Avon Cosmetics Limited v Dalriada Trustees Limited & Ors

This judgment is given in proceedings which raise questions as to the validity of one of a number of amendments made to a pension scheme referred to as the Avon Cosmetics Pension Plan (the “Plan”).

In short, the trust documents under which the Plan operated contained a power of amendment (the “Power of Amendment”). That Power of Amendment was subject to a proviso or fetter (the “Fetter”). The Fetter prevented any amendment which, at the date it was made, affected prejudicially (a) any pension in payment at that date or (b) any rights accrued or secured up to the date on which the amendment was made.

There is a point regarding the proper effective date from which the amendments took effect but I need not go into that point of detail in this judgment.

The relevant amendment that I was asked to consider comprised but one part of a series of amendments going well beyond simply affecting accrued rights falling within paragraph (b) of the Fetter. I am asked to assume that one part of that series of amendments (the “relevant CARE amendment”) did however prejudicially affect rights falling within paragraph (b) of the Fetter.

The change in question affected (or purported to affect) the accrued benefit rights of certain persons who, at the date of amendment, were continuing in service (i.e. employment). By “accrued rights” I mean pension rights that had already been acquired, though not then in payment. Such entitlements arose in respect of the relevant members’ service before the date of amendment.

At the time of the amendments, the accrued benefit rights of the relevant persons were measured by reference to (eventual) final salary. The change made by the relevant CARE amendment was to sever this link to final salary. Instead, under the amendment, the relevant member was treated as having retired at or about the date of the amendments. For these purposes, that employee’s then salary was taken as being their “final salary”. That benefit, that is, the amount to which the member would have been entitled if they had then retired, with his or her then (frozen) final pay, was to be valued and thereafter treated like a deferred benefit, being made subject to statutory revaluation each year by the applicable rate of the Retail Price Index (the “RPI”) (subject to the maxima set by the Pensions Schemes Act 1993). I refer to this ongoing changed position as involving benefits being determined on the “Revaluation Basis” and the position prior to amendment as being that benefits were determined on the final salary basis or “FS Basis”.

As I have said, the amendments as a whole affected more than the accrued rights, as at the date of the amendments, of those in continuing service and whose benefits had, prior to such amendment, been linked to final salary. As regards what had been the Final Pay Section of the Plan, that part of the Plan was closed, and so, as regards the future, benefits in respect of future service were also no longer based upon final salary but upon Career Average Revalued Earnings (“CARE”). The amendments as a whole have been referred to before me as the “CARE Amendments”.

More recently, it has been identified that, as regards accrued benefits as at the date of amendment for those whose accrued benefits were, prior to the amendment, to be valued on an (eventual) FS Basis, the CARE Amendments are capable of prejudicing some of these members. In the case of each relevant member, the precise effect on that member’s accrued benefits of the CARE Amendments (if valid) can only be assessed when their benefits crystallise on retirement or other leaving of employment. Nevertheless, the current actuarial assessment is that, in financial terms, certain persons will be better off under the CARE Amendments and that others will be worse off. In the first case, this will be a result of, during the relevant period of service after the date of the amendments, the rate of inflation being higher than the increases in salary of the relevant persons. In the latter case, it is the result of the member’s salary increases being greater than the rate of inflation.

The first category of members, that is those who are better off under the relevant CARE amendment, has been referred to before me as “Revaluation Winners” . The second category of members, that is those who are worse off under the relevant CARE amendment, has been referred to before me as “FS Winners”. As I have said, it is only possible to know precisely what the financial effect of the relevant CARE amendment (if valid) will have been on any particular individual member when their rights crystallise. That is because it is only at that point that the precise applicable RPI position will be known and the relevant final salary will be known.

In the case of FS Winners, it appears that the CARE Amendments are ones that, on the face of things, have or may have prejudiced the relevant members’ accrued rights in a manner that is not permitted by the Fetter. The ultimate question raised by the proceedings is therefore what the consequence is of the Power of Amendment being exercised in such a way as results in a situation that may not have been permitted by the Fetter.

As originally conceived, the proceedings were to answer the question of whether or not the CARE Amendments were invalid such that members’ rights in respect of the relevant accrued benefits could not be determined in all cases on the Revaluation Basis. If the Fetter had that effect then it was considered (on advice) that the consequence would be that the relevant accrued rights would be calculated on the Revaluation Basis (as per the CARE Amendments) but with what was called an “Underpin”, namely that if a calculation based on the unamended FS Basis yielded a greater value that would be substituted for the value achieved by applying the Revaluation Basis. Put another way, members’ entitlements in respect of the relevant accrued benefits would be provided by way of an underpin entitling them to benefits calculated on the basis of the better of the FS or the Revaluation Bases. Thus, it was considered that the court would be deciding whether the relevant amendment was in breach of the Fetter and, if it was, precisely how the FS Winners’ accrued entitlements as a consequence should be dealt with.

The broad basis for the assumption that the Underpin would apply if the Fetter operated to prevent the CARE Amendments taking effect, as regards relevant accrued rights at the date of amendment, in full was as follows:
(1) If the CARE Amendments prejudiced FS Winners, that was prevented by the Fetter. To that extent, under the Underpin, such persons would receive, as regards their accrued rights as at the date of amendment, the higher value by reference to the FS Basis;

However, in circumstances that I shall go on to explain, proceedings were intimated against various professionals which had been involved in advising the Trustees and the employer at the time of the CARE Amendments (the “Potential Part 7 Defendants”). They were defined in this way because proceedings in negligence under Part 7 CPR may be brought against such persons. In effect, by claim letter it has been said that the professionals in question failed to identify the potential problem, caused by the existence of the Fetter, in validly effecting the CARE Amendments.

The lawyers acting for one of the professionals noted that there was a duty to mitigate loss. At that stage, it had been estimated (as set out in a Watson Wyatt Report dated 25 September 2019, and taking a calculation date of 1 January 2018), that the extra costs of benefits, were a FS Basis to be applied to all relevant members’ relevant accrued rights, would be in the order of £12.3 million whereas if the Underpin applied, the extra costs would be in the region of £17.8 million. Those figures have since been updated to deal with the assessment of the position on the basis of the membership position as at 1 January 2021 and the financial conditions as at 31 December 2022. Approximate liabilities at that stage were assessed as being in the region of £6.7m with a reinstated final salary link and £12 million to provide members with either a final salary link or benefits on a Revaluation Basis, depending on which was higher. Although the extra costs of providing an underpin therefore remain in the order of over £5 million, the actual estimates of costs have changed quite significantly from the calculated position using members’ details as at 1 January 2018.

The Potential Part 7 Defendants noted that the claim form now before me, in its then form, theoretically raised the issue of whether an Underpin should apply or not, if the CARE Amendments were found to be invalid but that, looking at the representation orders sought, it did not appear that any person was going to be appointed on behalf of those interested to argue either for or against the Underpin. Put more precisely, the claim form, it was said, did not clearly envisage an argument as to whether or not the effect of the invalidity of the amendment as regards FS Winners (if established) was that the amendment was (i) totally ineffective to effect changes to the calculations of accrued benefits of both FS Winners and Revaluation Winners or (2) only invalid as regards FS Winners, the amendment being effective in respect of Revaluation Winners.

The claim form was accordingly amended to raise that issue, or raise it more clearly.

As I shall explain, various representation orders were also sought. I deal with those at the end of this judgment. Although put in terms of arguing “for or against” the Underpin, I find it helpful to unpack that description a little more. For the FS Winners, their interest is to argue that they are entitled to have the FS Basis applied to the valuation of their accrued benefits arising from pre-amendment service. It is also in their interest to argue that the Revaluation Winners should have their like accrued benefits valued on the FS Basis. This is because the application of the Revaluation Basis to Revaluation Winners would result in a much larger payment out from the Plan and thus, at least potentially, create strain as regards available assets of the Plan to meet such liabilities. The relevant argument for the FS Winners is that, if the relevant CARE amendments are invalid as against FS Winners’ relevant accrued rights then they also should not apply to anyone else (i.e. the Revaluation Winners) and the relevant CARE amendment is wholly invalid in its effect.

One could be forgiven for thinking that the validity of the CARE Amendments on the relevant accrued rights and their effect was a question that would most naturally be answered holistically by considering the position of each of the identified categories of persons with relevant accrued rights and, further, that the logical starting position might be to consider the effect and validity of the amendments on the FS Winners and, in light of that, to consider the position of the Revaluation Winners. That, however, is not the way in which the proceedings have developed.

It was, I am told, always envisaged that the position of the FS Winners would be capable of being resolved by a process of compromise, sanctioned by the Court. I am told that that compromise has been reached (or is largely agreed). The matter is envisaged as coming before the Court in February 2024 for the Court to give any necessary sanctions. In those circumstances, and because, in addition, it is not thought appropriate for the Court’s approach to be tainted by the disclosure to it of the compromise and its basis, nor any basis upon which such compromise is considered to be beneficial to particular persons including classes of member, I am invited to deal with the question of the effect of the relevant CARE amendment on the Revaluation Winners, assuming, but without determining its effect on FS Winners. The result is, in my view, somewhat unsatisfactory in the sense that the process I have to engage upon is a somewhat artificial or partial analytical process.

