Accounting: Pension plan

The costs of company pension schemes have to be recorded in financial
statements. Most finance directors would agree with that statement, but after
that the consensus over accounting for pensions starts to crumble.

Despite all the focus on the pension problem, one subject that has received
remarkably little attention over the past couple of years is how we actually
account for pensions. Since FRS17 burst onto the scene – persuading us all that
we were in a crisis – corporates have been so focused on handling the deficit
that there has been little innovative thinking over the actual accounting. FRS17
and its imitators may be the best accounting system that we have for pension
accounting, but that is not saying a lot. And while FRS17 has become the
standard in the UK, it is also being seen as the benchmark for both
international accounting standards and even the US Financial Accounting
Standards Board seems to be determined to fall in line.

FASB has decided (after some consultation, naturally) that, by the end of
2006, US companies will be forced to bring pension schemes and other
post-retirement benefits deficits onto their balance sheets. We’re talking big
numbers: according to one US analyst, General Motors’ liabilities for current
and future health benefits of existing and retired workers totals $60bn.

But even with the benefit of FRS17 and its international counterpart IAS19,
non-US companies should not be too complacent. Both in terms of measurement and
disclosure, pension accounting is still shaky. The Accounting Standards Board –
which has been adopting a more critical tone in recent months over the direction
of international accounting – sees the need for improvement. It has just
released a new financial reporting exposure draft (FRED) and a draft reporting
statement, Retirement Benefits – Disclosures, which complements the disclosures
in FRS17. The draft reporting statement sets out six principles that FDs should
consider when providing disclosures for defined benefit schemes by providing
information that will give users of financial statements information to help
assist them to evaluate the risks and rewards that an entity is exposed to over
its defined benefit schemes. The draft reporting statement is designed as a
formulation of best practice. It is intended to have persuasive rather than
mandatory force and leaves entities (in the ASB’s words) “with the flexibility
to make disclosures that are appropriate to the risks and rewards that the
entity is exposed to in relation to defined benefit schemes”.

One example of where improved disclosure is needed is in mortality. One of
the key factors in determining the pension liability of a corporate is when its
pensioners and any pensionable dependents are likely to die. At present, only
half the FTSE-100 disclose their mortality assumptions. Of those that do
voluntarily disclose there is a 22% variation among the FTSE-100. Most assumed
life expectancies for a 65-year-old ranged from 18 to 22 years. That is not
necessarily unreasonable – a given set of workers will shake off this mortal
coil at different stages. The lower the life expectancy assumption used, the
lower the value of a scheme’s liabilities will be. Revealing the stringency of
the mortality assumptions used by companies for their employees and pensioners
allows an evaluation of the prudence of the pension deficit calculation.

While more comprehensive disclosure is welcome, it still leaves the far
trickier issue of measurement. How do you measure liabilities? Standard setters
have been toying with the idea to procure a valuation for the present value of a
final salary pension, but it is impossible to look to the market to provide a
future value for salaries that may rise. At present, valuations are based on
estimating the final salary. But standard setters are suggesting one significant
move to use today’s salaries rather than estimated final salaries.

If measuring liabilities is hard, then so are rates – interest rates, the
return on assets and the discount rate that should be used for working out the
liability value of pension plans. When FRS17 was under construction, the ASB was
told to base the discount rate on the return on assets. But it rejected that
argument on the grounds that a pension fund invested in junk bonds doesn’t just
slash its liability. Equity yields were next up for consideration, but were
turned down on the grounds that there is no correlation between salary increases
and share prices. There is a growing feeling that we should use the risk-free
discount rate, an idea that has been around for 30 years. That seems attractive
as it would have the advantage of breaking the current obsession with AA bonds.

It appears that the IASB is more in favour of adopting short-term solutions,
rather than going for a complete revamp. Within the context of the convergence
programme with US standards that is understandable, especially as the US is
trying hard to play catch up. Perhaps pension accounting is one example ASB
chairman Ian Mackintosh had in mind when he said convergence was good, but not
convergence at any price. Accounting for pensions could and should be improved.