IBM United Kingdom Pensions Trust Ltd v IBM United Kingdom Holdings Ltd & Ors [2012] EWHC 2766 (Ch) (12 October 2012)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just Â£1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

England and Wales High Court (Chancery Division) Decisions

You are here: BAILII >> Databases >> England and Wales High Court (Chancery Division) Decisions >> IBM United Kingdom Pensions Trust Ltd v IBM United Kingdom Holdings Ltd & Ors [2012] EWHC 2766 (Ch) (12 October 2012)
URL: http://www.bailii.org/ew/cases/EWHC/Ch/2012/2766.html
Cite as: [2012] EWHC 2766 (Ch)

[New search] [Printable RTF version] [Help]

		Neutral Citation Number: [2012] EWHC 2766 (Ch)
		Case No: HC11C02799

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION

Royal Courts of Justice
Strand, London, WC2A 2LL

12/10/2012

B e f o r e :

MR JUSTICE WARREN
In the matter of THE IBM PENSION PLAN
____________________

Between:

	IBM UNITED KINGDOM PENSIONS TRUST LIMITED	Claimant
	- and -
	(1) IBM UNITED KINGDOM HOLDINGS LIMITED (2) IBM UNITED KINGDOM LIMITED (3) GEORGE METCALFE	Defendants

____________________

Jonathan Gaisman QC, Jonathan Evans and Edward Sawyer (instructed by Nabarro LLP) for the Claimant
Andrew Simmonds QC, Paul Newman QC and Joseph Goldsmith (instructed by Dickinson Dees LLP) for the First and Second Defendants
Nicolas Stallworthy QC (instructed by DLA Piper LLP) for the Third Defendant
Hearing dates: 9th,10th,11th, May 2012, 14th,15th,16th,17th,18th, May 2012, 21st, and 22nd May 2012 and, 30th, 31st May 2012 and 1st June 2012
____________________

HTML VERSION OF JUDGMENT
____________________

Mr Justice Warren :

Introduction

This case concerns a pension scheme known as the IBM Pension Plan ("the Main Plan") which was originally established in 1957. The Main Plan is a large occupational pension scheme with both defined benefit and defined contribution sections. These proceedings relate to part of the defined benefit section, know as "the C Plan". The claimant ("the Trust Company") is the present trustee of the Main Plan. The first defendant is the Principal Employer as defined in the Main Plan. The second defendant is the largest participating employer of the Main Plan. I will refer to them as "Holdings" and "IBM UK" respectively. I will refer to them together as "IBM" although on occasions I will use "IBM" to refer to the IBM group in the UK generally. The third defendant, Mr Metcalfe, has been joined as a member of the C Plan. He supports the Trust Company's case.

The Trust Company seeks an order rectifying the deed of amendment dated 14 December 1983 ("the 1983 Trust Deed and Rules") which created the C Plan and subsequent deeds of amendment. Unusually for a claim of this sort, it is brought by the Trust Company as trustee and not by a member or members of the C Plan. There are perfectly sensible reasons for this, but I do not need to rehearse them in this judgment. Nor do I need to rehearse why Mr Metcalfe is a party to the proceedings given that the Trust Company is prosecuting the rectification claim. His role is, however, restricted as a result of directions which I gave at an earlier interim hearing to protecting the interests of the class to which he belongs by monitoring the conduct of the case by the Trust Company and by lending such assistance as he can.

Representation by Counsel is as follows: the Trust Company is represented by Jonathan Gaisman QC, Jonathan Evans and Edward Sawyer; IBM is represented by Andrew Simmonds QC, Paul Newman QC and Joseph Goldsmith; and Mr Metcalfe is represented by Nicolas Stallworthy QC. Mr Stallworthy responsibly restricted his cross-examination and both his written and oral submissions in the light of the essentially supervisory function of Mr Metcalfe, his main input, I am sure, being by way of discussion with the Trust Company's legal team.

I received (i) on behalf of the Trust Company a long written opening and a very long written closing (I shall refer to them as "the Trust Company's opening" and "the Trust Company's closing") (ii) on behalf of Mr Metcalfe short written opening submissions and (iii) on behalf of Holdings and IBM UK a skeleton argument and a substantial written closing, the latter of which I will refer to as "IBM's closing".

The principal question at issue is whether, when the Trust Company and Holdings created the C Plan as a new section of the Main Plan in 1983, it was intended to be on terms that members of the C Plan would have a right to retire on an unreduced pension at any age between 60 and 63 without any employer consent. That is the sense in which I will use the phrase "flexible retirement" although it is right to point out that one of the principal witnesses, Mr Cawley, has from time to time used it in a different sense as will be seen.

A related question raises a similar question in relation to early leavers, that is to say persons leaving service with a participating employer without becoming entitled to an immediate pension: it is whether it was intended that early leavers should be entitled to draw an unreduced pension at any time after attaining the age of 60.

Those questions might appear to be relatively straightforward to answer. However, they give rise to a number of issues of considerable complexity. Not the least of these is resolution of the factual issues given that the events in question took place in the early 1980s, some 30 years ago. Much of the documentary material no longer survives or, if it does, it has not been found and made available. The recollection of the many witnesses I have heard from is, hardly surprisingly, not in some respects complete; and some persons whose evidence would have been highly relevant are either no longer alive or are not able to give evidence because of health reasons. Given those evidential problems, the claim might be seen as ambitious. But high achievement often requires ambition and it is my task to decide whether that ambition is fulfilled.

I wish to record here, so that I do not need to repeat, that in preparing this judgment I have re-read all of the witness statements in the light of, and with a focus on, the cross-examination which took place. I refer expressly to only a limited amount of the evidence but have taken account of all of it in forming my assessment of the witnesses and in reaching the conclusions which I do.

In relation to the central claim for rectification of the provisions of the C Plan, it is alleged that a mistake was made in the drafting of the 1983 Trust Deed and Rules by which the C Plan was formally introduced into the documentation of the Main Plan. I say formally introduced, because, as we will see, the C Plan was actually in operation some months before the 1983 Trust Deed and Rules were executed. Since 1983, there have been a number of amendments to the Main Plan. It is the Trust Company's case that the error in the 1983 Deed and Rules was repeated in subsequent versions of the rules of the Main Plan and that these, too, should be rectified. The result would be that the current trust deed and rules dated 24 April 1997 ("the Current Trust Deed and Rules") should be amended to reflect what was originally intended in 1983.

The Current Trust Deed and Rules provide, in summary and insofar as relevant to the current proceedings that:

i) for active C Plan members, retirement between ages 60 and 63 is subject to consent, but any pension payable from 60 is unreduced: see rule 2(1)-(3) of the Rules in Schedule D;

ii) for deferred C Plan members, retirement between ages 60 and 63 is subject to consent for deferred members still in Service (but not other deferred members) and any pension payable will be "subject to actuarial reduction": see rules 2(1)-(3) of the Rules at Schedule F.

If rectification is granted in relation to active members but not deferred members, issues arise under certain statutory requirements known as the Preservation Requirements, originally contained in section 63 and Schedule 16 Social Security Act 1973 and continued by the Pension Schemes Act 1993. I will consider these provisions in some detail later.

The Trust Company also advances a contractual claim on behalf of members of the C Plan whose rights under other sections of the Main Plan were converted into C Plan benefits. And IBM asserts a contractual counterclaim so far as concerns new joiners (that is to say members of the C Plan other than those who transferred from other Plans within the Main Plan whom I will refer to as "New Joiners") restricting the benefits which the C Plan would otherwise provide assuming that the claim for rectification succeeds.

There is also a claim pleaded which relates to a different section of the Main Plan, namely the M Plan. In the light of the way in which the evidence developed at the hearing, there is no longer any dispute specific to the M Plan which I need to decide. IBM has made a concession in the light of which I say no more about this claim. I anticipate that the parties will be able to agree the relief which follows from the concession in the light of my other findings in this judgment. The concession is as follows:

If a C Plan member otherwise has a right to an unreduced pension at age 60, IBM does not contend that such a member who elected to transfer to the M Plan in 2006 thereby lost that right. The effect of the concession is that IBM will consent to the early payment of M Plan benefits in cases where C Plan benefits are payable as of right. Accordingly the issues which arose on the pleadings in relation to the M Plan are no longer live.

The Law

I consider it helpful to address the law concerning rectification at this early stage of my judgment in order to establish the focus of factual enquiry. I will also deal with the Preservation Requirements. I will postpone dealing with the extent to which contractual terms are capable of restricting benefits otherwise arising under a pension scheme to a later stage of this judgment.

Rectification generally

The most helpful starting point in considering the law of rectification is to be found in the opinion of Lord Hoffmann in Chartbrook Ltd v Persimmon Homes Ltd [2009] AC 1101. Although his discussion of this topic was strictly obiter, it can be taken as an accurate statement of the law, especially given the concurrence of the other Law Lords hearing the appeal. This was the approach taken by the Court of Appeal in Daventry DC v Daventry District Housing Ltd [2012] 1 WLR 1333 ("Daventry") albeit with reservations being expressed by the Master of the Rolls and Toulson LJ. At [48], Lord Hoffmann said this:

"The requirements for rectification were succinctly summarized by Peter Gibson LJ in Swainland Builders Ltd v Freehold Properties Ltd [2002] 2 EGLR 71, 74, para 33:

'The party seeking rectification must show that: (1) the parties had a common continuing intention, whether or not amounting to an agreement, in respect of a particular matter in the instrument to be rectified; (2) there was an outward expression of accord; (3) the intention continued at the time of the execution of the instrument sought to be rectified; (4) by mistake, the instrument did not reflect that common intention.'"

In Daventry at [80], Etherton LJ (dissenting in the event but with whom Lord Neuberger MR agreed in this respect at [227]) explained Lord Hoffmann's clarification concerning "outward expression of accord" and "common continuing intention" and suggested that Peter Gibson LJ's summary could be re-phrased. He said this:

"80. Lord Hoffmann's clarification was the required "common continuing intention" is not a mere subjective belief but rather what an objective observer would have thought the intention to be: see Chartbrook at [60]. In other words, the requirements of "an outward expression of accord" and "common continuing intention" are not separate conditions, but two sides of the same coin, since an uncommunicated inward intention is irrelevant. I suggest that Gibson LJ's statement of the requirements for rectification for mutual mistake can be rephrased as: (1) the parties had a common continuing intention, whether or not amounting to an agreement, in respect of a particular matter in the instrument to be rectified; (2) which existed at the time of execution of the instrument sought to be rectified; (3) such common continuing intention to be established objectively, that is to say by reference to what an objective observer would have thought the intentions of the parties to be; and (4) by mistake, the instrument did not reflect that common intention."

I accept that as an accurate statement of the law which I should follow. It is consistent with what Mummery LJ had said in Munt v Beasley [2006] EWCA Civ 370 at [36], where he identified the "trend in recent cases to treat the expression "outward expression of accord" as more as an evidential factor rather than a strict legal requirement in all cases of rectification". This is not to be read as indicating that rectification is dependent on a continuing common subjective intention: see Christopher Clarke J in The Aktor [2008] 2 All ER (Comm) 784 at [48].

Since an outward expression of accord and an objectively established common intention are two sides of the same coin, proof of one will almost inevitably result in proof of the other. It is not correct that a burden lies on the party seeking rectification to prove two entirely unrelated facts.

There is no doubt that, in appropriate circumstances, it is possible to rectify a document by which trustees of a pension scheme exercise a power of amendment or consent to the exercise by an employer of a power of amendment: see my own decisions in Drake Insurance plc v MacDonald [2005] EWHC 3287 (Ch) at [33] and ZF Lemforder UK Ltd v Lemforder UK Pension Trustees Ltd [2005] EWHC 2882 (Ch) at point 1 of [41]. In point 2 of [41] of Lemforder, I quoted from [66] of the judgment of Lawrence Collins J (as he then was) in AMP v Barker [2001] PLR 77 ("AMP"). I do not need to set it out again, but draw attention to two points which he made:

i) There needs to be cogent evidence of the intentions of both the trustee and the employer where the power of amendment requires the consent of both. Following Chartbrook, it must now be taken as clear that the intention of each party must be objectively manifested but I would maintain, if the point is relevant, that there does not need to be evidence of an accord between the employer and the trustee. If the evidence shows what each of them, objectively, intends and if they both execute the relevant amending instrument with the same intention, even if not communicated to each other, that is enough. This is not a surprising result. In a case such as Chartbrook or Daventry, what is sought to be rectified is a contract; it makes sense that, in order to displace the contract actually made by rectifying it, there should be found a consensus, albeit not one giving rise to a legally binding agreement. In contrast, in a case such as the present, no sort of agreement is required for there to be a valid deed of amendment. What is needed is an exercise of the power of amendment by the trustee and the consent of the employer to the exercise of that power. If that is to be called a consensus, so be it, but it is a different animal from the agreement or consensus which is relevant in a contractual case.

ii) Where one is considering the intention of a collective body such as a group of trustees or a committee of a board it is their collective intention which is relevant, and it would be a very odd case if that collective intention were not objectively manifested.

In Gallaher Ltd v Gallaher Pensions Ltd [2005] EWHC 42 (Ch) at [117], Etherton J (as he then was) followed AMP and held that an objective manifestation or outward expression of the continuing common intention is not a separate requirement. Nothing in Chartbrook casts any doubt on the points made by Lawrence Collins J or the decision of Etherton J.

It follows from the first of the points which I have drawn from [66] of Lawrence Collins J's judgment in AMP that the search in the present case is not for a consensus between IBM and the Trust Company about what it was that the 1983 Trust Deed and Rules or any of the later versions of the Trust Deed and Rules should provide. Rather, the search has a different emphasis: it is to establish two matters: first, what, objectively, the Trust Company, as the person in which the power of amendment was vested, intended each of the versions of the Trust Deed and Rules should provide; and secondly, what, objectively, Holdings intended to consent to when executing those Deeds and Rules.

It follows from the second of those points that it would be very odd if there were not an objective manifestation of the collective intention of the decision makers within each of Holdings and the Trust Company because, in order to reach a decision, there would have to be communications between the decision makers in each company which would almost of necessity provide the evidence to establish objectively what the intentions of Holdings or the Trust Company were.

There is thus a significant difference of approach in the contractual cases and a case such as the present. In the contractual case, it is necessary to establish a continuing common intention objectively manifested. Where one of the contracting parties is a single individual, there will be no objective manifestation of his intention if it is kept locked in his mind. The intention remains uncommunicated or unadvertised and is not relevant to establishing the necessary accord which is the foundation of the claim for rectification in a contractual case. The same approach applies, of course, where the contracting party is a company or body of trustees. In such a case, it may be possible to establish beyond any doubt what the directors or the trustees intend when they enter into a contract by reference to internal material such as board minutes or communications with lawyers. But those are not relevant or admissible in establishing a common continuing intention or outward expression of accord because they are not communicated to the other contracting party. They are in principle, so far as concerns the counter-party to the contract, no different from the subjective intention locked in the mind of an individual.

In the case of the exercise of a power of amendment by a trustee of a pension scheme and the giving of consent to the amendment by an employer the position is different. The objective evidence, such as board minutes, will be very important in establishing precisely what it was that the relevant decision makers and approvers intended. This evidence may not be evidence of an expression of intention communicated to any third party since the minutes or other evidence may remain entirely internal to the trustee or the employer and not be communicated to each other or indeed to anyone else. But that is not a requirement when it comes to the trustee seeking to prove that the amending document did not in fact reflect the intention of the employer or, indeed, its own intention.

There may, in such a case, be a tension, however, between what the internal documents show and what is in fact communicated to the other party. Take, for example, the present case so far as concerns the 1983 Trust Deed and Rules. Suppose that it were clear (it is not as will become apparent when I address the facts in detail) from the internal documents of Holdings that it did not intend to permit flexible retirement; but suppose also that, innocently, the clear message had been communicated to the Trust Company that flexible retirement would apply; and suppose that the Trust Company had then executed the 1983 Trust Deed and Rules with the intention that flexible retirement would apply. Could the Trust Company then obtain rectification of the 1983 Trust Deed and Rules which, as we know, did not in fact provide for flexible retirement? On one view it should be able to do so because, on the face of it, the requirement for rectification set out by Lord Hoffmann in Chartbrook would be satisfied. Thus

i) the parties had a common continuing intention that flexible retirement would apply;

ii) that intention existed at the time of execution of the 1983 Trust Deed and Rules;

iii) such common continuing intention is established objectively, that is to say by reference to the communication from Holdings to the Trust Company, from which an objective observer would have been able to establish the common intention; and

iv) by mistake, the instrument did not reflect that common intention.

But on another view, and in my judgment the correct view, the intention of Holdings for the purposes of rectification would be that which is established by the internal documentation. On the supposition under consideration, it was not Holdings' intention to provide for flexible retirement; on that supposition, the Trust Company could not obtain rectification since the 1983 Trust Deed and Rules as rectified would not reflect Holdings' intentions. That is not to say that the 1983 Trust Deed and Rules would stand as a valid exercise of the power of amendment in its unrectified state. The Trust Company obviously has a strong case for saying that it never intended to make an amendment in the form of the unrectified 1983 Trust Deed and Rules and that its apparent exercise of the amending power is vitiated. It may have a remedy on that basis: but if it does have a remedy, it is not rectification. I say nothing about a case where the communication passing from an employer to a trustee involves some sort of sharp practice, such as a deliberate intention to mislead, when entirely different considerations will come into play. That is obviously not the present case.

I have so far referred only to the intentions of Holdings and the Trust Company. In fact, it was the Management Committee (as to which see paragraph 94(iv) below) which made all the decisions concerning the C Plan on behalf of Holdings. It may therefore, strictly, be the intentions of the Management Committee which are relevant in establishing the intentions of Holdings. I will consider this aspect further in my discussion of the decision-making process within IBM generally.

I will need to say something more about the law in the context of the claim for rectification of the later versions of the Trust Deed and Rules. It is more convenient to deal with this when dealing with that claim. Similarly, I will deal with the law concerning contractual displacement of the terms of a pension scheme later on.

The Preservation Requirements

There has for many years been provision in the legislation governing occupational pensions schemes designed to protect the benefits of members of a pension scheme who cease to be active members before reaching the age at which they can draw their pension, typically when they leave employment to take a job with an unconnected employer. The requirements (known as "the Preservation Requirements") were originally contained in section 63 and Schedule 16 Social Security Act 1973 and have been repeated, with modification, in successor legislation ever since.

For the purposes of the present litigation, I need to set out some of the provisions of the Pension Schemes Act 1993. Section 70 is a definitions section. The definition of "long service benefit" is relevant. It means, in effect, the benefits which would become payable as a matter of legal obligation to a member on the assumption that he remains in pensionable employment "until he attains normal pension age".

The definition of "normal pension age" was, and remains, to be found in section 180. In the context of the Main Plan, it means "the earliest age at which the member is entitled to receive benefits (other than a guaranteed minimum pension) on his retirement from [employment with IBM]"; but any scheme rule making special provision as to early retirement on grounds of ill health or otherwise is to be disregarded. It is generally accepted and is in my view correct, although this has never been the subject of direct judicial authority, that a rule which allows for early retirement but only with the consent of the employer is such a rule.

Sections 71, 72 and 74, so far as material, provided as follows (as originally enacted):

"71 Basic principle as to short service benefit

(1) A scheme must make such provision that where a member's pensionable service is terminated before normal pension age and—

(a) he has at least 2 years' qualifying service, or

(b) a transfer payment in respect of his rights under a personal pension scheme has been made to the scheme,

he is entitled to benefit consisting of or comprising benefit of any description which would have been payable under the scheme as long service benefit, whether for himself or others, and calculated in accordance with this Chapter.

(2) The benefit to which a member is entitled under subsection (1) is referred to in this Act as "short service benefit".

(3) Subject to subsection (4), short service benefit must be made payable as from normal pension age or, if in the member's case that age is earlier than 60, then from the age of 60……"

"72 No discrimination between short service and long service beneficiaries

(1) A scheme must not contain any rule which results, or can result, in a member being treated less favourably for any purpose relating to short service benefit than he is, or is entitled to be, treated for the corresponding purpose relating to long service benefit……."

"74 Computation of short service benefit

(1) Subject to the provisions of this section, a scheme must provide for short service benefit to be computed on the same basis as long service benefit

……

(7) Where long service benefit is related to a member's earnings at, or in a specified period before, the time when he attains normal pension age, short service benefit must be related, in a corresponding manner, to his earnings at, or in the same period before, the time when his pensionable service is terminated."

To put it shortly, the Preservation Requirements required a scheme to provide a member by way of "short service benefit" a pension calculated on the same basis as his ordinary retirement pension. The short service benefit was thus, broadly speaking, a pension based on the period of service to, and the pensionable remuneration payable at, the date of leaving service. This pension was not payable immediately. Instead, it was deferred pension which only became payable at the member's "normal retirement age" which would commonly have been the age at which the ordinary retirement pension would have become payable.

However, there was one qualification to that which needs to be mentioned. It is that a deferred pension did not have to be paid until age 60 even if there was an entitlement, under the rules of the scheme, to payment at an earlier age. Consider a member of a conventional defined benefits scheme with a normal retirement age of 55 and an accrual rate of 1/60ths. Suppose that he commenced service aged 35 with 20 years' potential service and a maximum potential pension, therefore, of 1/3^rd of his final pensionable salary. He leaves service at age 45 having accrued 10 years of pension. He is entitled by way of short service benefit to a deferred pension of 1/6^th (ie 10/60ths) of his salary at the date of leaving service (that date being the result of section 74(7)). As a result of section 71(3), the deferred pension did not actually have to be paid, however, until age 60.

As I have just mentioned the Preservation Requirements were originally given effect by section 63(1) and (2) and Schedule 16 Social Security Act 1973. Section 63(1) introduced Schedule 16, Part I of which contained the Preservation Requirements; and section 63(2) provided that "this section" was to have effect for securing that occupational pension schemes conform with the Preservation Requirements. Although paragraph 6 Schedule 16 stated that a scheme "must provide" the relevant benefits for an early leaver, the Preservation Requirements were not overriding requirements: it was declared in section 63(3) that nothing in sections 63(1) or (2) or in Schedule 16 was to be taken to apply with direct effect to a scheme, or to the rights or liabilities of any person in, under or by virtue of a scheme.

This does not mean that the Preservation Requirements had no teeth. Section 63(4) provided that it was the "responsibility of the trustees and managers of the scheme……to take such steps as are open to them for bringing the scheme into conformity … with" the Preservation Requirements. And section 63(7) provided that, if the Occupational Pensions Board ("the OPB") determined that a scheme did not conform with the Preservation Requirements, they were empowered to order the trustees or managers to exercise such powers as they possessed for modifying the scheme with a view for bringing it into conformity; and in the absence of such powers to modify the scheme as required by the OPB, the OPB could confer the necessary power on any person or themselves modify the scheme.

The Preservation Requirements were continued, without any modification of substance which is relevant to the present proceedings, up to and including the Pension Schemes Act 1993. In particular, section 71(1) provided that a scheme "must make such provision" as would ensure compliance, section 131 contained a declaration that nothing in Chapter 1 of Part IV (ie the Chapter containing the Preservation Requirements) applied "with direct effect to any scheme or to the rights or liabilities of any persons in, under or by virtue of a scheme", section 132 continued the obligations of the trustees and managers to ensure conformity and section 133 continued the powers of the OPB.

It is worth noting at this point that, although the Preservation Requirements did not have direct effect, a member had certain rights on the footing that they had been complied with. In particular, section 94 Pension Schemes Act 1993 provided, and still provides, that a member should acquire a right to a cash equivalent on the termination of pensionable service, being the cash equivalent of benefits accrued to or in respect of him under "the applicable rules". Those rules include "any provision which the rules of the scheme do not contain but which a scheme must contain if it is to conform with [the Preservation Requirements]".

The Pensions Act 1995 established the Occupational Pensions Regulatory Authority ("OPRA") and abolished the OPB with effect 6 April 1997. The functions and powers which the OPB had in relation to preservation (sections 133-140 Pension Schemes Act 1993) were repealed but without replacement: those functions and powers did not transfer to OPRA.

I note for completeness (although nothing I think turns on it) the submission made on behalf of the Trust Company (based on what was said in the course of Parliamentary debates on what became the Pensions Act 1995) that it appears that the reason why the powers of the OPB in relation to securing compliance with the Preservation Requirements were not transferred to OPRA when the OPB was abolished by the Pensions Act 1995 was, over-optimistically, that since the Preservation Requirements had been in force for so long, all schemes would have already achieved conformity, such that the OPB's powers in this respect were no longer needed.

Significant amendments and changes were made, including changes to the Preservation Requirements, by the Pensions Act 2004.

First, the Act created the Pensions Regulator ("tPR") and dissolved OPRA. Among the many powers given to tPR are the powers to issue an improvement notice under section 13 and to issue a third party notice under section 14. The Trust Company's submission is that these provisions enable tPR to bring about compliance with the Preservation Requirements. I am far from clear that that is right in all cases since these notices can only be given to a person who is in contravention of one or more provisions of the pensions legislation. No obligation is, however, imposed on an employer under the Preservation Regulations to bring a scheme into conformity, so it is not easy to see how the sections are engaged by a failure of an employer to consent to an amendment which the trustees or managers wish to make pursuant to their own statutory duty. This is subject to the separate issue, dealt with later, whether an employer can, in a particular case, be said to be within the meaning of "trustees and managers" of a scheme.

In any event, even if the sections are engaged, the contravention must, in my view, be a contravention of the pensions legislation (which includes Pensions Schemes Act 1993) as it subsists at the time when the notice is given. I do not consider that tPR has power to impose on a person a requirement to bring a scheme into compliance with the Preservation Requirements as they stood at some time in the past.

The position today, therefore, is that the trustees and managers of a scheme which does not conform with the Preservation Requirements must exercise their powers, insofar as they are able, so that it does conform with those Requirements as they currently provide. But there is no power vested in tPR (or any other body) to compel any person who is neither a trustee nor a manager, to exercise powers which he has to bring about conformity, and no power for tPR (or any other body) itself to effect any necessary modifications. Nonetheless, cash equivalents are still calculated on the footing that the applicable scheme rules were compliant with the Preservation Requirements so that there could be a disconnection between the benefits to which a deferred member is entitled and the benefits by reference to which his cash equivalent is to be calculated.

Secondly, the Pensions Act 2004 brought about modifications to the Preservation Requirements themselves. In particular, section 71 was modified by the substitution for sub-section (3) of the following:

"(3) Subject to subsection (4), short service benefit must be made payable as from an age which is no greater than—

(a) the age of 65, or

(b) if in the member's case normal pension age is greater than 65, normal pension age."

Subject to any temporal limitation on the effect of this amendment, it is no longer necessary, therefore, for a scheme to provide for short service benefit to become payable under the age of 65 even where normal pension age is a younger age. It is permissible, of course, for an age under 65 to be adopted.

An amendment was also made by section 263 Pensions Act 2004 to section 72 by the addition of a new sub-section (4) in the following terms:

"(4) This section is subject to subsections (3) and (6) of section 71 (age at which short service benefit is to be payable)."

This subsection makes it clear, in my judgment, that the anti-discrimination provisions of section 72 are not infringed simply by virtue of the fact in a particular case that the scheme provides for a normal retirement age of, say, 60 but provides for short service benefit to be payable only at age 65. It might otherwise be argued that a member would be treated less favourably in relation to short service benefit (payable at age 65) when had he continued to serve he would have been entitled to his full pension at age 60.

Consider another example, that of a member of a conventional defined benefits scheme with a normal retirement age of 60 and an accrual rate of 1/60ths. Suppose that he commences service after 2004 aged 40 with 20 years' potential service and a maximum potential pension, therefore, of 1/3^rd of his final pensionable salary. He leaves service at age 50 having accrued 10 years of pension. He is entitled by way of short service benefit to a deferred pension of 1/6^th (ie 10/60ths) of his salary at the date of leaving service (that date being the result of section 74(7)). As a result of section 71(3) as amended, the deferred pension does not actually have to be paid, however, until age 65. As a result of section 72(4), it cannot be suggested that this result breaches the provisions of sections 72(1) – (3).

What, if any, temporal limitation is to be applied to these changes? Mr Simmonds submits that the meaning of the legislation is clear. The Pensions Act 2004 contains no transitional provisions and none is to be implied. Since the Preservation Requirements apply from time to time and are satisfied or not satisfied as the case may be on a member-by-member basis, there is no reason why the provisions should not be given their natural meaning. On that basis, no distinction is to be drawn between different periods of pension accrual – before and after commencement of the relevant amendment on 6 April 2005.

The Trust Company submits "Not so". In the present case, so far as active members are concerned, the construction of the legislation for which IBM contends would be to deprive them of an accrued right, namely a right in the event of leaving pensionable service before age 63 to an unreduced pension at age 60, a right to which effective implementation of the Preservation Requirements in the Current Trust Deed and Rules when they were made would have entitled them. It is scarcely credible that Parliament would have deprived members of such a right without the clearest language, language which is not found in the Pensions Act 2004.

The result, it might also be said, would not be consistent with the clear protection which was given to accrued rights by section 67 Pensions Act 1995 prior to the Pensions Act 2004. Section 67 has been replaced: after the coming into force of section 262 Pensions Act 2004, we find a substituted section 67 introducing the subsisting rights provisions which provide a more nuanced protection for members in the context of scheme amendments. A "subsisting right" at a given time (see the new section 67A(6) introduced by section 262) includes any right which has accrued to or in respect of the member at that time "under the scheme rules" and, by section 67A(8), scheme rules includes any provision which the rules of the scheme do not contain but which the scheme must contain if it is to conform with the requirements of Chapter 1 of Part 4 of the Pension Schemes Act 1993 (ie the Preservation Requirements). The draftsman, it can be seen, has gone out of his way for the future to protect from adverse amendment benefits which a member would be entitled to under the Preservation Requirements after the coming into force of the relevant provisions of the Pensions Act 2004; in the light of that, it would be very odd, so the argument runs, if he were to have ignored the accrued rights which a member would have if the Preservation Requirements as they stood before that date had been properly implemented.

The conclusion which I am asked to draw by the Trust Company is that the change effected by the Pensions Act 2004 should apply only in respect of service after 6 April 2005. As I understand the submission, the correct procedure in the case of a member leaving service with a deferred pension would be as follows: first, to calculate the total amount of the deferred pension in the usual way by reference to the total period of service and pensionable pay at the date of leaving service; secondly, to time-apportion that amount between the periods before and after 6 April 2005. The member would have the right to draw that part of the deferred pension attributable to the earlier period at normal pension age (assuming normal pension age is 60 or more) with the balance being payable unreduced only at age 65 (or some lower age provided by the rules of the particular scheme in question). Since different commencement dates may not be possible or desirable, the member could either delay taking his pension until age 65, with the part attributable to earlier period being increased, it is said, in accordance with regulations 7-11 of the Preservation Regulations (The Occupational Pension Schemes (Preservation of Benefits) Regulations 1991, SI 1991/167 as amended) or he could take his deferred pension at normal pension age with the part attributable to the later period being actuarially reduced.

That is certainly one solution for which a draftsman expressly addressing the point might have provided. But it is a solution which is not entirely consistent with the idea that accrued rights are to be protected. In different situations, "accrued rights" can mean different things. But in the context of the Pensions Act 1995, and in particular the protection afforded by section 67, accrued rights at a particular time are ascertained as if he had opted out of pensionable service at that time. In other words, what that provision would protect would be a deferred pension based on pay at the date of a proposed amendment falling within section 67. It is not immediately apparent, to me at least that, assuming a temporal limitation is to be found at all, it should have the effect for which the Trust Company contends which would protect the benefit based on salary at the actual date of leaving service rather than protecting simply the level of deferred pension had the member opted out of the scheme on 6 April 2005. It is in any case surprising to me that, if the intention was to provide for a split of this sort with different commencement dates, the drafting did not say so expressly, but there is not even a hint to be found.

It is then said on behalf of the Trust Company that the temporal limitations, if any, of the Pensions Act 2004 so far as concerns the Preservation Requirements are unclear and give rise to ambiguity in the sense which justifies recourse to Hansard in accordance with Pepper v Hart principles. Mr Simmonds says this is not a Pepper v Hart case and that I should, as he puts it in his written closing submissions

"follow the example of Mann J in BT Pension Scheme Trustees Ltd v British Telecommunications PLC and bear firmly in mind the observation of Lord Hoffmann in Robinson v Secretary of State for Northern Ireland in the passage quoted at [92] of the BT case, to the effect that it will be 'very rare indeed' for an Act to be construed as meaning something different from what it would mean to a member of the public who was aware of all the material forming the background to its enactment but who was not privy to what had been said by the minister during debates in Parliament."

What Lord Hoffmann said is perhaps an elegant way of making the point that recourse to Hansard is not to be allowed in order to establish an ambiguity which does not exist where a statute can be construed applying ordinary canons of construction. I find myself able to construe the relevant provisions of the Pensions Act 2004 applying ordinary canons of construction; in doing so, I have not detected the sort of ambiguity which justifies recourse to Hansard.

In any case, appropriate transitional provisions could have been included in the statutory instrument bringing the provisions into force, but this was not done.

As a matter of construction, there is, in my judgment, no temporal limitation in relation to the changes to sections 71 and 72 introduced by the Pensions Act 2004. They apply to the entirety of the deferred pension of a member leaving service before normal retirement age. As I have already said, it would be very surprising if the result for which the Trust Company contends had not been spelt out in the legislation at least to the extent that the changes were to apply only in relation to future accrual.

But quite apart from that, there are other pointers to the conclusion which I have reached. First of all, section 322(5) Pensions Act 2004 provides that the power to make orders under section 322 (commencement orders made by the Secretary of State) includes power to make transitional adaptations or modifications (i) of the provisions brought into effect by the order, or (ii) in connection with those provisions of (among other matters) the Pension Schemes Act 1993 and the Pensions Act 1995. That power is, I consider, wide enough to have allowed the introduction of just the sort of arrangement which the Trust Company submits that the Act should be construed as providing. The interpretation for which the Trust Company contends is, in substance, a transitional provision and could reasonably be expected to be found in secondary legislation.

Secondly, the draftsman might well have thought that there was no need to make any transitional provisions. If one starts from a position where scheme documentation had already been brought into line with the Preservation Requirements as they stood before the Pensions Act 2004, an amendment to a scheme could only be made in accordance with the substituted section 67 and the new section 67A of the Pensions Act 1995. A member's subsisting rights would be protected so that the part of his deferred pension accrued up to the date of the amendment, on the assumption that he had then opted out of the scheme, could not be adversely affected without his informed consent. The draftsman might reasonably have assumed that members would be sufficiently protected in that way.

Thirdly, by treating in some provisions (eg section 67A) the scheme rules as including rules which the scheme does not include but ought to include in accordance with Preservation Requirements, the draftsman clearly recognises that, absent such a deeming provision, the Preservation Requirements would not be directly enforceable and that rule amendment is needed properly to give effect to them. If he had had in mind at all the perhaps unusual situation of a scheme which, on the commencement date of the amendments to the Preservation Requirements, did not comply with the pre-existing Preservation Requirements, he could have been expected to have made express provision for the trustees and managers to ensure that the scheme complied with the pre-existing Preservation Requirements in respect of past service.

In that context, it seems to me that the Trust Company's case makes an unjustifiable assumption which is that, by changing the age condition for the payment of deferred benefits, the Act would be impacting adversely on the rights of members. But it would not be doing so: rather, it would be impacting on the obligation of the trustees and managers to make changes to the scheme to bring it in line with the pre-existing Preservation Requirements. It is in relation to that obligation that the Trust Company must find some temporal limitation if it is to succeed. I find it impossible, however, to see how, on the wording of section 132 Pension Schemes Act 1993 and the change in wording to section 71, the trustees and managers can be under a statutory obligation, after 6 April 2005, to bring the scheme into conformity with the pre-existing Preservation Requirements. At most, it seems to me, they might be guilty of breach of trust in having failed to do so before the new provisions came into force; but there is nothing in that argument in a case where the trustees and managers either did not have the power to do so (because of the need for employer consent) or did not appreciate that the circumstances might obtain where they were under such an obligation (because they had no idea that the scheme was not compliant with the Preservation Requirements, being unaware of a right to rectify the scheme in relation to the long service benefit on which the short service benefit depended).

Accordingly, the effect of the Pensions Act 2004, in my judgment, is that a scheme will be compliant with the Preservation Requirements if it provides for short service benefit to become payable no later than age 65 in cases where normal pension age is 65 or under. It is not necessary for the scheme to provide for short service benefit in respect of pre-6 April 2005 service to become payable at normal pension age (or age 60 if later). This is not to say that, in all cases, the pre-existing Preservation Requirements are irrelevant. If the trustees and managers have failed to bring their scheme into line with the Preservation Requirements at a time when they could and should have done so, they might well be obliged to do so even after 6 April 2005, not as a matter of statutory obligation but as a matter of fiduciary or some other duty. But that is not the present case.

An issue has also arisen in relation to the Preservation Requirements whether Holdings was or is a "manager" of the Main Plan within the meaning of section 67 Social Security Act 1973 and successor Preservation Requirements legislation. If it was, then it was under a statutory obligation to bring the C Plan into conformity with the Preservation Requirements. If rectification of the 1983 Trust Deed and Rules and subsequent versions to provide for flexible retirement is granted, it can be argued that IBM failed to comply with that obligation. I will come back to this issue later.

The Witnesses

I have heard oral evidence from seventeen witnesses on behalf of the Trust Company and read witness statements from a further three, one of whom Mr Simmonds did not wish to cross-examine and two of whom were too ill or unable to attend. All of these witnesses were able to speak of their own knowledge about the events leading to and surrounding the introduction of the C Plan from 1981 to 1983. In addition, there was one email from a Mr Tony Ford, which was the subject matter of a Civil Evidence Act notice. I heard oral evidence from four witnesses on behalf of IBM, two of whom, Stephen Gooch and Helen James, were actuaries with Clay & Partners who provided actuarial advice and services to IBM and to the Trust Company in relation to the Main Plan generally. They had some involvement with the introduction of the C Plan as will become apparent. I also read witness statements from two other witnesses whom it was not necessary for Mr Gaisman to cross-examine.

It is not sensible, or even possible, for me to attempt to review the evidence of each of these witnesses in detail. Some witnesses are, of course, more important than others and will feature to a greater extent in my discussion than others. I will, in any case, be referring to relevant parts of the evidence in reaching my findings of fact. It is right, nonetheless, that I should say something at this stage about some of the witnesses and their respective backgrounds and roles.

I make a general comment about all of the witnesses, subject to the limited caveats which I include below, so that I do not repeat the comment in relation to each of them. It is clear to me that, subject to those caveats, all of the witnesses were doing their best to assist the court. None of them was in any way trimming his or her evidence to suit the case of either side. Each of them was honest and straightforward. All of them suffered, however, from the same disadvantage, namely that the events in question took place over a period between 29 and 31 years ago. It is all too easy for a witness to think that he or she remembers something which is in reality reconstruction; and in the case of some of the more elderly witnesses, there is inevitably a degree of confusion or inconsistency. But none of them can be accused of deliberately making things up.

Sir Leonard Peach ("Sir Leonard") is a central character who gave evidence in support of the Trust Company's claim. As Mr Peach, he had been employed by IBM from 1962 to 1985. He joined as a member of the UK Personnel Department becoming director of personnel in 1971. From 1972 to 1975 he was Group Director of Personnel for EMEA (as to which see paragraph 94(viii) below) on assignment in Paris. He returned to the UK as Director of Personnel and Corporate Affairs, a role he held from 1975 to 1985. From his career path it can be seen not only that he was successful but also that he was in a position to gain an enormous understanding of the culture of IBM globally and how Paris and Armonk (as to which see paragraphs 94(i) and (viii) below) thought and operated. Not only was he in a position to gain such an understanding but, having heard his oral evidence, I am sure he did so.

During the relevant period, he was a director of Holdings and of the Trust Company. He was very much involved in the proposals for the C Plan as they developed from time to time. Sir Leonard is clear in his own mind that flexible retirement was intended by the Management Committee (a committee of the board of Holdings to which wide powers had been delegated: see in more detail at paragraph 94(iv) below), and thus by Holdings itself, and the Trust Company to be a feature of the C Plan and he believes firmly that the 1983 Trust Deed and Rules are in error in failing to reflect that intention. Whether his recollection of events some 30 years ago is now as good as he suggests is not so clear. He was not, unsurprisingly, wholly consistent in all aspects of his answers and displayed an unwillingness to engage in answering some of Mr Simmonds' question concerning the extent to which Holdings had a discretion, consistent with "powers reserved", to alter proposals which had been approved by Armonk.

Nonetheless, the evidence in his two witness statements was not shown to be clearly incorrect in any material respects. And it is right that I should record that flexible retirement of some sort was clearly seen by Sir Leonard at the time as a key selling point of the C Plan to employees. The material which was distributed to employees under his name must therefore have reflected what he at least intended the C Plan should provide. There is, quite properly, no suggestion that he or his fellow members of the Management Committee or the board of Holdings were behaving in any sort of underhand or misleading way.

