On June 8, 2021, S&P Global Ratings published a request for comment (RFC) on its proposed criteria, "Request For Comment: Financial Institutions Rating Methodology."

Following feedback from market participants, we finalized and published our criteria, "Financial Institutions Rating Methodology" (FI criteria), on Dec. 9, 2021.

We'd like to thank investors, issuers, and other intermediaries who provided feedback. This RFC Process Summary provides an overview of the written comments and certain other feedback we received from the market on the proposed criteria, the changes we made following the RFC period, and the rationale for those changes.

Several of the responses that we received to this RFC included comments on the proposals in "Request For Comment: Banking Industry Country Risk Assessment Methodology And Assumptions," June 8, 2021. We address all comments relating to the BICRA proposals in "RFC Process Summary: Banking Industry Country Risk Assessment Methodology And Assumptions," Dec. 9, 2021.

In the RFC, we asked:

• What are your views on the methodology we have discussed in the article?
• Are there any other factors you believe we should consider in the proposed criteria?
• In your opinion, do the proposed criteria contain any significant redundancies or omissions?
• Is the structure of the methodology clear, and if not, why?
• Do you believe the framework places too much emphasis on any particular rating factor, and if so, do you believe this emphasis could be mitigated by the use of the adjustments?
• Do you believe we are appropriately capturing risk and agree with the manner in which we propose to assess this risk? If not, what alternative(s) would you propose?

Respondents addressed these questions and other issues, and several respondents expressed a general appreciation for the clarity and transparency of the criteria and specific enhancements to our approach.

Written Comments Received From Market Participants That Led To Significant Analytical Changes To The Final Criteria

We did not receive any external written comments from market participants that led to significant analytical changes to the criteria.

Written Comments Received From Market Participants That Did Not Lead To Significant Analytical Changes To The Final Criteria

Comments on the structure of the framework

Scope.  We received a comment expressing surprise at the timing and scope of the RFC, and disagreeing with the proposal to include nonbank financial institutions (NBFIs) in a single framework with banks. Another comment welcomed the consolidation of multiple criteria articles into one because they saw this as enhancing transparency. The RFC did not represent a recalibration of, or fundamental change to, our rating framework for FIs--rather our objective was consolidating 10 criteria articles (several of which were over seven years old) into one piece to enhance consistency and transparency.

While combining banks and NBFIs in one article leads to a longer article with some sections that are less relevant to specific types of FIs, the drivers of creditworthiness are similar for FIs, even allowing for differences in regulation and business models (differences that can also apply within the global banking sector). The analytical frameworks for banks and NBFIs therefore follow the same format and fit well together. We also consider that the bank and NBFI sectors are often interrelated--for example customer, product, and group interactions, and competitive dynamics and regulatory changes in one sector can affect other sectors.

We also received a comment suggesting that the criteria framework would be more comprehensive if we incorporated the methodology for FI hybrid capital instruments. We agree that this format could further consolidate the number of criteria articles relevant for an FI, but have decided to keep the criteria for FI hybrid capital instruments in the same hybrid criteria article as for other sectors. This is because the principles of our hybrid capital criteria are similar across sectors, and so we provide our methodology for assessing the credit characteristics of hybrid capital instruments as an asset class in a single criteria article ("Hybrid Capital Methodology And Assumptions," July 1, 2019).

The role of environmental, social, and governance (ESG) factors.  We received a comment appreciating the incorporation of ESG credit factors--specifically climate risk--more clearly in the RFC. However, the respondent also recommended that we should give climate risk, and how management teams are addressing it, greater prominence in our methodology (for example, by recognizing entities that become industry leaders in addressing climate change by enhancing business position, risk position, or our assessment of earnings). We agree that ESG credit factors, including climate risk, are becoming more visible as drivers of creditworthiness and that their impact on creditworthiness should be clear and transparent under our rating methodologies (as highlighted in "Environmental, Social, And Governance Principles In Credit Ratings," published Oct. 10, 2021). We believe the criteria provide an effective framework for incorporating the influence of ESG credit factors on creditworthiness for FIs, so we have not made changes to reflect this comment.

We confirm that ESG credit factors can influence our assessments of business position under our FI criteria. For example, within the business position section, we:

• Cite effective management of risks related to climate-related developments as favorable for our view of management, governance, and strategy;
• Mention a business model or practices that create opportunities related to climate-related factors as a potential benefit for our view of business stability; and
• Refer to how our view of diversification could be more successful if an FI develops expertise and becomes an industry leader in a climate change or transition-related niche that strengthens revenue stability.

These enhancements to business position can also lead to a more positive view of earnings quality and stability. We also confirm that vulnerability to climate transition risk and physical risk factors can affect our view of risk position, for example through our assessment of concentrations and of risks not captured within our risk-adjusted capital (RAC) ratio.

The respondent also suggested that we could incorporate banks' public commitments to the Paris Climate Agreement, as well as monitor their track record toward achieving net zero greenhouse gas emissions goals in operations, supply chain, and financing activities. We agree that these are relevant indicators of how an FI is addressing climate transition risk factors but have not incorporated them as specific metrics in the criteria. The criteria establish our principles for assessing creditworthiness for FIs globally, but we do not list all potentially insightful metrics, especially when their correlation with creditworthiness can vary. We also publish complementary research and commentaries to provide greater transparency on how sectors, including FI, are addressing ESG credit factor-related issues.

