S&P Global Ratings has published a request for comment (RFC) on proposed changes to its rating methodology for insurers' risk-based capital adequacy (see "Request For Comment: Insurer Risk-Based Capital Adequacy--Methodology And Assumptions," published Dec. 6, 2021).

The main reasons for the proposed changes to our criteria include:

• Incorporating recent data and experience since our last update of the insurance capital model criteria;
• Enhancing global consistency in our risk-based capital (RBC) analysis;
• Increasing risk differentiation in capital requirements where relevant and material to our analysis, as well as reducing complexity where it does not add analytical value;
• Improving the transparency and usability of our methodology, such as with our proposal to supersede 10 related criteria articles; and
• Supporting our ability to respond to changes in macroeconomic and market conditions by introducing sector and industry variables.

A list of changes to our proposed criteria appears in the RFC and in the appendix of this article.

Frequently Asked Questions

How will the proposed criteria, if adopted, affect ratings?

We believe that, based on our testing and assuming entities in scope maintain their credit risk characteristics, the proposed criteria could lead to credit rating actions on up to 10% of ratings in the insurance sector. We estimate the majority of rating changes would be by one notch, with more upgrades than downgrades. We expect the proposals to have a more material impact on our capital and earnings assessment, with changes in this key rating factor for up to 35% of insurers. These score changes could affect up to 20% of stand-alone credit profiles. The lower potential impact on ratings compared with components of our ratings reflects the application of the insurance ratings framework, our group rating methodology, and sovereign rating constraints.

We anticipate potential improvements in capital adequacy for some insurers, primarily due to our proposal to capture diversification benefits more explicitly and due to increases in total adjusted capital (TAC), owing to the removal of various haircuts to liability adjustments and not deducting non-life deferred acquisition costs. On the other hand, some insurers could face declines in capital adequacy because of factors including changes to our methodology for including hybrid capital and debt-funded capital in TAC, as well as the recalibration of our capital charges to higher confidence levels.

How do the proposed capital model criteria affect S&P Global Ratings' insurance rating methodology?

We are not proposing any changes to our insurers rating methodology (see "Insurers Rating Methodology," published July 1, 2019). However, if the proposed criteria are adopted, we will make changes to the guidance for the insurers rating methodology. See Appendix III in the request for comment for specific details of the proposed changes.

Our insurance capital model remains an integral part of our overall analysis of insurance ratings. See the chart below, which shows rating components that could be affected.

How are you changing the methodology for the calibration of insurer risks, and what is the rationale for the different confidence levels chosen?

We are proposing to recalibrate the risk charges to higher confidence levels. The new calibration uses a one-year basis for the confidence levels, in alignment with the one-year stress for measuring the potential volatility in risk drivers. We determined the confidence levels using analytical judgment--informed by the confidence levels we use in other sectors--and using modeled outcomes based on observed ratings performance.

The objective in selecting the confidence levels was to ensure outcomes that preserve rank-ordering and risk differentiation and that are consistent with the calibration of our insurance ratings framework. We also believe calibrating the lowest confidence level at a one-year 99.5% value at risk (VaR), for example, will aid transparency and improve our dialogue with insurers, given the similarity with many regulatory regimes.

How do the proposed criteria incorporate diversification?

The current criteria incorporate diversification benefits explicitly through the application of correlation matrices and implicitly through the use of index and industry-level data.

In the proposed criteria, we intend to make the diversification benefit more explicit by revising the selected confidence levels and adopting an explicit, multilayered diversification approach that assesses risk dependencies using correlation assumptions between various risks--allowing for diversification within lines of business, within risk categories, and between risk categories.

The proposed approach continues to implicitly assume some level of diversification by using index and industry-level data in our calibration of individual risk charges. Where we see these diversification assumptions not being met, potentially leading to a material understatement of capital requirements, we reflect these specificities through our insurers rating methodology. We also continue to capture geographic diversification benefits in the competitive position assessment within our insurance rating methodology.

The recalibration of the model to higher confidence levels has generally led to an increase in the underlying risk charges. However, we believe that the proposed changes to correlation matrices and diversification benefits, along with other methodological changes, may in many cases offset those increases. We believe that less diversified companies may not benefit as much, and diversification benefits for these companies may not fully offset the proposed increase in capital requirements from the higher charges.

How does the proposal incorporate criteria that will be superseded?