I made clear that I could only proceed to determine the rights of the Revaluation Winners if I were to make certain legal/factual assumptions about the rights of the FS Winners. The precise factual/legal basis upon which I act and the precise questions asked of me, as set out in the (already amended) claim form, can be formulated in the light of this judgment. For present purposes, however, it is agreed that I should assume that the effect of the Fetter on the position of the FS Winners is that they are entitled to be treated on a Final Salary Basis and the relevant Care amendment as affects them transgresses the Fetter and is thus invalid as against them.

For present purposes, it suffices to say that (on the assumption that appropriate representation orders would be granted), Mr Keith Bryant KC leading Ms Naomi Ling, argued that the effect of the Fetter was to invalidate the CARE Amendments, so far as they purport to affect relevant accrued rights, across the board. In particular, they submitted that it was not possible to sever the effect of such amendments as regards different categories of member. This was on two main bases, first that the two groups identified by the concepts of Revaluation Winners and FS Winners are not sufficiently discrete and secondly that an element of the test for severance in this context is whether or not the same decision would have been taken had the decision maker known of the legal invalidity, flowing from the Fetter, of what was purportedly effected by the CARE Amendments and that that test is not met.

Mr Hitchcock KC leading Ms Seymour, on the other hand, argued that the Fetter only potentially invalidates the Care Amendments as regards FS Winners and that the Care Amendments are valid so far as they affect the rights of Revaluation Winners.

Mr Newman KC for the trustees took a neutral stance, though assisting with the law. An important potential role for him will be in identifying practical problems from certain conclusions that the Court may otherwise reach.

I am grateful to all Counsel for their helpful and focussed submissions, both oral and written and to the solicitors involved, not least for their preparation of the helpful evidence and bundles.

As I shall explain, this judgment deals primarily with the legal issue of what, assuming the relevant CARE amendment involves what is often referred to as an “excessive exercise” of power, is the consequence for Revaluation Winners. That is, on the assumption that the power to make the relevant CARE amendment has been exceeded by purportedly moving FS Winners to the Revaluation Basis. For these purposes, the concept of an “excessive exercise” is that the relevant power (of amendment in this case) has been exercised in a manner that was not conferred by the power in a jurisdictional sense. In other words, the arguments focus on the scope of the power and the jurisdiction it grants to make the CARE Amendments so far as affects relevant accrued rights or, looked at another way, the vires to make the relevant amendment.

Although not arising in this case, at least as argued before me, it is also necessary to consider and distinguish two other principles which may result in an exercise of a power being open to challenge. This is because some of the cases in question raise these heads of potential challenge and discuss them. The two principles in question are the “inadequate reasoning” basis of challenge (labelled as such by Lord Walker in Pitt v Holt and referring to the principle that if a power is exercised by a fiduciary and the fiduciary fails to take into account a relevant consideration or takes into account an irrelevant consideration, the exercise will be liable to be avoided as an abuse of power and breach of fiduciary duty) and the proper purposes doctrine (or “fraud on a power” as it is sometimes referred to) dealing with the concept that a power must be exercised for the purposes for which it is conferred and not a collateral purpose.

At this point I should also clarify two further points.

The Claimant, Avon Cosmetics Limited (“Avon”) is the current Principal Employer under the Plan. It is proposed that a representation order is made under CPR r19.9 so that Avon acts as representative of the Revaluation Winners. The interest of the Revaluation Winners on this first hearing before me is to argue in favour of the validity of the Care Amendments in relation to them. More particularly, in relation to their pre-amendment service, it is to argue that the relevant accrued rights as at the amendment date should be calculated on the basis of their final salary at that date and thereafter revalued on the Revaluation Basis.

Mr Hitchcock KC leading Ms Seymour acts for the Claimant but it is proposed that a representation order be made so that the Claimant represents the interests of the Revaluation Winners.

The first to sixth defendants (as they currently are) are the current trustees of the Plan and are strictly neutral subject to pointing out any practical difficulties of particular rulings by the Court and assisting the Court.

The seventh defendant, Mr Richard Pinnock, is a former member of the Plan and is represented by Mr Bryant KC and Ms Ling. It is proposed that he be appointed to represent the position of the FS Winners. I will deal with his personal position in a little more detail when considering representation orders.

The Plan was established from 1 January 1960 by an Interim Trust Deed dated 23 December 1959.

At the time relevant to these proceedings the power of amendment applicable to the Plan was contained in a Definitive Deed and Rules dated 29 November 1999. Clause 22 of the Definitive Deed provided (so far as relevant) as follows:

As I have said, the Fetter for present purposes is that in clause 22(4)(a)(ii).

I have already explained, in broad terms, the manner in which the Definitive Deed operated as regards relevant accrued rights and how the relevant CARE amendment affected their position.

At the time of the CARE Amendments the Plan had three sections: a “Final Pay” section (a defined benefit section in which pension benefits were calculated by reference to final salary (the “FS Section”)), a “money purchase” section (a defined contribution section) and a “Career Average” section (a defined benefit section in which pension benefits was calculated by reference to pensionable pay in each year of pensionable service, revalued to the date of retirement). This last section was introduced, with effect from 1 April 2003, for new “joiners” by an interim Amending Deed dated 27 March 2003. That Deed also closed the FS Section to new members.

The CARE Amendments were contained in a Deed of Amendment dated 3 October 2006. By clause 3 of that Deed it was provided (so far as relevant for present purposes):

As later summarised by the Watson Wyatt Report of 25 September 2019:

As regards the circumstances in which the CARE Amendments came to be made, they are as follows.

In February 2006, the well-known firm of Watson Wyatt Limited produced a paper looking at the advantages and disadvantages to a number of different possible changes to the Plan. In the same month they also produced an estimate of the cost effect of 5 different changes to the Plan. They then made a presentation to the Avon Board in March 2006. The urgent need for change is explained on page 7 of the slide presentation. Leaving aside urgency, the need for change was to reduce current levels of pension cost to previous (sustainable) levels. The preceding slide referred to significant company contributions (some £13m in 2005) and the increased accounting pension cost between 2004 and 2005 and its projected increase in 2006 to more than double what it had been in 2004 because of lower discount rates and a historic deficit position. As part of the proposals it was recommended that the timetable for changes should allow for the clarification of any legal constraints.

As regards legal constraints on change, the Trustees sought legal advice from Baker & McKenzie on whether four potential options for change to the Plan could be introduced and attached an “Options for the future plan design”. In March 2006 as regards “Option 2” (which is in effect what became the relevant CARE amendment), it was said that a Plan amendment would be needed but that “Mechanically there is nothing to prevent such an amendment being made”. However employment law considerations would have to be taken into account.

Also in March 2006, DLA Piper were asked to advise the Company as regards a number of options including what became the change effected by the relevant CARE amendment. The advice given was that the changes could be made but that an amendment to the governing plan documentation would be needed (as well as regards certain matters, employee consultation to take place).

At a meeting of the Trustees on 26 April 2006, one of the points raised by Baker & McKenzie was the need to ensure that members’ accrued rights were not adversely affected by the proposed changes. The trustees, having discussed the proposals, are noted as having requested assurance that there was no adverse affect to [sic] accrued benefits.

A Watson Wyatt Note of June 2006 regarding Trustees’ responsibilities in relation to the plan changes noted (among other things), under the heading “Are the proposed changes permitted?”

At a Trustees’ meeting on 24 July 2006, Baker & McKenzie’s advice that the proposed changes were acceptable is recorded but various other assurances from Baker & McKenzie were sought.

Ultimately, Baker & McKenzie were instructed to and did draft the Deed which contained the relevant CARE amendment.

There is witness evidence further confirming that positive advice was given that the proposed changes were lawful and that relevant amendments that were made to the Plan could be made.

Mr Ibbotson, a trustee of the Plan at the time, made a witness statement dated 26 July 2023 which also gave evidence about what “hypothetically” might have happened if advice had been received that the option adopted might not be or would not be legally effective. He said:

I turn now to consider the following topics:

I understood the general principles to be agreed.

First, the general law as to construction of documents applies in the field of trust law as much as in, for example, contract law or conveyancing law (see Re Courage Group’s Pension Schemes [1987] 1 ALL ER 528 at537f-g; Barnardo’s v Buckinghamshire & ors [2018] UKSC 55; [2019] Pens. L.R. 4 at [13].)

A short-hand summary of the principles revisited in recent years by a number of House of Lords and Supreme Court cases has helpfully been set out by Carr LJ (as she then was) in Network Rail Infrastructure Ltd v ABC Electrification Ltd [2020] EWCA Civ 1645 (“Network Rail”) at [18] and [19], which I gratefully adopt:

In the field of Pensions law, the general rules of construction are applied taking into account the relevant context and the following factors have to be borne in mind:

As Lord Walker put it in Pitt v Holt, one of the strands of legal doctrine that was plaited together so as to produce the rule which was for long known as the rule in Hastings-Bass, named after the case of that name, was:

As will become clearer when I consider the pension cases in the context of excess of jurisdiction in more detail, the courts have reached for the concept of severance when deciding whether an exercise of a power of amendment under a Pension Plan which is not permitted by the power of amendment may result in only a partial invalidity of the amendment in question. As will be clear from the short summary of the arguments before me that I have adverted to, Mr Bryant and Ms Ling rely upon relevant principles of severance to say that in this case, it is not possible to sever the exercise of the power of amendment so that, as regards Revaluation Winners, the relevant CARE amendment will be valid. It is therefore necessary to consider the applicable principles of severance.

