The IASB held an education session on 19 May 2015 to consider the application of the variable fee approach to mutualisation, revenue and transition, and the accounting model that would apply to contracts that would not be eligible for the variable fee approach, but for which the cash flows vary with the return on assets (defined in the IASB Staff paper as “indirect participating contracts”).
The variable fee approach views the insurance contract as creating an obligation to pay to the policyholder an amount equal to the value of the underlying items less a variable fee for service.
In addition the IASB considered whether to provide an accounting policy choice for the presentation of changes in discount rate in other comprehensive income (OCI) or profit or loss for both direct and indirect participation contracts. Finally the IASB was provided with an update on the interaction between IFRS 9 Financial Instruments and the Insurance Contracts project.
Application of variable fee approach: Mutualisation (agenda paper 2A)
This paper provides background on the IASB’s tentative decisions on the level of aggregation, discusses what is mutualisation and when does it occur, and describes the implications of mutualisation on the determination of the contractual service margin (CSM) under the variable fee approach, and when gains and losses are recognised in profit or loss.
The Staff noted that the level of aggregation does not affect the measurement of the fulfilment cash flows, but it does affect the measurement of the CSM and, therefore affects when underwriting gains and losses are recognised in profit or loss.
In some insurance contracts with participation features, policyholders have agreed that they will bear the risks of other policyholders that share in the same pool of underlying items and share the risks of the expenses arising from them, including the cost of any guarantees written to other policyholders in the pool. This reduces the direct exposure of the insurer to the collective risk of the group, which this paper refers to as mutualisation. When the total returns of the underlying items are insufficient to pay the expenses, even after reducing the payments due to other policyholders through mutualisation, the entity will be responsible for those expenses.
Mutualisation arises from a subset of contracts with participation features that qualify for the variable fee approach, and occurs when the contract specifies the returns from the underlying items that the policyholder participates in and that the returns that are finally passed to the policyholder may be reduced by any guarantees to other policyholders. One of the criteria for the variable fee approach is that the contract specifies that the policyholder participates in a clearly identified pool of underlying items.
If a policyholder enters into a contract that qualifies for the variable fee approach, and that contract specifies mutualisation with other policyholders, the transfers between groups of policyholders could be reflected in the determination of the CSM. Consequently, if the level of aggregation is determined taking into account the mutualisation no losses are recognised in profit or loss when a group of policies becomes onerous if another set of policyholders bears those losses, and losses are only recognised in profit or loss when the underlying items in the fund as a whole are insufficient to bear those losses and the portfolio becomes onerous to the insurer. Consequently, some think that there should be additional requirements to increase the transparency of the risks arising from mutualisation agreements. Additional disclosures could increase the transparency of the risks arising from mutualisation arrangements.
An IASB member supported the suggestion that there should be disclosure about losses being absorbed by other policyholders, and felt that the information needed for such disclosures should be available to management as they should be tracking such losses. Other IASB members questioned whether this would provide useful information. The Staff noted that any risk of losses being in excess of those that could be borne by other policyholders and hence becoming an onerous contract for the entity should be included in the existing risk disclosures.
In response to a question by an IASB member the Staff confirmed that the principle of recognising in profit or loss the CSM on maturity of a contract still applies, but because of mutualisation the need to recognise losses is unlikely to arise.
An IASB member expressed the view that where an insurer exercised discretion to reduce the amount credited to a policyholder in order to meet losses of other policyholders, this resulted initially in a profit to the insurer, who then decided to use this profit to meet losses on other contracts. Consequently he felt that economically this was not mutualisation. Another IASB member expressed the view that there was the danger of concentrating on the contract terms at the risk of missing the underlying economics.
An IASB member questioned the proposed conditions regarding the entity’s interest in underlying items being viewed as a variable fee for service, whereby the entity expects that a substantial proportion of cash flows from the contract will vary with changes in underlying items and the entity expects the policyholder to receive an amount representing a substantial share of the returns from the underlying items. He asked whether instead of ‘substantial’ a specified proportion could be the criteria for the variable fee approach to be applied. The Staff will consider this point further.
Application of variable fee approach: revenue (agenda paper 2B)
This paper explains the application of the revenue proposal for the variable fee approach used to measure insurance contracts with direct participation features.
The entity would recognise revenue in each period as it satisfies the performance obligations, and exclude from revenue any investment component. Insurance contract revenue could be determined using changes in the liability for the remaining coverage or by reference to the fulfilment cash flows and the CSM.
