Forgot Password?
Getty Images
President Barack Obama greets Rep. Barney Frank (D-Mass.) and Sen. Chris Dodd (D-Conn.) after signing the Dodd-Frank Wall Street Reform and Consumer Protection Act at the Ronald Reagan Building on July 21, 2010, in Washington.
Tweet This
Tweet This
Tweet This
Magical thinking: the latest regulation from the Financial Stability Oversight Council Download PDF
Subscribe to the Financial Services Outlook series
The Financial Stability Oversight Council (FSOC), following directions in the Dodd-Frank Act, has been pursuing the idea that a financial crisis can be prevented by stringent regulation of large “interconnected” financial institutions. Its latest proposed regulation goes further than ever in establishing the importance of interconnectedness among financial institutions as the source of potential systemic instability. However, the events following the Lehman Brothers bankruptcy show that Lehman’s failure had no significant knock-on effects, casting doubt on the idea that Lehman-like nonbank financial firms are too big to fail. The financial crisis was caused not by Lehman’s failure but by a common shock to all financial institutions that were holding privately issued mortgage-backed securities based on subprime loans. The way to prevent future financial crises is to prevent future common shocks. The FSOC’s pursuit of the bogus interconnectedness theory will, if it results in the designation of certain nonbank financial institutions as systemically important financial institutions, impair our competitive financial system while failing to prevent another financial crisis.
Key points in this Outlook:
One of the key underlying ideas of the Dodd-Frank Act (DFA)—an idea that has gone viral in the community of financial regulators around the world—is that a goal of financial regulation after the financial crisis was to prevent “systemic risk.” But there is a major problem with this idea: one can always imagine a systemic collapse, but identifying the factors that might cause it—that is, systemic risks—is quite another matter. Shortly after the financial crisis, the idea became popular among regulators worldwide—appearing first in a report by the Group of Thirty in January 2009—that large financial institutions such as Lehman Brothers could cause a systemic breakdown if they failed.
Little evidence existed to support this idea, but government officials suggested—and the media widely reported—that these and other institutions were “interconnected,” so that the failure of one would cause systemic instability by bringing down others. This notion underlay the rescue of Bear Stearns in March 2008, and it gave rise to the idea that these firms—called systemically important financial institutions (SIFIs)—were too big to fail because their failure would cause another financial crisis. Accordingly, the argument ran, they should be identified and specially regulated and, if they still failed, wound down through a special resolution process outside the bankruptcy system.
That this would give regulators more power over the financial system as a whole should be duly noted; it clearly had a role in attracting support for the DFA, particularly at the Fed. In any event, the notion that that SIFIs existed became generally accepted, and the Obama administration quickly adopted the idea as it framed what became the DFA. However, as this Outlook will show, Lehman’s failure cast a different and unfavorable light on these ideas. Despite the chaos that followed, the collapse did not result in any significant knock-on effects, providing robust evidence of the DFA’s misplaced focus on designating and regulating SIFIs.
As the legislation that was to become the DFA moved through Congress, members of the Obama administration made grandiose claims about the promise of systemic risk regulation. For example, Michael Barr, then an assistant treasury secretary and still a believer in the idea, noted recently, “Today, while the regulatory infrastructure is far from ideal, with too many divided responsibilities, the Financial Stability Oversight Council (FSOC) is accountable to identify threats to financial stability and to address them.”1 The FSOC, in this view, would identify systemic threats in advance and take the necessary action to head them off.
"Without the interconnectedness idea, there is no need to identify SIFIs, no need for special regulation of nonbank financial institutions, and no need for a special resolution process to resolve large nonbank financial institutions."--Peter Wallison
Key sponsors of the DFA as it was being drafted made statements to the same effect. For example, Senator Chris Dodd (D-CT), the principal Senate sponsor of the DFA, said in a speech on the Senate floor on April 29, 2010, “We create an early warning system. . . . I think we all agree that to have the ability to watch and monitor what is occurring, both domestically and internationally, is very important. We have established what we call a systemic risk council that will allow us to observe what is occurring on a regular basis so we can spot these problems before they metastasize and grow into, as we have seen, problems that created as much harm for our economy as the present recession has.”2 Again, the underlying notion was that systemic risk was something that could be recognized in advance, confronted, and eliminated.
