Book yield and effective yield approaches to presenting interest expense in profit or loss and Illustrative examples of book yield and effective yield approaches

In a set of papers Staff described the book yield approach (where the yield is directly related to the yield on underlying items) and the effective yield approach (where the yield is indirectly related to the yield on underlying items). The Staff’s proposed approaches were then illustrated in a number of scenarios.

The objective of the book yield approach is to reduce accounting mismatches between the presentation of interest expense in profit or loss and interest income from the underlying items when there is an economic match between the underlying items and the insurance liability. In order to apply a book yield approach an entity would identify the underlying items that determine the policyholder cash flows that vary with investment returns. It would then determine the basis of the accounting return for the specified underlying items to construct a yield curve based on the book yield at each reporting date. This yield curve would be characterised by including the yields from those items for the duration comprehended within that of the projected cash flows of participating contracts.

The Staff proposed that the IASB should consider only a book yield approach in which the book yield minimises accounting mismatches with underlying items. Consequently, the Staff’s proposed book yield approach would not be permitted when (i) equity instruments are measured at FVOCI, and (ii) investment properties are measured at cost and the policyholder receives a share of the capital gains. It would be permitted when the underlying items are, for example, (i) bonds that are accounted for at cost, FVOCI or FVPL provided that the effect in profit or loss of expected credit losses on the bonds accounted for at cost or FVOCI is reflected in the book yield, (ii) investment properties that are accounted for at cost when policyholders only benefit from a share of rental income, and (iii) when the underlying items are accounted for at FVPL.

The book yield approach would ensure that at inception of the contract, the yield curve for the presentation of the effect of discounting recognised in profit or loss is the same as the yield curve used for the measurement of the liability on the balance sheet.

The effective yield approach is a form of the effective interest method which is used to calculate the amortised cost of financial instruments and to allocate the interest income or interest expense in profit or loss. This approach would mean that profit or loss would isolate the underwriting result, while changes in market interest rates would be presented in OCI. Two variations of the effective yield approach are considered in this paper, being the level yield method and the projected crediting method. An effective yield is calculated on initial recognition of a contract as a single rate that exactly discounts estimates of expected future cash flows to the carrying amount of the liability determined on an amortised cost basis. The effective yield is reset when there are changes in estimated investment returns. The reset effective yield is the rate required to accrete amortised cost liabilities measured immediately prior to a change in estimated cash flows to equal the revised expected cash flows when they take place on a level basis. The effective yield is indirectly affected by the returns from the underlying items because the expected cash flows that are discounted reflect the estimated returns on the underlying items.

The projected crediting method calculates discount rates on a basis that reflects the entity’s projected crediting rates (i.e. the rates that the entity intends to use to determine the policyholder cash flows).

The potential for further refinements to the effective yield is discussed in the paper, but in the Staff’s view such amendments would increase the complexity of determining the effective yield, and on balance they did not recommend that the effective yield approach should be modified.

The Staff recommended that the projected crediting method should be adopted because of its ability to reduce mismatches between investment income and interest expense when there are changes in estimates. The Staff considered that interest based on crediting rates is closer to an incurred cost view of interest expense, as well as being more likely to mirror investment income when it is accounted for on an amortised cost basis.

The IASB members first discussed the book yield approach. It was noted that there are issues and trade-offs between the two approaches, and that volatility would be created if there were asset/liability duration mismatches. One IASB member questioned whether instead of using the OCI it would be better to unlock the shareholders share of the contractual service margin to reflect the effect of changes in discount rates. He also questioned how many insurers would use the book yield approach when FVTPL was available to them.

Another IASB member stated that in a perfect world the use of OCI would not be required, but insurance liabilities often extend beyond the duration of the backing assets, which is why the insurance industry feels that the use of OCI is preferable to avoid net profit being distorted.

Other IASB members expressed support for the book yield approach as it largely eliminated accounting mismatches.

There was some discussion over the circumstances in which it would be appropriate to permit or require the use of the book yield approach. The Staff confirmed that they were looking into using approaches both with and without the use of OCI.

The IASB members then discussed the effective yield approach. In response to a question raised by one of the IASB members, the Staff confirmed that underlying items were not restricted to asset returns but also included all variable returns e.g. the benefits of mortality gains and cost savings.

One IASB member stated that some financial assets would be accounted for at FVPTL because of the characteristics of the instruments, therefore use of the effective yield method would create accounting mismatches.

Another IASB member noted that the refinements to the effective yield considered in the first paper was effectively the book yield approach but without the use of different discount rates. He felt that such refinements did not make the approach too complicated, but his view was not shared by the other IASB members who commented on the same subject.

Use of OCI for contracts with participating features; should there be a book yield approach for determining interest expense in profit or loss?

In the first of its next couple of agenda papers, the Staff considered whether the effects of changes in discount rates should, or may be, presented in OCI, the scope of the book yield and effective yield approaches, and whether adaptations are needed to the general model for (1) participating contracts that the IASB believes should not be accounted for using either the book yield or the effective yield approaches, and (2) participating contracts that the IASB believes could be accounted for using the effective yield approach but should not be accounted for using the book yield approach.

The Staff recommended that:

1. For contracts with participating features, an entity should choose to present the effect of changes in discount rate in profit or loss, or in other comprehensive income, as its accounting policy, and should apply that accounting policy to all contracts within similar portfolios.
2. An entity should apply the OCI approach applicable to contracts with no participating features to insurance contracts where the cash flows that vary with investment returns on underlying items are not a substantial proportion of the total benefits to the policyholder over the life of the contracts.
3. An entity should apply the effective yield approach to contracts where the cash flows that vary with investment returns on underlying items are a substantial proportion of the total benefits to the policyholder over the life of the contracts, and:
   1. the returns to be passed to the policyholder do not arise from the underlying items that the entity holds;
   2. the policyholder will not receive a substantial share of the total return on the specified underlying items; or
   3. applying the book yield approach would create or increase accounting mismatches in profit or loss between the insurance contract and the underlying items.

