These questions are important because banks receive copious government subsidies and bailouts (Ueda and de Mauro 2012). This support is necessary because combining lending and deposit-taking makes for a very risky way to organize a business.

The most conspicuous form of government support comes in the form of deposit insurance. For banks whose depositors are thus protected, the insurance serves as a form of credit enhancement that enables the banks to fund their loans and other assets at rates far lower than their nonbank competitors. Other subsidies take the form of access to the Federal Reserve’s discount window, direct access to the Federal Reserve’s payment system (Kwast and Passmore 1997), and access to cheap loans from Federal Home Loan Banks (Judge 2024). And, it seems, the bigger the bank, the bigger the subsidy, as banks considered “too big to fail” not only borrow at lower rates of interest, but they also take bigger risks (IMF 2014). Fortunately, bailouts themselves are uncommon.

The subsidies and bailouts routinely provided to banks prevent the critical process of “creative destruction” from working in the banking industry, weeding out obsolete ways of doing business to make way for newer ones. Banking could be an obsolete industry, but people really wouldn’t know of its obsolescence.

This article argues that banks are, indeed, obsolete. As markets and technology have evolved over the past several decades, close substitutes for banks have emerged that provide services that are likely equivalent or superior to the services banks provide. The subsidies and bailouts that are routinely and predictably provided to banks have prevented these alternatives from competing with banks on a level playing field.

Why Is Banking Inherently Unstable?

Good business practice requires that companies match the term (maturity) structure of their assets and liabilities, but banks do the opposite: They borrow short term and lend long term. Common sense dictates that a borrower seeking to finance a project that is not expected to generate revenue for 18 months should not promise to repay the loan immediately or at any time the lender wants its money back because it is likely there won’t be funds available to repay the loan when it is called in by the lender.

Another distinct feature of banks is that the liquidity of their assets and of their liabilities are highly asymmetrical: Bank assets are highly illiquid, while bank deposits are extremely liquid. Banks’ liabilities are predominantly in the form of deposits that are available to depositors on demand. In other words, deposits are the equivalent of cash. The Federal Reserve Bank of New York recently estimated that “deposits make up roughly 80 percent of bank liabilities.” Moreover, “the bulk of deposits are demand deposits, available for withdrawal without notice. This suggests that deposits are a short-maturity, floating-rate liability” (Luck et al. 2023). In contrast, on the asset side of banks’ balance sheets, in 2022 over 44 percent of the assets of US commercial banks were in the form of loans, according to the data website Statistica. Unlike deposits, loans are highly illiquid. There is no ready secondary market for most commercial loans, and they are considered financially opaque in the sense that it is difficult to estimate their value. Banks’ balance sheets often are dominated by assets that cannot be readily sold at a price close to their fair market value as estimated by a discounted cash flow analysis.

A third unusual feature of banks is that they are very thinly capitalized relative to other sorts of firms, where capitalization is defined as the amount of equity relative to debt. Specifically,

banks tend to have very little equity relative to other firms. Although it is not uncommon for typical manufacturing firms to finance themselves with more equity than debt, banks typically receive 90 percent or more of their funding from debt. (Macey and O’Hara 2003)

These fundamental characteristics of banks render them inherently unstable and essentially unworkable in the absence of government-sponsored deposit insurance:

Deposit insurance, created during the Great Depression in 1933, has sharply reduced the frequency of bank runs that once were common in the U.S. … [A]bout 40% of all U.S. banks disappeared between 1929 and 1933: They failed, closed, or were absorbed by other banks. That happened because there were massive runs, bank runs, where people lost confidence in the banks and pulled out their money…. The ones that were closed couldn’t make loans, obviously, and the ones that survived became extremely cautious, being very reluctant to make loans. (Gorton and Zhang 2023)

The primary source of bank instability is the collective action problem facing bank depositors. Depositors well know that, as a group, it is impossible for them to obtain repayment of their deposits simultaneously for the simple reason that if all of them ask for their funds at the same time, the bank will not have enough money on reserve to repay them. Too many requests for repayment by depositors likely will require the bank “to take value-reducing actions such as liquidating commercial loans at distress prices or calling loans prematurely” (Fischel et al. 1987). The collective action problem takes the form of a prisoner’s dilemma in which “everyone rushes in to withdraw their deposits before the bank gives out all of its assets” (Diamon and Dybvig 1983).

