Daniel K Tarullo: Tailoring community bank regulation and supervision

Speech by Mr Daniel K Tarullo, Member of the Board of Governors of the Federal Reserve System, at the Independent Community Bankers of America 2015 Washington Policy Summit, Washington DC, 30 April 2015.

The views expressed in this speech are those of the speaker and not the view of the BIS.

On numerous occasions over the last year, I have discussed the importance of differentiating prudential regulation and supervision based on the varying nature of the risks posed by different groups of banks.¹ This differentiation needs to be explicit both in the analytic foundations of our prudential system and in the application of that system to banking organizations.

At an analytic level, we need to be clear that prudential aims vary with the risks posed by diverse groups of banks. To take an obvious example, the risks to the entire financial system that would arise from the failure of a very large, universal bank call for different prudential objectives from those suitable for a $200 million community bank. At the level of application, we need to tailor prudential regulations and supervisory practices with an eye both to those varying objectives and to the characteristics of differing groups of banks.

This morning I would like to revisit, and develop a bit more, the concept of regulatory and supervisory "tiering" as it pertains to community banks. First, I will lay some analytic foundations by identifying what I believe to be the most salient characteristics of community banks for determining suitable regulatory objectives and prudential policies. Second, I will turn to the implementation of those objectives by explaining how we have tried in practice to shape supervisory policies appropriately. Third, I will touch on the implications of those objectives for some regulatory issues, including capital requirements.

Prudential objectives for community bank oversight

At the outset, let me note that in speaking of "community banks," I am generally using the term as defined in the Federal Reserve's supervisory portfolio. That is, it refers to banks with $10 billion or less in total assets. But it is worth making two observations about this definition. First is the fact that the vast majority of community banks have less than $1 billion in assets. Second is the fact that, for purposes of establishing regulatory objectives, a bank with $12 billion in assets might not be readily distinguishable from one with $8 billion in assets. Of course, lines will always have to be drawn in applying regulations and establishing supervisory practices. But it is useful to keep these facts in mind in thinking about prudential tiering.

The two most important characteristics of community banks for purposes of establishing regulatory objectives and supervisory practices are their size and their business model. My reference to size is obviously a bit tautological, but size is worth emphasizing precisely because of what it says about the risks community banks do not pose. The possible failure of a community bank self-evidently poses no risks to the financial system. And while individual community banks in smaller communities provide sources of credit that would be hard to replace, their limited size means their failures would not result in credit contraction in significant swaths of the country.

The business model of nearly all community banks is grounded in the most traditional form of commercial banking - lending to businesses and households with funds predominantly obtained from deposit accounts. And, as this audience well knows, lending by community banks is built substantially on relationship banking. While community banks over the years have found it increasingly difficult to compete with larger banks in the types of lending that can be efficiently scaled through larger volumes and standardized credit models, they maintain a competitive advantage relative to larger banks through knowledge of their local communities and their individual borrowers. As a result, community banks play a unique role in their local economies, particularly with regard to lending to small- and medium-sized businesses. The relationships these institutions have with their customers oftentimes mean they can look beyond traditional credit factors to consider unique borrower characteristics when making credit decisions and to reduce information failures about borrowers' willingness and capacity to repay loans. Numerous studies have documented these advantages and their value to economic development. One recent study found that loans extended by rural community banks to small businesses default less frequently than similar loans granted by their urban counterparts, and that the performance advantage is greater when the bank and the borrower are located in the same county.²

The traditional intermediation provided by community banks, along with the relatively small scale on which each community bank provides these services, suggests that the principal relevant prudential aim should be the equally traditional one of protecting the deposit insurance fund. A second aim should be to help protect the availability of credit in geographic areas or to particular classes of borrowers unlikely to be served by larger banks. It is important to note that this second aim implies not only that we want community banks to be operated in a safe and sound manner, but that we want them to be able to operate successfully. Just as rural customers would be denied credit if their bank fails, they would also be denied credit if their bank's costs make it an unviable business proposition.

Supervision of community banks

Let me turn now to describing how our supervisory policies reflect these regulatory objectives. First, of course, the Federal Reserve has long organized supervision into four portfolios of banking organizations, based principally - though not exclusively - on asset size. Since the financial crisis and the passage of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act), supervisory policies for the largest firms have become increasingly more data-driven, horizontal, and centrally coordinated. The enhanced prudential standards applicable to the largest firms, such as supervisory stress tests and quantitative liquidity regulation, have increased the differences in supervisory practice among our four portfolios. We do not regard the supervision of community banks (or, for that matter, smaller regional banks) as a diluted version of large bank supervision, but as an entirely separate undertaking. We believe there is a distinct "state-of-the-art" approach to each of the four portfolios, including community banks.

