Eoduard, thank you for that very kind introduction. Before I start, I wanted to thank the Banque de France for hosting and Chairwoman Zilioli and the members of the committee for inviting me to participate in the conference this weekend. I feel very honored to be here and am looking forward to our discussion on recent challenges in the banking sector.
Today I will share an attorney’s perspective on the Federal Reserve’s Bank Term Funding Program (or the BTFP), which helped stabilize the banking system following the failures of Silicon Valley Bank (or SVB) and Signature Bank. I will also offer some observations based on those failures, and will assume familiarity with facts already in the public record. Specifically, I will propose three areas of improvement: (i) contingent liquidity preparedness, (ii) follow-up on supervisory findings, and (iii) the importance of effective challenge in corporate governance.
Throughout, the views that I express are my own and do not necessarily represent those of the Federal Reserve Bank of New York or the Federal Reserve System.1
Bank Term Funding Program
The New York Fed, which is my client, did not supervise either SVB or Signature Bank. We actively monitored both situations, and were involved with Signature as a lender of last resort. When the Federal Reserve established the BTFP, the lawyers of the New York Fed played an important role in facilitating its rapid implementation. I was responsible for coordinating among my team of attorneys at the New York Fed and the Board of Governors to ensure that our actions complied with applicable statutes and regulations.
Over the weekend of March 11 and 12, the Fed designed the BTFP to support the stability of the broader financial system by providing a source of financing for banks with Treasury, Agency and other eligible holdings whose market value had significantly diminished given interest rate increases. Market participants were very focused at the time on banks with large, unrealized losses in their holdings of these sorts of securities. The rapid deposit growth of both SVB and Signature in the wake of the pandemic resulted in these banks holding quite large Treasury and Agency positions relative to their overall balance sheet. Where a bank is holding securities on its books as “held to maturity”, changes in market value do not affect book equity or regulatory capital.2 However, the unrealized losses may—and in recent cases, did—affect the ability of the bank to sell or pledge those securities to generate liquidity. Selling these securities, as we saw with SVB, creates a realized loss on the balance sheet. Borrowing against the securities would not have forced the bank to realize a loss, but the value of these securities as collateral would have been calculated using market value, not face value. Accordingly, these securities may not have generated as much liquidity as the bank needed, desired, or anticipated. BTFP was an innovative way to create an additional source of liquidity using these high-quality securities, calming market fears over the consequences of fire sales.3
Along with the goal of enabling financing up to the face value of these securities, the Fed concluded it would be helpful to the borrowers for the term of the loan to be set at one year. The thinking was that committed financing of these securities would significantly mitigate any concerns market participants had that a bank would need to realize losses on these securities. If a bank decided it did not want the financing for a full year, the BTFP loan could be repaid before maturity without penalty.
Having decided the facility should be set up to lend against the face value of the securities and for up to a year, the next question was, “How could the program be up and running by Monday morning?”
Not exactly a small ask.
There was not enough time to set up special purpose vehicles as the Fed had done for some of the pandemic programs. The only way to have the program up and running so quickly was to leverage our discount window facilities. However, our discount window authority—Section 10B of the Federal Reserve Act—does not authorize reserve banks to lend for a period greater than four months.4 In addition, under traditional discount window policies, loans are extended against collateral that is assigned a lendable value by the reserve bank based on a haircut applied to the asset’s fair market value.5 Under the circumstances we faced, traditional discount window operations could not fully meet the acute needs of the banking sector.
As a result, we turned to Section 13(3) of the Federal Reserve Act, which authorizes specialized lending in unusual and exigent circumstances.6 The BTFP extends the maximum term of lending from the Section 10B limit of four months up to a special limit of one year. Additionally, unlike traditional discount window operations, the BTFP authorizes banks to borrow against eligible holdings up to their par value rather than their market value less a haircut. By limiting eligible BTFP collateral to essentially Treasury and Agency securities, the BTFP was targeting issues arising from interest rate-driven mark-to-market losses. Allowing other forms of collateral would have potentially introduced questions about the credit or underwriting of the collateral, which would have unnecessarily complicated and increased the risks of the program. To reduce arbitrage opportunities, only securities owned by the pledging bank as of March 12 are eligible for BTFP loans.7
BTFP loans are recourse to the borrowers, so the lending reserve bank is not limited to just looking to the BTFP collateral for repayment in the event of a default. Instead, the reserve bank would have a claim for repayment in full against the borrower. After applying BTFP collateral, the lending reserve bank may also recover against other assets, if any, pledged in connection with regular discount window arrangements or otherwise.8 If this additional collateral is inadequate and the borrower cannot otherwise repay the BTFP loan, the Treasury Department has agreed to reimburse the Federal Reserve for aggregate losses on BTFP loans up to $25 billion.9
As I previously mentioned, the U.S. banking sector stabilized thanks, in part, to the BTFP assuring markets that banks with underwater government securities portfolios can use those to generate liquidity on the basis of their par value. It helped avoid a fire sale. As of May 31, the total outstanding amount of advances under the Program was approximately $107 billion, while the collateral pledged to secure the loans was approximately $129 billion.10
Three Lessons for Banks and Supervisors
While I consider the BTFP to be a necessary and successful short-term response to the events in the U.S., there are different measures we should begin to incorporate over the longer term into the architecture of banking and supervision. I want to focus today on three, drawn from the failures of SVB and Signature Bank, and offer my views on a path forward.
