Continuing my examination of the way in which pension transfers are calculated and the effect this has on pension transfer recommendations, I noted that four of the major considerations in converting a future required pension into a lump sum as being:
The level of future interest rates.
Assumed life expectancy.
The level of increases to pensions in payment.
Death benefits payable to surviving dependants.
This week, I will consider the impact of life expectancy on preserved defined-benefit pensions and the advisability or otherwise of a transfer to a private pension arrangement.
By way of an example, I have been looking at the case of John, who has a revalued preserved pension of around £21,000 a year, which is likely to cost around £400,000 at age 60 if we assume a level pension in payment and a 5 per cent underlying interest rate.
But from where did we get the figure of £400,000 as a purchase price of a £21,000 annuity? Well, I arrived at this figure from looking at standard annuity tables for a 5 per cent interest rate. These tables in turn have been calculated using life expectancy assumptions relating to mortality tables constructed up to the year 1990.
Mortality tables are historical records of life expectancy and are not intended as a precise indication of future life expectancy. I have already noted that male life expectancy has improved by around three years (for a male aged 60) over each of the last three decades.
If this improvement continues in future years, annuity rates will continue to fall. It must be remembered, as with future fluctuations of interest rates, that the gain or loss within final-salary pension schemes will be borne or enjoyed by the scheme itself. The scheme makes a promise of the level of benefits it will pay and the eventual cost of those benefits (interest rates, life expectancy and so on) is therefore the responsibility of the scheme.
The more expensive it becomes to buy those benefits, the worse for the scheme. Conversely, if the costs become less expensive (increasing interest rates, falling life expectancy and so on), a gain will arise for the scheme against its assumptions.
Looking next at the transfer scheme – a money-purchase private pension arrangement of some description – the possible continued increase in future life expectancy is a risk which should be brought to the preserved pension client's attention.
Increased life expectancy will lead to falling annuity rates (other factors, such as interest rates, remaining equal) which, in turn, means that, for this transfer alternative to produce the same annual pension as the current employer's final-salary scheme, a greater fund will be needed than may have been anticipated at the date of transfer.
If this need for a bigger fund could have been anticipated at outset, this would have resulted in a higher critical yield which, in certain circumstances, could have turned a positive transfer recommendation into a recommendation to retain benefits with the existing scheme. This future life expectancy trend cannot be overlooked nor underestimated by transfer advisers.
Thus, we now have two pension transfer risk warnings from this stage in the calculation of a transfer value (falling interest rates and increasing life expectancy) although, of course, these also represent two features which could mean that a transfer recommendation produces an even better-than-expected benefit for the client in the case of higher interest rates and/or falling life expectancy.
Another aspect I want to look at is the level of increases to the pension in payment, with particular relevance to schemes which are in the practice of paying at least part of these increases only at the discretion of the trustees.
Avoiding detailed technical discussion, suffice it to note that although schemes are obliged to pay a minimum level of increase in respect of certain pension benefits (LPI for post-1997 benefits, RPI or 3 per cent escalation for post-1988 GMP and so on), no escalation is required by law for other types of benefit .
Although many schemes voluntarily guarantee to pay escalation on this latter benefit, many others do not. These are split fairly equally between schemes which pay no escalation, because they do not have to, and those which leave annual declarations of escalation to the discretion of the trustees. Let us concentrate on those schemes paying some or all of their escalation on a discretionary basis.
Take, as an example, revalued preserved pension benefits payable at a starting level of, say, £21,000 a year which are set to increase only at the discretion of the trustees. Let us also assume that the trustees have consistently, over each of the last 30 years, increased pensions in payment broadly in line with increases in the rate of price inflation and there is no indication that they are likely to discontinue this practice in the foreseeable future.
Now, for the purposes of calculating the transfer value, should the actuary value the benefits as if they are due to be payable on a level basis – the only guarantee within the scheme – or should the valuation be based on an assumption of future escalation?
The former assumption would lead to a calculation of, say, £250,000 while the latter assumption would lead to a calculation of perhaps £325,000.
There are a number of implications for transfer analyses from this decision. If the discretionary increases are taken into account in the calculation, the transfer value will represent a fair value for an increasing pension in payment and this could represent a benefit of a transfer. This view is based on the fact that, if the client retains his benefit with his previous employer's scheme, he is by no means certain to benefit from increases to his pension in payment by the time he starts to draw his benefits. This is because the trustees may cease their current practice of granting discretionary increases.
The client could cut and run with a transfer value which represents fair value for an increasing pension in payment and, once transferred, even if the trustees cease paying such increases to preserved pensions, they cannot reclaim the overpayment, as one might view it, with the benefit of hindsight.
If, on the other hand, the discretionary increases are not taken into account and the transfer value is calculated on the basis of a level pension in payment, then one could argue that a pension transfer is attempting to match a pension which is likely to increase when in payment, with a transfer value which has been calculated on a level pension basis.
If, in these circumstances, the transfer analysis calculation is made by assuming that the existing scheme pays an increasing pension, the critical yield will be very high and will in all probability strongly discourage any thought about transferring the benefits.
So, in respect of a scheme paying discretionary increases, what should the pension transfer adviser do? I would strongly suggest that, if the history of discretionary increases is sufficiently well established and frequent, the adviser should treat them as if they were guaranteed, leading to a higher critical yield than if the scheme was treated as paying benefits on a level pension basis. Then, in the recommendation and reasons-why letter, they should bring to the attention of the client the fact that these increases could be discontinued at any time and the implications of such a future development.
Due to an error, this piece should have preceded last week's article. We apologise for any confusion
Keith Popplewell is managing director of Professional Briefing