Insurance firms often create brand images touting their financial reliability. This strategy has clear appeal. After all, consumers make insurance claims in the future and depend on insurers to honor their earlier commitments. Reservesthe financial assets in which insurers invest their premium incomeare often the most tangible sign of an insurance firm's staying power. Thus, it was not surprising that policymakers made reserves a centerpiece of the federal deposit insurance program.* Unfortunately, the laws governing reserveswhich link reserves to the setting of insurance premiumsactually prevent the Federal Deposit Insurance Corp. (FDIC) from managing deposit insurance in a sound manner. In particular, the rules governing assessments, in conjunction with current high levels of reserves, have led the FDIC to not charge an insurance premium to 95 percent of insured banks. An insurance firm following this pricing strategy tempts bankruptcy and, likewise, the no-assessment deposit insurance regime encourages excessive risk taking by banks.
But the problem runs deeper than the current link between premiums and reserves. Legislators, regardless of the policy at the time, look to reserves when assessing the current fiscal condition of deposit insurance and its ability to meet future financial objectives. Yet reserves are inherently backward looking as they reflect past premium payments and insurance claims. Thus, reserves are a particularly poor means of gauging the current or future financial performance of deposit insurance, particularly relative to an examination of the premiums that the FDIC charges. This misallocation of attention is all the more costly because the FDIC's ability to make good on the deposit insurance contract derives from its status as an agent of the federal government, not from the financial assets it has accumulated. Nor would reserves be necessary to encourage prompt payoff of depositors at insolvent banks if the federal insurer had expanded borrowing authority. In total, there is a very strong rationale for eliminating the link between premium setting and reserves, and good reason, more generally, to eliminate the holding of reserves.
While it is clear the current system has serious flaws, reforms to the FDIC's funding system would also raise a number of policy challenges. An illustrative example of an insurance system without reserves follows the analysis of the current system. It is presented to further discussion on how to correct existing failures.
Reserves and the FDIC's targeted policy
Private insurance firms offer contracts obligating them to make payments to the insured if a prespecified event occurs, such as a death, theft or failure to make payment on a bond or a mortgage. The insurance firm receives premiums in return for these commitments. The firm then invests the premium income in financial assets, called reserves, that it will liquidate in the future in order to honor its contracts. The ability of the private firm to make good on its promise depends on the adequacy of its reserves and whatever original cash infusion it received, among other factors.
The FDIC was established in 1933, received its inflow of initial funding, and was organized, according to the FDIC's official history, like "a typical casualty insurance company."1 Like a private firm, the FDIC began to charge premiums and accumulate reserves in its accounts. From the onset, government officials and analysts looked to the reserves to ascertain the FDIC's ability to meet its commitments. Congress most recently formalized the role of reserves in 1991, requiring the FDIC to maintain the ratio of fund reserves in the Bank Insurance Fund (BIF) to insured deposits at 1.25 percent. The BIF reached this target in the summer of 1995 and has exceeded the target since then. Now, the FDIC collects insurance premiums from only 5 percent of insured banks because of its interpretation of the legislative mandate to "maintain" the BIF at its target reserve ratio as well as an act of Congress that eliminated a minimum insurance assessment. ["A brief history of reserves and premiums" provides more detail on the path taken to current policy.] In other words, current premium setting is based largely on the level of reserves.
Should premium setting be linked to reserves?
Well-run insurance programs try to set actuarially fair premiums. Roughly speaking this means that the insurer should try to charge assessments so that the premiums paid by the insured equal the cost of insuring them. More precisely, the discounted value of the premium (discounting adjusts for the fact that a dollar today is worth more than a dollar delivered in the future) should generate revenue for the insurer that equals the expected discounted costs to the insurer from claims made by the insured over the life of the insurance contract.2 The insurer should also incorporate the uncertainty associated with predicting expected losses when determining how much to charge.
The policy of setting premiums based on the size of current reserves prevents, almost by definition, the setting of actuarially fair premiums. Reserves are backward looking, informing policymakers about past premiums and past payments to the insured. In contrast, the setting of fair premiums is an exercise in forecasting future losses that may not reflect what happened last year or the year before that. In this case, the FDIC must currently give away insurance to the vast majority of banks because reserves exceed the targeted level. As a result, the FDIC essentially charges banks one flat rate for deposit insurance just as it did over the vast majority of its history, even though Congress ostensibly required premiums to vary by the risk of a bank failing in 1991.