Contingent liquidity preparedness
The first lesson is to prepare for contingencies—especially given the speed and magnitude of recent bank runs. In past practice, small and midsize banks have often relied primarily upon a single contingency funding source. Under modern economic and technological conditions, reliance on a single source carries too great of a risk that borrowers will be unable to generate sufficient liquidity from their investment positions in a short window of time. Instead, modern banks of all sizes need multidimensional, regularly tested funding sources to ensure emergency access to credit. This is not just for the benefit of individual banks, but also for the stability of the financial system.
I draw this lesson principally from the collapse of Signature Bank—which, in my view, was too reliant on the Federal Home Loan Bank of New York as its main source of contingency funding. When Signature reached a critical point in its fight for survival, its personnel were unfamiliar with the operational processes and requirements to use the New York Fed’s discount window. It had been five years since Signature last tested its discount window operations. Staff at Signature did not understand the discount window collateral eligibility and diligence standards. It had not worked through the diligence requirements necessary to pledge its commercial real estate loans. Signature was also unfamiliar with the process and time constraints to transfer pledged securities from the Federal Home Loan Bank of New York to its pledge account at the New York Fed.
With Signature facing multiple friction points, my team at the New York Fed identified an ad hoc solution that enabled Signature to borrow late on Friday, March 10. Under Section 10B, discount window loans must be sufficiently secured.11 Normally, the Fed establishes and perfects its security interest in Fedwire-eligible securities, such as Treasuries, Agencies, and Agency MBS, by requiring the borrower to transfer them to the borrower’s pledge account at the New York Fed. In this case, however, we worked with the Federal Home Loan Bank of New York and with Signature to enter subordination and pledge arrangements that provided the New York Fed with a perfected priority security interest without transferring the securities.12 This satisfied the Section 10B requirement.
Though we tailored a solution under these circumstances, banks would be very unwise to assume the Fed will be able to provide special, individualized attention and creative solutions in the event of another liquidity crisis. Bank management must take ownership of their operational preparedness before a liquidity crisis occurs. At a minimum, banks should conduct tabletop exercises to identify gaps in their capabilities. Banks should refamiliarize themselves with the requirements to access the discount window. Most important, I urge banks to take the extra step of exercising these borrowing capabilities regularly to facilitate access to funds should the need arise.
On our end, the New York Fed has begun a constructive dialogue with small and midsize banking firms centered on emergency liquidity preparedness, including contingency fundraising and borrowing. New York Fed staff have been meeting with banks to explain discount window operations and the BTFP. We actively promote use of the discount window as it offers the most reliable source of liquidity in a crisis.
Unfortunately, the discount window remains stigmatized in the banking community.13 I hope that, in light of recent events, banks will take advantage of this moment to establish a regular practice of using the discount window. It would redound to the benefit of banks and their customers if we overcome discount window stigma and prioritize sound risk management.
Follow-up on supervisory findings
The second lesson from these events is that supervisory findings need to resonate with bank leadership—the board and senior management in their respective roles.
Vice Chair for Supervision Michael Barr’s report on the failure of SVB (the Barr Report) concluded, in essence, that internal governance and risk management failures at SVB contributed substantially to conditions that fomented the run on the bank.14 His report also faulted supervisors for not having done more to draw attention to those problems and compel their remediation.
As a threshold matter, I want to note that it is not at all clear that SVB would still be in business even if Fed examiners had acted sooner or more forcibly. Regulators and supervisors can minimize the impact and probability of bank failures, but they cannot prevent all banks from failing. The decisions of bank managers will ultimately play the most important role in determining the success or failure of a given bank.
