Why have one way of calculating a pension deficit when you can have three?
Each of the methods stems from a completely different perspective. The deficit
is in the eye of the beholder: there is no “right” view as to how big it really
is. Where you stand determines what you see.
The three methods are:
• The funding valuation or actuarial calculation of the fund’s value, which must
be carried out every three years, with the actuaries being commissioned by the
• The accounting calculation, as set out by FRS17 and IAS19, and which lies in
the hands of the finance director of the company sponsoring the scheme (and of
course the company’s auditors); and
• The full buyout or section 75 calculation (the two are different beasts, but
amount to the same thing, as we will see). This last is in the hands of the
pension buyout providers.
Already we see three completely different communities of people: independent
actuaries and pension trustees; finance directors and auditors; and insurance or
capital markets companies. Each of the three communities has a subtly different
starting point and end goal, so the fact that the three calculations throw up
wildly different figures for the fund value (be it surplus or deficit) should be
no surprise to anyone.
“Scheme valuations look at the cash flowing into the scheme from member
contributions throughout the life of the scheme and from asset valuations, and
they look at outflows in terms payments to retired members to try to place a
single value on all those future cash flows,” explains Donald Fleming, head of
pensions at KPMG.
But, he adds, “if you are trying to compare the three calculations on the
basis of which is more subjective, the funding valuation has the greatest leeway
for subjectivity on key assumptions such as the scheme discount rate, average
longevity of the membership and the inflation rate over the life of the scheme.”
Not a good start especially given that it is this calculation that determines
the amount of cash to be paid by the employer into the scheme.
While all the assumptions made in carrying out the actuarial calculation of
fund value will have an impact, the assumption that dwarfs all the rest, in
terms of its ability to push and pull the final figure about, is the discount
The yield on long-term gilts is the usual starting point, though the yield on
high-quality corporate bonds has also been used. Sarah Farrant, a consulting
director and actuary with Deloitte Total Reward and Benefits, says the discount
rate used to calculate the liabilities may be adjusted according to the expected
returns from the investment strategy adopted for the assets. “In this way, they
can allow for expected higher returns that are hoped to be achieved on the more
risky assets held by the scheme (such as equities or property).”
She adds that the shape of the yield curve has an impact, too, depending on
the maturity of the pension scheme. A scheme whose members are mostly pensioners
or near to retirement age will use a shorter-dated gilt yield than a scheme
whose members are mostly young. At present, that would mean using a lower
discount rate in the former example than in the latter.
Further adjustments to the discount rate might be made depending on the
trustees’ views of the financial position of the sponsoring company itself ie,
the strength of the employer’s covenant and so its ability to make ongoing
contributions into the scheme.
Add on to this your cash flow model that determines your view of the scheme
outgoings. This contains at least three very subjective judgements: (i) How much
will employees’ salaries increase up to the point where the last member
retires? (ii) How much will prices increase by the time the last member dies?
(iii) How long will members live once they’ve collected the gold watch? (Think
about this: how many people in 1929 would have accurately predicted today’s
prices and salaries? The point here is that 80 years is about the lifespan
remaining to the youngest member of most schemes, given continuous mortality
creep: life expectancy is already increasing by 12 minutes an hour.)
There are other, more refined assumptions that you need to fiddle with, such
as the percentage of people in the scheme that are going to be married when they
retire, since the scheme will have to pay out 50% of the benefit to the
surviving spouse if the member dies first, which adds a further set of
calculations and liabilities.
It’s up to the trustees to decide on the whole range of assumptions that will
be used to value their scheme and work out the deficit or surplus. But as
Farrant says, “The trustees are required to take advice from their actuary and
reach agreement or at least consult with the employer on the assumptions to
be used.” The assumptions have to be prudent, though that doesn’t mean they
eliminate all risks from the calculations. What it does mean is that “the
trustees need to ensure they are more likely to overestimate rather than
underestimate the liabilities.”
