The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
EPR FEDERAL RESERVE BANK OF NEW YORK ECONOMIC POLICY REVIEW Why Do Supervisors Rate Banking Organizations? James Bergin and Kevin Stiroh Volume 27, Number 3 November 2021 Why Do Supervisors Rate Banking Organizations? James Bergin and Kevin Stiroh OVERVIEW • The rationale for assigning ratings to banking organizations may seem straightforward, but the process involves considerably more complexity and nuance than many would recognize. • Ratings provide discipline to the regulatory process and a means for clear communication of supervisory assessments to firms, regulators, and other stakeholders. If done well, ratings facilitate remediation among supervised firms while bolstering confidence in both the supervisory process and the banking system. • The authors highlight three channels by which the assignment of a supervisory rating, in the context of its associated consequences, can influence the behavior of a financial firm: as a communication tool, as a direct risk mitigant, and as a broad incentive mechanism. • Understanding these channels, and their implications, is important for both developing effective supervisory assessments and considering the optimal design of a rating framework. Federal Reserve Bank of New York T his article addresses a question that at first may appear simple: Why do supervisors rate banking organizations? Supervisors of banking organizations periodically summarize their views of a banking organization into a confidential assessment or rating—essentially, a grade for the organization’s safety and soundness and compliance with law. To banking organizations and their supervisors, the assignment and use of supervisory ratings by their examiners is so familiar that it may seem pedestrian to pause to consider why it is done. At the same time, to outsiders to financial supervision, the process is another opaque aspect of a dimly understood craft. In this article, we try to shed some light on the practice in a way that will be illuminating for both insiders and outsiders. We believe that there is considerable complexity and nuance incorporated in the concept of rating banking organizations and that it warrants a rigorous discussion of why supervisors assign ratings and how they advance the statutory and regulatory goals of supervision. The recent Board of Governors and FDIC joint notice seeking comment on the depository institution rating system is an invitation to join this discussion.1 Looking at the historical record and current practices, we believe ratings have been an important tool for supervisors for two primary reasons. One, ratings add discipline to the supervisory and regulatory process. Supervisors and James Bergin is deputy general counsel in the Legal Group of the Federal Reserve Bank of New York. Kevin Stiroh is senior adviser in the Supervision and Regulation Division of the Federal Reserve Board. Emails: james.bergin@ny.frb.org, kevin.stiroh@ny.frb.org. Disclaimer: The views expressed in this article are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System.To view the authors’ disclosure statements, visit https://www.newyorkfed.org/ research/epr/2021/epr_2021_bank-ratings_bergin. Economic Policy Review 27, no. 3, November 2021 1 Why Do Supervisors Rate Banking Organizations? regulators are assigned significant power under law to make judgments about the soundness of private firms and to influence their activities. Ratings are an important means of ensuring that supervisors and regulators make these judgments rigorously and consistently across heterogeneous firms and over time. First developed as a sort of coordination mechanism, ratings have been further deployed over time by Congress and regulators to help the official sector make decisions with the best information possible. They now provide a meaningful link between supervisory assessments of the current condition of a firm and official judgments about such matters as what areas supervisors should focus on and whether a firm should be allowed to expand or offer new financial services. This adds to the legitimacy and fairness of the supervisory and regulatory process and can help build confidence in the financial system. Two, ratings facilitate communication of supervisory assessments to the many stakeholders of the examination process—directly to the supervised firms and other regulators, and indirectly to Congress and the public at large. Ratings allow supervisors to summarize and convey the complexity of a financial firm assessment to each of the stakeholders. Among other things, they enable supervised firms to anticipate the potential regulatory consequences of their supervisory condition. This awareness facilitates remediation and behavioral changes that ultimately support the underlying supervisory objectives, while also adding to the legitimacy of the process. Communication to Congress and the public is more attenuated because ratings are generally confidential by law, but the practice of ratings is an important element of generating reliable and consistent assessments internally within regulatory agencies, which allows these institutions to effectively communicate with public stakeholders.2 With these purposes in mind, we then contemplate a number of questions about the proper design of a rating framework. We examine how a supervisory rating actually influences firm behavior in a way that is consistent with supervisory objectives. We consider three channels of potential influence with respect to how ratings can affect regulatory judgments: (1) as a communication tool to raise awareness of the most salient risks and supervisory concerns; (2) as activity restrictions that act as a direct risk mitigant by reducing the specific behavior of concern; and (3) as activity restrictions that operate as a broad incentive mechanism to change behavior. Discussing ratings practices and renewing our understanding of their purpose and impact should provide clarity on the overall approach and hopefully inform the design of the optimal supervisory ratings framework. 1. The Practice and Origin of Ratings 1.1 The Practice of Ratings Supervision involves monitoring and overseeing financial firms to assess whether they are in compliance with applicable laws and regulations and whether they are engaged in unsafe and unsound practices. It involves visitorial powers and privileges that are not a feature of many industries, and the authority to instruct firms to correct problems.3 As Vice Chair Randal Quarles discussed in a recent speech, the banking industry is subject to a special form of government oversight because its crucial role in promoting economic growth means that it Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 2 Why Do Supervisors Rate Banking Organizations? warrants a government safety net that other industries do not. Supervision of financial firms goes beyond compliance with a rulebook. “[I]t isn’t enough to set the rules and walk away like Voltaire’s god. The potential consequences of disruption in the financial system are so far-reaching, and the erosion of market discipline resulting from the government safety net sufficiently material, that it is neither safe nor reasonable to rely entirely on after-the-fact enforcement to ensure regulatory compliance.”4 Supervision is a process that enables the government to respond to the idiosyncratic circumstances of individual firms, to order corrections if merited under law, and to otherwise guide firms toward prudent behavior. Financial supervision has a different character than ensuring compliance with pre-set regulation or a rulebook. A banking organization can comply with the letter of every relevant statute and regulation, and still act in an unsafe and unsound manner. Therefore, supervision and regulation are complementary tools to achieve the same broad objective of a banking system that provides critical financial services in a sustainable way. This judgmental role of discerning whether an organization’s practices pose undue risk is assigned to financial supervisors. To make their assessments, supervisors are granted the authority to go where they need to go and to direct firms to make changes if necessary.5 This type of governmental authority over private firms is not without its tensions, and places special focus on ensuring that supervisory assessments have as much rigor as possible and are as well-understood by the regulated firms as possible. To make their assessments, supervisors conduct a wide variety of activities requiring analysis of multiple inputs, by many people, with different skillsets, across time. For a large and complex firm, supervision may involve credit quality review of loan files; assessment of appropriate calibration of anti-money laundering filters; analysis of liquidity stress scenario modeling; evaluation of the credibility of resolution plans in a bankruptcy; and appraisal of managerial capability. Even for smaller and less complex firms, supervisory assessments require a wide range of inputs and perspectives. Supervisors generally sum their views into an overall rating at a certain point in the cycle, and then communicate this rating to the management of a firm.6 In this article, we will focus primarily on supervisory ratings of a firm’s prudential condition—its overall health and robustness to stress—but there are other regulatory ratings. This supervisory rating is typically a numeric or categorical distillation of a supervisor’s assessment of a firm, which serves as a shorthand expression of a supervisor’s view about the firm. Ratings can provide a “composite” view of a firm’s aggregate condition (for example, Bank ABC is “3” or “fair” overall), or they can focus on a particular component of a firm’s performance, such as capital or earnings (for example, the earnings of Bank XYZ are “3” or “fair”).7 Importantly, ratings are, for the most part, confidential. Because ratings reflect the assessment of supervisors, they are considered confidential supervisory information (“CSI”) that is owned by the regulatory agencies. As such, banking organizations are generally not permitted to share ratings information with other entities without the permission of the appropriate regulator and are even required to be careful about how much they share ratings information internally.8 The rationale for the confidentiality of ratings, as with other CSI, comes from the Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 3 Why Do Supervisors Rate Banking Organizations? plenary access that supervisors have to the information of supervised firms, and the necessarily subjective judgments that they are required to make about whether the risks they take are sound ones. Public disclosure of a funding or operational risk, for example, could exacerbate the