One of the many objectives of the proposed criteria is enhancing the transparency and usability of our insurance capital adequacy criteria. We are consolidating several criteria articles into a single criteria article, thereby embedding in the new criteria key principles of the superseded criteria. For example, we are proposing to integrate and update our criteria for assessing property/casualty insurers' loss reserves in the new criteria. We are also proposing to supersede certain other criteria, such as that for assessing insurers' economic capital models.

In other cases, we are superseding country-, sector-, or asset-class-specific criteria--for example, our methodology for assessing capital charges for commercial mortgage loans held by U.S. insurance companies--to enhance global consistency.

In the U.S., we are proposing to remove explicit capital charges for convexity risk and regulatory closed blocks, which are currently addressed by separate criteria articles.

How has your approach to debt-funded capital and hybrid capital changed?

The principles underlying our treatment of debt-funded capital and hybrid capital remain the same. As well, we are not proposing any changes to the hybrid capital criteria; rather, the proposed changes relate to the quantitative limits applied to eligible hybrid capital.

The proposed changes are intended to:

• Clarify the principles and conditions for inclusion of debt-funded capital in TAC; and
• Ensure global consistency by aligning the definition of capital and the formula by which tolerance limits for debt-funded capital and hybrid capital are determined.

Under this proposal, we will continue to include debt-funded capital in TAC (subject to the proposed limits) in those circumstances where the proceeds of a nonoperating holding company (NOHC) debt issuance are downstreamed as common equity or S&P Global Ratings-eligible hybrid capital, and where we consider structural subordination to be high. Where the issuer of a debt instrument is within the regulatory perimeter, these conditions are not met and the proceeds will not be eligible for inclusion in TAC.

We include debt-funded capital in TAC given its equity-like characteristics, including the ability to absorb losses for the benefit of the group's senior creditors (typically policyholders). Where the requirements are met, the debt or debt-like instrument is transformed into an equity-like instrument. Although structural subordination is present between almost all regulated operating entities and their NOHCs, it is only where the NOHC is outside of the regulatory perimeter that we consider the structural subordination to be high enough to effect this transformation. For instance, we consider Bermuda-based NOHCs, which are within the scope of group solvency calculations and group supervision, to be within the regulatory perimeter, and we are proposing to align their treatment with that of European NOHCs.

With this proposal, we are aligning the tolerance limits on a global basis, while at the same time revising the formula by which tolerance limits are calculated. The proposed formula will now be based on adjusted common equity (ACE). The proposal will also remove the "sliding scale" tolerance limit, for which, under the current criteria, the more debt-funded capital or hybrid capital issued, the higher the amount of eligible nonequity capital in TAC.

The use of ACE will apply a cap to the amount of eligible nonequity capital included in TAC. The updated tolerance limit for intermediate equity content hybrids, at 33% of ACE, is broadly equivalent to the current 25% of TAC. Although this represents an increase in hybrid tolerance for U.S. and Bermudian issuers, it will be more than offset for some issuers by the reduction in the total tolerance limit for nonequity capital in TAC. The use of a common denominator in all regions for determining the tolerance limits may also reduce the amount of eligible nonequity capital in TAC for insurers where we currently use U.S. generally accepted accounting principles (GAAP) equity. We are proposing these changes to better enable global consistency.

How do the proposed criteria recognize an insurer's individual exposure to interest rate risk and its ability to manage it?

We believe interest rate risk is a complex risk to model, particularly for life insurance companies with long-term products and guarantees. We use the duration mismatch as a risk factor to measure interest rate risk, acknowledging that the duration mismatch is only a proxy for an insurer's exposure to interest rate risk.

We derived standard duration mismatch assumptions based on benchmarking and analytical judgment. However, we will determine a company-specific modified duration mismatch, where applicable, to reflect the individual risk profile of an insurer.

Our goal is to capture the economics of potential losses, when material, utilizing various metrics and approaches that may be used across the industry for both regulatory reporting and internal risk management.

What is your approach to natural catastrophe and pandemic risks?

For natural catastrophe risk, we have expanded the quantification of this risk to all lines of business as opposed to only property-related lines. We are proposing to use different return periods for each stress scenario for natural catastrophe risk, in line with other risk drivers.

We have also developed a charge to capture excess mortality losses due to pandemic risks. This capital charge comes in addition to our mortality charges and is designed to better capture event risk.

Can you still make analytical adjustments to the outcome of the capital model?

