For better or worse, over the course of my career, I have had the opportunity to participate in the response to three financial crises – the thrift and banking crisis of the 1980s and early 1990s as a member of the staff of the Senate Banking Committee, the Global Financial Crisis of 2008 as Vice Chairman of the FDIC, and the three large regional bank failures in the spring of 2023 as FDIC Chairman.

As I look back on those experiences, I am struck by the commonality of the causes of those crises –interest rate and liquidity risk, concentrations of assets and deposits, leverage, rapid growth, inadequate capital, new activities and products whose risks were poorly understood, interconnection with non-bank financial companies, poor bank management, and failures of supervision and regulation to identify and address those risks, and in some cases exacerbating them.

As I leave the FDIC, I thought there might be value in sharing some of the lessons of that experience as we head into a period of uncertainty about the future path of financial regulation in the United States and globally. In particular, I offer the observation that while innovation can greatly enhance the operation of the financial system, experience suggests it be tempered by careful and prudent management and appropriate regulation and supervision.

Thrift and Bank Crisis of the 1980s

Let me start by going back to 1980, when the banking and thrift industries had experienced more than four decades of stability.

After the reforms of the Great Depression, which included the creation of the FDIC in 1933, banking became a steady, perhaps even boring, business. From the end of World War II to 1979, only 160 depository institutions failed, or fewer than five per year. This stability was in large part due to a combination of laws that heavily restricted competition in banking, including caps on deposit rates, restrictions on branching and interstate banking, and restrictions on what types of products could be offered. In fact, the history of the FDIC can be neatly divided into its first 45 years, from 1933 to 1978, and the 45 plus years since.

Beginning in the late 1970s, banks and thrifts were facing a new type of competition – from what we would later call “shadow banking” or nonbank financial institutions. Newly formed money market mutual funds (MMMFs) offered deposit-like products but paid higher interest rates than banks were allowed. In 1975, MMMFs held just $3 billion in assets, and by 1981 they held almost $150 billion.¹ Compared to the roughly $200 billion of deposits at thrifts and $1 trillion of deposits at banks, this was a significant amount of competition.

In 1980, Congress passed the Depository Institutions Deregulation and Monetary Control Act (DIDMCA), which removed many of the interest rates caps that limited what banks and thrifts could pay on deposits. The Act was in part a response to interest rate increases by the Federal Reserve to control inflation. While the Act helped banks deal with MMMF competition in a rising interest rate environment, it also made it easier for banks and thrifts to take on brokered deposits and other forms of “hot money.”

Thrift Crisis 

Thrifts in particular faced difficulties due to their large portfolios of fixed-rate mortgage loans. Many of them were insolvent on a mark-to-market basis. In 1982, Congress passed the Garn-St. Germain Act, which removed many of the safeguards on thrifts to try to help improve their profitability. The Act let thrifts increase their investments in areas where they had little experience, including commercial real estate and consumer loans. Many thrifts “gambled for resurrection” by expanding rapidly and taking on risk in an attempt to earn their way back to solvency. Thrifts were also allowed to count so-called “net worth certificates” and increased amounts of goodwill toward their regulatory capital even though they carried no real value.²

Unfortunately, the agency that supervised thrifts, the Federal Home Loan Bank Board, was largely ineffective. Approximately 1,300 thrifts – or almost one-third of the industry – failed between 1980 and 1994.³ Deposit insurance for thrifts in those days was in a separate fund administered by the Federal Savings and Loan Insurance Corporation, or FSLIC, and these failures cost the taxpayers an estimated $132 billion.⁴

While certain deregulatory measures were appropriate for the thrifts to navigate their interest rate-induced losses, it is clear in retrospect that the manifestation of risk in one area cannot be dealt with by deregulating other types of risk-taking. Moreover, in a deregulatory period, strong and effective supervision is indispensable. The forbearance and accounting manipulations that were permitted simply created greater future challenges.

Bank Crisis 

As the thrift crisis unfolded, banks also were encountering their own challenges. Many of them had over-invested in commercial real estate and the energy sector and were severely affected by regional recessions of the early 1990s. Between 1980 and 1994, more than 1,600 banks failed.⁵ In 1990, for the first time, the FDIC’s deposit insurance fund balance went negative.

The 1980s also saw the failure of Continental Illinois National Bank and Trust Company, which is often considered to be the first “too big to fail” bank. It was the seventh-largest bank in the United States, and had grown rapidly by participating in risky loans, particularly in energy, and funding this activity through brokered deposits, Eurodollars, and other “hot money.” Most of these deposits were uninsured and started running quickly when questions emerged about Continental’s viability.

