The Federal Deposit Insurance Corporation Improvement Act of 1991 requires institutions that propose to close branches to notify regulators and submit supporting material. This notification invites rather routine inquiry into lending and other activities in troubled areas.
Those charged with Community Reinvestment Act compliance may find it instructive to review Ameritrust Corp.'s experience with branch closings and related consolidations. Ameritrust was taken over by another Ohio banking company, Society Corp., in early 1992.
Ameritrust applied to the Office of the Comptroller of the Currency in late 1991 for, among other things, permission to consolidate a handful of inner-city Cleveland branches into two new, upgraded branches in the same general area.
Claims of Redlining
The mayor of Cleveland, and community groups and activists, filed protests with the OCC, asserting that Ameritrust was abandoning the inner city. These protests claimed that it engaged in systematic discrimination, or redlining, against the African-American community.
The protests focused particularly on the number of home-purchase loans made in the inner city versus more affluent neighborhoods and suburbs.
Like many large urban areas, the inner city of Cleveland is predominantly African-American -- over 90% in many neighborhoods. More affluent portions of the city and its suburbs are predominantly white.
Overlaying loan-distribution data with racial-composition data, the protesters believed the emerging pattern could be explained only by discrimination against African-American borrowers.
Deeper Look at Data
These protests, and the resulting OCC special examination, compelled us to examine in minute detail Ameritrust's lending data, together with data Ameritrust was not accustomed or not required to keep on hand.
We were convinced that Ameritrust deserved the satisfactory CRA rating it received after its routine 1991 examination. Nevertheless, we faced the very real possibility that our branch applications would not be approved and, worse, that Ameritrust would end its independent existence with a black eye.
The data were troubling at first blush, so it was a challenge to prove that they did not mean what the mayor and his fellow protesters said they meant. Happily, we prevailed.
In a little more than three weeks after the protests were filed, we amassed material from outside resources such as the Census, local economic opportunity organizations' analyses of local and statewide poverty data, and the county recorder's compilations of home-sale figures.
Marshaling Internal Sources
Much more came from internal sources. We compiled data on Ameritrust's lending for home purchases (both conventional and government-assisted), home improvement, small business, and other consumer purposes; as well as deposit analyses and numerous other pieces of information.
These data indicate the average age of homes in Cleveland's many neighborhoods and the median housing values in those neighborhoods. They include the number of sales:
* In each such neighborhood in 1990 (the year particularly at issue).
* Consummated without the use of any financing.
* Employing government-assisted mortgage financing.
Also included are the number of households that own their own homes, as opposed to rent, and the number of person over the age of 18 in these homes.
A customer usage survey conducted in early 1990 supported the proposition that our branch consolidations would have little impact on customer convenience. This survey also helped us lay a strong foundation for the economic justification of our branch closings.
We examined geocoded home-purchase and home-improvement loans, and we geocoded 1990 small-business-lending data.
Although not a single protester mentioned small-business lending -- let alone complained about it -- it was an important piece of the total mosaic.
We argued successfully that the role of small-business lending in stimulating economic activity in a given area, with consequent creation of jobs and tax revenue, deserved at least as much attention as home-purchase lending.
Protesters asserted that the percentage of home-purchase loans made by Ameritrust in 1990 in the inner city compared very unfavorably with home-purchase lending in outlying portions of the city and the suburbs.
There was indeed such a disparity in loan distribution in 1990, whether viewed in isolation or in comparison with other leading local banks.
A closer examination, however, revealed that:
* Ameritrust dominated the home-purchase loan market in the city generally.
* An extremely large number of home-sales transactions in the inner city were consummated without any financing, because the value of inner city homes tends to be extremely small. (In some African-American neighborhoods there were no conventional home-purchase mortgages made in 1990 - not one.)
* The vast majority of home-purchase loans in the inner city were government assisted.
* The government-assisted lending market is dominated by mortgage brokers, not banks.
* Ameritrust had only recently redoubled efforts to qualify lending personnel to arrange government-assisted mortgage loans, with all their complex and labyrinthine requirements.
We also pointed out that Ameritrust was a lending home-improvement lender in the inner city. Because market share figures for this product were not available, we were compelled to present Ameritrust's data in a vacuum. But the magnitude of the figure refuted the proposition that Ameritrust was a discriminatory lender.
Three-year trend data also demonstrated that the emphasis of Ameritrust management on home-improvement lending was apparent throughout the county and state, not merely in the inner city.
Given the demonstrable difference in the average age and value of homes in the inner city versus outlying neighborhoods in the city and county, we made a convincing argument that Ameritrust was a very active participation in the stabilization of Cleveland's most impoverished neighborhoods.
One CRA assessment factor is "evidence of prohibited discriminatory or other illegal credit practices." Clearly, though, a crude analysis of the distribution of one type of loan only -- which ignores the underlying reasons -- should not be sufficient to suggest failure of CRA compliance, let alone be taken as "evidence" of discrimination.