Second, supervision within the community bank portfolio has not been centralized in the way that large firm supervision has been. On the contrary, with 837 state member community banks and several thousand community bank holding companies, it would make little sense to do so. While we strive for a well-considered overall supervisory program for our community bank portfolio, we will continue to rely on the examiners in the 12 Federal Reserve Banks to carry out that program in accordance with local conditions and based on the varying situations of community banks.

I believe this intention reflects the preference of most community bankers, who worry that more centralization of supervisory practice would lead to delays in getting answers to questions and decisions on applications. At the same time, members of the Board do, from time to time, hear complaints from community bankers that supervisory practice may not always be consistent across Reserve Banks. To some degree, of course, this is an inevitable byproduct of an approach that leans toward bottom-up supervision. Nonetheless, our staff at the Board will continue to assess practices across all the Reserve Bank districts to promote overall consistency without micromanaging the supervision of individual banks. And I want to emphasize that, if community bankers believe that significant variation in certain practices or policies is having an unfair or deleterious effect on one or more of their banks, they should feel more than comfortable raising these concerns with Board staff. We do not regard such observations as criticism of our examiners, but instead as opportunities to test, and possibly adjust, our policies to promote overall consistency.

A third way in which our community bank supervision differs from that of other banks is the particular effort made to limit the length of examinations and the amount of time bank personnel must spend on them. Community banks have much smaller balance sheets over which to amortize the resources they spend on regulatory and supervisory requirements. Of late, we have increased off-site supervisory activities, which can reduce the burden on community bank personnel. For example, we can conduct some aspects of the loan review process off site for banks that maintain electronic loan records. While off-site loan review has burden-reducing benefits for both bankers and examiners, some bankers have expressed concern that increasing off-site supervisory activities could potentially reduce the ability of banks to have face-to-face discussions with examiners regarding asset quality or risk-management issues. Accordingly, we remain flexible and will continue to work with community banks that may prefer their loan reviews to be conducted on site.

More generally, the Federal Reserve has invested substantial resources in developing technological tools for examiners to improve the efficiency of both off-site and on-site supervisory activities. These measures should lead to greater consistency and more efficient, effective, and risk-focused examinations by better enabling staff to tailor the scope of examinations to the activities and risks of individual banks. The automation of various parts of the community bank examination process can also save examiners and bankers time, as a bank can submit requested pre-examination information electronically rather than mailing paper copies to a Reserve Bank. Through these efforts, the Federal Reserve aims to strike an appropriate balance between off-site and on-site supervisory activities to ensure that community banks are subject to supervision that is both high-quality and resource-efficient.

Fourth, our supervisory oversight of community banks is itself grounded in the traditional relationship banking model of funding local lending with customer deposits. There are risks in this model, of course, since geographic and portfolio concentration can make community banks vulnerable to local economic downturns, as we have witnessed during and after the Great Recession. Even though the number of banks on the Federal Deposit Insurance Corporation's "Problem List" has fallen from a peak of 888 at the end of the first quarter of 2011 to 291 at the end of 2014, this number is still nearly three times the historical average. But these risks are understood by supervisors and bankers. Traditional regulatory and supervisory methods are well suited to deal with them. Indeed, it is noteworthy that most community banks adhering to this model continued to thrive even during the worst years of the financial crisis.³

In contrast, community banks that moved beyond their traditional business model and entered unfamiliar or more complex lines or markets experienced difficulties. For example, small banks that turned to a more transactional model and funded construction loans, often outside of their local markets, with borrowings rather than core deposits were more likely to fail. This experience helps explain why our supervisory intensity will often increase for banks, including community banks, that embark upon unfamiliar activities.