When Continental suddenly collapsed in May 1984, rather than place the bank into receivership, it was supported by an equity injection from the FDIC and a consortium of other banks, extensive borrowing from the Federal Reserve’s Discount Window, and a blanket guarantee on its uninsured deposits and general creditors by the FDIC. Not for the last time, regulators were forced to make a difficult choice between averting systemic risk on the one hand and making shareholders and creditors of failed banks bear losses on the other.

Policy Responses

The Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA) abolished the FSLIC, transferred the deposit insurance responsibilities for thrifts to the FDIC, and created the Resolution Trust Corporation to manage the assets received from the failed thrifts.⁶

In the Federal Deposit Insurance Corporation Improvement Act of 1991, or FDICIA, Congress reined in the use of forbearance by the regulators on both large and small banks. FDICIA created the Prompt Corrective Action framework, which forces regulators and banks to take corrective and remedial actions when capital ratios fall. It also put in place the least-cost test, which limited the FDIC’s ability to protect uninsured depositors and other creditors. FDICIA also created a framework in which the banking industry is responsible for recapitalizing the Deposit Insurance Fund, so that taxpayers would no longer be on the hook.

After this turbulent period, the industry recovered and expanded in the mid-1990s. When I joined the FDIC Board in 2005, we were in the midst of more than two years without a bank failure, the longest such period in the FDIC’s history at that time. The number of problem banks was approaching historic lows. Strong loan growth helped insured banks set six consecutive annual earnings records from 2001 through 2006.

Banking conditions seemed so favorable that friends asked me at the time if I really wanted to become a member of the FDIC Board—after all, nothing was happening in the banking industry and I might be bored.⁷ Unfortunately, this tranquility masked an enormous increase in risk-taking that would soon lead to the deepest financial crisis since the Great Depression.

Global Financial Crisis of 2008–2010

In retrospect, it became clear that the Federal banking agencies – the Federal Reserve, the OCC, and the FDIC – did not appreciate the full extent of these risks, and the bank regulatory framework was ill-suited for the challenge. Risks were amplified and masked through banks’ interactions with less-regulated nonbank institutions. The pre-crisis rules allowed, and in some cases encouraged, activities that fueled a housing bubble and contributed to the collapse of the financial system.

Gramm-Leach-Bliley Act

Before the crisis, there was a movement to expand the activities of both securities firms and bank holding companies.

Investment banks became publicly traded companies and expanded their business model from advising and underwriting equity and debt to also include trading, investment, securitization, and derivatives.

The Gramm-Leach-Bliley Act of 1999 removed restrictions across the financial sector. The largest bank holding companies (BHCs) became major players in investment banking by establishing or acquiring securities firms as subsidiaries. Some argued that financial institutions had strong incentives to protect shareholders and would regulate themselves by managing their own risk. There was opposition to imposing limits on bank activities or growth. From 1998 to 2007, the assets of the five largest U.S. banks grew from $2.2 trillion to $6.8 trillion.⁸

As the scope of financial institutions’ activities grew under the Gramm-Leach-Bliley Act, banking regulators were limited in their ability to directly examine or require reports from subsidiaries regulated by other agencies.⁹ These policies made it difficult for any single regulator to see the entirety of a financial firm’s activities. Furthermore, regulatory mandates primarily focused on risks to individual institutions, rather than on risks that develop across and between markets and institutions.¹⁰

These regulatory factors contributed to some of the key drivers of the 2008 Global Financial Crisis including deterioration of lending standards – particularly in the mortgage market, amplification and concentration of exposures to the mortgage market through securitization and derivatives, inadequate capital, and the deep interconnectedness of the financial system.

Mortgage Crisis

The nationwide housing expansion of the early 2000’s was rooted in a combination of factors, including a prolonged period of low interest rates. By mid-2003, both long-term mortgage rates and the federal funds rate had declined to levels not seen in at least a generation. One response to low interest rates was an acceleration in U.S. home price appreciation to double-digit rates for the first time since 1980. Another response was a series of mortgage market developments that dramatically weakened credit standards in mortgage lending.

As housing prices increased, mortgage underwriting standards deteriorated. Mortgage origination became less of an income-based credit decision and more reliant on continued appreciation in housing prices. The subprime mortgage securitization pipeline allowed risky mortgages to be packaged and sold throughout the financial system. The originate-to-distribute model undermined accountability for the long-term viability of mortgages.

Modest attempts to curtail risky lending activities through regulatory guidance were met with significant resistance from the industry and Congress. Such efforts were often thought to stifle innovation.

