so protected, the firm while alive may obtain funding at lower cost, both giving it a competitive advantage over smaller firms that are not expected to be protected and incentizing it to increase its risk exposure. The benefits of being classified as TBTF tend to be seen first and the costs only later. Moreover, regulators tend to overestimate the benefits and underestimate the costs, particularly in crisis periods. This results in greater and more frequent use of TBTF resolutions. But as costs increase and become more visible, opposition has increased and the continuation of TBTF resolutions has become a major public policy issue. But TBTF has been difficult to kill for good, despite numerous attempts to do so. Like vampires, it is likely to haunt us for years to come.

ABBREVIATIONS

DFA: Dodd-Frank Act Wall Street Reform and Consumer Protection Act

FBC: Federal Bankruptcy Code

FDICIA: Federal Deposit Insurance Corporation Improvement Act

GAO: Government Accountability Office

GFC: Great Financial Crisis

LCR: Least Cost Resolution

OLA: Orderly Liquidation Authority

RFC: Reconstruction Finance Corporation

SRE: Systemic Risk Exemption

TARP: Troubled Asset Relief Program

TBTF: Too Big to Fail

TBTNR: Too Big to Not Rescue

doi: 10.1111/ecin.12169

REFERENCES

Acharya, V., D. Anginer, and A. J. Warburton. "The End of Market Discipline?" Working Paper, Stern School of Business, New York University, June 2014.

Arnold, M. "Archbishop of Canterbury Warns Banks Are Still 'Too Big to Fail'." Financial Times, June 18, 2014.

Artill, S., D. Hou, and A. Sarkar. "Components of U.S. Financial Sector Growth, 1950-2013." Economic Policy Review (Federal Reserve Bank of New York), March 2014.

Barth, J. R., G. Caprio Jr., and R. Levine. Rethinking Bank Regulation. Cambridge, UK: Cambridge University Press, 2006.

--. Guardians of Finance. Cambridge, MA: MIT Press, 2012.

Bliss, R. R., and G. G. Kaufman. "Resolving Insolvent Large Complex Financial Institutions: A Better Way." The Banking Law Journal, 128(4), 2011, 339-63.

Board of Governors of the Federal Reserve System and FDIC. "Agencies Provide Feedback on Second Round Resolution Plans of 'First-Wave' Filers." Joint Press Release, Washington, DC, August 5, 2014.

Caprio, G., Jr. "Financial Regulation after the Crisis." Special Paper 226, Financial Markets Group, London School of Economics. November 2013.

Conover, C. T. "Testimony." Inquiry into the Continental Illinois Corp. and Continental Illinois National Bank: Hearing before the Subcommittee on Financial Institutions Supervision, Regulation, and Insurance of Committee on Banking, Finance, and Urban Affairs. 98-111. U.S. House of Representatives, 98th Congress, 2nd Session, September 18-19 and October 4, 1984.

Dudley, W. "Solving the Too Big to Fail Problem." Speech, Federal Reserve Bank of New York, New York, November 12, 2012.

Financial Stability Board. Report: Progress and Next Steps towards Ending "Too-Big-to-Fail" (TBTF), September 2, 2013.

Frank, B. "Testimony." Subcommittee on Financial Institutions and Consumer Credit, U.S. House of Representatives (No. 122-37), Washington, DC, June 14, 2011.

Gambacorta, L., J. Yang, and K. Tsatsaronis. "Financial Structure and Growth." BIS Quarterly Review, March 2014,21-35.

Garcia, G. G. H. "The U.S. Financial Crisis and the Great Recession--Counting the Costs," in The Great Financial Crisis: A Retrospective, edited by James Barth and George G. Kaufman. Singapore: World Scientific and Now Publishers, forthcoming.

Kaufman, G. G., and A. G. Malliaris "The Financial Crisis of 2007-2009: Missing Financial Regulations or Absentee Regulators," in Lessons from the Financial Crisis, edited by R. W. Kolb. Hoboken, NJ: John Wiley & Sons, 2010, 337-44.

LaWare, J. "Testimony." Economic Implications of the "Too-Big-to-Fail" Policy: Hearing before the Subcommittee on Economic Stabilization of Committee on Banking, Finance, and Urban Affairs. U.S. House of Representatives, 102nd Congress, 1st Session, May 9, 1991.

Levine, R., N. Loayza, and T. Beck. "Financial Intermediation and Growth: Causality and Causes." Journal of Monetary Economics, 46(1), 2000, 31-77.

Reinhart, C. M., and K. S. Rogoff. This Time Is Different. Princeton, NJ: Princeton University Press, 2009.

Strahan, P. "Too Big to Fail: Causes, Consequences, and Policy Responses." Annual Review of Financial Economics, 5, 2013, 43-61.

Tarullo, D. "Confronting Too Big to Fail." Speech, Board of Governors of the Federal Reserve System, Washington, DC, October 21, 2009.

--. "Regulatory Reform since the Financial Crisis." Speech, Board of Governors of the Federal Reserve System, Washington, DC, May 2, 2012.

U.S. Government Accountability Office (GAO). "Large Bank Holding Companies: Expectations of Government Support." GAO-14-621. GAO, Washington, DC, July 2014.

Wallison, P. "Too Big to Ignore: The Future of Bailouts and Dodd-Frank after the 2012 Election (Policy Studies)." American Enterprise Institute, Washington, DC, October 24, 2012.

(1.) A recent study by the Government Accountability Office (GAO) found that large bank holding companies had a funding advantage over smaller bank holding companies during the financial crisis, but the advantage appears to have disappeared afterwards (U.S. Government Accountability Office 2014).

(2.) A good review of TBTF appears in Strahan (2013).

GEORGE G. KAUFMAN, This article is a revised version of ray presidential address to the Western Economic Association in Denver, Colorado, on June 29, 2014.

Kaufman: John Smith Professor of Finance and Economics, Department of Finance, Loyola University Chicago, Chicago, IL 60611. Phone 312-915-7075, Fax 312-915-8508, E-mail gkaufma@luc.edu; Consultant, Federal Reserve Bank of Chicago, Chicago, IL