ESBG, Rue Marie-Thérèse, 11 - B-1000 Brussels
ESBG Transparency Register ID 8765978796-80
4 September 2019
>> See .pdf version
We welcome the opportunity to comment on EFRAG'S draft comment letter on the IASB's ED/2019/4 amendments to IFRS 17. The European Savings and Retail Banking Group (ESBG) brings together savings and retail banks of the European Union and European Economic Area, which represent approximately one third of the retail banking market. ESBG helps savings and retail banks in 20 European countries strengthen their unique approach. An advocate for a proportionate approach to banking rules, ESBG unites at EU level some 1,000 banks, which together employ 810,000 people driven to innovate at 60,000 outlets. ESBG members have total assets of €7.1 trillion, provide €500 billion in SME loans, and serve 190 million Europeans seeking retail banking services.
ESBG is aware that IFRS 17 seeks to significantly increase the comparability in accounting for insurance contracts between companies from different countries and business models, as well as to enhance the quality of financial information. We continue to support a high quality standard for insurance contract accounting, however, we firmly believe that the standard does not correctly reflect certain contracts issued by our members that represent long-term life-saving products managed under cash flow matching and, to a certain extent, participating contracts, through its measurement nor its presentation requirements.
In particular, the amendments included in the Exposure Draft are insufficient to capture the particularities of the life insurance business model in Spain and France (see Appendix 2 and 3 for more detail), so there are still critical technical issues that remain unresolved through the proposals of the Exposure Draft and which we believe merit additional consideration by EFRAG:
- The level of aggregation requirement in the long-term saving-products is not justified and consistent with the principles that underpin the insurance business, it leads to an artificial variability in the adjustments of the CSM in senior cohorts and increases the scope of potential onerous cohorts. Hence, modifying the standard on the level of aggregation of insurance contracts and on annual cohorts is crucial for a proper representation of the performance of insurance contracts and to better fit with the way insurance contracts are managed by insurers (intergenerational mutualisation, offsetting risks across generations from the perspective of the insurance company reflecting the link between aggregation and pricing determination). See our response in “Additional issues –annual cohorts" for more information.
- Under the Fair Value approach in Transition, current requirements lead to an accounting mismatch in the accumulated amount of OCI in products without direct participation features but managed under cash flow matching techniques to be measured under the general model. Under local commercial law of certain countries, this impact could prevent companies from distributing dividends to shareholders.
Additionally, under the Modified Retrospective Approach, further simplifications are necessary. We believe that a number of additional modifications are required if the modified retrospective approach is to be applied in practice. See our response in Question 8 for more information.
-In relation to interim financial statement requirements: paragraph B137 needs to be amended to avoid inconsistencies between group and solo reporting that are solely due to differences in reporting frequency. Otherwise, entities will incur in unnecessary costs of dual accounting. See our response in Question 5 for more information.
ESBG has also commented on proposed changes in the Exposure Draft to address areas of concern, that do not fully resolve the identified issues,. This include the following:
- The definition of investment return service introduced by the Exposure Draft requires the existence of an investment component, while we firmly believe that certain products without investment component provide this service to the policyholder, leading to accounting differences when contracts are economically identical. See our response in Question 3 for more information.
- Risk mitigation is a critical element of insurance business. As such, we propose an extension of the risk mitigation option to include non-derivative financial instruments and to be applicable to all insurance contracts (not only contracts accounted for under the variable fee approach). See our responses in Question 6 for more information.
- ESBG believes there has been a positive development from the IASB in reinsurance topics. However, additional amendments are necessary to avoid mismatches between insurance contracts and reinsurance (in particular, non-proportionate reinsurance contracts held and the contract boundary topic should also be considered). See our response in Question 4 for more information.
- Presentation: we believe that receivables and payables, as claims payable, collateral deposits (reinsurance), between others, should be presented in the balance sheet separately from insurance liabilities, as this would provide more useful information. See our response in Question 5 for more information.
Finally, ESBG also believes that the proposed deferral of the effective date to 1 January 2022 is insufficient and that an additional deferral to 1 January 2023 is needed for a successful implementation of the standard. A delay of one additional year will not be disruptive nor defer the implementation efforts of the companies, as this time is extremely necessary to implement the IT systems and face all the operational challenges arising from the significant changes introduced by IFRS 17, as well as to prepare the information that will be presented to the market.
Having made these general remarks, please find in the appendix 1 below, ESBG's responses to the specific questions posed by EFRAG in its draft comment letter, including our view on the IASB's questions included in the exposure draft.
Detail and responsibility of the ESBG person who signs the letter
(a) Paragraph 7(h) proposes that an entity would be required to exclude from the scope of IFRS 17 credit card contracts that meet the definition of an insurance contract if, and only if, the entity does not reflect an assessment of the insurance risk associated with an individual customer in setting the price of the contract with that customer.
Do you agree with the proposed amendment? Why or why not?
ESBG agrees with the exclusion of certain credit cards that provide insurance coverage from the scope of IFRS 17 and supports the comments made by EFRAG in its draft comment letter. In particular, we are concerned that the term “credit card“ excludes other types of payment cards which have similar clauses as the credit cards in the scope exclusion, so we believe that the scope exclusion should make reference to “payment cards“ in general.
(b) If not excluded from the scope of IFRS 17 by paragraphs 7(a)–(h), paragraph 8A proposes that an entity would choose to apply IFRS 17 or IFRS 9 to contracts that meet the definition of an insurance contract but limit the compensation for insured events to the amount required to settle the policyholder's obligation created by the contract (for example, loans with death waivers). The entity would be required to make that choice for each portfolio of insurance contracts, and the choice for each portfolio would be irrevocable.
ESBG agrees with the proposed amendment and supports the comments made by EFRAG in its draft comment letter, with no additional comment.
Nevertheless, ESBG believes there is another relevant scope issue, related to contracts that change in nature over time. The insurance industry issues products that change significantly in nature during their life due to the execution of an option by the policyholder (for example, products with a savings phase with profit sharing that may become an annuity). As the classification between general model and variable fee approach occurs at inception and is irrevocable, certain products may have to be accounted for under the variable fee approach, whereas, after the execution of the option, the variable fee approach model is not suitable and not comparable to similar products with a different 'history'. We propose a solution that treats a significant change in the nature of a contract due to the execution of an option by the policyholder as a contract modification. The 'new' contract post execution of the option by the policyholder could be reassessed and treated under the appropriate measurement model for its new features.
10. B.4.1.9.E of IFRS 9 allows to consider a regulated interest rate as a proxy for the time value of the money in doing the SPPI test, under certain conditions. EFRAG understands that in some countries the insurance element is not required by the regulation and, as a consequence, the financial instrument could fail the SPPI test and would have to be measured at fair value through profit or loss. How prevalent are these concern within your jurisdiction?
