Remarks

by

Director, Office of Thrift Supervision

at the

Newport, Rhode Island

June 17, 1999

This is a wonderful opportunity to bring together many of the issues that are high on OTS’ agenda and are of great importance to me personally.

The four Federal financial regulatory agencies have worked hard to make the revised CRA regulations a success. These interagency efforts have included everything from joint examiner training sessions to the publication of the CRA Questions and Answers, which serve as the primary vehicle for guidance and interpretation of the regulation. The transition to the new regulations has gone well and the general consensus is that the new regulations have accomplished their goal of being more performance-oriented. But as you have heard hints of this morning and, surely, will hear more about this afternoon, there are still some difficult issues to grapple with as we move forward with our administration of CRA.

I have organized my comments around three concepts in the CRA regulations that are designed to help us measure an institution’s performance:

• the definition of an institution’s assessment area,
• the quantity of an institution’s lending, investment or service activity, and
• the quality of those activities.

Let’s tackle the most difficult subject first: the assessment area issue. My apologies to the next panel – aptly titled 21^st Century Trends and the Unimportance of Assessment Areas – if I cover some of the points they intend to discuss.

The assessment area is the geographic area within which the agencies evaluate an institution’s CRA performance. So it is, by definition, a fundamental part of measuring an institution’s CRA performance. For most institutions, the regulation’s requirement to delineate assessment areas that correspond to commonly recognized metropolitan areas or political subdivisions that surround their branches or deposit-taking ATMs is adequate. But more and more institutions are using alternative product delivery systems rather than – or in addition to – the traditional brick-and mortar branch structure. It is not unusual now for an institution to use mail, telephone, loan production offices, agent relationships, affinity relationships, and the Internet to market and deliver banking services. The reach of these systems transcends the regulation’s focus on assessment areas in evaluating CRA performance. And, even with its intended flexibility, the revised rule challenges an examiner’s ability to measure the CRA performance of these institutions.

The problems with the assessment area definition are as varied as the business strategies of financial institutions. Some institutions choose to be very visible in the area where their main office is located. These institutions often wish to focus their attention on the low- and moderate income areas within that community, much the same way that wholesale and limited purpose institutions are allowed to do under the community development test. Other institutions have little or no presence within the area where they are headquartered. Internet banking where, in its purest form, all deposit and loan related activity would take place in cyberspace, vividly illustrates a business strategy that does not look to any one local area for its identity. Rather, these institutions view the entire nation, and sometimes the world, as their "community."

And, of course, there are an increasing number of traditional retail institutions that are expanding the geographic reach of their deposit and credit markets by introducing alternative service delivery systems as increasingly important adjuncts to their underlying branch networks, or as separate, independent business strategies. Just a few weeks ago, for example, First Union announced its intention to use the Internet to expand its business rather than its previous strategy of mergers.

The first and most important point to make about non-branch based institutions is that, with limited exceptions, all banks and thrifts have a CRA obligation; the issue we’re wrestling with is how to evaluate their performance of that obligation. One obvious question is whether it is time to expand the traditional notion of community.

The regulation’s definition of an institution’s assessment area is based on the idea of community we have worked with since 1977. It was reasonable then to assume that branches were the primary means by which institutions gathered deposits and so should define the area in which they should provide credit and services. But if we evaluate non-branch based institutions solely on the basis of assessment areas delineated around a single home office or branch, those institutions would have little incentive to meet the needs of low- and moderate-income borrowers in areas where they lend but have no branches. We need to guard against implicitly fostering a perverse, unintended outcome of actually lessening the CRA obligations of those who operate more and more in nontraditional ways. It makes no sense to create a system of CRA evaluations that distinguishes, not on the basis of the credit products offered by an institution, but on the method by which those products are delivered.

Alternatives

So what’s the answer? The agencies have developed several alternatives to consider; undoubtedly there are others, which we’d like to learn about.