These ideas now appear to have been abandoned, with the administration taking the position that the objective of the FSOC is simply to identify and shore up SIFIs against failure, through stringent regulation. This became clear in an exchange between Treasury Secretary Timothy Geithner and Senator Richard Shelby (R-AL), the ranking member of the Senate Banking Committee, at a committee hearing on October 6, 2011. Senator Shelby asked whether the FSOC had made progress in identifying the threats to the stability of the financial system that had been advertised as a purpose of the council during the framing of the DFA. Geithner responded that this was not what the FSOC was doing: “The basic strategy that underpins the design of this [systemic risk] framework is not a strategy that depends on the wisdom and foresight of government officials. It is a strategy that relies on building much stronger shock absorbers, safeguards in the system so that we are protected against a whole range of potential risk.”3
In other words, the FSOC will not be seeking the scent of systemic risk but instead will simply be attempting to identify nonbank financial firms that are SIFIs and ensure that they are effectively regulated to prevent another systemic breakdown. Thus, the questions now are whether the DFA, with its focus on SIFIs, has created an effective structure for preventing a financial crisis in the future and, more particularly, whether the failure of a large financial institution can cause a financial crisis. Like so many other accepted ideas about the financial crisis, this idea has not been explored in any depth. In this Outlook, I will show not only that the interconnections theory is wrong, but that the absence of significant knock-on effects from Lehman’s bankruptcy demonstrates that nonbank financial institutions—even those designated as SIFIs—are highly unlikely to cause a financial crisis if they fail.
This means that there is no need to identify SIFIs, no need for special regulation of nonbank financial institutions, and no need for a special resolution process to resolve large nonbank financial institutions, an idea explored fully in my August-September Outlook.4 For all practical purposes, these issues are relevant only for nonbank financial institutions, the focus of the following discussion. The DFA has already determined that any bank or bank holding company with assets greater than $40 billion will be considered a SIFI. (This is the magic number because every Federal Reserve Bank district contains at least one bank or bank holding company of that size. The fundamental impulses of the regulatory bureaucracy never change, lending credence to the idea that the SIFI notion is yet another way for regulators to enlarge their authority.)
The Notice of Proposed Rulemaking (NPR) the FSOC promulgated on October 11, 2011, is fully consistent with Geithner’s statement about FSOC strategy. This NPR is the FSOC’s second attempt to clarify the standards and criteria it would use to identify SIFIs. The first was figuratively hooted down because it did little beyond repeating the statutory language. The new version, while more detailed, still misses the mark. It makes clear that the FSOC has not fully understood, or even seriously considered, what happened after the bankruptcy of Lehman Brothers and how that aftermath should affect its analysis.
“Common Shock” and the Financial CrisisThe August-September Outlook fully addressed the causes of the post-Lehman chaos, but I will summarize them here. According to data gathered by Professor Robert Shiller of Yale University, a massive ten-year housing price bubble began to develop in the United States in 1997.5 By 2007, when it began to deflate, the bubble was almost ten times larger than any previous bubble. Because of US housing policy, which mandated that, by 2008, mortgage credit be made available to borrowers at or below the median income level, half of all mortgages in this bubble—27 million—were subprime or otherwise weak and risky loans.6 As is true of all bubbles, this bubble’s growth suppressed delinquencies and defaults; people who could not meet their mortgage obligations were able to refinance and extend their obligations or sell their homes for more than the principal amount of the mortgage.
By 2002, investors were beginning to notice that these high-yielding subprime mortgages were not showing the delinquencies and defaults normally associated with such low-quality loans. In other words, they had very favorable risk-adjusted returns, setting off strong investor demand, especially among banks and other financial institutions, for privately issued mortgage-backed securities (PMBS) based on these subprime and other low-quality loans. This demand resulted in the securitization and sale of about 7.8 million subprime loans, backing almost $2 trillion in PMBS and totaling slightly less than 30 percent of the 27 million subprime and other risky loans outstanding. The rest of the loans were on the balance sheets of US government agencies such as Fannie Mae and Freddie Mac, which had made these loans in compliance with the US government regulations noted above.