Another comment suggested that good corporate governance be added as an assessment factor in the ratings framework. We agree with the respondent's observation that implementing good corporate governance affects FI's risk exposure. We have not added an additional factor though because governance is already part of our assessment of business position. In particular, governance, management, and strategy is now the first subfactor that we consider when assessing business position.

The role of qualitative adjustments and analytical judgement.  One respondent expressed concern that the potential for qualitative adjustments could make it more difficult for market participants to clearly understand why each rating factor is scored as it is, and what actions or events could lead to a change in the assessment of that factor and, by extension, the ratings. In particular, the respondent was concerned that the combination of a potential adjustment to the capital and earnings assessment--the most quantitative of the stand-alone credit profile (SACP) factors--and a potential comparable ratings analysis (CRA) adjustment could lead to more subjectivity and less transparency in ratings. Another respondent welcomed the capacity to provide more flexibility and granularity to analytical judgment on qualitative considerations.

We agree that transparency and consistency are key, and we expect that our rating reports will explain why we've made an adjustment to arrive at the final capital and earnings assessment and the rationale for any CRA adjustment. We also expect to clearly state the impact of these adjustments--if any--on the rating. In accordance with our criteria, the capital and earnings assessment will typically depend on the initial score (based on the specific capital or leverage metric). We expect to use the CRA adjustment in a minority of cases. Although, it could be used more frequently than the "setting the ICR" adjustment that took place at the issuer credit level (ICR) level under the previous bank criteria, which was used in a low proportion of cases. We will continue to support analytical consistency through our committee and analytical oversight procedures.

A respondent supported applying judgment, where appropriate, to arrive at a forward-looking assessment of financial condition and performance so as not to rely too heavily on backward-looking data about risk appetite, earnings, or profitability. Another comment highlighted that the degree to which an FI received state aid in past crises may be relevant. We agree with both views. Our scoring is forward looking and considers how an FI has responded to or dealt with past events.

Another comment welcomed the increased capacity to choose an appropriate peer group for an FI--rather than, for example, referring to entities in the same geography that may not be useful peers--to enhance transparency about the rating drivers. Given the capacity for judgement about the makeup of the peer group, the respondent suggested that our reports disclose the specific peer groups used for an FI, including by each factor (for example, an FI might be a peer for business position but not for funding and liquidity). We agree that transparency about peer groups is important and will include this in our full analyses as relevant.

Specificity of the descriptors and the use of subfactors.  One comment discussed how the descriptors for the factor scores appear broad and not, in the respondent's view, very useful in understanding the underlying considerations, while another comment wanted to see subfactor scores reintroduced because their removal could reduce the ability of an FI to challenge our ratings opinion. The respondent felt they would have less visibility on how the assessments are determined. We agree that transparency about the underlying considerations of our assessment is important, and we'll continue to provide, in our rating reports, explanations on how we determine our assessments.

Even though we will not assign subfactor scores, we anticipate that we will describe our views on the strengths or weaknesses and organize our comments around the subfactor themes as relevant. We have decided not to use subfactor scores because the factor score will take into account the interplay between the subfactors and focus on the subfactors that drive the overall factor assessment. We also believe that this can help communicate the rationale more clearly by focusing on the characteristics that affect an FI's creditworthiness and not on characteristics that are less relevant.

Anchor for banks and NBFIs.  We received two comments recommending that the anchor be adjusted to reflect regional differences in economic or industry risk within a country. We agree that regional differences can be an important influence on the creditworthiness of an FI but have decided to include this in the scoring of the FI-specific factors. We assign anchors at the country level and do not have regional anchors within a country, except for specific cases such as Hong Kong (where that territory has separate legal, regulatory, and reporting frameworks from other parts of China).

When the countrywide anchor does not represent the risks for that FI, whether due to economic or industry risk characteristics, we use the entity-specific scores. In our new criteria, we have, for example, increased the granularity of the scoring for several factors versus the previous criteria. We have also made edits to the descriptors for the categories that enhance the capacity to incorporate in-country differences.

We assess business position and risk position each on a relative, rather than an absolute, basis. For example, a bank may have a good franchise in a part of a country but have a weaker business position score because of its heavy concentration or because where it operates has less attractive aspects as a place in which to do business. If the region has better operating conditions than the country as a whole, this could boost the business position score for a bank, if the industry risk score sufficiently overestimates the market challenges for that bank. Similarly, a bank could have a weaker business position score if the industry risk score is boosted by regulatory approaches that do not apply to that type of bank. In the same way, a bank will have a weaker risk position score if the economic risk score for that country underrepresents the risks in that part of the country in its home region (and vice versa). The new criteria do not limit the proportion of banks or FIs that can be assigned a given score for a category (for example, the proportion of banks that can be scored as strong or very strong risk position).

Another comment expressed concern that setting the preliminary finance company anchor always three notches below the bank anchor in the same country means that lower-risk NBFIs cannot be rated at the same level as their banking peers. We agree that lower-risk NBFIs should be able to be rated at the same levels as their banking peers and therefore have made several edits to the criteria to clarify the role of the country/sector-specific adjustment. We believe the final criteria article more clearly states that we can use this adjustment to move the anchor for an NBFI to the same level as the bank anchor when that NBFI sector does not face higher risks than the banking sector in the country. We also agree with the respondent that there are many different types of NBFIs and that some have greater or lower risk characteristics than banks.

