Federal Deposit Insurance Corporation
100th Annual Meeting
The Conference of State Bank Supervisors
Traverse City, Michigan
May 18, 2001
It is a pleasure to be here with you today, and to celebrate with you the organization's 100th anniversary.
We've seen a lot of changes in the past century. A hundred years ago, we had about 15,000 banks. They had a total of $14 billion in assets and $11 billion in deposits. Today we have fewer than 10,000 FDIC-insured institutions -- but they hold about $7 trillion in assets and $5 trillion in deposits.
One thing hasn't changed, however: Our job as bank supervisors, ensuring the safety and soundness of the banks we supervise. I want to commend your outgoing Chairman, Tom Curry, for the outstanding job he has done. Tom has been a leader among state regulators in his efforts to ensure fair competition, especially for community banks, without impeding industry growth. He has been a powerful advocate for achieving needed but balanced consumer protection in the areas of predatory lending, privacy and Internet banking.
And I want to congratulate your new Chairman, Elizabeth McCaul. She's Wonder Woman. Ten years as an investment banker at Goldman, Sachs. During her first days on the job at the New York banking Department she had to deal with the failure of NATIONAR, which provided liquidity and processing for hundreds of institutions. She successfully resolved that failure. She's been going strong ever since. She has overseen the operations of the Holocaust Claims Processing Office. She is an active director on many boards, including the Harlem Community Development Corporation.
On top of all this she is raising six children - amazing.
She has the intellect, the perspective and the energy to make CSBS' second century even better than the first. I know that Elizabeth, like Tom, will work to ensure the continued vitality and strength of the dual banking system.
I want to talk today with you about deposit insurance reform.
As you know, the FDIC has issued recommendations for reform. Our proposals fit well with the work we all do in examining and supervising banks. The safety of consumers' deposits and the soundness of the banks are best achieved if strong supervision is combined with economic incentives. If adopted, the FDIC reform proposals will provide important support for your work in keeping the banking system safe and sound and serving its communities.
There has never been a better time to address these issues. The banking industry's return on assets has been above 1 percent for eight years. It has never reached that height before in the history of the FDIC.
But good times don't last forever.
As you recall, just last year several large banks experienced earnings setbacks. The industry's net income, as a result, declined slightly from 1999's record level. More recently, 22 of the top 25 banking companies reported an increase in nonperforming assets during the first quarter of this year.
Even so, times are still good. And we need to fix the flaws in our system now, while conditions are good. We need to do this because the current system could exacerbate an economic downturn by raising premiums at the worst possible time -- pulling funds from banks and limiting credit availability when it is needed most.
A premium increase then would hit when banks are less healthy and losses are depleting the insurance funds, taking a toll on the safety and soundness of the banking industry.
All these reasons argue for taking action now.
The first problem we must address is the way insurance premiums are charged. Today, 92 percent of the industry gets federal deposit insurance for free. Free insurance distorts market discipline. Financial regulators around the world are coalescing around the idea that government should create incentives for the industry to manage risk appropriately. The U.S. deposit insurance system is out of step with that effort.
All insured institutions should pay for insurance. And they should pay based on risk -- that is to say, their probability of failure.
So how would the FDIC differentiate based on risk?
In our recommendations, we talk about one potential approach which uses a statistical failure-prediction model as the basis for a scorecard. That scorecard assigns depository institutions into risk categories by means of CAMELS ratings and financial ratios. The scorecard will incorporate data such as equity, income, nonperforming loan and other real estate ratios, funding and liquidity ratios, and growth.
The scorecard is similar to credit scoring models you are all familiar with and it appears to work well for the great majority of depository institutions. It could form a sound basis for risk differentiation for the thousands of small banks whose historical experience underlies the analysis.
This approach would have to be modified to assign large, complex institutions to appropriate risk classes. Their risk characteristics do not fit neatly into the same statistical profiles, as do small banks. Moreover, there is information available on larger institutions that is not available for their smaller counterparts. One, they have continuous on-site supervision. Two, they are monitored by ratings agencies. And, three, they have debt and equity whose prices may reflect useful market signals about risk.
The scoring system must accommodate these inherent differences. But the FDIC does not believe it is appropriate to otherwise differentiate systematically among institutions on the basis of asset size alone in setting premiums.
Let me put it this way: If small banks currently rated 1A -- our top insurance category -- were divided into three risk groups, large banks rated 1A should be similarly divided.
