Larry Summers, who was one of President Obama’s key economic advisors when the Dodd-Frank Act of 2010 was enacted, recently decried what he called “excessive populism” in portions of that legislation. This might seem surprising; Dodd-Frank’s technocracy-on-steroids approach (848 pages! 390 separate rulemaking requirements!) might seem like the antithesis of bust-up-the-banks populism. “My administration is the only thing between you and the pitchforks,” President Obama once famously told the nation’s leading bankers.
But Summers was referring to several specific Dodd-Frank provisions that curtailed the federal government’s financial rescue powers. During the financial crisis of 2007-2008, the Federal Reserve, the Treasury Department, and the FDIC relied to a large extent on a set of long-dormant statutory provisions to create an array of programs to stem the panic and stabilize the financial system. Dodd-Frank carved back these emergency powers in significant ways. Summers is not the only critic of these carve-backs. Ben Bernanke, Hank Paulson, and Tim Geithner—the key orchestrators of the U.S. crisis response—are on record lamenting Dodd-Frank’s restrictions on the federal government’s crisis-fighting toolkit. “If you want peace, prepare for war,” they write.
What should the government’s financial-crisis-response toolkit consist of? How should we think about its optimal scope and design? In Guarantor of Last Resort, Kate Judge offers a novel perspective on these questions. At a high level she agrees with Summers, Bernanke, Paulson, and Geithner that the existing toolkit is inadequate. In this respect she joins a number of other legal scholars and commentators. For example, Hal Scott has argued forcefully that the central bank needs robust liquidity support powers and that Dodd-Frank unwisely restricted them. Jeffrey Gordon and Christopher Muller have advocated creating a “Systemic Emergency Insurance Fund” to provide capital support to the financial sector during a panic. And Eric Posner has proposed expanding the federal government’s financial emergency powers to include among other things the power to make unsecured loans and equity investments.
Judge has a different angle on these questions. She proposes creating a new Emergency Guarantee Authority (EGA) to complement existing crisis-response powers. The proposal is innovative and carefully engineered. The EGA would empower the Treasury Secretary to provide time-limited emergency guarantees to financial institutions to halt runs by their short-term creditors and other claimants. Unlike the traditional lender of last resort, the EGA would not be constrained by a solvency requirement. The EGA would be designed to stop the panic and then give policymakers time to resolve the underlying weaknesses that gave rise to the crisis. While the Treasury Secretary would have significant discretion in implementing the guarantees, Judge envisions procedural constraints and reporting requirements to bring a measure of transparency and legitimacy to the exercise of this authority. And locating the EGA within the executive branch would promote democratic accountability.
Would creating an EGA be a good idea? To avoid impairing market discipline, Judge intends for the EGA to be used only for risks that threaten the functioning of the financial system as a whole, rather than for idiosyncratic risks at specific institutions. But the distinction between these things is not always clear.
In retrospect, most people agree that the Lehman Brothers bankruptcy presented a serious risk to the financial system as a whole. But this was not entirely obvious ex ante. At the time of Lehman’s failure many experts thought the U.S. financial system could withstand the firm’s collapse. David Wessel’s book In Fed We Trust describes a conference call between senior government officials just before Lehman’s bankruptcy: “With Lehman clearly struggling for survival, Paulson and Bernanke assured each other—and others on the call—that all the companies and traders that did business with Lehman had been given time to protect themselves from a possible Lehman bankruptcy.”[1] The day after Lehman’s bankruptcy, Ken Rogoff—among the world’s leading experts on financial crises—wrote an op-ed titled “No More Creampuffs.”[2] He applauded regulators for letting Lehman fail and “forc[ing] some discipline onto the system.” (To be fair, Rogoff acknowledged that “the risks are very real” and that “there really is no telling where the unprecedented failure of a big investment bank might lead”—but this is exactly my point.) Another prominent economist, Vincent Reinhart, opined that same day that “Lehman did not cast a long enough shadow over markets to warrant support.”[3] It is easy to identify risks in hindsight, much harder ahead of time.
Because a large, interconnected financial institution’s failure may imperil the financial system as a whole, such an institution would seemingly always be eligible for EGA support if on the brink of default—simply by virtue of its size and interconnectedness. The distinction between “idiosyncratic” and “systemic” comes close to collapsing in these cases. The larger and more interconnected the institution, the more likely it will get a lifeline. And if the likelihood of support is an increasing function of size and complexity, firms have incentives to get bigger and more complex. This is a problem not just for Judge’s EGA proposal but for any system of discretionary public support for financial firms.
This raises the question of moral hazard, which Judge views as probably the most serious objection to EGA.[4] Guarantees induce risk-taking because someone else bears some or all of the downside. Judge responds, sensibly, that the moral hazard question needs to be considered against an appropriate baseline—one in which “market participants already expect significant government support in the event of systemic distress.” The breadth and structure of EGA, she contends, “might actually make it easier for regulators to take the chance of allowing an institution to fail” because the government could use the EGA to limit knock-on effects. Thus the EGA “may well reduce the aggregate moral hazard in the system.”
I’m not persuaded that the creation of a new guarantee authority would stand much of a chance of reducing moral hazard (though Judge’s proposal hardly turns on this question). This is especially true given that Judge cites the EGA’s “breadth”—including its “capacity to address problems that arise outside the regulated sphere”—as one of its key virtues in relation to the existing toolkit. While it is no doubt true that adding a new tool may affect how existing tools are used, it would be hard to convince me that adding a new guarantee facility to the mix would not add to moral hazard, at least to some degree. It seems likely that, say, a large and highly leveraged hedge fund would welcome the EGA’s adoption as an additional potential governmental lifeline in the event of distress. If prime brokers agreed with this assessment, the hedge fund would presumably receive cheaper and less risk-sensitive funding, as well as favorable derivatives pricing (owing to lower counterparty credit risk) and greater leeway for risk-taking.
