Communications of the ACM
The Fall of Silicon Valley Bank
Silicon Valley Bank's closure represents the second-largest bank failure in U.S. history, and the largest since the 2008 financial crisis.
Credit: Nikolas Kokovlis/NurPhoto/Getty Images
On March 10, 2023, Silicon Valley Bank (SVB) failed after investors and customers tried to pull $42 billion in deposits from the bank the day before. This run on the bank's deposits represented about 24% of the bank's deposits at the time, and the bank could not meet these obligations. This ultimately led the California Department of Financial Protection and Innovation to close SVB.
SVB's closure represents the second-largest bank failure in U.S. history, and the largest since the 2008 financial crisis.
The collapse of SVB was a major shock to the financial world and raised important questions about risk management, banking regulations, the impact of social media on public companies, and a host of other issues.
SVB catered to technology startups, many of which had millions of dollars in deposits and loans from the bank. The bank grew rapidly in recent years as startups depended on the bank for funding. But a recent decline in startup funding and rising interest rates led to a decrease in the value of SVB's investments. Some experts believe SVB's overreliance on startups was a major factor in its demise.
Candice Nonas, Managing Consultant at RGP, who advises banks on risks and regulatory issues, said, "As far back as 2016, there was this desire among banks to grow and grow rapidly. Some banks, including Silicon Valley Bank, took a strategy that brought about concentration risk, whether it was exposure to cryptocurrency and Bitcoin, private equity, and high net worth depositors."
The Federal Deposit Insurance Corporation (FDIC) insures deposits up to $250,000, but many Silicon Valley Bank customers had funds much higher than this amount with the institution. In fact, more than 90% of SVB's deposits were uninsured—the highest rate of uninsured deposits of any U.S. bank, according to an analysis by S&P Market Intelligence. When news broke that SVB was raising capital to shore up its assets, customers quickly began pulling funds from the bank out of fear of losing the uninsured portion of their deposits.
According to the Federal Reserve's review of the SVB failure:
The combination of social media, a highly networked and concentrated depositor base, and technology may have fundamentally changed the speed of bank runs. Social media enabled depositors to instantly spread concerns about a bank run and technology-enabled immediate withdrawals of funding.
Silicon Valley Bank was one of several major banks recently shuttered, with assets and deposits moving to other institutions. New York City-based Signature Bank was closed just three days after SVB, while First Republic Bank was taken over by regulators less than a month later. According to The New York Times, the assets of these three banks totaled $532 billion, more than the combined assets of all 25 banks that failed during the 2008 banking crisis.
The FDIC ultimately covered Silicon Valley Bank's customer deposits, but the failure impacted thousands of companies. Large and high-profile firms were not immune from the bank's failure; Etsy said it could not send daily deposits to sellers on its platform, which impacted thousands of small businesses. Crypto company Circle tweeted that it had $3.3 billion in deposits held at the bank, while Roku said it had 26% of its funds with the bank.
Regulatory bodies, including the FDIC and Federal Reserve, faced criticism for their perceived lack of adequate oversight, especially given SVB's over-reliance on high-growth tech startups and heavy use of algorithmic risk assessment models. Regulators were aware of SVB's unique business model but underestimated the potential fallout from a cooling tech market and the inherent limitations of the bank's risk-assessment algorithms. "Bank supervisors were a little derelict in their duties, if you will, to take action based on the information and evidence they were seeing," said RGP's Nonas.
In response, regulatory bodies are now considering more stringent oversight of banks that heavily rely on specific industry sectors or technologies for risk assessment. This includes increased scrutiny of their risk-management practices and the necessity for more diversified portfolios. Moreover, regulators are also exploring ways to strengthen the risk-assessment parameters for artificial intelligence (AI) and machine learning models, ensuring these technologies are used responsibly and prudently in the banking sector.
Nonas believes regulators can leverage AI, too, to avoid future bank failures, but it ultimately comes down to bank supervision. "I think that bank supervisors can gain insight into the trajectory of a bank's business by using artificial intelligence. But as I look back forensically at what happened at these banks, aside from the quick transmission of data, which in my opinion we're not going to be able to stop, it was more human error. Core management and a weak infrastructure were the Achilles heels of these four banks that recently went out of business."
Mark Broderick is a banking and payments industry analyst with Panoramic Research.
No entries found