Sections

Commentary

Op-ed

Rethinking bank oversight in the wake of recent banking turmoil

dominos against stack of coins

The Federal Reserve, the Federal Deposit Insurance Corporation, and the Treasury Department recently took extraordinary actions to help prevent the collapse of Silicon Valley Bank (SVB) and Signature Bank from spilling over and igniting a potential financial contagion. The Fed created a new lending program and took other steps to infuse liquidity into banking institutions in distress. In addition, the three agencies invoked the “systemic risk exception” as modified in the Dodd-Frank Act to protect all deposits held by the two banking institutions.

Government officials have since attempted to clarify when the systemic risk exception might or might not be used. While congressional action likely is required for any broad expansion of deposit insurance protection, it is our view that in light of recent events, for all intents and purposes, policymakers have placed themselves in the exceedingly and politically difficult position of not letting even small banks suffer deposit runs in the future. As a result, the federal government arguably has become the ultimate guarantor of all deposits, large and small.

Market purists are bemoaning this state of affairs, but the genie of all deposit protection is out of the bottle. It can’t be put back in. The speed of the run on SVB—the bank was closed in less than 24 hours after the run was mounted—has exposed the fact that “market discipline” seemingly imposed by uninsured depositors can be so destabilizing to the banking system as a whole that it will be impossible to run the experiment of not protecting them, at any bank of any size, in the future.

All this calls to mind General Colin Powell’s famous caution in 1991, warning policymakers about having military forces going all the way to Baghdad in the war back then: “If you break it, you own it.” Although the federal government now does not “own” the nation’s banks, it now implicitly owns the risks associated with all depositors, who fund roughly 90% of the $23 trillion dollars of assets in the banking system. In our view, those risks are inextricably linked with the monetary policies the Fed pursues, specifically monetary tightening, which can impair banks in two ways: by increasing the likelihood of credit losses and by causing the market values of their assets to decline, potentially by more than the offsetting market values of their deposits as interest rates increase.

Although SVB had other problems—its deposit base was almost entirely uninsured and was heavily concentrated among venture capital investors—it is hardly alone in having an “interest rate risk” problem due to the Fed’s anti-inflation campaign. Public reports indicate that the entire U.S. banking system has over $600 billion of unrealized losses in marketable securities alone, not counting any additional unrealized losses in banks’ loan portfolios, representing a substantial portion of the $2.2 trillion in reported capital, in the aggregate, of all roughly 5,000 banks.

What then, should federal regulators do now? In the coming weeks, the government will rightfully turn its attention to evaluating and intensifying various regulatory and supervisory requirements for banking institutions. The Fed also will examine its supervisory oversight process to identify any of its own failures. These efforts all make sense. But in light of the speed of the deposit run at SVB and the prospect and dangers of similar fast runs at weak banks in the future, reviews should do more than simply taking the well-trodden road of increasing regulatory requirements.

We believe that one important lesson from the SVB failure is that the world of banking now moves at a different speed than the regulatory framework that was originally designed to oversee it. For instance, in this age of social media and instant news, information spreads with breathtaking speed and scale, potentially adding another accelerant to deposit runs. For their part, the banking agencies appear to acknowledge these changes. Rohit Chopra, the head of the Consumer Financial Protection Bureau, stated at a recent industry conference that “[t]here is no question that … one viral social media posting really could have a contagious effect. […] So I’m not really sure there’s a clear solution of that other than us as regulators and industry accepting that faster communication is an opportunity and a risk.” In his recent testimony to Congress, Michael Barr, the Fed’s vice chair for supervision, also acknowledged how regulators must “evolve” their “understanding” of the banking business, and implicitly their regulation and supervision of it, given “changing technologies”.

In this brave new world, banking regulators and policymakers more broadly must think outside the box and build and use new tools. Regulators need the capability to gather timely intelligence and respond more quickly and systematically, even if we are wrong in our view that now all deposits at all banks are de facto insured.