Although, in the particular context of Pension Plan amendments, it is clear that the courts have not applied a full contractual analysis of severance as it applies in, for example, the context of illegal restraints of trade, it is helpful briefly to set out that position so that the discussion in the pension cases makes more sense.

The general position as regards covenants in restraint of trade is helpfully summarised by Chitty on Contracts (35^th Edn, 2023) at paragraph 19-203 (leaving out the footnotes):

As regards severance, in paragraph 19-264, Chitty on Contracts commences a detailed discussion on severance in this context as follows (again omitting footnotes):

After the decision of the Supreme Court in Egon Zehnder Limited v Tillman [2019] UKSC 32; [2020] AC 254, the issue of severance in the contractual context depends upon the application of three criteria: (a) the “blue pencil test”; (b) the “remaining adequate consideration” test and (c) that the contract remains of the same character.

Of these three criteria, Lord Wilson (with whose speech the other Justices agreed) said the following:

Lord Wilson also made clear that two other criteria which earlier cases had identified were not part of the relevant test for severance. The two elements were, first that the covenant should in effect be a combination of different covenants and secondly, that the part proposed to be removed should be no more than trivial or technical. These criteria had been identified in Attwood v Lamont [1920] 3 KB 571, but largely been abandoned in recent years. However, the first requirement had been resurrected as part of the third criterion (identified in paragraph 33 above) by the Court of Appeal in the Egon Zehnder case. Of these two elements, Lord Wilson said:

I turn to the issue of severability in the context of the exercise of powers to legislate which are exceeded. For present purposes I can concentrate on DPP v Hutchinson [1990] 2 AC 783. In that case Lord Bridge (with whom the majority agreed) considered that there were traditionally two tests that applied to the question of severability. First, there was the “blue pencil” test of “textual severability” and secondly there was the test of “substantial severability”. As regards these tests:

Lord Lowry, in the minority, considered that the two tests, of textual severability and substantial severability had both to be satisfied before severability was possible. (Subject to this he agreed with Lord Bridge). Lord Bridge (and the majority) disagreed on there necessarily being two tests, both of which had to be met:

In considering the test for substantial severability, Lord Bridge noted that in some jurisdictions the test had been put in terms of the intentions of the persons exercising the lawmaking power and whether they would have had an intention to make the law in its proposed severed form. Thus in the United States, in the Employers Liability Cases (1908) 207 U.S. 463 the Supreme Court judgment of White J talked in terms of:

However, in Australia the test had been expressed in a slightly different form:

As I read it, Lord Bridge adopted the objective approach to substantial severance. That is, the court should not ask the hypothetical question of what the person exercising the power actually would have done in the situation where they had known that they could not validly exercise the power in the manner that they had done (because of the fetter on them so doing), and in particular whether they would have exercised the power so as to create the proposed severed valid provision. Rather, the court should consider whether or not the proposed severed valid part of the provision would be a substantially different law. It might be said that the objective intention of the exerciser of the power is ascertained in this way, rather than there being some investigation into their subjective intention. As Lord Bridge put it more elegantly:

In the Hutchinson case, the relevant legislator had power to make bye-laws regulating the use of land appropriated for military purpose, including the power to prohibit intrusion on and obstruction of such land. However, the power was subject to the proviso or fetter that the lawmaker could not take away or prejudice rights of common. Bye-laws were made making it a criminal offence to enter or remain in an area of Greenham Common set aside for military purposes without authority. Non-commoners were convicted for breach of the bye-laws. The question was whether the bye-laws could be severed such that they were invalid as regards commoners exercising common rights but valid as regards others. The House of Lords determined that severance was not possible, applying the principles that I have outlined:

It is first necessary to construe any power to decide what jurisdiction is thereby conferred. The principles of construction that I have referred to will apply.

Once the jurisdiction to exercise the power is identified and delimited, the next question is whether or not a particular exercise is or was within the terms or scope of the power, as a matter of jurisdiction or vires. If the person with the power errs by going beyond the scope of the power given to them, the error has commonly been referred to by lawyers as a case of “excessive execution of the power”. The scope of a power can be exceeded where, for example, its terms themselves are exceeded but also where, for example, a rule of law prohibits the or a particular end result flowing from the exercise in question. These are but examples. As Lloyd LJ put it in Pitt v Holt:

In this case it is assumed that the Fetter applies to prevent relevant accrued rights based on final salary being altered to the Revaluation Basis where the existing rights confer greater economic benefit and thus the rights would be prejudiced by the amendment.

Further questions may then arise, such as whether the manner in which the power has come to be exercised, will affect the validity of the exercise. One example, is where the trustee fails to give proper consideration to relevant matters in making a decision which is within the scope of the relevant power. This was labelled by Lord Walker, in Pitt v Holt, as a case of “inadequate deliberation”.

A third relevant category of invalidity may flow from an exercise of a power which is ostensibly within the scope of a power but made for an improper purpose. This is, in trusts law, often referred to, somewhat inappropriately, as a “fraudulent exercise” of a power. In the Company law context of breaches of fiduciary duty by directors, it is usually referred to as a breach of the proper purpose rule (see e.g. Eclairs Group Ltd v JKX Oil & Gas Plc [2015] UKSC 71; [2016] 3 All E.R. 641). In that case, Lord Sumption pointed out at paragraph [15]:

What was left open by the Supreme Court in the Eclairs case was whether or not, in a case of multiple purposes of which only one or some were improper, the search is for the primary or dominant purpose or some other test is to b applied. Lord Sumption (with whom Lord Hodge agreed) advocated a test based on causation:

The majority of the Justices left the issue open as there had not been full argument on the point and the case could be decided on the basis that, on the facts, there was but one improper purpose in mind.

I should also note that the question of the purpose for which a power was exercised may raise both questions of a possible excess of power and also a breach of duty in exercising powers for an improper purpose (see e.g. British Airways plc v Airways Pension Scheme Trustee [2018] EWCA Civ 1533; [2018] Pens. L.R. 19).

I turn to the main authorities to which I was referred.

re Hastings Bass

The simplified facts in re Hastings-Bass [1975] Ch 25 were as follows. In March 1958, in purported exercise of their powers of advancement under s32 of the Trustee Act 1925, trustees of a 1947 trust transferred funds out of that settlement to the trustees of a 1957 settlement, to be held on the trusts of the 1957 settlement.

Subject to a proviso regarding the maximum amount of the sums that could be advanced (calculated by reference to the presumptive or vested share or interest of the relevant person in the trust property to whom or for whose benefit the advancement was being made), s32 of the Trustee Act 1925 permitted trustees to pay or apply capital monies in such manner as they in their absolute discretion thought fit for the advancement or benefit of any person entitled to the capital of the trust property.

On the facts of the case, although the advancement was valid as regards a life interest taken in relation to such funds (under the 1957 Settlement) by a family member of the Hastings-Bass family called William, it was otherwise void (regarding other interests) as infringing the rule of law known as the rule against perpetuities. The Inland Revenue asserted that the effect of the transaction was not to create any new beneficial interests in the fund in question with the result that a large sum in tax fell due upon the death of another Hastings-Bass family member (William’s father).

A number of propositions can be identified from the judgment of Buckley LJ in the Court of Appeal, as explained in Pitt v Holt at both Court of Appeal and Supreme Court level.

A large part of the understanding revolves around what has been identified as a fourth reported proposition of the Inland Revenue in the Hastings-Bass case. That “Fourth Proposition” was that to validly exercise a power of advancement the trustees must weigh the benefits to the advancee under the sub-settlement against the other interests affected and for that purpose must have a proper understanding of the effect of the sub-settlement. If they do not, they have not validly exercised the power at all. Unless the conglomerate benefit to the advancee is found, the weighing operation cannot be carried out. (See generally Pitt v Holt [2012] Ch 132 at paragraphs [50] and [52] for full text of the Fourth Proposition and its further explanation.) In Hastings-Bass, it was submitted by the Inland Revenue, the trustees exercising the power of appointment had not appreciated that the exercise would only be valid as regards the creation of the life interest in favour of William. Hence, they did not have a proper understanding of the exercise of the power and hence the exercise was (wholly) invalid.

As Lloyd LJ pointed out in Pitt v Holt in the Court of Appeal, the fourth proposition of the Inland Revenue was rejected by Buckley LJ in Re Hastings-Bass. Lloyd LJ identified the ratio of Hastings-Bass being as follows (see paragraph [64]):

It is important to note that the reason that the question of intention to benefit entered into the question of the validity of the exercise of the power was because the power, in vires terms, could only be exercised with the intention of benefitting the advancee, as set out in s32 of the Trustee Act 1925. As such, Hastings-Bass was about whether or not a power of advancement had been exercised as a matter of vires or jurisdiction. It was not about whether or not there had been an abuse of power or breach of duty by the trustees in the manner in which they had exercised a power that there was jurisdiction to exercise.

