Published May 1, 2008 | May 2008 issue
*The authors thank David Fettig, Art Rolnick, Phil Strahan, Dick Todd, David Torregrossa and Niel Willardson for their comments.
In this essay, we explain why the government’s response to the 2007–08 financial turmoil, although justified, expanded the safety net and exacerbated the existing too big to fail (TBTF) problem. A larger TBTF problem is costly, having the capability to sow the seeds of future financial crises, which means we should begin now to develop a new approach to manage TBTF.
We believe recommendations we had already crafted to address TBTF would effectively address the safety net expansion and position policymakers to respond more effectively to “the next Bear Stearns.” We describe the recommendations briefly and explain their relevance in today’s environment in the second half of the essay. Because our approach and recommendations are spelled out in our 2004 book, Too Big To Fail: The Hazards of Bank Bailouts, we conclude with excerpts from it summarizing our arguments in a bit more detail.
The Federal Reserve’s expansion of the safety net was not subtle or implied. The Federal Reserve took on risk normally borne by private parties when it supported JPMorgan Chase’s purchase of Bear Stearns. The Federal Reserve also opened the discount window to select investment banks (i.e., primary dealers).
One could describe the former action as one-time and the latter program as temporary. But such a characterization obscures the message these actions send. Through these efforts, the Federal Reserve sought to limit the collateral damage or spillovers caused by the failure of a large financial firm. And these spillovers can take many forms. In a simple example, the failure of a large financial firm means that other large financial firms might not have loans paid back or otherwise receive funds owed to them by the failing entity. In another case, the failure of a large financial firm could prevent it from providing critical services to financial market participants such as clearing and settlement of financial transactions. In both examples, the shock to financial firms could impair their normal operations, which could injure their customers and the rest of the economy. If the threat of such spillovers presented itself again, and spillovers frequently define a financial crisis, many large-firm creditors would anticipate another extraordinary action or resurrection of a special lending program.
To be sure, Bear Stearns’ equity holders—including many employees of the firm—took significant financial losses. This was an appropriate outcome. And doesn’t this action sufficiently curtail expectations of government support in the future and thus fix whatever problem such expectations create? The short answer is no. The long answer requires a brief summary of why we care about safety net expansion and TBTF in the first place.
The bigger the government safety net, the more the government shifts risk from creditors of financial firms to taxpayers. With less to lose, creditors have less incentive to monitor financial firms and to discipline risk-taking. Consider an extreme but simple case where nominally uninsured depositors at the largest U.S. commercial banks come to expect complete government support if their bank fails. These depositors have essentially no reason to pull their funds even if these banks take on so much risk that they doom themselves to failure.
Now, this dulling of the depositors’ senses has the welcome effect in our example of stopping runs on the largest banks. Such runs can spread into panics and significant economic downturns. The prevention of such ill effects, as noted, motivated the Federal Reserve’s safety net expansion and is the reason government support during a crisis should never be categorically ruled out.
But the same stickiness of deposits has a major downside, which is the point of our example. The large bank that fleeing depositors would otherwise close remains open to continue or increase its risky bets. If it does not get lucky, the bank’s losses actually grow. In this way, the safety net encourages risk-taking that exposes society to increasing losses, with their associated instability.
Of equal concern, TBTF wastes society’s resources. Financial firms allocate capital, and when they work well, they ensure that high-return projects are funded. But excessive government support warps that allocation process, sending too much money to higher-risk projects.
We focused deliberately on depositors in our example; we could have mentioned other short- or long-term holders of interest-bearing investments, insured or uninsured. For it is the reduced vigilance of depositors and other debt holders—lulled by implied government support—that leads large financial institutions to take on too much risk and underlies TBTF. Policymakers face a TBTF problem even if equity holders fully expect to suffer large losses upon failure of the firm in question.
And policymakers faced a TBTF problem even before recent safety net expansions; the TBTF problem we described in 2004 has grown since then.1 Some very large banks and financial firms (e.g., Countrywide Financial) faced significant pressure during the 2007–08 market disturbance. Reporting on these cases, sometimes months before the run on Bear Stearns, had at times explicitly raised the specter of government support. The initial rescue in 2007 and later nationalization of Northern Rock in 2008 by the British government may have contributed to the speculation. Nationalization occurred in a country viewed, like the United States, as having a low propensity to support uninsured creditors and involved a financial institution that supervisors did not apparently treat as if it posed significant systemic risk.
