[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 Holdings Ltd & Anor v Dalgleish & Ors (Rev 2) [2015] EWHC 389 (Ch) (20 February 2015) URL: http://www.bailii.org/ew/cases/EWHC/Ch/2015/389.html Cite as: [2015] EWHC 389 (Ch)

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

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION

		Royal Courts of Justice Strand, London, WC2A 2LL
		20/02/2015

Mr Justice Warren :

Introduction

1. Following the handing-down of my judgment dated 4 April 2014 ("the Judgment") and the delivery of a short judgment at a directions hearing on 13 June 2014 ("the June Judgment"), I have heard further submissions over the course of 9 days dealing with the remedies for the breaches of duty on the part of IBM established by the Judgment. I shall use the same terms and abbreviations in this judgment as I used in the Judgment and the June Judgment.

The Judgment ? Clarification

2. There has been disagreement between IBM, on the one hand, and the Trustee and the Representative Beneficiaries, on the other hand, about the meaning of what I said in certain parts of the Judgment and about precisely what conclusions I have reached as a matter of binding decision. The authorities establish, I consider, that I am able, before judgment is entered (that is to say, before any order is sealed) to give further clarification of the Judgment and to correct any errors and eliminate inconsistencies. The court has an untrammelled inherent jurisdiction to amplify its reasons and clarify the meaning of its judgment at any time prior to the sealing of its order: see for instance Paulin v Paulin [2009] EWCA Civ 221; [2009] 3 All ER 88. Indeed, there is even jurisdiction to alter an actual decision (see In re L (Children) (Preliminary Finding: Power to Reverse) [2013] UKSC 8; [2013] 1 WLR 634) although it is not, as will be seen, necessary for me to exercise that jurisdiction in the present case. That case shows that the exercise of this power is not restricted only to exceptional circumstances.
3. The lack of clarity in the Judgment arises out of my consideration of the separate elements which went to make up Project Waltz. In carrying out that exercise, I considered the 2009 NPAs "in isolation". I used those words in different ways in different paragraphs of the Judgment. In doing so, I have caused some confusion. As well as causing confusion, I may have said, in using those words, certain things which read literally do not reflect my true views. The difficulties were raised at the hearing in June. I did not consider it appropriate to deal with these difficulties at that hearing, considering that the matter should be left for fuller argument at the remedies hearing, although I did give some indications of my then thinking. Having heard submissions about what the parties consider I did mean by what I said and what is, and is not, open for me to clarify or even review, I am now in position to deal with the confusion which I engendered. I propose to clarify matters by considering each of the paragraphs of the Judgment which might be seen as giving rise to confusion or even inconsistency and, in so doing, state, or restate, my conclusions. I will refer to paragraphs of the Judgment thus: Judgment [xxx].
4. In carrying out this task, I occasionally use the phrase "Project Waltz as a whole", a phrase which I also used in the Judgment. By that, I mean (and meant in the Judgment) to include not only the 5 elements identified at Judgment [19], but all the surrounding circumstances of Project Waltz, the purposes which it was designed to achieve and the method and strategies involved in attempting to achieve those purposes. In this context, the 2010 and 2011 NPAs form part of Project Waltz.
5. At Judgment [1511], I said that there can be no doubt that the Project Waltz changes conflicted with the Reasonable Expectations which I had identified. There were three factors which I mentioned expressly in support of that conclusion: active members of the DB Plans would (i) cease to accrue further DB benefits from 6 April 2011 (ii) obtain no salary increases in the future (notwithstanding that Mr Riley had announced increases for 2009) unless they entered into the 2009 NPAs albeit that, after the initial announcement, the "future" in this context was limited to 2009 to 2011 with a review thereafter and (iii) be subject to the detrimental change in ER policy. Enhanced M Plan members would also be subject to (iii).
6. Mr Newman suggests that item (ii) should be omitted from that list because there was no Reasonable Expectation that there would be any salary increase at all. But that is not the point: the point is that if a salary increase were granted, it would not be pensionable. Following Soto, members had a Reasonable Expectation that benefit accrual would continue at least until April 2011 so that, if salary increases were in fact awarded, they would be pensionable in part. To put that in different words, the members had a Reasonable Expectation that salary increases would be pensionable; the 2009 NPAs displaced (to use a neutral word) that Reasonable Expectation even if it is right to say that the member was not entitled to any salary increase and had no Reasonable Expectation that he would receive any. This is not to say that this displacement was a breach of duty. That is a question which I addressed later in the Judgment. I should add that I did not, in the Judgment, identify the pensionability of salary increases as giving rise to a separate Reasonable Expectation. It is, however, a necessary part of the Reasonable Expectation that benefit accrual would continue in relation to salary, because the increase is part of the salary.
7. At Judgment [1512], I concluded that it was right that the Project Waltz changes should be viewed as a whole in assessing whether there was any breach of duty. I then went on to say that the individual elements of those changes must be considered separately. I stated that this was for two related reasons:

i) The first is that one or more of the individual elements may of themselves give rise to a breach of duty. It is important to know whether that is so.

ii) The second is that it is important, when it comes to considering any remedy, to know what Holdings could properly have done to avoid a breach of duty if it has committed one, and in that context it is important to know what, if any, individual element would have given rise to a breach of duty taken by itself.

8. I did not, in that paragraph, use the words "in isolation" although the words "separately" and "by itself" have much the same flavour. In referring to each element separately, I was intending to identify each element of Project Waltz with a view to considering whether that element, viewed apart from Project Waltz as a whole, would have given rise to a breach of duty. I was therefore drawing a distinction between each element in the context of Project Waltz as a whole and (counterfactually) that same element apart from Project Waltz. I was not, however, at this stage of the Judgment, drawing a distinction between two different scenarios: the first ("Scenario A") is one where, ignoring the other elements of Project Waltz, each element is to be considered in the context of the Reasonable Expectations which were established; the second ("Scenario B") is one where not only are those other elements to be ignored but also each element is to be considered on the assumption that no Reasonable Expectations at all had been established. Under Scenario B, there might be a breach of duty quite apart from any Reasonable Expectation. This could have been so, for instance, in relation to the earlier 2009 NPAs (and the same goes for the 2011 NPAs) which were unlimited in duration: this is the caveat mentioned in Judgment [1521]. Of course, a breach of duty under Scenario B would necessarily be a breach of duty in the context of Project Waltz as a whole. When I said that each element must be considered separately, the enquiry could embrace both Scenario A and Scenario B.
9. At Judgment [1515] I dealt expressly with the 2009 NPAs in the context of Project Waltz as a whole and concluded (on the hypothesis there under consideration that there were no global or local strands) that Holdings' conduct in relation even to the Later 2009 NPAs was a breach of duty.
10. I then went on to refer in Judgment [1516] to the 2009 NPAs "in isolation". I did not distinguish between Scenario A and Scenario B in that paragraph. The words "in isolation" would subsume both Scenarios. The rest of that paragraph and also Judgment [1517] and [1518] are concerned with Mr Tennet's submission that there was a separate Reasonable Expectation engendered by Ocean and Soto that members would not be treated differently so far as concerned pay increases depending on their election in 2006. I stated that if there was an entitlement to pay increases such as Mr Tennet maintained, it did not arise as the result of a separate Reasonable Expectation but would be the result of the Reasonable Expectations actually established (in the context of and taking into account Project Waltz as a whole). I left that issue for a further hearing. Viewing the different elements "in isolation" did not feature in this part of my reasoning.
11. At Judgment [1519] (continuing to examine the position in the absence of the global and local strands) I stated that my conclusion (ie that there was no separate Reasonable Expectation engendered by Ocean and Soto about pay increases) was consistent with Mr Tennet's own submissions. He accepted that, had it not been for the Reasonable Expectations on which he relied (ie not only the ones which I have held to be established but also the separate one which I rejected), Project Waltz would not have given rise to a breach of duty. I did not either agree or disagree with his concession: I simply used it to demonstrate that there was no inconsistency between his case and my conclusion.
12. I could have stopped there, but I went on to say something more. I said that it followed from his concession that the 2009 NPAs would not have given rise to a breach of duty "whether [they] were viewed as part of the overall changes or viewed in isolation". In using the words "in isolation" in that passage, I did not need to distinguish between Scenario A and Scenario B since the hypothesis here was that there were no Reasonable Expectations at all engendered by Ocean and Soto. In effect, "in isolation" was a reference to Scenario B.
13. The last sentence of Judgment [1519] uses the words "in isolation" again. It is this which gives rise to some confusion. What I was attempting to say was that, given the counterfactual hypothesis that the Reasonable Expectations summarised in Judgment [1510] had not been established, and given that the further Reasonable Expectation concerning salary increases had not been established, there would be nothing on which to base the assertion that the 2009 NPAs gave rise to a breach of duty "in isolation". If, as I fear is the case, it is not clear what I meant by "in isolation" at that point of the Judgment, I should say that this should be taken as a reference to Scenario B. I did not intend to say anything about the position under Scenario A still less about the position taking the 2009 NPAs in the context of Project Waltz as a whole. In any event, whatever uncertainty there may be arising from the words I actually used, it is clear, reading Judgment [1518] and [1519] together that I was not saying anything in the latter to preclude the argument which I had identified at the end of the former. As to Judgment [1521], what I said here was a caveat to the proposition that, absent any Reasonable Expectations and so that the words "in isolation" refer to Scenario B, there would be no breach of duty.
14. As I said in the June Judgment, there appears to have been an editorial error on my part in that Judgment [1519] and [1520] are substantially repetitions. I should not have included both of them. Judgment [1520] should be treated as deleted.
15. I draw attention to my actual finding in Judgment [1528]: no reasonable employer would have adopted the Project Waltz proposals in the form which they took. I held that Holdings was in breach of its Imperial duty so far as concerns accrual of benefits and change in ER policy and of its contractual duty of trust and confidence so far as concerns the 2009 NPAs. It was, I consider, implicit in that, and if it was not, I now make it an express finding that the 2009 NPAs, viewed as a separate element of Project Waltz but in the context of Project Waltz as a whole, gave rise to a breach of the contractual duty. In other words, if one asks whether the NPAs gave rise to a breach of contract, the answer was "Yes" in the context of Project Waltz as a whole, and not in isolation under either Scenario A or Scenario B. I did not state in this paragraph that the 2009 NPAs viewed under Scenario A gave rise to a breach of the contractual duty although in Judgment [1515] I had already held that to be the position absent the global and local strands. I indicated in Judgment [1528], although this was merely obiter at that stage, that the NPAs would not have given rise to a breach of the contractual duty under Scenario B.
16. There are some clarifications I need to make arising out of the Answers to Issue 4 appearing in Judgment [1533] to [1535]. Judgment [1534] is my primary finding that Project Waltz as a whole gave rise to breaches of the Imperial duty and the contractual duty. Judgment [1535] then goes on to consider each element of Project Waltz viewed "in isolation" and I answered Issues 4.1 to 4.5 in that context. However, that was not what Issues 4.1 to 4.5, as formulated by the parties in the List of Issues were directed at, whether on Scenario A or Scenario B. Rather those Issues were directed at whether each element of Project Waltz constituted a breach of duty when viewed in the context of Project Waltz as a whole and in the light of established Reasonable Expectations. By diverting myself to a slightly different question, namely whether each element was a breach of duty when viewed "in isolation" (whether I was thinking of Scenario A or Scenario B is something I will come to), I failed to answer the actual Issues raised. However, to be fair to myself, I think that the answers to those Issues are implicit in Judgment [1534] in the light of the lengthy analysis which I had carried out earlier in the Judgment; in particular, Judgment [1528] discussed above shows that I considered the 2009 NPAs to give rise to a breach of the contractual duty.
17. The words "in isolation" appear in the introductory words to Judgment [1535]. But that is not to say that the sub-paragraphs are all to be taken as referring to the same Scenario. It is clear that paragraph i) is referring to Scenario A since my reasoning was that the exercise of the Exclusion Power was contrary to the established Reasonable Expectations.
18. Mr Newman submits that, in paragraph ii), I held that the Later 2009 NPAs did not involve Holdings in any breach of duty. I do not agree. What I said there was, to repeat, in the context of viewing each element of Project Waltz "in isolation" without identifying whether I was considering Scenario A or Scenario B. Clearly, I was not intending to say anything contrary to the analysis which I had carried out earlier, in particular Judgment [1528]. In saying that the Later 2009 NPAs did not breach Holdings' duties, I was addressing the position under Scenario B. I accept that this is not clear from what I actually wrote, but I certainly did not intend to say that, taking account of the established Reasonable Expectations, there was no breach (although I did not decide to the contrary either). To read what I said as deciding that there was no breach even taking account of the established Reasonable Expectations (ie Scenario A) would be inconsistent with what I said in Judgment [1517] and [1518] leaving open the argument (for a further hearing) that DB members had an entitlement to the same pay increases as DC members.
19. There is one other clarification I must make in relation to Judgment [1535 ii)]. At the end of Judgment [1521], I said that I would return to the caveat which I had raised. I did not expressly mention the caveat later in the Judgment. The first sentence of Judgment [1535 ii)] was intended to deal with the caveat. In other words, the Initial 2009 NPAs would under Scenario B (and a fortiori under Scenario A) have involved Holdings in a breach of its contractual duty of trust and confidence.
20. Paragraph iv), like paragraph i), is clearly looking at the position under Scenario A.
21. I said in the June Judgment that it was open to argument what I meant by the words "in isolation" in Judgment [1535]. I have now explained what I meant and what I think my words actually meant. But if there is doubt about it, I clarify the Judgment by making clear that I did not decide whether the Later 2009 NPAs would, under Scenario A, have given rise to a breach of the contractual duty. Mr Stallworthy, however, draws attention (see footnote 22 of his written opening submissions for this hearing) to the June Judgment [14] where he records me as expressing this provisional view:

"All I said [in Judgment [1519] was that (i) absent any reasonable expectations there would have been no breach of duty, and (ii) with the Reasonable Expectations and the 2009 NPAs in isolation, there would have been a breach of duty……"

22. I think that (ii) is an error on my part. On a more detailed consideration, I do not consider that Judgment [1519] can be read as expressing any conclusion (or indeed view) about the Later 2009 NPAs under Scenario A.

Duration of Reasonable Expectations

23. At Judgment [1053] I stated that Ocean did not give rise to any Reasonable Expectation concerning future benefit accrual enduring past April 2011. That was not of much relevance because Soto came along. In relation to that, I stated that the Reasonable Expectation would have endured until at least April 2011. I did not expressly articulate there, or indeed anywhere else in the Judgment, an end date beyond which the Reasonable Expectation would not endure. It was necessarily implicit, however, in my findings at Judgment [1511 i)] (in relation to benefit accrual) and [1535 i)] (in relation to the Exclusion Notices) that the Reasonable Expectation endured for some period beyond 6 April 2011, a period sufficient for there to have been a breach of duty (and thus not for a simply de minimis period). It was, and is, my view that that is so and I should have said so expressly. However, I did not need to decide the question of how much later the Reasonable Expectation could be said to endure and there is nothing express or implied in the Judgment to indicate my conclusion about that. I will refer to the date beyond which the Reasonable Expectation concerning future benefit accrual did not continue as "the Longstop Date". I will address later in this judgment the question of the Longstop Date: see at paragraph 426 below, although I mention that it is common ground that (on the basis of the Judgment and therefore subject to any appeal) this is no later than 31 March 2014.
24. The "implied obligation of good faith" is how Sir Nicolas Browne-Wilkinson V-C described the contractual duty of trust and confidence in Imperial Tobacco: see at p 597. This was the term referred to by Knox J in Hillsdown Holdings which I referred to in Judgment [353]. I described a different phrase "the implied duty [rather than obligation] of good faith and, perhaps begging a question, described that as a duty arising in a pensions context, using words reflecting the contractual duty.  In doing so, I am guilty of some inaccuracy in Judgment [353] and [354] since Knox J was clearly referring to the contractual duty whereas the reader of the Judgment might think he was referring to the implied obligation under a pension scheme. But as I said in Judgment [354], Mr Simmonds himself used the words "Imperial duty" and that was clearly in contradistinction to the contractual duty. When I used those words, I certainly intended them to be a reference to the implication into the trust deed and rules of the limitation described by the Vice-Chancellor. I was, I can now see, eliding to a greater extent than perhaps I should have the content of the contractual duty with the limitation on the employer's powers. I do not think anything turns on that elision for the purposes of the Judgment since the issue was always the extent to which the limitation was not observed, that is to say whether Holdings had conducted itself "in a manner calculated etc….". But the matter is now brought into sharper focus by the submission which Mr Simmonds makes to the effect that the Imperial duty is not a duty at all but simply a limitation on the exercise of powers, a submission which has an impact on the availability of any financial remedy for failure to observe the limitation. I will be dealing with that submission in due course. I shall, nonetheless, continue to use the phrase "the Imperial duty", intending it to refer not only to the limitation itself but also to the obligations, such as they are, which are imposed on the employer as a result of that limitation when it comes to the exercise of a power under the trust deed and rules.  Further, as explained in paragraph 431 below, the same limitation is to be implied in relation to other functions of an employer under the pension scheme so that, in using the phrase "the Imperial duty", I intend it to include, also, that limitation and the obligations arising from it.

The Issues and representation

25. A. The parties have agreed a List of Issues, running to over 30 separate items with several sub-issues, although there is overlap between some of them. The text of each Issue appears in each relevant section of this judgment. At the beginning of the List of Issues are a number of definitions which I will use in the Judgment. I set out the entire document prepared for the hearing at Annex A.
26. B. The legal teams have been under considerable pressure in preparing for the hearing. The legal team acting for the Representative Beneficiaries could not, for various reasons, realistically prepare submission on all relevant issues. Accordingly, the arguments on some of the issues have been prepared and presented by the Trustee's legal team.

Issues 1 to 12: the NPAs

27. Issues 1 to 12 all relate in one way or another to the NPAs. They divide into two categories:

i) issues which concern whether NPAs which were not dealt with at the trial (that is to say the 2010 and 2011 NPAs) were entered into in breach of IBM's duty to members: Issues 9 and 11; and

ii) issues which concern what remedies flow from the breaches which I have held in the Judgment to have been committed or which are established by my decision in this judgment.

28. To recap, the position in relation to the 2009 NPAs is that, viewed in the context of Project Waltz as a whole and in the light of the Reasonable Expectations which I have held to arise, both the Initial and the Later 2009 NPAs were procured by Holdings in breach of its contractual duty of trust and confidence: see Judgment [1528] and [1534]. Further, the Initial 2009 NPAs viewed in a counterfactual context where there was no Project Waltz and no Reasonable Expectations (ie Scenario B described in paragraph 8 above) involved a breach of the contractual duty by Holdings because of the indication or threat that there would be no salary increases at any time in the future for members who did not sign up to the 2009 NPAs: see Judgment [1535ii)]. That finding was quite independent of the established Reasonable Expectations. I made no decision about whether the Later 2009 NPAs would have given rise to a breach of duty under Scenario A (ie looked at in the context of the Reasonable Expectations established but ignoring the other elements of Project Waltz). I made no decisions, at least expressly, about the 2010 and 2011 NPAs.
29. However, no distinction had been sought to be drawn by the parties between the Later 2009 NPAs and the 2010 NPAs; and I expressly reserved the position in relation to the 2011 NPAs (see Judgment [1535] iii)). It is not surprising that the parties did not draw any such distinction. The whole concept of viewing various elements of Project Waltz in the way I did – that is to say as something independent of Project Waltz itself – was mine. Having used the concept in the way which I did, it is only right that both sides should be entitled to address such arguments as they wish in relation to the validity of the 2010 and 2011 NPAs under either Scenario A or Scenario B. I will return to this aspect later.
30. It is important to remember that both Scenarios A and B are counterfactual. In fact, each and every element with which I am concerned took place as part of Project Waltz and in the context of the established Reasonable Expectations. Issues 9 and 11 are directed at whether the 2010 and 2011 NPAs involve a breach of duty in the context of Project Waltz as a whole; they are not directed at whether they would be breaches of duty under either Scenarios A or B. However, the position under each of Scenarios A and B may be relevant to the remedies to which the RBs are entitled and so needs to be considered.
31. When it comes to remedies, Mr Stallworthy submits that the right approach is to view the Project Waltz changes as a whole and in the light of the established Reasonable Expectations consistently with the approach to the issues of breach of duty. I agree with that submission. It reflects with what I said at Judgment [1512] (which I have mentioned at paragraph 8 above where I explained why it was appropriate to consider each element of Project Waltz separately). This was not to cut down, in any way, the circumstances where a breach of the contractual duty of trust and confidence or of the Imperial duty might be found. Rather, there were two reasons to consider the elements of Project Waltz separately: first, because an element of Project Waltz might, of itself, give rise to a breach of duty (as I have held to be the case in relation to the Initial 2009 NPAs); and secondly because, in considering remedies, it is important to know what Holdings could have done to avoid a breach of duty.
32. In this context, Mr Stallworthy is quite correct when he says that an employer's duty of trust and confidence can be breached by its actions cumulatively (even where the "last straw" on its own is not unreasonable or blameworthy): see eg Waltham Forest LBC v Omilaju [2004] EWCA Civ 1493, [2005] ICR 481 at [20]. As Mr Stallworthy puts it:

"To 'salami-slice' a package of measures and determine hypothetically the worst that an employer could do in respect of one element of that package ('in isolation') without breaching the duty of trust & confidence does not mean that all such 'worst case' measures would not breach the duty of trust & confidence when put back together and suffered by employees in aggregate as a package. The remedies fall to be determined having regard to Project Waltz as a package just as much as breach fell to be determined having regard to Project Waltz as a package."

33. He submits that it is a wrong approach, one which is neither appropriate or of assistance, to consider, as I did in the Judgment, the different elements of Project Waltz in isolation (ie under either Scenario A or Scenario B). He is, I have no doubt, right to say that the remedies to which the RBs are entitled must be assessed by reference to Project Waltz overall and its cumulative effect. In introducing into the Judgment consideration of each element "in isolation" I certainly did not intend (and do not consider that I actually did so) to say that the remedy for breach of duty was to be assessed by an accumulation of separate remedies for each element which, by itself, would have given rise to a breach (whether under Scenario A or Scenario B). But that is not to say, in my view, that it is inappropriate and of no assistance to see what the effect of each element might be taken by itself particularly if such examination reveals that an element, taken by itself, constitutes a breach of duty quite apart from its context within Project Waltz.
34. For instance, it might be said that it is only the cumulative effect of the elements of Project Waltz which have given rise to any breach of duty. It could then be argued that by removing one objectionable element (eg by setting that element aside) the remaining elements would cease to be objectionable and no further remedy would be needed in relation to them. In contrast, if an element of Project Waltz is, under Scenario B, of itself a breach of duty, then no amount of elimination of other objectionable elements will render that element unobjectionable. This is the case, for example, in relation to the Initial 2009 NPAs which I have held to give rise to a breach of duty under Scenario B (ie even in the absence of any Reasonable Expectation and ignoring the other elements of Project Waltz).
35. Further, in assessing contractual damages, a claimant is to be put in the position in which he would have been but for the breach, and the defendant is generally entitled to have damages assessed on the footing that he would have taken the least onerous course in fulfilment of his obligations. Knowing whether a particular element of Project Waltz would have constituted a breach under either of Scenarios A and B may inform the correct approach to remedy.
36. It is considerations of that sort which led me to say at Judgment [1512] that it was necessary to consider the various elements of Project Waltz separately. It has resulted in a larger amount of time and effort being spent by all parties on these issues than I had anticipated or intended. Perhaps (in the case of the RBs) this was out of fear that I would not focus on the real question (ie what is the remedy for the breach to which Project Waltz as a whole gives rise in the context of the established Reasonable Expectations) but instead look for a remedy in relation to each element separately; and (in the case of Holdings) that it was with a view to improving its position by persuading me to adopt precisely that (in my view erroneous) latter approach.
37. So far as the facts are concerned, I do not need to repeat what I said at Judgment [1323] to [1328] in relation to the 2009 NPAs. I should add two matters. The first is that an employee who registered acceptance through the online tool received an email which read as follows:

"Subject: Confirmation of acceptance of non-pensionability of future salary increases in relation to DB Benefits

Dear colleague, This e-mail has been sent to you to confirm that you have submitted your acceptance that any further salary increases that you receive from IBM will be non-pensionable for Defined Benefit purposes.

Please keep this e-mail for you records."

38. The second matter is that salary increases were in fact awarded to the affected members. I do not know whether, when the salary increase was notified to employees, they were told, again, that the increase was not pensionable or that it was made on the basis of the NPA which the members had signed as confirmed by the email.
39. With that introduction to Issues 1 to 12, I turn to Issue 1 (and its sub-issues). This issue is closely related to Issue 3 and the discussion will, to some extent, overlap. But in accordance with the parties' approach, I will take them in order.

1. Are the Initial 2009 NPAs:

1.1 void and/or unenforceable;

1.2 voidable and liable to be set aside, and if so:

1.2(a) • can they only be avoided and set aside by the Court in relation to all or a group of members collectively,

• and if so will the Court now avoid them and set them aside;

1.2(b) • can they be avoided and set aside at each member's individual election,

• and if so can members now avoid them and set them aside, and what would they have to do or prove in order to do so; or

1.2(c) • are they liable to be avoided and set aside in some other manner, and if so what manner;

1.2(d) • with effect from what date (in particular from the date on which they were purportedly entered into or from the date of election to avoid and set aside);

1.3 valid?

40. The starting point in addressing these issues is that all of the 2009 NPAs took place in the context of Project Waltz. In that context they all gave rise to a breach of duty as held in the Judgment. In addition, I have, earlier in this judgment, clarified the Judgment in relation to the Initial 2009 NPAs to indicate that the Initial 2009 NPAs would "in isolation" under Scenario B (and a fortiori under Scenario A) have involved Holdings in a breach of its contractual duty of trust and confidence: the breach here was "presenting the members with the choice of signing non-pensionability agreements or receiving no pay increases at all in future". As I shall explain below, it is common ground, and I accept, that the NPAs did not themselves amount to binding contracts. A contract came into being only when an actual salary increase was offered and accepted. The implementation of the NPAs in the context of actual salary increases subject to the non-pensionability term was as much part of Project Waltz as the NPAs themselves and was all part and parcel of the conduct which gave rise to the breach of duty. In particular, it was a breach of duty for Holdings (for all of the 2009 NPAs in the context of Project Waltz and for the Initial 2009 NPAs even in isolation) to offer salary increases only on terms that they would be non-pensionable and, accordingly, the actual introduction of such a term into the varied contracts of employments was also part and parcel of that breach of duty.
41. The issue is therefore not just whether the NPAs themselves are void, voidable or unenforceable: another important point is whether the non-pensionability term applicable to the actual salary increases is void, voidable or unenforceable.
42. Mr Newman suggests that I did not, in the Judgment, make any finding that the variation to the employment contract, or the collateral contract as he would have it, gave rise to a breach of the implied duty of trust and confidence. I did not say so in so many words. This was because the whole focus of the debate before me was on the NPAs themselves. It goes without saying, in my view, that if the procuring of the NPAs was a breach of duty in the context of Project Waltz as a whole, so too was the procuring of the actual contracts giving effect to the protocol which the NPAs laid down. It also goes without saying, in my view, that if the NPAs gave rise to a breach of duty in isolation (whether under Scenario A or Scenario B) then the actual contracts would do so too. If it does not go without saying, then I say it now and hold that a contractual variation following on from an NPA was as much a breach of duty (or component of the overall breach of duty arising as a result of Project Waltz as a whole) as the NPA itself.
43. The Trustee's primary contention is that the 2009 NPAs are unenforceable. Its fallback position is that they are void or voidable and liable to be set aside. Quite what these words (void, voidable, unenforceable) mean in the context of the NPAs is something I will come to in a moment. Holdings' position is that the NPAs are all valid (and this is so notwithstanding that they give rise to breaches of duty in the context of Project Waltz as a whole and, in the case of the Initial 2009 NPAs notwithstanding that they would give rise to breaches of duty even in isolation). Mr Newman submits that, in any event, it would not make any difference to the outcome if they were void or voidable. As to salary increases, the Trustee's position is that salary increases actually awarded to affected members are pensionable whereas Holdings' position is that they are non-pensionable since that was the basis on which they were awarded.
44. It needs to be borne in mind throughout the discussion that the Trustee was not a party to the NPAs. So far as the Trustee is concerned, the benefits to which members are entitled are to be found in the provisions of the Plans of which the affected members were members (taking account of the validity of otherwise or the Exclusion Notices in accordance with my answers to Issues 13 to 18). It is only if, and to the extent that, the Trustee is obliged or is able to give effect to a contract made between Holdings and an affected member (ie the terms of any salary increase subject to the NPAs) that the actual benefits payable to a member can properly be other than the benefits for which the Plans provide.
45. It is Holdings' position that the NPAs of themselves were not contracts. This is said to be so because Holdings gave no consideration for them. The Trustee did not contend that that was not a possible analysis. The RBs presented an alternative analysis that the NPAs were contracts supported by consideration that members would not be rendered ineligible for salary increases. I consider Holdings' position to be correct. I am not sure that much turns on this since the important point is to establish the terms on which the salary increases were in fact offered. Even if there was a contract, as the RBs contend, and even if that contract is void or voidable as a contract, it does not follow that the terms of the (void/voidable) contract cannot, nonetheless, form the basis on which an actual salary increase is offered. Ordinarily, a salary increase once awarded (unless and until it is somehow set aside) gives rise, I consider, to a variation in the contract of employment: it is not a free-standing contract that, in return for continuing to work, the employee will receive payment of a sum of money ie the amount of the increase. In my view, it makes no difference that, in the present case, a salary increase is awarded to a member on terms that the increase is subject to the non-pensionability term. Further, the non-pensionability term itself is also to be given effect, if it is valid, as part of the contract in accordance with which the member receives the salary increase and therefore as a variation of the terms of the employment contract.
46. I have referred above to a variation in the terms of the employment contracts. I do not overlook Mr Newman's submission that this is not a correct categorisation of the offer and acceptance (on the terms of the NPAs) of a salary increase. He says that there is simply a collateral contract. The reason he draws the distinction between a variation and a collateral contract is because he perceives it as strengthening Holdings' position when it comes to remedies: if there is only a collateral contract, then it is easy to set the whole contract aside if the Trustee is correct in saying that the non-pensionability term cannot be relied on by Holdings whereas this is more difficult in the context of a variation to the employment contract itself. I do not, myself, think it makes any difference which analysis is correct. An important part of the Trustee's case is that the consideration for the salary increase is not only the non-pensionability term but the employee continuing to provide his services. Mr Newman's case, it appears to me, sidelines that point by focusing on an alleged collateral contract which does not itself include that element of the consideration. If that is wrong, and the analysis does make a difference, it is my judgment that the salary increase and the non-pensionability term take effect as variations to the contract of employment and not as merely a collateral agreement.
47. To explain that last proposition, the Trustee's case (with which I agree) is that, where a salary increase is accepted as part and parcel of a valid package including a term that it will be non-pensionable, it is at that stage that there is a binding contract. I do not understand Mr Newman to disagree with that, his point (which I have just rejected) being that there is a collateral contract rather than a variation of the original contact. In South West Trains Ltd v Wightman [1998] PLR 113 ("South West Trains"), the salary increase was offered on terms, and the acceptance of the offer gave rise to a contract under which the salary increase was subject to those terms (there being no issue of invalidity similar to that in the present case). The position was the same in Bradbury v British Broadcasting Corporation [2012] PLR 283 ("Bradbury") although I have recently held a further hearing on the question whether the BBC was in breach of its duty of trust and confidence; judgment is reserved. Similarly, in The Trustees of the NUS Officials and Employees Superannuation Scheme v Pensions Ombudsman [2002] PLR 93 (the "NUS case"), the employee, Mr Allen, received a salary increase which was given to him on the express terms that it would not be pensionable. The Pensions Ombudsman concluded that, unlike in South West Trains, there had been no binding contract between Mr Allen and his employer. Lightman J disagreed. He held that Mr Allen had received an offer with

"two integral and interdependent elements, namely the increase in wages and the provision about pension entitlement. The terms could not be severed. It was not open to Mr Allen to accept one and not the other."

