The explicit costs of government deposit insurance.

AuthorHogan, Thomas L.
PositionReport

The Diamond-Dybvig (DD) model is often cited as a theoretical justification for government deposit insurance. In the model, rational agents find it in their interest to withdraw their bank deposits if they suspect other depositors plan to do likewise. When a sufficient number of agents are expected to liquidate their accounts, a bank run ensues. Guaranteeing deposits through a system of government-administered deposit insurance removes the temptation to run on the bank and thereby precludes the need to ever use the deposit insurance. As Thomas Sargent makes clear, deposit insurance enters the model as a costless solution:

The good news in the Diamond-Dybvig model ... is that if you put in government-supplied deposit insurance, that knocks out the bad equilibrium. People don't initiate bank runs because they trust that their deposits are safely insured. And a great thing is that it ends up not costing the government anything to offer the deposit insurance! It's just good all the way around [Rolnick 2010: 31].

Diamond and Dybvig (1983: 44) conclude, "Government deposit insurance can improve on the best allocations that private markets provide."

In practice, however, government-provided deposit insurance is not a costless solution. It is frequently invoked to cover the losses of failed banks. In the United States, deposit insurance provided by the Federal Deposit Insurance Corporation (FDIC) is administered as a "rainy-day fund." Each member bank pays a premium based on its risk rating and on the amount of insurable deposits held by the bank. Premiums are held in a deposit insurance fund (DIF) administered by the FDIC. When a member bank becomes insolvent, debts to its depositors are paid out of the DIF and premiums for all banks increase until the fund is restored. Annual premiums reflect the cost of operating and administering the system and recent losses due to bank failures. The cost of guaranteeing deposits through such a system is decidedly nonzero.

The costs of providing insurance are sufficiently high as to warrant their inclusion in any reasonable model of deposit insurance. If the corresponding benefits of deposit insurance were sufficiently large and the alternative means of acquiring these benefits were either nonexistent or sufficiently costly, then ignoring the real-world costs of government-provided deposit insurance is perhaps appropriate. However, we do not believe this is the case. The benefits are not so large that one need not be concerned with costs, and there are potentially superior alternatives to government-provided deposit insurance. If one is to consider alternatives by engaging in comparative institutional analysis, a better understanding of the costs of government deposit insurance is required.

In what follows, we explore the explicit costs of government-provided deposit insurance. We focus on the FDIC as a specific example. First, we review the DD model and show how FDIC deposit insurance differs from the model in several key respects. Second, we discuss the history of the FDIC, paying particular attention to how the maximum amount covered, number of bank failures, and cost of managing the deposit insurance fund have changed over time. Third, we briefly discuss private deposit insurance and other risk-constraining mechanisms as alternatives to government-provided insurance. Finally, we offer some concluding remarks.

Theory of Deposit Insurance

The DD model demonstrates that although banks can reduce individual risk by acting as financial intermediaries, they create systemic risk in the potential for bank runs. Under certain conditions, financial contagion can cause all banks, even solvent ones, to be run upon simultaneously. Diamond and Dybvig (1983) propose that government insurance can costlessly limit the risk of bank runs by guaranteeing the values of customer deposits. Here we review the DD model to demonstrate how deposit insurance under FDIC differs from that posited by Diamond and Dybvig.

Banks in the DD model provide insurance against some form of uncertainty about the future. Suppose there is a group of agents with only one type of opportunity for production over three periods 0, 1, and 2. For each unit of capital invested in the production process in period 0, any agent can earn a return of R > 1 in period 2 or withdraw his original investment of 1 unit in period 1. Many analogies have been used for this scenario: planting corn that grows in the future but provides a meager yield if harvested early (Selgin 1993); a business project where investors' time horizon is uncertain (Diamond 2007); a real estate fund that may be relinquished early at a discount or held to maturity (Sebastian and Tyrell 2006). The asymmetry of future payments in each of these cases makes their payoff patterns suboptimal. Because the agents invested in these technologies tend to be (or at least are assumed to be) risk averse, they would prefer to accept a reduction of their high potential payment in period 2 in return for a small increase in their potential payment in period 1.

