Introduction

On July 27, 2023, the federal banking regulators—the Office of the Comptroller of the Currency (OCC), the Federal Deposit Insurance Corporation (FDIC), and the Federal Reserve (Fed)—jointly issued a proposed rule that revises large bank capital requirements.¹ The proposal (1) implements the so-called Basel III Endgame (the regulators' final phase of regulations intended to implement policy consistent with the latest recommendations of the Basel Committee on Banking Supervision [BCBS]),² (2) implements some of the recommendations that Fed Vice Chair Michael Barr proposed in his "holistic capital review,"³ and (3) responds to issues that arose in the 2023 large bank failures.⁴

This report focuses on summarizing and analyzing the proposal, and it assumes the reader has some familiarity with bank regulation and capital standards. The textbox below provides basic background on these issues. For further background, see CRS Report R47447, Bank Capital Requirements: A Primer and Policy Issues, by Andrew P. Scott and Marc Labonte.

Basel III Endgame

Forty-five bank regulators worldwide, including those in the United States, formulate and agree to apply generally consistent bank capital requirements through an international standard-setting body called the Basel Committee on Banking Supervision. While the Basel frameworks (also referred to as "accords") have no legal force in the United States, it is helpful to understand the BCBS approach because domestic regulators have chosen to closely align their rules—which are implemented through the notice-and-comment rulemaking process pursuant to the Administrative Procedure Act—with these frameworks. This choice is arguably predictable given that members of the BCBS negotiate the rules together and that U.S. bank regulators are, as the representatives from the world's largest economy and financial system, influential members. As such, BCBS standards likely reflect, at least in part and possibly significantly, the views and preferences of the U.S. regulators.

Setting capital requirements through Basel is an iterative process—requirements have repeatedly been tweaked over the decades as problems emerge or policy priorities change, with three landmark revisions (known as Basel I, II, and III). In 2010, the BCBS issued new prudential standards to address problems exposed by the 2008 financial crisis, revising and updating the Basel II standards. These new standards came to be known as Basel III. Some of the more technical details were filled in later when the BCBS agreed on and issued the final major set of Basel III standards in December 2017. As with Basel III generally, many of the details in the 2017 proposal were in response to problems that arose during the financial crisis. This last round of Basel III reforms is sometimes colloquially referred to as the Basel III Endgame. According to the BCBS

A key objective of the revisions … is to reduce excessive variability of risk-weighted assets (RWAs) … [and] help restore credibility in the calculation of RWAs by: (i) enhancing the robustness and risk sensitivity of the standardised approaches for credit risk and operational risk, which will facilitate the comparability of banks' capital ratios; (ii) constraining the use of internally-modelled approaches; and (iii) complementing the risk-weighted capital ratio with a finalised leverage ratio and a revised and robust capital floor.⁵

The recent U.S. proposal would implement all of these changes except the revision to the leverage ratio, which is being addressed through separate rulemaking.⁶ With the issuance of the proposed rule, U.S. regulators are now on par with some other major economies but still lagging behind others in implementing the Basel III Endgame provisions.⁷

Basel participants are committed to applying Basel standards to internationally active banks (a subset of large banks), and U.S. regulators have applied some—but not all—of the standards to all domestic banks. Divergence between U.S. capital standards and the Basel III agreements broadly comes in the form of either allowing small banks to opt out of Basel III requirements or imposing more stringent requirements on large banks. As a result of this divergence, all U.S. banks calculate their RWA using the standardized approach, and the largest, most systemically important, and internationally active U.S. banks are also currently required to calculate RWA using advanced approaches. The more complex advanced approaches are determined using internal models specific to each bank, referred to as the "internal ratings-based approach" to calculating RWA. Advanced approaches banks must calculate the two different ratios required of both approaches: capital/standardized approach RWA and capital/advanced approaches RWA. To determine whether they meet their minimum requirements, advanced approaches banks are required to apply the lower of their two ratios to their capital requirements.

Holistic Capital Review

In 2022, Fed Vice Chair Barr announced a "holistic review of capital standards" for large banks, which he described as "not looking only at each of the individual parts of capital standards, but also at how those parts may interact with each other—as well as other regulatory requirements—and what their cumulative effect is on safety and soundness and risks to the financial system."⁸ One motivation for the review was to evaluate whether policy goals are still achieved given the interaction of multiple large bank capital requirements. In a speech in July 2023, he announced the outcome of that review, which resulted in recommendations for future rulemakings.⁹ Barr did not recommend fundamental changes in large bank capital requirements and announced that several requirements would not be changed at all. He highlighted implementation of the Basel III Endgame as "an important aspect of my proposals," and the rule significantly modifies various aspects of how large banks calculate their capital requirements.