When the housing bubble began to deflate in 2007, the unprecedented number of delinquencies and defaults on both the government-induced and the privately securitized mortgages drove down the values of houses nationwide, as well as the PMBS backed by subprime mortgages on these now-devalued homes. Investors, shocked by the unprecedented number of defaults, fled the PMBS market, dropping PMBS prices to near zero. Mark-to-market accounting then came into play for banks and other financial institutions, requiring them to write down the value of their PMBS holdings to market values. This substantially reduced their regulatory capital, and because PMBS could no longer be used for financing purposes the liquidity of these firms was also impaired. No wonder the media dubbed these securities “toxic assets.”
This produced what scholars know as a “common shock”—a condition in which a large number of financial institutions are hit by the sudden loss of value in a widely held asset.7 In this case, many of the world’s largest financial institutions were subject to this shock, and the writedowns in their assets required by mark-to-market accounting made them appear weak and unstable. The widespread instability among financial firms made it appear to investors and other market participants as though the whole financial system was disintegrating.
In March 2011, with investors and other market participants unsettled and near panic, the US government rescued Bear Stearns, the smallest of the large investment banks on Wall Street. This briefly calmed the markets, but it also caused enormous moral hazard by leading many firms to believe that they did not have to dilute their shareholders by raising significant amounts of additional capital. It also led many investors and other market participants to believe that the US government had established a policy of rescuing any large financial institution in financial distress. When the government did not rescue Lehman in September 2008, investors and market participants were shocked; financial institutions, fearing that investors and depositors would seek return of their funds, began to hoard cash; banks stopped lending to one another. This is what we know as the financial crisis.
The Administration and Congress Get It WrongThis brief history shows that three elements were at the heart of the financial crisis—a mortgage and housing price meltdown as the bubble deflated, mark-to-market accounting for financial institutions, and a resulting common shock to all financial institutions that held substantial amounts of PMBS. If the framers of the DFA had properly understood this sequence of events, the legislation would have been markedly different. They would have chosen to take steps to prevent another financial crisis by reducing the likelihood of common shocks and modifying mark-to-market accounting rules so as not to exacerbate any common shock that does occur. In a forthcoming paper, for example, Roberta Romano of Yale Law School posits that regulation itself, by reducing diversity among regulated institutions, could be a source of widespread breakdowns similar to a common shock.8 Exploring ideas of this kind, rather than racing willy-nilly to impose new regulations, would have resulted from a more careful look at what happened after Lehman’s bankruptcy.
The Obama administration and the framers of the DFA clearly did not see common shock as a cause of the financial crisis. Creating an agency (the FSOC) to designate certain nonbank financial firms as SIFIs and consign them to stringent Fed regulation makes sense only if the administration and its allies in Congress believed that the failure of Lehman was the cause of the financial crisis. To draw this conclusion, they uncritically accepted the idea—initially used to justify the rescue of Bear Stearns—that large nonbank financial institutions were interconnected and that the failure of one could drag down others.
Lehman—although unquestionably the kind of large nonbank financial institution that would be classified as a SIFI—did not cause any insolvencies among other financial institutions when it failed. The only institution that suffered serious losses was the Primary Reserve Fund, a money market mutual fund that continued to hold large amounts of Lehman’s commercial paper, in the mistaken belief—probably induced by moral hazard—that the US government would rescue Lehman as it had rescued Bear. The fund “broke the buck” (was unable to redeem its shares at the promised $1.00 per share) and in the panicky condition of the market at that time set off an unwarranted run on other money market funds.