The judgment of Lloyd LJ in the Court of Appeal in Pitt v Holt was confirmed by the judgment of Lord Walker on further appeal ([2013] UKSC 26;[2013] 2 AC 108).

Lord Walker also helpfully identified what Hastings-Bass did not decide. At paragraph [23] he pointed out that Hastings-Bass:

In development of the point that Hastings-Bass did not involve any inquiry into what was actually in the minds of the trustees when exercising the power of advancement, Lord Walker pointed out (see paragraph [24], Pitt v Holt) that, as Lloyd LJ had explained it, the Court of Appeal in Hastings-Bass had decided that case:

In Hastings-Bass itself, Buckley LJ set out a statement of principle described by Lord Walker as follows (see Pitt v Holt paragraph [24]):

As regards this statement (which I shall refer to as the “Buckley LJ formulation”), Lord Walker agreed with Lloyd LJ that limb (1) covered the whole ground and that limb (2) added nothing (see Pitt v Holt paragraph [25]). He added:

Indeed, as Lloyd LJ pointed out in the Court of Appeal in Pitt v Holt, the Hastings-Bass case was later developed by a line of authority to extend well beyond the actual decision in Hastings-Bass. As Lloyd LJ explained the way in which Hastings-Bass was later developed:

Lloyd LJ also considered the “inadequate consideration” breach of trustees’ duties. As regards this he said (see paragraph [131]) that the correct principle was not that developed from Hastings-Bass through Mettoy and later cases, but rather that set out by him in paragraph [127] of his judgment:

Part of this statement of principle repeated what he had already said at paragraphs [124] and [125]. In the latter he said:

These views were confirmed by Lord Walker on appeal (see e.g. paragraphs [43], [80] –[83] and [93]-[94]), who also drew the important distinction between professional advice which will not affect the validity of a decision made in excess of power (where the professional advice may, at the most, be relevant to relief under s61 of the Trustee Act) and that which may prevent there being a breach of fiduciary duty in the context of a claim of breach of fiduciary duty by reason of inadequate deliberation.

Lord Walker also discussed whether a breach of fiduciary duty under the “inadequate deliberation” principle would only come into play if the trustees “would” have made a different decision or whether it was enough that, if properly advised, they “might have” done so (see paragraphs [56], [91] and [92]). He pointed out that Lloyd LJ had not resolved this issue. Lord Walker considered that the Buckley formulation (of “would not” have made a different decision) could not be regarded as clear and definitive guidance. It might be that in practice the court may sometimes think it appropriate to proceed on the basis that for family trusts “would not” would be the appropriate test and “might not” would be the appropriate test for pension trusts:

The next case, chronologically, to which I was referred is that of re Courage Group’s Pension Schemes. I have already referred to the rules of construction iterated by Millett J (as he then was) which were adopted by him in deciding a question as to whether, on the facts in that case, trustees had a discretion or not as to concur in relevant deeds altering the schemes, assuming such amendments could otherwise be made.

He also decided that proposed amendments (regarding substitution of another company as “the company” under the schemes) were ultra vires as manifestly altering the main purpose of the scheme on two slightly different bases. The committee of management was, therefore, not at liberty to execute the proposed amending deeds substituting a different company as “the company”.

He also decided that the proposed deeds could not be executed as they would involve a breach of relevant provisos to the powers of amendment in the relevant trust deeds. Such provisos were in terms of (in the case of two pension funds) not reducing accrued pensions of employed members, except in certain circumstances, and, as regards one pension scheme, not to vary or affect relevant secured benefits without the member’s consent in writing. However, under the proposed amending deeds there was no guarantee of the benefits to be provided under the transferee scheme and the amendments did not carry with them the provisos and restrictions necessary to ensure that the powers conferred could not be exercised in a manner which would reduce the benefits secured by past contributions.

In those circumstances, there was no question of severance or partial validity of the proposed amendments. Hastings-Bass does not seem to have been cited to the Judge according to the law report in the All England Law Reports.

In Doyle v Manchester Evening News Limited [1989] Pens LR 47, Hastings Bass was referred to. The case concerned a representative action by which (in effect) journalists sought to establish first that certain payments made by Manchester Evening News Ltd were part of their salary for the purposes of the relevant group pension scheme and secondly that any amendment of the scheme to exclude those payments from the definition of salary with retrospective effect would be invalid.

The Vice-Chancellor of the County Palatine of Lancaster (Blackett-Ord V-C), first determined that the payments in question were “salary” for the purpose of the relevant pension scheme (see paragraph [63]) and rejected arguments that relevant agreements in this respect were to be relieved in equity on the grounds of mistake, that there should be rectification or that the payment could in effect be excluded from the definition of “salary” in the scheme by a certification process under the scheme (see paragraphs [64]-[69]).

He then went on to consider the question of whether a retrospective amendment of the pension scheme rules could be made to like effect. He noted that there were five heads of attack put forward by the journalists, of which the fifth was that such a change would be precluded by article 15.02(iii) of the relevant trust deed which prevented any amendment which would “affect or prejudice any pension already being provided to a member at the date of such alteration or qualification taking effect”. The relevant amendment would, it was said, affect pensions already being paid. The factual position was that the amendment was to have retrospective effect (the relevant amending resolution to effect the change was passed on 16 November 1983 and was expressed to take effect from 1 April 1983). Retrospective amendments were permitted under the relevant trust deed. Between April and November 1983 two journalists retired who were entitled to a pension based on salary which included the specific payments that the Vice-Chancellor had held were to be treated as “salary” for the purposes of the pension scheme in question.

The Vice-Chancellor noted the submission of the journalists that the rule change was invalid. However, in his judgment the effect of the relevant provisions of the Trust Deed was simply:

As regards the reference to Hasting-Bass, that was made by the Vice-Chancellor earlier in his judgment and not in connection with the question of the jurisdiction to amend and the limit placed upon such power by article 15.02.

The first four heads of attack by the journalists had attacked the exercise of a power of amendment and/or the way in which the trustees had promulgated the same showing that they were (it was said) themselves mistaken as to the position.

As regards this, the Vice-Chancellor considered that, in the absence of fraud, the reasons for the exercise by a trustee or his or her discretion in any particular case cannot be gone into but that if their motives could be analysed, the Buckley Formulation applied (see paragraph [70]). He went on, in paragraph [71]:

He then went on to consider the evidence and what it was the trustees had taken into account.

As regards the first two heads of attack of the journalists, that the amendments were discriminatory and unfair and that they sought to overturn an agreement reached between the journalists and the relevant company, he held that whilst factually this was to some extent true it was not unfair and it did not involve an improper exercise of the trustees’ powers under the deed (see paragraph [79]).

He then turned to the third head of attack, which was that the notice of the resolution gave a misleading impression of the effect of the change and that the trustees were themselves mistaken as to the position. Having considered the notice in the factual context he held that objections based on the wording of the notice failed. Although it was “confused”, no-one was misled and the trustees were well aware of the nature of the journalists’ case (see paragraph 80).

He then turned to the fourth head of attack, that the rule change was considered as a package with other changes rather than separately and rejected that complaint (see paragraph [81]).

Mettoy Pension Trustees Ltd v Evans [1990] 1 WLR 1587, has since been fully explained in Pitt v Holt. In the Court of Appeal, Lloyd LJ dealt with the case at paragraphs [68] to [72]. The question was as to the validity or otherwise of a deed made in 1983 by the company and the relevant pension trustees.

In the Supreme Court, Lord Walker agreed that Hastings-Bass and Mettoy were about two different legal principles:

I have dealt with other points arising from the consideration of this case in Pitt v Holt.

AMP (UK) Plc v Barker

In AMP (UK) plc v Barker [2001] Pens. L.R. 77, Lawrence Collins J (as he then was) had to consider an amendment made to a pension scheme which had introduced substantial new benefits for early leavers where no such effect had been intended. The court granted relief by way of rectification but would also have been prepared to set aside the amendments for mistake and would also have held the amendments void on the basis that the trustees failed to give adequate consideration to the consequence of the amendments and would not have passed the resolution they did if they had.

As regards the inadequate deliberation ground of attack on the alteration, the Judge referred to Hastings-Bass and Mettoy as well as subsequent cases. He considered that the authorities were to “strongly suggest” that the application of the rule in Hastings-Bass led to the act being void rather than voidable (which has since been shown in Pitt v Holt not to be the position in law) but that resolution of the issue did not matter on the facts of the case. He also held that it did not matter whether the test required the person seeking to set aside the trustees’ act to show that the trustees would have acted otherwise or simply might have acted otherwise had they been properly informed. On the facts it did not matter which formulation applied. The resolution would not have been passed if the relevant matters had been taken into account.

