Federal Deposit Insurance Corporation
American Bankers Association
September 18, 2000
A lot has changed in the 125 years since the ABA was founded, and deposit insurance - an anchor of stability in an uncertain world - was one of the fundamental changes. In many ways, our system of deposit insurance itself has changed over time. And, as you know, the FDIC is undertaking a comprehensive review of our current system to see where we can make it even better. I am here to discuss that effort generally, and, in particular, I'd like to examine two questions that the ABA has raised more than once: How large should the insurance fund be? And the related question: Are rebates possible?
There are two ways to look at how you pay for deposit insurance. One way is what we can call the "user-fee" arrangement. Under this view of the deposit insurance system, the government bears all the risk for the costs of bank failures. Premiums are regarded as "user fees" that banks pay to compensate the government for its risk-bearing.
There can be no consideration of rebates, because premiums are a payment -- a fee -- for a valuable service rendered - our guarantee.
A user-fee system has advantages and disadvantages. For example, on the plus side, premiums for the industry can be relatively steady over time. Why? The government absorbs fluctuations in insurance losses. If losses in a given year exceed the revenue collected, the government is responsible for the difference. This is similar to the FDIC system that was in place from 1934 to 1950, when the average premium was set at 8.33 basis points and the fund absorbed losses exceeding this amount. Because we cannot look that far into the future, we can never know for certain whether the rate selected is appropriate -- we cannot see into the long run. Therefore, because the rate is stable under a user-fee system, the insurance funds are likely to build up substantially over long periods or shrink substantially during prolonged periods of losses.
And that brings us to the minus side: Under a user-fee system, there will always be some concern about mispricing. This was essentially the situation for the FDIC system in 1950. The fund had built up quite substantially -- to $1 billion! -- leading to concerns that banks were being overcharged. These concerns were addressed by tying pricing more closely to current loss experience. From 1950 through 1989, the FDIC continued to charge the statutory rate of 8.33 basis points, but refunded up to 60 percent of excess assessment revenue in any given year. This refund was from current-year premiums only - not past premiums or interest earnings on the fund. Even under the refund system, premiums were still relatively stable.
But in the late 1980s concerns arose about undercharging for insurance, and Congress linked premiums directly to the fund ratio as well as risk, and, in so doing, Congress provided the framework for the system we have today.
This system has its virtues. By giving the FDIC a call on the industry's capital -- as necessary to maintain the target -- it provides explicit protection for taxpayers. But this system also creates the problems we face today: It moves us away from stable premiums, requiring you to pay the most when banks can least afford it.
In good times, banks pay zero, distorting incentives and allowing deposits to enter the system without any contributions to the insurance fund. In short, by looking at the origins and history of FDIC insurance, some might characterize our system as a user-fee arrangement - one in which banks have no claims on past premiums. However, our current system provides few of the benefits of that approach. By tying premiums to a hard target and giving the FDIC a substantial call on industry capital, we have introduced the potential for significant premium volatility.
We can either step back and make some repairs to this system or choose a different approach. And that is where the second view of funding arrangements for deposit insurance comes in.
Under this alternative view, the industry is seen to bear the costs of bank failures "mutually." That is to say, the resources of the industry are available to cover the costs of insuring deposits. Some resources are held in the insurance fund; others are available through a call on industry capital.
In this way of looking at deposit insurance, past premiums are viewed essentially as industry capital the government holds in trust. A key question here is: How much should be held in the insurance fund? The answer: Any target for the fund should have a sound analytical basis; for example, measuring risk exposure relative to the size of the fund. This approach is fundamentally different from the current system, where the DRR (Designated Reserve Ratio) remains a fixed percentage unless the FDIC foresees large losses in the near term.
We as regulators and insurers would criticize any banking organization that looked at capital adequacy in this manner. The FDIC should not be forced to use such an approach for the deposit insurance funds. Unlike the user fee system, a mutual model of deposit insurance opens up the possibility for rebates. Banks would have a claim on past premiums, but a more rigorous approach to fund adequacy than what we now have would be absolutely necessary.
As you are aware, the FDIC has asked the risk-management consulting firm of Oliver, Wyman & Company to assist us in analyzing the adequacy of the BIF using the same methods they would employ for any large, complex financial institution. This work will serve to enhance further discussions regarding a sound analytical approach to fund adequacy.
We believe that rebates should be conditioned upon the best available analysis showing that the fund is more than adequate relative to the risk the FDIC is asked to bear. This is a high hurdle and requires a strong faith in the results of quantitative risk analysis.
But rebates could be appropriate in a mutual system when the fund level clearly exceeds the level required to cover risk.
There would be benefits to determining and allocating rebates separately from the pricing of risk. If the rebates were to be allocated based on past contributions, the cash flow between any bank and the insurance fund would have two components: one reflecting past premiums, and the other reflecting current risk exposure. Some institutions might find that their cash flows with the deposit insurance fund net to zero - the same as under the current system. But I would argue that, from an insurance perspective, not all zeroes are equal: A system in which banks always are charged explicitly for the marginal risk they pose is preferable to a system without this feature.
Charging explicitly for marginal risk is critical if premiums are to provide appropriate incentives and reduce moral hazard.
I also want to call your attention to another possibility under a mutual system - that is, banks having explicit claims on the insurance fund. I'm sure you all are somewhat familiar with the National Credit Union Share Insurance Fund, which is one example of a mutual system. Credit unions are required to maintain a one percent deposit in the insurance fund on an ongoing basis. This component of their contribution adjusts proportionally over time to reflect deposit growth or shrinkage.
One could envision a similar feature for the FDIC system to address our concern that institutions now can bring in substantial new deposits without contributing to the fund. The credit union system has raised concerns due to a double-counting of the one percent deposit; it is treated as an asset for both the member credit unions and the share insurance fund. A mutual approach would require careful consideration of how to structure any claims. The structure of the claims will determine the extent to which FDIC losses pass though to the books of your institutions. This again raises the specter of increased volatility, along with complex operational issues.
We look forward to sorting through these issues with you in the weeks ahead, as we continue to explore the relative merits of a user-fee approach versus a mutual approach.
Now, I don't want to leave you with the impression this morning that I am advocating rebates, or that I am advocating explicit bank claims on the fund, or any other feature of deposit insurance reform. Yes, the FDIC is undertaking a comprehensive review of the deposit insurance system, because it is clear that there are latent flaws in the system. Yes, we've held roundtables and industry outreach meetings on reform, and we have published our options paper that describes various ways in which the system might be improved. And, yes, we intend to bring recommendations early in the next Congress.
But, I want to stress that, to date, with the exception of our longstanding support of merging the BIF and the Savings Association Insurance Fund, we have not endorsed anything - although we are looking hard at the issues. That means that rebates are on the table - just as increased coverage is.
Deposit insurance reform doesn't necessarily mean you have to open your wallets, but we all need to open our minds to the possibilities of change.
The FDIC has and will continue to encourage participation by the industry - and others - in our review. We could have developed a reform package of recommendations behind closed doors and then presented it to the world, but we know that the issues are complex - and that we don't have a monopoly on knowledge and wisdom. We know that, everything else being equal, an open deliberative process that includes all that have a stake in the system will result in better policymaking. And we know we have a community of interest - in the stability of the banking industry, in protecting the taxpayer, and in assuring the financial integrity of the insurance funds.
Banking has changed over the past 125 years -- fundamentally and dramatically. Change will continue, either by choice or by circumstance. That is as true for deposit insurance as it is for anything else. Today we have an extraordinary opportunity to strengthen the system. Seize the day.