risk. Moreover, CSI can contain proprietary information about a firm’s strategy and positions that would be inappropriate to share publicly.9 While commentators have suggested reform in this area, the confidentiality of ratings remains fundamental to the current practice of supervision, and has important implications for how ratings are used by regulators and firms and how the consequences of ratings are understood by the public.10 1.2 The Origin of CAMEL and BOPEC Assigning a numeric rating seems to be a practice of long vintage. Rating systems date back to at least 1926, when the Federal Reserve Bank of New York used a simple system to categorize the 900 state member banks then under its supervision. Other federal supervisors appear to have been using their own individualized rating scales by the 1930s, although enthusiasm for their use seems to have waxed and waned over time.11 Uniformity in ratings received its real boost in the 1970s when Congress and policymakers began to worry about the difficulty of effectively supervising increasingly complex bank holding companies. Bank failures had increased from the early to mid-1970s, and a number of the banks that failed had holding company affiliations that were considered to have contributed to their problems. Some observers pointed out that the Federal Reserve, in particular, had not identified issues that bank holding companies under their purview were causing those banks before they failed and was not coordinating integrated supervision of the firms. Further, some observers noted that banking regulators differed in how they determined which banks deserved special supervisory attention, and that these differences hampered their ability to coordinate with one another.12 It was felt that the lack of effective coordination would only grow worse “as the holding company movement spreads and as the banking industry becomes more sophisticated and complex.”13 In response, Congress passed the Financial Institutions Regulatory and Interest Rate Control Act (FIRA) in 1978. Among other things (such as increasing the Federal Reserve’s enforcement powers), FIRA formalized a new Federal Financial Institutions Examination Council (FFIEC) of the bank regulators, including at the time the Federal Reserve, the Office of the Comptroller of the Currency (OCC), the Federal Deposit Insurance Corporation (FDIC), and the National Credit Union Administration. Among other things, the FFIEC was charged with prescribing “uniform principles and standards for the Federal examination of financial institutions” by bank regulators and to “make recommendations to promote uniformity” in their supervision. The statute did not mandate the creation or use of ratings or any particular rating system as such, but it did call for FFIEC recommendations to identify firms “in need of special supervisory attention.”14 The new FFIEC adopted the Uniform Financial Institutions Rating System for depository institutions later in 1978. It quickly became known as the CAMEL rating, after its component parts (Capital, Asset quality, Management, Earnings, and Liquidity). It has been a durable instrument, despite modifications over time. John Heimann, the Comptroller of the Currency, Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 4 Why Do Supervisors Rate Banking Organizations? who served as the chair of the FFIEC as it developed the CAMEL rating, described two purposes for the rating system: “The rating system proposed by the Council has a two-fold purpose. First, it is designed to reflect in a comprehensive and uniform fashion an institution’s compliance with applicable laws and regulations, and its overall soundness. Second, the rating system the Council proposes is meant to assist the public and the Congress in assessing the aggregate strength and soundness of the financial system.”15 Later in the adopting interagency release, the first purpose was extrapolated and emphasized: Heimann said that the “primary purpose [emphasis added] of the uniform rating systems is to help identify those institutions whose financial, operating or compliance weaknesses require special supervisory attention and/or warrant a higher than normal degree of supervisory concern.”16 At the beginning, it appears that the focus was on consolidating information in order to make informed decisions about supervisory resources. To do this, supervisors must assess firms in a “comprehensive and uniform” way—taking advantage of all information to hand, and in a manner that is comparable across firms and across regulators. The primary purpose of this assessment is to direct scarce supervisory resources to the right ends, in accordance with FIRA’s goals. Second, there is a focus on communication—providing a means for disseminating these confidential assessments to those who need to know about them. In Heimann’s telling, at the outset of CAMEL, the external stakeholders who needed this confidential knowledge were Congress and the public at large. Congress had held hearings in the 1970s because of concerns about large bank failures, including Franklin National Bank in the Second Federal Reserve District, and whether regulators were up to the task of preventing them; so it makes sense that communication with external stakeholders was considered important.17 We can infer, by its absence, that communicating effectively to the supervised firms was less of a priority at the time. Importantly, CAMEL enabled communication within the supervisory community. The great benefit of CAMEL was its uniformity, so that the multiple regulators involved in overseeing a complex entity could communicate efficiently and coordinate their activities.18 The OCC’s adopting release emphasized that uniform ratings will enable identification of issues “in such a way that does not depend solely upon the nature of its charter…or the identity of its primary Federal regulator.”19 The adoption of the CAMEL system was followed closely by the Federal Reserve’s adoption of a rating system for bank holding companies—the “BOPEC” system (named for its component parts: Bank subsidiaries, Other nonbank subsidiaries, Parent company, Earnings, and Capital, along with F/M – Financial composite and management composite). What did the Federal Reserve think it was doing with these two rating systems? Testifying before the Senate banking committee shortly after the introduction of CAMEL and BOPEC, Governor Charles Partee sounded notes similar to Heimann’s and focused on bringing discipline to the supervisory process and directing resources to problem institutions. CAMEL “should help us identify more precisely those banks in need of particularly close supervisory attention” and BOPEC “standardized the evaluation of the financial condition of holding companies and has helped to identify those companies with significant financial problems.”20 Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 5 Why Do Supervisors Rate Banking Organizations? 2. The Evolution of Ratings and Their Purposes The rating systems for banks and bank holding companies have each been revised a number of times in response to changes in regulatory philosophy and in the industry, and regulators have also created entirely new rating systems. It is instructive to review what regulators—and here we will focus on the Federal Reserve—have articulated as the purpose of these rating systems each time they changed or adopted them. In some ways, one can see the evolution of rating systems as indicating changes in supervisory policy and priorities, even while the main goal of the prudential rating system has remained consistent over time. We focus particularly on the bank and bank holding company rating systems here, but there are many others. This review shows consistency with the underlying drivers of discipline and communication but also substantive variation that reflects the evolving environment. See Table 1 for an overview of rating system evolution over time. 2.1 CAMEL to CAMELS BOPEC and CAMEL were each modified over the course of 1995 and 1996. In 1995, the Federal Reserve required the assignment of a new rating for risk management at the bank and bank holding company level. This “reflect[ed] the view that properly managing risks has . . . become even more important as new technologies, product innovation, and the size and speed of financial transactions have changed the nature of banking markets.”21 In 1996, these factors led CAMEL to become CAMELS, as sensitivity to market risk was added as a sixth factor. One can see the evolution of rating systems as indicating changes in supervisory policy and priorities, even while the main goal of the prudential rating system has remained consistent over time. The interagency adopting release reflected on the original articulated purposes of CAMEL and largely affirmed them. Similar to the views of Heimann and Partee, it focused on three main attributes of ratings: that they enable “comprehensive and uniform” evaluations; that they direct supervisory resources to firms exhibiting “weaknesses or adverse trends;” and that they “assist Congress in following safety-and-soundness trends.”22 2.2 BOPEC to RFI/C(D) The next significant change to the bank holding company rating system came in 2004. The passage of the Gramm-Leach-Bliley Act in 1999 enabled bank holding companies to enter new areas of business and expand their footprints. The Board’s new system, RFI/C(D) to reflect Risk management, Financial condition, Impact on depositories/Composite (Depository institution rating), was intended to focus less on evaluation of legal entities and more on an Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 6 Why Do Supervisors Rate Banking Organizations? evaluation of a firm’s ability to manage risk and implement sound controls across business lines. RFI/C(D) was said to reflect a recognition of how the Federal Reserve’s supervisory processes had evolved over time, together with the changes in the banking industry.23 The adopting release for this new rating system made some changes to the Heimann/ Partee explanation of why the Federal Reserve rates firms. Ratings are a “management information and supervisory tool” that “serves three primary purposes in the supervisory process:” “First and foremost, the BHC rating provides a summary evaluation of the BHC’s condition for use by the supervisory community. Second, the BHC ratings form the basis of supervisory responses and actions. Third, the BHC rating system provides the basis for supervisors’ discussion of the firm’s condition with BHC management.”24 Discipline and communication continue to characterize these purposes, but there are shifts in emphasis. We see that ratings are intended to facilitate communication with the management of firms, rather than with Congress and the public. It seems obvious now that this should be a major purpose of ratings, but we do not see it in the adopting releases until 2004. In addition, there is a subtle re-characterization of the old “primary purpose,” moving from a focus on problem firms to a more neutral description of ratings forming the basis of supervisory responses and actions. This may reflect the increasing use of ratings as a statutory and regulatory decision-making criterion during this period, or the consequences of Gramm-Leach-Bliley. 