The second agenda paper in this set considered whether the book yield approach should be used to determine the interest expense for contracts where the cash flows that vary with investment returns on underlying items are a substantial proportion of the total benefits to the policyholder over the life of the contracts, and (1) the returns to be passed to the policyholder arise from the underlying items that the entity holds, (2) the policyholder will receive a substantial share of the total return on the specified underlying items, and (3) applying the book yield approach would reduce or eliminate the accounting mismatches in profit or loss between the insurance contract and the underlying items.

The Staff recommended that:

1. When an entity presents the effect of changes in discount rates in other comprehensive income, the use of the book yield approach for the presentation of interest expense should not be allowed.
2. Consequently, an entity would determine the interest expense in profit or loss using an effective yield approach for all contracts in which the expected cash flows of the contract that vary with returns on underlying items are a substantial proportion of the total benefits to the policyholder over the life of the contracts.

One IASB member repeated his view that OCI should not be used for insurance contracts either with or without participating features. He felt that it was wrong for contracts without participating features, and although less wrong for contracts with participating features (if the refinements to the effective yield were to be adopted), this approach would be complex.

Another IASB member stated that there was an accounting policy choice for insurance contracts but there was less freedom for assets given the criteria set out in IFRS 9. He therefore felt that some criteria may be necessary to determine the most appropriate accounting policy for insurance contracts in order to minimise accounting mismatches.

Finally, another IASB member stated that if it was decided not to allow the use of OCI for contracts with participating features it would be necessary to do likewise for contracts without participating features. However, an alternative tool would be to unlock the contractual service margin. The Staff confirmed that if there were to be unlocking of the contractual service margin the use of OCI would become redundant.

Premium allocation approach: revenue recognition pattern

In the penultimate paper Staff considered whether to provide guidance on the pattern of recognition of the insurance contract revenue in the premium allocation approach (PAA), based on the provision of insurance coverage. The Staff recommended that under the premium-allocation approach, an entity should allocate insurance contract revenue in profit and loss in a systematic way that best reflects the transfer of services, and that the IASB should clarify that the transfer of services occurs:

1. On the basis of the passage of time and the expected number of contracts in force; but
2. If the expected pattern of release of risk differs significantly from the passage of time, then on the basis of the expected timing of incurred claims and benefits.

The Staff stated that guidance was needed on the recognition of revenue in order to prevent diversity of approach and a lack of comparability.

One IASB member noted that it was the value as well as the number of contracts that was relevant, and the Staff confirmed that they would clarify this.

Several IASB members questioned what was meant by the transfer of services, and whether this was limited to claims or also included other costs that are incurred over the duration of the contract and the stand-ready obligation to meet claims.

When called to vote by the Chairman, all 14 IASB members were in favour of the Staff recommendations.

Determination of interest expense in the premium allocation approach

This last paper discussed how to determine interest expense for the liability for incurred claims in the PAA, and in particular whether this discount rate should be the rate locked-in at the inception date of the contract, or the rate locked-in at the date when the claim was incurred.

The paper noted that if the rate at inception of the contract is used, a catch-up adjustment must be recognised in OCI to reflect the effect of changes in discount rates between the date of the contract inception and the date when the claim is incurred. This catch-up adjustment may be difficult to explain because no gains or losses would otherwise be separately recognised in the statement of comprehensive income relating to any changes in assumptions between the date of inception of the contract and the date the claim is incurred.

In the Staff’s view this catch-up adjustment would mean that using a discount rate locked-in on a date other than the date the claim is incurred would add complexity for users to understand financial statements without bringing significant benefits. The Staff recommended that, when an entity presents the effects of changes in discount rates in other comprehensive income, the discount rate that is used to determine the interest expense for the liability for incurred claims in the premium-allocation approach should be the rate locked-in at the date the claim was incurred.

When called to vote by the Chairman, all 14 IASB members were in favour of the Staff recommendations.

Additional feedback from separate ASAF meeting

In this month’s ASAF session on insurance contracts, the IASB staff sought further feedback on an alternative accounting model for participation contracts as presented at the September IASB meeting. Many ASAF members requested the IASB Staff to illustrate with examples how the proposed model would be applied to contracts with participating features when the returns to policyholders also includes participation in the insurer’s underwriting or operating performance such as mortality surplus or cost savings.

The ASAF members also debated the conditions for a contract to be deemed participating. The condition which requires the policyholder to receive a substantial share of the total return on underlying items resulted in a widespread concern among ASAF members on how ‘substantial’ would be defined and applied in practice. One of the ASAF members commented on his preference that no bright lines should be introduced in the new Standard as this could trigger accounting arbitrage.

In addition, the IASB staff debated the insurance contracts transition requirements with the ASAF, especially around the determination of the contractual service margin (CSM) when data is not sufficient to calculate it retrospectively. Three alternative approaches to determine the CSM on tradition were explored: 

1. Setting the CSM to zero which received no support from the ASAF,
2. Using fair value to determine the CSM which also received very limited support on the basis of feasibility to derive a fair value for the CSM and,
3. Using the premium that the entity would have charged the policyholder if it had entered into a contract with equivalent terms at the date of transition which was seen as a more practicable approach.

The ASAF members noted that they would not oppose other pragmatic methods that would result in achieving the goal of providing the closest proxy to a retrospective application.