One possibility is to expand the scope of the community development test. The new regulations define two categories of institutions that can avail themselves of the community development test: wholesale and limited purpose. Wholesale institutions are those that are not in the business of extending home mortgage, small business, small farm, or consumer loans to retail customers. Limited purpose institutions are those that offer only a narrow product line (such as credit card or motor vehicle loans) to a regional or broader market. These institutions, such as CEBA credit card banks, were difficult to evaluate under the old regulations. As wholesale or limited purpose institutions, they are now evaluated on the basis of their community development lending, qualified investments and community development services in their assessment area and the broader statewide or regional area that includes the assessment area. Moreover, where the institution adequately meets the needs of its assessment area in this way, examiners also consider its community development activities outside of its assessment area or broader statewide or regional areas that include the assessment area. This feature of the community development test is particularly helpful to institutions with a national deposit or lending reach.

However, because they offer a full array of credit products to the retail public, many of the nontraditional institutions that are presenting assessment area issues under the regulation do not qualify for treatment as either wholesale or limited purpose institutions. One possible avenue for addressing the assessment area problems of these institutions might, therefore, be to broaden the eligibility for the community development test to encompass a larger variety of nontraditional institutions.

On the other hand, the CRA regulation, at its core, is based on the notion that a financial institution that provides retail services should help meet the needs of its entire community by providing loans and service in that community. There are many non-branch based institutions lending on a regional or nationwide basis to all borrowers, including those with low or moderate incomes. Should these institutions be able to demonstrate their commitment to CRA with their lending performance in all income segments in all lending areas without regard to where they may have an office?

For example, the agencies could expand upon the regulatory definition of assessment area by allowing institutions to delineate areas not only where they have their main office, branches and deposit taking ATMs, but also where they either gather a substantial amount of their deposits or make a substantial portion of their loans. If the agencies were to take this approach, they would have to find a way to prevent institutions from delineating only those areas where their CRA performance looks good. However, I am sure that precautions against this kind of "cherry picking" can be designed.

Another possible approach is suggested by the statutory assessment area provided for institutions that are established primarily to serve military personnel or their dependents. As Congress recognized several years ago, the notion of a geographic assessment area is irrelevant to evaluating the CRA performance of an institution whose customers are spread across the world. Their solution: create a customer-based assessment area. This solution is one that, by analogy, may be applicable to other non-branch based institutions. It is certainly an alternative worth exploring, particularly as more institutions are being chartered as a means of servicing pre-existing customers of a credit union or an affiliated financial services provider, such as a brokerage or insurance company.

Sometimes I think there are as many different options for solving this problem as there are business strategies of financial institutions. Which brings me to the strategic plan option. The regulations give all institutions the option of operating under an approved strategic plan.

The strategic plan option requires that an institution consult with community representatives on an informal basis in developing a plan for helping to meet the credit needs of its community. Once the plan is developed, the institution must formally notify the community that it is available for review for a period of at least thirty days. Plans must then be submitted for approval to the institution’s primary regulator. The regulations require that strategic plans have annual, measurable goals. They may cover a period of up to five years.

Why would an institution pursue the strategic plan option? It provides more certainty in the evaluation process for an institution willing to spend the time to develop a plan in consultation with its community. For institutions with unusual business strategies, the strategic plan provides flexibility, because it allows institutions to tailor the criteria used in their evaluation to their business strategies. The plan option allows institutions to emphasize investments or service over lending or to combine lending, investment and service goals in any way that is justified by the institution’s business plan and the needs of its community.

The strategic plan recently approved by OTS for Household, FSB offers a good example of the flexibility institutions have in tailoring their CRA obligation to their particular business strategies. In that case, Household established goals for community development activities within the Chicago metropolitan area, where its home office is located, as well as lending goals related to its consumer loans with the AFL - CIO. Household’s business arrangement with the AFL - CIO illustrates how the customer-based relationship can substitute for a branch-based assessment area. Household’s plan demonstrates that serving consumers with a particular affinity—in this case AFL-CIO membership -- can effectively address an institution’s obligation to help meet the credit needs of the low- and moderate-income segments of a nationwide market.

Unfortunately, there have been very few strategic plans approved since the new regulations became effective. It seems that the option is perceived as more work than it is worth.