Our concern about the preexisting TBTF problem led us to suggest policy reforms, as detailed in our book. We now turn to summarizing our approach, explaining why it applies to the current situation and why it is preferable to other options.
While safety net expansion has increased TBTF concerns, the essence of the problem and underlying cause of TBTF have not changed since 2004: Policymakers support large-bank creditors to contain or eliminate spillover effects, but the support creates an incentive for too much risk-taking in the future. Our approach is straightforward. If spillovers lead to government support, then policymakers who want to reduce creditors’ expectations of such support should enact reforms that make spillovers less threatening. Reforms that fail to address this fundamental issue will not change policymaker behavior and will not convince creditors that they face real risk of loss. We provide more details on this approach in excerpted summaries from our book following this section.
So what should policymakers do to address concerns over spillovers? We recommend a three-pronged approach (again, a few more details follow in the excerpts with many more details in the book itself). Policymakers should
For each of the three strategies, we recommend that policymakers broadly communicate the actions they’ve taken to reduce expectations of bailouts. We detail the form and benefits of potential communication elsewhere, but the basic point is simple.3 Creditors will not realize that the spillover threats have declined and will not change behavior unless informed through effective communication.
Put together, this approach offers at least the potential for a positive cycle. Policymakers limit the need for government support by managing underlying sources of instability. Reduced expectations of government support lead to less risk-taking and greater stability.
Our approach contrasts with some other alternatives policymakers might adopt. Some observers suggest that policymakers try to manage the expanded safety net, for example, by extending rules that procedurally make it more difficult for policymakers to support creditors. For example, the Federal Deposit Insurance Corporation Improvement Act of 1991 (FDICIA) requires on-the-record support from a variety of policymakers before the FDIC can provide extraordinary support to bank creditors (FDICIA subjected such extraordinary support to other reviews and reforms as well). Policymakers might apply these strictures before providing support to creditors of any financial firm.
While we do not oppose expanding the types of firms covered under the FDICIA regime, we doubt the changes would materially reduce the support provided to large-firm creditors. Why? These procedural changes do not reduce the underlying reason policymakers provided support in the first place. Consider that the intervention with Bear Stearns involved the type of on-the-record voting and consultations across agencies that FDICIA would mandate.
Pledges of “no bailouts” from policymakers or general prohibitions against bailouts are even less credible unless accompanied by action. And such prohibitions and related jawboning are unwise. Policymakers will face circumstances where, even accounting for distortions to future behavior, the provision of government support has benefits exceeding costs.
Observers also suggest that enhanced supervision, or regulations like those found in Basel II, might curtail the risk-taking of financial firms. While supervision and regulation have an important role to play, these tools may not adequately curtail the risk-taking encouraged by TBTF. Supervisors with discretion, for example, cannot easily limit firm risk-taking before the damage is done. Minimum capital rules also seem one step too slow; that is, regulators cannot readily institute capital rules that link minimum capital levels to current bank risk-taking.
None of this is to suggest that our recommendations are beyond reproach. Some of the specific recommendations we made in 2004 deserve a second look given the events of 2007 and 2008. For example, we suggested that policymakers consider implementing a form of “coinsurance” for uninsured creditors, whereby such creditors must take some loss if their financial firm becomes insolvent. While our proposal differs from the use of coinsurance for insured depositors in England, some observers attribute part of the Northern Rock crisis to this feature, suggesting it deserves reconsideration.
Our recommendations have received more general critiques as well. Some critics focus on the inability of our recommendations, or any recommendations for that matter, to anticipate the source of the next major disruption. These observers argue that the idiosyncratic nature of each financial disruption means that policymakers can, at best, fight the last war and cannot take steps that limit future spillovers. Who could have foreseen, critics might ask, that losses originating in subprime mortgages would ultimately lead to a freeze in the secured funding markets on which Bear Stearns and others relied?
The manner in which Bear Stearns imploded certainly caught most observers and market participants by surprise. But it was no surprise that a failure of one of the largest U.S. investment banks posed spillover risks or raised TBTF concerns. Indeed, Paul Volcker, in the foreword to our book, raised a similar point.