In National Grid Co plc v Mayes [2001] UKHL 20; [2001] 1 WLR 864, one issue was whether arrangements made by the principal employer (as defined in a pension scheme) with regard to an actuarial surplus were within the terms of the power (under clause 14(5)) to make arrangements conferred by the relevant pension scheme. The arrangements involved payment of some sums to employees (and members) but also a reduction in the payments into the plans in question by the principal employer. The House of Lords were of the view that although the power under clause 14(5) was on its face in unrestricted terms, it was to be construed in accordance with clause 41(2) which prohibited certain amendments. As construed the House did not consider that the effect of the relevant exercise did fall within the clause 41(2) prohibition or fetter. Mr Newman submitted that in this case the rules as construed together resulted in a limitation in one rule being read into another. He submitted that it is a short step to imply a limitation from the rules into the terms of an amendment to the rules.

I agree that this technique is available but it seems to me that the issue is when is it appropriate to deploy it. In my judgment, and as I go on to explain, it seems to me that it is not simply a question of looking at the wording but also a question of the substantive effect on the amendment.

Bestrustees v Stuart [2001] Pens. L.R. 283, was a case about the validity of attempted pension scheme changes to equalise benefits to provide for sex equality following the European Court of Justice’s decision in Barber v Guardian Royal Exchange [1991]1 QB 340.

In that case, thethen trustee of a pension scheme and the principal employer issued an announcement on 26 April 1994 which purported to change Normal Retirement Date (NRD) under the Scheme to age 65 for men and women. Previously it had been age 65 for men and age 60 for women. In May 1996, new rules were brought in force to give effect to the changes in the announcement. One of the limits on the power of amendment under the scheme trust deed was that no amendment could be made which would affect in any way prejudicially the accrued rights in respect of pension benefits accrued to any prospective beneficiary (the proviso to clause 16, the “Proviso”). I focus on the effect that such a fetter was alleged to have and do not deal with other issues that arose.

The case raised various issues regarding the effect of the proposed alterations as between certain dates. The first relevant one for my purposes is the period between 26 April 1994 to 22 May 1996, which, in the Judgment of Neuberger J (as he then was) is referred to as the “Third Period”. The relevant legal problem was that the alteration was said to take effect from 6 April 1994 whereas, if made, it was only made on 26 April 1994. Neuberger J found that the relevant purported change in this respect was ineffective for a number of other reasons. He went on to say that if otherwise effective, the alteration would have fallen foul of the Proviso. This was because the change for women to 65 years would have affected those who had been in employment earning salary in the period between 6 and 26 April 1994. If the change had not been invalid for other reasons, he would have held that the Proviso did not prevent the change taking effect, but only prospectively from 26 April 1994.

As regards this, he considered the issue of whether the “excessive execution” of the power to amend was, in this case, good in part or bad in part or whether it did not amount to an execution at all. Neuberger J then applied two tests: first, was the valid part of the exercise and the invalid part of the exercise of the power separable, as a matter of concept rather than language, and secondly whether there was anything which leads one to believe that had the trustee been told it could not exercise the power retrospectively, it would not have exercised the power prospectively at all:

He applied the same analysis to the Fourth Period identified in the judgment, that from 23 May 1996 onwards. The relevant rule change purported to be retrospective to 1994 and therefore fell foul of the Proviso, in the same way that the relevant change effected by an announcement as regards the Third Period was also ineffective: “The invalid part of the Rule is excisable, and I am satisfied that it should be excised”.

However, there was also an issue because the rule change as made by a 1996 Deed, did not fully remove discrimination. Under the rule change women were given a penalty-free opportunity for retirement between the ages of 60-65 with the consent of the principal employer only, whereas men seeking to retire between these ages also required the consent of the trustee as well as the principal employer. Neuberger J dealt with this problem by treating the rules as removing the discriminatory provision applying to men in accordance with the principles which he derived from case law as to how equal pay was to be effected.

In this context, he also referred to an argument on behalf of the Second Defendant in the case that a change effected by a 1996 Deed was ineffective so far as it required men (but not women) to obtain the consent of the trustee for retirement between the ages of 60 and 65. The argument was based on Hastings-Bass but apparently as developed in Mettoy. Neuberger J considered that his decision as to how the rules were to be read meant that the Hastings-Bass principle had no operation but, if the 1996 Deed would otherwise be ineffective, he would have reached the same result on the basis of the reasoning in Hastings-Bass (see paragraphs [58]-[60]). In this context Hastings Bass was apparently being called in aid to excise the offending part of the rule change (see discussion after judgment at paragraphs [73]-[74]).

I find the reasoning on this point to be difficult to understand, as it appears to be compressed. As I read it, the argument for the second defendant was not that the trustee in amending the rules had wrongly acted on the assumption that the requirement for men to obtain the consent of the trustee to retire between the ages of 60-65 would be valid (whereas it was not) and had thereby taken into account irrelevant considerations. Rather, it seems to have been an argument that the power had been exercised unlawfully as regards imposing a discriminatory condition that men of the relevant age had to obtain the consent of the trustee as well as the principal employer to retire. That (unlawful) condition could be severed following the Hastings-Bass rule. Neuberger J said that he did not need to resort to Hastings-Bass to sever the unlawful part of the rule change, as he had already achieved the same result through his application of Equal Pay principles. However, if the relevant rule change regarding men was otherwise invalid, he would have severed the change and left the rest of the change as valid (so that the same provision applied to men and women). My interpretation seems to fit in with Neuberger J’s earlier reliance on Hastings Bass as regards the retrospectivity points that had arisen and his reference in paragraph [48] to the fact that he would be dealing with the case in more detail later in the judgment.

The confusion stems from the reference to Mettoy which is now seen as an invocation of the inadequate deliberation principles rather than excess of jurisdiction principles.

Betafence v Veys [2006] Pens. L.R. 137 was another case following changes to a pension scheme which attempted to equalise benefits to provide for sex equality following the decision in the Barber case (supra).

In 1991 (as held by the Judge) the pension scheme had been altered so that between the ages of 60 and 65 men and women were entitled to retire without any reduction of pension benefit. Previously there had been a condition applicable only to male members of the scheme, namely that employer consent was first necessary.

In 1993, amendments were made to the pension scheme rules. Rule 23 allowed for amendments but, unless a Member had consented to the amendment, contained a proviso that “no amendment shall be made so as to affect prejudicially the benefit secured in respect of any Member up to the date of the amendment”.

The 1993 amendment expressly imposed a requirement of employer consent to a retirement between the ages of 60 and 65 if there was to be no deduction from the Member’s entitlement. When executed, the 1993 amendment was intended to do no more than formalise what was understood (wrongly) to have been the position following the 1991 amendment that I have referred to. There was no intention to change the 1991 rules.

The 1993 amendments imposing consent requirements for retirement between the ages of 60 and 65 were invalid by reason of the proviso to the amending power as regards any period prior the date of amendment, which was 17 November 1993, the date of the relevant amending deed. In this context the issue was whether the consent requirements were valid as regards benefits accruing after this date or whether the consent requirements were incapable of severance and so wholly invalid. As regards this, Lightman J said:

The next issue he considered was the fact that the trustees had made a mistake in thinking that in respect of the 1993 amendment they were merely formalising the 1991 amendment, whereas the 1993 amendment in fact substantively altered the position brought about by the 1991 amendment. Having decided that the discretion had been exercised, Lightman J considered that the Hastings Bass rule (regarding inadequate deliberation) applied. Although questioning whether it was correct to apply the same rule to pensions as in private family trusts he felt bound to follow the guidance of Lloyd LJ in Sieff v Fox. Accordingly, (a) if the trustees were free whether or not to exercise their discretion, their exercise could only be challenged if the trustees but for the mistake would have acted differently and (b) if the trustees were bound to exercise their discretion, the exercise could be challenged if the trustees but for the mistake might have acted differently. In the instant case, (a) applied. On the evidence it could not be said that, but for the mistake the trustees would have acted differently. The amendment was therefore upheld.

The facts in IMG Pension Plan HR Trustees Ltd v German [2010] Pens.L.R. 23 were similar to those in this case insofar as the case concerned a conversion of rights under a pension plan from a defined benefit scheme or final salary scheme to a defined contribution or money purchase scheme. The change involved (as here) the conversion of accrued rights from defined benefit to defined contribution. The relevant conversion had purportedly been made from 1 January 1992 by a Deed dated 3 March 1992. The power of amendment under the governing documentation had been changed so as to remove a restriction initially contained within it. For present purposes the relevant restriction was that no amendment should have the effect of reducing the value of the benefit secured by contributions already made. Among other issues in the case were (a) was the effect of the amendment to the power of amendment (removing the previous fetter regarding reduction in value of benefit) valid; (b) if the fetter still applied, did that mean no conversion of defined benefits at the date of conversion could be effected and that the benefits could not be decoupled from any increases linked to final salary?

For present purposes the first relevant issue was whether the amendment to the power of amendment to remove the fetter was in excess of power. Arnold J considered that it was. He accepted the Existing Members’ contention that the original power of amendment could not be construed as permitting an amendment to remove the fetter which was part of the terms of the power, limiting its scope. Otherwise, the fetter could be circumvented by first amending the power to amend by removing the fetter and then exercising the amended power to amend in a manner not permitted by the fetter. The newly stated, revised power of amendment was wholly invalid. Unlike the position in Bestrustees (supra) the amendment to the power of amendment could not be partially valid because it was not possible to distinguish conceptually between the valid part of the exercise of the power and the invalid part. It was not therefore necessary for the Judge (Arnold J as he then was) to consider a further argument that the exercise of the power to amend was invalid as a “fraud on the power”.