Bank Failures

After experiencing zero U.S. bank failures in 2021 and 2022, the spring of 2023 witnessed the second (First Republic), third (Silicon Valley Bank), and fifth (Signature) largest failures in history as measured by asset size in nominal dollars.¹⁰ Combined, these failures are expected ultimately to impose tens of billions of dollars of losses on the FDIC's Deposit Insurance Fund. To avoid a broader run on the banking system, the FDIC invoked the rarely used systemic risk exception to guarantee all uninsured depositors at two of the banks.¹¹ While the Basel III Endgame agreement predates the failures, regulators have pointed to the failures as a rationale for applying most elements of the proposed rule to banks with over $100 billion in assets. Each of the three large banks that failed had over $100 billion in assets.

As will be discussed below, the role that unrealized losses on securities played in the failures may have contributed to the proposed rule's inclusion of those losses in regulatory capital for all banks with over $100 billion in assets. That capital treatment was first proposed for all banks in 2012¹² but was not part of the 2017 BCBS Basel III Endgame document. Other factors that played a role in their failures—such as rapid asset growth, reliance on uninsured deposits, supervisory forbearance, and concentration risk—are not specifically addressed by the proposal.

The Joint Large Bank Capital Proposed Rule

The proposed rulemaking would revise the capital framework for large banking organizations. The Acting Comptroller of the Currency approved the joint proposal, and the FDIC board voted 3-2 and the Fed's board voted 4-2 for the joint proposal. The comment period was extended from November 30, 2023, to January 16, 2024. The sections below summarize the scope of the proposal and its technical provisions.

According to the proposal, its purpose is to improve the consistency of capital requirements across banks, better match capital requirements to risk, reduce their complexity, and improve transparency of banks' financial condition for supervisors and the public.

Scope and Timing

The proposed rule applies to insured depository institutions (IDIs), which include commercial banks and savings associations; bank holding companies (BHCs); savings and loan holding companies that are not substantially engaged in insurance; and foreign banking organizations with over $100 billion in assets (hereinafter, collectively referred to as "banks," except where a distinction is necessary). As of the date of the proposal, the total number of affected institutions are 25 U.S. BHCs, 12 intermediate holding companies (IHCs) of foreign banks, and 62 IDIs (including IDIs of holding companies with over $100 billion in assets).¹³

The proposal would replace advanced approaches with a new expanded risk-based approach and extend those requirements to Category III and IV banks. Since the Fed's 2019 rule implementing the enhanced prudential regulation tailoring required by the Economic Growth, Regulatory Relief, and Consumer Protection Act (EGRRCPA; P.L. 115-174, enacted May 24, 2018 and often referred to by the bill number, S. 2155), only the nine Category I and II banks are required to use advanced approaches, but any other bank may voluntarily opt in.¹⁴ (See the text box below for more details on EGRRCPA and large bank categories.) Unlike the 2019 rule, however, most of the proposed rule would also apply to IDIs that are not subsidiaries of holding companies. Signature Bank did not have a holding company.

In addition, the market risk provisions (described below) would apply to banks with $100 billion or more in total assets or $5 billion or more of trading assets plus trading liabilities (increased from $1 billion or more under current regulation) or trading assets plus trading liabilities equal to 10% or more of total assets (unchanged from current regulation).

Under the proposal, Category IV banks would also be subject to the countercyclical capital buffer and the supplementary leverage ratio (SLR). Currently Category I-III banks are subject to the countercyclical buffer, which requires large banks to hold more capital than other banks when regulators believe that financial conditions make the risk of losses abnormally high. In normal times, the countercyclical buffer is to be set at zero, but in high-risk circumstances, it could be set as high as 2.5%.¹⁶ In practice, it has always been set at zero since inception but could potentially be raised in the future.

Currently, Category II and III banks must meet a 3% SLR, and Category I banks must meet a higher SLR. Under the proposal, Category IV banks would also have to meet an SLR equal to 3% of Tier 1 capital/(total assets+off-balance sheet exposures). Like the leverage ratio, the SLR uses Tier 1 capital in the numerator and unweighted assets in the denominator. The difference between the leverage ratio and the SLR is that the SLR includes off-balance-sheet exposures in the denominator. Thus, the numerator is the same, but the denominator is larger. The SLR is intended to ensure that the bank is adequately safeguarded against off-balance-sheet losses that are not captured in the leverage ratio.