For example, banking agencies’ guidance on risks and concentration in subprime lending issued in 1999¹¹ and 2001¹²was met with severe criticism for being overly restrictive. Relatively unchecked, subprime mortgages grew from 8.2 percent of mortgage originations in 2003 to 23.5 percent of mortgage originations in 2006.¹³

Capital

The demand for securitized mortgage products increased and the capital held against these exposures was entirely inadequate. Rules for risk weights, combined with inflated credit ratings, allowed banks to reduce the capital they held against risky mortgages.¹⁴ Furthermore, banks did not hold capital against off-balance sheet exposures to the housing market, for example through their sponsorship of structured investment vehicles. This fueled growth in the private-label mortgage-backed securities market and allowed the banks to increase their leverage.

From 2000 to 2007, risk-weighted capital ratios were virtually unchanged for the largest banks, even as risk grew, and leverage ratios declined from 7 percent to 5.5 percent.¹⁵ Capital requirements outside of commercial banking were even weaker and institutions were more highly levered. The balance sheets of the five largest investment banks expanded by $1.8 trillion from 2004 to 2007,¹⁶ heavily increasing their exposures to the mortgage market and reliance on short term funding. This resulted in leverage of more than 40 to 1, by one measure, at year-end 2007.¹⁷

Capital remained overstated as the housing bubble began to burst because accounting rules enabled institutions to mask exposures and losses. Mark-to-market accounting rules in 2007¹⁸ required financial institutions to write down losses to certain assets. However, many of these assets were illiquid and had little to no discoverable market prices, increasing the uncertainty of their valuation. Marking to model or marking to “make-believe”, as some called it, allowed banks to overstate their capital position.

Securitization and Derivatives

At the same time, unregulated growth in the shadow banking sector further fueled the boom in the mortgage market. The use of mortgage-related securities as collateral in the asset-backed commercial paper and repo markets increased the demand for mortgage-backed securities to support these short-term funding markets and made those markets vulnerable to freezes. When property prices began to soften, uncertainty in the valuation of the MBS used as collateral rose, and financial institutions reliant on the collateral suddenly lost access to short-term funding.

As the mortgage market became over-extended, continued demand for highly-rated assets and declining demand for the riskier tranches of mortgage-related securities incentivized financial engineering of new products – such as collateralized debt obligations (CDOs), CDO-squared, synthetic CDOs, and credit default swaps (CDS) – that fueled the demand for continued securitization of non-prime mortgages. At the time this financial engineering was considered a form of innovation.

Lack of regulation of the over-the-counter (OTC) derivatives market, as a result of the enactment of the Commodity Futures Modernization Act of 2000,¹⁹ allowed many firms trading these products to operate with no capital, margin, or reporting requirements, and exposures grew rapidly. For example, the notional amount of CDS grew by more than 100-fold from 2000 to 2008, to over $60 trillion.²⁰

Interconnectedness

In hindsight, the interconnectedness of the system was significant and underappreciated. As the housing market collapsed in the fall of 2008, the financial crisis deepened. Investors and counterparties struggled to understand the opaque distribution of mortgage-related securities losses across the financial system. The opaqueness of the nature and size of exposures, particularly derivative exposures, combined with the heightened leverage embedded in the system, resulted in a dramatic seizing up of credit markets as market participants looked to minimize or eliminate exposure to potentially vulnerable counterparties.

Crisis Response

The systemic threat posed by the financial crisis called for a forceful public response. The government’s initial plan for the Troubled Asset Relief Program (TARP) to buy mortgage securities to support the market proved to be less effective than hoped. U.S. authorities looked for other tools to help stem the panic. On October 13, 2008, the FDIC, Federal Reserve, and Treasury announced a package of three unprecedented actions. The first action called for the Treasury to use the TARP program for capital injections rather than asset purchases. The second action was for the Federal Reserve to establish a widely available commercial paper facility to facilitate liquidity.

The third action called for the FDIC to use the systemic risk exception under the Federal Deposit Insurance Act to establish the Temporary Liquidity Guarantee Program (TLGP). This program included a Debt Guarantee Program (DGP) which guaranteed certain senior unsecured debt issued by eligible institutions and a Transaction Account Guarantee Program (TAG), which fully guaranteed noninterest-bearing transaction deposit accounts above $250,000.²¹

These actions helped to calm market fears and restore financial stability.

