Doing mortality to death
To invest time following the progress of International Accounting Standards Board's (IASB) pensions project is generally to waste part of an otherwise productive existence. But recent board deliberations on the issue of mortality disclosure have rendered the process somewhat more worthwhile, with IASB member Jim Leisenring questioning the "financial geniuses" to draw worthwhile conclusions from the voluminous information that some of them now demand.
Mortality is perhaps the ‘black box' of pensions accounting, probably more so than discount rates, the focus of much recent board attention.
The likely response from constituents if - and it is a big if - the IASB ever publishes an exposure draft on its Phase I pensions account ing project, will see the user community demanding more disclosure around mortality, consultants arguing for disclosure tempered with restraint, and preparers broadly opposing any additional burden.
Summing up the outcome of a 19 May board discussion of the issue, IASB member Warren McGregor suggested the board should "ask a question in the exposure draft (ED) as to whether or not people feel there is a need for more detailed information about mortality assumptions."
Earlier, bringing mortality to the table, staff explained that "to address concerns raised, we have proposed a more principles-based approach and also guidance on materiality and disaggregation similar to that already found in IFRS7 ED10 and IFRS1."
Whatever form any mortality disclosure might take, it looks certain to lack an explicit materiality requirement, a notion that Jim Leisenring had some difficulty applying to the concept of disclosure. "I'm not sure that it's as applicable or so easily applied to disclosure as it is to recognition and measurement... It's easy to see that X amount is kind of material to the financial statements taken as a whole, whether some of the disclosures are material or not, I'm not even sure how I'd judge that."
As IPE highlighted in March, the debate around the volume of pension plan disclosures rages between the endpoints of disclosures that would fit on a single sheet of A4 paper and just about anything that anyone could ever imagine about pensions and then some.
"I sympathise a little with the side of the staff camp that thinks that if anything we have overkill on disclosure," said Leisenring. "And I think the overkill is a carry-over from the black box we invented with amortisation, corridors and all of the other garbage."
Indeed, under the US GAAP, with the introduction of FAS87, smoothing and deferral mechanisms, the quid pro quo was that preparers would provide users of accounts with sufficient information in the footnotes to unpick the effects of deferred recognition.
Leisenring continued: "I don't understand why we have enormously more detail about this obligation and the funding aspect of this obligation than we do about anything else, particularly now if we are going to make the accounting for it transparent. I think disclosure compensated for atrocious accounting at both the FASB level and here, but I think we ought to be looking for a dramatic reduction in some of them."
The thrust of the changes to IAS19's existing disclosure objective is largely driven by demands from some users for additional information about an investment target's exposure to pension plan risk factors.
As explained at the board meeting: "It is to do with the problems they see in measuring the obligation, partly, and also they believe, rightly or wrongly, that these are special in their own way and it represents a long-term nature liability similar to insurance, and also that in countries it is seen as a social good and should be more transparent in some way."
As for the specifics of any mortality disclosure, it rapidly becomes apparent from the official IASB recording of the 19 May meeting that the board has made precious little progress in enhancing its understanding of the issue.
"Looking at the data, every year there's a four per cent swing in the average liability," noted IASB chairman David Tweedie. "That's quite big. And when you look at the US statistics, for example, they estimate a 65-year-old man is going to live for about 18 years, I'm going to live for 19, [his wife] Jan's going to live for another four... while in France, because it's an egalitarian society, everybody lives 23 years."
There was, Tweedie said, "something slightly odd about all this, and so I just wonder why we are not going to try and do something. Now, if you had a mining industry and they said everyone's going to live for 40 years after they've retired at 65, that's just nonsense. It's probably the most important piece of information. It may be difficult but I think they should do it."
To a great extent, Tweedie and his board members are prisoners of the information presented to them by the staff team. For example, could differences in French mortality be down to any particular bias in the sampling used to compile the French tables? Are French pension plan sponsors simply relying on the best available evidence, rather than obscuring the true picture from investors?
And faced with shortcomings in actuarial tables, might not entities find in their own workforce an insufficiently wide statistical basis for making meaningful adjustments to that table?
The debate could have been better informed had the staff presented evidence along the lines proposed in the now infamous 2008 discussion paper. "The board intends to review the disclosures required about post-employment benefit promises in a later stage of this project. As part of that review, the Board intends to consider best practice disclosures in various jurisdictions."
That research might have led the staff to Germany, for example, where as a rule, German multi-nationals use the same actuarial tables for their pensions accounting as they use for tax purposes, the so-called Heubeck tables.
Had the staff - or any analyst for that matter - looked at the Heubeck AG website, they might have learned that "the new actuarial tables naturally also take into account... the clear tendency towards increases in life expectancy".
As for the possible impact of the updated 2005G actuarial tables, Heubeck AG notes: "In the event that you would like an estimate of the likely impact on your entity, we would be happy to be of service."
The suspicion is that, working with the best evidence available to them, German pension plan sponsors and Heubeck AG might between them have given more thought to the concerns voiced by Tweedie than he realises. They might not, of course, but until the IASB asks them we will never know.
Although the staff claims to have consulted members of the board's employee benefits working group, that approach is at least as vulnerable to sampling bias as any mortality table. And it did not appear to have left staff able to address Tweedie's questions promptly.
Curiously absent from the debate is any effort by the analysts demanding more information to explain precisely what it is that they would do with that information. Also, the IASB does not appear to have examined the cost-benefit of providing ever-increasing amounts of disclosure. The elephant trap awaiting the board is perhaps one of a disclosure requirement of such complexity that users of financial statements are no better informed.
One sceptic is Leisenring, whose comments from the follow-up 21 July board discussion on disclosure provide a blueprint for the critical analysis of the analysts: "I just have this nagging feeling that we are sometimes confusing whether we are reporting the plan or the beneficial interest in the plan, and whether the information that I would like to see if I were analysing a plan is, or is not, the same as that I would want to know about the employer that sponsors the plan. And I don't know the answer to that question."
And then, addressing his comments to one of the three former analysts who now sit on the board, he continued: "So there's just a variety of things [about sensitivity analyses] that seem to me to be fuzzy as to whether they are as relevant as they appear to be. And frankly, I'm damned cynical about some of this actuarial information that can in fact tell me that 100 basis points swings in interest rates does the following to the stock market.
"I think those are some of the same geniuses that have been risk managing banks and insurance companies recently, and I'm not necessarily confident that I'd be providing relevant information and not being deceptive. I'm just going to shut up, because I didn't agree with the sensitivity stuff in [IFRS] 7 because I don't think some of it can be done. I think we may not be getting what we think we're getting."