The proposal includes a three-year phase in (until June 30, 2028) of the change to RWA under the expanded risk-based approach, discussed in the next section, and to the effect on capital of changes in accumulated other comprehensive income, discussed in the section below entitled "Changes to the Definition of Capital, Including AOCI." Other parts of the proposal would apply upon the effective date of the rule.

Changes to Risk Measurement

As discussed above, advanced approaches banks currently face two requirements for calculating their RWA—the standardized approach and the advanced approach. The proposal replaces the advanced approach with the expanded risk-based approach. Other banks with over $100 billion in assets would be required to calculate RWA under two approaches for the first time.

The standardized approach would remain focused on general credit risk and market risk whereas the expanded risk-based approach would comprise credit risk, credit value adjustment (CVA) risk for over-the-counter derivatives, operational risk, and market risk. Thus, banks that were previously subject to only the standardized approach would be facing for the first time some of the major requirements that advanced approaches banks currently face, such as operational risk requirements. The internal models (or ratings-based) approach used in advanced approaches gives banks some latitude to assess risk in their own ways. The proposed rule would replace the internal ratings-based approach¹⁷ for credit and operational risk with a second standardized approach (the expanded risk-based approach). Market risk and CVA risk would use a revised internal model-based approach that would "better account for tail risk" and "require a more rigorous model approval process."

Under the proposed framework, the RWA under the new approach "would equal the sum of risk-weighted assets for credit risk, equity risk, operational risk, market risk, and CVA risk" less certain adjustments for credit losses. The RWA under the standardized approach would have a revised approach to determining market RWA. The bank would apply the RWA measure that yields the lowest capital ratio under each calculation.

The individual elements that face revisions are discussed in more detail in the following sections.¹⁸

Credit Risk

Credit risk is the risk that a borrower or counterparty will fail to meet a financial obligation. The proposal would replace the internal models for credit risk with a new expanded risk-based approach. The new approach would retain many of the same definitions and some of the risk weights from the current framework. The proposal intends to provide "more transparent capital requirements for credit risk exposures" than the internal models approach and "more granular risk factors" than the current standardized approach. For example, the expanded risk-based approach would include exposures to banks and credit unions; subordinated debt; and retail, real estate, and corporate entities.

Securitization Framework

Securitization is the process of pooling together certain types of assets and packaging them as securities that pay income to investors based on an ordered priority. The securitization framework provides the capital requirement element related to the repayments on securitized assets. The proposed rule would retain much of the existing capital rule but would modify requirements for certain securitizations, and it would establish a standardized approach to replace the existing supervisory formula approaches. It would also prohibit using the framework for certain credit derivatives (contracts where payment obligations are determined by changes in creditworthiness of some reference entity) and establish a new treatment for certain derivative contracts and securitization exposures.

Equity Risk

Given that equity exposures "represent an ownership interest in the issuer of an equity instrument and have a lower priority of payment or reimbursement," equity risk is the risk of loss in the event that the issuing entity fails to pay all of its debts.¹⁹ Currently, advanced approaches banks can use internal models for both publicly and nonpublicly traded equity derivative contracts and exposures. The proposed rule would eliminate the internal models approach. Instead, publicly traded equity exposures and certain equity exposures to investment funds would be subject to the market risk framework described below. A standardized approach would be implemented for calculating the capital requirements for illiquid and infrequently traded equity exposures. The remaining sections of the current capital rule that do not rely on models would generally remain the same.

Operational Risk

Currently, advanced approaches banks calculate RWA for operational risk—the "risk of loss resulting from inadequate or failed internal processes, people, and systems, or from external events"—using an internal model called the advanced measurement approaches. The proposed rule would eliminate this model and introduce a standardized capital requirement for operational risk. The operational risk capital requirements under the expanded risk-based approach would be based on a banking organization's business volume and historical losses. According to Barr, these indicators are good predictors of future operational losses, and "large banks have experienced significant losses due to operational weaknesses over the past two decades."²⁰

Market Risk

The current framework allows the use of internal models for calculating capital requirements for market risk—exposure to asset price movements. The proposal would introduce a risk-sensitive standardized methodology for calculating RWA. This measure would be the default methodology for banks that face market risk requirements. According to Barr, "The aim of the revised market risk framework is to comprehensively address the lessons of the global financial crisis. The revised framework would permit banks to use their own models to compute elements of the market risk capital requirements only when such risk can be modeled well."²¹ The measure would include "a sensitivities-based capital requirement that would capture non-default market risk based on the estimated losses produced by risk factor sensitivities under regulatorily determined stress conditions." It would also include a standardized default risk and residual risk capital requirement. Limited use of internal models would be permitted when they can "appropriately capture risk."²²

Standardized Output Floor

The proposed rule would implement a "standardized output floor" at 72.5% of the sum of the bank's credit, equity, operational, and CVA RWA as calculated under the expanded risk-based approach plus market RWA under the standardized approach. The purpose of this floor is to set a limit on how much banks can reduce their RWA (and hence, capital requirements) using internal models for market risk.