Too Big to Fail

In addition to these extraordinary interventions to prevent the failure of large institutions and the financial system, from 2008 through 2013 nearly 500 banks failed, including the largest failure in the FDIC’s history, Washington Mutual with $300 billion in assets. These failures would eventually cost the Deposit Insurance Fund approximately $69 billion. Most of the failed institutions were community banks, often in parts of the country where the subprime mortgage crisis and the recession made real estate problems more severe than elsewhere.

In addition to the enormous economic and human cost of the crisis – almost nine million of lost jobs, 12 million homeowners facing foreclosure and an estimated $10 to 15 trillion in lost GDP²² – it also highlighted two related aspects of how failing banks were handled in the U.S.

The first and most obvious is the problem of too-big-to-fail. When the largest banks were on the brink of failure, the authorities provided them extraordinary government assistance on an open-bank basis rather than allow them to fail. This meant that their shareholders and creditors were not exposed to losses and senior management was not held accountable. It also meant that uninsured depositors at these banks were fully protected.

The second and related issue was the disparate treatment between large and small banks. When community banks and smaller regional banks were undercapitalized, they were closed. Their shareholders and creditors were wiped out and management was replaced. And while in the majority of these cases the acquisition by another bank under the least-cost test resulted in no loss to uninsured depositors, this was not always the case. So, for practical purposes, only uninsured depositors at smaller banks faced the risk of loss.

Post-Crisis Response

The Dodd-Frank Act, enacted in 2010, addressed many of the regulatory gaps that surfaced in the crisis and, in many ways, expanded the options for crisis management. ²³

• Title I of the Dodd-Frank Act subjected large bank holding companies (BHC)²⁴ to heightened prudential standards including higher risk-based capital requirements and leverage limits, liquidity and risk management requirements, and resolution planning.
• Specifically with regards to resolution, Title I required the largest BHCs to provide a plan for their rapid and orderly resolution under the U.S. Bankruptcy Code.²⁵ Title II provided the FDIC with dramatically expanded authorities to manage the orderly failure of a U.S. Global Systemically Important Bank (GSIB), or for that matter any financial company whose failure was deemed to pose a risk to U.S. financial stability.²⁶
• Title VII also required the clearing of standardized derivatives through central counterparties and established significantly strengthened margin requirements for most OTC derivatives. ²⁷

It is important to recognize that Basel III is an effort by the U.S. banking agencies to strengthen the banking system because our nation’s largest, most systemically important financial institutions were found to be woefully undercapitalized and over-leveraged in 2008. The early round of Basel III changes significantly improved the quantity and quality of capital held by internationally active banks. The four critical areas of risk addressed under the remaining final phase of Basel III – credit risk, market risk, operational risk, and risk associated with financial derivatives – are a direct response to the experience of 2008.

Regional Bank Failures of 2023

The focus on GSIB capital standards and resolution, while entirely appropriate post-crisis, meant that less attention was paid to the risks associated with the failure of a large regional bank. In 2019, I gave a speech here at Brookings pointing out that the resolution of a regional bank could cause significant systemic risk, that regional banks were highly reliant on uninsured deposits, and underscored the importance of devoting appropriate attention to their supervision and resolution.²⁸

In 2018, however, Congress passed the Economic Growth, Regulatory Relief, and Consumer Protection Act (EGRRCPA). This rolled back some of the safeguards that the Dodd-Frank Act had put in place after the Global Financial Crisis by granting the Federal Reserve discretion to increase the asset threshold for many prudential requirements from $50 billion to $250 billion. Capital, liquidity, and stress-testing requirements were substantially reduced for banks in that range.

In addition, the threshold for submitting Title I resolution plans, or “living wills,” was raised to $250 billion under the authority given to the Fed.²⁹ The FDIC also placed a moratorium on bank-level resolution plans for regional banks.

In March of 2023, Silicon Valley Bank of California (SVB) with over $200 billion in assets, then the sixteenth largest bank in the U.S., experienced a bank run. SVB relied on a deposit base with 90 percent uninsured deposits, invested in long-term government securities, and had expanded rapidly, tripling in size in the preceding two years. As interest rates rose, the market value of the bank’s assets fell. When the bank sold its portfolio of securities at a large loss to raise liquidity, it experienced a big hit to its capital and uninsured depositors withdrew en masse.

