ChatGTP-4o image of customers at a teller’s window in a nineteenth century bank.
The United States has many more commercial banks than do other high-income countries. At the end of 2024, there were about 4,000 commercial banks in the United States. In contrast, in 2024, there were only 35 commercial banks in Canada, 20 in South Korea, and fewer than 400 in Japan, France, Germany, Spain, Switzerland, Norway, Sweden, and the United Kingdom.
But even 4,000 commercial banks is many fewer than the peak of 29,417 commercial banks in the United States in 1921. The following figure shows the number of commercial banks in the United States from 1781, when the Bank of North America became the first bank to operate in the United States, through 2024. The grey line shows the total number of banks; the blue line shows the number of state banks—banks with a charter from a state government; and the orange line shows the number of national banks—banks with a charter from the federal government.
A key to the rapid expansion in the number of banks in the United States—there were already 1,600 in 1861—was legislation in most states that restricted banks to a single location. Without the ability to operate branches and with travel slow and expensive, most banks collected deposits and made loans only to businesses and firms in a small geographical area. Research by David Wheelock of the Federal Reserve Bank of St. Louis has shown that in 1900, of the 12,427 commercial banks in the United States, only 87 had any branches. In addition, a series of federal laws limited the ability of banks to operate in more than one state. The most recent of these laws was the McFadden Act, which Congress passed in 1927. With their loans concentrated in one geographic area—and often in a single industry—unit banks were inefficient because they were exposed to greater credit risk. If a bank is located in a town in down-state Illinois where most local businesses depend on agriculture, a drought could force many farmers to default on loans, causing the bank to suffer heavy losses and possibly fail.
The figure shows a surge in the number of banks during the World War I period. The total number of banks rose from 24,308 in 1913 to 29,417 in 1921—an increase of 21 percent. Research by David Wheelock and Mathew Jaremski of Utah State University discusses the reasons for the sharp increase in the number of banks through 1921 and the rapid decline in the number of banks in the following years. World War I led to problems in European agriculture, which drove up agricultural prices worldwide. U.S. farmers responded by expanding production and new banks opened to meet farmers’ increased demand for credit. Newly opened banks were particularly aggressive lenders. When European agriculture revived following the end of the war in 1918, agricultural prices declined sharply and many farmers defaulted on their loans. As unit banks, the assets of many newly opened banks were highly concentrated in loans to local farmers. When defaults on these loans sharply increased, many banks failed.
The number of banks declined slowly through the 1920s, reflecting in part the continuing problems of U.S. agriculture during those years. The Great Depression, which began in August 1929, led to a series of bank panics that reduced the number of banks from 25,125 in 1928 to 13,949 in 1933.
Over time, legal restrictions on the size and geographic scope of banking were gradually removed. After the mid-1970s, most states eliminated restrictions on branching within the state. In 1994, Congress passed the Riegle–Neal Interstate Banking and Branching Efficiency Act, which allowed for the phased removal of restrictions on interstate banking. The 1998 merger of NationsBank, based in North Carolina, and Bank of America, based in California, produced the first bank with branches on both coasts.
Rapid consolidation in the U.S. banking industry has resulted from these regulatory changes. While in 1984, there were 14,496 commercial banks in the United States, in 2024, as noted earlier, there were only about 4,000. This consolidation is what we would expect in an industry with substantial economies of scale when firms are free to compete with each other. When an industry has economies of scale, firms that expand have a lower average cost of producing goods or services. This lower cost allows the expanding firms to sell their goods or services at a lower price than smaller rivals, driving them out of business or forcing them to merge with other firms. Because large banks have lower costs than smaller banks, they can offer depositors higher interest rates, offer borrowers lower interest rates, and provide investment advice and other financial services at a lower price. (We discuss many aspects of the history and economics of banking in Chapter 10 of Money, Banking, and theFinancial System.)
Even though over the past 30 years there has been tremendous consolidation in the U.S. banking industry, 4,000 banks is still many more banks than in most other countries. So, it seems likely that further consolidation will take place, and the number of banks will continue to dwindle.