A bank can be created to reduce the cost of uncertainty by smoothing the potential future payoffs. Agents that invest in the bank receive a deposit contract that allows them to choose between future payoffs of return [r.sub.1] in period 1 or [r.sub.2] in period 2 where 1 1. If too many agents redeem their deposits in period 1, the bank will not have sufficient capital to fulfill its obligations, and the bank will go into default. Once it becomes known that the bank may default, all agents have an incentive to redeem their deposits immediately in period 1 since no capital will be left in period 2. This flood of simultaneous redemptions constitutes a bank run. The danger of bank runs is most poignant when consumer preferences are unknown. Because each agent fears that the others may withdraw early, bank runs become a self-fulfilling prophecy: any indication that there may be a bank run can itself cause a bank run. (1)

Diamond and Dybvig (1983) propose that a system of government-provided deposit insurance can mitigate the danger of bank runs. The authors assume that the government has an advantage over private banks because it can enact its desired policies ex post once the optimal allocation of resources is known. "In particular, it can tax those agents who withdrew 'early' in period T = 1." By contrast, "a private insurance company is constrained by its reserves in the scale of unconditional guarantees which it can offer" (Diamond and Dybvig 1983: 413). They conclude that "this asymmetry 'allows a potential benefit from government intervention" (p. 414). Once it is known that the government will redistribute any undeserved gains from bank runs, agents no longer have an incentive to run on the bank. Hence, the government's commitment to providing deposit insurance precludes the possibility of a bank run and guarantees that deposit insurance payouts will never be necessary. In this way, government deposit insurance becomes a costless solution to the problem of bank runs.

Other works extend the DD model to examine the optimality of deposit insurance under a variety of assumptions. Dowd (1988), Wallace (1990), Selgin (1993), and Green and Lin (2000) propose alternative measures, such as proper capital allocation and suspending deposit redemptions to improve upon government deposit insurance. Peck and Shell (2003) show even those optimal contracts may be subject to runs. Others study the effects of signaling and information on the potential for runs (e.g., Samartin 2003; Andolato, Nosal, and Wallace 2007). However, each employs a model applicable only under a strict set of assumptions, with little consensus as to which is the most useful representation of deposit insurance in practice.

Diamond and Dybvig (1983) acknowledge the implementation of deposit insurance is likely to be suboptimal. They note that their model produces "a very strong result (which may be too strong) about the optimality of government deposit insurance" (p. 414). The costs of actual deposit insurance deviate from the DD model in several ways. Taxes assessed on banks to fund deposit insurance will cause the provision of insurance to be suboptimal because there are real costs to assessing and collecting taxes. Diamond and Dybvig (1983: 415) state that "if a nonoptimal tax must be imposed, then when t is stochastic there will be some tax distortions and resource costs associated with government deposit insurance." (2) Furthermore, to the extent that failures resulting from bank runs are indistinguishable from other types of failure, government-provided deposit insurance is more likely to be under- or oversupplied.

Government deposit insurance departs markedly from that proffered in Diamond and Dybvig (1983). To further illustrate the differences between deposit insurance in theory and practice, and to more clearly understand the actual costs, we examine deposit insurance offered under the FDIC. (3)

FDIC Deposit Insurance

The FDIC was established by the Banking Act of 1933 primarily in response to widespread bank failures in that year. Prior to the FDIC, deposit insurance was provided at the state level. However, rural bank failures during the economic downturn of 1921 and the decade of crop failures that followed proved too much for these funds to handle, and all had ceased operations by 1930. Despite initially opposing federal deposit insurance, President Franklin D. Roosevelt signed the bill into law on June 16, 1933. By January of the following year the program was up and running. (4)

The history of the FDIC has been presented in much greater detail elsewhere. (5) Since we are ultimately concerned with the cost of deposit insurance in practice, our aim is limited to expositing the size, scope, and function of the FDIC over time. Specifically, we consider the maximum amount covered, number of bank failures, and the cost of managing the deposit...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT