All new IAS19: the verdict
KPMG’s London-based corporate advisory pensions team has had a busy start to 2014. As if the traditional year-end rush were not enough to cope with, Naz Peralta, a director at KPMG, has just signed off on the 2014 Pensions Accounting Survey, a round-up of the year’s accounting trends. Under the microscope are corporates reporting under International Financial Reporting Standards (IFRS) or UK/US generally accepted accounting principles (GAAP) at 31 December 2013.
The survey looks at the experiences of 295 KPMG clients with UK defined benefit pension obligations. It is one of the first snapshots of the impact of the International Accounting Standards Board’s revisions to International Accounting Standard 19, Employee Benefits. The IASB published the revisions in mid-2011.
The changes to IAS19 apply to annual reporting periods beginning on or after 1 January 2013. Most corporates, other than those with a June or September year-end, have now adopted the new regime. In summary, the 2011 revisions brought in the net-interest approach on the face of the income statement, as well as a raft of new disclosure requirements. These, the board said, would “highlight risks arising from [a] defined benefit plan”.
The new disclosure regime adds up to a major shake-up of the pensions footnote: out goes the boilerplate, in comes a series of disclosure objectives that are intended to provoke an intelligent conversation with the marketplace about risk. For example, paragraph 145 of IAS19 now requires entities to disclose: “A sensitivity analysis for each significant actuarial assumption at the end of the reporting period, showing how the [defined benefit obligation] would have been affected by changes in the relevant actuarial assumption that were reasonably possible at that date.”
As for the impact of these changes, the KPMG survey found that:
• 61% of sponsors used a discount rate within 15 basis points of the 2012 median, with the figure climbing to 81% this year;
• 75% of companies used an inflation assumption within 15bps of the median in 2013 rather than 65% in 2012, and;
• 80% of companies used a longevity assumption within a three-year range, marking a 4% increase.
And so, as Peralta explains, “a lot of the information we are seeing is not so much about pension cost – pension schemes are increasingly a legacy issue – but, rather, about tackling pension risk.” The picture, he argues, is broadly one of risk, rather than cost-control.
Generally, he continues, the new disclosures have, indeed, pushed companies into a dialogue about the risks that they face and their de-risking strategy. The context for this discourse is also important – particularly where risks are peaking and levelling off and plan sponsors are looking to eliminate as much risk as possible.
Peralta says: “Most clients will have a time horizon by which they would like to exit from their pension liabilities and therefore take pensions off the balance sheet. This horizon will differ from one company to another. For some it might be within five years but, for others, a deadline as tight as that might not be realistic. Or we could be looking at 20 or 30 years. Irrespective of the timetable, there are strategies that companies can implement to either reshape the nature of the risk, or take it out altogether.
“We have several large FTSE clients going through similar exercises at the moment and, as we will see in the next 12 months or so, the financial reporting framework will give them the opportunity to make positive statements to the markets. Some of these exercises reduce liabilities, if the terms offered are less than cost neutral, while some remove liabilities – for example, those involving transfer options. Most will take out inflation or longevity risk. It all helps towards the ultimate end game.”
Due diligence by trustees
The AstraZeneca 2013 annual report is the one of the early signals that the 2011 revisions to the IAS19 disclosure objective have had an impact. The company’s description of its largely UK-based scheme is broadly unchanged on its 2012 disclosure exercise; the real changes comes with its discussion of the risks that it faces through its DB exposure.
On the face of it, AstraZeneca’s tabular approach to risk disclosure and analysis draws on the company’s disclosure strategy for other types of risk in earlier financial reports. For example, the company adopts a similar approach in its 2012 annual report to its disclosure of product pipeline risks. In that case, AstraZeneca identified risks such as regulatory approval and product development roadblocks and analysed their likely impact.
In relation to its DB exposure, the company has compiled a table identifying the four main risks that its schemes face: volatile asset returns; changes in bond yields; inflation risk; and longevity. It then describes each risk and the steps that it has taken to mitigate that exposure.
So, for longevity, the company notes that a rise in life expectancy could – unsurprisingly – increase its DB liability. It explains that the UK scheme entered into a longevity swap in 2013 to cover $3.8bn of liability in respect of 10,000 current pensioners. It also reveals that a one-year increase in life expectancy will lead to a $178m in plan assets.
• The KPMG survey can be found at http://bit.ly/1kFNyR9
But what about those key pensions assumptions such as inflation and longevity? “In our view this has been a fairly low-key year in terms of major trends in pensions assumptions, certainly compared with the peaks and troughs during the financial crisis,” says Peralta.
“There has been a big uptick in long-term inflation assumptions – a 40-basis-point jump between December 2012 and December 2013 – but that was mainly because the ONS decided not to change the RPI index calculation after the market had priced in a change, and a fall in long-term RPI.” Positive asset returns during 2013 have helped to keep liabilities in check.
Peralta also highlights a few subtleties around inflation. He says the use of inflation risk premia has stabilised at around 20bps, with a range of up to 30bps. He also notes a longer-term trend that puts consumer price inflation (CPI) at around retail price inflation (RPI) minus one percentage point as almost the norm. Some 49% of respondents have used a one-percentage-point difference between RPI and CPI. In previous years, there was more of a spread. “There have generally been fewer issues than in previous years about aggressive assumptions,” as Peralta puts it.
Unsurprisingly, mortality assumptions continue to vary from country to country. “In Germany, for example, there is almost universal use of a single table,” Peralta comments. “Similarly, in the US, the introduction of tables which allow for generational improvements could change this. In the UK, meanwhile, although there is an established pattern of picking up the actuarial profession’s latest CMI model projections, there is still an almost infinite combination of base tables and improvement factors in use to reflect scheme-specific mortality.
“The median assumed long-term rate of improvement has nudged up from 1% to 1.25%, which broadly means 1.25% fewer deaths at the same age in each future year, which makes quite a difference to assumed life expectancy and, therefore, liability,” says Peralta. “We are anticipating only small tweaks over the next few years.” Against this backdrop, he adds, the market has recorded an increased interest in longevity swaps, typically trading at 3-5% of the liability.
So if 2013 is the year that was, where is pensions accounting heading? “More companies will be making bold statements about eliminating risk through one route or another. Assumptions will be of less interest to analysts as they shift their focus to risk-reduction strategies,” Peralta believes. “This year end was the first opportunity for companies to report under the new requirements, I expect that disclosures will develop as they pick up on examples of best practice and look to understand how their peer group is approaching certain issues.”