As insurance revenue represents the amount of the release from the liability for remaining coverage, an entity needs to separate the liability for the remaining coverage from the liability for incurred claims. It would also need to separate from the liability for remaining coverage any amounts related to losses previously recognised in profit or loss (when the contract was loss making). An entity would need to track this amount for recognition purposes because the IASB had tentatively decided that if there are favourable changes in the estimates an entity needs to reverse previously recognised losses before it rebuilds the CSM.
In some circumstances insurance risk and expected insurance claims are not significant for contracts with direct participation features (e.g. investment contracts with discretionary participation, in the scope of the IFRS for insurance contracts). In those circumstances, revenue for the period would be mainly related to the release of the CSM and the risk adjustment and the allocation of acquisition costs and other expenses.
In response to a question raised by an IASB member the Staff explained why the 2013 Exposure Draft did not propose a similar requirement to that contained in IFRS 15 whereby the amount of revenue recognised in profit or loss be constrained to the extent that it is highly probable that significant reversals in the amount of cumulative revenue recognised will not occur. This is because a current expected value model had been chosen for insurance contracts, and revenue was unlikely to be recognised in full until the contract matures if the IFRS 15 “high probability” constraint applied.
Application of variable fee approach: transition (agenda paper 2C)
This paper considers whether the IASB’s previous tentative decisions relating to transition would need to be modified if the IASB were to require that entities apply the variable fee approach to measure insurance contracts with direct participation features.
The IASB has proposed that an entity should measure and present insurance contracts at the beginning of the earliest period presented retrospectively if practicable. When this is not practicable, a simplified retrospective approach would be applied that would enable entities to approximate retrospective application for the measurement of each component of the fulfilment cash flows at initial recognition of the insurance contract. The simplified retrospective approach might still be impracticable if the entity could not obtain information about the cash flows that occurred between the initial recognition of the contract and the beginning of the earliest period presented. Accordingly, the IASB tentatively decided that in such circumstances an entity should use a “fair value approach” for determining the CSM at initial recognition of the insurance contract. An entity would determine the CSM at the beginning of the earliest period presented as the difference between the fair value of the insurance contract and the fulfilment cash flows measured at the beginning of the earliest period presented, and determine interest expense, and the related amount of OCI accumulated in equity, by estimating the discount rate at the date of initial recognition using the simplified retrospective approach.
To comply with the principle of retrospective application, the entity would need to determine the CSM at the beginning of the earliest period presented as if the new IFRS had always been applied. The entity would not require historical information to determine the variable fee for service, but it would require historical information to determine the cumulative amounts of the CSM/(loss) recognised in profit or loss. The Staff believes that retrospective application would often be impracticable for entities applying the variable fee approach because estimating that information would require the use of hindsight. Furthermore, the Staff notes that the simplified retrospective approach could not readily be applied to contracts that are accounted for using the variable fee approach because this estimates the CSM at the beginning of the earliest period presented as if all changes in estimates before that date were known at initial recognition. Those changes include changes in the entity’s share of the fair value of underlying items (that would have unlocked the CSM) at initial recognition without the use of hindsight.
The Staff proposed that the IASB considers two approaches to address this difference. Option 1 is to not provide additional simplification for the variable fee approach. The Staff notes that this would mean that an entity applying the variable fee approach would generally apply the fair value approach, but the objective of this approach is not to estimate a value of CSM had the standard always been applied. Consequently applying a fair value approach would mean that there would be reduced comparability between contracts written before and after the earliest period presented. Option 2 would provide an additional simplification for the variable fee approach to the retrospective simplified approach. In order to estimate how much CSM would have been released in each period between initial recognition and the beginning of the earliest period presented, the entity could assume that the CSM at initial recognition was released according to the pattern of services. The Staff believes that results achieved using this method could differ from those calculated using the retrospective approach only because an entity would assume that all changes in the variable fee for service (resulting from changes in the fair value of the underlying items and fulfilment cash flows) were known at initial recognition of the contract. This is similar to the simplified retrospective approach for non-participating contracts and on that basis the Staff believes that this method would provide a reasonable approximation of the retrospective approach.
Historical information might be needed to estimate the accumulated balance of OCI recognised at the beginning of the earliest period presented when the current period book yield approach is applied. The Staff believes that this would be often impracticable without using hindsight. Consequently, the Staff proposed that the IASB provides a simplification to enable entities to approximate the cumulative OCI balance for insurance contracts.
The entity would assume that there are no differences in the accumulated balance of OCI for the insurance contracts and the underlying items because of differences in timing between the initial recognition of the insurance contracts and the underlying items. It would also assume that the accumulated balance of OCI for the insurance contract is determined as follows:
- when underlying items are measured at FVPL, there would be no amounts accumulated in OCI for both underlying items and insurance contracts;
- when underlying items are measured at FVOCI, the accumulated balance of OCI for the insurance contracts would be equal and opposite to the accumulated balance of OCI recognised for the underlying items; and
- when underlying items are measured at amortised cost, the accumulated balance of OCI for the insurance contracts would be the difference between the amortised cost of the underlying items and their fair value.