We often hear that institutions are concerned about giving away confidential information in their business plans or that community groups will ask for more than institutions feel they can commit to in a plan. We don’t believe that these concerns should drive institutions away from exercising the strategic plan option. The regulations specifically allow institutions to provide confidential information to the agencies separately from the strategic plan. All that we ask is that the goals contained in the plan are measurable and specific enough for the public and the agency to judge its merits.

None of the institutions that submitted plans to the agencies indicated that working with community groups to develop the plan was particularly difficult. And, quite frankly, our review of public comments submitted in the formal comment phase leads us to believe that the concern over public participation is overstated. Moreover, nothing in the regulations or in any interagency guidance we have issued on the strategic plan suggests that the agencies would require an institution to meet the demands of any particular commenter. On the contrary, we are looking for reasoned comments from the public on credit needs in the community and appropriate responses on the part of institutions to those comments.

Even so, for most traditional lenders the strategic plan may be more work than is necessary. But for institutions that are using non-branch delivery systems to do business, this option may be well worth the trouble and time to do a plan. The rewards are these: your CRA obligation can be met using the expertise that you have within your institution, with goals that reflect your institution’s overall business, and you can rest well knowing exactly what needs to be done to accomplish that satisfactory or outstanding CRA rating at the next examination.

Let me reiterate: no matter what the right answer is to the assessment area question, one thing is certain: institutions have an obligation to the low and moderate-income segments of the communities they are chartered to serve. That obligation is not going away. Our goal is to be sure that financial institutions have the flexibility to meaningfully meet their CRA obligation within the context of their business and the communities they truly serve.

In the meantime, OTS is working within the framework of the regulations with the increasing number of thrifts that are using non-branch delivery systems. Recently, we have addressed these issues with insurance company applicants for thrift charters. Beginning with the Travelers application, and continuing through the State Farm application, we have interpreted the CRA regulation to capture the level of performance of an institution throughout the markets in which it does credit business, not just in its main office assessment area, but well beyond. We believe we have given practical meaning to the Congressional mandate to assess an institution’s record of meeting the credit needs of its entire community.

Because details of broad CRA obligations can be shrouded in the confidentiality of proprietary business plans, let me enunciate the rule we have applied to applicants proposing non-traditional retail operations:

• Applicants for a thrift charter that propose to engage
• in nationwide or super-regional
• home mortgage or multi-product consumer lending to the retail public
• through non-traditional means with a single main office or branch
• must demonstrate the capacity to achieve satisfactory performance of its CRA obligations
• in lending, investment and services
• (1) by at least adequately addressing the needs in its main office assessment area, given the performance context of its operations in that area,
• (2) by showing that the prospects for its retail products penetrating low- and moderate-income markets in the regions it reaches outside its assessment area are favorable,
• and (3) by demonstrating that its community development lending, qualified investments and community development services provide appropriate levels of benefit to appropriate markets throughout the scope of its thrift operations.

This is precisely the rule behind the result in the State Farm application. Although there is a misconception that State Farm’s thrift’s CRA performance will be limited to the MSA surrounding its Bloomington, Illinois headquarters, that is emphatically not the case. Yes, that MSA is the only assessment area mandated by the CRA regulation, but that assessment area is neither the limit of State Farm’s thrift’s proposed performance, nor the limit of OTS’ evaluation of State Farm’s thrift’s record of helping to meet the credit needs of its entire community. OTS will evaluate State Farm’s thrift in the rest of Illinois when its operations reach there. OTS will follow State Farm’s thrift to Missouri and Arizona when the projected expansion to those states is finally realized. In the further future, OTS will evaluate this institution’s performance as it grows and expands to other states and regions. All of this will happen even if there is still only one main office assessment area. What is more—State Farm knows this and has demonstrated its capacity to satisfactorily meet these expectations.

Now some of you are probably wondering what magic is this? Let me assure you that this policy is well founded in the CRA regulation and the Q&A commentary that has been issued on an interagency basis. OTS is not breaking with its fellow banking regulators. We are still committed to resolving the thorny issues around assessment areas with an interagency consensus. However, in the meanwhile, OTS must apply the existing regulation to the real situations presented to us, taking what guidance has been provided and making our best judgment about how to accomplish the statutory mandate contained in the CRA to evaluate an institution’s record of meeting the credit needs of its entire community, not just one local community, but all local communities it is chartered to serve.

