England and Wales High Court (Chancery Division) Decisions

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION

		Royal Courts of Justice Rolls Building, Fetter Lane, London, EC4A 1NL
		21 October 2015

Mrs Justice Asplin :

Introduction

1. This Part 8 claim raises a single issue. It concerns the construction of the rule relating to increases to pensions in payment to which certain members of the occupational pension scheme now known as the FDR Limited Pension Scheme ("the Scheme") are entitled in relation to benefits accrued in respect of service prior to 20 June 1991.
2. The Claimants are the present trustees of the Scheme ("the Trustees") and the Defendant, which is a company incorporated in Delaware, is the present Principal Employer ("the Company"). In fact, the Company is the only employer participating in the Scheme. It is intended that the Trustees represent those members (and persons claiming through them) who would benefit from the issue being determined in a manner which in actual terms provides a higher rate of increase to pensions in payment, and that the Company represent those members (and persons claiming through them) who would benefit from the issue being determined in the manner which has the effect of providing a lower rate of increase. Although it will be necessary to determine the precise terms of the representation orders, in the light of the nature of the issue, I consider it proportionate and appropriate to make the appropriate representation orders pursuant to CPR 19.7(2).

Scheme background and documentation

3. It is not necessary to set out the entirety of the background to the Scheme and its documentation. I will set out only the relevant context. The Scheme was established by The Joint Credit Card Company Limited by an Interim Trust Deed dated 1 May 1972 with effect from that date. The first Definitive Trust Deed and Rules was dated 3 February 1977 (the "1977 Definitive Deed"). It contained a power of amendment which was subject to a proviso. The proviso protected both pensions in payment and members' accrued rights from being affected prejudicially. The proviso contained in Clause 16 of the 1977 Definitive Deed was as follows:

"PROVIDED ALWAYS that no such alteration or addition shall (1) operate so as to affect in any way prejudicially (a) any pension already being paid in accordance with the Rules of this Deed at the date such alteration or addition takes effect or (b) any rights or interests which shall have accrued to each prospective beneficiary in respect of pension benefits secured under the Scheme up to the date on which such alteration or addition takes effect . . . "

(the "Proviso")

4. The Rules of the Scheme contained provision for pensions in payment to be increased. Pensions commencing before 1 May 1974 were required "to be increased at each anniversary by an amount equal to 2.5 per cent" and pensions commencing on or after that date were to be "increased at each anniversary by an amount equal to 3 per cent", subject in both cases to Inland Revenue maxima.
5. It seems that new Rules were adopted in April 1980 (the "1980 Rules"). But for one immaterial exception, they were expressed to apply from 1 November 1977. Rule 16(e) provided that:

"A pension payable under these Rules shall, subject to the provisions of Rule 17 and except as provided below, be increased at each anniversary of the date of its institution by 3 per cent. compound. For this purpose the date of its institution shall be regarded as the date on which a pension became payable to the Member under the Scheme or the date of the Member's death if earlier."

("Old Rule 16(e)")

The reference to exceptions and to Rule 17 are not relevant for these purposes.

6. Thereafter, by a Deed of Amendment dated 8 June 1989 a new Trust Deed and Rules was adopted (the "1989 Definitive Deed and Rules"). The power of amendment at Clause 15 of the 1989 Definitive Deed and Rules remained in the same terms as it had appeared at Clause 16 of the 1977 Definitive Trust Deed. Further, Rule 16(e) of the 1989 Definitive Deed and Rules relating to increases in pensions in payment was in the same terms as Old Rule 16(e).
7. A further Deed of Amendment dated 20 June 1991 (the "1991 Deed") was executed. By Clause 2 it purported to delete Rule 16(e) of the 1989 Rules (which had been in the same form as Old Rule 16(e)) and replace it with the following:

"The amount of a pension payable under these Rules at each anniversary of the date of its institution of that pension falling after 1^st January 1991 shall, subject to Rule 17 and except as provided below, be increased by the lesser of 5% of that amount or such increase in the Government's Index of Retail Prices since the immediately preceding anniversary date. For this purpose the date of its institution shall be regarded as the date on which a pension became payable to the Member under the Scheme or the date of the Member's death if earlier."

(the "Amended Rule 16(e)")

It is the construction of this provision subject to the Proviso which is the issue before the court.

