Five Thoughts on the California DFPI Report on the Failure of SVB
Even with the benefit of hindsight, it is difficult to identify any one action or set of actions that regulators could have taken that would have been guaranteed to prevent SVB’s collapse.
So begins the “Key Findings” section of the California Department of Financial Protection and Innovation’s review of its oversight and regulation of Silicon Valley Bank.
The review is fairly high-level and appears to be intended to reach a broader audience. Nonetheless, after plowing through some general background sections on what banks do (“Banks are privately-owned financial institutions that, generally, are chartered to accept deposits and make loans”) and how California supervises them (“Starting in 1857, banking enterprises in California were granted charters under the General Corporation Laws….”) the report does get around to making a few notable points. This post highlights five of them.
The Report Tries to Emphasize the DFPI’s Secondary Role
The report explains that under the dual banking system the “DFPI works with the appropriate primary federal regulator to supervise and examine California-chartered banks.” The DFPI then takes pains to note, however, that though it works alongside the federal regulators, this does not mean responsibilities are equally distributed, particularly when it comes to banks the size of SVB.
“The DFPI and SVB’s primary federal regulator, the Federal Reserve Bank of San Francisco (FRBSF), had divided their respective oversight and supervisory activities over SVB in such a way that the FRBSF had assumed a lead role for many supervisory activities.”
“After SVB entered the Large and Foreign Banking Organization (LFBO) portfolio in February 2021, the FRBSF took the lead in examinations.”
“The FRBSF took the lead role in the examination of issues regarding SVB’s liquidity with a particular focus on Liquidity Risk Management.”
“In November 2022, the FRBSF, which was in the lead due to its expertise in this area, alongside the DFPI found that SVB’s Interest Rate Risk (IRR) simulations were not reliable and required improvements.”
“SVB’s liquidity positions were examined independently by the FRBSF due to their expertise and resources available.”
“For the 2022 supervisory cycle, the FRBSF was responsible for the evaluation of SVB’s hedging activities due to its expertise in reviewing market risk and conditions.”
“The [DFPI’s] Dedicated Examiner-in-Charge oversaw the DFPI’s supervision of SVB. While the Dedicated Examiner-in-Charge’s responsibilities typically include developing the examination plan, assigning work to examiners, and communicating with bank management and federal counterparts, the Dedicated Examiner-in-Charge assigned to SVB did not develop the examination plan. Instead, the Dedicated Examiner-in-Charge reviewed the examination plan developed by the FRBSF and provided feedback given the FRBSF’s expertise and resource available.”
The report explains that initially some of this deference to the FRBSF was driven by resource constraints.
For the DFPI’s part, a total of 12 DFPI examiners participated in supervising SVB, but only two were specifically dedicated to SVB. In contrast, the FRBSF had a team specifically dedicated to SVB that grew to approximately 20 examiners, with subject matter expertise examining large banks in excess of $100 billion in total assets.
And when the DFPI team asked for additional resources, the initial response was that the FRBSF had already made additions to its team supervising SVB, such that further additions to the DFPI team were unnecessary.
Toward the end of 2021, the DEIC and the DFPI Financial Institution Manager assigned to SVB elevated the need to divert resources to SVB due to its complexity and large asset size. These discussions did not result in assigning additional staff or an Operations Manager being allocated to the DFPI exam team, as the DFPI determined that the FRBSF had already brought in more examiners and covered the reviews typically assigned to an Operations Manager.
To his or her credit, the Dedicated Examiner-in-Charge continued to request additional resources. and DFPI management eventually agreed. But even after they agreed, they found that “examiners with the necessary experience and skill sets were already assigned to key roles in other bank examinations,” and so the allocation of additional staff was delayed. In the end, prior to the collapse of SVB the DFPI was still planning to add an additional full-time examiner but due to “competing staff demands, turnover, and employee development” had not yet been able to do so.
Going forward, the report includes a commitment from the DFPI to “review internal staffing processes … for banks with assets above $50 billion commensurate with accelerated institution growth or increased risk profile” and says it will “evaluate future resource needs through the state budget process.”