Runs at Silicon Valley Bank and others emerged quickly and drove steep losses in regional bank equity values. Regulators shouldn’t have been caught by surprise, but at least they should take lessons from the shock. The subsequent ad hoc fixes to deposit insurance and assurances that the banking system is sufficiently well-capitalized don’t yet suggest a serious policy focus on those lessons. Calls for much higher levels of bank capital and tighter financial regulation notwithstanding, deeper questions about bank regulation merit greater attention.
Runs are a feature of banking. Banks transform short-term, liquid (even demandable) deposits into longer-term, sometimes much less liquid assets. Bank capital offers a partial buffer against the risk of a run, though a large-scale dash for cash can topple almost any institution, as converting blocks of assets to cash quickly to satisfy deposit withdrawals is almost sure to bring losses. The likelihood of a run goes up with bad news or rumors about the bank and correlation among depositors’ on their demand for funds back. That’s what happened at Silicon Valley Bank. Think also George Bailey’s impassioned speech in the classic movie It’s a Wonderful Life,explaining how maturity and liquidity transformations can unravel, with costs to depositors, bankers, and credit reductions to local businesses and households. The bank examiner in the movie, eager to get home, didn’t see it coming.
While bank runs and banking crises can be hard to predict, a simple maxim can guide regulation and supervision: Increase scrutiny in areas and institutions in which significant changes are occurring over a short period. On an aggregate level, the sharp, rapid increase in the federal funds rate since March 2022 should have focused attention on asset values and interest rate risks. So, too, should the fast potential compression in values of office real estate in many locations as a consequence of pandemic-related working shifts and rate hikes. At the bank level, significant inflows of deposits—particularly uninsured deposits—merit closer risk review. This approach isn’t limited to banking. A recent report of the Brookings-Chicago Booth Task Force on Financial Stability, co-chaired by Donald Kohn and me, put forth a similar change-based approach to scrutiny of nonbank financial institutions.
Such an approach would have magnified supervisory attention to Silicon Valley Bank and First Republic Bank . It also suggests the desirability of greater scrutiny and stress testing of midsize banks generally facing interest rate and commercial real estate risk. Those stress tests can give the Fed and the Federal Deposit Insurance Corp. an indication of capital adequacy concerns that could give rise to mergers or bank closures.
Even with this enhanced regulatory and supervisory attention, two major questions remain: For bank liabilities, what role should deposit insurance play in forestalling costly runs? For bank assets, what role should banks play in commercial lending?
Actions taken since Silicon Valley Bank’s collapse have effectively increased deposit-insurance guarantees for troubled institutions. But the absence of a clearer policy framework for dealing with uninsured deposits dragged out the unwinding of First Republic Bank and threatens other institutions experiencing rapid deposit increases and interest rate risk. When regulators asserted in the wake of the runs that the status quo of a $250,000 limit remained unchanged, they lacked both credibility and a means to reduce uncertainty about future policy actions in a run. That makes runs at vulnerable institutions both more likely and more severe. One reform would be to increase deposit insurance limits for transaction accounts of households and small and midsize firms, as recently proposed by the FDIC. Of course, even this reform raises concerns about implementation, how to price the enhanced coverage, and whether supervision will shift toward the “focus on the changes” framework I outlined earlier.
Retaining a more modest role for deposit insurance raises a larger question: What role should banks play in business lending for working capital, investment, and commercial real estate? The FDIC is mandated to resolve bank failures at the least cost to the deposit-insurance fund, but following that path may lead to more mergers of vulnerable institutions into the nations’ largest banks. While consolidation may mitigate risks for depositors with greater diversification of deposits, it leaves open effects on the mix of lending. Knowing local borrowers better, small and midsize banks have a prominent local lending presence in commercial and industrial loans and real estate. Whether such projects would be financed in a similar mix by local branches of megabanks is a question. Congress should consider whether other alternatives might be reasonable. It might permit consolidation among smaller institutions, even if more costly in resolution in the near term to taxpayers. Or nonbank institutions could be permitted to play a role in resolving troubled banks. The latter mechanism should be considered, as nonbank asset managers like Blackstone or BlackRock could fund loans originated by local banks.