8. In order to complete the chronology, I should add that the Company became the Principal Employer under the Scheme by a Deed of Change of Principal Employer and of Scheme Name and Deed of Variation dated 13 November 1991. Thereafter, a new Definitive Trust Deed and Rules dated 29 March 1994 was adopted the terms of which are not relevant for these purposes. The Scheme was closed to further accrual of benefits with effect from 30 June 2009 and thereafter, the 2014 Definitive Deed and Rules was adopted on 24 September 2014. It was only at the stage at which the drafting of the 2014 documentation was underway that the question of the construction of Amended Rule 16(e) came to light.
9. In fact, since 1991, the Scheme had been administered on the basis that Amended Rule 16(e) was valid in respect of service both before and after 20 June 1991 and as a result, pensions in payment have been increased by the lesser of 5% of that amount or such increase in the Government's Index of Retail Prices since the immediately preceding anniversary ("5% RPI"). It is now accepted that this was incorrect and that as a result of the Proviso, it was not possible to reduce increases in relation to pensions in payment in respect of periods of service before the date on which the 1991 Deed took effect, being 20 June 1991.
10. As at 31 March 2014 the defined benefit section of the Scheme had 2,200 deferred members and 1,588 pensioner members, a large number of whom had accrued benefits in respect of service prior to June 1991. The Scheme Actuary, Mr. Andrew Barnes has calculated that as at May 2014, depending upon how the Amended Rule 16(e) subject to the Proviso should be construed, the potential increase in the liabilities of the Scheme measured on a technical provisions basis under the Pensions Act 2004 is between £5m and £17m. The Scheme is already in substantial deficit.
11. In order to complete the picture I should add that it is not in dispute that prior to the Pensions Act 1995 there was no general statutory obligation to increase a pension once it came into payment although limited statutory increases applied to guaranteed minimum pensions accrued between April 1988 and 1997. However, the 1990 Actuarial Valuation in relation to the Scheme records discretionary increases granted in the years 1986 to 1989 and the intended effect of the Social Security Act 1990 providing for mandatory increases to pensions in payment. In fact, those provisions were not brought into force.

The Issue in more detail

12. As I have already mentioned, it is common ground that the 1991 Deed was valid in so far as it operated prospectively and therefore was effective to alter the rate at which pensions in payment would be increased in respect of pensionable service accrued after 20 June 1991. It is also not in issue that Amended Rule 16(e) breached the Proviso in respect of benefits accrued in relation to pensionable service prior to 20 June 1991 to the extent that it purportedly operated to remove the right to increases in pensions pursuant to Rule 16(e) in the 1989 Definitive Deed and Rules in a way which was prejudicial. As Rule 16(e) in its 1989 form was identical to Old Rule 16(e), I shall refer to it as such.
13. In relation to benefits accrued in respect of pre 20 June 1991 service it is necessary therefore to determine the relationship between the entitlement to increases in pensions in payment of 5% RPI (as a result of Amended Rule 16(e)) and the entitlement to increases at the rate of 3% p.a. compound under the Old Rule 16(e) to the extent that it is preserved by the Proviso.
14. Both the Trustees and the Company describe the present state of affairs which arises as a result of the Proviso as one giving rise to an "underpin" of 3% p.a. compound in relation to benefits accrued in respect of pre-20 June 1991 pensionable service. The issue between the parties is how that "underpin" is to be applied and has been approached as a matter of administration and mathematics. There are three competing approaches which have been put forward: the Annual Approach favoured by the Trustees; the Alternative Approach which is the Trustees' fallback position; and the Modified Cumulative Approach which is favoured by the Company.
15. In summary, the Annual Approach requires the pre 20 June 1991 element of a pension in any given year to be increased by the greater of 3% per annum and 5% RPI on each anniversary of the commencement of the pension. Such an approach results in the greatest increase in Scheme liabilities.
16. Under the Alternative Approach the yearly increase in relation to the pre-20 June 1991 element of a pension is at least equal to 5% RPI. Even if the right to a 3% increase operates as a cumulative underpin, which the Trustees do not concede, it is said that the right to an increase each year of 5% RPI must be preserved. The calculation is carried out by taking the higher of: the value of that element of the member's pension as at the date of retirement increased year on year by 3% p.a. compound, to and including the year in which the increase is to take effect; and the value of that element of the member's pension paid in the year immediately prior to the increase taking effect increased by 5% RPI.
17. The Modified Cumulative Approach which produces the cheapest result in terms of additional Scheme liabilities, and which is favoured by the Company requires the pre 20 June 1991 element of a pension to be paid in any given year to be determined by taking the higher of the value of that element of the member's pension as at the date of retirement increased year on year by 3% p.a. compound to and including the year in which the increase is to take effect; and the value of that element of the member's pension as at the date of retirement increased year on year by 5% RPI compound to and including the year in which the increase is to take effect, subject to a floor of 0% to avoid the effects of any negative retail prices increase. It is not disputed that this approach requires two separate individual member by member annual pension calculations first of the cumulative 5% RPI increases since the inception of the pension and secondly, of the cumulative 3% increases.

