Skip Navigation LinksHome > News > 1999 > September 1999 > Deferred Annuity Costs on Winding Up UK Pension Schemes

Deferred Annuity Costs on Winding Up UK Pension Schemes

Senior partner Huw Wynne-Griffith discusses the problems of deferred annuity costs for the trustees of pension schemes that are winding up.

The problem of deferred annuity costs for the trustees of pension schemes that are winding up

  1. When a final salary pension scheme is being wound up the trustees are faced with the need to buy from insurance companies annuities which match the pensions that members have earned in the scheme. For retired members the pensions are matched by immediate annuities and for those who have not yet retired the pension is matched by a deferred annuity which starts at the member's normal retirement age.
  2. There are issues in connection with the cost of the immediate annuities but that is a matter left to be dealt with elsewhere. This paper is concerned with the cost of the deferred annuities for members who have not yet retired.
  3. The premium for a deferred annuity is made up of :-
    • The cost of the assets required to provide the future income
    • the cost of administering those assets and the actuarial reserves
    • the cost of paying the annuity in due course
    • the additional insurance company reserves to allow for "prudence"
    • the profits of the life office
    The pension fund has to meet the first three items just as the insurance company would. However, the insurance company also has to meet the final two items whereas the pension fund does not.
  4. The consequence is that whilst that the pension fund can be shown to be capable of meeting its obligations the assets are not enough to meet the final two cost items as well. Even if a pension scheme were to be in surplus on winding up the effect of the last two items is to reduce the additional benefits available to members as a result of that surplus.
  5. The insurance company's profits are a matter of commercial pressure and pension fund trustees should have to accept this cost as a fact of life. However, the additional reserves for "prudence" are a matter of statutory regulation as well as actuarial judgement and it is this latter point that is of most concern to trustees.
  6. Having said that, there is an interaction between the "prudent" reserves and the profit element. This arises because there might be only a few life offices that are capable of absorbing the extra reserves caused by the new deferred annuities (the "new business strain"). This would be the case if there were a large volume of new business - for example, when a large pension scheme winds up. This serves to reduce competitive pressures.
  7. The main problem for trustees, therefore, is the requirement that insurance companies must maintain "prudent" reserves.
  8. The main difference between a pension fund and a life office is that the life office effectively guarantees that it will pay the policy proceeds whereas the trustees of a pension scheme do not offer such a guarantee and, in extremis, are reduced to allocating between beneficiaries only those resources available to them.
  9. One could argue that in the absence of the support of the Policyholders' Protection Act the same can be said of a life office - it only has available to it the resources of its life fund and its shareholder fund (if any) and these are also limited.
  10. The problems of trustees, therefore, reduce to the cost of the life office having to guarantee the benefits that the trustee do not have to guarantee.
  11. Currently, the cost of that guarantee is high - frequently too high for trustees to be able to afford.
  12. It is open to the trustees to say something on the lines of :- The actuary tells us that in his opinion there are sufficient assets for the scheme to be able to meet your pension promise. However, the are not in a position to guarantee this. If you want a guarantee then only an insurance company can provide this. At present the cost of the guarantee is high and as a result your pension, if it is to be guaranteed, will only be 80% (or whatever) of the non-guaranteed scheme pension.
  13. The members' expectations will be that the scheme will provide the full pension and talk of a guaranteed pension being only 80% (or whatever) of that will be seen as a frustration of the expectations of members. It is for this reason that trustees are seeking to find a way out of the problem of having to pay the high cost of the guarantee.
  14. The only way that a guarantee can truly be given is to invest in a way that provides exactly the right income at exactly the right time. That is; to follow a matched investment strategy.
  15. There are many problems related to such a requirement. One - that of income reinvestment has been partly resolved by the introduction of stripped gilts. However, other problems remain.
  16. First, there is the problem of the term of the investment. The liabilities will stretch far into the future. For example, the surviving widow pensioner of a member who is now, say, 35 years old could still be paid her pension in 50 or even 60 or more years from now. No gilt is available for such a long term (although I understand that a 50-year Swiss bond was issued recently).
  17. The second problem is that of improving mortality experience. Allowances can be made for this but one can never be certain that such allowances are adequate. The difficulty of coping with improving mortality is beyond the scope of this paper.
  18. Because the term of available investments is too short to match the projected outgo then an assumption has to be made as to the investment conditions that will prevail when the maturing gilts are to be reinvested when the time comes. The assumption will have to be a cautious one in order to ensure that the guarantee can be met. This is where the problem lies. That element of caution costs too much from the trustees' point of view.
  19. Very long term bonds will be needed to apply this solution to the problem - much longer than have been issued hitherto. Who would issue such bonds?
  20. It would appear that the government is currently out of the market (and this could well be the case for some time if the beneficial economic climate continues). There is a case for arguing that the government should continue to issue long term gilts in order to meet the demands of society for such investments to back funded pension promises (and anyway it is the Government that sets the reserving requirements for insurance companies and pension schemes). The money raised need not be ploughed back into short term investment but could be parcelled in PFI-type projects. For example, a government backed loan issued by London Underground to finance the refurbishment of tunnels, track and rolling stock.