In response to this feedback, we added references to how we can make the NBFI country/sector-specific adjustment at the subsector level. This change makes the adjustment applicable at the subsector level if an NBFI subsector within a country has elements in common that are markedly different from the wider sector in that country. This allows for a more granular anchor adjustment, but because it's a refinement, we don't consider it a substantive change to our analytical approach.

An NBFI anchor cannot be higher than the bank anchor for the same country. But, we can use our FI-specific factors to assign an NBFI a rating higher than that of a bank in the same country--for example, due to positive assessments of its capitalization, funding, business model, and risk position.

Comments on FI-specific factors

Business position.  We received a comment noting that diversification of business activities does not equate to lower risk because that depends on the quality of the governance and enterprise risk management of an FI. We agree that a well-managed entity that focuses on and has deep expertise in a particular product set should not automatically be considered higher risk or subject to greater revenue volatility than a more diversified entity. We haven't made any changes to the criteria, however, because we believe that the criteria appropriately emphasize the importance of governance and risk management of an enterprise and its business lines and because the business position assessment is not constrained by a headline view of the level of diversification.

Governance, management, and strategy is the first subfactor that we look at when considering business position, and it may reinforce or weaken or view of the other aspects for an FI, such that weaker capacity to manage diversification will affect the business position assessment. We also agree with the respondent's observation that the criteria consider that diversification for its own sake may be a weakness--for example, if an FI weakens its overall business position by entering new products and countries where it has limited expertise and lacks critical mass to be a real competitor to the incumbent market leaders, or where the new areas are riskier than its traditional business.

Another comment asked for more specificity about how we measure revenue stability. In response, we edited our definition of this concept in the glossary to clarify that we are assessing the volatility and predictability of overall revenues and its components. There is no single measure or narrow list of metrics because actual composition of revenues differs by entity so that breakdowns of revenue may be more insightful than aggregate data depending on the entity. Our assessment is forward looking, but informed by past experience where relevant.

Capital and earnings.  We received several comments on capital and earnings, focusing on:

• The role of regulatory capital measures and metrics other than the risk-adjusted capital (RAC) ratio in the capital and earnings assessments;
• The ability to make an adjustment to the initial capital and earnings score, and the implications of doing so;
• RAC outcome sensitivity to BICRA scores;
• Temporary capital surpluses;
• Debt-to-EBITDA assessments for securities firms; and
• Specific considerations for BDCs.

Several comments recommended that we expand the role of regulatory capital measures in the capital and earnings assessment, making them either explicit alternatives to RAC or more specific complementary measures or secondary metrics that drive the assessment. One comment suggested removing RAC entirely, while another was skeptical about its capacity to differentiate risk insightfully. Most comments welcomed the wider ability to make a qualitative adjustment to the initial capital and earnings score, although several comments raised concerns about transparency and potential incentives for entities to weaken their capitalization.

We agree with the comment that forthcoming enhancements to bank regulatory capital standards may provide greater comparability of risk-weighted assets across jurisdictions, and that enhanced consistency could obviate the need for our RAC analysis. We continue to believe that comparability constraints mean that it would be premature to stop using the RAC ratio in this global criteria article but agree that regulatory capital metrics can have a wider role in determining the final capital and earnings score than in the previous criteria. (We discuss the comparability issues in more detail in "The Basel Capital Compromise For Banks: Better Buffers, Elusive Comparability," June 3, 2021, and "Basel III Bank Capital Rules In Europe: Delayed And Diluted," Oct. 28, 2021.)

In line with the RFC proposals, we have wider capacity to make an adjustment to the initial capital and earnings score to reflect the information provided by indicators such as regulatory capital outcomes. We have decided not to introduce specific scoring ranges for regulatory capital metrics (for example, deciding not to state that a certain regulatory ratio level would, in a specific country, lead to a positive or negative adjustment) given their remaining inconsistencies. But, we have edited the text to make it clearer that other leverage and capital metrics can be considered in the adjustment to the initial score.

Another comment raised concerns that a bank with significant regulatory capital excess over regulatory minimums, including under more conservative measures, and that has good capital outcomes under regulatory stress tests, could be penalized under the RAC ratio and we would consider it to have scarce capital. Another comment discussed a case where an FI has access to resources that are not included in the RAC ratio but that would provide financial flexibility in a stress scenario. We agree that indicators other than the RAC outcome can be insightful, and therefore the adjustment to the capital and earnings score can take into account such factors.

One comment expressed concerns about the sensitivity of the RAC ratio to BICRA scores, noting that the impact of a higher-risk economic risk score (within the BICRA analysis) makes it very difficult for a bank in such a country to meet a RAC threshold of 10% or more, even if the bank's regulatory capital ratios and track record show resilience and strength. The respondent saw BICRA-determined risk weights as overly penalizing at times, compared with regulatory norms, and potentially inducing volatility in the rating when BICRAs change. Another comment considered that the RAC model inputs and risk charges may be overly influenced by experiences in regions such as North America and Western Europe.

We agree that RAC risk weights may not reflect fully the position of a specific FI and therefore now have greater capacity than under the previous criteria to incorporate other information in the capital and earnings assessment. We also continue to use the risk position assessment to refine our view of an FI's risks beyond the capital and earnings analysis.

