Insurance contracts

Date recorded:

The IASB held an education session to discuss the possible use of other comprehensive income (OCI) for presenting certain changes in an insurance liability. As this was an education session, the Board was not asked to make any tentative decisions. Instead, the staff intends to bring this issue back to the joint IASB and FASB meeting in April 2012.

Background

In accordance with the IASB’s Exposure Draft ED/2010/8 Insurance Contracts (Insurance ED), insurance contracts would be measured on the current basis with all income and expense included in profit or loss. Remeasurement through OCI would not be permitted or required.

However, while many constituents preferred the accounting for insurance liabilities to be a current measure, certain constituents expressed concern with the resulting volatility in profit or loss. Others expressed a view that short-term gains and losses due to market movements were not relevant to financial statement users given that they ignored the long-term nature of insurance contracts and obscured underlying long-term performance. Some felt that the underwriting results were overshadowed by market volatility.

Still others felt that it was important to reflect ‘economic volatility’ in profit or loss but not the volatility from accounting mismatches arising from insurer’s financial instruments held at amortised cost or at fair value through OCI. As a result, many constituents suggested application of a fair value option to minimise the perceived accounting mismatch with the use of OCI to present market volatility outside of management’s control.

Given that the IASB agreed to re-open IFRS 9 Financial Instruments (classification and measurement) to consider a third classification of fair value through OCI (FVOCI) for financial assets backing insurer’s liabilities, the staff presented a paper researching potential uses of OCI for insurers.

Potential use of OCI

The staff presented background information surrounding the possible use of OCI to present changes in the insurance liability arising from changes in discount rates and interest rate sensitive cash flow assumptions (e.g. embedded options, minimum return or interest guarantees, lapse assumptions and inflation adjustments).

This was presented as two options: to isolate in OCI only the changes in discount rates from contract inception or to present in OCI changes in interest sensitive cash flows as well as changes in the discount rate. The staff noted arguments for the use of OCI, including:

  • accounting mismatches are reduced assuming assets are also measured at FVOCI
  • short-term movements in the discount rate are removed from profit and loss given that they do not reflect the long-term nature of insurance
  • underwriting results are not overshadowed by market movements and continue to be reported in profit or loss
  • information about economic mismatches (e.g., duration, options and guarantees) is presented in a transparent manner in OCI, although several Board members noted that transparency is only available if viewed in conjunction with net profit or loss (i.e., transparency is present if considering total comprehensive income as opposed to considering profit or OCI in isolation).

Board members cited many arguments against the use of OCI, including:

  • economic mismatches arising from duration mismatches, credit spreads and options / guarantees are presented in OCI rather than profit and loss
  • a general concern as to use of OCI when the conceptual basis for presentation of items within profit or loss or OCI has not been adequately established
  • potential inconsistencies with the IFRS 9 Financial Instruments model regarding use of OCI. Specifically, multiple Board members noted that the IASB recently agreed to consider making improvements to IFRS 9 and, in particular, to consider the interaction with the insurance contracts project. One area planned for consideration in the IFRS 9 improvements project is the possible expanded use of OCI or a third business model for certain debt instruments. Board members sought consistency in any decisions taken on use of OCI in that project with decisions taken herein.

One Board member saw constituent concerns as being driven by a desire to have a current measure of the insurance liability without the effects of any changes through profit or loss. In considering the implications of such a decision with guidance in IFRS 9, he noted that for those assets requiring fair value measurement, OCI would not provide any relief. For those measured at amortised cost the use of OCI would shift a liability generated mismatch from profit or loss to equity. Therefore, he believed that to avoid creating additional mismatches, insurers using OCI for insurance liabilities should be precluded from using an amortised cost model for financial assets and instead should be limited to two categories: fair value through profit or loss or FVOCI:

  • complexity in disaggregating changes in discount rates and interest rate sensitivity cash flows from other assumptions for purposes of presentation in OCI as compared to profit or loss. Some Board members saw some conceptual merit in isolating only the impacts of discount rates, while others were concerned both with operational difficulties and the resulting meaningfulness of the profit and loss and OCI. Specifically, one Board member was concerned that the disaggregation of changes in discount rates on interest rate sensitive cash flows could not be measured reliably. The staff indicated that the practical implications of disaggregating changes would require a preparer to run multiple liability models to isolate the effects of specific assumptions and inputs
  • if OCI was used to isolate both the changes in discount rates and interest sensitive cash flows, there were concerns that this would result in mixing the changes in Building Block 2 – discount rates and changes in Building Block 1 – cash flows. There were several Board members concerned that these cash flows were real cash flows that would not naturally reverse over time (if, say a guarantee is triggered) and will potentially never be reflected or recycled in profit or loss; thereby obscuring the economic performance of insurance contracts
  • reducing the accounting mismatch through the use of OCI may result in a mismatch in equity for assets measured at cost
  • concern that whilst the OCI approach is being proposed to reduce short-term volatility in profit or loss it would instead remove all volatility from profit or loss (except for experience adjustments and changes in assumptions). This Board member preferred to record all changes in the liability in profit or loss, with segregation of the changes in profit or loss so that users could understand the source of the short-term volatility.

Following from the above concerns, the staff noted that the use of OCI to record changes in the insurance liability arising from changes in discount rates and interest rate sensitive cash flow assumptions would require further consideration in a number of areas. Those areas included whether the use of OCI should be permitted or required and at what level should the unit of account be determined (e.g., contract level consistent with the fair value option or portfolio level consistent with the measurement of the liability).

While no formal decisions were taken, little support was visible for the accounting through OCI for changes in the insurance liability arising from changes in discount rates and interest rate sensitive cash flow assumptions.

Loss recognition test

The staff noted that some constituents consider a contract to be loss making when asset returns are lower than expected (e.g., if an insurer prices its contracts assuming that it can earn interest of 7 per cent on its investments, but now market returns are only 3 per cent, this may be indicative that a contract is loss making). Therefore, the staff intends to formally discuss in a future meeting the possible inclusion of a loss recognition test in the insurance contracts proposals.

A loss recognition test would accelerate the recognition in profit or loss of future insurer’s potential losses arising from a mismatch between insurance liabilities and assets backing them. The accelerated recognition would take place in the period in which management is first aware of those losses. This would provide signalling information that the insurer will have to draw on its capital to fulfil the insurance liability.

Board members expressed many concerns with this test, including:

  • a test that considers the returns of the assets is inconsistent with the Boards’ decision that the discount rate for the liability should reflect the characteristics of the liability
  • a view that this test represents a unique from of onerous and impairment tests as those tests do not consider the performance of other assets or liabilities.

The staff presented multiple alternative loss recognition test triggers to the Board. These alternatives included:

  • (Alternative 1) (liability discounted at today’s discount rate less liability discounted at the rate at inception) > margin
  • (Alternative 2) (liability discounted using return on investment less liability calculated using the discount rate at inception) > margin, and
  • (Alternative 3) when qualitative factors, to be specified, indicate that the expected return on the assets is lower than the liability’s discount rate at the inception.

In summarising his views, one Board member noted that none of the alternatives were attractive. Although not having a loss recognition test results in net income which lacks meaning because it would use a discount rate locked in at inception and would not reflect the changes in the insurance liabilities’ current value or the changes in the insurer’s assets, he specifically noted that:

  • Alternative 1, for situations in which an entity has ‘blown through its margin’, may result in an entity having no underwriting margin going forward and a rather misleading net investment income. For situations in which the entity has not ‘blown through its margin’, any reduction in interest rates would result in a loss. Therefore, losses would ultimately be reflected in profit or loss and gains would be reflected in OCI.
  • Alternative 2 results in a linkage to asset returns, which is contrary to the principle in IFRSs that liabilities are not measured based on assets.

This Board member concluded that gains and losses should be recorded in profit or loss as opposed to OCI and that this would immediately alleviate any concerns for the need of a loss recognition test.

No support was visible for the use of a loss recognition test. However, no formal decisions were taken.

Correction list for hyphenation

These words serve as exceptions. Once entered, they are only hyphenated at the specified hyphenation points. Each word should be on a separate line.