  21. The private sector might be encouraged to offer bonds to investors instead of equity if the appetite of pension funds for matching investments to liabilities continues. This could happen if the Minimum Funding Requirement stays on the statute books (wrongly, in my view) and if the membership of schemes continues to age. In order for corporate bonds to replace gilts there would have to be some alteration to the MFR rules and to the rules for calculating insurance company reserves for the effect to be significant.
  22. Even if the market in corporate bonds were to grow substantially I would expect that the term of such bonds would be too short. Moreover, pension scheme trustees are necessarily cautious and so long corporate bonds would probably not offer adequate security (unless there was a very large spread of available corporate bonds covering a wide variety of companies and industries). The conclusion appears, therefore, to be that only gilts could satisfy the needs of trustees for very long bonds
  23. However, the problem could also be eased if insurance companies were able to treat equities as a matched investment for long term deferred annuities. At present equities are not seen in this light and so a "mismatch reserve" has to be maintained which contributes to the high cost of the guarantee. To some degree there appears to be a mismatch already in that the pricing of annuities seems to be based on long gilts whereas the investment strategy actually adopted is to invest in other than such bonds.
  24. It is difficult to find other than anecdotal of this (the published data is not that detailed). According to market comments life offices generally have recently been large purchasers of long gilts in order to reduce the level to which they have unmatched liabilities. The longest gilt in issue, the Treasury 6% 2028, is itself only some £8 billions whereas the life offices are said to have been buyers of long gilts to the tune of £13 billions (and this only covers an estimated half of their unmatched liabilities).
  25. There should be some concern that life offices and pension schemes are unable to buy investments to match their liabilities. Perhaps the definition of "match" should be revised.
  26. Even on a conservative basis of valuation a portfolio of equities should have a value that exceeds that of a corresponding portfolio of gilts - all experience shows this to be the case over the longer term (and we are concerned only with the longer term by definition). The current regulations concerning insurance company reserving do not allow credit to be taken for the expected additional return from investing in equities.
  27. If the authorities could be persuaded to alter this approach then the cost of deferred annuities would almost certainly fall. However, the element of prudence would evidently have been reduced and this would have to be acknowledged and dealt with in some way in order to maintain the good name and reputation of the insurance industry.
  28. The reduced "prudence" could be paid for in the form of a weaker guarantee. Policies issued under this regime would be similar perhaps to equity-linked policies but with some element of guarantee.
  29. This would be an important characteristic of a policy of this kind - that the level of guarantee as to the provision of the benefit would be consistent with the premium that was paid: the higher the premium, the greater the guarantee and vice versa. The range would cover the entire spectrum from the current "100%" guaranteed non-profit deferred annuity down to a simple money purchase benefit with no guarantee at all.
  30. Trustees would then choose the insurance company by reference to the level of guarantee allied to the premium cost rather than just to the premium cost as at present (with the level of the guarantee being the same at, in effect, 100% across the whole market). The market in annuities is small in this country. There are few insurers that are capable of taking on any large block of business on a winding up because of the potential cost of financing the "new business strain". The dearth of participants results in a less than competitive market where market forces as such are too weak to overcome the combined effect of the reserving requirements imposed by the authorities and the lack of pressure on the insurers that they will lose the business to "this month's competitor".
  31. An arrangement like this would similar to the mooted "Central Discontinuance Fund" but instead of being a single fund the same liabilities would be shared around the insurance market generally.
  32. There is a big difference between the rights of pension scheme members when a scheme winds up and their expectations. This has always been the case and it will take a lot of education before the two are accepted as being different (if ever). If this long process could be short-circuited in some way then that would ease matters for trustees in that the expectations of members could be re-defined as something other than the pension earned to the date of the winding up. Currently, expectations could be frustrated if there were inadequate funds. Therefore, these expectations are not guaranteed. Replacing them with annuities would add the value of the guarantee that is currently absent. To this extent therefore the pension scheme member on a winding up would getting the added bonus of the guarantee and without having to pay for it.
  33. One very different approach to providing winding up benefits would be to allow wider pension commutation rights in such a situation. At present, commutation is allowed only at retirement and then only within limits and the lump sum is tax free.
  34. On a winding up the members who have not yet retired could be allowed the option to take some of their pension rights as an immediate cash sum (possibly free of tax up to a limit).
  35. The benefits up to some defined level would have to be provided in the form of a non-commutable pension but the balance could be commuted for cash. The trustees could require that an option of this kind would only be available in the context of the remaining benefits being on a money-purchase basis. This would then remove the need to provide the member with a defined deferred pension by way of a guaranteed deferred annuity.
  36. The present situation is undesirable. Members feel disappointed that they cannot always have their deferred pension guaranteed in full. This is especially galling for trustees because the reason has to do with the system of reserving for mismatched investments. The trustees could transfer their responsibilities quite adequately to insurance companies in return for policies which were guaranteed at a lower level if the insurance companies could invest in equities (without having to hold mismatch reserves).

For further information, please contact Huw Wynne-Griffith at the London office.

Huw Wynne-Griffith, September 1999.


Share this