All in all, though, it is a fairly underwhelming picture of bank supervision in the largest state in the United States.
From one perspective this would not be a huge concern. If we were starting from scratch, it would make a lot of sense that supervision of banks in the United States should generally be managed at the federal level, with state-level supervisors playing at most a supporting role (if any). And although the report is obviously written from a self-interested point of view, I do not take issue with the DFPI’s decisions to defer to Reserve Bank staff that had, or at least should have had, more expertise in examining large banks.
But an explicitly federal-led bank supervision system is not what we have. Instead, as the report explains, we have dual banking system that “is unique to the United States.” Obviously this is not changing any time soon, but if you are going to contend, as the report gamely does in one place, that “[i]n certain situations, a state banking regulator such as the DFPI may be more nimble or responsive than a federal regulator” because it can “offer a unique ‘local’ perspective,” then it seems incumbent on the states to make sure that their regulators have sufficient resources to carry out this work.
California is Making a Change to Supervision That May Not Be Helpful
The report describes a few “supervisory enhancements” that the DFPI has already put into motion. For example:
The DFPI will add another level of supervisory review to exam reports before they are issued. Supervisors will assess the exam reports to ensure that the examiner has identified risks appropriately and that concerns are escalated if necessary. This will allow the DFPI to ensure it is identifying as many risks as possible and will help to ensure timely remediation of concerns. Additionally, this step will help state examiners in holding bank management to swifter action.
The DFPI describes this as a way of addressing its key finding that SVB was “slow to remediate regulator-identified deficiencies,” including because “regulators did not take adequate steps to ensure SVB resolved problems as fast as possible.”
So I guess the idea here is that adding another layer of supervisory review will give more senior-level DFPI staff the ability to review the work of front-line examiners, and to ask questions about whether those front-line examiners are being sufficiently demanding of the banks they supervise in terms of things like remediation plans and timing.
My worry, though, is that based on the Federal Reserve Board’s report on SVB and the FDIC and NYDFS reports on Signature Bank, there is a serious risk this could wind up being counterproductive. For instance, among various other supervisory shortcomings it documents, the Board’s report on SVB tells the story of an MOU that “took more than seven months to develop.” It also relates an anecdote about a delay until August 2022 in issuing supervisory ratings from 2021 that “illustrates how the normal supervisory practices did not keep up with SVBFG’s rapid expansion.” Per the FDIC and DFS reports, exam findings were similarly delayed for Signature Bank.1 How much would it have helped to add another layer of review to either of these processes?
Also, maybe this is me unfairly taking issue with the loose drafting of a single sentence, but should the goal of an examination really be to “identif[y] as many risks as possible”? The DFPI report, like the Board report before it, makes clear that quite a few risks were identified at SVB prior to its failure, but many of these risks had little or nothing to do with SVB’s failure, and those that did prove to be extinction-level risks were not appropriately treated as such.
The DFPI Plans Changes to Supervision of California-Chartered Banks (One of Them, At Least)
As another supervisory enhancement, the DFPI intends to increase its focus on uninsured deposits.
The DFPI will increase its focus on banks’ uninsured deposit levels, in addition to continuing to monitor key indicators such as banks’ concentration of uninsured deposits by industry. DFPI will review requirements for California-chartered banks and take steps to require these banks to provide the DFPI with a plan for more proactive management of uninsured deposits, and will make them subject to enhanced testing, in closer collaboration with the DFPI’s federal partners.
Banks over $50 billion in assets will get special attention.
In particular, for banks with assets over $50 billion, the DFPI will review the circumstances in which these banks must provide a written assessment evaluating their management of uninsured deposits, including their historic and projected levels, whether these deposits are concentrated in a specific depositor or industry group, and the bank’s plan to mitigate any liquidity risk associated with such concentration.
As a technical matter there is only one California-chartered bank that (1) had over $50 billion in assets as of year-end 2022 and (2) has not failed,2 but I would expect a few banks below this threshold will be subject to a similar analysis.
SVB’s Responses to Exam Findings and Remediation Work Were “Positive”
In a post today criticizing the Federal Reserve Board’s report on SVB, the Bank Policy Institute asks about what SVB did in response to supervisory findings and, more importantly, how the Board reacted to those efforts.
