The average American family lost one-third of its net worth during the Great Recession. One in ten families lost their homes. One in ten workers lost their jobs.[1] The consequences of the crisis still reverberate today, reflected in distrust of large financial institutions,[2] dissatisfaction with politics as usual,[3] and concern that capitalism is no longer working for the American people.[4]

It is possible to draw a line from the crisis to the election of Donald J. Trump as the 45th President of the United States. Further, in the most recent presidential election cycle, much of Senator Elizabeth Warren’s case for her electability was tied to her work in the Recession—arguing that she, unlike Republicans (and many Obama Administration officials), was focused on putting consumers first after the worst downturn since the Great Depression.[5]

While much has been written on the ways in which financial regulation has been overhauled since the crisis,[6] little research has been done on whether this overhaul was sufficient, or whether the system is still at risk. This Article steps in to fill this void. I argue that, despite regulators’ statements to the contrary, the vulnerabilities that led to the Recession remain in our financial sector. Without a course correction, the next time will be the same, and the consequences for ordinary Americans likely even more dire.

This Article differs substantially in tone from the calm espoused by those in the financial regulatory community of late. For example, in a speech in July 2019, Federal Reserve Vice Chair Randal Quarles announced that “banks have now built enough capital to withstand a severe recession.”[7] He further stated that it was now appropriate to deregulate large financial institutions because, since the crisis, large banks have addressed the “substantial deficiencies in their ability to measure, monitor, and manage their risks”—deficiencies that led to the Great Recession.[8]

Vice Chair Quarles is not alone in his optimism. Federal Reserve officials have claimed repeatedly in recent years that financial crises are behind us due to substantial reforms enacted in the aftermath of the Recession.[9] These reforms include decreasing banks’ ability to make risky bets and designing a plan for how to unwind large financial institutions with minimal harm to consumers. Perhaps most importantly, banks are now subject to annual stress tests that are intended to measure their ability to cope with a crisis-like event. For the last several years, all large financial institutions have cleared the stress tests with flying colors, suggesting that the system today is well equipped to weather the next storm.

At the same time, the market has not been so sanguine. In fact, in August 2019, only weeks after passing the stress tests, all large banks lost ten percent of their market value. Their probability of defaulting on borrowers, assessed from the cost of buying insurance that pays out if the firm defaults, skyrocketed. Analysts attributed this steep decline in value to an increase in bank risk: The business of banking involves borrowing short term and lending long term. In 2019, revenues from lending (long-term interest rates) fell below the costs of borrowing (short-term interest rates). This threatened the business model of large financial institutions and also prompted fears that a recession was imminent.[10] Concerned credit analysts downgraded financial firms,[11] and large financial institutions themselves advised their clients to begin to prepare for a recession.[12]

While industry participants, observers, and market signals were sounding alarms, regulatory measures of bank health were static. It is plausible that the market overreacted in August (in fact, it experienced a partial recovery in subsequent weeks), but it is unlikely that the risks in the financial sector were unchanged during this period as regulatory measures of bank capital suggested. Market measures provide a more dynamic assessment of the evolution of financial stability during this period. Regulators can, and should, monitor this information. Yet they do not.

Since the 1980s, capital regulation has been the primary form of bank regulation. Banks fail when the total amount of money they owe (their liabilities) exceeds the total value of the assets they have. The difference between a bank’s assets and liabilities is known as equity capital. Capital helps banks absorb losses that decrease the value of their assets and is measured by regulators as the difference between book values of bank assets and liabilities, known as book or “regulatory” capital.

However, this information is reported only quarterly and is prone to manipulation by sophisticated firms.[13] Because regulators rely solely on backward-looking, static, and manipulatable measures of capital, their assessments paint an inaccurate picture of bank health. I show this empirically in two ways.

First, I subject large banks in the United States to a hypothetical “market-based” stress test based on the value financial markets assign a bank’s business. These measures are dynamic and forward-looking, unlike the book-capital measures regulators have relied on historically. The results of a market-based stress test demonstrate that large financial institutions would experience cataclysmic losses in the event of a crisis like the Recession. Despite policymakers’ statements to the contrary, it is unlikely that these banks would be able to continue to intermediate as usual in the absence of substantial government assistance (that is, bailouts) during the next crisis.

Second, I document the failure of regulatory capital measures during the Great Recession and show that these measures were lagging indicators of bank health. Only days before some of the largest banks in the country failed, capital ratios indicated that all was well. In the case of Bear Stearns, Securities and Exchange Commission (“SEC”) Chairman Christopher Cox even testified before Congress after the firm had failed that it was healthy and well capitalized based on regulatory measures.[14] In contrast, market measures of bank health signaled cause for concern an entire year before the bankruptcy of investment banking giant Lehman Brothers sent the economy into free fall. In addition to being slow moving, regulatory capital measures also proved inaccurate: it was impossible to distinguish between healthy and doomed banks based on their reported capital levels. In contrast, there was significant divergence in the market’s perception of risk at these institutions, and its prediction of bank failures proved prophetic.