The CRA's primary purpose is to ensure equitable distribution of loans in low-and moderate-income areas. It is not intended to ensure that Afrian-Americans obtain as many loans -- relative to total population -- as whites -- relative to theirs.
Branch Usage Patterns
A 1990 survey of more than 10,000 customers helped us demonstrate that the branch consolidation and associated closing at issue would have little or no impact on customer convenience.
The survey enable Ameritrust to delineate each branch's primary trade area, based on branch-usage patterns of actual customers. Plotting patterns on a map offered stark visual evidence that a number of branches were competing for the same customers. In some cases, trade areas of individual branches completely overlapped.
The graphic depiction of trade areas made them appear rather unnatural. But the detailed, credible survey of customers helped us lay a strong foundation for the decision to close branches.
We were then able to focus on hard data such as deposits and population trends. In addition to complete overlap in at least one case, and greater than 50% in others, we noted a clear and longstanding decline in the subject branches' deposits and in the population in surrounding areas.
Bars Against Expansion
These were some of the oldest and poorest sections of Cleveland and, as one would expect, the facilities were old and in need of repair. In many cases the facilities were unsuitable for continued or expanded use because of limited parking space, limited ability to put in drive-through ATM facilities, and similar problems.
Again, market share data allowed us to demonstrate that Ameritrust was by far the dominant bank presence in the inner city and would continue to be so even after the branch closings.
These market-share data also provided concrete proof of inefficiency. In one case, Ameritrust had more branches in a particular neighborhood -- 54% -- than all of its competitor combined, but only 31% of deposits. deposits had declined over 30% over a three-year period in the same neighborhood, while population data showed a similar decline.
It is important to define trade areas on a basis that is supportable, preferably bearing some relation to actual customer usage. Some research should be done into market share, population, deposit trends, and the like if they are not otherwise readily available.
Ameritrust's closing predated the branch-closing provision of the Federal Deposit Insurance Corporation Improvement Act. Nevertheless, we faced the possibility that these branch consolidations would have been disapproved without adequate support. It is perhaps no coincidence that our applications were approved on the very day -- Dec. 19, 1991 -- the improvement act became law.
Home Mortgage Disclosure
The protests against Ameritrust were filed with the OCC before the disclosure of nationwide Home Mortgage Disclosure Act data on loan denial rates to minorities. As it happened, denial rates were not part of the protest, but I am confident they would have been if the HMDA data had been related earlier.
Ameritrust, like many banks, had denied home mortgage loans to African-Americans at a greater rate than to white loan applicants. So we sampled Ameritrust's loan application files to prepare for a protest that seemed highly likely, but ultimately never occurred.
I would expect any future Community Reinvestment Act protests to focus as a matter of routine on loan-denial rates.
The Federal Financial Institutions Examination Council's late 1991 policy statement on analysis of geographic distribution of lending under the CRA states: "An analysis by each covered institution of its own HMDA data is a necessary element of an analysis of that institution's geographic lending patterns" for purposes of CRA compliance.
Ultimately, CRA protests of the type we faced must address hard data that banks have or can easily obtain, and know how to use. Banks are also in a position to preempt protests in the first instance in a number of ways.
Not least among these efforts are, of course, good-faith efforts to identify segments of low-and moderate-income neighborhoods that are creditworthy and underserved by local banks.
This might require:
* More creativity and persistence to identify creditworthy customers in low- and moderate-income neighborhoods.
* Deviation from the bank's historical product mix and underwriting criteria.
* Holding more such loans in portfolios, rather than for resale.
If a deficiency is identified by the bank during a protest, frank acknowledgment can lend credibility to its entire CRA compliance efforts.
For example, we made clear in our response to the branch protest that we undertook for the first time an analysis of the loan-to-deposit ratio of each branch in each of Cleveland's many neighborhoods.
A number of inner-city branches turned out to be well below the average loan-to-deposit ratio (excluding business deposits, jumbo deposits, and nonconsumer loans).
Routes to Improvement
Thus, we identified a concrete example of how we might improve our CRA performance. We then undertook the ratio for each branch on a quarterly basis and made improvement of the ratio a part of each branch's performance review.
A high-visibility compliance program may be more likely to produce an outstanding CRA rating than actual success in reinvestment act lending, but this could change.
A new direction could require banks to report routinely on certain loan-distribution analyses of the type Ameritrust and others have undertaken.
Banks could also be required to conduct routine samplings and analyses (using credit scoring for all types of loan applications) of their loan-application files in order to present hard proof that African-American loan denial or foreclosure rates are not the product of discrimination.
This is the direction in which the recent examination council's policy statement and recent Federal Reserve comments on the disclosure act data are heading.
If the Community Reinvestment Act is worth having at all, it is worth making it work. And I don't know how you know whether it works unless you undertake a disciplined analysis of the facts.
I wouldn't be at all surprised to see the regulators, community leaders, and activists up the ante in the future over CRA compliance, requiring more hard proof of compliance and public reporting of some or all of the results.
I can envision a standardized system for testing compliance that would allow comparisons among institutions, as well.