Fundamentals: Key Insights on Banks, Banking Crises, and Policy Solutions

This policy brief presents one way of thinking about banks and banking crises in the United States. After using an accounting balance sheet to depict how individual banks work, I turn to history to shed light on the structure of the US banking system and show that the United States has historically had many banks and many banking crises. I close with a discussion of ways to make banks more resilient through simple, solvency-focused capital regulation, contingent liability, or both to lower the likelihood of crises before they might happen.

Understanding Risk Through a Balance-Sheet View of Banks

Banks are financial intermediaries that facilitate payments between their customers and counter- parties. Banks also use a variety of funding sources, including the deposits people make into their bank accounts, to acquire different assets, including the loans they originate and hold. 

The generic balance sheet in table 1 shows a way of thinking about banks that I will use to dem- onstrate the simple bank capital policy options discussed later. On the liabilities or funding side, a bank might use customer deposits—some FDIC insured, some uninsured—to acquire assets. These deposits can be prone to runs, in the sense that depositors, upon hearing negative news concerning their bank, may decide to withdraw simultaneously, which might cause the bank to fail. Banks can also use their capital to acquire assets. Capital includes convertible debt—debt that can be converted to equity if the bank experiences distress—and equity, which, unlike deposits, are not prone to runs. On the assets side, a bank might hold mostly illiquid loans as well as a variety of more liquid securities such as Treasuries, mortgage-backed securities (MBS) and municipal bonds, and reserves held at the Federal Reserve (Fed). 

This balance sheet reveals the two primary sources of risk in banking: asset risk and funding risk. Credit risk arising from risk of default and severity of losses can affect loans and other assets that banks hold. Interest-rate risk and liquidity risk can also affect loans, Treasuries, and other assets that banks hold. Looking to the funding side, deposits and other short-term debt funding can become scarce, leading to funding-liquidity risk. Increased leverage, which occurs when the funding mix includes more debt—especially deposits and other forms of short-term debt—than equity, also puts a bank at greater risk of default. In the United States, banks generally have the lowest estimated asset volatility across all industries, but banks are also the most leveraged, due to their extremely high fraction of debt.[1] 

Another Source of Risk: The Structure of the US Banking Industry

Beyond bank-specific asset risk and leverage, the US banking system’s stability through the Great Depression was also adversely affected by its fragmented nature. Figure 1 depicts the number of banks since 1834, showing the extent of fragmentation in the US banking system. It shows that the number of banks since 1834, showing the extent of fragmentation in the US banking system. It shows that the number of banks increased almost exponentially through 1921 and then declined. Figure 1 suggests that there has been a bias against large banks for most of US history. Prohibitions against interstate banking and branching in many states prevented the growth of large banks and led to the proliferation of many smaller banks as seen in the peak in 1921.[2]

The causes of fragmentation

The fragmentation of the US banking system and the rise in the number of banks through 1921 can be traced back to the US constitution, which prohibited states from issuing their own cur- rency and taxing interstate commerce.[3] In response to the prohibition against states issuing their own currency, some states instead generated revenues from bank chartering (with more charters being more lucrative) and from investing in banks. In response to the prohibition on taxing inter- state commerce, states instead prohibited out-of-state banks from operating in their jurisdiction. Competition with the Office of the Comptroller of the Currency (OCC), established in 1863 to increase state charters, also contributed to the rise through 1921.[4] The large decline in the number of banks after 1921 was the result of the large number of banking failures through the Great Depression years.

Fragmentation led to costly crises

The fragmented nature of the US banking system helps illuminate why the United States had eleven major banking panics and crises. The first nine were in 1819, 1833, 1837, 1839, 1857, 1873, 1893, 1907, and 1930–33.[5] After the Great Depression, major crises, including the Savings and Loan Crisis and the 2007–09 Crisis, have occurred less frequently but have tended to be prolonged and costly.[6] Major crises have considerable costs, whether measured in terms of (1) the administrative costs, including the fiscal cost of resolving failed banks or the bailout costs from assisting distressed banks during a crisis, or (2) the economic costs, measured as forgone real GDP.[7]

Figure 2 shows how these costs have evolved throughout US history. The top figure shows that the fiscal costs of crises have increased exponentially since the 19th century. In terms of cumulative economic costs, the bottom figure shows that since the Panic of 1893, major banking crises have happened less frequently but have been more severe. Among these major crises, the banking crisis of 1930–33 during the Great Depression remains the costliest in US history, as the cumulative losses amounted to 41.3 percent of one year’s real GDP.

Correctly assessing why banks have failed also has implications for the appropriate policy response, which I turn to next.