Our very recent experience illustrates the consequences of misplaced reliance on book capital as a measure of bank health, yet the post-crisis overhaul of financial regulation did not include a rethinking of the role this information plays in our assessments of large financial institutions. There has been no move toward incorporating more accurate market information into the regulatory regime.

This is an unforced error with significant repercussions. Large banks remain vulnerable to a crisis, and these risks are unacknowledged by the regulatory community. The singular focus on regulatory capital has also fueled a misunderstanding of the causes of the Great Recession and the tools policymakers had at their disposal to address them at their onset.

Policymakers typically offer two responses when asked about their failure to act more aggressively in the early stages of the crisis—that is, before Lehman’s bankruptcy—to forestall the catastrophe that ensued. The first is that the crisis could not be foreseen; illustrated, for example, by former Treasury Secretary Henry Paulson’s 2018 statement that his “strong belief is that these crises are unpredictable in terms of cause or timing or the severity when they hit.”[15] The second is that regulators lacked the legal authority to bolster struggling institutions. For example, as former Treasury Secretary Timothy Geithner stated in 2014: “The Fed didn’t have the legal authority to force Bear Stearns, Lehman Brothers, or other investment banks to raise more capital. We couldn’t even generate stress scenarios bleak enough to force the banks we regulated to raise more capital.”[16] Neither of these explanations, however, is accurate.

As to the unpredictability of financial crises, as this Article will show conclusively, substantial time existed between the first tremors in financial markets in the summer of 2007 and their eventual collapse in the fall of 2008. I assemble data on a variety of market-risk measures (including stock-price volatility, credit default swap (“CDS”) spreads, and market-based capital measures) and compare these with regulatory capital indicators. Market-based risk measures for large financial firms raised red flags for an entire year before the system collapsed. However, regulatory measures are slow to update—thus, failing banks were well above regulatory requirements for minimum capital ratios, not because they were healthy, but because these measures are flawed.

As to the lack of legal authority to intervene with financial institutions, this Article reviews the substantial legal authority at the disposal of financial regulators and demonstrates that lack of authority was not the binding constraint to action. Some pieces of evidence from this novel analysis are especially dispositive: First, regulators in fact did rely on their substantial legal authority to strengthen small financial institutions once risks emerged in the financial sector in the fall of 2007 and early 2008. This same authority could have simultaneously been wielded to bolster large, systemically important financial firms. Second, once the crisis was underway, regulators found ways to intervene and prevent even worse damage. By 2009, they forced banks to stop paying dividends and to raise new capital, which prevented additional failures. No new legal authority emerged between 2007 and 2009, which proves that lack of authority cannot explain the failure to respond more aggressively to the crisis at its onset.

In fairness to regulators, hindsight is twenty-twenty. It was impossible to predict with certainty in the summer of 2007 that a Lehman-size catastrophe was a year away. However, in the months leading up to Lehman, and especially after the collapse of Bear Stearns in the spring of 2008, the probability of a systemic collapse increased dramatically. In February 2008, academics presenting to Federal Reserve officials estimated that the losses that followed the collapse of the housing market would total about $500 billion, with half being borne by large and heavily leveraged financial institutions. This, they estimated, would imply a $2.3 trillion contraction in bank balance sheets—a substantial decrease in lending to households and businesses that would have immediate real consequences.[17] The way to prevent this contraction was to increase banks’ capital levels so they would not fail or to stop lending to households and businesses when imminent losses began to accumulate.

Yet instead of hoarding and raising capital to buffer against imminent asset losses, more than $100 billion of bank capital left the financial system in the form of dividend payouts to bank shareholders in the year before Lehman’s catastrophic bankruptcy. In fact, Lehman increased its dividend by thirteen percent in January 2008—six months before it collapsed and months after industry observers were aware of significant problems at the firm.[18] This is akin to deflating an airbag exactly when the risks of a crash are rising. The same occurred in the lead-up to the COVID-19 crisis—regulators allowed capital to be paid out to shareholders in the form of dividends and share buybacks at the same time monetary and fiscal authorities were contemplating economic interventions of unparalleled scope.

If not wanting for time or authority, what caused regulators to underreact to the initial stages of the crisis? This Article attributes this failure to reliance on regulatory capital, which painted (and continues to paint) an overly optimistic picture of financial stability.

