The FDIC's Report on the Failure of Signature Bank
Earlier today I wrote up my initial thoughts after reading through the Fed’s report on Silicon Valley Bank. This post does the same thing after reading through the FDIC’s report on Signature Bank.
Mistakes Were Made
In the opinion of the FDIC’s chief risk officer tasked with producing the report, the FDIC’s approach to supervising Signature Bank left some room for improvement:
“In retrospect, FDIC could have escalated supervisory actions sooner”
“Additionally, examination work products could have been timelier”
“[C]ommunication with SBNY’s board and management could have been more effective”
“The FDIC could have been more forward-looking and forceful in its supervision”
“In retrospect, the FDIC could have acted sooner and more forcefully to compel the bank’s management and its board to address these deficiencies more quickly and more thoroughly”
“With respect to the February 2023 meeting with SBNY’s board and management, the lessons learned review concluded the [New York Regional Office] could have delivered a stronger message regarding repeat funds management weaknesses as well as preliminary findings from ongoing targeted reviews and trends of concern.”
A few examples of the report’s findings supporting the above conclusions:
The passing grades given to Signature Bank by the FDIC in terms of supervisory ratings “appeared incongruent” in some instances with the actual findings included in the reports of examination accompanying those ratings. Pages 22-23 of the report quote a few excerpts of ROEs that began by saying “Board and management performance remain satisfactory,” and then went on to describe various practices that, in retrospect, look quite unsatisfactory.
As an example of an examination work product that could have been timelier, the report tells a bleak story about a 2021 roll-up exam that made findings as of December 31, 2021 but for which a transmittal letter with those findings was not submitted to Signature Bank until mid-December 2022. Along similar lines, a corporate governance targeted review was conducted in March 2022, but the supervisory letter formally conveying the FDIC’s findings was not delivered until January 2023. This, along with an extension of time to respond granted by the FDIC at Signature Bank’s request, meant that the bank failed in March 2023 before the FDIC ever received management’s response to these findings which dated from as early as 2021.
The report notes that the FDIC’s “[Large Bank Supervision] Branch in headquarters provides nationwide oversight and quality control of significant examination activities at institutions in the CEP.” But the report then explains that this oversight and quality control does not necessarily manifest itself in the way you might think.
In 2022, an analyst in LBS recommended a further downgrade in one of Signature Bank’s supervisory ratings in light of its “increasing risk profile of the institution, rapid capital call lending growth, increased funding stress, and digital assets-related deposit activity.” The New York Regional Office, however, “did not accept this recommendation.” No further explanation is offered, and the report takes no position on which group was right. (The report does note that the NYRO went on to downgrade the bank in a later quarter.)
Also in 2022, LBS recommended that the bank’s management rating be downgraded from 2 to 3 as part of the 2021 roll-up exam. Again the NYRO disagreed with LBS’s “approach and timing,” including because the NYRO “wanted to maintain the linear integrity of the 2021 examination cycle.” That is, NYRO management was concerned about 2022 findings bleeding into the 2021 ratings, which had already been delayed by an “extremely extended timeline.”
What caused this?
As discussed earlier today, the Federal Reserve Board attributes its supervisory failures in relation to SVB, in part, to “a shift in the stance of supervisory policy [that] impeded effective supervision by reducing standards, increasing complexity, and promoting a less assertive supervisory approach.”
The FDIC report takes a different tack. To the extent any culprit at all is blamed for the FDIC’s failure to be appropriately “forward-looking and forceful,” the report mainly points the finger at inadequate resourcing at the New York Regional Office and a related failure by headquarters to address resourcing needs.
NYRO management is responsible for ensuring that banks in the region are adequately supervised. While resource shortages were a significant factor in the supervision of SBNY, NYRO management is ultimately responsible for prioritizing and risk-focusing the use of scarce resources […]
RMS headquarters is responsible for ensuring that the regions have sufficient examination resources with the necessary skillsets and experience to effectively supervise their portfolio of institutions. […]
RMS headquarters is also responsible for addressing regional office resource needs; making prioritization decisions across the regional offices and reallocating examination resources based on a nationwide view of risk; and ensuring examination programs, such as the CEP, are effective and achieve their intended objective
I do not think this proves that the disempowered supervisors thesis discussed in the earlier post on SVB is wrong, but it does suggest questions worth exploring.
If it is indeed the case that supervisors at the Board and its Reserve Banks were cowed into taking a less assertive supervisory approach, why did FDIC supervisors not succumb to the same influences? An easy answer would be that the FDIC and the Board were led by different individuals, but this distinction between the banking regulators is not one that many were making at the time, which makes me skeptical of attempts to use it to backfill an explanation.