48. Further, Lightman J concluded that Mr Allen had accepted the offer because "he accepted payment of the increase and both he and the RMT continued to make pension contributions on the basis that the increase did not constitute part of Mr Allen's basic salary for pension purposes." I agree with Mr Spink when he submits that the contractual analysis must be that the agreement reached between Mr Allen and his employer varied his contract of employment; the variation included (i) an agreement by the RMT to pay an increased salary and (ii) an agreement by Mr Allen that the increase should not be pensionable. The salary increase clearly takes effect as a variation and since Lightman J held the two terms to be interdependent and not susceptible to severance it must follow that the non-pensionability agreement also became a term of the employment contract as varied.
49. Like Mr Spink, Mr Newman says that there is no contract because there is no consideration moving from Holdings to the affected member: by signing an NPA, the employee is not getting anything to which he or she was not otherwise entitled. That is not to say, on Mr Newman's case, that the NPAs are not valid for what they are ie a statement of Holdings' intention: it is simply that they are not valid and enforceable as contracts. The NPAs are in the nature of a protocol or convention being laid down by Holdings as employer as to the interpretation of salary increase offers in the future. Whilst employee agreement was sought, because the formulation of salary increase offers is a matter for Holdings alone, no such agreement was in fact legally necessary. As to that, I accept, of course, that no such prior agreement was necessary: as South West Trains, Bradbury and the NUS case show, the contractual relationship can arise at the same time as the award of the salary increase itself.
50. I have already referred to the 27 October 2009 emails relating to the Initial 2009 NPAs and the acceptance via the online tool. Mr Spink has alternative interpretations of the emails (assuming that no breach of duty were involved):

i) The NPAs represent the member's advance agreement to the terms on which salary increases will be offered. A contract is completed as and when a salary increase is granted: it is only at that stage that Holdings gives any consideration.

ii) Alternatively, the member's agreement to the 27 October 2009 email is to be seen as a unilateral offer from the employee that is accepted by Holdings when a salary increase is granted.

In my view, the first of those approaches is to be preferred although, analytically, there is probably an offer by Holdings (which is made by an actual offer or by an actual increased payment); that offer includes the non-pensionability term and is accepted by the member accepting the payment and continuing to work. I do not think anything turns on which interpretation is correct; under both of them, the eventual contract takes effect as a variation of the contract of employment.

51. It follows from the fact that the NPAs are not contractually binding that members are free to retract their agreement at any time. Mr Spink qualifies this by saying "at least before the first salary increase is given" although for my part I cannot see why each proposed salary increase should not be treated independently for this purpose. I will come to this in more detail later, but one of Mr Spink's submissions is that the members who sent protest emails did vary or qualify their agreement so that Holdings cannot say, in relation to them, that salary increases were given only on the basis that they were non-pensionable. He also submits (again I will come to this in more detail) that the NPAs cannot be relied on by Holdings because they gave rise to a breach of duty on the part of Holdings. He relies on what I said in Bradbury, namely that "the BBC ought not to be able to rely on a contract between it and a Member of the Scheme, the imposition of which would have given rise to a breach of any of the Implied Duties". What I did not say there, however, was anything about the impact on the salary increase awarded if the recipient of it subsequently sought to prevent the employer from relying on the agreement: it might be said that if Mr Bradbury sought to prevent the BBC from relying on his side of the bargain, so too the BBC ought not to be held to its side of the bargain.
52. Mr Newman seeks to derive a diametrically opposed conclusion from the fact that the NPAs are non-contractual. He submits:

i) the effect of the NPAs is to render non-pensionable any salary increase which might be offered by Holdings;

ii) but Holdings was not obliged to offer any salary increase;

iii) in particular, the employees' contracts of employment contain no express or implied rights to such increases;

iv) in the absence of such a right, the employment relationship does not give rise to any promise that an employee will receive any salary increases or even be considered for salary increases: see Lavarack v Woods of Colchester Ltd [1967] 1 QB 278; Murco Petroleum Ltd v Forge [1987] IRLR 50.

53. He submits, therefore, that the NPAs cannot be disregarded as part of the factual background against which any salary increase offer should be construed. And this is so, notwithstanding the breach of duty which has been established. He also submits that the NPAs cannot be declared void or voidable since, not being contracts, there is nothing to declare void or voidable. However, even if he is right about that, the actual non-pensionability term imposed when a contract does come into being might be void, voidable or unenforceable; that is a question which I have to address.
54. Before I turn to address the issues of voidness, voidability and unenforceability of the NPAs or actual contracts, I should say that I am willing to proceed on the basis that the NPAs are not themselves contracts. On the facts of the present case, I think that is so. But I should not be taken as accepting the proposition that any non-pensionability agreement which one might come across in relation to other employers and other pension schemes and which is entered into (like the NPAs in the present case) in advance of any actual salary increase proposal, is necessarily non-contractual. The agreement might form part of a larger agreement in relation to which consideration does move from the employer.
55. Questions as to whether the NPAs or ensuing contracts are void or voidable have been raised, but the central issue, it seems to me, is whether they are unenforceable and, if so, what the consequences of that are. Indeed, the Trustee's primary case is not that they are void or voidable but that they are unenforceable. The answer to the question of unenforceability is informed, however, by considerations which go to whether the NPAs and ensuring contracts are void or voidable. I propose, therefore, to deal with the questions of void or voidable first notwithstanding that these are fall-back positions for the Trustee.

Void/voidable?

56. Issues 1 and 2 relate to the NPAs themselves. They do not relate to the variations to the contracts of employment which follow on from the NPAs although similar issues arise which I will need to address in considering whether the NPAs themselves can be said to be void, voidable or unenforceable. As to the NPAs, I have held that Holdings was in breach of duty in procuring them (that is what I meant by saying that the NPAs gave rise to a breach of duty), the relevant breaches being summarised in paragraph 26 above. Issues 1 and 2 ask whether the Initial 2009 NPAs are void, voidable, unenforceable or valid. It is necessary to examine quite what the words "void", "voidable" and "unenforceable" mean in this context.
57. In the case of a document which creates or varies (or purports to do so) legal or equitable rights and obligations, it makes sense to talk about the document, or certain of its provisions, being void or voidable or unenforceable. A document which is void is a nullity in terms of creating those rights and obligations; it is as if the document had never been entered into and as a result the rights of third parties may be adversely affected. A document which is voidable may be voidable in whole or in part and it may, to some extent, be set aside on terms; third party rights may be protected. A contract may purport to create a right which is in fact wholly unenforceable: thus a covenant in restraint of trade will not be enforceable if it is of too wide an ambit. What is common to all these concepts is that they qualify apparent legal or equitable rights and obligations. An apparent right or obligation is held not to exist (void) or is set aside (voidable) or cannot be enforced. But even a void, voidable or unenforceable agreement can be relied on for some purposes. The relevant document still remains a document and it still has content. The fact that it is void or is set aside or contains terms that are not enforceable does not mean that it cannot be referred to for other purposes. It may, for instance, be highly relevant to the construction of some other document (by being expressly referred to in that other document) or to establishing the legal effect of other conduct (for instance in establishing what other legal or equitable rights and obligations arise out of the same transaction as the void/voidable/unenforceable document or out of some related transaction).
58. It is not at all easy, however, to apply these concepts in the case of an arrangement which does not give rise to any legal or equitable rights or obligations. If no rights or obligations are created, there is nothing to avoid or set aside (in the sense that those concepts are conventionally applied) and nothing which is unenforceable for there is nothing that is potentially enforceable in the first place. The NPAs are such arrangements since they are not contractual documents and therefore no party derives rights directly from them.
59. Although the NPAs do not create rights or obligations, what they do, or would do if valid and may do even if they are invalid, is to specify one of the terms on which a salary increase will be paid if and when an increase is awarded. Thus, if there were no NPA and if Holdings were to grant a salary increase without imposing any condition, then clearly the increase would be pensionable. In contrast, if the NPAs were valid, the increases would be non-pensionable because both Holdings and the members would be contractually bound by an agreement to that effect. I do not find the concepts of void/voidable/unenforceable transactions at all helpful in the context of the NPAs.
60. Mr Spink has relied on three authorities in relation to the void/voidable aspect of the NPAs: Mattu v The University of Coventry and Warwick NHS Trust [2013] ICR 270, Attrill and Annar v Dresdner Kleinwort [2012] EWHC (Owen J), [2013] EWCA Civ 394 and Johnstone v Bloomsbury Health Authority [1991] ICR 269. Given that I do not find those concepts helpful in the context of the NPAs, these authorities are not helpful either and I do not propose to address them in any detail. I wish only to say that, even if those concepts were helpful, those authorities do not take Mr Spink where he would like to go. In my view, there is no helpful parallel between Mattu, Attrill and Johnstone, on the one hand, and the present case on the other.
61. Rather than questions about void, voidable and unenforceable, better questions to ask, it seems to me, are whether the agreement of a member signing an NPA that any increase should be non-pensionable is vitiated (or is rendered invalid) as a result of Holdings' breach of duty; and if so, whether that agreement is then to be treated as never having been given or is to be treated as having been withdrawn at some later time and, if so, when.
62. Although he has argued the case in the language of void/voidabale, Mr Spink's submission can be expressed in the language of invalidity/vitiation. He submits, in effect, that the NPAs are invalid and not to be treated for any purpose as providing assent to any salary increase taking effect only subject to the non-pensionability term: the members' agreement is to be treated as never having been given. Accordingly, the terms of the relevant NPAs are not incorporated into any subsequent offer or into any contract under which the salary increase is effected. On this basis, the question whether the non-pensionability term would be void, voidable or unenforceable as a term of the varied contract of employment does not arise because it never finds its way into any binding contract in the first place.
63. The principal contrary argument, it seems to me, is that, valid or not, the terms of the NPA formed part of the offer of the salary increase and that it is only in the context of the contract of employment as varied (to include both the salary increase and the non-pensionability term) that the validity of the non-pensionability term falls to be assessed. In other words, the position is no different from what it would be if there had been no NPA in the first place but simply an offer (as in South West Trains and the NUS case) of a particular salary increase on terms, albeit an offer on terms which gave rise to a breach of Holdings' contractual duty of trust and confidence.
64. The answer to these rival approaches is informed by consideration of the sort of offer of salary increases which Holdings might have made in the light of the Initial 2009 NPA which was current at the time of the first salary increase. There are two courses (differing from what it actually did) which it might have taken:

i) First, Holdings could have written to members on the occasion of the first salary increase expressly offering the increase on the basis that it would be non-pensionable. Either that condition would be valid or it would be invalid, but it would be clear that the offer was made (and accepted) on that basis.

ii) Secondly, instead of doing that, Holdings could have made the offer expressly by reference to the agreement already made. In that case, it would have been clear that the offer was being made on the basis that the increase would be non-pensionable; the reference to the agreement could only be taken as a reference to the terms of the first option contained in Mr Ferrar's email of 27 October 2009 which the members had elected to adopt. This, in my view, would be so even if it were subsequently held that the NPAs had been procured in breach of duty. This is because the NPAs would not be relied on as a source of rights or obligations, but only as a document containing terms incorporated by reference in the offer. The fact that the agreement to those terms may be invalid having been procured in breach of duty, does not mean that the terms themselves, contained in Mr Ferrar's email to members on 27 October 2009, could not form the terms of an offer made by Holdings any more than they could not have formed the terms of an offer under i) above. As under i) above, either that condition would be valid or it would be invalid, but my conclusion is that the offer would have been made (and accepted) on that basis.

65. Holdings did neither of those things; at least, the mass of evidence which I had at the trial did not, so far as I recollect or have been able to find, discloses that it did either of those things. Instead, as I understand matters, Holdings simply implemented salary increases, no doubt in the belief that the NPAs would govern the terms on which the increases were awarded because, in its view, there was nothing at all objectionable in the salary increases being made subject to the non-pensionability term.
66. In those circumstances, what are the members to be treated as having agreed to by accepting the salary increases? And, if they are to be treated as having agreed to the non-pensionability term as one of the terms of receipt of that increase (by way of variation of their contracts of employment), what is the impact of the breach of duty?
67. In answering those questions, it is helpful to address the contractual position had there been no breach of duty. The starting point (see paragraph 48 above) is that the NPAs represent the members' advance agreement to the terms on which salary increases will be offered: a contract is made as and when a salary increase is granted since it is only at that stage that Holdings gives any consideration. The next step in the contractual analysis is that by implementing the salary increase, Holdings is making an offer to the members that, in consideration of their continuing to serve, they will receive the salary increase. But clearly the members cannot accept the salary increase without accepting also that the increase will be non-pensionable. The reason that that is so is because it must be taken to be implicit in the offer that the terms of the (valid – my hypothesis in this paragraph being there is no breach of duty) NPA are incorporated into it; by accepting the salary increase, the members accept the non-pensionability term. It is not clear (and in the context of a valid NPA does not matter) whether the non-pensionability term, when given contractual force, is properly to be seen as an express term or an implied term. It might be said to be express because the implicit term of the offer is itself express in the sense that there is an express agreement (the NPA) that any salary increase will be non-pensionable. On the other hand, it might be said to be implied because there is no difference between something being implicit and being implied.
68. Considering, by analogy with the case of a valid NPA, the contractual position on the facts of the present case (but putting aside for the moment the position of those members who sent emails of protest), the starting point, again, is that the NPAs represent the advance provision of the members' agreement but with the important difference that that agreement has been obtained only as a result of a breach of duty on the part of Holdings. What is not clear, however, is that the terms of the NPA obtained from members in breach of Holdings' duty, can be relied on as part of the offer which the members accept by receiving the salary increases awarded.
69. In my judgment (ignoring for the moment the position of those members who sent protest emails), the terms of the NPAs are incorporated into the terms of the offers and, by accepting the salary increases, the members are to be treated as accepting offers on those terms. The validity of the non-pensionability term is therefore to be assessed in the context of the contracts of employment as varied. Members may have been – and many clearly were – unhappy about Holdings' conduct in procuring the NPAs but there is no doubt that they actually made the elections, which they did pursuant to the online tool, and thereby signified their agreement to the non-pensionability of salary increases. Members did not withdraw their election or agreement prior to the salary increases: it is clear that Holdings intended the salary increase to be subject to the non-pensionability agreement. It is not possible, in my view, to interpret the offer as one of a salary rise subject to the non-pensionability term if and only if the NPAs involved no breach of duty. I therefore consider that the members' challenge to the validity of the non-pensionability term must be made within the context of the contracts of employment as varied. This is an aspect I deal within when addressing Issue 3.
70. I need to say something more about the members who did send protest emails. A typical email (and I think the example raises all the points which can be taken) contained the following:

"In summary, I do not feel that I have been given sufficient information or time to make an informed decision. I feel that l am being pressured into making a choice which may be irrevocable. I believe that the lack of information and the pressure to make a decision in a very short timescale amounts to a breach of the implied term of trust and confidence in the employment relationship.

I wish it to be placed on record that I feel unable to make an informed decision, and have completed the choices tool only because of the time pressures I am being put under. I do not consider my action in completing the choices tool to be informed consent to any change in my terms and conditions, or the terms governing my pension benefits. I therefore reserve the right to change my position when I have received, and been able to properly consider, your response to the points I make above.

For the avoidance of doubt, I continue to consider the Trustees bound to provide benefits as set out in the rules of the I Plan and do not believe that IBM has the right to vary the terms on which I am a member of the I Plan except by the amendment power contained in the rules of the I Plan."

71. I find it impossible to read that email other than as a clear statement that the member concerned does not accept that any further salary increases will be non-pensionable. The only basis on which Holdings could maintain that the member concerned would be subject to the non-pensionability term when he came to be awarded a salary increase is because he or she had ticked the acceptance box. The express term of this email must, in my view, be effective to override the tick in the box and to make clear the member's lack of agreement. I do not disagree with the proposition that, had I held the NPAs to involve no breach of duty (either in isolation or as part of Project Waltz), then a member sending an email such as this could not take the benefit of the salary increase without being subject to the non-pensionability term. But that is not to say that, for such a member, he is bound by it given the breach of duty which I have held to be involved. If Holdings, in the face of an email such as this, chose to give a salary increase, it was open to the member to continue to serve and to claim the increased salary as a valid variation of his contract of service. It was Holdings which took the risk that the NPAs might be held to involve a breach of duty. In terms of offer and acceptance, the members' emails demonstrate that there was no NPA vis à vis such members so that the offer made by granting the salary increases (without at the same time expressly incorporating the non-pensionability term) was not an offer which included that term. This conclusion feeds into Issue 5.
72. I consider this result to be correct whether the member made his election via the online tool before or after the email. If it was made before, then his election was clearly qualified and could not sensibly be taken by Holdings as giving the consent which was sought; if it was made after, it withdrew the consent which had been given – clearly a member was entitled to withdraw consent.
73. But having said that, I do not consider that the I Plan members who sent emails of this sort can say that any salary increase is wholly pensionable. Although the 2006 Partial NPAs were not reflected in an amendment to the I Plan (in contrast with the position under the Main Plan), it is obvious, if I may say so, that in referring to the obligation of the Trustee to provide benefits in accordance with the terms of the I Plan, the member did not have in mind the 2006 Partial NPAs; thus the email says the member "continue(s)" to consider that the Trustee is bound to provide benefits under the terms of the I Plan when, had he thought about it, he would have realised that the Trustee was only bound to provide benefits which reflected the 2006 Partial NPAs. If the analysis is that the members sending such emails were not consenting to the non-pensionability term, then the 2009 NPA was effectively being rejected. There is no reason to conclude that the terms of the 2006 Partial NPAs were abrogated vis à vis these members.
74. Other members sent a different form of objection. In particular, clarification was sought of Mr Ferrar's first email leading to the Initial 2009 NPAs and seeking further time in which to make a decision. The typical email ended with this paragraph:

"In the circumstances, I look forward to receiving clarification of the ultimatum and an indication I may delay responding to it until the Trustee's application to the court has been dealt with. If I do not receive such clarification then any selection (or lack of) I may make will be under duress and I would demand the right to alter any choice in the future, once full clarification has been provided and in the light of the outcome of the Trustee's application to the court."

75. This email demonstrates a different attitude from the email referred to at paragraph 68 above. In the case of the latter, the writer demonstrates that agreement is not being given and that benefits should continue to be paid under the I Plan. But in the case of the former, the writer appears to acknowledge that his agreement is given but that it has been procured under duress. Accordingly, the terms of the NPA are incorporated in the contract of employment when the salary increase is accepted, but the member concerned remains able, in my view, to rely on the email to demonstrate the pressure (described by the Trustee as duress) under which he was acting in relation to the contractual variation.
76. Accordingly, the conclusion in relation to Issue 1 and its various sub-issues is that they ask the wrong questions and are not proper issues at all. The Initial 2009 NPAs would not of themselves give rise to any enforceable rights and obligations even absent any breach of duty so that to ask whether they are void or voidable is not a proper question. Those concepts are apposite only to documents which do give rise to such rights and obligations. As to whether they are enforceable or not, that is again, in my view, to ask a non-question. The issue is not whether the NPAs are, of themselves, unenforceable. The issue is whether the terms of the NPAs are effectively incorporated into the contractual offer and acceptance in relation to the salary increases and, if so, whether those terms can be relied on by Holdings.
77. As to that, my conclusion is that the non-pensionability term did form part of the offer which was accepted by the members continuing to serve and accepting the salary increase. Whether that term is then void, voidable or unenforceable is a matter I address under Issue 3. That conclusion is subject to the position of members who sent protesting emails such as I have set out at paragraph 68 above. For those members, I do not consider that the contract arising as the result of the salary increase and its acceptance by the member included the non-pensionability term.
78. It follows that Issue 1 is to be answered in the following way. The Initial 2009 NPAs are not void or voidable or unenforceable. But, having been procured in breach of duty, the non-pensionability term does not form a term of the varied employment contract of a member who sent a protest email along the lines of the email from which I have quoted at paragraph 68 above. As to members who did not make any sort of protest, the non-pensionability term did form part of the salary increase offer which was made and accepted. Whether that term, as part of the contract of employment, is void, voidable or unenforceable is a matter I deal with under Issue 3. In relation to Issues 1.2(a) and (b), I would only add that the effect of the NPAs is to be considered on a member-by-member basis.

Issue 2: the Later 2009 NPAs ? this issue raises similar issues as Issue I mutatis mutandis

79. The analysis which I have carried out in relation to the Initial 2009 NPAs applies equally to the Later 2009 NPAs. The Later 2009 NPAs give rise to a breach of duty on the part of Holdings when viewed in the context of Project Waltz as a whole. The difference between the two categories is that the Initial 2009 NPAs would give rise to a breach of Holdings' duty even if viewed in isolation (under either of Scenarios A or B). But that hypothetical breach is not the actual breach of duty for which a remedy is sought. The difference may feature, however, when it comes to considering the consequences of the actual breach which I turn to under Issue 3. The answers to Issue 2 are therefore, mutatis mutandis, the same as the answers under Issue 1.

Issue 3: As regards any 2009 NPAs that are void and/or unenforceable, or are avoided and set aside:

80. 1 • are members entitled to keep all or part (and if so, which part) of the salary increases they were awarded subsequent to entering the agreements,

• and are they entitled in the future to continue to be paid salary incorporating such increases;

81. 2 are all or part of such salary increases (and if so, which part) to be treated as pensionable for the purposes of the DB Plans?
82. The Trustee's position is that the affected members can retain the salary increases which they have been awarded and can continue to enjoy them into the future and the increases are all pensionable. In effect, it is said that the salary increases take effect but that the non-pensionability term is either not a term of the contract of employment as varied or, if it is, it is void, voidable or unenforceable.
83. Holdings' position is that the affected members are entitled to keep the salary increases but only on terms that they are all not pensionable for the purposes of the DB Plans: those were the terms on which the salary increase contracts were made so that the salary increases stand or fall with the terms on which they were awarded.
84. Although those are the stated positions of the parties in the context of Issue 3 as formulated, it will be clear from my analysis under Issue 1 that I perceive the real question as being directed at the effect on the pensionability term – as a term of the varied contract of employment – of Holdings' breach of duty.
85. I make the immediate observation that the so-called salary increase contracts referred to by Holdings take effect as variations to the contracts of employment as already explained. This is important because I am not concerned simply with a contract for salary increases where, in consideration of an increase, the member agrees that it will be non-pensionable. Rather, I am concerned with a larger contract, namely the contract of employment: the consideration for the salary increase is, principally, the agreement to continue to provide services with the non-pensionability term forming only part of that agreement. Further, if I am right in my conclusions in relation to members who sent protest emails, the non-pensionability term did not become a term of the contract of employment as varied at all.
86. In this context, although Holdings may not have been under any legal obligation to award any salary increase, it is perfectly clear, I consider, that, when it is granted, the increase can be enforced by the employee.
87. The Trustee's case is that the salary increases do not stand or fall with the non-pensionability term. First, it says that on a proper construction of the terms of the agreement under which the salary increases were awarded, those increases are not repayable; and secondly (which is really the point I have just made), this is not a case where the whole of the consideration for the salary increase has failed. I think the Trustee's case can be stated in this way (although it is not quite how Mr Spink puts it): Holdings has awarded salary increases which take effect as variations to the contracts of employment; it has attempted to attach conditions to the increases; although it has obtained the agreement of the members to those conditions, that agreement was obtained in breach of duty; therefore the conditions are invalid; and so the salary increases take effect free from the conditions. Alternatively, the non-pensionability term is more than a condition attached to the salary increase, but is a term which is, taken by itself, invalid: it can and ought to be set aside without the need to set aside the whole of the contractual agreement which resulted in the variation of the contract of employment. In either case, the risk of the non-pensionability term being held invalid was Holdings' risk. In this context, to adopt Holdings' position would be to leave the salary increases in place only if the non-pensionability term is given effect; otherwise, the members are not entitled to any salary increases. This would, it is said by the Trustee, be unfair to members since they would have no increases, which would be to condone the breach because it would implement the default option under the online tool.
88. To put this another way, as he does, Mr Spink says, in the context of Issue 3.1, that the members did exactly what they had been asked to do in that they had accepted that their salary increases would not be pensionable. They had not provided a warranty that this would be the case, nor agreed to any term whereby they would repay or lose the salary increases if the NPAs were unlawful or ineffective. Further, the members performed their part of the agreement by working for Holdings on the basis of the agreed salary: that is labour that they have given and, quite obviously, cannot take back. The position is entirely different from that considered by Lightman J in the NUS case where the non-pensionability term involved no breach of duty on the part of the employer so that the terms on which the salary increase was offered became terms of a valid and enforceable agreement.
89. It follows from my analysis under Issue 1 that there is no question of the variations to the contracts of employment being void. If they were, it would follow that no salary increases had been awarded to anyone which, I have to say, would be an unacceptable and surprising result. Nor, I think, is there any question of the voidness or setting aside of any free-standing contract in which, in consideration of Holdings awarding a salary increase, the members agreed that that increase would not be pensionable. In the first place, there would not appear to have been any such free-standing contract: rather, there was an overarching variation of a member's contract of employment as already explained with consideration moving each way – on the member's part to continue to serve for the increased salary (with the increase being non-pensionable) and on Holdings' part to pay an increased salary for that service. Secondly, even if there had been such a free-standing contract, it was superseded by, and merged into, the contract of employment. If the members have a remedy, it is relation to the enforceability or otherwise of the non-pensionability term as a term of the employment contract.
90. Voidable: I consider that a contract procured as the result of a breach of the contractual duty of trust and confidence is voidable; the innocent party is not restricted to a claim in damages. There is no good reason why, as a matter of principle, this should not be so any more than in cases of duress, fraud or mistake.
91. Normally, a contract can only be set aside where it is possible to put the parties back into the position in which they would have been if the contract had not been made; in other words, full restitutio in integrum is required. This is no doubt the basis on which Holdings contends that a claim by members for their salary increases to be pensionable requires the salary increase to be foregone for the future and repaid for the past: the contracts resulting in the variation of the contracts of employment have to be set aside in their entirety.
92. The question then is whether the non-pensionability term can somehow be abrogated without the members losing their salary increases. In my judgment, it can, and should be. This, subject to the separate issue relating to the effect of the 2006 Partial NPAs, is for the following reasons which I develop in the paragraphs below:

i) First, the non-pensionability term is not enforceable in the sense that Holdings would not be entitled to injunctive relief to prevent the members from claiming their full pensions.

ii) Secondly, the 2009 NPAs gave rise to a breach of duty on the part of Holdings. The implementation of those NPAs by the inclusion of the non-pensionability term as a term of the contracts of employment is to be treated as giving rise to a breach of Holdings' duty just as much as the 2009 NPAs themselves. The non-pensionability term is severable from the other provisions of the contracts of employment. The inclusion of that term was of the essence of the breach of duty and is invalid. The remainder of the contracts of employment can properly stand even in the absence of the non-pensionability term.

iii) Thirdly, the enforcement of the non-pensionability term by Holdings would give rise to a further breach of the duty of trust and confidence and should be restrained.

93. Mr Spink made a number of arguments about why the NPAs should be regarded as unenforceable. And Mr Newman made arguments about why they should not be. I can transpose those arguments in the light of my analysis under Issue 1, so as to address the enforceability of the non-pensionability term as a term of the contracts of employment.
94. The arguments on behalf of the Trustee start with the passage from my judgment in Bradbury quoted at paragraph 49 above. As I will explain in a moment, I stand by what I said there. However, I said nothing there about the ability of Mr Bradbury to rely on the pay rises which he obtained from the BBC which thought, when it awarded them, that they would not be pensionable. I do not think that anything which I said in Bradbury gets Mr Spink to where he wants to be.
95. The submission made by Mr Spink which I now wish to deal with is in essence that Holdings' case has to rely on an implied term to the effect that members will not claim that their salary increases are pensionable ("the disputed term"). The NPAs do not contain an express term to that effect but only provide an agreement by the members that the salary increases will not be pensionable and therefore it is necessary to rely on the implied term. It is not possible, in his submission, to imply the disputed term when to do so would be to condone Holdings' breach of duty in procuring the NPAs.
96. I reject that submission. In South West Trains Neuberger J (as he then was) accepted my submission as counsel (which I then was) that it was implicit in the binding agreement between the employer and the employees (ie the express term that the salary increase would not be pensionable) that the employees would not claim pensions at a higher level than agreed. He saw no difficulty in implying a term into the contract to that effect either because such a term went without saying or because the contract simply would not work without such an implied term. He would obviously have reached the same conclusion applying the more comprehensive test stated by Lord Hoffmann in A-G of Belize v Belize Telecom [2009] 1 WLR 1988:

"… 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 provision would spell out in express words what the instrument read against the relevant background, would reasonably be understood to mean."

97. It followed that, were the employees to seek to claim the higher pension, the employer could obtain an injunction restraining them from doing so.
98. Lord Hoffmann's test for the implication of terms (if implication is relevant at all) which I have just set out is the one which I should apply. The traditional test for implication has been one of necessity, and the traditional tests – for instance "goes without saying" and "necessary to give business efficacy" – are only tools which different judges in different circumstances have used to find the answer but they are not additional or different tests. Some care must, however, be taken in applying what Lord Hoffmann said out of context. The use of the words "reasonably be understood to mean" does not substitute a test of reasonableness for one of necessity. As Lord Hoffmann said in addressing the "necessary to give business efficacy" test:

"The second [important point], conveyed by the use of the word "necessary", is that it is not enough for a court to consider that the implied term expresses what it would have been reasonable for the parties to agree to. It must be satisfied that it is what the contract actually means."

99. It is on the basis of what Neuberger J said about the implication of a term that Mr Spink submits that Holdings needs to rely on the disputed term being implied into the contracts of employment as varied. And, given the need to identify an implied term, he goes on to present three arguments which he says strongly militate against the implication of the disputed term.
100. Mr Newman says that this is to ask the wrong question. He says, following Lord Hoffmann, that the question of implication of terms is essentially one of construction taking into account all of the relevant background and asks what a reasonable person, having that knowledge, would consider the contract to mean. Applying that approach, there can be no doubt the disputed term is to be implied into contracts where the non-pensionability term is to be found (ie as I have explained, the contracts of employment). The correct approach is to ask (i) whether the breach of duty on the part of Holdings is such as to justify imputing a different intention to the parties (ii) whether it is sufficient to disregard what the parties actually thought the terms of the contract meant and (iii) whether it is sufficient to require one of the parties to be held to a bargain it never agreed to and which the other party never thought was on offer. Mr Newman's argument, therefore, is that the correct approach is to start off by identifying what the terms of the contract are without reference to extraneous matters of enforceability. He goes so far as to say that the disputed term should be implied even if (contrary to his position) an express term to the same effect would be held unenforceable because of the breach of duty; in other words, the disputed term should be implied even though it would not be enforced.
101. Why does this superficially rather arid debate matter? It might be said that it does not matter because:

i) an express term that the affected member will not seek to claim his full entitlement from the Trustee would either be valid or it would be unenforceable;

ii) if it would be valid, there is no reason not to imply the disputed term;

iii) if it would be unenforceable, it does not make any difference whether the disputed term is implied or not, since the member could claim his full benefit and the Trustee would have to provide it;

iv) so the real issue is whether an express term to the same effect as the disputed term would be valid.