Specifically, regulators failed to act in the early stages of the crisis because the default rule was inaction until book-capital levels signaled distress. Many looked at banks’ high regulatory capital ratios and concluded that there were few risks in the system: in the month before Lehman’s collapse, one Federal Reserve official guessed “that the level of systemic risk has dropped dramatically and possibly to zero.”[19] Others believed that, although it would be helpful for banks to have more capital, they were unlikely to do so while well above regulatory capital minimums, pointing to banks’ assertions that “now is not a good time” for equity-raising. Still others believed that acting aggressively—for example, by restricting banks’ dividend payments—would fuel a panic rather than prevent one.[20]

It is inaccurate and unfair to equate today’s regulatory regime to that in place in the summer of 2007. Capital requirements are higher, so banks have more of a cushion in place to bolster themselves when their assets begin to lose value. But the exercise of stress testing highlights the vulnerabilities that remain—that is, should a situation arise in which losses are so large that banks need to recapitalize, regulators will be slow to force them to do so because our tools of measuring banks’ risk, despite their known unreliability, have yet to be overhauled.

This Article provides a way forward, arguing that supplementing our understanding of financial stability with market information will paint a fuller picture. It also makes a case for automating regulatory action when banks appear undercapitalized—either based on regulatory or market measures. If in place during the crisis, such a regime would have forced banks to hoard and raise new capital in the year leading up to Lehman Brothers’ collapse, decreasing the need for costly government bailouts. The regulatory innovations advanced in this Article will prevent the next recession from becoming a “Great” Recession.

I propose different approaches to incorporating market information into the financial regulatory regime. The most extreme form would automate an aggressive response to market indicia that distress is imminent. This approach, which I label “dynamic capital regulation,” would quickly recapitalize banks the market deems to be on the brink. This recapitalization could be accomplished through: (1) a market-based stress test whereby failure requires new capital-raising; (2) the requirement that banks purchase capital insurance; (3) the conversion of some proportion of bank debt to equity, which eliminates the risk that creditors will be able to withdraw funds and push the bank to failure; or (4) a market trigger that forestalls capital leaving the financial system when bank equities experience drastic moves.

These market-based approaches will increase the dynamism and the transparency of financial regulation. However, dynamic capital regulation will also highlight concern about death spirals—that is, that market speculators will short financial firms when dilution appears imminent. Properly designed regulation can address these concerns, as I describe.

Still, dynamic capital regulation is not a panacea. The result will be fewer Great Recessions but also more false positives, which create unnecessary pain for the financial sector and its shareholders. For example, banks may be disallowed from paying dividends in periods when distress is not actually imminent, despite market signals to the contrary. However, concerns about false positives may be overblown: the analysis in this Article demonstrates that the simplest market-based indicator (bank stock performance) correctly identifies the two financial crises that have occurred since 1990 and results in no false positives. Deciding on the type of errors we prefer—false positives that are unfairly harsh to banks and their shareholders versus false negatives that result in costly losses to the government and taxpayers—is a tradeoff that requires thoughtful deliberation.

This Article favors dynamic capital regulation based on a premise that our regulatory regime should favor the protection of ordinary citizens over the protection of bank shareholders. Incidentally, given the more extreme alternatives, this approach is also likely to be favored by large financial institutions; it will allow them to intermediate efficiently with low levels of capital in normal times and only require them to bolster themselves in extraordinary moments when distress appears likely. In contrast, approaches like a thirty percent capital requirement proposed by Professors Anat Admati and Martin Hellwig,[21] or even more extreme discontinuation of financial intermediation full stop, as proposed by Professor Adam Levitin, are less efficient and more punitive.[22]

The right approach to bank capital is ultimately a question of policy, which regulators must decide. The main objective of this Article is to force a debate that is currently missing in the financial regulatory community due to misplaced confidence in regulatory measures of bank health. Given the known failure of these measures to provide useful and timely indicia of distress during the Great Recession, our continued sole reliance on them is puzzling. Market data are plentiful and informative; ignoring them would be extremely ill-advised for our regulatory regime.

This Article proceeds as follows. Part I begins by demonstrating the importance of bank capital to the financial system and describes how financial crises begin. Part II tells the story of the Great Recession, arguing that the severity of the crisis could have been mitigated by more aggressive regulatory action in 2007 and 2008. Although authority for intervention existed, inaction was the consequence of a regulatory regime that fails to respond until regulatory measures of bank health—which are static and often inaccurate—signal cause for concern. Part III calls for overhauling the regulatory default to make action, rather than complacency, the automatic response to the early stages of a downturn. This approach would have forced banks to stop paying dividends and required raising new capital at the beginning of the financial crisis. This approach will prevent the next downturn from being “Great.” Part IV concludes.

View Full PDF