An IASB member commented that he had no objection to further simplification for transition as a whole. He was in favour of a single objective for transition, and expressed doubts as to whether retrospective application would result in greater certainty or comparability given the estimates and assumptions that would be involved.
Another IASB member stated that the simplification to enable entities to approximate the cumulative OCI for the insurance contracts and the underlying items seemed sensible, but she needed to understand this proposal better in order to understand why the approach labelled “option 2” is a good proxy for retrospective application.
An IASB member stated that in many respects the fair value approach has a clearer objective, but would result in a different CSM where market conditions had changed since the inception of a contract. Another IASB member expressed the view that the fair value approach may be more appropriate as this would be more relevant and reliable than trying to apply retrospective application where data did not exist or was not reliable. A further IASB member questioned whether retrospective application would be suitable for long duration contracts when the business model had changed since the date of inception of a contract.
Proposed accounting for indirect participation contracts (agenda paper 2D)
An indirect participation contract has cash flows that vary with the returns on assets, but does not create an obligation to pay the policyholder an amount based on the underlying items less a variable fee for service. Accordingly, an entity would not be able to apply the variable fee approach to these contracts.
Measurement - Application of previous tentative decisions for measurement
When a contract includes asset-dependent cash flows, the initial estimate of the fulfilment cash flows is determined using the entity’s estimate of the expected cash flows, discounted using a discount rate that reflects the extent of any dependence on asset returns.
After initial recognition, the fulfilment cash flows could change. The changes in estimates that arise as a consequence of changes in asset gains or losses, and the corresponding change in the discount rates would be recognised in the statement of comprehensive income. Changes in estimates of the participation percentage affect the consideration the entity will receive in return for undertaking the obligations provided by the contract. Consequently the entity would recognise changes in the estimates of the profits for future services as an adjustment to the CSM, and changes in the profits for services in the current and past periods immediately in profit or loss. The CSM is not explicitly re-measured in the IASB’s tentative decisions.
The IASB has concluded that the discount rate used to determine the change in fulfilment cash flows that adjust the CSM should also reflect the characteristics of the cash flows of the insurance contract, and should be determined at the date of initial recognition of the insurance contract. The measurement of the CSM subsequent to initial recognition reflects the accretion of interest on the CSM using locked-in rates, as the IASB was persuaded that this is conceptually correct, and would result in more useful information because it retains a clean separation of underwriting and investment results.
Implications for indirect participation contracts of the variable fee approach
The CSM after initial recognition would differ between the general model and the variable fee approach. In the variable fee approach, changes in the estimates of the variable fee for future services that adjusts the CSM reflect the current period’s estimate of asset returns. This means that the adjustment to the CSM is determined using the discount rate at the date of the change in estimate, the rate used to accrete interest on the CSM is a current interest rate and the opening balance of the CSM is re-measured to reflect changes in discount rate. In the general model the adjustment to the CSM is determined using the discount rate at initial recognition, the rate used to accrete interest on the CSM is the interest rate at initial recognition and the opening balance of the CSM reflects the interest rate at initial recognition.
Application of the IASB’s tentative direction for the disaggregation of interest expense for indirect participation contracts into an amount presented in profit or loss and in OCI
The IASB has considered three variations to the approach to disaggregate interest expense. These are a level yield method that would determine the interest expense in profit or loss using a single discount rate that exactly reverses out any amounts recognised in OCI over the life of the contract, a projected crediting rate that reflects the pattern of expected crediting rates, and a modified effective yield approach that would address the accounting mismatches that might arise between interest expense and investment income when an effective yield approach is applied where the underlying items are a mix of assets measured at FVPL and cost, the underlying items measured at cost are sold and a realised gain or loss is presented in profit or loss (without a corresponding change in amounts credited to policyholders).
Implications for the effective yield approach arising from the current period book yield approach
The Staff proposed at the previous meeting that the current period book yield approach should only apply when there is no possibility of an economic mismatch, i.e. when the entity’s obligation is to pay to the policyholder an amount equal to the value of the underlying items less a variable fee for service and the entity holds the underlying items. The Staff proposed that an effective yield approach for determining interest expense should apply when a contract does not qualify for the current period book yield approach.