A similar pattern obtains in the Federal Reserve's supervision of consumer compliance in community banks. It is axiomatic that all financial consumers deserve the same protection, whether they are doing business with a very large bank, a regional bank, or a community bank. But community banks do not have large systems with hundreds or thousands of employees in contact with consumers on a daily basis. So, while the substantive goals of consumer protection remain the same across banks, the way in which we check compliance at community banks takes account, again, of their size, less complex structure, and business model. These characteristics of community banks informed the revisions to our risk-focused consumer compliance supervision framework, which was implemented in January 2014.⁴

Under the revised program, examination intensity is explicitly based on an individual community bank's risk profile, including its consumer compliance culture and how effectively it identifies and manages consumer compliance risk. Examiners do more comprehensive risk assessments before they arrive on site, permitting them to focus on areas of higher risk at individual banks. The program also calls for examiners to spend less time on low-risk compliance issues at community banks. Of course, in cases where a firm's risk profile is high or where it changes materially as a result of the addition of more complex or higher-risk strategies, more frequent contact may be appropriate.

Early feedback from bankers and examiners regarding the program has been largely positive. For example, bankers have noted that examiners seem to better understand their institutions, that the exam work undertaken better reflects the individual institutions being examined, and that the time spent by examiners at the banks has decreased. We will be conducting a review of the program this summer to determine if adjustments are needed.

Our examiners have observed increased consumer compliance risk where community banks move into products that are less congruent with bank-customer relationships in the traditional community bank business model, such as prepaid cards or credit card add-on products. Similarly, risks can arise where a bank relies upon third-party providers for core banking services or products that the bank does not have the resources or expertise to undertake in-house. Of course, when properly chosen and managed, the use of third-party providers is not only acceptable, but desirable for important services such as cybersecurity. But reliance upon third parties to broaden a bank's scope of operations can pose supervisory concerns when the bank itself does not have adequate capacity to vet and monitor these vendors.⁵ For example, a bank that relies upon a third party to provide a loan processing system or to conduct compliance audits may face increased risks because the bank no longer has direct control over these activities.

Simplifying the regulatory regime for community banks

In the wake of the financial crisis, the overall aim of the Dodd-Frank Act was to strengthen the regulation of banks and certain other parts of the financial system. A key regulatory innovation in Dodd-Frank was regulatory tiering - the creation of different classes of banking organizations, based dominantly though not exclusively on asset size, to which different regulations were to apply. Underlying this tiering was the principle that progressively more stringent regulation should apply to the different classes of banks based on their relative importance to the financial system, and thus the harm that could be expected to the system if they failed. As with supervision, an important consideration in regulatory tiering is to weigh the compliance cost of a particular rule against the safety and soundness benefits to be gained by that rule with respect to specific groups of banks. Of particular significance in performing this assessment for community banks is, as I mentioned earlier, the fact that their smaller scale means that they have a more limited base of activities across which to amortize certain fixed compliance costs.

There are two complementary ways to implement a tiered approach to prudential regulation. One is to apply specific regulations only to those classes of banking organizations whose activities and scale require those measures. The other is to tailor the application of generally applicable measures based on the size or other salient characteristics of banking organizations. When regulatory agencies adopt regulations under their general prudential authority, they usually have a choice as between these approaches. When the agencies are implementing statutory mandates, their discretion to exclude or tailor may be constrained in some respects.

In implementing the Dodd-Frank Act and in modifying our regulations under general prudential authorities, we have tried to include appropriate tiering in accordance with the considerations I mentioned a few moments ago. But over the last few years there emerged a fairly widespread view that some fine tuning of regulatory tiering was warranted. Let me now offer a few of my own thoughts on this point.

First, I want to mention how pleased we were that last December Congress amended the statutory provision that had prevented us from raising the threshold for application of our Small Bank Holding Company Policy Statement. I will not repeat here the purpose and history of this statement, which I described in advocating this change last fall.⁶ I will note that earlier this month, the Board issued a final rule implementing the statutory change.⁷ The rule expands the scope of application of the statement from bank holding companies with less than $500 million in total consolidated assets to those with less than $1 billion, and adds savings and loan holding companies. As a result, our statement now covers nearly 90 percent of bank holding companies. We had already taken steps to relieve the regulatory reporting burden for the impacted firms - specifically, by eliminating quarterly and more complex consolidated financial reporting requirements for all of these institutions and eliminating certain regulatory reporting requirements entirely for savings and loan holding companies with less than $500 million in assets.⁸