One final caution. Although OTS has created a conceptual framework to deal with evaluating the non-traditional thrift, the ultimate examination methods and practical judgments about how to assess such a thrift’s performance must be dealt with. Time, experience and dialogue with the institutions we regulate and the communities they serve are needed to develop effective and consistent practices in a varied and dynamic environment.

Let me now turn to issues that surround the second measure of CRA performance that I mentioned: the QUANTITY of an institution’s lending, investment and service activities. It is inevitable that a regulation designed to be performance-based should have quantitative measures. It is also inevitable that the institutions subject to the regulation will try mightily to determine "How much is enough?" As you might guess, I am not here to answer that question. Rather, I would like to offer my observations on how the regulation’s quantitative criteria affect the lending, investment and service activities of financial institutions.

The quantitative criteria take on a couple of forms in the CRA regulations. First, there are the criteria within the separate tests that directly evaluate the number and dollar amount of loans and qualified investments or, in the case of services, the distribution and accessibility of existing branches and the number of branches opened or closed within the examination period. In each of these tests we also evaluate the proportion of these activities within the assessment area and the distribution of loans, investments and services among borrowers and geographies by income segments.

Second, the regulations impose a quantitative measure when arriving at an overall rating of the institution’s performance. In simple terms, the lending test is weighted at 50 percent of an institution’s overall grade; the investment and service tests count for 25 percent each. This formula was established to ensure the primacy of lending when evaluating CRA performance. The agencies built several controls into the formula to assure that an institution could not receive an overall rating of "satisfactory" without having satisfactory performance on the lending test.

Data analysis software affords significant assistance when measuring CRA performance under the lending test. Not only does it afford a multitude of comparisons of loan performance, but it also enables examiners and others to develop detailed understandings of an institution’s performance context. Unfortunately, comparative data on consumer loans is less available than data on home mortgages and small business and farm loans. Efforts to create good measures of investment and service performance are also hampered by a lack of comparative data.

As I mentioned, institutions are understandably interested in finding out how much is enough when it comes to the quantitative criteria in each of the performance tests. This has led to some interesting results. For example, because the lending test for large, retail institutions considers both originations and purchases of loans, institutions have an incentive to purchase loans in order to hit the numbers they believe are necessary to achieve a good lending test rating. This incentive can lead to the purchase and sale of the same loans over and over again among institutions covered by the CRA. Such activity does not lead to more dollars in the communities where credit is needed, however.

The regulations consider purchased loans in the lending test because sometimes there are good reasons for purchasing loans. Purchased loans are an important source of liquidity, especially where one or more institutions that are not very good at making some of the more difficult loans needed in the community team up with others that have such expertise. But I believe that the agencies must find ways to discourage the churning of loans from one institution to another for the sake of hitting the numbers they believe are necessary to achieve a good lending test rating. The regulation’s emphasis on lending in the overall rating, combined with its equal consideration of loans that are purchased as well as originated during the examination period, may have encouraged some institutions to count loans and buy market share rather than to creatively meet the needs of their communities, including needs for services and investments as well as loans.

I am also concerned about how the quantitative criteria in the regulation affect institutions that are "large" for examination purposes, but, in a relative sense are dwarfed by their competition. Smaller, community-based institutions face stiff competition from larger entities with a national marketing reach. Some are having a difficult time keeping their loan rates in line or in processing loan applications as efficiently as their competitors. These competitive difficulties challenge these institutions to find creative ways to support their neighborhoods – whether through different loan products better matched to community credit needs, or through deposit or other services. Currently, our examiners consider these contextual factors as part of a CRA evaluation. Even so, the CRA regulation should encourage and reward this creativity better than it does now.

I would be remiss at a conference covering both CRA and Fair Lending topics, if I did not mention an important issue with respect to monitoring fair lending performance. As you know several agencies, including the OTS, have urged the Federal Reserve Board staff to amend Regulation B by removing the prohibition against collecting monitoring information (race, gender, national origin) on other than real estate loans. I believe that allowing collection of this information would enable institutions to better monitor their fair lending performance. If the Federal Reserve implements our suggestion, I urge all institutions to take advantage of this opportunity to gather the information necessary to better measure how well you are serving all market segments and to assure that your products are reaching all parts of your communities in a non-discriminatory manner. In the interim, I commend our Q&A’s on Special Credit Programs, also on our website.