Michael Cawley is another key witness who gave evidence in support of the Trust Company's claim. He is a former Employee Benefits Specialist. He was employed by IBM UK to work in its Compensation and Benefits Department. He commenced his employment with IBM UK in 1970, fulfilling a number of roles within the Personnel function. In 1981, he was asked by Sir Leonard to take a role in the design and implementation of what became the C Plan. He describes that role as Benefits Policy Manager. He was not a board member of any group company but he says, and I accept, that he was of sufficient seniority within the Personnel Department to have a clear understanding of the decision-making process which I describe at paragraph 94ff below. He was well aware that the policy decision to create the C Plan fell within "powers reserved" which I describe at paragraphs 94(viii) – (xii) below. He understood that senior UK management (that is to say the Management Committee) had a discretion as to how to implement policies decided on or approved by Armonk; but once the design of the C Plan had been decided on by the Management Committee, it was not open to any individuals or more junior managers in the UK to change that design, the more so once that design had been announced to employees in the UK. He therefore understood, and I am clear that he was right to have understood, that he personally did not have any authority to make changes of substance to the terms of the proposed C Plan once they had been, as he put it in his first witness statement, "approved by those higher up the chain of command than me". Mr Cawley's evidence was, with one exception, clear in explaining what he did and did not understand and what he could and could not remember. Whether he is entirely free from criticism in the way in which he dealt with the inclusion in the C Plan of a requirement for employer consent to retirement between ages 60 and 63 at a very late stage is not to the point: the point is the reliability of Mr Cawley as a witness. My assessment of him as a witness is not at all affected by the possible criticism which might be made of what he actually did or did not do in 1983.

The exception which I have just mentioned concerns Mr Cawley's use of the phrase "flexible retirement" in his evidence. In some respects, his explanation of his understanding of the concept was hard to follow and, to my mind at least, he seemed to use this critical phrase in different ways: it is not unkind, I think, to describe his understanding of the concept as itself "flexible".

The Trust Company's closing includes this submission:

"As the person principally responsible for the design of the revised proposal that was put to the UK Management Committee and the Trustee, it is to be expected that Mr Cawley knows what that proposal did and did not contain. His evidence that it did contain a right to an unreduced pension at 60 without the need for company consent should be accepted. He was firm about this central aspect and it is simply not credible that he would have designed the entire employee communication process in the terms he did if the C Plan had been intended to include a consent requirement for retirement between 60-63. His evidence on this key point is fully supported by the contemporaneous documents and was not undermined in cross-examination."

I will need to consider exactly what his evidence actually is later in this judgment. But even if the submission is correct in relation to the communication process which took place before the execution of the 1983 Deed and Rules, it does not follow that a last-minute problem could not have lead to a change in the detailed provisions of the C Plan.

Another important witness is Sam Ellis. He was called on behalf of IBM although whether his evidence, in the end, was more supportive of IBM than the Trust Company is perhaps debatable. The Trust Company's closing does not see the point as debatable at all, rather it sees his evidence as the dagger to the heart of IBM's case.

Mr Ellis joined IBM UK in 1978 as a pensions specialist. He was actively involved in the introduction and implementation of the C Plan in 1983. Importantly, he was responsible within IBM for the drafting, or for supervision of the drafting, of the 1983 Trust Deed and Rules. His evidence about how the consent requirement which is at the heart of these proceedings came to be included is significant although, unfortunately for me, it does not provide the short, sharp resolution of the case which I would like. If his evidence represents a dagger, it is something of a bendy one. He was, nonetheless, a helpful witness and I would not dissent from the description of him in the Trust Company's closing as "a thoroughly reliable witness".

David Barrett was called on behalf of the Trust Company. The odd thing about Mr Barrett's position is the contrast in how he perceived his role in the formulation and approval of the proposals on the one hand and the virtual absence of any mention of him by other witnesses and the total absence of any mention of him in any of the surviving documentation on the other hand. He gave an account of the culture within Armonk which is somewhat at odds with that given by other witnesses, a specific matter to be dealt with later.

Mr Barrett did not provide a witness statement but there is a long letter from him dated 29 November 2009 to Tony Ford in which he set out his recollection of events and his involvement. This letter was subject to a Civil Evidence Act Notice which stated that Mr Barrett was not willing to provide a witness statement. He was, however, called as a witness and appeared entirely willing to give evidence. He stated that the reason why he had not provided a witness statement was because he had never been asked. This is all slightly odd but nothing turns on it.

The information given by Mr Barrett in his letter is as follows. He joined the Legal Department of IBM UK in 1978, working out of its Portsmouth headquarters. He says that he was a member of the Legal Department with responsibility for advising IBM UK (and specifically its personnel division) on the introduction and implementation of the C Plan. He thinks his involvement started in 1981 with an overall review of benefits. By 1982, his main role in the Legal Department had moved to more field-oriented activities by which I understand him to mean that he was no longer directly involved in the C Plan proposals. He recalls two key elements in what he referred to as "our planning" namely "the ability for members to retire at 60 (or at least in one's early 60s) with no diminution in benefit" as of right, and the ability to retire in one's 50s, albeit with reduced benefits, this being discretionary (ie with employer consent). He went on in the letter, however, to describe how the scheme which was developed was approved by UK and Europe but was "in part vetoed by Armonk at least in terms of retirement at 60 without loss of benefit". All that happened was, according to him, that the normal retirement age was moved from 60 to 63. I will explain in due course how that evidence fits into the jigsaw and address what Mr Barrett said in his oral evidence. I will also address what Mr Barrett saw as the cultural reasons for that veto and also his use of the phrase "smoke and mirrors" (as to which see more at paragraphs 211- 212 below).

I would only add at this stage that the letter which Mr Barrett wrote was not a witness statement, it was not prepared with a view to its use in litigation and it was not prepared with a knowledge of the case which IBM was presenting and the points it would make. Two things follow from that. First, the letter was not prepared with the care or attention to detail which would be necessary for a witness statement; I rely, therefore, more on what Mr Barrett said in oral evidence. Secondly, he would not have known of the importance which Armonk's involvement would take in the light of the way IBM developed its case and in particular the relevance of the cultural and other reasons for Armonk's rejection of the initial proposal. I agree with what is said in the Trust Company's closing that this suggests that Mr Barrett was doing his best to give as full and accurate an account of events surrounding the introduction of the C Plan as he was able. I reject the suggestion made in cross-examination that his evidence was crafted in an attempt to provide support for the Trust Company's case.

David Iltyd Caddick, another witness called on behalf of the Trust Company, was a senior manager in IBM UK. He was at material times Personnel Staff Director of IBM UK and a member of the board of the Trust Company. He was able to give first hand evidence about the execution of the 1983 Trust Deed and Rules.

Helen James was a partner in Clay & Partners, a firm of consulting actuaries, from 1977 to 1993, when it became part of what is now AONHewitt. She worked with a more senior partner, Stephen Gooch, who also gave evidence. They together gave advice in relation to the Main Plan including conducting periodic actuarial valuations and advice in connection with the introduction of the C Plan. Mrs James was called to give factual evidence about the events surrounding the creation of C Plan. It was, nonetheless, inevitable that she needed to express some views of an expert nature in order for her factual evidence to be understood. I will need to look at her evidence in some detail later in this judgment in particular in relation to the difference between her evidence and Mr Cawley's evidence concerning what, if anything, Mr Gooch told him about the costing of the improvements to benefits effected by the C Plan as compared with the benefits of the N Plan.

The Trust Company's closing describes her as a thoroughly unreliable witness. It is noted, correctly, that on the first day of her cross-examination (a Friday), she accepted that she had limited recollection of the events of 1982-83; she was unable to recall any specific meetings. When pressed with the proposition that the reason why her witness statement made no mention of a meeting in November 1982 (at which, according to IBM's case, Mr Cawley was told of the need to include in the Rules of the C Plan a requirement for the consent of Holdings to retirement before normal retirement age) was because she had no recollection of any such meeting, her answers were, it is said, evasive. Whether or not evasive, they were certainly confusing; and she seemed to be reconstructing what she thinks would have happened rather than saying what she actually recollected as she was asked to do on several occasions. I will return to all of this in more detail in due course.

Mrs James reflected over the weekend on the evidence which she had given and had made some notes. Before Mr Gaisman recommenced his cross-examination, Mrs James spoke for some minutes to enlarge with clarity on what her position was and why she wanted to say something before cross-examination recommenced. She said initially that her purpose was to correct some things she had said on the Friday but later modified this to say it was to correct some things in her witness statement. She had had, over the weekend, a "surge of recollection". She claimed to remember details of a meeting with Mr Cawley in November 1982 that she had not previously been able to remember and had not covered in her witness statement. She explained that this was because "when I hear something, it takes time, sometimes more than a day, for me to connect it to whatever happened before. So it means I'm a very, very slow thinker". She claimed to have been confused during her evidence on the Friday: "I was confused, or my thinking was confused, on Friday afternoon and it has sort of taken all the weekend and still things change a bit, or I remember more things". She said that she had tried to think hard but her change of evidence was in reality unexplained: "it is something buried deep and I don't know".

One might be forgiven for thinking that she would have thought hard in the production of her witness statement and in preparing for cross-examination and for being surprised that those processes did not produce the "surge of memory" rather earlier than after a period of cross-examination.

The Trust Company's closing describes this evidence as a "most egregious example of recovered memory syndrome (a notoriously unreliable source of truth)". It is said that this was not recollection at all: it was reconstruction assembled over the weekend. And it is said, indeed, that Mrs James effectively accepted that her new evidence was not recollection; over the weekend, she had managed to convince herself that what she thinks must have happened did happen. I do not propose to set out in this judgment the passages from the Transcript of evidence set out in the Trust Company's closing to be found on pages 13, 15 and 21 of the Transcript for Day 9. Her answers are littered with "would have" and "probably". When asked whether her evidence was that she remembered the consent issue being raised by Mr Gooch she replied "No, I feel certain that Stephen Gooch or I would have raised it".

And so it is said that she herself accepted that her evidence was not reliable: "All I can say is sometimes I feel things in this case and I don't know the reason for it, and sometimes they turn out to be right and sometimes they turn out to be wrong". It is also pointed out that she accepted that her apparent recollection of these events might be mistaken. But to be fair to her, any witness speaking to events so long ago would be sensible to accept that possibility, and Mrs James rejected the proposition put by Mr Gaisman that her recollection was in fact as likely as not to have been mistaken.

These are just some of the examples which lead me to treat Mrs James' evidence, particularly where it conflicts with the evidence of any other witness, with the utmost caution. I do not suggest that Mrs James is being other than entirely honest in giving her evidence. But I do think that there is a serious risk that what I have heard is to a considerable extent reconstruction and not recollection although to describe her as "a thoroughly unreliable witness", and to describe the example given as "egregious" goes beyond what is fair or necessary. It is a reconstruction to explain what she and Mr Gooch as competent professionals should have done and therefore what they would have done. But for Mrs James, who at that stage of her career at least was something of an expert technician, I perceive a real risk that she is confusing what may have been discussed internally with Mr Gooch with what was discussed with Mr Cawley.

Stephen Gooch was a partner of Clay & Partners from 1974 to 1993. As well as working with Mrs James on giving advice in relation to the Main Plan including conducting periodic actuarial valuations and advice in connection with the introduction of the C Plan, he subsequently became the Scheme Actuary in succession to his partner Alan Fishman. Like Mrs James, he was called to give factual evidence about the events surrounding the creation of C Plan. And, as with her, it was, nonetheless, inevitable that he needed to express some views of an expert nature in order for his factual evidence to be understood.

The Trust Company's closing gives Mr Gooch a very poor report. He was, it is said, "an unreliable and truculent witness, whose stance was motivated by self-justification and an animus towards the Claimant and/or its solicitors for having conducted the 14.04.10 and 09.06.10 interviews and then not showing him the notes". To explain that submission, I need to explain what is being referred to by the interviews. Mr Gooch's witness statement is dated 12 March 2012. He had, however, attended an interview in April 2010 with representatives of the Trust Company. These were David Newman, the Pensions Trust Manager and Secretary to the Trust Company, two partners of Nabarro Nathanson ("Nabarro", the Trust Company's solicitors) namely Neal Gibson and Jennifer Bell and an assistant, Alison Birkitt who was responsible for taking the main note of the interview. There was a further telephone interview in June 2010 with Mr Gibson also attended by Ms Birkitt, again responsible for taking a note. The notes of these two interviews were finalised after discussions between the representatives of the Trust Company involved in each case. But in neither case were the notes shared with Mr Gooch who therefore had no opportunity to comment on or correct them. It is said on behalf of the Trust Company that the contents of the notes are in important respects inconsistent with the contents of Mr Gooch's witness statement. I will say more about this later.

Mr Gooch's witness statement mentions that the consent issue was raised with Mr Cawley "in discussions I and Helen James had with Mr Cawley". He gave no details of precisely what was said, how it came about that the issue arose or when and where any relevant discussion took place. It is reasonable to assume that Mr Gooch had given careful consideration to his evidence in preparing that witness statement and that he would have included everything relevant which he could remember. In answer to questions asked by Mr Simmonds in the course of a short examination-in-chief which I allowed, he said that he could not recall the dates and that he could not recall the contents of any discussion beyond what was said in his witness statement. He then immediately corrected himself. It is worth setting out his answer in full:

"Well, as you can see, this witness statement was prepared two and a half months ago, and at that point in time it reflected the totality of what I could recall with sufficient clarity to put my name to and to state on oath.

During the intervening period up to now, a number of things have cropped up, a number of documents have been shown to me, which meant that I have spent a great deal of time over the last two or three months racking my brains as to whether there were any other points I could remember, again, with sufficient clarity to state them on oath.

One has occurred to me. It occurred, I suppose you could say, out of the blue during this racking of the brains, but it came to me with complete clarity and it relates to a meeting with Mike Cawley. It was after we had been given some documentation that had been sent to members, which we hadn't seen up until then, and the documentation and what concerned us did not refer to consent of the employer to early retirement. And that is the basis obviously for my statement on which we did our costings.

During the meeting with Mike Cawley, I said to him - and we had a very brief dialogue which went along the lines of, "Mike, you have not included in the communication material the fact that early retirement is with company consent". And he said, "Okay, we will include it next time"."

Mr Simmonds had not put this alleged conversation to Mr Cawley: I assume that he did not know what Mr Gooch was going to say when he cross-examined Mr Cawley. But it is impossible to think that Mr Cawley would have agreed with this evidence in the light of what he actually was asked and responded to. Mr Gooch was, in any case, unable to place any sort of accurate date on this discussion which, according to him, lasted no longer than a minute. Asked in cross-examination whether his recollection of a conversation which came to him out of the blue after 30 years might be mistaken he replied "No". When asked why be said "Because it was a very, very important point and that is why I recalled it" and although the point had been put to bed, as he put it, in the light of the later employee handbook he remembered it "because at the time it was very significant". Again, I must treat with care the proposition that a very, very important point which was very significant at the time could be so clearly recollected when it had not been remembered at the time when he prepared his witness statement dealing, among other matters, with the specific issue of consent.

Before leaving this general survey of some of the witnesses I comment that what is unfortunate from my perspective as a judge having to make findings of fact is the total absence of evidence from anyone to support any positive case presented by IBM in relation to Armonk and about what was or was not approved. I do not know whether this is because the relevant people are no longer living or are too ill to give evidence or make witness statements or whether it is because they cannot remember or because what they can remember is not helpful to IBM. Quite apart from having no evidence about how Armonk dealt with the question of approval for the C Plan, I do not even have evidence from anyone connected with Armonk at the time, or indeed at any time, who could speak to the cultural reasons, if any, which might have led to a rejection of the proposals. I must therefore rely on the evidence, in particular, of Sir Leonard and Mr Barrett to establish that culture if and to the extent that it is of any relevance.

The IBM Group and decision making

At this stage, I turn to describe the decision making process within IBM in the UK. Sir Leonard has given evidence about the structure of the IBM group and the decision-making hierarchy at the relevant time in 1981 to 1983. I do not think that there is any dispute, save where otherwise identified, about the following:

Group structure

i) IBM was and is a global computer technology and IT consulting group. The top company of the IBM group was International Business Machines Corporation ("IBM Corporation") which was listed on the New York Stock Exchange. It had its headquarters at Armonk in White Plains, New York State. Headquarters global management were variously referred to as "IBM Corporation", "Armonk" or "White Plains". "Armonk" has been most frequently used in the hearing before me and that is how I will describe the global management there.

ii) Holdings, as its name suggests, was the holding company of the UK IBM group. Holdings held certain group responsibilities including acting as Principal Employer in relation to the Main Plan. Holdings itself was an indirect subsidiary of IBM Corporation.

iii) IBM UK was the main operating company and a wholly-owned subsidiary of Holdings. It employed the majority of the members of the Main Plan which was IBM's main UK occupational pension scheme.

iv) IBM's business in the UK was run through a management committee ("the Management Committee") which was made up of the heads of the various divisions of IBM's UK business, for instance Finance, Legal, Manufacturing, Personnel, Marketing and Administration. The Management Committee was chaired by Edwin Nixon (who was knighted in 1984 and to whom I will refer as "Sir Edwin" regardless of the time in question). Sir Edwin was, at material times, the Chairman and Chief Executive of IBM UK as well as chairman of the board of Holdings. Sir Edwin died some years ago.

Decision making: the Management Committee

v) The Management Committee was the body which dealt at executive level with the making of business decisions concerning Holdings, IBM UK and other group subsidiaries in the UK. Although, in appropriate cases, decisions could be referred back to the boards of the relevant companies, Holdings operated in a relatively passive way. Sir Leonard states and, lest there be any doubt in case it is contentious, I accept that all substantive decisions of Holdings in relation to the matters with which these proceedings are concerned were made by the Management Committee and actioned accordingly.

vi) In other words, the Management Committee was the competent executive decision-making body for Holdings and its UK subsidiaries in relation to the matters with which these proceedings are concerned. In describing matters in that way, I should make clear that the Management Committee is to be seen as making decisions concerning the management of the relevant company in the same way as a board of directors, or a senior manager or even a more junior employee tasked with the certain duties. There would, of course, have had to be express or implied authority from the board for the Management Committee to act, just as such authority would need to be found for managers or staff to be able to bind the company.

vii) I should, however, point out that there is no documentary evidence before the court recording any formal conferring of such authority on the Management Committee. But there is nothing to suggest that the Management Committee did not in fact have the authority to act in the way in which it did. It cannot possibly be contended that the boards of Holdings and IBM UK were unaware of the pension changes being considered and implemented in the period 1981 to 1983 or were unaware that actual decisions were being made by the Management Committee. And yet there is no hint that the boards had any concern that matters were proceeding without their assent. This is not at all surprising, of course, given that there was a large overlap in the memberships of the board of Holdings and of the Management Committee.

Decision making: "powers reserved"

viii) IBM worldwide was divided into global regions of which IBM Europe Middle East Africa ("EMEA") was one. IBM in the UK was part of the EMEA region. EMEA had its headquarters in Paris and "Paris" was used as a shorthand for EMEA general management in a similar way to "Armonk" for global general management.

ix) In addition to general management in Paris, Sir Leonard says EMEA had its own main board (ie the EMEA board) in Armonk. I see no reason to doubt that although there has been some question during the course of the trial about what a body referred to as the "EMEA Executive Committee" actually is, a matter I will refer to later. IBM Corporation was responsible for high level policy. EMEA, and through it Holdings, IBM UK and other UK subsidiaries, were subjected to that policy through the agency of the EMEA board. Accordingly, in what follows, like Sir Leonard, I do not distinguish between IBM Corporation and the main EMEA board and use "Armonk" as including both.

x) The IBM group consisted of a large number of separate companies in different jurisdictions. Although the group was run as a single organisation, there was in practice a considerable amount of devolution of power and a considerable amount of autonomy at national level. However, for certain matters, Armonk retained "powers reserved" in relation to which certain decisions had to be approved by Armonk. These "powers reserved" included decisions in relation to Retirement Plans, Executive Compensation, Accounting, Corporate Governance and Real Estate among others. Accordingly, the introduction of the C Plan was subject to "powers reserved" and, if the system operated in accordance with the policy, should only have been implemented with the approval of Armonk. A matter referred to Armonk under "powers reserved" would not necessarily be dealt with at the board level of IBM Corporation. There is no material available to indicate at what level any particular matter for decision would be dealt with, and the evidence of the witnesses in the present case does not indicate at what level the approval of the C Plan would fall to be dealt with.

xi) In relation to the UK, the process for obtaining approval from Armonk for a matter within "powers reserved" was this. First, the Management Committee would formulate a proposal. Secondly, the proposal would be referred by the Management Committee to Paris. Thirdly, Paris would consider the proposal and decide whether it should be referred to Armonk for approval either as it stood or with amendments discussed with the Management Committee.

xii) If, after referral to Armonk, the proposal was not approved, that would be an end of the matter. If the proposal was approved, the actual decision whether to implement it would remain with the relevant local decision making body, that is to say in practice, in the UK, the Management Committee. There was no obligation on the Management Committee to decide to implement a proposal simply because it had been approved by Armonk (although I think it would be fair to add that it would be an unusual case – such as a relevant change in circumstances – in which the Management Committee would go to the trouble of obtaining approval unless the eventual decision was an almost foregone conclusion).

Sir Leonard's description of the corporate structure and decision making process is supported by the witness statement of Barrie Morgans who was the Finance Director of Holdings for a number of years from 1982 (having previously been Financial Controller of IBM UK). He was a member of the Management Committee at all relevant times. He was a director of the Trust Company from the end of 1980 until 2002. He was too ill to attend for cross-examination but I do not believe this aspect of his evidence is really open to challenge.

A detailed account is also set out in the witness statement of Derrick Taylor. He became the Treasurer of IBM UK in 1975. He was a director of the Trust Company from the end of 1980 until March 1990. He was not a member of the Management Committee but reported to Mr Morgans who was a member. He was familiar with the Management Committee's workings and attended some meetings.

I have mentioned that the evidence, whether generally or in relation to the C Plan, does not indicate the level within Armonk at which a decision under "powers reserved" was to be taken. But what, I consider, the evidence does establish so far as the present case is concerned, is that it was in accordance with the "powers reserved" procedure for the Management Committee to submit their request for approval to Mr Kofmehl in Paris for onward transmission, if necessary, to Armonk. This was what was done. In my view, it must be right that the Management Committee, if and when they were to receive an approval, should be entitled to assume that such approval had been validly obtained in accordance with "powers reserved". It would be for Paris to refer the matter to Armonk by transmission to whoever it, Paris, considered to be the relevant person in Armonk and for the matter then to be referred up the chain of command until it reached the appropriate decision-making level in Armonk.

The Trust Company's closing identifies contemporaneous documents which demonstrate the distinction between (i) Armonk/Paris approving a proposal and (ii) local management actually implementing the transaction in question:

i) In the bundle is to be found an "Organization Letter" dated 2 March 1982, containing a "corporate instruction" from Armonk applicable within the IBM group. This stated that "voluntary employee benefit plans and programs" which are "proposed by … EMEA" are "to be reviewed for concurrence by the IBM Corporate Personnel Staff Executive prior to implementation or revision". The words "reviewed for concurrence" ie for approval and "implementation or revision" contrasts a proposal with the actual implementation.

ii) The proposal for the creation of the C Plan (that is to say the Initial Proposal described in paragraph 113 below) drew a distinction between "EMEA approval" and "implementation".

iii) In the covering letter dated 18 August 1982 for the Initial Proposal, addressed to Paul Kofmehl, Directeur Générale des Opérations in Paris, Sir Edwin sought "your approval to the proposal for the introduction of a contributory pension and life assurance plan". And Mr Cawley's accompanying letter to his opposite number stated that "EMEA approval" was required on 14 December 1982 so that announcements (ie to the employees) could be made in the week ending 24 December 1982. This accorded with the timetable set out in the Initial Proposal itself.

iv) A minute of the Executive Committee of the board of Directors of IBM World Trade Europe/Middle East/Africa Corporation (as to which see paragraphs 153 and 154 below) dated 14 December 1982 refers to "concurrence" in a "proposal". This Committee did not purport itself to take the decision to proceed with the C Plan.

The letter from Sir Edwin to Mr Kofmehl is significant also in demonstrating the channel of communication with Armonk in relation to the C Plan. As we shall see, it was from Paris that Sir Edwin received what he perceived as a communication of approval from Armonk, on the basis of which announcements were sent to employees.

It is clear, in my judgment, that "powers reserved" did not deprive Holdings of its power and authority to make its own corporate decisions. See paragraph 104 below.

In the present case, however, the decisions made in relation to the proposals for an implementation of the C Plan were taken by the Management Committee. I will come in due course to consider what, if anything, was decided upon or ratified by the boards of Holdings and the Trust Company. But on the assumption, at least in relation to Holdings, that there was no relevant decision by the board and no ratification by the board of the decisions of the Management Company, the issue arises whether the intention of the Management Committee is the relevant intention for the purposes of rectification in accordance with the law discussed at paragraphs 14ff above.

As I have said, the actual decision to implement a proposal approved by Armonk remained to be taken at local level. Such a decision, as with any other high-level executive decision, would be taken by the Management Committee and implemented accordingly. Acting within their delegated powers, there is no reason to think that the Management Committee would be acting beyond the scope of their delegated authority or that the decision taken would not be binding on Holdings.

But what would be the position if a decision on a proposal falling within "powers reserved" were made at local level without Armonk actually having approved a proposal?

If such a decision were to be taken by the board of Holdings, it would, in my judgment, be valid and binding on Holdings. Thus, if the board of Holdings had resolved to approve an amendment to the Main Plan and that amendment had been effected by the Trust Company, the amendment would not be invalidated by reason of an absence of approval by Armonk. The adoption of a policy by Armonk – that certain decisions should be subject to "powers reserved" – was precisely that, a policy, and could not, as a matter of English law applied to an English company, prevent the board from exercising its powers even contrary to that policy. In any case, even if it were possible for a board to release its powers to the extent necessary to ensure compliance with the retention of "powers reserved" by Armonk, this was not done. There is nowhere to be found a resolution or other instrument which purports to restrict the powers of the boards of Holdings (or IBM UK for that matter) so as to reflect "powers reserved". Nor, I might add, to the constitutions of either company contain any such restriction.

The same result would be reached if the decision were taken, not by the board of Holdings, but by the Management Committee, provided that it did so with the approval or authority of the board or provided that the board subsequently ratified the decision. However, if the Management Committee were to take the decision without further reference to the board, it might be argued as follows:

i) "powers reserved" was a feature of the decision-making process within Holdings even though it was not binding as a matter of law rather than policy on Holdings or its board.

ii) The Management Commitee can only be acting pursuant to a delegated authority from the board of Holdings.

iii) Since there is no evidence about what was actually delegated to the Management Committee, inferences have to be drawn about the scope of the delegated authority.

iv) Having acted as the executive decision making body in relation to all aspects of the activities of Holdings without any apparent objection from the board, it might have to be accepted that the Management Committee had been vested with very wide powers by way of delegation.

v) But even if it is to be inferred that wide powers of management had been delegated to the Management Committee, it is not to be inferred that they could make a decision, without reference back to the board, which breached a fundamental aspect of the decision-making process laid down by Armonk as a matter of policy, namely "powers reserved".

vi) Accordingly, a decision made by the Management Committee in relation to a matter within "powers reserved" would be ineffective if made in the absence of approval from Armonk since it would not fall within the scope of the delegated authority of the Management Committee.

I reject such a line of argument on the facts of the case. As I have said in paragraph 94 above, consent to the proposals for the C Plan was sought by the Management Committee (acting within in its authority from the board of Holdings) in accordance with the appropriate procedures. A response was received in accordance with those procedures. Provided that the Management Committee then acted in accordance with the response received (I deal with this later), there is, in my judgment, no scope for an argument along the lines which I have just set out. The Management Committee were entitled to rely on the response which it received.

In any case, I consider that the wide scope of delegated authority which the Management Committee possessed was sufficient to enable them to make decisions on behalf of Holdings even where "powers reserved" did apply and was not complied with, with the possible exception of a case where they knew, or ought to have known, of this non-compliance. Those are not the facts of the present case.

Trustee decisions

Mr Taylor's witness statement contains evidence about how the Trust Company viewed its role. He describes the relationship between it, Holdings, IBM UK and the Management Committee as close; this was because the board of the Trust Company had strong employer representation including Sir Edwin, Sir Leonard and other directors of Holdings. Through that representation, the board of the Trust Company was always well aware of the employer's decisions on matters of personnel and benefits policy in so far as they had an impact on the Main Plan. The board was also aware of the significance of "powers reserved".

He states that the Trust Company had no involvement in the preparation of such policy. Its role was seen as being to administer the Main Plan, to protect members' accrued benefits, to manage the investments, and, working with the actuary, to advise the employer about contributions. It did not seek to decide the level or type of benefits to be provided for employees' future service as this was regarded as a matter for the employer. Accordingly, the approach of the Trust Company to changes suggested by Holdings in relation to future benefits would generally to be to agree the change if it was beneficial to members. The Trust Company would, he states, consider any proposal carefully and not let it go through simply on the nod. It would not have attempted to influence the employer in any direction as regards future benefit provision. That evidence, as far as it goes, was not challenged and I accept it.

The evidence of Mr Ellis, Mr Taylor and Mr James Lamb (the current Chairman of the Trust Company at the time these proceedings were issued) is also entirely consistent with the description of the approach of the Trust Company. Without being a rubber stamp, it would always, or almost always, accept Holdings' proposals so far as benefits were concerned. In the context of the C Plan, it would have been likely, and this is in fact what happened, that it would agree to the introduction of the C Plan as Holdings proposed. And although the power to make the necessary amendment was vested in the Trust Company itself, the decision to exercise that power would ordinarily follow the wishes of Holdings. It is not, in any event, suggested by IBM in the present case that the 1983 Trust Deed and Rules are somehow ineffective or vitiated because no proper decision was made by the Trust Company; the issue is simply one of its intention when it did so and whether that intention supports the claim for rectification.

The circumstances leading up to the creation of the C Plan; its approval and its implementation

I turn now in more detail to (i) the events leading up to the creation of the C Plan during 1982 and 1983 (ii) the extent of the approval which was obtained from Armonk and Paris to the terms of the C Plan and (iii) the implementation of the C Plan up to July 1983 when it went "live".

Drawing from the evidence of a number of witnesses and from the documents, the following conclusions can be drawn:

i) By the end of the 1970s, IBM (in the UK) was operating a number of pension plans, the most significant of which was the N Plan section of the Main Plan. This was a non-contributory arrangement.

ii) At the beginning of 1980, IBM decided that it wished to carry out a full review of its existing pension arrangements, in part with a view to improving the benefits of the Main Plan. It was a stated objective of IBM that it would seek and maintain in its remuneration and benefits programmes for employees a position amongst the leaders in British industry and commerce. It had become clear by 1981 that, for all IBM's achievements, there was growing dissatisfaction with the level of benefits provided by the Main Plan in that they were falling behind other leading employers. There was also a current debate about shortening the duration of a working life (strange as that may seem to 2012 minds) and about equalising retirement ages for men and women (something now long since introduced by law).

iii) By that stage, the UK business faced a human resources problem. Sir Leonard states, and I see no reason to doubt this, that IBM was also becoming concerned to have greater room for manoeuvre in its manpower management. It had grown from 3,000 employees in 1962 to some 18,000 in 1981/82. It was committed to a practice (albeit not an announced formal policy) of "no redundancy", a policy to which it remained committed for years after the introduction of the C Plan. Labour turnover had fallen from 12% pa in the early/mid 1960s to 3% pa in the early 1980s.

iv) This lack of turnover meant that IBM had an increasingly mature workforce in an industry which depended on innovation and where it was perceived to be valuable to promote or bring in new talent. Management was therefore keen that any future need to reduce the size of the workforce should be helped rather than hindered by the terms of the Main Plan. Such encouragement would, it was perceived, result from changes to the Main Plan which would allow members to retire early, to do so without actuarial reduction and to do so without employer consent. It would encourage early retirement without associated redundancy costs and without the scope for employee disenchantment which often accompanied redundancy programmes. Such changes would, as Mr Cawley put it, encourage "attrition" at senior levels and enable "vitality" hiring to continue at younger ages.

v) The three drivers for changes were, in summary, first, the opportunity to equalise retirement ages for men and women; secondly, to bring benefits into line with those provided by competitors; and thirdly, most importantly, to encourage turnover of staff to get rid of dead wood and bring in new talent.

vi) In 1981, the Management Committee agreed that a small working party should be set up led by Personnel and assisted by Finance to formulate proposals for improvements to be submitted to Paris and Armonk. The initial timetable for the project including obtaining approval was 15 months from September 1981 to December 1982.

vii) The general pensions review was, at a senior level, under the control of Sir Leonard. He asked Mr Cawley to assist with the design of a new section providing improved benefits over those of the N Plan, which eventually became the C Plan. Mr Cawley led the review. It was agreed that Sir Leonard and Mr Morgans would provide what Sir Leonard describes as "outline strategic guidance and extra help as needed". In the event, Mr Morgans had very little to do with the project.

viii) The team led by Mr Cawley included Mr Ellis. Mr Ellis reported to Mr Cawley during the project. Together they had responsibility for putting together a proposal for a new section of the Main Plan. Mr Cawley was the primary author of the proposal with the duty of seeing the project through.

ix) The first stage of the project was to prepare a proposal to put to Paris and Armonk for approval. I do not think that I need to go into the detail of how this proposal was developed. Briefly, it involved numerous meetings with Mr Cawley attended by representatives of the various interested IBM areas (such as Legal and Finance). Actuarial advice was sought from Clay & Partners who provided cost data on a number of variations of cost and design about which I will need to say more later. Mr Cawley says, and I accept this, that he also discussed from time to time during this period of development the design of the C Plan with some members of the Management Committee.

x) Mr Cawley relates – and this was not challenged – that since pensions matters fell within "powers reserved", IBM Corporation had a "blueprint" for his team to work to which determined the form and content of any proposed document. That "blueprint" (which I have not seen) set out procedures that would need to be followed to present a proposal and the information required in support. It was IBM practice at that time to carry out comparisons of compensation and benefits with other leading UK companies. Apart from ICL which was, at that time, the only large computer company with which IBM could make comparisons, the survey covered other industries.

xi) Mr Cawley also relates – and again this was not challenged - that he did indeed ensure that the proposal as drafted satisfied the framework of the blueprint. One matter he had particular regard to was normal retirement date ("NRD"). The N Plan had an NRD of 65 for men and 60 for women. It was the general approach of UK management (supported and shared by Armonk) that NRD should be equalised for men and women. In that regard, an NRD of 60 for all members was initially suggested by Mr Cawley and Mr Ellis following a review of pension provision in the UK market place. The rationale was to make the C Plan attractive in comparison with the N Plan (even though it would be contributory). There would be no change in NRD for women and the reduction in NRD for men would give them the opportunity to retire earlier. It was part of IBM's hope and expectation that as many N Plan members as possible would transfer across (transfer always being proposed to be voluntary).

xii) It is to be noted that the C Plan was to be the section of the Main Plan which new employees would join. The main motivation behind the introduction of the C Plan was to help IBM in the UK address its personnel issues and, clearly, the flexibility and attractiveness of the C Plan were important in relation to new employees as much as to existing members of the N Plan.

xiii) The work carried out by the team led by Mr Cawley resulted in a 95 page document which has been referred to in these proceedings as "the Initial Proposal". I will come to the detail of the Initial Proposal in the next section of this judgment starting at paragraph 113.

xiv) The final version of the Initial Proposal was approved by Sir Leonard and Mr Morgans to be put to the Management Committee for approval before submission to Paris. I accept Sir Leonard's evidence that the usual process at the time was for the appropriate staff group, usually Personnel or Finance or perhaps Legal, to test the view of the relevant staff in Paris and/or Armonk before submitting a formal proposal document. A proposal could then be refined in the light of comments received. In the present case, the Initial Proposal was submitted to Paris rather than straight to Armonk.

xv) Sir Leonard explains the design and objectives of the C Plan quite briefly in paragraph 54 of his witness statement. I accept what he says which I can summarise as follows:

a) The C Plan was designed so as to be an improvement on the N Plan which, as I have explained, was seen as falling behind the benefits offered by other top-ranking UK employers. By including favourable retirement terms, the C Plan would encourage turnover of employees.

b) The C Plan would be contributory, unlike the N Plan. It was to be introduced with an employee contribution rate of 5% with the balance of cost being met by the employers. The employers' rates of contribution would exceed those under the (non-contributory) N Plan in the light of the cost of improved benefits.

c) The C Plan would provide for equality for men and women in accordance with the then Government's stated aims. The original proposal was that all employees would be able to retire at age 60, which was already the retirement age for women. There was no issue about a need for consent to retire at age 60: that age was simply to be the normal age at which members would retire, just as it was already 60 for women and just as it was currently 65 for men under the N Plan.

d) The intention and hope of IBM was that the take up for the C Plan would be as high as possible: the personnel objective which I have described, what Mr Cawley referred to as "attrition", would not otherwise be achieved. Women already enjoyed an NRD of 60; that factor militated strongly against the introduction of a later NRD firstly because women would not be encouraged to transfer from the N Plan to the C Plan and secondly because it would give rise to contractual employment issues. Considerations of equality therefore led to the adoption of a common NRD of 60.

e) Early retirement between age 53 and age 60 was to be a feature of the C Plan, but only with the consent of Holdings. In this way, IBM would be able to control the turnover of its key skill groups. As Sir Leonard put it: "It would recognise the importance of allowing flexibility (for employees) in the age of retirement so enabling management to use the attractive retirement terms in the pension scheme as an asset in influencing labour turnover or one-off reductions". And the operation of the flexibility of the C Plan would provide an attractive alternative to compulsory redundancy should the need arise.

The Initial Proposal

The copy of the Initial Proposal in the trial bundle (and the only edition which has been made available) is dated August 1982. Its page numbering (to be found in the bottom left-hand corner) reads "IBM Confidential August 1982 Page [x] of 95". Pages 24 and 25 of 95 are missing. It is unlikely that they contain anything relevant to the present proceedings. The front page of the Initial Proposal is headed in this way:

"IBM UNITED KINGDOM HOLDINGS LTD

RECOMMENDATION TO IBM CORPORATION

FOR 1983 IMPROVEMENT TO:

IBM PENSION AND LIFE ASSURANCE PLAN

IBM SUPPLEMENTARY PENSION AND LIFE ASSURANCE PLAN"

The Initial Proposal contains 5 sections: Present pension and life assurance plans (ie the Main Plan); Proposed pension and life assurance plans (which includes the C Plan); timetable for implementation; Justification; and Cost estimates.

Section 1, the present plans, sets out in tabular form over the first six pages, the benefits of the various Plans contained within the Main Plan. The first column deals with the N Plan which is all I need be concerned with for present purposes. The column is headed "N Plan (non-contributory)". It contains a long list of entries including the following:

1.1 DEFINITIONS

Normal Retirement Age	Male: 65 Female: age 60
Contributions: Employee Company	Nil 14.25% (15.25% in 1983…) of previous year's pensionable earnings plus frozen initial liability of £4.446m for 1981 improvement

1.2: NORMAL RETIREMENT

Pension formula	[this contains a formula incorporating integration with the state pension: nothing turns on the detail]
Increases to normal retirement pensions in payment:	Irregular ex-gratia reviews at company discretion

1.3 LEAVING IBM SERVICE

If less than 5 years qualifying service completed	No entitlement – premium paid to state to secure requisite minimum benefit (see attachment A)
If more than 5 years qualifying service completed	Deferred pension entitlement payable at normal retirement date and calculated on normal retirement formula……
Increases to deferred pensions during deferment	Statutory minimum portion revalued at mandatory 5% pa compound. Offset against plan entitlement attributable to service after April 1978. See attachment A
Increases to deferred pensions in payment	Statutorily no less favourable treatment than normal retirement pensions in payment

1.4: EARLY RETIREMENT

Eligibility	5 years service and: Male: Age 55+; Female: Age 50 +.
Pension formula	Immediate pension calculated as deferred pension on leaving service and reduced by early payment discount of 0.25% of deferred pension per month earlier than age 60

There are three points to make arising from the table as it appears in the Initial Proposal. First, the reference to the frozen initial liability ("FIL") in £s against Company Contributions is almost certainly an error and should be US$s. It represents the cost of certain benefit improvements made in 1981 (mainly the introduction of an accrued benefit for surviving spouses) resulting in an increase in the contribution rate. Secondly, no mention is made of early payment of a deferred pension. Thirdly, early retirement is permitted only within 10 years of NRD (hence the different eligibility ages) and, although no mention is made of this, the consent of Holdings to early retirement was a requirement of the Rules.

Section 2, the proposed plans, set out in similar tabular form the benefits of the three proposed new plans, the C Plan, the E Plan (an executive plan) and the A Plan (an additional voluntary plan). The first column deals with the C Plan which is, again, all I need be concerned with for present purposes. The column is headed "C Plan (contributory)". Like the table relating to the existing plans, it contains a long list of entries including the following:

2.1 DEFINITIONS

Eligibility	All permanent employees age 25 or over and who join IBM prior to age 55
Norman Retirement Age	All members: Age 60
Contributions: Employee Company	5.00% (current pensionable earnings less state lower earnings limit) 19.50% of current pensionable earnings plus frozen initial liability $53.523m (including FIL for 1981 improvement)

2.2: NORMAL RETIREMENT

Pension formula	For each year of service: 2.25% FPE – 2% State Pension Plus supplement if applicable (see note III) Note……[nothing turns on the detail]
Commutation	$1.00 pa pension surrender for: Male: $10.18 tax free lump sum Female: $11.00 tax free lump sum [maximum lump sum also specified]
Increases to normal retirement pensions in payment:	Minimum 3% pa compound subject to rate of price inflation being at least 4% pa ….plus irregular ex-gratia reviews at company discretion

2.3 LEAVING IBM SERVICE

If less than 5 years qualifying service completed	No pension entitlement, contributions with interest (5% compound) refunded less certified amount (see attachment A) and less 10% tax. Premium inclusive of certified amount paid to state to secure requisite minimum benefit
If more than 5 years qualifying service completed	Deferred pension entitlement (including supplement) payable at normal retirement date and calculated on normal retirement formula……
Increases to deferred pensions during deferment	Statutory minimum portion revalued at mandatory 5% pa compound (see attachment A). Excess over statutory minimum portion increased at 3% per annum compound, subject to rate of price inflation being at least 4% pa.
Increases to deferred pensions in payment	Statutorily no less favourable treatment than normal retirement pension in payment

2.4: EARLY RETIREMENT

Eligibility	Within 10 years of Normal Retirement Date and with 5 years IBM service.
Pension formula	Immediate pension (including supplement) calculated as deferred pension on leaving service and reduced by early payment discount of 0.25% deferred pension per month early.