Our four recommendations below would improve banking system oversight by creating a monitoring system with built-in processes that routinely collect, analyze, and interpret information in real time and thus provide early warnings about the health of regulated institutions, individually and as a whole. More specifically, we recommend:

  • 1. The development of a centralized and automated real-time financial monitoring system (RFMS) for regulators. This system would require, at minimum, every banking institution above $100 billion in assets to input a daily set of key financial data, including other comprehensive income (OCI). In addition, the RFMS would provide a real-time, automated analysis of financial standings of individual banks and the system as a whole in order to identify weaknesses. Results from this RFMS can augment, and also be used to challenge, the supervisory process.
  • Currently, supervisory oversight is largely a backward-looking exercise, a challenge in a fast-moving environment. Bank examinations are inherently a point-in-time health check. And regulators often have to work with information that is somewhat dated or point-in time snapshots. And by the time supervisory analyses are rolled up into exam reports and CAMELS ratings months later, the information is already dated. “Supervisory monitoring” —designed to supplement examinations—relies heavily on quarterly regulatory reporting data (“Call Reports”). By the time, regulators get access to the Call Reports, however, the bank data is more than 30 days old and, in the case of holding companies, more than 45 days after the quarterly snapshot. We acknowledge that the regulators do have a limited set of daily and weekly data collection processes in place. However, these processes often apply only to the very largest banking organizations and/or seek only a limited, specific set of data. There is no tool currently available to the regulators and supervisors alike that would provide a real-time, unvarnished (e.g., changes in OCI), and comprehensive picture of the financial health of institutions above $100 billion. Real-time monitoring will correct for these limitations.
  • 2. Along with the RFMS, the development of a robust, automated early warning system to help regulators identify and flag banking institutions with a significant risk profile. The early warning system could trigger automatically if certain conditions are met, including deteriorating financial data from the RFMS, individual bank or industry trends that might indicate significant risk-taking behaviors such as unusually fast growth or unhedged mismatches in the durations of banks’ assets and liabilities (which would permit the regulators to calculate the market values of at least all the banks’ marketable securities), or direct warnings from the supervisory process. We also believe that the automated feed of real-time data and real-time analysis with built-in early warnings would make it easier for—and essentially force—the regulators to recognize and deal with emerging issues.
  • Additionally, the SVB failure again demonstrated that fast-moving or major shifts in economic policy (such as the significant fiscal and monetary policy actions taken during the pandemic emergency) can have consequences. While the Fed’s supervisory stress tests are designed to measure the capital adequacy of large banking organizations under hypothetical stress conditions, the process, by design, does not work well in fast-moving macro environments. For example, the supervisory tests run annually, with test scenarios released in the first quarter of any given year. With a real-time monitoring system of banks in place, the Fed’s monetary policymakers and regulators alike can independently conduct their own capital and liquidity stress tests with any additional scenarios they determine as necessary.
  • 3. Given the “all deposits are protected” regime that may be effectively in place, if supervisors find unsatisfactory results after investigating early warnings, they should be able to take the set of escalating supervisory actions that is already allowed under the law without as much (or any) fear of contagious bank runs once word of those actions becomes public. These supervisory interventions could range from informal and formal enforcement actions to ultimately closing the institution. With a real-time monitoring system in place, coupled with strengthened resolve by the supervisors to take these timely interventions, supervisors will be much better positioned to prevent those institutions that are now at or near market value insolvency from “gambling for resurrection” by taking additional risks in between exams.
  • 4. Finally, a real-time regulatory framework will require nimble and coordinated response among different regulators. The current regulatory framework, by design and history, has many layers of review and interagency consultations. That framework offers the benefits of checks and balances as well as multiple perspectives. However, it also makes it challenging to get a clear picture of what’s happening in the banking system with different banking regulators. For example, federal and state regulators may not always see eye to eye and therefore may need to negotiate to find common ground on a myriad of details, such as what sample data to review, who to interview during exams, or what exam outcomes should be.
  • Even within the same agency, there are internal processes—within individual regional offices and between regional offices and agency headquarters—for reviewing exam reports and enforcement actions. Trying to get them to communicate and agree on how to deal with fast-moving events can be complicated. In particular, banks deemed to be operating in good condition (CAMELS ratings of 1 or 2) would typically receive a written communication to correct deficiencies within a fixed timeframe and to submit the results for review before the regulators consider escalating actions.
  • In fast-moving situations—rapid growth in assets combined with rapid changes in interest rates, as SVB experienced—such a methodical approach may work against the regulators taking timely action. Therefore, the RFMS and early warning system described above must be accompanied by a streamlined and nimble decisionmaking process that requires banks to correct problems quickly as well as by simplified communication and messaging protocols across the bank regulatory agencies.

There will be challenges to pivoting the existing regulatory framework towards real-time monitoring. For example, designing such a monitor would require strong safeguards against cyber and information security risk. How to handle the transition, among other details, also would need to be ironed out. Our recommendations do not address shadow banks that continue to operate outside the regulatory purview, despite their growing risks to the financial system. But getting a much better handle on the risks in the regulated banking system through adoption of the foregoing recommendations at least will represent a major step toward a safer and more resilient financial system.