What is wrong with this policy? Unlike a flat assessment, a correctly priced premium discourages insured banks from taking on excessive risk. Two FDIC economists have described the problem clearly:
With deposit insurance, insured depositors hold an asset whose value is independent of the solvency position of the bank, and so no credit-risk premium is required by these depositors. Moreover, under a flat-rate premium structure, banks' insurance costs will be the same regardless of their risk position. As a result, banks may take on additional risk without having to pay higher interest rates on deposits or higher insurance premiums. The risk/return trade-off has been altered such that the price of assuming greater risk has been reduced and, consequently, the bank has an incentive to move to a riskier position.3
The absence of a fair premium thus creates a dynamic where insured banks take on excessive risk and the insurer could face higher losses than they might expect otherwise. Poor pricing also encourages firms paying too little to effectively seek out additional insurance by increasing their funding through insured deposits, for example. Those banks paying too much could try to limit their insurance coverage and premiums by funding themselves through nondeposit means. The end result of this adverse selection process is a higher-risk insurance pool.
Charging fair premiums shifts the costs of risk taking from the FDIC to the insured bank and reduces the distortions caused by improper pricing. As a result, risk-based pricing, a central tenet of private insurance, allows insurers to pay for their losses and manage the risk-taking proclivities of the insured. Because of these characteristics, this Federal Reserve bank has made the establishment of risk-based premiums a central tenet of its plan for increasing the market discipline on insured banks [See the Minneapolis Fed's 1997 Annual Report, published in The Region, March 1998.] And, despite the fact it linked reserves to assessment setting, Congress also has demonstrated an understanding of the flaws of a flat-rate premium. Congress noted in 1991 that, "The current system of flat-rate deposit insurance premiums fails to provide any disincentives for insured depository institutions to engage in unsound and risky activities, and actually penalizes conservatively managed institutions."4
Of course, one should not underestimate the difficulty in setting fair deposit insurance premiums. But, attempting this task became unavoidable after the banking crises of the 1980s and 1990s, when policymakers expressed a preference for sound management of deposit insurance over virtually uncontrolled losses.
It seems clear that decoupling reserves and premium setting has benefits. However, an examination of the reserve setting raises a broader question. Why does the FDIC hold reserves?
Should the FDIC hold reserves?
Regulators and policymakers have suggested at least three reasons why the deposit insurance system should hold reserves. First, like a private firm, the FDIC needs reserves to honor its insurance contract. Second, holding reserves reduces the likelihood of forbearance (the regulatory practice of not closing insolvent financial institutions). Third, reserves provide a useful measure of bank financing, the degree to which the premiums paid by banks will repay future claims made by banks.
These rationales do not hold up under scrutiny. In fact, the FDIC's repayment capacity flows from its federal nature, not from its reserves. Moreover, the FDIC need not hold reserves to limit forbearance if granted enhanced borrowing authority. Finally, the focus on reserves obscures and even prevents the sound fiscal operation of deposit insurance, particularly in the area of pricing.
Claims paying ability
Recent mega-mergers raised a question for policymakers. Could the FDIC make good on its contracts after the failure of one or many of the newly created banking behemoths? In addressing that question, the acting chairman of the FDIC, Andrew Hove, noted in April 1998 that
... Our staff has analyzed the sufficiency of the deposit insurance funds and has concluded that the resources available to the federal deposit insurance system are sufficient to withstand several years of failures at the highest historic levels experienced by banks. The resources of the funds also appear sufficient to handle the failure of a large insured institution. Unprecedented failures of a number of very large financial institutions simultaneously would be more problematic, but it is questionable whether it would be appropriate to maintain insurance funds that are large enough to address a worst-case scenario.5
Perhaps unintentionally, this type of question and answer suggests that the FDIC's reserves are fundamental to the insurer's capability to pay claims (an alternative interpretation of the acting chairman's statements is discussed below). However, the ultimate repayment capacity of the FDIC comes from the financial strength of the U.S. government. It had previously been articulated and understood that the full faith and credit of the United States backs insured deposits, even though this policy only became law in 1987.6 Indeed, former FDIC Chairwoman Ricki Helfer described the creation of federal deposit insurance as " ... placing the full faith and credit of the federal government behind bank deposits..."7
This view reflects the fact that, since its inception, deposit insurance has been part and parcel of the federal government, unlike any truly private firm. Thus, it is hard to imagine the federal government not financing the deposit insurance commitment if needed. Depositors certainly hold this view. They did not run troubled banks even after July 1991 when the FDIC reported that the liabilities of the insurance fund exceeded its assets by $7 billion.8
The BIF's balance sheet provides final confirmation of the almost fictional nature of reserves.** In 1997, 94 percent of the assets of the BIF were IOUs issued by the Treasury to the FDIC.9 In other words, the assets that this government agency holds to make good on its full faith and credit backing for insured deposits are more federal assurances. This is the equivalent of an insurance firm holding the bonds it issued as reserves. But, there is, of course, an important difference between the private firm and the FDIC. The taxing power of the federal government ultimately allows it to make good on the IOUs the FDIC holds and the insurance contract the FDIC issues.