Response to EFRAG:
We believe that most contracts that provide an additional cover, which improves the minimum cover required by regulation, would have only a “de minimis" effect on their contractual cash flows arising from the payment card. In consequence, we believe that this additional cover should not affect the classification of the financial asset in accordance to paragraph B.4.1.18 of IFRS 9.
Nevertheless, if the IASB is aware that the SPPI test could fail in certain circumstances and jurisdictions, an exemption to these payment cards should be considered in order to avoid measuring their balances at fair value through profit or loss, which we believe would not be appropriate and provide useful information to users.
Question 2—Expected recovery of insurance acquisition cash flows (paragraphs 28A‒28D, 105A–105C, B35A–B35C and BC31–BC49)
Paragraphs 28A–28D and B35A–B35C propose that an entity:
(a) allocate, on a systematic and rational basis, insurance acquisition cash flows that are directly attributable to a group of insurance contracts to that group and to any groups that include contracts that are expected to arise from renewals of the contracts in that group;
(b) recognise as an asset insurance acquisition cash flows paid before the group of insurance contracts to which they are allocated is recognised; and
(c) assess the recoverability of an asset for insurance acquisition cash flows if facts and circumstances indicate the asset may be impaired.
Paragraphs 105A–105C propose disclosures about such assets.
Do you agree with the proposed amendments? Why or why not?
ESBG agrees with the proposed amendment. However, contrary to the comments made by EFRAG in its draft comment letter, ESBG believes that the allocation of acquisition costs to expected renewals should be optional, not mandatory.
Being the allocation a requirement, not an option, the amendment might introduce the obligation each year to demonstrate, in case there is no allocation to renewals, that the expected renewals have effectively not been considered in the decision to incur in certain acquisition cash flows. In order to avoid this complexity and costs, ESBG would prefer the allocation of acquisition costs to expected contract renewals to be optional, not a requirement, as the relief is particularly useful for P&C business, but should not create additional work to life-insurers that issue short-term insurance contracts (measured under the PAA model).
Amending this topic as an option, not a requirement, would also solve the impairment in comparability that would introduce the use of FV approach in Transition (there would not be an asset recognised for this item) in relation to any of the retrospective approaches (in which there might).
18. Insurance contract renewals are not a defined term which may lead to diversity in practice when allocating insurance acquisition cash flows. Do you consider that insurance contract renewals should be defined in order to achieve comparability and, if so, how would you define them?
Response to EFRAG:
ESBG believes that it is not necessary to develop a definition of renewals. The renewals to be considered in the allocation of acquisition costs will be entity–specific, as it will depend on the expectation of contract renewals considered by the entity in the decision to incur in a certain initial amount of acquisition costs. Defining contract renewals would be inconsistent with the approach taken in IFRS 15 and introduces the risk of achieving a restrictive definition that could limit the benefit of the amendment (better reflecting the economic substance of the transaction and providing more relevant information to users of financial statements).
Question 3—Contractual service margin attributable to investment-return service and investment-related service (paragraphs 44–45, 109 and 117(c)(v), Appendix A, paragraphs B119–B119B and BC50–BC66)
(a) Paragraphs 44, B119–B119A and the definitions in Appendix A propose that an entity identify coverage units for insurance contracts without direct participation features considering the quantity of benefits and expected period of investment-return service, if any, in addition to insurance coverage.
Paragraph B119B specifies criteria for when contracts may provide an investment-return service.
ESBG believes the proposed amendment is an improvement to the current requirements in IFRS 17. The CSM established at inception for these types of contracts includes expected profit from both insurance and investment related activities. We agree that the profit from these services should be recognised in line with the service provision over the life of the contract. In our view the proposed amendments significantly improve the relevance of the income statements for these contracts.
However, we also believe that the definition of an investment return service is unduly prescriptive and too narrow as it would result in economically similar contracts having different accounting results.
In particular, it restricts the use of the investment return service as coverage unit when the insurance contract includes a non-distinct investment component or the policyholder has the right to withdraw an amount. Consequently, the amendment does not work for (i) deferred annuities without payment on death in the accumulation phase or the payout phase (or in both), and (ii) deferred capital during the term agreed (accumulation period) without death benefit.
Additionally and in broader terms, ESBG is also concerned that, based on this amendment, any type of long-term life contract whose surrender value is linked to the market value of certain underlying assets (contracts not eligible under the VFA) could not qualify as providing an investment return service depending on the interpretation that is made of the “expected positive return".
We believe that an investment return service is present where the contract provides the policyholder with a positive expected investment return, and that this distinguishes contracts which provide investment return and insurance services, from contracts which provide only insurance services.
We therefore suggest that the wording of B119B is revised as follows:
“Insurance contracts without direct participation features may provide an investment-return service if, and only if:
a) the contract provides (on an expected basis at group level) a positive investment return (which could be below zero, for example in a negative interest rate environment); and
b) the entity expects to perform investment activity to generate that positive investment return."
We observe that under this definition an investment return service would either be absent, or present throughout the lifetime of a contract, and so the operational difficulties associated with coverage units changing once investment return services are deemed to have ceased, or investment management expenses being only partly included in the fulfilment cash flows are avoided.
The above proposed definition should be considered on the assumption that policyholders will exercise their options only when they are economically beneficial for them.
(b) Paragraphs 45, B119–B119A and the definitions in Appendix A clarify that an entity is required to identify coverage units for insurance contracts with direct participation features considering the quantity of benefits and expected period of both insurance coverage and investment-related service.
ESBG agrees with the proposed amendment and supports the comments made by EFRAG in its draft comment letter.
Nevertheless, we would like to highlight the introduction of IFRS 17.B119A (along with IFRS 17.BC61) as we are concerned that this requirement does not correctly reflect the economics of some saving contracts (especially on the French market; please refer to Appendix 3 for further details regarding these contracts). Indeed, for these contracts, the investment returns are shared among all generations of policyholders and can either be distributed to current or future policyholders, as assets are not dedicated to a specific generation of contracts and there is a sharing and transfer of wealth (returns of underlying assets) between the generations of policyholders. Therefore, there is no specific distinction between current and future policyholders regarding the payments of amounts related to the investment-related service. The strict application of this modification to IFRS 17.B119A would lead to results which do not correctly reflect the economics of these contracts.
(c) Paragraph 109 proposes that an entity disclose quantitative information about when the entity expects to recognise in profit or loss the contractual service margin remaining at the end of a reporting period. Paragraph 117(c)(v) proposes an entity disclose the approach used to determine the relative weighting of the benefits provided by insurance coverage and investment-return service or investment-related service.
Do you agree with the proposed disclosure requirements? Why or why not?
ESBG is concerned about the removal of the option in paragraph 109 of IFRS 17 to provide only qualitative information in relation to the expected recognition in profit or loss of the contractual service margin remaining at the end of the reporting period. We believe there is not a similar requirement of future performance disclosure in other industries so this fact should be considered before removing this option under IFRS17.