Ironically, then, there was some real value in the catastrophe of the Lehman collapse. It provided strong empirical evidence that a large nonbank financial institution can fail without dragging down others. Previously, in cases like those of Bear Stearns and Long-Term Capital Management, the government—contending that the collapse of large financial institutions would bring on a financial meltdown—had taken or encouraged others to take steps to rescue large, failing financial institutions. The resulting severe moral hazard made the too-big-to-fail problem worse. It is worth noting at this point that the government’s determination that a financial institution could not be allowed to fail is not evidence that this is true. Regulators are notoriously fearful that they will be blamed for chaotic conditions in the market they regulate and thus lean heavily in favor of bailouts and rescues."If an agency like the FSOC is to have any value for preventing systemic risk, it would be to spot the cases where the risk of a common shock is developing."--Peter Wallison
Even the credit default swap (CDS) market—cited by many as the very “shadowy” heart of the interconnections problem—continued to function normally after Lehman’s failure. Lehman was a major player in the CDS market, and at the time of its bankruptcy had 900,000 outstanding CDS contracts. Most of these were canceled by Lehman’s counterparties—as provided in bankruptcy law—and the remaining ones became claims against the bankrupt estate. CDSs written on Lehman itself were settled five weeks after its bankruptcy by the exchange of less than $6 billion among hundreds of counterparties. As far as we know, no Lehman counterparty failed because Lehman failed or could not make good on its CDS obligations.
Some will argue that the steps the US government took after the Lehman collapse prevented other failures. The Troubled Asset Relief Program (TARP) and the Federal Deposit Insurance Corporation’s (FDIC) guarantee of loans to or from financial institutions are examples of this assistance; certainly they helped stabilize the financial system in that chaotic period. However, the goal and result of these efforts was not to shore up the finances of financial institutions weakened by Lehman’s inability to meet its obligations, but rather to restore market confidence. This confidence had been lost in the common shock brought on by the mortgage meltdown and was further eroded by the government’s irrational failure to rescue Lehman after it had rescued Bear.
That the government’s actions were aimed only at restoring market confidence is shown by the fact that the first funds made available under TARP were not provided until October 28, 2008, about six weeks after Lehman’s failure. By that time, any firm that had been mortally wounded by Lehman’s collapse would have filed for bankruptcy. Moreover, most of the TARP funds were repaid by the largest institutions—with Fed approval—only eight months later, in mid-June 2009. If the firms that took these funds had needed them to cover losses on Lehman, they could not have recovered from those losses in only eight months. Similarly, the FDIC’s guarantees allowed otherwise solvent firms to borrow but did not fill in holes in their balance sheets caused by losses on Lehman.
So the interconnections theory is simply wrong. Interconnections undoubtedly exist among financial institutions, but it appears that they are not so significant that the failure of a large nonbank financial institution like Lehman will drag others down. This evidence is especially robust because Lehman’s failure occurred at a time when the entire financial system was in the midst of a widespread common shock, when the failure of one large institution like Lehman could have been expected to have the most devastating effect. If Lehman’s failure did not have knock-on consequences then, it is highly unlikely that it would have any serious consequences during more stable times. From this, we can conclude that nonbank financial institutions are not as a general matter too big to fail and should be allowed to go into bankruptcy like any other insolvent institution.
Accordingly, if an agency like the FSOC is to have any value for preventing systemic risk, it would be to spot the cases where the risk of a common shock is developing. By recognizing these conditions, such an agency could take appropriate action. However, the latest NPR shows that the FSOC is barking up the wrong tree by pursuing its misguided interconnections theory.
The Magical Thinking of the FSOCThis is easily demonstrated by an analysis of the NPR, where the interconnections idea has a central role. The proposal creates a three-stage “determination process” for deciding whether a nonbank company is a SIFI and requires stringent Fed regulation. Stage 1 narrows the search to a smaller set of companies by using quantitative thresholds (for example, size) to identify the nonbank financial institutions that deserve further scrutiny. In Stage 2, “the Council will conduct a comprehensive analysis of the potential for the identified nonbank financial companies to pose a threat to U.S. financial stability.”9 The firms that remain after this “comprehensive analysis” will then be more thoroughly examined, with the use of “quantitative and qualitative information collected directly from the nonbank financial company” in Stage 3.