Policy Responses: Liquidity Views vs. Solvency Views

Policy debates concerning banking crises have for some time centered around the liquidity vs. the solvency views of banking crises.[8] The liquidity view focuses on liquidity as the source of the problem and calls for lender-of-last-resort programs that swap illiquid bank assets with safe as- sets or provide loans to shore up a bank’s funding. The solvency view focuses on solvency as the source of the problem and calls for banks to add more non-run-prone sources of funding, such as equity, during a quiet period, so that they can remain resilient during periods of distress.

These two views can be used to classify a variety of federal government policies. For instance, in terms of liquidity policies, Section 13(3) of the Federal Reserve Act prohibits the Fed from lending to insolvent institutions.[9] If the Fed lends to a bank, it must be because the Fed deems the bank illiquid but not insolvent. There are also liquidity regulations. One such regulation is the Basel III Liquidity Coverage Ratio, which attempts to ensure banks have sufficient liquid assets over the next 30 days. Another liquidity regulation is the Net Stable Funding ratio, which aims to ensure a bank has sufficient non-run-prone funding for the next year.[10]

The solvency view focuses on banks having sufficient capital—convertible debt and equity—and here the debates concern how much equity funding is sufficient to ensure that banks remain sol- vent. Debates over the best way to formulate bank solvency regulation have been ongoing, especially since the Basel Capital Adequacy standards were unveiled after the 1988 Basel Accords. It has been debated how to measure bank capital (including equity and convertible debt), how much capital banks should fund with (10 percent, 15 percent or more), and relative to what (equity relative to assets, equity relative to liabilities, or something else).[11]

Another Simpler Policy Option: Synthetic Contingent Liability

One last policy option to consider relates to figure 2 and why the fiscal costs of crises were so low prior to the creation of the Federal Reserve. Historical research shows that before the Federal De- posit Insurance Corporation (FDIC) was created, banks were subjected to double, triple, and even unlimited liability.[12] That meant bank owner-shareholders could not only lose their entire investment—as owners of single-liability common-equity shares can today—but they could be made to pay additional penalties to creditors. This gave them an incentive to close a weak bank early, before it failed, which in turn dramatically reduced the fiscal costs of banking crises. With the creation of the Federal Reserve came the Federal Reserve’s Discount Window, which gave weak banks reasons to stay open longer. This in turn explains the dramatic loss of banks between 1921 and 1934.

Fast forward to 2010. Section 956 of the Dodd–Frank Act calls for clawback regulation to avoid inappropriate risk-taking incentives in compensation and to introduce penalties for executives of failed banks. One way to do this could be to recreate contingent liability synthetically, in the sense of recreating the payoffs of contingent liability without having to change the liability of the underlying shares, and it could apply to options compensation as well.[13]

Concluding Thoughts

The focus on liquidity as a source of crises means policy tends to assume that banks during a crisis are not at risk of default, but instead just need ex post government assistance through a variety of costly government programs. If policy instead focused on making banks more resilient ex ante, through capital regulation or clawbacks by synthetic contingent liability, we might once again reduce the fiscal costs of banking crises, as well as their frequency and severity.

About the Author

Stephen Matteo Miller is a senior research fellow at the Mercatus Center at George Mason Uni- versity. His research focuses on financial market crashes and crises and financial regulation. 

Notes

[1] For a comparison of asset volatility and risk across industries, see table 1 in Stephen Matteo Miller, “How Much Would It Cost to Guarantee Debt for All Publicly Traded US Corporations?” Contemporary Economic Policy 42 (2024): 604–22.

[2] One estimate from 1990 suggested that if the US had a banking system like Canada’s but scaled for population, there might be as few as 75 banks in the US, while if it were more like California’s there might be 3,700, slightly below the current number (see David Mengle, “The Case for Interstate Branch Banking,” Federal Reserve Bank of Richmond Economic Review, November/December 1990, 3–17.

[3] For more background on state reactions to constitutional prohibitions see Richard Sylla, John Legler, and John J. Wallis, “Banks and State Public Finance in the New Republic: The United States, 1790–1860,” The Journal of Economic History 47 (1987): 391–403.

[4] For a discussion of the competition to create more bank charters, see James R. Barth and Stephen Matteo Miller, “Benefits and Costs of a Higher Bank ‘Leverage Ratio,’ ” Journal of Financial Stability 38 (2018): 37–52.

[5] I calculate the number of major and minor banking crises from several sources. For the Panic of 1819, see Hugh Rock- off, “Oh, How the Mighty Have Fallen: The Bank Failures and Near Failures That Started America’s Greatest Financial Panics,” The Journal of Economic History 81 (2021): 331–58. For a discussion of the 7 major crises and 20 minor crises between 1825 and 1929, see Andrew Jalil, “A New History of Banking Panics in the United States, 1825–1929: Construction and Implications,” American Economic Journal: Macroeconomics 7, no. 3 (July 2015): 295–330. For the Great Depression, see Milton Friedman and Anna Schwartz, A Monetary History of the United States (Princeton University Press, 1963), 1867–1960.