102. The reason why it might matter whether (a) the disputed term should be implied even if it is unenforceable rather than (b) it should not be implied at all, is because the remedies may be different. If the disputed term is not to be implied at all, there remains, nonetheless, a contract between the affected members and Holdings which provides (i) for a salary increase and (ii) for that salary increase to be non-pensionable; but which does not include (iii) the disputed term. In contrast, if the disputed term is to be implied only to be held unenforceable, then its unenforceability by Holdings may have a consequence on the enforceability by the member of the salary increase awarded under the very same contract which includes the invalid (implied) disputed term.
103. Quite what the effect of (ii) in the preceding paragraph would be in the absence of (iii) is not at all clear to me. How could (ii) be enforced other than by (iii)? I suppose that, if a member were to receive benefits from the Trustee on the basis that salary increases were pensionable, Holdings could in turn obtain damages for breach of contract. Surely, it can be asked, receiving the benefit must breach the agreement that the salary increase would not be pensionable? But if that is so, why would the court not be willing to grant an injunction to restrain the breach of contract in the first place? And if such an injunction could be obtained, what is the difference between that and the term which, on the hypothesis now under consideration, is not to be implied?
104. In South West Trains, both Mr Christopher Nugee QC (as he then was) and I suggested that there was a difference between something being implicit and being implied. Neuberger J did not find that attractive. But the present case demonstrates that such a distinction might be valid, although the distinction becomes clearer if one uses the word "inherent" rather than "implicit". In other words, it is part of the nature of an agreement providing for a salary increase to be non-pensionable that a pension cannot be claimed in respect of it. It is not because the disputed term is to be implied that the member will not be able to make a claim; rather, the member will be unable to do so because there is no claim in the first place. Accordingly, the validity of the NPAs – or rather, of the express term of the contractual variation of the employment contracts – is to be judged by reference to the validity of a single term (ie the non-pensionability term) which properly construed contains within it an agreement not to claim a pension based on the salary increase. In other words, the disputed term is inherent in the non-pensionability term.
105. That is the approach which I favour. But if that is not open to me because of the approach of Neuberger J in South West Trains, I think that the conundrum can be solved in a different way which leads to the same result. The solution, it seems to me, is this. The reason why, in cases of valid non-pensionability agreements, a term is implied (to the effect that the member will not seek to obtain from the pension scheme a benefit based on his full salary) is because that is implicit; and the item in which it is implicit (or inherent) is the express term that the salary increase will be non-pensionable. If one asks what is meant by an increase being non-pensionable it means, quite simply, that as a matter of contract between employer and employee, a particular increase will not count towards pension; it would be inconsistent with such an agreement for a member to be able to claim the full benefit for which the rules of the scheme provide.
106. But there is a risk of such inconsistency in the present case, on Mr Spink's approach, if Holdings' breach of duty precludes the implication of the disputed term. Notwithstanding the express term of the contract arising as a result of the NPAs (ie the term that any salary increase will be non-pensionable), the member will be able to claim the full pension from the Trustee with the result that it will, after all, be pensionable. The inconsistency must, in my view, be eliminated; if it cannot be eliminated by the implication of the disputed term, it must be eliminated in some other way.
107. To see how this can be done, it is necessary only to note that I have held Holdings to be in breach of its duty of trust and confidence in procuring the making of the NPAs and, a fortiori, it is in breach of duty in attaching the non-pensionability condition to the salary increases it actually awards. This breach of duty has an impact not only on the question of implying the disputed term but on the enforceability of the express term which gives rise to the implication, namely the agreement between the employer and the employee that the salary increase should not be pensionable. The factors which would preclude the implication of a term may well preclude reliance by Holdings on the express non-pensionability term.
108. That last sentence can be expressed more strongly the other way round: factors which go to the enforceability of the express term will also be relevant to whether the disputed term should be implied. I go further than that and conclude that the only reason not to imply the disputed term would be because the express term on which it is essentially parasitic is itself unenforceable as a result of Holdings' breach of duty. The way this works in terms of enforcement of the contractual obligations is this:

i) A member seeks a full pension taking into account the salary increases.

ii) Holdings asserts a valid term of the varied contract of employment including the disputed term. Absent Holdings' breach of duty, Holdings would succeed.

iii) The member responds that the non-pensionability term is invalid and that therefore the disputed term is not to be implied. The member needs to assert the invalidity of the express term because, if the express term is valid, there is no reason, in my view, not to imply the disputed term. It does not matter, in terms of what the member can recover by way of pension, whether the analysis is that the disputed term is not to be implied or whether it is to be implied but, like the express term, is unenforceable. Perhaps, like Schrödinger's cat, the disputed term is both alive (implied) and dead (not implied) and it is only when Holdings seeks to apply it that it becomes inapplicable.

iv) On this basis, the potential inconsistency identified above is eliminated. There is no inconsistency between the member claiming his full benefit if the express term which would otherwise give rise to the implied term is itself unenforceable. In contrast, if the express term is enforceable, there is no reason not to imply the disputed term and again inconsistency is not present.

v) Whether the affected member actually can claim his full benefit is a different question. The issue here is whether, assuming the express non-pensionability term to be unenforceable, the salary increase itself is vulnerable, a matter to which I will come.

109. What this shows, it seems to me, is that it does not matter much whether Mr Newman is right in saying the disputed term is implied even if it is unenforceable or whether Mr Spink is right to say that it is not to be implied in the first place or whether I am right in saying that it is inherent in the express term that the full benefit will not be claimed. In all these cases, the affected members need to allege that the non-pensionability term, as a term of the varied contract of employment, is invalid. If they do not allege that, then they must accept that the express term is valid and if it is valid there is no basis on which they can say that the disputed term should not be implied even if it not already inherent in the express term. Whether the express term is unenforceable seems to me to raise precisely the same questions as whether, were the disputed term an express term, it too would be unenforceable.
110. In the light of that analysis, there is no need, in my view, to address Mr Spink's three arguments which are deployed to demonstrate that the disputed term should not be implied. As explained, the affected members need to show that the express non-pensionability term of the varied contracts of employment when applied to a particular salary increase is unenforceable. The answer to the question of whether that term is enforceable is the same whether or not the disputed term is inherent in it.
111. It is necessary now to address the question whether the express term as to non-pensionability of a particular salary increase in the contracts of employment as varied is unenforceable. In my judgment, it clearly is unenforceable in the sense that the affected members could not be prevented from claiming benefits based on salary increases being pensionable: Holdings would not be entitled to injunctive relief against the affected members preventing them from doing so. Given that the 2009 NPAs gave rise to a breach of Holdings' duty and – more importantly – given that the actual non-pensionability condition attached to salary increases gave rise to breach of duty, a court of equity ought not, in my judgment, to compel the affected members to give effect to an agreement which it was a breach of duty on the part of Holdings to obtain. Holdings' position is not helped by the absence of proper consultation of a non-trivial nature (as to which see Judgment [1585]).
112. This analysis does not address how the refusal of injunctive relief impacts, if at all, on the express non-pensionability term as a matter of contract. By seeking the full amount due, the affected members would be breaching that contract unless the non-pensionability term can be set aside rather than not being enforced by injunction. Holdings might, I suppose, then have a damages claim although determining the correct measure of damages is likely to give rise to significant difficulties but this has not been the subject of any submissions and I say no more about it.
113. Mr Spink also relies on the principle of construction that a party is not able to take advantage of his own wrong, referring to Alghussein Establishment v Eton College [1988] 1 WLR 587 at 595, where Lord Jauncey said that there is a "presumption that it was not the intention of parties that either should be entitled to rely on his own breach in order to obtain a benefit". In order to displace this principle there must be "a clear contractual intention to be gathered from the express provisions of the contract". The conclusion in the present case, consistent with that statement, is that there is no clear contractual intention that Holdings should be able to rely on and enforce the term as to non-pensionability irrespective of whether it was procured, or itself amounts to, a breach of the contractual duty.
114. It is to be emphasised, however, that the principle exemplified in Alghussein Establishment v Eton College is a principle of construction. As such, it can be displaced by the express terms of a contract. Mr Newman refers to BDW Trading Ltd v JM Rowe (Investments) Ltd [2011] EWCA Civ 548 where Patten LJ sets out the relevant principle at [31] of his judgment:

"Although there has been a certain amount of academic discussion as to whether the principle has the status of a rule of law which is imposed upon the parties to a contract almost regardless of what they have agreed, it is now clear as a matter of authority that the application of the principle can be excluded or modified by the terms of the contract and that its scope in any particular case will depend upon the construction of the relevant agreement."

115. Mr Newman also refers to [35] where Patten LJ refers to the judgment of Potter J in Richco International v Alfred C. Toepfer International [1991] 1 Lloyd's Rep. 136. Referring to the statement of the principle in Chitty on Contracts (30^th ed) at 12.082 which stated the principle in this way:

"It has been said that, as a matter of construction, unless the contract clearly provides to the contrary it will be presumed that it was not the intention of the parties that either should be entitled to rely on his own breach of duty to avoid the contract or bring it to an end or to obtain a benefit under it."

Patten LJ went on to say this:

"I am prepared to accept the principle as stated in Chitty subject to the reservation that as an exercise in construction the requirement of 'clear express provisions to the contrary' should not be read as meaning more than a clear contractual intention to be gathered from the express provisions of the contract."

116. I have to say that I do not really see the scope for the application of the principle in the present case. The relevant contract is the contract of employment as varied by (i) an increase in salary and (ii) the inclusion of a term that the increase in salary is not to be pensionable. Its meaning is clear. For Holdings to seek to enforce that latter term is not, subject to one point, a breach of the contract of employment at all: quite the reverse, Holdings would be acting in accordance with the terms of the contract and there would be no wrong of which Holdings would be seeking to take advantage. The principle is not, I think, applicable where the wrong complained of is the procuring of the very term which is said to be unenforceable.
117. Mr Spink perceives some support for the conclusion that the members are entitled to retain the benefit of the salary increases awarded (both for the past and the future) and to claim that they are pensionable without facing any cross-claim in damages by Holdings in the cases concerning duress. He submits that the courts have recognised that it is no answer to a plea that a contract is voidable and should be rescinded for economic duress that full restitution of benefits received by the innocent party under the contract may not be possible. He refers to Borrelli v Ting [2010] UKPC 21. In that case, liquidators sought successfully to set aside a settlement agreement which had been procured as a result of illegitimate means on the part of the defendant. It was argued on the part of the defendant that the parties could not be put back into the position which they occupied before the settlement agreement. It is not entirely clear why this was said to be so, but this can only have been because the defendant had lost his bargaining position since the liquidators had already enjoyed the benefit of the settlement agreement, namely the withdrawal of his effective veto on a certain scheme which the liquidators needed to implement. That veto had been improperly exercised, which represented the illegitimate means by which the defendant had procured the settlement agreement giving rise to an allegation by the liquidator of economic duress. And so, in relation to the submission that the liquidators could not set aside the settlement agreement because the parties could not be put back in the position they occupied before, the Privy Council said this at [39]:

"The main difficulty with that submission in the context of a case like the present is that it necessarily relies on the unacceptable proposition that because the parties cannot be restored to the position created by the illegitimate means employed by James Henry Ting, which resulted in the Settlement Agreement, the Liquidators are bound by that agreement. In truth the avoidance of the Settlement Agreement by the Liquidators did not amount to making a new bargain for the parties, but instead put the parties in the position they would have occupied had James Henry Ting not resorted to illegitimate means in order to secure the Settlement Agreement."

In other words, absent the illegitimate means, the parties were in fact put into the position they would have been in had the defendant acted properly.

118. I do not find this a particularly helpful authority in the context of the present case. It does, I accept, demonstrate some flexibility in the application of the restitutio in integrum principle where a defendant has acted unlawfully but it does not establish a wide power in the court to set aside parts of a contract where the parties would not be put back into anything like their positions before the contract. And although there were what have been referred to by Mr Spink as threats in the present case, the RBs' claims are not, at least in the current action, based on economic duress. In the present case, if Project Waltz as a whole and the NPAs in particular had not been proceeded with, it is entirely unclear, at present, what Holdings would have done. It is certainly not clear – indeed Holdings suggests that precisely the opposite is clear – that Holdings would have given the same salary increases nonetheless.
119. Of more assistance, I think, is the decision of Mr Jonathan Sumption QC (as he then was) in Marshall v NM Financial Management Ltd [1995] 1 WLR 1461 upheld by the Court of Appeal at [1997] 1 WLR 1527. In that case, a covenant in a commission was found to be an unlawful restraint of trade and therefore unenforceable. The issue was whether the claimant was entitled to retain the commission agreement. At first instance (see at p 1047E), the Judge said this at p 1446B:

"The principles on which unlawful provisions of an agreement may be severed from the rest of the terms are reasonably clear. In Sadler v. Imperial Life Assurance Co. of Canada Ltd. [1988] I.R.L.R. 388, 392, Mr. P. J. Crawford Q.C, sitting as a deputy judge of the High Court, was concerned with a contract which was in many ways very similar to the one before me. He summarised the conditions for the severance of void or unlawful terms in this way:

"1. The unenforceable provision is capable of being removed without the necessity of adding to or modifying the wording of what remains.

2. The remaining terms continue to be supported by adequate consideration.

3. The removal of the unenforceable provision does not so change the character of the contract that it becomes 'not the sort of contract that the parties entered into at all.'"

120. In the Court of Appeal, Millett LJ gave the leading judgment. It is worth setting out a few passages:

At p 1531E-H:

"It is obvious that where the invalid restraint of trade provides the only consideration for the promise, the promisee cannot enforce it. He has given no (valid) consideration for the promise which he seeks to enforce. Shorn of the "adventitious trappings" of the contract, this was the position in Vancouver Malt and Sake Brewing Co. Ltd. v. Vancouver Breweries Ltd. [1934] AC 181. It was also the position in Wyatt v. Kreglinger and Fernau [1933] 1 K.B. 793 where on the eve of his retirement an employee who otherwise had no entitlement to a pension was granted an ex gratia pension conditional on his not competing with his former employer. The condition was held to be void and the pension not payable. At the other extreme are cases where the invalid restraint is merely an incident of a larger transaction which could "survive without difficulty the elimination" of the invalid restraint: see Romer L.J. in Goodinson v. Goodinson [1954] 2 Q.B. 118, 126. This was the position in In re Prudential Assurance Co.'s Trust Deed; Home v. Prudential Assurance Co. Ltd. [1934] Ch. 338 and Bull v. Pitney-Bowes Ltd. [1967] 1 W.L.R. 273. In both cases a contract of employment entitled the employee to a pension with a proviso that upon retirement after the qualifying period of service the employee would not compete with his former employer. In both cases the pension would continue to be payable even though the proviso was void. Most cases fall in between these two extremes. The invalid restraint is only part of the consideration for the promise, but it is an important part, for without it the promise would probably not have been given."

And at p 1532 C-H:

"But it is often and perhaps usually the case that the promise would not have been given but for the invalid restraint, yet this does not prevent the contract from being enforced without the invalid provision: see Alec Lobb Q (Garages) Ltd. v. Total Oil (Great Britain) Ltd. [1985] 1 WLR 173, 181, per Dillon L.J. ……

The true explanation of the Amoco case, in my opinion, is that given by Dillon L.J. earlier, on p. 180, that is to say that in the Amoco case the invalid tie was the sole object or subject matter of the contract, which would have made no commercial sense without it. The test has been variously described in the cases. The contract will be upheld unless the invalid restraint forms "the real consideration" or "the main consideration" or "the whole or substantially the whole consideration" for the promise: for the last formulation see Bennett v. Bennett [1952] 1 K.B. 249, 261, per Denning L.J. I doubt that there is any real difference between these different formulations, so long as it is recognised (i) that the avoidance of the contract is not limited to the case where the only consideration for the promise, apart from the invalid restraint, is a technical or nominal consideration; (ii) that the court does not attempt to assess the relative values of the various considerations for the promise. But for my own part I prefer the formulation expressed by Denning L.J., since it appears to me to put the point of balance in the right place. The contract will be upheld even if the consideration for the promise of the promisee includes an invalid restraint. It will be struck down in its entirety only if, in substance, and regardless of its form, it is an agreement for an invalid restraint."

121. In my judgment, the non-pensionability term is to be treated in much the same way as the courts have treated invalid covenants in restraint of trade. In the case of a covenant in restraint of trade its validity is assessed against the admissible background and, if found to be an excessive restraint, it will be unlawful and unenforceable. Whether the remainder of the contract can be relied on or whether the whole contract has to be avoided is then to be assessed in accordance with the principles just explained. Similarly, in the present case, the validity of the non-pensionability provision must be assessed against the whole background. I have held that the 2009 NPAs, and thus the ensuing contractual variations to the contracts of employment, gave rise to a breach of Holdings' duty. In my judgment, the non-pensionability term is therefore invalid.
122. The question then is whether, applying the approach in Marshall, the variation to the contracts of employment can stand without the non-pensionability term. In my judgment, the non-pensionability terms are not "the whole or substantially the whole consideration" for the variation in the contract of employment. The main consideration is, in my view, the provision of the labour and services for which the members are being paid. Accordingly, the salary increases took effect; Holdings cannot reclaim past payments or decline to make future payments in response to claims by members that they are entitled to have the salary increases treated as pensionable.
123. Holdings says that it would not have offered the salary increases in 2009 onwards other than on terms as to their non-pensionability. That may be true but it does not show that the real consideration was other than I have just identified. The fact that the members had earned and deserved their salary increases was acknowledged in the Compensation Summary Statements, dating from autumn 2009, received by members. They include the following paragraph:

"IBM aims to reward employees according to the value of their jobs, skills and contribution to IBM and in the marketplace. Annually managers review each employee's compensation and make decisions based on comparisons to the market, individual performance and available funding. This statement shows the results of this review by your manager. If you have questions regarding this statement, please discuss with your manager."

124. Further, Holdings had in fact already planned salary increases for 2009. Thus one sees Mr Ferrar in the email leading to the Later 2009 NPAs stating that unless non-pensionability terms were accepted, members "will not receive any salary increases planned for 2009". Other documents show clearly that increases were planned and determined before the members were asked to sign NPAs. However, as Mr Newman points out, the increases were planned only after Holdings had decided to proceed with Project Waltz and thus only in the context that they would not be pensionable. That may be so, but overall, it is clear that Holdings regarded the members as deserving of some additional reward for their services. But rather than give an option for a member to take some other reward (eg a smaller, pensionable, salary increase of the same value (actuarially assessed) as the proposed non-pensionable salary increase), Holdings simply gave no alternative option (other than the option of no salary increase at all).
125. I conclude, therefore, that the non-pensionability term is not enforceable against members and they are entitled to retain salary increases, both for the past and the future.
126. There is another reason for reaching the same conclusion. I have said before that the inclusion of the non-pensionability term in the varied contracts of employment gave rise to a breach of duty just as much as the NPAs themselves. It does not eliminate that breach of duty (although it may have an impact on the remedy) for Holdings to say that, but for the term, it would not have provided the salary increase in the first place. And so Mr Spink submits that for Holdings to seek to enforce that term now or in the future (whether by attempting to restrain the Trustee from paying the full benefit or by asserting a claim to damages) would give rise to a further breach of the duty of trust and confidence.
127. As I see it, the result of the contrary view (ie that there would no further breach of duty) is unacceptable and so I accept Mr Spink's submission. The contrary view would mean that, in spite of Holdings having procured the NPAs and contractual non-pensionability term in breach of duty, the members would in fact be bound by it unless, of course, the whole salary increase contract was avoided. But in that case, the members would find themselves, 5 years after the Initial 2009 NPAs, with no salary increases since 2009 and the possible obligation to pay back the increases which they have already received. That would not be to put the members in the position in which they would have been if the 2009 NPAs had never been imposed in the first place. If they had not been imposed, it is quite possible (even if Mr Newman is right to say that there was no obligation on Holdings to do so) that members would at least have been considered for salary increases and, in practice, Holdings may have thought it right not to discriminate between DC and DB members; albeit that the salary increases might have been smaller, there would have been a real prospect that DB members would obtain some increase even absent a non-pensionability term. It is manifestly not Holdings' position that members should, retrospectively, be considered for salary increases if the increases actually awarded are abrogated. Quite the reverse: Holdings' position is that a further consultation in relation to Project Waltz is not necessary because the result would be a foregone conclusion: the result which Holdings considers to be a foregone conclusion is that the DB members should have salary increases from 2009 only if they are subject to a non-pensionability term (and perhaps not even then). Perhaps more importantly, the members will have worked for the additional 5 years for a reward (old salary and therefore no pensionable increases) less than that which even Holdings considered appropriate (increased salary but increase not pensionable). In these circumstances, I agree with Mr Spink's submission that an attempt by Holdings to enforce the non-pensionability term would give rise to a breach of duty which the members are entitled to restrain. I anticipate – at least I hope that I can anticipate – that were I to make a declaration to that effect, Holdings would not, in fact, seek to enforce the non-pensionabilty term. If Holdings is to escape from the conclusion that to enforce the non-pensionability term would give rise to a further breach of duty, it needs, at the least, to provide the members with something equivalent to that which they might have expected to receive had Holdings acted properly in 2009, by which I do not mean the minimum possible applying Lavarack principles.
128. There is a related issue concerning the validity of the 2006 Partial NPAs. These are the elections made by individual members following Project Soto to remain in the DB Plans on terms as to partial non-pensionability. The members were presented with two choices. By way of example, C Plan members were required to accept one or the other of the following options:

"*I want to remain in the C Plan for future service benefits.

I understand that:

I will not have another opportunity to join the Enhanced M Plan.

The Company's compensation policy (which will not be amended in this respect without prior consultation) is to award future salary increases as a base salary increase which is pensionable plus, for employees who agree, a non-pensionable salary supplement equal to 50% of the base salary increase. I agree to receive such non-pensionable supplements as part of future salary increases including any which take effect from 6 June 2006.

*I wish to join the Enhanced M Plan for future service benefits from 6 July 2006. I understand that:

[details of the Enhanced M Plan were then set out]."

129. Members were told that anyone who did not make an election would remain in their current DB pension plan and would not receive any non-pensionable salary supplement, in other words, they would receive only 2/3rds of the proposed salary increase.
130. In addition, by Deed of Amendment dated 30 March 2006, a proviso was added to the definition of Pensionable Earnings in the DB Rules of the Main Plan as follows:

"(vi) in the case of a Member who gives express consent to his Employer in writing that any future increase in his gross basic remuneration or monetary remuneration (as the case may be) will include a non-pensionable supplement equal to one-half any future increase in his pensionable gross basic remuneration or pensionable monetary remuneration (as the case may be), such non pensionable supplement shall be treated as falling within the above Exclusions for the purposes of calculating his Pensionable Earnings by reference to the Defined Benefit Rules'

131. No equivalent amendment was made to the I Plan Rules.
132. According to the Trustee's case, Holdings abandoned this compensation policy as part of Project Waltz. In particular, it ceased to award a pensionable base salary increase and a non-pensionable salary supplement. Instead, the policy became one of awarding a salary increase that was entirely non-pensionable by reference to the 2009 NPAs. The Trustee's case, therefore, is that no pension supplements were awarded so that there was nothing after the 2009 NPAs for the 2006 Partial NPAs to operate on. The actual increases are, it is said, therefore wholly pensionable and are not prevented from being wholly non-pensionable by the 2006 Partial NPAs.
133. This is obviously a very unattractive and opportunistic argument. I must, however, deal with it on its legal merits, such as they are.
134. The Trustee's case is that the 2006 Partial NPAs did not comprise an agreement (still less a contract that was binding from the moment that the member made the election) by the member that all future pay increases would be treated as including a non-pensionable supplement equal to 50% of the base salary increase, even if the amendment to the DB Rules anticipated that agreements would take such a form.
135. At most, they comprised an agreement by the member to receive non-pensionable supplements as part of future salary increases for so long as it remained Holdings' compensation policy to award salary increases in that fashion. The Trustee does not suggest that they were not valid and effective for so long as Holdings maintained that policy and awarded pensionable base salary increases together with non-pensionable supplements.
136. However, having abandoned that policy (albeit in favour of a new policy that was flawed) Holdings ceased to make awards of 50% non-pensionable supplements. Holdings has accepted in correspondence that there are no documents from 2009 – 2011 which make reference to supplements. As a result, there have since been no such non-pensionable supplements in relation to which the 2006 Partial NPAs can take effect.
137. One thing which is clear is that, prior to the Initial 2009 NPAs, members who had entered into the (valid) 2006 Partial NPAs had no expectation, reasonable or otherwise, that salary increases would be wholly pensionable. They knew that the total increase was divided into two components: a pensionable base component and a non-pensionable supplement. But this was a notional division. Members were to receive a single salary increase albeit comprising those two components: they did not receive two separate increases. Thus members agreed "to receive such non-pensionable supplements as part of future salary increases" (my emphasis).
138. The effect of the Initial 2009 NPA for those who accepted it would, if it had been valid, have been to eliminate the relevance of, and need for, the notional division into two components. It does not follow from that proposition that, even when the inclusion of the non-pensionability term in the varied contracts of employment is found to give rise to a breach of duty, the 2006 Partial NPAs or the amendment to the DB Rules in the Main Plan are not relevant. In my judgment, they remain of relevance and have continuing effect. They are to be interpreted, adopting a purposive construction, so that the actual salary increase is to be notionally divided into a basic component and a supplement. This is one natural meaning (indeed, in my view the more natural meaning) of the words used. Thus the amendment refers to a member agreeing that any salary increase will include the supplement: accordingly, if an increase in salary of £3X is granted without any more being said, that is to be treated as including a non-pensionable supplement of £X with only the increase of £2X being pensionable. This is the effect of the actual rule in the case of the DB Rules of the Main Plan; and it is the effect of the 2006 Partial NPA in the case of the I Plan where, as I have already concluded, that NPA continues to apply even in the case of members sending protest emails.
139. To put that last point another way, my analysis under Issue 3, of why Holdings cannot rely on the non-pensionability term demonstrates that that term is to be severed from the rest of the varied contract of employment, leaving the salary increase intact. However, if the non-pensionability term is severed in this way and is held to be invalid, the question needs to be asked: What are the terms on which any increase is awarded? Quite clearly, it is awarded on the same terms as the pre-existing salary, for instance as to timing and frequency of payment. I see no reason why one of the terms on which it is awarded should not include the terms on which it would have been awarded but for the invalid non-pensionability term. Indeed, one might expect that it would be awarded on the same terms as previous salary increases so that there would be a "pool" of salary increases divided into pensionable and non-pensionable components. I reach this conclusion as a matter of interpretation of the contractual arrangements between the parties.
140. Yet another way of putting essentially the same point is to focus on the impact of the breach of duty which is, in accordance with my conclusions under Issue 3, that the non-pensionability term, once severed from the rest of the contract of employment, was invalid. If a term is invalid, that means that it is ineffective to alter the pre-existing contract. If the non-pensionability term had not been included in the variation of the contracts of employment in the first place, that term would clearly not displace the 2006 Partial NPAs. The 2009 NPAs themselves do not displace the 2006 Partial NPAs: if the provisions of the 2009 NPAs do not become valid terms of the varied contracts of employment, then the offer which Holdings makes, and which is accepted by the members accepting the salary increase and continuing to serve, must by necessary implication include the pre-existing terms on which salary increase offers were agreed to be made, that is to say the 2006 Partial NPAs.
141. My conclusion is that the 2006 Partial NPAs continue to apply to salary increases awarded following the 2009 NPAs so that any salary increase is divided into a pensionable element and a non-pensionable supplement (in the ratio 2:1).
142. The answers to Issue 3 and its sub-issues (reformulated in the light of my approach to the NPAs) are:

i) there is no non-pensionability term to be found in the varied contracts of employment arising out of the Initial 2009 NPAs and the Later 2009 NPAs in the case of affected members who sent the protesting emails referred to in paragraph 68 above;

ii) the non-pensionability term in the varied contracts of employment arising out of the Initial 2009 NPAs and the Later 2009 NPAs cannot be relied on by Holdings in the case of other affected members (and indeed in the case of the class within i) above if I am wrong in my conclusion in relation to them);

iii) the members are entitled to keep all of the salary increases which have been awarded;

iv) the members are entitled in the future to continue to be paid salary incorporating those increases; and

v) such salary increases are to be treated as pensionable for the purposes of the DB Plans in accordance with the 2006 Partial NPAs.

Issue 4: are members in principle entitled to damages or equitable compensation from Holdings for having purportedly entered into the 2009 NPAs or to some other remedy and if so what remedy?

143. I agree with the Trustee's position that damages are, in principle, available. Holdings breaches, in relation to the 2009 NPAs and consequential variations of the contracts of employment, were breaches of the contractual duty of trust and confidence implied in those contracts. There is a right to damages on ordinary contractual principles: see Malik v BCCI [1998] AC 20. Issue 4 (which raises a broad general issue) does not raise the many difficult issues which would arise in relation to the correct measure of damages in various different cases eg members electing to enter in to NPAs, members failing to elect at all, members leaving service as a direct result of the options improperly offered to them. I do not drill down into those dark wells.
144. I do not accept Mr Newman's case to the contrary. He argues that members who entered into the 2009 NPAs are not in fact entitled to damages (or any other compensation) because they cannot demonstrate that the breach caused them loss. This is because employees had no right in advance of any particular salary increase offer to be awarded or to be considered for a salary increase. Thus, consistently with the Lavarack principle (see Lavarack v Woods of Colchester Ltd [1967] 1 QB 278), damages must be nil as Holdings could have declined to give any salary increases at all. If Mr Newman is right, then Holdings' case overall (much of which I have already rejected in relation to Issues 1 and 2) is that its breach of duty has absolutely no consequence at all; the members have no remedy. On Holdings' case, not only is the non-pensionability term valid and enforceable (a proposition I have rejected) but also the breach of duty in imposing that term carries no sanction.
145. I find that submission almost as unattractive as the Trustee's submission concerning the 2006 Partial NPAs; but, as with that, I will address it on its legal merits.
146. Lavarack is authority for the proposition that damages for wrongful dismissal cannot confer on an employee extra benefits which the contract did not oblige the employer to confer even though the employee might reasonably have expected his employer to confer them on him in due course. But in cases where the employer is exercising a discretion, the position is different. Thus in cases of bonus awards, the court, when assessing damages for a wrongful failure to award any bonus, seeks to evaluate what the employer itself would have been likely to do. It is not to be assumed that the discretion would have been exercised so as to give the least possible benefit to a claimant if such an assumption would be unrealistic on the facts: see Rutherford v Seymour Pierce Ltd [2010] EWHC 375 (QB), [2010] IRLR 606 at [33] and Horkulak v Cantor Fitzgerald International [2004] EWCA Civ 1287, [2004] IRLR 942 at [48] and [64] to [72]. At [37] of Horkulak, Potter LJ, giving the judgment of the Court, cited with approval from the decision of Timothy Walker J in Clarke v BET plc [1997] IRLR 348. In that case, the relevant service agreement provided that "the executive's salary shall be reviewed annually and be increased by such amount if any as the Board shall in its absolute discretion decide". It also provided "In making their decision the board shall consider a comparative group of companies ...". I set out part of the same passage:

"From the wording of the contract, in particular from the fact that the word "shall" plainly governs "be increased", it seems to me that there was here a contractual obligation on BET to provide, and a contractual right on Mr Clark to receive, an annual upward adjustment in salary. It is only the amount (if any) that is in the absolute discretion of the Board. In assessing the amount there is an obligation to consider figures from a comparative group of companies.

Accepting the principle that a defendant in an action for breach of contract "is not liable for not doing that which he is not bound to do" per Diplock LJ in Lavarack v Woods of Colchester [1967] 1 QB 278, 294B–C, if the board had capriciously or in bad faith exercised its discretion so as to determine the increase at nil and therefore to pay Mr Clark no increase at all, that would have been a breach of contract. This is a case where BET have repudiated their obligation altogether and the question I have to decide on the facts is what position Mr Clark would have been in had BET performed this obligation: see the judgment of Banks LJ in Abrahams v Herbert Reiach Ltd [1922] 1 KB 477, the passage at p.480 with which Scrutton LJ agreed at p.481. Nor should I assume that any discretion would have been exercised so as to give the least possible benefit to the plaintiff if such an assumption would on the facts be unrealistic."