As an example, insurance groups with a banking business line or financial conglomerates are required to disclosure information related to their net interest margin (sensitivity analysis to interest rates changes, for example), but not quantitative information about its future financial performance by time-period buckets. It is a common practice to provide guidance to the market about the future trend of the NIM, but we are not providing quantitative amounts in time buckets as the insurance sector would be required. ESBG is concerned that providing this information by time bands may go far beyond the information that companies intend to provide to investors with their market guidance.
In relation to paragraph 117(c) (v), ESBG does not agree with the proposed amendment because, depending on the driver used to amortize CSM, it may be difficult to identify which part of the CSM corresponds to the insurance coverage and which part corresponds to the investment return service, while the disclosure would not provide significantly useful information to users of financial statements.
35. EFRAG has been informed of possible fact patterns of deferred annuities for which there is no investment component as defined by the ED, nor a right to withdrawal; however, the insurance entity performs asset management activities, revenues of which would not be captured in the CSM release. For example, for particular Deferred Annuities, there is an accumulation phase followed by the annuity phase. The policyholder's beneficiaries receive no return if the policyholder dies during the accumulation phase. During the annuity phase, a surviving policyholder receives a fixed annuity amount based on premiums/technical provisions. In these deferred annuities the policyholder does not have a right to withdraw during either the accumulation phase or the annuity phase. Do you have additional examples of investment activities that are not captured by the proposals in the ED?
36. Entities have to provide quantitative disclosures on the expected recognition in profit or loss of the contractual service margin remaining at the end of the reporting period, in appropriate time bands. Do user constituents agree with this disclosure requirement? Do preparer constituents consider that this information is commercially sensitive? Please explain.
35. See answer to Question 3 (a). ESBG proposes a more review of the eligible criteria to assess whether the insurance company provides an investment return service, as is concerned about two specific types of contracts that provide the policyholders with access to an investment return, even they do not qualify for an investment return service as defined by the Exposure Draft .
36. See comments on answer to Question 3 (c). ESBG believes this information is commercially sensitive and that the existence of similar performance disclosures must be considered before removing the possibility of a qualitative disclosure.
Question 4—Reinsurance contracts held—recovery of losses on underlying insurance contracts (paragraphs 62, 66A–66B, B119C–B119F and BC67–BC90)
Paragraph 66A proposes that an entity adjust the contractual service margin of a group of reinsurance contracts held that provides proportionate coverage, and as a result recognise income, when the entity recognises a loss on initial recognition of an onerous group of underlying insurance contracts, or on addition of onerous contracts to that group. The amount of the adjustment and resulting income is determined by multiplying:
(a) the loss recognised on the group of underlying insurance contracts; and
(b) the fixed percentage of claims on the group of underlying contracts the entity has a right to recover from the group of reinsurance contracts held.
ESBG agrees with the proposed amendment and supports the comments made by EFRAG in its draft comment letter. However, we believe that the proposed wording is insufficient to address the issue in all situations.
In particular, we do not agree with the calculation of the reinsurance adjustment in paragraph B119D, as this amendment can result in the recognition of reinsurance income that does not reflect the expected profits or losses on a reinsurance contract. We propose that a principles based approach to calculating the amount recorded in profit or loss is included in the standard. Such an approach would allow entities to develop their own methodology to meet the rationale included in the Basis for Conclusions.
Additionally, we do not share the definition of a reinsurance contract held that provides proportionate coverage, as the explicit mention to a right to recover a fixed percentage is very restrictive in practice.
Moreover, “non-proportionate" reinsurance contracts held, which are part of the risk management policies of insurers, are excluded from this amendment. We are concerned that there would be a misalignment between the different reinsurance contracts held “natures" (proportionate and non-proportionate) within an insurance company. Not applying this amendment to reinsurance contracts held other than “proportionate" would not appropriately reflect the risk management policies of insurers and create accounting mismatches between underlying insurance contracts and some categories of reinsurance contracts held (the non-proportionate ones).
There are also restrictions on application of the proposed amendment depending on when the reinsurance contract is recognised.
Another issue related with reinsurance is that ESBG does not agree with the ineligibility of the VFA model to reinsurance contracts issued, especially when the underlying contracts are eligible to the VFA. Such reinsurance contracts issued exist in some jurisdictions such as in France. This misalignment between the underlying contracts and reinsurance contracts issued would create accounting mismatches as both contracts are measured under different models (the underlying contracts are measured under the VFA model in which financial variations are absorbed by the CSM whereas the reinsurance contracts issued are measured under another model (i.e. the General Model) in which the financial variations are accounted in P&L or OCI as the CSM cannot absorb them).
We believe these mismatches do not reflect the economics of the contracts and lead to economic inconsistencies in the financial statements for contracts with similar characteristics. Moreover, an insurer can have reinsurance contracts held and reinsurance contracts issued (based on underlying contracts eligible to the VFA) for which the accounting requirements will be different, creating accounting mismatches.
45. For proportionate reinsurance contracts, please provide fact patterns that are not captured by the amendment but for which the solution proposed by the IASB would be relevant.
46. The IASB has not addressed non-proportionate reinsurance contracts. A peculiarity of such contracts is that there is no one-to-one relationship between the direct underlying contract and the reinsurance contract held, for example because there are many underlying contracts that are covered by a single excess loss reinsurance contract held. Addressing non-proportionate reinsurance may therefore require the need to identify a “link" between the reinsured risk and the underlying contracts. EFRAG understands that any accounting mismatch for non-proportionate contracts may, in practice, be reduced due to the impact on the risk adjustment rather than on the CSM.
47. In your view:
(a) Should non-proportionate reinsurance contracts be treated similarly to proportionate reinsurance contracts, i.e. gains in profit or loss when a loss is recognised on underlying contracts? If yes, please provide information about (i) the prevalence of such contracts, including volumes and jurisdictions where the issue arises and (ii) the cash flow pattern of these non-proportionate reinsurance contracts.
(b) How would an accounting solution for non-proportionate reinsurance work?
ESBG supports the comments made by EFRAG in its draft comment letter. However, as mentioned before, the amendment is insufficient to address the issue on all situations.
(EFRAG CL – Appendix 2, Topic 4: Reinsurance contracts: contract boundary)
172. Do Constituents support the IASB's tentative decision not to amend IFRS 17 for the contract boundary of reinsurance contracts held?
173. Do Constituents that are Users consider that CSM for the reinsurance contracts held which reflects future expected contracts would provide useful information? Please explain.
174. EFRAG understands that there is no material impact on the balance sheet and probably not a significant impact on profit or loss (until certain events occur as explained in paragraph 169 above). Please explain the prevalence of holding reinsurance contracts that relate to underlying contracts that have not yet been issued, including volumes and the jurisdictions where the issue arises.