Stress Capital Buffer

As described above, advanced approaches banks must calculate their capital requirements using two methods. Banks must maintain a capital buffer on top of minimum capital requirements to avoid restrictions, such as limits on share buybacks and dividends. Currently, the stress capital buffer, which is based on stress test losses, is used only under the standardized approach, and a fixed capital conservation buffer of 2.5% is used for advanced approaches. The proposal would require large banks to use the stress capital buffer in both the standardized approach and the new expanded risk-based approach.

Disclosure Requirements

The proposal would revise certain "qualitative disclosure requirements" and introduce new disclosure requirements to "facilitate market participants' understanding" of banks' financial condition and riskiness. The proposal would also transfer most of the existing quantitative disclosures to regulatory reporting forms, which would be coordinated through the Federal Financial Institutions Examination Council, an interagency body composed of depository regulators.

Changes to the Definition of Capital, Including AOCI

The proposed rule would extend certain requirements that affect the definition of capital that currently apply to advanced approaches banks to banks with over $100 billion in assets. First, the proposal would increase the amount of mortgage servicing assets (MSAs) and temporary difference deferred tax assets (DTAs) that are deducted from a bank's capital. Currently a bank can hold MSAs or DTAs equal to up to 25% of its common equity Tier 1 (CET1) capital before a deduction from capital is made. Under the proposal, the threshold would be lowered from 25% to 10%. Second, the proposal would require covered banks to disclose to investors that capital could be subordinated to the U.S. government in receivership or bankruptcy when the banks issue new capital. Third, the proposal would limit the inclusion of allowances for credit losses in capital for Category III and IV banks and change how the limit is calculated for Category I and II banks.

Finally, these banks would have to include most parts of an accounting category called accumulated other comprehensive income (AOCI) in CET1 capital, which would align capital rules with the treatment of AOCI under generally accepted accounting principles. One component of AOCI to be included is unrealized capital gains and losses on available for sale (AFS) debt securities (e.g., corporate and government bonds).²³ Doing so would have the effect of increasing a bank's CET1 levels when it had unrealized capital gains and reducing CET1 when it had losses.

Changes to AOCI are not part of the 2017 Endgame proposal from the BCBS, and U.S. proposals to do so originally predate that document.²⁴ In 2012, the original Basel III proposal would have applied this requirement to all banks (and BHCs). The regulators argued that "unrealized losses could materially affect a banking organization's capital position … and associated risks should therefore be reflected in its capital ratios."²⁵

Facing criticism from banks that this treatment would cause capital levels to be too volatile, the version of the rule finalized in 2013 applied the requirement only to advanced approach banks—at the time, banks with at least $250 billion in assets or $10 billion in on-balance sheet foreign exposure. All other banks could permanently elect to opt out of this requirement. Doing so is sometimes referred to as the "AOCI filter."²⁶

In its 2019 regulation implementing EGRRCPA, the Fed reduced the number of banks subject to various EPR requirements, including limiting the AOCI requirement to Category I and II banks—applying it to only the nine Category I and II banks.²⁷

The 2023 proposal would extend the AOCI requirement to any U.S. bank, BHC, or IHC with over $100 billion in assets. As with earlier reforms, the treatment of trading and held-to-maturity (HTM) securities would not change.

As seen in Figure 1, recognizing unrealized gains and losses would lead to higher capital in some years and lower in others for banks overall, but unrealized losses have increased rapidly beginning in 2022, equaling $232 billion on AFS securities and $284 billion on HTM securities in the first quarter of 2023. This compares to $4 billion in realized losses in the first quarter. The proposal only partially addresses the current problem for two reasons. First, it does not apply to unrealized losses on HTM securities (the rationale being the bank does not intend to sell those securities), which account for over half of banks' unrealized losses. Second it does not apply to banks with less than $100 billion in assets, whereas banks of all sizes have experienced unrealized losses. According to the FDIC, community banks had unrealized losses of $59.2 billion in the first quarter of 2023, and their securities holdings (22% of total assets) are comparable to other banks (24%).²⁸

Figure 1. Unrealized Gains and Losses on Securities Held by FDIC-Insured Depository Institutions 2008:Q1-2023:Q1

Source: FDIC.

