People line up outside of a Silicon Valley Bank office on March 13, 2023 in Santa Clara, California. Days after Silicon Valley Bank collapsed, customers are lining up to try and retrieve their funds from the failed bank. The Silicon Valley Bank failure is the second largest in U.S. history.
Image: Justin Sullivan/Getty Images
The bank run that led to the stunning collapse of Silicon Valley Bank late last week continues to send shivers through the American financial system. SVB, the Santa Clara, California-based bank that catered to the tech industry, was the biggest US lender to fail since the 2008 global financial crisis—and was the second-biggest to fail ever.
Analysts say SVB was largely unprepared for the Federal Reserve’s aggressive interest rate increases, which shrank the value of its investments. As word spread quickly online that the bank could be in trouble last week, customers withdrew $42 billion in a single day, leaving the bank with a $1 billion negative balance, according to a regulatory filing by the company.
While financial regulators have announced that the US will guarantee all deposits at SVB, its collapse has spooked customers at other banks and raised concerns about other financial institutions. We asked Harvard Business School faculty who study banks: What does the failure of SVB say about the current state of the banking industry? Here’s what they said.
Victoria Ivashina: Banks are ‘fundamentally fragile.’
Much has been said already about the textbook nature of the deposits run on SVB, and the subsequent run on other regional banks. Many observers postulate that the vulnerability was hiding in plain sight (an unsettling thought), a result of a combination of COVID government stimulus followed by a series of rates hikes. I would add other contributors: general uncertainty exhaustion after several years of dealing with surprises ranging from shortages and runs on basic goods to the Fed’s limited ability to control or even forecast inflation. Something will also have to be said eventually about the profitable business of inciting runs that some hedge funds have been up to. While all of these factors likely played a role, this narrative oversimplifies a few points and plays into the panic.
Banks are fundamentally fragile, and as such, are prone to self-fulfilling prophesies. Deposit insurance has been effective in reducing deposits runs, but the truth is that—once the confidence is eroded—banks tend to face revolving credit runs and market funding runs. The deposits run in the US might feel like a thing of the past, but the history of the 2008 crisis saw many such examples. Regulation and supervision help moderate the depth of the shock that banks can withstand before a run could be unleashed, but they cannot eliminate the possibility of such a run. The relevant question is then: What has triggered the SVB run? This is where things get more complicated, and—the good news—they also get more idiosyncratic.
We actually do not know much about SVB’s exposures, since they fell below the Fed’s threshold for annual collection of Form FR Y-14A Capital Assessments and Stress Testing. At the moment, we also cannot track what fraction of SVB’s deposits was connected to the local venture capital (VC) ecosystem that it was serving. But we do know that that this bank was different, and it is highlighted in its name and in the history of its public statement.Also read: SVB and Credit Suisse failure, and how it will affect your equity portfolio
For example, it is very clear that there is no other bank with over half of its loan portfolio dedicated to private equity subscription lines—that is, lines secured by VC and private equity fund commitments but ones used to fund investments in the short- to mid-term in lieu of capital call. It appears that this unusual specialized business model might have introduced significant correlation between large deposits and its business of revolving credit, on which it had substantial exposure. To add to that, the SVB leadership’s unorthodox approach to management of liquidity pressures clearly backfired.
The bottom line: The art is not to predict whether a run might take place, but when. The triggers of the run on SVB were no more apparent than the runs we have seen in the crypto space. The swift government actions to back regional banks, in my view, are commendable, although more targeted to managing general confidence than reflecting the isolated nature of the SVB troubles. They should be effective in reestablishing confidence in the broader financial system.
As to SVB, its credit specialization in the VC space is leaving an important void. I am less concerned about subscription lines, although this might temporarily bounce into pension liquidity. These events, however, are likely to have other lasting effects on the VC industry side, a subject for another day.
Victoria Ivashina is the Lovett-Learned Professor of Finance and head of the Finance Unit.Also see: Silicon Valley Bank: The rise and fall of tech industry's favourite finance house
Erik Stafford: Interest rate exposure and customer withdrawals were devastating
SVB was unique in some ways, but also did some pretty normal bank activities. The business of banking involves investing in assets that are a little bit longer-term and a little bit riskier than the liabilities used to fund these investments. Banks invest in assets like US Treasury bonds, mortgages, and corporate loans. The liability side of their portfolio is a combination of checking and savings deposits, CDs, and debt.
Most deposit rates reset fairly often, so they represent relatively short-term borrowing for the bank. The equity receives the net cash flows and realizes the net gains and losses from value changes. Bank annual reports emphasize that bank earnings or cash flows are relatively insensitive to changes in interest rates.
Additionally, the stability of net interest margin (interest received minus interest paid) is pointed to as evidence that banks have little interest rate exposure. Many bank loans to businesses have floating interest rates, but most bank loans are mortgages, which tend to have fixed interest rates, and US Treasury bonds have fixed interest rates, too. Of course, because the interest payments are fixed, the value of these assets is sensitive to changes in market interest rates.
From an investment perspective, the value sensitivity to changes in interest rates is the most relevant. Note that it is typically difficult to hedge both the value and cash flow interest rate exposure, so the fact that the cash flows appear insensitive to interest rate changes suggests that the value-based interest rate exposure has not been hedged.
Banks are highly levered, which magnifies the asset risk exposures for the equity. Suppose that bank assets resemble two-year bonds, interest rates increase quickly from 1 percent to 5 percent, and banks have equity of $20 for every $100 of assets. In order for the two-year bonds to earn the new market rate of 5 percent, their value will need to fall by about 7.5 percent, so assets of $100 now have a market value of $92.50. Because of the leverage, this $7.50 loss per $100 of assets represents a loss of $7.50 per $20 of equity, or around -38 percent.
One unique feature of SVB was their large uninsured deposit base. It is often argued that the present value of the ongoing profits that will be generated from deposit accounts effectively hedge the value-based interest rate exposure of the bank assets. In other words, the value of these accounts to the bank increase when interest rates increase, which offsets, at least partially, the interest rate exposure of the assets described above. One interesting feature of the SVB situation is that the value of their deposits fell as interest rates increased in a somewhat unexpected way. Well before the bank run last week, their deposit customers had been withdrawing money to pay their operating expenses, while their deposit inflows from new VC funding slowed, which produced large net withdrawals of deposits.
In this case, deposits appear to have added, rather than hedged, value-based interest rate exposure. The bank run was devastating for SVB, but the real problems that triggered this event were the underlying interest rate exposure and the slow withdrawal of deposits. SVB was forced to issue a large amount of equity, which brought a lot of attention to their situation. There is now a lot of attention on the situation at all banks. The underlying interest rate exposure is common across banks, so some drop in bank equity values is appropriate. An important question to be determined is whether there are other banks who have been experiencing a slow withdrawal of customer deposit balances. This is quickly being investigated by investors and regulators.
Erik Stafford is the John A. Paulson Professor of Business Administration.
[This article was provided with permission from Harvard Business School Working Knowledge.]