The Staff thinks that contracts that are not eligible for the current period book yield approach can be classified between those contracts where the cash flows do not necessarily reflect the cash flows of the assets and those where they do reflect the cash flows of specified assets. For the former category of contracts, modifying the effective yield approach so that the investment expense reflects the investment income on the assets held would not portray an accurate depiction of the relationship between the assets and the insurance contract. The latter category of contracts would be eligible for the current period book yield approach, provided that the entity held the specified assets. However, where an entity did not apply this approach e.g. because it no longer met the criteria for the application of the current period book yield approach, it may be justified to amend the effective yield approach to eliminate accounting mismatches between the cash flows of the insurance contract and the cash flows of the assets. However, in the Staff’s view, the effective yield approach should not be modified to reduce the accounting mismatches because if the effective yield approach is to apply to all contracts in which the cash flows vary with changes in investment returns, then adjusting the effective yield on the insurance contract to reflect differences that arise in the timing of the recognition of gains and losses on the insurance contract would only increase the complexity of determining the effective yield and make it more difficult to understand its objective. Further, the investment expense in profit or loss should report on an accruals basis the investment expense incurred in the period, regardless of the pattern of crediting/notifying the policyholder of its entitlement to those payments. Determining investment expense based only on the crediting rates for the period, and ignoring the effect that changes in investment returns will have on future crediting rates, is inconsistent with the expected cash flow principles in the proposed IFRS.
The Staff has revised its previous view when they recommended the projected crediting rate version of the effective yield. The Staff now recommends that the IASB should use an effective yield approach, on the assumption that there would be a revised scope of the effective yield approach, and an OCI approach specifically when there is a clear link between underlying items and the cash flows of the insurance contract.
An IASB member stated that unlocking the CSM as a result of exercising discretion was good, but it was necessary to set the definition of discretion more clearly.
Another IASB member felt that the most important issues to tackle are the scoping of the fair value approach and the application of the indirect participation contracts approach, and the slicing of contracts into different pools. Examples would be helpful to make the approaches more understandable. He questioned whether having different approaches was worth the additional complexity, or whether it would be better to have one approach.
An IASB member felt that the CSM could be re-measured, and this was implicitly being done when the variable fee approach is applied. Several IASB members expressed a preference for using a current rate to unlock the CSM, as they considered that this was more meaningful. An IASB member considered that not using a current rate would result in an accounting mismatch which may drive insurers to using the OCI solution. The Staff felt that this may be an economic mismatch rather than an accounting mismatch.
An IASB member asked whether there would be any criteria for when an insurer would cease to qualify for the current period book yield approach resulting from the absence of 100% matching with the underlying assets. The Staff stated that this would be a matter of judgement, but an IASB member felt that some guidance on this issue would be needed.
Presentation of interest expense for contracts with participation features – whether to provide an accounting policy choice (agenda paper 2E)
The IASB had tentatively decided that an entity should choose as its accounting policy to present the effect of changes in discount rates in profit or loss or in OCI. Furthermore, that accounting policy choice would apply to all contracts within a portfolio. In reaching this conclusion the IASB sought to balance the competing demands of understandability and comparability. Consequently, entities would need to apply judgement as to the benefits and costs in deciding whether to present changes in discount rate in OCI or in profit or loss.
Effective yield approach
There are many similarities between the effective yield approach and the use of locked-in rates for contracts without participating features. Applying the effective yield approach to contracts without participation features where there are no cash flows that vary with asset returns would result in the same outcome as applying locked-in rates. Furthermore, an entity applying the effective yield approach to present the effect of changes in discount rates in OCI would need to make additional calculations to derive separate amounts on profit or loss and in OCI, and permitting the effective yield approach creates additional complexity for users.
Permitting the same choice of accounting policy for indirect participation contracts and contracts without participation features would avoid differences in the outcome if an entity was required to apply the effective yield approach to indirect participation contracts, and an entity could choose as its accounting policy whether to apply the locked-in discount rate approach to contracts without participation features. Such differences would be difficult to explain; therefore the Staff believes that an entity should choose as its accounting policy to present interest expense either all in profit or loss, or in profit and loss and OCI using the effective yield approach.
Current period book yield approach
Under this approach the entity determines the investment expense on the insurance contract liability as equal and opposite in amount to the investment income on the underlying items that are reported in profit or loss. Any difference between the interest expense reported in profit in profit or loss and the interest expense determined on a current basis would be reported in OCI. The key advantage of this approach is that it fully eliminates any accounting mismatch between insurance investment expense on the liabilities and the investment income on the assets when there could be no economic mismatch between them. In addition, the Staff thinks that this approach is less complex to apply than other approaches.