Having determined the terms of the applicable power of amendment, Arnold J went on to consider what was the effect of the amendment regarding conversion of accrued rights in terms of (a) the proposed conversion and (b) the date at which any conversion took effect (i.e. was the amendment retrospective in effect). Leaving aside the issue of retrospectivity, he identified four different possibilities:

For present purposes the issue that I have to determine is akin to the distinction between the third and fourth possibilities canvassed by Arnold J, it being assumed for present purposes that the amendment to accrued benefits from FS basis to Revaluation Basis is not valid insofar as that would prejudice members with such rights by reason of the value of such benefits being less.

The broad issue that I have to assume was decided by Arnold J on the basis that the answer to the first two alternatives was in the negative. However, he concluded that the effect of the fetter was to render ineffective amendments which reduced the value of benefits, and in particular future final salary benefits, which had accrued to members by virtue of their service down to the date of amendment. An amendment to convert such benefits from a final salary entitlement to a money purchase entitlement was possible but only subject to an underpin which preserved the future monetary value of the proportion of final pensionable pay which the member had accrued in respect of pre-amendment service. He thus concluded that the third alternative applied and not the fourth (see paragraph [141] and [149]). In reaching this conclusion, he accepted that (as in this case) the value of a member’s benefit could not be determined at the date of the amendment but that this was the inevitable consequence of protecting the value of a benefit, such as a final salary benefit, which is inherently prospective in nature (see paragraph [140]).

This case follows, in time, the decision in the Supreme Court in Pitt v Holt. The first instance decision of Warren J (4 April 2014) is reported at [2014] Pens. L.R. 335; the Court of Appeal decision of 3 August 2017 ([2017] EWCA Civ 1212) is reported at [2018] ICR 1681. The decision again concerns pension schemes. Its key importance for present purposes is that Mr Bryant and Ms Ling, submit that the Court of Appeal lay down a rule that when considering a valid severance of an amendment which has been effected in excess of power, it is necessary for the court to be satisfied that the person exercising that power would have made the same decision to amend had it known and appreciated the fetter on its power of amendment.

The case concerned two pension schemes of UK IBM businesses. These provided either final salary, defined benefit (“DB”) or defined contribution (“DC”) benefits. The DB parts of the schemes had been closed to new members for years but still had about 4,000 employee members, IBM decided to close the DB parts of the schemes altogether, except for some employees who had a contractual entitlement to such benefits. The exercise was known as Project Waltz. It involved three essential elements: (a) excluding members from the DB parts of the schemes as regards future service from April 2011; (b) bringing an end to a long-standing policy of allowing early retirement from age 50, on beneficial terms for employees, as from April 2010 and (c) ensuring that future increases in salary would not count towards the employees’ final pensionable salary for the purposes of DB rights referable to past service up to April 2011. The last point was to be implemented by asking employee who were members of the DB sections to sign non-pensionability agreements (NPAs), expressed to agree that future increase in their pay would not count towards their pensionable salaries for the purposes of the plans. The reason for this was that the termination of DB accrual would mean that future service would not count for the calculation of a member’s DB pension but, without more, the pension attributable to past service would be calculated by reference to a member’s final salary and would therefore be affected by pay rises (the final salary link).

One major area of contention and submission was the scope and effect of any duty of trust and confidence implied into employment contracts and the similar duty of good faith imposed on a party such as the principal employer under a pension scheme which has a non-fiduciary discretionary power under an occupational pension scheme (referred to in the case as the Imperial duty after the decision in Imperial Group Pension Trust Limited v Imperial Tobacco Limited [1991] ICR 524, 531-535). That area I need not consider.

For present purposes, the issues relevant to the proceedings before me were two-fold. First, the scope of a power (conferred on the principal employer) to direct that any specified person or class of person should cease to be a member (referred to for convenience as the “exclusion power”) and whether or not this power was validly introduced by way of amendment into one of the schemes, the Main Scheme, and secondly, whether on the facts the power to exclude members in reliance on such power was exercised for an improper purpose.

In the judgment of Warren J that I have referred to (there were in fact several judgments which ultimately were the subject of or relevant to appeals) he identified Issue 1 before him as being whether or not the relevant Exclusion Power had been validly introduced.

In that context, one sub-issue was: if the Exclusion Power could be exercised in a manner that would infringe one or more fetters on the then power to amend set out in 1983 documents (the “1983 Amendment Power”), was the Exclusion Power void in toto or should it be construed as subject to an overriding limitation to protect the relevant beneficiaries from two identified adverse effects of the relevant amendment (see paragraph [143]).

The relevant power of alteration in the 1983 Trust Deed and Rules allowed the Trustee to make amendments to the trusts powers or provisions of the deed or the rules and with retrospective effect but subject to a number of provisos.

Although there was a suggestion that the Trustee was acting under a fundamental mistake as to the legal effect of the relevant amended Trust Deed and Rules conferring the Exclusion Power, because the Trustee did not appreciate such a significant new power was being introduced, it was accepted that such a mistake would not render the decision void or voidable under the then current law, following the Court of Appeal decision in Pitt v Holt, and no further submissions were made following the Supreme Court decision in that case (see paragraph [153]).

One element of the Exclusion Power that was identified as breaching the fetter identified in the provisos that I have referred to, was a break in a final-salary link between accrued service and future salary increases. The question was whether the introduction of the Exclusion Power was for this reason wholly invalid or valid pro tanto and to be given effect to in relation to service after the date of the amendment whilst preserving the final salary link in relation to service prior to the amendment (see paragraph [158]).

Having decided that the interests and retirement benefits protected under the provisos under (d) and (e) included the final salary link (see paragraph [184]) and that the Exclusion Power prima facie prejudiced those rights, as was conceded (see paragraph [186]), although the power could prima facie be exercised in a manner that would be valid as only affecting future accrual (see paragraphs [189]-[190]), Warren J went on to consider the question of whether the change made was wholly invalid or could be saved in part (see paragraph [192]). The principal employer, arguing for partial validity, relied on Bestrustees, Betafence and HR Trustees Ltd (see paragraphs [199]). Warren J accepted that conclusion (see paragraphs [208]-[209]). In his summary at paragraph 289(iii) he refers to the limitation on the validity of the introduction of the Exclusion Power as being one that is implied:

It is to be noted that in discussing Bestrustees, Warren J referred simply to the conclusion of Neuberger J and not the reasoning, whereas in referring to Betafence, he referred to both the dicta regarding how the amendment was to be “construed” and the “only requirement” being the need for clarity as regards the distinction between what was and was not objectionable and the meaning and application of what was unobjectionable was clear.

I can ignore Issue 2 for present purposes.

Issue 3 was identified as being whether the purported exercise of the Exclusion Power in the manner envisaged in the relevant notices involved the exercise of the Exclusion Power (and/or other relevant powers) for an improper purpose (see heading above paragraph [274] and that paragraph itself).

At the risk of oversimplification, it was submitted that the Exclusion Power (assuming it to have been properly introduced) was actually exercised for an improper purpose, not to bring an end to the membership of the plans but only of certain sections of the plans in order to force members into different sections of the plans (see paragraphs [282], [284]) and to close relevant plans to defined benefit accrual without triggering a winding up (see paragraph [286]). Warren J concluded that no improper purpose was involved (see paragraphs [285], [286] and 289(iv)).

In the Court of Appeal, the Court’s judgment was delivered by Sir Timothy Lloyd. Two elements of alleged improper purpose were relied upon on the cross-appeal. The first arose from the notice exercising the Exclusion Power. It was said that the notice which on the one hand excluded members of the plan but on the other hand also offered the possibility of re-joining a different part of the scheme was a contradiction in terms and shows on its face that the power was sought to be exercised for a purpose for which it was not conferred (see paragraph [165]). The second was the intention to break the final salary link which, it was said, was evinced by the notices and which the Judge had held could not be done by reason of the relevant proviso to the amendment power. Therefore, it was said, the power was sought to be used for a power beyond its ambit.

The Court of Appeal rejected the ground of appeal that there was an alleged improper purpose of excluding members but at the same time inviting them back in. The statement in the notices about the ability to re-join (a different part of) the plan was collateral to the function of the notices, which was to exclude from membership of the plan and the notices were neither self-contradictory nor internally inconsistent. The Judge’s conclusion was upheld (see paragraphs [171]-[172]).

As regards the intention (which could not be given effect to) to break the final salary link, the Court of Appeal characterised this argument as being one of excessive execution of the power rather than improper purpose. This, the Court said, was quite unlike classic cases of improper purpose where the defect lies not in the terms of the execution of the power but in the motive lying behind it. In other words, motive was (and is) irrelevant to excessive execution questions. As the Court pointed out, the Judge had considered the relevant authorities on excessive execution when considering the question of whether the Exclusion Power had been validly introduced into the Main Scheme trust deed and had referred to the authorities including Bestrustees. The Judge had held that the exclusion power had been validly introduced but was subject to the implied limitation that it could not be used to break the final salary link (see paragraph [173] of the Court of Appeal judgment).