Second, I want to reiterate my view that a few Dodd-Frank Act provisions might usefully be amended to exclude community banks entirely from their coverage. As I have said before, the concerns addressed by provisions such as the Volcker rule and the incentive compensation requirements of section 956 are substantially greater at larger institutions. In the unusual case in which a small bank is engaged in proprietary trading that could pose a risk to the deposit insurance fund or has in place a compensation system that incentivizes excessive risk, the supervisory process would remain available to address these risks. I recognize that statutory revisions of this sort would not be a major reduction in compliance burden. But they would reflect the fact that, absent a change in this regard, community banks must expend scarce compliance resources to conform to the requirements of such regulations. There is, in my view, no need to make particularized prudential requirements of this sort applicable on a mandatory basis to thousands of community banks. Indeed, the Volcker rule and the Dodd-Frank Act incentive compensation provisions present almost prototypical cases in which minimal potential safety and soundness benefits are outweighed by the compliance costs faced by those thousands of banks. It would be preferable to relieve both supervisors and community banks from examining compliance with these kinds of requirements in order to concentrate resources on the real issues presently faced by these institutions, such as cybersecurity and interest rate risks.

Third, I want to address concerns that community bankers have expressed regarding the revised standardized capital requirements issued in 2013. In particular, we have heard that the framework's separate risk weight for high-volatility commercial real estate (HVCRE) loans is burdensome in part because it requires firms to classify which loans within their corporate loan portfolios are HVCRE using a complex definition that includes calculation of each such exposure's loan-to-value ratio. I should first say that, unlike proprietary trading or distorted compensation packages, HVCRE lending has posed real and substantial risks for community banks in recent years. These kinds of loans performed much worse in the recession than other commercial real estate loans and, in fact, were a significant factor behind many of the roughly 500 bank failures during this period.

So, we do need to take account of HVCRE risks, given that commercial real estate lending accounts for a significant percentage of the assets of many community banks. But we can look for ways to simplify the specific capital requirements while ensuring that appropriate capital buffers exist. For example, it might be possible to determine applicable capital requirements by reference to the proportion of a bank's real estate loans that are HVCRE, as defined more simply, rather than requiring more elaborate loan-by-loan calculations.

More generally, it is worthwhile for the bank regulatory agencies to consider possibilities for a broader simplification of capital requirements for smaller community banks. As important as it was to make the Basel I risk-based requirements more risk sensitive, it is also important to recognize that the greater detail of the new rules requires a degree of categorization, recordkeeping, and reporting that may be particularly costly, in relative terms, for smaller community banks. So we should explore whether we can achieve the safety and soundness purposes of capital regulation in a simplified way.

In doing so, of course, we must be mindful of the Collins Amendment, which requires that minimum leverage and risk-based capital requirements be established for all insured depository institutions and bank holding companies, and that the minimum applicable requirements not be less than "generally applicable" requirements.⁹ Still, there are some possibilities that may be consistent with the Collins Amendment. For example, one idea I have heard is to allow smaller community banks to opt into a simpler set of risk-weighted capital requirements in exchange for a higher minimum required ratio than under the more risk-sensitive, but more complicated, standardized risk-weighted requirements finalized in 2013. I believe the concept of a "simpler" set of requirements is meant to describe something much closer to Basel I in terms of the detail and number of risk categories. Because so many smaller community banks maintain capital levels well above minimum regulatory levels anyway, the tradeoff of higher requirements for a simpler approach may be promising

This is only one idea, and there may be other approaches that could simplify smaller community bank capital requirements in a manner consistent with the Collins Amendment. I look forward to discussions within the banking agencies, with the industry, and with other interested parties on possibilities for developing a sensible and feasible approach.

Conclusion

Having just become chair of the Federal Financial Institutions Examination Council, I hope to make the required decennial review under the Economic Growth and Regulatory Paperwork Reduction Act (EGRPRA) a productive one. A productive exercise in this context will, among other things, be one that results in changes in the regulations and supervisory practices of the banking agencies so as to yield significant reduction in compliance costs for community banks. It should not be a merely bureaucratic exercise in formal fulfillment of a statutory requirement. There are numerous issues beyond those I've mentioned already that almost surely can be profitably addressed, including reporting requirements and examination practices.

The chances of this outcome occurring will be maximized if regulators work to put ourselves in the place of community banks in order to understand in concrete ways the impact of particular regulatory and supervisory practices upon them and if, reciprocally, bankers understand the responsibility of regulators to protect the safety and soundness of banks and, thus, the deposit insurance fund. So, for example, although it is always useful to hear why regulated entities think a particular practice is unduly burdensome, it is even more useful to hear concrete suggestions for how that practice might be modified so as to reduce compliance burden while still achieving the ultimate regulatory objective.

Thank you very much.

---