The Evidence

18. The factual background in this matter is largely common ground and I need not refer to it in further detail. The expert actuarial evidence is set out in a report of Mr. Ronald S. Bowie, FIA, of Hymans Robertson LLP, dated January 2015 on behalf of the Company and a report by Mr. N.J.H. Salter, FIA, of Barnett Waddingham LLP, dated 20 March 2015 on behalf of the Trustees. The experts also produced a joint statement of May 2015. I should mention that they proceeded on the basis of the Modified Cumulative Approach which was first suggested by Mr. Salter in his report and which was adopted by the Company in preference to the Cumulative Approach shortly before the hearing.
19. Before turning to the detail, I should mention that the experts agreed that whilst the 3% annual underpin in the Annual Approach sense is "relatively common", a cumulative underpin in accordance with Cumulative Approach is "very rare" and in fact, it is "rare" for any increase underpin to apply on a cumulative basis. They also agree that it would be possible to automate a cumulative increase underpin at a cost estimated by Mr. Bowie at £50,000, but that was unlikely to have been the case in 1991, at which time such calculations might have been undertaken by the Scheme Actuary with a consequent increase in ongoing cost. Mr. Salter raised concerns in relation to the potential for inaccuracy prior to automation had the Cumulative Approach been adopted although it was agreed that if increases were calculated by the Scheme Actuary the risk would have been much lower. Not surprisingly, it was also agreed that Member communications would be more complex and more onerous if the Cumulative Approach were applied, because at the same anniversary date different pensioners would be receiving different percentage levels of increase (or, in the case of some, perhaps no increase at all). They accepted however, that it ought to be possible to communicate increases under the Cumulative Approach in a way that members would understand. Mr. Salter also added that in the current market insurers would be reluctant to secure pensions with a cumulative increase underpin and that those that did would be likely to apply a significant premium and this was likely to have been the position in 1991.
20. Both experts produced tables containing worked examples on the basis of a starting position of a notional pension of £1,000 per annum. Mr Bowie compared the Cumulative and the Annual Approach over the period from 2008 to 2014. In that period the annual RPI increase fell below 3% in three years. In fact, in 2009 it was actually -1.4% and in 2012 and 2014 it was 2.6% and 2.3% respectively. His table also reveals that on the Cumulative Approach the actual pension increases granted in the seven years under consideration would be 5%, 1%, 3%, 4.1%, 2.6%, 3.2% and 2.3%. However, if the Annual Approach were adopted the actual increase in any year does not fall below 3%. Over the given years, the value of the pension received based upon the notional £1,000 per annum is £1,232 if the Cumulative Approach is adopted and £1,300 under the Annual Approach. A further sheet containing three tables which Mr Newman QC who represented the Company handed up and to which I shall refer, includes a value of £1,249 for the same period if the Modified Cumulative Approach is adopted. Mr Bowie calculates that if the pension had been increased by 3% per annum it would have amounted to £1,230 in 2014 and if it had been increased purely by 5% RPI it would have amounted to £1,232 which is the same as under the Cumulative Approach. As I understand it, the figures are not in dispute.
21. Mr Salter's table dealt with four years from retirement during which RPI was respectively 4%, 3.5%, 0.5% and 3.5%. On the Annual Approach he calculated that a pension of £1,000 per annum would be worth £1,147 at year 4 whereas using the Cumulative Approach it would be worth £1,126. On the Alternative Approach he calculated that the pension would be worth £1,131.
22. Mr Newman stated that the three tables on his additional sheet were extracted from the experts' reports. One of the additional tables upon which Mr Newman's submissions were concentrated compares the Annual Approach, the Alternative Approach and the Modified Cumulative Approach over the same seven years used for the purposes of the comparative table in Mr Bowie's report. The value of the pensions payable in year 7, 2014, having taken a notional pension of £1,000 per annum is shown to be £1,249 on the Modified Cumulative Approach, £1,262 on the Alternative Approach and £1,300 on the Annual Approach. Given that the same rate of increase is applied, the same figure of £1,230 is reached both as a result of the 3% annual and the 3% cumulative calculations.