There are three points to make arising from this table as it appears in the Initial Proposal. First, the reference to the frozen initial liability in the entry against Company Contributions is to the cost of the benefit improvements of the C Plan as compared with the N Plan in respect of past service (assuming all N Plan members transfer to the C Plan) and includes the cost of the 1981 benefit improvements which I have mentioned above. Secondly, as with the N Plan, no mention is made of early payment of a deferred pension. Thirdly, the wording of the table is based on that of the table in section 1 with changes to the wording to reflect the changes in benefits; in particular, the adoption of a common NRD for men and women and resultant in changes to "Eligibility" in section 2.4.

Section 3 of the Initial Proposal sets out the timetable for implementation. Section 3.1 contains the Announcement Timetable. The first entry is "14 December 1982 – EMEA approval secured". It is clear from the evidence, indeed I think it is clear from this entry standing alone, that such approval was seen as being all that the Management Committee needed to proceed. In other words, any approval from Armonk, rather than Paris alone, would have been obtained in accordance with "powers reserved". But Paris, as the evidence shows, was to be the channel through which the Initial Proposal was to be fed to Armonk and was, consistently, also to be the channel through which approval would be fed. The date of 14 December 1982 was selected because it was known that there would be a meeting of some decision-making body within IBM in the US on that date at which the question of approval could be formally addressed.

It was envisaged, under this timetable, that the following would occur: an announcement to the workforce would be made in the week ending 24 December 1982; by the end of January 1983, comprehensive briefings would be given to personnel managers in operations groups and a "mini-statement" of benefits would be given to each eligible employee; during the period early February to the end of May 1983, there would be "road shows" (Mr Cawley's later description of the programme of information dissemination) to all UK business locations; there was to be a closing date of 3 June 1983 for applications to join the new plans; and implementation of the new plans was to take place on 6 July 1983 followed by provision of a full statement of benefits to each eligible employee.

Section 3.2 deals with Communication which was to take place in a number of ways. The initial announcement was to be by way of Notice Board bulletins and Management Information Letter ("MIL"). The MILs actually sent are of very great importance and I will be dealing with them in detail in due course.

Section 3.2 deals with entry to the proposed new plans. It is made clear that entry for all current employees is to be optional. The proposed plans represented a change in employees' contracts of employment and had to be voluntarily accepted.

Section 4 contains the justification for the proposals. This section runs from page 34 to page 49 of the Initial Proposal. I do not attempt to address its contents in any detail. Section 4.1 addresses broadly the external environment, including social security and political aspects. It discusses a number of specific issues under the headings Disclosure of Information, Representation on Trustee Boards, Solvency, Equal Status for Males and Females, Index Linking of Pensions, Treatment of Early Leavers, Contracting Out and Pension Fund Regulation. Of those, Treatment of Early Leavers takes on some added significance in the light of the way in which the proceedings have developed. This section draws attention to the report of the OPB published in 1981 and its recommendations concerning the revaluation of vested pension rights during the period of deferment. It is noted that the then Government had declared a preference for voluntary action in this area but that opposition parties were pressing for legislation. The section ends with this:

"This potential legislation could have significant impact on IBM and it is essential that IBM minimises the impact by taking voluntary action now."

The Initial Proposal makes no mention in this section or anywhere else of the Preservation Requirements. This is not surprising since, even if they had crossed anyone's minds, the benefit structure of the C Plan, as then proposed, would have conformed with them: the deferred pension would have been payable at age 60 (which was the earliest age at which short service benefit had to become payable) and the amount of the deferred pension of an early leaver was sufficient to conform.

This section was summarised in the following way in section 4.1.6:

"There is considerable activity in the pensions area, with pressure being applied from many sources. IBM must remain sensitive to these pressures and must seize opportunities to satisfy the pressures and to meet employees' needs and expectations as appropriate. Furthermore, the pressure for further legislation is sufficient, we believe, to take the position that legislation will happen but it is uncertain when it will happen."

Section 4.2 addresses the internal environment. Key issues were identified as these:

i) the short term requirement significantly to increase attrition to meet manpower targets:

ii) longer term requirements to facilitate separations which will be required to meet the vitality objectives; and

iii) present requirements to maintain and improve morale and motivation within management and employees.

I understand ii) to be management-speak to encourage employees to resign (facilitate separations) but while retaining key staff and recruiting dynamic innovative new recruits (meeting the vitality objective).

Section 4.3 describes the rationale. The section starts with a general review. The major concern with the existing plans is said to be the perceived low level of pension resulting after a full career with adverse reaction from members approaching retirement with a consequential increase in interest among younger members. Pensions strategy needed rethinking: any delay could have a serious impact on employee morale and motivation. This is all gone into at greater length, ending with this:

"The proposal is an ambitious one. However, it should be seen as a forward thinking proposal designed to meet the business challenges of the 1990's and beyond. This proposal presents an ideal opportunity to close the 'S' Plans and harmonise male/female conditions in a major step towards single status. Furthermore, it places IBM in a strong competitive position whilst achieving flexibility, through cost sharing, to meet future funding issues."

The description of the proposal as "ambitious" was both realistic and prescient: realistic because it could not be assumed that the proposal would pass muster with Armonk and prescient because it did not do so, as we will see.

Section 4.3.3 gives an outline of the proposals. Under the heading Normal Retirement Date, it is explained that the principal objective under this heading is to achieve equality between male and female employees. Reference is made to proposals published by the Government and supported by the National Association of Pension Funds and the Institute of Personnel Management, to harmonise at age 63. It is pointed out that in the UK, a low retirement age is perceived as better than high retirement ages, foreshadowing a cultural issue which has surfaced in the present proceedings. Various arguments for reducing the common pension age to 60 are then put forward.

Discussion of the benefits proposed proceeds under a number of headings, the detail of which is not of any materiality for present purposes save under the headings "Entry Age", "Full Career", "Leaver Benefits" and "Early Retirement".

i) Taking "Entry Age" and "Full Career", it is pointed out that a full contributor career will be 35 years: an Entry Age is to be selected and NRD is to be 60.

ii) Under the heading "Leaver Benefits", it is explained that these will continue to reflect minimum legal requirements which are that those with at least 5 years' service must have a "vested rights pension" on the same terms as retirements and hence these rights must become payable at the normal retirement date of age 60. The proposal to guarantee increases in deferment reflects the trend in UK practice and is required to enable IBM to meet the potential legislation which had already been referred to and which I have mentioned above. This is an accurate reflection of the Preservation Requirements as they would have operated in relation to a scheme with an NRD under its rules of age 60.

iii) Under the heading "Early Retirement", it is explained that the early retirement conditions broadly follow the present plan conditions. However, the effect of adopting an NRD of age 60 has the effect of harmonising the earliest eligible age for early retirement. I would add that, taking this section together with the description of the benefits of the N Plan earlier in the Initial Proposal, the reader would understand that both men and women would be able to retire on an immediate pension with the consent of Holdings at any time between ages 50 and 60.

Section 5 deals with Cost Estimates. Obviously Mr Cawley and his team had had the benefit of actuarial advice from Clay & Partners. The cost estimates provided in the Initial Proposal are, it is stated, supplied by the Actuary (ie Clay & Partners in the person of Mr Gooch and Mrs James). It is interesting, although perhaps slightly depressing to a member of the UK public, to note in passing that the costings are based on an exchange rate of £1 sterling = US$1.71. The estimates assume that 100% of the existing employees will join the proposed plans.

At section 5.4 is included an estimate of the Improvement Cost (ie the added cost of providing the benefits in respect of past service for those transferring into the new plans). The FIL is $390.585m or some £228m; this is amortised over 10 years to an annual rate of $49.077m or £28.7m. Additional funding, for ongoing accrual, at 4.25% is $15.501m. The whole of the cost of the improvements ranks for relief from Corporation Tax at 52%. The annual cost after tax is $30.997m as compared with $15.499m under the old arrangements.

In section 5.5 is to be found a table headed "Cost Estimates – Key Elements". I do not need to set out the full table, but the following is helpful:

	TOTAL PENSION PLAN Funding F.I.L. & PE ($000's)
Present Plan Costs (1983)	15.25% 4,466
Key Elements Formula Retirement Age Death in Service Total Additional Cost: Employee Contribution: Additional Cost for Company:	4.60% 25,992 3.80% 23,085 0.45% NIL 8.85% 49,077 4.60% NIL 4.25% 49,077
Potential Savings – Supplementary Plan Net Additional Cost	(0.48%) NIL 3.77% 49.077

Section 5.7 is headed Cost Justification and gives an example to demonstrate why the proposed change to NRD will result in future savings which exceed the additional cost of the proposal. It is, one might speculate, an example carefully chosen to present the proposal in the most favourable light. But what it does show is that in the long term there could actually be savings and provides considerable support for the proposition that, even if there are no net savings, the actual cost is not likely to be anything like the net additional cost shown in the Table of Cost Estimates. The conclusion is this:

"We believe saving will be realised at around 25 years after implementation provided that anticipated lower average levels are achieved. A more balanced population structure should result in lower average levels in the long term."

The only other part of the Initial Proposal which I want to mention is Attachment C which sets out the Pension Plan Valuation Basis. Within the valuation assumptions made appears the following: "Retirement: 25% of males and all females retire at age 60; residual males retire at 65: proposal all at age 60". This was also the basis on which the Actuarial Valuation as of 31 December 1982 and signed on 14 July 1983, was prepared, a basis adopted by the Actuary after discussion with Holdings/IBM UK. It seems that this basis is likely to have had some connection with reality.

Events after the finalisation of the Initial Proposal

Having dealt with the Initial Proposal, I can pick up the narrative from the time of its finalisation. On 18 August 1982, Sir Edwin wrote to Mr Kofmehl in Paris. His title was "Directeur General des Operations". He was a senior figure in EMEA in Paris. It was through him, it is clear, that Sir Edwin sought the approval from Armonk which was required under "powers reserved". The letter is not long and it is easiest if I set it out in full:

"Dear Paul

May I request your approval to the proposal for the introduction of a contributory pension and life assurance plan, with improved benefits, as detailed in the attached. [The attached can only have been the Initial Proposal as is confirmed by a letter from Mr Cawley of the same date which I come to in a moment].

The existing plans, introduced in 1975, provide a basic pension on a non-contributory basis, with the option to supplement the single life benefit through a voluntary contribution.

Over the period since 1975, our competitive position has deteriorated, the voluntary element proving to be inadequate, and increasing employee demands for improved pensions lead us to propose a comprehensive contributory plan.

The proposal would be compulsory for new hires at age 25, and optional for current employees. If we were to achieve a high level take-up and avoid a significant dual standard of pension provision, we must offer an attractive package of improvements.

Our consulting Actuary has determined the cost of this proposal in accordance with Corporate Instructions. The employee contribution meets over half of the future funding costs, with the balance together with past service costs being met by IBM. The contributory concept will permit greater flexibility to meet future improvement costs. The proposal is very expensive, but it should be borne in mind that the longer implementation is delayed, the more expensive it will become.

We plan to announce the improvements during the week ending 24 December 1982, with implementation on 6 July 1983. The intervening period is required to fully communicate with employees and to encourage maximum participation.

The proposal has the support of IBM United Kingdom Country Management.

Yours sincerely

ER Nixon"

On the same day, Mr Cawley wrote to Claude Baudot at Compensation and Benefits Planning, IBM Europe (at the same address as Mr Kofmehl). He attached two copies of what can only have been the Initial Proposal. He reported that five copies had been sent to Mr Kofmehl "as required by the International Benefits Planning Manual, to allow simultaneous review". He informed M Baudot that "Our consulting actuary, Mr Ellis, and myself" would be available to visit on 1, 2 or 3 September to provide any additional input required. It is not clear whether that was a misdescription of Mr Ellis as a consulting actuary (which seems unlikely) or was an indication that the actuary (presumably Mr Gooch) would be able to attend in addition to Mr Ellis. He added that, in order to meet the planned announcement during the week ending 24 December 1982, "we require EMEA approval on 14 December 1982". I understand that 14 December was mentioned because it was known that there was to be a meeting of some decision-making body within IBM in the US on that date.

One thing which is clear, I consider, from Sir Edwin's letter and Mr Cawley's letter is that they, and I am willing to infer everyone involved in the UK, considered that the need for approval under "powers reserved" would be satisfied by EMEA's approval following submission of the Initial Proposal to it. It would be for EMEA to make the running in obtaining approval from Armonk (if EMEA was any different from Armonk for this purpose) if and insofar as it was required. I have discussed at paragraph 94(viii) to (xii) above the manner in which approval under "powers reserved" would be sought and obtained. In accordance with that discussion, it is my judgment that all of those concerned in the UK – Holdings, the Management Committee and the individuals from Sir Edwin and Sir Leonard to Mr Cawley and Mr Ellis – were entitled to assume that approval from Paris in response to any revised proposals was approval for the purposes of "powers reserved".

The detail of what happened next is unfortunately very unclear. There is an almost total absence of documentation and inevitably memories at this distance of time are hazy. The material evidence, such as it is, really comes only from Sir Leonard, Mr Cawley and Mr Ellis. I have not received evidence from anyone involved at the Paris end of the approval process nor from anyone from Armonk who was involved. What is clear is that Armonk did not approve the Initial Proposal. Quite when, if at all, it ever formally refused approval is not clear. But as will appear in a moment, it became clear that Armonk was unwilling to accept an NRD of age 60 and may have had concerns about the overall cost, as a result of which revised proposals were formulated.

Sir Leonard says that, following submission of the Initial Proposal, there was a process of discussion, in the first instance with EMEA staff in Paris, with Mr Cawley acting as the contact point for the purposes of answering questions and then later, as the process approached fruition, with Sir Edwin, Sir Leonard and Mr Cawley attending meetings in Paris "to field Armonk's response" as Sir Leonard put it. Let me now set out what Sir Leonard says about these meetings in his witness statement:

"56. Unfortunately there are no records of these discussions and I am unable to reinstate the agenda purely from memory. However, the process, as with matters requiring reference up by the Management Committee through the EMEA structure would have taken a well established course. To the best of my recollection, the meeting took place sometime in November 1982, and since we received Armonk's final views on the Initial Proposal in early December, it is reasonable to infer (as would have been expected) that the Paris meeting stage dealt with Armonk's reaction.

57. As I recall, and is borne out by the events, we in the UK had received reports on the discussions which had taken place between the two EMEA headquarters (i.e. Paris and Armonk), with indications that Armonk had refused "retirement for all at 60", but had taken up a secondary reference in the Initial Proposal to the then current DHSS discussions based on the equalisation of retirement ages at 63 (this is referred to at pages 37 and 62 of the Initial Proposal…..). This was seen as less intrusive on male members' benefits, they previously having enjoyed a normal retirement date of 65 under the N plan."

That is Sir Leonard's position and he was not really shaken on it inj cross-examination. I accept this evidence. There is no doubt that Armonk did refuse approval of an NRD of 60 but did not object to an NRD of 63.

That account is consistent with what Mr Cawley says. He says that, after submission of the Initial Proposal, he liaised with personnel in the Paris office. In turn, Frank O'Keefe, who was the Paris employee benefits manager, liaised with Armonk on behalf of the UK to seek Armonk's opinion. He says that once Armonk's views were known, there were further meetings in Paris attended by Mr Cawley himself, Sir Edwin and Sir Leonard. Mr Cawley's evidence, which I see no reason to doubt, is that he certainly recalls travelling to Paris with Sir Edwin and Sir Leonard although he cannot place a precise date on this. He spent part of his time in discussion with Mr O'Keefe, his opposite number at staff level, whilst Sir Edwin and Sir Leonard were in discussion with Mr Kofmehl, their opposite number at board level.

Mr Cawley confirms that Armonk rejected an NRD of 60. Like Sir Leonard, Mr Cawley's recollection is that the thinking in the US at that time was to move to an older, rather than a younger, retirement age. He states that he had the impression that the higher age was considered in the US to be potentially more advantageous as it allowed employees to build up more pensionable service and a larger pension. It was felt that a 5 year reduction for men was too much as this was potentially adverse to men who wished to continue working. The equalised retirement age concept was, however, accepted as, at some stage, was an NRD of 63.

Mr Barrett also adverted to this cultural difference between the UK and Armonk. His perspective of the US attitude was, however, slightly different. It was not so much a concern about male employees who might wish to work longer and build up more pension, rather it was that men should not (whether this was for paternalistic or corporate reasons he did not explain) be encouraged to retire at what was perceived as a young age.

Mr Ellis's evidence is also consistent with all of this. However, he adds recalling that one of the factors leading to rejection of the Initial Proposal was cost. As a result, work was done in connection with the presentation of a revised proposal of some further costings in the light of the changes made. He was not, however, involved closely with this and was unable to give any detail.

There is no doubt that Armonk did reject an NRD of 60 but accepted an NRD of 63. Whether the cultural reasons adverted to by Sir Leonard and Mr Cawley were the reasons for such refusal or whether, as Mr Simmonds suggests, it was a concern about cost and, indeed, whether the cultural reasons were more along the lines indicated by Mr Barrett, I have no doubt, however, that there was a philosophical difference between the approach to retirement age and that was one reason why Armonk was opposed to an NRD of 60.

The objection to NRD was not the only comment made by Armonk. The Initial Proposal was also modified by a reduction in the accrual rate, the replacement of guaranteed increases to pensions in payment with discretionary increases, reduction in commutation rates and removal of non-statutory revaluation of deferred pensions in the period of deferral.

There is a dearth of documentation between the time of submission of the Initial Proposal and 14 December 1982 when some sort of approval was obtained by Sir Edwin from Armonk by way of Paris. Before considering what the witnesses have to say, I wish to consider the documents.

The first document is not a complete document (which is not available) but consists of a few pages of a larger document which has the appearance of a modification of the Initial Proposal. I shall refer to these pages "the Available Revisions". There can be no doubt that a revised proposal was submitted for approval by Armonk at some time prior to 14 December 1982 and, it is reasonable to assume in the absence of any contrary evidence, that the revised proposal was, like the Initial Proposal, submitted to Paris which then dealt with Armonk. Whether the Available Revisions are part of the revised proposal as finally submitted to Paris there is no way of telling. It is highly likely that the revised proposal went through different drafts making modifications to the Initial Proposal and there can be no certainty that the Available Revisions in precisely the form which I have described were in the final version.

The Available Revisions replace the table forming section 2 of the Initial Proposal. The page numbering, against in the bottom left-hand corner reads "IBM Confidential November 1982 page [x] of 99". The major changes are to the proposed C Plan, although there are also changes to the proposed E Plan. The Available Revisions include, in relation to the C Plan, the following:

2.1 DEFINITIONS

Eligibility	All permanent employees age 25 or over and who join IBM prior to age 58 [change from 55 in Initial Proposal]
Normal Retirement Age	All members: Age 63 [change from 60 in Initial Proposal]
Contributions: Employee Company:	5.0% (current pensionable earnings less state Lower Earnings Limited) 18.15% of current pensionable earnings plus frozen initial liability $29.241m (including FIL for 1981 improvement) [Reduction in Company contribution from Initial Proposal reflecting reduction benefits from Initial Proposal]

2.2: NORMAL RETIREMENT

Pension formula	For each year of service: 2.20% FPE – 2% State Pension Plus supplement if applicable (see note III) Note……[nothing again turns on the detail] [This is a reduction from 2.25% FPE in Initial Proposal]
Commutation	$1.00 pa pension surrender for: Male: [maximum lump sum also specified]
Increases to normal retirement pensions in payment:	No formal review process Irregular ex-gratia reviews at company Discretion [This removes the guaranteed increases under the Initial Proposal and effectively reverts to the N Plan provision]

2.3 LEAVING IBM SERVICE

If less than 5 years qualifying service completed	No pension entitlement, contributions with interest (5% compound) refunded less certified amount (see attachment A) and less 10% tax. Premium inclusive of certified amount paid to state to secure requisite minimum benefit
If more than 5 years qualifying service completed	Deferred pension entitlement (including supplement) payable at normal retirement date and calculated on normal retirement formula…… [the wording remains unaltered by the change in NRD from the Initial Proposal and results in the benefit being paid 3 years later]
Increases to deferred pensions during deferment	Statutory minimum portion revalued at mandatory 5% pa compound (see attachment A). Value of excess over statutory minimum portion awarded dividends from excess of actual investment performance over assumed performance
Increases to deferred pensions in payment	Statutorily no less favourable treatment than normal retirement pension in payment

2.4: EARLY RETIREMENT

Eligibility	Within 10 years for Normal Retirement Date and with 5 years IBM service [no change from Initial Proposal]
Pension formula	Immediate pension (including supplement) calculated as deferred pension on leaving service and reduced by early payment discount of 0.25% deferred pension per month earlier…

The commutation rate is shown in a box similar to that of the Table in the Initial Proposal. The figures of $10.18 tax free lump sum for males and $11.00 tax free lump sum for females in the Initial Proposal have been replaced by $9.48 and $10.36 in typescript; but manuscript amendments have been made striking through both of those figures and inserting a single entry of $9.63 applicable to both. The italicised words in the box against increases to deferred pensions during deferment have been struck through in manuscript. There is also a manuscript amendment to the box against Commutation in the Early Retirement section, but nothing turns on the detail. These manuscript changes demonstrate that the revisions to the Initial Proposal must have gone through different drafts. It suggests that the Available Revisions are not pages from the revised proposal as submitted as it is inherently unlikely that such an important document would contain manuscript revisions.

The words "earlier than age 60" in the box against Pension formula replace the word "earlier" in the Initial Proposal. It is to make clear that, as with the N Plan, a reduction for early payment of an immediate retirement pension is made only where the retirement is under age 60.

The next document is headed "Minutes of the Regular Meeting of the Executive Committee of the board of Directors of IBM World Trade Europe/Middle East/Africa Corporation" ("the December Executive Committee Minutes"). None of the witnesses was able to identify quite what IBM World Trade Europe/Middle East/Africa Corporation (let me call it "World Trade") was. Mr Barrett perhaps comes closest although it is not clear that even what he says is entirely accurate. He says that this would not have been Armonk which appears to be the view of all the witnesses. He refers to IBM Europe which he describes as an organisation based in Paris, which was a management and support company incorporated under French law that provided all the senior management and support functions for the whole EMEA group. But it was not a holding company. There was, according to him, a holding company, which was World Trade Europe, Middle East and Africa Corporation. The December Executive Committee Minutes are in his view minutes of a meeting of that entity. Mr James Lamb confirmed that the entity was not EMEA and explained that World Trade was a committee of senior executives in Europe. Mr Taylor said much the same: World Trade was as he put it the legal face of Paris. It was not Armonk (in the sense in which that word has been used to describe senior executives in IBM Corporation). Rather World Trade comprised some executives within the Paris structure and some executives within the Armonk structure. This is all consistent with Sir Leonard's evidence that EMEA had a presence in Armonk as well as in Paris. I am unable to resolve the status and function of World Trade and I do not consider that it is necessary to do so.

The meeting held on 14 December 1982 took place at 590 Madison Avenue New York, New York. Whatever the body actually was, the Executive Committee was clearly comprised of very senior IBM figures. The meeting was presided over by Mr KV Cassani, who was head of IBM Europe and a member of the board of Holdings. According to the December Executive Committee Minutes, the meeting was attended by, among others, Mr Kofmehl and by Mr JG Maisonrouge, a previous head of IBM Europe then working out of Armonk. Also present, as Secretary to the Committee, was Mr JS Frey, who was General Counsel for IBM Europe. The December Executive Committee Minutes record a number of resolutions of no relevance to the present proceedings. They then contain a heading "Pension Plans – Concurrence" and concerns two such plans, one in Norway and one in the UK. In relation to the UK, they record as follows:

"The Chairman presented a recommendation to improve the Pension Plan of IBM United Kingdom. Under the improved plan:

Employees aged 25 or more will contribute to the plan with 5% of their salary in excess of the Social Security lower earnings limit (now $2,500). However, in the first year of the plan, July 1, 1983 to June 30, 1984, the employee's contribution will only be 2.5%. Employees who do not wish to participate will remain covered by the current plan which is entirely financed by IBM United Kingdom.

The normal retirement age will be 63 and the earliest retirement age will be 53 for participating employees.

The normal retirement pension will be equal, for each year of contributory service, to 2.2% of the average salary in the best three of the last ten years, less 2% of the Social Security pension for a single person (now $2,524 per year). Between ages 63 and 65, a temporary supplement equal to 2% of the Social Security pension per year of service will be paid to male employees, as the Social Security pension is not payable to men before age 65.

The spouse of a deceased employee will receive a pension of 50% of the projected retirement pension.

After discussion, on motion duly made and seconded, the following resolution was unanimously adopted:

RESOLVED, that the Executive Committee hereby concurs in the proposal to improve the Pension Plan of IBM United Kingdom as presented at this meeting."

It is obvious, of course, that the meeting had before it far more than the available revision since it must have had a complete proposal; and, as already pointed out, the Available Revisions may not have been precisely what was actually in the proposal presented. It should be noted that there is nothing in the Initial Proposal or in the Available Revisions which states that employee contributions for the first year would be limited to 2.5% of the relevant earnings, again demonstrating the incompleteness of the material at the meeting. The December Executive Committee Minutes also show that, in presenting the proposal, Mr Cassani must have had a better understanding (quite possibly, I speculate, from some sort of briefing note) of the details of the proposal than would have been apparent from the Initial Proposal and the Available Revisions: for instance, the December Executive Committee Minutes record at paragraph 3 more than the completely uninformed reader of the Initial Proposal and the Available Revisions would have been able to state.

Both the Initial Proposal and the Available Revisions identify the early payment reduction under age 60. In the Initial Proposal, that is achieved by a reduction "per month early", that is to say by reference to the age at which the pension would ordinarily become payable which, under the Initial Proposal is the NRD of 60. In the Available Revisions it is achieved by a reduction "per month earlier than age 60" thus reflecting that no reduction would be made if the early pension was taken between age 60 and 63, the age of 63 being when the pension would ordinarily be payable. The December Executive Committee Minutes, however, say nothing at all about early retirement at any age and nor does it say anything about the amount or timing of the commencement of a deferred pension.

One matter on which the Initial Proposal, the Available Revisions and the December Executive Committee Minutes are all silent is the question of consent to early retirement, not only between ages 60 and 63 but also between ages 53 and 60. It is clear from the evidence, however, that nobody ever considered under any possible scenario that a member would be entitled to retire, without consent, on an immediate pension, whether or not reduced, under age 60. And so, it can be suggested that the benefit described in section 2.4 (Early Retirement) of the Initial Proposal is concerned with a benefit which is only payable with the consent of Holdings. If that is correct, then the change in NRD in the Available Revisions is, it can be suggested, similarly dealing only with benefits payable with consent. To put it another way. Section 2.4 of the Initial Proposal is dealing with an Early Retirement pension, that is to say one payable earlier than NRD (age 60 in the Initial Proposal); if a benefit is to become payable earlier than NRD, the consent of Holdings is required. That continues to be the case when, as under the Available Revisions, NRD is changed to age 63.

In this context, Mr Simmonds suggests that some help can be derived from how the Initial Proposal describes the benefits of the N Plan. There, too, is to be found in the box against "Pension formula" reference to an immediate pension but reduced by an early payment discount under age 60. In the case of the N Plan, there were existing rules to which it was possible to refer to see what the actual benefit was, namely an early retirement pension payable before NRD (60 for women, 65 for men and reduced if payable earlier than age 60 in either case) but only with the consent of Holdings. And so, it is argued, the use of a similar formula in the Initial Proposal and in the Available Revisions in describing the benefits of the C Plan shows that the benefit is payable only with consent and not as of right.

Although there is some force in that argument, it places too much reliance, in my judgment, on what is only a summary of the provisions of the N Plan and the intended C Plan as identifying all the significant components of each benefit. The summary of the N Plan early retirement benefit is accurate as far as it goes. A reader of the Pension formula for the N Plan with knowledge of the actual provisions of the N Plan could say to the draftsman that he has inaccurately described the benefit. The draftsman could respond in at least two ways: First, he could say that he did not intend to describe the benefit in detail but only wished to draw attention to the amount of the benefit when it came to be drawn; secondly, he could respond by amending his draft to include words such as "but only with the consent of Holdings". What he could not do, in my view, is say that as a matter of construction of the Pension formula for the N Plan, a consent requirement is to be implied. Rather, the benefit is misdescribed and it is necessary to look at the actual provisions of the N Plan to see what it actually is.

In the case of the proposed C Plan, there is no subsisting set of rules by reference to which the actual benefit can be ascertained. The reader of the Initial Proposal, knowing of the provisions of the N Plan and seeing that the benefit is misdescribed, might then ask the draftsman whether he intended that the proposed C Plan benefit was to be subject to a consent requirement to which the answer would clearly have been that, relating to benefits payable under age 60, it was intended to be subject to such a consent requirement. He could then give the same two answers as he could give in relation to the N Plan. But what he could not do is suggest that it was to be implied, as a matter of construction, that there was a consent requirement in relation to the C Plan. In a different scenario, his answer might have been that members should be entitled to retire without consent at age 55; and if that had been his answer, there would be nothing inconsistent with the Pension formula for the C Plan being amended by the addition of words to that effect.

When attention is switched to the revised Pension formula under the Available Revisions, it is clear that it would not be possible to read the Pension formula, again as a matter of construction and without reference to the N Plan formula, as including a consent requirement. Nor, without reference to the N Plan Pension formula would there be the slightest justification for implying a consent requirement into the description of the benefit in the Available Revisions. Although implication is now seen as little, if anything, more than an aspect of construction (see A-G of Belize v Belize Telecom Ltd [2009] 1 WLR 1988), none of the well-used tests for implication (for instance a term necessary to give efficacy to the contract) is present. As to implication, Lord Hoffmann, at [16] to [28] of his Opinion, emphasised that the object of the task of interpretation including the implication of terms is to discover what the instrument means. The implication of the term is not an addition to the instrument. It only spells out what the instrument means. And, as he points out, the proposition that the implication of a term is an exercise in the construction of the instrument as a whole is not only a matter of logic (since a court has no power to alter what the instrument means) but also well supported by authority. In every case in which it is said that some provision ought to be implied in an instrument, the question for the court is whether such a provision would spell out in express words what the instrument, read against the relevant background, would reasonably be understood to mean.

Applying that approach, there is, in my judgment, no scope for the implication into the Available Revisions of a consent requirement in relation to retirement between the ages of 60 and 63. I doubt very much that a consent requirement could be implied into the Initial Proposal in relation to the C Plan even in relation to retirement under the age of 60; but it is not necessary to decide that issue. The important point is that the wording of the Pension formula in relation to the N Plan comes nowhere near justifying the conclusion that a requirement for Holdings' consent to retire between ages 60 and 63 is to be found, whether as a matter of construction or of implication, in the Pension formula for the C Plan found in the Available Revisions. And that would be so even if it were correct that a consent requirement in respect of the C Plan is to be found in the Initial Proposal in relation to retirement under the age of 60.

In expressing those views about the Available Revisions, I have considered them in isolation from the document of which they formed part, that is to say an entire draft of some revised proposals. Clearly, any revised draft would have included significant amendments quite apart from section 2 itself. Parts of section 4 (Justification) would have been inaccurate as they appear in the Initial Proposal. Whether section 4.3.3.12 (dealing with Early Retirement) would have been amended only by replacing reference to age 60 with reference to age 63, or whether something more would have been said, it is not possible to say.

I need to make plain that at this stage I am looking at the meaning of the documents and not at what was approved at the meeting recorded in the December Executive Committee Minutes. Although formally what was approved was the revised proposal placed before the meeting, it is possible that the decision was taken on the basis that it was clear one way or the other whether there was intended to be a consent requirement: this could be because other provisions of the proposal actually before them made the position clear, or because Mr Cassani made it clear or for some other reason.

Actuarial valuation and the Cost of the C Plan improvement

It is convenient at this stage to say something about the cost estimates of the revised proposal. I have already referred to the costs estimates in the Initial Proposal and to the fact that revised estimates must have been prepared for the revised proposal which eventually went to Paris and Armonk. There cannot really be any dispute that Mr Gooch and Mrs James did in fact carry out such revised estimates or that the revised FIL of $29.241m (which can be found as part of the Company contribution figure in the available revisions) was one of the revisions which they produced.

On 9 May 1984, Mr Gooch signed the 1983 actuarial valuation report ("the 1983 AVR"), valuing the Main Fund as of 31 December 1983. The 1983 AVR should have been produced on the basis of the 1983 Trust Deed and Rules which, in their unrectified form, required the consent of Holdings to retire between ages 60 and 63. Mr Gooch's evidence is that he did value the Main Fund on that basis.

Moreover, it is clear that he assumed, in the 1983 AVR, that deferred pensioners would draw their pensions at NRD ie at age 63. However, if there was a right for an active member to retire at age 60 without the need for any consent, the Preservation Requirements would have produced a normal retirement age for those purposes of 60 in contrast with the NRD under the C Plan of 63. This would have meant that deferred members would have been entitled to draw their (unreduced) pensions at age 60. Although it is not possible to dismiss the possibility of an individual, for some special reason, deciding to defer drawing his deferred pension beyond the age of 60, it is quite simply inconceivable that it would be appropriate, in conducting a valuation of the C Plan, to make the assumption which Mr Gooch made namely that all deferred members would take their pensions at NRD. There are only two possible explanations. Either Mr Gooch did value the benefits of active members on the basis that consent was required to retire before NRD; or he overlooked the Preservation Requirements (as indeed Mrs James must have done too, and she was a stickler for detail and a very professional lady as Mr Cawley acknowledged).

As to the second of those possibilities, it is not asserted that Mr Gooch was not a professional and highly competent actuary; indeed, Mr Cawley agreed that Mr Gooch was reliable and competent. Of course, even the most competent of professionals can sometimes get things wrong, even negligently wrong. But one thing which is clear to my mind is that both Mr Gooch and Mrs James were well aware generally of the Preservation Requirements. I would take a great deal of convincing that either of them, let alone both in connection with one valuation, would have overlooked the Preservation Requirements. There is nothing to support an allegation that, in conducting the 1983 AVR, either Mr Gooch or Mrs James in fact did overlook them or failed to take account of them. The Trust Company has not produced any evidence, expert or otherwise, to suggest that the value placed on liabilities was inconsistent with an assumption that consent for early retirement was required.

Mr Gaisman also suggested that the requirement for Holdings' consent was inconsistent with the assumption that 100% of females (but not males) would retire at age 60. Mr Gooch explained that the 100% assumption for females was chosen (after discussion with IBM staff I would add) because of other factors (such as previous retirement planning in the N Plan and State pension age) which had a different impact on males and females. This might be seen as a cautious assumption but it did not entail that all females definitely would retire at 60 and indeed, as we will see, the road-show presentations of the C Plan specifically mentioned the possibility of female members staying on after age 60.

Whatever the validity of any criticisms which are made of the evidence of Mr Gooch and Mrs James, I have reached the firm conclusion that both of them were well aware at all times (including in the period leading up to the World Trade board meeting on 14 December 1982) of the Preservation Requirements and that their understanding was reflected in the 1983 AVR. I conclude that the 1983 AVR was carried out on the assumption that the consent of Holdings to early retirement between age 60 and 63 was required.

So much for the 1983 AVR. But what about the assumptions made in relation to the revised proposals (which were prepared over a year before the effective date of the 1983 AVR and even longer before the preparation of the 1983 AVR)? The 1983 AVR itself records that the C Plan costings were carried out on a similar basis to that employed in the AVR. Similar does not, of course, mean identical. Whether a valuation on the basis of no consent is similar to one on the basis of consent is perhaps a nice question. But the issue of the basis of the costings is resolved, if Mr Gooch's evidence is accepted: he says that they were carried out on the basis that there was a consent requirement. As with the 1983 AVR, and for similar reasons, I am sure that Mr Gooch and Mrs James were well aware of the Preservation Requirements when they produced the revised cost estimates. And again, as with the 1983 AVR, there is no evidence to suggest that the one costing (the FIL of $29.241m) which we do find in the Available Revisions is inconsistent with a consent requirement. Accordingly, I find as a fact that the revised cost estimates found in the available revisions were carried out on the basis that consent was required to early retirement between the ages of 60 and 63.

Whether Mr Gooch and Mrs James were justified in making the assumption which they did, and whether, and if so when, either of them communicated the existence and effect of the Preservation Requirements to Mr Cawley or Mr Ellis (or indeed to Sir Leonard) are different matters.

As to communication, a considerable degree of clarification was obtained as a result of cross-examination of Mr Gooch and Mrs James as well as Mr Cawley. Mr Simmonds now makes this concession:

It does not appear that either Mr Cawley or Mr Ellis understood in late 1982 that the effect of granting a right to active members to retire at age 60 without company consent would be to confer the right for deferred pensioners to take their deferred pensions at age 60. It is no longer contended that Mr Cawley understood the need for an employer consent requirement in order to avoid the impact of the Preservation Requirements and so it is not contended that he informed members of the Management Committee of this.

This concession means that it is not necessary to decide what, if anything, Mr Gooch and Mrs James said to Mr Cawley or Mr Ellis about the Preservation Requirements prior to the end of 1982 and, if they did say anything, whether they thought that they had adequately conveyed the consequence of the absence of a requirement for consent. Mr Simmonds, in defence of Mr Gooch and Mrs James, submits that not only were they alive to the point but also thought that they had conveyed it to Mr Cawley. I make no finding about whether they said anything at all to him about it. All that it is necessary to say is that I hold, not only as a result of Mr Simmonds' concession but also as my own finding of fact, that, whatever Mr Gooch and Mrs James did or did not say, neither Mr Cawley nor Mr Ellis was aware that the Preservation Requirements might have any impact on the revised proposals which they were developing nor were they alive to the consequences of the Preservation Requirements on those proposals. There is no suggestion that the Management Committee, and in particular Sir Leonard, were aware of the point or that they would have become aware of it other than through Mr Cawley and Mr Ellis. Accordingly, when the C Plan was announced to the managers for transmission to the workforce (which, as we will see, was on 17 December 1982), no-one within the decision-making bodies appreciated that the Preservation Requirements would apply.

The concession also makes it unnecessary to resolve the dispute about what Mr Gooch said in the meeting held with Nabarro and Mr David Newman on 14 April 2010 and what he said in the later phone call with Mr Gibson and Ms Birkitt, which I have referred to at paragraph 90 above. There are full notes of that meeting and phone call, but in neither case was Mr Gooch given the opportunity of commenting on the notes or correcting them. The Trust Company seeks to rely on two things which the notes record. Firstly, it was recorded that Mr Gooch suggested– but here it is important to appreciate that he was only speculating – that the revised costings were on the basis of an NRD for deferred members of age 63 and an NRD for actives of age 60, Clay & Partners not appreciating that the Preservation Requirements would result in an NRD for deferred members of age 60 also. Quite apart from that being only speculation, it is wrong since we now know that the costings assumed only that 75% of men would retire from active service at age 60, with the balance retiring at age 63, which would be inconsistent with the valuation assumption of an NRD of 60.

In the absence of the complete revised proposal including its narrative, it is not possible to say how it addressed the costs savings as compared with the Initial Proposal. There is, however, no suggestion that Mr Gooch and Mrs James carried out any work to arrive at estimates based on the assumption that no consent was required for early retirement between ages 60 and 63. It is reasonable to assume, therefore, that the revised proposal contained the cost estimates actually made by Mr Gooch and Mrs James that is to say on the basis that 100% of deferred members would draw their pensions at age 63 (and that is how the benefit is described in the Available Revisions) and that 75% of men and 100% of women retiring from active service would do so at age 60, with the remainder of the men retiring at age 63. But no-one apart from the actuaries would have appreciated that the latter assumption might be inconsistent with good actuarial practice or at least the view of Mr Gooch of good actuarial practice: as he said, an actuary might, but he would not, adopt such an assumption where retirement at age 60 was a matter of entitlement. The only figure which can be established from the available material is the FIL of $29.241m which Mr Gooch and Mrs James arrived at on the basis that consent was required to retire between ages 60 and 63.

The narrative continued

Returning to the meeting held on 14 December 1983, the senior executives attending that meeting must have thought that they had a real purpose to fulfil in relation to the pensions issues before them concerning Norway and the UK. They can have been fulfilling one of only two possible functions:

i) Either they were approving the revised proposal in a capacity as Paris for onward transmission to Armonk (in the sense of management groups rather than physical places) which alone would make the decision under "powers reserved".

ii) Or they were themselves giving the approval of the revised proposal required by "powers reserved" without the matter needing to go to a different decision-making body, Armonk.

They must surely have been acting in the second capacity. The conclusion which I draw from the evidence of Mr Barrett and Mr Taylor is that World Trade was neither Armonk as such or Paris as such. Rather, it, or its board, was a body of persons drawn from senior executive in both the US and Europe. This body cannot simply have been carrying out the function which it was for Paris to carry out, that is to say to submit the Initial Proposal and any revised proposal to the United States for approval. Although "powers reserved" has been described as powers reserved to Armonk, there is no reason why Armonk itself should not have been satisfied with an approval given by a group of senior executives from both Armonk and Paris. The irresistible conclusion, in my judgment, from the incomplete evidence which is available is that this body of persons was the one which was empowered to make the decision whether or not to approve the revised proposal. And this is so even though it may have been doing nothing more than had been agreed in principle between senior executives in the UK, Paris and Armonk.