2.3 RFI/C(D) to LFI/RFI The Board significantly changed the rating system in 2018 with the adoption of the new LFI (large financial institution) rating system. The rationale for this new rating system was explained not in terms of changes to industry structure, as with the adoption of RFI/C(D), but rather in terms of changes to supervisory practice. The Board determined that its supervisory program for systemically important firms had changed so materially since the financial crisis of 2008 and the Dodd-Frank Act of 2010 that the RFI/C(D) system was no longer useful. Instead, the LFI system evaluates capital, liquidity, and governance and controls on a firmwide basis, with much less specific emphasis on the bank. “The explanation of the purpose of the new rating system continued to evolve as well. The adopting release explains LFI’s intentions in three parts. It is designed to: [a]lign with the Federal Reserve’s current supervisory programs and practices; [e] nhance the clarity and consistency of supervisory assessments and communications of supervisory findings and implications; and [p]rovide transparency related to the supervisory consequences of a given rating.”25 With the LFI ratings, the communication theme from the 2004 release evolves further with an emphasis on enabling clarity not only of supervisory messages, but also their implications and consequences. Vice Chair Quarles observed:26 Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 7 Why Do Supervisors Rate Banking Organizations? “Ratings are an essential vehicle for supervisory feedback—a clear, concise way to convey whether a firm meets expectations, with tangible, predictable consequences for those that fall short. Our ratings system for large institutions had remained unchanged since 2004, even as our supervision of those institutions evolved significantly after the crisis. The new rating system will better align ratings for these firms with the supervisory feedback they receive, and will focus firms on the capital, liquidity, and governance issues most likely to affect safety and soundness.” The release goes on to discuss explicitly what effect a given rating might have on the Federal Reserve’s posture toward an enforcement action against, or an application by, a given firm. And again, transparency to the public is not part of the articulated justification. The rationale for the LFI rating system also, and importantly, has a more backward-looking component, namely, the alignment with current supervisory practices for the largest firms. Following the financial crisis, the Federal Reserve introduced its revised framework for large bank supervision.27 This framework focused on capital, liquidity, and governance, and the LFI rating system followed along. In addition, the new framework took a much more macroprudential perspective with consideration given to both the probability of distress at a large firm and the potential impact of the distress on the broader financial sector and economy.28 Table 1 Evolution of Certain U.S. Bank and Bank Holding Company Rating Systems Rating system Period Notes Various Pre-1978 A number of systems seem to have been employed by different regulators, at least informally, for many years. A simple system was in use at the New York Fed since at least 1926. CAMEL 1978–1996 Established by the Federal Financial Institutions Examination Council. Formally the Uniform Financial Institutions Rating System, it quickly became known as the CAMEL rating, after its component parts: Capital, Asset quality, Management, Earnings, and Liquidity. BOPEC 1978–2004 The Federal Reserve’s standardized rating system for bank holding companies, named for its component parts: Bank subsidiaries, Other nonbank subsidiaries, Parent company, Earnings, and Capital (along with F/M – Financial composite and management composite). CAMELS 1996–Present CAMEL system expanded to include a sixth factor: Sensitivity to market risk. RFI/C(D) 2004–2018 With bank holding companies entering new business areas, the Fed adopted RFI/C(D) to gauge Risk management, Financial condition, Impact on depositories/Composite (Depository institution rating). The new system put less focus on the evaluation of legal entities and more on an evaluation of a firm’s ability to manage risk and implement sound controls across business lines. LFI 2018–Present The Fed determined that its supervisory program for systemically important firms had changed so materially since the financial crisis of 2008 and the Dodd-Frank Act of 2010 that the RFI/C(D) system was no longer useful. Instead, the LFI (large financial institution) system evaluates capital, liquidity, and governance and controls on a firmwide basis, with much less specific emphasis on the bank. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 8 Why Do Supervisors Rate Banking Organizations? 2.4 Other Rating Systems This article focuses on prudential ratings, particularly those that relate to banks and bank holding companies, but there are other regulatory rating systems. In some ways, one can track a path through evolving regulatory priorities by their reflection in existing and new rating systems. A recent example is the rating system developed by the Federal Reserve that applies to nonbank entities that have come under federal financial supervision, such as financial market utilities. This responds to a post–Dodd-Frank Act focus on overseeing the systemic risk that exists within clearing and settlement systems.29 A different example is the role that Congress asked financial regulators to play in assessing the success of banking institutions in meeting the credit needs of their local communities. Under the Community Reinvestment Act, first passed in 1977 and significantly revised in 1989, supervisors are required to assess whether certain banking institutions are “meeting the credit needs of its entire community, including low- and moderate income neighborhoods.”30 These ratings are required to be taken into account when assessing certain applications for regulatory approval from such institutions. In significant contrast to prudential ratings, these ratings are required by statute to be made public. The public availability of these ratings is very much the point: Congress wanted examiners to make public judgments on which depositories were doing better than others in enabling access to credit. Many papers have been published on whether the incentive scheme produced by this approach is helpful or harmful.31 For this article’s purpose, we will just observe that the incentive scheme is different than that for prudential ratings. 3. How Are Ratings Used? Ratings are a powerful tool for making sure that supervisors are disciplined in their assessments and communicate those assessments effectively. The communication of these assessments then enables the supervised firms to anticipate the probable consequences. Firms can then take informed decisions and appropriate actions—whether it be to invest to address the issues, to exit certain activities, or to avoid actions that would exacerbate the issues. This section examines how ratings affect regulatory and supervisory judgments to constrain financial institutions. From the banking industry’s perspective, ratings are a predictor of how tightly the regulators will constrain their activities. While organic growth is usually possible even without permission from a regulator, firms often find it onerous to be prevented from expanding by acquisition or into a new activity or business line when they might consider it strategically desirable. From a supervisor’s perspective, these constraints can serve a variety of purposes. Sometimes, they are a means of directing a firm’s attention to material weaknesses. Sometimes, the constraints are required by underlying statutes to be imposed when an institution’s condition degrades – as measured by a downward adjustment in the rating. And sometimes supervisors may use ratings as a convenient and intelligent way to make a hard regulatory decision based on the best information to hand. Again, we will focus primarily on Federal Reserve requirements, and we will look at three broad types of the potential consequences of ratings: to inform supervisory prioritization and Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 9 Why Do Supervisors Rate Banking Organizations? expectations; to assess the use of enforcement tools; and to inform regulatory decisions such as those related to enforcement or permissible activities. 3.1 Supervisory Implications From the beginning, regulators have stated that ratings are a tool for directing scarce supervisory resources to the right ends. Ratings, in other words, are a means for making robust decisions about the amount and type of supervisory attention to devote to a specific firm based on its risk profile and financial condition. This is described in the ratings definitions themselves. For example, the RFI rating system states that firms rated “fair” are “vulnerable and require more than normal supervisory attention and financial surveillance” and that firms rated “marginal” “require close supervisory attentional and substantially increased surveillance.”32 How this process of devoting more supervisory resources to weaker firms actually operates can be a little opaque in public materials, although presumably it informs prioritization decisions. The recent release asking for comment on the use of CAMELS states that “the agencies increase supervisory activities, which may include targeted examinations between regularly scheduled examinations, if an institution’s CAMELS ratings are less than satisfactory.”33 A public example of supervisory prioritization can be found in a recently revised provision of Regulation H, required by a recent regulatory reform law, which provides that state member banks with assets of $3 billion or less will be examined on an eighteen-month cycle, rather than a twelve-month cycle, as long as they have a CAMELS rating of at least “1” or “2.”34 Stronger firms with higher ratings presumably receive less supervisory attention—although one could reasonably question just how much less, especially given the increasing focus on regular horizontal exercises in large bank supervision, such as stress testing and resolution planning. How ratings affect enforcement decisions