IASB mulls over average salary

"We can't pretend that we're not doing something that we're doing," Mary Barth told her fellow IASB members on 19 September. "I think we have to be honest with the world about what we're doing as opposed to pretending that we are not changing things when we are."

The bad news - there isn't any good news - is that the IASB is mulling the idea of shifting the classification of average-salary defined-benefit pension plans from defined-benefit and project unit credit accounting to the defined-return category, and fair value measurement.

"A small subset of defined-benefit promises," explained project manager Jenny Lee, "such as cash-balance plans and plans that guarantee fixed or variable returns ... [are]
troublesome using the projected unit credit methodology."

Lee continued that the IASB has two choices: "We can can either carve out those plans and call them defined-return [DR] promises, or we can make the cut slightly differently and have all benefit promises with a promised return as defined return and everything else as defined benefit."

"We had wanted to clarify," she continued, "that guaranteed fixed returns were included in DR promises because of a comment that was made at the last board meeting; for instance, if you just have a lump sum plan is that DR or not? And it is difficult to say variable returns because strictly speaking salary increases are variable returns, but I do think that there is some work that we can do on that."

Also under this rationale, defined-contribution plans are now defined-return plans.

Average-salary pension plans have a champion in the shape of newly appointed board member Stephen Cooper.

But his failure to sway the argument might leave little hope for anyone hoping to persuade the board that defined-contribution plans with a small performance guarantee should escape defined-return classification.

"I think this will capture many plans," he began. "There are some very large UK plans … that will move from DB into DR. And given the fair value measurement, I think that is going to present problems in trying to persuade people of the merits of this." You can say that again.

So he did, in even more detail: "My understanding is that the reassessment ... is a reassessment based on either inflation or inflation up to a limit ... I don't believe that those
have the indexing to some equity index."

What he takes to be a troublesome plan, he explained, is "where you compound at the equity rate and discount at the bond rate and you clearly get a stupid answer".

But it made no difference, not even when he pointed out an internal inconsistency in the reasoning of those promoting change: "My concern would be that the definition of DR is actually inconsistent with something in para 13 in that the objective of Phase I should be limited to the troublesome plans."

The same might be said of what IASB member Warren McGregor called "plain vanilla defined-contribution"; defer the payment of the contribution to any significant degree, and you'll soon discover that "we have said that you are going to account for that on a fair value, well it's not fair value, it's a different basis," explained McGregor.

Fair value, it seems, is the new fundamentalism: "I think if you had a straightforward average-salary plan that was just 5%of the career average of your salary, I think you could argue that projected unit credit works as well for that as it does for final salary. But I think in practice most plans are not just flat career-average salary, they are career-average salary with some sort of revaluation, and it is the revaluation part that makes it defined return, and that makes it difficult to value," the staff team explained.

Staff continued that if they were "to carve out career-average plans from the troublesome plans, ... it would be difficult to think what the criteria for that would be. It would be those that could be expressed as having future salary increases plus a defined return are not part of the thing that we are carving out. And it is conceptually more robust to carve out anything that has a defined return in it."

One rationale the IASB might find were the heavy hints they dropped at the start of their project that they would leave untouched the twin endpoints of defined-benefit and defined-contribution accounting.

Measured against the marker of the Phase I project's initial remit, traditional measurement methodologies have taken quite a battering. Career-average salary plans could receive fair value accounting on the basis of their link to an external measure such as an inflation index. The rationale appears to be that whereas a salary increase is within the control of the sponsoring entity, inflation, like an investment index, is not.

Ahead of publishing a formal due-process discussion paper next year, IASB staff reminded IPE that the board's official positions are stated in its standards and interpretations. Which leaves it open to IPE's readers to either act on or disregard entirely the content of the these board's preparatory discussions on the content of its now delayed paper.

Tim Reay, with consultants Hewitt, wonders whether the upheaval is by design: "The career-average plan seems to have been unintentionally caught up in this definition. It just goes to show how difficult it is to develop working definitions without running into unintended consequences. Part of the challenge is that there is such a huge variety of plan designs out there that have various overlapping features."

The 19 September meeting was also staff's first opportunity publicly to wrap up the board in the consequences of its own decision making: "We're in a bit of a quandry here," IASB research chief Wayne Upton explained, "because the staff came to you and said let's be very targeted, let's make sure we don't disturb final-salary plans, let's make sure they turn up as defined-benefit, and the board very unanimously said ‘no, we don't want to do that'. So now we've got what you told us to do."


At some point during this discussion, Tatsumi Yamada, another board member, made the mistake of speaking out of the agenda running order. He wanted to know what the measurement implication was of classifying a defined-benefit plan as defined return: "How do you calculate it? I have no idea how to calculate it? Even though there is [an] example in [Agenda Paper] 6B."

Many will see this as a sensible starting point. After all, why sign up for a fundamental reworking of pensions accounting if you are clueless when it comes to the hit on the bottom line. Jenny Lee began launched into explanation; someone stopped her, leaving Yamada, presumably, in the dark.

His colleague Philippe Danjou, a former securities regulator, commented that he was "not ready to accept something if I don't know what the consequence of it is".

A chink of light was cast on the "controversial" issue of fair-value measurement of defined-return promises.

Senior project manager Anne McGeachin explained that the current understanding of fair value restricted her team's room for manoeuvre - particularly on performance and credit risk:

"We don't want in this paper to get into a discussion on how performance risk ... is or is not in fair value. That is for the fair-value measurement project. We just wanted to try, by describing the unit of account as the benefits as they are stated at the balance sheet date, we wanted to be clear that in this phase of this project that is what we are fair valuing."

Her comment was a response to a charge from Warren McGregor that "what you are measuring is not fair value". He reasoned that although there were elements of fair value in the calculation, what staff propose is not fair value.

Gilbert Gerrard said: "We have to refrain from calling it fair value"; Tatsumi Yamada pointed out that to arrive at a sound answer, account must be taken of actuarial gains and losses, although "if we take take into account the actuarial aspect [of] the calculation, I think it is not called fair value."

But, countered John Smith, fair value in this case "is the difference between the plan versus the cashflows". Perhaps it was a coincidence, but suddenly the US GAAP specialists came into their own. "If this is the unit of account that you want to fair value, there is nothing wrong in calling this fair value," reasoned Wayne Upton.

"The document that we have out on fair value for exposure, and accounting generally, does a lousy job of unit of account issues and general and liabilities in particular...If you define this as the unit of account that you want to fair value, then there is nothing wrong with calling what you are doing ‘fair valuing it'."

As IASB groupies would expect, Upton's intervention - whether you agree with it or not is another matter - cuts right to the issue. None of this tightly reasoned argument figures in the September IASB Update, despite the fact that it identifies the issue most likely to offend IASB's constituents.

He continued: "Most of the argument I've heard is the people who say that it's not the fair value of something else. No, it's not. The difficulty is ... the document [on fair value] goes into considerable detail on trying to identify a restriction on an asset and a restriction on a holder. It doesn't carry any of that discussion into liabilities...None of us really understand performance risk. We are going to ask the board to have a very serious look at all elements of performance risk, including credit risk, as to whether or not we think they are part of the fair value"

And here, courtesy of Upton, is the warning for anyone who thinks that taking a close interest in pensions accounting is about nothing more than watching a single project: "But that's for down the road and that's for the other project."

As intrusion into private grief, this one at least hints where the body might be buried.

Have your say

You must sign in to make a comment