Post-employment benefits

Date recorded:

Research Project: Post-employment benefits

Recap

The existing requirements of IAS 19 do not always provide relevant information when the risks of post-employment plans do not fall solely on an entity or solely on employees—ie it is a hybrid, reflecting some characteristics of a defined benefit plan and some characteristics of a defined contribution plan.  An example given by the staff is a plan where payments depend on the return of plan assets but the employer guarantees a minimum return.

The IASB has a research project that is seeking to identify a conceptually sound and robust model for accounting for post-employment benefits and to gather information about the trends in pension plans.

At this meeting, the staff will present three agenda papers but is not asking the IASB to make any decisions.

Global trends in pensions (Agenda Paper 15A)

The staff analyses that the use of hybrid plans is increasing around the world while traditional defined benefit plans are on the decrease. The agenda paper contains details of the information gathered and how the staff derives their analysis.

Potential models that might address the issue of contribution-based promises and other hybrid plans (Agenda Paper 15B)

The staff presents an analysis of several models that could potentially solve the issues around hybrid plans. Those models are:

(a)    The current IAS 19 model:

Under this model, the entity would apply the projected unit credit method to hybrid plans. This leads to a mismatch between expected returns on plan assets and discount rates based on (usually lower) bond rates and might not always provide relevant information for hybrid plans.

(b)   Models consistent with the proposals in the Conceptual Framework project:

        i. A fair value model:

Under this model, liabilities for contribution-based promises would be accounted for at fair value, assuming the terms of the benefits do not change. If the employee is guaranteed the higher of two or more possible outcomes (as in the ‘guaranteed return’ example), this ‘higher of’ option would be accounted for separately. The staff find that under the reasoning of the Exposure Draft of the revised Conceptual Framework, fair value would not be the best answer for pension accounting because the entity usually fulfils the liability rather than transferring it to another party.

         ii. A customised fulfilment value model:

Under the IASB’s approach to measuring insurance contracts, measurement for an obligation to pay net future cash outflows (fulfilment cash flows) would include a current, unbiased estimate of the cash flows expected to fulfil the contract, an adjustment for the time value of money and an adjustment for risks and uncertainty. As insurance contracts and pensions share similarities, this model could be applied to pensions as well. The staff conclude that this model would solve the issues relating to hybrid plans.

(c)    Other practical models suggested in the past:

         i. The D9 model:

This model was proposed in 2004 as Draft Interpretation D9 by the IFRS Interpretations Committee. The model suggests measuring benefits with a variable return at the fair value of the underlying reference assets and benefits with a fixed return using the projected unit credit method. A “higher of” option (see above) would be measured at the intrinsic value. While the staff acknowledge that this model would be an improvement to IAS 19 they think it is not possible to set a scope for it.

         ii. Bifurcation model: 

Under this model, a contribution-based promise would be separated into a defined contribution component and a component for any guaranteed return. The guaranteed return would be measured at fair value. The IASB had previously rejected this model as it would mix different measurement bases for one obligation and might provide opportunities for accounting arbitrage. The staff shares those concerns.

          iii. Mirroring model:

Under this model, the fair value of the plan assets would be the present value of the related obligations. The staff concede that it would be difficult to scope the model and that it would be conceptually challenging to not measure some assets at fair value.

(d)   A new practical model:

        i. A ‘capped’ ultimate costs adjustment model:

This model would cap the rate of asset returns used to estimate future benefits at a rate equal to the discount rate specified under IAS 19. While the staff acknowledge that this would conflict with the ‘best estimate’ requirements and would not provide the most useful information for a ‘higher of’ option, they think the model would be relatively simple to develop and give a cost-beneficial short-term solution for some of the problems relating to hybrid plans.

Potential impact of the Agenda Consultation and other IASB projects (Agenda Paper 15C)

The staff expects constituents to provide comments on the pension research during the current Agenda Consultation. They also think that it might be desirable to wait for the current revision of the Conceptual Framework to be completed before carrying out further work in this research project.

Board discussion

There was general agreement amongst Board members that IAS 19 is becoming more and more outdated and its financial reporting requirements are not capturing the characteristics of many of the schemes that have been introduced in recent years.

One Board member asserted that all liabilities that were associated with asset-liability-management should eventually be accounted for at fair value. He said that the fair value model and the customised fulfilment value model would not be significantly different, except for some assumptions. He therefore proposed that the Board pursue one of these models to ensure that assets and liabilities associated with a pension scheme are accounted for under the same measurement model. He said that, otherwise, the information provided would not be the most decision-useful information.

Another Board member proposed to wait until the accounting requirements for insurance contracts would be finalised. She said that the model developed for insurance liabilities might be useful when developing a model for pension liabilities. She also noted that under the bifurcation model the separate components would have to be consistently identified.

One Board member expressed support for the ‘capped’ ultimate costs adjustment model as a short-term solution. He acknowledged that this model was not conceptual but said that developing a single model that was conceptually sound would present a larger challenge and therefore it would take some time to develop.

As regards the interaction with the Agenda Consultation, there were mixed views amongst Board members. While one Board member welcomed the staff’s recommendation to wait for the agenda consultation, a fellow Board member said that the problem had existed for several years and presented a major issue in some jurisdictions. He concluded that the problem should be addressed regardless of the outcome of the Agenda Consultation.

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.