Two lessons for regulation loom large. The first is that attention should be paid to policy risk management as well as bank risk management in identifying areas of concern. Think easy money and the reach for yield, inflationary fiscal and monetary policy during the pandemic, and the Fed’s rapid-fire increase in short-term rates to combat stubborn inflation. Second, regulators and Congress need to be wary of both too much deposit insurance (with likely increased risk-taking and pressure on supervision) and too little deposit insurance (with likely jumps in banking concentration and disruption of local credit to businesses).
One can reasonably anticipate additional erosion of capital in non-money-center banks from rising interest rates and lower office real estate collateral values, hopefully motivating a quick grasp of these lessons. While banks don’t have to mark assets to market if current and can survive turbulence until monetary policy eases, potential runs can upset this equilibrium. Declining regional bank stock prices make this risk clear. Only good fortune or a more thoughtful policy stand in the way of additional bank distress and attendant credit supply reductions.
Rumors spread about the financial state of a bank. Some depositors begin to withdraw funds from their accounts. Suddenly a wave of withdrawals occurs and regulators step in and close the bank. A description of a run on a bank in New York City in the fall of 1930? No. This happened to Silicon Valley Bank, headquartered in Santa Clara, California and the sixteenth largest bank in the United States, on Friday, March 10, 2023.
Background on Bank Runs
In Macroeconomics, Chapter 14, Section 14.4 (Economics, Chapter 24, Section 24.4) we describe the basic reasons why a run on a bank may occur. We describe bank runs in greater detail in Money, Banking, and the Financial System, Chapter 12. We reproduce here a key paragraph on the underlying fragility of commercial banking from Chapter 12 of the money and banking text:
The basic activities of commercial banks are to accept short-term deposits, such as checking account deposits, and use the funds to make loans—including car loans, mortgages, and business loans—and to buy long-term securities, such as municipal bonds. In other words, banks borrow short term from depositors and lend, often long term, to households, firms, and governments. As a result, banks have a maturity mismatch because the maturity of their liabilities—primarily deposits—is much shorter than the maturity of their assets—primarily loans and securities. Banks are relatively illiquid because depositors can demand their money back at any time, while banks may have difficulty selling the loans in which they have invested depositors’ money. Banks, therefore, face liquidity risk because they can have difficulty meeting their depositors’ demands to withdraw their money. If more depositors ask to withdraw their money than a bank has money on hand, the bank has to borrow money, usually from other banks. If banks are unable to borrow to meet deposit withdrawals, then they have to sell assets to raise the funds. If a bank has made loans and bought securities that have declined in value, the bank may be insolvent, which means that the value of its assets is less than the value of its liabilities, so its net worth, or capital, is negative. An insolvent bank may be unable to meet its obligations to pay off its depositors.
The Founding of the Fed and the Establishment of the FDIC as a Response to Bank Runs
The instability of the banking system led to a number of financial crises during the 1800s and early 1900s, culminating in the Panic of 1907. Congress responded by passing the Federal Reserve Act in 1913, establishing the Federal Reserve System. The Fed was given the role of lender of last resort, making discount loans to banks that were experiencing deposit runs but that remained solvent. The failure of the Fed to stop the bank panics of the early 1930s led Congress to establish the Federal Deposit Insurance Corporation (FDIC) to ensure deposits in commercial banks, originally up to a limit of $2,500 per deposit, per bank. The current limit is $250,000.
Deposit insurance reduced the likelihood of runs but increased moral hazard in the banking system by eliminating the incentive insured depositors had to monitor the actions of bank managers. In principle, bank managers still have an incentive to avoid making risky loans and other investments for fear of withdrawals by households and firms with deposits that exceed the dollar deposit limit.