The criteria do not place any limits on the proportion of FIs that can be assessed as strong or very strong for capital and earnings and risk position, and we highlight that risk position is a relative assessment. For example, we state in table 12 that a risk position assessment of very strong indicates we believe that the FI-specific risks or attributes mean that it is materially better placed to withstand economic stress than the capital and earnings assessment indicates. Very strong for risk position does not mean that we consider the FI to be one of the lowest-risk FIs that we rate. In line with the proposals, we calibrate the impact of a capital and earnings assessment on the SACP for banks with higher-risk BICRAs. Table 3 shows how we reduce the negative impact of a weaker capital and earnings assessment on the SACP (and, in some cases, raise the positive impact) when the bank anchor is 'bb+' or lower.

Although we received several comments welcoming the ability to use the adjustment to the initial capital and earnings score more widely, including some in strong support, two respondents liked the clarity of the previous capital and earnings scoring thresholds, which they thought helped FI management teams be more confident about the ratings impact of significant decisions. These respondents recommended that our forward-looking view of capital--both the expected capital ratio and how we come to that evaluation--be transparent and visible, and recommended that banks should not have a weaker incentive to boost capital. We agree that our scoring rationale should be transparent, and that the impact of a particular capital level or strategy on the capital and earnings score should be clear.

We stress that more lax capital management could affect our forward-looking view of capital for an entity. The potential for a negative adjustment to the initial capital and earnings score, in our view, creates an incentive for an issuer to maintain capital ratios that are comfortably within a particular category. Also, the limited number of points on the assessment scale does mean that some FIs will have the same capital and earnings score as an FI that has lower capitalization, although the criteria enable us to differentiate where these differences are material.

Two comments expressed concerns about whether the adjustment to the capital and earnings score would often be used in practice, and one of these asked whether there was a de facto limit on how close to a threshold an FI would need to be for an adjustment to be made. We confirm that we use the adjustment when, in our view, it leads to a capital and earnings score that better reflects our view of whether (and the degree to which) capital and earnings are positive, neutral, or a weakness to the SACP. However, we state in the criteria that the metric used for the initial score will typically be the main factor determining the capital and earnings assessment. We also recognize that the adjustment should be used consistently. There is no de facto limit on how close an FI needs to be to a scoring threshold for us to make an adjustment, although we expect that--all else equal--an adjustment may be more likely when the FI is closer to a threshold because the additional information may be more persuasive to move the FI into a different category.

One of these respondents also asked whether adjustments to the capital and earnings score would always be offset by a change in the risk position score under the new criteria. We believe that in some cases an FI may now qualify for an adjustment to the capital and earnings score that leads to an offsetting change in the risk position score, but this will not always be the case because our risk position assessment incorporates a wider variety of considerations.

A respondent strongly supported the role of the adjustment to the initial score so that capital and earnings assessments could focus more on earnings generation, capital flexibility, earnings quality and consistency, and the ability to build capital under actual or hypothetical adverse scenarios. We agree that these are relevant considerations that could lead to an adjustment but have not made edits to the text because we believe that these are already addressed in the principles laid out in the criteria. Another comment highlighted that ownership considerations can lead to an FI targeting regular, rather than profit-maximizing, earnings levels. We confirm that our analysis looks at risk-adjusted earnings and not just headline ratios.

One comment suggested that a debt-to-adjusted equity metric could be appropriate for assessing capital and earnings for finance companies and leasing companies and suggested several other metrics to allow multiple quantitative metrics to collectively influence the final capital and earnings assessment. We agree that such metrics can be useful, and they can be assessed as part of the adjustment to the initial capital and earnings score, with some metrics more insightful for certain FIs than others. We have edited the text about this adjustment to highlight that we can consider a variety of leverage and capital metrics (regulatory and otherwise).

One respondent expressed concern that temporary excess capital may not be reflected in ratings, especially if it is not included in our measure of additional loss-absorbing capacity (ALAC). As was the case in the previous criteria, we don't place as much weight on capital or ALAC resources that we don't expect to be maintained on the balance sheet. However, we take a prospective view of both capital and ALAC and therefore can take into account future issuance and capital management plans, and their credibility and flexibility.

Another comment suggested that we incorporate return on tangible common equity (ROTCE) in our analysis of earnings within the capital and earnings assessment. We agree with the respondent's comment on the analytical advantages of looking at tangible common equity metrics. We can use this metric in our analysis of earnings, although earnings metrics that use equity denominators (such as return on equity and ROTCE) can be flattered when equity levels decline. For this reason, we look at other metrics as well.

One comment suggested that a debt-to-EBITDA ratio should also be used for the initial capital and earnings score for securities firms that have substantial asset management businesses, off-balance-sheet assets, or limited balance sheet risk. We have not changed our proposals based on this comment because we typically see securities firms as bearing balance sheet risks that favor a capital-based approach over a cash flow-based approach. However, we can look at cash flow-based metrics when considering whether to make an adjustment to the initial capital and earnings score, and we can make an adjustment if we consider that the RAC ratio overestimates risks--such as operational risks--associated with the asset management/off-balance-sheet portion of the business.