Professor Frederic Mishkin of Columbia University, and formerly director of research at the Federal Reserve Bank of New York, has correctly argued that reserves could provide benefits even if they do not play the same role for the FDIC as they do for a private firm. However, there are other ways to achieve the same benefits that Mishkin highlighted.
More specifically, Mishkin noted that "... a lack of resources in the insurance fund makes it very likely that the deposit insurance agency will engage in regulatory forbearance and not close down insolvent institutions. ... The rapid recapitalization of the Bank Insurance Fund of the FDIC should be seen as a major success of [the Federal Deposit Insurance Corporation Improvement Act of 1991]..."10 As the quote indicates, even with the government standing behind a deposit insurance pledge, Congress has been unwilling at times to provide regulators with the cash they need to close failing insured institutions. The insurer for savings and loans left failing thrifts open longer than prudent when starved of funds during the 1980s. The insurer and regulators then allowed so-called zombie institutions to roll the dice with insured deposits and create greater losses. Alternatively, regulators created closure methods that did not require the insurer to contribute cash. Many of these techniques relied on unsound accounting and resolution practices that raised the cost of closing down the failing savings and loans.
Ultimately, the insurer honored its contracts and depositors of failed institutions received funds during the 1980s. But, it had become clear that reliance on congressional appropriations to finance prompt closure could increase the cost of failures. Policymakers have an incentive to engage in such long-run, harmful behavior because, in the short run, they do not want blame for a banking crisis and a concomitant increase in federal spending.
Yet, because of its unique federal status, the FDIC has methods at its disposal other than reserves or appropriations to fund bank resolutions. In particular, the FDIC can borrow funds, including up to $30 billion under a line of credit from the Treasury (in response to the banking crisis of the 1980s, this line of credit was increased from $5 billion in 1991). Before the Treasury releases funds under the line of credit, the FDIC and Secretary of the Treasury must agree to a repayment schedule funded through assessments, and consult with Congress. The FDIC can also borrow so-called working capitalmoney that will be repaid as the assets of failed banks are soldequal to 90 percent of the value of failed bank assets.11 The FDIC can raise working capital from a lending arm of the Treasury, called the Federal Financing Bank, or from insured banks themselves.
Congress could convert the FDIC's existing rights into permanent and indefinite borrowing authority from the Treasury to completely eliminate the potential that the FDIC runs out of cash to close failing banks, as could occur under current arrangements. This one-time legislative action by policymakers would give the FDIC an unlimited and easy-to-access ability to borrow funds to resolve failing firms. Assumedly, Congress would continue to require that the FDIC repay such borrowing through premium income. Congress should also set up review processes to ensure that the FDIC's borrowing is consistent with the public interest. Drawing on the Treasury could not only address forbearance but could, as part of a reformed system discussed in "An illustrative example of a federal deposit insurance system without reserves", help policymakers determine if deposit insurance meets its financial goals. Unfortunately, policymakers' current focus on reserves distracts them from effectively evaluating the financial management of the insurance program.
Measuring bank financing
Policymakers have made clear that they want insured banks, and not taxpayers, to finance deposit insurance. In practice, this policy means that the premiums paid by banks should equal the costs that the deposit insurance system incurs. (Despite the intended split between banks and taxpayers, bank consumers could bear some of the cost of the premiums, through higher fees or lower interest rates on deposits, even though banks write the premium check.)
Policymakers and regulators use the amount of reserves in the BIF to judge if banks have paid an adequate amount to cover future insurance payments. They view the insurance system as falling out of balance if it appears that reserves cannot cover losses. It is in this context that the acting chairman of the FDIC may have linked reserves to the FDIC's ability to pay claims.