In Stage 3, the FSOC will apply what it calls an “analytical framework” consisting of six factors: size, interconnectedness, substitutability, leverage, liquidity risk and maturity mismatch, and existing regulatory scrutiny. It is immediately clear that in this framework only size, interconnectedness and substitutability are important for the SIFI determination. The other three factors—leverage, liquidity risk and maturity mismatch, and existing regulatory scrutiny—are important only after size, interconnectedness, and substitutability set the institution apart as a potential SIFI. As the proposed regulation says, “leverage, liquidity risk and maturity mismatch seek to assess the vulnerability of a nonbank financial company to financial distress.”10 Thus, a firm that has fifty-to-one leverage would not be considered a SIFI—despite its excessive leverage—unless that leverage level might create some threat to the stability of the financial system, which would be realistic only if it had one or more of the other three defining factors: size, interconnectedness, or substitutability. For purposes of this analysis, however, we can leave out substitutability, which refers to the ability of an institution to provide a critical function or unique service to the market—by definition a limited category. If the lack of a substitute for an institution’s services is a problem, Congress could authorize the Fed to decide whether the market could function without these services, and if not, after appropriate safeguards such as court review, the Fed could impose the regulation it believes is necessary.
So we are down to interconnections and size. We know what size is, but what are “interconnections,” and why are they important? This term clearly refers to mutual obligations of various kinds. These could be credit arrangements—with one institution a creditor of another—or to counterparty arrangements, in which institutions are obligated or exposed to one another through a credit default swap or similar transaction. Exposure is important because the NPR posits only three “channels” through which a SIFI’s “material distress” might be transmitted to the financial system as a whole: exposure, asset liquidation, and furnishing a critical function or service. Again, we can eliminate the final criterion and focus on exposure and asset liquidation as the two ways that a SIFI’s material distress could affect the financial system at large.
The NPR argues that the material distress of a failing firm may cause losses to other firms because of the interconnections that are associated with exposure: “A nonbank financial company’s creditors, counterparties, investors, or other market participants have exposure to the nonbank financial company that is significant enough to materially impair those creditors, counterparties, investors, or other market participants and thereby pose a threat to U.S. financial stability” [emphasis added].11 This is the classic interconnections scenario but does not reflect in any way what actually happened when Lehman failed.
In describing the adverse effects of asset liquidation, the NPR outlines the following hypothetical situation: “A nonbank financial company holds assets that, if liquidated quickly, would significantly disrupt trading or funding in key markets or cause significant losses or funding problems for other firms with similar holdings due to falling asset prices.”12 In other words, a SIFI’s material financial distress could cause it to liquidate assets. This in turn would have significant adverse effects on the value of the assets of other firms. Again, this is not what happened when Lehman was in danger of failing; if it did sell assets, its sales were not significant nor what caused PMBS values to decline. Indeed, it was the earlier decline in the value of these assets that caused Lehman (and others) to look weak, illiquid, and possibly insolvent.
It is worth noting, however, that the asset liquidation scenario acknowledges that the loss of value in widely held assets can have a substantial adverse effect on other firms—the very definition of a common shock—but does not seem to recognize that a common shock arising out of the collapse of the PMBS market well before Lehman failed was the triggering event of the financial crisis.
Remarkably, the FSOC also recognizes that a SIFI’s failure could be more destructive in an environment created by a common shock. “For purposes of considering whether a nonbank financial company could pose a threat to the U.S. financial stability,” the NPR states, “the Council intends to assess the impact of the nonbank financial company’s material financial distress in the context of a period of overall stress in the financial services industry and in a weak macroeconomic environment. The Council believes this is appropriate because in such a context, a nonbank financial company’s distress may have a greater effect on U.S. financial stability.”13
Yet, even though the Council apparently understands the significance of a “weak macroeconomic environment” in determining the likelihood that a failed or failing SIFI would create a financial crisis, it did not reach the obvious conclusion—that in the Lehman case, a SIFI had failed in the midst of a seriously distressed macroeconomic environment without any knock-on effects. Instead, it continues to pursue an idea based on nothing more than rank speculation. This is the very essence of magical thinking—pursuing an idea that has no evidentiary or empirical support while ignoring obvious facts that show it is heading in the wrong direction."If the FSOC continues on this course, it will make no contribution to preventing a future financial crisis while causing serious harm to the competitiveness of the US financial system."--Peter Wallison
In summary, if we examine what actually happened after Lehman, we can see that neither the exposure of counterparties to Lehman nor its asset liquidation caused substantial losses to other firms—or, at least, losses that caused those other firms (with the exception of the Reserve Fund) to become materially impaired. Because Lehman’s failure had no significant knock-on effects, the only conclusion we can reasonably draw is that nonbank financial institutions, even those large enough to be SIFIs, can be allowed to fail without danger to the financial system. They are not, therefore, too big to fail.