Given that conclusion, it is not open to the Trust Company to suggest that Armonk in fact approved something different and to ask me to infer, because of the paucity of the evidence about what actually happened, that Armonk must have had drawn to their attention, and approved, the absence of any need for consent to early retirement between ages 60 and 63. But equally, it is not open to IBM now to suggest that Armonk in fact approved nothing or, at least, that there is no evidence that it did approve anything.

I turn now to the non-documentary evidence about what was discussed between the UK and Paris and between Paris and Armonk about the Initial Proposal and revisions to it once it had become clear that the Initial Proposal was unacceptable. The evidence of Mr Cawley is important because he was the architect of the C Plan. He himself considered that if there was something which he did not know about the C Plan, then it did not form part of it. I am not sure that is wholly correct since even he accepts that there were discussions in Paris at senior level to which he was not privy. But since it is the case of the Trust Company itself, and since IBM does not disagree, that the idea of "flexible retirement" came from Mr Cawley, what he has to say about that concept is highly significant.

However, it is not only his concept of flexible retirement as he explained it in his witness statement and the somewhat different concept which came out of his cross-examination which is significant. Of even more significance is his explanation of his thinking to Sir Leonard and other members of the Management Committee since it is their understanding, as relevant decision makers, of Mr Cawley's suggestions as adopted by them which is relevant.

Before turning to the relevant evidence, I need to emphasise what I have already made clear. It is that I use the phrase "flexible retirement" to refer to the concept of entitlement to retire between ages 60 and 63 without the need for employer consent and without any reduction. Although flexible retirement as a concept arose in connection with the revised proposals, the use of that phrase to describe the concept first appears only in the communication exercise with the workforce once the decision had been made to implement the C Plan.

Taking Mr Cawley's evidence first, he gave this account in his first witness statement. As already noted, notwithstanding the rejection of the Initial Proposal by Armonk, the concept of an equalised pension age for men and women was accepted in discussions with Paris and it was Mr Cawley's impression, at least, that Armonk had no objection to this. It was "at this juncture" as he put it, but was unable to pin down the time with any precision, that the concept of "flexible retirement" was developed by him with input from Mr Ellis. He explained the concept of "flexible retirement" as giving employees (ie active members of the C Plan) the right to retire at 60 without employer consent on an unreduced pension but with an NRD of 63 giving employees the option to continue working until 63. He says that it was recognised (I infer by the Management Committee, and by himself and Mr Ellis) that giving employees the right to retire at 60 was crucial to achieving employee attrition and workforce vitality. Giving the right to retire at 60 was necessary to enable females to retire at 60 as under the N Plan. A requirement of employer consent would have meant that women would not be entitled as of right to retire at 60 and might make them reluctant to agree to transfer to the C Plan. The absence of early retirement factors over age 60 was also necessary to maintain the attractiveness of the C Plan over the N Plan which only imposed reduction factors before age 60.

Mr Cawley went on to say that this "flexible retirement" concept was discussed with Sir Edwin, Sir Leonard and Mr Morgans who all supported and approved it as an important element of the C Plan. However, he did not recall whether the concept was formally submitted to Armonk. His perception was that the US management were concerned with ensuring that NRD was not too low. He explained that, in any case, discussions went on informally with his superiors, such as Sir Edwin, so that he may not have been party to all of the thinking and debate on the point. The clear outcome, he says, was that senior UK management had decided on the "flexible retirement" by early December 1982, adding that "At the conclusion of the discussions with Armonk and Paris in early December 1982, I recall that Sir Edwin Nixon received confirmation that the C Plan was approved". In his second witness statement he said that he clearly recollected that Sir Edwin had received confirmation from Mr Kofmehl approving the C Plan. Mr Cawley recollects a handwritten note in which Sir Edwin had congratulated the team ("well done chaps" or words to that effect). He clearly remembers Mr Kofmehl's letter as a milestone in being able to forward the C Plan project.

Mr Cawley also refers to a feature of the C Plan which the US wanted to be incorporated into the C Plan. This was a scheme cap on benefits. It was close to the then Inland Revenue limits but was scheme specific and intended to provide a maximum benefit of two-thirds of the member's final pensionable earnings during his or her last 12 months of service, an amount which would usually be less that the Inland Revenue maximum. The calculation of the benefit cap (which became known as the C Plan Maximum) differentiated between retirement from ages 60 to 63 and retirement prior to age 60, with a clear intention that there would be a difference in the benefit design for the two categories.

Mr Cawley was cross-examined at some length by Mr Simmonds on his understanding of "flexible retirement". He submits that the conclusion to be drawn is that in Mr Cawley's mind there was a clear distinction between flexible retirement as it was likely to have been communicated to Paris/Armonk and flexible retirement as communicated to the UK workforce after approval had been granted. Mr Simmonds interprets Mr Cawley's evidence in this way: Mr Cawley saw the core element of flexible retirement as being an unreduced benefit payable at age 60; the issue of whether company consent was required was essentially a matter to be communicated to the UK workforce but was not part of the concept. The importance of this interpretation is that it opens up the possibility that the approval of the C Plan was given without Armonk and Paris being aware that members would have a right to retire from age 60 without the consent of the employer. What Armonk and Paris may well have understood was that a person retiring after age 60 would receive an unreduced pension and that was the core content of "flexible retirement"; the absence of a need for consent was only the way in which the C Plan would be sold to employees.

There are certainly answers given by Mr Cawley to Mr Simmonds' questions which suggest that the issue of consent might not have been discussed with Paris or Armonk. In IBM's closing, over 5 pages are devoted to setting out questions and answers to demonstrate this. It must be borne in mind that the answers on which Mr Simmonds relies must be read in the context of longer answers or other answers which Mr Cawley did not see as inconsistent in any way. I do not propose to repeat what is set out in IBM's closing. I do, however, need to say something about this part of his evidence. What is set out in IBM's closing really falls into two sections, the first dealing with what was discussed by Mr Cawley with Paris and internally in the UK and the second dealing with what he meant by "flexible retirement". The questions and answers in the first section used the phrase "flexible retirement" in which context Mr Simmonds clearly included both elements ie retirement without consent and retirement without reduction. Mr Cawley's answers only make sense if that is how the phrase is to be understood. On that basis, the following emerged:

i) Mr Cawley acknowledged that a right to retire without consent after age 60 was a separate matter from retiring after that age without any actuarial reduction.

ii) Although at one point he stated that flexible retirement was "not part of the lexicon in terms of discussion with Paris" and that it came up in the context of selling the plan to employees, yet what was discussed with Paris was normal retirement stretching over an age range "allowing individuals to choose whether they want to retire at 60, 61, 62 or 63".

iii) He accepted, of course, that the table in the Available Revisions did not refer to the consent aspect, but noted that the narrative (which was not available) might have explained matters. Mr Simmonds observed that the narrative would have reflected the table, an observation with which Mr Cawley agreed. I do not think that that observation really takes matters further. On any footing, the table is incomplete – it does not for instance refer to the reduced employee contribution in the first year to 2.5% - and the fact that a particular item does not appear does not mean that it was not discussed.

iv) He accepted that it was quite possible that the consent aspect never got into the revised proposal at all. Whilst myself acknowledging that that was a possibility, I would add that the absence of any express mention of consent either way does not conclude matters in favour of IBM or even assist it much.

v) But whilst it may not have got into the revised proposal, Mr Cawley's evidence was that it was certainly discussed by him at staff level and he imagined by Sir Edwin and Sir Leonard at their level in discussions with Paris. In this context, it is worth setting out part of what Mr Cawley said in that respect:

"A. ……But at staff level, as I said, the discussions I had were that this window, which we were calling, really, "normal retirement" for the provisions of a benefit specialist that would understand that, had this range of ages. And that was, I think -- as I said yesterday, included in that discussion was, "Well, why should 60 be a pivotal age under this new regime? What is the rationale for having it as a pivotal age?" You didn't need to have it. And when I explained the rationale -- and we discussed yesterday the maximum pension opportunities that people would have, between 60 and 63, and the female issue of encouraging them to come into the scheme without actually having to have consent and having an actual reduction, Paris staff accepted that as a very logical approach.

The discussion as to -- they didn't then say to me, "Ah, but do you mean that in every case you are going to have to get company consent?" I don't remember that discussion. They didn't ask me that question. Whether Mr Peach and Mr Nixon had that question asked of them, I couldn't comment."

Q. So on the basis of your knowledge, it may be that the consent issue never crystallised at that stage?

A. It may be. It wasn't necessarily crystallised with me in terms of my staff liaison. I don't know about the other –"

vi) The actual phrase "flexible retirement" does not appear to be one which was used at this stage ie in November 1982 and up to the date of the World Trade Executive Committee meeting on 14 December 1982. It was not always clear, to me at least, that when Mr Cawley spoke about the absence of discussions about flexible retirement, he was talking about the label or the substance.

It seems to me that what Mr Cawley was saying in this part of his testimony was that the discussion with Paris was on the basis that there would be a window during which an employee would be able to retire the rationale for which was as he had explained. It may be that the issue of consent was not expressly addressed. But one thing would have been obvious without being stated which is that consent to retire at age 63 would not be required. And in the context of a discussion about a range, or window, of retirement ages, one might think that no distinction was to be drawn between the retirement at one age rather than another. Further, judged against the rationale which Mr Cawley identified and discussed with Paris, it is not easy to see how consent to retirement between ages 60 and 63 would have been understood to be a requirement imposed before a member could exercise his or her choice.

That is particularly important in relation to females whose NRD under the N Plan was 60. It is clear that one concern of Mr Cawley and Sir Leonard, and indeed of the Management Committee, was that females would retain the right to retire at 60. This indeed was one of the reasons given in the Initial Proposal for adopting an NRD of 60. There was also the accompanying concern that unless that right was preserved, there might not be the necessary uptake of membership by females. It borders on the inconceivable, to my mind, that the Paris staff who had been involved in considering the Initial Proposal before its submission to Armonk and involved in considering what modifications needed to be made would not have appreciated the importance of this point or have thought that it was being abandoned in the revised proposal. The other aspect which they cannot possibly have thought was being abandoned was the equal treatment of men and women. I cannot take Mr Cawley's evidence as an acceptance that, because the issue of consent may not have been expressly discussed as a separate item with Paris in relation to the revised proposal, Paris might have misunderstood the intention of Sir Edwin, Sir Leonard and Mr Cawley and have thought that Holdings could refuse to allow a C Plan member to retire at any age between 60 and 63.

The second section of the part of Mr Cawley's evidence which I am presently addressing is where he dealt with his understanding of precisely what flexible retirement involved. Up until this point in the litigation, everyone, so far as I know, had been using the phrase to mean retirement between ages 60 and 63 without the need for any consent and without any reduction in benefit. At one stage, Mr Simmonds asked a question which provoked an intervention by Mr Gaisman saying that Mr Cawley had referred to "flexible retirement" as a label but Mr Simmonds was using it as a concept. The following exchange took place between Mr Simmonds and Mr Cawley:

"Q. I have made it perfectly clear what I mean. I mean the right to go at 60 and without company consent. And you understand that, do not you [sic], Mr Cawley?

A. Well, no, I'm not sure I do understand that from the point of view that -- flexible retirement means that between 60 and 63, that is a spread of normal retirement ages. Whether it's by right or whether it's by company consent is a different situation, in the way that I looked at it, and it becomes a communication situation.

You can either say, as we did, as part of the sale of the plan, "Look, we are empowering you". It's actually happening. Whether it's company consent or not, it's happening. So between 60 and 63, if you go, then you will get an unreduced pension.

A separate issue to that is: we empowered individuals within the communications that you could choose to go between 60 and 63. The issue is: should we have layered over that a company consent clause? That would have meant that in the communications we would have had that sort of umbrella over it. And frankly, if we had known at the time that we were going to fall foul of potential regulations and so on then I believe the UK company would have said, "Look, come up with a form of words that will still make the flexible retirement concept attractive, but we need to put some kind of caveat on that, probably for exceptional business reasons".

So from a communications point of view, we could have got round the problem without too much of an issue. And I know speculation perhaps is not something that I should indulge in, but I suspect if we would have been forced to go down that route we would have probably still seen the same kind of take-up as we saw -- because people would have trusted the company. They would have said, "Fine, you know, let's in fact accept that, for good business reasons, there might have been a company consent put over -- for that lid (sic), but we trust the company, we know what they are trying to do, we will still join the C Plan"…..

……

So the flexible retirement and then what you do with it in terms of consent or otherwise, they are two separate issues, in my view.

Q. There is an important point that comes out of that answer, Mr Cawley.

I put it to you in relation to your evidence of what happened in or around November 1983, that you saw no material distinction between a formal company consent requirement on the one hand, where the company was going to give consent and a so-called right, a legal right, of flexible retirement?

A. Yes.

Q. And I think you agreed with that, and your explanation of the thought process that went into the concept back in 1982 seems to be based on that same approach --

A. Yes.

Q. -- that actually -- this has become a big issue in these proceedings but at the time you didn't see any difference between having a formal company consent requirement and not having one?

A. Oh, I did see a difference. It was on the saleability of the scheme. I mean, I felt much more comfortable standing up in front of people and saying, "Look, here is great news. You are being empowered by the company to choose when you decide to retire between 60 and 63."

If I had had to stand up and say, "The flexible retirement concept is being introduced to you, but recognise there is still a company right to actually say yes or no when you come up to 60 or 63" rather dampens the saleability of that, but it doesn't make it a showstopper from the point of view we wouldn't have still gone ahead with flexible retirement.

We would simply have said, "As a matter of legal formality, we have to say to you there has to be company consent and therefore we will have a process through line management where that consent will need to be ticked off", but people trusted the company and the company knew full well that if we were offering something and we were telling them something then it would actually be met."

In relation to this topic, I put it to him that some of his answers suggested that he might consider flexible retirement to exist even where there was a requirement for consent, with which he agreed.

Mr Simmonds relies on these answers to show that it was no part of Mr Cawley's thinking that company consent to retirement between ages 60 and 63 would not be required: thus it was not part of the benefit structure that consent would not be required, it was simply part of a communications exercise with the workforce.

I do not accept that submission or anything like it. It is quite clear that Mr Cawley did see a difference between having a formal company consent since as he put it the difference "was on the saleability of the scheme". So whether or not consent was an aspect of "flexible retirement", consent was, to Mr Cawley, an important aspect of the C Plan. It is true that he speculated what might have happened if it had been realised that his concept would have fallen found of the Preservation Requirements; some way round the problem would have been found by the use, to adopt a phrase of my own, some mealy-mouthed drafting. But the fact of the matter is that no such problem was remotely in his mind. So, in seeing an important difference in terms of saleability, he must have seen the C Plan concept (whatever his concept of "flexible retirement") as involving early retirement between ages 60 and 63 without the need for consent.

I should add this. As will be seen later, the C Plan was in fact promoted to the workforce on the basis of a right to take early retirement between ages 60 and 63 without employer consent. That promotion involved a number of documents and road shows which I will describe in some detail. The preparation of some of those documents, in particular the announcements in December 1982 following closely on from the approval of the C Plan, were drafted on the basis of "no consent". It must have been known before approval was obtained on 14 December 1982 what the announcements would say. Mr Cawley was involved in the drafting of these announcements or certainly knew of their contents. There is no suggestion, and nor could there be, that Mr Cawley was deliberately allowing one thing to be said to Paris and Armonk which was at best ambiguous (ie a revised proposal intentionally refraining from mentioning the aspect of consent) whilst preparing communications (ie announcements providing for an early retirement without consent) and then allowing another thing to be said to the work force.

I therefore conclude that Mr Cawley's intentions prior to 14 December 1982 were consistently that the C Plan would provide for early retirement between ages 60 and 63 without the consent of Holdings or the employers.

I now turn to Sir Leonard's evidence about his concept of flexible retirement in the period leading up to 14 December 1982. I have already referred to Mr Cawley's evidence that the "flexible retirement" concept was discussed with Sir Edwin, Sir Leonard and Mr Morgans who all supported and approved it as an important element of the C Plan. That was contained in Mr Cawley's first witness statement where he had explicitly used "flexible retirement" to include the right to early retirement between ages 60 and 63. In the light of what he said in cross-examination about his concept of flexible retirement, it is not entirely clear from his evidence that he expressly mentioned the absence of any need for employer consent to his superiors.

Sir Leonard's evidence as contained in his first witness statement on that aspect is clear. He states unequivocally in paragraph 42 that the final proposal for the C Plan, as adopted by the Management Committee on behalf of Holdings, was that NRD would be 63 "but that retirement between 60 and 63 was available without either consent or actuarial reduction". Later in that witness statement, he dealt with the discussions with Paris and I have already set out what he said in paragraph 56 and 57. In paragraph 58, he explained the relevant retirement and early retirement provisions of the N Plan (NRD females 60, males 65 with no early retirement reduction for males retiring between 60 and 65). He had already, in paragraph 54.4, identified, in explaining the Initial Proposal, the need for take-up of the C Plan to be as high as possible, moving the NRD for females to an age greater than 60 would run counter to that and the need to treat males and females in the same way meant that the ability to retire as of right at age 60 was to be extended to males. Accordingly, these factors (the N Plan structure and the matters identified in paragraph 54.4) meant that it would be necessary to maintain a corridor of retirement between 60 and 63 with no actuarial reduction and no consent requirement. This was the so-called "flexible retirement" which he had described earlier at paragraph 23.

Sir Leonard explained that the solution had emerged although he does not anywhere say that the phrase "flexible retirement" was, at least prior to submission of the revised proposal to Paris, ever used. But he was clear that the solution – retirement with no actuarial reduction and no employer consent – did emerge before that time and that once it had emerged it obtained the full consent of Sir Edwin and Mr Morgans.

Sir Leonard was not undermined in this aspect of his evidence in cross-examination. Indeed, the factors which he identified in paragraph 58 of his first witness statement are very strongly supportive of the conclusion that he, at least, intended the C Plan to allow for retirement between ages 60 and 63 without actuarial reduction and without employer consent. One thing which I am perfectly certain that Sir Leonard did not contemplate was the removal of the right of a female to retire at age 60 as under the N Plan. And given his clear intention that men and women should be treated equally, the inevitable conclusion is that he, at least, intended the C Plan to provide the same right for men to retire at any age from 60 without consent.

I do not have the benefit of evidence from Sir Edwin. But I have no reason to doubt, and every reason to think in the light of the evidence of Sir Leonard and Mr Cawley, that his position was no different from that of Sir Leonard himself.

Sir Leonard's first witness statement in relation to discussions with Paris is not so categorical. In it, he said that it was Sir Edwin who dealt with much of the high-level discussion with Paris (often on the telephone). Although he is sure that "the necessary agreements were obtained", that is to beg the question of precisely what agreement was sought. He explains that Mr Cawley was not a party to all of the discussions held with Paris by Sir Edwin or Sir Leonard, and this may be the reason why Mr Cawley was unable to remember whether Paris and Armonk specifically agreed to the "flexible retirement" aspect of the C Plan. The implication of that is that Paris and Armonk did specifically agree to it, but I am unable to take Sir Leonard as giving first-hand evidence that either he remembers doing so himself or that Sir Edwin told him that he, Sir Edwin, had done so.

But even if the question of a requirement, or lack or requirement, for consent to retirement between ages 60 and 63 was not expressly discussed with Paris, I would adopt what I said in relation to Mr Cawley's discussions at staff level. It borders on the inconceivable, to my mind, that the Paris executives, including Mr Kofmehl, who had been involved in considering the Initial Proposal before its submission to Armonk and involved in considering what modifications needed to be made, would not have appreciated the importance of the need to retain the right for women to retire as of right at age 60 or would have thought that it was being abandoned in the revised proposal. Nor can they possibly have thought that there was to be a move away from the equal treatment of men and women. I do not consider that Paris might have misunderstood the intentions of Sir Edwin, Sir Leonard and Mr Cawley and have thought that Holdings could refuse to allow a C Plan member to retire at any age between 60 and 63.

Mr Ellis' evidence is consistent with and supportive of that conclusion. He was heavily involved in the development of the Initial Proposal. He was less involved in the modifications which followed on from the rejection by Armonk of the Initial Proposal. Mr Cawley took a greater lead in the revision of the Initial Proposal while he, Mr Ellis, was more involved in developing the anticipated communications programme. He did not draft a revised proposal but had some involvement in its production. In particular, he had discussions with Mr Cawley about NRD and the features around it. He agreed with Mr Cawley that the flexible retirement concept was developed by Mr Cawley with input from him. In that context, I should add, flexible retirement was being used in the sense which involved a right to retire without consent. He was clear that the revised proposal as developed by him involved flexible retirement in that sense. He did not remember any consent requirement ever being mentioned by Mr Gooch or Mrs James or other advisers (for instance MPA): if it had been mentioned, it would have been important because it would have cut across a central feature of the C Plan as developed by Mr Cawley and himself and he would surely have remembered it.

I accept all of that evidence from which it follows that the revised proposal developed by Mr Cawley and Mr Ellis included a right to retire between ages 60 and 63 without employer consent.

Mr Ellis also thought that the proposal put to Paris included flexible retirement in that sense and that the eventual approval was of a proposal which involved flexible retirement in that sense, such approval being that of the UK (ie the Management Committee), Paris and Armonk. However, he was not personally involved in the detail of precisely what was sent to Paris nor could he know of his own knowledge the content of the discussions between senior UK management and senior Paris management nor of the discussions between Mr Cawley and his counterparts in Paris at meetings which he did not attend.

As to Mr Barrett, I have given his background and involvement at paragraph 77ff above in the section of this judgment dealing with the witnesses and have set out the material parts of his letter dated 22 November 2009 to Tony Ford. Mr Barrett pointed out that the letter was not prepared for the purposes of litigation; indeed, he says he did not know that litigation was on the cards and thought that he was providing information to help Mr Ford in a dispute which he was having with management. However, he adopted the letter as his evidence in his examination-in-chief. He may have regretted adopting such an easy approach because Mr Simmonds was able to confront him with a number of matters in respect of which what was by then his testimony was inaccurate. Accepting that, and accepting that Mr Barrett's memory is not as good as he would like to think, he is, in my view, more likely to be accurate rather than inaccurate in his evidence about flexible retirement.

What is missing from Mr Barrett's evidence is how he himself was instructed in relation to the C Plan. He did not explain to the court, since he was not asked, who sent him information. He did not explain to the court, again he was not asked, who instructed him to consult with Mr Jack Frey (General Counsel for IBM Europe) or the purpose of his discussions with him. Neither Mr Cawley nor Mr Ellis mentions him in their witness statements nor do they advert to any involvement by IBM's legal department. I must therefore treat his evidence with a certain degree of circumspection.

He stated that he recalled two elements of "our plan". He identified a number of the people referred to by "our" including Mr Cawley and Mr Ellis. That plan involved two key elements, the ability of members to retire at 60 or at least in one's early 60's, with no reduced benefits and the ability to retire early in one's 50's, albeit with reduced benefit. The former was to be as of right, the latter discretionary. He referred, in his letter, to the concept of flexible retirement (ie including the right to retire between 60 and 63 without employer consent) and stated that it was part of his remit as the relevant IBM lawyer to ensure that the concept of flexible retirement was legally possible. He says that he discussed the C Plan at meetings in Paris; and that he discussed it at length with the general counsel of IBM Europe (I think this was a reference to EMEA), Mr Frey, and his deputy, Ian Mavor. He remembers going to meetings with them in Paris where senior executives, including Mr Cassani, were present. His evidence is that the following were people who knew about the flexible retirement concept: Sir Edwin, Sir Leonard, Mr Morgans, Mr Cawley, Mr Caddick, Mr Taylor, Mr Tony Green (Mr Barrett's boss), Mr Cassani and Mr Kofmehl. In fact, Mr Caddick, according to his own evidence, was not involved so, to that extent at least, Mr Barrett must be wrong. Moreover, Mr Barrett's evidence left me with no clear understanding of how he knew that these individuals were aware that the C Plan was intended to provide for flexible retirement in the sense described by Mr Barrett other than the fact that they were involved in C Plan development to a greater or lesser extent; nor was I left with a clear understanding of when all of these people were said to be aware for the first time of the inclusion of flexible retirement and in particular whether it was before or after approval of the C Plan in one form or another by Armonk. What can be said, however, is that there is nothing at all in his evidence to suggest that anyone concerned at any stage ever had a positive intention that retirement prior to age 63 should be allowed only with employer consent.

Mr Barrett was asked by Mr Gaisman whether there was a document which embodied revisions to the Initial Proposal. His recollection that that there was a full slide presentation. He thought that there was a revision to the original document put together by Mr Cawley and Mr Ellis. And he professed to remember various supporting documents, one of which was an opinion written by him which went out under the name of his superior, Mr Green. He was not sure whether the opinion referred in express terms to a right to retire without consent. It is another curious feature of the evidence overall that nobody else mentions such a slide presentation (although there was one later in connection with the road shows) nor such an opinion.

Although Sir Leonard, Mr Cawley, Mr Ellis and Mr Barrett all considered that cultural differences between the US and the UK formed at least one of the reasons for rejection of an NRD of age 60, the emphasis of the first two of them was different from that of Mr Barrett. I am not clear what, in fact, Mr Ellis considered the difference to be. Sir Leonard and Mr Cawley considered that the US objection was that employees should not be compelled to retire too young and that a reduction (at least in the case of men) from age 65 (under the N Plan) to age 60 (under the Initial Proposal) was too large a reduction. In other words, and to use my own words, the concern was to protect the individual employee who might wish to serve to a greater age. In contrast, Mr Barrett saw the US culture as being to avoid encouraging employees to retire at a young age.

It might then be said, as Mr Simmonds would argue, to be inconsistent with that to give members the right to retire at 60 on unreduced pensions even if they could choose to serve to a later age. Indeed, Mr Simmonds asked Mr Barrett a question directed specifically at that point. Mr Barrett then speculated. His explanation was given in terms of "smoke and mirrors"; but that is not as sinister as it may sound. It was simply Mr Barrett's way of describing a compromise under which IBM could achieve its business objectives without parading an NRD of 60. There may be something in that. After all, as Sir Leonard drew attention to, IBM was profitable and there were very good relations between UK management and US management. Everyone concerned would have wanted to reach an agreed solution.

I need to remember, however, that I have no firm evidence about these competing cultures whether at a general level (US as compared with UK) or at a group specific level (Armonk as compared with IBM in the UK). If I had to choose, I consider that Sir Leonard's views carry far more weight than those of Mr Barrett. He was experienced in business at a high level and had worked both in the UK and Paris. His field of expertise was in human resources. In contrast, Mr Barrett was a comparatively young and inexperienced employee whose expertise was legal. But having no firm evidence about what Armonk's reasons for objecting to an NRD of 60 were, I cannot accept the argument that a right to retire between ages 60 and 63 would run counter to Armonk's culture.

As to cost saving being one of the drivers for the refusal by Armonk of approval of the Initial Proposal, I have no way of being certain that that is so. Mr Simmonds can, I accept, find some support from that in the modification of the benefits contained in the Initial Proposal to make them less generous (a reduction in the accrual rate, the replacement of guaranteed increases to pensions in payment with discretionary increases, reduction in commutation rates and removal of non-statutory revaluation of deferred pensions in the period of deferral). Why, it might be asked, should this be done unless the benefits cost too much? There may be an answer to that, but without the help of expert evidence it cannot be said what the answer is or, equally, that there is not an answer. This is not a question of a burden of proof. Mr Simmonds raises the question to make a forensic point, but he had not laid the ground for it or given an opportunity to the Trust Company to meet it with evidence. It may be the case, I do not know, that the change from 2.5% FPE to 2% FPE for the accrual rate is accounted for by the extra 3 years of potential service from age 25 to the latest age, 63, to which a member may serve. It may be that the change from 60 to 63 as the maximum age for ordinary retirement accounts for a change in the commutation rates. It may be that removal of non-statutory revaluation of deferred pensions reflected a cultural attitude to early leavers. And it may be that concerns about the risk of the cost of guaranteed increases in pensions escalating with no mechanism to control those increases was the reason for removing the guaranteed increase, or perhaps it was, again, a cultural issue. But I simply do not know whether any of this speculation is correct. I do not, therefore, decide whether cost was or was not a reason for refusal to approve the Initial Proposal.

My conclusions on the facts up to and including the meeting of World Trade on 14 December 1982, in the light of all the evidence and argument, are as follows:

i) All those within IBM in the UK with any significant degree of involvement in the preparation of the Initial Proposal and the revised proposal (including in particular Sir Leonard, Mr Cawley and Mr Ellis) as well as some not so closely involved (including in particular Sir Edwin and Mr Morgans) understood that the revised proposal as discussed with and submitted to Paris intended the C Plan to provide for retirement between the ages of 60 and 63 for men and women without a requirement for consent from the employer and without actuarial reduction.

ii) That aspect of the C Plan was not simply locked in the minds of those persons but was expressly discussed between Sir Leonard, Mr Cawley and Mr Ellis and effectively communicated to Sir Edwin and Mr Morgans.

iii) There were discussions in Paris at staff level (between Mr Cawley and his staff counterpart). It is quite probable that the term "flexible retirement" was not used and even possible that the question of consent to retirement between ages 60 and 63 was not expressly raised as a point for discussion. Nonetheless, in the light of the various explanations of, and rationale for, the C Plan given in the Initial Proposal, it must have been the understanding of Paris, at staff level, (i) that women would retain an entitlement to retire at age 60 and (ii) that men were to be treated in the same way as women. Accordingly, even if this did not dawn on the Paris staff, the discussions resulting in the revised proposal, and the proposal itself, could only have been understood as providing an entitlement to retire between ages 60 and 63.

iv) There were discussions between Mr Barrett and Mr Frey, General Counsel for IBM Europe. I think it more likely than not that the discussions concerning the benefit structure did address whether members would be able to retire from age 60 without consent, but even if that was not the case, Mr Frey could not simply have assumed that consent would be required given the contents of the Initial Proposal.

v) There were discussions at senior level involving variously Sir Edwin and Sir Leonard on the one hand and Mr Cassani and Mr Kofmehl (and possibly others) on the other hand. In those discussions, Sir Edwin and Sir Leonard certainly understood that the revised proposal would include an entitlement for members to retire without consent between ages 60 and 63. It seems to me to be unlikely that the counter-parties to the discussions did not also understand that to be the case. The revised proposal did not come out of thin air but was a response to the refusal of Armonk to approve an NRD of 60 (and possibly to other objections if there were any). As I have said more than once already, essential elements of the Initial Proposal were that women would not lose their entitlement to retire at 60 and that men and women would be treated equally. It is inconceivable to my mind that if there was to be a departure from either of those two elements, there would not have been a full discussion about it. The central characters, Sir Leonard, Mr Cawley and Mr Ellis, could not all have failed to recollect discussions to that effect. And yet there was not a hint of any such discussion in their evidence. I am certain that if any of them had remembered such a discussion, they would have said so in their evidence.

vi) Although I do not need to attach any weight to Mr Barrett's evidence in reaching those conclusions, what he said is entirely consistent with them. I do, however, attach some weight to his evidence about his discussions with Mr Frey and his attendance at meetings in Paris with him and senior executives. What he says is confirmatory evidence of what is, in any event, obvious, namely that the C Plan received serious and detailed consideration at all levels. It seems to me virtually impossible that, after full and detailed discussion, including discussion with Sir Edwin and Sir Leonard at high level and with Mr Cawley at staff level, Paris could have had a positive understanding that members could retire between ages 60 and 63 only with the consent of the employer. Quite the reverse. I consider it more likely than not that Paris well understood that consent was not required. It is, I suppose, just possible that the question was not addressed at all; but if that were so and someone had asked "By the way, does a member have an entitlement to retire from ages 60 to 63 without any consent?", the answer could only have been "Of course".

There is, alas, a total absence of any evidence, except the December Executive Committee Minutes, of what happened in the US. It is inconceivable – I use that word again – that there were no discussions between Armonk (at whatever was the appropriate level of management) and Paris about the revised proposal. It would, to my mind, be surprising if Paris did not communicate to Armonk all the important features of the revised proposal including retention of the right for women to retire at 60 and the equal treatment of men and women. Women would therefore be able to retire from ages 60 to 63 and the same would therefore apply to men. It would be particularly surprising given that Mr Cassani and Mr Kofmehl were involved in the Paris discussions and must have appreciated this aspect of the C Plan; and it appears that both of them were involved in consideration of the revised proposal in the US.

The evidence, which I accept, of Sir Leonard and Mr Cawley is that approval of the C Plan was communicated to Sir Edwin by Paris. According to Mr Cawley, and I see no reason to doubt this, approval came in the letter from Mr Kofmehl which, I understand, came after the meeting of World Trade on 14 December 1982. Clearly Sir Edwin understood that approval to be of the revised proposal including the right to retire from age 60. And given that Mr Kofmehl knew that the revised proposal included that right, he must have intended to communicate such an approval.

What we do not know is how Armonk itself made the decision to give approval. Were it not for the December 1982 Committee Minutes, there would be a total absence of evidence on the point, and the only possible inference from the communication of the revised proposal by Paris and the subsequent communication of approval by Mr Kofmehl to Sir Edwin, would be that Armonk had indeed approved the C Plan including the right to retire from age 60.

However, we do have those Minutes. In my judgment, they do not detract from the conclusion that Armonk did give approval to the C Plan including the right to retire from age 60. There are two scenarios: either World Trade was the body whose formal consent was the necessary consent under "powers reserved" or it was not. If it was not, then the position is as stated in the preceding paragraph. If it was, then it needs to be determined precisely what was decided. It does not appear from those Minutes that there was detailed consideration of a lengthy document such as the revised proposal. If there had been, one might have expected to see a rather fuller minute. While not suggesting that the meeting was simply there to rubber-stamp any proposal, it does not appear from the Minutes or, indeed given the 2.00 pm start and the number of items on the agenda, from the nature of the meeting that the meeting would be doing anything than giving formal approval to matters which had been considered by others.

Thus we see present at the meeting both Mr Cassani and Mr Kofmehl who, for reasons already given, were I consider well aware of the right to retire from age 60 without consent. Mr Cassani is reported as having presented a recommendation – whether a full revised proposal was before the meeting is not recorded – and it is to be presumed in the absence of any contrary evidence that he presented to the meeting all the features of the C Plan which he considered relevant to the giving of approval even if not recorded in the Minutes. One might think that he would have said something about cost, but nothing is recorded. One might think that he would have said something about the absence of actuarial reduction if retirement took place from age 60, but nothing is recorded. Nothing, therefore, can be read into the absence from the minute of mention of retirement from age 60 being an entitlement rather than discretionary. And, as I have mentioned, the fact that the recommendation as presented included an element (that is to say an initial contribution rate by members of only 2.5%) which was almost certainly not contained in the lengthy revised proposal shows that the absence of express mention of an entitlement to retire after age 60 is certainly not conclusive, and in my view not even an indicator, that consent was required.

I note also that Mr Frey was present at the meeting. He too, like Mr Cassani and Mr Kofmehl, had been involved in discussions in Paris (as Mr Barrett relates) and understood, as I have found, the intention that men and women would have an entitlement to retire at any age between 60 and 63 without the need for any consent. He surely would have corrected any impression to the contrary which might have been held by any of those attending the meeting.

My conclusion, therefore, is that Armonk did approve the C Plan including an entitlement for male and female members to retire at any age between 60 and 63 without the consent of Holdings or of the member's actual employer and without actuarial reduction.

But even if that is wrong, it is clear, in my judgment, that the revised proposal was put to Paris on the basis of such an entitlement and was understood by Paris to contain such an entitlement. In communicating approval of the C Plan to Sir Edwin, Mr Kofmehl was signifying approval of that which had been submitted to Paris and thus recognising that the C Plan could properly include that entitlement. For reasons which I have already given, that approval was obtained in accordance with "powers reserved" and was therefore an effective approval authorising IBM in the UK to implement the C Plan including that entitlement.

I should add here that I find as a fact, on the basis of the evidence of Sir Leonard and Mr Cawley, that Mr Kofmehl did communicate approval of the C Plan to Sir Edwin following the meeting in New York on 14 December to which I have referred. I also find as a fact that the C Plan as thus approved included a right to retire between ages 60 and 63 without employer consent and without actuarial reduction.

I have not mentioned, but should for completeness do so, the position of the Management Committee. I have no doubt whatsoever that Sir Edwin and Sir Leonard were acting in relation to the Initial Proposal and the revised proposal, and in their discussions with Paris, with the knowledge and approval of the Management Committee. Mr Cawley and Mr Ellis were also acting within the scope of their authority as employees answerable to Sir Leonard in acting as they did. It is not possible, at this distance of time, to give chapter and verse of meetings or discussions which demonstrate the knowledge and approval of the Management Committee. But that is Sir Leonard's clear evidence. It is also clear that Mr Morgans approved of what was happening as did Sir Edwin who was in fact personally involved. The members of the Management Committee were in regular contact as executive managers of IBM in the UK. It is impossible to think that what was being done in relation to the C Plan was being done other than with full Management Committee approval.

As to the involvement of the Trust Company in the creation of the C Plan, I have already referred to Mr Taylor's evidence (which I have accepted) about the approach of the Trust Company to its duties. In accordance with that approach, the policy decision to introduce the C Plan and the preparation of the Initial Proposal and the revised proposals were matters for Holdings/the Management Committee and the employers rather than for the Trust Company. The Trust Company was reactive. As it happens, it was willing to approve the revised proposals. The implementation of those proposals required the Trust Company to make the necessary amendments to the Main Plan with the consent of Holdings, a matter which was not dealt with until much later, in December 1983. Nonetheless, the approval of the Trust Company was needed before the programme of announcements and communication could be commenced.

This approval was given at a meeting of the board of the Trust Company in Portsmouth on 17 December 1982. Present, among others, were Sir Edwin, Mr Morgans, Sir Leonard and Mr Taylor. Mr Gooch and Mr Cawley were in attendance. Sir Edwin and Sir Leonard, of course, knew all about the intended provisions of the C Plan and the revised proposals. For the benefit of others, Mr Cawley gave an explanation. The minutes record that he gave details of changes to an improvement in the Main Plan resulting from the (future) introduction of a contributory scheme in 1983. He outlined the intended introductory programme. The minute records that the improvements and the arrangements were noted. It is also recorded that the effect on Main Plan funding would depend on take-up into the new C Plan and that a further report on this would be made at the next meeting (although it does not appear that this was done). The minute does not record any detail whatsoever about the changes nor does it even record approval of the board to the introduction of the C Plan, although it is not suggested that approval was not given. Mr Cawley's evidence, which I accept on this point, is that one of the key elements he described was "the flexible retirement feature of the C Plan, whereby employees would have the right to retire from age 60 to 63 without employer consent on an unreduced pension". He is there, of course, using the term flexible retirement in its wide sense and not in the more limited sense which I have discussed above in considering some of his oral testimony.

Before moving on to what happened after approval had been received by Sir Edwin from Mr Kofmehl, it needs to be remembered what everyone involved up to this time intended the benefits of early leavers were to be. Under the Initial Proposal, deferred benefits would be payable at age 60 because that was the proposed NRD. When NRD was moved to age 63, it is clear that the intention was that deferred pensions should be paid at age 63 and not before (at least not without consent and not without actuarial reduction for early payment). As we will see, that is the basis on which the C Plan was presented to the workforce. This creates a serious tension when it comes to rectification of the 1983 Trust Deed and Rules (which I will need to address later) since rectification to bring about what the Trust Company maintains was the intention of the parties (ie an entitlement retire at any age between 60 and 63) will bring about a result, by reason of the Preservation Requirements as they then stood, which was contrary to the intention of the parties (ie an intention that deferred benefits should not be payable until age 63). An issue will be whether rectification should be granted in those circumstances.

Implementation of the C Plan following approval

Following approval of the C Plan as communicated to Sir Edwin by Mr Kofmehl, the plan for communication and implementation was put in train. I can take the documentation up until shortly before the execution of the 1983 Trust Deed and Rules quite shortly. It starts on the same day, 17 December 1982, as the approval by the board of the Trust Company, with a MIL (Management Briefing 63) bearing that date sent out to managers above the name of Sir Edwin. After explaining briefly the background to the introduction of the C Plan, and noting the importance that staff understand the new plan and its enhanced benefits, Sir Edwin emphasised four key elements. I need only mention the first which was this:

"Firstly, the plan provides flexibility for male and female employees to retire between age 60 and 63. This accords with our equal status principles and is also in line with published Government proposals to harmonise the State retirement age."

There is no hint of any need for company consent. Indeed, if it had been intended to communicate a need for consent, the drafting, (in which Mr Cawley and Mr Ellis, both of whom knew that consent was not required, were involved) would have had to word things differently since, obviously, consent would not be required at age 63 on any footing. So it would have been necessary to add a phrase such as "but not under the age of 63 without company consent".

On 20 December 1982, a further MIL (Personnel 557) was sent out to managers above the name of Sir Leonard (again, this was probably drafted by Mr Cawley or Mr Ellis). The MIL was expressed to be supplemental to Sir Edwin's earlier management briefing. Attached were (i) a comparative summary of the present plans (in particular the N Plan) and the new plan (the C Plan) in a similar format to that of the tables in the Initial Proposal and (ii) a Question and Answer set. Managers were required to pass this information to staff as soon as possible.

Managers were instructed to study the information carefully and to hold a departmental meeting in January to discuss its contents with staff. Attached, for use at such meetings, were two foils: the key benefits and the communications strategy. The implementation timetable set out envisaged: departmental meetings with managers in January; mini-statements to eligible members in February; road-shows over February to May at most UK locations; final cut-off date of applications to join to be published in May; and a full personal statement of benefits in September. Updates to the Employee Handbook section on pensions would be published.

The "Summary of Key Features of 'C' Plan" contained the following:

"Retirement Age

The 'C' Plan has been designed to allow both males and females to retire at any time between age 60 and 63 without reduction to the service-related pension. The latest age of retirement from IBM will be 63 for all employees. This is in line with published government proposals to harmonise state pension age.

