Pensions Accounting: Appeal of a lesser evil
The International Accounting Standards Board has agreed to investigate a last-ditch attempt to address the challenge of hybrid-risk plans. Stephen Bouvier explores the issues
Staff at the International Accounting Standards Board (IASB) say they have devised a solution to the challenges presented by a limited group of hybrid-risk plans. This solution is known as the ‘capped’ ultimate-costs-adjustment model and it would apply to pension promises that vary according to the level of returns on specified assets.
The method works by capping the cash flows that are included in the measurement of the pension liability. In other words, the cash flows would ultimately reflect a return that does not exceed the discount rate applied to the liability. The potential genius in the approach is that it would not require a new plan classification within International Accounting Standard 19 (IAS 19). Instead, any changes could be brought in by amending the paragraphs of the standard that deal with defined benefit (DB) plan assumptions.
There are several caveats. First, staff concede that they have yet to explore the model in detail to establish whether it would reduce the accounting anomaly caused by the inconsistency between the cash flows and the discount rate on hybrid-risk plans.
Similarly, they have yet to assess the cost-benefit trade-off for both the IASB and preparers of developing the approach, exposing it for public comment and, eventually, implementing it. Finally, staff must investigate whether the approach has any unintended consequences.
In a recommendation to the board on 17 May, the staff said: “We believe that our further investigation should focus solely on that approach. If that approach turns out not to be viable, we would recommend to the board doing no further work on post-employment benefits.” Some nine board members supported the proposal to mount the limited-scope research effort.
Several factors have led to this decision. First, the board has recently launched a formal agenda consultation exercise. The feedback to that process recorded moderate support for a pensions project. Certainly, those jurisdictions that could be in line to benefit from the project’s successful conclusion were vocal in their support.
There is also the long-standing view that the numbers produced under IAS 19 are unreliable. Staff told the 17 May board meeting that actuaries have been qualifying their IAS 19 valuation reports on the basis that the resultant values are grossly misleading.
In its response to the agenda consultation, the South African Institute of Chartered Accountants argued that IAS 19 produces bigger deficits than economically exist. In similar vein, the International Organization of Securities Commissions (IOSCO) wrote that it was concerned that “recent developments in employee benefit promises do not fit well within the existing accounting requirements”.
One IASB member, in particular, was sceptical about the projects’s chances of success. Mary Tokar said: “We have tried several times to have a surgical approach on only hybrid plans.… I think we should just cut our losses. As soon as we start talking about the accounting mismatch between [the] obligation and [the] discount rate, we’ll get people who are not in hybrid plans saying ‘Well, we have an accounting mismatch, too’.”
But Martin Lowes, a consultant actuary with Aon Hewitt, says the problem the IASB will investigate is clearly and widely understood: “All the way through, the problem is that DB plans are valued with a discount rate that includes an arbitrary credit risk premium, whereas defined contribution (DC) plans are accounted for with no credit risk. But when you try to look at hybrid plans that are somewhere in between, at some point you are leaping from no credit risk to a corporate bond rate.
“I think this would be an improvement from where we are, accepting that it is not perfection by any means. It would reduce some diversity and it would avoid some clear anomalies, but that doesn’t mean resulting valuations can’t be criticised”
“If you try to deal with the plans in the middle, you are always drawing a line between the two approaches. Every time they tried to do it, they’ve come across this quite fundamental issue that anything they do, unless you go back to first principles, will draw a line between DC-like accounting and DB-like accounting and you can’t get to a sensible dividing line that is any more sensible than the one they have now.
“The research project has really confirmed that they either fundamentally revisit DB pensions accounting, which is, in itself, not a priority, or they won’t be able to resolve all the issues with all hybrid plans. They are also struggling with the issue of fatigue. There have been a lot of changes to accounting in recent years and users and preparers are struggling with having gone through so much change so quickly. DB pension accounting looked at in isolation is fine.”
Eric Steedman, an IAS 19 specialist at Willis Towers Watson, says he also supports the decision: “It would be very easy to pick holes in it and say that it doesn’t solve this or that issue, but it will be almost a case of choosing the lesser evil. The proposed approach mitigates one issue and it leaves other issues. They will have to take a view as to whether, overall, that is better, while knowing that it isn’t perfection.”
He also says the fix could make life easier for preparers: “I don’t think this line of attack will make calculations more complicated. It might even make them simpler in some cases. This comes back to the fact that it is not trying to pick up the potential range of outcomes – it gets complicated where you try to model variability. You could say it’s simplistic for the reason that it doesn’t pick up that variability.
“I think this would be an improvement from where we are, accepting that it is not perfection by any means. It would reduce some diversity and it would avoid some clear anomalies, but that doesn’t mean resulting valuations can’t be criticised.”
One of the jurisdictions that could gain from the IASB’s move is Switzerland. Not only have the Swiss faced substantial challenges in applying the IAS 19 model to their pension plans, those very challenges have led some Swiss companies to ditch International Financial Reporting Standards altogether.
Nonetheless, André Tapernoux, a partner with Mercer’s Swiss practice says he, too, supports the IASB’s latest move, although he notes that it won’t solve every problem in Switzerland. IAS 19 presents Swiss plans with two difficulties in two main areas. First, there is the way the standard presents a high net liability that does not reflect the underlying economics of the plan.
Second, there is the question of complexity, which leads to high advisory costs and uncertainty flowing through from settlements, curtailment or changes in assumptions. On top of that, outside the scope of any work by the IASB, there are the further problems of the back-end loading of age-related interest credit under Swiss plans, as well as conversion rates on retirement.
Tapernoux explains: “These problems won’t be solved by the approach that the IASB has said it will examine, although the proposal might address the concerns with the overstatement of the liability. A typical plan has active members with savings and contributions each year. These build up as in a DC plan, with the difference that there is a minimum level of interest credit. The current legal minimum is 1.25% and the typical discount rate is 0.4%.
“So you project forward at those rates. This means the savings increase by around 10% without a specific reason for that increase. Under the IASB’s approach, you would just project at 0.4% and discount at 0.4%. You will then end up at the amount of savings.”