The Court of Appeal considered that by the same reasoning, there was no reason why exercise of the power should not be held valid to the extent permitted by the implicit limitation on the powers (see paragraph [174]). The Court went on however to consider the “object” and “principal reason” and the “intention” of the person exercising the power:

Pitt v Holt is not noted in the ICR report as having been referred to. However, as I understood it, all Counsel were agreed that Sir Timothy Lloyd would have been well aware of the case and had it in mind. Mr Hitchcock essentially relied on this factor as meaning that the judgment of the Court of Appeal in IBM should be read consistently with Pitt v Holt. Mr Bryant relied on this factor as being a reason why the IBM judgment is to be read as imposing on lower courts at least a requirement to consider intention and the like. I will come back to this issue later in this judgment.

In Wedgwood Pension Plan Trustee Limited v Salt [2018] EWHC 79 (Ch); [2018] Pens. L.R. 9, an attempt to break a final salary link was also in issue. Notices had been given to the trustee by the participating companies in a pension plan terminating their liability to contribute to the plan and the question was whether these were effective to close the plan to future accrual and break the final salary link with effect from 30 June 2006. The notices were given pursuant to a provision, rule 62, of some 2001 rules which permitted a participating company to stop contribution in respect of one or all employees by notice given to the trustee. However, the 2001 rules had replaced some 1995 rules in their entirety. The power of amendment in the 1995 rules (rule 48) contained a proviso prohibiting any alteration which “shall prejudice or adversely affect any pension or annuity then payable or the rights of any member”. The then equivalent of rule 62 was rule 45 which permitted any participating company (other than the principal company) to retire from the plan if from any cause it found it “impracticable or inexpedient” to continue to participate. The Trustee contended that rule 62 had been validly introduced and the notices were effective to close the plan to future accrual and to break the final salary link. A representative beneficiary argued that the fetter in the power of amendment had been breached and that the fetter protected future as well as accrued rights.

The Judge, Ms Penelope Reed QC sitting as a Deputy Judge of the Chancery Division, decided first the construction of the fetter established by the proviso to rule 45 of the 1995 rules. She determined that the proviso protected accrued rights but not benefits which might be obtained in the future as a result of future service with the employer. The proviso did not protect therefore “rights” to future accruals from further service (see paragraph [44]). The power introduced by rule 62 also, as a matter of construction, engaged the fetter that protected the final salary link for existing members (see paragraph [55]).

The next question was whether the exercise of the power of amendment, falling outside of the scope of the rule 48 power, resulted in the total invalidity of the introduction of rule 62 or whether the amendment should be held to be valid insofar as it did not infringe the fetter on the power of amendment. That would result in a requirement to show that the bar in rule 45 was met (see paragraph [56]).

The argument put forward by the trustee was that the 2^nd limb of the test set out by Neuberger J in the Bestrustees case did not have to be satisfied and that the question was not one of severance but of construction, implying into the exercise of the power a limitation in order to give it effect. It will be recalled that the Neuberger J test had two parts. First, the question had to be asked whether the invalid part of the exercise of the power could be separated conceptually from the valid part of the exercise. Secondly, the question had to be asked whether there was anything in the exercise of the power which leads one to believe that, had the trustee been told it was not entitled to exercise the power in the way it had done invalidly (in that case retrospectively) then it would not have exercised the power as it purported to do prospectively (the valid part of the exercise) or, in the alternative, in the way that it did.

As regards this, the Judge considered that Neuberger J in Bestrustees, was applying a test of excessive exercise first and that he then went on to consider the inadequate reasoning head of attack on trustees’ decisions. If the trustee would have exercised their power in any event (albeit with the excessive part of the exercise ineffective) then the rule in Hastings Bass would not apply (see paragraph [66]).

As I read it, the Judge considered that on the question of severance one looks first at the issue of conceptual separation of the valid from the invalid. However, in the case of trustees (but apparently not powers exercised by non-fiduciaries) she would say that one then goes on to consider the fiduciary duty to consider relevant factors before exercising the power and it is then material to ask whether the trustees would not have exercised the power at all, or would have exercised it differently, if they had been properly instructed as to the limits on the power. As I read it, the Judge was therefore considering first, excessive power and then the fiduciary duty of adequate reasoning. She considered that this was what the Court of Appeal had been doing in the IBM case. The “material to ask” question (whether the trustee would have acted differently) was then elevated into a requirement (see paragraphs [69] and [70] which I set out in full but with emphasis supplied by me):

The Judge therefore first considered the question of conceptual distinction:

She then went on to consider whether the trustees would have acted in the same way had they been aware of the limit on their powers. Here she looked at the evidence which did not focus on that question. Factors were identified which were said to militate against the Principal Company wishing to include such a limited power (see paragraph [75]) but the Judge concluded:

I consider first excessive execution of a power.

Where there is an excessive execution of a power by reason of the exercise, on its face, falling foul of a rule of law or the exercise going beyond the limits or jurisdiction of the power (e.g. because it ignores some limitation on the power) then, at least in the pension context, the court will naturally incline to uphold the validity of the exercise so far as it can. Where what has been done can be conceptually separated into a valid exercise and an invalid exercise and the two are in substance conceptually different then it is possible to save the valid part of the exercise of the power as a matter of construction by implication of a limitation on the terms of the exercise of the power so that it is read as only effecting the valid part of the exercise. This requirement is one that is mentioned, in slightly different terms, in most of the cases that I have considered. My description in this paragraph is an attempt at summarising and referring back to the earlier cases I have mentioned. It is a brief description not some definition.

As a matter of jurisprudential rationale, unless the valid and invalid parts of the exercise of the power can be clearly delineated then no construction to save an “unclear” valid part can be applied. An alternative way of looking at the matter is as a matter of severance of the valid from the invalid so that the valid survives. If that is the test then the cases that consider severance apply a specific severance test in this context which is not a blue pencil test and which is broadly the same as that which applies to what I have called the “construction route” as considered by me in the immediately preceding paragraph above.

The next question however is whether to save the valid part of an excessive exercise of a power, there is a further condition which is that the court must be satisfied that the person exercising the power, had they properly appreciated the true limits on the power, would have exercised it in the same way or at least as regards that part of the actual exercise which would otherwise be separable as valid from the invalid part. I refer to this as the “would still have exercised” test.

Whilst, the Neuberger J judgment in Bestrustees might be said to have been dealing with breach of fiduciary duty by reason of inadequate reasoning, that is not how I read his judgment. Further, it is not how the Court of Appeal considered his judgment or dealt with the “would still have exercised” test in IBM. Neuberger J treated his test as a composite test for severance of which the “would still have exercised test” was the second limb. Further in IBM, the Court of Appeal applied this test to the exercise of the exclusion power (though not its adoption, where they simply accepted Warren J’s reasoning, based solely on conceptual separation and construction). The application of the “would still have exercised” test was fairly and squarely said to be not as regards any breach of fiduciary duty regarding proper purposes but as being part of a consideration of excessive execution or exercise. In short, if Ms Reed QC in Wedgwood was saying that the “would still have exercised test” was being considered in the relevant cases as part of a consideration of the application of the “inadequate reasoning” breach of fiduciary test (or, I would add, if it is suggested that it is only relevant to a breach of the fiduciary duty relating to proper purposes or fraud on a power) then I would disagree.

In my judgment, it is necessary to go back to the cases prior to Hastings-Bass, as well as Hastings-Bass, as explained in Pitt v Holt. In summary:

Turning to the IBM case, the Court of Appeal was considering a breach of fiduciary duty not by reason of inadequate reasoning but in exercising the power in question for improper purposes. The Court, however, considered this aspect of the arguments from the perspective of excessive execution (see paragraph [173]) rather than breach of duty with regard to the proper purposes rule. (In any case, it is unclear that the giver of the notice in that case owed a relevant fiduciary duty in this respect). The court drew the distinction between examining “motive” (on a breach of duty case but not as regards excess of power) on the one hand, and “intention” and “object”. As in Hastings-Bass itself, the focus was what the exerciser of the power “would have done” but there was no discussion by either Warren J or the Court of Appeal of actual evidence of (subjective) intention or state of mind. In my judgment the court was considering the matter from an objective view point and, in effect, asking whether the power, so far as it was otherwise validly exercised, was being objectively exercised to effect the change in question, within the scope of the power or not. Looking at the various surrounding facts and documents, it was able to say that it was:

The way that I would put it, is that the objective ascertained intention was to effect the valid part of the exercise of the power, the intention to effect the invalid part was but incidental. The “would still have exercised” test is a way of showing the invalid exercise was but incidental. Looked at from the perspective of the public law cases on severance, it is akin to the need for the severing court to be satisfied that the part that is severed and survives would not result in a substantially different exercise of the power than the unsevered exercise.

From the perspective of contractual severance, the concept is that:

From the perspective of public law, the test of substantial severance requires the court to be satisfied that severance would be “effecting no change in the substantial purpose and effect of the impugned provision” (see DPP v Hutchinson [1990] 2 AC 783).