Alice Cho serves on the board of directors of First Interstate BancSystem and Globe Life, and she is a senior advisor at the Boston Consulting Group. The authors did not receive financial support from any firm or person for this article. Other than the aforementioned, the authors did not receive financial support from any firm or person with a financial or political interest in this article. Other than the aforementioned, they are currently not an officer, director, or board member of any organization with an interest in this article. The views expressed in this article are solely those of the authors.

Authors

  • Footnotes
    1. Federal Deposit Insurance Corporation, “The FDIC Quarterly,” February 2023. Data are as of 4th quarter, 2022. https://www.fdic.gov/analysis/quarterly-banking-profile/fdic-quarterly/2023-vol17-1/fdic-v17n1-4q2022.pdf. During the financial crisis of 2008-09, the government also took a wide-ranging set of significant actions both to stem the contagion and absorb its risks.
    2. Alexander Saeedy, “Investors See Higher Defaults Ahead as interest Rates, Inflation Take a Toll,” The Wall Street Journal, June 15, 2022, https://www.wsj.com/articles/investors-see-higher-defaults-ahead-as-interest-rates-inflation-take-a-toll-11655329404. For a concise explanation of the potential impact of monetary tightening on bank investment assets, see, Carl White, “Rising Interest Rates Complicate Bank’s Investment Portfolios,” Federal Reserve Bank of St. Louis, https://www.stlouisfed.org/on-the-economy/2023/feb/rising-rates-complicate-banks-investment-portfolios.
    3. SVB’s deposits were more than 90 percent uninsured, making the institution especially vulnerable to a run. That said, SVB’s run was remarkably fast, taking less than 24 hours; SVB’s ultimate fate was predicted and discussed openly by social media influencers, arguably fanning the flame. See, Vivian Giang and Mike Dang, “10 Days That Have Roiled the Markets: A Timeline of the Banking Chaos,” The New York Times, March 20, 2023,  https://www.nytimes.com/article/svb-silicon-valley-bank-collapse-timeline.html. As a point of comparison, the experience during the 2008-9 financial crisis was different. IndyMac experienced a run around June 26, 2008 and was closed by the FDIC on July 11, 2008; Washington Mutual experienced two runs lasting September 8 – 25, 2008. Wachovia’s runs began around September 15, 2008, and lasted until its announcement of acquisition by Wells Fargo on October 3, 2008. See Jonathan D. Rose, “Old Fashioned Deposit Runs,” Finance and Discussion Series, Federal Reserve Board, November 18, 2015, pp. 2 – 6. https://www.federalreserve.gov/econresdata/feds/2015/files/2015111pap.pdf.
    4. In September 2022, at the release of 2nd quarter banking performance data, the FDIC Chair, Martin Gruenberg, began warning about the rising risk of unrealized losses in banks’ investment portfolios. Had regulators been privy to real-time changes in the value of investments for individual institutions and for the industry as a whole, and been forced to reckon with it in the form of automated data monitoring, we believe the warning signal would have been gone up earlier for policymakers and supervisory personnel alike. See, “Remarks by FDIC Acting Chairman Martin Gruenberg on Second Quarter 2022 Quarterly Banking Profile,” The Federal Deposit Insurance Corporation, September 8, 2022, https://www.fdic.gov/news/speeches/2022/spsep0822.html.
    5. The regulators gather financial data from regulated institutions in a number of ways, including through the quarterly Federal Financial Institutions Examination Council (FFIEC) Call Report. And regulators can demand real-time information during times of crisis. However, with the key exception of daily liquidity reporting requirement (e.g., FR-2052a) for Global Systemically Important Banks (G-SIBs) and a subgroup of certain large organizations, regulatory data are largely collected monthly, quarterly, annually, or through periodic onsite examinations. (Note that the Fed also collects basic balance sheet information from a subgroup of institutions on a weekly basis; but the data are limited and do not include interest rate sensitivity information.)
    6. The Office of Financial Research (OFR) supports the work of the Financial Stability Oversight Council by preparing high-quality financial data and analysis. The OFR does not have real-time access to bank data.
    7. In addition, current stress tests are designed to focus on regulatory capital and pre-provision net revenue, and do not explicitly integrate funding risk or liquidity risk into the methodology.
    8. The primary regulator of most, if not all, of the largest bank institutions is the Office of the Comptroller of the Currency (OCC). The Federal Reserve has oversight of these banks through its role as supervisor of bank holding companies. For other banks, the primary federal regulator may be the OCC, Federal Reserve, or FDIC. In the case of Federal Reserve and FDIC banks, they are jointly supervised by the state banking departments, typically with federal and state regulators rotating to lead examinations.