ESBG believes that reflecting potential future insurance contracts in the reinsurance asset does not provide useful information. As such, having the CSM for reinsurance contracts held include future expected contracts would not provide useful information for investors (moreover, depending on the assumptions used for measuring future expected contracts, the information provided to investors could vary substantially and not be accurate).
We are therefore concerned that the contract boundaries for reinsurance contracts held are aligned with those of the underlying insurance contracts, to avoid providing inaccurate financial information to investors, given the expected:
accounting mismatches (i.e. regarding the calculation and amortization of the CSM which would be assessed based on different assumptions (expectations of future expected contracts would be included in the measurement of reinsurance contracts held but would not be in the measurement of the underlying insurance contracts)), and;
operational consequences (i.e. an entity would have to perform two separate calculations of fulfilment cash flows which would be based on different future assumptions and would have to be performed within very limited time, given the short reporting periods).
Question 5—Presentation in the statement of financial position (paragraphs 78–79, 99, 132 and BC91–BC100)
The proposed amendment to paragraph 78 would require an entity to present separately in the statement of financial position the carrying amount of portfolios of insurance contracts issued that are assets and those that are liabilities. Applying the existing requirements, an entity would present the carrying amount of groups of insurance contracts issued that are assets and those that are liabilities. The amendment would also apply to portfolios of reinsurance contracts held that are assets and those that are liabilities.
ESBG agrees with the proposed amendment but there are other presentational challenges still unaddressed.
We believe the standard should be amended to include premiums and claims on an accrual basis in the measurement of insurance liabilities, with separate premiums receivable and claims payable, balances included separately on the balance sheet. Benefits of this change include improvements in the quality of financial information presented and reduced implementation costs.
On the other hand, the requirement to remove insurance revenue and insurance service expenses relating to non-distinct investment components from the insurance service result adds complexity with limited benefits. We agree with the original objective of increase comparability with products in other industries, such as bank deposits, but further simplifications in the definition of investment components are needed in order to make the requirement more operative. In particular, we believe that comparability can be achieved by revising the definition of an investment component to include only contracts where the policyholder has the right to make withdrawals (the right to withdraw his deposit would be adjusted as appropriate by investment return added and fees deducted from the deposited amount).
We believe such a change would better meet the needs of users.. We believe the original intention of the IASB was that non-distinct investment components only need to be identified and measured when a claim occurs, while changes to the definition of an investment component have created uncertainty around when a non-distinct investment component is identified and measured. We ask the IASB to resolve the uncertainty around identification by clarifying that non-distinct investment components are identified based on facts and circumstances at initial recognition of the contract and are measured when the claim occurs.
Additionally, another identified issue related to presentation are the measurement differences solely due to frequency of reporting. The requirement in paragraph B137 of IFRS 17 that the CSM must be “locked-in" at interim reporting means that any differences in external reporting frequency between group and subsidiary entities would result in different CSMs at different levels of consolidation. This adds significant operational complexity in the production of financial statements in a group, with little impact on the financial information reported at group and subsidiary level. As such, we would support an amendment to IFRS 17 to require an annual “year to date" approach to be taken in the calculation of the CSM and other estimates, irrespective of the frequency of reporting. Such an amendment would prevent the need to calculate different CSMs and other estimates at different levels in the group consolidation only because of the different reporting frequency.
The issue related to interim financial statements raised by the discrete approach required by B137 of IFRS 17 goes beyond the difference in treatment between group and subsidiary financial statements. Indeed, under the current IFRS 17 requirement, two identical insurers with identical estimates of fulfilment cash flows and the same economic and non-economic factors will measure the CSM differently depending the frequency of their external financial reporting. As such, the IFRS 17 does not currently ensure comparability between industry competitors.
For example: Company X produces quarterly financial statements in compliance with IAS 34 while Company Y only produces annual financial statements. If all actual and expected fulfilment cash flows and other financial effects are identical for Company X and Company Y, users of the financial statements will see differences in the reported profit for Company X and Company Y purely due to the timing effect created by B137. Linked to the objective of not having differences in the measurement of the CSM between subsidiary and group level, ESBG also believes that a practical expedient should be provided so that, unless it is proved otherwise, the CSM at group level could be considered equal to the subsidiary's in relation to the acquisition costs included in measurement.
(EFRAG CL – Appendix 2, Topic 3: Balance sheet presentation: Non–separation of receivables)
161. Do Constituents support the presentation of separate information about premiums receivable? If so, should information about premiums receivable:
(a) be mandatory?
(b) be based on a predefined definition of “premium receivables" and , in this case, how should premiums receivable be defined?
(c) be provided on the face of the balance sheet or in the notes?
(d) be separated by insurance portfolio?
ESBG believes that receivables should be presented in the balance sheet separately from insurance liabilities as this would provide useful information. Premium receivables should be defined as all premiums that are due from the policyholder (and, therefore, excluded from the insurance liability) but not yet received.
The separate presentation issue (see question 5) is also relevant for other receivable/payables such as claims payable, collateral deposits (reinsurance).
Question 6—Applicability of the risk mitigation option (paragraphs B116 and BC101–BC109)
The proposed amendment to paragraph B116 would extend the risk mitigation option available when an entity uses derivatives to mitigate financial risk arising from insurance contracts with direct participation features. That option would apply in circumstances when an entity uses reinsurance contracts held to mitigate financial risk arising from insurance contracts with direct participation features.
ESBG agrees with the IASB proposal to extend the scope of the risk mitigation option to reinsurance contracts held, and not only in case of using derivatives, in order to avoid the accounting mismatch that arises using the VFA. However, we believe that further changes are needed to allow other financial instruments to be used as the hedging instrument and under the general model.
Our understanding is that under IAS 39 companies are hedging different risk components (such as interest rate exposure arising from providing interest rate guarantees to the policyholder) not only with derivatives and reinsurance contracts. This can occur both under the VFA and the general model. A mix between fixed rate and variable rate instruments together with swaps, options and IRS may be used to ensure that expected cash flows to be paid to the policyholder match the cash flows arising from the financial asset portfolios. In addition, we believe that the risk mitigation option should be available to all insurance contracts, rather than only to those with direct participation features.
The inability to use the risk mitigation option outside the variable fee approach may result in accounting mismatches, as the effects of changes on hedging instruments is not recognised in the same location as the changes on the hedged items. This significantly distorts the net result and creates misalignments between accounting results and risk management. Furthermore, insurance company risk management is typically organised at a macro level, covering both contracts accounted for under the variable fee approach and the general measurement model.
For example, for products accounted for under the IFRS 17 general measurement model, using the OCI option for changes in interest rates results in volatility in profit or loss caused by accounting mismatches if companies are not able to designate their derivatives under the criteria for hedge accounting. The effect of the derivatives used for economic hedging will be recognised in profit or loss, while the entire effect of interest rate changes will be recognised in OCI. Therefore, if OCI is elected, additional volatility in profit or loss from hedging will create a disincentive for companies to mitigate risk. Similar accounting mismatches occur in situations where an insurer applies the OCI option to a portfolio but only hedges a subset of contracts within the portfolio.