Another reason for not making a change to the criteria article is that we can already use a different approach for securities firms that operate a mixed business model with asset management activities. Our asset manager criteria use debt-to-EBITDA (cash flow-based) metrics, and we can therefore use a mixed-group analysis (where we derive separate SACPs for the securities firm and asset management businesses) when we consider the asset management activities to be substantial.

Another respondent asked if we cap the capital and earnings score at adequate for securities firms where we use a debt-to-EBITDA metric. We confirm that we do not cap the final capital and earnings score for such securities firms at adequate. Table 10 enables us to differentiate better between certain securities firms that have scored moderate or lower under table 9, potentially leading to a higher initial score than under table 9. We can also use an adjustment to the initial score to arrive at a higher final score where appropriate.

We also received a comment suggesting refinements to the capital and earnings assessment for BDCs due to the impact of the U.S. Small Business Credit Availability Act of 2018 that permits a BDC to adopt different leverage and asset coverage levels. Based on this feedback and our review of the impact of the legislation, we have made several changes from the RFC proposals, and we note that the BDC anchor has been in line with the preliminary U.S. finance company anchor since 2018 as a result of the changes arising from that legislation.

We now assess capital and earnings for a BDC in the same manner as for other finance companies, using table 11 to determine the initial capital and earnings score for a BDC. The adjustment to arrive at the final score includes the impact of earnings capacity and quality, which we assess in the same manner as finance companies with some additional considerations. We continue to consider three earnings metrics specific to BDCs (realized return on average portfolio investments, non-deal-dependent income interest coverage, and non-deal-dependent income coverage of both interest and dividends) and other earnings metrics considered for finance companies. We can also make an entity-specific adjustment to the anchor for a BDC to reflect that it has adopted a particular regulatory approach under the 2018 Act. We don't consider these significant changes to our analytical approach, given that they refine, rather than alter, our fundamental approach to looking at a BDC's capital and earnings.

Risk position.  Two comments asked for more specificity about the quantifiable risk measures that we use to assess risk position, including those used to assess complexity, and for more detail on how we assess market risk on a quantitative basis for larger banks and global systemically important banks. We added wording to the risk appetite section to highlight the use of risk measures other than those relating to credit risk. Paragraph 146 has a non-exhaustive list of factors that we consider when assessing complexity. We also added language to the complexity section regarding other metrics that we use to assess market risk. We confirm that we include quantitative market risk charges within the RAC ratio (see "Risk-Adjusted Capital Framework Methodology," July 20, 2017, for more details).

Another respondent asked that we reinstate the enterprise risk management section that was in the previous criteria. We agree that enterprise risk management considerations are key to how an FI manages its risks but have decided to have those considerations integrated into the main text that discusses the potential risks (taking into account the risk management approaches alongside the main causes of risk) rather than shown as separate considerations.

One respondent raised comments on the calibration of our RAC charges for operational risk and expressed concerns that they overestimate the potential regulatory charges or other operational loss payments that an NBFI would incur. We agree that our specific RAC charges may not always reflect the risks that would be borne by a specific FI. Where relevant, this can be addressed through an adjustment to our initial capital and earnings score or through the risk position assessment.

We received one comment suggesting that asset quality should play a greater role in determining FI ICRs. We believe that the new criteria enhance the capacity for the risk position assessment to reflect asset quality, given the refinements we have made to the text (for example, we state that we use risk position to assess factors other than those incorporated in the capital and earnings adjustment, such as asset quality) and because we can now use the capital and earnings adjustment to capture some items that were previously in risk position.

Funding and liquidity.  One comment asked whether the proposed changes in the funding and liquidity section will generally have a neutral impact on bank SACPs and ratings. We confirm that this is generally the case, although some changes may occur. Another felt that funding and liquidity would rarely have a positive impact on bank ratings under our approach. Our criteria framework allows for funding and liquidity to have a positive impact on the SACP, based on relative strength versus peers (with systemwide improvements reflected in the anchor through our BICRA analysis).

One comment suggested refining the text on deposit stability given observations on the stability of deposits from certain distribution channels. We made several edits to the text on deposit base stability in response to this feedback, mainly to highlight that this is a granular assessment instead of one based on preconceptions of how a particular segment may perform. Another comment welcomed our increased focus on the stability of funding sources instead of a predefined preference for specific types of liabilities. We also confirm that we do not automatically treat all wholesale funding as less reliable in our assessment, and that we look at the composition, maturities, and concentrations of wholesale funding to arrive at our view of funding risk.

One comment discussed the role of metrics like broad liquid assets to short-term funding (BLAST) and the stable funding ratio (SFR) and their limitations for assessing bank funding and liquidity. We agree that these have limitations and note that they are not mechanistic determinants of our assessments. We also agree with the two comments that regulatory funding and liquidity metrics such as liquidity coverage ratio (LCR) and net stable funding ratio (NSFR) are insightful for banks. However, we have decided not to base our assessments specifically on these regulatory measures, as recommended by the comments, because they are not available globally and may have some inconsistencies. We expect them to be increasingly useful metrics within our assessments as supporting disclosure continues to improve.

A respondent asked for us to provide estimated SFR ranges for a finance company or BDC to receive a funding assessment of moderate and weak. We have decided not to do so because we expect other characteristics, such as the funding of long-term illiquid assets and access to funding markets, to be the main determinants for an assessment in these categories.