Losses on securities played a major role in the failure of Silicon Valley Bank (SVB). At the end of 2022, SVB had $1.9 billion in AFS losses that would have been recognized in capital under AOCI, although most of SVB's securities were classified as HTM and so would not have been affected by the proposal.²⁹ SVB had over $100 billion in assets and would have been subject to the Endgame proposal.³⁰ The Fed also reports that SVB would have had to start complying with the AOCI requirement in 2021 as an advanced approach bank had the 2019 tailoring rule not limited the AOCI requirement to Category I and II banks.³¹

Concerns About Specific Aspects of the Proposal

Because the proposal primarily has the effect of changing the way that specific RWA are calculated, the proposal disproportionately affects specific asset classes most affected by the modified risk weights. Banks holding those assets, in turn, would be most affected by the proposal.

The goal of risk weights is to assign a weight that appropriately reflects an activity's actual risk. Appropriate risk weights incentivize banks to pursue activities where risk and reward are properly balanced. If regulators raise risk weights because previous weights were too low relative to actual risk, then the safety and soundness of the banking system would be improved by reducing the incentive to pursue that activity. By contrast, if regulators raise risk weights to be higher than is commensurate with the activity's actual risk, then banks would be too disincentivized to engage in an activity, and economic efficiency would fall (or the activity would migrate out of the banking system). Further complicating matters is the fact that riskiness changes unpredictably over time. Regulators may be able to retroactively estimate the approximate riskiness of an activity based on historical data, but that historical relationship may not hold in the future.

The new proposal has received some criticism from industry.³² A few specific complaints about affected asset classes have been raised so far,³³ although the volume of complaints is not necessarily commensurate with the most affected asset classes.

Tax Equity

One concern is the way that equity is being treated in the proposal. For example, some interest groups, such as the renewable energies industry, have voiced concern over the way tax equity³⁴ would be treated. The current framework applies a 100% risk weight to nonpublicly traded equity under 10% of a bank's total capital. Banks with values that exceed 10% would apply a 400% risk weight. The proposal would remove the 10% threshold, and it would result in banks facing the higher risk weights irrespective of their exposure levels. Industry advocates have suggested that the removal of this threshold is likely to disincentivize investment in renewable energies, which often rely on tax equity project financing.³⁵

Residential Mortgages

Compared to the current standardized approach, which assigns a 50% risk weight for prudently underwritten mortgages that are current or a 100% risk weight otherwise, the proposal would assign risk weights that rise in increments from 40% to 125% depending on the loan-to-value (LTV) ratio and whether the mortgage is dependent on cash flows from the property, such as rent. For example, a mortgage with a 60%-80% LTV that is not dependent on cash flows would be assigned a 50% risk weight. Mortgages with a higher LTV would receive a higher risk weight, as would those with the same LTV that are dependent on cash flows. On average, risk weights on residential mortgages would be expected to rise.³⁶ Risk weights under the proposal are also higher than the BCBS's standards in the 2017 Endgame document.³⁷

Another proposal that might affect mortgage markets would change how banks make deductions from their capital. In the current regime, Category III and IV banks must limit deductions of certain assets such as mortgage servicing assets (or rights) that individually exceed 25% of CET1 capital. The proposal would lower that threshold to 10%, on par with the threshold facing Category I and II banks.

Capital Markets Activities

New market risk requirements cause the largest increase in RWA for Category I and II banks compared to the status quo and also increase for Category III and IV banks. This could increase the cost of banks providing capital markets services such as securities trading, market making, and underwriting. For most types of securities, underwriting activity is dominated by Category I banks. Category III and IV banks that were not previously subject to CVA rules and Category I and II banks that did not previously face advanced approaches as a binding constraint would now be bound by CVA risk rules that could cause the cost of derivatives offerings to rise. The new operational risk requirements are also based partly on trading activities.³⁸

Fee-Based Activities

Newly applicable operational risk requirements for Category III and IV banks would have the largest effect on RWA for those banks. (By contrast, operational risk requirements would reduce RWA for Category I and II banks compared to advanced approaches but would be newly binding for those banks that switch from having the standardized approach to the proposed enhanced risk-based approach as the binding constraint.) Operational risk requirements are based in part on fee and commission income received and expenses paid from "advisory and financial services, including insurance income." Those fee-based services include fiduciary activities, brokerage, investment banking, interchange and wire transfer fees, and underwriting.