Nevertheless, the FSOC plunges ahead. If it continues on this course, it will make no contribution to preventing a future financial crisis while—for reasons outlined below—causing serious harm to the competitiveness of the US financial system.
(Wrong) Ideas Have (Bad) ConsequencesIn September 2009, former Federal Reserve chairman Paul Volcker appeared before the House Financial Services Committee and declared his opposition to designating specific firms as SIFIs:
The approach proposed by the Treasury is to designate in advance financial institutions “whose size, leverage, and interconnection could pose a threat to financial stability if it failed.” Those institutions, bank or non-bank, connected to a commercial firm or not, would be subject to particularly strict and conservative prudential supervision and regulation. The Federal Reserve would be designated as consolidated supervisor. The precise criteria for designation as “systemically important” have not, so far as I know, been set out. However, the clear implication of such designation whether officially acknowledged or not will be that such institutions, in whole or in part, will be sheltered by access to a Federal safety net in time of crisis; they will be broadly understood to be “too big to fail.”14
This testimony by the respected Volcker failed to dissuade the committee or Congress from empowering the FSOC in just this way. Volcker’s principal concern expressed that day was that the designation of nonbank financial firms as SIFIs would extend the financial safety net beyond banks. He added, “What all this amounts to is an unintended and unanticipated extension of the official ‘safety net,’ an arrangement designed decades ago to protect the stability of the commercial banking system. The obvious danger is that with the passage of time, risk-taking will be encouraged and efforts at prudential restraint will be resisted. Ultimately, the possibility of further crises—even greater crises—will increase.”15
However, a more important issue exists. Volcker is completely correct that the government’s designation of certain nonbank financial institutions as SIFIs will mean that they will have been declared too big to fail. Despite the lack of obvious justification for such a policy, it will clearly signal to the world that the government will take steps to prevent the failure of these special firms. This will give them a funding advantage over their smaller competitors because creditors will immediately recognize that their loans to SIFIs are safer bets than loans to others, as government regulation will reduce their risk taking and be more likely to bail out their creditors if they fail.
These funding advantages have already appeared in banking, where differences of fifty to eighty-nine basis points have already been found between the funding costs of large and small banks. If these large-firm advantages are spread to the other financial industries through the designation of insurers, finance companies, holding companies, hedge funds, and others as SIFIs, it will change the competitive nature of our financial system. We have seen this movie before. Because of their perceived government backing, Fannie Mae and Freddie Mac were able to borrow funds at rates that enabled them to drive all competition from the secondary mortgage market.
These competitive advantages can extend beyond funding. In the highly competitive insurance market, for example, many small and midsize property and casualty firms compete effectively with the largest companies. Imagine what the industry will look like if the FSOC declares that two or three insurers or their holding companies are too big to fail and subjects them to special regulation by the Fed. Customers will feel more secure buying insurance from these large, federally regulated and protected firms than from others, seriously distorting competition in the insurance market. The same insidious process will occur in every other market where the FSOC declares one or more companies a danger to stability if they fail.
Some will argue that the Fed’s stringent regulation will actually be so costly to these large companies that they will get no benefit from the SIFI designation. Indeed, from all indications, large companies are busily making presentations to the FSOC, arguing that they are not too big to fail—certainly a signal that they do not see any advantages in the designation. Unfortunately, that is not very comforting. All firms would prefer to avoid regulation, especially regulation that the DFA terms “stringent,” but if these firms are correct and the added regulation is actually more costly than the funding advantages, conditions could be even worse.
Read Full Article »