Potential comparable ratings analysis (CRA) adjustment.  Three commenters raised questions about the potential subjectivity and visibility of the CRA adjustment, although another comment welcomed the incorporation of this adjustment. We agree with the comments that transparency about the reasons for a CRA adjustment, and the composition of the peer group used for the peer analysis supporting the decision, is important to demonstrate that we use it consistently.

One of these comments asked whether we would use the CRA adjustment rarely or more frequently, while another of these asked for an estimate of the percentage of cases in which it would be used. We don't place a predetermined upper or lower limit on how often the adjustment should or could be used. As stated earlier, we expect to use it in a minority of cases--given that we use it to incorporate additional credit factors that the criteria do not separately identify or other factors that are not fully captured in the other SACP factors, and we believe that the criteria framework addresses credit factors comprehensively. However, we expect that it could be used more frequently than the "setting the ICR" adjustment was under the previous criteria.

Another comment asked whether the peer group for the CRA adjustment would be domestic or global. Table 4 confirms that we use peers with the same or a similar SACP (one notch higher or lower) and that the comparison can include entities in other countries or FI sectors--for example, when there are not enough domestic banks or NBFIs or when similarities exist with banks or NBFIs in other countries.

Comments on extraordinary external support

External support from governments.  One respondent expressed concern that the proposals put increased emphasis on potential ratings uplift from government support and that this diminishes banks' motivations to improve their business fundamentals, while noting that many governments are establishing resolution frameworks to deal with distressed banks. The respondent also considered that the matrix approach in tables 21-23 (the FI government support tables) represented a change in our approach to assigning bank ratings.

We agree that many governments are implementing resolution frameworks to move banking systems from a "bail-out" to "bail-in" world. As a result, the number of countries that we now classify as support uncertain, from a government support perspective, has increased (from 2011, when we published our previous bank criteria and introduced the government support tables), and we include ALAC to reflect extraordinary support that would be provided by bailing in various bank liabilities. The criteria do not significantly alter the potential to incorporate government support in bank ratings relative to the previous criteria. The main change is to enable us to consider both GRE status and systemic importance in certain cases, but we have not altered our criteria for assessing potential uplift from either of these two items. The ability to make a government support-related adjustment at the ICR-level replaces the previous "setting the ICR" adjustment such that it will make rationales for using this adjustment more transparent.

The respondent suggested that we include government support in ratings only if, for example, an FI's SACP is below 'bbb-', to prevent an FI being less motivated to improve its business fundamentals. Our ratings are opinions on relative default risk, and we do not regulate or direct the behavior of management teams. Our ratings are based on credit factors that drive default risk, and we observe that the potential for government support can be a relevant factor in relative default risk even when an SACP is higher than 'bbb-'. For example, if two banks have SACPs of 'bbb' but one has a sufficiently lower risk of default because of potential government support, then the ICR on that bank is higher than the ICR on the other bank.

We have decided to maintain the approach in tables 21-23 given that they do not represent a change in our analytical approach, have performed clearly, and reflect potential government support as a demonstrated driver of differences in default risk. The matrix approach is also consistent with the approach that we take for government support considerations in "Rating Government-Related Entities: Methodology And Assumptions," March 25, 2015 (our GRE criteria). We do not use the FI government support tables to determine ratings in countries that we classify as support uncertain but we do use them to determine ratings in countries where the likelihood of extraordinary government support is still a key driver of default risk.

The comment also expressed concern that government support could be included in the potential CRA adjustment. This would only potentially be the case for ongoing government support that the FI is currently receiving, which we include in the SACPs for banks globally (usually this benefits the funding and liquidity assessments). For example, banks globally have benefited from ongoing funding and liquidity support from governments and central banks throughout the pandemic, as well as regulatory capital accommodations. The potential for extraordinary future government support is not included in the CRA adjustment because this is not part of our SACP assessment.

Another comment suggested that we should be able to incorporate more government support in the ratings on systemically important banks, noting that the maximum uplift under the FI government support tables is less than that possible under GRE criteria and that the respondent believed the likelihood of government support could be stronger in some countries given the focus on financial stability. We agree that some governments may have even stronger willingness to provide support to a specific FI than the FI government support tables indicate, but we can reflect this either by assigning a higher GRE status to that FI (noting that government ownership is not a necessary condition for GRE status) or could use the government support-related adjustment in some cases.

Additional loss-absorbing capacity (ALAC).  Three comments disagreed with the proposal to remove excess total adjusted capital (TAC) from ALAC and adjust the ALAC thresholds. One focused on the impact this could have for some existing ratings. The other two discussed the potential for:

• Banks to be incentivized to switch to weaker quality capital or maintain lower amounts of ALAC;
• Good quality capital to be overlooked in our analysis;
• "Cliff effects," whereby a certain amount of capital may or may not benefit a rating, depending on the capital and earnings score; and
• Going- and gone-concern capital may not always be clearly demarcated in practice.