Finally, we note that another key way that one might assess whether supervisors were appropriately focused on actual risk (and not just risk processes) is to evaluate what views the Federal Reserve expressed to SVB about its effort to remediate the Fed’s supervisory findings – that is, SVB’s responses to the Fed’s MRAs and MRIAs. This material would illuminate whether supervisors were focused on ensuring that SVB was actually mitigating underlying risks as it improved its processes, as opposed to simply ensuring that SVB was checking the relevant compliance boxes that examiners had established. Unfortunately, the Federal Reserve has made none of that material publicly available, nor does the report describe it in even general terms.
Like the Board’s report, the DFPI report spends very little time on this question, but even so there are a couple of passages that were sort of interesting to read this afternoon with the above complaint by BPI in mind.
On December 1, 2021, SVB responded to the November 2, 2021, supervisory letter and provided a Management Action Plan. SVB took numerous steps to address MRIA #1. Actions taken included hiring consulting firm Ernst and Young to provide an independent risk management assessment and hiring a new head of liquidity risk management.
On December 29, 2021, SVB submitted a follow-up letter to demonstrate that it believed its actions satisfied the requirements to update its project plan and gap assessment and ensure sufficient staffing and resources were allocated to execute the plan by the December 31, 2021, due date. While a new plan and staffing were in place, MRIA #1 would remain open until regulators could determine whether the plan and the staffing were effective. The risk management plans and the staffing to implement them required testing over a longer period of time to identify their response to changes in the market.
On March 31, 2022, SVB requested and was granted an extension to October 31, 2022, to fully remediate MRIA #2. According to a May 4, 2022, letter from the FRBSF to SVB, the FRBSF did not object to the extension, partly due to SVB’s management taking prompt action and submitting materials that showed progress towards remediation. […]
Through the continuous monitoring process, SVB kept the FRBSF and the DFPI apprised of its continued efforts to remediate the MRIAs and MRAs from November 1, 2021, and discussed in the August 12, 2022, letter. In internal discussions leading up to the 2022 CAMELS ratings, SVB’s progress toward remediating the MRAs and MRIAs was described as “positive” by the FRBSF.
None of the above is smoking gun stuff exactly, but these passages from the DFPI report, high-level though they are, seem at least a little inconsistent with a story of FRBSF examiners desperately wanting to hold SVB more accountable and being prevented from doing so by political or due process considerations.
The Chief Risk Officer Vacancy - Still a Mystery to the DFPI
I admit I had missed this until reading the BPI post earlier today, but the GAO report from a few weeks ago on the failures of SVB and Signature Bank sure seems to suggest that SVB’s decision to replace its Chief Risk Officer was driven in one way or another by the Federal Reserve:
Although Federal Reserve staff stated that the Federal Reserve’s supervisory actions compelled SVB to take steps including replacing the Board Chair, Chief Risk Officer, and Treasurer and revising its incentive compensation program to incorporate risk management as a formal assessment criteria, its supervisory actions were inadequate given the bank’s known liquidity and management deficiencies.
DFPI’s report also brings up the CRO vacancy but, as with certain other aspects of the report, mostly does so only to make clear that they had nothing to do with it and that any questions about it should be directed elsewhere.
SVB employed a Chief Risk Officer (CRO) for almost six years from 2017 to April 2022. The DFPI was notified in February 2022 that SVB intended to terminate the CRO’s employment in April, but the Department was not informed that they remained at SVB in a transition role until October.
The DFPI does not know the circumstances that led to their termination or continued employment in a transition role. As discussed above, issues with risk management and the board’s failure to hold senior management accountable for executing sound risk management plans were the basis for MRIAs, MRAs, and ratings downgrades.
Thanks for reading! Thoughts, challenges, criticisms are always welcome at bankregblog@gmail.com.
The NYDFS report is a report about the New York state regulator’s supervision of an FDIC-supervised state nonmember bank and so is not directly relevant to California or SVB, but there are also California state nonmember banks supervised by the DFPI and the FDIC, to which some of the dynamics discussed therein could be relevant.