Contagion, Moral Hazard, and the Too-Big-to-Fail Policy
But if these depositors fail to monitor risk taking by bank managers or if a bank’s loans and investments decline in price even though they weren’t excessively risky at the time they were made, the FDIC and the Fed face a dilemma. Allowing banks to fail and large depositors to be only partially paid back may set off a process of contagion that results in runs spreading to other banks. Problems in the banking system can affect the wider economy by making it more difficult for households and firms that depend on bank loans to finance their spending. (We discuss the process of contagion in this post on the Diamond-Dybvig model.)
The Fed and the FDIC can stop the process of contagion if they are willing to ensure that large depositors don’t suffer losses. One mechanism to achieve this result is facilitating a merger between an insolvent bank and another bank that agrees to assume responsibility for meeting depositors withdrawals from the insolvent bank. But stopping contagion in this manner with no depositors suffering losses can be interpreted as amounting to deposit insurance having no dollar limit. The result is a further increase in moral hazard in the banking system. When the federal government does not allow large financial firms to fail for fear of damaging the financial system, it is said to be following a too-big-to-fail policy.
Silicon Valley Bank and VCs
Runs on commercial banks have been rare in recent decades, which is why the run on Silicon Valley Bank (SVB) took many people by surprise. As its name indicates, SVB is located in the heart of California’s Silicon Valley and the bank played an important role in the financing of many startups in the area. As such, SVB provided banking services to many venture capital (VC) firms. As we note in Chapter 9, Section 9.2 of the money and banking text, venture capital firms play an important role in providing funding to startup firms:
VCs such as Sequoia Capital, Accel, and Andreessen Horowitz raise funds from investors and invest in small startup firms, often in high-technology industries. In recent years, VCs have raised large amounts from institutional investors, such as pension funds and university endowments. A VC frequently takes a large ownership stake in a startup firm, often placing its own employees on the board of directors or even having them serve as managers. These steps can reduce principal–agent problems because the VC has a greater ability to closely monitor the managers of the firm it’s investing in. The firm’s managers are likely to be attentive to the wishes of a large investor because having a large investor sell its stake in the firm may make it difficult to raise funds from new investors. In addition, a VC avoids the free-rider problem when investing in a firm that is not publicly traded because other investors cannot copy the VC’s investment strategy.
An article on bloomberg.com summarized SVB’s role in Silicon Valley. SVB is
the single most critical financial institution for the nascent tech scene, serving half of all venture-backed companies in the US and 44% of the venture-backed technology and health-care companies that went public last year. And its offerings were vast — ranging from standard checking accounts, to VC investment, to loans, to currency risk management.
Note from this description that SVB acted as a VC—that is, it made investments in startup firms—as well as engaging in conventional commercial banking activities, such as making loans and accepting deposits. The CEO of one startup was quoted in an article in the Wall Street Journal as saying, “For startups, all roads lead to Silicon Valley Bank.” (The Wall Street Journal article describing the run on SVB can be found here. A subscription may be required.)
SVB’s Vulnerability to a Run
As with any commercial bank, the bulk of SVB’s liabilities were short-term deposits whereas the bulk of its assets were long-term loans and other investments. We’ve discussed above that this maturity mismatch means that SVB—like other commercial banks—was vulnerable to a run if depositors withdraw their funds. We’ve also seen that in practice bank runs are very rare in the United States. Why then did SVB experience a run? SVB was particularly vulnerable to a run for two related reasons:
1. Its deposits are more concentrated than is true of a typical bank. Many startups and VCs maintain large checking account balances with SVB. According to the Wall Street Journal, at the end of 2022, SVB had $157 billion in deposits, the bulk of which were in just 37,000 accounts. Startups often initially generate little or no revenue and rely on VC funding to meet their expenses. Most Silicon Valley VCs advised the startups they were invested in to establish checking accounts with SVB.
2. Accordingly, the bulk of the value of deposits at SVB was greater than the $250,000 FDIC insurance limit. Apparently 93 percent to 97 percent of deposits were above the deposit limit as opposed to about 50 percent for most commercial banks.