is not hard-wired, but ratings are understood to play a role in the supervisory judgment as to whether an action is merited. Related to prioritization decisions is the use of ratings to inform the expectations that supervisors will have for firms. A public example from guidance is the BHC Supervision Manual’s discussion of funding expectations, which says that BHCs with less-than-satisfactory ratings should be asked to prepare specific action plans for reducing short-term obligations without undermining their affiliated banks.35 Similarly, an insured bank is required, by regulation, to have an audit at the bank level, rather than the holding company level, when its rating falls below a 1 or a 2.36 3.2 Enforcement Tools Ratings may play a role in supervisors’ decision making about when to employ enforcement tools with respect to individual firms, ranging from informal nonpublic actions such as Board Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 10 Why Do Supervisors Rate Banking Organizations? resolutions and memorandums of understanding (MOUs) to formal and public actions such as written agreements and cease and desist orders. How ratings affect enforcement decisions is not hard-wired, but ratings are understood to play a role in the supervisory judgment as to whether an action is merited. The recent LFI rating is relatively forward leaning in discussing how ratings are tied to enforcement tools. It says that there is a “strong presumption” that a firm with a “Deficient-1” rating will be subject to an informal or formal enforcement action and that a firm with a “Deficient-2” rating will be subject to a formal enforcement action. The recent CAMELS release states that “composite and component ratings…are significant indicators of the need for heightened supervisory attention including enforcement actions for more problematic issues.”37 Bank regulators’ statutory authority to order banks to correct unsafe and unsound practices creates a presumption that an insured depository institution may be deemed to be engaging in an “unsafe and unsound” practice if “in its most recent report of examination, [it received] a less-than-satisfactory rating for asset quality, management, earnings, or liquidity.”38 Other official statements do not directly connect enforcement decisions with rating decisions, although the circumstances that lead to a poor rating may be similar to those that lead to an enforcement decision. For example, the FFIEC BSA examination manual provides guidance on what level of program breakdown will lead an examining agency to take an enforcement action, but does so without referring to the rating process.39 Less-than-satisfactory ratings also allow regulators to use prompt corrective action to require conservation measures by adequately or undercapitalized state member banks. If the Board has determined, after notice and opportunity for hearing, that in the most recent examination of the bank, the bank received and has not corrected a less-than-satisfactory rating in certain CAMELS categories, then the Board can impose requirements such as capital distribution limits or limits on growth.40 3.3 Regulatory Judgments In certain cases, ratings inform regulatory judgments, and in other cases, the rating itself has direct regulatory consequences by operation of statute or regulation. Although the distinction with the supervisory judgments category described above can be a little blurry at times, we think it is useful to describe regulatory judgments separately. Permissibility decisions Ratings are crucial determinants of regulatory judgments about when supervised firms are entitled to the privilege to acquire, hold, and retain subsidiaries that engage in certain activities. A prominent example is the provisions of Section 4 of the Bank Holding Company Act added by the Gramm-Leach-Bliley Act, which permits financial holding companies (FHCs) to engage in certain nonbanking activities as long as, among other things, those firms are well capitalized and well managed.41 “Well managed” is defined by the statute (and further implemented by regulation) with specific reference to ratings assigned by regulators. The consequences for failing to remain well managed mean that financial holding companies, Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 11 Why Do Supervisors Rate Banking Organizations? among other things, have to enter into an agreement—commonly referred to as a 4(m) agreement—with the Board of Governors and lose the ability to make new investments in FHC permissible activities without Board approval. This is a matter of consequence for firms and may be one of the least desirable aspects of a poor rating from their perspective.42 Expansion decisions Ratings inform regulatory judgments about when firms may expand or reorganize their activities. A number of regulations use ratings to assess how much expansion a firm can undertake without seeking specific regulatory approval. For example, Regulation H, which governs the activities of state member banks, uses management ratings to determine both how much banks may invest in their own premises and whether they may make public welfare investments, without specific approval.43 Ratings are also frequently used as a decisional factor when a formal regulatory application approval is sought. For example, Section 3 (and Regulation Y) of the Bank Holding Company Act requires an application to the Federal Reserve when a bank holding company seeks to acquire a bank, and the Federal Reserve is then required to consider (among other things) the financial and managerial resources of the applicant, its CRA performance, and its record of Bank Secrecy Act compliance. The Federal Reserve considers ratings in making each of these decisions, including ratings assigned by the regulator of subsidiary depository institutions.44 In 2014, the Federal Reserve issued a guidance letter, SR 14-2, to “enhance transparency in the . . . applications process and provide . . . better insight into the issues that could prevent the Federal Reserve from acting favorably on a proposal.” This letter placed strong emphasis on ratings in making these decisions. It established a high bar for favorable consideration of applications from organizations “that are rated less than satisfactory.” It stated that the Federal Reserve will consider applications from organizations with one or more component ratings of “3” or a composite rating of “3” only in “very limited circumstances” and where they could demonstrate, among other things, that the acquisition would strengthen the organization. Consumer compliance and CRA ratings are similarly given great weight in this letter.45 Notably, the industry has argued that the relatively hard boundaries in this letter have worked to make supervisory decisions about ratings practically binding and rule-like in their consequences for the supervised firms.46 Special restrictions for problem firms Consistent with the original Heimann “primary purpose,” ratings are used to inform regulatory judgments about special restrictions that firms should be subject to when they are in weak condition. For example, Regulation O, which governs loans to insiders, provides for a lower loan limit when the firm is not in a satisfactory condition.47 Less-than-satisfactory ratings lead to a number of restrictions meant to protect the FDIC’s Deposit Insurance Fund (DIF). The DIF’s assessment levied upon bank deposits uses a bank’s supervisory rating as part of its calculation methodology—the worse your rating, the more you pay in.48 Discount window access can also be restricted by a bank’s rating. A Federal Reserve Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 12 Why Do Supervisors Rate Banking Organizations? Bank faces certain statutory restrictions in lending to an “undercapitalized institution,” which includes, among other things, an institution that received a composite “5” rating or its equivalent.49 Not by statute, but by policy, Federal Reserve Banks use ratings to discern which institutions are eligible for primary and secondary credit.50 Firms with insured depository institutions (DIs) that are in “troubled” or worse condition,51 or that have a composite CAMELS rating of 4 or 5, are restricted, by statute and implementing regulation, from making (or agreeing to make) “golden parachute payments” to institution-affiliated parties.52 Firms with insured DIs that are in troubled or worse condition (or meeting certain other conditions) are required, by statute and implementing regulation, to submit prior notice of changes in their senior executive officers and directors to the Board of Governors, and the Board can disapprove of those appointments.53 3.4 Growth of Consequences of Ratings over Time It has been observed that the consequences of a rating seem to have grown since CAMEL was first formalized.54 This seems to be true and represents a mix of choices by Congress and by regulators to make use of this assessment tool to more closely link constraints to the current condition of a firm. Since the formalization of the ratings process in the late 1970s, Congress has explicitly made use of the availability of rating systems a number of times as a means to determine which financial firms should be subject to certain limitations, penalties, incentives, and other consequences because their condition had either improved or declined. These include a number of the measures discussed above. For example, in 1991, Congress amended the Federal Reserve Act to include the limitations about advances to “undercapitalized” state member banks, and it defined “undercapitalized” by reference to the composite CAMEL rating of 5.55 In 1994, Congress streamlined audit requirements for highly rated banks, as discussed above.56 In 1996, Congress amended the Bank Holding Company Act to provide for expedited processing of nonbanking proposals by bank holding companies and subsidiary depository institutions that are “well managed,”57 and defined the term by linking it to ratings. In that same 1996 legislation, Congress also amended the Federal Reserve Act to eliminate the requirement that a state member bank seek prior approval before making an investment in bank premises if it meets certain criteria, one of which is a CAMEL composite rating of 1 or 2.58 The Gramm-Leach-Bliley and the Dodd-Frank Act each saw Congress add the statutory provisions under which a firm cannot engage in a number of activities unless the holding company and its depository institutions are “well managed.”59 Regulators also made use of ratings as a mechanism for guiding and constraining their discretion to make certain decisions with respect to a number of regulatory consequences. For example, in 1990, the Board revised Regulation Y to implement the statutory requirements of