Fed knew or should have known about Silicon Valley Bank risks: independent review of failure causes

A new independent review of the Silicon Valley Bank collapse concluded that Federal Reserve supervisors knew or should have known about the bank's key vulnerabilities as early as March 2022—about a year before its failure. However, the regulator did not achieve sufficiently rapid reduction of the risks that later played a critical role in the crisis.

The initial findings of the review, conducted by outside firm Starling Advisory Group, were presented on September 18 by Fed Vice Chair for Supervision Michelle Bowman. The analysis is meant to answer the main question left after the 2023 banking crisis: if Silicon Valley Bank's problems were visible in advance, why didn't supervision force the bank to address them before a massive deposit run began.

Silicon Valley Bank was closed by regulators on March 10, 2023. At the end of 2022, its assets were about $209 billion, and deposits were more than $175 billion. The bank was especially closely tied to technology companies, startups, and venture capital funds.

According to the new findings, SVB simultaneously accumulated several dangerous vulnerabilities. Unrealized losses on its securities portfolio exceeded the bank's capital, 94% of the deposit base was uninsured, and deposits themselves were highly concentrated among companies linked to the venture and technology sector.

An additional problem was liquidity. The review indicates that the bank was not operationally prepared to quickly obtain needed funding through the Fed's discount window—a mechanism through which U.S. banks can borrow money from the central bank in a stressful situation.

The mechanism of the crisis was relatively simple. During a period of low interest rates, SVB invested a significant portion of incoming deposits in long-term securities. When interest rates rose sharply, the market value of those assets fell. As long as the securities were not sold, the losses remained largely accounting losses, but with a massive outflow of deposits, the need to quickly raise cash turned these losses into a real threat to the balance sheet.

It is here that the new review makes one of its most important findings: Fed supervisors, according to Starling's assessment, knew or should have known about these vulnerabilities as early as March 2022. Despite that, they did not take sufficiently quick and decisive action to force the bank to reduce interest rate risk and the concentration of other vulnerabilities.

This does not mean that responsibility for the failure is shifted entirely to the regulator. The Fed's own internal review in 2023 concluded that Silicon Valley Bank's management and board of directors themselves failed to manage risks. That same report also acknowledged supervisory problems: Fed staff were too slow to escalate pressure on the bank even after serious signals appeared.

Thus, the fundamental novelty of the current analysis is not in the admission of supervisory mistakes themselves. The Fed acknowledged those three years ago. The new dispute concerns the reasons why the regulator acted so slowly.

The 2023 review linked part of the problem to the easing and tailoring of bank regulation after the 2018 law, as well as to a less stringent approach to supervision. The new independent review reaches a different conclusion: according to its preliminary assessment, the delays were not caused by the requirements of that law or by directives from prior Fed leadership to reduce supervisory intensity.

Instead, Starling points to the internal culture of the supervisory mechanism itself. According to Bowman, staff often considered it safer not to take action until they were fully certain that the chosen measure was correct. The situation was aggravated by unclear allocation of authority: staff did not always understand who exactly had the right to make a final decision and take responsibility for it.

Another unusual finding concerns the role of social media. After SVB's failure, one popular explanation for the speed of the bank panic was information that spread instantly among investors and customers through social media. However, an analysis by Charles River Associates conducted for Starling found no evidence that social media triggered or substantially accelerated the depositor run. According to the presented data, 96% of social media discussions of the crisis appeared after the bank's failure had already become virtually inevitable.

For the Fed, the review already has practical consequences. The regulator stated that it is changing its approach to bank supervision: examiners should focus earlier on threats that could cause substantial damage to a bank's financial condition or to the stability of the system, rather than excessively focusing on procedural violations.

In addition, supervisory teams will be required to report monthly to management on cases where examiners doubt whether there are sufficient grounds for intervention. The idea is to allow concerns to be escalated more quickly, without waiting months for additional evidence to accumulate.

For bank customers and investors, the new review does not change deposit insurance rules or interest rates. Its significance is broader: the SVB story shows how dangerous a combination of large uninsured deposits, customer concentration in one sector, losses on long-term assets, and too-slow supervisory response can be.

That is why the Silicon Valley Bank analysis remains relevant more than three years after its failure. The question now is not only why one large bank made critical mistakes, but also whether the regulator is capable of recognizing a similar combination of risks early enough and intervening before the problem turns into a new banking crisis.

At this point, these are still the initial findings of the independent review. The Fed calls Starling's work the first stage of a broader analysis of the 2023 bank failures, so the final assessment may be supplemented as subsequent materials are published.

Sources: U.S. Federal Reserve, Independent Review of the 2023 U.S. Bank Failures, Fed 2023 review, FDIC