147. I disagree with Mr Newman's proposition that the members had no right even to be considered for a salary increase. I doubt that it is often right that an employer can say that it need never consider granting salary increases. But even if that is wrong as a universal proposition, it is clear, in my view, that Holdings did have a duty to consider salary increases for DB members. This is because it did actually consider salary increases for DC members and actually awarded increases. Although Holdings might have been able to justify awarding no increases to DB members had it acted wholly properly and not in breach of duty in the context of Project Waltz, that is very far from clear. At the very least, it would have had to consider whether or not to do so and a refusal to do so could be seen as giving rise to a breach of the contractual duty of trust and confidence and perhaps even as irrational and perverse.
148. Given that Holdings would have had to consider granting salary increases, the case would fall, in my judgment, to be approached in the same way as the judges approached the question in Clarke and Horkulak. Accordingly, in assessing damages the court would have to evaluate what Holdings would have been likely to do. I do not, therefore, consider that Mr Newman is right to say that members who entered into the 2009 NPAs cannot as a matter of principle obtain damages. Subject to proof of loss in individual cases and the application of the normal rules of causation and remoteness, I consider that, in principle, damages are recoverable.
149. Mr Newman has another argument which would qualify that conclusion. He submits that, in respect of the case against Holdings based on the disappointment of Reasonable Expectations, no remedy should be given going beyond the extent of the Reasonable Expectation. He makes four points going to the calculation of damages, which I take in turn.
150. The first is that there can have been no Reasonable Expectation in 2009 that all future salary increases would be pensionable. This is because the 2006 Partial NPAs made clear that 1/3rd of those salary increases would not be pensionable. Further, the Main Plan DB Section was amended following Project Soto to incorporate this limit. The Trustee does not dispute this. It is to be taken account of in assessing the measure of damages.
151. Mr Newman's second point is that, in addition, by analogy with the position regarding future benefit accrual, there can have been no Reasonable Expectation that the policy of Holdings set out in the 2006 Partial NPAs would never change. Accordingly, there must be a point of time at which Holdings could properly adopt a policy of not awarding pensionable salary increases. Again, the Trustee does not dispute the general proposition but observes that, among other matters (i) there would be a need for consultation and (ii) such a proposal might give rise to issues of unlawful age discrimination.
152. Mr Newman's third point is that members who joined the DB Plans after the conclusion of Projects Ocean and Soto cannot have had Reasonable Expectations based on those projects. This is relevant only to the I Plan since the C and N Plans were closed to new members at the material times. In my view, the conclusion does not follow from the fact that these members joined the DB Plans after the projects had been implemented. It all depends on what those new members were told about the projects and whether what they were told could reasonably be taken as representations to them. It is not a matter in which I have ruled in the Judgment. It is not a matter which I can take any further in this judgment: the answer will all depend on the facts applicable to the individual concerned.
153. Mr Newman's fourth point is that the members cannot get more by way of damages than they could be entitled to under the Rules of the Plans. It is evident from Judgment [289(iii)] that the effect of the implied limitation on the Exclusion Power preserving the final salary link is to protect benefits accrued up to the date of the service of the Exclusion Notices (ie 28 May 2010). Accordingly, accrual between 28 May 2010 and 6 April 2011 would not attract the benefit of the implied limitation being referred to, and a member cannot through an attack on the NPAs alter this position (assuming the Exclusion Notices to be valid). As will be seen later, my view is that the Exclusion Notices are voidable, so this point falls away in relation to members who elect to avoid the Exclusion Notices vis à vis themselves. It is, in any case, based on a rather literal reading of what I wrote. I think the meaning of what I wrote is clear, namely that the implied limitation has effect from when the Exclusion Notice takes effect, not from when it is physically served. If it was not clear, let me make clear now by way of decision that that is when the implied limitation has effect from.
154. One further issue which might arise is whether a member who asserts the invalidity of the non-pensionabilty term is also able to obtain damages as a cumulative remedy. In this context, the Trustee does not contend that damages are payable in circumstances where the 2009 NPAs and the non-pensionability terms in the contracts of employment as varied are void, unenforceable or set aside. I assume that this will be its position in relation to the status of the non-pensionability term in the light of my conclusions under Issue 1 to 3. However, the Trustee's position is subject to two caveats where:

i) a member may have suffered consequential losses (such as adverse tax consequences); and

ii) a member has left employment in response to the breach of the contractual duty and sustained loss as a result.

155. As to the first of those, I doubt very much that on the facts any member would be able to show such a loss. Consequential losses can only be viewed as part of the losses arising as a result of the breach. It is necessary to compare the position of a member if there had been no breach of duty with his actual position in the context of the breach of duty. In purely financial terms, the invalidity of the non-pensionability term will almost certainly over-compensate the member since it is unlikely that Holdings would, acting properly and in the absence of the NPAs, have awarded members collectively as high a salary increase as was in fact awarded to those who had agreed to the NPAs. Such over-compensation would, in my view, have to be set off against the consequential losses claimed.
156. As to the second of those, I foresee considerable difficulties of causation and remoteness in the way of a member claiming such loss.
157. Those are only indications of my view. I decline to make any ruling on those two points. But my conclusion is that, in principle, damages can be claimed in addition even though the non-pensionability term cannot be enforced by Holdings.
158. There are two additional points to make under Issue 4. The first is that, where pensions have already come into payment, they must be adjusted, with retrospective effect, to take account of any salary increases which have, up to now, been excluded as a result of the non-pensionability term.
159. The second point is that it is Holdings' case that a member who joined the I Plan after the conclusion of Ocean and Soto cannot have had Reasonable Expectations based on those projects. This is not an issue which arose at the trial. It may be the case that some, or possibly all, of the members in this class did not in fact hold any relevant Reasonable Expectations. But that cannot be determined without a factual enquiry. It all depends on what such members were told and by whom. Accordingly, I make no determinations about this class.
160. The answer to Issue 4 is that members are entitled, in principle, to damages from Holdings for having entered into the 2009 NPAs.

Issue 5: what is the measure of damages (in relation to the 2009 NPAs)

161. Holdings' position is that I should not say anything about the measure of damages at this stage. The ordinary principles of damages for breach of contract apply and it would be speculation for me to say anything about the measure of damages that should apply in potentially numerous, and at this stage hypothetical, separate proceedings each of which would, if brought, turn on its own facts. There is some sense in that but, on the other hand, I anticipate – I certainly hope – that it would help with the ultimate resolution of the disputes in this case, so as to give closure to a number of members who are not so young, if I were to deal with at least some of the points.
162. A preliminary observation is this. Although Issues 4 and 5 are focused on the 2009 NPAs (and under my analysis the ensuing salary increase contracts), it is Holdings' overall conduct in relation to Project Waltz which gives rise to the breach of duty. The purpose of any award of damages is to compensate the members for that breach. It is true that I have held that the Initial 2009 NPAs (and also the 2011 NPAs) give rise to a breach of duty quite apart from Project Waltz, but they too form part of Project Waltz and any damages remedy must reflect that.
163. As is trite law, the general principle behind an award of compensatory damages for breach of contract is to put the claimant in the position in which he would have been but for the breach. For an example of the application of this principle in the case of a breach of the implied duty of trust and confidence, see Transco plc (formerly BG Plc) v O'Brien [2002] ICR 721 at [22].
164. I have given some attention to Lavarack, on which Mr Newman relies, in my discussion under Issue 4, where I mentioned Horkulak and Clarke v BET on which Mr Spink relies. To those authorities I add Durham Tees Valley Airport Ltd v bmibaby [2010] EWCA Civ 485, [2011] 1 All ER (Comm.) 731 where Patten LJ commented that:

"the court is not required to make assumptions that the defaulting party would have acted uncommercially merely in order to spite the claimant. To that extent, the parties are to be assumed to have acted in good faith although with their own commercial interests very much in mind."

That was not said, of course, in the context of an employment case, but it states a clear and wise approach.

165. I considered in my discussion of Issue 4 what Holdings might have done had it acted properly, concluding that it would at least have had to consider whether or not to grant salary increases to DB members. I also concluded that in assessing damages the Court would have to evaluate what Holdings would have been likely to do. In Horkulak, Potter LJ said this:

"The broad principle that a defendant in an action for breach of contract is not liable for doing that which he is not bound to do will not be applicable willy-nilly in a case where the employer is contractually obliged to exercise his discretion rationally and in good faith in awarding or withholding a benefit provided for under the contract of employment."

166. So too, in the present case, I do not consider that the broad principle is applicable willy-nilly where Holdings is contractually obliged to consider whether DB members should be awarded salary increases in a context where DC members carrying out the same or equivalent work are being awarded salary increases. Indeed, I do not think that it should be applied willy-nilly even if Holdings was not, contrary to my view, under any duty even to consider whether DB members should be awarded salary increases. It may well, in either case, be that the requirement to act rationally and in good faith would preclude Holdings from awarding no salary increase at all. The court, in assessing damages, must then evaluate what would have been likely to happen: and in carrying out such an assessment, the court is not constrained to determine the very lowest figure which Holdings could have awarded without giving rise to any breach of any duty. However, the court must consider all of the reasonable courses which Holdings might have taken (or at least ones which Holdings presents to the court on an enquiry as to damages).
167. In that context, Mr Newman submits that Holdings could, away from the context of Project Waltz, have offered a particular salary increase on terms that it would not be pensionable; such an (isolated) offer would not have given rise to any breach of duty. I agree with that submission as far as it goes; indeed, it is entirely consistent with my conclusion that the Later 2009 NPAs would not have given rise to a breach of duty under Scenario B. What is said to follow from that is that this possibility should form the basis of the assessment of damages: there would, in fact, be no damage because this is precisely what those who accepted the Later 2009 NPAs actually got.
168. It may well be that, viewed in isolation (ie Scenario B) in this way there would be no breach of duty; this possibility must therefore be considered as a potential outcome. But it is only a possibility. It does not provide the answer to the evaluation which the court must carry out following the Horkulak approach.
169. Clearly, that evaluation is not to be carried out as if the only breach of contract flowed from the 2009 NPAs and the consequential salary increase contracts. The rest of Project Waltz and its underlying purpose would still fall to be regarded as part of the context in which the hypothetical offer of a salary increase is to be assessed: the offer would, in effect, be part of Project Waltz designed to disappoint the members' Reasonable Expectation concerning continued benefit accrual and would be as objectionable as the actual NPAs and non-pensionability terms themselves.
170. Instead, the evaluation, has to be carried out as part of a wider evaluation namely what would Holdings have been likely to do between 2009 and the Longstop Date (whatever that may be) so far as concerns salary increases in a way which did not breach its duty of trust and confidence. In other words, the question is: what could or would Holdings have done if Project Waltz had never seen the light of day and if Holdings had not had an underlying purpose which conflicted with the Reasonable Expectations established?
171. Those questions raise a number of other questions, the ones which immediately spring to mind being precisely why Holdings would actually require salary increases to be non-pensionable (and whether an offer along the lines of the hypothetical offer would be a likely possibility at all) unless it was the first step in an (objectionable) proposed course of conduct to do precisely what Project Waltz was designed to achieve. This judgment is not the place to begin to answer those questions. The fact that they arise demonstrates to me that Mr Newman cannot be right to suggest that the measure of damages is to be assessed applying his approach to Lavarack and on the basis that Holdings would have been able properly to make the hypothetical offer. I would not, even on his approach, in such an evaluation attach much, if any, weight to what Holdings now says it would have done (rather than could have done). It may not be admissible evidence. But even if it is, I venture to suggest that it is almost impossible for Holdings to form a sensible conclusion about what it would have done had there been no Project Waltz and were it to take account only of matters which it could properly have taken into account in making decisions about employee remuneration consistently with all Reasonable Expectations. Holdings has shown its colours already by saying that a further consultation is not necessary because its decision, even following a consultation, is a foregone conclusion, namely to implement Project Waltz or something as near to it as can practically be achieved.
172. Further (although this may be no more than to put the same point in a different way), we are concerned with damages for Holdings' breach of the implied duty of trust and confidence, a breach stretching over a period of time and involving several components. The search is not for the loss flowing from the NPAs and the non-pensionability term in the salary increase contracts by themselves, but is rather for the loss flowing from Project Waltz as a whole, of which the NPAs and the non-pensionability term form only a part, albeit an important part. If one is to apply the general principle of damages (ie to put the claimant in the position in which he would have been but for the breach), members are to be put into the position which they would have been if there had been no Project Waltz at all. It may well be that Holdings could justifiably say that damages should be assessed on the basis that there would be no pensionable salary increases after the Longstop Date (whatever that date may be). But before that date, what Holdings would have been likely to have done in terms of salary increases is to be assessed on the footing that Holdings would not have suggested a course of conduct designed, in practice, to circumvent the Reasonable Expectation.
173. Mr Spink also referred me to the decision of the House of Lords in Rigby v Ferodo Ltd [1988] ICR 29 which he contends supports the conclusion that Lavarack principles do not come into play. In that case, the defendant was found to have acted in breach of contract in reducing the plaintiff's wage; the contract of employment had not been terminated nor had it been varied by acceptance of the reduced wage. On the question of damages, the defendant placed reliance on the "minimum obligation rule". The argument was that, since the defendant could lawfully have terminated the plaintiff's employment by 12 weeks' notice, it must be treated as if it had done so with the result that the claim to damages must be limited to the shortfall in wages during that period. That argument was rejected. Lord Oliver said this:

"….the argument advanced is, in my judgment, based on a fallacy. It assumes the very proposition which has already been rejected, namely, that the employment under the contract of service has come to an end. If it had, then no doubt there would be room for an inquiry at what date the employer could first lawfully have terminated it. But that is not the case with which this appeal is concerned. What your Lordships are concerned with here is a claim for sums due under a continuing contract which never was terminated, either lawfully or unlawfully, and there simply is no room in such a case for the application of the principle referred to."

174. In relation to the members who did agree to the NPAs, Holdings did, in fact, award salary increases and did so as a reward for the members' services. On the basis of what Lord Oliver said in Rigby, it is submitted that there is no room for the Lavarack approach (or, I suppose, the Horkulak approach either) because Holdings' argument is also based on a fallacy: it assumes the very proposition which has been rejected namely that no salary increase was awarded.
175. I cannot accept that submission. In Rigby, the contract clearly required the full wage to be paid; it was earned for the actual period for which the claim was made because the plaintiff was actually employed during that period. The defendant could not limit its liability by reference to a counterfactual situation which it might have brought about and which would have prevented its contractual obligation arising. In the present case, the breach of contract on which the damages claim is based is not a failure to grant the actual salary increase in a way which was pensionable: the matter could not be put that way because there never was such an obligation. Rather, the breach is the procuring of the NPAs and the non-pensionability term: the damages flowing are to be assessed as if that breach had not occurred. It is therefore necessary to evaluate what Holdings would have done since it cannot be said, without further evaluation, that it would have awarded the salary increases which it in fact awarded.
176. In reality, this whole issue is likely to be of little relevance to those members who agreed to the NPAs. Assuming I am right in my conclusions on Issues 1 to 3, such members will be able to claim their salary increases as pensionable. If such a member decides, instead, to stick with the non-pensionability of the salary increases and to sue for damages, then (if he can assert such a claim at all which is not clear given his election) there can be no question of claiming damages to compensate him for the salary increase not being pensionable.
177. Mr Spink also presented an argument based on age discrimination in saying that Holdings could not have adopted the course of refusing pay increases to all members of the DB schemes who did not agree to the NPAs. The argument here is that DC members were, predominantly, younger than DB members. Giving pay increases to DC members but not (in the case of the hypothetical offer) to DB members rejecting the non-pensionability term would amount to a prohibited disadvantage. I am not prepared to address that issue. It raises a number of questions which, in my view, cannot possibly be dealt with on the basis of the factual findings which I have already made or which I could make on the basis of the evidence which I received. This discrimination issue will have to be resolved on a case by case basis if and when claims are brought by members for damages.
178. If I am wrong on all or any of Issues 1 to 3, then there is no right to set aside the non-pensionability term and the members will be bound by it. If it is binding, there can be no claim for damages to compensate for the fact that the salary increase is not pensionable. The contractual claim will be for breach of duty in the imposition of the NPAs and the non-pensonability term; and here the measure of damage is ascertained by carrying out a comparison of the members' positions as matters stand and as they would have stood but for the breach of duty.

Exemplary damages

179. The Trustee contends that Holdings' behaviour in relation to Project Waltz and the breach of duty which it involved was so serious as to warrant an award of exemplary damages. I am not willing to deal with this issue in the context of a Part 8 claim seeking declaratory relief. Although wishing to be as helpful as possible in bringing finality to the disputes, there is a point beyond which it is inappropriate to go. I do not consider that it is appropriate that I should, in the face of opposition from Holdings, determine on a group basis a hypothetical issue, hypothetical because it may be that no claim for damages is ever brought. Mr Simmonds made some powerful points at the beginning of his oral submissions. I have already gone considerably further (and will do so under other issues too) than he considers I should go. I obviously take a different view, but I do agree with him that the issue of exemplary damages is one which should be dealt with only in the context of litigation where they are actually claimed.
180. My reluctance to deal with the issue is made all the greater by these three considerations. The first, which I hope is correct and uncontentious, is that exemplary damages have never been awarded in relation to a breach of contract rather than in respect of tortious liability. The second, which may be debateable, is that exemplary damages are anomalous in that they are punitive and punishment has no part to play in the adjudication of the rights and obligations of parties in civil litigation. The third, which is only a tangential consideration, is that this judgment is already very long and overlong in its preparation. A judgment on the issue of exemplary damages would, of itself, be a significant undertaking. It is more important that I produce this judgment than that I deal with exemplary damages. However, that is not necessarily an end of the point: see paragraphs 185 and 186 below under the heading "A supplemental note on exemplary damages/compensation".

Exemplary compensation in equity

181. Mr Evans has addressed me on the proposition that an exemplary award of equitable compensation can be made. As in the case of exemplary damages, and for similar reasons, I do not propose to address this issue. It would be quite wrong to attempt to decide this issue, not least because equitable compensation for breach of the Imperial duty will, in terms of the measure of compensation, be influenced by the correct approach to damages for breach of the implied duty of trust and confidence. If exemplary damages are not available, that is not a good starting point for those asserting that there can, nonetheless, be an exemplary award of equitable compensation and it is certainly not an issue which should be decided unless the issue of exemplary damages is decided. Further, those asserting the power of the court to award such a measure of financial recompense will have to circumvent the oxymoronic concept of compensation which punishes rather compensates. It is not simply a matter of using a different phrase – for instance equitable damages – since the description "equitable compensation" reflects the jurisprudential basis on which a financial award has, in the past, been made.
182. At the end of the discussion above, it is not possible (in contrast with other Issues) to give a short answer to Issue 5 (save the non-answer in relation to exemplary damages). My conclusions appear from that discussion.

Issue 6: are the answers to Issues 1 to 5 any different for those members who purportedly entered into 2009 NPAs having sent emails of protest substantially in the form of the email referred to at paragraph 68 above or other protests to similar effect?

183. I have dealt with the emails at paragraphs 69 - 71 above. Save as otherwise indicated in that discussion, I do not think that the answers to Issues 1 to 5 are different for those members.

Issue 7: to what, if any, remedy are the members who did not enter into the 2009 NPAs entitled?

184. This issue is addressed principally to members who remained in employment. For those members, it seems to me that the breach of Holdings' duty is to be found in the Project Waltz proposals overall. Members in this category were as much the victims of Holdings' breach of duty as those who elected to enter into the NPAs, each group being faced with an unpalatable choice but with different groups taking different views.
185. I do not consider that this group of members can claim (pensionable) salary increases equal to the (purportedly non-pensionable) salary increases awarded the group entering into the 2009 NPAs. Indeed, I do not see how the court can have any jurisdiction to compel Holdings to award any salary increase. All that the court can do is award damages for breach of contract. I have discussed, under Issue 5, the correct approach to the measure of damages. The same approach applies to this group. Accordingly, Holdings is not able to rely on Lavarack to launch an argument that no damage has been suffered because Holdings was not obliged to grant any salary increase at all. Rather, damages are to be assessed on the basis of an evaluation by the court of what the likely outcome would have been if Project Waltz had not been formulated and implemented. I agree with the way that it is put in the Trustee's closing written submissions:

"….members who did not enter into the 2009 NPAs are entitled to damages to reflect the salary they would have received in the ordinary course of events, but for Project Waltz, and the loss of pension and other rights as a result of those salary increases not having been granted."

186. Having said that, the most obvious way to put the members into the position in which they would have been is to provide them with actual salary increases (with retrospective effect) equal to those which they would have received (although it is, I suppose, unrealistic to think that Holdings and the members would ever be able to agree a figure) with concomitant pension accrual.
187. As to those members who left employment, the Trustee's case is that those who did so in response to a breach of the contractual duty are entitled to damages, flowing from that dismissal on ordinary contractual principles. Quite what "in response to" means, I am not sure. And to use the word "dismissal" is to beg a question: a member would be able to obtain damages if Holdings' breach of contract amounted to a repudiatory breach justifying the member in treating the contract of employment at an end. In the present case, that requirement is, in my judgment, satisfied and the begged question answered accordingly. Even so, the member would have to establish that he had suffered loss. There would be the usual issues concerning the taking of alternative employment; and Holdings would be able to argue, in any particular case, that the member concerned had decided to leave service in any case and that the breach of contract was not therefore causative of any loss. Further, since Holdings could have terminated service by giving an appropriate period of notice, it may be that Lavarack principles apply with damages in relation to lost earnings being limited by that consideration, although the pension rights attached to termination without cause would need to be taken into account.
188. The Trustee submits that any damages awarded to any members would need to be grossed up to take account of tax. Holdings takes the exactly opposite view, saying they would have to be "grossed down" to use Mr Newman's apt words. I do not propose to resolve that point other than to say that, in principle, the amount of a damages award subject to whatever tax that award has to bear, should equate to the amount which the member would have received taking into account whatever tax those amounts would have borne or would (so far as concerns future payments from the DB Schemes) in the future have to bear.

A supplemental note on exemplary damages/compensation

189. Although I have declined to embark on a general consideration of exemplary damages and equitable compensation, I will, if all parties ask me to do so, consider revisiting that refusal to a limited extent. I may be willing, if all parties are in agreement, to address all or any of the following issues:

i) whether as a matter of law, exemplary damages are available for breach of the implied contractual duty of trust and confidence and, if they are, the criteria for determining whether such exemplary damages should be awarded;

ii) whether as a matter of law, an exemplary financial award is available in relation to breach of the Imperial duty and, if it is, the criteria for determining whether such an exemplary award should be made;

iii) whether, on the findings of fact and the evidence referred to in the Judgment, the threshold for an award of exemplary damages under the second category in Rookes v Barnard has been crossed.

190. As to the first of those issues, I can see that a clear "No" to that (and similarly in relation to ii)) might save further litigation (subject to an appeal). All parties might be helped by an answer to i) and ii).

Issue 8: In the light of the answers to Issues 1 to 7, should the Trustee administer the DB Plans on the footing that salary increases awarded following the 2009 NPAs (and/or any salary increases that members who did not enter the 2009 NPAs are entitled to be treated as having) are:

8.1 non-pensionable or are to be treated as not having been awarded;

8.2 pensionable;

8.3 to be treated in some other way and if so what way?

191. In my judgment, the Trustee should administer the DB Plans on the footing that salary increases awarded following the 2009 NPAs are pensionable but only on the footing that the 2006 Partial NPAs remain of continuing effect. The Trustee is not concerned, in administering the Plans, with the salary increases which members who did not enter into the 2009 NPAs would or might have received if Holdings had not acted in breach of duty.

The 2010 and 2011 NPAs

Issues 9 and 11; breach of duty

Issues 9: Was it a breach of the Imperial duty or the contractual duty of trust

and confidence for IBM to procure that members enter into the 2010 NPAs?

Issue 11: Was it a breach of the Imperial duty or the contractual duty of trust

and confidence for IBM to procure that members enter into the 2011 NPAs?

192. Before I deal with Issues 9 and 11, I propose to address the questions whether the 2010 and 2011 NPAs taken by themselves give rise to any breach of duty, addressing each of Scenarios A and B. After that discussion, it will be easier to see precisely what Issues 9 and 11 are driving at.
193. Holdings' position (subject of course to any appeal against the Judgment) in relation to the 2010 and 2011 NPAs is as follows:

i) First, it accepts that the 2011 NPAs involved a breach of the contractual duty of trust and confidence because, like the Initial 2009 NPAs, the freeze on salary increases for members who did not sign was indefinite. Judgment [1535 ii)] is, for reasons already explained, to be read as a holding that the Initial 2009 NPAs involved a breach of duty even under Scenario B. The same must therefore go for the 2011 NPAs.

ii) Secondly, it is contended that the 2010 NPAs involved no breach of duty. They are to be treated like the Later 2009 NPAs. Holdings' position is that my finding in [1535 ii)] of the Judgment was to the effect that the Later 2009 NPAs did not, "viewed in isolation", involve any breach of duty. I have already clarified what I did, and did not, intend by that paragraph. In the light of that clarification, the argument only runs under Scenario B since I did not hold that the Later 2009 NPAs would not have given rise to a breach of duty under Scenario A. I agree with Mr Newman this far, namely that, under Scenario B, there is no material distinction between the Later 2009 NPAs and the 2010 NPAs and therefore reach the conclusion that, under Scenario B, there would be no breach of duty involved in the 2010 NPAs.

194. But Mr Newman's argument does not run under Scenario A because, in my judgment, (i) as I will explain in due course, the Later 2009 NPAs would give rise to a breach of duty under that Scenario and (ii) there is no material distinction between the Later 2009 NPAs and the 2010 NPAs.
195. As to paragraph i) of paragraph 189 above, Holdings does not accept different reasons for saying that the 2011 NPAs gave rise to a breach of duty, namely because they are contrary to any Reasonable Expectation. Whether that is right depends on the Longstop Date. If the Longstop date was prior to the date of the 2011 NPAs, then Mr Newman's argument is right; but if the Longstop Date was after that, then it is wrong.
196. Mr Stallworthy advances two arguments which have been referred to as the "Soto accrual argument" and "the continuum argument" in support of the conclusion that all of the NPAs gave rise to a breach of duty by Holdings when viewed on their own. These arguments can only be applied to the position under Scenario A (in relation to which I have not yet made any decision): under Scenario B I have already decided that the Later 2009 NPAs and the 2010 NPAs would not have given rise to a breach and it is not open now to advance further arguments contrary to that conclusion. In any event, for reasons which will become apparent, I would reject those arguments when applied to Scenario B. Further, the Soto accrual argument proceeds on the basis of the established Reasonable Expectations, whereas under Scenario B there are (it is to be assumed) no Reasonable Expectations at all.
197. The Soto accrual argument, if it is correct, goes much further than supporting the view that the relevant NPAs gave rise to a breach of duty: it would also establish that the DB members could not be discriminated against as compared to DC members when it comes to salary increases. Indeed, it is because of their entitlement to be treated in the same way as DC members that the DB members can say that the relevant NPAs involve a breach of duty. This is because DC members have in fact been given salary increases; DB members should have been given the same increases; for those that signed NPAs, the result has therefore been a disappointment of their Reasonable Expectations, a disappointment so serious as to give rise to a breach of the duty of trust and confidence.
198. I should say at this point that Mr Newman contends that these two arguments are ones which the RBs should not be allowed to run. They should have raised them at the trial and it is now too late to do so. To deal with that submission, it is necessary to understand precisely what the arguments are. So, perhaps somewhat illogically, I will consider the arguments first and then decide whether I should allow them to stand as part of the RBs' case. Before turning to those arguments, it is helpful to recap on some background and to deal with a preliminary argument.

Recap on background

199. There is no doubt that Holdings had the ultimate objective of bringing DB accrual to an end from 6 April 2011 (originally from 2010 but that changed). On the way to reaching that objective, Holdings' strategy was to ensure that future salary increases, including the 2009 salary increases, would be non-pensionable; the 2009 NPAs (and indeed the later NPAs) were part of that strategy. I consider that what I have already said in the Judgment is sufficient to establish that conclusion: see for instance, Judgment [18], [19], [26], [1307], [1324] - [1325], [1521], [1525], [1535 ii)] and [1587] and June Judgment [10]. Mr Stallworthy has mentioned, in his written opening submissions, a great deal of material which evidences this strategy. I agree with him that it demonstrates that IBM and Holdings had, from the beginning of Project Waltz, a strategy to make future pay increases permanently non-pensionable. I reject Mr Newman's submissions to the contrary. I do not propose to go through Mr Stallworthy's written opening in detail although I mention in particular:

i) Mr Riley's email to all employees on 7 July 2009 explicitly warned that "There will be no further pensionable salary increases for Defined Benefit members while they are active members of the Defined Benefit Plan." which I quoted at Judgment [18].

ii) The PCC presentation dated 30 July 2009 maintaining this stance.

iii) A comment to Declan Murphy on the draft Q&A document dated 4 November 2009 which reveals this:

"Declan I'm wondering if to lessen the impact of this we should think about saying that we will go back each year and re offer the opportunity to get a salary increase on the basis that it is non pensionable. It makes the message sound less "forever", and us more reasonable although it will probably mean you get less people signing up to accept at this time? Might also need to check KPMG are OK with this from a savings perspective".

The Soto accrual argument

200. In essence, the Soto accrual argument is that it was a breach of Holdings' duty of trust and confidence to seek to procure NPAs subverting and undermining the Reasonable Expectation of continued accrual in accordance with the Soto changes. Mr Stallworthy's main line of argument is that, Holdings and the Trustee having agreed the Soto changes (under which salary increases would be at least 2/3rds pensionable), it was likely seriously to damage employees' trust and confidence in Holdings for it to turn round three years later and insist on an agreement that future salary increases had to be entirely non-pensionable otherwise no increases would be given. While DB members may not have had any Reasonable Expectation of any particular level of salary increases in any particular year, it was implicit, he submits, in the relevant Reasonable Expectation (that future salary increases would be at least 2/3rds pensionable) that Holdings would not declare DB members ineligible for any salary increase whatsoever unless they entered NPAs making salary increases entirely non-pensionable. That subverted and undermined the 'deal' struck under the Soto Changes and the elections which members made under Soto on the faith of IBM's assurances. What distinguishes the present case from the typical non-pensionability agreement (such as is found in South West Trains) is that in the present case, but not in the typical case, there is a relevant and important Reasonable Expectation.
201. Mr Stallworthy goes on to identify four further factors which make Holdings' conduct even more likely to damage the members' trust and confidence:

i) Even the Later 2009 NPAs were preceded by what he describes as threats that members would not receive pay increases if they did not sign the NPA. I am not assisted, as I have said before, by the use of pejorative words such as "threat". But whatever word one uses, the reality was that it was being made clear that no NPA meant no salary increase.

ii) Holdings' requirement for members to enter into NPAs was part of its strategy that pay increases should be entirely non-pensionable. The Later 2009 NPAs and the 2010 NPAs were put to members as part of the same strategy: although expressed to be comparatively short-term agreements, they were part of a series of agreements which, from their inception, were intended to achieve the objective of making all future salary increases wholly non-pensionable. Mr Stallworthy suggests (although Mr Newman rejects this suggestion) that the short-term nature was cosmetic because Holdings believed (incorrectly as it turns out) that the Exclusion Notices were valid and would permanently break the final salary link: the 2011 NPAs, at least, were only put in place because Holdings had become concerned about the validity of the Exclusion Notices.

iii) Holdings' strategy unreasonably, perversely and capriciously discriminated between (i) DC members (including Enhanced M Plan Members who had elected to move to the M Plan under Soto), who would be eligible for salary increases and whose salary increases would be fully pensionable (and as regards Enhanced M Plan Members, even for DB accrual) and (ii) DB members, who would be entirely ineligible for salary increases unless they submitted to any salary increase being entirely non-pensionable, contrary to the 'deal' struck under Soto.

iv) Holdings made matters worse by restricting access to the M Plan. It sought to maximise its costs savings: it sought to deter members from leaving the DB plan in order to obtain revaluation on their deferred pensions by preventing them from joining the M Plan, leaving them with no employer contributions: see Judgment [1163], [1190] and [1328]. Mr Stallworthy describes the effect as being "to narrow members' already bleak options further, compelling members to sit through two salary cycles (2009 and 2010) with neither pensionable salary increases nor statutory revaluation of their accrued benefits."