the Financial Institutions Reform, Recovery, and Enforcement Act (FIRREA), to require firms in “troubled condition” to provide prior notice of director and senior executive changes.60 It did so by defining “troubled condition” with reference to the most recent exam rating.61 In 1992, when the Board amended Regulation H to implement the prompt corrective action framework under the Federal Deposit Insurance Corporation Improvement Act of 1991, the Board defined “unsafe or unsound practice” to mean that the agency has determined, after notice and opportunity for hearing, that in Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 13 Why Do Supervisors Rate Banking Organizations? the most recent exam, the firm received and has not corrected a “less-than-satisfactory rating” for any of the CAMELS component ratings.62 In 1994, the Board revised Regulation O to include the higher loan limit for state member banks with deposits of less than $100 million if the bank, among other things, receives a satisfactory composite rating in its most recent exam. Outside of rulemaking, regulators also employed ratings in guidance documents in order to provide guidance to firms regarding how they will evaluate certain judgments.63 The increased interest of administrative law scholars in the process of bank supervision over the past several years has increased scrutiny of the link between ratings and their consequences, and how comfortably that links sits within administrative law. What exactly is a supervisory rating for purposes of administrative law? It represents the considered judgment of the official sector overseer and it has consequences, whether they be direct and immediate— such as changing deposit insurance assessments—or indirect and subject to further judgment—such as what type of new supervisory intensity is appropriate. We do not take a position on this question in this article, but note that it is an area of increasing inquiry.64 4. How Do Ratings Influence Outcomes? The ultimate goal of supervision is to influence the behavior of supervised financial firms so that outcomes are consistent with statutory and regulatory objectives such as the safety and soundness of a particular depository institution, the efficient and sustained provision of financial services to the real economy, consumer protection, and the stability of the financial system as a whole. An obvious next question is how, precisely, does the use of supervisory ratings within the regulators’ toolkit influence firm behavior to achieve those goals? Again, this might seem to be a simple question, but we think it is important to be clear in order to design the most effective ratings regime. As we describe above, a rating is a supervisor’s assessment, using all available information and judgment, of the current condition of a firm. In all cases, the rating itself is the same—a summary statistic that reflects a wide range of perspectives about the safety and soundness of a financial firm. The supervisor’s assessment is then used, in either a mandatory or permissive way, to make statutory and regulatory judgments about the constraints that a firm should be subjected to or relieved from. These constraints, naturally, inform the incentive structure for the firm. We identify three distinct channels by which a supervisory rating can influence the behavior of a financial firm: as a communication tool, as a direct risk mitigant, and as a broad incentive mechanism. It is helpful to distinguish between these channels so that we can employ a supervisory rating in the most effective way and design the rating framework accordingly. We note that this framework is relevant for both microprudential and macroprudential objectives. That is, all three channels could influence how firms respond to concerns about idiosyncratic behaviors and broader macroprudential ones. See Table 2 for a summary of these channels. 4.1 Channels of Influence Supervisory ratings can influence behavior as a communication tool to inform senior managers and boards of directors about supervisory concerns. This raises awareness of the most salient issues Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 14 Why Do Supervisors Rate Banking Organizations? and helps both financial firms and supervisors allocate resources and focus in order to address the most material risks. Supervisors and firms have complex and ongoing interactions, and the rating distills and focuses the supervisory message so that concerns can be properly understood. Challenges to communication can be especially hard at large, complex firms. In 2017, the Board proposed guidance to clarify the role of directors of large financial institutions, including how supervisors should communicate with them.65 Large financial firms are complex and ratings offer a simple, clear signal of supervisory concerns that can inform managers and directors of their concerns. The LFI rating system is a notable step forward in terms of communication because it allows more focused discussions with bank management about supervisory assessments of the most critical supervisory findings related to key areas such as capital, liquidity, and governance and controls. Boards can and do take actions to change management based on supervisory messages. It is also an effective tool for senior management to prioritize and communicate remediation efforts. As discussed above, communication to Congress and the public seems to have been de-emphasized as a stated purpose of ratings over the past forty years, while there is an increased focus on communication with the firm. For good reasons, supervisory ratings— again, with the important exception of CRA ratings—are not made public and are considered one of the most sensitive assessments of a financial firm. As such, the communication channel in its current form is primarily about influencing firms directly through interaction with senior management and boards of directors, rather than the public at large or Congress. A second channel for how a supervisory rating influences behavioral outcomes for a financial institution is as a direct risk mitigant—that is, by linking the rating, categorically or persuasively, to consequences that constrain the specific activity that creates the risk or fails to support supervisory objectives. In this view, supervisors identify a material risk that leads to a supervisory rating, and this identification leads to direct constraints on the underlying, and undesired, behavior. The direct constraints could take the form, for example, of supervisory prioritization decisions, enforcement tools, or regulatory judgments. A firm with identified weaknesses in governance and controls that create a material risk may receive a less-than-satisfactory rating in that area. The associated rating of the deficient area, and the supervisory judgment that underlies it, could lead to consequences that mitigate Table 2 Channels of Influence: Three Functions of Supervisory Ratings Function Action Goal Communication tool Supervisors use ratings to directly express concerns to a financial firm’s senior managers and board of directors. Clarify the most salient issues and the level of supervisory concern; help both the firm and its supervisors allocate resources to address the most material risks. Direct risk mitigant In their rating assessment, supervisors iden- Mitigate the harms caused by the identified tify a material risk that leads to constraints deficiency, supporting regulators’ prudenon the firm’s undesired behavior. tial or consumer protection mandate. Broad incentive mechanism A weak supervisory rating triggers a wider Provide management with a broader set of restrictions that aren’t directly linked incentive to fix the firm’s problems and to the underlying supervisory finding or induce changes in behavior in a range of risk-generating behavior. areas. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 15 Why Do Supervisors Rate Banking Organizations? the harms caused by the deficiency. For example, it could occasion an increase in supervisory intensity in that area, a potential enforcement action that restrains expansion of related activities, or the firm could be prevented from expanding into certain new activities related to that activity. Imposing one or more of these constraints reduces the chance that the weak control environment would affect a related set of activities that undermine safety and soundness objectives. Similarly, a firm with consumer compliance issues might receive a low supervisory rating and be restricted from expansion via new branches. Again, this constraint directly reflects the risk that would be imposed on consumers because of the weakness in this area and supports the regulator’s consumer mandate. A third channel for how a supervisory rating can influence firm behavior is as a broad incentive mechanism. In this view, a weak supervisory rating and the resultant activity restrictions impose a general cost on the firm, but the restriction and cost need not be directly linked to the underlying supervisory finding or risk-generating behavior. In this framing, the purpose of the consequences of the rating regime is to provide management with a broader incentive to fix its problems. For example, a 4(m) agreement might limit a wide set of expansionary activities, including activities that are not directly related to the negative supervisory assessment. This can constrain expansionary activities that are executed well or even those that are potentially risk-reducing via increased revenue streams or diversification because the firm is not considered “well managed” as a whole. By contrast, the successful remediation of a supervisor-identified risk and the resultant rating upgrade may relax a constraint and allow the firm to expand. The recent LFI rating system requires that a firm meet expectations in all three pillars to meet the “well managed” standard. To the extent that low-rated firms face more activity constraints that firms would like to avoid, this approach imposes a strong incentive for firms to change behavior in a way that is consistent with supervisory objectives, irrespective of the underlying issue. While not stated explicitly, this appears to be most consistent with the broader incentive mechanism approach rather than a direct risk mitigant approach. The key difference with the risk mitigation view is that a supervisory rating and resultant restrictions may act as a sufficiently punitive constraint to induce changes in behaviors in other, unrelated areas. As discussed below, this distinction has implications for optimal policy design. 