We agree that our rating approach should recognize the reduced default risk when a bank has a higher level of good quality capital resources. Under the previous criteria, this excess TAC (the amount that the bank had in excess of the amount needed to achieve its capital and earnings score) could only be recognized in the SACP in the very limited cases where we could raise the initial capital and earnings score. We have decided to maintain the proposal to exclude excess TAC from ALAC because under the new criteria, we can consider this excess amount a driver of a positive adjustment to the capital and earnings score (or to prevent a negative adjustment). In some cases, the capital policy of a bank that maintains higher-quality capital ratios could also influence other aspects of the SACP analysis, including the potential for a positive CRA adjustment to the SACP. In these ways, the excess TAC could benefit the SACP of the bank, which was not the case under the previous criteria, and therefore would not be ignored. Not all banks will see an improvement in their SACPs due to this change in the criteria, but we think that the SACP is the best place, on balance, to incorporate capital strength.

We also agree with the comment that the line between a regulatory determination that a bank still has going-concern status versus gone-concern status could, in a stress scenario, be a grey zone. We reflect this in the criteria in two ways. First, we can make an adjustment to the ALAC uplift thresholds when we consider that the headline quantitative metrics may over- or underestimate an FI's loss-absorbing capacity, for example if we consider that the amount of ALAC included in the numerator sufficiently underestimates the actual loss-absorbing capacity. Second, when we consider a resolution may be a near-term outcome, we can use an additional support adjustment to indicate the specific recapitalization that we expect to take place based on the bail-in of on-balance-sheet liabilities.

We acknowledge the concern regarding potential "cliff effects" as a bank's metrics move it from one scoring category for capital and earnings or ALAC into another. We have decided to maintain the RFC proposals because we believe that the new more holistic approach toward assessing capital and earnings, as well as the role of the ALAC thresholds, reduces the potential for "cliff effects." The previous criteria, in our view, had greater capacity for a strengthening balance sheet not to lead to a higher rating because it involved elements of TAC moving into one category (capital and earnings) from another (ALAC) based on more mechanistic scoring thresholds. We also score both capital and earnings and ALAC on a forward-looking basis.

We believe that another comment did not consider the increased capacity we had proposed to modify the ALAC thresholds would enhance clarity or consistency. We have maintained the capacity to make these adjustments in cases where the headline metrics in our view over- or underestimate the loss-absorbing capacity of the bank but acknowledge the importance of transparency. Our reports will be clear about the existence and size of any adjustment, as well as the rationale, and we will continue to use our committee and analytical oversight processes to support analytical consistency.

A respondent recommended that we alter our approach of reserving two notches of ALAC uplift for FIs with SACPs of 'a-' or higher. We maintain the limits on ALAC uplift at higher points in the rating scale. This is because the ratings express differentiated default risk. At such SACPs, banks are far from a potential point of nonviability or entry into a resolution process, so larger ALAC buffers have a diminishing benefit on the issuer's likelihood of default. For similar reasons, we typically compress other forms of external support ratings uplift at higher SACP levels.

This links with another comment that asked what the benefit to a bank would be of holding excess ALAC over the amount that would lead to a rating uplift. We expect that a bank will make decisions about the level of loss-absorbing capacity instruments to hold with many stakeholders in mind, including regulatory requirements. Higher levels of on-balance-sheet ALAC may reinforce ALAC ratios in times of stress or reduced market access.

One respondent asked whether, when determining the remaining maturity of ALAC instruments, we will look at the legal maturity date if the regulator has an explicit or implicit right to block early redemptions. We confirm that, under both the existing and new criteria, if an FI has the option to redeem an instrument earlier than the maturity date, we treat the legal maturity as the effective maturity if the regulator has oversight giving it the ability to prevent redemption at the call date, assuming that the instrument does not have a step-up.

We received one comment on our approach to assessing ALAC for groups that would go through a multiple-point-of-entry (MPE) resolution. While the respondent felt that the proposals treated such MPE groups in a more accurate way than the previous criteria, such as through the treatment of resolution perimeters, they still felt the criteria were too vague, could penalize some groups unnecessarily, and should focus more on a solo analysis that also included the benefits of excess TAC held within the group.

We agree that it is important to include the detail of how such a group would be approached within a resolution, and therefore have retained capacity in the criteria to reflect the idiosyncrasies of MPE groups and of the resolution solutions that resolution authorities can apply to them. We can do this by using consolidated or solo approaches and vary the entity or subgroup level that we use for our assessment depending on the group structure and resolution approach, and using a threshold adjustment where relevant. We expect that our reports on such groups will be transparent about our assumptions regarding the resolution strategy and the details of our approach. We have added a reference to subgroups to the section on MPE to highlight how we assess these groups at the relevant levels for their resolution models.

Two respondents welcomed that we would no longer exclude from ALAC instruments maturing between one and two years that exceed 0.5% of expected S&P Global Ratings risk-weighted assets (S&P Global RWAs).

Comments received on issue credit ratings

NBFI issue credit ratings.  We received one written comment requesting that we review our notching approach for NBFI issue credit ratings, which it saw as inconsistent with our approach for corporate issuers, both in cases where we notch down and because we do not notch up. The headline percentage thresholds of priority debt to adjusted assets at which we deduct one or two notches from the ICR to arrive at the issue credit ratings for senior unsecured debt differ materially for NBFI and corporate issuers. We have not changed the headline thresholds because we look at corporate balance sheets differently from those of banks and bank-like NBFIs (that is, those that are in scope of the FI criteria).