The Law

23. Although submissions for the most part were concentrated on the tables, the expert evidence as a whole and the administrative practicality of the three approaches, all of the figures were produced purely by way of illustration and I treat them as such. The real question is how the Amended Rule 16(e) subject to the Proviso is to be construed. There is no dispute about the principles which apply in relation to the construction of provisions of a pension scheme. They are set out succinctly in the judgment of Arden LJ with whom Auld and Waller LJJ agreed, in Stevens & Ors v Bell & Ors [2002] OPLR 207 at paragraphs 26 – 32 as follows:

"The interpretation of pension schemes

26. … There are no special rules of construction but pension schemes have certain characteristics which tend to differentiate them from other analogous instruments. I mention some of those characteristics in the following paragraphs.

27. First, members of a scheme are not volunteers: the benefits which they receive under the scheme are part of the remuneration for their services and this is so whether the scheme is contributory or non-contributory. This means that they are in a different position in some respects from beneficiaries of a private trust. Moreover, the relationship of members to the employer must be seen as running in parallel with their employment relationship. This factor, too, can in appropriate circumstances have an effect on the interpretation of the scheme.

28. Second, a pension scheme should be construed so as to give a reasonable and practical effect to the scheme. The administration of a pension fund is a complex matter and it seems to me that it would be crying for the moon to expect the draftsman to have legislated exhaustively for every eventuality. As Millett J said in Re Courage Group's Pension Schemes [1987] 1 WLR 495 at p.505F-G:

"its provisions should wherever possible be construed to give reasonable and practical effect to the scheme, bearing in mind that it has to be operated against a constantly changing commercial background. It is important to avoid unduly fettering the power to amend the provisions of the scheme, thereby preventing the parties from making those changes which may be required by the exigencies of commercial life."

In other words, it is necessary to test competing permissible constructions of a pension scheme against the consequences they produce in practice. Technicality is to be avoided. If the consequences are impractical or over-restrictive or technical in practice, that is an indication that some other interpretation is the appropriate one. Thus in the National Grid case, to which I refer below, where there was a choice of possible constructions, Lord Hoffmann held that the correct choice depended "upon the language of the scheme and the practical consequences of choosing one construction rather than the other" (see [2001] OPLR 15 at p.28, paragraph 53).

29. Third, in pension schemes, difficulties can arise where different provisions have been amended at different points in time. . . . . Pension schemes are often subject to considerable amendment over time. The general principle is that each new provision should be considered against the circumstances prevailing at the date when it was adopted rather than as at the date of the original trust deed: see per Millett J in Re Courage Group's Pension Schemes, above, at pp.505-506. Likewise, the meaning of a clause in the scheme must be ascertained by examining the deed as it stood at the time the clause was first introduced. …

30. Fourth, as with any other instrument, a provision of a trust deed must be interpreted in the light of the factual situation at the time it was created. This includes the practice and requirements of the Inland Revenue at that time, and may include common practice among practitioners in the field as evidenced by the works of practitioners at that time. It has been submitted to us that the factual background is only relevant if the document is ambiguous. I do not accept this submission, which is inconsistent with the approach laid down by Lord Hoffmann in Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896. In Lord Hoffmann's words "[i]nterpretation is the ascertainment of the meaning which the document would convey to a reasonable person having all the background knowledge that would reasonably have been available to the parties in the situation in which they were at the time of the contract" (p.912H). Lord Hoffmann also distinguished (at p.913B-D) the meaning of the words to be found in dictionaries from the meaning of documents:

"(4) The meaning which a document (or any other utterance) would convey to a reasonable man is not the same thing as the meaning of its words. The meaning of words is a matter of dictionaries and grammars; the meaning of the document is what the parties using those words against the relevant background would reasonably have been understood to mean. The background may not merely enable the reasonable man to choose between the possible meanings of words which are ambiguous but even (as occasionally happens in ordinary life) to conclude that the parties must, for whatever reason, have used the wrong words or syntax: see Mannai Investments Co Ltd v Eagle Star Life Assurance Co Ltd [1997] AC 749."

31. Fifth, at the end of the day, however, the function of the court is to construe the document without any predisposition as to the correct philosophical approach. . . . . .

32. Sixth, a pension scheme should be interpreted as a whole. The meaning of a particular clause should be considered in conjunction with other relevant clauses. To borrow John Donne's famous phrase, no clause "is an Island entire of itself"."