Finally, the QUALITATIVE criteria. The agencies, in response to the many comments we received after the first proposal to revise the CRA regulations, included qualitative considerations among the regulation’s evaluative criteria. We agreed that counting up the numbers of loans, investments and services alone would not be a true indication of CRA performance. At a conference like this, it is worth remembering that no matter how sophisticated our data analysis capabilities become, CRA performance is measured by more than crunching numbers. By including qualitative criteria, such as the responsiveness, innovation or complexity of a particular activity, we had hoped to encourage the small dollar activities that have an impact within a community far beyond what the numbers indicate.

That was our intent. All too often, however, we are given reason to doubt that our intentions have been understood. We hear, for example, that institutions are reluctant to continue with programs or community partnerships that work because they have been done already and, therefore, are no longer "innovative." And we still hear the tales of frustrated bankers who are in their communities doing the hard work, who believe that the agencies are more likely to reward large dollar investments in targeted mortgage-backed securities. I want you to know that’s not OTS’ policy. But we clearly need to do more to clarify the investment test for institutions and our own examiners, emphasizing that both innovation and creativity and successful – even if tried and true – community partnerships will be rewarded.

Another aspect of qualitative consideration that I believe is relevant to the CRA evaluation but has largely been ignored up to now, is the terms upon which credit is made available to low- and moderate-income borrowers. I addressed issues related to the development of risk-based pricing models in the mortgage market at length in a speech I delivered in November – which you can find on our Web site: www.ots.treas.gov. I don’t want to get into a discussion of the sub-prime market here except to say that I believe that properly administered risk-based pricing can, in fact, broaden the market and improve homeownership opportunities. Risk-based pricing has the prospect of bringing mainstream lenders into lower tiers of the credit market that until now have had to rely on very high-priced, often predatory, alternative institutions.

Nevertheless, predatory lending does occur, particularly in connection with debt consolidation, home improvement financing or home equity financing. How should we evaluate, in the context of CRA, loans to low- and moderate-income individuals that are made at a cost so high that they may actually be counterproductive to meeting the community’s credit needs? At this point we have no interagency answer to this question. But I think it only fair, having raised the issue, to give you some idea how I feel about it.

First, I understand consumers have an obligation not to fall for deals that are too good to be true, and community groups have some responsibility for helping their neighbors understand what’s being offered and why it might not be such a great deal. But the difficulties humans have resisting temptation is well documented – starting with Genesis. Those who offer products that the target audience can’t understand and would be horrified about if they could, bear the bulk of the responsibility.

Let me caution the thrifts in the audience that there are enough potential violations of enough laws involved in these practices so that any OTS-regulated thrift even thinking about engaging in them – and I must say so far we’ve seen somewhere between few and none – better think again. Examiners who identify unethical or predatory practices as part of an institution’s penetration of low- and moderate-income markets will note this behavior and consider it unfavorably when describing and rating its performance as part of the CRA public evaluation. Moreover, we will make referrals to the Justice Department if we find fair lending violations and will take prompt supervisory action to correct violations of any other laws or regulations we find.

There are plenty of good loans out there at decent spreads. Responsible lenders can use sound risk-based pricing technology to successfully and safely compete with predatory lenders and drive them out of the marketplace. It will come more easily if you provide good banking services to the target market, making them feel welcome as depositors and check cashers as well as borrowers. I would encourage you to do just that. The subprime market is a viable source of good business and worthwhile CRA performance, if engaged in wisely.

In conclusion, I invite you to share your observations about the issues that I have raised with me or with your primary regulator. There is no reason to wait until the year 2002, when the agencies promised to conduct a full review of the CRA regulations, to raise issues or suggest possible answers. Between now and then, we will be tackling a number of CRA measurement challenges. The creative input of industry and community representatives, examiners, and others is welcome as we formulate solutions to these challenges.

I thank you for your attention and look forward to further discussion of these issues.