Retirement between age 53 and 60 will be available at company discretion, subject to an actuarial reduction on the service-related pension."

Although Mr Simmonds did at one stage argue that this was to be interpreted as a requirement for consent to retirement after the age of 60, this submission fell by the wayside in the light of a concession which he eventually made, which I will come to later. But for completeness I must record my own view that that argument is wrong. Even if the first paragraph, on its own, might leave the reader wondering whether consent was required, the reference to company discretion in the second paragraph makes the position clear by drawing a contrast between retirement between ages 60 and 63 and retirement between ages 53 and 60.

The table of comparative benefits runs to five pages. I draw attention only to the following:

i) Normal retirement age is 63 for all members.

ii) There are differences between the C Plan as described in the MIL and the Available Revisions which I do not need to set out in detail. I do not know which, if either, version found its way into the revised proposals eventually submitted to Armonk.

iii) The early payment formula, like the Available Revisions, makes no mention of consent one way or the other but does provide for reduction below the age of 60.

iv) Deferred pensions for early leavers are payable at NRD, that is to say age 63 with only statutory revaluation in deferment.

Reading the table in isolation might lead a reader to question whether consent was needed to retire between ages 60 and 63, particularly as the unreduced pension at age 60 is found in the part of the table dealing with early retirement. But the table is not to be read in isolation since it must be read with the body of the MIL which, for reasons just given, makes clear that such consent is not required.

The Question and Answer set adds nothing of relevance.

There is, in the bundle, a two page document dated January 1983. The first page is headed Communications Support and sets out a timetable similar to one I have already mentioned. The second page is headed Key Features. One sees here for the first time in a document the use of the phrase "flexible retirement". Thus one finds:

"FLEXIBLE RETIREMENT BETWEEN AGE 60 AND 63 (LATEST)

MALES AND FEMALES"

Employee contributions are shown at 2.5% approx of base pay for the first year and 4.5% approx thereafter. I am not clear how these relate to the figure of 5% contained in the Available Revisions.

The exact provenance and intended use of this document is not clear. It could have been produced as part of the material for managers to use in meetings with staff or for use in the road-shows.

There is also a page comprising a time-table and locations of road-shows these running from the end of February 1983 to the middle of May 1983. 33 road-shows are indicated with Mr Cawley conducting 11 of them and Mr Ellis 10 of them.

There is also included in the bundle a set of documents which I think are print-outs of slides. As Mr Simmonds says, the evidence did not nail down who prepared these slides or what they were used for. Mr Leesmith thought they may have been used as a presentation, probably to European headquarters and corporate headquarters although this was something of a guess.

I do not need to refer to most of them as they are uncontroversial and do not assist in the resolution of any of the issues. But I do refer to the following:

First, there is a slide which is headed "LEAVING SERVICE" for members with over 5 years' service. It refers to the deferred pension entitlement being payable under the present plans (eg the N Plan) "at 60(F), 65(M)" and under the new plan (eg the C Plan) "at 63". Worked examples are then given in relation to the present and the new plans for a female joining at age 26 and leaving at age 55. The benefit emerging from the calculation under the C Plan is referred to as "Payable at 63" in contrast with the benefit under old plan as "Payable at 60".

Sir Leonard was rather evasive in his answers when it was pointed out to him that this was a disadvantage to females who, under the N Plan, would have been entitled to receive a deferred pension, albeit of a smaller amount, at age 60. He was reluctant to accept that that was the necessary result having had the point raised without any notice or time to consider it. Ironically, as it turns out, had the point arisen in 1983 and had expert advice been taken, the effect of the Preservation Requirements might have been realised so that the female would obtain the larger benefit but payable at 60 after all.

However, the example is clear and, ignoring the Preservation Requirements, correct and indicates an understanding on the part of the draftsman of the slide that the C Plan was to provide for payment of an unreduced deferred benefit only at age 63.

Another slide is headed "RETIREMENT". It gives worked examples calculating the amount of pension emerging on the assumptions of the examples for pension at age 60 and pension at age 63 for a male joining at age 26. These examples are given, in the same format, for both the present plans and the new plans. No mention is made of any consent requirement. It does not follow from the slide, however, that consent is not required between ages 60 and 63 since no distinction is drawn in the presentation between the present plans and the new plans; and yet the former requires consent to retirement at both ages 60 and 63.

On 7 January 1983, a management announcement appeared over the name of Sir Edwin. I am not clear why this additional announcement was made. The first bullet point in the announcement under "some of the key points" read as follows:

"Retirement flexibility between age 60 and 63, without reduction in service-related pension, for male and female employees. Early retirement will be permitted from age 53 at IBM's discretion."

As with the first MIL, it is not strictly necessary to decide what this has to say about the need for consent to retirement between ages 60 and 63. But as with that MIL, I express a view which is that this announcement cannot be read as requiring consent to be given in order to retire from age 60. Although the second sentence of the bullet point does not, in contrast with the MIL, say that early retirement will be permitted "between ages 53 and 60", it does draw a different distinction which is between "retirement" and "early retirement". It seems to me that "Retirement flexibility" gives a member the choice of when to retire and if that choice is made in the age range 60 to 63 that is to be seen as ordinary retirement. In contrast, retirement as young as 53 is seen as "early retirement" for which consent is needed.

There was a special issue of IBM UK News (an internal newspaper) on 19 January 1983 (Issue 280) which described the Plan Highlights using language similar to that which we have already seen:

"Flexible retirement age

Retirement at any age between 60 and 63.

Latest retirement age will be 63

………

Early retirement

All employees may be eligible to retire from 53"

Those producing the newspaper, it can be seen, had got the message that flexible retirement (ages 60 to 63) was to be a choice for members in contrast with early retirement for which there would merely be eligibility indicating that an employee would not have a unilateral right to claim an early retirement pension from 53.

The bundle also contains another set of documents which I understand to be print-outs of slides used at the road shows and which accompanied a script or tape, the text, a copy of which is also in the bundle. One such slide is headed "INLAND REVENUE LIMITS". Against the subheading "'C' Plan" one finds this:

Flexible retirement

Age 60-63 66.7%

Early retirement MAXIMUM x (SERVICE TO ERD)

Age 53-60 AT 60 (SERVICE TO 60)

This again draws a distinction between flexible retirement which is seen as ordinary retirement and early retirement which is seen as retirement, in this slide, before age 60. The fact that the Inland Revenue maximum is based on a proportion of the pension payable at age 60 suggests that at age 60 the pension would be payable as of right.

The script includes the following in relation to a young male employee:

"The new plan allows flexible retirement between ages 60 and 63. Our young man could retire at any age from 60 with no early payment discount. He could work on to 63, the latest age at which he can retire under the new plan."

Then the script turns to the example of a woman who will have completed 34 years service by age 60. She is asking why she has to work on to 63. In answer to that we find the following:

"The new plan does not force her to work for 3 more years. It can provide her with a high pension at 60 and at significantly less cost than the present plans. So she can still retire at 60 but now she has the choice of working to 63 and increasing her pension."

At the end of the script is a summary of the key elements of the C Plan. It includes:

"A flexible retirement age allowing you to choose when you retire between age 60 and 63….

Early retirement from age 53 with the company's agreement…"

Nothing could be clearer than that the person preparing this script understood that it was an essential element of the C Plan that both men and women would have the right to retire between ages 60 and 63 without the need for consent from any employer. I do not know who prepared the script. It may be that Mr Ellis did so or had some input into it; or indeed that Mr Cawley had some input. Whoever prepared it, I have no doubt at all that it reflects the messages which had come from Sir Edwin and Sir Leonard and reflects the intentions of the Management Committee and the Trust Company in approving the C Plan.

The start date for the C Plan was fixed for 6 July 1983. The formal documentation was not in place, but the Main Plan was administered on the basis of the intended C Plan from that date. Contributions were collected on that basis and future benefits were earned on that basis. New joiners became members of the Main Plan as if the C Plan were on foot. I imagine that it is the case that there were retirements between then and December when the December 1983 Trust Deed and Rules were executed and it would be surprising if some were not between age 60 and 63. There is no record of any person complaining that he or she was unable to retire between those ages in that period: this is perhaps not surprising given IBM's practice under the N Plan of giving consent. Whether members retired in this period in that age range because they were entitled to do so or because their managers agreed to it probably did not arise as an issue.

Once the C Plan was up and running from July, the Trust Company took over responsibility, including responsibility for the preparation of the formal C Plan documentation. Mr Cawley said that his job was done and he had no further input save in relation to a new version of the employee handbook to reflect the pension changes. Quite what supervisory function he held in relation to Mr Ellis after that time is not clear. I accept that he had no such role in relation to the preparation of the 1983 Trust Deed and Rules but he did have some role in relation to the handbook. In the end, as will be seen, I do not need to determine the scope of his responsibility.

On 18 October 1983, a further MIL was issued above the name of Mr Taylor. This announced a further opportunity for N Plan members to transfer to the C Plan. It does not refer to "flexible retirement". Indeed, all it says in relation to retirement other than at age 63 is this:

"Latest retirement age for the C Plan is 63.

Earliest retirement age (with company consent) for the C Plan is age 53".

I do not consider that this MIL assists one way or the other. It is entirely consistent with flexible retirement and is entirely consistent also with consent to retirement being required even after age 60. The description of the benefit is very brief and does not even mention that retirement from age 60 would be without actuarial reduction. I say no more about it.

Subsequent deeds

The November 1983 Handbook and the 1983 Trust Deed and Rules

IBM provided all of its employees with an Employee Handbook. It dealt with all aspects of an employee's terms and conditions of service. As part of the exercise of implementation of the C Plan, the existing Employee Handbook needed revision. The then current edition dated from February 1982. In a section dealing with the N Plan it was provided as follows:

"You may retire with a reduced pension before normal retirement date:

a)…..

b) with the consent of the company, after completion of at least 5 years service, and having attained age 50 (female) or age 55 (male)."

It should be noted that, read in isolation, this statement is not entirely accurate since the pension of a man who was allowed to retire early over the age of 60 was not reduced for early payment.

A new edition ("the November 1983 Handbook") was produced in November 1983. It was addressed to an employee as "you". Like the previous version, it was divided into a number of sections, with sections grouped together under three headings "Introduction", "Terms and Conditions of Employment" and "Benefit Schemes and Other Information". The majority of the sections simply repeat what had gone before. But under the third heading appeared section 9 dealing with Pension and Life Assurance Plans.

Section 9.01 was an introductory section dealing with all the Plans; section 9.02 dealt with the C Plan. It explained that Retirement and Life Assurance Benefits were provided under a trust deed "to protect the rights of members". The Trust Deed and Rules were "the formal legal documents which give effect to the plans". They "override statements in this handbook and all other statements" and were available for inspection on application. The different plans within the umbrella of the Trust Deed and Rules were described briefly. It was stated that the plans were expected to continue indefinitely but the right to change or terminate all of part of all of the plans was reserved to Holdings "subject to the provisions of the Trust Deed and Rules". On page 5 of section 9.01 it was stated that "It is a condition of your employment with IBM that you join the Pension and Life Assurance Plans as soon as you are eligible to do so….".

The statement to the effect that the formal documents overrode the Employee Handbook was not a particularly helpful explanation so far as concerned the C Plan since the 1983 Trust Deed and Rules did not exist at the date of publication of the November 1983 Handbook and there was no other formal documentation governing the C Plan.

In section 9.01 it was also explained that

"the age at which you can expect to retire depends on the plan in which you are a member. Early retirement may, in certain circumstances, be allowed. Full details are contained in subsequent sections."

The phrase "early retirement" is not defined in section 9.01. It is not necessarily to be equated as retirement before NRD. Indeed, flexible retirement was seen, as Mr Cawley explained, as providing a range of ages within which a member could chose to retire. His or her choice did not result, in his view at least, in early retirement; rather there was a range of ordinary retirement dates.

Section 9.02 was concerned with the C Plan. On page 2 of that section, certain terms appearing later in the section 9.02 were defined, corresponding to definitions eventually appearing in the C Plan itself. In particular, Normal retirement date (referred to as NRD) was "the first day of the pay month coincident with or next following your 63^rd birthday".

On page 3 of section 9.02.3 appears the heading "Retirement". It provided as follows:

"You can expect to retire not later than age 63 (your normal retirement date

The policy of the IBM United Kingdom group of companies is for 'C' Plan members to retire not later than age 63

You may retire with a reduced pension before age 63:

on account of incapacity; or

with the consent of the company, after completion of at least 5 years service and having attained your 53^rd birthday"

On page 4 of Section 9.02, there appears a heading "Retirement Benefits". Under this heading is a first marginal heading "How much is my pension if I retire between 60 and 63?", against which was set out a formula for the calculation of C Plan benefits and referred to Example A at the end of the section. That example was set out against a marginal heading "Normal Retirement Pension". It concerned the case of an employee retiring at age 60. The example said nothing either way about whether consent had been given to the retirement, but it produced a figure which was not reduced because it was payable before age 63.

Under this heading there is a third marginal heading (the second is of no real significance) "How much is my pension if I retire before age 60?". Against this marginal heading it is provided

"With the consent of the company, you may retire before age 60 but not before age 53. You can receive a reduced pension payable from your retirement date calculated as follows:"

There then followed a description of the calculation. The pension at age 60 (assuming service to that age) was to be calculated, the resulting figure being reduced to reflect the actual number of years service to retirement and reduced for early payment prior to age 60. Reference was made to Example B at the end of the section which concerned the case of an employee retiring at age 55. That example was set out against a marginal heading "Early retirement Pension".

The paragraphs quoted above from page 3 of section 9.02 are, on any view, in error in referring to a reduced pension apparently in all cases of retirement under age 63. Even on IBM's case, it is accepted that there is no reduction in the pension in cases where consent is given to retirement between ages 60 and 63.

Further, the absence of any reference to consent in relation to retirement between ages 60 and 63 on page 4 of section 9.02 and in Example A in contrast with a such a reference in relation to retirement under age 60 on that page and in Example B is pretty poor drafting if the intention was that all retirements under the age of 63 would require consent.

It is said by Mr Simmonds that the description of the retirement benefit under the heading "Retirement" in section 9.02 of the November 1983 Handbook shows that retirement before the age of 63 requires, in all cases, the consent of the company. I do not consider that that is a sensible reading of section 9.02. As matter of construction, approaching the issue as a lawyer and applying ordinary canons of construction, section 9.02 is not, in my judgment, to be read as requiring employer consent to retirement between ages 60 and 63. The words on which Mr Simmonds relies do not distinguish between (i) consent and (ii) reduction so far as age is concerned. It is, accordingly, very difficult to argue that a reduction applies only under age 60 but that consent is required under age 63. Example A makes clear, however, that there is no reduction for retirement between ages 60 and 63; that points to there being no consent requirement between ages 60 and 63, a conclusion which is strongly supported by the absence of any reference to consent in relation to the first marginal heading and the inclusion of such a reference in relation to the third marginal heading. Further, the words in the introduction against the marginal heading "Retirement" on page 6 of section 9.01 show that consent really features only in relation to early retirement: in the context of the C Plan, the marginal headings to Examples A and C show that retirement between ages 60 and 63 is normal retirement whereas only retirement under age 60 is early retirement.

I should place one caveat on that conclusion. If it were permissible to construe the November 1983 Handbook against a factual matrix which included the terms of the 1983 Trust Deed and Rules, the argument that consent to retirement between ages 60 and 63 does appear from the Employee Handbook itself is stronger. I do not understand how that can be so given that the November 1983 Handbook pre-dated the 1983 Trust Deed and Rules.

That is the answer which I reach as a matter of construction. But even if that answer is wrong, it is certainly not the case that it is clearly wrong. The provisions of section 9 are, at best from the point of view of Mr Simmonds' argument, ambiguous. I have reached my answer as a lawyer addressing the question as one of construction in the way that the court addresses the interpretation of any document the meaning of which is in dispute.

But ordinary members of the C Plan would not come to the November 1983 Handbook with the eyes of a lawyer, balancing one phrase against another. Such persons would, I suggest, be left with the clear impression that there was an unfettered right to retire from age 60, particularly persons who had attended one of the road-shows and had therefore correctly understood that he or she would have such an unfettered right. It is difficult to think that such a person, coming to the November 1983 Handbook, would detect that its provisions relating to the C Plan meant anything different from that which had previously been presented.

Why is this important? It is important because, on Mr Simmonds' argument, the November 1983 Handbook reflected an intention at the time of its publication that members of the C Plan should not have an unfettered right to retire at age 60. It shows that that was the intention of decision makers (Holdings and the Trust Company) at that time whatever may have been their intention at an earlier time. Or, if that is wrong, it is evidence of the intention of those decision makers when the 1983 Trust Deed and Rules were executed. Since, on my interpretation of the November 1983 Handbook, it does not demonstrate an intention on the part of anyone that a member of the C Plan could not retire between ages 60 and 63 without the consent of the employer, Mr Simmonds' argument fails in limine. But in case I am wrong, and he is right, about the message conveyed by the November 1983 Handbook, I propose to examine the evidence about why the drafting took the form which it did.

The drafting of the 1983 Trust Deed and Rules was being carried out at the same time as preparation of the revisions to the Handbook. Since the sequence of events, insofar as it can be established at all, is of relevance, I consider the events leading up to the publication of the November 1983 Handbook and the execution of the 1983 Trust Deed and Rules together.

I deal first with Mr Cawley's evidence about this. In his first witness statement, he said that it was a mystery to him why the consent requirement on page 3 of section 9.02 was included since, so far as he was concerned, the revisions were simply meant to be a follow-up to the employee communications earlier in the year and were meant to contain the same message. He considered that the consent requirement was an error and speculated that it was an inappropriate carry over from the provisions of the Employee Handbook as they previously applied to the N Plan. In referring in his evidence to the consent requirement, he was referring to the words in the description of the retirement benefit "You may retire with a reduced pension before age 63….with the consent of the company….". But he made the point (which I have already identified) that the description was incorrect read in isolation because, on any view, the pension was not to be reduced where retirement was after age 60.

Enlarging on this in his second witness statement, he explained that he supervised Mr Ellis in the preparation of an updated section of the existing handbook to reflect the introduction of the C Plan (ie section 9 of the November 1983 Handbook). The N Plan remained in place because not all members transferred to the C Plan. Presentationally, he and Mr Ellis wanted to treat the C Plan and the N Plan as consistently as possible. Thus their starting point for the part of section 9 dealing with the C Plan was to follow the approach already adopted for the N Plan. It is this which he thinks has led to what he calls an ambiguity arising from the language of section 9.02.

He stated that the wording was certainly not a deliberate attempt to depart from the principle of flexible retirement (ie without the need for company consent). He does not believe, and certainly does not recollect, any discussion with more senior management within the Trust Company or the Management Committee to depart from that concept. On that point, at least, I am sure he is correct. It is confirmed by Sir Leonard and there is no evidence to suggest that the contents of the November 1983 Handbook were ever discussed with him or presented to the Management Committee for approval before publication or for any other reason. Likewise, I am sure that the matter was not raised with the board of the Trust Company. Sir Leonard was on the board and would have known if it had been. Mr Taylor, who was at all material times a director of the Trust Company says he knew nothing of it. I find as a fact that the inclusion of a provision into the November 1983 Handbook, which would have resulted in a requirement for the consent of the company to retirement between ages 60 and 63, was not raised with the Management Committee or the board of the Trust Company before its publication. None of the members of those bodies had any idea that a handbook had been produced which departed from the concept of flexible retirement which they approved as part of the C Plan.

In relation to the drafting of the 1983 Trust Deed and Rules, Mr Cawley's evidence was as follows. In his first witness statement, he stated that he had no recollection of and no participation in the preparation of the 1983 Trust Deed and Rules. By the time of his second witness statement, he was able to recall one relevant piece of information having considered the parties' cases and the terms of IBM's Defence. He was able to recall what he describes as a short conversation with Mr Ellis concerning advice given by MPA, a benefits consultancy based in Chichester, about preservation (ie the Preservation Requirements). He recalled that

"Mr Ellis mentioned to me that MPA had spotted that by not extending the "right" to flexible retirement to the deferred pensioners, the Pension Scheme could fall foul of the preservation requirements. This had not been pointed out to us at an earlier stage."

It should be noted here that MPA must, at the time it drew attention to the point, have known that the C Plan was being presented to the workforce on the basis that consent was not required and must have believed that that reflected IBM's intention. Representatives of MPA attended some, at least, of the road-shows and answered queries from people present about their entitlement, including statements to the effect that retirement from age 60 was available without employer consent. I make a finding of fact based on the evidence of Mr Ellis which was not in any way contradicted to the effect that MPA knew that the C Plan was being presented on that basis and that they believed that basis to reflect IBM's intention concerning the benefits of the C Plan.

Mr Cawley did not state in his second witness statement when this conversation took place. In his cross-examination, he said that he was certain that it was after the November 1983 Handbook had been published and therefore at the end of November 1983. Mr Cawley went on to explain that MPA had suggested a solution namely to insert a consent requirement for all retirements. Mr Ellis came to Mr Cawley (who according to Mr Cawley himself was not formally involved in the preparation of the 1983 Trust Deed and Rules) to ask him for advice. Mr Cawley told him to follow MPA's advice. His explanation might be thought to be surprising; and Mr Cawley himself accepts, to use my words, that his suggestion can be seen in retrospect as not, perhaps, the best advice he has ever given. He explained his position in this way:

"At that time, I did not think deferred members were intended to benefit from the flexible retirement concept. In contrast, Holdings had, via the employee communications, already committed to giving employees the option to retire from age 60, and from all the discussions I had previously had with company management on the design of the C Plan back in 1981-1983, I believed it would be their intent to continue to do so. I therefore saw the consent clause simply as a formality to get round the preservation issue."

In cross-examination, Mr Cawley added this

"Knowing that the company management and the trustees mirrored one another, my expectation was that the lawyer, having expressed the view to Sam, and Sam having expressed that to me, was going to go in front of the trustee board with a whole series of points that he would want to raise with the trustees before that trust deed and rules were actually produced. But in November I was not in any way involved with the trustees and my job had really come to an end. So from that point of view I was just relaxed and too keen to get out of the project, frankly."

Whether Mr Cawley actually thought all of this at the time, or whether it is the best explanation he can give without actually confessing that he simply got things wrong I find it hard to say. But it does not matter, for present purposes, which it was. It will be seen why I take this view once I have reviewed Mr Ellis' evidence about his discussion with MPA and his conversation with Mr Cawley and after I have reviewed the evidence of Mr Ellis, Mr Gooch and Mrs James about a discussion which they say took place with Mr Cawley on the same topic.

As for Mr Ellis, he agreed, and this is in any case common ground, that he was involved in the preparation of the November 1983 Handbook. He said, and I accept, that this exercise was being carried on in parallel with the preparation of the 1983 Trust Deed and Rules in which he was also involved.

Mr Ellis's evidence was that he had a discussion with Adrian Lamb of MPA during the course of his drafting of the 1983 Trust Deed and Rules. The problem arising out of the impact of the Preservation Requirements was identified. As Mr Ellis explained it, Mr Adrian Lamb floated the idea of a solution which was that IBM could require consent to early retirement. The rationale was that the Preservation Requirements might apply if the company did not have a formal consent requirement written into the rules; an automatic right to draw a pension at 60 would have applied to the payment of deferred pensions, increasing the funding cost of the C Plan. Another reason given for introducing a consent requirement would be to give IBM room for manoeuvre or to adapt in the event that the circumstances which had given rise to the introduction of the C Plan in the first place were to change.

Mr Ellis said that this idea was explored between him and MPA in some detail. He saw no apparent downside to the proposal. Neither did he see any impact on the internal communications regarding the C Plan which had already been sent to employees, (and on which, I might add, they had already acted by joining the C Plan) "because the position was (at that time) that consent would be automatically granted". He said that he did turn his mind to the conflict between inclusion of a consent requirement and the communications which had been sent to employees it being important to ensure that the scheme rules were consistent with the impression created by those communications. He recalled speaking to Mr Cawley about it and had a clear recollection that they both agreed that a consent requirement was consistent with the concept of consent being automatically granted as under the existing practice. He imagined that there might be a time when consent would be withdrawn because of the changing needs of the business. He said that he would not have agreed to the inclusion of a consent requirement without first discussing the position with Mr Cawley.

Mr Ellis was at first unable to say whether it was before or after the November 1983 Handbook had been finalised and published. But towards the end of his cross-examination he said that it was before the November 1983 Handbook had been published. But then he accepted that he could not recall the timescale.

He did not say that the consent requirement appeared in the revisions to the November 1983 Handbook as a result of his conversation with Mr Cawley. Indeed, he acknowledged that his drafting of the November 1983 Handbook did not look as though he had been intending to reflect a consent requirement. The overwhelming evidence from a number of witnesses, including Mr Ellis himself, was that a significant change – and this would undoubtedly include the introduction for the first time of a consent requirement – would have been carefully communicated to the work-force, almost certainly by way of a further MIL. It is clear, however, that there was no further MIL and that the only communication with the members of the C Plan was the November 1983 Handbook. This all points to the 1983 Handbook having been finalised and published before the conversation between Mr Cawley and Mr Ellis.

If the identification of the preservation issue and the conversation between Mr Cawley and Mr Ellis preceded the finalisation of the November 1983 Handbook, the drafting of it is pretty astonishing. Even if Mr Ellis had thought that the introduction of a consent requirement did not change the nature of a member's entitlement to take early retirement, he would surely have ensured that the drafting made clear that consent was required, even if it was stated that consent would be given automatically. It does not do so: whether I am right or wrong as a matter of construction of the November 1983 Handbook, it is, as I have said, at best ambiguous. Everyone who knew Mr Ellis and his work seems to accept that he was a careful and competent draftsman. If he had intended the November1983 Handbook to indicate to the reader that consent was required, I am sure he would not have allowed it to appear in the form which it did.

Nonetheless, what actually happened is surprising to say the least. It is not disputed that Mr Cawley and Mr Ellis themselves had no authority to alter a significant term of the C Plan as approved by the Management Committee and the Trust Company. Each of them was perfectly well aware of that. The only explanation for Mr Ellis' conduct (and possibly the same goes for Mr Cawley) is that he thought that the requirement for consent made no difference. Indeed, it was precisely because the practice had been to give consent under the N Plan that he considered that consent was a formality. In his witness statement, he said that he imagined there might come a time when consent would be withheld. But he also repeatedly referred, both in his witness statement and in his oral testimony, to "automatic" consent. In answer to a question from me, he explained that "automatic" meant that the company would have to give its consent and would not be entitled to withhold its consent. It would, as I put it, be as if hidden away in a schedule to the Rules there was a provision that said, in the case of retirement between the ages of 60 and 63, consent could not be withheld.

The evidence of Mr Adrian Lamb confirmed what Mr Ellis has said. There is no reason to think that he, unlike Mr Cawley and Mr Ellis, was aware of the basis on which the C Plan had been developed and no evidence about whether he knew what was in the November 1983 Handbook.

There was, it can be seen, a serious element of confusion in Mr Ellis' thinking. One aspect of his thinking was this: he knew that the C Plan had been developed and communicated on the basis of an entitlement to retire at ages 60 to 63 and he did not want to cut across that. Another aspect of his thinking was this: he perceived advantages from the point of view of IBM of a consent requirement both in terms of the problem arising out of the Preservation Requirements and in terms of potential flexibility in the future.

However, if consent was to be a genuine consent in the sense that it could be refused, those two aspects of his thinking could not be reconciled. His attempt at reconciliation was the idea of "automatic" consent, that is to say a consent which would have to be given if asked for and could not be refused. Whilst he might have been able to convince himself that a form of words apparently requiring consent but which could not ever be refused might get round the preservation problem, it is not possible to see how an ability for IBM to alter things in the future could constitute anything other than a significant departure from flexible retirement as he understood it, a departure which he knew he had no power to authorise.

I should next say something about the evidence of Mr Gooch and Mrs James in the context of what they did or did not say about the Preservation Requirements during the course of 1983. In paragraph 13 of his witness statement, Mr Gooch referred, without giving details of time or place, to the issue of consent to retirement being raised in discussion between Mrs James and himself with Mr Cawley. It became apparent in cross-examination that this discussion was probably in November 1982. It was separate from what is a more significant discussion to which I have already referred in paragraphs 91ff above and which I do not repeat. It concerned the need for company consent to retirement. I note that it is not suggested that Mr Ellis was present at this conversation. I do, however, need to say a little more about this evidence.

The discussion, if it took place, must have been in 1983, and not 1982, because Mr Gooch had by then been provided with some documentation that had been sent to members, which he had not seen before. Precisely what this documentation was remains unknown, but it is likely to have been the MILs and accompanying documentation. Mr Gooch said that "we" (presumably Mrs James and himself) were concerned about the absence of any reference to employer consent. He then gave the evidence in the last paragraph quoted in paragraph 91 above.

Mr Gaisman pressed Mr Gooch, of course, on whether he was sure that this dialogue took place: he said he was. When asked how he could be so sure, he replied, as I have already quoted, that it was "a very, very, important point and that is why I recalled it". Rather than summarising what came next, I set out the exchange between Mr Gaisman and Mr Gooch:

" Q. Have you not recalled it for the past 30 years?

A. No, because it was resolved soon afterwards.

Q. How was it resolved soon afterwards?

A. By the employee handbook.

Q. So you had no reason, as far as you were concerned, the matter was sorted out there and then, was it?

A. It was sorted out when we saw that employee handbook, yes.

Q. So you had no reason, as it were, to remember it. The matter had been put to bed, as far as you were concerned.

A. The matter had been put to bed, yes.

Q. But nonetheless you still remember it 30 years later?

A. Yes, because at the time it was very significant.

Q. For a short time?

A. For a period of months, yes.

Q. I thought you said you didn't know when it took place? For a period of months? So you are now saying it was a period of months before the employee handbook, are you?

A. What I'm saying is that I know clearly that the comments -- the question and answers that I had with Mike Cawley took place after some communications had gone out to members.

Q. Yes. You did say that.

A. And that the handbook was produced, I believe, late in 1983.

Q. It was.

A. So there is a period of months between when I said that to Mike Cawley and the employee handbook emerging.

………..

Q. At any rate, as far as you were concerned, the matter, do I understand this, was made clear and therefore you could stop worrying about the point, insofar as you were worrying about it, by the contents of the handbook?

A. I could -- I stopped worrying about it when we had seen the handbook, yes."

Mr Cawley seemed willing to accept that a conversation with Mr Gooch when preservation was mentioned might have taken place, although he did not remember it. He said that it would have been after publication of the November 1983 Handbook. If it had been before then, the November 1983 Handbook would have been the perfect opportunity to communicate the issue to the members of the C Plan. Being unable to recollect a meeting, he could not say what he did or did not say. But it would have been very odd if he had said anything like "We will put it right" in relation to a benefit which he considered had been properly identified in the road shows and other communications.

Some other points also need to be made.

i) First, as Mr Gooch himself said, the conversation was very short. It is astonishing to me that Mr Cawley would simply have said "Okay, we will include it next time" or words to that effect. Mr Cawley, after all, knew and appreciated the importance of flexible retirement including the entitlement to retire at age 60; he knew it was a selling point of the C Plan. And yet here he is supposedly accepting without more ado that a consent requirement would be inserted "next time" whatever that may have meant, as though its omission had been an error. Surely his response would have been "What on earth do you mean: the whole concept of the C Plan is an entitlement to retire at age 60 without consent". He might then have gone on to have the sort of conversation which Mr Ellis relates took place when he, Mr Ellis, raised a consent requirement with Mr Cawley.

ii) Secondly, according to Mr Gooch, the point was "very, very important" and "significant" but not, apparently, important or significant enough to be recalled in the context of preparing a witness statement itself dealing with conversations about preservation.

iii) Thirdly, Mr Gooch does not suggest that preservation was in fact discussed or mentioned: his evidence was simply that he pointed out to Mr Cawley that there was no reference to company consent. Mr Gooch might have appreciated that the impact of the Preservation Requirements was serious but does not appear to have told – or as he might have it, to have reminded - Mr Cawley about them.

iv) Fourthly, it would be very surprising indeed to my mind if Mr Cawley had not discussed the point with anyone else, particularly Mr Ellis and Sir Leonard, if he had agreed with Mr Gooch that the Employee Handbook would be corrected (unless either he had forgotten the point or had decided to keep it to himself for whatever reason). There is not a shred of evidence that he did discuss the point with anyone else. I find as a fact that he did not do so and that the only discussion he ever had on the point with anyone within IBM was with Mr Ellis following Mr Ellis' own discussion with MPA.

v) Moreover, one might expect that he would have hot-footed it from the meeting to do just that, to discuss the issue with others, in order to put in place a correction to a very serious piece of misinformation which had been communicated to the workforce not only in the written communications but also in the road shows in the preparation of both of which he had been involved and, in the case of the road shows, which he himself had presented on several occasions.

vi) There is here a marked contrast with what happened after his conversation with Mr Ellis. In that case, there was a discussion between them as a result of which, whatever the shortcomings in their actions, they decided that the issue could be dealt with by the insertion of a consent requirement. In contrast, it is not alleged even by Mr Gooch and Mrs James that there was a similar discussion. It was simply that Mr Cawley agreed that it would be "put right next time". If what Mr Gooch said is correct, then another oddity is that Mr Cawley would have had the conversation which he did with Mr Ellis once MPA had identified the issue. Surely he would at least have mentioned his earlier conversation with Mr Gooch; it has not been suggested that he did so and that was not put to Mr Ellis.

vii) Fifthly, Mr Cawley did not in fact follow up what Mr Gooch reported him as agreeing to: there was no communication with the workforce in which they were told that company consent to retirement would, after all, be required. Further, if it is said that that is precisely what the November 1983 Handbook did and what Mr Cawley, contrary to his evidence, in fact intended it to do, it would be little short of dishonest for Mr Cawley deliberately to have dealt with the matter in such an opaque way.

viii) I find Mr Gooch's position that his concerns were addressed by the November 1983 Handbook as remarkable. I may be right, I may be wrong about its correct construction. But one thing which is clear is that a person (ie Mr Gooch) who was concerned about the point could not reasonably see the November 1983 Handbook as not only providing the clarity which was necessary but also as providing it to a person who had been told the contrary by the road shows and in respect of whom it was particularly important for clarity to be achieved.

Turning to Mrs James, she said, in paragraph 25 of her witness statement that she recollected becoming aware that one or more employee communication documents distributed by IBM did not adequately describe the fact that IBM consent to retirement before age 63 was required. That she believed that to be a requirement I do not question, although it is not clear to me precisely what led her to hold that belief other, perhaps, than a misunderstanding between herself and Mr Gooch on the one hand and Mr Cawley on the other hand about the basis on which the costings for the revised proposal were to be carried out. She recalled bringing the issue to Mr Gooch's attention. She then referred to a meeting at which she was present when Mr Gooch spoke to Mr Cawley reminding him of this consent requirement and that it had underpinned the costings that had been done. She remembered Mr Cawley indicating that communications would be amended in the future. The conversation took place prior to the issue of the November 1983 Handbook but she could not remember when. In cross-examination, she said that Mr Cawley

"said that he wasn't concerned that it was a big issue because he would be able to correct it in subsequent communications and it wasn't a big issue because IBM would let -- would allow active members to early retire in the foreseeable future. I remember when the November 1983 handbook came out, looking at that and noticing with disappointment that it would have been clearer if that had -- if the paragraph on early retirement between 60 and 63 had been clearer, that it needed company consent. It said it at the beginning and then -- said it at the beginning for early retirement, then it did the 60/63 -- it didn't mention it. And then before 60, it did mention the need for consent. And I mentioned that to IBM and they said that they didn't see it as a big issue and it would be corrected in the next issue. And I also remember noticing that it wasn't corrected in the next issue.

Q. Who did you mention that to at IBM?

A. I'm not sure….."

It is difficult not to share her disappointment if the November 1983 Handbook was intended to reflect an intention on the part of IBM that consent would be required for retirement below age 63. It does not appear to be a disappointment shared by Mr Gooch whose evidence, as I have said, was that he stopped worrying about the point when the November 1983 Handbook had been published.

At this point, it is appropriate to mention another meeting referred to by Mr Ellis at paragraph 47 of his witness statement. He recalled that there were one or two errors in the MILs which had been issued in relation to the C Plan although he could not remember precisely what they were. In his oral evidence, he said, and I accept, that they were nothing to do with a consent requirement. He said that they were pointed out to him and Mr Cawley, he thought by Mrs James. It is suggested in the Trust Company's closing that the evidence of Mr Gooch and Mrs James is (at best) defensive reconstruction and that the likely explanation is that they are muddling the conversation about the consent requirement with this conversation.

That is certainly a possibility. I am bound to say that I do not think that the recollection of Mr Gooch or Mrs James is as good as they suggest. Without suggesting that they are attempting to mislead the court I do think that there is an element of reconstruction taking place which makes it very difficult to accept that the conversation about consent took place, if it took place at all, before the publication of the November 1983 Handbook.

I do not, however, need to make findings necessary to resolve the evidential dispute between Mr Gooch and Mrs James on the one hand and Mr Cawley on the other. This case is not about whether Mr Gooch and Mrs James fell short, in some way, of their professional duty nor is it about whether Mr Cawley and Mr Ellis have failed in their own duties to IBM. The issues at the moment are what the intentions of IBM and the Trust Company were when they executed the 1983 Trust Deed and Rules and, peripherally, in issuing the November 1983 Handbook. The essential point to note is that it is not what was said to Mr Cawley which is important but what was said by him (or by Mr Ellis) to others. As to that, I am perfectly satisfied that nothing was said by Mr Cawley or by Mr Ellis to anyone senior: there was no communication up the line from Mr Cawley. I find as a fact that the need for a consent requirement to avoid the consequences of the Preservation Requirements was not communicated by Mr Cawley or Mr Ellis to the Management Committee, the board of Holdings or the board of the Trust Company (or, so far as the available evidence goes, to any members of those bodies).

What then were the intentions of the decision makers in relation to the 1983 Trust Deed and Rules? By "decision makers" I mean, of course, the Management Committee (as delegates of the board of Holdings) and the board of the Trust Company. Given my finding of fact that the need for a consent requirement to avoid the consequences of the Preservation Requirements was not communicated to the Management Committee, the board of Holdings or the board of the Trust Company (or, so far as the available evidence goes, to any members of those bodies) it follows that there would have been no reason for the decision makers, in executing the 1983 Trust Deed and Rules, to have departed from their previous intentions as objectively identified, that is to say an intention that the C Plan should provide an entitlement, without the need for company consent, to retire at any age between 60 and 63.

The 1983 Trust Deed and Rules were executed by Holdings and by the Trust Company on 14 December 1983. The only surviving record of its discussion is in the minutes of the board of the Trust Company held on that date. The board, as I have mentioned, included many senior IBM personnel. Those present at the meeting included Sir Edwin, Mr Caddick, Mr Morgans, Sir Leonard and Mr Taylor. I note that Mr Gooch was in attendance. The minutes record the following:

"The Secretary explained the principal changes which were proposed to the [Main Plan] Trust Deed and Rules and the reasons for them (mainly due to the introduction of the new contributory pension plan) and confirmed that the changes were satisfactory from a Secretarial standpoint, and that the revised Trust Deed and Rules had already been sealed by the Principal Employer."

It was resolved to ratify and execute the "revised Trust Deed and Rules" and the deed was then duly executed. Nothing is known about the execution of the deed by Holdings save that it appears to have been duly executed and that it is recorded in the Trust Company minutes as having been executed.

There is no record of any sort of discussion by the Management Committee, the board of Holdings or the board of the Trustee Company about the terms of the C Plan in the period from July to December 1983, that is to say from the date when the C Plan commenced to the date of the 1983 Trust Deed and Rules. There is no evidence of any consideration of the terms of the C Plan by the Trust Company since the end of 1982 when it had approved the C Plan including flexible retirement. Nor is there any evidence of such consideration by the Management Committee or the board of Holdings since approval of the C Plan was received from Paris in December 1982, although it is intrinsically unlikely, I think, that the Management Committee did not note receipt of the approval and also confirm the implementation of the C Plan. It is highly unlikely that any member of the Management Committee was unaware that the C Plan was being presented to the workforce as including an entitlement to retire between ages 60 and 63 without any requirement for consent.

The evidence of Sir Leonard, Mr Taylor and Mr Caddick, which was not challenged, was that the assumptions of the boards of Holdings and the Trust Company in December 1983 was that, since the C Plan was already up and running, there was no need to revisit its terms, which had been announced to Members. They were set in stone by this stage. Mr Morgans says much the same thing, although I am reluctant to place reliance on his evidence since he was not available for cross-examination. In more detail:

i) Sir Leonard points out that the C Plan had been up and running for almost 6 months by 14 December 1983. There was no question, he said, of having to revisit matters which had been decided in December 1982 and communicated to employees. The board of the Trustee Company had been informed of the take-up of C Plan membership at meetings on 21 July and 18 October 1983 so that it was clear, at the December meeting, that they were dealing with an already-established section of the Main Plan.

ii) Mr Taylor was surprised that "flexible retirement" was not reflected in the 1983 Trust and Rules. He was at the meeting of the board of the Trust Company on 14 December and was party to the approval given. He had no detailed recollection of the content. It was, however, his understanding that the Rules were not a vehicle in themselves for generating IBM policy but were designed to give effect to the policy decisions already taken by IBM as reflected in the MILs and other employee communications. It will be recollected, as I have already dealt with, that the Trust Company saw its role as administration and proper custodianship with benefit design being the primary role of Holdings. Mr Taylor did not recall any discussion about revisiting the rights of members in the run up to the execution of the 1983 Trust Deed and Rules. He did not recall anyone drawing attention to anything to suggest that they might be materially inconsistent with the MILs and other communications. He is sure that the board did not intend to approve the execution of a document that was inconsistent with them.

iii) Mr Caddick, who was also present at the board meeting of the Trust Company on 14 December 1983, stated his understanding was that it was intended to amend the Main Plan to reflect the introduction of the C Plan in the form communicated to employees at the end of 1982 and the beginning of 1983. He had no recollection of any departure from the terms of the C Plan as previously communicated to employees being discussed at the meeting.