24. Mr Rowley QC on behalf of the Trustees also referred me by analogy to Rainy Sky S.A. v Kookmin Bank [2011] UKSC 50, [2011] 1 WLR 2900 at paragraph 21 per Lord Clarke for the proposition that where there are two possible constructions, the court is entitled to prefer the construction which is consistent with business common sense. The case was concerned with a commercial contract. It seems to me that the principle is equally applicable in the pensions context but that in any event, it adds little to the approach outlined by Arden LJ at paragraph 28 of her judgment in Stevens v Bell, to which I have referred.
25. In the context of this case, I take from the authorities that overall, Amended Rule 16(e) subject to the Proviso should be construed so as to give reasonable and practical effect to the Scheme; that technicality should be avoided and it should be borne in mind that if the consequences of a construction are impractical or technical in practice, it is likely that the interpretation is not the right one. Further, the amendment should be considered against the circumstances and factual situation prevailing in 1991, together with the relevant background which would reasonably have been available. Having taken into account the matters to which I have referred, it is necessary to determine the true construction of the Amended Rule 16(e) subject to the Proviso, being the meaning it would convey to a reasonable person with all of the background knowledge available to the parties. In doing so, it is necessary to consider those provisions in the context of the 1991 Deed as a whole and to approach the matter without any particular predisposition as to the correct philosophical approach.
26. In relation to the nature of the rights which subsist as a result of an exercise of an amendment power in a way which falls within a proviso to that power, I was referred first to Re Courage Group's Pension Scheme [1987] 1 All ER 528, by way of background. It was a case in which Millett J (as he then was) amongst other things was required to construe the proviso to an amendment power which protected "benefits already secured by past contributions." Millett J held that final salary linkage, being the connection between a member's pensionable service down to the date of an amendment and his pensionable salary at the later date of his retirement or leaving service if earlier, was preserved by the proviso in that case.
27. Mr Rowley also referred me to HR Trustees Ltd v German [2010] PLR 23 per Arnold J in which it was held that the conversion of members' past service final salary benefits to a money purchase entitlement was permissible but, by virtue of that scheme's Courage style proviso, subject to an underpin which preserved the future value of the benefits accrued down to the date of the amendment. In fact, the proviso in that case was expressed in terms of preserving the "value" of the benefits secured by contributions already made. Arnold J whilst accepting the wording of the proviso or fetter was different from those in other cases and must be construed on its own terms and in its context stated at [136] that he was not persuaded that the inclusion of the words "the value of" made a fundamental difference to the construction. He concluded at paragraph [141] that the effect of what he described as "the Fetter" was "to render ineffective amendments which reduce the value of benefits, . . . which have accrued to members by virtue of their Service down to the date of the amendment." He concluded therefore, that an amendment was permissible only subject to an "underpin which preserves the future monetary value of the proportion of Final Pensionable Pay which the member has accrued in respect of pre amendment Service."
28. I was also referred to similar passages in IBM UK Holdings Ltd. v. Dalgleish [2014] P.L.R. 335, at paragraphs 185 to 213 and 289-293. Warren J. held that the exercise by the employer of a power to exclude active members from future membership of the scheme, and hence prima facie terminating final salary linkage at the date of exclusion, was subject to an implied limitation such that those members' benefits, when they came into payment, would be the greater of (i) those benefits calculated at the date of his exclusion statutorily revalued under the Pension Schemes Act 1993 and (ii) an underpin based on salary when the member actually retired or left service, but excluding statutory revaluation over that period.
29. In fact, Mr Rowley accepted that the issue both in the HR Trustees case and the IBM case was different from the one with which I am concerned. In both cases, only a one off calculation was necessary in order to preserve the pre- amendment right and to determine the member's entitlement on retirement. By contrast, in this case it is necessary to determine how to give effect to the Amended Rule 16(e) subject to the Proviso in order to determine the nature of the actual right to increases in respect of pre June 1991 service on an annual basis.
30. Mr Newman helpfully took me to Bestrustees v Stuart & Anr [2001] PLR 283. The case was concerned with an attempt to amend a scheme in order to equalise normal retirement ages for men and women as a result of the decision of the European Court of Justice in Barber v Guardian Royal Exchange Assurance Group [1991] 1 QB 340 in circumstances in which there was a fetter or proviso to the amendment power in precisely the terms under consideration here. Neuberger J (as he then was) approached the matter at [48] by severing that part of the amendment rendered invalid by the proviso to the amendment power from that which remained valid and as a result, separating out the periods during which Normal Retirement Date remained different or was equalised either as a result of an effective or ineffective exercise of the amendment power or as a result of the operation of law effected by the Barber decision.
31. Mr Newman also referred me to Betafence Ltd v Veys & Ors [2006] PLR 137, which was also a case in which there was an attempt to equalise benefits following the Barber decision. The amendment power also contained a proviso to the effect that no amendment could be made which affected prejudicially the benefit secured in respect of any member up to the date of the amendment. Lightman J held at [69] that an amendment effected in 1993 had to be construed as "having effect subject to the overriding limitation on the power of amendment contained in the proviso." In that case the question of severance did not arise.
32. Although the cases to which I have been referred are helpful background and of assistance when determining how to approach an amendment which falls foul of a proviso protecting previously accrued benefits and rights, it seems to me that for the most part, they turn on their own facts and do not address a construction issue of the kind with which this case is concerned.