Because it is far cheaper for bank depositors simply to withdraw their funds from their bank rather than investigate the quality of the bank’s assets, rational uninsured depositors will quickly withdraw their money from banks on the slightest hint of a problem. For this reason, bank runs that cause the failure of an entirely solvent bank can easily occur.

The Economic Rationales for Combining Lending and Deposit-Taking

The argument that the strange structure of banks’ balance sheets survives only because of the existence of the regulatory safety net must confront the fact that banks made loans backed by demand deposits long before there was a government safety net. In fact, banks funded medium- and long-term loans with demand deposits long before deposit insurance.

The fact that banks survived without bailouts and subsidies prior to the New Deal would support the proposition that their combination of lending and deposit-taking is efficient. However, prior to regulation and bailouts, banking panics were regular events in which “even solvent banks could not [meet] the demands of depositors trying to withdraw funds at one time” (Gorton 1984). Banking panics, in other words, are a “nearly universal experience” in countries with banks, and the recurring incidence of banking panics “has led many governments to regulate the banking industry” (Gorton 1988).

Describing the pre–New Deal banking system as “efficient” means only that it was better than the next best alternative at the time. Bank failures were a regular occurrence in the United States before deposit insurance was introduced in the summer of 1933, with bank runs and panics occurring regularly (1819, 1837, 1873, 1907, and 1929). Thus, while it is true that banks combined deposit-taking and lending for eons before deposit insurance came on the scene, the banks placed financial systems in a constant state of peril. In other words, the traditional thinking is that banking is a necessary evil, and that deposit insurance is required to mitigate the social costs of having a system (Diamond and Dybvig 1986).

Checking account hypothesis / Theories about why banks are organized as they are “have centered on information and information flows” (Nakamura 1993). The “fundamental proposition, developed by Fisher Black and elaborated on by Eugene Fama, is that information is crucial to bankers both in determining risk premiums which set the price of credit, and in collecting loans, ensuring that lenders are repaid as much as possible within the terms of the contract.” The implication of the theory, of course, is that the lenders with the best information will perform better and thus outcompete rivals with inferior access to timely, high-quality information.

In a nutshell, then, the economic theory of banking as presented in the standard models is that “banks act on behalf of depositors to monitor borrowers, a role known as delegated monitoring.” The theory notes that banks have a cost advantage over non-banks because their borrowers are also depositors:

Banks have a cost advantage in making loans to depositors. The ongoing history of a borrower as a depositor provides information that allows a bank to identify the risks of loans to depositors and to monitor the loans at lower cost than other (non-bank) lenders. The inside information provided by the ongoing history of a bank deposit is especially valuable for making and monitoring the repeating short-term loans (rollovers) typically offered by banks. Information from an ongoing deposit history also has special value when the borrower is a small organization (or individual) that does not find it economical to generate the range of publicly available information needed to finance with outside debt or equity. (Fama 1985; see also Black 1975)

The observation that banks obtain information from their borrowers’ checking accounts that enhances their ability to make and monitor their commercial loans has been dubbed the “checking account hypothesis” (Gendreau et al. 1992). The core claim is that banks have a competitive advantage over nonbanks as commercial lenders because of their privileged access to the information that can be derived from observing the ebbs and flows of money in and out of customers’ checking accounts. As Macey and Miller 1992 wrote:

By examining the checks written against a commercial customer’s checking account and the deposits into the account, a bank loan officer can determine the size of the payroll, the salaries of the firm’s key personnel, the amount of money paid for supplies, the identity of the firm’s major customers, and the seasonal pattern of the business’s receipts. With this information readily available, banks obtained a significant competitive advantage over other sorts of financial intermediaries.

However, there is reason to doubt the continued validity of this hypothesis. Improvements in technology have made it possible for nonbank lenders to access the same high-quality information that traditionally has only been available through checking accounts (Macey and Miller 1995).

First, borrowers of all sizes, including sole proprietors, increasingly are using online and web-based systems to manage their finances and accounting, and to provide budgeting, payroll, invoicing, automatic payments, tax preparation, and other business needs (Akcigit et al. 2023). These systems provide information easily accessible by any lender that is more detailed and of higher quality than the information available from checking accounts. It includes not only account balances and records of cash flows in and out of bank accounts but also information about sales trends and seasonality of cash flows, profitability over time, and invoices and cash flow forecasts. In fact, it is now possible for individuals and small businesses to apply for loans with banks and other lenders directly from their accounts at accounting software companies like QuickBooks. Thus, lenders often no longer obtain information about prospective and current borrowers from the information gleaned from resident checking accounts, but from information gleaned from online accounting and financial applications.