However, the Staff noted that complexity could arise once a portfolio of insurance contracts ceases to qualify for the current period book yield approach after inception because it no longer holds the underlying items. In these circumstances the Staff thinks that the entity should be able to apply the approaches for presenting changes in discount rates for contracts that do not qualify for the book yield approach, i.e. to select an accounting policy for recognising the changes in discount rates in either profit or loss or in OCI using the effective yield approach.
The Staff therefore believes that for a contract that qualifies for the current book yield approach at initial recognition, the entity should have an accounting policy choice to determine interest expense in profit or loss using the current period book yield approach, or the effective yield approach, or current discount rates (the entire effect of discount rate changes being in profit or loss under this last approach).
An IASB member questioned why the use of an effective yield approach should be permitted when an insurer qualified for the current period book yield approach as he felt that the effective yield is an inferior approach. The Staff responded by stating that the effective yield approach may be a more appropriate approach when it was expected that the insurer may not qualify for the current period book yield approach at a future date, as this would avoid the need to transition to the effective yield approach.
An IASB member questioned whether the need to transition from the current period book yield approach would arise when it was recognised that following a change in circumstances the insurer would not be able to continue to hold the underlying asset at some future date. The Staff stated that the intention was that it would only be possible to cease to apply the current period book yield approach once the insurer ceased to hold the underlying items.
An IASB member expressed concern at the number of options that were being considered, and sated that she would prefer choices only when it was essential to address fundamental concerns rather than to attempt to find solutions that were as equitable as possible. Another IASB member stated that tensions were created when criteria and ‘bright lines’ were introduced.
An IASB member felt that the application of the OCI solution for indirect and participating contracts was less likely because of asset liability management, which also strengthened the arguments for using the current period book yield approach. He also stated that decisions were needed that will make the financial statements understandable to preparers and users. The Staff asked whether this indicated a desire to mandate the use of the current period book yield approach. An IASB member responded to this by expressing the concern at creating different models and the inability to continue to use the current period book yield approach even where the economics had not changed significantly.
Update on the interaction between IFRS 9 Financial Instruments and Insurance Contracts project (agenda paper 2F)
The effective date of the new insurance contracts Standard can no longer be aligned with the effective date of IFRS 9 because of the IASB’s intention to allow approximately three years between publishing the new insurance contracts Standard and the date when that Standard becomes effective, and the effective date of IFRS 9 is 1 January 2018. Accordingly, the IASB tentatively decided to confirm the transitional reliefs proposed in the 2013 Exposure Draft that will enable entities to align classification elections under IFRS 9 with how insurance contract liabilities are accounted for both when IFRS 9 is initially applied and when the new insurance contracts IFRS is initially applied at a later date, and to consider providing further transitional relief to permit or require an entity to reassess the business model for managing financial assets when the new insurance contracts Standard is initially applied.
At the March 2015 meeting of the Accounting Standards Advisory Forum (ASAF) the European Financial Reporting Advisory Forum (EFRAG) requested the IASB to reconsider its position of not deferring IFRS 9 for entities that issue insurance contracts, as implementation of IFRS 9 without simultaneous implementation of the new insurance contracts Standard could create disruption for users of financial statements of entities that issue insurance contracts, and make understanding of the financial position and performance of insurance businesses more difficult without commensurate benefit. This request was supported by some of the European ASAF members. Non-European ASAF members indicated that any further transitional reliefs for entities that issue insurance contracts should be optional rather than mandatory so as not to penalise entities that had already begun implementing IFRS 9.
On 4 May 2015, EFRAG issued a draft endorsement advice (DEA) that recommended IFRS 9 endorsement in the European Union without further delay. In the DEA, EFRAG also repeated its concerns as expressed at the ASAF meeting, and advised the European Commission to ask the IASB to defer the effective date of IFRS 9 for insurance businesses and align it with the effective date of the new insurance contracts Standard.
The Staff will continue to monitor the developments and provide updates to the IASB, including updates on further insights and evidence of the potential effects of IFRS 9 being implemented in advance of the new insurance contracts Standard.
An IASB member noted that the above issue had been discussed at the Insurance Accounting Working Group within EFRAG and at other meetings with the insurance industry. He expressed the view that although the volatility of asset values had been discussed, there had not been an adequate discussion that also considered insurance liabilities.
The Staff stated that the information that had been provided to them is very high level.
The expected publication of the new insurance contracts Standard is now likely to be in the middle of next year.
The IASB will consider a system to facilitate testing and implementation of the new insurance contracts Standard. Deloitte believes that this system could result in the creation of a Transition Resource Group for Insurance Contracts similarly to those created for Revenue Recognition and for Impairment of Financial Instruments.