Submissions

33. Mr Rowley on behalf of the Trustees submits that the Annual Approach is consistent with both the language and nature of the right contained in Old Rule 16(e). He submits that the right to have a pension referable to pre June 1991 service "increased at each anniversary of the date of its institution by 3 per cent compound" forms a floor, below which the additional annual increase right conferred by the 1991 Deed cannot drop. He says that removal of the right to 3% increases on each pension anniversary date would breach the terms of the Proviso and that this is the effect of the Modified Cumulative Approach. He emphasises that the right is to a 3% increase on the pension in payment each year. He also points out that Mr Bowie's Table reveals that under the Modified Cumulative Approach, the actual increase granted at any anniversary fluctuates and can fall below 3% which he says is contrary to the Proviso.
34. He also submits that the Annual Approach avoids the complexities of differing rates of increase, member-specific communications and the significantly more detailed calculation exercises described in the experts' reports, all the more so, back in 1991 when the sophisticated computer technology now used by pension scheme administrators did not exist. He also points out that the Annual Approach avoids the difficulties that would arise in the event that the Scheme's benefits, or any part of them, were to be bought out, and the antipathy of insurers to cumulative underpins to which Mr Salter refers. He also says that such an approach is consistent with business common sense.
35. In relation to the Alternative Approach it is said that its virtue is that it is not necessary at each anniversary in relation to each calculation to move from the 3% cumulative calculation to the 5% RPI calculation.
36. Mr Rowley comments that Mr Newman's written submission that what determines the correct underpin is the function which the underpin is designed to serve and accordingly, that the Modified Cumulative Approach is to be preferred, is a use of a mere label and relies upon circular reasoning. It is also premised as Mr Newman states in his written submissions, on the basis that the 1991 Deed made no change to increases in respect of pensionable service prior to the amendment. In fact, in oral submissions, Mr Newman put forward a different and more subtle approach. He contended that the right to increases is now derived from two sources the first being the Proviso which creates an underpin or protects the benefits subject to the Proviso and secondly, the 5% RPI derived from the 1991 Deed and Amended Rule 16(e). He said that it was necessary to keep those two streams or sources separate and not to allow one to contaminate the other. He described the right to increases immediately after the amendment as no less than that payable under the pre 1991 regime and that which was in fact paid under the post 1991 regime.
37. Mr Newman described the Annual Approach and the Alternative Approach as resulting in unjustified windfalls for the Members and pointed out that the Modified Cumulative Approach generated pension figures which were closest in value and never less than a pension increased solely by 3% per annum compound from the date of retirement. This is the comparison between Mr Bowie's illustrative figure of £1,230 in year 7 for 3% increases whether annual or cumulative and the next closest figure in year 7 which is £1,249 on the Modified Cumulative Approach, both the Annual and the Alternative Approach producing greater values. Mr Rowley on the other hand says that use of terms such as "windfall" and "contamination" are merely inflammatory and should be avoided. If, in fact, the provisions of the Scheme should be construed in such a way, no windfall or contamination arises.
38. Mr Newman did not pursue the point made in his skeleton in relation to the Inland Revenue guidance in force at the time of the 1991 Deed and since, concerning the increase of pensions in payment as it preserves the purchasing power of a pension over time. Mr Rowley accepted that IR12 envisages that a cumulative approach is possible. As there were no provisions in force for the increase in pensions in payment in 1991, I do not find the reference to IR12 to be of any real assistance in any event.
39. In addition, Mr Newman says that the expert evidence does not suggest that the complexities caused by the Modified Cumulative Approach are in any way insurmountable and that any added administrative cost will fall at the door of the Company in any event. In this regard, Mr Rowley points out that complexity is an indicator against a possible construction and reminded me that it is said that in 1991 the Modified Cumulative Approach would have required more intervention from the actuary in order to administer the Scheme. He says that although an automative approach may be possible now, it is the position in 1991 which is the benchmark.
40. As I have already mentioned, Mr Newman says that the Modified Cumulative Approach is closer to the intention of the parties in 1991 because he submits that prior to the 1991 Deed increases were at a rate of 3% and thereafter, they were intended to take effect at 5% RPI. The Modified Cumulative Approach requires each to be calculated from the inception of the pension without reference to the other and the better to be applied. In fact, there is no evidence of the intention of the parties in 1991 before the Court, nor as Mr Rowley points out, is there a claim for rectification of the 1991 Deed. The effect of the 1991 Deed is to impose 5% RPI increases prospectively and retrospectively and that may, actually have been the intention. However, what may have been the intention of the parties is not relevant to the task in hand and I approach the matter purely on the basis of the legal test to which I have referred.
41. In any event, Mr Newman submits that it is necessary to compare 3% per annum increases which was the right protected by the Proviso and provides a floor and the 5% RPI increases as a result of the Amended Rule 16(e) without reference to the Proviso. He emphasises that those two sources of the right to increases are separate and concurrent and run in parallel. He says that they should not be mixed and that the rule should not be construed in a way which entitles the member to both streams at once.
42. He says that his additional table illustrates the fact that the Annual Approach does not use the 3% per annum increases preserved by the Proviso as a floor but in fact, mixes the two sources of the right to increases. The calculation under the Annual Approach takes the previous year's pension and multiplies it by the greater of 3% or 5% RPI. As a result the figure at any one anniversary might already comprise an element which has been increased at 5% RPI rather than purely at 3%. This is what he describes as "contamination". He submits therefore that this creates a windfall for the member and that what is being preserved exceeds the original right to annual increases at 3% and therefore, attributes a greater effect to the Proviso than it has. He says that the Alternative Approach suffers from the same flaw but is less egregious because in the worked examples the increase at a rate above 3% occurs a year later than in the Annual Approach calculation. However, he submits that it also based on contamination and confuses two elements of the comparison. What is to be protected is the 3% per annum increases from the inception of the pension in payment in existence before the amendment.
43. Mr Newman submits that Mr Rowley's emphasis upon the increases being "annual" is simplistic. Lastly, he says that any issues in relation to buy out are irrelevant and that the difficulties in relation to years of negative inflation have been dealt with by imposing a 0% ultimate floor to the Cumulative Approach so as to produce the Modified Cumulative Approach.