It is sometimes suggested that the 'through profit or loss' approach in the IFRS 17 general measurement model, together with using the fair value option for the financial assets, would be sufficient to address this problem. However, if those options are elected profit or loss could still show short-term volatility from mismatches that otherwise would have been more usefully reported in OCI. Some of these mismatches are fundamental, as credit spread changes on assets are not necessarily reflected equally in the IFRS 17 liability. Other mismatches could result from a company decision, based on ALM objectives, not to hedge financial risks in full.
ESBG also believes that the risk mitigation option should be applied retrospectively on transition. Without such a change, the economics of existing hedging arrangements cannot be accurately reflected on transition.
64. EFRAG has heard that the extension of the risk mitigation option should be widened, for example, to include non-derivative instruments such as when hedging of interest rate risk is carried out using a combination of swaps, swaptions and fixed interest securities.
65. Please explain the prevalence including volumes and jurisdictions involved, of the risk mitigation strategies identified in paragraph 64 above.
See answer to question 6. We believe that the risk mitigation option should be widened to include non-derivative instruments and that the responses to “EFRAG Hedge accounting questionnaire for insurers" will provide relevant information on the current risk mitigation strategies.
Question 7—Effective date of IFRS 17 and the IFRS 9 temporary exemption in IFRS 4 (paragraphs C1, [Draft] Amendments to IFRS 4 and BC110–BC118)
IFRS 17 is effective for annual reporting periods beginning on or after 1 January 2021. The amendments proposed in this Exposure Draft are such that they should not unduly disrupt implementation already under way or risk undue delays in the effective date.
(a) The proposed amendment to paragraph C1 would defer the effective date of IFRS 17 by one year from annual reporting periods beginning on or after 1 January 2021 to annual reporting periods beginning on or after 1 January 2022.
ESBG believes that the IASB proposal to defer the effective date of the standard by one year is a step in the right direction, but considers it is insufficient and at least 2 years of deferral is needed to a successful implementation of IFRS17.
A delay of one additional year until 1 January 2023 will not be disruptive nor defer the implementation efforts of the companies, as this time is extremely necessary to implement the IT systems and face all the operational challenges arising from the significant changes introduced by IFRS 17, as well as to prepare the information that will be presented to the market.
(b) The proposed amendment to paragraph 20A of IFRS 4 would extend the temporary exemption from IFRS 9 by one year so that an entity applying the exemption would be required to apply IFRS 9 for annual reporting periods beginning on or after 1 January 2022.
Do you agree with the proposed amendment? Why or why not?
ESBG believes that is necessary to extend the temporary exemption from IFRS 9 to be aligned with IFRS 17's final effective date so that companies can apply both standards together and supports the view that any additional delay of the effective date of IFRS17 should result in a postponement of IFRS 9.
73. Do you consider that the proposed deferral of the effective date to 1 January 2022 is sufficient or would you support an additional year (i.e. 1 January 2023)?
74 Arguments in favour of accepting the proposed effective date of 1 January 2022 include:
(a) Further delaying the application of IFRS 17 beyond 2022 will be disruptive, as will increase the costs of their implementation processes; and
(b) A delay beyond 2022 may encourage entities to defer their implementation efforts rather than using the extended period to better implement the Standard.
75 Arguments in favour of further delaying the effective date to 1 January 2023 include:
(a) Some entities, mainly small and medium sized ones, often rely on third IT systems providers and so far there are no IT solutions for IFRS 17 available on the market, thereby making it difficult to meet the proposed 2022 effective date;
(b) The IASB expects to finalise the amendments by mid-2020. As a result, there will only be six months before the comparative period for IFRS 17 starts and this may be challenging for some entities; and
(c) Entities that would like to apply IFRS 17 earlier would be able to do so.
See answer to Question 7(a). ESBG believes that the proposed deferral of the effective date to 1 January 2022 is insufficient and that an additional deferral to 1 January 2023 is needed for a successful implementation of the standard. A delay of one year will not be disruptive nor defer the implementation efforts of the companies, as this time is necessary to implement the IT systems and all the significant changes introduced by IFRS 17.
Question 8—Transition modifications and reliefs (paragraphs C3(b), C5A, C9A, C22A and BC119–BC146)
(a) Paragraph C9A proposes an additional modification in the modified retrospective approach. The modification would require an entity, to the extent permitted by paragraph C8, to classify as a liability for incurred claims a liability for settlement of claims incurred before an insurance contract was acquired.
Paragraph C22A proposes that an entity applying the fair value approach could choose to classify such a liability as a liability for incurred claims.
We agree with the proposed amendment of the IASB and also support the comment made by EFRAG in its draft comment letter. However, the amendment should also apply to the post transition period in order to increase transparency and comparability with other portfolios and entities.
Additionally and in broader terms, we believe that the amendments proposed by the IASB for transition are insufficient insofar as they do not solve the main issues of entities when applying for the first time IFRS17.
In particular, under the fair value approach, the option to set OCI to nil, when an entity chooses to disaggregate insurance finance result between PL and OCI in accordance with IFRS 17.88(b), is not available to the related assets accounted at fair value through OCI. Setting OCI on the liabilities to nil at transition, whilst maintaining the historical OCI on related assets in products managed under cash flow matching will distort equity at transition and results going forward significantly. This problem affects especially long term insurance contracts, where interest rates at the transition date can be very different from interest rates at initial recognition of the contracts.
On the other hand, determining the amount in OCI retrospectively (in accordance with IFRS 17.C24(a)) introduces also a distortion in OCI in portfolios of underlying assets that have been restructured during the life of the policies, leading to a significant change in the overall interest rate of the portfolio
This issue would be solved by establishing the locked-in rate at the date of transition for the fair value methodology based on the rate of the underlying assets. This proposed approach minimises any accounting mismatch in equity as a similar discount rate is used in assets and liabilities and the OCI of the liabilities offset the OCI of the assets.
In more specific terms, ESBG proposes to amend paragraph C24(c) so this option would also be available for contracts measured under the general model and managed under cash flow matching techniques and not only for insurance contracts with direct participation features to which paragraph B134 applies.
(b) The proposed amendment to paragraph C3(b) would permit an entity to apply the option in paragraph B115 prospectively from the transition date, rather than the date of initial application. The amendment proposes that to apply the option in paragraph B115 prospectively on or after the transition date, an entity would be required to designate risk mitigation relationships at or before the date it applies the option.
ESBG agrees with the proposed amendment but also supports the comment made by EFRAG in its draft comment letter, related to the necessity to allow the retrospective application of the risk mitigation relief.