4.2 Design Implications Understanding how supervisory ratings and their consequences actually influence firm behavior is important because these different channels have different implications for the optimal design of the rating framework. The communication tool approach, for example, has implications for the disclosure of ratings. As mentioned above, communication of supervisory ratings is now aimed primarily at firm managers and boards. If the goal is to send a powerful signal about supervisory views, however, public disclosure, at least on an aggregate level, could amplify that clarity and impose additional market discipline to promote the desired behavioral change. This is the judgment that Congress made in the context of CRA ratings, for instance. Effective communication can also provide additional transparency and help build the legitimacy of the supervisory process. The Federal Reserve has made a concerted effort recently to Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 16 Why Do Supervisors Rate Banking Organizations? increase transparency of the supervisory process. Since November 2018, the Board of Governors has begun to produce a semi-annual Report on Supervision and Regulation designed to increase transparency to the public and increase the legitimacy of the supervisory program.66 These reports now notably include aggregate information on ratings. This report does not reveal firm-specific assessments, so it does not directly promote market discipline, but it does increase transparency and provide an industry-wide view on the issues that supervisors deem to be most important. This can serve as an effective discipline on the supervisory process itself, and thus helps build the legitimacy, trust, and effectiveness of the supervisory process. Public disclosure, of course, is not a panacea, and one must consider a wide range of potential implications. Any disclosure regime needs to be robust to a wide range of possible outcomes. For example, would supervisors be comfortable disclosing that a supervisory rating reflected underlying issues with cybersecurity controls or difficulty meeting its funding requirements? That type of disclosure could be counterproductive and even induce exactly the wrong type of response from counterparties, clients, or bad actors. Aggregate reporting of rating information on a periodic basis, as is currently done, may be the right balance to strike. The agencies could also consider disclosing individual rating information with an appropriately long lag, although that raises different issues. Returning to the consequences of ratings, under the direct risk mitigant view, any restriction imposed as a consequence of that rating should be tied directly to the underlying risk. This approach likely has the most beneficial impact for customers or consumers because the consequence directly stops, or prevents the expansion of, the harmful behavior. One implication, however, is that the ratings framework would need to clearly identify each behavior and potential supervisory concern. This might create complexity that could work to undermine other objectives, such as clear communication. It also requires a high degree of precision in terms of both issue identification and remediation. One can also consider the incentive structures that are created for a firm if they perceive certain regulatory consequences to be more or less onerous than others. For example, if the consequence of poor capital planning is a regulatory halt on dividends, and the consequence of poor BSA/AML controls is the payment of a fine, then, at least in theory, a firm may decide that it cares more about the former than the latter. In a world of scarce resources, will this create incentives for the firm to focus resources on capital and not BSA/AML? Would a supervisor want that outcome? By contrast, the broad incentive mechanism implies a less direct link to the underlying behavior. That is, a regime could be designed where a low supervisory rating restricts a range of activities, not just those that create the fundamental concern that drives the rating. This creates a potentially broader set of options for the design of the consequence framework because the most powerful incentives can be utilized across a full set of risk concerns. As a hypothetical example, if firms find restrictions on capital distributions to be a particularly painful constraint, then that restriction could be used as a penalty for any supervisory issue, not just those related to capital adequacy or the capital planning process. Similarly, if poorly rated firms find it burdensome to seek prior approval when replacing senior executives, there is a powerful incentive for them to change behavior, exit that rating category, and have the constraint lifted. As a more positive example, the benefits that highly rated firms get from the expedited processes for applications (for example, in domestic branching or international banking) may serve as a positive inducement to invest broadly in remediating supervisory concerns. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 17 Why Do Supervisors Rate Banking Organizations? To be clear, this incentive mechanism does not free supervisors from the need to be clear about the underlying issues and expectations for remediation, but it does sever the direct link between the concern and the consequence. This perspective also severs the direct link with harmed parties if the impact of the incentive is miscalculated and the underlying behaviors continue. An advantage is that supervisors may want firms to build up their resiliency and risk management across the entire firm because supervisors (and bank management) do not always know where a breakdown in controls will manifest next. This overall risk management perspective would argue for a broader incentive-based framework, rather than the direct risk mitigant approach. This distinction between the direct risk mitigant approach and the broad incentive mechanism approach also has implications for how supervisors conduct their processes. In a world where the information content of ratings degrades quickly and ratings are slow to adjust, it may be less appropriate to impose broad penalties.67 This is particularly true when penalties are nondiscretionary. It also suggests that supervisors need to ensure that their rating judgments have not grown stale when they are acting as a particularly powerful, broad-based constraint, by refreshing examination work on a regular basis. It seems appropriate that firms facing broad, disruptive penalties should have the expectation that the penalty be removed quickly once the underlying issue is remediated and validated by supervisors. The consequences suggest that supervisors have an obligation to ensure the rating systems are working as intended.68 A final observation is that if ratings have real consequences for financial firms, then supervisors should actively survey that package of consequences to ensure they are consistent with the underlying policy objectives. The recent CAMELS release calls for public feedback on how ratings are used in considering applications and enforcement actions, which seems like a healthy step. It appears, for example, that the consequences of ratings have increased since the CAMEL/BOPEC/LFI framework was first implemented, but it is not clear whether the link between ratings and consequences has been revisited in a comprehensive way in order to understand the whole package of incentives facing the supervised firms. A core part of that assessment is to be clear about the desired channel and to understand how information and constraints flow through to affect firm behavior. Supervisors face a trade-off between making ratings uniform and consistent versus recognizing the broad range of idiosyncrasies that both amplify and mitigate risk. Finally, this observation introduces the familiar policy question of rules versus discretion. In this context, should the supervisory consequences from a particular rating be fixed and hard-wired ex ante or judgmental and left up to the discretion of the supervisor ex post? Both approaches feature in different places in our current statutes and regulations. Either is a valid approach for any of the channels of influence described in the prior section and the choice reflects additional considerations linked to uncertainty, transparency, and fairness. A supervisory rating, for example, is a summary statistic that will be an imperfect indicator of supervisors’ assessments. In developing the approach, supervisors face a trade-off between making ratings uniform and consistent versus recognizing the broad range of idiosyncrasies Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 18 Why Do Supervisors Rate Banking Organizations? that both amplify and mitigate risk. If ratings are too blunt and fail to capture those firm-specific differences, one might prefer additional discretion when it comes to imposing consequences. That discretion, however, reduces transparency and potentially calls into question the fairness and legitimacy of the entire supervisory process. 5. Conclusions Supervisory ratings are a core part of the current and historical practice of bank supervision. While the emphasis and stated goals have evolved over time, the underlying rationale for their use has remained centered on providing discipline to the supervisory process and communicating complex assessments to a broad range of stakeholders. These are constructive goals that help build the trust in and legitimacy of the supervisory process and confidence in the banking system. Understanding current views on the purpose and use of ratings is a necessary step in designing the most effective framework. Supervised firms need to understand the framework so they can respond appropriately to supervisory expectations and supervisors need to be clear on the framework to develop and maintain an internally consistent and efficient approach to oversight. This requires assessing complex topics such as the precise channel of influence, communication and transparency goals, and the need to design and implement a framework that is transparent and fair in a world with underlying uncertainty and vast heterogeneity across firms. This article contributes to that assessment, but there is surely more work to be done. To facilitate further work, we conclude by raising several open issues that warrant further investigation. It appears that the statutory and regulatory consequences that are linked to ratings have grown over time. Is this an intended or unintended consequence? One view is that if the tool is a good one, it should be used more frequently when important decisions are to be made. Moreover, linking consequences to ratings promotes the communication channel, enabling firms to anticipate what will happen if their underlying condition declines. On the other hand, this places considerable weight on the tool, perhaps more pressure than the tool was designed to withstand. One can also question how a confidential assessment should be linked to public consequences. For example, one sees this tension when banks seem to choose not to expand, even when there is no information in the public domain that would seem to prevent them from doing so. As a related point, it is useful to recall Goodhart’s law. This is the idea that if a measure becomes a target, it ceases to be a good measure, for the simple reason that those subject to the target will be motivated to take actions that undermine the measure’s usefulness.69 If firms care too much about how they are rated, and if they learn too much about how the ratings are determined, it is reasonable to ask whether they will take steps to “game” their rating without actually improving their safety and soundness or compliance with law. This could be a problem, although there are a number of reasonable mitigants, such as the judgment of trained supervisors, the breadth of what is being assessed, and confidentiality. Understanding the strengths and limitations of such mitigants seems like a fruitful topic for further inquiry. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 19 Why Do Supervisors Rate Banking Organizations? Supervisory ratings represent a confidential judgment based on confidential information. What can be learned by a comparison with other assessment models? For example, the ratings produced by credit rating agencies are mostly public judgments based on mostly public information. As another example, CRA ratings are public judgments. A more formal comparison of the pros and cons of these different assessment models would likely inform our understanding of the supervisory approach. Finally, the focus on stress testing following the financial crisis of 2008–09 marked a fundamental shift for supervision. Capital stress testing incorporates a supervisory viewpoint into an assessment of a banking organization’s balance-sheet capacity to withstand a range of severe but plausible shocks. An essential element of stress testing is that it involves a substantial amount of disclosure to the public, even as the precise nature of the disclosure has evolved over the years. It has involved a public consequence as well, which raises questions about the information content of the public stress test results relative to the confidential supervisory ratings. This involves many of the trade-offs mentioned earlier and raises further questions about the consistency of the disclosure regimes for different supervisory tools. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 20 Why Do Supervisors Rate Banking Organizations? Notes Acknowledgements: The authors thank Alexandra Rubin, Angela Sun, and Won Chai for their assistance with research for this article. In addition, the authors appreciate the review and comments of a number of helpful colleagues, including João Santos, Dianne Dobbeck, Anna Kovner, Bev Hirtle, Nicola Cetorelli, Michael Hsu, Benjamin McDonough, Lauren Hargraves, William Brodows, HaeRan Kim, Rosanne Notaro, Janine Tramontana, Richard Charlton, and Michael Held, and others they may have neglected to mention here. 1 See “Request for Information on Application of Uniform Financial Institutions Ratings System.” Federal Deposit Insurance Corporation and Board of Governors of the Federal Reserve System (October 18, 2019) [hereinafter CAMELS Release]. See also Randal K. Quarles, “The Eye of Providence: Thoughts on the Evolution of Bank Supervision,” remarks at the Federal Reserve Board, Harvard Law School, and Wharton School Conference: Bank Supervision: Past, Present, and Future (December 11, 2020). https://www.federalreserve.gov/newsevents/ speech/quarles20201211a.htm. 2 For example, the FDIC publishes aggregate information about troubled banks with FDIC insurance. See the FDIC’s Quarterly Banking Profile. https://www.fdic.gov/bank/analytical/qbp/. 3 Various statutory examination and reporting authorities provide the federal banking agencies with broad authority to oversee and exercise visitorial powers with respect to depository institutions and their holding companies. For example, the Federal Reserve’s authorities to examine and supervise bank holding companies can be found in the Bank Holding Company Act, 12 U.S.C. §§ 1841 et seq. The federal banking agencies’ authority to issue cease and desist orders to correct unsafe and unsound practices can be found in 12 U.S.C. § 1818(b)(1). Federal banking agencies also rely on 12 U.S.C. § 1818(b)(1) to issue matters requiring attention (MRAs) based on safety-and-soundness matters in their reports of examination. For a broader discussion of financial supervision, see, among many other sources, Gerald Dunne, “The Legal Basis of Bank Supervision,” in Bank Supervision (Federal Reserve Bank of St. Louis, 1963). 4 Randal K. Quarles, “Spontaneity and Order: Transparency, Accountability, and Fairness in Bank Supervision,” remarks at the American Bar Association Banking Law Committee Meeting 2020 (January 17, 2020). https://www. federalreserve.gov/newsevents/speech/quarles20200117a.htm. 5 See Cuomo v. Clearing House Association, 557 U.S. 519. When the National Bank Act was enacted in 1864, “visitation” was accordingly understood as “[t]he act of examining into the affairs of a corporation” by “the government itself.” 2 J. Bouvier, A Law Dictionary 790 (15th ed. 1883). Lower courts understood “visitation” to mean “the act of a superior or superintending officer, who visits a corporation to examine into its manner of conducting business, and enforce an observance of its laws and regulations.” First Nat. Bank of Youngstown v. Hughes, 6 F. 737, 740 (CC ND Ohio 1881). 6 For a description of how ratings operate as part of the supervisory process, see Thomas Eisenbach, Andrew Haughwout, Beverly Hirtle, Anna Kovner, David Lucca, and Matthew Plosser, “Supervising Large, Complex Financial Institutions: What Do Supervisors Do?” Federal Reserve Bank of New York Staff Reports No. 729 (May 2015); Jose A. Lopez, “Using CAMELS Ratings to Monitor Bank Conditions,” Federal Reserve Bank of San Francisco Economic Letter (June 1999). 7 For example, in the CAMELS framework, firms are assigned a rating of 1 (strong); 2 (satisfactory); 3 (fair); 4 (marginal); or 5 (unsatisfactory). See Commercial Bank Examination Manual (Federal Reserve Board, May 2019). 8 See, for example, Rules Regarding Availability of Information, 85 Fed. Reg. 57616 (Sept. 15, 2020) (codified at 12 CFR Part 261); Interagency Advisory on the Confidentiality of the Supervisory Rating and Other Nonpublic Supervisory Information (Feb. 28, 2005). https://www.federalreserve.gov/boarddocs/press/bcreg/2005/20050228/ attachment.pdf. 9 A good description of the rationale for protecting the confidentiality of the supervisory process can be found in In Re: Subpoena Served Upon the Comptroller of the Currency, 967 F. 2d 630 (D.C. Cir. 1992): “Bank safety and soundness supervision is an iterative process of comment by the regulators and response by the bank. The success of the supervision therefore depends vitally upon the quality of communication between the regulated banking firm and the bank regulatory agency. This relationship is both extensive and informal. It is extensive in that bank examiners concern themselves with all manner of a bank’s affairs: Not only the classification of assets and the review of financial transactions, but also the adequacy of security systems and of internal reporting requirements, and even the quality of managerial personnel are of concern to the examiners.” Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 21 Why Do Supervisors Rate Banking Organizations? Notes (Continued) 10 See Annette L. Nazareth and Margaret E. Tahyar, “Transparency and Confidentiality in the Post Financial Crisis World – Where to Strike the Balance?” 1 Harvard Bus. L. Rev. 145 (2011); Clifford S. Stanford, “Toward a Coherent and Consistent Framework for Treatment of Confidential Supervisory Information,” 22 N.C. Banking Inst. 41 (2018); and Peter Conti-Brown, “The Curse of Confidential Supervisory Information,” (2019). https://www. brookings.edu/research/the-curse-of-confidential-supervisory-information. 11 See George R. Juncker, “A New Supervisory System for Rating Banks,” Federal Reserve Bank of New York Quarterly Review (Summer 1978); Statement of Brenton C. Leavitt, Board of Governors of the Federal Reserve System, before the Commerce, Consumer and Monetary Affairs Subcommittee of the House Committee on Government Operations, Federal Reserve Bulletin, 62, No. 2 (February 1976). 12 See “Federal Supervision of Bank Holding Companies Needs Better, More Formalized Coordination,” Government Accountability Office, GGD-80-20 (1980); “Federal Supervision of State and National Banks: A Study by the Comptroller General of the United States,” OCG-77-1, Section 6-8 (1977); “Staff Analysis of Testimony Presented by Comptroller General Elmer B. Staats before the Commerce, Consumer and Monetary Affairs Subcommittee of the House Committee on Government Operations and the Financial Institutions Supervision,” Regulation & Insurance Subcommittee of the Senate Committee on Banking, Finance & Urban Affairs (February 1, 1977); see also Mark Greenlee, “Historical Review of ‘Umbrella Supervision’ by the Board of Governors of the Federal Reserve System,” Federal Reserve Bank of Cleveland, Working Paper 08-07 (October 2008). 13 Financial Institutions Regulatory Act of 1978, H.R. Rep No. 95-1383, cited in Michael Barr, Howell Jackson, and Margaret Tahyar, Financial Regulation: Law and Policy (2018): p. 835. 14 92 Stat. 3694, § 1002.1006(b)(1). 15 Press release, “Uniform Financial Institutions Rating System,” Federal Financial Institutions Examination Council (November 21, 1979). https://fraser.stlouisfed.org/files/docs/historical/frbdal/circulars/frbdallas_circ_19791129_ no79-191.pdf [hereinafter UFIRS Release]. 16 UFIRS Release. 17 See “Oversight Hearing into the Effectiveness of Federal Bank Regulation (Franklin National Bank Failure),” House of Representatives Subcommittee of the Committee on Government Operations, February 1976; Report to Accompany Financial Institutions Regulatory Act of 1978, Senate Committee on Banking, Finance and Urban Affairs, 95 H.R. Rep. No. 95-1383 at 17 [hereinafter FIRA Report].” 18 See FIRA Report, p. 24. 19 “Uniform Financial Institution Rating System,” Office of the Comptroller of the Currency, 1979 OCC CB LEXIS 5, p. 6. 20 Statement of Governor Charles Partee before the Senate Banking Committee, Federal Reserve Bulletin 65, No. 6 (June 1979): 463. 21 Rating the Adequacy of Risk Management Processes and Internal Controls at State Member Banks and Bank Holding Companies, Federal Reserve Board SR Letter 95-51. 22 Board of Governors of the Federal Reserve System, FRRS 3-1575 (1996). 