Economics writer Noah Smith notes that SVB required that startups it was lending to keep their deposits with SVB as a condition for receiving a loan. (Smith’s discussion of SVB can be found on his Substack blog here. A subscription may be required.)
The Reasons for the Run on SVB
When the Fed began increasing its target for the federal funds rate in March 2022 in response to a sharp increase in inflation, longer term interest rates, including interest rates on U.S. Treasury securities, also increased. For example the interest rate on the 10-year Treasury note increased from less than 2 percent in March 2022 to more than 4 percent in March 2023. The interest rate on the 2-year Treasury note increased even more, from 1.5 percent in March 2022 to around 5 percent in March 2023.
As we discuss in the appendix to Macroeconomics, Chapter 6 (Economics, Chapter 8) and in greater detail in Money, Banking, and the Financial System, Chapter 3, the price of a bond or other security equals the present value of the payments the owner of the security will receive. When market interest rates rise, as happened during 2022 and early 2023, the value of the payments received on existing securities—and therefore the prices of these securities—fall. Treasury securities are free from default risk, which is the risk that the Treasury won’t make the interest and principal payments on the security, but are subject to interest-rate risk, which is the risk that the price of security will decrease as market interest rates rise.
As interest rates rose, the value of bonds and other long-term assets that SVB owned fell. The price of an asset on the balance sheet of a firm is said to be marked to market if the price is adjusted to reflect fluctuations in the asset’s market price. However, banking law allows a bank to keep constant the prices of bonds on its balance sheets if it intends to hold the bonds until they mature, at which point the bank will receive a payment equal to the principal of the bond. But if a bank needs to sell bonds, perhaps to meet its liquidity needs as depositors make withdrawals, then the losses on the bonds have to be reflected on the bank’s balance sheet.
SVB’s problems began on Wednesday, March 8 when it surprised Wall Street analysts and the bank’s Silicon Valley clients by announcing that to raise funds it had sold $21 billion in securities at a loss of $1.8 billion. It also announced that it was selling stock to raise additional funds. (SVB’s announcement can be found here.) SVB’s CEO also announced that the bank would borrow an additional $15 billion. Although the CEO stated that the bank was solvent, as an article on fortune.com put it, “Investors didn’t buy it.” In addition to the news that SVB had suffered a loss on its bond sales and had to raise funds, some analysts raised the further concern that the downturn in the technology sector meant that some of the firms that SVB had made loans to might default on the loans.
Problems for SVB compounded the next day, Thursday, March 9, when Peter Theil, a co-founder of PayPal and Founders Fund, a leading VC, advised firms Founders Fund was invested in to withdraw their deposits from SVB. Other VCs began to pull their money from SVB and advised their firms to do the same and a classic bank run was on. Because commercial banks lack the funds to pay off a significant fraction of their depositors over a short period of time, in a run, depositors with funds above the $250,000 deposit insurance limit know that they need to withdraw their funds before other depositors do and the bank is forced to close. This fact makes it difficult for a bank to stop a run once it gets started.
According to an article in the Wall Street Journal, by the end of business on Thursday, depositors had attempted to withdraw $42 billion from SVB. The FDIC took control of SVB the next day, Friday, March 10, before the bank could open for business.
The Government Response to the Collapse of SVB
The FDIC generally handles bank failures in one of two ways: (1) It closes the bank and pays off depositors, or (2) it purchases and assumes control of the bank while finding another bank that is willing to purchase the failed bank. If the FDIC closes a bank, it pays off the insured depositors immediately, using the bank’s assets. If those funds are insufficient, the FDIC makes up the difference from its insurance reserves, which come from payments insured banks make to the FDIC. After the FDIC has compensated insured depositors, any remaining funds are paid to uninsured depositors.
As we write this on Sunday, March 12, leaders of the Fed, the FDIC, and the Treasury Department, were considering what steps to take to avoid a process of contagion that would cause the failure of SVB to lead to deposit withdrawals and potential failures of other banks—in other words, a bank panic like the one that crippled the U.S. economy in the early 1930s, worsening the severity of the Great Depression. These agencies hoped to find another bank that would purchase SVB and assume responsibility for meeting further deposit withdrawals.