Yes, we may make company-specific adjustments to TAC and RBC requirements if we consider them material to our analysis. To determine TAC, we will also continue to make routine adjustments to equity (surplus) for differences in valuation assumptions for assets and liabilities, including for different accounting standards.

When will the new capital model criteria be in place, and what data will be required?

Following the request for comment period, we will consider the comments we receive before publishing our final criteria. We will utilize an updated capital model with the release of the new criteria. Regardless, any rating actions we take always solely reflect the criteria in place at the time of the rating committee.

The data required will remain a mix of publicly available and confidential data received from issuers. This allows us to better capture differences in risk profiles and exposures between our rated issuers.

For example, we may derive value-in-force (VIF) based on either publicly available or confidential data from issuers, and we generally use information that is subject to an independent third-party review (e.g., by an auditor, regulator, or actuarial consultancy). However, we may include less than 100% of VIF when, for example, we determine the methodology or assumptions used to calculate VIF are aggressive, or where the information we use to determine VIF is not subject to an independent third-party review.

Are you publishing a model with the RFC?

No. When we publish the final criteria, a new version of the capital model will be available. With the new version, we expect to consolidate several capital models into a single model.

How do the proposed criteria address different accounting standards and potential accounting changes?

Under the proposed criteria, we will continue to make adjustments to reported equity for differences in the valuation of assets and liabilities under different accounting standards. These adjustments also enable us to address recent and ongoing financial reporting changes. Examples include updates under International Financial Reporting Standards (IFRS), U.S. GAAP, or other accounting regimes.

The proposed criteria also indicate we may use regulatory-basis financial statements if they provide more financial information that we believe is relevant to our capital analysis.

In circumstances where certain information is no longer disclosed or publicly available, we may use confidential information to supplement our analysis. We believe that insurance entities, even those reporting under IFRS 17, will be able to provide necessary information on a confidential basis even where they do not provide it as part of financial disclosures.

Another example is the contractual service margin (CSM), which is a measure of future profits, under IFRS 17. We therefore see the CSM as well as the life risk adjustment as eligible under VIF. We would apply the same approach to the excess amount of policy reserves based on the Zillmer method under Japanese GAAP.

What is the process for submitting comments?

Interested market participants should submit their written comments on the proposed criteria by Feb. 18, 2022, to https://disclosure.spglobal.com/ratings/en/regulatory/ratings-criteria/-/articles/criteria/requests-for-comment/filter/all#rfc, where participants must choose from the list of available Requests for Comment links to launch the upload process (you may need to log in or register first). We will review and take such comments into consideration before publishing our definitive criteria once the comment period is over. S&P Global Ratings, in concurrence with regulatory standards, will receive and post comments made during the comment period to https://disclosure.spglobal.com/ratings/en/regulatory/ratings-criteria/view-criteria-comments. Comments may also be sent to CriteriaComments@spglobal.com should participants encounter technical difficulties. All comments must be published, but those providing comments may choose to have their remarks published anonymously or they may identify themselves. Generally, we publish comments in their entirety, except when the full text, in our view, would be unsuitable for reasons of tone or substance.

Appendix: Proposed Changes From Previous Criteria

The proposed criteria incorporate analytical changes that improve our ability to differentiate risk, enhance the global consistency of our methodology, and improve the transparency and usability of our methodology. If adopted, these criteria will supersede 10 criteria articles that we currently use to assess an insurer's capital adequacy. We plan to maintain separate capital adequacy criteria only for assessing bond insurers. However, these proposed changes affect the assessment of asset-related risks for bond insurers. If we adopt these criteria as proposed, we will also make changes to a related guidance document (see Appendix III, "Proposed Changes To Guidance For Insurers Rating Methodology," in the RFC).

More specifically, we are proposing the following changes to TAC:

• Revising our calculation of TAC to reduce complexity and align with proposed changes to our measure of an insurer's RBC requirements, including i) removing various haircuts to liability adjustments (such as non-life reserve surpluses and allowing for up to 100% credit for life value-in-force), ii) not deducting non-life deferred acquisition costs, iii) updating our approach to non-life reserve discounting, and iv) updating, simplifying, and clarifying the approach to unconsolidated insurance subsidiaries, noninsurance subsidiaries, associates, and other affiliates;
• Revising our methodology for including hybrid capital and debt-funded capital in TAC--although we are not proposing any changes to our hybrid capital criteria--by i) clarifying the principles for determining the eligibility of debt-funded capital in TAC, ii) introducing a definition of high structural subordination, iii) aligning globally the hybrid capital and debt-funded capital tolerance limits, and iv) introducing a new metric (adjusted common equity, or ACE) for determining the amount of hybrid capital and debt-funded capital that is eligible for inclusion in TAC;
• Clarifying how we adjust equity for life insurers when there is a mismatch between the balance-sheet valuation of assets and liabilities;
• Updating our treatment of certain equity-like reserves to enhance global consistency;
• Using a narrower definition of policyholder capital that is eligible for inclusion in TAC, along with making enhancements to our criteria for assessing risks relating to ring-fenced participating business;
• Consolidating and updating the analytical principles relating to property/casualty loss reserves and U.S. life insurance reserves; and
• Clarifying that all adjustments to determine TAC are net of the related tax impact, and all capital requirements are pretax.

We are proposing the following changes to RBC requirements:

• More explicitly capturing the benefits of risk diversification in RBC requirements by revising the confidence levels that we use to calibrate risk charges to 99.5%, 99.8%, 99.95%, and 99.99% from 97.2%, 99.4%, 99.7%, and 99.9%, respectively, and proposing updated correlation assumptions and additional risk pairings;
• Updating capital charges for almost all risks based on the revised confidence levels and incorporating recent data and experience;
• Using a single set of charges for each risk with country- or region-specific charges as warranted to reduce complexity and enhance global consistency in the treatment of similar risks;
• Removing the adjustment to the capital model output resulting from our review of insurers' economic capital models (the "M factor") because of proposed changes to these criteria, such as the update to our approach to assessing interest rate risk to better capture an insurer's risk exposures;
• Changing our methodology for determining credit risk charges on bonds (and certain other credit assets) to capture only unexpected losses, rather than total losses;
• Increasing risk differentiation in our credit risk capital requirements for bonds and loans to capture i) variations in loss given default based on sector, creditor ranking, and collateral features and ii) differences in potential losses for structured finance assets, compared with assets in other sectors based on our correlation and recovery assumptions;
• Introducing globally consistent assumptions for determining the rating input for bonds and loans to better differentiate risk, including the use of assumptions for unrated exposures that vary by sector and economic risk group;
• Enhancing global consistency in assessing capital requirements for residential and commercial mortgage-backed securities and mortgage loans;
• Updating our methodology for assessing interest rate risk to enhance global consistency, better capture an insurer's risk exposures, and increase risk differentiation in our interest rate stress assumptions by country, as well as proposing to i) use liabilities as the exposure measure for life and non-life liabilities in all countries, ii) enable use of a company-specific duration mismatch assumption under certain conditions, iii) apply an assumption based on the mean term of non-life liabilities to measure the duration mismatch for non-life business, and iv) reduce the risk of understating capital requirements by introducing floors in our mismatch assumptions and limiting the ability to offset losses in one business segment with gains in another segment;
• Aligning our methodology for life technical risks (in particular, longevity, lapse, expense, and operational risks) across all countries, along with introducing additional risk differentiation for assessing the extent of longevity risk embedded in certain products;
• Introducing explicit capital requirements to capture morbidity risks on disability and long-term care products outside the U.S.;
• Revising the conditional tail expectation (CTE) levels we use to determine capital requirements for variable annuities (VAs), consistent with the updates to our confidence levels, and increasing the amount of credit we include for VA hedging to up to 75% from 50%;
• Introducing capital charges to capture pandemic risk and contingent counterparty credit risk relating to reinsured catastrophe exposures;
• Replacing the flat one-in-250-year posttax property catastrophe capital charge with a pretax natural catastrophe (i.e., across all non-life business lines) capital requirement that varies from one-in-200 to one-in-500 years at different stress scenarios;
• Enhancing consistency in assessing liability-related risks by aligning the treatment of mortgage insurance, trade credit insurance, and title insurance with other non-life business lines;
• Introducing a scaled risk charge on life value-in-force (VIF) to capture the potential change in VIF in stress scenarios (this change is related to our proposal to include up to 100% of life VIF in TAC);
• Removing explicit capital charges for convexity risk and regulatory closed blocks in the U.S.;
• Removing capital charges for assets under management and deducting the investment in asset management businesses to determine TAC to increase the consistency of our approach to noninsurance businesses; and
• Clarifying that we make company-specific adjustments only where they are material to our analysis.