202. These are all very powerful arguments in support of the conclusion (which I reached in the Judgment) that Project Waltz as a whole gave rise to a breach of Holdings' duty of trust and confidence and its Imperial duty. But his argument directed at the Later 2009 NPAs and the 2010 NPAs is not without problems. I deal with his four supporting points before looking at his main argument. In doing so, it must be remembered (I need to repeat) that the present exercise is being carried out in the context of Scenario A since Scenario B has already been adequately addressed.
203. Point iii): I have already held that there is no Reasonable Expectation that DB members would not be discriminated against as compared with DC members: see Judgment [1518]. Under Scenario A Project Waltz is to be ignored and we simply start with the position under which (i) there are two classes of member, DB and DC members and (ii) there is a Reasonable Expectation that salary increases will be at least 2/3rds pensionable. Even if it would be a breach of contract to award no pay increase at all (this is point i) to which I will come in a moment), I can see no reason why discrimination between DB and DC members is necessarily unreasonable or perverse or capricious. If in the context of a conventional South West Trains situation, a non-pensionability agreement such as the Later 2009 NPAs and the 2010 NPAs would not be objectionable (as, in my view, would be the case) notwithstanding a difference in treatment from that afforded to DC members, I do not consider the fact that there is a Reasonable Expectation regarding pensionability of salary increases leads to the conclusion that a DB member nonetheless has an entitlement to the same increase in salary as a DC member.
204. Point iv): this is something which can be taken into account in determining the remedy for the overall breach to which Project Waltz gave rise. But in the context of consideration of whether the NPAs, under Scenario A, gave rise to a breach of duty, point iv) is not, in my view, relevant. To bring into the equation the restriction on access to the M Plan would be to bring into account another element of Project Waltz. The object of my analysis is to ascertain the position on the hypothesis of Scenario A. Whereas I think there is assistance to be gained, in addressing remedies, from establishing the position under Scenario A, I do not think any assistance is to be gained by addressing a Scenario A Max (ie Scenario A plus exclusion from M Plan), under which almost by definition Mr Stallworthy gets what he wants. It really adds nothing to the actual finding I have made that all of the NPAs, in the context of Project Waltz as a whole and taking account of the established Reasonable Expectations, involve Holdings in a breach of duty.
205. Point ii): under Scenario A, the NPAs are looked at divorced from their context in Project Waltz. The strategy is of the essence of Project Waltz and cannot feature in an analysis of the NPAs under Scenario A. However, this point is taken account in my further considerations and conclusions below.
206. Point i): Although there was a serious "threat" (I will use Mr Stallworthy's word but ignore its pejorative connotations) in relation to the Initial 2009 NPAs and the 2011 NPAs, that is not the case in relation to the Later 2009 NPAs and the 2010 NPAs. If this were a conventional South West Trains situation, periods of non-pensionability of that sort would be unlikely to give rise to any breach of an employer's duty. Taking, counterfactually, the Later 2009 NPAs and the 2010 NPAs out of the context of Project Waltz (ie Scenario A again), I do not consider that this point demonstrates a breach of duty when, without it, there would not be one. The (short-term) threat is said to be objectionable because the only reason it was short-term was because Holdings believed the Exclusion Notices would achieve the desired end. But that, it seems to me, is once again to set those NPAs in the context of Project Waltz which is contrary to the assumption of Scenario A. As with point iii), I do not think any assistance is to be gained by addressing a Scenario A Max (Scenario A plus long-term intention) which again really adds nothing to the actual finding I have made that all of the NPAs, in the context of Project Waltz as a whole and taking account of the established Reasonable Expectations, involve Holdings in a breach of duty.
207. I come, then, to Mr Stallworthy's main line of argument as set out in paragraph 196 above. This is an attempt, as I see it, to establish a further Reasonable Expectation namely that some salary increase would be awarded. I do not remember Mr Tennet raising this possibility at the main hearing although he did, of course, raise the possibility of a Reasonable Expectation of non-discriminatory salary increases, a possibility I rejected. There was nothing in the evidence which I received at the trial to suggest that future salary increases were discussed: there was no agreement, or even a non-binding expressly stated understanding, that any salary increase in the future would be made. There could, therefore, have been no Reasonable Expectation generated unless it was necessarily implicit in the established Reasonable Expectations. Under Scenario A, I do not consider that it would be implicit. A member would no more be entitled to an increase than he or she would be if the NPAs had not been made; whilst members may have had a hope, or even an expectation (but not one engendered by Holdings) of future pay increases, it cannot be said that they had a right to such increases.
208. In any case, if there were a right to an increase, that would immediately raise the question "How much?" to which, I think, there is no answer and shows that there cannot really be a right in the first place. What can be said, however, is that a member might have a right to be considered for a pay increase (subject to any valid NPA which would absolve Holdings from considering the case further during the currency of the NPA). This is a point which I will address further in the context of remedies.
209. Accordingly, I reject the Soto accrual argument as articulated by Mr Stallworthy if I apply it under Scenario A.

The continuum argument

210. Mr Stallworthy submits that, even if the Later 2009 NPAs and the 2010 NPAs do not give rise to a breach of duty by Holdings viewed "in isolation" (he does not distinguish between Scenario A and Scenario B), it would be artificial and inappropriate to divide up the different categories of NPA, divorcing Later 2009 NPAs and/or 2010 NPAs from the Initial 2009 NPAs and the 2011 NPAs. He submits that the NPAs should all be viewed together as a single continuum: they all formed part of Project Waltz, which Holdings admits. He also submits that they were all part of Holdings' strategy to make future salary increases non-pensionable for all time.
211. That strategy, he submits, was itself a breach of duty. The question should not be whether each NPA gave rise to a breach of duty but whether all of the NPAs together were a breach of duty. The various categories of NPA were a continuum effecting that strategy (just as much as if, in pursuit of such a strategy, an employer came back every year successively requiring an NPA limited to a single year for any pay rise ever to be granted). Since the reality is that Holdings' strategy was to impose a complete, permanent freeze to DB accrual, rendering all future salary increases entirely non-pensionable, the number of several NPAs by which this was sought to be achieved should not alter the conclusion that the result was (as I held) a breach of duty.
212. I come back to the same point which I made in relation to the Soto accrual argument. It is, in different words, that my introduction of the "in isolation" aspect of the case was not to qualify in any way the breaches of duty on the part of Holdings. It is the entirety of Project Waltz which gave rise to the breaches of duty and the remedies to which the RBs are entitled must reflect those breaches of duty. I had focused on each element of Project Waltz in isolation (ie sometimes Scenario A and sometimes Scenario B) because I considered, and still consider, that some assistance may be given in determining the appropriate remedies from consideration of the breaches, if any, to which any particular element gives rise in isolation. I can see that it might also be of assistance to consider whether the NPAs taken as a whole, but independently of the other elements of Project Waltz gave rise to a breach of duty (and whether the position would be different if the Reasonable Expectations are or are not taken into account) but that is not what I had in mind when referring to the various elements in isolation.
213. My own analysis is slightly different from that of Mr Stallworthy. In the Judgment, I held that DB members had a Reasonable Expectation that benefit accrual "in accordance with the Soto Changes" would continue at least until April 2011. The only relevant changes under Soto were that future salary increases would consist of a base salary increase plus a 50% non-pensionable supplement (resulting, in effect, in increases in salary being only 2/3rds pensionable). I held that the Project Waltz changes conflicted with that Reasonable Expectation because DB members would become ineligible for salary increases in the future unless they entered into NPAs: for slightly more detail, see Judgment [1511]. And I also held (see Judgment [1528]), that the 2009 NPAs were procured in breach of duty in the context of Project Waltz as a whole.
214. The three major elements of Project Waltz were identified in Judgment [1511] (cessation of benefit accrual from 6 April 2011, no salary increases unless NPA signed and detrimental changes in ER policy). I have already explained that the 2009 NPAs displaced the Reasonable Expectation that salary increases would (at least in part, following the Soto changes) be pensionable without at that point deciding whether the disappointment of this Reasonable Expectation in relation to the Later 2009 NPAs is sufficiently serious, by itself, to give rise to a breach of the contractual duty of trust and confidence. In my judgment, it is. The Reasonable Expectation that benefit accrual would continue until a Longstop Date (on any view not before April 2011) was an extremely important part of the agreements and understandings which were part and parcel of the Soto changes. Part of that Reasonable Expectation was that salary increases (if any) granted before that date would be pensionable and the pensionability of salary increases was an important aspect of the Reasonable Expectation. To disappoint that part of the Reasonable Expectation without justification (as I have held to be the case) seems to me to be a very serious matter. The same goes for the 2010 NPAs and the 2011 NPAs.
215. It is no answer to that, in my view, for Holdings to say that there could be no disappointment of the relevant Reasonable Expectation because it did not have to award any salary increase. The point is that if a member did sign an NPA, any salary increases actually awarded would not be pensionable. As a matter of fact, such a member could reasonably hope for a salary increase and, indeed, increases were awarded to those who did sign. If it is right that Holdings was able to decline to grant any salary increases, that may go to remedy: it does not go to breach.
216. Accordingly, I consider that under Scenario A, the Later 2009 NPAs and the 2010 NPAs gave rise to a breach of duty by Holdings. A fortiori, if those NPAs are viewed in conjunction with the other NPAs, there is a breach of duty.
217. That is a conclusion based on the disappointing of the relevant Reasonable Expectation. It is not, however, just a matter of disappointing the Reasonable Expectation. The Initial 2009 NPAs and the 2011 NPAs are objectionable quite apart from the Reasonable Expectation because no reasonable employer would make a blanket decision refusing for all time not to grant any pay increase unless the member concerned entered into an NPA. Since, as I have held, all of the NPAs were part of a strategy to ensure that future pension increases would be non-pensionable, it is appropriate, in my judgment, to assess the validity of the Later 2009 NPAs and the 2010 NPAs by reference to their place in the overall scheme of things, as part of the strategy. This is, in effect, Mr Stallworthy's continuum argument. I accept it. I reject Mr Newman's submission that there is no factual basis for this conclusion. Indeed, I think that the paragraphs in the Judgment which I have mentioned earlier in this judgment contain relevant findings. If those are wrong, on the basis that there was no evidential basis for them, that is a matter for correction by the Court of Appeal.
218. I now return to Issues 9 and 11. These Issues ask whether it was a breach of the Imperial duty or the contractual duty of trust and confidence for Holdings to procure that members enter into the 2010 NPAs and the 2011 NPAs respectively. The answer is clearly that there was a breach of the contractual duty in each case in the context of Project Waltz as a whole.
219. However, when it comes to considering remedies, it may be important to understand why, conceptually, there is a breach. To illustrate the point which I wish to make, I go back to the proposition (see paragraph 30 above) that an employer's duty of trust and confidence can be breached by its actions cumulatively. This can be so where none of the relevant actions, if it were the only action, would amount to a breach of duty. So imagine a case where the employer takes 5 steps none of which by itself and taken out of context would give rise to a breach of duty but which cumulatively do give rise to a breach of duty. If one asks whether step 3 was a breach of duty, the response might be for further clarification of what the question means. Does it ask whether step 3 would be a breach of duty if it was the only step, to which the answer is "No". But if the question is whether step 3 forms part of a course of conduct which gives rise to a breach of duty, then the answer is "Yes"; and this could be so even if the cumulative effect of steps 1, 2 and 3 was not to give rise to a breach of duty, but steps 4 and 5 were needed to complete the breach. On the facts of the case, it may be that steps 1 to 5 were not simply a sequence of steps having a cumulative effect, but formed part of a pre-planned scheme or strategy: the question whether step 3 was a breach of duty might then be focusing on whether, when it was taken, it was a breach of duty in which case the answer would, again, be "Yes". The answer is, a fortiori, "Yes" if step 3 would have been, standing by itself, a breach of duty.
220. In the present case, the 2010 NPAs gave rise, for reasons already stated, to a breach of duty under Scenario A, that is to say viewed apart from Project Waltz. This was because of the serious disappointment of the Reasonable Expectation that salary increases would continue to be pensionable until at least April 2011. In the context of Project Waltz as a whole, it was an essential element in a larger scheme that disappointed the Reasonable Expectation and was thus a breach viewed in the context of Project Waltz as a whole. This is because a threat that no salary increases would be considered at all in the future unless an NPA was signed would be a breach of duty; and the 2010 NPAs were part of a strategy to achieve that end with no genuine intention that the position would be reviewed each year. At least, there was no evidence of a genuine intention to do so and I am entitled to infer from the totality of the evidence which I did receive that there was no such intention.
221. The answer to Issue 9 must therefore be that Holdings was in breach of its contractual duty of trust and confidence in procuring the members to enter into the 2010 NPAs. However I do not consider that procuring the members to enter into the 2010 NPAs gave rise to a breach of Holdings' Imperial duty. It is true that the way in which effect is given to a valid NPA (as in South West Trains) involves the trustees of the scheme giving effect to the contract between the employer and the employee and thus overriding the formal provisions of the scheme. If the trustees threaten to ignore that contract, the employer's remedy is to seek injunctive relief to restrain payment to the employee of that to which the rules strictly entitle him. The economic effect of the contract, it can be seen, is the same as a valid rule amendment if one could be made. However, in my view, the procuring of the contract does not engage the same sort of duties as the exercise of a power exercisable by an employer under the scheme.
222. Accordingly, I consider that procuring the members to enter into the 2010 NPAs did not give rise to a breach of Holdings' Imperial duty. In practice, I doubt that the imposition of the Imperial duty on Holdings would in any case add anything to its contractual duties of trust and confidence; indeed, the contractual duty can sometimes be more onerous than the Imperial duty.
223. In relation to the 2011 NPAs, the position is slightly different. By the time that these were entered into, the duration of the Reasonable Expectation concerning benefit accrual was of course shorter. The Longstop Date was no further away than 2014 and may have been earlier. On its own, an NPA which disappointed that Reasonable Expectation for the period remaining might not be seen as giving rise to a breach of duty. However, the threat of no salary increases being considered at all in the future unless an NPA was signed was again a breach of duty. Had the 2011 NPAs been genuinely short-term, and not such as to give rise under either Scenario A or Scenario B to a breach of duty, the position might have been different. But that is counterfactual. On the facts, the 2011 NPAs were objectionable.
224. The answer to Issue 11 must therefore be that Holdings was in breach of its contractual duty of trust and confidence in procuring the members to enter into the 2011 NPAs. But as in relation to Issue 9, I consider that procuring the members to enter into the 2011 NPAs did not give rise to a breach of Holdings' Imperial duty.
225. Holdings' case is, as I have said, that the Soto accrual argument and the continuum argument are not open to the RBs at this stage of the proceedings for various reasons, some of which, at least, would apply to the way in which I now put the case. In essence, it is said that the arguments are barred by a res judicata arising on three grounds. I set out those grounds and my decisions on them:

i) The first relates to Judgment [1535 ii)]. I have already dealt with this fully and do not consider that there is anything in this point.

ii) The second ground is that the RBs did not argue at trial that the Soto changes engendered a Reasonable Expectation about pensionability of future salary increases which the NPAs confounded; and that it is too late to do so now. For reasons already given, I do not consider that Mr Stallworthy's arguments that the failure to provide any salary increase would be perverse or capricious, let alone his argument that DB members should have the same salary increases as DC members, are good ones which makes it unnecessary to decide whether I should admit them. But so far as the NPAs give rise to a breach of duty on the basis that salary increases would be non-pensionable, I have already said that this Reasonable Expectation is a necessary part of the Reasonable Expectation that benefit accrual would continue in relation to salary, because the increase is part of the salary. There is no res judicata preventing me from dealing with this point.

iii) The third ground is that the Soto accrual argument cannot be run without giving Mr Ferrar an opportunity to give further witness evidence, which was not feasible, and did not take place, at the July hearing. In the light of my observations about the Soto accrual argument, there is nothing in the argument as applied to Scenario A. I do not, however, consider that further evidence from Mr Ferrar was necessary in order to address the argument properly.

226. In relation to the continuum argument, the whole issue concerning the treatment of each element or a discrete collection or elements together arises only as a result of what I said in the Judgment. Submissions arising out of my approach could not reasonably have been anticipated at the hearing. The only question is whether anything which I have said in the Judgment is inconsistent with the continuum argument and, in particular, whether any res judicata arises. In my judgment there is no such inconsistency and there is no res judicata. It is in my judgment entirely appropriate to allow this argument to be raised.

Issues 10 and 12

227. These issues follow on from Issues 9 and 11 and ask the same questions mutatis mutandis, in relation to the 2010 and 2011 NPAs respectively as Issues 1 to 8 in relation to the 2009 NPAs. Since I have held, in effect, that the 2010 NPAs are to be treated in the same way as the Later 2009 NPAs, my answers are the same. And since I have held, in effect, that the 2011 NPAs are to be treated in the same way as the Initial 2009 NPAs, my answers are again the same.

The Exclusion Notices – Issues 13 to 20

Issue 13

228. At Judgment [1534] I held that Holdings acted contrary to its Imperial duty in exercising the Exclusion Powers. Issue 13 asks the following questions:

13 Were the Exclusion Notices:

13.1 •    void and/or

???•    unenforceable

13.2 voidable and liable to be set aside,

and if so:

13.2(a) •    can they only be avoided and set aside by the Court in relation to all or a group of the Affected Members collectively,

???•    and if so will the Court now avoid them and set them aside;

13.2(b) •    can they be avoided and set aside at each Affected Member's individual election, and if so can Affected Members now avoid them and set them aside,

???•    and what would they have to do or prove in order to do so;

13.2(c) •    are they liable to be avoided and set aside in some other manner, and if so in what manner;

13.2(d) •    may any power to avoid or set aside the Exclusion Notices

be exercised independently of any such power in relation to the NPAs (so that, for example, the NPAs are avoided but the exclusion is affirmed);

13.3 valid?

Issue 13.1: Were the Exclusion Notices notice void and/or unenforceable?

229. The Trustee's primary case is that the Exclusion Notices were void. Being void, they are also unenforceable. Also, I add, if they are void, that will be the case in relation to all members: it will not be possible for any members to elect that the Exclusion Notices should be treated as valid vis à vis them if, for some reason, it suits their economic interests better. In contrast, Holdings' position is that the Exclusion Notices are simply voidable; they should not be avoided but, on the contrary, should be validated on terms (as to which for more detail see at paragraph 263 below) the effect of which, in summary, would be that they would take effect according to their terms but Affected Members would be afforded the opportunity to be placed in the position that they would have been in if pensionable service in relation to the DB Plans had not been terminated by the Exclusion Notices until the Longstop Date (for the meaning of which, see paragraph 23 above).
230. The Trustee has two principal arguments for saying that the Exclusion Notices are void which can be summarised as follows:

i) First, the Imperial duty acts as a restriction on the scope of an employer's powers under a pension scheme so that a purported exercise of a power (in the present case, the Exclusion Powers) in breach of the Imperial duty is outside the scope of the power and, therefore, ultra vires and void.

ii) Secondly, a power can only be exercised for what has sometimes been referred to as a "proper" purpose. I will say something more about that in a moment. But what the Trustee says is that a "proper" purpose cannot include a purpose which gives rise to a breach of the Imperial duty when the power is exercised in a particular manner. The effect of the purported exercise of a power for an improper power is again that the exercise is void.

Although these arguments are addressed separately in the parties' submissions, I see them as very closely interlinked since they both require me to consider the nature of the Exclusion Powers. I propose, therefore, to deal with them together.

231. Mr Evans submits that the correct approach to the so-called "duty" is to see it as a limit on, or disability affecting, the manner in which the power may be exercised rather than as a positive duty to act. Accordingly, a purported exercise of a power in a manner which breaches the Imperial duty is an exercise beyond the limits of the power and thus outside of its scope. This, he submits, is consistent with the origins of the duty in the employment contract where the duty of trust and confidence operates as an overriding restriction on the literal scope of an employer's powers. This analysis of the Imperial duty is regarded as plausible by the author of Thomas on Powers (2^nd ed): see at 10.207.
232. Mr Evans suggests that the characterisation of the Imperial duty as a limitation (or fetter) on the exercise of a power is clear from my own approach in the Judgment. He relies on the following:

i) My acceptance of Holdings' submission that the Imperial duty is an implied term which:

"operates to constrain the exercise by the employer of any power vested in it under the governing provisions of the scheme, insofar as such exercise affects members of the scheme." (see Judgment [358])

That is not quite accurate. What I did in that paragraph was to record Mr Simmonds' acceptance that in a pension scheme, the implied term "operates etc" as in the quote above. Of course it constrains the employer: but that does not mean that, when the employer breaches the constraint, its acts are void rather than voidable.

ii) My statement:

"the effect of the Imperial duty is to place a limitation on the manner in which an apparently unfettered power may be exercised." (see Judgment [372])

However, I went on in the same paragraph to say the consequences are not clear and that, in some cases, it may not be proportionate to strike down the exercise of the power at all and that there may be other cases where the result is partial, rather than total, invalidity of the exercise of the power.

iii) My statement:

"Indeed, if there are fetters on the power, the task for those alleging a breach of the Imperial duty may be harder than if there had been no fetters at all. The presence of fetters may be seen as at least some indication that the power is intended otherwise to be entirely unfettered." (see at Judgment [425])

Mr Evans suggests that this shows that I clearly considered the Imperial duty to operate as an "implied fetter" on the scope of a power. I do not agree. As reference to Judgment [424] shows, the discussion was about fetters on the power in the sense of restrictions: it was a discussion about what could properly be done in the exercise of a power and not one about the remedy where the power has purportedly been exercised in a way inconsistent with the fetter or restriction.

233. Mr Evans goes on to say that his approach is fully consistent with both Imperial and the later authorities which have considered the principle. He has referred to the following:

i) Imperial itself, where Sir Nicolas Browne-Wilkinson V-C, when formulating the principle, stated that the rights and powers of the employer "can only be exercised in accordance with the implied obligation of good faith". Mr Evans notes that the declaration made by the Vice-Chancellor was that the employer's right to give or withhold its consent "is subject only to the restriction that such right is not validly exercisable in breach of the obligation of good faith". Accordingly, he submits, any exercise in breach of such obligation would be outside of the power.

ii) In Hillsdown Holdings Plc v Pensions Ombudsman [1996] PLR 427 Knox J referred to the Imperial duty as acting to "restrain" the employer in the manner of exercise of its powers.

iii) Holdings' own skeleton argument for the main hearing where the Imperial duty was described as operating "to place limitations on the discretionary powers".

234. Mr Evans submits that this approach is consistent with (and supported by) the need for a power to be exercised "in good faith". This goes to the scope of the power so that an exercise in bad faith is outside the scope of a power: see Snell's Equity (32^nd ed) at 10-019. It is the final paragraph of that section which encapsulates the point but the discussion in Snell at this point is concerned with fiduciaries. It reads as follows:

"The requirement of good faith limits what a fiduciary may or may not do lawfully in exercise of his powers: good faith goes to the scope of a power. Put the other way, bad faith action is outside the scope of a power."

235. And so Mr Evans maintains that the correct characterisation of the Imperial duty as a disability or limitation on the scope of a power means that a purported exercise of a power in breach of that duty will be void. This is because there is simply no power to effect the purported exercise (ie it is ultra vires).
236. In the same vein, he refers to the decision of the Court of Appeal in Pitt v Holt [2011] EWCA Civ 197 at [96], where Lloyd LJ explained, in relation to a power vested in trustees, that: "The purported exercise of a discretionary power on the part of trustees will be void if what is done is not within the scope of the power." And in the Supreme Court, Lord Walker said that:

"The rule, [the so-called rule in Hastings-Bass]astings-Bass)Hstings-Bas properly understood, depends on breach of duty in the performance of something that is within the scope of the trustees' powers, not in the trustees doing something that they had no power to do at all."

He agreed that a breach of that "rule" resulted in the exercise of the power being voidable, not void.

237. While considering Pitt v Holt, I note that Lord Walker also had something to say about fraudulent appointments (see at [62]):

"There is Court of Appeal authority that a fraudulent appointment is void rather than voidable: Cloutte v Storey [1911] 1 Ch 18. In that case the appointee under an improper appointment had charged his equitable interest as security for a loan (and in doing so made two false statutory declarations as to the genuineness of the appointment). It was held that the lender had no security, even though it had no notice of the equitable fraud. It is an authority which has bedevilled discussion of the true nature of the Hastings-Bass rule. Lightman J in Abacus Trust Co (Isle of Man) v Barr [2003] Ch 409 , para 31 found the judgment of Farwell LJ problematic and Lloyd LJ shared his reservations [2012] Ch 132 , para 98. So do I. It is hard to know what to make of Farwell LJ's observations [1911] 1 Ch 18 , 31:

'If an appointment is void at law, no title at law can be founded on it; but this is not so in equity: the mere fact that the appointment is void does not prevent a Court of Equity from having regard to it: eg, an appointment under a limited power to a stranger is void, but equity may cause effect to be given to it by means of the doctrine of election.'

The decision in Cloutte v Storey may have to be revisited one day. For present purposes it is sufficient to note that a fraudulent appointment (that is, one shown to have been made for a positively improper purpose) may need a separate pigeon-hole somewhere between the categories of excessive execution and inadequate deliberation. "

238. Later on (see at [93]) he expressed his agreement with Lloyd LJ's holding (at [99] of his judgment) in the Court of Appeal that:

"if an exercise by trustees of a discretionary power is within the terms of the power, but the trustees have in some way breached their duties in respect of that exercise, then (unless it is a case of a fraud on the power) the trustees' act is not void but it may be voidable at the instance of a beneficiary who is adversely affected."

239. It is also important to note something else which Lloyd LJ said when he was considering Cloutte v Storey. At the end of [98], he said this:

"…… It is not necessary to go into the point in more detail for present purposes, but although we are bound to hold that the effect of an appointment being found to have been made in fraud of the relevant power is that it is void, not merely voidable, I am not willing to apply that decision more extensively, by analogy, to cases to which it does not relate directly as a matter of decision."

240. That hostility, if I may use the word, to the voidness, rather than the voidability, of the exercise of a power is further reflected by Lloyd LJ at [101], articulating a rather wider policy imperative:

"In principle, cases where an act done by trustees which appears to be within their powers can be held to be void ought in my judgment to be kept to a minimum, just as at common law the cases where a transaction is void, rather than voidable, are few and far between. Robert Walker J's reference in the passage cited at para 90 above to "damaging uncertainty as to what has and has not been validly decided" is very much in point here, as is what Lightman J said in his judgment in Abacus Trust Co (Isle of Man) v Barr [2003] Ch 409 , para 32: see para 88 above. By contrast, if the defect renders the act voidable, the availability of equitable defences and the court's discretion as to the remedy make it easier to reach a just outcome while recognising the defect in the transaction."

241. Holdings' case is that the Exclusion Notices are voidable rather than void. Mr Simmonds' principal submission is that the nature and effect of the Imperial duty is to impose a fetter upon the manner in which an employer may exercise an otherwise unfettered discretionary power conferred upon it in the context of an occupational pension scheme. It is not a limit on the scope of the power. In this regard, the fetter imposed by the Imperial duty on an employer is analogous to the fetter on trustees in respect of their discretionary, dispositive powers: the exercise of such powers is fettered in the sense that they may not exercise such a power in a manner that would constitute a breach of trust/fiduciary duty (as a result, for example, of their failing to consider matters that they are duty-bound to consider). What the Imperial duty does not do, however, is impose a fetter upon the scope of the power conferred upon the employer. A breach of the duty does not render the exercise of the power in breach of the Imperial duty exercise ultra vires.
242. Mr Simmonds, too, refers to passages in the Judgment to show that what I said is consistent with his approach, referring to Judgment [358] and, like Mr Evans, to Judgment [372]. I do not consider that either side can derive any assistance from what I said in the Judgment. Nowhere, I think, did I say anything inconsistent with either approach. It is, in any case, clear that my intention, in the Judgment, was to avoid saying anything substantial about remedies; and what I said in the various paragraphs relied on by the parties is not to be taken as displaying a departure from that intention.
243. Mr Simmonds submission is that the consequences of an exercise in breach of the Imperial duty of a power conferred upon an employer under an occupational pension scheme ought to be the same as the consequences of an exercise in breach of trust or fiduciary duty of a power conferred upon a trustee under a private trust. The two situations, he says, are obviously analogous, in the sense that the manner of the exercise is, in each case, within the scope of the power conferred upon the employer or the trustee (as the case may be) but is nevertheless vitiated by the fact that such exercise is in breach of the Imperial duty (in the case of an employer) or in breach of trust/fiduciary duty (in the case of a trustee).
244. Therefore, the consequences of an exercise in breach of the Imperial duty of a power conferred upon an employer under an occupational pension scheme ought to be the same as the consequences of an exercise in breach of trust/fiduciary duty of a power conferred upon a trustee under a private trust. Mr Simmonds notes (as is apparent from the passages I have already quoted) that in Pitt v Holt, the Supreme Court has confirmed that the consequence of such a breach is that the exercise is voidable rather than void. He submits that, by analogy and parity of reasoning, an exercise of a power in a manner that constitutes a breach of the Imperial duty ought to render such exercise voidable rather than void.
245. Furthermore, he submits that, given that the Supreme Court has confirmed in Pitt v Holt that an exercise by a trustee of a dispositive power in a manner that constitutes a breach of trust/fiduciary duty on his part is voidable rather than void, it would be extremely odd and illogical if it were the case that an exercise by an employer of a power conferred upon it under an occupational pension scheme in a manner that breached the non-fiduciary Imperial duty were void rather than voidable. There is no sensible reason for imposing – indeed, it would be perverse to impose – a more draconian remedy in respect of the exercise of a power in a manner that breaches a non-fiduciary duty than the remedy that arises in respect of the exercise of a power in a manner that is in breach of trust/fiduciary duty.
246. Mr Simmonds has referred to [101] of Lloyd LJ's judgment in Pitt v Holt and the policy which it discloses in favour of voidability rather than voidness. The reasoning, he says, applies with equal force to decisions made by employers in respect of occupational pension schemes.
247. Furthermore, treating the Exclusion Notices as void rather than merely voidable would risk lack of flexibility and the possibility of unintended consequences. On the facts of the present case, two such unintended, or unwelcome, consequences are, according to Mr Simmonds, immediately apparent. First, it follows necessarily from the Trustee's contention that the Exclusion Notices are void ab initio that there cannot be any right on the part of a given member to elect under the Pension Decision Tool to remain in his Chosen DC Arrangement even if that would be more beneficial to him than continuation of pensionable service for the purposes of the DB Plans. Secondly, a consequence of a finding that the Exclusion Notices are void would be that Hybrid Deferred status would simply fall away so that any new exercise of the Exclusion Powers (it being common ground that the Longstop Date is no later than 31 March 2014) would result in C Plan Over 60 Members being unable to access an unreduced pension on early retirement.
248. I agree with the submissions made in the preceding two paragraphs. I attach no weight to the way in which the matter was put in Holdings' position paper or to Mr Evans' expression of surprise on the part of the Trustees on learning of Holdings' current stance (which, as Mr Simmonds' points out, was the Trustee's own position before it took on the mantle of arguing the RBs' case). Further, I reject Mr Evans' argument to the effect that Holdings' argument fails to recognise the effect of the Pension Decision Tool. Mr Simmonds' point, with which I agree, is that the Tool itself envisages an affected member waiving the entitlement to be treated as having continued in pensionable service under the DB plan in the event that the Court held (as I did) that members were so entitled. That provision of the Tool is consistent only with the Exclusion Notices being voidable rather than void (or indeed unenforceable). Unless I am compelled by an overwhelming analysis, or bound by authority, to conclude that a breach of the Imperial duty leads to the Exclusion Notices being rendered wholly void, it is not the conclusion which I would reach.
249. I do not consider that I am compelled or bound in that way. The Imperial duty is a comparatively recent invention of the Courts. It is itself derived from another comparatively recent invention, namely the implied duty of trust and confidence between employer and employee in an employment context. A new duty may be expressed in terms of established concepts (eg in the employment context as an implied contractual term), but it does not follow that the remedy for its breach has to follow the identical remedies which apply to those concepts. The Imperial duty may be expressed as a restriction or a fetter on an employer's power, but to state that does not answer whether the fetter is on the scope of the power, as Mr Evans submits, or on the manner of its exercise, as Mr Simmonds submits.
250. Mr Evans argues that the Exclusion Notices constitute a fraud on the power. He relies on:

i) Millet J's decision in Re Courage [1987] 1 WLR 495 at p505 (in relation to a power of amendment):

"It is trite law that a power can be exercised only for the purpose for which it is conferred, and not for any extraneous or ulterior purpose. The rule amending power is given for the purpose of promoting the purposes of the scheme, not altering them."

ii) Lord Cooke's statement in Equitable Life Assurance Society v Hyman [2002] 1 AC 408 at p460 in the following terms:

"No legal discretion, however widely worded…can be exercised for purposes contrary to those of the instrument by which it was conferred'."

iii) Pitt v Holt in the Court of Appeal at [96] and [97] of Lloyd LJ's judgment (which I do not propose to set out).

