The High Court has handed down an important judgment on equalisation of guaranteed minimum pensions (GMPs) in Lloyds Banking Group Pensions Trustees Limited v Lloyds Bank plc and others (Lloyds).
The pensions industry now faces the cost and complexity of grasping one of its fiercest nettles - how to implement the equalisation of GMPs in practice. This Insight summarises what the problem is with GMPs , for those who are not immersed in the topic.
In addition, our experts have distilled the judgment in Lloyds to bring you a key point summary, a plain English explanation of the issues and suggestions on next steps for employers and trustees.
Key points on why GMPs are unequal
GMPs are intended to replicate certain state benefits
GMPs are intended to replicate members' State Earnings-Related Pension Scheme (SERPS) benefits to try to ensure that he or she would not be worse off as a result of having contracted out of SERPS.
Working life for GMP purposes ends at age 60 for women and age 65 for men
GMP works out average earnings over the 'working life'. Working life is defined as the period starting from age 16, and ending at age 60 for women but 65 for men. This is where the inequality comes from.
GMPs are unequal for a number of reasons
Women can accrue the same GMP as a man in a shorter time. This was designed to reflect a shorter working life for women. In addition, women's GMP becomes payable five years earlier (i.e. at age 60 compared to a man's GMP which becomes payable at age 65). This is further complicated because of how GMPs interact with the rest of their scheme benefits (the excess) when pensions are revalued in deferment and increased in payment.
Anti-franking legislation adds another layer of complication
So-called "anti-franking" legislation operates to prevent the excess being used to satisfy the obligation to revalue the GMP during the interval between the member leaving contracted-out service and the GMP coming into payment.
What is the problem with GMPs?
Unequal accrual rate and payment date
GMPs accrued between 1978 and 1997 for members of occupational pension schemes which were contracted out of SERPS.
The GMP was intended broadly to replicate the member's SERPS benefit to try to ensure that he or she would not be worse off as a result of having contracted out of SERPS.
A GMP is a career average benefit. It works out average earnings over the "working life" (or, to be more accurate, the part of the working life that falls after 1978 when SERPS started) and pays a pension in respect of them.
Working life is defined as the period starting from the tax year in which the individual achieved age 16, and ending at the tax year in which they achieved age 60 for women but 65 for men (reflecting the old state pension ages). This is where the inequality comes from.
Most occupational pension schemes had unequal retirement ages for men and women, usually 60 and 65, which they have since had to equalise. However, the career average nature of the GMP benefit, and its interaction with the overall benefit payable by the pension scheme, makes GMP equalisation considerably more complicated than equalisation of most non-GMP benefits.
To understand why, we need to understand how a GMP is calculated. For ease of reading, the following simplified description uses words whereas the legislation uses formulae. This description only applies to GMPs accrued since 6 April 1988.
To calculate the GMP, a scheme must take all the member's earnings that are pensionable for SERPS purposes, and revalue them up into today's money. That total is then divided by the length of the member's working life. In determining the length of working life, any part of it that falls before 1978 is ignored and there is a minimum length of 20 years. This division gives the average annual earnings in today's money. That amount is multiplied by 20%, and the result is the GMP.
As a woman's "working life" is five years shorter than an equivalent man's (ending at age 60 instead of age 65); when a woman's total earnings are divided by the length of the working life, the result will be bigger than it would be for a man who had the same earnings history and date of birth. Dividing by a smaller number gives a bigger result.
Put simply, a woman can accrue the same GMP as a man in a shorter time.
There is logic to that; because a woman's working life was shorter, she needed to be able to accrue her GMP faster in order to get the same overall benefit.
The problem is then compounded by the fact that the woman's GMP not only accrues faster, it becomes payable five years earlier, at age 60, whereas a man's GMP payment age is age 65. Again, this reflects the old state pension age.
A woman's GMP would often be postponed (e.g. if her retirement age in the pension scheme was 65) but she is then entitled to an increase in respect of the postponement. The man's GMP comes into payment at 65 with no such enhancement.
Interaction between the GMP and the overall scheme benefit
It would seem from the above that the inequality in GMPs favours women over men, but that is not in fact necessarily the case. This stems from the fact that the GMP is not a standalone benefit; it is a guaranteed minimum to be compared against the (usually bigger) pension payable by the pension scheme.
At the point in time when the member ceased being in contracted-out pensionable service, the GMP simply an underpins the main benefit. However, from that point on, the main benefit is divided into two separate elements of pension: the GMP, and the remainder of the pension above the GMP (commonly called the "excess").
These two elements are usually increased differently, both in respect of revaluation during the period (if any) between leaving service and drawing a pension, and increases due when the pension is in payment. The increases payable on the GMP may be less generous than the increases payable on the excess. That is more likely to be the case in respect of increases payable after the pension starts to be paid, but it varies.
So, counterintuitively, giving somebody a bigger GMP could make them worse off because the total pension is still the same; and by making the GMP bigger, one simply makes a larger proportion of the pension subject to the less generous pension increases.
It becomes even more complicated where there is an interval between the member leaving contracted-out service and the GMP coming into payment. So-called "anti-franking" legislation operates to prevent the excess being used to satisfy the obligation to revalue the GMP during that interval. The operation of this legislation is notoriously convoluted and the details are beyond the scope of this note.
For present purposes, it suffices to say that the legislation requires an "anti-franking" test to be carried out at the commencement of payment of the GMP. As that test needs to be carried out five years later for men than for women, it adds to the potential for inequality. This element tends to favour men as, when the test is applied five years later, there is five years' more GMP revaluation to be protected by it.
The above is (believe it or not) a simplified explanation, but it demonstrates that this is not a simple issue. It is the complexity in the benefit design that makes equalising GMPs difficult, not only in terms of doing the work, but even to identify what steps should be taken to redress the inequality.
The position in any particular pension scheme may deviate from the position set out above, but the above will apply to most pension schemes which were contracted out in the relevant period.
For a summary and analysis on what the Lloyds case now requires to be done about this, and the implications, please see our Insight on GMP equalisation - what now for trustees and employers.
 Although the GMP can and does differ from the corresponding SERPS benefit in certain circumstances.
 The calculation was different from 1978 to 1988, but that does not matter for present purposes as the duty to equalise between men and women only applies to benefits deriving from service from 17 May 1990.
 but not always - the benefit design of the scheme in question always needs to be specifically checked