ABIR Response: IAIS Memo On Global Insurance Capital Standard (ICS) MAY 28, 2014
June 9th, 2014 ABIR RESPONSE: IAIS MEMO ON GLOBAL INSURANCE CAPITAL STANDARD (ICS) MAY 28, 2014 To: firstname.lastname@example.org email@example.com About ABIR: The ABIR represents 21 international insurers and reinsurers that protect consumers around the world. These groups reported 2013 global gross written premium of $70.1 billion on a capital base of $95.4 billion. With headquarters or underwriting operations in Bermuda; and with operating subsidiaries around the world, these carriers derive business income from 150 countries and employ nearly 16,000 in the US, nearly 1,600 in Bermuda, more than 8,600 in Europe and nearly 35,000 worldwide (CY 2012). Q1. Do you have any comments on the proposed approach to the development of ICS? ABIR appreciates the IAIS’ call for comments and feedback from Observers and more importantly applaud the IAIS in publishing an interim memo which will further spur on discussion. However, in the absence of more details and specificity around field testing results and observations, it is challenging to provide specific feedback on some of the questions posed. Whilst we understand that there are increasing calls for new global insurance regulatory standards, we have seen no analysis that shows clear gaps in regulation or provides any kind of rigorous cost/benefit analysis of the new proposals. Is such analysis available? In order to achieve consistency with the IAIS approach applied to its Comprehensive Framework for the Supervision of Internationally Active Insurance Groups (IAIGs) (ComFrame), ABIR is concerned that the approach envisioned by the IAIS for the ICS may be overly prescriptive. The ICS should be principles based as is the basis of ComFrame. Fundamentally, ABIR would ask the IAIS that as it has done so in its Insurance Core Principles (ICPs), that it applies the same approach in the development of an ICS so that there are “principles” which govern the application of an ICS. Paragraph 13 makes reference to the definition of qualifying capital resources which is set against the criteria of policyholder protection and loss-absorbency. We believe that there is still work to be done in making the determinations of qualifying capital resources under ComFrame and is an area that needs further consultation. We are in agreement with paragraph 23, “To ensure that the complete assessment of solvency is considered, the ICS development should be seen as the joint development of the capital requirement measure and the capital resource assessment”. Whilst acknowledging the need for consistency and comparably, we are strongly in support of the statement in paragraph 15, “The ICS will be an input to a capital adequacy assessment by supervisors; it will not be a capital adequacy assessment by itself.” No single measure will fit all companies. To test and understand the true impact of any ICS approach, it should be tested alongside the process by which it would be used, i.e. alongside the complete capital adequacy assessment. In the development of a group capital assessment we would support a simple, minimum, base-line risk based capital measurement that can provide an illustration of a group’s minimum regulatory capital needs, affords a basis of comparison amongst international groups and focuses on a minimum or floor capital requirement that if breached would be the basis for regulatory discussion. Q2. With regard to implementation steps and potential transition arrangements (referred to in paragraph 17 above), do you have comments on the intended implementation timeline for ICS, bearing in mind that it is expected to take place from 2019 onwards? ABIR remains concerned that the proposed timeline for implementation as projected may be too ambitious given the uncertainty that remains surrounding a yet to be agreed approach to the development of the ICS. We understand that it is this lack of agreement that is the basis that prompted this discussion via memo. Other key solvency/capital requirement regimes such as Solvency II and Basel III were progressed over a prolonged development phase made up of several field testing and quantitative impact studies to arrive at appropriate capital measures. In addition, reasonable transitional/implementation phases were granted that would allow firms to adjust to the new requirements. We note that the memo suggests that the testing and development of the ICS will be completed by 2016 and then testing and refinement during 2017-2018 with implementation to begin in 2019. It is important that sufficient time be allowed to test and assess the implications of any proposal prior to implementation. Paragraph 17 refers to transition arrangements “from local group-wide capital requirements” suggesting that ICS may replace local jurisdictional group wide capital requirements. ABIR understands the IAIS and the Financial Stability Board’s interest in developing a global ICS and we note that the Bermuda Monetary Authority (BMA) already has in place a group risk based capital standard for the insurers which it supervises as a group supervisor. Around the world there are sophisticated solvency regimes in place and under development. For these jurisdictions it is vital that the ICS sets principles for balance sheet valuation and capital requirements that allow for the various approaches taken by these regimes without duplication where they meet the ICS principles. ABIR would not support development of an additional capital standard that requires capital in addition to existing group capital requirements, such as those imposed by the Bermuda Monetary Authority. Duplicative or redundant group capital standards would be inefficient and counterproductive to the functioning of consumer insurance markets. It is vital that the principles of proportionality and materiality are embedded