Another possibility was that the FDIC would declare that closing SVB, selling the bank’s assets, and forcing depositors above the $250,000 deposit limit to suffer losses would pose a systemic risk to the financial system. In that circumstance, the FDIC could provide insurance to all depositors however large their deposits might be. As discussed earlier, this approach would increase moral hazard in the banking system because it would, in effect, waive the limit on deposit insurance. Although the waiver would apply directly only to this particular case, large depositors in other banks might conclude that if their bank failed, the FDIC would waive the deposit limit again. Under current law, the FDIC could only announce they were waiving the deposit limit if two-thirds of the FDIC’s Board of Directors, two-thirds of the Fed’s Board of Governors, and Treasury Secretary Janet Yellen agreed that failure of SVB would pose a systemic risk to the financial system.
According to an article on wsj.com posted at 4 pm on Sunday afternoon, bank regulators were conducting an auction for SVB in the hopes that a buyer could be found that would assume responsibility for the bank’s uninsured deposits. [Update evening of Monday March 13: The Sunday auction failed when no U.S. banks entered a bid. Late Monday, the FDIC was planning on holding another auction, with potentially better terms available for the acquiring bank.]
Update: At 6:15 pm Sunday, the Treasury, the Fed, and the FDIC issued a statement (you can read it here). As we noted might occur above, by invoking a situation of systemic risk, the FDIC was authorized to allow all depositors–including those with funds above the deposit limit of $250,000–to access their funds on Monday morning. Here is an excerpt from the statement:
“After receiving a recommendation from the boards of the FDIC and the Federal Reserve, and consulting with the President, Secretary Yellen approved actions enabling the FDIC to complete its resolution of Silicon Valley Bank, Santa Clara, California, in a manner that fully protects all depositors. Depositors will have access to all of their money starting Monday, March 13. No losses associated with the resolution of Silicon Valley Bank will be borne by the taxpayer.”
[Update Monday morning March 13] As we discussed above, one of the problems SVB faced was a decline in the prices of its bond holdings. As a result, when it sold bonds to help meet deposit outflows, it suffered a $2.1 billion loss. Most commercial banks have invested some of their deposits in Treasury bonds and so potentially face the same problem of having to suffer losses if they need to sell the bonds to meet deposit outflows.
To deal with this issue, Sunday night the Fed announced that it was establishing the Bank Term Funding Program (BTFP). Banks and other depository institutions, such as savings and loans and credit unions, can use the BTFP to borrow against their holdings of Treasury and mortgage-backed securities and agency debt. (Agency debt consists of bonds issued by any federal government agency other than the U.S. Treasury. Most agency debt is bonds issued by the Government Sponsored Agencies (GSEs) involved in the mortgage market: Federal National Mortgage Association (Fannie Mae), Government National Mortgage Association (Ginnie Mae), and the Federal Home Loan Mortgage (Freddie Mac).) The Fed explained its reasons for setting up the BTFP: “The BTFP will be an additional source of liquidity against high-quality securities, eliminating an institution’s need to quickly sell those securities in times of stress.” You can read the Fed’s statement here.
On Sunday, Signature Bank was closed by New York state banking officials and the FDIC. As with SVB, the Fed, FDIC, and Treasury announced that all depositors, including those whose deposits were above the $250,000 deposit limit, would be able to withdraw the full amount of their deposits.
Shareholders in SVB and Signature Bank lost their investments when the FDIC took control of the banks. On Monday morning, investors were selling shares of a number of regional banks who might also face runs, fearing that their investments would be lost if the FDIC were to seize these banks.
President Biden, speaking from the White House, attempted to reassure the public that the banking system was safe. He stated that he would ask Congress to explore changes in banking regulations to reduce the likelihood of future bank failures.
HubbardOBrienEconomics.com Blog 