We’ve been here before
Is the International Accounting Standards Board (IASB) about to launch a pensions-accounting project? First and foremost, it is early days. The IASB has an active research agenda split between short-to-medium term and long-term projects. Pensions accounting falls into the latter category.
The fact that the IASB is discussing pensions accounting at all should not come as a surprise. When the board published its 2011 revisions to International Accounting Standard 19, Employee Benefits (IAS19), it was clear that there were issues for the board to fix as part of a wider, more fundamental review of employee benefits.
Recent developments – such as they are – represent progress over the past 12 months. Speaking during an IASB meeting on 16 September 2013, IASB’s research director, Alan Teixeira, said the pensions research effort was “on the longer-term research plan” but had no staff allocated to it. He also noted that no standard setters had “expressed an interest in the project”.
The completion of the IASB’s 2011 tinkerings with IAS19 coincided neatly with the board’s first public consultation on the shape of its future work agenda in the summer of 2011. A statement followed in December 2012. This explained that the board had no plans to issue either a discussion paper or a research document on post-employment benefits “within the next three years”.
Comments made at a board meeting on 22 September suggest that the board might be about to breach that commitment and publish some sort of due-process pensions document before the end of 2015.
In their introduction to the board, the staff explained that they “have not yet decided” whether they should publish a discussion paper. They do, however, plan to issue a research paper during 2015. This document will explore:
• A “conceptually sound and robust measurement model” for pension plans; and
• The cost-benefit analysis of any such accounting model given recent trends in plan design.
Staff also explained that if the research paper identifies “enough evidences [sic] to consider a fundamental amendment to [IAS19], we may propose to publish a discussion paper.”
In the board’s sights are pension promises that fall outside IAS19’s two plan classifications of defined benefit (DB) accounting and defined contribution (DC) accounting. In the case of the former, a plan sponsor projects forward using the relevant plan assumptions and discounts back using a AA-corporate bond discount rate.
Although accounting for DC plans is a relatively simply exercise in expensing pension cost in the current period, the standard has felt the strain in recent years when it comes to the challenges thrown up by so-called intermediate-risk plans or contribution-based promises.
At a glance
• The IASB plans to issue a research paper on IAS19 some time next year.
• This will look at a “conceptually sound and robust measurement model” for pension plans, and take into account trends in plan design.
• A full-blown discussion paper could follow.
• Previous attempts to revisit IAS19 have stalled over the issue of contribution-based promises.
• Work on discount rates for non-traditional DB plans is also at an early stage.
Many pension plan sponsors have attempted to de-risk their pensions exposure by implementing plan designs that transfer risk, at least in part, to their employees. A plan might, for example, offer a minimum investment guarantee or even the higher of a specified rate of return or the outperformance of an equity index.
The IASB and its interpretive committee have made several attempts over the past decade to address pensions. The International Financial Reporting Interpretations Committee published the so-called IFRIC D9 approach for contribution-based promises in 2004.
The committee called a halt to that work when, in 2006, the IASB launched its unsuccessful bid to develop a new accounting methodology for contribution-based promises. That effort resulted in the board’s ill-fated 2008 discussion paper, which proposed a new fair-value measurement methodology.
IFRIC’s successor, the IFRS Interpretations Committee, decided to take another look at contribution-based promises in 2013. That effort stalled earlier this year. Both D9 and the IASB’s 2008 discussion paper try to address the problem of applying the IAS19 projected unit-credit model and discounting approach to non-traditional DB plans.
What of the board’s work on discount rates? Again, the research is at an early stage and forms part of a limited-scope project. In a June 2014 board paper, staff summed up the effort as a “review of existing requirements” and the board describes the effort as sitting in the short-to-medium term.
In a meeting paper presented to the IASB, staff wrote: “We propose that research on discount rates should focus on reviewing measurements that already require use of present value technique and for which the objective of measurement is not fair value.”
Staff described the review as a “desktop study of requirements and related literature, supplemented by interviews with a small selection of stakeholders.” Despite the limited scope of the project, discounting was a topic of some interest to the respondents on the board’s 2011 agenda consultation. Some respondents urged the board to launch a comprehensive project to “define and describe how to determine discount rates”.
But what might a future pensions project look like? On the evidence of the 22 September meeting, the IASB appears to be heading towards a project that will explore the accounting for all categories of pension promise.
IASB member Stephen Cooper said: “I think that’s the only way to go. I think if you try and do contribution-based promises, as we’ve discovered in the past, it just becomes impossible and you get these bright lines. The only way to solve this is to look at the whole range of things.”
He added: “The first question I would like to answer with pensions is: does the insurance valuation model work for pensions. Because the similarity is there and surely that has to be the first question, what differentiates insurance from pension to warrant a different valuation model.”
So, does this imply that a future pensions-accounting model under IFRS will include a discount rate for both defined benefit promises and defined contribution promises? Not necessarily says Diana Scott, former IASB employee benefits working group member and Towers Watson consultant. “Measurement issues, including the discount rate, go to the heart of what measure is most relevant and best serves the user of the financial statements. On one hand it would seem that the same benefit promise by two companies to identical populations should result in the same measure of the benefit obligation,” Scott says.
“On the other hand, one could argue that consideration should be given to credit or non-performance risk on obligations that haven’t yet been settled,” she argues, adding that there are “many other” considerations, among them liquidity risk, that also affect measurement.
As for any overlap with insurance, Scott says: “I appreciate that one could analogise the measurement issues around retirement benefits to an insurer’s measurement of its insurance liabilities, but I tend to think of the retirement benefit obligation more as a form of debt than insurance.”
And Aon Hewitt consultant Simon Robinson questions whether the research effort will actually deliver the goods in the shape of a final pensions standard. “It does feel as if we have been here before, with the UK Accounting Standards Board’s 2008 discussion paper and the IFRS IC’s work on IFRIC D9 in 2004.”
He continues: “I support the IASB pushing ahead with a research paper, but I think it will be hard to find a solution within the current structure of IAS19. I suspect the research paper will provide some interesting discussion, but ultimately the measurement of hybrid plans will be put back on to the too-difficult pile.”