Certainly, the amount of reserves could provide some information to policymakers and analysts on the likelihood that bank payments will fund insurance losses. But, as the banking crises made clear, reserves represent a second-best measure of self-financing. From the late 1970s until 1987, the reserve ratio stayed between 1.1 percent and 1.24 percent, within the targeted range at the time. However, during that same period banks took the risks that nearly depleted the fund three years later. Reserves represent what has happened in the past and thus cannot help but doing a poor job in forecasting the potential draw on taxpayer support.
As noted earlier, Congress needs to focus on premium setting to determine the ability of banks to pay for future losses. And the current links between reserves and premiums are an immediate impediment to setting premiums fairly. Congress, though, based the legislation restricting premiums on the long-held view that reserves, not premiums, accurately gauge the financial management of federal deposit insurance. Low reserves inherently tempt policymakers to judge premiums as being underpriced, even though the insurer could have correctly set assessments. Low reserves demonstrate the effect of past losses but do not inform about future needs. High reserves encourage policymakers to believe that the federal insurer has set premiums too high.12 Indeed, the insured may even convince policymakers to rebate the "excess premiums" even though the premium charged represents the fair value for the insurance. Thus, even if policymakers removed the current link between targeted reserves and premium setting and gave the FDIC more discretion, the mere existence of reserves may continue to dominate policymakers' approach in managing deposit insurance.
Better off without reserves
Policymakers have long held implicit and explicit expectations that the deposit insurance fund hold reserves. Because of their link to the premium setting process, reserves have actually become a major impediment to setting premiums based on risk, a cornerstone of prudent insurance operations for both private firms and the government. As a result, banks have an incentive to take on excessive risk and the insurer is more likely to face higher losses. More generally, the federal government need not hold reserves to ensure payment on insurance claims. In this regard, the government is profoundly different from a private firm. Other methods besides reserves also exist for limiting forbearance, another possible justification for holding reserves. In total, it appears that the deposit insurance system would be better off without reserves.
* References in this paper to banks apply to commercial banks and references to deposit insurance apply to the insurance for commercial bank deposits unless otherwise noted. Many of the concerns expressed in this paper, however, also exist for federal insurance for the deposits of savings and loans.
** Noted banking consultant Carter Golembe also described the deposit insurance fund as a fiction when recently interviewed in the June 1998 Region, although his concerns are different than those identified in this article.
1 Federal Deposit Insurance Corp. 1984. The First Fifty Years. Washington, D.C.: FDIC, p. 55.
2 This definition is taken from Fissel, Gary S. 1994. Risk Measurement, Actuarially-Fair Deposit Insurance Premiums and the FDIC's Risk-Related Premium System. FDIC Banking Review 7 (Spring/Summer), p. 18.
3 Blair, Christine E. and Fissel, Gary S. 1991. A Framework for Analyzing Deposit Insurance Pricing. FDIC Banking Review (Fall), p. 26.
4 House Report 102-330 (November 19, 1991). U.S. Code Congressional and Administrative News, vol. 3, 102nd Congress, First Session, 1991, p. 1922.
5 Testimony of Andrew C. Hove. Jr., Acting Chairman, Federal Deposit Insurance Corporation before the Committee on Banking and Financial Services, United States House of Representatives, April 29, 1998, p. 3.
6 Federal Deposit Insurance Corporation. 1997. History of the Eighties: Lessons for the Future. Volume 1, An Examination of the Banking Crises of the 1980s and Early 1990s. Washington D.C.: FDIC, p. 99.
7 Remarks of Ricki Helfer, Chairman, Federal Deposit Insurance Corporation, before the Bank Administration Institute, December 11th, 1996, p. 6.
8 Federal Deposit Insurance Corporation. 1991. FDIC Annual Report. Washington D.C.: FDIC, pp. 43 and 67.
9 In 1997, about $27 billion of the BIF's roughly $29 billion in assets were Treasury securities. Budget of the United States Government, Fiscal Year 1997Appendix, p. 1077.
10 Mishkin, Frederic S. 1997. Evaluating FDICIA. Research in Financial Services: Private and Public Policy 9, p. 22.
11 House Report 102-330 (November 19, 1991). U.S. Code Congressional and Administrative News, vol. 3, 102nd Congress, First Session, 1991, pp. 1908, 1927-1928.
12 Testimony of Lawrence H. Summers before the Committee on Banking and Financial Services Committee, United States House of Representatives, April 23rd, 1998, p. 6.