We constructed a number of examples to show how all this might work.
The results: Under current economic conditions, about 50 percent of current 1A banks -- or about 4,500 banks -- would pay 1 basis point.
Thirty percent -- or about 3,000 banks -- would pay 3 basis points.
And 20 percent -- or about 2,500 banks -- would pay 6 -- before rebates.
Our recommendations incorporate the use of rebates to be paid based on past contributions to the insurance funds. Using rebates, we can collect a regular charge for risk while preventing the insurance fund from growing too large. Rebates would be given only if and when the FDIC determines, year by year, that the fund exceeds a size needed to cover expected losses.
In today's economy, the aggregate rebate to the industry would be about $1.4 billion. Therefore, the 50 percent group paying 1 basis point is likely to get more money back from the FDIC in rebates than they paid in premiums. The 30 percent group would net out to about a 1 to 2 basis point premium. And the 20 percent group would net out to about 4 to 5 basis point premiums.
As you know, that's not too different from the historical average of premium costs for all banks.
The second problem we must address is volatile premiums when banks are in bad shape. Today, average premiums for most banks would jump from zero to 23 basis points if the Bank Insurance Fund were to drop so low that it could not get its reserve ratio back to 1.25 percent within a year.
If that were to happen, the law requires that we charge average premiums of at least 23 basis points. As you know, that's the kind of premiums banks paid in the early 1990s. Today, such a premium would drain almost $9 billion from insured banks and thrifts - and could lead to a $65 billion contraction in industry lending.
We, at the FDIC, are not expecting any such drop in the BIF anytime soon. However, a confluence of events in the future could bring down the reserve ratio. New deposit growth, combined with an increase in bank failures could cause such a drop fairly quickly.
It is for this reason that we recommend small but steady premiums over time. This would make it unlikely that premiums would climb so high during bad times.
Let me give you an example of how this would work.
In the simulations we ran, a "typical" institution in the best-rated category would pay about 10 or 11 basis points in a significant downturn, rather than 23 basis points under the current statutory framework.
Consider a community bank with $100 million in assets. At 11 basis points, it would pay about $110,000 in assessments. And at 23 basis points, it would pay more than $230,000 -- a $100,000 difference.
Similarly, a large bank with $150 billion in assets would pay about $90 million versus $190 million -- $100 million difference. And, given the numbers, we think this plan is better for the fund and better for individual banks.
And there's one other issue we want to address with our proposal for steady risk-based premiums: the inflow of new deposits covered by the insurance program, without any corresponding payment for that coverage - the so-called "free-rider" problem.
This inflow is not new. It has been occurring over the past five years. It was more pronounced than ever last year -- deposits grew at 6 percent - compared to a historical growth of 3 percent annually in recent years. In the past year, a couple of brokerage firms have swept more than $84 billion into their affiliate banks. Most of those funds are insured, and have reduced the Bank Insurance Fund reserve ratio by about 3 basis points.
And the inflow of deposits may continue - in various ways - as new products come on line. While this deposit growth is a successful and legitimate activity, there is no denying that, older banks can rightfully say that they are subsidizing the deposit insurance costs of new and fast-growing banks. In total, about 900 banks have never paid for their insurance coverage.
In addition, these new deposits could, over time, contribute to the dilution of reserves sufficient to trigger premium payments by all banks. This issue has gotten the attention of many bankers interested in change.
So there you have it. Deposit insurance reform is an opportunity -- an opportunity to enhance our safety and soundness efforts as bank supervisors. To make the most of that opportunity, we must act now - without the pressures and distraction that an economic downturn would bring, or the urgent demands that would arise in a crisis.
I want to close on a personal note.
In 1983, I was honored to become a member of CSBS. Sydney Bailey of Virginia was the CSBS President then. Larry Kreider was the Executive Vice President.
As FDIC Chairman, it has been an honor and privilege for me to work, once again, with CSBS, and all of you. Serving as FDIC Chairman is the experience of a lifetime.
I want to leave you today with one thought about the FDIC and CSBS. The law of the United States prescribes a five-member board for the FDIC. The law does not require that any member of the FDIC Board of Directors be a banker. The law does not require that any member of the FDIC Board of Directors be an attorney. The law makes only one stipulation as to experience. And that is that at least one member of the FDIC Board must have "state bank supervisory experience."
Think about that.
That is how much your perspective is valued. That requirement will ensure that the close ties between the CSBS and the FDIC continue for many years to come.