Annex B - Actuarial advice to the investigation
Jump to
Introduction
1. This annex sets out the technical scope of – and gives further detail about – the actuarial advice I commissioned to help me to investigate complaints about the various changes to the actuarial basis of the MFR test and an alleged failure to inform scheme members and trustees of the effects of those changes.
2. The report produced by my advisers is over 60 pages long. This annex therefore can only be a brief summary of their advice.
3. The work done for me by my advisers included:
(i) an assessment of whether the MFR – at both its original level and at those after the subsequent changes to it – provided to non-pensioner members of a final salary scheme a ‘reasonable expectation’ of achieving equivalent benefits through a personal pension on wind-up;
(ii) an assessment of whether the principal changes to the MFR were significant; and
(iii) an assessment of whether those changes had the effect of strengthening or weakening the protection afforded to the accrued pension rights of final salary schemes as measured on the MFR basis.
Technical scope of advice
4. This work involved the application of a stochastic model developed by my advisers (and also used by them in relation to their other work).
5. It also involved the use of statistical methods to generate a large number of possible outcomes for asset values - in this case equity and bond values - over a given term.
6. My advisers generated sets of possible outcomes at the four key dates in the development of the MFR basis: January 1996, April 1997, June 1998 and March 2002.
Assumptions used
7. At each date, they also generated the outcomes based on the economic conditions at the time, as an attempt to avoid the use of hindsight - to put them in the position that they might have been in, if they had been assessing the MFR in this way at that point in time.
8. My advisers recognise that this calibration of their modelling to the historic economic conditions cannot be perfect. The stochastic outcomes were generated by a model that itself is based around a set of long term assumptions.
9. They therefore used the same long term assumptions at each of the four dates in order to avoid introducing distortions to the results, although they advise me that, if required, they could have provided analysis on the effect of varying the assumptions.
10. As a result, there might have been minor differences in the probabilities had my advisers selected different long term assumptions.
11. I am told that it is therefore more helpful to focus on the movements in the probabilities between the different dates - rather than the absolute level of the probabilities at the starting date.
Process of modelling
12. The starting value (the initial investment) and the target value (the amount required to provide the deferred pension given up when the transfer value was taken) are both set with reference to values derived on the MFR basis.
13. For simplicity, the starting value was the MFR transfer value, i.e. dependent on the equity MVA, which would have been paid in respect of an accrued annual pension of £1,000, which is index-linked in payment and payable from age 65.
14. This starting value was then run through the stochastic model to generate 5,000 simulated fund values (net of investment fees and expenses) that might result from the investment of this amount, on a 100% equity basis, in a personal pension vehicle.
15. The model also allowed for the switch from equity to bond investments over the 10 years before retirement age, as implied by the MFR basis.
16. My advisers then assessed the percentage of these fund values that exceeded the MFR target value.
17. The MFR target value was calculated as the accrued pension increased, between the date the transfer value was assumed to be taken and retirement age, in line with the MFR assumption for deferred benefit revaluation (4% p.a.).
18. This projected amount was then converted to a value using the MFR pensioner basis, which is linked to gilt yields at the relevant date.
19. This then provided an assessment of the probability of the MFR transfer value being sufficient to provide the pension being given up.
Summary of advice
20. In relation to whether the MFR delivered the ‘reasonable expectation’ policy intention – defined as a 50% chance - for non-pensioners at the time of its conception and at the time of its implementation, my advisers tell me that, at those times, on the basis of their analysis as outlined above it is their opinion that the design of the MFR was consistent with that policy intention.
21. They also tell me that, although the MFR was a measure which was intended to test the position upon the ‘discontinuance’ of a fund, it was derived from an actuarial valuation method used for funds which were ‘ongoing’. For this reason, the MFR had a number of design flaws which meant that, whilst it delivered the policy intention at the time of its introduction, it was not robust to changing market and other economic conditions.
22. Such changes included:
- lower inflation;
- lower nominal interest rates;
- changes in the make up of the All Share Index;
- changes in annuity buyout costs;
- the removal of tax credits on dividends; and
- changes in dividend policy.
23. In addition, they advise me that the MFR contained a ‘discontinuity’ in that the amount required to be held immediately before retirement was often less than the amount required to be held immediately afterwards.
24. In relation to the ability of the MFR to deliver the policy intention for non-pensioner members over time, my advisers firstly considered the position if there had been no changes to the MFR.
25. They calculated the change in the probability that a non-pensioner would achieve their full accrued benefits if their scheme wound up. Their advice is that (if there had been no changes to the MFR basis) the initial probability asserted by the Government of an even chance (i.e. 50%) would have fallen by March 2002 as follows:
a) (ignoring the detrimental effect of rising annuity buyout costs) to a little under 45%; and
b) (including that detrimental effect) to around 35%.
26. The advice to me is that the MFR’s inability to adapt to changing market conditions alone would have resulted in it reaching levels where it failed to achieve the policy intention ascribed to it.
27. My advisers then tested the effect of the changes made in 1998 and 2002. They advise me that the changes had the effect of reducing the probabilities above to just above 35% and potentially to less than 30%, respectively.
28. It is the opinion of my actuarial advisers that at that time the MFR was already failing to deliver the Government’s policy intention and that the changes exacerbated that failure.
29. My advisers then considered whether the changes were ‘significant’. They told me that the MFR was a minimum. To varying degrees, depending on the wording in schemes’ documentation, trustees had powers to enforce contribution rates higher than the minimum.
30. However, I am told that, in fact, while most trustees could take steps to reduce investment risk and thereby increase the MFR minimum contributions, few did so. In practice therefore a significant number of schemes only enforced the minimum required by the MFR.
31. I am told that, for some companies, particularly where the employer set the contribution rate, the reduced protection afforded by the changes and the 2002 extension to the period over which deficits could be corrected meant that, in practice, lower employer contributions were paid than would otherwise have been the case and some other steps available to companies and trustees were not considered.
32. I am advised that the impact of these changes would have been very scheme specific - for example, for well funded schemes they might have had very little impact.
33. However, where the MFR ‘bit’ and where employers went insolvent, it is the opinion of my advisers that the combination of the changes to the MFR formula and deficit recovery periods did constitute a significant weakening of protection to members.
34. For the Government’s policy intention to be met, my advisers inform me that the changes made by it would have needed to be in the opposite direction.