All of that evidence is, of course, consistent with the evidence from Mr Cawley and Mr Ellis that nothing was communicated by them up the chain of command. There was simply no way in which the point would have come to the notice of the decision makers. It is also consistent with this feature: as I have noted, Mr Gooch was present at the meeting. He said, as I have recorded, that he regarded the consent requirement as "very, very important". If the issue of consent had arisen at the board meeting of the Trust Company on 14 December 1983, he would surely have pointed out the need for consent to avoid problems in relation to the Preservation Requirements. It is clear, and I find as a fact, that he did not do so. This suggests that consent was not raised at all.

I do not propose to review the evidence of other witnesses, such as Mr Jonas, Mr Leesmith, Mr Fitzgerald and Mr Martin James, all of whom were consistent about the absence of any announcement of any qualification to the MILs and road shows which had demonstrated an entitlement to flexible retirement. I find as a fact that there was no announcement or MIL prior to the execution of the 1983 Trust Deed and Rules which qualified that entitlement in any way.

If there had been a change of policy on the part of Holdings and the Trust Company to introduce a consent requirement, it would follow that such a change was introduced surreptitiously without informing all those affected and who had transferred to the C Plan (to repeat, a Plan which was already up and running by 14 December 1983). It would have involved a significant breach of good faith by senior management with the knowledge of Mr Ellis and Mr Cawley. There is no evidence to support the suggestion that such underhand behaviour took place. Indeed, all of the evidence from witnesses at senior level (Sir Leonard, Mr Taylor, Mr Caddick and Mr James, as well as Sir Anthony Cleaver and Mr Morgans) is that this would have been completely out of the question and counter to IBM culture. I do not believe for a moment that this is what happened.

So far as the board of Holdings is concerned, there is no record of a board meeting at which execution of the 1983 Trust Deed and Rules was approved. It was, however, executed and it is almost inconceivable that its execution was not properly approved. Precisely what was approved is not recorded. But given that Sir Edwin, Sir Leonard and Mr Morgans were on the boards of both Holdings and the Trust Company, it is scarcely credible that they, at least, would have approved the 1983 Trust Deed and Rules as providing for flexible retirement in their capacity as board members of the Trust Company but requiring consent in their capacity as board members of Holdings (and the Management Committee too).

My conclusion, therefore, H His that it was the intention of the boards of both the Trust Company and of Holdings in authorising the execution of the 1983 Trust Deed and Rules that the C Plan should provide an entitlement for members to retire between the ages of 60 and 63 without the consent of Holdings or any employer. The intention is, in each case, objectively established and is not simply an intention hidden away in the mind of each individual who was a member of each board (or the Management Committee).

But I go further than that. It is consistent with that conclusion that it was also the intention of Holdings and the Trust Company that the 1983 Trust Deed and Rules should give formal effect to the way in which the C Plan had been implemented in July 1983. I find as a fact that the intention of Holdings and the Trust Company was to do precisely that. It is clear that the C Plan was up and running from that time, and certainly well before 14 December 1983. The benefits under the C Plan were, therefore, not simply intended benefits but ones which were already part of the structure of an existing Plan. The benefits included flexible retirement: the position was not simply that Holdings and the Trust Company intended to provide that benefit when the C Plan came to be formally adopted. Quite the reverse: the position was that the intention of Holdings and Trust Company when executing the 1983 Trust Deed and Rules was that its formal provisions should accurately reflect the benefits which formed the basis on which members transferred into the C Plan.

This finding has significance in the context of whether effect should be given, by way of rectification, to the intention to provide flexible retirement.

The next issue I want to consider is whether the C Plan benefits were intended to be provided to joiners as well as to members transferring to the C Plan from other sections of the Main Plan. The answer to this is, in my judgment, clear and it is that, as of 14 December 1983, there was no intention to distinguish between transferring members and new joiners.

The reason it is clear goes back to the Initial Proposal which itself made clear that the C Plan was to be not only available to, but compulsory for, new joiners. Of course, the benefit structure was not identical to that provided for transferring members because special provision had to be made for transferring members in respect of their previous employment. But so far as future service benefits were concerned, new joiners and transferring members were to be treated in the same way. Sir Leonard and Mr Crawley were clear in their evidence to that effect and were not challenged on it. Mr Morgans says much the same thing. A difference in treatment would have resulted in a complex and to my mind rather illogical structure. In any case, there is not a shred of evidence to suggest that transferring members and new joiners were to be treated any differently so far as concerns an entitlement to retire between ages 60 and 63.

My conclusions concerning flexible retirement take account of the way in which the C Plan was operated after December 1983. As appears from Gallaher at [141], "it is permissible in claims for rectification to have regard to events after the transaction is entered into as evidence of the parties' intention at the time of the transaction and (where required) as objectively manifesting that intention". I do not, however, consider that the way in which the C Plan was actually operated so far as concerns the obtaining of consent to retirement between ages 60 and 63 is relevant. That is because the evidence of the intention at the date of the 1983 Trust Deed and Rules is clear and cannot be displaced by evidence about the subsequent operation of the C Plan. In my judgment, it is clear on the evidence that flexible retirement was intended by Holdings and the Trust Company when authorising the execution of the 1983 Trust Deed and Rules to be included as a feature of the C Plan. There is also a clear explanation as to why a consent requirement was in fact included. Even if the C Plan had thereafter been administered (perhaps even with the knowledge of the decision makers), that goes nowhere near displacing the clear evidence of intention which I have just mentioned. All that it would show is that the C Plan was being administered according to the mistaken provisions of the C Plan. The evidence is compelling, in my view, and more than adequate to satisfy the requirements for a successful claim to rectification.

Members with deferred benefits

I turn now to consider the position of deferred members of the C Plan by which I mean early leavers with an entitlement to benefits commencing or payable at some time in the future. The Trust Company's primary case is that there is no evidence that the decision makers on the Management Committee and the board of the Trust Company deliberately decided to discriminate against deferred members or to deprive them of the right to take their pension at age 60 which (given the intention regarding actives) was required to be conferred by the Preservation Requirements.

The alternative case is that the intended right to flexible retirement was not limited to actives but was also intended to survive into deferment. It is said that, objectively speaking, the parties to the 1983 Trust Deed and Rules intended to give effect to the MILs and employee communications issued in December 1982 and the first half of 1983. It is said that these communications never expressly stated that the right to draw an unreduced pension at age 60 would be inapplicable if C Plan members left IBM service and gave the impression that females (who under the N Plan would have had a right to a deferred pension at 60 if they left pensionable service) would be better off under the C Plan.

I find the whole idea of "flexible retirement" in the context of deferred benefits something of an oddity. The concept of "flexible retirement" is one under which a person may leave service and take an immediate pension within a range of ages. That offers him or her a real choice about how long to go on working and thus accruing further benefits. But for a person who has left service and is entitled, prospectively, to a pension, it is not at all easy to see why, in December 1983, it might be thought that such a person would ever start to draw an unreduced pension later than the earliest date on which it could become payable. In those days, there was no revaluation of deferred pensions in excess of GMP so that there would have been no reason for delay given that there would be no further benefit accrual.

Let me start with the Initial Proposal. At that stage, "flexible retirement" was not in the lexicon although the concept of retiring between ages 60 and 63 as of right certainly was. Since NRD under the Initial Proposal was age 60, there was no issue about when a deferred pension would become payable: it would be at NRD that is to say age 60. This would have maintained the position for women and would have brought forward the age for men from 65 to 60.

Moving on to the period after the rejection by Armonk of the Initial Proposal, the Available Revisions state that the deferred pension entitlement would be payable at NRD that is to say age 63. Nothing is said about any right or option to claim early payment whether reduced or not. It is particularly unfortunate in relation to this benefit that the narrative in the revised proposal is not available as it would be bound to have qualified in some way what was said in the Initial Proposal.

It is fair to say, however, that the material used in the road shows does not appear to address the payment date of deferred pensions in any detail. The only reference in the road-show script is to a male's benefit which was payable "two years earlier at age 63". In that case, there was no suggestion that it might, at the member's election, be payable at 60.

These documents all point to an intention that leaving service benefits would be payable only at age 63 and not earlier.

Mr Simmonds relies heavily on Mr Cawley's evidence to show that there was no intention that deferred benefits might be payable unreduced from age 60; indeed, he describes Mr Cawley's evidence as pivotal in determining what the actual intentions of Holdings and the Trust Company were in 1982/83. Mr Cawley's evidence was to the following effect. Although he was not a member of the Management Committee he was the architect and primary author of the C Plan proposal. The concept of "flexible retirement" was his idea. The Management Committee approved what he had designed and if he did not know about a particular feature, it was not part of the C Plan (although for reasons already given, I think that it is putting it too high: see paragraph 181 above). It was his job to liaise generally with the Management Committee in relation to the formulation of the C Plan. He was clear that "flexible retirement" was only intended to apply to actives and his briefings to more senior management would have covered this. Nobody in senior management ever proposed to him that deferred members should have a right to an unreduced pension at 60.

It is clear from Mr Cawley's evidence, which I accept, that it was never his intention or understanding, subjectively, that early leavers should have a right to receive their deferred benefits prior to NRD, the common age of 63 for both men and women. I am also satisfied that his design of the C Plan was consistent with that intention and that he never suggested to anyone more senior (or indeed to Mr Ellis) that early leavers should become entitled at any younger an age.

This conclusion is supported by this consideration: if it had ever been his intention that deferred pensions should be payable unreduced prior to age 63, then his conversation with Mr Ellis at a much later stage about a consent requirement for active members would have made no sense at all. The only reason to have inserted a consent requirement at that time would have been to avoid the consequence that deferred pensions would become payable at age 60. If it had been the intention to confer an entitlement of a deferred pensioner to draw an unreduced pension at age 60, no issue would ever have arisen about introducing a consent requirement for actives to draw their pensions unreduced between ages 60 and 63.

So far as his communication with more senior management is concerned, I can do no better than set out a short part of his cross-examination, in relation to which I accept what he said:

"Q. Okay. Can we just look at a few aspects of that description of early retirement? First of all, you are very specific in your statement that this idea related to active C Plan members only?

A. Correct.

Q. There was never any intention that this concept would apply to early leavers?

A. Correct.

Q. And would you agree with me that that is hardly surprising, because the concept that you explain in those five lines is the flexibility is you either retire and take a pension, or you stay with IBM and accrue further benefits; and that was never an option for an early leaver?

A. That's quite correct.

Q. And related to that, there was never any intention that early leavers should be able to draw an unreduced pension at 60?

A. No intention.

Q. You obviously liaised with/briefed your superiors, if I can call them that on the Management Committee, Mr Peach and, where relevant, Mr Morgans, about what was going on?

A. Yes.

Q. Do you think that you explained to them that this concept was dealing with active members only?

A. I'm not sure that I would specifically have said that this is for active members only. I probably would have simply said, "Look, we have been knocked back from having a normal retirement age of 60. We want our active members -- our active employees to go at 60, if we can possibly achieve it, and so this would be a way of doing that." I wouldn't, in that conversation, have talked about early leavers at all. That wouldn't have been part of any discussion.

Q. I see, because they weren't really on the agenda?

A. No.

Q. Would it be right, if one takes the reverse of that, that nobody in senior management proposed to you that deferred members should be given that right?

A. Nobody proposed that to me, no."

Mr Simmonds also relies upon Sir Leonard's evidence, which I accept, to the effect that the application of flexible retirement benefits to deferred members was never discussed by anyone that he knew of, including the Management Committee.

Culture concerning early retirement within IBM

There is a dispute about the IBM culture in relation to early leavers. Mr Simmonds submits that the culture was one under which early leavers were not viewed as particularly worthy of generous treatment. The position of the Trust Company before me is that by early 1982, IBM's desire to encourage employee turnover was such that it did not want to discourage potential early leavers from leaving IBM and thus becoming deferred members of the Main Plan.

Support for the Trust Company's position is, it is said on behalf of the Trust Company, to be found in the minute of a meeting of the board of the Trust Company on 29 March 1982. Sir Leonard highlighted

"the case of male employees, particularly long-serving IBMers, who retired and then died within a few years before reaching age 55 (i.e having not had the opportunity to opt for an immediate early pension). Their next-of-kin at present received only a benefits based on the legal minimum required by the Social Security Pensions 1975 Act, which would generally amount to a minimal sum. This also might act a deterrent for potential early leavers….."

If support is to be found for the Trust Company's case in that minute, it is, I consider, the most minimal support. Sir Leonard was there addressing the particular problem facing the dependants of an early leaver who died before age 55. He was certainly not addressing the benefits of early leavers generally or when those benefits would become payable. Rather, he was addressing a particular problem so that the reference to deterrent in that context lends scant support to the proposition that there should be encouragement for members to leave early, that is to say in the cases of men before age 55. This particular matter was addressed in the Initial Proposal by providing for an immediate spouse's pension on death in deferment, a pension far more generous than the statutory minimum. That aspect was retained in the Available Revisions and in the MIL dated 20 December 1982. But I do not see that as having anything to do with encouraging early leaving in contrast with ameliorating an identified hardship. Further, what Sir Leonard said about deterrent for potential early leavers does not demonstrate a policy of encouraging early leavers. As a matter of policy, IBM might not want to encourage people, generally, to leave early, but having identified a person who it is wished to encourage to leave (eg someone no longer perceived as pulling his or her weight), IBM would not want to find an obstacle in the path to persuading such a person to leave.

It is recorded that Mr Webb later on advised the meeting of the case of a member who had left service at age 57 without taking an immediate pension – it is not revealed whether this was because he chose not to ask for one or because a request for consent to retire was refused. The member was seeking payment of his pension, together with increases awarded to pensions in payment since he left service. Sir Leonard is recorded as saying this:

"…if such a request [for increases] were granted it would establish a new and perhaps unwarranted precedent. The normal practice was for members to take an early pension on retirement or on reaching age 55….."

The Trust Company relies on that as showing that early payment of a deferred pension would normally be permitted at age 55 indicating that there was no discrimination against deferred members. I am reluctant to take that single sentence as an indication of the policy applied by the Trust Company. But even assuming that it does so, there is nothing to suggest that the pension would not be reduced from NRD age 65 in the case of males or that the reduction would be other than cost neutral in contrast with the position in relation to immediate early retirement pensions which were reduced only from age 60 and at a more generous discount rate of only 3% per annum.

More assistance is to be found in Sir Leonard's memorandum on early retirement dated 28 April 1982, that is to say in relation to the N Plan before the C Plan proposals had been formulated. The policy document was produced as the result of a case where a leaver had been promised early payment of his pension at age 55 and was produced to clarify IBM's policy on the commitment to early payment of pensions for employees who leave service. There were two categories to consider: those leaving within 10 years of NRD (ages 50 to 60 for women and ages 55 to 65 for men) and those not within 10 years (under age 50 for women and age for 55 men).

In relation to the first category, those with 5 years' service would be entitled to a pension at NRD. The employee had the option to request the Trust Company to pay the pension early "ie commencing at the date of leaving". Employees in this category should be informed of this option prior to leaving. The option was subject to company consent but "generally, the company will give its consent, except where the employee is dismissed".

In relation to the second category, those with 5 years' service would be again entitled to a pension at NRD. No commitment to pay the pension early should be given. In particular cases, "where the separation is mutually beneficial" the Director of Personal and Corporate Affairs (that is to say Sir Leonard himself at that time) may authorise a commitment to early payment of the pension. A procedure for use in such cases was set out. Ex-employees "retain the right to request the Trustee to pay their pension early". Generally the Trustee will decline such requests unless there are extenuating circumstances.

This policy document draws a clear distinction in the treatment of members in the two age brackets:

i) Those over age 50/55 will normally be granted an immediate pension if they request it. If a leaver does not request it, or a request is refused, there is nothing to suggest that he or she will have any entitlement other than as an early leaver in accordance with the rules of the N Plan; in other words, they would be able to ask the Trust Company at some time in the future for immediate payment of the deferred pension. There is no reason to think that the policy for agreeing to such payment would be any different from that applicable to persons in the second category, as to which see ii) below.

ii) Those under age 50/55 were not to be given, on leaving, a commitment to payment other than at age 60/65 save in special situations agreed on a case by case basis. Such an early leaver would be able to ask the Trust Company at some time in the future for immediate payment of the deferred pension but generally this would be declined unless there were extenuating circumstances. Although the policy document does not state this, in such a case the pension would be discounted for early payment from age 60 for women and age 65 for men at cost-neutral discount rates.

The Trust Company submits that the differential treatment of actives and deferred members only applies in the period 10 years or more before NRD and that the difference in treatment is not all that marked since both actives and deferred members are subject to a requirement for consent which will be given only in limited cases.

I do not consider that that is a possible reading of the policy document. The differential treatment is not between actives and deferred members. Rather, it is a differential treatment of different classes of active members. Some will ordinarily be able to retire on an immediate pension, some will not be able to do so: it all depends on the age at which the member leaves service. Once it is established which category a leaver falls into, the policy document sets out the treatment perfectly clearly.

In that context, the difference is very marked, contrary to the submission of the Trust Company. A member leaving before age 50/55 with a deferred pension can only draw that pension early in extenuating circumstances. A member leaving after age 50/55 will ordinarily be given consent to take an immediate pension.

Moreover, the discounts for early payment are entirely different. There was built into the rules a preferential discount rate of 3% per annum for active members taking a pension under age 60, as compared with a deferred pensioner being allowed to take early payment of his or her pension before NRD (age 60/65) where the Trust Company had a discretion, in practice exercised to provide cost-neutral (and therefore less generous than 3%) discounting and where the discount would be from age 65 for men.

It is true that the policy document does not in terms deal with the position of a member who leaves after age 50/55 but who does not become entitled to an immediate pension. I have considered the position of such a member above, concluding that he or she would simply receive the benefits which the rules provide, which would be the same as those provided to a member leaving service before age 50/55.

I therefore agree with Mr Simmonds' submission that the clear meaning of the memorandum is that early leavers were treated less favourably than those seeking to retire early. Even though Sir Leonard was reluctant to accept that this discrimination existed, his own memorandum shows that it did. He said he was surprised that he had issued the document, thereby at least accepting, contrary to the Trust Company's submission, that the proper reading of the policy document is as Mr Simmonds submits. I consider that Sir Leonard's memorandum is more likely to be an accurate reflection of IBM policy than his recollection 30 years later of the general treatment afforded to members of the N Plan in different categories. I also consider it is a far more reliable guide than the one sentence from the minute of the meeting on 29 March 1982.

Further in relation to the attitude of IBM to early leavers, Mr Simmonds submits that the presentation in the Initial Proposal suggests that IBM did not view early leavers favourably. The benefits for such persons were presented as either the minimum legal requirement or that they would become legislative requirements in the near future. Mr Cawley agreed that it was fair to say that the way the Initial Proposal was presented was to meet the expected reaction from Armonk.

The Trust Company draws precisely the opposite conclusion from the Initial Proposal. It identifies two main benefit improvements for deferred members in the Initial Proposal: (i) guaranteed pension increases including in deferment and (ii) an immediate 50% spouse's pension upon death in deferment.

As to (ii), I have already considered this when dealing with the meeting on 29 March 1982 referred to in paragraph 339 above. It was in any case vetoed by Armonk so that no reliance at all can be placed on this factor to support the view that Armonk, at least, did not view deferred members in a favourable light. I do not consider (ii) supports the Trust Company's case.

As to (i), the Trust Company relies on the rationale given in the Initial Proposal which was essentially that it was important to include revaluation to meet the perceived injustices perpetrated on the early leaver with frozen vested rights. The Initial Proposal thus, as it is put on behalf of the Trust Company, "addressed issues which inhibit employees from seeking employment outside IBM (ie treatment of vested rights pensions), thus assisting attrition and vitality. Payment of deferred pensions at 60 plus revaluation in deferment would assist by removal of focus on the present 'frozen' nature of vested rights pensions".

I accept that the Initial Proposal shows that the Management Committee was not positively hostile to giving early leavers a more generous benefit than hitherto. The improved treatment was, however, part of a package designed to provide better pension provision overall for the workforce. There was no suggestion that existing deferred members were to be entitled to transfer to the C Plan and thus obtain better rights than they already had. Nor was there anything in the Initial Proposal to suggest that leavers under age 50 were to be treated in a favourable way or that a leaver over that age who did not take an immediate pension would be treated in a favourable way. So although there was, in IBM in the UK, no positive hostility to deferred members, nor were they viewed particularly favourably. Thus UK management did not, so far as the evidence disclosed, seek to hold out against Armonk when it came to the removal of guaranteed increases to pensions in excess of GMP during deferment in the revised proposal.

The position in relation to Armonk is different. I consider that the evidence establishes reasonably clearly that Armonk was not at all favourably disposed to early leavers by which I mean members who have left service other than on an immediate pension. Whether it is correct to describe this as "hostility" does not matter; the point is that the culture within Armonk at the time of the formulation and presentation of the Initial Proposal and thereafter was not favourable to deferred members and was not one under which a person who had actually left service without an immediate pension should be able to draw the pension early (leaving aside the question of whether early means before age 63 or before age 60). In that context, the witnesses have given evidence that early leavers who had moved to work for competitors were sometimes viewed as traitors (to use Mr Barrett's description) at least at a senior level or as second class citizens (to use Mr Ellis' description). As Mr Cawley put it, Armonk was "not very supportive of such members": Armonk would wish to see provided only that which was a legal requirement.

Mr Gooch's evidence on this point is interesting. In paragraph 12 of his witness statement he said that, as a consequence of his work with IBM over a period of time, he had formed the impression that, culturally, the US had certain views about pension provision, recalling being told at some time well before by an American management figure that IBM was not at all well-disposed to early leavers and would have preferred to avoid the need to pay those early leavers pensions. I cannot really place any weight on that since it is too vague and amorphous. But it is at least consistent with Mr Gooch's own understanding that deferred members were not to be entitled to draw their benefits at age 60.

Although I have been referred to the evidence of Mr Leesmith and Mr Martin James by the Trust Company, I do not derive any assistance from them on this cultural issue. What I can say, however, is that their evidence does not, in my view, assist the Trust Company in establishing a positive case to show an objective intention on the part of either Holdings or the Trust Company that the 1983 Trust Deed and Rules should include an entitlement for deferred members to take their benefits at age 60.

The strongest evidence in support of the view that deferred members were intended to have an entitlement to take their benefits at age 60 without any need for consent is to be found in the witness statement of Mr Morgans. I have to view that with some circumspection since he was not available for cross-examination and none of the points raised above could be put to him.

Moreover, it is not at all clear precisely what time or times he is referring to in relation to some of the matters which he deals with. Beginning at paragraph 74 of his witness statement is a section headed "Preservation". That paragraph states that the personnel policy (ie to encourage staff turnover or attrition as it has been referred to) which underpinned retirement at age 60 did not apply with the same force to deferred members (who by definition had already left employment with IBM). Mr Morgans then stated explicitly (at paragraph 75) that the effect of the preservation requirements was not overlooked "at the time". A reading of the entirety of the whole paragraph leaves me with the clear impression that "the time" he referred to was in the period leading up to the Initial Proposal. As he stated, and as has been recognised on all sides including by Mr Cawley, the Initial Proposal itself drew attention to the fact that deferred members had to be treated in the same way as members retiring at age 60, the NRD under the Initial Proposal. The question, however, is not what the Initial Proposal provided for but what the intention was when the revised proposal came to be implemented in July 1983.

As to that, in paragraph 76 of his witness statement, Mr Morgans stated that, at some point "prior to the establishment of the C Plan", he did advise Paris that "deferred members would have these rights". And he remembered in particular discussing the issue with Sid Handler (Controller of IBM Europe) in Paris. When Paris queried the position, he recalled explaining that he had advised that the cost of granting the rights to deferred members was minimal. He was able to do so because:

i) he recalled that Clay & Partners prepared costings and it was established that the cost of granting the benefits to deferred members was minimal and

ii) having approved the cost of the C Plan with an NRD of 60 under the Initial Proposal, he was not concerned that the revised proposal would also have the effect of granting deferred members the right to take their benefits at age 60.

It is not difficult to see that Mr Morgans would have been challenged on much of this and in particular the detail of the timing. It may be that "these rights" was an ability to take an early pension only with consent. It may be that he was confusing the rights arising under the Initial Proposal with those arising under the revised proposal. Even if a conversation did take place with Mr Handler in the context of the revised proposals, he said that he was under the impression that the cost of the deferred benefit (by which I think he must have meant the cost as compared with the N Plan for deferred benefits) would be minimal. Clearly, however, neither Mr Gooch nor Mrs James would have told him that the cost was minimal other than on the basis that the benefit would be payable at age 63, and even then there is no evidence to suggest that the cost of C Plan deferred benefits at age 63 was a minimal increase over the cost of N Plan deferred benefits at 60 (females) and 65 (males).

Mr Morgans did not state the source of his understanding. There is in fact no evidence at all that Mr Morgans ever spoke to either Mr Gooch or Mrs James about this, and no evidence that he spoke to anyone else at Clay & Partners about it. There is nothing to suggest that he could obtain any knowledge about the revised proposals and their cost from anyone other than one or more of Sir Leonard, Mr Cawley and Mr Ellis. Each of them, however, must have thought (if they thought about it at all) not only (although perhaps erroneously) that the revised costings reflected the cost of flexible retirement for active members but also (correctly) that they reflected the cost of deferred benefits payable at age 63 since each of them thought that active members would have an entitlement to retire without consent between ages 60 and 63 but would have no right to a deferred pension until age 63.

If Mr Morgans' evidence is to be read as saying that he expressly discussed with Mr Handler an intention that deferred members should be entitled take their benefits at age 60, it would be very surprising indeed if he had not discussed that with one or more of Sir Leonard, Mr Cawley and Mr Ellis, but they say they knew nothing of it.

Finally, I note that Mr Morgans stated in paragraph 77 of his witness statement that he made that point to Paris that it would be inconsistent with IBM's "general policy of non-discrimination" to grant a benefit to active members but not to deferred members. For reasons already given, I do not consider that there was such a policy, which calls into question whether Mr Morgans actually did make the point.

In any case, even if Mr Morgans did raise the point with Paris, there is no evidence to suggest that the point was raised by him with the Management Committee, the board of Holdings or the board of the Trust Company or, indeed, with anyone at all. The evidence of others, in particular Sir Leonard, Mr Taylor and Mr Caddick is inconsistent with the conclusion that he did so. I find as a fact that Mr Morgans did not say anything to the Management Committee, the board of Holdings or the board of the Trust Company or to Mr Cawley or to Mr Ellis which would have led them to think that he, Mr Morgans, understood the intention of the C Plan was to provide benefits for deferred members at age 60 and nothing which would have led them to change their intention that those benefits would be payable at age 63.

My conclusion in the light of all of the evidence and the discussion above, and I find as a fact, is that it was the intention of Holdings and of the Trust Company, when authorising the execution of the 1983 Trust Deed and Rules, that deferred members would be entitled to their benefits, at least if they were to be paid without actuarial reduction, only at age 63. I find that that was the intention of each board objectively ascertained. Even if it was Mr Morgans' subjective intention that the benefit should be payable at age 60, this was not communicated to those boards or to the Management Committee.

It follows from that conclusion that I reject both limbs of the submissions of the Trust Company namely (i) that Holdings and the Trust Company had a positive intention to include an entitlement for deferred members to take their benefits at age 60 and (ii) that Holdings and the Trust Company had no positive intention to exclude that benefit. I would only add that, even if I had some doubt about the correctness of my conclusion (ii), I have no hesitation in rejecting the Trust Company's submission on (i).

Consequences of findings of fact

What follows from my conclusions as a matter of law? On the one hand, the objectively ascertained intention was that active members should be entitled to retire on an unreduced pension from ages 60 to 63 without employer consent. On the other hand, the objectively ascertained intentions were (i) that a deferred member should be entitled to draw his or her pension only at age 63 and (ii) that in cases where early payment of a deferred pension was permitted by the company and the Trust Company, the pension would be actuarially reduced from age 63. But those intentions were, in the light of the Preservation Requirements, inconsistent and irreconcilable.

For Holdings, Mr Simmonds argues that the Trust Company's claim for rectification should fail in its entirety even if the requisite common intention is established in relation to active members. The argument is as follows:

i) The first step of the arguments runs this way. Holdings and the Trust Company had a composite or overarching intention in relation to retirement under the C Plan. There was flexible retirement age for active members, ages 60 to 63, but only one age for deferred members. Any claim for rectification of the 1983 Trust Deed and Rules must be in a form which gives effect to the totality of Holdings' composite or overarching intention. There is no basis on which the Court can "prefer" one element over the other or treat one element as more "important" that the other.

ii) I interpose here to add that perhaps retirement age is the wrong label: it is a correct use of language to describe leaving service with an entitlement to an immediate pension as retirement and the age at which it is done as retirement age, and this is so even if the member concerned takes another full-time job with another employer. He has retired from IBM but has not retired 'full stop'. In relation to a deferred member, leaving service is not to be equated with retirement in any real sense and when the age comes at which the member draws his pension, he may long since have ceased gainful employment or, in contrast, he may have been employed during the entirety of the period of deferment and remain employed after he starts to draw his deferred pension from the C Plan. At that time he neither retires from IBM nor does he retire 'full stop'. The concept of retirement is different in each case.

iii) The second step of the argument is this. In 1982/83 the Preservation Requirements had the effect that, if active members had an unqualified right to an unreduced pension at 60, the Trust Company was under a statutory duty to take such steps as were open to it to ensure that members enjoyed the same rights in deferment. Even though the Preservation Requirements were (and are) not overriding, there was in 1982-83 an enforcement mechanism available which would have enabled the Trust Company to secure the extension of such rights to deferred members even in the absence of co-operation by the employer, namely by means of a modification order made by the Occupational Pensions Board: see paragraph 36 above.

iv) The third step of the argument is this. The only way, consistently with the Preservation Requirements, for a pension scheme in 1982/83 to have provided an unreduced pension at 60 for actives but not for deferred members was to ensure that actives' rights were subject to a consent requirement (as eventually recommended by MPA in the present case). This would have ensured that the relevant members' "normal pension age" for preservation purposes was not 60 but the stated NRD of 63 (because the consent requirement was a "special provision" as to early retirement which fell to be "disregarded": see paragraph 31 above).

v) The fourth step of the argument is this. Rectification of the 1983 Trust Deed and Rules to remove the company consent requirement in respect of actives retiring between ages 60 and 63 would conflict with the actual intentions of Holdings and the Trust Company at the time in respect of deferred members.

Accordingly, the reality is that the parties' actual unvarnished intentions were legally incapable of fulfilment in the 1983 Trust Deed and Rules. For that simple reason, the rectification claim should fail.

Mr Simmonds adds the following in support of this line of argument:

i) On the evidence he submits that it seems likely that, if the preservation issue had come to light and been addressed in 1982 in relation to the revised proposals, the 1983 Trust Deed and Rules would still have been executed in their unrectified form, incorporating the company consent requirement.

ii) Mr Cawley's evidence is said to be pivotal. Flexible retirement was his idea and there is no reason to think that his approach would not have been adopted by senior management. Mr Cawley saw the matter principally as a communications issue. If IBM had known that a company consent requirement was necessary for technical reasons (ie to ensure that deferred members did not acquire similar rights), it would have announced this in the MILs. This would not have been a problem because employees would have trusted IBM to deliver the benefits summarised in the MILs and the take-up rate would have been unaffected. As Mr Cawley said in response to a question from me:

"If we had had to, for legal reasons, overlay that with a company-consent requirement, it wouldn't have in any way affected the benefit structure."

iii) Mr Simmonds suggests that it is significant that, at no stage in his evidence, did Mr Cawley consider for a second that the solution to the preservation problem would have been to extend enhanced early retirement rights to deferred members.

iv) But this, in my view, carries no more weight than Mr Ellis' perception of "consent" namely that it was to be automatic in the sense that it could never properly be refused. The truth is that what Mr Simmonds here relies on is no more than Mr Cawley's speculation. If the matter had been addressed and taken into account, it may be that the C Plan would have been structured in an altogether different way, perhaps with a single retirement date of, say 62, with an option to leave early with the consent of the employer and an option to serve longer (either with or without consent). But that is even more speculative than Mr Cawley's speculation. None of it assists in answering what is, in the end, a legal issue once it has been decided as a matter of fact that two inconsistent intentions were held.

The Trust Company says that Mr Simmonds' argument is misconceived. Even if (as is my finding of fact) Holdings and the Trust Company had a positive intention (in the objective sense required for rectification) that deferred pensions should come into payment only at age 63, all that would mean, it is said, is that the intention in 1982-83 (incorrectly recorded in the 1983 Deed) was to confer a right of flexible retirement on actives but not deferred members. There is nothing to stop the Court from rectifying the 1983 Trust Deed and Rules to reflect those intended terms, ie a right of flexible retirement for actives but not deferred members. The grant of such rectification would bring the formal documents into conformity with the true terms intended in 1982-83. The fact that, both in 1982-83 and now, the Preservation Requirements would then give rise to consequences in relation to deferred members is irrelevant; that is simply a consequence of the terms that were intended in 1982-83. That consequential effect would have arisen in 1983 if the terms had been recorded correctly in the 1983 Trust Deed and Rules, just as much as it would arise now if they are rectified so as to correct the erroneous drafting. There is no inequity in rectifying them.

Consider, then, the position in, say, October 1983. The C Plan had by then been operative for some months from July 1983. Members had transferred from the N Plan to the C Plan. Contributions had been deducted from salary. It is clear that members were expecting, and in my view were entitled to expect, that the benefits which they had been told they would receive would be payable in due course. One of those benefits was an entitlement to an unreduced pension from an age selected by the member between ages 60 and 63. As I have already noted. this was a benefit which was already a feature of the C Plan months before the 1983 Trust Deed and Rules came to be executed. It was not simply that Holdings and the Trust Company intended to provide that benefit when the C Plan came to be formally adopted.

It was a consequence of that entitlement, even before execution of the 1983 Trust Deed and Rules, that the Preservation Requirements applied with the result that members should have been provided with early leaving benefits payable at age 60. It would not have been open, in my judgment, to Holdings to say to a member in October 1983 that he or she would not, after all, be entitled to such benefits because the impact of the Preservation Requirements had not been appreciated so that the benefit had been provided by mistake. Nor could Holdings justify such a stance making effectively the same point by saying that the member was only intended to receive a deferred benefit at age 63, a result incompatible with flexible retirement for active members. Holdings could not, in my judgment, have said that that had been provided by mistake. On the contrary, it had been provided quite deliberately. The mistake was in the failure to appreciate the consequence of providing it.

That is essentially the argument which is put in relation to the 1983 Trust Deed and Rules themselves. The (objective) intentions were that there should be flexible retirement for actives but payment only at age 63 of pensions for deferred members. The conceptually perfectly compatible provisions were prevented only by the Preservation Requirements. Rectification in relation to flexible retirement for actives should be granted, leaving the law to impose whatever consequences flow for early leavers from the provision for flexible retirement for actives.

In my judgment, the intention that the C Plan should provide for payment of deferred pensions only at age 63 does not preclude rectification to provide flexible retirement for active members. I reach that conclusion based not only on my findings concerning the objective intention that it was the intention of Holdings and the Trust Company not only to include flexible retirement in the 1983 Trust Deed and Rules but also their intention to reflect the way in which the C Plan had in fact been implemented in July 1983. Such implementation was in accordance with the December 1982 announcement and the MILs which provided for flexible retirement. I reject Mr Simmonds' argument that there was a composite or overarching intention under which the two benefits were necessarily linked. There is no evidence that that is how anyone thought of matters. It is not an argument which can survive my analysis of the intention as including an intention to reflect the actual implementation and operation of the C Plan since July 1983. On my analysis, the intention to provide flexible retirement must prevail.

But even if I am wrong in that analysis, I would still prefer the Trust Company's argument relying only on the fact that it was the intention to provide flexible retirement. If the C Plan were rectified to provide for flexible retirement, it would be left, as a matter of drafting, still providing for the payment of deferred pensions only at age 63 (subject to earlier payment with consent and actuarially reduced). The Preservation Requirements were not, and are not today, overriding. Even if the C Plan had been rectified on, say, 1 January 1984, a member would have had no right to claim, as against the Trust Company, payment of his or her deferred pension until age 63.

That would not have been an end to the story: at that time, it would have been possible for the C Plan to have been amended even without the involvement of Holdings since the OPB had power to authorise the Trust Company to amend it, even if Holdings was not a "manager" of the Main Plan, or itself to do so (see paragraph 36 above). But unless and until such an amendment was made, the only enforceable right which the member would have had against the Trust Company was payment of the deferred pension at age 63.

Accordingly, I do not consider that the Preservation Requirements provide a defence to the Trust Company's rectification claim to include flexible retirement.

New Joiners (1)

During the course of argument, I raised some concerns about the position of New Joiners (that is to say, members of the C Plan joining it after execution of the 1983 Trust Deed and Rules, other than those who transferred, whether in 1983 or later, from the N Plan). I wish for the moment to focus on New Joiners who joined the C Plan before the replacement of the 1983 Trust Deed and Rules on 1 October 1990. During the period December 1983 to October 1990, the Employee Handbook went through new editions in October 1984 and September 1985. I deal with these in a little detail in paragraphs 479ff below. The provisions of the October 1984 Handbook and the September 1985 Handbook in relation to the C Plan were unchanged from the November 1983 Handbook.

A new Pension Booklet appeared in October 1987. It contained the following under the heading "A summary of your benefits if you join IBM at around Age 25" which included the following:

i) "On retirement at 63": a pension of "around 2/3 of your final pensionable earnings";

ii) "You may also …… retire at any time from 53 if you have the consent of IBM";

iii) Under the heading "When you may retire" it was provided that "Normal retirement is at 63. With the consent of IBM, you can retire early, provided…you are at least 53";

iv) The Trust Deed and Rules were to "override" the provisions of the Pensions Booklet.

In May 1988, a supplement to the Pensions Booklet was issued called "Your Pension Choices". Against the marginal heading "Early retirement pension" appears "Payable anytime between age 53 and 63 with IBM's consent". The supplement was "for guidance only". A further supplement "Your Pension Choices in IBM" appeared in August 1990; this provided that the Trust Deed and Rules were overriding.

It is clear that the reader of the Pensions Booklet and the supplements would, absent any other material, understand that consent was needed for any retirement under age 63, something which would have been confirmed by reference to the then extant 1983 Trust Deed and Rules.

In the light of those documents, I can restate my concern in this way. Even if the intention in December 1983 was that the C Plan should provide for flexible retirement, it in fact did not do so. A New Joiner would not have thought, by looking at the 1983 Trust Deed and Rules, that he or she might be entitled as of right to retire at any age after 60 on an unreduced pension. A New Joiner joining before October 1987 might have thought that there was such a right if the November 1983 Handbook was to be taken at face value, but that expressly stated that the Trust Deed and Rules took precedence and those, in their unrectified form at the time, contained a consent requirement. New Joiners from and after October 1987 would have received the Pensions Handbook which referred to a consent requirement. I was concerned that there must have come (i) a time when the individual staff members within IBM whose functions included agreeing, on behalf of IBM, employment contracts with New Joiners, included only individuals who had no knowledge of the right to flexible retirement and (ii) a time when even the decision makers (ie the Management Committee and the board of the Trust Company) for the time being were persons who did not appreciate that there had ever been an intention to provide the flexible retirement option. It seemed odd to me that the Trust Company should be able to obtain a remedy, rectification, at least vis a vis New Joiners after those times, which would give those New Joiners a right to a pension which they had never expected and which the staff acting on behalf of IBM in employing them and, more importantly, the decision makers themselves, did not appreciate they would be entitled to.

The Trust Company has two responses to my concerns.

i) The first response is that New Joiners became members on whatever the true provisions of the C Plan were, and (subject to any defences) are accordingly affected by rectification of the 1983 Trust Deed and Rules whether it is adverse or advantageous to them. On the hypothesis that the rectification case succeeds in principle, I would have found that the true intention of the parties to the 1983 Trust Deed and Rules when it was executed was that C Plan members should enjoy a right of flexible retirement, so there is no windfall or inequity in New Joiners having this benefit also.

ii) The second response addresses my concern on the basis that there might have been an extrinsic contract (typically a contract of employment) to inhibit and modify a flexible retirement right. It is said that those are not the facts of the present case and that the question raised by me does not arise on the facts. Essentially, this is to say that IBM's contractual counterclaim is misconceived. Whether or not there is a contract is the issue which arises on IBM's counterclaim which I will come to later. But jumping ahead, my conclusion is that IBM fails to establish any contracts under which employees are precluded from claiming an unreduced pension between age 60 and 63 without employer consent. Insofar as my concern was based on the possible existence of a contractual provision, I therefore agree that there is no basis for that concern.

In relation to the first response, it is said, that the intention to provide flexible retirement for transferring members applied equally to New Joiners. I have already held that to be the case at the time of the execution of the 1983 Trust Deed and Rules. Had this claim for rectification been brought a few months after the execution of the 1983 Trust Deed and Rules, then it would be very difficult indeed for Holdings to maintain that it had changed its intentions from those which it had at the time of execution or that the relevant personnel and decision makers were unaware of the intention. In that sense, the Trust Company is clearly correct that, if rectification were granted, there would be no unexpected windfall for New Joiners to obtain the flexible retirement right that was always intended for them and no injustice to IBM.