Additionally, ESBG believes that the main issue on transition remains unresolved (see answer to question 8.a)
(c) Paragraph C5A proposes that an entity that can apply IFRS 17 retrospectively to a group of insurance contracts be permitted to instead apply the fair value approach to that group if it meets specified criteria relating to risk mitigation.
ESBG supports EFRAG's response in its draft comment letter: the proposed amendment is a step in the right direction but a retrospective application of the risk mitigation of the risk mitigation relief would provide more relevant information.
However, as mentioned before, ESBG believes that the main issue on transition remains unresolved (see answer to question 8.a)
EFRAG additional questions to constituents
94. Do Constituents agree with the approach suggested by EFRAG, i.e. to prefer retrospective application of paragraph B115 instead of supporting the two consequential amendments? Please explain why.
95. If you expect to apply the risk mitigation retrospectively under the approach proposed by EFRAG, how would you find the required evidence in practice? What would be the starting point for collecting the evidence and what process would you use?
As mentioned before, ESBG agrees with the retrospective application of the risk mitigation relief as suggested by EFRAG. The risk of hindsight would be mitigated if appropriate documentation on risk management strategies exists and entities can prove with reasonable and supportable information that the conditions in paragraph B116 were met in the past.
(EFRAG CL – Appendix 2, Topic 2: Modified retrospective approach and fair value approach)
155 Please provide specific prevalent fact patterns where the application of the modified retrospective approach is proving particularly challenging in practice. This would assist EFRAG in understanding better the interpretation difficulties arising in obtaining reasonable and supportable information and in estimating missing information that is required to apply the modified retrospective approach.
We believe that the modifications currently permitted under the modified retrospective approach are too restrictive and do not make retrospective application possible in practice. In particular, the modified retrospective approach is considered impracticable for long-term life-saving products due to the large amount of high-quality historical data necessary for estimating the remaining amount of CSM. This data is not available in the companies' datapool (neither can be reconstructed based on own accounting historical data or based in previous SII cash flows) and we believe that reasonable and reliable estimates cannot be generated with the information currently available. In other words, we understand we would not be able to fulfil the “reasonable and supportable information" criterion. Examples of critical missing data, or not available without undue cost or effort, are the real cash flows previous to the transition date together with the historical assumptions used in measurement (actuarial or expenses) and how they have changed until the Transition date. Not all the changes in historical cash flows and assumptions are stored in the companies' data system.
Therefore, we believe that amendments are needed either under the modified retrospective or under the fair value approach in order to ease transition requirements and better portray the financial situation of Spanish life-saving business and French participating contracts.
In addition to the necessary amendment under the Fair Value Approach at the transition described above (answer to Question 8(a)), under the Modified Retrospective Approach, further simplifications are necessary. We believe that the IASB should to allow reasonable approximations and greater flexibility for insurers. As a minimum the following simplifications are necessary:
· Permitted to estimate the future cash flows at the date of initial recognition as the amount of the future cash flows at the transition date without adjusting by cash flows known previous to the transition date, and to apply a retrospective calculation only when it is possible (for example, for estimating the IFRS 17 discount rate).
· The cumulative OCI at transition date should be consistent with the financial assets associated, for all the portfolios managed by cash flows matching, as long as they are classified in the FV-OCI portfolio.
Question 9—Minor amendments (BC147–BC163)
This Exposure Draft also proposes minor amendments (see paragraphs BC147–BC163 of the Basis for Conclusions).
Do you agree with the Board's proposals for each of the minor amendments described in this Exposure Draft? Why or why not?
ESBG supports the proposed amendments with the following exceptions:
If paragraphs 24 and 28 are to be amended, we believe that paragraph 22 has also to be amended to refer to “contracts initially recognised more than one year apart".
We do not agree with the amendment to paragraph B128, which is not effective if underlying items include a mixture of assets and liabilities. In consequence, we propose that this amendment is excluded from the final version of IFRS 17.
99. Do Constituents consider that there are any unintended consequences arising from the minor amendments? Please explain.
100. EFRAG has heard two concerns which are described in the following paragraphs.
B128 of the amended IFRS 17
101 Paragraph B128 of the amendments to IFRS 17 clarifies that changes in the measurement of a group of insurance contracts caused by changes in the fair value of underlying items should be treated as changes in investments and hence as changes in the time value of money and financial risk. The concern is that there would be a misclassification between insurance service result and finance result requiring the presentation of non-financial items in the financial result.
Paragraph 28 of the amendments to IFRS 17 and paragraph 22 of IFRS 17
102. Paragraph 28 of the amendments to IFRS 17 indicate that in recognising a group of insurance contracts in a reporting period an entity shall include only contracts that individually meet one of the criteria set out in paragraph 25 of the amendments to IFRS 17. That is, based on:
(a) the beginning of the coverage period of the group of contracts;
(b) the date when the first payment from a policyholder in the group becomes due; and
(c) for a group of onerous contracts, when the group becomes onerous.
103. However, in paragraph 22 of IFRS 17, an entity shall not include contracts issued more than one year apart in the same group.
104. Using the issue date in paragraph 25 of the amendments to IFRS 17 instead of the recognition date for the grouping would have implications on, for example, for the discount rate and could create difficulties in terms of data availability causing operational issues and undue costs.
105. If you agree with either of the above two issues, please explain why this is an issue for you and the prevalence of the issue, including volumes and jurisdictions where the issue arises?
See our previous comments in question 9. We believe both issues must be addressed.
This Exposure Draft proposes to add to Appendix A of IFRS 17 the definition 'insurance contract services' to be consistent with other proposed amendments in this Exposure Draft.
In the light of the proposed amendments in this Exposure Draft, the Board is considering whether to make a consequential change in terminology by amending the terms in IFRS 17 to replace 'coverage' with 'service' in the terms 'coverage units', 'coverage period' and 'liability for remaining coverage'. If that change is made, those terms would become 'service units', 'service period' and 'liability for remaining service', respectively, throughout IFRS 17.
Would you find this change in terminology helpful? Why or why not?
ESBG does not have a view on this question at this point in time.
110 Do Constituents consider that there may be any unintended consequences arising from the proposed change in terminology? Please explain.
ESBG is not aware of any unintended consequences at this point in time.
Additional Issues - Annual Cohorts
The annual cohort requirement adds undue operational complexity to the measurement of insurance contracts. Many insurers issuing long term insurance contracts do not manage their business on an annual cohort basis and measuring contracts using groups that are inconsistent with the way the contracts are managed will not generate useful information.
In particular, ESBG believes that in long-term life contracts like annuities, IFRS17 requirement of annual cohorts is not consistent with current management practices and actuarial estimates and would lead to excessive granularity, complexity and costs (one-off and increasing on-going costs as the number of cohorts becomes larger over time).