251. The submission really comes back to the points already made that the purposes of the trust deed and rules of a pension scheme include promoting trust and confidence between an employer and the members rather than undermining those concepts. Accordingly, an employer's powers under a pension scheme must be exercised for a purpose which is consistent with maintaining trust and confidence rather than undermining them. If, instead, an employer purports to exercise its powers for a purpose which would undermine trust and confidence (and thereby breach its Imperial duty), then such exercise would be contrary to the purposes for which the powers had been conferred. And so it is said that, because the exercise by Holdings of the Exclusion Power (in serving the Exclusion Notices) was a breach of the Imperial duty, it thereby constituted an exercise of the Exclusion Power for an improper purpose. That improper purpose was to seek to implement the Project Waltz changes in order to achieve IBM's own financial objectives in a manner which confounded the members' Reasonable Expectations.
252. Mr Simmonds' response is that the Trustee's position is entirely at odds with the finding in the Judgment that the service of the Exclusion Notices did not constitute an exercise of the Exclusion Powers for an improper purpose: see Judgment [289(iv)] and [293(iii)]. However, I expressly stated at the end of Judgment [289(iii)] and at Judgment [294] that my conclusions were subject to any "good faith" argument and without prejudice to any issues arising in relation to the Imperial duty. Whether or not the point was open to the RBs to run at the earlier hearing, the focus in relation to the Exclusion Notices was rather different. I am sure that the fair and proportionate approach on the present hearing is to allow the Trustee to run whatever points it wishes to raise about the Exclusion Notices giving rise to a breach of Holdings' contractual duty of trust and confidence and the Imperial duty.
253. He has a further argument, namely that if the Trustee's argument in this regard were correct then every exercise of a power by an employer in breach of its Imperial duty would be contrary to the purpose of the power and, hence, void. By analogy, every exercise by a trustee of a discretionary power in a manner that constitutes a breach of trust/fiduciary duty would also be void – but that is inconsistent with the decisions of the Court of Appeal and the Supreme Court in Pitt v Holt. The fallacy in the Trustee's argument he seeks to identify is the assertion that the purpose for which the power was conferred included maintaining trust and confidence between employer and member. That is the purpose of the duty that imposes a fetter upon the power; it is not the purpose of the power itself (not least because, if it were, there would be no need for the Imperial duty to impose such a fetter). Mr Evans' retort is that this argument all turns on equating breach of the Imperial duty with the sort of breach of trust which only renders the exercise of a power voidable and is therefore to beg the question.
254. One thing which is clear, in my judgment, is that a breach of the Imperial duty is not, of itself, a fraudulent exercise of the relevant power in the sense of a fraud on the power as addressed in the authorities. Those cases are concerned with benefiting non-objects of the power. But that is not the present case: the purpose of the Exclusion Power was surely to enable Holdings to bring an end to benefit accrual if it considered it desirable to do so even in its own interests. The objection to the Exclusion Notices is not that this is what they purported to do (ie to exclude the affected members in Holdings' own interests), but that they did so in circumstances which resulted in a breach of the Imperial duty. As Mr Simmonds submits, the power (to exclude affected members) and the duty (not to act in a way which breaches trust and confidence) are not to be confused. If it is right to look at the purpose of the duty at all, it can only be described in an entirely circular way as the purpose of maintaining the relationship of trust and confidence. If Mr Evans' line of argument is correct, I do not see why it would not apply equally to a breach of fiduciary duty in relation to the exercise of a discretionary power: one of the purposes of such powers, on his line of argument, would be to preserve the fiduciary relationship between the trustees and the beneficiaries so that any breach of trust becomes a fraud on the power (because, ex hypothesi, it is beyond the scope of the discretionary powers to be exercised in a way which breaches the fiduciary duty). Such a line of argument is clearly wrong and, in any case, is not admissible in the light of Pitt v Holt in the Supreme Court.
255. The most that Mr Evans can properly do, I think, it to draw an analogy between breach of the Imperial duty and a fraud on the power. Lloyd LJ expressed himself as unwilling to apply Cloutte v Storey more extensively, by analogy, to cases to which it does not relate directly as a matter of decision. Breach of the Imperial duty is a case to which that decision does not relate directly. Like Lloyd LJ, I am unwilling to apply that decision to the present case. In my judgment, the Exclusion Notices are voidable, not void.
256. I add one further point which is perhaps only an aspect of the policy point. It would to my mind be very surprising if an affected member, in the case of breach of the Imperial duty, could not waive the breach and seek to rely on that which the employer has actually purported to do. And yet, if the exercise of the power is void, the employee could not make such an election even if it were to his financial benefit to do so.
257. I should also add that I have not overlooked the arguments which Mr Evans raised based on Air Jamaica v Charlton [1999] 1 WLR 1399 and Mr Simmonds' response to them. I do not propose even to summarise the arguments. I do not consider that Air Jamaica supports the Trustee's case.
258. Mr Evans' fall-back argument is that, even if the Exclusion Notices are not void in the sense of being within the scope of the power, they are nonetheless unenforceable by Holdings. This is said to be because I have found that the Exclusion Notices were served in breach of duty and the affected members would be entitled to obtain injunctive relief restraining Holdings from enforcing the Exclusion Notices. Such injunctive relief would be based on Holdings' breach of both its contractual duty and of the Imperial duty.
259. Mr Evans draws an analogy with South West Trains. In that case, Neuberger J considered that an employer could obtain an injunction to restrain a member from seeking to claim a higher level of pension under the rules of the scheme than the level of pension the member had separately agreed with the employer. In the same way, he says that in the present case an affected member could obtain an injunction to restrain Holdings from relying on the Exclusion Notices since reliance on the Exclusion Notices would be contrary to Holdings' contractual obligations.
260. As Neuberger J explained, the obligation not to seek a greater payment from the trustee of the scheme than that which a member had agreed with the employer could arise from an implied term to that effect contained in the separate contract between the member and employer. In the present case, Mr Evans submission is that it could equally be said that the employment contracts of the affected members contained an implied term that Holdings would not seek to enforce any act which it committed in breach of the employment contract.
261. The affected members would, according to Mr Evans, also be entitled, subject to the Court's discretion, to obtain an injunction against Holdings to restrain a continuing breach of the Imperial duty. The duty imposes an equitable obligation and it follows that, in principle, an injunction could be obtained to enforce compliance. As Lloyd LJ stated in the Court of Appeal in Pitt v Holt in the context of an act performed in breach of duty:

"If the act of the trustees which purports to alter or bring to an end the interest of a beneficiary is affected by a breach of fiduciary duty, then the beneficiary is entitled to restrain the trustees from acting on it, and to have it set aside, subject always to equitable defences and discretionary factors…"

262. In a similar way, an affected member would, it is submitted, be entitled to obtain an injunction against Holdings restraining it from acting on its breach of the Imperial duty.
263. For my part, I do not consider that there is a half-way house of the Exclusion Notices being unenforceable somewhere between their being void and being voidable. In South West Trains the agreement between the employer and the employees did not purport to amend or alter the provisions of the pension scheme. In contrast, the Exclusion Notices (unless they are void or voidable and set aside) do precisely that: they render the affected members ineligible for further benefit accrual as a result of the direct application of the provisions of the Main Plan and the I Plan. If they are void, then no question of unenforceability arises. If they are voidable and are totally set aside, there is no scope for an unenforceability argument. If they are voidable but are set aside only in part or not at all, then to the extent that they are not avoided, I will have decided that the breach of duty (whether of the contractual duty or the Imperial duty) is adequately remedied, in which case there would be no case for further injunctive relief to give what could then be described as an excessive remedy.
264. Mr Simmonds encapsulates the same, or a similar point, in his written closing submissions where he submits that it is not apposite to speak in terms of "unenforceability" at all. The exercise is either valid, in which case it has the effect of altering proprietary rights according to its terms; or it is void (whether ab initio or by reason of the avoidance of a voidable exercise), in which case there is no alteration in proprietary rights. Once it has been determined that the exercise of the power is effective to alter proprietary rights, there are no proper grounds for preventing the beneficiary of such an alteration from enforcing (in the sense of asserting and, if necessary, coming to Court for vindication of) those proprietary rights.
265. The answer to Issue 13.1 is therefore that the Exclusion Notices were neither void nor unenforceable.

Issues 13.2 and 14

266. It follows from the discussion of Issue 13.1 that, in my judgment, the Exclusion Notices are voidable. It follows that, in principle, they are liable to be set aside. The various sub-issues of Issue 13.2 address the possible options as to how and by whom such (total or partial) setting-aside might be effected. In the light of my answer to Issue 13.1, Issues 14.1 and 14.2 do not arise. Issue 14.3 arises if the Exclusion Notices are avoided or set aside and asks:

"If the Exclusion Notices are avoided or set aside

(a) can and should the Court grant relief on terms, so that the Exclusion Notices are treated as taking effect at some date after 6 April 2011 and as terminating such pensionable service at that date,

(b) and if so what is that date;

(c) what, if any, other terms should apply to the setting aside:"

267. Mr Simmonds' written closing submission addresses Issues 13.2 and 14.3 together. Holdings' primary position is that, although the Exclusion Notices are, in principle, voidable they should not in fact be avoided but should be validated or affirmed on terms. Mr Simmonds (perfectly fairly) does not present a fully-drafted formulation on those terms but presents a summary (subject to detailed drafting and the execution of any necessary documents and the passing of resolutions) of such terms as follows:

i) Affected members are to have a window of opportunity in which to exercise any statutory right of cash equivalent transfer (or to effect a transfer under a scheme power) to transfer their benefits from their Chosen DC Arrangement between 5 April 2011 and the Longstop Date into the DB Plans in return for transfer credits, such credits to mirror the benefits that the member in question would have accrued between those dates on the footing that the Exclusion Notices had been invalid to the extent necessary to give effect to such members' Reasonable Expectations;

ii) the conditions in paragraph 5.2 generally of the Pension Decision Tool be implemented in respect of any member making such an election. This would result in (i) affected members' DC funds being transferred to the relevant DB Plan so far as they match what would have been employee and employer contributions to the DB Plan for the relevant period and (ii) any shortfall in employee contributions be made up; and

iii) arrangements be made in the DB Plans to re-activate pensionable service of affected members between the date of implementation and the Longstop Date and, subject to any necessary consents being given by the Trustee and subject also to any subsequent events that would have caused cessation of Hybrid Deferred status in respect of any given member, to give Hybrid Deferred status (by which is meant Hybrid Deferred status in the form that was communicated to members in 2009) to Affected Members from the Longstop Date.

268. Mr Simmonds puts forward this solution as, in effect, a voluntary offer, as he puts it, "to mitigate the mischief of the Exclusion Notices, namely their confounding of Affected Members' Reasonable Expectations as to future accrual in accordance with the Project Soto changes". The terms are offered as the basis upon which Holdings invites me, in the exercise of my discretion, to validate the Exclusion Notices.
269. This is, I think, a rational proposal from Holdings, so far as concerns the role of the Exclusion Notices in Holdings' overall breach of duty. Mr Simmonds' point is that the Exclusion Notices would have been valid if they had operated from the Longstop Date rather than from 6 April 2011; Holdings' offer is to put the affected members in the same position as they would have been in if it had acted consistently with its duties.
270. Mr Simmonds acknowledges that the same economic effect would be achieved by avoiding (rather than validating) on terms that the Exclusion Notices are treated as valid from the Longstop Date but not earlier. Holdings considers, however, that there may be technical difficulties in unravelling matters in a way which puts affected members in the position they ought to be in if the Exclusion Notices are avoided on terms.
271. Mr Simmonds submits that it is clear that the Court has jurisdiction to validate or avoid the Exclusion Notices on terms, placing reliance on Pitt v Holt at [99] to [100] where Lloyd LJ refers to "the availability of equitable defences and the court's discretion as to the remedy making it easier to reach a just outcome while recognising the defect in the transaction". He says that I indicated a similar approach at Judgment [372] where I recognised that "there may be cases where it would not be just or proportionate to strike down the exercise of the power at all; and there may be cases where the result is partial, rather than total, invalidity of the exercise of the power". And he refers to Judgment [1532 i)] as well. I flag here that account needs to be taken, in relation to the duration of the DB accrual, not only of the Longstop Date but also of the requirement for consultation. Holdings submits that no further consultation is needed; the Trustees submit to the contrary. I deal with this aspect at paragraph 689ff below.
272. Mr Simmonds draws an analogy with the grant of other equitable remedies. It is the general practice for the grant of an injunction in interim proceedings to be accompanied by cross-undertakings given to the Court. Similarly, the grant of the discretionary remedy of rectification can be granted conditionally or on the imposition of terms: see Sargeant v Reece [2007] EWHC 2663 (Ch) at [86]; Konica Minolta Business Solutions (UK) Ltd v Applegate [2013] EWHC 2536 (Ch) at [49] – [53], decisions I see no reason to doubt. Finally, he refers to Pitt v Holt in the Supreme Court. In that case, HMRC were concerned that the settlement should not be set aside because, among other reasons, the Court would not know what proprietary claims would vest in the estate of the disponor against third parties (such as the trustees of the settlement). In order to eliminate that concern, the disponor's executors confirmed by letter an irrevocable intention not to make any claim on behalf of the estate against the trustees or recipients of distributions. Lord Walker observed that, but for that letter, "the court might, if minded to grant relief, have required an undertaking to the same effect as the one that [the executors] have volunteered".
273. There is the further point made by Mr Simmonds that, as the decision of Neuberger J in Bestrustees shows, it is possible to sever, in appropriate cases, the exercise of a power into valid and void components and to uphold the valid and strike down the void. In my view, it must surely follow (again in appropriate cases) that the court can take the same approach in relation to a voidable exercise of a power which can be severed in the same way. But the point goes further; it can be argued that the court should recognise the flexibility of the discretionary remedy by, in effect, severing the good from the bad by imposing terms or accepting undertakings.
274. Then comes the critical part of Mr Simmonds' submissions on the facts. He submits that there cannot have been a breach of the Imperial duty in the context of Project Waltz but for the disappointment of Reasonable Expectations on the part of members. For that, he relies on various parts of the Judgment, in particular Judgment [389], [391], [1030] and [1519] which show that the disappointment of Reasonable Expectations was the foundation of the breach of the Imperial duty. That is right subject to two caveats. The first is that the Initial 2009 NPAs and the 2011 NPAs would have given rise to a breach of the Imperial duty because of the permanent nature of the pensionable pay freeze; and the second is the failure to consult, the consequences of which I come to later. However, it is possible to take account of both of those caveats in fashioning the appropriate remedy (or in the formulation by Holdings of its proposals). But subject to those caveats, the Reasonable Expectation in relation to benefit accrual extended only to the Longstop Date and to treat the Exclusion Notices as valid beyond that date would be to provide a remedy for a period in respect of which there was no breach of duty, a result which he says would be neither just nor proportionate.
275. The Trustee addresses these issues by reference to the separate paragraphs of Issue 13.2. The questions in each paragraph are premised on the footing that the Exclusion Notices are voidable and liable to be set aside.

Issue 13.2(a): (i) can they only be avoided and set aside by the Court in relation to all or a group of the affected members collectively? (ii) And if so will the Court now avoid them and set them aside?

276. As to (i), the Trustee's position is that the Exclusion Notices are avoidable at each individual affected member's election rather than only liable to be avoided by the Court in relation to all or a group of Affected Members. I do not understand Mr Simmonds to argue for a different conclusion. Indeed, it is inherent in Holdings' own proposals that it is for each affected member to elect. In any case, the Trustee's position is, in my judgment, the correct one. Whatever may or may not be severable, I see no reason why the Court should not be able to treat each affected individual separately. The Imperial duty is, after all, a duty owed to each member just as the contractual duty of trust and confidence is owed to each employee. This is not a case where it is critical, or even important, to Holdings that all affected members are treated in the same way so there is no scope for an argument that it would be unfair and prejudicial to Holdings to avoid the Exclusion Notices in relation to some but not all of the affected members. It follows from the submission that I am not being invited actually to avoid the Exclusion Notices at all at this stage but only to determine that each affected member has the right to avoid vis à vis themselves. The answer to (i) is therefore "No" and to (ii) is that I am not being invited to set them aside. Whether, and if so how and to what extent, an affected member can actually elect to set aside the Exclusion Notice will appear from my answers to the other Issues.

Issue 13.2(b): (i) can they be avoided and set aside at each affected member's individual election? (ii) and what would they have to do or prove in order to do so?

273. Subject to Mr Simmonds' submissions that the Exclusion Notices should not be invalidated at all because of Holdings proposals designed to remedy any breach of duty, it must follow from what I have already said that an affected member is entitled to elect to set aside the Exclusion Notice to the extent discussed below. In strict terms, the position as I see it is that it is the Court which would set aside the Exclusion Notice but would do so only in relation to affected members who elect for that course. It is therefore open to me to impose conditions for the exercise of my power in terms of the giving of notice. That would include the timing of such an election. A similar question of timing relates to the connected issue of election discussed under Issue 19 at paragraph 417 below. For the same reasons as given in that paragraph, I propose to address the question of the timing of any election at a further hearing after the hand-down of this judgment. The answer to (i) of this issue is "Yes". I do not answer (ii) at this stage. My order subject to the caveat at the start of this paragraph) will be one which orders the setting aside of the Exclusion Notice in relation to the individual affected member on the making of the election: I do not need to be further involved at that stage.

Issue 13.2(c): are the Exclusion Notices liable to be avoided and set aside in some other manner, and if so in what manner?

277. Nobody suggests that the Exclusion Notices should be set aside other than following an individual election nor that written notice is unnecessary. The answer to this issue is "No".

Issue 13.2(d): may any power to avoid or set aside the Exclusion Notices be exercised independently of any such power in relation to the NPAs (so that, for example, the NPAs are avoided but the exclusion is affirmed)?

278. It is common ground that this Issue should be answered "Yes".

Issue 13.3: were the Exclusion Notices valid?

279. The answer to this issue, in the light of the above, is "No". That leaves open the answer to Mr Simmonds' submission that they should be validated.

Issues 14.1 and 14.2

280. These issues ask a number of questions on the footing that Exclusion Notices are void or unenforceable. They do not arise. I do not propose to address them.

Issue 14.3: this is set out at paragraph 262 above.

(a) Granting relief on terms

281. The Trustee's position is that, for an affected member who makes the election to have the Exclusion Notices set aside, I should not impose any term that the Notices should be treated as taking effect from some particular date after 6 April 2011. The Trustee's fall-back position is that if such a term is imposed, the relevant date should be (i) the date on which I hand down this judgment or (ii) if that is rejected, 6 April 2014 or (iii) if both of those are rejected, such other date as late as possible but before 31 March 2014 (but falling on the day immediately following the end of a Pay Month, a date chosen for purely practical and pragmatic reasons).
282. The answer to the debate about the imposition of terms is also informed by consideration of the consultation issue, that is to say the failure of Holdings to consult properly in relation to the Project Waltz proposals and which I have held to be a breach of the contractual duty of trust and confidence. The Exclusion Notices were served in the context of proposals on which there should have been, but there was not, a proper consultation so that the lack of consultation was not just a failure to observe the relevant Regulations. It is one thing to allow the exercise of a power to stand, in whole or in part, where there has simply been an inadequate consultation; it would be quite another thing to allow such an exercise to stand where a consultation was carried out which was not merely inadequate but constituted a breach of the duty of trust and confidence. Mr Simmonds says that all that is needed to remedy the breach of duty resulting from the service of the Exclusion Notices is to treat them as taking effect from some later date. That fails to recognise the breach of duty arising from the lack of proper consultation. I return to this theme later.
283. At Judgment [1578] I made the point that the failure to consult would, in some respects, have also given rise to a breach of the Regulations commenting that this added nothing to the RBs' claim based on breach of contract and that, in any case, breach of the Regulations did not give them a remedy in this Court. Mr Simmonds' response to that was that the only remedy for breach of the Regulations was that provided in the Regulations themselves so that there could not be any parallel contractual claim. I do not agree with that, for two reasons. The first reason is that I do not see the claim under the Regulations as precluding a contractual claim. Suppose that Holdings had entered into an express contract with an affected member to comply with its obligations under the Regulations. I see no reason why a breach of the Regulations should preclude the member from obtaining a contractual remedy including the recovery of damages if any could be shown. Regulation 18 does not, in my judgment, have that effect but is there only to make clear that the breach of statutory duty carries only the statutory sanction. The second reason is that it is not the mere breach of the Regulations which has given rise to the breach of duty; the facts giving rise to the breach of the Regulations are part of, but not all of, the facts giving rise to the contractual claim which can, in my judgment, be asserted by the RBs in parallel with the breach of the Regulations and is a claim in respect of which a remedy can be provided quite apart from the remedy under the Regulations. I enlarge on this when considering the consultation breach under Issue 29.
284. The Trustee contends that the approach to "voidability" should be the same as the approach which courts have adopted where a transaction is procured by misrepresentation, undue influence or fraud. Mr Simmonds' analogy with equitable remedies in other situations (eg claims for injunctions, rectification) is therefore inapposite. Mr Evans relies on what Lloyd LJ said in Pitt v Holt in the Court of Appeal, referring to [99] and [100], paragraphs on which Mr Simmonds also relies. He accepts that where a court exercises a discretion to avoid an act or rescind a transaction, it may impose terms as a condition of granting relief. But this is not, he contends, a licence for the court to rewrite the parties' bargain or the impugned act (ie it is no licence to rewrite the Exclusion Notices) or to impose whatever terms it thinks fit in order to achieve a result that is, at some abstract level, "fair". Mr Evans nonetheless does accept that the specific aim of rescission or avoidance is to undo the relevant act or transaction and "thereby" (his word) put the parties back into the position in which they were or would have been absent that act or transaction.
285. Mr Evans has provided detailed submissions concerning rescission in equity, citing passages from O'Sullivan, Elliot & Zakrzewski The Law of Rescission (2008) at 19.15 and Snell's Equity (32^nd ed) at 15-014 each to the effect that the discretion to impose terms is designed to ensure substantial restitution and is not to be used to reform the transaction or otherwise to achieve a result that the court considers fair.
286. As to the authorities, Mr Evans has referred me to TSB v Camfield [1995] 1 WLR 430 in which Nourse LJ rejected the approach of Robert Walker QC (sitting as a judge of this division) in Bank Melli Iran v Samadi-Rad [1995] 1 FCR 465 at p 477. Mr Walker had said this:

"If a wife has been induced to create security by the misrepresentation that it is good for £60,000…… and she claims relief against the mortgagee, equity can compel her, as a condition of relief, to recognise the security as good for the limited sum."

Mr Walker's decision was itself reversed by the Court of Appeal (see [1995] 3 FCR 735), in reliance on Camfield. And Camfield was again followed at Court of Appeal level in Castle Phillips Finance v Piddington (1995) 70 P&CR 592. Mr Evans has also referred me to the decision of Colman J in Molestina v Ponton [2001] CLC 1412 where the cases are reviewed. It is worth setting out his conclusion that it is clear as a matter of English law:

"that the principle that there cannot be partial rescission is part of the wider requirement that there cannot be rescission unless there can be restitutio in integrum. Further, that requirement is the conceptual consequence of the basic nature of the remedy of rescission which is to discharge all the parties from the bargain into which the misrepresentor has induced them to enter. It is not and never has had the function of providing compensation for the misrepresentation or some hybrid solution to reflect what would be fair between the parties having regard to the nature of the representation and the extent to which one party has been misled by another. Consistently with that, the court has no power to create a new bargain for the parties. What has been induced is the original bargain and it is the purpose of the remedy to return the parties to their position before that particular bargain was made. There is therefore no room for any form of equitable engineering directed to re-constructing the fabric of the original contract… "

287. In the present case, the imposition of a term giving effect to the Exclusion Notices from some date after 6 April 2011 would not, on Mr Evans' case, be in order to achieve any sort of restitution or anything like it, which should be restored by the affected members were the Exclusion Notices to be otherwise wholly avoided. As he puts it, the aim of avoidance, like rescission, is to undo the relevant act or transaction and not to rewrite it. But that is precisely what I would be doing if I acceded to Mr Simmonds' submissions.
288. Next, it is said that the Court cannot be satisfied that Holdings would have served the Exclusion Notices with effect from a later date had it complied with its Imperial duty. The earliest possible date for such a notice would have been the Longstop Date. I did not, in the Judgment, identify the Longstop Date (other than that it was after 6 April 2011 and, implicitly, at least as long after that date as to render the disappointment of the Reasonable Expectations before that date a breach of the Imperial duty). Nor did I make any finding that Holdings would, or would not, have served Exclusion Notices on or about 28 May 2010 (when the Exclusion Notices were in fact served) or at any time thereafter taking effect at a permissible future date. The facts established at trial do not, Mr Evans submits, justify a finding that Holdings would have served valid Exclusion Notices.
289. Mr Evans' next argument relates to the defects in the consultation which Holdings carried out in relation to Project Waltz. He submits that the Exclusion Notices were in their entirety "infected" (as he puts it) by the breach of duty in connection with the consultation. To allow the Exclusion Notices to take effect from a later date would amount to the Court condoning the breach.
290. Mr Simmonds does not accept the argument that the Exclusion Notices must stand or fall in their entirety. He contends that the Court is able to craft a just and proportionate remedy. He says that the proposals put forward by Holdings provide me with the just and proportionate approach which I should adopt.
291. The cases relied on by Mr Evans are, he says, ones where there was a bilateral transaction – a contract – and where the court found itself unable to rewrite the parties' bargain. In contrast, the present case, he says, is a case of unilateral transaction where Holdings was simply exercising a power vested in it and exercisable without the need for any consent. There was no transaction or bargain, with members or with the Trustee. Referring to O'Sullivan, Elliot & Zakrzewski, he cites a passage at 19.07: "Because it is concerned with bargains, the rule against partial rescission does not apply to gifts and other unilateral dispositions". Mr Simmonds accepts that gifts are obviously different from the Exclusion Notices; but he submits that the approach should be the same. Given that there is no bargain, the objection about rewriting the parties' bargain does not arise. As Mr Simmonds puts it, all the court has to do "is identify to what extent the unilateral transaction is valid". This chimes with the test in Bestrustees (as to which see below) the question being whether it is possible to identify conceptually two different elements of the exercise of the power. And so Mr Simmonds asks rhetorically why it should not be possible to impose terms on the parties to produce a comparable result to the one that would follow if the Exclusion Notices were valid.
292. In any case, he submits that Holdings had a unilateral right to terminate accrual from the Longstop Date because that would not interfere with any Reasonable Expectations. The Exclusion Notices should be set aside only partially, assuming that they should be set aside at all, to reflect that unilateral right and to give them validity to that extent. It is in the context of the partial avoidance/validation that Holdings makes its proposals which, in its submission, provide a just and proportionate remedy. If I do not accept what is proposed, Holdings is prepared to consider alternative proposals to meet any inadequacies in the proposals which I perceive.
293. Mr Evans makes two points in riposte. First, he says that the service of the Exclusion Notices is not to be equated in any way with a true unilateral transaction such as a gift. It has to be seen in the context of the mutual rights and obligations between Holdings and the members. The service of the notice varies the rights as between them – for instance it has an impact on the obligations of the members to pay contributions by bringing an end to their obligations to do so, and the service of the notice determines the quantum of the members' benefits. There is therefore a bilateral element to the Exclusion Notices. I do not consider that this riposte answers Mr Simmonds' point. The impact on members of the service of invalid notices is a factor to be taken into account in determining the terms on which they can be (wholly or partially) set aside. But the case remains, in my view, outside the realm of bilateral transactions relevant to rescission.
294. Mr Evans' second point is that the materials relied on by him do not relate only to bilateral transactions.

Discussion of arguments on voidability

295. It is important to make one preliminary point. It is that the issue for me is whether the Exclusion Notices themselves can be set aside in part (or whether, rather than setting them aside, I can accept undertakings from Holdings to implement its proposals). In answering that question, it will be necessary to examine what action Holdings could properly have taken at the time ie in May 2010. The issue is not whether, at some later time, Holdings could have served effective Exclusion Notices. It would not, in my judgment, be right to tailor the remedy for the breach of duty in serving the Exclusion Notices by taking account of subsequent events so as to justify a closure of the DB Plans to further accrual from some later date. If it is right to take account at all of what Holdings could properly have done, that matter must be judged as at May 2010 and not some later date.
296. Mr Simmonds' argument has a great deal of attraction producing as it does what, at least superficially, appears to be a fair result. However, ultimately I reject it for essentially two reasons. The first is that, in principle, I do not consider that I have the power, in the exercise of my discretion, to impose conditions on the avoidance of the Exclusion Notices to achieve their validity from the Longstop Date; and not having that power, I do not consider that I have power either to accept undertakings or other commitments from Holdings to achieve the same result. The second reason is that, even if that is wrong, I am not satisfied that Holdings' proposals would, in fact, provide a remedy for its breaches of its Imperial duty in the context of Project Waltz as a whole. I enlarge on those reasons in turn.

(i) power to achieve Holdings' proposals

297. I am unpersuaded by Mr Simmonds' analogy with the exercise by the court of other discretionary powers such as the grant of injunctive relief and the making of an order for rectification. Typically, cross-undertakings in damages are given when interim injunctions are granted: the price of obtaining the injunction is that, should it turn out that the injunction should not have been granted, the claimant will compensate the defendant for any loss occasioned. Cross-undertakings of that sort are not appropriate, of course, in the case of final injunctions. If the court does impose terms (or accept undertakings from a claimant) where a final injunction is granted, which must be rare, it could only be because the terms reflect the right which the claimant is seeking to enforce. In the present case, however, the "right" (if one is to continue with the analogy) which the RBs are seeking to enforce is the right to have the DB Plans properly implemented in accordance with their provisions. What Holdings seeks to do is to implement a change to the Plans which has not been properly implemented by giving partial effect to an otherwise invalid exercise of its powers.
298. As to conditions imposed on the making of an order for rectification, Mr Simmonds relies on Sargeant and Konica. In Sargeant, the deputy judge made it a condition of his order that Hilda (the person seeking rectification) return to her brother's estate a sum which she had received under the agreement before its rectification. It may very well have been the case that the estate would have had some sort of restitutionary claim; but even if it did not, the restoration of the money was within the scope of the sort of terms which could be imposed in cases of misrepresentation, undue influence or fraud to achieve restitutio in integrum. The additional concerns of the judge relating to the continued occupation of the relevant farmland by the executors indicate that he might have been prepared to impose other terms but, as I read what he says, those terms would have been directed at ensuring that the rights of the estate were not prejudiced. In Konica, the same deputy judge made an order rectifying the terms of a pension scheme as a result of which some beneficiaries would cease to be entitled to certain amounts under the unrectified provisions. As a matter of fact, it appeared that there were no beneficiaries who had actually received any sums which, under the rectified provisions, they would not be entitled to receive. But in case there were, the deputy judge indicated that he would provide in his order that any affected beneficiary could apply to the court to argue that the rectification order should have been made only on terms that the beneficiary did not need to refund any sums so received. It does not appear to me that the judge was intending to give such a beneficiary the opportunity to retain that which he was not entitled to retain. The beneficiary might, for instance, have a change of position defence to any claim by the pension scheme for return of the overpayment. The judge was saying no more, I think, than that a right such as that could properly be given effect to by imposing a term as part of the order for rectification.
299. Mr Simmonds also relies on the statement by Lord Walker in Pitt v Holt referred to at paragraph 268 above. Lord Walker made his remarks in a situation far removed from the present case. What the executors had done was to eliminate a concern of HMRC that the order setting aside the settlement would create uncertainty. There might, without the elimination of that concern, have been a defence to the claim to set aside the settlement. That is very different from the present case where the breach of duty has been established – there was no adequate defence – and where the search is now for the appropriate remedy for that breach.
300. On the other hand, the approach applied in cases of rescission for misrepresentation is not, I think, to be applied without a measure of flexibility which recognises the very different circumstance of misrepresentation (and indeed undue influence and fraud, the other matters to which Mr Evans refers by way of analogy) on the one hand and breach of the Imperial duty on the other. Having said that, I do not consider that it is open to the court to rewrite history by giving a wholly different effect to an act of the employer in order to remedy a breach of the Imperial duty. Subject to one caveat, it is not, in my judgment, open to me, as a matter of principle, to give effect to the Exclusion Notices by treating them as validly taking effect from some future date; nor is it open to me to refuse to exercise my discretion to set aside the Exclusion Notices on Holdings' undertaking to implement its proposals.
301. The caveat is this. Just as a transaction which is alleged to be altogether void can, in appropriate circumstances, be severed (as in Bestrustees plc v Stuart [2001] PLR 283 and Betafence Ltd v Veys [2006] EWHC 999, [2006] PLR 137) so as to separate the good, or valid, parts from the bad, or void, parts, so too it must, in my view, be the case that such severance is permissible where the transaction is potentially voidable rather than void. I can think of no rational argument for saying that severance is possible where a transaction is potentially void but not where it is only potentially voidable. The question then arises of whether the Exclusion Notices are severable. If they are severable, then a separate question may arise which is whether terms can be imposed when the severable part is in fact avoided. But if severance is not possible, I do not consider that the court can achieve severance, by the back door as it were, by rescinding the entire transaction but only on terms that it be given partial effect. If the Exclusion Notices cannot be severed, then they cannot be rescinded either even on terms that they take effect only from the Longstop Date.
302. The argument in favour of the view that severance is possible, it seems to me, has to be that the Exclusion Notices have a continuing temporal effect relating to the status of an affected member: the Exclusion Notice is invalid insofar as it purports to affect his status as a member up until the Longstop Date (or perhaps some other date) but is valid insofar as it purports to do so after that date.
303. I do not consider that the Exclusion Notices are severable in this way. In my judgment, the exercise of the power does not comprise two components which can be separated, conceptually, in the way that could be done in Bestrustes and Betafence. Instead, there was a single component, namely the change of status of the affected members on 6 April 2011. It is wholly artificial to view the Exclusion Notices as having a continuing temporal effect which can be divided into two components, one before and one after the Longstop Date (or some other date). Since severance is not possible, it follows from what I have said in the preceding paragraphs that there cannot be rescission on terms.

