The collapse of Silicon Valley Bank is the largest bank failure in the United States since the global financial crisis. The bank’s vulnerability was the result of having a high proportion of uninsured deposits and a large proportion of deposits invested in hold-to-maturity securities.
The failure of Silicon Valley Bank (SVB) in March 2023 has sent ripples through the global banking industry. Its collapse represents the second largest failure in US banking history (see Figure 1). It was quickly followed by the failure of Signature Bank, a New York-based bank that was heavily involved with the cryptocurrency industry.
Figure 1: Top ten largest US bank failures
Source: Quartz, 2023 and Federal Reserve, 2023
Note: The assets values are inflation-adjusted
SVB was founded in 1983. Headquartered in Santa Clara, it was banker to about 50% of all venture capital-funded technology and life sciences companies in the United States. It also had a subsidiary in London, serving tech companies in the UK and continental Europe.
In terms of size, it was the 16th largest bank in the United States at the end of 2022 (see Figure 2). Nevertheless, it was relatively small compared with the four behemoths of US banking – JP Morgan Chase, Bank of America, Citibank and Wells Fargo – and it had less than 1% of US bank assets.
Figure 2: Largest 20 US commercial banks by total assets ($ billion) at end of 2022
Source: Federal Reserve, 2023
Why are banks prone to failure?
Banks are intermediaries. They take in deposits – which are both their chief liability and chief source of funds. Banks then take these deposits and lend them to borrowers and invest them in a range of fixed-income securities such as government bonds. Figure 3 is a simple bank balance sheet that illustrates why banks are vulnerable to failure.
The first source of vulnerability is that demand deposits are redeemable on request. In other words, depositors can withdraw all their deposits from the bank at the push of a button. Time deposits can usually be withdrawn immediately, but there is typically a penalty for doing so, such as forgone interest payments.
The second source of vulnerability is that bank assets such as loans and securities are illiquid. In other words, it is costly to turn them into cash at short notice.
Banks take in deposits that have a short maturity and invest them in securities and loans that have a long maturity. The difference in interest rates between these two is how they make their money.
Figure 3: Example bank’s balance sheet
Assets (£) | Liabilities (£) | ||
Securities | 45 | Demand deposits | 45 |
Loans | 45 | Time deposits | 45 |
Cash | 10 | Capital | 10 |
100 | 100 |
But what happens if lots of depositors want to withdraw their money at the same time? Our example bank in Figure 3 holds cash to meet excess withdrawals, but this can be quickly exhausted. The bank would then have to sell its securities to meet withdrawals.
The issue with having to sell securities is that their values may have fallen since they were purchased by the bank. This can happen simply because interest rates have increased in the interim. A rise in interest rates pushes down the price of fixed-income securities. This could potentially mean that the bank does not have enough assets to cover its deposits.
This is where shareholder capital comes in: it helps to absorb the fall in the value of securities, acting as a buffer. But what happens when this capital is exhausted? At this point, the bank is insolvent.
Central banks and regulators have long recognised the fragility of banks and their susceptibility to bank runs. In the United States, standard deposit insurance is up to $250,000 per depositor, per insured bank (Federal Deposit Insurance Corporation, FDIC, 2023). The equivalent figure in the UK is £85,000 (Bank of England, 2023).
This reduces runs by small depositors such as happened to Northern Rock during the global financial crisis of 2007-09 (Shin, 2009). At that time, deposit insurance in the UK only insured 90% of deposits up to an upper limit of £35,000. As a result, withdrawing money from, or ‘running’, Northern Rock made sense for small depositors.
Why did SVB fail?
SVB was a banker to many tech and life sciences companies. This meant that its deposit base consisted of many customers whose deposits were well in excess of the $250,000 deposit insurance upper limit.
As Figure 4 shows, 93.9% of SVB’s deposits were uninsured and it had the second highest ratio among large US banks. This was its first vulnerability.
SVB’s deposit growth since 2020 had been impressive. The tech boom of 2020-21 meant that many of its tech customers were raising lots of money from venture capitalists and private equity investors, and depositing this money with SVB. As a result, its deposits grew from $65 billion in 2019 to $189 billion in 2021.
Figure 4: Top 10 large US banks by proportion of uninsured deposits at end of 2022
Bank | Proportion of uninsured deposits (%) | Loans + HTM securities / total deposits (%) |
Bank of New York Mellon | 96.5 | 31.2 |
Silicon Valley Bank | 93.9 | 94.4 |
State Street Bank and Trust Co | 91.2 | 40.1 |
Signature Bank | 89.7 | 93.3 |
Northern Trust Co | 83.1 | 54.5 |
Citibank | 77.0 | 64.6 |
CIBC Bank USA | 73.2 | 87.1 |
HSBC Bank USA NA | 72.5 | 47.4 |
City National Bank | 70.4 | 93.6 |
First Republic Bank | 67.7 | 110.6 |
Source: S&P Global
The boom in SVB’s deposits was mirrored in its share price, which rose from $150 in March 2020 to $730 in January 2022 (see Figure 5). This share price rise during 2020-21 matched the boom in share prices that took place in the US tech bubble (Quinn and Turner, 2021). The subsequent fall in the bank’s share price during 2022 also coincided with the bust in the share prices of US tech companies.