Grouping by annual cohorts would not correctly portray the business performance of long-term saving-contracts managed with matching adjustment technics, so it would not provide additional value to users of financial statements. For such contracts, it may happen that the annual cohorts with greater seniority survive with a reduced number of policies. The small volume of contracts in these cohorts would introduce variability in the adjustments to the CSM and would increase the scope of potential “onerous" cohorts or “artificial" CSM.
This variability (positive or negative) is not a result of a negative deviation of actuarial assumptions, but derived from having a reduced number of contracts in senior cohorts. Actuarial calculations need of a sufficient number of policyholders in order to not suffer deviations in the expected future cash flows. Probability weighted future cash flows use probabilities of death/survival of the policyholders and will never be equivalent to the real cash flows of one individual policyholder. The more reduced is the number of policies, the more difference would arise by this effect when adjusting the CSM for changes in estimates of the present value of future cash flows (B96(b)) caused by an experience adjustment (for example, the death of a policyholder), even if there is no change made to the underlying assumptions (mortality or longevity).
We propose that the requirement to group contracts into annual cohorts is removed, on the condition that contracts issued in different years can only be aggregated if they were in the same profitability group at inception. The revised level of aggregation requirement should include a principle that requires an entity to set the unit of account based on the nature of its business and risk management.
If the requirement of annual cohorts cannot be removed, at a minimum, relief from the use of annual cohorts is needed (i) for in force business at transition, regardless the measurement model and the transition approach, (ii) for long-term saving-contracts under cash flow matching and (iii) for contracts with direct participating features that share a significant part of returns on underlying items across generations.
(EFRAG CL – Appendix 2, Topic 1: Annual cohorts)
140. For contracts with cash flows that affect or are affected by cash flows to policyholders of other contracts:
(a) EFRAG is suggesting to the IASB to provide an exception to the requirement to restrict the grouping of contracts using the annual cohorts. Would Constituents agree with this proposal? Please explain why or why not.
(b) Please provide fact patterns - and their prevalence - for which the application of the annual cohorts requirement results in added complexity that is not justified and, as a consequence, should be captured in such an exception. For example:
(i) Contracts to which the VFA applies compared to other contracts;
(ii) Contracts with full sharing of risks compared to other contracts that only share a substantial or significant part of the risks;
(iii) Contracts that share all risks or only particular risk types; and
(iv) Contracts with sharing of asset returns of underlying pools compared to other contracts.
141. As reported in paragraph 129, the exception should meet the reporting objectives of IFRS 17 (i.e. depicting profit trends over time, recognising profits of contracts over the duration of those contracts and timely recognising losses onerous contracts).
With reference to the pattern of recognition of the CSM, EFRAG in its case study received mixed results as to whether the resulting information would be impacted by the removal of the annual cohorts.
In your opinion, how would you ensure that the CSM release pattern would be in line with the IFRS 17 stated objectives? Do you envisage any loss of information as contemplated by the IASB in paragraph BC177 of the ED? If so, how would you address that loss of information?
142. Are there other types of contracts in the life insurance business, other than the contracts with cash flows that affect or are affected by cash flows to policyholders, that create similar complexity?
143. Some have observed that when a grouping approach broader than annual cohorts is applied, there is a benefit in providing additional information about trends in profitability. Such disclosure could include:
(a) Reconciliations for the CSM of those groups from the opening to the closing balances (according to paragraph 101 of IFRS 17)
(b) Disclosure on profitability trends by presenting the CSM effect of new business joining the groups, extracted from (a), as a series of historical data (for example, the last 3 years);
(c) Disclosure of the actuarial techniques applied for computing the CSM effect of new business joining the group as well as disclosure about the method used for assessing the profitability referred in (b).
Would Constituents consider it appropriate to include these additional disclosures?
142. See previous comments on the annual cohorts issue. ESBG believes that annual cohorts are not justified in long-term life-saving contracts managed with matching adjustment technics, as they do not provide useful information to users but introduce unduly variability in the adjustments of the CSM caused by experience adjustments in senior cohorts with a reduced number of policies.
These contracts are an example, additional to the one mentioned by the EFRAG in its draft comment letter, where the annual cohorts requirement must be re-considered. In particular, ESBG believes that the annual cohorts requirement should be removed, on the condition that contracts issued in different years can only be aggregated if they were in the same profitability group at inception. The revised level of aggregation requirements should include a principle that requires an entity to set the unit of account based on the nature of its business and risk management.
If a principle-based approach cannot be achieved, ESBG proposes that a relief from the use of annual cohorts is provided for long-term life-saving products measured under the general model and managed with matching adjustment techniques (not on an annual cohort basis).
A relief should also be provided for contracts with direct participating features that share a significant part of returns on underlying items across generations.
Additional Issues – Volatility in OCI introduced by IFRS 17 discount rates
ESBG is concerned about the variability in OCI introduced by IFRS 17 for Spanish long-term life-saving contracts that are not eligible to be measured under the variable fee approach.
Under the general measurement model (both PL and FV-OCI option) changes in the IFRS 17 discount rate after initial recognition do not lead to a remeasurement of the CSM, given that the CSM is measured at inception with the locked-in rate and not remeasured to reflect changes in this rate.
Even if the expected cash flows from an insurance contract are economically and perfectly matched with non-contractually disclosed financial assets that replicate those cash flows, including any long-term interest rate guarantee, an insurer will recognise in P&L/OCI amounts that go beyond the credit risk spread. This arises as a consequence of the CSM not being remeasured at each reporting date for changes in the discount rate.
The fact that the CSM is not remeasured for changes in the IFRS 17 discount rate is equivalent to having a portion of the insurance liability not measured on a current basis, giving rise to amounts recognised in P&L/OCI that do not offset completely (assuming there is not a spread credit risk) with the remeasurement at fair value of the corresponding financial instruments.
ESBG believes such a difference in measurement leads to an accounting mismatch that does not portray the economic net financial situation of Spanish long-term life-saving products. Spanish insurers will mainly apply the OCI option for the presentation of the insurance finance result, as their related assets will be mainly classified in FV-OCI portfolios under IFRS 9. In this context, ESBG is significantly concerned about the variability that will be recognised in OCI for these products under the general measurement model. It is important to highlight that Spanish users of insurers' financial statements place much emphasis on understanding the trend and evolution of the profit and loss and OCI statements, not expecting significant variability for the current business model under an economically matched balance sheet.
In order to solve this variability, a re-measurement of the CSM at each reporting date for changes in the discount rate should be permitted, including the effect in OCI , while keeping the other IFRS 17 current requirements unchanged. Such re-measurement would mitigate these accounting mismatches in OCI between IFRS 9 and IFRS 17. This proposal would apply to companies that apply the OCI option under the general measurement model, and some type of conditions or constraints could be set up to limit the remeasurement to certain types of insurance contracts (managed under matching adjustment techniques, for example).