(ii) insufficient remedy

304. If I am wrong in my conclusions in relation to the extent of my discretion when setting aside the Exclusion Notices, I would consider that I should nonetheless set them aside in their entirety.
305. My principal reason for reaching this conclusion is that I am not satisfied that in May 2010 Holdings could (or would even if it could) have served the Exclusion Notices but specifying some later date consistent with the Reasonable Expectations concerning benefit accrual.
306. As to whether it could have served such a notice, I do not accept the Trustee's submission that Holdings' breach of duty was (merely) that it served the Exclusion Notices too early so that the Exclusion Powers could not permissibly have been exercised at all at the time when they were in fact exercised. Clearly an Exclusion Notice can specify a future date; I see no reason, as a matter of construction, to restrict that future date in any way and reject any suggestion to the contrary. However, whether Holdings could properly have made a decision in May 2010 that affected members should cease to be members at some far distant date is a different question which could only be answered after a close examination of the justification for such a decision.
307. Mr Evans made some submissions about the effect (the infection as he puts it) of the defects in the consultation which, as I have held, gave rise to a breach by Holdings of its contractual duty of trust and confidence. He is right to draw attention to those defects. In addressing the question of remedies it is not right, in my view, only to ask in abstract whether Holdings could have served valid Exclusion Notices in May 2010 taking effect at some future date. The issue is whether the actual Exclusion Notices which were served in the context of Project Waltz and in the context of a defective consultation can be allowed to stand even in part. In my judgment, they cannot be allowed to stand at all. And this is so even if, contrary to my view, the Exclusion Notices are otherwise severable: the failure to consult applies as much to the "good" element as to the "bad".
308. But even if that is wrong, I am not satisfied that Holdings would (and could) have served Exclusion Notices specifying the Longstop Date or some later date as the date from which the affected members would cease to be members. In this context, it seems to me that the burden is on Holdings to demonstrate that it would have served Exclusion Notices since it is Holdings which is relying on what it says would have been possible if it had complied with its duties. Further, it is not, in my judgment, enough for it to establish that it would have been able properly to serve Exclusion Notices; and then, in reliance by analogy on the general principal in damages claims to say that the RBs' remedies should be constrained by the minimum which Holdings could have done to comply with its duties. That general rule is not applicable to cases where an employer has failed properly to exercise a discretion in relation to bonuses or other benefits: see for instance Rutherford v Seymour Pierce Ltd [2010] EWHC 375 (QB), [2010] IRLR 606 at [33]. In such cases the court, in assessing damages, puts itself in the position of the employer and assesses what the likely level of bonus would have been. In the case of a binary decision, as in the present case between the service or non-service of an Exclusion Notice, the court needs to be persuaded at least on a balance of probabilities that Exclusion Notices would have been served. This is a matter I have already addressed at greater detail at paragraph 142 above.
309. My findings of fact at the trial are insufficient to support a finding that Holdings would, in May 2010, have served Exclusion Notices with an effective date on or after the Longstop Date. On the evidence which I received, I do not consider that I would then, or should now, reach the conclusion that it would have done so. The context in which it would have had to consider the question would have been one where it would have appreciated that Project Waltz as a whole would give rise to a breach of duty and would have appreciated also that the NPAs – all of them – were part of a series of transactions comprised in those breaches of duty. Holdings would have realised that, consistently with the Reasonable Expectations, it was unable to bring about a situation under which DB accrual between 2009 and the Longstop Date could be terminated (other than by, perhaps, the nuclear option of winding-up the DB Plans). Project Waltz was flawed; I cannot conclude, in the hypothetical scenario where that flawed project was not taking place, that this single element – the service of Exclusion Notices effective from the Longstop Date – would have been implemented.
310. Further, one vital element of Project Waltz was delivering the 2010 EPS targets imposed by CHQ. Booking the curtailment gains in the 2009 accounts was central to that object. It is accepted by Mr Evans that specifying a later effective date in the Exclusion Notices would still have allowed curtailment gains to be booked in the 2009 accounts. But there is a dispute about the impact of a change of effective date on the amount which could be booked. Mr Simmonds says that it would have been only slightly different.
311. Mr Evans has this to say about that submission. First, he says, correctly, that I made no findings about this. More importantly, the matter was not explored in any detail at the trial. I agree that it is not appropriate for me now to make detailed findings without the benefit of further expert advice which neither side has sought to adduce for the purposes of this hearing. Nonetheless, it is possible, I consider, to say that the amount might, and probably would, have been very different. The curtailment gain represents an acceleration, for accounting purposes, of the savings achieved by Soto. This was because the amortisation period for the prior service credit was brought forward as the result of the closure of the Plans. In effect, the part of the prior service credit which would otherwise have been spread over the years after closure is recognised in one year. In the present case, the amount accounted for in 2010 and 2011 before closure was unaffected, but the amount that would have been accounted for after closure was brought forward and it was recognised in 2009 on the footing that the Exclusion Notices were valid. Mr Evans has shown me some of the material which was in evidence, material emanating from IBM, which shows that for the Main Plan there would be a curtailment gain of £66 million in respect of the period from closure to the end of the third quarter of 2014 booked for 2009 on the assumption of accrual ceasing in 2011. However, as he points out, if the closure had taken effect from 31 March 2014 the only amount which would have been booked in 2009 was the residual amount properly referable to the period after that date (being the date of closure). So the curtailment gain booked in 2009 would have been very much smaller. I agree that postponing the closure to 31 March 2014 would have had a significant effect on the amount of the curtailment gain. This would have had an impact on the ability to deliver the 2010 EPS targets which were required by Project Waltz. That impact was significant.
312. Accordingly, there is a real doubt (to put it at its lowest) that an important element of Project Waltz was capable of delivery by the premature announcement of termination of DB accrual. In those circumstances, it is far from clear – certainly I am not satisfied – that Holdings would nonetheless have served any Exclusion Notices. It is more likely, it seems to me, that it would have worked up some alternative proposals and have waited to see how matters developed in the future. As it happens, things moved on quite a lot; in particular, in May 2010, IBM announced that non-operational retirement related costs would no longer form part of its EPS reporting and targets given to investors. And of course, much later, Holdings would have known that the Exclusion Powers could not break the final salary link.
313. At Judgment [1591] I said that:

"…it cannot possibly be said that the outcome of a new consultation would be a foregone conclusion. A consultation today might result in IBM deciding that the Project Waltz changes are no longer needed in order to fulfil business objectives".

314. In my view, adopting the same reasoning that has resulted in that conclusion, it cannot possibly be said that in May 2010, Holdings would have chosen to exercise the Exclusion Powers prospectively in advance of the Longstop Date.
315. It follows, in my judgment, that the Exclusion Notices are to be avoided in their entirety. I do not regard this as an unfair or unjust result. Holdings' appropriate course of conduct is not to seek to implement something different from that which it attempted to achieve (by giving partial effect to the Exclusion Notices) but is to implement the provisions of the DB Plans correctly. The answer to Issue 14.3 is that the Exclusion Notices should be set aside and that the Court should not grant such relief on terms.

Issue 14.4: Hybrid M Plan and Hybrid Deferred status

316. Issue 14.4 arises only if the Exclusion Notices are to be treated as taking effect from some date after 6 April 2011. I have held that they are not to be so treated. Accordingly, Issue 14.4 does not arise. However, the complications surrounding the Hybrid M Plan and Hybrid Deferred status arise in the context of Issue 15.3 and I discuss them there. In the context of Issue 14.4 it does not appear that there is much if anything between the parties because, if contrary to my finding, the Exclusion Notices can be partially set aside by being treated as taking effect from a future date, Holdings' proposals provide the affected member who elects to accept the proposals, with Hybrid Deferred status. It is common ground that Hybrid Deferred status would be subject to any necessary consents being given by the Trustee and subject to any subsequent events that would have caused cessation of such status in respect of any given member. There would, however, be a dispute between Holdings and the Trustee about the form of the Hybrid Deferred status in question, Holdings' case being that it means the form communicated to members in 2009 in particular in respect of early retirement. I deal with this further under Issue 15.3.

Issue 14.5: given my answer to Issue 14.3, Issue 14.5 is whether affected members are to be treated (a) as having continued in pensionable service under the DB Plans after 6 April 2011? And (b) as still being in such pensionable service at present (subject to any events which would have caused a cessation of pensionable service apart from service of the Exclusion Notices?

317. The answer to each of those questions is, in my judgment, "Yes". I do not understand Holdings to contend otherwise.
318. Coming to Issue 15, this raises questions (in the light of my answers to Issues 13 and 14) about Holdings' power to terminate future DB accrual.

Issue 15.1: Is Holdings' power to terminate such future pensionable service now inhibited by the Reasonable Expectations engendered by Project Ocean and Project Soto (and if so to what date)?

319. It is common ground that the Longstop Date is at latest 31 March 2014 (a date now in the past). The Trustee therefore accepts (and I agree) that Holdings' power is not now inhibited by the Reasonable Expectations. The answer to Issue 15.1 is therefore "No". The Trustee reserves the right to argue on appeal that the Longstop Date is later than 31 March 2014.

Issue 15.2: If Holdings wishes to effect termination of pensionable service under the DB Plans by excluding members from the DB Plans would it have to serve new exclusion notices?

320. On the basis, which I have held to be correct, that the Exclusion Notices are to be set aside in their entirety, it follows, in my judgment, that Holdings will have to serve fresh Exclusion Notices with prospective effect if it wishes to effect termination of pensionable service by excluding members from the DB Plans. The answer to Issue 15.2 is "Yes".

Issue 15.3: Would the members be entitled to Hybrid M Plan membership and Hybrid Deferred status as from the date of termination of pensionable service (subject to any necessary consents being given by the Trustee)? In particular, if Holdings were required to and did serve new exclusion notices in order to terminate pensionable service under the DB Plans, would it be obliged to offer all (or any, and if so what) elements of Hybrid Deferred status to the members in question (subject to any necessary consents being given by the Trustee)?

321. If, as I have held, the Exclusion Notices are to be avoided in their totality without the imposition of any terms, the Trustee's position is that the affected members would be entitled to Hybrid Deferred status and that Holdings must offer it to them. The whole issue of Hybrid Deferred status is addressed by the Trustee in a 61 paragraph Appendix to its closing written submissions. In spite of my own approach in dealing at length with nearly all of the submissions made (as shown by the length of the Judgment), even I have baulked at dealing with that Appendix in any detail in this judgment. I have, however, read it and take account of it in reaching my decisions. In summary the Trustee's case is:

i) The members have legal entitlement to Hybrid Deferred status as from the date of being excluded from the DB Plans. This entitlement arises from an unqualified and enforceable agreement reached between the Trustee and Holdings in December 2010 to execute a Deed of Amendment to create Hybrid Deferred status on the terms previously communicated between IBM and the Trustee. That agreement is binding and enforceable because:

a) it is a specifically enforceable contract pursuant to which Holdings is obliged to execute a Deed to give effect to Hybrid Deferred status;

b) (as a consequence of the first point), the agreement is effective as an amendment on the basis of the equitable maxim that "equity looks on that as done which ought to be done";

c) the agreement gave rise to an executory trust obliging IBM and the Trustee to execute a formal Deed of Amendment to give effect to Hybrid Deferred status.

ii) That agreement was not conditional or liable to be unravelled as a result of my findings in the Project Waltz proceedings. Hybrid Deferred status was not itself alleged to be a breach of the duty of good faith and thus there is no reason why the parties' 'unqualified' agreement to create Hybrid Deferred status was liable to be unravelled as a consequence of my other findings.

iii) Further, on the proper interpretation of the communications from Holdings to the Trustee and members in the summer and autumn of 2009, it is apparent that Hybrid Deferred status would apply from the date of closure of the DB Plans to accrual, whenever that is and even if new Exclusion Notices have to be served (rather than only on the basis that accrual ceased on 6 April 2011).

iv) As a further or alternative analysis, the members have a contractual claim against Holdings to compel it to provide Hybrid Deferred status on the terms it had agreed with the Trustee. This claim arises from the terms of the Pension Decision Tool (which had contractual force) issued to members on 10 December 2010 immediately after the agreement between the Trustee and IBM was reached, which included an obligation on IBM to take the necessary steps to create Hybrid Deferred status on the terms that it had agreed with the Trustee. This obligation is also reinforced by (or arises from) Holdings' Imperial duty or its implied contractual duty of trust and confidence which behoves it to take those steps.

322. Holdings, unsurprisingly, has a different view. Holdings' position is that there is no pre-existing right or entitlement on the part of members to Hybrid Deferred status. Holdings' offer of such status was conditional upon Project Waltz being implemented so as to terminate DB accrual from 5 April 2011. Given the findings in the Judgment, that condition has not been satisfied and, accordingly, the original offer falls away.
323. The one exception to this is that members who elect pursuant to paragraph 5.2 of the Pension Decision Tool to remain in their Chosen DC Arrangement are contractually entitled to Hybrid Deferred status. Ms Marsh's email of 10 December 2010 (which introduced the Pension Decision Tool), and page 4 of the Information Pack cross-referred to in it, each offered Hybrid Deferred benefits "on cessation of DB accrual on 5 April 2011".
324. Holdings is prepared, however, to offer Hybrid Deferred status as part of the terms set out above in the context of validation of the Exclusion Notices.
325. Returning to the point mentioned under Issue 14.4 about the terms of Hybrid status, Holdings relies in particular, with respect to early retirement, on Factsheet 3 which contained the following passages:

"Hybrid Deferred Members will be able to request early retirement on the terms available to active members and the new early retirement policy will apply. The new early retirement policy will be applied by the Company in deciding whether it will grant its consent to early retirement applications (and is subject to change at any time at the Company's sole discretion). This is likely to reduce the number of instances where consent is granted to early retirement.

The same early retirement provisions will apply to Hybrid Deferred members as those relating to active members and detailed in Factsheet 1, except that any reference to your Pensionable Service shall be a reference to your Pensionable Service accrued until the date of the C Plan closure. However, it should be noted that all applications for early retirement which require Company consent will be subject to the Company's new early retirement Policy in relation to applications made after the closure of the Early Retirement Window (see below) and the minimum age from which you can retire on these terms will be 55 from April 2010. If this potential modification is implemented the above will differ from the position set out in Factsheet 2 in that if you are able to early retire from Hybrid Deferred status you will remain eligible for the ERDFs applicable to active members but only if Company consent is granted to early retire (under the new policy)."

326. I recognised the fact that consent to early retirement from Hybrid Deferred status was intended by Holdings to be governed by the New ER Policy at Judgment [41 iv)(c)].
327. Mr Simmonds suggests that the Reasonable Expectations of members regarding early retirement relate to – and only to – retirement from active status: the Reasonable Expectations concerned Holdings' policy in respect of its power to give consent to early retirement in cases where consent was needed. This power – and, hence, any Reasonable Expectation in respect of it – was relevant only to retirement from active status. Once the Longstop Date has occurred and valid Exclusion Notices have taken effect, however, members will not be in active status and, accordingly, any Reasonable Expectations in relation to the early retirement policy will no longer be of relevance. Furthermore, Hybrid Deferred status itself has never been subject to any Reasonable Expectations on the part of members; it was a concession offered by IBM during the course of the Consultation. In particular, there was no separate Reasonable Expectation as to what Hybrid Deferred status would consist of. Therefore, if Hybrid Deferred status is to apply to affected members from some future date, it must be Hybrid Deferred status in the form that was communicated to members in 2009 (including, in particular, the application of the New ER Policy).
328. Furthermore, even if (contrary to those submissions above) the Trustee is correct in its assertion that Holdings is now somehow 'locked in' to providing Hybrid Deferred status generally by way of some contractual agreement, then that must also entail provision of the status in the terms agreed in that contract, ie Hybrid Deferred status in the manner communicated to members in 2009. It cannot plausibly be anything different.
329. It follows from the above analysis that if (contrary to Holdings' submissions) the Exclusion Notices are avoided in their totality, Hybrid Deferred status falls away entirely. If, in this context, Holdings were now to serve new Exclusion Notices (the Longstop Date having now passed), it would not be bound by any terms offered or concessions made the first time around and it would not be faced with any Reasonable Expectations on the part of members.
330. In my judgment, Mr Simmonds' submissions are correct and I reject the Trustee's case. For reasons given below, I do not consider that there would be any enforceable agreement between Holdings and the Trustee and/or the members which would entitle an affected member to membership of the Hybrid M Plan from the date of cessation of DB accrual following the service of valid Exclusion Notices. If that is wrong, then the terms on which an affected member would be entitled to benefit include the New ER policy.
331. The Appendix to the Trustee's written closing submissions which I have mentioned contains a detailed history of the events leading up to the agreement between Holdings and the Trustee to amend the Trust Deed and Rules to give effect to Hyrid Deferred status. I can pick it up at 22 October 2009 when there was a TMM at which the Trustee agreed to the necessary amendment subject to the Court confirming that IBM had not acted in breach of good faith and had met members' reasonable expectations. That, of course, did not give rise to a binding contract and, in any case, the condition was not fulfilled as I have held that there was a breach of the duty of good faith. The minutes of the TMM record Mr Ferrar as asking the Trustee to agree to two proposals: (i) the creation of a new category of member (hybrid deferred) "for all members who ceased accrual on 5 April 2011" and (ii) "the opportunity [my emphasis] to admit Members into M Plan for future service". It was recorded that the new, more stringent, early retirement policy from 6 April 2011 would be implemented. It was also recorded that Mr Lamb clarified that the Trustee was not prepared to introduce a hybrid deferred category or to allow Members to enter the M Plan without the direction of the Court. On the next day, Mr Newman wrote to Mr Ferguson to confirm the Trustee's position:

"Following advice from Leading Counsel and lengthy consideration of the issues, the Trustee has concluded that it is prepared to agree to amend the Trust Deed and Rules of each Plan to include provisions relating to "hybrid deferred members" and their membership of the M Plan on the terms set out in your letter and with effect from the date specified, provided that the Trustee has obtained confirmation from the Court that the Company has acted in good faith and that the future service proposals meet the legitimate expectations of those members.

The Trustee does not consider that it can give agreement in relation to any element of the Company's proposal in isolation of the other elements or without the sanction of the Court."

332. The letter referred to was one from Mr Ferguson setting out the terms of Holdings' proposals. He had written "to set out the changes the Company is prepared to make if the Trustee (a) agrees to implement them by amending the Plans and (b)…..". The changes included cessation of DB accrual on 5 April 2011. The requested amendments were clearly to give effect to the proposals, that is to say changes taking effect as from 5 April 2011. And, as Mr Ferrar made clear at the TMM, the amendments were to facilitate the proposals – there needed to be in place a structure pursuant to which Holdings would be able to implement its proposals (which, to repeat, included the New ER terms). The understanding, of course, was that Holdings would in fact implement the proposals: it would hardly have taken all the time and trouble which it had if that was not its firm intention. But there was no obligation, at that stage, to do so and, had there been a change of heart for some reason before service of the Exclusion Notices, nobody could have compelled Holdings to provide M Plan membership.
333. The Trustee's written closing suggest that Holdings renewed, in November 2010, "its requests that the Trustee agree to amend the DB Plans to confer Hybrid Deferred status on the excluded members (and to give them an option to join the M Plan section of the Main Plan)". That, I think, is not strictly accurate. The amendments (requested in Dickinson Dees' letter dated 2 November 2010) were not requested in order that the amendments should of themselves confer Hybrid Deferred status or to give the option; rather, the amendments were requested in order to provide the structure under which an affected members could achieve Hybrid Deferred status and under which effect could be given to such an option. The amendments were needed in order for Holdings to be able to offer an affected member the benefits of Hybrid Deferred status as from 6 April 2011.
334. Following that request, Holdings issued an announcement to members on 5 November 2010 stating that it wished to give them the option "in relation to their pension benefits in April 2011" of joining the M Plan and of having Hybrid Deferred status, but that this required the Trustee's agreement which it did not currently have. This communication was about Holdings wishing to have the ability to implement its proposals; it was clearly, in my view, about putting in place the necessary changes to the Trust Deed and Rules to enable this to be done.
335. Nabarro replied to Dickinson Dees' letter on 29 November 2010. The letter identified Holdings' request as being for:

"the Trustee to agree to:

1. allow the current DB Main Plan members to join the M Plan prior to judgment in the High Court proceedings; and

2. amend the Main Plan and the I Plan in order to provide hybrid deferred benefits."

336. The response was that the Trustee would "broadly speaking" be prepared to do this. But it did not want there to be any pre-judging of what the Court might rule and wanted to see put in place an interim arrangement which was capable of being unravelled. The letter contained the following:

"As for offering members admission to the M Plan and hybrid deferred status, the Trustee's view of the matter is as follows:

(1) If it turns out that Project Waltz is invalid, then DB accrual will have continued after 6· April 2011 and all Affected Members will have remained members of the Main Plan or the I Plan. Therefore questions of M Plan membership and hybrid deferred status will not have arisen.

(2) On the other hand, if it turns out that Project Waltz is valid, then as from 6 April 2011 the Affected Members will have been excluded from their respective Plans. In that event, the Trustee would have wished to give Affected Members the option to join the M Plan and to amend the Plans to confer hybrid deferred status. It appears to the Trustee that this would have been beneficial to the members (albeit less beneficial than continued DB accrual)."

337. On 2 December 2010, Dickinson Dees wrote again. The Trustee relies on various passages from that letter in support of its case that the eventual commitment by Holdings was that Hybrid Deferred status would be available, as of right, to any member whenever his DB accrual terminated as the result of the exercise of the Exclusion Powers. Two of the passages relied on are these:

"It remains our client's view that Main Plan members have the potential to suffer significant detriment if they cannot be offered those options which are dependent on Trustee consent, and that your client can, and should, now agree without qualification to:

1. permit current Main Plan DB Members entry to the M Plan; and

2. amend the Plans in order to make available the proposed Hybrid Deferred benefits."

and

"If we have not received your client's unqualified agreement to this proposal….. our client will be unable to offer current Main Plan members the option to join the M Plan on 6 April 2011, and will also be unable to confirm to members that all the Hybrid Deferred benefits will be available to them."

338. It is submitted that the "without qualification" and "unqualified agreement" requirement supports the Trustee's argument. I do not agree: one needs to ask what it was that Holdings wanted the Trustee to agree to without qualification, the answer to which was to give its consent to DB members entering the M Plan and to amend the Plans in order to make available Hybrid Deferred status. There was nothing in this that would oblige Holdings, as a matter of contractual obligation to the Trustee or to the RBs, actually to make to the affected members the offer which it was proposing to make. This conclusion is consistent with another passage relied on by the Trustee which in my view, far from supporting its argument, actually supports Holdings' case:

"Please confirm that your client is prepared to produce a first draft of the proposed Interim Deed of Amendment assuming agreement between our respective clients to allow M Plan entry on the terms of this letter, on the basis of the information provided previously about the nature of the Hybrid Deferred benefits (including in our letter dated 2 November 2010)."

This shows plainly, in my view, that what was contemplated was an amendment which would merely allow, and not actually produce, Hybrid Deferred membership for any given individual. Something else would need to be done actually to give rise to such membership within the amended structure which allowed for it. That "something else" would be (i) the offer of such status by Holdings and (ii) the election by the affected member to adopt one of the three offers which Holdings was proposing to make.

339. On 6 December 2010, Nabarro wrote to Dickinson Dees confirming the Trustee's unqualified agreement ie to permit current Main Plan DB Members entry to the M plan and to amend the plans in order to make available the proposed Hybrid Deferred benefits. I make the same points in relation to that as I have already made in relation to the earlier correspondence. The Trustee is not assisted, in my view, by the passage in that letter in which Nabarro notes an "apparent risk" that Holdings would withdraw the option of M Plan membership if the Trustee failed to give the agreement sought. It is not possible to derive from that the proposition that the Trustee was agreeing to the amendment only on the basis of a contractual commitment to admit the affected members to Hybrid Deferred status and to membership of the M Plan. But even if that is wrong, the contractual commitment was, in my judgment, only in the context of a valid termination of DB accrual on 6 April 2011 (although if that date had shifted by a few days or even weeks, no doubt Holdings would have continued with the proposals and not declined to make the offer which it was, at that stage in December 2010, intending to make). If there was a contractual commitment at all, it was only one which would have obliged Holdings to admit the affected members to Hybrid Deferred status and to M Plan membership as a consequence of the termination of the service of the Exclusion Notices contemplated by the proposals (and not, therefore, of some different and later Termination Notices served following the determination of the Court that the actual Exclusion Notices were to be set aside).
340. On 10 December 2010, Holdings circulated, by email, the Pension Decision Tool to all affected employees. Jane Marsh's covering letter included the following:

"I am pleased to Inform you that the Trustee has agreed that affected members of the Defined Benefit (DB) section of the Main Plan and I Plan ("affected members") will be granted Hybrid Deferred type benefits and will have the option to join the M plan, the Defined Contribution (DC) section of the Main plan, from 6th April 2011. This agreement means that affected members now have the full choice of pension benefit options which IBM wanted to provide, on cessation of DB accrual on 5th April 2011." (underlining in original)

341. The letter went on to say that Holdings "will offer affected members the following options in relation to their benefits from 6^th April 2011" and there followed the three options, the first of which was to "Join the M Plan".
342. There was a link to the Pension Decision Tool which included a section headed "IMPORTANT INFORMATION" which included the following:

"2. It is clear that the current court proceedings will not be resolved before the cessation of DB accrual. You are therefore choosing your future IBM pension arrangements that will be in place from 6th April 2011 until resolution of the court case. ….. [I]f the final court judgment means IBM has to reinstate DB accrual and you have chosen to join a DC Plan, any contributions paid into your DC Plan will be used to offset the cost of DB accrual in accordance with paragraph 5 below. Completing this form and joining a DC arrangement will have no influence on the outcome of the Court decision and it will not prevent you from continuing DB accrual should the Court decide that it is appropriate for you to do so."

343. Paragraph 5 explained the need to have legally binding agreements in place and stated:

"5. By choosing to join [the M Plan], you agree:

……

5.2 that if, following the final outcome of the court proceedings, you are entitled to be treated as having continued in Pensionable Service under your DB Plan (unless you choose to waive this entitlement and remain in your chosen DC Plan);

344. The slides accompanying the Pension Decision Tool explain Hybrid Deferred status in some detail. The Trustee relies on one slide which explains that "IBM will provide "Hybrid Deferred" benefits to members". There is no underlining of "will" on this occasion to support the idea that there was a contractual commitment to provide this status even if the Court ruled against Holdings on the validity of the proposals. What the Trustee's written closing submissions do not draw attention to is the heading of this part of the slide "From 6 April 2011: DB accrual will cease".
345. In my judgment, these communications do not establish any binding obligation on Holdings to provide M Plan membership or Hybrid Deferred status other than in the context of a valid termination of DB accrual on 6 April 2011. The contractual obligation to provide this membership and status to a particular individual would have arisen only after the member had elected, under the Pension Decision Tool for M Plan membership. That election would have taken place in the context of amendments effected, or to be effected, to enable such membership and status to be effected. If the Court did not give its approval, both the prior communications with the Trustee and the Pension Decision Tool (in particular paragraph 5.2) demonstrate that the member would be entitled, retrospectively, to DB accrual (unless that entitlement were waived). It is, in my view, necessarily implicit that, if the Court ruled against validity, all bets were off and that matters would revert to the status quo ante.
346. It is no answer to that to say that the creation of Hybrid Deferred status and admission to M Plan membership were not breaches of any duty on the part of Holdings, although they certainly formed part of the mechanics of implementation of the Project Waltz proposals. The reason it is no answer is because the arrangement clearly contemplated (subject to waiver of the entitlement) that members would be entitled to be treated as having continuing Pensionable Service under the DB Plans. They could not, at the same time, accrue benefits under the M Plan or, even if they could, they were clearly not intended to do so. The continuing accrual of DB benefits meant that the affected members simply remained as members of the DB Plans as though the Exclusion Notices had not been served: they could not, with that status, also have the status of Hybrid Deferred members.
347. In these circumstances, it is impossible to reach the conclusion, in my judgment, that if and when the Exclusions Powers are exercised in the future – not as part of Project Waltz but as a separate exercise and after proper consideration of the exercise of those powers by Holdings – the then affected members have an entitlement to Hybrid Deferred status or to membership of the M Plan.
348. If that is wrong, then I consider that the terms on which an affected member would be entitled to benefit include the New ER Policy. This is a short point. Holdings' position is, in effect, that it is clear that Hybrid Deferred status was only ever intended by Holdings to be on the basis of the New ER Policy and, if there was any sort of binding contract at all, it can only have been on terms that Holdings would be able to exercise its powers in relation to members with Hybrid Deferred status in accordance with the New ER Policy.
349. The Trustee's position is that there was only one policy at any given time. The New ER Policy was intended by Holdings to apply equally to those retiring from active service and to persons retiring from Hybrid Deferred status after 6 April 2011. The adoption of the New ER Policy as early as 6 April 2011 was a change, or was part of a wider set of changes, which gave rise to a disappointment of Reasonable Expectations and a breach of Holdings' Imperial duty. The New ER Policy was accordingly invalid: it is as much invalid in relation to Hybrid Deferred members as in relation to active members.
350. I reject the Trustee's submissions on this point. It is clear, in my view, that the communications with the Trustee and the Pension Decision Tool make clear that Hybrid Deferred status is being conferred on the basis that the New ER Policy would be adopted in relation to the Hybrid Deferred members. There is nothing wrong, in principle, in Holdings adopting different ER policies in relation to different groups of people, provided that the difference is not illegally discriminatory or irrational or perverse. Holdings could, therefore, have decided to retain the Old ER Policy for those retiring from active service, but have expressly adopted a different policy in relation to those retiring from Hybrid Deferred status. I see no reason why that "different policy" should not have been the New ER Policy. Accordingly, if Holdings were now to serve valid Exclusion Notices taking effect from the Longstop Date (or some later date) and assuming (contrary to my finding) that the members affected by such Notices were entitled to Hybrid Deferred status, there is nothing to prevent Holdings from applying the New ER Policy to them.
351. In any case, the end date for the Reasonable Expectation concerning early retirement has now passed. Holdings is now able to adopt a new ER policy even in relation to members retiring from active service; it must be able to do so in relation to members retiring from Hybrid Deferred status. Unless there is some reason why the adoption of the New ER Policy would now give rise to a breach of the Imperial duty, that is one policy which could properly be adopted. I know of no reason why it would give rise to such a breach.
352. A point has been raised, of narrow compass, concerning one of the terms of Hybrid Deferred status. It relates to the revaluation "underpin" applicable to Hybrid Deferred benefits. This is dealt with under Issue 30. It does not make any difference to the answer to Issue 15.3 given the result of my analysis.
353. The answer to Issue 15.3 is that the affected members would not be entitled to Hybrid M Plan membership or Hybrid Deferred status as from the date of termination of pensionable service.