Second, we now live in a world of “open banking” (sometimes called “open finance”) in which third-party developers access financial data in traditional banking systems through so-called “Application Programming Interfaces.” The data can then be freely shared with anyone that borrowers and consumers select to be a recipient of such information.

Open banking fundamentally undermines the traditional “closed system in which … personal and transactional information resides within legal, technological, and economic vaults to which only incumbents enjoy access” (Awry and Macey 2023). The “entrenched (competitive) advantage over both their own customers and any potential competitors” that banks’ exclusive access to customers’ financial data once gave them has disappeared. The essential information about borrowers, to which banks had access through their administration of borrowers’ deposit accounts, is no longer the exclusive domain of banks. This essential information is increasingly freely available to banks’ competitors.

Sticky deposits hypothesis / In addition to obtaining information about their loan customers from their window into their borrowers’ checking accounts, another important reason the capital structure of banks is viable is the so-called “deposit franchise” (Drechsler et al. 2023). Essentially, it holds that bank deposits are “sticky” in the sense that bank depositors are insensitive to changes in interest rates or monetary policy. The idea is that, while depositors have the legal right to access their funds on demand, in practice they do not do so. As a result, “even though deposits are short term, funding via a deposit franchise resembles funding with long-term fixed-rate debt.” The existence of a deposit franchise is crucial to the hypothesis that the simultaneous deposit-taking and commercial lending done by banks is a viable business model because it suggests that “deposit-taking and long-term lending have important synergies” and “should not be separated.”

Deposit stickiness is important for bank stability because, to the extent that deposits are sticky, rising interest rates and other macroeconomic events that depress the market value of a bank’s assets do not immediately and automatically lead to bank runs (Koont et al. 2023). The notion that deposits are sticky directly confronts the received wisdom that banks are inherently unstable because of the mismatch in the term structure of their assets and their liabilities. The idea is that, while bank assets have an average estimated duration of 3.7 years as compared to an average duration of only 0.3 years for their liabilities (Drechsler et al. 2023), this mismatch is more apparent than real because depositors do not withdraw their money in response to changes in the interest rate. As Drechsler et al. 2023 write, “Aggregate bank cash flows are insensitive to interest rate changes.”

The introduction of mobile banking has made deposits less sticky, and this “has reduced the franchise value of deposits and consequently (reduced) the stability of the banking sector” (Drechsler et al. 2023). Simply put, technology in the form of digital banking has drastically reduced the transaction costs of withdrawing money from a bank. Thus, bank runs are getting cheaper. As mobile banking becomes ever more prevalent, we should expect to see banks become increasingly fragile and susceptible to runs, thereby further undermining the case for banks. As Koont et al. 2023 point out, “digitalization,” or depositors’ ability to use technology to more easily access their funds and move them out of banks and into brokerage accounts and money market funds, “has important consequences for bank stability.” In particular, the presence of digital platforms and the ability of depositors to digitally link their bank accounts with their brokerage accounts “have made deposits more sensitive to increases in the deposit spread (the spread between interest rate paid on deposits and the interest rates available elsewhere) than before.”

Commitment lending / Regarding the existence of synergies between deposit-taking and making commitments to provide funding for borrowers in the future, Kashyap et al. 2002 observe that both providing loan commitments and deposit-taking require banks to hold large balances of liquid assets, or at least to have large balances of liquid assets available to fund loans and make good on deposit obligations when called upon by borrowers and depositors. Kashyap et al. point out that a major role played by lending institutions involves providing corporate clients with lines of credit and other sorts of future loan commitments that give a borrower the right, but not the obligation, to receive a loan on demand during a contractually specified period.

Kashyap et al. argue that there are efficiencies to combining lending and deposit-taking because there are synergies in offering both loan commitments and deposits “as long as deposit withdrawals and commitment takedowns are not too highly correlated.” The idea is that synergies arise to the extent that deposits flow into banks more-or-less contemporaneously to periods when commercial borrowers in need of liquidity emerge to take down their previously committed lines of credit. This idea is creative and interesting. Kashyap et al. provide an additional argument that “the institutional form of the commercial bank may be attributable to real considerations or economic efficiency, rather than simply to historical accident or the distortions inherent in policies such as deposit insurance.”