Economic/Capital Impact

The proposed rule provides a five-page qualitative cost-benefit analysis of the impact of the rule on covered banks and the broader economy in contrast to a numerical estimate of the overall net impact. In their analysis, the regulators state that they expect the benefits of the proposal to outweigh the risks because "better alignment between capital requirements and risk-taking helps to ensure that banks internalize the risk of their operations."³⁹ The regulators expect capital requirements to increase "modestly" for lending activities and "substantially" for trading requirements. According to Barr

The bulk of the rise in required capital anticipated in the proposed rule is attributed to trading and other activities besides lending—activities that have generated outsized losses at large banks and areas where our current rules have shortcomings. The estimated increase in capital required for lending activities on average—inclusive of both credit risk and operational risk requirements—is limited. Such a rise might be expected to increase the cost to banks for funding the average lending portfolio by up to 3 basis points—0.03 percentage points.⁴⁰

Although the proposed rule would not increase the required capital ratios banks face, it would increase the amount of capital that banks would have to hold, primarily because it would increase their RWA (i.e., the denominator of risk-weighted capital ratios). The regulators estimate that the proposal would increase the average binding CET1 capital level that large banks are required to hold by 16% and Tier 1 capital by 9%.⁴¹ Note that (1) this estimate is an average, and the effects on any particular bank would differ; and (2) this is an estimate based on past data, and the actual effect would depend on future actions by the banks, including how they responded to the rule. For example, banks could reduce their trading activities to lower the capital impact. The largest and most complex banks would see the greatest impact (CET1 requirements would increase 19% for Category I and II banks, 6% for Category III and IV domestic banks, and 14% for IHCs), though the effects would vary by bank. As discussed above, the effect would be greatest for banks with the most affected activities, such as significant trading activities and fee-based income. The regulators estimate that all banks already hold enough capital to meet the new capital requirements except for five Category I or II holding companies, which would need to raise capital "between 16 and 105 basis points relative to their risk-weighted assets" to meet the proposed requirements.

The regulators also estimate that the proposal would increase the average total loss absorbing capacity (TLAC) requirement by 15.2%, leading three banks to have a shortfall, and the average long-term debt requirement by 2.0% for Category I banks.⁴² Currently, TLAC and long-term debt requirements apply only to Category I banks, but a separate proposed rule would extend long-term debt requirements to all banks and BHCs with over $100 billion in assets.⁴³

In addition, the changes to the definition of capital in the proposal would increase how much capital banks have to hold by modifying what qualifies for the numerator of the capital ratio. The regulators estimate, based on 2015 to 2022 data, that the inclusion of AOCI in capital would increase average capital in the long run, as summarized in Table 1. In any given year, the effect would be larger if banks have unrealized losses and smaller if banks have unrealized gains.

The proposal would also reduce the G-SIB surcharge for Category I banks by 0.16 percentage points, partially offsetting the capital increases required by other parts of the proposal.

In addition, the proposal imposes a regulatory burden through compliance costs. The regulators estimate that the proposal would increase "estimated annual burden hours" by over 900,000 hours cumulatively for all of the banks subject to the proposal.⁴⁴

Policy Issues

The policy debate over capital requirements is long-standing, with proponents of stronger prudential regulations often citing past financial turmoil as evidence that the existing framework is insufficient. Opponents to more stringent regulation point toward risks to bank profitability and competitiveness with nonbanks. This represents the perennial friction between safety and soundness facing bank regulators.

A better capitalized and safer banking system could make future banking failures and crises less likely but could also make credit more expensive and less available.⁴⁵ Depending on the marginal change in those two potential effects, stricter capital rules could improve or worsen economic outcomes on net over the long term. Proponents of strict capital rules argue that they are necessary to avoid bad outcomes such as those seen during and after the 2008 financial crisis, when hundreds of U.S. banks failed and even more received government assistance. That financial crisis caused the deepest and longest recession since the Great Depression. According to the proposal, many of its changes draw from the lessons of the financial crisis and, to a lesser extent, the 2023 large bank failures. Critics argue that the banking system, which has already been operating under stronger regulatory requirements since the financial crisis, has proven resilient through the COVID-19 pandemic disruptions and multiple rounds of stress testing and thus does not need stricter requirements.