Conclusions

44. First, it seems to me that the Amended Rule 16(e) subject to the Proviso amounts to a single right to annual increases in pensions in payment but derived from more than one provision or source. The Proviso preserves what is necessary of the Old Rule 16(e) to avoid the member being prejudiced. However, I do not find that conclusion of particular assistance in the construction exercise. In my judgment, the approach adopted by Lightman J in Betafence v Veys is of assistance in this case. Until 20 June 1991, increases in pensions in payment were governed by Old Rule 16(e). It is not in dispute that as a result of the Proviso the right to increases under Old Rule 16(e) could not be taken away in respect of accrued service. As a result of the Proviso, increases at the rate of 5% RPI introduced both prospectively and retrospectively by the 1991 Deed were invalid to the extent that they prejudicially affected the right to increases which already existed in respect of benefits accrued in respect of service to the date of the amendment. To put the matter another way, the 1991 Deed must be construed as having effect subject to the limitation contained in the Proviso. I find it easier therefore, to view the position as a composite or blend of the Amended Rule 16(e) and the Old Rule 16(e) to the extent necessary to prevent prejudice to the member rather than an "underpin". Although I approach the rule as a composite, in my judgment it would make little difference if one were to approach the task of interpretation on the basis that the 3% per annum increases provided in the Old Rule 16(e) operates as an underpin, the 5% RPI as an overlay or that the 5% RPI should apply subject to the limitation imposed as a result of the Proviso. It seems to me that approaching the right created as a result of the Amended Rule 16(e) subject to the Proviso as a blend avoids the danger of falling into a predisposition in favour of one increase rule over the other. The consequence of such an imbalance may lead to the failure to give proper effect to the preserving effect of the Proviso or to exceed it. However, probably not much turns on the difference in terminology.
45. I also derive no assistance from terminology such as "windfall" or "contamination". The right which came into existence after the 1991 Deed is just that. It arises from the Rules themselves. The amendment to introduce 5% RPI increases together with the effect of the Proviso which was at all times a part of the then Definitive Deed and Rules, results in a single right to increases, no more, no less. No question of windfall or contamination arises.
46. It also seems to me that there can be little doubt but that the composite right to increases must be construed in a purposive and practical way which results in a reasonable and practical effect and takes into account the relevant background at the relevant time which in this case is 1991. I also note that complexity may be an indicator that a particular construction is incorrect.
47. Overall, in my judgment, the Amended Rule 16(e) should be construed to mean that that part of a pension which accrued as a result of pre 20 June 1991 service must be increased annually by the greater of 5% RPI and 3%. That is the most natural meaning of the Amended Rule 16(e) subject to the Proviso which prevents the member from being prejudiced should the 5% RPI increase prove less beneficial than the entitlement under the Old Rule 16(e). Both the Old Rule 16(e) and the Amended Rule 16(e) require the pension in payment on the anniversary date to be increased and that is what the Annual Approach does. Such a construction also avoids technicality and would have been straightforward to operate in 1991. It also provides for annual increases in pensions at no less than 3% per annum compound increased at each anniversary of the pension having come into payment which in my judgment is consistent with the effect of the Proviso, given the meaning it would convey to a reasonable person with all of the background knowledge available to the parties in 1991. The fact that such a construction results in a pension which is more valuable than that to which the member was originally entitled is hardly surprising.
48. By contrast, it seems to me that the way in which the Company submits that the Amended Rule 16(e) subject to the Proviso should be construed (the Modified Cumulative Approach) amounts to increases in relation to pensions accrued in respect of pre 20 June 1991 service being an annual increase of such percentage as would lead to the pension being an annual amount equal to the lower of (a) the amount it would have been had each annual increase since its commencement been at the lesser of 5% RPI for each year since commencement of the pension and (b) the amount it would have been had each annual increase since its commencement been at the rate of 3%. In my judgment this is an unnatural construction and not the meaning which would be conveyed to a reasonable person with all the relevant knowledge and background.