In important work following Kashyap et al., Gatev et al. 2009 provide a powerful explanation for why this happens. They show that banks continue to receive funding inflows when liquidity dries up, and this is not because combining lending and deposit-taking is efficient but because of regulatory subsidies. They write:

Why do banks enjoy funding inflows when liquidity dries up? First, the banking system has explicit (government) guarantees of its liabilities. Second, banks have access to emergency liquidity support from the central bank. Third, large banks such as Continental Illinois have been supported in the face of financial distress. Thus, funding inflows occur because banks are rationally viewed as a safe haven for funds. … Thus, government safety nets can explain why banks generally receive funding during crises.

This explanation is important because the Kashap et al. synergy story does not always hold up: Sometimes, both depositors and borrowers decide that they want access to their cash simultaneously. Kashyap et al. ignore the implications of this inconvenient fact; the possibility of a run on the bank does not exist in their analysis. But bank runs are a real problem because not all bank liabilities are insured.

Most importantly, Gatev et al. show banking is not efficient without a government safety net. This is because, in the real world, banks fail when deposit-taking and commitment takedowns become “too highly correlated.” And this happens whenever depositors want their money at the same time that borrowers want banks to make good on their loan commitments. In other words, the Kashap et al. model ultimately depends on the deposit franchise acting effectively to keep depositors from withdrawing their funds, and that depends on a credible government safety net that features deposit insurance, subsidized emergency lending to banks, and of course bailouts. To say that banks are efficient if and only if they operate in a world with deposit insurance and other government subsidies of banks is to say that they are not efficient at all.

The future is now / Already, non-depository institutions are neck-and-neck with banks in lending. In 2022, nonbanks accounted for over half (50.9 percent) of mortgage loans, and they accounted for 60.7 percent in 2021. Nonbank lenders similarly account for approximately half of the new credit extended to small businesses. At some points in recent history, as much as 70 percent of home mortgages were funded by nonbanks, and banks’ (including credit unions) share of the market was as low as 28 percent. Moreover, it has long been known that securitization of income-producing assets such as commercial and residential mortgage loan debt, credit card receivables, student loans, and automobile loans enables banks to move loans off their balance sheets and into the hands of nonbank investors.

Both housing finance and the financing of non-mortgage consumer debt and non-financial corporate debt are increasingly coming from nonbanks. Business lending by nonbanks involves every segment of the market, including syndicated corporate loans, loans to middle-market firms, and loans to small businesses (Covas and Fernandez Dionis 2022). Thus, “banks no longer play as important a financial-intermediation role as they once did” (Levitin 2016). Interestingly, this seems particularly true for Black- and Hispanic-owned firms and other non-traditional borrowers (Wiersch et al. 2022).

The success of nonbank lending is particularly impressive because nonbank lenders are competing with traditional bank lenders whose costs are lower because they receive cheap funding from depositors whose claims are backed by the government-sponsored deposit insurance program.

Conclusion

The traditional economic justification for combining deposit-taking and lending is that the information that banks glean from customers’ transaction accounts is secret and highly valuable because it enables bank lenders to monitor borrowers in ways that their nonbank rivals cannot. Open banking (also called open finance) is an important emerging banking practice that provides nonbank lenders and other bank competitors with open access to consumer and business banking, transaction, and other data. Open banking makes information that traditionally was in the exclusive possession and control of banks available to nonbank competitors, thereby undermining the core assumption that there are synergies in combining lending and deposit-taking that cannot be duplicated by nonbanks.

Those who argue that combining lending and deposit-taking is efficient ignore the significant costs and overstate the efficiencies associated with combining the two functions. Those who argue that traditional banking is efficient do not consider the fact that maintaining this system requires massive regulation to deal with the instability (deposit insurance distortions, recurring runs, panics, bailouts) imbedded in this industrial structure.

Banks are alleged to benefit from combining lending and deposit-taking either because of their superior access to information or because of their ability to attract cheap funding, either through sticky deposits or deposits that are available to fund banks’ promises to make credit available to borrowers in the future. But consumers do not appear to get much out of the deal, other than the opportunity to park their money in a very risky enterprise that offers near-zero returns and relies on a massive panoply of government support for survival.