The accounting valuation is dictated by accounting standard setters who get a
nervous tick just thinking about allowing FDs as much leeway to play with as
trustees have. So the accounting valuation which binds FDs but not scheme
trustees tries to be much more stringent in constraining the assumptions to be
made about the discount rate, even if it can’t do a great deal about mortality
assumptions except to urge FDs to err on the side of prudence.
Not surprisingly, the mortality assumptions of the FTSE-100 as a collective,
lag mortality statistics by an appreciable amount, and FDs will argue
vociferously against using “impossibly long-tailed” mortality assumptions since
the impact of even six additional months’ worth of longevity on their scheme
liabilities is severe. It takes much pushing and prodding by scheme actuaries
and company auditors to bump them further along the track.
Back to the discount rate and the accounting calculation, rather than allow
FDs to second-guess market returns over 35 years or more since they would all
plumb for in excess of 10%, deflating the present value of liabilities the UK
accounting standard FRS17 and the international standard IAS19 specify the rate
as that equivalent to AA-rated investment grade corporate bonds. This is
normally a well-behaved figure that tends not to wander much above 0.5% to 0.75%
above government bonds.
Unfortunately for prudence, now, with most financial stocks being AA-rated
investment grade corporate bonds, and with no one wanting to buy those debts,
the yields have gone through the roof (see page 9). Happy days for FDs, tough
times for auditors who are having some difficult conversations trying to rein
their clients back to the path of prudence. Just to put this in context, Fleming
points out that an increase of 1.7% in the AA-corporate bond yield would
depreciate a scheme’s liabilities by as much as 30% to 40%.
Will this change? The UK Accounting Standards Board issued a discussion paper
last year in which it floated two ideas that would have opposite effects on the
As far as the discount rate is concerned, the paper suggested that perhaps
there was no reason to discount liabilities by anything other than a rate that
reflects the time value of money for example, the risk-free rate on gilts, not
The discussion paper also suggests it would be better not to provide for
future (discretionary) salary increases, but instead to calculate the liability
each year based on current salary levels. “The liability to pay benefits that is
recognised should be based on the benefits that the employer is presently
committed to provide,” the paper argues.
Of all the figures, the buyout rate is the most prudent and the most serious
figure, since it is the one that the scheme buyout providers strike in order to
buy a scheme’s liabilities. In other words, it is the only one of the three
where the market is actually determining the value to be put on a scheme’s
liabilities. The only reason for a provider to do this is that it will generate
value for their shareholders, so the sum reached has to be both prudent and
At the same time, providers are in a competitive market, so the quotes they
give for scheme liabilities have to be competitive with one another. That said,
the full buyout price for scheme liabilities remains too expensive for most FDs
and most companies to contemplate. The provider will usually tweak the AA-bond
rate downwards and will be very prudent on mortality assumptions.
Included under this buyout heading is the statutory Pensions Act 1995,
section 75 liability, which is triggered if there is a corporate transaction
that amounts to the employer walking away from the pension scheme. The law says
that they can only do this by, in effect, transferring the whole scheme to an
insurance company, which is where the buyout value comes in.
The Pension Protection Fund also uses a buyout-based calculation to determine
the levy that pension schemes have to pay to the PPF. The actual PPF deficit
figure will differ from the buyout calculation to some extent, however, because,
as Farrant points out, the PPF only provides cover for a restricted amount of
According to Fleming, the best rule of thumb for evaluating the three methods,
one against the other, is that if a scheme valuation is calculated at £100m by
the accounting standard’s approach, it will be somewhere between £80m and £120m
on the actuarial valuation, and between £110m and £150m on the buyout valuation.
Between 2005 and 2007, FTSE-100 pension schemes’ deficits were very similar
on a funding basis and an accounting basis. Since mid-2007, however, the
funding deficit figure has been tracking sharply downwards towards the much more
negative buyout deficit level. “This has given companies and trustees a
headache,” says Farrant.
“Companies are currently looking at a more healthy funding position in their
company accounts, while trustees are finding their schemes have much larger
deficits and are asking companies to increase their cash contribution.”
for more on Financial Director’s final salary schemes.