The FDIC experiences the post-COVID labor market
Given the report’s conclusion that resource issues contributed materially to some of the FDIC’s supervisory failures, there is a lengthy discussion of the FDIC’s attempts to staff up the New York Regional Office. Like a lot of employers during this time period, the FDIC evidently found that good help was sometimes hard to find:
“The [Examiner-in-Charge] position is of critical importance. The NYRO had difficulty finding qualified staff to serve that role. … The NYRO advertised the EIC vacancy two separate times during late 2021 and early 2022 but identified no qualified applicants.”
“Several times during this period, the NYRO advertised for multiple team vacancies and filled only one vacancy due to weak rosters or lack of interest”
“The New York region has 61 authorized LFI positions. Since 2020, an average of 40 percent of the LFI positions have been vacant or filled by temporary staff.”
“The use of temporary staff in supervising LFIs is problematic. Multiple NYRO officials indicated that it takes at least six months to become familiar with large institutions such as SBNY and the use of temporary staff, while necessary, is inefficient and presents continuity challenges.”
“NYRO and headquarters RMS officials identified multiple reasons for LFI staffing challenges including the high cost of living in New York, competition from other regulators and private sector firms that can pay more for talent than the federal government, and competition for LFI staff from other FDIC Divisions and headquarters, which may offer greater work-life flexibilities or higher-graded position”
“[G]iven the nature and difficulty of LFI work, particularly the EIC role, a sentiment exists that there are other similarly-graded positions available within the FDIC that are less difficult or that come with less responsibility”
All of the above makes a certain degree of sense, even if you might reasonably ask questions about how the FDIC, which does not rely on Congressional appropriations for funding, found itself in this position.
There was also paragraph in this section that I had more trouble wrapping my head around. The report notes:
Further, SBNY was frequently in the media spotlight, which required frequent interaction with RMS headquarters to respond to numerous information requests and media articles, further exacerbating resource challenges.
Making Up For Lost Time
The report observes that in light of Signature Bank’s “fundamental and recurring liquidity control weaknesses, … unrestrained growth, management’s slow response to address findings, and management deficiencies in other areas,” it “would have been prudent” to downgrade the bank’s management rating to unsatisfactory “as early as the second half of 2021.”
Appendix 5 to the report describes how the FDIC’s NYRO eventually concluded that downgrades were indeed appropriate and tried their best to make up for lost time.
First, on March 10, 2023, the NYRO “began preparing interim ratings downgrades to Liquidity (‘3’ to ‘4’), Management (‘2’ to ‘3’), and the Composite (‘2’ to ‘3’).”
But things continued to deteriorate. Thus, on March 11 the NYRO and New York Department of Financial Services decided that further downgrades were in order. They moved to “expand[] the in-process interim ratings changes and that night downgraded Liquidity, Management, and the Composite ratings to ‘5’ and Capital to ‘3’ based on management’s inability to properly identify, measure, monitor, and control the bank’s liquidity position.”
“At 5:30 p.m. EDT on Sunday, March 12, 2023, the NYSDFS closed SBNY and appointed the FDIC as receiver.”
SBNY’s Data Issues
The DFS has explained that its decision to shut down Signature Bank stemmed not, as Barney Frank has asserted, from concerns about or animus toward crypto but instead from the bank’s failure to “provide reliable and consistent data,” which “creat[ed] a significant crisis of confidence in the bank’s leadership.”
The FDIC’s report today gives further details about what was going on at the time.
As mentioned earlier, during 2021 and 2022, SBNY increased lending in the form of capital call/subscription loans. […] SBNY intended to pledge these loans to the Federal Reserve Bank of New York (FRB-NY) as collateral for Discount Window lending. However, FRB-NY would not accept the loans as collateral because they were not eligible as many of them had foreign limited partners. SBNY pursued efforts to pledge these loans for months, hiring two law firms to make the case for FRB-NY to accept the loans. During the weekend SBNY failed, management again tried, unsuccessfully, to pledge this portfolio to FRB-NY. SBNY also unsuccessfully tried to identify alternate entities that would accept the portfolio as collateral for a borrowing line. Even though SBNY management knew they did not have a formally confirmed avenue to obtain liquidity from this portfolio, they continued to try to include these loans in collateral calculations just hours before the institution failed.