Banks' past losses can inform capital rules, but future losses are often from new and unpredictable sources. This is to say that the likelihood of specific capital requirements being exactly correct is close to zero, as it is largely an (informed) exercise where policymakers seek to balance risks to safety and soundness and profitability across the industry. Given that there are several different requirements and over 4,500 unique banking institutions in the United States, there is little chance a standard approach is perfectly suitable to any one institution. Rather, capital regulations (in particular standardized rules) seek to address the industry's total risk, and policymakers attempt to provide tailored regulation where they determine it is needed.

Against the backdrop of these tradeoffs and uncertainty surrounding them, regulators have little incentive to lower capital requirements, as their mandates skew toward safety and soundness, and banking industry leaders generally push for lower requirements, as they feel constrained by any regulation that makes them use more expensive funding sources and potentially impedes their profitability.

In addition to the overall economic effects of the proposal, critics have raised some more specific points that are discussed in the rest of this section.

"Gold-Plating"

Critics complain that the United States has unnecessarily "gold-plated" several of the requirements originally proposed by the BCBS—within both the Endgame proposal and previous rules implementing Basel standards—in ways that make U.S. standards more stringent.⁴⁶ Basel sets out a framework for the regulators to implement. The banking industry has been vocal that the U.S. regulators' proposal exceeds many of these requirements, particularly when compared to international peers such as the European Union.⁴⁷ For example, a number of requirements in the proposal, such as mortgage risk weights, exceed those proposed by the BCBS.⁴⁸ However, U.S. regulators argue that the largest U.S. banks (to which the gold-plated standards apply) have unique characteristics that exceed the standard risks of most banks domestically and internationally.⁴⁹ Meanwhile, the European Union has proposed requiring its affected banks to hold less capital than the BCBS proposal (but still more than they currently hold).⁵⁰ Gold-plating is disadvantageous from an international competitiveness perspective but may or may not be helpful from a safety and soundness and financial stability perspective.

Standard Rules vs. Internal Models

As noted above, one goal of the proposal is to reduce complexity for Category I and II banks. For other banks with over $100 billion in assets, complexity would be increased, as they would face a new set of requirements. There is a long-standing debate over whether simple or complex regulations better reduce risk in the presence of uncertainty in general, as simple rules are more easily understood and applied and may be hard to game, but complex rules can more effectively address the complicated risks and activities undertaken by banks.

Despite the goal of reducing complexity, both the internal models and their proposed replacements are highly complex and technical. Internal models have the advantage of being created and calibrated by the bank, which has the most inside knowledge of the risks it faces. But internal models have the disadvantage of lacking transparency and being potentially susceptible to internal manipulation to reduce capital requirements. According to the regulators, in practice internal models have resulted in "unwarranted variability across banking organizations in requirements for exposures with similar risks."⁵¹

The proposal replaces the largest banks' ability to use internal models to assess risk in most cases and more heavily controls the way that banks can apply internal models to assess market risk. If regulators have decided that banks cannot properly assess risk, some may question whether regulators can assess risk better. Particularly with respect to the bank failures of 2023, existing prudential regulatory requirements did not result in supervisory actions by supervisors that prevented the banks from failing.

Overlapping Approaches

The standardized approach that applies to small and large U.S. banks includes some deviations from the standards issued by the BCBS. Both the advanced approach and its proposed replacement, the "expanded risk-based approach" applying to large banks, hew more closely to the BCBS's standards.

U.S. regulators have set themselves two policy goals that they are navigating between: (1) setting rules that align with the BCBS's standards to promote a "level playing field" in the United States and abroad or (2) setting rules that are different from Basel standards but tailored to address U.S.-specific concerns.⁵² Rather than choosing between these two goals, U.S. regulators have chosen to apply both sets of rules to advanced approaches banks (sometimes referred to as "two capital stacks") simultaneously—and then use only the binding one—in order to achieve both goals. The proposal would extend this dual system to all banks with over $100 billion in assets.

Requiring large banks to calculate their capital ratios using two methods also ensures that large U.S. banks do not face lower requirements than small ones do. The tradeoff is more burdensome regulatory compliance for large banks. This system requires compliance departments at large banks to ensure that each activity undertaken by the bank is compliant with two sets of capital rules, even though only one set of rules ends up being binding. One could debate whether regulators are imposing unduly burdensome regulations on large banks—especially if the new expanded risk-based approach turns out to consistently be the binding one—instead of choosing between their own potentially incompatible policy goals.⁵³ While it is often suggested that large banks have different levels of scale and complexity than community banks do, it is potentially unclear why regulators would then retain the original standardized approach for these banks. As discussed above, the dual system has also allowed "gold-plated" U.S. standards in some cases, which could pose challenges for one of the policy goals—the level playing field internationally.

