Mike Otsuka’s extraordinary diligence in understanding the pension scheme he is a member of, is an example to us all. In this article he delves into the archives and uncovers some disturbing statements. They shed light on the current “unaffordability” of USS arose.
Scheme actuary declared 14% employer contribution rate insufficient to cover future service
Those who have been following USS are aware of the controversy regarding the claim that employers took a ‘contribution holiday’ between 1997 and 2009, during which period employer contributions went down from 18.55% to 14%. Universities UK denies that employers have ever taking a ‘contribution holiday’. But this denial appears to be based on the claim that they continued to pay 14% as opposed to taking a contribution-free holiday. Fair enough. So let’s substitute the phrase ‘contribution skiving’ for ‘contribution holiday’, where skiving is consistent with a reduction in contributions below what one ought to be paying, without necessarily going completely AWOL.
I believe that UUK would maintain that their reduction from 18.55% to 14% in 1997 was justified on the following grounds: Prior to 1997, employers paid 4.55% above and beyond their 14% regular contributions for ‘future service’ (i.e., pension accrual corresponding to a worker’s future years of employment). This 4.55% consisted of payments to make up for a past funding shortfall in the Federated Superannuation Systems for Universities (FSSU) that USS superseded when it was established in 1975. These payments were the equivalent, under older pensions legislation, of the deficit recovery contributions that are called for by current pensions legislation. By 1997, employers had completed the necessary payments to make good this funding shortfall. Hence, from 1997, employers simply carried on making the same 14% employer contributions that they had been making in previous years.
It may seem, therefore, that employers are beyond reproach for their decrease in contributions from 18.55% to 14%.
Not so fast, UUK!
I’ve stumbled upon the following in USS’s 31 March 1999 full actuarial (triennial) valuation.* The scheme actuary writes:
6.1.1. The method used [for this valuation] is known as the projected unit method. This is the same method as was adopted for the 1996 valuation and I consider it to be an appropriate method to adopt….
8.1.1. Under the projected unit method, the normal contribution rate required from the Institutions for future service benefits is 16.3 per cent of salaries. This compares with the normal rate at the 1996 valuation of 14.6 per cent.
10.2.1 Based on the method and assumptions described in this report it has been agreed that the Institution contribution rate will be maintained at 14.0 per cent of salaries. To fund the reduction of 2.3 per cent below the Institution contribution rate of 16.3 per cent of salaries for the next 11 years will require the use of £561.3m of the Main Section surplus….
In other words, the scheme actuary reported that 14% is now well below the amount necessary to fund future service but that it had been agreed to run down the scheme’s surplus by over half a billion pounds in order to keep regular employer contribution rate down to 14%. They ought instead to have kept this surplus untouched as a rainy day fund to get the scheme through difficult times.
[*] USS’s actuarial valuations prior to 2011 are no longer available on their website. The 1999 valuation to which I link above was one that I had previously downloaded from their website.
Mike is quite right to do some historical digging.
I’m puzzled why actuarial reporting standards (the little known TAS 300, for example, which even TPR refer to in their most recent annual DB statement, and there are other standards, some of which used to be call TAS-R) don’t require more than just a reconciliation of movements since the previous triennial.
If the forward duration of a scheme is, say, at least 20 years (and the USS may be longer, for all I know), why are trustees not shown in summary the historic funding assumptions’ development over at least the previous five, six or seven triennials? (I think quoted companies at least summarise their last ten years, and pensions are a longer game.)
Over to you, FRC, as you are responsible for these standards, although actuaries must have views on this too.
I hope we get some comments on this George. I’ve noted elsewhere that reporting standards favoured employers in former times, now they are being used as an excuse to transfer risk to lecturers. Can you have double(accounting) standards?
I think most people are enjoying the Bank Holiday sunshine, Henry?
Accountants have had double standards for centuries, starting with double entry, and again that’s in the FRC’s gift to fix, or to let sleeping dogs lie ….
…. but I was specifically thinking of actuarial reporting standards this time.
I have noticed, tho’, that pensions accounting disclosures used to show more than 2 years of trend information and also “best estimates” of expected returns for each asset class, but these disclosures seem to have disappeared on the FRC’s watch.
Professional small print was always good at limiting the period for which opinions can be considered “current”, but if Mike’s right then it’s extending to cleaning up the archives to prevent light being shone back on earlier assumptions.
A comment on this statement:
” it had been agreed to run down the scheme’s surplus by over half a billion pounds in order to keep regular employer contribution rate down to 14%. They ought instead to have kept this surplus untouched as a rainy day fund to get the scheme through difficult times.”
– back in the nineties, the Inland Revenue required any surplus exceeding 5% to be eliminated (IR12 para 13.19 et seq). With hindsight now of course this policy was mistaken, but it was a reason why in some schemes the employer contributions were reduced, even to zero for a while some cases. I don’t have the USS figures to hand, but I’m guessing this could have been the reason for this decision.
With hindsight, the funding assumptions which were used back in the days of the Inland Revenue 105% rule were imprudent.
I seem to recall investment return assumptions of up to 9% nominal, 2 or 3% real, and of course we now know the mortality assumptions used then had failed to keep up with emerging improvements.
It would have been possible to stay within the 105% ceiling by increasing the estimated liabilities. All too often, however, the discussions between trustees and sponsors were framed in terms of contribution holidays for the sponsors (and sometimes the active members as well) and benefit increases for pensions in payment and deferred pensions.
I am sure Ian that this was a primary contributor, though I think the point of the article is different. Pension schemes aren’t finite and employers who fund them must accept that there are times when funding are easier than others. Mike’s historical perspective makes this point well.