In February 2023, examiners questioned SBNY’s regulatory reporting of pledged securities, which required SBNY to refile its year-end 2022 Call Report with a multi-billion upward adjustment to pledged securities. Prior to and during the weekend of SBNY’s failure, examiners informed SBNY that pledged securities were still misstated on the bank’s liquidity monitoring reports, because management continued to under-report pledged securities and overstate on-balance sheet liquidity on information provided to examiners. This issue persisted until the day SBNY failed, when management finally produced an accurate report of unpledged securities.
Reputation Risk
In the first half of the 2010s, a few FDIC employees perhaps got a bit carried away with their supervisory approach to banks providing services to payday lenders.
The FDIC’s concerns about this sort of activity and the related actions taken by supervisors were grounded, in part, on reputation risk, which an Inspector General report at the time described as the “risk that potential negative publicity regarding a financial institution’s business practices could cause a decline in the customer base, costly litigation, or revenue reductions.”
In prior posts I have voiced open skepticism of claims that the banking regulators’ approach to crypto represents an Operation Choke Point 2.0. And I continue to believe that these claims are, at best, dramatically overstated.
On the other hand… the FDIC’s report includes a section with the heading, “Reputation Risk and Contagion from Crypto Industry Turmoil,” the first sentences of which say:
SBNY’s board and management employed a strategy of rapid growth and expansion into the digital asset markets. The strategy exposed SBNY to greater susceptibility to liquidity, reputation, and regulatory risk due to the uncertainty and volatility of the digital asset space.
Later on, the report quotes a supervisory letter sent to Signature in January 2023 saying the same thing.
I have no argument (how could anyone?) with the conclusion that Signature Bank’s association with cryptocurrency ended very poorly for the bank.
My objection instead is with the reliance here on the concept of reputation risk, which I have never been convinced actually adds much to the discussion. A law review article by Professor Julie Andersen Hill articulates this point better than I could:
The nature of reputation risk makes it difficult to regulate in a way that adds meaningful value to the regulatory system. Reputation risk is often a derivative risk. Because bank regulators have broad powers over other more direct risks, reputation risk often does little work. When Wells Fargo employees illegally opened unauthorized accounts, they violated the law and created reputation risk. When banks violate anti-money laundering laws, they create reputation risk. When banks have credit quality problems, they create reputation risk. Enforcement actions in those situations do not need to be grounded in reputation risk. Enforcement can be grounded in other law. Reputation risk adds little to the regulatory toolbox.
I think the Signature Bank example illustrates this well. Were Signature’s troubles really a result of its reputation as a banker for the crypto industry, and thus its association with “public announcements regarding fraud and law enforcement investigations”? Or were they instead the result of the fact that its depositors realized, belatedly but correctly, that Signature had “funded its rapid growth through an overreliance on uninsured deposits without implementing fundamental liquidity risk management practices and controls”?
To see what I mean, take this paragraph from the report:
SBNY was also frequently associated with Silvergate in media reports, as these two banks were seen as most closely tied to the crypto industry. Following the March 1, 2023, announcement by Silvergate regarding the delay in filing its year-end 2022 financial statements and comments about its ability to continue as a going concern, SBNY once again experienced negative media attention, which raised questions about its liquidity position. The announcement on March 8, 2023, that Silvergate intended to self-liquidate placed additional pressure on SBNY’s liquidity.
I guess I understand how you could describe this as in some sense a reputational risk, but the “questions about [Signature’s] liquidity position” were good ones! They were not questions that had as their root cause “negative media attention” but rather were questions with a basis in the hard facts of Signature’s assets and liabilities. It was not Signature’s “reputation” of being a banker to a volatile industry that was experiencing a bank run and contagion that was the problem, but the fact that Signature Bank actually was one of the key bankers for that industry.
***
So long as we are talking about reputation risk, there was another part of the report I thought was funny or interesting. In Appendix 2, the FDIC traces through the bank’s supervisory history and describes the findings of various exam reports. The report notes that in the context of the 2020 exam cycle, Signature Bank had outstanding supervisory recommendations relating to “borrower reputation risk analysis.” According to the FDIC’s summary, the same concern remained outstanding during the 2021 exam cycle.
Unlike the Federal Reserve Board with respect to Silicon Valley Bank, the FDIC has not made publicly available the Signature Bank exam reports themselves, so it is not clear for which borrower(s) the FDIC had these concerns. You might have a few guesses, though.
Thanks for reading! As always, thoughts, challenges, criticisms are always welcome at bankregblog@gmail.com. I am sure there are interesting things this post missed.