The implications of [the TBTF book] … go beyond the world of commercial banking. Witness the officially encouraged (if not officially financed) rescue a few years ago of Long-Term Capital Management, a large but unregulated, secretive, speculative hedge fund. The fact is the relative importance of commercial banks in the United States has been diminishing steadily. Consequently, the lessons and approaches reviewed in Too Big To Fail have wider application.4
Moreover, we do not need to forecast the event that brings down systemically important firms to make progress against TBTF. Instead, we need to consider the spillovers that failure might cause. Would that failure, for example, eliminate the availability of important clearing and settlement services? If so, what can we do today to facilitate continued provision of those services? Would that failure impose large losses on other firms potentially seen as TBTF? If so, what actions today would help policymakers quickly quantify potential exposures and assess counterparties’ management of that risk? Of course, this approach is sure to miss some potential spillovers or risks. While not perfect, this approach is superior to efforts that do not focus on spillover potential or which react to instability once a firm fails.5
In conclusion, we think the recommendations we made several years ago have stood the test of time. They offer a structure and specific steps that policymakers can take to better manage the safety net and the TBTF problem. Due to its recent expansion, such safety net management should, in our view, take a considerably higher priority with policymakers than it has in the past.
1 See Stern and Feldman (2004). Mishkin (2006) provides a detailed summary and critique of our book. Analysis published after the book including, but not limited to, Morgan and Stiroh (2005), Rime (2005), and Deng et al. (2007) continues to find evidence of a TBTF problem. For Moody’s related assessment of the likelihood that select large banks in the United States would receive government support, see American Banker (2007). Acharya and Yorulmazer (2007) discuss a phenomenon somewhat similar to TBTF.
2 These two examples are discussed in Stern and Feldman (2006).
3 See Stern (2007) and Stern and Feldman (2005a, b).
4 See Stern and Feldman (2004, ix).
5 Without implying agreement between our proposal and more recent alternatives, other parties have also suggested that policymakers respond to safety net expansion by focusing on broad stability-related issues. For one example, see Nason (2008).
Viral Acharya and Tanju Yorulmazer. 2007. “Too Many to Fail: An Analysis of Time Inconsistency in Bank Closure Policies.” Journal of Financial Intermediation 1, 1–31.
American Banker. 2007. “Big-Bank Safety Net.” March 6.
Saiying Deng, Elyas Elyasiani, and Connie X. Mao. 2007. “Diversification and the Cost of Debt of Bank Holding Companies.” Journal of Banking and Finance 8, 2453–73.
Frederic S. Mishkin. 2006. “How Big a Problem Is Too Big to Fail? A Review of Gary Stern and Ron Feldman’s Too Big to Fail: The Hazards of Bank Bailouts. Journal of Economic Literature 4, 988–1004.
Donald P. Morgan and Kevin J. Stiroh. 2005. “Too Big To Fail after All These Years.” Federal Reserve Bank of New York Staff Report 220.
David G. Nason. 2008. “Remarks on Treasury’s Blueprint for a Modernized Regulatory Structure.” Press release. April 29.
Bertrand Rime. 2005. “Do ‘Too Big To Fail’ Expectations Boost Large Bank Issuer Ratings?” Presentation at Banking and Financial Stability: A Workshop on Applied Banking Research. Bank for International Settlement. Vienna, Austria, May 9.
Gary H. Stern. 2007. “Addressing the Trade-offs: Market Discipline, Stability and Communication.” The Region. December.
Gary H. Stern and Ron J. Feldman. 2006. “Managing Too Big To Fail by Reducing Systemic Risk: Some Recent Developments.” The Region. June.
Gary H. Stern and Ron J. Feldman. 2005a. “Constructive Commitment: Communicating Plans to Impose Losses on Large Bank Creditors,” in Douglas D. Evanoff and George G. Kaufman (eds.), Systemic Financial Crises: Resolving Large Bank Insolvencies (Hackensack, N.J.: World Scientific Publishing).
Gary H. Stern and Ron J. Feldman. 2005b. “Addressing TBTF When Banks Merge: A Proposal.” The Region. September.
Gary H. Stern and Ron J. Feldman. 2004. Too Big To Fail: The Hazards of Bank Bailouts (Washington, D.C.: Brookings Institution Press).