When SVB was closed mid-morning on Friday, March 10, 2023, the FDIC initially planned to pay out its relatively limited amount of insured deposits, provide uninsured depositors access to a portion of their funds, and begin marketing the rest of the bank’s operations. This plan was in line with the Federal Deposit Insurance Act’s requirement that the FDIC choose the least costly manner of resolving a firm.³⁰

However, the prospect that uninsured depositors at SVB would experience losses alarmed uninsured depositors at several other regional banks, and depositors began to withdraw funds. Signature Bank of New York, in particular, experienced heavy withdrawals, and was closed on Sunday, March 12, 2023. Faced with growing contagion in the financial system, a systemic risk exception (SRE) was invoked for both SVB and Signature Bank protecting uninsured depositors. This allowed the FDIC to organize bridge banks for their operations and buy some time to find potential buyers.³¹

These actions calmed the market. When First Republic Bank of California, which had 70 percent uninsured deposits, failed on May 1, 2023, the FDIC was able to resolve the bank using its ordinary processes without recourse to the SRE. These bank failures were the second, third and fourth largest in U.S. history, although it is worth pointing out that this is because when some of our largest institutions were at risk of failure in 2008 they were bailed out.

The decision to recommend a systemic risk determination was not an easy one, although I would note that SVB and Signature Bank were allowed to fail, shareholders were wiped out, and the boards and management teams of those institutions were replaced.

The deregulatory environment of the time did not help. Silicon Valley Bank would not have been in compliance with the full Liquidity Coverage Ratio as it had been applied prior to the implementation of the 2018 law.³² It was not required to undertake company-run stress testing, and the transition rules under the 2018 law delayed its supervisory stress test despite its rapid growth. Its holding company was not large enough to require a Title I resolution plan.³³ The 2018 law also had a chilling effect on supervisors at the time, as documented in the Federal Reserve’s analysis of the SVB failure.³⁴

During that period, supervision should have been emphasized more, not less, especially due to SVB’s rapid growth and balance sheet concentrations.

Policy Response

In response to the Spring 2023 regional bank turmoil, the FDIC, Federal Reserve, and OCC have pursued a number of policy responses to address the identified risks.

• After examining our pre-failure supervisory activities, all three agencies acknowledged gaps and the need for more intensive and timely supervision.³⁵ Interest rate risk, concentrations of unrealized losses on assets, concentrations of uninsured deposits, and rapid growth have been a focus of attention for all three agencies.
• The FDIC updated its insured depository institution resolution planning rule in June 2024 to require a comprehensive plan and resolution strategy from banks with at least $100 billion in total assets, and a more limited informational filing from banks with at least $50 billion in total assets.
• The three banking agencies also jointly proposed a requirement for IDIs with more than $100 billion in total assets to maintain a minimum amount of long-term debt that could absorb losses in resolution ahead of uninsured deposits. This would reduce the incentive of uninsured depositors to run and perhaps reduce the likelihood of failure. In the event of failure, it would increase the prospect of the FDIC having resolution options beyond liquidation and reduce the cost of failure to the Deposit Insurance Fund.
• The FDIC issued a report on Options for Deposit Insurance Reform.³⁶ The report identifies Targeted Coverage–which would allow for higher or unlimited insurance for business payment accounts–as having the greatest potential for meeting the fundamental objectives of deposit insurance relative to its costs.

Lessons for the Future

As I indicated at the outset, as I look back at these three episodes of financial system disruption, I am struck by how many common threads run through them, even as the specific context and details differ.

Interest rate and liquidity risk, reliance on uninsured deposits and wholesale funding, inadequate capital, leverage, rapid growth, poorly understood new financial products and companies, sheer size, inadequate risk management by the banks, and accommodating supervision and regulation have repeatedly forced the hand of the U.S. government to intervene and protect different types of creditors, and the firms themselves.

I am particularly concerned with the proliferation of activities of non-bank financial institutions, which I believe pose financial stability risks. The Financial Stability Oversight Council has repeatedly pointed out risks growing outside the regulatory perimeter ranging from hedge funds to private credit lenders to non-bank mortgage servicing companies.

I think that we have made progress since the 1980s with improved capital and liquidity requirements, strengthened regulation of derivatives markets, and resolution planning. On the other hand, the largest banks are bigger, more complex, and deeply interconnected domestically and internationally.

I am concerned that memories are short. We should not allow the current relative stability of the banking and financial systems to lull us into a false sense of complacency. Not only are many people not familiar with the thrift and banking crises of thirty years ago, some seem to have lost sight of the experience of the Global Financial Crisis of 2008 and even the regional bank failures of the spring of 2023.

New technologies, new financial products, and new kinds of financial companies are part and parcel of the evolution of the financial system that we have experienced before. But we should not kid ourselves into believing that they do not present risks that need to be carefully supervised and, if necessary, regulated. That to me is the core lesson of these three financial crises to which I hope we pay close attention.

ENDNOTES