But the position of New Joiners soon after the execution of the 1983 Trust Deed and Rules was not quite the concern which I expressed. I was more concerned with the much later times just identified when nobody involved with pensions within IBM had any idea that it had ever been intended to provide flexible retirement.

The Trust Company relies on the decision of Lawrence Collins J in AMP v Barker [2000] EWHC 42 (Ch). It is put this way in the Trust Company's closing:

"This is the converse of the point considered by Lawrence Collins J in AMP v Barker. In that case, it was argued that new joiners should take free of an equity to rectify (which was adverse to members but favourable to the employer). Lawrence Collins J rejected the argument at [79], holding that where a pension scheme is rectified, members should be treated as purchasers of "such rights as were properly granted under the rules", by which he meant under the rules as rectified. The same reasoning applies in the present situation: new joiners signed up to such rights as were properly granted under the rules, and if, as in the present case, the rights properly granted were more favourable than those erroneously recorded in the Deeds, the new joiners should be entitled to enjoy those rights."

I do not consider that, on a careful reading, what Lawrence Collins J is in fact supportive of the Trust Company's case. In that case, AMP, an insurance company, acquired NPI, another insurance company. The rules of the NPI pension scheme for its own staff had been amended to provide incapacity benefits but the change of rule unintentionally provided enhanced benefits to all employees leaving with 2 years' service. There was, however, no suggestion that the intention of the relevant decision makers might have changed since the making of the amendment. And what is more there was no reason to think that the sixth defendant, the representative defendant, was a new joiner of the sort I am considering (ie a person who became a member after the amendment in question) although he was joined to put all the arguments against AMP's claim. The argument addressed in [79] of the judgment was that members had acquired vested or accrued rights, albeit as a result a mistake, for value, rights which could not be removed by rectification. It was accepted, however, that before the mistake was discovered, early leavers or those in service had no expectation or intention of receiving enhanced benefits as a result of the rule changes. The Judge acknowledged that they gave consideration for their pension rights, but held that, having given no additional consideration for the "rights" which the rule changes mistakenly conferred on them, it was wholly unrealistic to treat them as purchasers of anything other than the rights properly granted to them.

It was not possible, therefore, as I read the judgment, for the members (who were not, it will be recognised, parties to the amending document) to resist rectification on the basis that they were purchasers of an interest and should take free of the equity of rectification. There is no parallel to that in the present case in relation to my concern. Rather, the hypothesis which I am now considering is that neither the employee nor anyone within IBM had a positive intention that New Joiners should have a flexible retirement option and those who were concerned with pension matters would have had no reason to think that the 1983 Trust Deed and Rules were mistaken. This is entirely different from the position in AMP v Barker where everyone concerned would have known that the amendment was mistaken once its knock-on consequences for all early leavers was appreciated. I do not consider that AMP provides an answer to my concerns.

I have not found this an easy point to resolve but in the end I do not consider that my concerns are justified. They certainly do not justify refusing rectification altogether. It is clear, to my mind, that the 1983 Trust Deed and Rules should be rectified at the very least to provide flexible retirement for members of the C Plan who transferred from the N Plan or other Plans within the Main Plan. And the same must apply to New Joiners for some time after the execution of the 1983 Trust Deed and Rules.

For New Joiners joining after the times which I have identified (ie when nobody concerned with pensions within IBM knew of an intention to provide flexible retirement), there are two reasons why my concerns are not justified. The first is the impossibility of identifying the relevant time with any accuracy. Indeed, even by the time of the 1990 Trust Deed and Rules, Sir Leonard was still on the board of the Trust Company so that it cannot be said that everyone concerned was unaware of the original intention. The second reason is that my concern goes to a defence to a rectification claim. The starting point is that the Trust Company has succeeded in its case concerning the intentions of the Trust Company and Holdings in December 1983 and the question is whether rectification to give effect to that intention should be refused. The onus is not, in my view, on the Trust Company to demonstrate a positive continuing intention from time to time on the part of the Trust Company and Holdings that there should be a right to flexible retirement; rather, the onus would be on Holdings to show a positive case that, had the mistake been realised, New Joiners would not have been afforded a right to flexible retirement. In that context, it is worth noting that there is no evidence at all that the decision makers from time to time (the Management Committee and the boards of Holdings and the Trust Company) ever positively considered the issue prior to October 1990. Accordingly, they never formed a positive intention, contrary to the previously existing intention, either that flexible retirement should not be available to New Joiners or, to put it the other way round, that early retirement should be available but only with employer consent.

I should add that even if my conclusion (as to which see the discussion later) that IBM's contractual case fails is wrong, then the Trust Company should still be entitled to obtain rectification. This does not mean that IBM is thereby deprived of its contractual claim. Rather, it is that that claim would fall to be assessed against the terms of the 1983 Trust Deed and Rules, as rectified.

Conclusion on rectification of the 1983 Trust Deed and Rules

My conclusions on the issue of rectification of the 1983 Trust Deed and Rules are (i) that they should be rectified to include flexible retirement for all members whenever they joined the C Plan but (ii) they should not be rectified so as to give deferred members the right to payment of their deferred benefits from age 60; they are payable only from age 63 under the express provisions of the 1983 Trust Deed and Rules.

In consequence, subject always to IBM's contractual claim, the Preservation Requirements would have applied. This may have given the Trust Company or the members of the C Plan remedies against IBM, but rectification was not one of them.

Rectification of the later deeds

In the light of my conclusions concerning the 1983 Trust Deed and Rules, the question then arises whether subsequent Trust Deeds and Rules should be rectified as well. As I have explained at paragraph 262 above, the focus is on the 1990 Trust Deed and Rules.

IBM's position is that there was a wholesale change in the personnel involved with pension matters within both Holdings and the Trust Company between 1983 and 1990; that is not quite accurate, as we will see. But there is no doubt that the decision-makers in 1990 were significantly different from those in 1983 and it is not necessarily the case that the intentions (where not expressly changed in relation to matters actually addressed) in 1983 continued unchanged. No-one has suggested that there is such a thing as corporate memory for the purposes of rectification which would impute to the human decision makers knowledge of the intentions of their predecessors on the board of Holdings and the Trust Company and on the Management Committee. When the search for the objective intentions of the relevant individuals from late 1990 onwards is carried out, it is not necessarily the case that, objectively, that which was objectively established as the intentions of their predecessors is necessarily to be carried forward unless it is shown positively that a different intention had been formed.

In a nutshell, the parties' positions are these:

i) IBM's approach is that the decision makers in relation to the 1990 Trust Deed and Rules as a collective body were unaware that the contents of the 1983 Trust Deed and Rules did not accurately reflect the intentions of Holdings and the Trust Company. They had no intention of providing for something of which they did not know and which was not contained in the governing documentation. Therefore the 1990 Trust Deed and Rules reflected their actual intentions and should not be rectified.

ii) The Trust Company's approach, in contrast, is that neither Holdings nor the Trust Company intended to make any material change to the early retirement provisions of the C Plan. Their intentions, on an objective basis, are to be ascertained by reference to the provisions of the 1983 Trust Deed and Rules as Holdings and the Trust Company had intended them to be and thus to replicate in the 1990 Trust Deed and Rules the rectified provisions of the 1983 Trust Deed and Rules.

The Trust Company's closing puts its submissions in slightly greater length in this way:

i) the mistake in the 1983 Trust Deed and Rules went unnoticed by the Trust Company and Holdings;

ii) like the 1983 Trust Deed and Rules, the 1990 Trust Deed and Rules were executed by the Trust Company and Holdings;

iii) the 1990 Trust Deed and Rules were presented as, and intended to be, an administrative consolidation exercise to tidy up the trust documents of the Main Plan; it was not the occasion for reinvention or redesign of the benefits under the C Plan;

iv) the decision-makers at board level who decided to execute the 1990 Trust Deed and Rules never reconsidered flexible retirement or discussed changing it;

v) the parties to the 1990 Trust Deed and Rules had no intention to introduce a consent requirement for the first time;

vi) the intention of the parties to the 1990 Trust Deed and Rules was not adversely to affect or change existing rights under the C Plan but (subject to making limited changes which were not relevant to flexible retirement) was to carry existing rights forward, whatever the true nature and extent of those rights was, and not to reduce or restrict those rights;

vii) consequently, the 1990 Trust Deed and Rules are now liable to be rectified so as to preserve and provide for the right of flexible retirement which had been intended under the 1983 Trust Deed and Rules and which is contained in that Deed, once rectified.

In relation to IBM's case, it is observed by the Trust Company that what it describes forensically as the adventitious effect of nullifying a successful claim to rectify the 1983 Trust Deed and Rules derives from the happenstance that, as a matter of form, the later versions of the Trust Deed and Rules were all definitive deeds and rules which were expressed to replace the previous provisions in their entirety. But the reality, the Trust Company submits, is that, in substance, the later versions of the Trust Deed and Rules were amending instruments and their effect could have been achieved by keeping the 1983 Trust Deed and Rules in place and making amendments from time to time. If that had happened throughout, there would, it is correctly said, now be no way in which IBM could assert that the effect of rectification was wiped out from 1990 onwards. I note as an aside that there would be no reason to rectify the later versions in relation to deferred pensions either, leaving the time for payment of an unreduced deferred pension at age 63.

Mr Simmonds submits that the approach of the Trust Company confuses two different types of mistake which have different legal consequences. There is the sort of mistake which engages the remedy of rectification, that is to say a mistake in the terms of the document as a result of which it fails to reflect the relevant intention, in the present case, the objective intentions of the Trust Company and Holdings in the sense which I have discussed at paragraphs 24ff above. Then there is the sort of mistake which is not a mistake in the document itself but is a mistake about the circumstances in which the document is made and which results in the document having consequences different from that which it would have had if the mistake had not been made.

Mr Simmonds says that the error in the present case falls into the second category. When the 1990 Trust Deed and Rules were executed, the only mistake made by the board of the Trust Company and the Management Committee was, he says, that they acted in ignorance of the rectifiable error. They intended to execute the 1990 Trust Deed and Rules in the form in which they were in fact executed intending the members to have the benefits which they in fact provided. If they had known about the mistake in the 1983 Trust Deed and Rules, they might have executed the document in a different form. The word "might" is my word. Mr Simmonds actually says that they "would" have executed it in a different form, and he is almost certainly right about that because, so far as members of the C Plan who transferred from the N Plan were concerned, it would not have been permissible to affect adversely their accrued rights. But it seems to me (and hence "might") what they would have thought about New Joiners is a matter of speculation. They may have decided, having learnt of the intentions of the decision-makers in 1983, that there was no reason at all to depart from those intentions; or they may have decided that things had changed since 1983 and that New Joiners should not be entitled to flexible retirement.

It does not matter which, since they did not appreciate the point and it was not addressed. Mr Simmonds goes on to say that, until recently, this sort of error might have been categorised as a classic Hastings-Bass error but, following Pitt v Holt [2011] EWCA (Civ) 197, [2011] 3 WLR 19, that approach may not be open (but I add that it may be open: I do not propose to go into the ramifications of Pitt v Holt in this judgment, interesting as the topic is). He submits that the mistake might nonetheless vitiate the 1990 Trust Deed and Rules (in whole or in part) but it does not give rise to a claim for rectification of the 1990 Trust Deed and Rules.

Mr Simmonds relies on the summary of the law as set out in Snell's Equity (32 ed) at 16-015 on pp 457-8:

"There can thus be no rectification if the omission of a term was deliberate, even if this was due to an erroneous belief that the term was unnecessary or that it was sufficiently dealt with in the antecedent oral agreement, or that the term was illegal, or a breach of covenant, and similarly if the instrument intentionally contains a provision which in fact means something different from what the parties thought it meant. Rectification ensures that the instrument contains the provisions which the parties actually intended it to contain, and not those which it would have contained had they been better informed."

His argument is that the same approach applies where (as he asserts is so in the present case) the document intentionally contains a provision in the erroneous belief that it reflected the members' rights, erroneous because nobody realised that the 1983 Trust Deed and Rules were themselves in error. He refers to Frederick E. Rose (London) Ltd v William H Pim Jnr & Co Ltd [1953] 2 QB 450 as the classic illustration. Lord Hoffmann in Chartbrook at [60] referred to that case, describing what Denning LJ said in his judgment at p 461 as the "clearest statement" of the correct approach. In Rose v Pim, the plaintiff and the defendant thought that "feveroles" meant "horsebeans" and that "horsebeans" meant "feveroles". They contracted, on that mistaken basis, for the sale and purchase of "horsebeans". The parties' common intention, objectively ascertained, was to buy and sell "horsebeans". This mistake was not a sufficient basis for rectification of the contract because (see at p 462 second paragraph):

"There was, no doubt, an erroneous assumption underlying the contract – an assumption for which it might have been set aside on the ground of misrepresentation or mistake – but that is very different from an erroneous expression of the contract, such as to give rise to rectification."

"On the assumption that "feveroles" are different from "horsebeans", it cannot be said that the parties agreed on the sale of a commodity of the separate existence of which they had no knowledge."

And so in the present case, he argues that the parties to the 1990 Trust Deed and Rules executed it intending that it should contain the provisions which it did on the (mistaken) assumption that it accurately reflected the "true" provisions of the C Plan. The document therefore reflected their actual intentions. The mistake may give rise to some remedy, but it is not rectification.

It might be thought that the decision of Lawrence Collins J in AMP is a bar to such an argument. Mr Simmonds submits that it is not. In that case, AMP, an insurance company, acquired another insurance company, NPI. The NPI pension scheme for its own staff had been amended to provide incapacity benefits, but the change of rule unintentionally provided enhanced benefits to all employees leaving with two years' service. This was because it had been overlooked that the calculation of leaving service benefits was to be calculated as if the member was leaving service on account of incapacity. The Judge found that the (objectively ascertained) common intention of the Employer and the Trustee was to make a very specific and limited improvement to incapacity benefits. As he put it at [52] of his judgment:

"It was put to the Trustees that they had no intention to break the link between early leaver benefits and incapacity benefits. But they had no knowledge of any such link and they had no intention in relation to it, except that they intended to improve incapacity benefits and nothing else……"

Rectification was granted even though the amending deed contained the very wording which was intended. The Judge identified the relevant principle in this way at [71] of his judgment:

"…rectification may be available if the document contains the very wording that it was intended to contain, but it has in law or as a matter of true construction an effect or meaning different from that which was intended."

So the wording used did not, because of the link to deferred benefits, have the effect which was actually intended (on an objective basis) by the parties which was to change only the incapacity benefit. By contrast, Mr Simmonds submits that in the present case, the 1990 Trust Deed and Rules did have the effect which was actually intended (on an objective basis). I will consider in due course whether that is the correct conclusion to draw from the evidence.

Mr Simmonds also dealt with two unreported decisions, Premier Farnell plc v Canham (unreported, 24 March 2010, HH Judge Simon Brown QC, Leeds District Registry) and DAS Services Ltd v Gait-Parker [2010] EWHC 2592 (Ch), on which the Trust Company relies where later pension scheme trust deeds which replicated a rectifiable error in an earlier deed were themselves rectified in like manner. Both decisions were on uncontested summary judgment applications. The issue raised by IBM in these proceedings was not argued. He says that these cases are therefore of no real assistance. Moreover, I add, that the facts of each of those cases are very different from those of the present case. The decisions do show that a serial error contained in successive deeds is capable of elimination by serial rectification but they do not establish that what is a rectifiable mistake in one set of rules is a mistake, let alone a rectifiable mistake, in a successor set of rules. I have found them of no assistance in reaching my conclusions.

The intention of the Trust Company

So far as the evidence goes, it was agreed that the evidence of Mr Gamble and Mr James Lamb could be taken as representative of what all directors of the Trust Company in a similar situation would say. As it was explained to me by Mr Gaisman:

"We [that is to say Mr Gaisman and Mr Simmonds] have agreed that [ie not repeating questions to future witness already put to Mr Gamble and Mr Lamb] not on the basis that the questions are not the appropriate questions to have asked, because on the basis that it's a waste of time because it's almost inevitable that he will get the answers that he wants, I will get the answers that I want, and it is quite clear that the real issue that divides the parties in relation to that is a legal issue, and therefore we have made a pact that it is to be deemed to be the case that all subsequent witnesses to the subsequent deeds will give the same evidence."

As to that evidence, Mr Gamble was not in fact involved in the 1990 Trust Deed and Rules. He became involved only later when he was appointed as an independent trustee director on 1 February 1994. But what he said in relation to his role in later versions of the Trust Deed and Rules can be taken as the evidence of other directors, mutatis mutandis, in relation to the 1990 version. His evidence was that until the issue had been raised by a member in 2009, he was not aware of the detailed early retirement provisions of the C Plan. He said that subject to any material changes which were pointed out to him, he would have expected the new version of the Trust Deed and Rules to replicate what was there before. Nothing was in fact pointed out about early retirement. Reconstructing, but not recollecting, matters, he did not intend to alter the substantive rights of members or to subtract from "the true nature and content" of members' rights. On the counterfactual hypothesis that he had been aware that there was a significant mismatch between the version of the Trust Deed and Rules under consideration and the employee communications, he would have wanted to reflect the latter.

Mr James Lamb gave similar answers. He said this in his witness statement:

"Although the 1995 deed and the 1997 deed were in the form of new Definitive Trust Deeds, in reality they were only meant to make specific changes to the IBM pension plan and otherwise simply repeated the existing provisions without change…….

…..

I confirm that at no time were the early retirement provisions governing the C Plan tabled for discussion at the time of the 1995 or 1997 Deeds. My fellow Trustees directors and I did not revisit these provisions or discuss making any changes to them. They were part of the old provisions which we intended to carry forward into the new deed."

He qualified that in cross-examination by saying that the word "intended" was too strong since the matter was not actually considered:

"A. ……..if 'intended' means that I had thought about this particular paragraph, given it consideration and therefore had a firm intent, a firm purpose to leave it unchanged, that would be somewhat of a stronger statement than I would wish to make.

Q……..The intention – and I would suggest that you used the word correctly – the intention that you would have had was, insofar as the deed generally is -- insofar as changes are not brought to our attention, the deed generally should replicate the previous one.

A. Okay, so you are saying because we didn't give it any attention, it didn't change? I agree with that."

In re-examination, we find these questions and answers:

"Q. Was it or was it not your intention in agreeing to execute the 1995 and 1997 deeds to alter the substantive accrued rights of -- no, the substantive rights of members under the C Plan?

A. It was not our intention to change them.

Q. Was it your intention to subtract from the true nature and content of those rights whatever the true nature and contents of those rights were?

A. We had no intention to detract from them…."

The Trust Company also relies on other evidence. I mention it for completeness although in the end I do not think it goes to the heart of the issue which is what the decision makers intended – their intentions being objectively ascertained – when the 1990 Trust Deed and Rules were executed. But it does explain the background leading up to the 1990 Trust Deed and Rules. Accordingly, some of what follows may seem to be somewhat tangential to the main argument.

It is, I think, common ground (and I proceed on that basis) that in around January 1988, the Trust Company approached Nabarro with a view to instructing them to draft a new Definitive Deed for the Main Plan. The main point of contact between the Trust Company and Nabarro was Mr Ness, a legal assistant in Nabarro's pensions department. Mr Ness explains in his witness statement that Nabarro were instructed "to consolidate all of their existing Deeds into a more modern, 'user friendly' document". The partner at Nabarro who supervised Mr Ness, was Mr John Quarrell.

The detailed chronology of events relating to the drafting of the 1990 Deed is narrated in Mr Ness' witness statement. I see no reason to doubt what he relates. His counterpart at IBM was Mr Furnell, a pensions specialist working in the Trust Company. He was working at the administrative level and not at (or anywhere near) the board level. Mr Furnell's witness statement, the contents of which I accept, also explains the background to the drafting of the 1990 Trust Deed and Rules.

Mr Ness and Mr Furnell both give evidence, which I accept, to the effect that Nabarro's instructions were to consolidate the existing Rules of the Main Plan into "a new, modern, single document, using plain English where possible". Mr Ness also said, and again I accept this, that he knew nothing about the communications which had been made to IBM staff nor could he comment on the intentions in relation to flexible retirement in 1982-83 but simply followed the terms of the 1983 Deed.

In her witness statement Mrs James explained that she became involved in the drafting process. She picked up the omission from an earlier draft of the 1990 Trust Deed and Rules of a consent requirement. Mr Furnell explained in his oral evidence that the omission of the consent requirement from the draft was just an accidental omission as part of the process of converting "a particularly tortuous" Deed into a readable one. It was just a drafting slip of no relevance to the present proceedings. Mr Ness explained in his witness statement that he corrected the draft in the light of Mrs James's comments after further correspondence on the subject. Mr Furnell stated in his witness statement that he did not keep the board members of the Trust Company informed about the drafting process and did not copy them in on the correspondence with Mrs James. In relation to the issue of the consent requirement, Mr Furnell confirmed in his oral evidence that "it wasn't an issue that I needed to raise because there were so many changes going through….".

I accept all of that and conclude, as is submitted in the Trust Company's closing that, whilst Mrs James specifically pointed out that (in her opinion) there ought to be a consent requirement, this was not reported up the management chain to the board of the Trust Company. The details of the drafting were dealt with at the administrative level; the decision-makers at board level were not involved. It follows that Mrs James's comments are irrelevant to the ascertainment of the intention behind the 1990 Trust Deed and Rules.

A draft of the 1990 Trust Deed and Rules (either in their final form or materially in that form) was presented to a board meeting of the Trust Company on 20 September 1990. Of the board members of the Trust Company who attended this meeting, most had not been in office in 1982/83 when it was agreed by the board to implement the C Plan. However, Sir Leonard was in office at all material times. And Mr (now Sir Anthony) Cleaver ("Sir Anthony") was a member of the Management Committee at all material times.

The minutes of the meeting show that Mr Quarrell made a presentation about the draft 1990 Trust Deed and Rules; a written report had clearly been prepared by Nabarro. It was recorded that a copy of the new Definitive Trust Deed and Rules would be distributed to "the members of the committee for their detailed consideration". The reference to the committee is clearly to the Trustee Management Committee, that is to say the board of the Trust Company. It is not recorded in the minutes that the draft was approved nor that an authorisation was conferred on anyone to execute the final Deed. The date of the next meeting was not scheduled until 21 March 1991. The minute of that later meeting indicates that the draft Deed and Mr Quarrell's report had indeed been distributed.

Mr Quarrell's report ran to nine pages. It explained that the 1983 Trust Deed and Rules had become out of date due to changes in statute and changes to benefit structures and Plan definitions. It identified a number of the statutory changes (including revaluation of deferred benefits and a number of taxation changes). It then summarised the content of the various schedules to the Deed. At paragraph 13, the report described the retirement rules (contained in Schedule G to the 1990 Trust Deed and Rules), explaining that Schedule G "also details how the Plans calculate benefits both on early retirement (should this be permitted by the Principal Employer) late retirement and ill-health early retirement". However, the report does not indicate that any change was being made to the "early retirement" provisions of the C Plan, again to my mind something which is hardly surprising since Mr Quarrell thought that the new version of the Trust Deed and Rules was replicating the 1983 Trust Deed and Rules save in respect of amendments his firm had been instructed to reflect in the drafting (which did not include any change to early retirement conditions). Nor, again unsurprisingly, does it say anything about the concept which has been described in these proceedings as "flexible retirement". The description of the deferred pension rule (Schedule I) is silent about payment of deferred pensions before NRD.

The 1990 Trust Deed and Rules were executed on behalf of the Trust Company bearing the date 1 October 1990, although there is no evidence about when it was actually executed. It reflected Nabarro's style and plain English approach and effected a complete rewrite of the 1983 Trust Deed and Rules. In so doing, it broadly reflected the substance of the provisions of the 1983 Trust Deed and Rules regarding retirement before NRD under the C Plan. In particular, it included a consent requirement for retirement before NRD from active status, with no reduction for retirements between ages 60 and 63: see rules 2(1), and 2(3) of Schedule G. It also included a Company consent requirement for retirement before NRD from deferred status, and provided that such payment of deferred pensions would be "subject to actuarial reduction": see rule 2(3)(b) Schedule I.

The 1990 Trust Deed and Rules were ratified at the board meeting of the Trust Company held on 21 March 1991. They were no doubt ratified because there does not appear to have been any formal approval at the meeting on 20 September 1990 for the execution of the 1990 Trust Deed and Rules; indeed, one might think that the actual decision would be made only once the board had read the draft Trust Deed and Rules and Mr Quarrell's report which were to be circulated. The minutes of the 21 March 1991 meeting confirm that a copy of the new Definitive Deed and of Mr Quarrell's report had been circulated to all members of the board of the Trust Company. Mr Morgans, who was present at the meeting on 21 March 1991, and who was at all material times a member of the board would therefore have received a copy. The minutes do not suggest that there was any discussion of the provisions of the 1990 Trust Deed and Rules concerning retirement from active status before NRD or the drawing of deferred benefits before NRD at either the September 1990 meeting or the March 1991 meeting. There is no evidence that they were discussed. Indeed, the evidence is that they were not discussed and I make a finding to that effect.

Sir Leonard's evidence, which I accept, was that it was only the end result of the drafting process which was presented to the board of the Trust Company for consideration at their meeting on September 1990. Although he himself was aware of the intention in 1982/83 to provide for flexible retirement, the consent requirement was not brought to the attention of the board by Mr Quarrell or anyone else. That again is hardly surprising to my mind since the consent requirement had (mistakenly) been included in the 1983 Trust Deed and Rules and the drafting of the 1990 Trust Deed and Rules in its final form (and the only form seen by the board) included the same consent requirement. This conclusion is consistent with the evidence of Mr Ness and Mr Furnell.

Sir Leonard was unsure but doubted that he read Mr Quarrell's report. If he had done, one might think that an alarm bell would have rung. But whether he read it or not, the important point (which I find to be the case as a matter of fact) is that he did not raise any issue about consent with anyone at any stage.

Sir Leonard said that his intention was not to withdraw the existing right to flexible retirement. He was one of the few people at decision-making level on the scene in 1990 who knew that the intention behind the 1983 Trust Deed and Rules was to implement a C Plan including flexible retirement. I can readily accept that he had no positive intention to include a consent requirement and that, had it been drawn to his attention that a consent requirement was present, he would have realised that the 1990 Trust Deed and Rules represented a departure from the intention behind the C Plan. But I do not accept from his evidence that he had a positive intention, when addressing the draft 1990 Trust Deed and Rules in September 1990, or when ratifying the executed Deed on 21 March 1991, that it should not include such a consent requirement. If he had had such a positive intention, one might think that he would have checked to see what Mr Quarrell had to say about this topic and would have at least cast an eye over the provisions of the draft deed. The truth, it seems to me, is that he did not address his mind to consent at all. However he, unlike most others, had at one stage had a positive intention that there should be flexible retirement and thus that there should not be a consent requirement. He had not at any stage changed or even reviewed that intention. His subjective intention, therefore, at the relevant times (from 20 September 1990 through to 21 March 1991) was that there should be no consent requirement. But he did not communicate that intention to the other members of the board of the Trust Company or the Management Committee.

He also said that it was not the intention, so far as he was aware of his fellow trustee directors, any more than it was his own intention, to withdraw the existing right of flexible retirement. I can readily accept that it was not the positive intention of his fellow directors to include a consent requirement: the evidence is clear that the issue was not discussed. And I can accept that he could say that, so far as he was aware, the other directors did not intend to withdraw flexible retirement in cases where the directors concerned knew that the C Plan was meant to provide for flexible retirement. But that is as far as it goes. A director who did not know that the C Plan had ever been intended to provide for flexible retirement could not form an intention one way or the other about its withdrawal. To have asked at the time "Do you intend to withdraw flexible retirement" would not have provoked the answer "Yes" or "No" but rather "What do you mean? Are you telling me that the C Plan does provide flexible retirement already?".

Sir Anthony thought that he did read Mr Quarrell's report. He was asked whether he would have queried the passage in the report referring to the need for company consent for early retirement. He said not, as he saw nothing in this that was inconsistent with his understanding:

"I don't think I would in the sense that my understanding is -- and I think always was -- that those who wished to retire early could retire potentially from the age of 53 or 50 with company consent and that would appear to cover that situation."

The Trust Company asserts that the witnesses confirm that, when approving the 1990 Trust Deed and Rules, they had no intention to change the provisions of the C Plan except in relation to the specific changes brought to their attention, and otherwise that they intended to carry forward the existing rights and entitlements under the C Plan. They all say that no changes regarding the C Plan's provisions for retirement before NRD or flexible retirement were drawn to their attention. In particular, it is plain that Mr Quarrell's report did not do so. I agree with all of that with two qualifications.

The first qualification is the substance of the provisions of the 1990 Trust Deed and Rules for early payment of deferred benefits was not absolutely identical to that of the 1983 Trust Deed and Rules. The requirement for the consent of the Trust Company and Holdings in the 1983 Trust Deed and Rules was changed to a requirement for the consent of the Trust Company alone. None of the board-level witnesses, Sir Leonard, Sir Anthony Cleaver and Mr Cruttenden, can recall this change being mentioned. It was a change which was not adverted to in Mr Quarrell's report. Mr Ness expressed the view in his witness statement that it was not material. No discussion of it is recorded in the minutes of the board meetings of the Trust Company on 20 September 1990 and 21 March 1991. It is clear from all of the evidence, and I find as a fact, that the question of consent, whether in relation to early retirement or in relation to early payment of deferred benefits, was not discussed by the board of the Trust Company at any time in the context of the approval and execution of the 1990 Trust Deed and Rules. The change in the consent requirement was not inserted as the result of any decision of the board of the Trust Company appreciating that there was a change. It cannot be inferred from this change that the board of the Trust Company had in fact given careful scrutiny, or indeed any scrutiny, to the consent provisions.

The second qualification, which goes to the root of the dispute, is that it is not clear, (and it is a matter which I must address,) what the provisions of the C Plan were in this context.

On behalf of the Trust Company it is said that important points which the witnesses made include these:

i) If the 1990 Deed took away rights that had been given to members in 1983, they would have expected this to be brought to their attention so that it could be considered and discussed.

ii) Taking away rights from members (even if only prospectively) would have been regarded as a serious step and would have had to be announced to members.

iii) It was never suggested by anyone at the time of the 1990 Deed that members' rights were being changed adversely.

iv) There was no suggestion that a consent requirement was being introduced; had there been it would have had to be discussed and carefully considered.

In those circumstances, it is submitted that it is unreal to think that the parties agreed to do anything other than to carry forward existing rights without changing them (except for the particular changes brought to their attention which they specifically intended to make). That is particularly so given that the 1990 Trust Deed and Rules were presented to the board of the Trust Company as merely an updating document produced as part of an administrative exercise to consolidate previous changes. And it is all the more so when the last thing that IBM wished to do during the 1990s was to restrict members' rights to retire before NRD; as has been seen, IBM wanted employees to retire as soon as possible.

The four factors identified are, it seems, not really supportive of the Trust Company's case. Each of those points could equally validly be made even if the board of the Trustee Company had had a full appreciation of the wording and effect of the provisions of the 1990 Trust Deed and Rules concerning early retirement. Let me explain why I consider that to be so.

Suppose, then, that the provisions of the 1990 Trust Deed and Rules concerning early retirement had been gone through at the meeting of the board of the Trustee Company on 20 September 1990 with a careful comparison being made with the terms of the 1983 Trust Deed and Rules. And suppose that those provisions had been expressly agreed with a full understanding of their purported effect.

In that scenario, Mr Simmonds' submissions would, to my mind, be compelling. The board of the Trust Company's clear intention would have been (a) to adopt the wording which it did adopt and (b) for that wording to have the effect which, as a matter of ordinary language, it does have. The case would be entirely different, as he submits, from a case such as AMP where the wording deliberately adopted had an unforeseen consequence entirely counter to the intentions of those making the amendment in that case. In the present case, in this hypothetical scenario, the provisions of the 1990 Trust Deed and Rules concerning early retirement would have taken the shape they did because of a mistake; but the mistake would not have been that the document failed to reflect the intentions of those who made it, rather it would have been a mistaken understanding of the facts which lead to the formation of the intention which they actually did have, namely that the 1983 Trust Deed and Rules were themselves free of any relevant error.

Each of the four factors identified in paragraph 439 above would still obtain in that scenario.

i) As to the first factor, it would still be the case that the witnesses would have expected it to be brought to their attention that the 1990 Trust Deed and Rules took away rights that had been given to members in 1983. The fact that they were not informed was because nobody involved knew of the error in the 1983 Trust Deed and Rules. The witnesses would have acted under a mistake, namely a lack of appreciation of the error. But the intention which they formed would have been clear, namely to include a consent requirement.

ii) Similar considerations apply in relation to the second factor. The fact that there was no announcement is explained, again, by the absence of any appreciation that the 1983 Trust Deed and Rules contained an error. The absence of an announcement was not due to a mistaken translation of the intentions of the board of the Trust Company into the 1990 Trust Deed and Rules. Rather, it was due to a mistake in forming the intention in the first place.

iii) As to the third factor, the absence of a suggestion at the time of the 1990 Trust Deed and Rules that members' rights were being changed adversely would, again, have been due to the absence of any appreciation that there was an error in the 1983 Trust Deed and Rules, and not because of the same mistaken translation of the intentions of the board.

iv) Similar considerations apply to the absence of any suggestion that a consent requirement was being introduced.

The scenario which I have just been addressing does not, however, represent the facts. It is clear, and I find as a fact, that there was no discussion of the early retirement provisions or of the deferred pension provisions of the 1983 Trust Deed and Rules at the meetings of the board of the Trust Company on 20 September 1990 and 21 March 1991 nor was there any other discussion concerning those aspects of the C Plan more generally. There is no evidence of the subjective understanding of any of those attending the board meeting about the contents of the 1983 Trust Deed and Rules in relation to early retirement or about the time for payment of deferred pensions. Indeed, if either Sir Leonard or Sir Anthony had appreciated that the 1983 Trust Deed and Rules contained a consent requirement in relation to retirement between ages 60 and 63, one might expect that they would have drawn this to the attention of the board since it would not have reflected what they, at least, appreciated was the intention behind the C Plan. I am unwilling to infer that any of them did appreciate, whether by virtue of their position as board members or otherwise, what the provisions of the 1983 Trust Deed and Rules actually were in relation to those matters. I do, however, find as a fact that none of them, other than Sir Leonard and Sir Anthony, knew that the intention behind the creation of the C Plan was to provide for flexible retirement and there is certainly nothing which would justify an inference, assumption or presumption that they did know.

It follows from these considerations that none of the individuals concerned had a positive subjective intention, one way or the other, specifically directed at the inclusion of a consent requirement in relation to pensions payable between ages 60 and 63. Nor, in my judgment, can it be said that an objective intention, one way or the other, is established. The fact that there is no evidence to demonstrate an appreciation of the terms of the 1983 Trust Deed and Rules and the absence of any discussion about early retirement shows, I consider, objectively that there was no such intention.

Ordinarily, of course, the fact that a person has executed a document, be it a contract or, as in the present case, a deed exercising a power of amendment, will bind him to it. Prima facie, he is to be taken as intending that which he has signed. If rectification is sought, the person seeking rectification must displace that prima facie starting position and show positively that the person executing the document intended something different.

How, then, can it be said that the 1990 Trust Deed and Rules should be rectified when the Trust Company cannot demonstrate that the relevant individuals had a positive intention that there should be a right to early retirement without consent between ages 60 and 63, the onus being on it to establish the intention necessary for rectification? The answer to that is that a different intention may be sufficient. Thus, if it were clear that the intention was that the 1990 Trust Deed and Rules should reflect the entitlement which members of the C Plan had as a matter of law, it would follow that the 1990 Trust Deed and Rules ought also to reflect those rights; if the 1983 Trust Deed and Rules were subject to a valid claim for rectification, then the 1990 Trust Deed and Rules ought to reflect that claim and themselves be rectified to give effect to the intention. In contrast, if it were clear that the intention was that the 1990 Trust Deed and Rules should do no more than reflect, in new language, the provisions of the 1983 Trust Deed and Rules continuing the substance of those provisions as they stood at the time of the 1990 Trust Deed and Rules, a claim to rectify the 1990 Trust Deed and Rules would fail. Remedies may be available to ameliorate the effect of the mistake in the 1983 Trust Deed and Rules, but rectification would not be one of them.

The question for me is really to which of those contrasting situations the present case is closer. The evidence is critical, of course; that is an obvious statement but it must not be lost sight of. So although counsel on all sides seem to agree that the matter is essentially one of law (see paragraph 416 above) as applied to clear facts, I consider that the issue is really one of interpretation of the evidence and thus a question of fact.

To repeat, this is not a case where the individuals concerned read and understood the provisions of the 1983 Trust Deed and Rules and decided that those provisions should continue. If those had been the facts, then Mr Simmonds' submissions would, as I have already said, be compelling.

Instead, without focusing on precisely what the provisions, if any, concerning early retirement or deferred pensions actually were or were thought to be, those involved authorised or confirmed the execution of the 1990 Trust Deed and Rules with the intention of continuing the pre-existing provisions of the C Plan. Since those individuals did not address and discuss the provisions of the 1983 Trust Deed and Rules and did not adopt the wording of the 1990 Trust Deed and Rules as accurately reflecting the previous wording, there is no reason think that their intentions, objectively ascertained, were to continue the effect of the 1983 Trust Deed and Rules as drafted rather than to continue the substantive provisions of the C Plan in accordance with the law, that is to say as rectified. Indeed, for what it is worth, the intentions of Sir Leonard and Sir Anthony were almost certainly to continue flexible retirement, albeit that their intentions, even if objectively established, cannot be taken as the intention of the decision making body, namely the Trust Company.

I interpret the evidence given by the various witnesses as establishing an intention that the 1990 Trust Deed and Rules should continue the substantive benefits which the C Plan provided immediately prior to the execution of the 1990 Trust Deed and Rules, that is to say, the C Plan as rectified; this was an intention that the substantive rights of members under the C Plan should continue save to the extent that changes were expressly brought to their attention and specifically approved. If, as has turned out to be the case, those substantive rights differ from the rights set out in the 1983 Trust Deed and Rules, it was continuation of the former which, in my judgment, it was the intention of those individuals to continue. Those intentions were, in my view, objective intentions when it comes to applying the relevant test for rectification. These were not simply intentions locked up in the minds of the individuals concerned. They were intentions to be derived from the objective facts in particular (a) the absence of any discussion of either the provisions of the 1983 Trust Deed and Rules concerning early retirement (b) the absence of any discussion of early retirement generally and (c) a shared understanding that the 1990 Trust Deed and Rules was, in effect, a redrafting in plain English of the C Plan which was not intended to alter the substantive rights which members should enjoy. Accordingly, the extent to which the 1983 Trust Deed and Rules were subject to a valid claim for rectification, that claim should, in my judgment, follow through and be available in relation to the 1990 Trust Deed and Rules as well.

Accordingly, my conclusion is that the objective intention of all the relevant individuals was that the members of the C Plan should have the same rights (save as specifically addressed and varied) under the 1990 Trust Deed and Rules as it had originally been intended they should have and thus the same rights as they would have under the 1983 Trust Deed and Rules as rectified. Since that was the objective intention of all the relevant individuals, it is not necessary to address the far from straightforward position where different individual members of the decision making bodies had different intentions and how, in such a case, the intentions of the bodies themselves are to be ascertained.

The Trust Company's case is that it should follow that the 1990 Trust Deed and Rules must therefore be rectified to provide for a right to retire between ages 60 and 63 without employer consent. There is, however, one difficulty facing that conclusion which I must address. It would follow that, if rectification were granted along those lines, the Preservation Requirements would have applied to the 1990 Trust Deed and Rules as rectified. The result would then be that the 1990 Trust Deed and Rules would have had to be amended from its rectified state so as to comply with those requirements following which it would not accord with one of the original intentions behind the 1983 Trust Deed and Rules, namely to provide for the payment of deferred pensions only at age 63.

In the case of the 1983 Trust Deed and Rules, I have concluded, for the reasons given, that the Preservation Requirements do not preclude rectification. Essentially, the reason for that conclusion is that the 1983 Trust Deed and Rules can be rectified so as to reflect, as a matter of language, what the Trust Company and Holdings intended, namely both flexible retirement and payment of deferred pensions only at age 63. The Preservation Requirements would then superimpose certain unforeseen obligations on the Trust Company and Holdings to alter the provisions of the 1983 Trust Deed and Rules to bring them in line with the Preservation Requirements.

However, that escape from the impact of the Preservation Requirements as a defence to rectification in relation to the 1983 Trust Deed and Rules is not, arguably, available in relation to the 1990 Trust Deed and Rules. The argument is this: in relation to the 1983 Trust Deed and Rules, the evidence was that the impact of the Preservation Requirements was overlooked. The decision makers therefore had two intentions which, unappreciated by them, were not legally consistent, namely flexible retirement and deferred pensions payable only at age 63. In relation to the 1990 Trust Deed and Rules there is no such evidence and I would not be prepared to infer that the effect of the Preservation Requirements was not understood by the decision makers in relation to the 1990 Trust Deed and Rules. Accordingly, it seems impossible to me to say that the decision makers in relation to the 1990 Trust Deed and Rules, unlike those making the original decision in December 1983, could have had the positive intentions both to provide flexible retirement and to provide for deferred pensions to be payable only at age 63. Had they focused on these two provisions of the C Plan, they would have realised that they were incompatible and that their two intentions were inconsistent. The argument would then be that there would be no reason to see one intention as paramount with the result that it would also be impossible to say that an intention to include flexible retirement was established with the certainty necessary to found a rectification claim.

