related documents

top stories

Comments from ACCA
September 2008

ACCA is pleased to have this opportunity to respond to the above Discussion Paper (DP), which was considered by ACCA's Financial Reporting Committee.

Overall comments

ACCA supports a replacement of IAS19 in the longer term, as this current standard includes a number of weaknesses – alternative accounting treatments and a lack of clear principles for accounting for different pension promises and for their measurement. We agree that such a replacement is likely to take some time to develop and that there should also be a shorter term limited revision of the standard, but one more restricted in scope than proposed in the DP. The existing deferral and non-recognition options (including the 'corridor') for actuarial gains and losses, as the most unsatisfactory feature of current IAS19, should be eliminated as proposed. No further changes to defined benefit (DB) scheme accounting can be justified at this stage. Neither should the proposal to create a new category – contribution based promises (CBP) – be pursued at this point because of the complexity and change that this would involve. The proposals include a new boundary for DB accounting which would reclassify many pension schemes (the so called hybrids) as CBP and would require a new form of measurement for them. This may not significantly assist the accounting for hybrids and would represent a much more complex approach for existing defined contribution (DC) schemes.

ACCA's answers to IASB's specific questions

Q1. Are there additional issues that should be addressed?

No, not in a short-term revision of IAS19. Indeed we think the number of revisions proposed in this DP should be reduced.

As noted above, however, in the longer term IAS19 should be replaced entirely by a new standard. While we did not agree with all of the proposals in the recent paper 'The Financial Reporting of Pensions' from the UK Accounting Standards Board and EFRAG, this was a helpful start for IASB to consider the main features of a new standard.

Q2. Are there factors the board has not considered in arriving at its preliminary views on DB schemes

• all changes in the value of plan assets and post-employment benefit obligation (PBO) should be recognised in the period in which they occur
• the return on assets should not be divided between an expected return and an actuarial gain or loss
• unvested past service costs should be recognised in the period of amendment?

We agree that all changes in the value of plan assets and PBO should be recognised in the period in which they occur. ACCA has commissioned research in this area among large European companies. This found that the corridor method was the most popular choice among relevant companies (51%) but by doing so equity was on average overstated by over 3% and the pension liability understated by 41%. A copy of the research report is attached.

We agree that unvested past service costs should be recognised in the period of amendment to complete the full recognition of the PBO.

The presentation of the changes in the plan assets should not be changed at this point, but should be considered as part of the longer term replacement of IAS19.

Q3. Which approach to the presentation of DB costs provides the most useful information to users?

Q4. How could the Board improve the approaches to provide more useful information?

As noted above the existing corridor approach should be eliminated in the short-term revision. No other change should be made for now to the presentation of pension cost. This would leave the existing option to recognise immediately all elements of that cost in profit for the year and the option to recognise actuarial gains and losses in other comprehensive income (OCI) in accordance with paragraph 93A. We are currently extending our existing research project to assess the usefulness of the current IAS19 methods to capital markets, including the Para 93A method and will be happy to share the results with you when available.

In reaching this conclusion we considered the three approaches put forward in the DP and have the following comments on each.

• Approach 1 we agree would be the right approach if there were to be in future a single performance statement then this. However in the absence of a new standard on performance reporting then this should not be the only method, on the grounds of consistency with the current reporting of comparable items. Approach 1 is an option under existing IAS19 which should remain available (though in our survey less than 3% of the companies used it).
• Approach 2 would exclude from profit all interest (discount) and income from plan assets and this would be inconsistent with users' expectations and other accounting standards.
• Approach 3 would not represent in our view a significant improvement on the existing model. Under the Para 93A method the major sources of volatility in pension costs are recognised via the OCI and that meets significant concerns of many preparers and users. To strip out one of these components (the difference between expected and actual return on plan assets) ahead of a fundamental revision of performance reporting will not be an improvement in the usefulness of accounts. Approach 3 would also not be consistent with other accounting standards such as IAS37 on provisions.

Q5 to 13 Contribution based promises

IASB should not pursue these proposals for a new category of pension obligation as part of a short-term revision of IAS19. In any longer term replacement of the standard we would like to see an end to this rules-based distinction between different pension obligations and the adoption of a more principles-based approach.

We are concerned that the definition of CBP is complex and hard to understand. The new boundary in the accounting would seem to be determined not by the incidence of risk as a whole, but based on salary risk alone. Nor are we clear where for example death-in-service benefits (a common feature of many UK schemes) would fall.

Many existing hybrid schemes in our view would be reclassified from the DB to the CBP model. All existing IFRS preparers in Europe and elsewhere who have such schemes will have addressed the DB accounting and so will be asked under these proposals to now do something different which will entail a cost. We are not yet convinced that the CBP model will produce a more useful result to overcome this cost, either in the calculation of the obligation or in its presentation of the whole pension charge as part of profit (see our comments on Q3 and Q4 above). Nor would the accounting model be noticeably easier to apply. The risk-adjusted discount rate proposed, including the recognition of changes in own credit risk, need to be considered as part of an overall reconsideration of pension accounting and not inserted now.

The majority of pension provision in the UK and elsewhere is by DC schemes. Their accounting treatment under IAS19 is currently easy to understand and apply. These proposals would mean DC schemes would be classified as CBP and they would as a result be faced with the much more complex measurement model set out at length in Chapter 7 of the DP.

Q14. Further disclosures to be reconsidered at a later date

Improvements in disclosure could be considered as part of a short-term project. The ASB/EFRAG paper contained some helpful proposals in that regard. We would simply note further at this stage that extra possible disclosures scheme-by-scheme could be very extensive. Some guidance on aggregation and segregation of different schemes could be helpful in avoiding overwhelming details and high compliance costs, including the case of subsidiaries and group pension schemes.

Back to top