We can draw on history here. In the late 1980s and early 1990s the securities industry lobbied hard for legislation that would give them access to Fed emergency loans.[5] The chairman of the Securities Industry Association, who was a Goldman Sachs partner, testified in a Senate committee hearing in April 1990.[6] Among other things, he argued that the very existence of this power would make its use unlikely: “Frankly, if the banks note that the Fed has the power to do that, we don’t expect it would ever be used because, with proper collateral, they will answer the phone and make the loan if they know the Fed is going to do it if they don’t.” Wall Street’s argument prevailed. In 1991 Congress greatly expanded the Fed’s ability to lend to nonbanks.
What were the consequences? These powers were tapped on a staggering scale in 2008-2009, of course. But consider also the incentive effects. Theory predicts that enhanced access to backstops will encourage growth, risk-taking, and reliance on unstable (but cheap) short-term debt funding. The resulting funding subsidies should also boost profitability and compensation levels. The financial sector’s post-1991 trajectory does not contradict these predictions. According to one study, U.S. “securities industry output” accelerated sharply in the early 1990s.[7] Nondeposit short-term debt markets entered a period of explosive growth at around the same time. In addition, according to one study, “Workers in finance earn[ed] the same education-adjusted wages as other workers until 1990, but by 2006 the premium [was] 50% on average. Top executive compensation in finance follows the same pattern and timing, where the premium reache[d] 250%.”[8] The 1991 liberalization of the Fed’s financial rescue powers may have played a big role in all these trends.
Moral hazard, then, is one component of a broader set of issues that arise when the financial sector becomes increasingly entangled with the state; parallel problems of rent extraction, distortions in industry size and structure, and regulatory capture also arise. I fear the EGA (and other proposals of this genre) would make these problems worse, notwithstanding the EGA proposal’s intelligent design and incorporation of discipline and transparency. Democratic accountability may reduce adverse political consequences from the EGA’s exercise. But when it comes to emergency rescue powers, there is no escaping the inherent tradeoff between disaster prevention and market discipline.
Is this level of entanglement between the financial sector and the government simply the price we must pay for financial capitalism, as Judge seems to suggest? My answer will not surprise her; she and I have been debating these issues for years. To oversimplify (but only just a bit), financial stability reform proposals come in two types: structural and technocratic. Antitrust lawyers’ distinction between “structural” remedies and “behavioral” or “conduct” remedies is perhaps analogous. The EGA is technocratic: it takes the basic structure of finance as a given and seeks to deal with that structure’s dangers and pathologies without altering the underlying rules of the game. And it relies on shrewd policymakers occupying the commanding heights to make high-stakes decisions at crucial moments. As Bernanke, Paulson, and Geithner contend (see above), this is about “war” preparations. Broad and open-ended powers are therefore essential.
Structuralists (like me) are allergic to these kinds of open-ended powers. Emergency support should, wherever possible, be coextensive with some regulatory perimeter. Judge notes that the financial sector’s runnable liabilities (basically short-term and demandable debt) are a major reason emergency support powers like the EGA are needed in the first place. It was not so long ago that this distinctive liability structure was unique to regulated banks. But it has since migrated on a huge scale to other parts of the financial sector. As Judge has described masterfully in other scholarship, remaining passive in the face of this migration was itself a policy choice. Banking has virtually always been characterized by restricted entry, which means legally defining the liability structure that is permissible to chartered banks but off-limits to everyone else. This regulatory boundary eroded to a shocking degree in the decades preceding the crisis. A structuralist might say: Shore it up!
Judge rules out this particular structural approach; she thinks it is infeasible, on account of the inherent dynamism of finance. Maybe she is right, though I don’t think so. But the broader menu of options seems to me essential in evaluating the EGA proposal. If we are considering technocratic fixes in isolation, with the status quo as the baseline—in other words, if structural reforms are ruled out—then Judge’s EGA seems worthy of serious consideration as a way of fortifying existing crisis-response powers. But the calculus changes when structural reforms are deemed admissible: The right kind of structural reform might render open-ended discretionary backstops for the financial sector less appealing to begin with.
ENDNOTES
[1] David Wessel, In Fed We Trust, 11.
[2] Kenneth S. Rogoff, “No More Creampuffs,” Washington Post, September 16, 2008.
[3] Vincent Reinhart, “Secretary Paulson Makes the Right Call,” Wall Street Journal, September 16, 2008.
[4] Judge at 759.
[5] Since 1932 the Fed has had the power to lend to nonbanks under “unusual and exigent circumstances,” but before 1991 this power was subject to very strict collateral limits that made Fed loans practically unavailable to Wall Street securities firms.
[6] Deposit Insurance Reform and Financial Modernization: Hearings before the Committee on Banking, Housing, and Urban Affairs, US Senate, 101st Cong. 1:101 (April 3, 1990).
[7] Robin Greenwood and David S. Scharfstein, “The Growth of Finance,” Journal of Economic Perspectives 27, no. 2 (2013): 9.
[8] Thomas Philippon and Ariell Reshef, “Wages and Human Capital in the U.S. Finance Industry: 1909–2006,” Quarterly Journal of Economics 127, no. 4 (2012): 1551.
This post comes to us from Professor Morgan Ricks at Vanderbilt Law School.