The answer to this difficulty, in my judgment, is to be found, again, in the identification of the precise intention of the decision makers in relation to the 1990 Trust Deed and Rules. As I have said, my interpretation of the evidence is that the decision makers intended to continue the substantive rights of the members under the C Plan as existing at the time of the 1990 Trust Deed and Rules without an appreciation of precisely what those rights would have appeared to be from a reading of the 1983 Trust Deed and Rules. The point, to be made yet again, is that they did not focus on the actual provisions at all but decided to continue what was already in place, whatever that may have been. They did not, therefore, have a positive intention to continue flexible retirement or a positive intention to continue a right to payment of a deferred intention only age 63; they had those intentions only as a result of the more abstract, higher level, intention to continue what had gone before.

Since the Trust Company succeeds in its claim to rectify the early retirement provisions of the 1983 Trust Deed and Rules but not the deferred pension provisions, the substantive pre-existing rights were (i) flexible retirement and (ii) a deferred pension payable only at age 63 albeit that the Preservation Requirements required the C Plan to be amended without having overriding effect. There is, in my view, no more inconsistency in the intentions of these later decision makers in seeking to continue what had gone before than there was in the intentions of the original decision makers in the first place.

My conclusion, therefore, is that the 1990 Trust Deed and Rules should be rectified so as to provide for retirement between ages 60 and 63 without employer consent and without any reduction but should not be rectified to include a right to receive payment of an unreduced pension from any age earlier than 63.

Concession

I have so far dealt with the question of rectification without mentioning a concession made in IBM's closing in the light of the evidence as it came out at trial. That concession has been further discussed in correspondence. There is, unfortunately, a dispute about precisely what was conceded in IBM's closing, in Mr Simmonds' oral submissions in closing and in the first part of the later correspondence. The Trust Company says that it understood the concession to be this:

IBM accepts that members of the C Plan who transferred from the N Plan, whether before or after execution of the 1983 Trust Deed and Rules who received the 1982/83 MILs and transferred in reliance on them are entitled to retire without consent from age 60 and to receive an unreduced pension. It is accepted by IBM that reliance is to be assumed. The concession does not extend to other members who may have joined the C Plan having seen the MILs and on the basis of them but who did not transfer from the N Plan.

In the later part of the correspondence, IBM wrote say that the concession never went that far. The concession that no evidence would be required applies only to those members who are offered voluntary redundancy or who are made compulsorily redundant in the ongoing restructuring programme currently being carried out. In relation to other members, IBM will require evidence that individual members relied on the 1982/83 MILs when transferring from the N Plan to the C Plan. The Trust Company does not know what evidence IBM says it requires in order to be satisfied of such reliance.

This dispute is not, in the light of my decision on rectification, of great importance. Even taking the extended concession which the Trust Company says was made, it is no concession at all but represents, in my view, the true legal position.

The Counterclaim

IBM makes a counterclaim by which it is asserted that members of the C Plan were employed under contracts of employment which define the pension benefits to which they are entitled. In particular, it is alleged that those contracts included terms (i) that a member would be entitled to retire before age 63 only with the consent of the employer and (ii) that deferred pensions would come into payment as of right at age 63. It is part of the claim in relation to (i) that in cases where consent was given, the pension would be unreduced where retirement took place between ages 60 and 63.

Consistently with the concession which I have just described, IBM does not pursue the Counterclaim in relation to the first claim (ie under item (i) against members of the C Plan who transferred into the C Plan from the N Plan in relation to early retirement). It does, however, continue to assert the second claim (ie under item (ii) in relation to deferred pensions). And it continues to assert the entire counterclaim against New Joiners.

As a matter of principle, it is possible for a contract of employment to restrict in some circumstances the benefits which a member of a pension scheme would otherwise be entitled to receive under the rules of the scheme: see for instance South West Trains v Wightman [1998] PLR 183 ("SWT") and Trustees of the NUS Officials and Employees Superannuation Fund v Pensions Ombudsman [2002] PLR 93 ("the NUS case"). But the decision of Arnold J in Re IMG Pension Plan; HR Trustees Ltd v German and International Management Group (UK) Ltd [2009] EWHC 2785, [2010] PLR 23 ("IMG") shows that this principle is not unrestricted. I considered these cases in my own decision in Bradbury v British Broadcasting Corporation [2012] EWHC 1369 (Ch).

It is important to note that IMG, like SWT, was concerned with changes to the benefits of existing members of the scheme. Arnold J referred to a number of factors which distinguished SWT and which led to the conclusion that they had not precluded themselves by contract from asserting their rights under the unamended scheme. These factors were as follows:

i) The members were unaware of the terms of the amending power, clause 7(i), which contained certain protections.

ii) They received no advice in relation to it.

iii) It was not clearly explained to them what was happening to their final salary benefits.

In particular

iv) They were not told how the transfer value and additional special contributions would be calculated and the assumptions which would be employed.

v) They were not given any real choice about whether or not to consent.

vi) They received the impression that they would not be adversely affected by the change.

None of those factors is of relevance in relation to New Joiners. There was nothing in the provisions of the C Plan which would have prevented the Trust Company and IBM amending its provisions so far as concerned New Joiners joining the C Plan after the date of an amendment or, indeed, in closing the C Plan to new entrants and providing new staff with benefits under a new section of the Main Plan. I conclude that, subject to a point concerning section 91 Pensions Act 1995 which I come to in the next paragraph, it was possible, in principle, for IBM to contract with New Joiners on the basis that they would be entitled to retire between ages 60 and 63 only with employer consent. Whether, on the facts, New Joiners were subject to such a contractual provision is a matter of fact which I will come to in due course.

It has been suggested that section 91 Pensions Act 1995 might have an impact on the validity of any such contract. As amended by the Welfare Reform and Pensions Act 1999, it provides, materially, as follows:

"Inalienability of occupational pension

(1) Subject to subsection (5), [where a person is entitled to a pension under an occupational pension scheme or has a right to a future pension under such a scheme] –

(a) the entitlement or right cannot be assigned, commuted or surrendered,

(b) the entitlement or right cannot be charged or a lien exercised in respect of it, and

(c) no set-off can be exercised in respect of it,

and an agreement to effect any of those things is unenforceable."

The wording which I have placed in square brackets in subsection (1) replaces the original wording which was "where a person is entitled, or has an accrued right to a pension under an occupational pension scheme". For that purpose, the meaning of "accrued rights" was to be found in section 124(2) Pensions Act 1995. Those rights at any time were "the rights which have accrued to or in respect of him at that time to future benefits under the scheme"; and where the member was still in pensionable service, his accrued rights were to be determined "as if he had opted, immediately before that time, to terminate that service".

I considered section 91 at some length in Bradbury v BBC [2012] EWHC 1369 (Ch) concluding that section 91 (even in its form after amendment by the Welfare Reform and Pensions Act 1999) would not be engaged by an agreement between a member and the BBC, the employer, restricting the amount of any future pay increase which would be pensionable even though the whole would otherwise be pensionable. As I put it at [84] and [85] of my judgment, the member might have realised that by refusing the BBC's offer, he would end up with no salary increase and therefore no increase in Pensionable Salary either. His right to a future pension based on the full amount of an anticipated pay rise was no right at all; and by agreeing to a pay increase only part of which would be treated as pensionable, he did not alienate anything to which he was even prospectively entitled. As the BBC's counsel, Mr Ham QC, put it, members had no right to a salary increase so that accepting a salary increase on terms that only part is pensionable would not involve a surrender of anything; the member would become entitled to a greater future pension, albeit one smaller than if the whole increase were pensionable.

The position of a New Joiner is, in my view, the same. Unless and until an individual becomes employed by IBM he would have no right of any sort at all in relation to the C Plan; he would not even be eligible to join it. If, as part of the contract under which he becomes employed for the first time, he agrees that his pension benefits should be restricted in some way, that is not a surrender of any right. Nor is it an agreement to surrender any rights in the future. Rather, the agreement defines what the future pension rights will be. Such an agreement is clearly not an assignment or an agreement to assign or the commutation of any right nor is it an agreement to assign or commute in the future. Accordingly, section 91 is not, in my view, engaged.

Even if that is wrong, it is clear that as originally enacted, section 91 would not have applied. It applied only where there was a surrender etc (or an agreement to effect a surrender etc) of an entitlement to a pension (ie where the pension was already in payment) or of an accrued right to a pension (ie a pension already earned at the date of the purported surrender or agreement). A New Joiner would have no such entitlement or accrued right. The relevant amendment to section 91 came into force on 1 December 2000. If the contract of employment of a New Joiner employed before that date provided for early retirement only with the consent of the employer, section 91 (in its original form) would not be engaged.

In my judgment, therefore, section 91, even in its form following amendment by the Welfare Reform and Pensions Act 1999, does not preclude the inclusion of a term in the original contract of employment of a New Joiner under which the consent of the employer is required for retirement between ages 60 and 63.

Returning to the facts, there are two periods which need to be considered separately so far as this contractual claim is concerned. The first is the period during which the description of the pension provision made available to employees was contained in the Employee Handbook; the second is the period during which that description had been removed from the Employee Handbook and was found in the IBM Pension and Life Assurance Plans Handbook; in the second period, the Employee Handbook contained no detail about the various Plans within the Main Plan but told the reader that the corresponding information was to be found in the Pensions Handbook. IBM's closing acknowledges that the pension provisions were withdrawn from versions of the Employee Handbook from and after June 1991; however, the first version of the Pensions Booklet appeared in 1987 and the first version of the Employee Handbook from which the pension provisions were withdrawn was issued in June 1991.

It is IBM's case that the contractual terms of employment of all employees of IBM are set out in each employee's contract of employment, which expressly incorporate in full the Employee Handbooks (and, subsequently, the Pensions Booklets). The Employee Handbooks were provided to all new hires on appointment. In the circumstances, it is IBM's case that it was a term of the contracts of employment of all new hires who joined IBM after November 1983 that the provisions of the C Plan applicable to such employees were as set out in the Employee Handbooks (and, subsequently, the Pension Handbooks). IBM does not submit that the level of benefit promised by the C Plan in accordance with the relevant pensions formula was a contractual term of employment; but what it does say is that it was a term of such contracts of employment that the employee in question could not retire before NRD without the consent of the Company, save in the case of incapacity.

The Trust Company points out that in the earlier versions of the Employee Handbook, there were separate sections: one section was headed "Terms and Conditions" and another was headed "Pension and Life Assurance Plans". It is suggested that a distinction is therefore drawn between contractual provisions, which were to be found in the former and were incorporated into an employees' contract of employment, and non-contractual provisions which were to be found in the latter which was merely descriptive. This distinction, it is said, is all the more clear in the second period during which the pension provisions were not to be found in the Employee Handbook at all. I do not consider that the distinction is as stark as the Trust Company suggests, not least because the pensions section of the early versions of the Employee Handbook contained, as I have already mentioned, a paragraph which stated that it was a condition of employment with IBM – clearly a contractual term – that the employee join the Pension and Life Assurance Plans as soon as eligible to do so.

Offers of employment were made to employees in letters. The trial bundle contains a number of such offer letters which can be taken as representative of the terms on which all employees were employed. None of them contains any detail about the terms and conditions of employment, but most of them state that the Terms and Conditions of employment are as set out in the Employee Handbook. Later offer letters noted that the section of the Employee Handbook which had previously dealt with pensions was to be found elsewhere. Thus, as Mr Simmonds notes, one typical letter contained this: "(please note that the section in the handbook referring to the Pension Plan is in the process of update and has therefore been superseded by the enclosed booklet 'Your Pensions Choices in IBM')". He submits that it would be very strange if a change made for the sake of administrative convenience effected a change in the contractual nexus, particularly where, as noted above, the later editions of the Employee Handbook and the subsequent offer letters cross-referred to (and, IBM submits, thereby incorporated by reference) the Pensions Booklets.

I have already dealt at some length with the provisions of the November 1983 Handbook. I have also mentioned sufficiently for present purposes most of the provisions of the Pensions Booklet "Your Pension Choices" noting that the reader of that booklet would take away the message that an early retirement pension at any age before 63 would require the consent of the employer. I have not yet dealt with the Employee Handbooks and Pension Booklets which post-date the 1990 Trust Deed and Rules. As to that, the documents now available are these:

i) an edition of Employee Handbook dating from June 1991. This contained nothing about pension provision save to say that the previous section 9 had been withdrawn with "the corresponding information" being contained in "the IBM Pension and Life Assurance Plans Handbook";

ii) an edition of the Pension Booklet dating from January 1993; and

iii) an edition of the Employee Handbook dating from January 1996 (which, like the 1991 Employee Handbook, does not contain a pensions section but states that "details of these Plans" are contained in the Pensions Booklet.

Each of the relevant publications – the various editions of the Employee Handbook and the Pensions Booklet– contain a provision in different wordings which make clear that it is not a definitive statement of the terms of the C Plan:

i) the November 1983 Handbook provided that "The Trust Deed and Rules override statements in this handbook and all other statements";

ii) the September 1985 Handbook provided that the Trust Deed and Rules "override" the Handbook;

iii) the October 1987 Pension Booklet provided that the Trust Deed and Rules "override" the Booklet;

iv) the May 1988 "Your Pension Choices" leaflet was expressed to be for "guidance only";

v) the August 1990 "Your Pension Choices in IBM" leaflet provided that the Trust Deed and Rules "overide" [sic] the leaflet; and

vi) the January 1993 Pension Booklet explained that the C Plan was "governed by a Trust Deed and Rules" and, in an amended form, provided that "in the event of any conflict the Trust Deed and Rules will prevail".

When Mr Simmonds submits that the provisions of the various Employee Handbooks, Pensions Booklet and leaflets were incorporated into the contracts of employment of members of the C Plan, I do not understand him to suggest that IBM itself came under an obligation to pay benefits in accordance with those provisions. Rather, the submission, as I understand it, is that IBM and the members concerned agreed, contractually, that the member should be entitled to retire between ages 60 and 63 only with consent. In other words, on the one hand, IBM agreed to provide a pension scheme (providing whatever level of benefit it actually did provide) and, on the other hand, the member agreed to work for IBM on the basis that he could retire between ages 60 and 63 only with consent. If the C Plan actually provides for a right to retire without consent (in the event because the C Plan falls to be rectified), the contract is to be given effect, in accordance with the law discussed at paragraphs 465ff above, so as to override the provisions of the C Plan.

In my judgment, IBM fails to establish any such contract. It is true that the offer letters provided that the employee was employed in accordance with the terms and conditions contained in the Employment Handbook (with reference, in later offer letters, to the Pensions Booklet). But reading the offer letter together with the Employee Handbook, the effect is in my view clear. What the Employee Handbook did was to describe the pension plans which were made available to employees at the time of the offer letter; it was a contractual term that the employee would join the C Plan as soon as eligible; and it is to be implied, no doubt, that IBM for its part would allow the employee to join the C Plan. There was, however, no contractual obligation on IBM to ensure that the C Plan would continue unamended or, indeed, at all: although it was IBM's intention that the C Plan should continue, it reserved the right to alter or terminate it. Further, the Employee Handbooks and the Pension Booklets made it clear that the Trust Deed and Rules were the formal governing documents and, in case of inconsistency with the Handbook or Booklets, were to prevail.

In those circumstances, I find it impossible to see how any sort of contractual obligation could rest on IBM to ensure provision of the early retirement benefits described in the Employee Handbook and Pension Booklets only with consent or how any sort of contractual obligation could rest on members to restrict themselves to taking such benefits as set out in the Employee Handbook and Pension Booklets only with consent. That would be to treat those documents as overriding in circumstances where, according to their own terms, they were expressed to be subsidiary. In my view, it is quite impossible to read the offer letter and the Employee Handbook/Pension Booklet together as an offer to employ the potential employee on terms that he or she will be entitled only to the benefits set out in those documents.

Mr Simmonds submits that it is no answer to the counterclaim to say that the Employee Handbooks contained wording to the effect that the Trust Deed and Rules were overriding. He points out that, when the contract of employment was entered into with a new employee, the Trust Deed and Rules were entirely consistent with the provisions of the Employee Handbook relating to early retirement. He identifies the Trust Company as arguing that the overriding Trust Deed and Rules for this purpose are the Trust Deed and Rules as rectified but that argument makes a nonsense of the objective approach to the construction of contracts. The only Trust Deed and Rules available upon joining IBM were the un-rectified documents and it is on the basis of these un-rectified documents that the terms of the contract of employment must be construed. The present case is, therefore, very different from Steria v Hutchison [2006] EWCA Civ 1551 on which the Trust Company relies, where the trust deed was inconsistent with the representations on which the member sought to base an estoppel. In the present case, the Trust Deed and Rules serve to reinforce the contractual counterclaim rather than defeat it.

I reject that attempt to eliminate the importance of the various statements that the Trust Deed and Rules were overriding. The importance, as I see it, of those statements is that is shows that the provisions of the Employee Handbooks and Pensions Booklets concerning pension benefits, and in particular the requirement for consent to retirement between ages 60 and 63, cannot be seen as giving rise to contractual obligations for the reasons which I have given. Mr Simmonds' argument assumes that a contractual provision concerning the level and conditions for payment of a benefit must be found somewhere and that it is to be found in the Employee Handbook and Pension Booklets which must be objectively construed against the provisions of the unrectified Trust Deed and Rules. In my judgment, however, the only contract between the employee and IBM was that the employee should be entitled and bound to join the C Plan. The fact that IBM and the Trust Company publicised its terms inaccurately (because it was not appreciated that the Trust Deed and Rules were themselves incorrect) does not result in any contract concerning the level and conditions for payment of a particular benefit.

Test the matter this way. Suppose that the 1983 Trust Deed and Rules had accurately provided for flexible retirement; but suppose also that the Employee Handbook had made it clear, mistakenly, that consent to retirement between ages 60 and 63 was required. In those circumstances, it would be clear, to my mind, that a member would have a right to flexible retirement and that it could not sensibly be argued that the Employee Handbook, expressly giving primacy to the Trust Deed and Rules, cut down the right given by the trust instrument. In my judgment, it makes no difference to the proposition that the provisions of the Employee Handbook did not give rise to contractual rights that the right to flexible retirement is included in the formal documentation only as the result of an order for rectification.

Nor can it be said (if it is said), in my judgment, that there were contractual obligations – on IBM to provide and on the member to accept no more than – the benefits set out in the Trust Deed and Rules at the time of employment. The Trust Deed and Rules were subject to a power of amendment; there was clearly no contractual provision that that power would not ever be exercised so as to reduce the future accrual of benefits. Quite the reverse: it is clear that there was no contractual restraint on the power of amendment although it would, of course, have to be exercised in accordance with the provisos to which it was subject and in accordance with IBM's duties in accordance with Imperial Group Pension Trust Ltd v Imperial Tobacco Ltd [1991] 1 WLR 589 (Ch D). I shall refer to these duties as the "Imperial duties". The absence of a contractual obligation on IBM not to reduce benefits in the future suggests very strongly to me that there was no contractual obligation on a member to accept a lesser benefit than that which the C Plan provided.

My conclusion is that the conditions (if any) under which a member would be able to retire between ages 60 and 63 was not a matter of contract between an employee and IBM as employer but that those conditions were to be determined as a matter of the terms of the C Plan's rules. The relevant rules, for this purpose, are those found in the Trust Deed and Rules as rectified.

I would reach the same conclusion even if the relevant Employee Handbooks and Pension Booklets had not expressed the overriding nature of the Trust Deed and Rules. It would remain the case, in my view, that the Employee Handbooks and the Pension Booklets were simply descriptive of the C Plan and its provisions and that there was no contractual term about the benefits which the C Plan would provide or the conditions under which they would become payable. It may well be that the Employee Handbooks and the Pension Booklets would then have amounted to some sort of representation by IBM about the level of benefit. But such a representation is not something on which IBM could rely to cut down the benefits to which a member is entitled. Contrast the position if the intention in 1983 had been that consent to retirement between ages 60 and 63 was required so that the 1983 Trust Deed and Rules would have been accurate; suppose that there had been a representation in the Employee Handbooks and Pension Booklets clearly stating that an employee would have the right to retire without consent. In such a case, and in the absence of a provision expressly giving primacy to the formal documentation, the member might be entitled to rely on the representation to claim the larger benefit. This would not, however, be by way of a contractual right.

The Trust Company also argues that the contractual claim should fail on the following basis. It is said that, even if the offer letter and the Employee Handbooks and Pension Booklets are to be construed as giving rise to a contractual obligation such as that which is alleged, that contract would have been made, on behalf of IBM, by personnel who had no authority to determine the pension provision for employees, a matter to be determined at senior level by the board of the Trust Company and the Management Committee or board of Holdings. Accordingly, the contract would not have been validly made. It is not necessary for me to decide whether this is a good further reason for rejecting the contractual argument. Given the almost total absence of evidence about the authority of those acting on behalf of IBM in agreeing service contracts, I do not consider that I should express any view on this question and say no more about it.

My conclusion is that IBM's counterclaim fails.

Consequences of rectification for the purposes of the Preservation Requirements

Once the Current Trust Deed and Rules have been rectified pursuant to my decision, there will be a right for a member to retire at any time between age 60 and 63 without employer consent; but provision will be made for deferred pensions to be payable without reduction only at age 63. The Trust Company argues that:

i) where a deferred pension does come into payment before age 63, no reduction is to be applied;

ii) in cases where IBM's consent is required under the Current Trust Deed and Rules, that is to say where the member concerned is still in service, IBM is required (either contractually or in accordance with its Imperial duties or pursuant to a statutory duty imposed on it by the Preservation Requirements) to take such steps as are necessary to achieve conformity with those Requirements; and

iii) this is so notwithstanding the changes to the Preservation Requirements made by Pensions Act 2004.

In addressing that argument, I start with the Pensions Act 2004. IBM submits that the Preservation Requirements are satisfied in the present case because the benefits of all C Plan members are payable before age 65. This argument relies on the fact that section 71(3) of the Pension Schemes Act 1993 was amended, as I have already mentioned in my general consideration of the Preservation Requirements at paragraph 45 above, so that short service benefit originally payable at normal pension age (or age 60 if later) thereafter became payable at age 65 (or normal pension age if later).

IBM's case is that provided that the C Plan makes provision for an unreduced deferred pension to be payable at age 63 for a member, whatever the period of service of that member, the Preservation Requirements, as they now stand, will be satisfied. The C Plan is not required to make any provision at all for early payment of a deferred pension. If it does, nonetheless, make such provision, there is no infringement of the Preservation Requirements if the pension is made subject to actuarial reduction; and this is so even if the member is able to claim payment of a deferred pension without consent from age 60.

The Trust Company's position is, however, that it is the quantum of the deferred pension (ie the amount after actuarial reduction), rather than the date of payment, that falls foul of the Preservation Requirements. Ignoring deferred members still in service (where it is accepted that consent to draw a deferred pension before age 63 is required so that no preservation issue arises unless they leave service before seeking to draw their pension) and focusing on a deferred member who is not in service (where consent is not required but where actuarial reduction will be applied), it is said that the Preservation Requirements are infringed by the actuarial reduction for early payment.

I have discussed in some detail both the Preservation Requirements in the form which they took under the Pensions Schemes Act 1993 and the effect of the amendments made by the Pensions Act 2004. I have concluded that the statutory obligation for a scheme to satisfy the Preservation Requirements is met provided that the scheme rules comply with the Preservation Requirements as they stand after amendment by the Pensions Act 2004. It is not necessary, in order to fulfil the statutory obligation, that the scheme comply, in respect of service prior to 6 April 2005 (the date on which the changes to sections 71 and 72 Pensions Act 1993 came into force) with the Preservation Requirements as they stood prior to amendment.

It follows that I accept IBM's submission that the C Plan currently satisfies the Preservation Requirements and that the trustees and managers are not under a statutory obligation to ensure, so far as they are able, compliance (in relation to any part of a member's service) with the Preservation Requirements as they stood prior to 6 April 2005.

This, it seems to me, provides a complete answer to the Trust Company's claim based on the Preservation Requirements unless (i) it can demonstrate a breach of some duty by IBM arising out of a failure (by the Trust Company itself and by IBM) either to have brought the C Plan into line with the Preservation Requirements before 6 April 2005 or to do so today or (ii) it can show that under the rules of the C Plan as they stand, IBM must agree to the payment of a deferred pension from age 60 without actuarial reduction.

I therefore need to address further arguments made by the Trust Company. The first argument relates to actuarial reduction. For its part, IBM accepts that its consent is not required to the early payment of a deferred pension in respect of a deferred member who is no longer in service with IBM: the Current Trust Deed and Rules allow for such a member to draw his deferred pension at any time after age 60 (indeed the Rules allow such a member to draw his deferred pension as early as age 50). However, consent is required where a deferred member is still in service. In all cases where payment does commence before age 63, Mr Simmonds submits that the effect of the Current Trust Deed and Rules is that the pension is to be subject to actuarial reduction. In contrast, the Trust Company submits that, in cases where a deferred pension comes into payment between ages 60 and 63, there should be no actuarial reduction. The first argument is that the reduction depends on the exercise of a discretion by the Trust Company which can be properly exercised only by applying no reduction.

The relevant rule (rule 2(3)(b) Schedule F) provides as follows:

"Deferred Benefits shall be payable from…any date not earlier than age 50 but subject in all cases to the prior agreement of the Principal Employer in the case of a Deferred Retiree who is in Service at the date the Deferred Retiree requests payment of pension (subject to actuarial reduction)^"

This provision can usefully be compared with the provision relating to late retirement (see rule 2(3)(c)):

"Deferred Benefits shall be payable from…any date not later than age 75 that the Deferred Retiree shall elect, provided that his Deferred Benefits shall be subject to increase on a basis intended to be cost neutral to the Plan and certified as reasonable by the Actuary having regard to the period of postponement after Normal Retirement Date and Clause 4 of Part XI."

I agree with Mr Simmonds when he submits that neither of these rules, as a matter of construction, gives the Trust Company any discretion whether or not to apply a reduction or increase. They are both mandatory provisions. I would reach that conclusion in relation to early payment even reading rule 2(3)(b) in isolation. But when the two rules, which are clearly meant to reflect each other, are read together, the answer is I consider clear.

The Trust Company relies on Universities Superannuation Scheme v Simpson [2004] EWHC 935 (Ch) in support of the contrary conclusion in relation to early payment. The wording of the relevant rule in that case was entirely different, providing for early payment to be on "such terms as the trust company, acting on actuarial advice… shall decide". That gave a discretion to the trustee company which had to be exercised by reference to the Preservation Requirements.

Further, it is not possible, in my view, to read the words "subject to actuarial reduction" as meaning "without actuarial reduction" in order to ensure compliance with the Preservation Requirements. That would stretch the elasticity of interpretation beyond breaking point; and although it is right that the provision must be construed in the light of the existence of the Preservation Requirements, it is to be observed that those Requirements are not overriding and should not lead to the adoption of an unduly strained construction.

I therefore reject the Trust Company's submission on this point.

The next argument which the Trust Company deploys is that IBM is obliged to consent to an amendment to the Current Trust Deed and Rules or otherwise to co-operate (for instance by always giving consent to early payment of a deferred pension without actuarial reduction) to achieve compliance with Preservation Requirements (as they stood, I should add, prior to the Pensions Act 2004). That obligation is said to arise either as the result of a contract or pursuant to IBM's Imperial duties.

The contractual case rests on the provisions of Clause 13 of Part III and Clause 12 of Part II of the Current Trust Deed and Rules ("Clause 13" and "Clause 12" respectively) which are in the following terms:

"13. In order to maintain Revenue Approval and to ensure that the Plan complies with the preservation requirements and the contracting-out provisions of the 1993 Act:

the Trustee with the consent of the Principal Employer may by Deed make any changes necessary in this Trust Deed and/or the Rules; and

the Trustee may give such undertakings to such regulatory authorities as are appropriate regarding the provision of benefits under the Plan as may be required, and may adjust the benefits under the Plan to comply with any such undertaking, and the provisions of any undertaking given under this Clause shall be treated as part of the Trust Deed."

"12. Each Employer shall do everything in its power, and give all information in its possession, which the Trustee may reasonably require in connection with the administration and management of the Plan".

The obligations imposed and powers created by those provisions apply from time to time. Clause 13 enables changes to be made which are necessary to ensure compliance with, among other matters, the Preservation Requirements and enables the Trust Company to give undertakings to the appropriate regulatory authority as is appropriate to ensure such compliance. It is clear, in my judgment, that neither the power to change, the Trust Deed and Rules nor any undertaking can go beyond what is needed to ensure compliance. Thus, even if, which is disputed by IBM in any case, IBM could be compelled to give its consent to a proposed change, it could not be compelled to agree to a more generous alteration than was needed to ensure compliance; nor could the Trust Company give a more generous undertaking. In my judgment, what is "necessary" or "appropriate" to ensure compliance is to be assessed by reference to the Preservation Requirements as they stand at the time when the change is to be made or when the undertaking is to be given. Since the Pensions Act 2004 contained no relevant transitional provisions, the Preservation Requirements are to be ascertained by reference to the amendments effected by that Act. There is, therefore, no subsisting power to amend the C Plan or to give an undertaking in relation to it so as to provide a deferred member with a right to an unreduced deferred pension before age 63 even where his service includes service prior to the operative date of the amendments effected by the Pensions Act 2004.

But if that is wrong, the position is more complex. The Trust Company submits that the requirement for IBM's consent is not in fact a substantive limitation on the power at all, but merely a procedural requirement; this is because the purpose of the power to amend contained in Clause 13 was to ensure compliance with the Preservation Requirements, a power which must have been intended to have some substance going beyond that of the general power of amendment. Having agreed to confer power on the Trust Company to amend the C Plan to achieve conformity, IBM cannot then prevent an amendment whose purpose and effect is to achieve that very object by refusing to consent to the exercise of the power. In effect, by giving the Trustee an additional power for a specific and limited purpose, IBM has implicitly committed itself to giving the consent which is formally required, in order to enable the power to be used for its stated purpose.

IBM's position is that this attempt to categorise the requirement for consent as a procedural requirement is nothing more than an assertion that IBM's consent is not required in spite of what is expressly stated. But it does not, so it seems to me, follow from that consideration that consent can always be refused so as to prevent the C Plan from complying with the mandatory statutory provisions. Mr Simmonds submits, however, that if an obligation to consent to an amendment or other method of complying with the Preservation Requirements exists, it is derived from IBM's implied duty of good faith not from any implied restriction on the ability to refuse consent for the purposes of Clause 13.

Without expressing a final conclusion, I incline to the view that, before the amendments effected by the Pensions Act 2004, Clause 13(1)(a) would have precluded IBM from refusing to consent altogether to the making of any amendment whatsoever to bring the C Plan in line with the pre-existing Preservation Requirements. Consider a case where the Trust Company had proposed modification A but that IBM had then rejected that modification, putting forward an alternative (reasonable) modification B. In those circumstances, the Trust Company might well not be able to enforce modification A. But in the absence of any alternative proposal from IBM and in the absence of the articulation of any objection to modification A, it may well be correct that the Trust Company could implement modification A without formal consent from IBM. The provisions of section 71(1) Pension Schemes Act 1993 is mandatory: a scheme must comply with the Preservation Requirements. IBM has signed up to the provisions of the C Plan. Just as a contract will be construed so far as possible in such a matter as not to permit one party to it to take advantage of his own wrong (see generally Lewison: The Interpretation of Contracts (4^th ed) section 7.10) – so that contracting party cannot take steps to prevent the satisfaction of a condition in a conditional contract so too it is strongly arguable that IBM cannot undermine the clear intention behind Clause 13 that the C Plan is to comply with the Preservation Requirements. The power to give or withhold consent must, on this argument, be used only for the purpose for which it is given which, on one view, is to prevent the Trust Company from imposing one modification rather than another but not to prevent it from amending the C Plan at all so as to achieve compliance.

But against that view, it can be pointed out that when what is now Clause 13 was first introduced into the C Plan provisions in 1990 (as part of the Nabarro drafting), compliance with the Preservation Requirements could be achieved by the OPB without the need for consent from an employer. There is a great deal to be said for the view that IBM could perfectly properly have refused its consent to an amendment since, absent consent, the OPB could intervene. That argument, however, is not available unless it is argued that the provisions of Clause 13 must necessarily take their meaning from the corresponding provisions of the 1990 Trust Deed and Rules.

The true meaning and effect of Clause 13 is, to my mind, a difficult issue. In the light of my earlier conclusions, I do not need to express a final view on it and I do not do so.

Clause 12 also creates some difficulty for IBM. In my judgment, however, Mr Simmonds is right when he submits that a general provision such as Clause 12 cannot override a specific provision such as Clause 13. As he observes, if the Trust Company's argument were correct, IBM's power to veto an amendment under the general amendment power at clause 1 of Part II would be lost whenever the Trust Company "reasonably" disagreed with the Company.

If I am wrong in my conclusions concerning Clause 13 and Clause 12, as set out in paragraphs 508 and 514 above, it is, nonetheless, clear that the claims of the Trust Company under Clause 13 and Clause 12 depend critically on (i) the claim for rectification succeeding and (ii) the order for rectification once made having retrospective effect for the purpose of imposing an obligation on IBM. However, whatever the extent of the retrospection which an order for rectification entails, it is clear that it was not until well after the commencement of the relevant provisions of the Pensions Act 2004 – indeed it would not have been until after the rectification claim was first suggested in 2009 – that the Trust Company could first have required IBM to agree to take any action to bring the C Plan into conformity with the Preservation Requirements on the footing that the C Plan should be rectified to provide for flexible retirement. It is one thing to say that IBM can reasonably be required to bring the C Plan into conformity with the Preservation Requirements as they stand at the time of a request by the Trust Company to do so; it is quite another to say that IBM can reasonably be required to bring the C Plan into conformity with the Preservation Requirements as they stood many years before the request. It seems to me that this point is closely related to the duty, if any, which rests on IBM pursuant to its implied duty of good faith. I propose to postpone saying any more about it until I turn to that duty.

Trustees and Managers

The Trust Company argues that IBM falls within the meaning of the words "the trustees and managers of the scheme" in section 71 Pension Schemes Act 1993, reliance being placed on Century Life v Pension Ombudsman [1995] PLR 135. The Trust Company's point is , as I understand it , a general one, namely that the phrase subsumes any employer under a pension scheme where that employer's consent to an amendment is required and does not turn on any special power which Holdings has in relation to the C Plan. I agree with Mr Simmonds when he says that that decision is not to the point: although the judge looked at the use of the expression 'trustees and managers' in various parts of the Pension Schemes Act 1993, he was concerned with an insurer rather than an employer. The role of an insurer in relation to an insured scheme is entirely different from that of an employer; I do not consider that a valid parallel can be drawn.

In any case, a clear distinction is made in various provisions of the Pension Schemes Act 1993 (as originally enacted) between the "trustees and managers" of a scheme and the scheme employer: see for instance sections 116(b); 121(2); 135(1)(a) and (b); 137; 139(2)(b); and 146(4). A highly relevant reference is given by Mr Simmonds, namely section 139(2) (now repealed), by virtue of which the OPB was empowered, in order to achieve compliance with the Preservation Requirements,

"to...authorise the trustees or managers, or other persons named in the order (and, in particular, in the case of an occupational pension scheme, any employer of persons in service in an employment to which the scheme applies) to make that modification"

That provision, in the very context of the Preservation Requirements themselves, demonstrated a clear distinction between "the trustees and managers" and an employer of persons in relevant service.

In my judgment, IBM does not fall within the meaning of "trustees and managers" in section 71 Pension Schemes Act 1993.

Breach of IBM's Imperial duties

I have adopted the expression "Imperial duties" as a shorthand to describe the implied duties of trust and confidence first described in a pensions context by Browne-Wilkinson V-C in Imperial Tobacco (supra). As it happens, I am due to hear, commencing in early 2013, another action involving the Trust Company and IBM in which the nature and extent of IBM's Imperial duties is central. I think it is right, therefore, that I say no more than is absolutely necessary in the present litigation about this topic. I have not, in any case, had as full submissions (either on the facts or on the law about this issue) as I would like. I make no criticism of anyone about that. The issue is only one of many issues in the present case, and was not perceived as one of the most important issues.

Quite apart from that, it seems to me that it is premature for me to decide whether IBM would be in breach of its Imperial duties if it failed to take steps (whether by way of amendment of the C Plan or by some formal commitment always to allow the early drawing of a deferred pension insofar as attributable to service before 6 April 2005) to produce the result for which the Trust Company contends. In my view, IBM ought to be given the opportunity to digest this judgment and the consequences of my decisions concerning rectification of the various versions of the Trust Deed and Rules. It is difficult to see how IBM is currently in breach of any duty (although in saying that, I do not intend actually to decide it). I say that because, until receipt of this judgment, IBM will not have known that the Current Trust Deed and Rules are to be rectified and would have no reason to agree to an amendment to provide for payment of an unreduced deferred pension other than at age 63. I do not know if the Trust Company would submit that IBM was acting in breach of its Imperial duties prior to the receipt of this judgment; nor whether it would say that, once I have made an order for rectification, IBM is to be treated as retrospectively in breach of those duties. If such allegations are not in fact made by the Trust Company, the issue becomes whether IBM would be in breach of its Imperial duties if, following my order for rectification, it were then to decline to agree to amend the C Plan (or to take any other appropriate course) so as to provide for flexible retirement.

Further, if I am correct in my analysis of the Pensions Act 2004 and how it impacts upon the Preservation Requirements, the C Plan is now compliant with the Preservation Requirements. Even if IBM would have been in breach of its Imperial duties following a successful claim for rectification in, say 1998 (I take a random date before the Pensions Act 2004), it does not follow that it would now be in breach of those duties in failing to provide now what it should have provided then. It would only be in breach of its Imperial duties if it were, following an order for rectification, required to bring the C Plan into conformity with the Preservation Requirements – not as they stand today but, so far as service prior to 6 April 2005 is concerned, as they stood before that date (when the relevant provisions of the Pensions Act 2004 came into force). I do not know what the Trust Company's case on this is.

Although its position in argument has been to reject any allegation of breach of its Imperial duties even if it were to refuse to amend the C Plan after failing in its defence to the rectification claim, IBM may, having reflected on this judgment and my various findings of fact, take a different view. Or it may not. Similarly, the Trust Company's position has been to assert that IBM is in breach of its Imperial duties in failing to give consent to an amendment but has done so in the light of its primary arguments concerning the effect of the Pensions Act 2004, arguments which I have rejected. It, and Mr Metcalfe with his advisers, may, having likewise reflected on this judgment and my various findings of fact, take a different view from that which it has so far been advocated. Or they may not. If the parties remain at odds, I will decide the issue whether, following rectification, IBM is bound to give consent to amendments to reflect the Preservation Requirements as they stood before the Pensions Act 2004 in respect of service before 6 April 2005. I will do so, of course, in the context of the present litigation but after receiving such further submissions as the parties wish to make. If this is necessary, I hope that it can be done on paper, but if any of the parties wishes to have a further, short, hearing, I will see that it is held as a matter of some urgency.

I make precisely the same observations about the point, addressed in paragraph 515, concerning Clause 13 of Part III and Clause 12 of Part II of the Current Trust Deed and Rules (see at paragraph 507 above), namely whether IBM is bound to give its consent to an amendment as a result of Clause 13(1)(a) or pursuant to its obligation to co-operate with the Trust Company in the administration and management of the Main Plan under Clause 12, although given the conclusions expressed in paragraphs 508 and 514 above, the point does not arise for decision.

Conclusions

The Trust Company's claim to rectify the 1983 Trust Deed and Rules and later versions, including the Current Trust Deed and Rules, to include flexible retirement for active members succeeds. The claim to rectify the provisions of those Trust Deeds and Rules relating to deferred pensions fails.

IBM's counterclaim fails.

The Trust Company's other claims in relation to the payment of deferred pensions from age 60 without actuarial reduction fail.

I leave unresolved for the present the issues arising out of IBM's Imperial duties and its obligations, if any, arising out of Clause 13 and Clause 12 if I am wrong in my primary conclusions concerning those Clauses expressed in paragraphs 508 and 514 above.

Postscript

In the Trust Company's closing, I was invited to make 39 findings of fact. In his closing written submissions, Mr Simmonds addressed those 39 issues, proposing his own answers. I do not propose to say any more about any of them than I have said in this judgment apart from noting that Finding 18 is agreed namely "Employees were invited to make an irrevocable election to join the C Plan on 06.07.83 on the basis of the description of the C Plan provided to them in the employee communication programme". I have answered them to the extent which I consider necessary for the purposes of my decision and have, I hope, covered sufficient ground for a court of appeal to reach final conclusions even if it were to disagree with me on any of my conclusions of law or the application of that law to the facts as found by me.

I have been invited by the Trust Company following receipt of my draft judgment to make some finding of fact about the addressees of the employee communications in 1982-83 and about who issued the MILs (in each case as set out in the Trust Company's opening). I decline to do so at this stage. If an issue should arise, the question will turn on the interpretation of documents and not on any oral evidence which I have received. The Court of Appeal will be able to interpret them as easily as me. I doubt very much that any issue will ever arise.

I would like to thank all counsel for their very helpful submissions, both in writing and oral, no doubt hugely aided by the solicitors instructing them. The length of those submissions has perhaps been slightly intimidating and accounts, to a large extent, for the length of this judgment, about which they will no doubt have similar feelings and for which I make no apology. I would also like to thank those behind the scenes for preparation of first-class bundles. I add that I have found the transcripts, achieved with commendable accuracy and supplied promptly, invaluable.

BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII
URL: http://www.bailii.org/ew/cases/EWHC/Ch/2012/2766.html

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Chancery Division) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Chancery Division) Decisions >> IBM United Kingdom Pensions Trust Ltd v IBM United Kingdom Holdings Ltd & Ors [2012] EWHC 2766 (Ch) (12 October 2012) URL: http://www.bailii.org/ew/cases/EWHC/Ch/2012/2766.html Cite as: [2012] EWHC 2766 (Ch)