23 See “Bank Holding Company Rating System,” Board of Governors of the Federal Reserve System, 69 FR 43996 (2004); Scott Alvarez, “Risk Management and Compliance from the Fed’s Perspective,” speech at the Second Annual Minnesota CLE Banking Law Institute (March 7, 2005), cited in Michael E. Bleier, “The Federal Reserve Board’s New Rating System for Bank Holding Companies and Financial Holding Companies” (2007) . 24 Board of Governors of the Federal Reserve System, FRRS 3-1575 (1996). 25 “Large Financial Institution Rating System; Regulations K and LL,” Board of Governors of the Federal Reserve System, 83 FR 58724 (Nov. 21, 2018). Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 22 Why Do Supervisors Rate Banking Organizations? Notes (Continued) 26 Statement of Randal K. Quarles before the Senate Committee on Banking, Housing and Urban Affairs (November 15, 2018). 27 See “Consolidated Supervision Framework for Large Financial Institutions,” Federal Reserve Board, SR Letter 12-17 (Dec. 17, 2012). 28 The Board kept RFI/C(D) in place for other banking organizations, on the basis that the nature of their supervision had not changed materially since the crisis. Its discussion of the purpose of the system bears much more resemblance to the 1996 releases than it does to the articulated purpose of the new LFI system (or even the 2004 release). Its purpose is consistency (“comprehensive and uniform”); identifying problem or deteriorating companies; and informing the Federal Reserve (not Congress or the public this time) about the aggregate strength of the industry. See “Supervisory Ratings System for Holding Companies with Total Consolidated Assets less than $100 Billion,” Federal Reserve Board, SR Letter 19-4 (Feb. 26, 2019) [hereinafter SR Letter 19-4]. 29 81 Fed. Reg. 58932 (Aug. 26, 2016). 30 12 U.S.C. 2903. 31 Compare, for example, Michael S. Barr, “Credit Where it Counts: The Community Reinvestment Act and Its Critics,” 80 N.Y.U. Law Review, 513 (2005), and Jonathan R. Macey and Geoffrey P. Miller, “The Community Reinvestment Act: An Economic Analysis,” 79 Va. L. Rev. 291 (2003). 32 SR Letter 19-4. 33 “Request for Information on Application of the Uniform Financial Institutions Rating System,” Board of Governors of the Federal Reserve System and Federal Deposit Insurance Corporation, 84 FR 58383 (Oct. 31, 2019). 34 See Economic Growth, Regulatory Relief, and Consumer Protection Act § 210, Pub. L. No. 115-174, 132 Stat. 1306, 1316 (codified at 12 U.S.C. § 1820(d)(4)(A)); 12 CFR § 208.64(b)(3). A second example is the BHC Supervision Manual, which, in discussing prioritization decisions for regional BHCs, states that Federal Reserve holding company supervisors should “heavily rely” on the ratings assigned by regulators of banks when the BHC is noncomplex and the CAMELS ratings are “1” or “2.” Similar to supervisory prioritization, the agencies have agreed on enhanced coordination of examinations for the lead bank within a BHC, when the lead bank is rated 4 or 5 or the BHC or lead bank is a composite whose financial condition has worsened significantly since the last inspection. Bank Holding Company Supervision Manual, Consolidated Supervision of Regional BHCs 1050.2.5.13 (July 2016). 35 See Bank Holding Company Supervision Manual, Funding (Bank Holding Company Funding and Liquidity), 2080.05.03 (July 2010). 36 12 CFR § 363.1 et seq. 37 CAMELS Release. 38 12 U.S.C. § 1818(b)(8). 39 See BSA/AML Manual, Federal Financial Institutions Examination Council, Core Examination Procedures for Assessing the BSA/AML Compliance Program and Appendix R: Interagency Statement on Enforcement of Bank Secrecy Act/ Anti-Money Laundering Requirements. The Government Accountability Office (GAO) recently criticized the Federal Reserve for not having more formal criteria for when to escalate supervisory judgments to enforcement actions. This criticism is similar to others the GAO has made in recent decades, recommending that the Federal Reserve adopt “tripwires” for the use of enforcement actions. See, for example, “Regulators Improved Supervision of Management Activities but Additional Steps Needed,” Government Accountability Office, GAO-19-352 (2019); “Lessons Learned and a Framework for Monitoring Emerging Risks and Regulatory Response,” Government Accountability Office, GAO-15-365 (2015); “Prompt and Forceful Regulatory Actions Needed,” GAO, T-GGD-91-15 (1991). 40 12 CFR § 208.43(c). 41 See 17 USC § 1843(m). Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 23 Why Do Supervisors Rate Banking Organizations? Notes (Continued) 42 12 CFR § 225.83. Similarly, state member banks may hold “financial subsidiaries” (subsidiaries with the power to invest in financial activities not permissible for the state member banks themselves) when they are, among other things, “well-managed,” and if their rating falls below that threshold, they need to execute a 4(m)-like agreement. 12 CFR § 208.74. 43 See 12 CFR §§ 208.21–208.22; see also SR Letter 13-7. 44 See 12 CFR § 225.13. 45 See SR Letter 14-2, “Enhancing Transparency in the Applications Process,” Federal Reserve Board (Feb. 24, 2014). 46 See Greg Baer and Jeremy Newell, “How Bank Supervision Lost Its Way,” Bank Policy Institute (2017). https://bpi. com/how-bank-supervision-lost-its-way. 47 12 CFR § 215.4(d)(2). As another example, the qualified financial contracts (QFC) rule provides that institutions with a troubled rating have to meet heightened supervisory expectations for how they keep records of QFC, to ensure that the QFCs will be readily identifiable in case of resolution. 12 CFR § 371.3. 48 12 CFR § 327.9. For a recent case where the translation of ratings into consequences had an effect on the analysis of the case, see Builders Bank v. FDIC, 846 F. 3d 272, 275 (7th Cir. 2017) (“[t]he effect of CAMELS ratings on insurance premiums creates a concrete stake that makes the current dispute justiciable”). 49 See 12 USC § 347(b)(D)(ii); 12 CFR § 201.3(c). 50 See Federal Reserve Discount Window Payment System Risk, General Information, Primary and Secondary Lending Programs. https://www.frbdiscountwindow.org/pages/general-information/primary-andsecondary-lending-programs#elig. 12 CFR § 201.4. 51 The FDIC’s implementing regulation defines “troubled” to mean, among other things, having a composite CAMELS rating of 4 or 5. 52 12 USC § 1828(k). 53 12 CFR § 225.71 et seq. 54 See Comment Letter from the Bank Policy Institute, “Substantive Review & Revision of the Uniform Financial Institution Rating System” (Jan. 10, 2020). 55 See Pub. L. No. 102-242 § 142, 105 Stat. 2279–81 (Dec. 19, 1991) (codified as amended at 12 U.S.C. § 347b(b)). 56 See Pub. L. No. 103-325 § 314, 108 Stat. 2221–22 (Sept. 23, 1994) (codified as amended at 12 U.S.C. § 1831(m)(i)). 57 See Pub. L. No. 104-208 § 2208, 110 Stat. 3009-405–3009-406 (Sept. 30, 1996) (codified as amended at 12 U.S.C. § 371d(a)). 58 See Pub. L. No. 104-208 § 2206, 110 Stat. 3009-405–3009-406 (Sept. 30, 1996) (codified as amended at 12 U.S.C. § 371d(a)). 59 See Pub. L. 106-102, 113 Stat. 1338, 1342–52 (Nov. 12, 1999) (codified as amended at 12 U.S.C. § 1843(k) et seq.); Pub. L. 111-203 § 606, 124 Stat. 1376, 1607 (codified as amended at 12 U.S.C. § 1843(l)(1)(C). 60 See Pub. L. No. 101-73 § 914, 103 Stat. 183, 484–47 (July 21, 2010) (codified as amended at 12 U.S.C. § 1831i). 61 See 55 Fed. Reg. 6787 (Feb. 27, 1990) (codified as amended at 12 C.F.R. §§ 225.71–225.73). 62 See 57 Fed. Reg. 44866-01 (Sept. 29, 1992) (codified as amended at 12 C.F.R. § 208.43(c)). 63 See, for example, note 46 above and the accompanying text; see also SR Letter 15-11/CA 15-9, “Examinations of Insured Depository Institutions Prior to Membership or Merger into a State Member Bank,” Federal Reserve Board (Oct. 13, 2015). Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 24 Why Do Supervisors Rate Banking Organizations? Notes (Continued) 64 See Builders Bank v. FDIC, 846 F.3d 272 (7th Cir. 2017). See also Daniel K. Tarullo, “Bank Supervision and Administrative Law,” Columbia Business Law Review (forthcoming), for a systematic consideration of how administrative law doctrines apply to banking supervision. https://custom.cvent.com/20310C03166C4C11B1AA63B0D6300264/files/ event/67aec69c628d459d8366466979e3f8af/28c97d0e935940c0b705ae832b63af55.pdf. 65 See “Proposed Guidance on Supervisory Expectations for Boards of Directors,” Board of Governors of the Federal Reserve System, 82 FR 37219 (2017); see also Clifford S. Stanford, “Towards a Coherent and Consistent Framework for Treatment of Confidential Supervisory Information,” 22 N.C. Banking Inst. 41 (March 2018). 66 Federal Reserve Board of Governors, “Federal Reserve Supervision and Regulation Report.” https://www. federalreserve.gov/publications/supervision-and-regulation-report.htm.” 67 For analysis of how long information revealed by supervisory ratings is likely to be superior to market signals, see, for example, Rebel A. Cole and Jeffrey W. Gunther, “A CAMEL Rating’s Shelf Life,” (November 2008), https://papers. ssrn.com/sol3/papers.cfm?abstract_id=1293504; Beverly J. Hirtle and Jose A. Lopez, “Supervisory Information and the Frequency of Bank Examinations,” Federal Reserve Bank of New York Economic Policy Review 5, no. 1 (1999). 68 See Randal K. Quarles, “The Eye of Providence: Thoughts on the Evolution of Bank Supervision,” remarks at the Federal Reserve Board, Harvard Law School, and Wharton School Conference: Bank Supervision: Past, Present, and Future, Dec. 11, 2020. https://www.federalreserve.gov/newsevents/ speech/quarles20201211a.htm. See also Daniel K. Tarullo, “Bank Supervision and Administrative Law,” Columbia Business Law Review (forthcoming). https://custom.cvent.com/20310C03166C4C11B1AA63B0D6300264/files/ event/67aec69c628d459d8366466979e3f8af/28c97d0e935940c0b705ae832b63af55.pdf. 69 See “Goodhart’s Law” in Oxford Reference. https://www.oxfordreference.com/view/10.1093/oi/ authority.20110803095859655. Federal Reserve Bank of New York Economic Policy Review 27, no. 3, November 2021 25 FEDERAL RESERVE BANK OF NEW YORK ECONOMIC POLICY REVIEW The Economic Policy Review is published by the Research and Statistics Group of the Federal Reserve Bank of New York. The views expressed are those of the individual authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Series Editor: Julian di Giovanni. Editorial Staff: Robert Powell, Trevor Delaney, Maureen Egan, Anna Snider, and Peter Stevens. Design Staff: Theresa Izzillo, Alyona Botogoeva, and Laura Gharrity. Economic Policy Review articles may be reproduced for educational or training purposes, provided they are reprinted in full; include credit to the author(s), the publication, and the Bank; and include the publication’s disclaimer. ©2021 The Federal Reserve Bank of New York www.newyorkfed.org/research Follow us on Twitter: @NYFedResearch