Readings

• Akcigit, Ufuk, et al. 2023, “Intuit QuickBooks Small Business Index: A New Employment Series for the US, Canada, and the UK,” NBER Working Paper Series no. 31350, June.
• Awrey, Dan, and Joshua Macey, 2023, “The Promise and Perils of Open Finance,” Yale Journal on Regulation 40(1): 1–59.
• Black, Fisher, 1975, “Bank Funds Management in an Efficient Market,” Journal of Financial Economics 2(4): 323–339.
• Covas, Francisco, and Gonzalo Fernandez Dionis, 2022, “Assessing the Decline in Bank Lending to Businesses,” Bank Policy Institute, September 14.
• Diamond, Douglas W., and Philip H. Dybvig, 1983, “Bank Runs, Deposit Insurance, and Liquidity,” Journal of Political Economy 91(3): 401–419.
• Diamond, Douglas W., and Philip H. Dybvig, 1986, “Banking Theory, Deposit Insurance, and Bank Regulation,” Journal of Business 59(1): 55–68.
• Drechsler, Itamar, et al., 2023, “Banking on Deposits: Maturity Transformation Without Interest Rate Risk,” Journal of Finance 76(3): 1091–1143.
• Fama, Eugene, 1985, “What’s Different About Banks?” Journal of Monetary Economics 15(1): 29–39.
• Fischel, Daniel R., et al. 1987, “The Regulation of Banks and Bank Holding Companies,” Virginia Law Review 73(2): 301–338.
• Gatev, Evan, et al., 2009, “Managing Bank Liquidity Risk: How Deposit-Loan Synergies Vary with Market Conditions,” Review of Financial Studies 22(3): 995‑1020.
• Gendreau, Brian, et al., 1992, “Debating the Checking Account Hypothesis,” Journal of Retail Banking 14: 35–51.
• Gorton, Gary, 1984, “Private Clearinghouses and the Origins of Central Banking,” Federal Reserve Bank of Philadelphia, Business Review, January–February.
• Gorton, Gary, 1988, “Banking Panics and Business Cycles,” Oxford Economic Papers 40(4): 751–781.
• Gorton, Gary B., and Jeffrey Y. Zhang, 2023, “Taming Wildcat Stablecoins,” University of Chicago Law Review 90(3): 909–971.
• IMF (International Monetary Fund), 2014, “Global Financial Stability Report, Moving from Liquidity to Growth Driven Models,” April.
• Judge, Katherine, 2024, “The Unraveling of the Federal Home Loan Banks,” Yale Journal of Regulation 41(3): 1011–1077.
• Kashyap, Anil, et al., 2002, “Banks as Liquidity Providers: An Explanation for the Coexistence of Lending and Deposit-Taking,” Journal of Finance 57(1): 33–73.
• Koont, Naz, et al., 2023, “Destabilizing Digital Bank Walks,” Chicago Booth/​Stigler Center for the Study of the Economy and the State Working Paper Series no. 328, May.
• Kwast, Myron, and Wayne Passmore, 1997, “The Subsidy Provided by the Federal Safety Net: Theory, Measurement and Containment,” Board of Governors of the Federal System, December.
• Levitin, Adam J., 2016, “Safe Banking: Finance and Democracy,” University of Chicago Law Review 83(1): 357–455.
• Luck, Stephan, et al., 2023, “How Do Interest Rates (and Depositors) Impact Measures of Bank Value?” Liberty Street Economics (blog), Federal Reserve Bank of New York, April 7.
• Macey, Jonathan, and Geoffrey Miller, 1992, “Bank Failure: The Politization of a Social Problem,” Stanford Law Review 45(1): 289–309.
• Macey, Jonathan, and Geoffrey Miller, 1995, “Deposit Insurance, the Implicit Regulatory Contract, and the Mismatch in the Term Structure of Banks’ Assets and Liabilities,” Yale Journal of Regulation 12(1): 1–24.
• Macey, Jonathan, and Maureen O’Hara, 2003, “The Corporate Governance of Banks,” Federal Reserve Bank of New York, Economic Policy Review 9(1): 91–107.
• Nakamura, Larry I., 1993, “Commercial Bank Information: Implications for the Structure of Banking,” in Structural Change in Banking, Michael Klausner and Lawrence J. White (eds.), Irwin Professional Publishing.
• Ueda, Kenichi, and Beatrice Weder de Mauro, 2012, “Quantifying Subsidies for Systematically Important Financial Institutions,” International Monetary Fund Working Paper WP/12/128, May.
• Wiersch, Ann Marie, et al., 2022, “Clicking for Credit: Experiences of Online Lender Applicants from the Small Business Credit Survey,” Federal Reserve Bank of Cleveland, August.