in the framework. In this regard, paragraph 17 suggests a transition away from jurisdictional group-wide requirements. ABIR would support an approach that recognizes existing jurisdictional group-wide capital requirements which is what paragraph 30 seems to be referring to. Paragraph 21 states that a common means to measure capital adequacy at the group consolidated basis will contribute to a level playing field but ABIR believes that the IAIS approach would create an un-level playing field between IAIGs and other insurers within domestic markets. It would be preferable for local regulatory regimes to meet the standards required by the ICS and as such be recognized within the framework. Q3. Do you have any comments on the proposed aims for the development of the ICS? We note that the main objectives of the ICS (paragraph 19) are the establishment of a globally comparable risk-based measure of capital adequacy, support of financial stability and policyholder protection. In order to achieve this aim, ABIR believes that a simple, minimum, base-line risk based capital measurement that can provide an illustration of a group’s minimum regulatory capital needs, affords a basis of comparison amongst international groups and focuses on a minimum or floor capital requirement that if breached would be the basis for regulatory discussion is the only way to achieve that goal. Otherwise it is difficult to see how the ICS will also support the development of consistent approaches to capital adequacy in emerging markets, facilitate cross-border insurance activities to advance financial inclusion. Q4. Valuation has long been a contentious subject based on the different approaches adopted in different jurisdictions. Comparability is at the center of the aims for the ICS. As stated in paragraph 24, a working assumption is that most of the exposure measures will come from the balance sheet and it is necessary that the balance sheet is comparable to give a comparable measure of qualifying capital resources. A number of statements have been made with respect to the ICS and also in development of the BCR that existing jurisdictional approaches to valuation should be used. What are your views on the development of a balance sheet valuation approach for the purposes of comparability in supervisory reporting and capital adequacy assessments, as mentioned in paragraph 24? In particular, what would be the best way to develop an ICS which is based on comparable capital requirements and resources? The ICS should look at existing efforts to enhance comparability between jurisdictions, like Solvency II and the recognition afforded to equivalent jurisdictions. Deviations from these should be considered in terms of the cost / benefit of creating yet another set of reporting standards and therefore used sparingly. Further, where there are deviations but where existing jurisdictional valuations exist that are materially representative of the intentions of the ICS valuation basis, then the existing jurisdictional valuation should be applicable as a reasonable proxy, again with the cost / benefit in mind of enforcing only marginally different valuation bases. Given the focus of the IAIS on Groups, efforts should be made to reflect the way Groups manage their balance sheets (including diversification effects). It is more appropriate for the wider range of exposures to be included in the ICS than the BCR, but arguably they are best picked up under an ORSA style requirement rather than as a Pillar 1 requirement. The lack of an agreed public accounting model by the FASB and the IASB for insurance accounting complicates regulatory accounting and thus complicates the development of a group capital measure since different accounting systems are currently in use. Until such time as an agreed public accounting model is instituted, regulations should respect the use of the public accounting models most widely in use and regulatory prudential deviations from those models should be sparingly created. For the ICS project, supervisors should be cautious in taking actions that compel creation of substantially modified insurance accounting. The focus on the level of capital must be one that looks to sufficiency to run off policyholder obligations and not a going concern model since the role of the regulator is to honor the contractual obligations to the policyholders. Q5. There were a significant number of comments received in the ComFrame consultation about the approach to capital resources. Field testing will provide further information on capital resources. The IAIS will decide the approach to capital resources based on both the input from consultation and field testing over the period of development of the ICS. Are there any additional points that you wish to make about capital resources that have not already been said in previous consultation submissions? ABIR appreciates that the IAIS is continuing to review its approach to capital resources and we look forward to continuing consultation following the field testing. The proposed rules around deduction of deferred tax assets (“DTA”) appear to be unnecessarily penal. DTAs maintain value for insurers even under stress conditions. Consideration should be given to a more principles based approach to the recognition of DTAs similar to that set out within Solvency II and interpreted in detail by the PRA in their Supervisory Statement SS2/14. Some form of transitional guidance / allowances may be required to accommodate capital instruments currently in issue that do not fully meet qualify as capital under the ICS. Q6. What are your views on the use of an internal model-based method to determine either the entire capital requirement or part of the capital requirement for the ICS? Should this be part of the ICS developed by 2016 or part of a longer-term goal for the ICS? Internal models often help