From what I can tell based on the public record, it appears that SVB management was, in a word, complacent. I find it fascinating that in his testimony before the Senate Banking Committee, when asked why action was not taken with respect to a finding on interest rate risk, the former CEO of SVB downplayed a “Matter Requiring Attention” (or MRA) as the “lowest level” of supervisory finding. The fact that an issue becomes a supervisory finding at all means it is important! Moreover, at the time of its failure, SVB had 31 outstanding findings—that is, MRAs and “Matters Requiring Immediate Attention” (or MRIAs)—on safety and soundness issues. This leads me to conclude that supervisory findings did not create a sufficient sense of urgency in SVB management.15 A key lesson from SVB’s failure, then, is that management can become desensitized over time to the seriousness of findings—especially where they are allowed to linger without consequences.
For supervisors, this is an opportunity to assess how we issue and monitor supervisory findings, and to embolden supervisors to employ extraordinary remedies when risk management deficiencies threaten a firm’s safety and soundness. There is, understandably, a good deal of self-reflection on why supervisors with an understanding of the severity of these MRAs failed to follow-up more tenaciously. Looking ahead, perhaps supervisors should prescribe more exact timetables for addressing MRAs and MRIAs—deadlines that are realistic, but firm. And, when we set those timelines, we could be clearer about the consequences if matters are not resolved on time. Finally, if findings linger, we need to actually impose the consequences we forecasted. The severity of those consequences—which could include a ratings downgrade or enforcement action—would depend on the nature of the finding and the reasons for delayed remediation. These are case-specific determinations, of course, best entrusted in the first instance to the judgment of experienced examiners and, ultimately, to the Board of Governors.
I believe there is a complementary oversight role for bank boards. Directors need to understand how bank management addresses supervisory findings, and hold managers to account—for example, through decisions about compensation and promotion. Directors might ask how the remediation of supervisory findings factors into variable compensation for those responsible for the finding—ultimately, senior management. If the answer is, “Not at all,” perhaps it is time to reassess. More generally, directors might want to assess their institution’s internal mechanisms for accountability to avoid the complacency that undermined SVB. And that leads me to my third and final point.
Effective challenge in corporate governance
The recent failures of SVB and Signature should remind us about the importance of effective challenge in corporate governance.
Both SVB and Signature enjoyed long runs of growth and profits. I can’t help but think that sustained success made it more difficult for their boards and senior management to recognize weaknesses. I don’t know enough about the internal workings of those firms to be more specific than that, and I’m not an organizational psychologist. But I’ve been a lawyer long enough to know that one of the most valuable contributions a lawyer can make is asking the questions that others want to avoid. Lawyers add a lot of value to organizations as “devil’s advocates”—not only instructing clients what’s permissible, but challenging assumptions and questioning whether business decisions are wise.
Within SVB, for example, I wonder, who was asking the tough questions? Who was challenging decisions about the bank’s balance sheet and interest rate risk—in particular, its reliance on large, uninsured deposits and the ballooning of longer-duration investment securities as a percentage of its total assets—and asking, candidly, can this hurt us? If so, how much? And how can we fix it?
This is, perhaps, where the lack of a chief risk officer translated into a real, negative impact on the bank. Though SVB had a risk function, it was headed in the absence of a chief risk officer by the chief executive officer. That does not strike me as an effective way to challenge the risks undertaken to promote growth, since growing a business is a chief executive’s job. I question whether, lacking a leader, SVB’s risk management team had a clear, credible voice capable of communicating critical feedback to the management team and, ultimately, to the board. The chief risk officer plays that role. Without that officer, perhaps no one was in a position to encourage the bank to slow down and focus on managing some important risks in the midst of its rapid growth.
Looking ahead, the lesson from SVB is not to undervalue effective challenge and the individuals who provide it. This is also an observation raised repeatedly over the nine years of the New York Fed’s culture initiative—our effort to focus industry attention on group norms that affect individual behavior within financial institutions. I realize it can be inconvenient to hear critical feedback, and it is easy to discount effective challenge when everything seems to be going well. But that can be when professionals like risk officers, compliance officers, and lawyers are most valuable—pointing out risks that are obscured by profit. Don’t assume your organization can do without a chief risk officer—someone with the stature to raise concerns about risk management at the highest levels of the organization.
In closing, it is clear that banks and supervisors can learn important lessons from the demise of SVB and Signature Bank. The Barr Report is evidence of the Fed’s eagerness to learn and willingness to grow—which makes me optimistic about our ability to deter or minimize similar problems in the future. I hope the industry follows suit, and that banks assess themselves against the examples of SVB and Signature Bank. I am also confident that, when problems arise in the future, the Fed will have gained valuable experience through the BTFP on how to minimize stability risks to the financial system.
Again, thank you very much for the opportunity to join you all today.