Figure 5: SVB share price (US $), 2018-2023
Source: S&P Global
During the boom, SVB took its new deposits and invested them in long-dated fixed-income securities such as US government bonds and mortgage-backed securities issued by US government agencies. Prevailing interest rates on these safe assets at the time were very low. SVB effectively locked away half of its assets for ten years by investing in these hold-to-maturity (HTM) securities.
As Figure 4 shows, SVB held 94.4% of its deposits in these HTM securities and loans. This was SVB’s second vulnerability.
The collapse of the tech boom from November 2021 onwards meant that many of SVB’s corporate clients were drawing down their deposits in 2022. The high proportion of assets tied up in loans and HTM securities meant that SVB had not left itself a lot of room for manoeuvre should deposit outflows increase. This meant that after exhausting its cash reserves and short-term liquid securities, the bank had to start selling its HTM securities.
The problem for SVB was that as well as the tech boom coming to an end, inflation was increasing and was not likely to be a transitory phenomenon. This meant that the US Federal Reserve (‘the Fed’, America’s central bank) and other central banks around the world had to raise interest rates to tame inflation. The US Federal Funds Rate rose from 0.25% in March 2022 to 4.75% by February 2023. This very steep rise in interest rates meant that the low-interest HTM securities held by SVB dropped in value.
SVB’s two vulnerabilities created the perfect storm as soon as depositors became aware of them. Tech companies that had more than $250,000 of deposits ran the bank to get their uninsured deposits out. Only one other large bank – Signature Bank – had this combination of a high proportion of uninsured deposits and deposits invested in loans and HTM securities (see Figure 4).
The runs on SVB and Signature meant that they quickly ran out of capital and failed.
Why were depositors bailed out?
The US Treasury, the Federal Reserve and the FDIC stepped in on 12 March to announce that all depositors in SVB and Signature Bank would be made whole – that is, they would receive all their uninsured deposits back.
One prominent beneficiary of this announcement was USD Coin, the world’s second largest stablecoin and fifth largest cryptocurrency. It had deposited $3.3 billion of its reserves with SVB. Before the bailout was confirmed, its one-to-one parity with the US dollar was broken, and one USD Coin was worth about 88 cents. But once it became apparent that uninsured depositors would be covered, the one-to-one parity was restored.
Why did the US government do this? First, because so many tech firms banked with SVB, it did not want the difficulties with the bank to spill over into the tech sector, with firms being forced into bankruptcy because they could not pay their bills.
Second, the fear of contagion meant that uninsured depositors of other banks would be tempted to run their banks. The joint statement by Treasury, the Fed and the FDIC was designed to reassure depositors of other banks that the US government would stand behind bank deposits.
In effect, the US government is revealing that there is no upper limit on deposit insurance. This creates a problem of ‘moral hazard’, whereby banks may take more risk with deposits than they should, in the full knowledge that the government will step in to bail out depositors.
Third, there may well be political economy reasons as to why the US government stepped in. Lobbyists and campaign contributions have for many years played a distortionary role in the US financial industry (Igan and Misfra, 2014; McCarty et al, 2015; Quinn and Turner, 2020). Campaign contributions and lobbying efforts by the tech and financial industries may have influenced the government to act in the way it did.
What is for sure is that lobbying by banks such as SVB played a significant role in the deregulation enacted by the US Congress in 2018. This rolled back the regulations introduced in the aftermath of the global financial crisis of 2007-09 – the Dodd-Frank Act of 2010.
This deregulation exempted banks with assets below $250 billion – such as SVB – from stress tests and tougher capital and liquidity requirements. In 2019, the Fed further reduced the regulatory burden for all but the largest banks. This allowed SVB to take the risks that it did with other people’s money.
The fourth explanation as to why SVB depositors were bailed out is that the monetary authorities would usually reduce interest rates and pump money into the banking system when it faces instability. But reducing interest rates and engaging in large-scale quantitative easing is not currently available given the Fed’s efforts to tame inflation. Nevertheless, the tensions between monetary policy and financial stability may well mean that central banks will put the brakes on rate rises for the time being in case other banks face the same fate as SVB.
Where can I find out more?
- Banks are designed to fail — and they do: Martin Wolf on the SVB failure.
- Silicon Valley Bank shows the perils of regulators fighting the last war: Gillian Tett on the need to learn from financial history.
- How Silicon Valley Bank collapsed in 36 hours: Wall Street Journal video on the SVB failure.
- SVB crisis 'speaks to the need for much stronger regulation' by authorities: MSBNC interview with Larry Summers.
- Baby busts and bank crashes: The Goodfellows (John Cochrane and Niall Ferguson) discuss the SVB failure.
Who are experts on this question?
- Jagjit Chadha
- John Turner
- John Wilson