Accounting Matters: Drawing a line in the sand
Standard setters’ focus on risk-sharing pensions recalls previous abandoned attempts to reconcile non-DB and DC accounting approaches
Just when you thought that pensions accounting couldn’t get any more complicated than it already is, up pops the Canadian accounting standard setter with a call for the International Accounting Standards Board (IASB) to tackle the accounting for so-called hybrid pension plans.
Speaking during the July meeting round of the IASB’s Accounting Standards Advisory Forum, the chair of the Canadian Accounting Standards Board (AcSB), Linda Mezon, urged the IASB to consider her staff’s research into pensions and “either add it as another dimension” to the IASB’s feasibility study into pension benefits that depend on asset returns, or to take it on as a separate project.
But while their intervention at this particular moment certainly complicates any work that the IASB might eventually do on pensions, the Canadians do have a point. Hybrid pension plans have proliferated in many countries over the past two decades as sponsors have ditched traditional defined benefit (DB) provision in favour of defined contribution (DC) promises with features such as an investment-return guarantee or some other derivative-type promise.
Inevitably, these types of pension arrangements can give rise to accounting challenges under International Financial Reporting Standards (IFRS). The main problem is that International Accounting Standard 19 (IAS 19) has a binary ‘DB or DC’ approach to both the classification and measurement issues of pension schemes.
That it falls down when it comes to more modern hybrid pension plan design, was apparent with the board’s efforts between 2006 and 2008 to develop an accounting model to address so-called contribution-based promises. This work led to the so-called contribution-based promises approach that forced sponsors to measure any sort of pension promise other than a pure DC or pure DB plan using a fair-value model.
Put simply, the IASB’s understandable pursuit of theoretical consistency meant that a large number of pension promises slipped into the new accounting definition and, accordingly, saddled sponsors with disagreeably large accounting liabilities.
Fortunately for German sponsors such as BASF, which stared aghast at the prospect of the new accounting regime, the IASB binned its proposals and instead diverted its attentions to the wider – and ultimately fruitless – task of converging its accounting rules with the United States.
But although pension plan sponsors around the world might have breathed a collective sign of relief as the IASB made a series of more limited amendments to IAS 19 – the oft-maligned corridor disappeared with barely a whimper – what we were left with was an accounting standard that was pretty much a prisoner of time.
A 2008 discussion paper was book-ended by the IFRIC D9 approach, which looked at pension promises based on a return on contributions or notional contributions, as well as a bid by the German national standard setter in 2012 to obtain guidance on pension promises with a return guarantee and a ‘higher-of’ component. Both attempts fell by the wayside. Indeed, the lesson from history is that any shake-up of IAS 19 produces winners and losers and the consequent accusation that it is ‘unfair’.
If anything, however, the 2012 German issue was a reminder – as if the IASB needed one – that the fundamental mismatch between IAS 19’s twin DB and DC pillars and the intricacies of modern pension plan design had gone nowhere.
So, when the board conducted its agenda consultations in 2011 and 2015, it will have come as no surprise that respondents flagged up pensions accounting as an issue that needed attention. In fact, in 2014, the IFRS Interpretations Committee launched a project that, to this date, rumbles on to clarify when DB sponsors can recognise a plan surplus in circumstances where a third party, such as a trustee body, can determine how that surplus might be distributed.
But nothing is ever simple when it comes to pensions accounting, and in June the IASB agreed with a staff proposal to align work on the IFRIC 14 project with its heavily trailed research work on pension benefits that depend on an asset return.
So, the challenge for the IASB now is to align its past experience with the outcome of the IFRIC 14 project, while not swamping the possible scope of its potential work on asset-linked pension promises. If nothing else, the task ahead illustrates the importance of history.
The yardstick, says Simon Robinson, a consultant actuary with Aon Hewitt, is replicability. “In other words, can you invest to guarantee that benefit?” he explains. “With pure DC, it is obvious how to invest to guarantee the benefit – you simply pay the contributions. But with most DB plans, it would also be possible to guarantee the benefit by handing money to a third party now by paying a contribution now. So it is really akin to a buy-in or buyout of a benefit.
“But if you construct a DB plan, you could go out and purchase an asset that matches that, so why isn’t your cost also the cost of the contribution for the matching asset? So when you have plans that have both DB and DC elements, you end up with contradictions. And this explains why the question is being put to the IASB and why they are thinking about it.”
“When you have plans that have both DB and DC elements, you end up with contradictions. And this explains why the question is being put to the IASB and why they are thinking about it”
Robinson continues: “At the moment, we have DB and DC and draw a line. If there are any guarantees of any sort, we draw the line there. But contribution-based promises moved that line somewhere else. It put the line between using a corporate bond rate or a risk-free rate around whether there was a link to salary or not. They split pensions into two camps one using a risk-free rate and one using a corporate bond rate but with a different cut.”
To be blunt about it: there is no theoretically certain place to draw the line between DB and DC plans. And that remains as true today as it was over a decade ago.
Lane Clark & Peacock partner Tim Marklew is downbeat in his assessment of the board’s prospects for success in the absence of a major overhaul of IAS 19. “Overall, we’ve seen this type of thing before. I think this is a good summary but there is nothing very new there.
“The difficulty coming up with a one-size-fits-all solution is that in order to really fit everything, it would need to address both current DC accounting and current DB accounting. Under the present rules, accounting for the two is significantly different. So bringing the two together would require the IASB to revisit the whole of IAS 19.
“That was broadly the point the IASB got to with contribution-based promises in 2009-10. Where they landed was to say that, in order to get a solution that works, they would need to review the scope of IAS 19 – which they didn’t have the appetite to do at the time. The only other viable option is piecemeal extra guidance, albeit with inconsistencies between them, which is perhaps where this paper will lead.”
The options facing the IASB and its staff are stark. Do they: catalogue a gargantuan array of plan designs and develop an array of workable accounting outcomes for each; target their efforts on certain categories of plan and potentially reduce comparability; or admit defeat now?