The evidence of Mr Ibbotson demonstrates the very problems of asking for factual evidence from a witness, after the event, of their answer to a hypothetical question. That problem is as identified in the cases that I have discussed with regard to statutory severance and is entirely consistent with the dicta in Pitt v Holt that actual intention is not what the court is looking to investigate. Indeed, if it was necessary to examine with witnesses what on the balance of probabilities they would have done, that itself may turn on the need to carry out a huge investigation as to what evidence would then have been available to the decision maker, if asked for (which might also involve what others would have recommended). It seems to me unlikely that a question of vires should turn on such an inquiry. In Mr Ibbotson’s case, it is of course a truism that the Trustees would not have carried out the amendments in the form they were made to the extent that they were if there was a relevant risk that they were invalid. To speculate as to what decision the trustees actually would have reached in that scenario would involve, no doubt, actuarial and business/economy projections not just of the effect of the Underpin but of wholly different options as well as a consideration of what was thought to be acceptable to employees and so on. It does not seem to me right that question of severability of the sort I am dealing with would necessitate a complete re-running of an options exercise carried out some years earlier. The (partial) validity of an amendment in this context should be capable of being determined from the context and the relevant instrument itself. As it has been put in the context of legislative severance, but changing the wording to the Trust context: the purpose of the amendment can only be inferred from the text of the amendments as a whole as applied to the factual situation to which their provisions relate. This assists in identifying a single approach whether it is looked ta through the lens of construction or the lens of severability.

Ultimately, I consider that whether the matter is viewed as one of construction or one of severability the same answer will apply (and at least does apply in this case). Even in the context of textual severability in the legislative field, the issue of construction requires not only the valid to be clearly separate from the invalid as a matter of concept but for the overall purpose of the legislation (as it remains) to be considered:

It seems to me that the concepts of FS Winners and Revaluation Winners and the “underpin” are clearly sufficiently different and identifiable, even if there may be a timing issue as to when categorisation of an individual relevant member into the relevant category is possible. Other cases have followed this course. The Trustees, through Mr Newman, have not identified any difficulty in this respect. Whether viewed as a question of construction or severance, there is no problem in this respect. Mr Bryant submitted that the concepts were not sufficiently different and identifiable but this seemed to me largely an assertion rather than being based upon any concrete factors.

Turning to the question of substantial purpose, looking the CARE Amendments as a whole, their substantial purpose was to remove the final salary link by closing the FS Section to future accruals and to move those with accrued rights to a FS link to the Revaluation Basis. If, as I am to assume, the Trustees were not able, because of limitation on the amendment power, fully to sever the FS link for one category, namely FS Winners, it seems to me objectively that what remains (as regards Revaluation Winners) was precisely within the overall objective intention. What is left, assuming one removes simply the FS Winners from the relevant CARE amendment is, in my judgment, a situation which “effect[s] no change in the substantial purpose and effect of the impugned provision”.

I turn briefly to consider “inadequate reasoning”. I can be brief. On the basis of the evidence it does not seem to me that there is any issue of possible breach of fiduciary duty in relation to inadequate reasoning based on the Trustees having (on the assumed basis) misunderstood the legal effectiveness of the relevant CARE amendment. Quite simply, they considered the relevant issue of jurisdiction and prejudice and acted on proper legal advice. In the light of Pitt v Holt, there is no question of breach. It seems to me that a test of “would they have still exercised the power” simply does not arise.

Finally I turn to improper purpose or fraud on the power. Again, as in IBM, it seems to me that there is no question of any improper motive. The motives were entirely proper. The fact that what was intended was (on the assumed facts) in part legally not possible is a question going to jurisdiction and excessive exercise not to proper purposes or fraud on the power.

Accordingly, on the assumption that the relevant CARE amendment is invalid as regards the accrued rights at the relevant time of FS Winners, it seems to me that it is nevertheless valid as regards Revaluation Winners.

I was asked to make representation orders with regard to the live issue before me in this case, and appointing:

For reasons that I go on to explain I decided to make such representation orders prior to handing down this judgment.

The application was made pursuant to CPR r19.9 which, so far as relevant, provides:

In each case it seems to me that the two classes in respect of whom the representation order should be made, FS Winners and Revaluation Winners each form a class of persons who have the same interest in the claim (at least as regards the issue in question), that the claim is about property subject to a trust and that to appoint a representative for each class would further the overriding objective. I am also satisfied that the requirements of CPR r19.9 are met as regards the applicant for the orders and service of the same.

I had two main concerns or issues that I considered I need to deal with in a little more detail.

The first concern was the late stage at which the application was sought to be dealt with, effectively as part of the substantive consideration of the relevant issue and under the same judgment. As a practical matter, it might be viewed as holding a gun (or perhaps a carrot) to the court’s head. If the court decides that a representation order is not appropriate or that there is no jurisdiction to make it, then a full hearing will have been wasted. This however is simply a practical objection. In response, it may be said that the Court will act correctly and not be (improperly) swayed by the gun (or carrot). I should add that I am told (and the reported cases show) that it is quite common in pension cases at least to make the representation order(s) at the very end of the relevant proceedings or part thereof.

Further in this context, I note that CPR r19.9(7) provides that a judgment or order will, subject to certain conditions, be binding on all persons represented in the claim, but only where the judgment or order is given in a claim in which a party is acting as a representative under the rule. However, had no representation orders been made before judgment was handed down, this judgment would have been given before any person was acting as an appointed representative. There might be a question whether the desired effect of CPR r19.9(7) will apply (a) because only with retrospective effect will the relevant party (assuming a representation order is made) at the relevant time have been acting as a representative appointed under CPR r19.9 and/or (b) the substantive Order giving effect to this judgment on the claim can be made immediately after the making of the representation orders. This was a point that I raised in the draft judgment as circulated prior to formal hand down. In that draft I invited the parties to consider whether they wished to present further argument before I handed down judgment on the substantive issue or whether they were content to address me after hand down with regard to the form of order. I should add that I have not been shown any case raising the issue that I now raise. Ultimately it is for the parties to be satisfied that there is not a problem about late (as I would put it) representation orders being made.

I received a very helpful note from Mr Newman on behalf of all parties. In brief, it was submitted that the position was covered by CPR r19.10(2) and that all the members and beneficiaries of the Plan (apart from the Seventh Defendant, who has been separately joined) are bound by the final judgment, unless and until the representation orders sought in these proceedings are made. Alternatively, it was suggested that I could make the representation orders prior to formally handing down judgment.

I do not consider that it is appropriate for me to make any determination on this point and as regards the further submissions made to me. However, it seemed to me sensible to make the representation orders, as suggested, prior to handing down judgment on a “belt and braces” basis. Accordingly, I made such an order the day before the hand down of this judgment, which can be treated on this point as setting out my reasons for deciding such orders were themselves appropriate.

The second concern that I had was that of “interest”. As regards the FS Winners, there are two sub-issues, first the personal position of the seventh defendant and secondly the relevant interests of the FS Winners. As regards the Revaluation Winners, the question is the position of the claimant.

As regards Mr Pinnock, he is an FS Winner who was a member in service at the date of the amendments. Although he has left the scheme he still has a real interest as a FS Winner. He has been advised and represented by well-known and extremely competent solicitors and counsel. I have no difficulty in making a representation order in relation to him.

As regards the claimant, it might be said that the claimant’s interest is in minimising any costs to the Plan flowing from the CARE Amendments. I am asked to assume that the relevant CARE amendment is invalid as regards the FS Winners. If the relevant CARE amendment is valid as regards Revaluation Winners, then, as I have indicated, the extra costs to the Plan have been estimated to be £5.5 million (the difference between the £12.3 and the £17.8 million figures that I mentioned earlier) or £5.3 million (on the updated figures).

Ultimately, this consideration did not cause me to refuse to make the representation orders sought in relation to the claimant and the Revaluation Winners. First, the figures I have given are only an estimate. They may turn out to be completely wrong in the light of the way in which the economy develops and the way in which final salaries in practice change. Further, and in any event, the application is under CPR r19.9 not 19.8 (where a “same interest” test applies). As the cases make clear (see e.g. Thompson v Fresenius Kabi Ltd [2013] Pens L.R. 158 esp. at paragraphs [14]- [16]; Punter Southall Governance Services Ltd (as Trustee of the Axminister Carpets Group Retirement Benefits Plan) v Hazlett [2022] Pens, L.R. 1)), the test is whether the court is satisfied that the represented class will be properly represented.

Again, the employer is advised by and represented by extremely well known and extremely competent solicitors and counsel. The Trustees (likewise advised) are satisfied that all members and beneficiaries are properly represented and that all arguments that can be put before the court will be advanced by legal teams and, the hearing now being complete, the Trustees have not suggested that their assessment has been disappointed or has changed. Further, I am satisfied that that assessment has proved to be correct. On that basis and in line with the authorities shown to me I was content to make the representation orders as regards the claimant and the Revaluation Winners for the purposes of this hearing.

It follows that I made the representation orders as sought.

I invite the parties to agree a form of order to give effect to this judgment. If the parties can agree such an order then there need be no further hearing regarding this judgment. Otherwise a short remote hearing should be capable of being arranged.