Our FI analytical framework focuses on issuers with balance sheet risks, rather than risks related to cash flows (which is the corporate approach). These FI entities typically are more confidence-sensitive than corporate issuers, and we observe that their balance sheets can change materially in times of stress. We typically expect that liabilities will be more subordinated during periods of stress than near- to medium-term financial metrics imply, and that the level of unencumbered assets could also fall significantly from what near- to medium-term metrics imply. We therefore have a different approach to contractual subordination for FI than for corporate issuers, as well as for situations where an NBFI senior bond is, in effect, subordinated due to the existence of priority debt. Our notching for NBFI issue credit ratings does not reflect the expected loss given default or expected recovery on an obligation but instead is an indicator of subordination. We continue to monitor how FI balance sheets change as an FI experiences distress and to consider what this means for our approach to notching for contractual and effective subordination in FI.

We have, however, made some edits to the final criteria to provide more detail on how we deduct such notches for subordination. We made a change to scenario B of the NBFI-specific subsection of the "Issue Credit Ratings On Instruments Other Than Hybrid Capital" section. This shows when we would deduct one notch, instead of two notches, under this scenario. This is not, in our view, a significant change to our overall framework because it builds on the RFC proposals for when a calculation is close to a threshold. We added text to highlight the prospective view that we take regarding the notching thresholds. We also added text indicating that this notching is not about the likelihood of default, to make the mapping onto national scale ratings clearer.

Comments received on in-use criteria

We received several comments on criteria that were not subject to this RFC or the BICRA RFC. We are treating these as written comments received on in-use criteria, which means that they will be addressed separately to this RFC process in accordance with our policies.

Hybrids.  We received one comment recommending that we use the SACP and not the ICR as the starting point for our hybrid notching when a bank has an ICR that is lower than its SACP due to sovereign rating-related constraints on its SACP.

We received another comment recommending that the ICR be the starting point for our hybrid notching when the issuer is a GRE that does not have a GRE status of very high or above.

We also received a comment asking for more explicit criteria on how we would rate TLAC securities. We confirm that the criteria for TLAC securities is in "Hybrid Capital Methodology And Assumptions." Table 2 and the associated text of that criteria describe how we arrive at a rating for a bank hybrid that is not included in Tier 1 or Tier 2 capital. TLAC securities are hybrids when they can absorb losses without causing a default on senior obligations.

Another comment suggested that we treat the legal maturity as the effective maturity for bank Tier 1 hybrid capital instruments when the regulator has to approve early redemptions.

RAC.  We have also flagged the comments on RAC operational risk charges for NBFIs, and on the overall calibration and transparency of the RAC model, as comments on in-use criteria.

Significant Analytical Changes To The Final Criteria That Did Not Arise From Market Feedback

We finalized and published the criteria without making any significant analytical changes that are unrelated to the external written comments or other market feedback we received.

However, we changed the earnings buffer metric to a pretax basis because the proposed RFC calculation mixed pretax and posttax inputs in the numerator of the calculation. Core earnings is a posttax number, but other elements of the proposed calculation were pretax. The revised calculation replaces core earnings (a posttax number) with preprovision operating income (pretax).

We don't consider this to be a substantive change to our approach because this revised calculation remains consistent with the purpose of the earnings buffer as described in the RFC and the final criteria: "We may use the earnings buffer to measure the capacity for earnings to cover normalized losses as part of our assessment of the quality of earnings and our expected ratios." In the final criteria, core earnings continue to be an important consideration when assessing whether an adjustment to the initial capital and earnings score is appropriate. As proposed in the RFC, we have aligned the earnings buffer metric for securities firms with that of other FI, but we have done so on a pretax rather than posttax basis.

We incorporated other editorial revisions, where appropriate, to improve the readability and clarify the intent of the criteria. For example, in addition to several edits to make the text read more clearly, we removed duplicate text (such as in table 16) and moved what were tables 28 and 29 in the RFC so that they're alongside other similar tables in the "Metrics Definitions" section of the appendix. These revisions are not significant analytical changes.

We also provided the following clarifications, none of which alters the analytical approach significantly on an individual or cumulative basis:

• We edited the text to make it clearer that a positive CRA adjustment cannot be used to raise an SACP above an SACP cap applicable to that FI based on its regulatory capital assessment.
• We clarified the treatment of maturities of exactly one year in our definitions of broad liquid assets, stable funding needs, and gross stable funding needs for securities firms, and in tables 28, 29, 31, and 32.
• In addition to the changes to the capital and earnings approach for BDCs, we moved language relating to BDCs into the body of the criteria, from the appendix where it was included in the RFC. We have maintained separate sections on BDCs when we adopt a different approach than for other finance companies (such as when we look at the regulatory capital requirements for BDCs). We flag in the section headings when we use the same sections for finance companies and BDCs.
• We clarified that we may include exposures to affiliates when considering a BDC's exposure to a single obligor.
• We removed a reference that implied the frequency with which specific factors could lead to an NBFI entity-specific adjustment because we do not limit the ability to make adjustments based on a hypothetical view of how often a factor may be relevant. We also added an example to this section to clarify the interaction with the country/sector-specific adjustment that can be made at the subsector level.
• We clarified the criteria that apply to entities that have banking licenses but that are in scope of other criteria, such as "Public-Sector Funding Agencies: Methodology And Assumptions," May 22, 2018.

Related Criteria

• Financial Institutions Rating Methodology, Dec. 9, 2021
• Banking Industry Country Risk Assessment Methodology And Assumptions, Dec. 9, 2021