Actuaries are only human

Neil Copeland

or Subscribe to Feed

Can we get something straight? Final salary pension liabilities are not linked to the yield on AA rated corporate bonds. Nor are they linked to returns on equities. Nor are they linked to the yield on gilts, index linked or otherwise.

Final salary pension liabilities are the aggregate of the payments promised to members and their beneficiaries over the life time of a scheme. For closed schemes this is a finite amount, but the actual amount can’t be known until after the death of the final member and beneficiary. Add up all the payments to members, beneficiaries, advisers, the PPF and other parties, and that’s your liability.

So if you have been rash enough to promise your employees final salary pensions and had your original vague aspiration to ensure those faithful and loyal employees who stuck with you through to their retirement had a reasonable income in retirement,  turned into a gold plated, inflation proofed guarantee for all and sundry, including that spotty herbert who worked in accounts for a couple of years in the late eighties and was worse than useless, well,  I’m afraid you can have no idea what it’s going to cost you in the long run.

Obviously employers don’t like uncertainty, and try to eliminate it from their businesses in as many areas as possible. Final salary pension schemes are wildly uncertain. The actual liability will depend on a wide range of factors including advances in medical science, pandemic killer diseases, governmental whims and climate change, to name just a few.

To help them try and make an estimate of the cost of final salary schemes, and set aside some sensible provision for the funding thereof, employers turn to actuaries.

Actuaries are usually very intelligent, highly professional, capable mathematicians, and in the words of that indispensible reference manual, the Hitch-hikers Guide to the Galaxy “Mostly harmless”.

However there are some important facts that need spelled out in relation to actuaries:

1) They cannot accurately predict what return your scheme is going to earn on its assets next year, never mind over the next 40 to 50 years
2) They cannot predict what price inflation is going to be over the same periods
3) They, not unreasonably, can’t assess how long your scheme members are going to live. At this point I would ask you to note that some scientists think that the first person to live for 1000 years has already been born which, as Bridget Jones might say is v.v. bad
4) Their models allow for only a relatively few variables and, given points 1 to 3 above, the answers they generate are very likely to be wrong. If an actuary actually divined the right answer at a particular valuation point, this would be due to what someone unfamiliar with the mathematics of very large numbers would call coincidence.

Actuaries have come up with a number of approaches to derive assumptions for valuations, all of which can be argued for or against depending on the interests of a particular group or individual. My favourite is in relation to the inflation assumption and goes along the lines of “Well the Bank of England’s target for CPI is 2%, add on a bit because RPI is usually higher and then add another bit just to be on the safe side.” Who said actuaries don’t have a sense of humour!

You would think the inherent inaccuracy of actuarial valuations is an obvious point, but the amount of time expended by employers and trustees debating whether the inflation assumption in their valuation should be 3.25% pa rather than 3.5% or asserting that the “right” discount rate is 6.75% pa  rather than 6.5% pa, suggests otherwise.

Employers can get certainty about their pension liabilities, not by eliminating risk, as this is impossible in this context, but by transferring it to another party by one of a number of methods, including:

1) Purchasing annuities, in which case the risk is transferred to an insurance company
2) Paying transfer values,  in which case the risk is transferred to individual members
3) Regularly culling older scheme members, in which case the risk is that someone would notice and prosecution swiftly follow. Unless you were in Switzerland!!

Of the options noted above, and assuming 3) has been discounted, paying transfer values, even at an enhanced level, is likely to be the most cost effective option from an employer’s perspective.

In an article in Professional Pensions, Martin Bogira of Prudential slightly misses the point of these exercises. In my experience most employers who carry out such an exercise, whilst fully appreciating the cost implications, do so out of a desire to reduce future uncertainty with regard to pension costs. The cost of enhancement will inevitably be high if taking on the risk is to be attractive to a member, but is likely to be less than the premium charged by an insurer for taking on the same risk. If this means employers see a one off funding loss in their accounts in a particular year, they usually feel that that is a small price to pay in return for increased certainty.

Neil Copeland

Post by Neil Copeland

Director, pensions consultant and adviser to trustees and employers on all aspects of work based pension schemes.