Capital Neutrality

One of the main intended purposes of the proposal is to better align risk with capital through modifications to risk weights. A perhaps unintended outcome of the proposal is to require large banks to hold more capital overall because of the increase in RWA—the proposal would effectively raise required Tier 1 capital by an estimated 16% and CET1 by 9%.⁵⁴ This occurs because the standardized approach goes from being the binding approach that requires more capital under current rules to the non-binding approach that requires less under the proposal for most large banks, which must simultaneously comply with both approaches. This may explain why large banks prefer the status quo.⁵⁵

In theory, regulators could offset the effect of higher RWA on required capital with reductions to ratios, making the proposal roughly capital neutral. Capital neutrality would be desirable if the overall required level of capital is currently achieving the goals of capital requirements (namely, safety and soundness and financial stability). If that were the case, effectively requiring banks to hold more capital would impose economic costs that outweigh the benefits.⁵⁶ There is little consensus on this topic, but several Fed and FDIC board members who voted against the proposal argued that evidence suggests that large banks already hold sufficient capital to achieve policy goals.⁵⁷ A drawback to making the proposal capital neutral is that it would require applying either different capital ratios to small and large banks or different ratios to the standardized approach and the new expanded risk-based approach.

Tailoring

EGRRCPA amended Title I of the Dodd-Frank Act to require the tailoring of EPR requirements for large banks, differentiating among large banks based on factors such as size, riskiness, complexity, and the activities they engage in. For banks between $100 billion and $250 billion in assets, it provides the Fed discretion to apply EPR requirements if it determines that the requirement is appropriate to prevent financial instability or promote safety and soundness.

The Endgame proposal is tailored in the sense that it entirely exempts banks with less than $100 billion in assets unless they are active in trading. But most of the provisions of the proposal apply uniformly to all banks with over $100 billion in assets, leading to criticisms that is not adequately tailored.⁵⁸ (Proposed changes to the SLR and the G-SIB surcharge, the latter of which was issued on the same day in a separate proposal, would leave those two provisions tailored differently for Category I banks compared to other large banks.) The largest bank the proposal applies to has over 30 times more assets than the smallest. Generally, Basel standards are meant to apply to internationally active banks, but international activity would no longer be used as a screening criterion.

It is unclear whether the proposal is reliant on any authority specific to Title I, as the regulators have broad authority to set capital requirements.⁵⁹ Nevertheless, regulators have repeatedly acknowledged that it is appropriate to tailor regulation generally. On the one hand, much of the added regulatory complexity of the proposal applies only if banks are participating in the affected activities. On the other hand, the case for the necessity of these changes for all of the banks subject to the rule is not clear. Proving that requirements do not apply to a bank can still impose compliance costs. A letter signed by the House Financial Services Committee chair and other majority Members on the committee argued that a lack of tailoring in this rule and others would lead to a "barbell" shaped banking system with no midsize banks.⁶⁰ Although the 2023 bank failures are invoked as a rationale for the lack of tailoring, outside of the changes to AOCI, the connection between Endgame provisions (which the BCBS issued in 2017) and the failures is unclear.

Transparency

On September 12, 2023, six major industry trade associations submitted a joint comment letter arguing that the proposal

would significantly increase capital requirements for larger banks. Yet in support of these substantial new requirements, the proposed rule repeatedly relies on data and analyses that the agencies have not made available to the public. This reliance on non-public information violates clear requirements under the Administrative Procedure Act that agencies must publicly disclose the data and analyses on which their rulemaking is based. To remedy this violation, the agencies must make available the various types of missing material identified below—along with any and all other evidence and analyses the agencies relied on in proposing the rule—and re-propose the rule.⁶¹

The letter refers to the nonpublic internal agency research that went into estimating the various parameters and formulas found throughout the proposed regulation.⁶² The proposal also contained an economic impact analysis (discussed in the section above entitled "Economic/Capital Impact") that critics argued was insufficient and flawed.⁶³ (On October 20, 2023, the Fed announced that it would collect more up-to-date and detailed data from the banks subject to the proposal to "further clarify the estimated effects of the proposal and inform any final rule."⁶⁴) While it is beyond the scope of this report to weigh in on the legal merits of the letter's claims, as context, the proposal takes up 316 pages of the Federal Register, with the first 156 pages a preamble that explains and justifies the proposal. The underlying documentation that the letter calls for would presumably be much longer still. By its nature, this underlying documentation would arguably be too technical to be useful or comprehensible to the general public, although it could be useful to industry and researchers.