understanding and increase risk sensitivity, but other simple tools can ask insightful, challenging questions. Regulators need a range of tools and the ability to exercise judgement based on their interpretation of “facts” – no single approach is “best” on its own – instead a range of approaches should be used, taking the relative strengths of a range of measures and allowing adjustment for their weaknesses. In terms of timelines for allowing for internal models and comparability when moving away from a standardized formula, we encourage regular dialogue to understand the unique features of a company, allowing adjustments to their “standard formula” where required as well as giving weight to an approved internal model, rather than applying a “one size fits all” approach. The use of an internal model-based method to determine the entire capital requirement should be an option as part of the ICS developed by 2016. Q7. What are your views regarding the approaches described in paragraph 32? ABIR is in favor of the fourth bullet point option i.e., that the use of internal models should be part of the ICS that is developed by the end of 2016, including supervisory approval and appropriate safeguards for the use of such internal capital models. Q8. If you believe either or both internal models or other risk-based methods of implementation should be allowed, how should comparability of outcomes be ensured and evidenced? Any group capital ICS that is created should not negate the impact of regulatory approved economic capital models for the calculation of group capital in jurisdictions where models are allowed. It is essential that from the start groups are given the opportunity to use either a standard model or an approved internal model which should fit into the 2016 timetable. Part of the internal model approval process should involve comparing and analyzing the outcomes of the two types of models e.g., standard and internal model. Paragraph 30 references that the IAIS within the ICS will consider what other risk-based methods of implementation jurisdictions may introduce to address additional risk coverage and/or maintain higher prudential targets than the standard, as long as these methods produce outcomes at least as robust as the ICS standard method. This suggests that consideration will be given to other jurisdictional capital requirements as long as they are comparable to an ICS standard. We would support this approach and suggest for pragmatic reasons this would be an essential element of any IAIS standard. Q9. What are your views on the ICS being developed as a group-wide PCR as proposed in paragraph 35? (Otherwise, how else should it be developed)? We note that the ICS will be developed to the Prescribed Capital Requirement (PCR) as opposed to a minimum capital level. Calibration is sought at either at least the 99.5% VaR over 1 year or at least 90 TVar or CTE over 1 year. These limits are high and are not consistent with a simple, minimum base line capital requirement. We would support creation of an ICS at the minimum capital level. To promote development of a PCR without sufficient field testing and without recognition of variability of products and risks underwritten in various jurisdictions would be punitive; and would be counterproductive to the IAIS’ goal of wide-scale adoption of its standard. Q10. Do you support the two alternative criteria on which information would be collected to enable calibration of the capital requirement? It is more commonly assumed that the 99.5% VaR and the 99% TVaR over 1 year give comparable outcomes, whereas paragraph 36 suggests the 90% TVaR. Is this an intentional divergence? Further, would there be more value in gathering information at two distinct points, such as the 99.5% VaR and the 90% or 99% VaR. This would enable calibration to be based on a wider range of information as well as including lower and typically more robust percentiles. However, neither measure is consistent with a minimum capital approach. Q11. What are your views on the development of the supervisory capital adequacy assessment process as proposed in paragraph 38? Should there be additional elements? What are your views on whether the capital adequacy assessment process to be developed in parallel with the ICS should be:
- One where clear enforceable supervisory actions are set out in that process which would provide some certainty regarding supervisory actions that can be expected if the ICS is breached; or
- One with a more flexible approach where possible supervisory actions are articulated but application of those actions is subject to supervisors taking into account the specific circumstances of the IAIG if the ICS is breached?
ABIR supports the second option of a more flexible approach. The condition of a company as well as wider market conditions could be expected to be at the least unusual, if not unique, in an instance where the ICS is breached. Attempting at this time to apply rigid rules to a company could cause greater loss to policyholder value as well as causing possibly damaging delays whilst the rules are investigated and applied. The existence of rigid rules (that will not necessarily suit all companies’ risk profiles) might also discourage good risk management and risk mitigation. Allowing a more flexible approach, where a supervisor can react to the specific situation in hand, is far more likely to achieve the best outcome for the policyholder. A supervisory capital adequacy assessment process should not duplicate efforts at the local level for local regulatory model use. It will require some work to develop an appropriate process given that there is a requirement to justify the model in the context of the standard model and if there are two different standard models (local and ComFrame) this is a further administrative burden and it is unclear what value this would add.