The above suggestion would not change other current IFRS 17 requirements (i) to use the locked-in rate to accrete interest on the CSM, and (ii) to use the same locked-in rate to determine the adjustments to the CSM for changes in non-financial assumptions that affect future cash flows.
At the same time, we believe it would not affect any core principle of IFRS 17. In particular, the amounts recognised in OCI would naturally reverse over time and the insurance service result would be shown separately from the insurance finance result.
One of the most widespread types of insurance contracts used to promote the long-term savings of population in Spain is in the form of life annuities, both immediate and deferred annuities, which were tested in a complete case study from EFRAG.
Compared to other countries, Spanish insurers mainly provide a long-term fixed guarantee on interest rate to policyholders that does not change over time even if interest rates change.
This guaranteed interest rate to the policyholder is fixed by companies based on the observable market yield of the investment portfolio assigned to the age of the policyholder when the contract is underwritten. That is, the pricing of each policy depends on the observable market rates when the offer is made and an expected duration of the policy based on the age of the insured person.
From a simplified view, and considering the above pricing methodology, Spanish insurers earn a constant financial margin in these annuities that is the difference between the internal rate of return of financial assets (expected to be measured at FV-OCI under IFRS 9) and the guaranteed interest rate to the policyholder, while they are exposed to other non-financial risks (basically, deviation from the assumptions used in pricing in relation to longevity risk, to the risk margin or to operating expenses) that would determine the overall margin.
It has been around 20 years that the Spanish regulation incorporated financial immunization and asset-liability management (ALM) as methodologies for covering interest rate and spread risks for this type of contracts. The experience is borne out by the effective role that they have played in the control of the interest rate provided to the policyholder and the spread credit risk assumed by life insurance undertakings even through different macroeconomic environments (high and low interest rates, different phases in the business cycle…).
Although these annuities are economically matched and have specific backing portfolios of debt instruments supporting the cash flows to be paid to policyholders, they may not be eligible to be measured under the variable fee approach (VFA), as the policy contractually does not specify in all cases the financial assets on which the guaranteed profitability is based. Furthermore, when contemplating guaranteed benefits, the variation in the market value of the assets may not have a significant impact on the benefits expected to be paid to the policyholders. In particular, only in the case of surrenders before the maturity date the policyholder would receive the fair value of the underlying assets. This leads to companies assuming basically only default risk and reinvestment risk if there are deviations from expected duration.
It is relevant to mention that Spanish annuities are designed to provide the policyholder with access to an investment guaranteed return for the premium paid for the whole life of the policyholder, covering therefore the longevity risk. The company links the surrender value to the market value of the assets in order to not incur in investment risk, but not with the objective to allow the policyholder to share the market value of the investments. In fact, certain products include a penalization over the capital gains in order to discourage surrenders and, in general, surrenders are very unusual in these products.
As mentioned in the cover letter, the Spanish industry believes that IFRS 17 should portray the asset and liabilities management and the interaction between liabilities and their supporting assets. While any economic mismatches should be reported to users of financial statements, the performance and financial situation depicted under IFRS 17 should be consistent with the business model companies have in place and the sources of profit earnings.
Life insurance contracts are saving contracts which are issued by insurers on the French market. These contracts are very widespread and subscribed by policyholders in France as they offer specific guarantees, tax advantages, permit surrenders and, in case of death, the accumulated capital is transferred to designated beneficiaries in the contract.
Under these contracts, policyholders can invest in two types of funds (contracts can either include one of the two following funds but generally include both):
The “general fund" that backs the financial risks as the initial invested capital (and accumulated interests) is guaranteed by the insurer to the policyholder. These contracts share a significant part of returns on underlying items across generations as policyholders participate significantly in the returns of a common underlying pool of items over time. Assets are not dedicated to a specific generation of contracts and there is a sharing and transfer of wealth (returns of underlying assets) between the generations of policyholders (all policyholders, whatever the date of entrance in the contract, share equally in the returns of the pool of underlying items that has been constituted over time with their premiums).
The returns on underlying items are credited to the policyholders, either at the end of a given year or within 8 years (in that case, a specific reserve is booked in the balance sheet of the insurer). The premiums arising from these contracts are broadly invested in bonds.
The unit-linked fund in which the policyholder's capital is not guaranteed as he supports all the financial risk (the insurer guarantees a specific number of units but not their value). Therefore, the value of these contracts varies depending on the financial markets (nevertheless, specific guarantees such as GMDBs (Guaranteed Minimum Death Benefits), may be subscribed by the policyholder).
Under IFRS 17, these contracts are direct participating contracts eligible to the Variable Fee Approach model, allowing financial variations to be absorbed by the Contractual Service Margin.
Nevertheless, given the characteristics and mechanisms of these contracts (specifically the general fund) major issues arise regarding these contracts under IFRS 17. The main issue concerns the level of aggregation as the requirements under IFRS 17 (segregation by portfolio x profitability x generation) does not reflect the contractual terms, economics of these contracts nor the way French insurers currently manage them.
Indeed, these contracts are currently managed as a whole without any distinction by generation and the specific financial mechanisms (previously described) are incompatible with the level of aggregation requirements under IFRS 17. Hence, even though IFRS 17 allows cash flows to be determined at a higher level than cohorts to reflect mutualization mechanisms between groups of contracts, the CSM nevertheless still needs to be allocated by cohorts. French insurers therefore consider that segregating life-insurance contracts by annual cohorts will lead to an artificial and discretionary segmentation which does not reflect the way contracts are currently managed nor their characteristics. This segregation would only be performed for the purposes of IFRS 17 and could lead to counter-intuitive results compared to the economics of these contracts.
Moreover, applying the annual cohort requirements and following this segmentation over time for these contracts will lead to operational complexities (as they are not managed operationally at annual cohort levels) requiring investment in IT, actuarial and accounting systems.
Therefore, stakeholders on the French market strongly consider that IFRS 17 should exempt contracts eligible to the variable fee approach that share a significant part of returns on underlying items across generations from the annual cohort requirement. This exception, intended to apply to French life-insurance contracts, will lead to results which are more representative of the economic performance of the portfolio of contracts.
European Savings and Retail Banking Group – aisbl
Rue Marie-Thérèse, 11 ￭ B-1000 Brussels ￭ Tel: +32 2 211 11 11 ￭ Fax : +32 2 211 11 99
Info@wsbi-esbg.org ￭ www.wsbi-esbg.org
Published by ESBG. 04.09.2019
 “For the purpose of applying paragraph B119, the period of investment-return service or investment-related service ends at or before the date that all amounts are due to current policyholders relating to those services have been paid, without considering payments to future policyholders included in the fulfilment cash flows applying paragraph B68".
 In case that all the insurance contracts in the group are issued by the subsidiary.
 That either dies or survives, but does not die with a certain probability.
 Adjustments of the CSM for changes in estimates of the present value of the future cash flows in the liability for remaining coverage (B96(b)).