49. This is underlined by the emphasis placed by Mr Newman upon the value of the ultimate pension paid and the comparisons set out in the additional table. There is no reference in the Old or the Amended Rule 16(e) to the ultimate value of the pension payable. Nor is there any justification for the added requirement of the further limitation in the form of 3% compound since the date of retirement. Both Old and Amended Rules refer to increases on pensions in payment at each anniversary. They do not refer to a test which includes a reference to a particular value or to any control calculation based on increases from the date of retirement.
50. It seems to me that such a construction is also in part a product of seeking to give more weight or importance to one source of the composite increase rule than the other. I can see no justification for such an approach. It follows that I also reject Mr Newman's submission that increases at 3% per annum must be kept entirely separate from the 5% RPI increases. As I have already said, I consider that the Amended Rule 16(e) subject to the Proviso is a composite and that the attempt to separate out the original sources leads to an attempt to give more weight to one than the other and in fact, to calculate increases from the date of retirement rather than on an annual basis. Furthermore, the fact that on the basis of the Modified Cumulative Approach the actual increase rate can fall below 3% seems to me to be a clear indicator that the construction is not correct. In my judgment, such an effect is clearly contrary to the Proviso which preserves Old Rule 16(e). The effect of the Modified Cumulative Approach would undermine the entitlement to have the pension in payment "increased at each anniversary of the date of its institution by 3 per cent compound".
51. Furthermore, in my judgment, the need to modify the Cumulative Approach in order to avoid the effects of years in which there is negative inflation in order to create the Modified Cumulative Approach is itself another indicator that the resulting construction is not the correct one. As a matter of construction of the Amended Rule 16(e) subject to the Proviso, I can see no basis for the modification. Without it, the Cumulative Approach produces an effect which may in any year produce an actual reduction in the pension paid rather than an increase. Such a result is obviously inconsistent with the rule itself.
52. I am also fortified in my conclusion by the fact that the construction which underpins the Cumulative Approach and the Modified Cumulative Approach leads to complex, technical and relatively impractical consequences. In particular, it would have made the administration of the Scheme particularly complicated in 1991. By comparison the construction which leads to the Annual Approach is simple to implement and would have been so in 1991.
53. It is not necessary therefore, to consider the Alternative Approach. For the sake of completeness, I should add however, that I consider that the Alternative Approach also does not arise from the meaning of Amended Rule 16(e) subject to the limitation of the Proviso which would be conveyed to a reasonable person with all the relevant knowledge and background. In my judgment, the effect is even more convoluted than the Modified Cumulative Approach. It seems to me that it requires one to construe the provision to mean that the member is entitled to an annual increase of 5% RPI for the relevant year or, if greater, such increase as would lead to the pension after such increase being an annual amount equal to the lower of (a) the amount it would have been had each annual increase since its commencement been at the lesser of 5% RPI for the previous year and (b) the amount it would have been had each annual increase since its commencement been at the rate of 3%. In my judgment, this cannot be the meaning it would convey to a reasonable person with all of the background knowledge available to the parties, taken in the context of the 1991 Deed as a whole.
54. To reiterate therefore, in my judgment, the Amended Rule 16(e) subject to the Proviso should be construed in a way which means that increases in pensions in payment which accrued in respect of pre 20 June 1991 service to be applied annually are the greater of 5% LPI and 3%. I should add that I did not hear any submissions in relation to the GMP element of the pensions in payment because they are dealt with separately under the Rules.