Issue 15.4: must IBM carry out a further consultation before the termination of pensionable service under the DB Plans?

354. The Trustee contends that further consultation before termination is required because:

i) The termination of accrual at a future date would be a change to the DB Plans which required consultation in accordance with Regulations 6(1), 7(3) and 8(1)(c) of the Consultation Regulations (the relevant parts of which are set out in Appendix E to the Judgment).

ii) No consultation has been carried out in respect of that change.

iii) In the circumstances, in particular the conduct of IBM's previous consultation process, a failure to consult would constitute a further breach of the contractual duty of trust and confidence.

355. Holdings' position is that I have no jurisdiction to order it to consult further and that there is no reason for it to do so in any event. Issue 15.4 does not, however, ask me to order that Holdings actually carry out a further consultation. It simply asks whether Holdings must do so before termination of DB accrual. I suppose one could quibble about what "must" means in Issue 15.4. For my part I consider that, if a statutory regulation provides that an employer may not take certain action unless he has consulted about it, it would be entirely apposite to say the employer "must" consult before taking that action. In the present case, the relevant Regulations of the Consultation Regulations (mentioned above) provide that an employer may not make a listed change unless there has been the consultation required by the Consultation Regulations. Clearly a termination of DB accrual would be a listed change; in my judgment the Regulations require Holdings to carry out the statutory consultation before it serves any further Exclusion Notice.
356. The only remedies for a failure to consult in accordance with requirements of the Consultation Regulations are (exclusively) those provided for in Regulation 18. That is why Holdings submits that I have no jurisdiction to order it to comply. I make two observations about that:

i) First, I sincerely hope that the point is academic. I am entitled to expect that a responsible and respectable UK corporation, part of a group with a global presence, will comply with the UK statutory provisions by which it is bound, provisions designed to protect the interests of members (in particular of its UK workforce) of a scheme and which are in need of precisely the protection which consultation might provide. I add, in parentheses as it were, that a deliberate failure to comply with those obligations notwithstanding the Judgment and the present judgment might itself be seen as a breach of one or both of the implied duty of trust and confidence and the Imperial duty.

ii) Secondly, I do not accept that this court has no jurisdiction to grant injunctive relief to prevent the service of an Exclusion Notice unless and until the consultation process has been complied with. It is one thing to say that I cannot grant a remedy for an actual breach; it is quite another to say that I cannot prevent a breach. I will return to this when considering consultation under Issue 29.

357. Adopting the ordinary sense of the word "must" which I have described above, Issue 15.4 is to be answered "Yes". Holdings "must" carry out further consultation before such termination.
358. However, if I am being asked whether Holdings "must" carry out a consultation in the sense that I can order it to do so (and ignoring for the moment the question of whether I can grant injunctive relief such as that just mentioned), it has to be accepted that I cannot do so, if Holdings' only obligation is to consult in accordance with Regulation 18. However, if there is a wider obligation, I may be able to compel compliance with that obligation (or at least to answer Issue 15.4 in the affirmative adopting that approach to "must").
359. As to such a wider obligation, Mr Simmonds submits the breach of duty which I held to exist in relation to the actual consultation was a breach of the contractual duty of trust and confidence, not of the Imperial duty: there is nothing in that duty which requires an employer to consult. In that context, he refers to Judgment [1554] and [1555]. I agree with that submission, but only up to a point. I accept that there is no general contractual duty to consult so that generally the failure to do so would not give rise to a breach of any duty of trust and confidence. But there may be cases where the fact of a failure to consult destroys or seriously damages the relationship of trust and confidence between the employer and the employee. In such a case, there is a breach of the implied contractual term. I do not consider, in such a case, that it would be correct to say the term which is ordinarily to be implied is excluded in relation to the failure to consult because the only remedy is that provided by the Regulations.
360. Suppose, for instance, that an employer had told its workforce that it was considering whether and how it might reduce its pension costs, but gave clear and repeated assurances that before deciding on any changes, it would carry out the consultation required by the Regulations. It seems to me that the employer might well be in breach of its implied duty of trust and confidence if it then failed to consult but nonetheless implemented changes to the disadvantage of the members. Now it may well be that Mr Simmonds is correct to say that the only remedies for that breach are those laid down in Regulation 18. But that does not mean that there has been no breach of the contractual duty: it simply means that that breach does not entitle the employee concerned to any remedy other than that laid down by the Regulations. In such a case, if one asks whether the employer "must" consult, the answer must surely be "Yes" otherwise not only will it be in breach of its statutory obligations, but also in breach of contract. Further, the case for injunctive relief in these circumstances would be stronger: such relief would not only prevent a breach of statutory duty but would also prevent a breach of contract.
361. A related, but separate, issue is whether the fact of the established breach of duty gives rise to an obligation to re-consult before any alternative proposals are decided upon. Mr Simmonds submits that there is nothing in the Consultation Regulations that confers upon the Court the jurisdiction to require an employer to re-consult. Regulation 18 sets out the exhaustive list of remedies which do not include any remedy that involves an employer's being ordered by the Court to carry out a process of consultation. I do not disagree with that. But it is not an answer to the point that Holdings needs to consult about what is (jumping ahead: see for further analysis the immediately following paragraphs of this judgment) an entirely new proposal, namely to serve Exclusion Notices in the future, some years after the original Project Waltz proposals and against an entirely different financial and economic backdrop.
362. Mr Simmonds might say that my comments in paragraph 353(i) above are inapposite. He submits that there is no need to carry out a further consultation and so could say that a failure to do so would not be a matter for any real criticism. Holdings' position is this:

i) It has already carried out a consultation exercise in relation to termination of pensionable service. That consultation resulted in a decision that such termination would occur in 2011.

ii) If new exclusion notices have to be served, the earliest date on which pensionable service could now terminate would be in 2014. The members would in that event have received at least three more years of pensionable service than would have been the case on the basis of the earlier consultation.

iii) This would not be a new proposal, giving rise to a fresh requirement to consult in accordance with the Consultation Regulations, but rather the implementation of the existing proposal on which a consultation has already been carried out but on terms that are now more generous to members in terms of the date of termination than those suggested at the time of the consultation.

iv) There is no need to consult on something more advantageous to members (as the current position is compared to what was consulted upon before the service of the Exclusion Notices). It would make no sense to require an employer to consult on the product of a consultation if that product were more advantageous to members than the original proposal. It is an outcome that, in its nature, is contemplated by the consultation process. Moreover, it could make no sense as a matter of policy to require employers to re-consult on concessions made during a consultation process. The resulting delay would act as a disincentive to making such concessions.

363. I do not accept that argument (although various points made in it are correct). The consultation which Holdings carried out was, in fact, defective for the reasons given in the Judgment. It was the inadequacy of the consultation which gave rise to a breach of duty, not simply the fact that the proposed date for termination of DB accrual was too early. Holdings cannot now rely on what was then inadequate as now adequate.
364. But quite apart from that, there are two other reasons why the earlier consultation cannot be relied on. The first is this. I do not accept the submission that new Exclusion Notices would not be a new proposal. It is true that part of the previous consultation was whether pensionable DB service should be terminated. But there has been no consultation about whether that termination should take place in the very different economic and financial circumstances of 2014 as compared with 2010. Moreover, there has been no consultation in the context of the change in IBM's EPS reporting to investors. Further, in the light of the Judgment, it is to be hoped that Holdings might view with slightly more independence from IBM Corporation than it did previously proposals emanating from Armonk with which it might not agree from a UK operating perspective. I draw attention to Judgment [1590] in this context.
365. The second reason is that the previous consultation was not a consultation simply about termination of DB accrual. It was a wider consultation about Project Waltz and it was only within that context that the termination of DB accrual was to take place. A new proposal to terminate DB accrual will not be in the context of Project Waltz. Whether there will be a free-standing proposal to end DB accrual by exercise of the Exclusion Powers, or whether that will be done in the context of a wider proposal, perhaps similar to Project Waltz, I simply do not know. But one thing which can, in my judgment, be said is that whatever is proposed hereafter will be a new proposal. It will not be possible to rely on the original consultation as an adequate consultation in respect of any component of such a new proposal.
366. The answer to Issue 15.4 is therefore that Holdings must carry out a further consultation before terminating DB accrual.

Issue 16: Having regard to the answers to Issues 13 to 15, are members in principle entitled to damages or equitable compensation from Holdings as a result of the service of the Exclusion Notices, or to some other remedy and if so what remedy?

367. Although this issue is raised in the context of the Exclusion Notices, it is not only in that context which the issue whether an award of damages or equitable compensation can be made is raised. I shall address the questions raised by the parties as a matter of principle in the context of the Exclusion Notices and then address later how the principle is to be applied in the different contexts of the ER Window and New ER Policy.
368. I have so far addressed the Exclusion Notices only in the context of Issues 13 to 15 where the focus has been on breach of the Imperial duty and the extent to which breach of that duty entitles affected members to have the Exclusion Notices set aside. There are, however, two related questions concerning the Exclusion Notices. The first is whether purely financial compensation can be awarded for breach of the Imperial duty: in other words, could an affected member affirm his status as a Hybrid Deferred member and member of the M Plan but claim financial compensation to reflect the extent to which his position is thereby worse than it would have been had the Exclusion Notices not been served. The second is whether a remedy by way of contractual damages can be given in relation to an element of Project Waltz which, by itself, did not give rise to a breach of contractual duty but only to a breach of the Imperial duty (eg service of the Exclusion Notices).
369. Holdings' position is that affected members are not entitled, even as a matter of principle, to damages, equitable compensation or any other financial remedy as a result of the service of the Exclusion Notices. Mr Simmonds accepts (subject of course to any appeal) that service of the Exclusion Notices gave rise to a breach of the Imperial duty. But this was not a breach of contract sounding in damages: there is no free-standing financial claim for breach of the Imperial duty even by a person who might be able to prove that a particular breach of the duty has caused him loss.
370. In the Judgment, I drew the distinction between the two duties, and their different content at various points in the Judgment: see for instance at [353], [354], [395], [397] and [421]. Mr Simmonds submits that the service of the Exclusion Notices (as well as the change in ER policy) were breaches of the Imperial duty but not of the contractual duty. That submission raises, I think, the question of how one is to view the Exclusion Notices. If one views them in isolation (under Scenario A ie in the context of Reasonable Expectations), they gave rise to a breach of the Imperial duty. I did not expressly deal with the question whether the Exclusion Notices, viewed in isolation, also gave rise to a breach of the contractual duty.
371. As to that, I have been referred to Imperial Tobacco at p 597 where, having identified the implied contractual term, Sir Nicolas Browne-Wilkinson V-C said:

"I will call this implied term "the implied obligation of good faith." In my judgment, that obligation of an employer applies as much to the exercise of his rights and powers under a pension scheme as they do to the other rights and powers of an employer. Say, in purported exercise of its right to give or withhold consent, the company were to say, capriciously that it would consent to an increase in the pension benefits of members of union A but not of the members of union B. In my judgment, the members of union B would have a good claim in contract for breach of the implied obligation of good faith: see Mihlensted v. Barclays Bank International Ltd.[1989] I.R.L.R. 522,525,531, paras. 12,64 and 70.

In my judgment, it is not necessary to found such a claim in contract alone. Construed against the background of the contract of employment, in my judgment the pension trust deed and rules themselves are to be taken as being impliedly subject to the limitation that the rights and powers of the company can only be exercised in accordance with the implied obligation of good faith…."

372. He then referred to Mihlenstedt again, pointing out that in that case it was not necessary to decide whether the employee's rights rested in contract alone or could be enforced under the trust deed, since the case could be decided in contract alone. After referring to the judgment of all three judges, he concluded that the case was not an authority which drove him to the conclusion that there could be no implied limitation under the trust deed.
373. In the light of that, the Trustee's case is that the contractual duty and the Imperial duty are capable of existing in parallel. Whilst Mr Simmonds accepts that the Vice-Chancellor "appears to be saying that there are contractual and trust claims in parallel" he submits that, when one examines the reasoning, the implied contractual term does not apply where the Imperial duty applies. He does not put it quite that way but it is, I think, the effect of his submission.
374. Mr Simmonds submits that Nourse LJ's reasoning starts from the premise that there was no trust law obligation on the bank. If a remedy was to be found, it could be found only in the contract of employment. He (Mr Simmonds) then goes on to say that the premise is wrong, since Imperial Tobacco has established a trust law obligation. He points out that Nourse LJ based his decision on an implied term rather different from the implied contractual term of trust and confidence. In contrast, Nicholls LJ did refer to something very similar to the implied duty of trust and confidence, saying at [64] of his judgment that it was necessarily implicit in the contract of employment that the bank agrees that it will discharge its functions under the trust deed in good faith. But why, Mr Simmonds asks, is it a necessary implication if there is an independent fetter in the trust deed?
375. I see the force of the point behind that question. There are, however, two reasons why these submissions are not an answer to the Trustee's case that the duties run in parallel. The first is that the remedy for the Imperial duty may not provide the remedy which justice demands. If Mr Simmonds is right in his submission that there is no financial remedy available to a member for breach of the Imperial duty, then the implied contractual term is needed to enable him to obtain a remedy for conduct on the part of the employer which destroys or seriously damages the relationship of trust and confidence. Indeed, on one view, the implied contractual term is more obviously appropriate given the Imperial duty than it would be without such a duty. Mr Simmonds' submission turns on the need to imply a term. But the implication of a term is not to be judged exclusively by necessity. Rather, the test is that which I have discussed at length earlier in this judgment in the context of the disputed term: see paragraph 91ff above. In applying the test, it is helpful to ask whether the term would go without saying. Surely, if Holdings and a member had been asked, they would have agreed that it went without saying that Holdings would exercise its powers in accordance with the limitation to which the powers are subject.
376. The second reason is that it I do not think that it is open to me (or if it is I do not consider that I should) to adopt Mr Simmonds' approach. It would require me to reject the reasoning in Milhenstedt which I think only the Court of Appeal should do. Even if Nourse LJ's decision can be side-lined on the basis that his focus was not on the implied obligation of trust and confidence but on a different implied term, Nicholls LJ clearly rests his decision in contract. He does so even though there might be a trust claim (the point which he left open) and must, therefore, have considered it possible to imply a term into the contract of employment even if there were a limitation on the exercise of the discretion.
377. Accordingly, I conclude that I must proceed on the basis that the contractual duty and the Imperial duty can co-exist, but I make two observations:

i) First, I have held in the Judgment that the test to be applied in cases of an alleged improper exercise of a power conferred on an employer in a pension scheme is the test of irrationality or perversity. If the alleged improper exercise of the power is also relied on as giving rise to a breach of the contractual duty, the same test should be applied. A claimant should not be able to succeed in contract by reference to a lesser test than applies to the Imperial duty. It would be unsatisfactory to have two different tests applying to a single set of facts.

ii) Secondly, a great deal of care must be taken if a person seeks cumulative remedies relying on the same acts as giving rise to a breach of each duty. There is a danger of a person being over-compensated if he is allowed to cherry pick. If it is possible to obtain financial compensation for breach of the Imperial duty then it may be possible to obtain a combination of remedies eg setting aside plus a measure of compensation. But if a person seeks to obtain both equitable remedies in respect of breach of the Imperial duty and common law remedies in respect of the breach of contract, matters are more complex. These are matters I come to in a moment.

378. It follows from that analysis that the service of the Exclusion Notices, viewed in isolation (ie under Scenario A for this purpose) gave rise to a breach of both duties. This is because I have already held that, viewed in isolation, there was a breach of the Imperial duty: it follows that there has been a breach of the contractual duty since the strict test of irrationality or perversity applicable to the Imperial duty was fulfilled and that, clearly, was sufficient to fulfil any test applicable to the contractual duty.
379. Quite apart from that, I held in the Judgment that Project Waltz overall gave rise to breaches of both the contractual duty and the Imperial duty. I did not say that each objectionable element gave rise to a breach of both duties. Indeed, conceptually it is not possible to say that an aspect of Project Waltz which did not involve the exercise of a power under the DB Plans (eg the NPAs) involves a breach of the Imperial duty. I consider that all one can properly say, in the context of Project Waltz as a whole, is that certain actions gave rise to a breach of contract, other actions gave rise to a breach of the Imperial duty and other actions may have given rise to a breach of both duties. The context of Project Waltz is important because it is only in that context which one can judge whether a particular action breaches whichever of the duties is under consideration. Thus an action which may not be a breach of duty viewed in isolation may still give rise to a breach in the context of Project Waltz overall and that is what I have meant by describing Project Waltz overall as giving rise to breaches of both duties.
380. It follows from this analysis that whether any particular act (eg the service of Exclusion Notices) gives rise to any breach must be judged in the context of Project Waltz as a whole – in that context, did the particular act destroy or seriously damage the relationship of trust and confidence – but the categorisation of the breach to which the particular act gives rise is to be determined by the focus of the act in question.
381. So far as concerns the Exclusion Notices, viewed in the context of Project Waltz as a whole, the service of them gave rise to a breach of the Imperial duty; and since a parallel duty existed in contract – all the affected members were active members – service of them also gave rise to a breach of the contractual duty.
382. The upshot of all this is that an affected member can seek a remedy to reflect service of the Exclusion Notices. He can seek remedies for breach of the Imperial duty (in particular setting aside and/or equitable compensation if available) or he can seek remedies for breach of contract (typically damages). Whether he can ever combine the two I leave for the moment.

Equitable compensation or damages for breach of the Imperial duty

383. The issue of parallel duties is of significance principally, if not exclusively, in relation to remedies. If the same, or a similar, financial remedy is available in relation to breach of the Imperial duty as is available for breach of contract, the issue ceases to have significance. The Trustee's case is that, even if the service of the Exclusion Notices gave rise only to a breach of the Imperial duty, it is nonetheless possible for an affected member to obtain a financial remedy. Either damages could be awarded under Lord Cairns' Act or equitable compensation could be awarded for breach of equitable duty.
384. It is argued that conduct on the part of an employer capable in principle of giving rise to a breach of the Imperial duty is not confined to conduct in relation to the exercise of a power vested in the employer or for the exercise of which its consent is necessary. But where the conduct does relate to the exercise of a power by the employer, the duty, as expressed by Sir Nicolas Browne-Wilkinson V-C, finds effect as a limitation on the exercise by the employer of that power: it can only be exercised in accordance with the implied obligation of trust and confidence. As Mr Simmonds submits, the effect of the Imperial duty is to impose a fetter on the exercise of the power. Breach of the Imperial duty in relation to the exercise of a power might render the exercise voidable but it does not follow, he says, that such a breach gives rise to any claim for damages or equitable compensation.
385. Mr Evans submits that the relevance of the wide scope of conduct which is subject to the Imperial duty is that the consequences of breach fall to be determined by reference to the nature of the breach. The breach may have occasioned financial loss and the remedy for the breach should include compensation for that loss. And so in the context of the exercise of a power, relief by way of financial compensation should not be excluded as a remedy simply because another remedy (eg setting aside) may be available.
386. To some extent this is, I consider, a bootstraps argument. Even on Mr Evans' case, the consequences of a breach of duty fall to be determined by reference to the nature of the breach; parity of reasoning suggests that the remedy also falls to be determined by the nature of the breach. Just as a breach of the Imperial duty can take many different shapes and forms, so too the equitable remedy should be crafted to reflect the nature of the breach. Rather than saying that, because one remedy is available for a particular sort of breach (eg financial compensation for one type of breach), that remedy should be available for all types of breach, I think that a more principled approach would be to match the remedy to the breach in accordance with established authorities applied by way of analogy where it is possible to do so.
387. The Trustee suggests two sources for a financial claim which the Trustee and/or RBs are able to assert. The first is damages under Lord Cairns' Act (section 2 Chancery Amendment Act 1858 now reflected in section 50 Senior Courts Act 1981). The provision (whether in its old or modern wording) bites where, among other cases, the Court has jurisdiction to entertain an application for any injunction. I am prepared to assume for present purposes, and I think it almost certainly correct, that the scope of the jurisdiction covers all unlawful conduct and so would cover a threatened breach of the Imperial duty.
388. The Trustee's case is that an act on the part of Holdings which amounts to a breach of duty constitutes the commission or continuance of a wrongful act, engaging the jurisdiction to award damages instead of or in addition to an injunction. A member could in principle, subject to the Court's discretion, obtain an injunction against Holdings to restrain a threatened or continuing breach of the Imperial duty. The duty imposes an equitable obligation and it follows that, in principle, an injunction could be obtained to enforce compliance.
389. So far so good. However, the breach of the Imperial duty in serving the Exclusion Notices is a matter of history. There is no continuing breach of duty and there is no injunctive relief which an affected member would be able to seek. In the Trustee's skeleton opening, it is said that this does not matter because damages under Lord Cairns' Act are available in respect of past loss caused by conduct that could have been the subject of a prohibitory injunction: reference was made to Leeds Industrial Co-Operative Society Ltd v Slack [1924] AC 851 at 861 to 863. I was not taken passages in oral submissions and no detailed submissions were made orally on the point. I have looked at the passage cited and do not see that it supports the proposition. In my judgment, there is no relevant injunction which could now be granted which would found an alternative claim to damages under the Act.
390. Quite apart from that, Lord Cairns' Act was really a procedural reform. It enabled the Court of Chancery to award damages (which previously it could not do) and it enabled the common law courts to award damages for future or repeated wrongs. It did not give any court the power to award damages whenever it thought it right to do so. There had to be a cause of action justifying the award. Thus I agree with Mr Simmonds that the Trustee's submission really begs the question of whether there is a financial cause of action for breach of the Imperial duty. If there is not, Lord Cairns' Act does not magic one out of the ether.
391. The Trustee's second source for a monetary remedy for breach of the Imperial duty is "equitable compensation": the Imperial duty is said to be an equitable duty so that, if it is breached, there is no reason in principle why the Court should not order monetary compensation.
392. The Trustee gives examples where monetary compensation has been awarded for breach of an equitable duty. I take these verbatim from the written closing submissions:

i) A personal monetary award may be made against a trustee for breach of trust. Even though this is properly analysed as an accounting remedy, in practical terms it is indistinguishable from a common law damages remedy and is essentially compensatory in nature: see Bartlett v Barclay's Bank Trust Co Ltd [198] Ch 515 at p545; Target Holdings Ltd v Redferns [1996] AC 421 at p439.

ii) Equitable compensation may be awarded for breach of fiduciary duty: Bristol & West Building Society v Mothew [1998] Ch 1. This is not restricted to fiduciaries who are custodians of their principal's assets (and in that sense in a position similar to a trustee): see Cia de Seguros Imperio v Heath REBX Ltd [2001] 1 WLR 112 (insurance brokers); solicitors can also be liable for breach of fiduciary duty without holding client money: Longstaff v Birtles [2002] 1 WLR 470.

iii) Equitable compensation may be awarded for breach of a (non-fiduciary) duty of skill and care: Bristol & West Building Society v Mothew [1998] Ch 1. There may be differences in the applicable rules of causation and hence the amount of loss recoverable in the case of breach of a fiduciary duty and a non-fiduciary duty, but equitable compensation is available in each case.

iv) Equitable compensation may be awarded for breach of confidence, an obligation existing only in equity: Seager v Copydex Ltd (No. 1) [1967] 1 WLR 923.

v) Commonwealth authority supports the award of equitable compensation for the exercise of a power in cases where there is a "fraud on the power": Houghton v Immer (No 155) Pty Ltd (1997) 44 NSWLR 46, a decision of the Court of Appeal of New South Wales.

vi) When confronted with a breach of a duty imposed by equity, the Court can in principle award equitable compensation, and should do so if and to the extent that is required in order to do justice. The contrary proposition, that the Court cannot grant any remedy beyond declaring the effectiveness or otherwise of the conduct in breach of duty, is unattractive and unsupported by authority.

393. Mr Simmonds does not agree with much, if any, of that, and certainly not the conclusions. He submits that in order to claim damages or equitable compensation, an affected member would have to demonstrate a recognised cause of action. Since we are here concerned with the Imperial duty, we are not concerned with contract; and no claim is made against Holdings in tort so there is no damages claim. In theory, it is possible to obtain equitable compensation for breach of trust or fiduciary duty but no such duty has been breached in the present case (a matter which is, I believe, common ground).
394. Mr Simmonds suggests that a helpful analogy can be drawn with the circumstances in which a voluntary disposition is rendered voidable by reason of mistake. In such a case, the mistake does not give rise to any claim for damages or equitable compensation on the part of any interested individual; it merely allows the Court (at the suit of such an individual) to avoid the disposition. Likewise, an exercise of a discretionary power in a manner that breaches the fetter imposed by the Imperial duty renders the exercise voidable but it does not give rise to a claim for damages, equitable compensation or any other individual remedy.
395. I do not find this analogy particularly helpful. In such a case (ie a voluntary disposition) it will be the disponor who is seeking to avoid his disposition on the basis that it was made as the result of mistake. There will be no question of any breach of duty owed to the recipient by the disponor in such a case. In contrast, the essence of the Imperial duty is that the employer is subject to certain obligations the breach of which impinges on the members.
396. Mr Simmonds contends that the "Imperial duty" is in itself a misnomer in that it imposes no positive duty at all. It is, he says, simply a fetter on the manner of exercise of discretionary powers under pension schemes. Since it does not impose a positive duty, there should not, he says, be a financial remedy for breach of the duty (by which he must mean acting contrary to the fetter).
397. [This paragraph is intentionally left blank].
398. I do not agree that "the Imperial duty" is a misnomer. It is true that it does not impose a positive duty on Holdings to take any particular action. But a duty does not have to be a duty to do something. One can have a duty not to do something, or a duty, if one does something, to do it, or not do it, in a particular way. In the case of pension schemes, there is a duty (or call it an obligation if you will) on an employer to exercise its powers, if it exercises them at all, in accordance with the limitation stated in Imperial Tobacco. Inherent in, indeed the essence of, the limitation is an obligation on the employer not to act in a particular way.
399. Connected with that point is another made by Mr Simmonds with which I agree. The Imperial duty has been fashioned in the context of the exercise of powers. It may be, as Mr Evans submits, that there are other situations in which the employer or the employee may take actions in the context of the pension scheme (in contrast with the contract of employment) which would destroy or seriously damage the relationship of trust and confidence. It may be that, by way of further extension of the implied contractual obligation of trust and confidence, a new duty should be imposed or the Imperial duty expanded. But that is not, simply in terms of language, what is meant by the Imperial duty. The remedy for breach of the Imperial duty must be assessed by reference to the scope of the duty which has, thus far in the development of the jurisprudence, been established. Having said that, I consider that the Imperial duty should apply to closely analogous functions (see paragraph 431 below). I do not perceive that sort of case as any real extension of the duty.
400. Returning to analogies, I find the analogies drawn by Mr Evans much more apposite than those drawn by Mr Simmonds, in particular the claim which an employer has against and an employee for breach of confidence. Further, Houghton v Immer is a case where the voidable exercise of a power led to an award of equitable compensation. English law is, I consider, consonant with that decision. It provides a good analogy and suggests strongly to me that equitable compensation is available in cases of breach of the Imperial duty. I so find in relation to the breaches of the Imperial duty to which Project Waltz gave rise.
401. The Trustee's position is that, in principle, to the extent that members are not fully compensated by the orders which I make under Issues 13 to 15, 18 and 19, they should be compensated by a monetary award – either by way of damages or equitable compensation, as set out above.
402. The answer to Issue 16 is that the members are in principle entitled to damages and equitable compensation as a result of service of the Exclusion Notices.

Issue 17: If members are entitled to damages or equitable compensation from Holdings, in principle what is the measure of damages for such compensation?

403. Holdings' position is that no damages have been suffered because an affected member can set aside the Exclusion Notices. As to equitable compensation, its position is the same. The affected members will have remained (or will be treated as having remained) in pensionable service notwithstanding service of the Exclusion Notices and benefits will be provided accordingly (with additional back-payments for those who have retired in the interim). In the circumstances, no affected member has suffered any loss. Further, if a member elects not to avoid the Exclusion Notices so far as concerns him, he has no claim for equitable compensation: he cannot pick and choose in that way.
404. Holdings therefore submits that Issue 17 does not arise. But if that is wrong, it submits that any affected member seeking damages or equitable compensation would have to prove (in separate proceedings and on a member-by-member basis), causation and loss in accordance with the usual contractual principles or analogous principles applicable to the assessment of equitable compensation. In these circumstances, Mr Simmonds seeks to dissuade me from laying down general guidelines as to the measure of damages or equitable compensation that should apply in what he says will be potentially numerous and, at this stage, necessarily hypothetical separate proceedings, each of which would (if brought) turn on its own particular facts. He gives three reasons why I should not do so.

i) First, to the extent that such proceedings may properly be framed as claims for breach of contract, the principles relating to the assessment of damages in respect of such claims are well established. In particular, there is a principle that damages are to be assessed on the footing that the contract of employment is terminated at the earliest lawful moment: see Lavarack v Woods of Colchester Ltd [1967] 1 QB 278); no useful purpose would be served by the Court giving further guidance in the abstract in that respect in these proceedings.

ii) Secondly, since such claims will, or at least may, be pursued before a different tribunal, it would be invidious for the decision of that tribunal to be straitjacketed in advance.

iii) Thirdly, the courts generally (and wisely) shy away from deciding points of law divorced from the facts – either as found or as agreed. No facts in relation to individual claims have been found or agreed and there are no test cases on foot.

405. I bear the second and third of those in mind; certainly, I should be circumscribed about what, if anything, I do decide. In relation to the first reason, I do not think that matters are as clear as Mr Simmonds suggests. Lavarack is authority for the proposition that damages for wrongful dismissal cannot confer on an employee extra benefits which the contract did not oblige the employer to confer even though the employee might reasonably have expected his employer to confer them on him in due course. But in cases where the employer is exercising a discretion, the position is, I think, different. Thus in cases of bonus awards, the court, when assessing damages for a wrongful failure to award any bonus, seeks to evaluate what the employer itself would have been likely to do. It is not to be assumed that the discretion would have been exercised so as to give the least possible benefit to a claimant if such an assumption would be unrealistic on the facts: see Rutherford v Seymour Pierce Ltd [2010] EWHC 375 (QB), [2010] IRLR 606 at [33] and Horkulak v Cantor Fitzgerald International International [2004] EWCA Civ 1287, [2004] IRLR 942 at [48] and [64] to [72]. See further the discussion at paragraph 142 above.
406. The Trustee's position is that both equitable damages under Lord Cairns' Act (which only arises if I am wrong in my conclusion that this route to recovery is not open) and equitable compensation for breach of Holdings' Imperial duty fall to be assessed on the same basis as damages at common law: see Attorney General v Blake [2001] 1 AC 268 at p281 and Cia de Seguros Imperio v Heath REBX Ltd [2001] 1 WLR 112 respectively. Accordingly, conventional principles of causation and remoteness apply. The Trustee contends that members should in principle be entitled to recover any financial loss that they would not have suffered but for IBM's breach of duty. Examples are given in the Trustee's written closing:

i) salary related losses (for example, basic salary, salary increases, variable pay, bonuses and other benefits);

ii) pension benefit losses; and potentially

iii) other losses (for example, state pension losses, loss in value of share options or an inability to maximise the investment strategy and tax efficiency of their pension arrangements).

407. Mr Simmonds says that there is no evidence that any member has in fact suffered any loss of those types.
408. There is also the possibility of a claim for equitable compensation, based on the same principles as exemplary damages, for breach of the Imperial duty which is developed over nearly 5 pages of the Trustee's written closing submissions. Mr Simmonds has two answers to that.

i) The first is that exemplary damages, otherwise known as punitive damages, are awarded against a defendant as a punishment, so that the assessment goes beyond mere compensation of the claimant. Whilst such damages are capable of being awarded in tort (albeit only in very limited circumstances), there is no right to recover exemplary damages for breach of contract. If any right to damages arises in the present case, it would be founded upon (or by analogy with) a cause of action in contract. Therefore, as a matter of principle, exemplary damages would not be recoverable in the present case.

ii) The second answer is that, even if such an award is available in principle, it should be by reference to the principles developed in tort and subject to the restrictions laid down in Rookes v Barnard [1964] AC 1129. The facts of the present case do not fall within those principles.

409. In my view, it is not appropriate that I should say anything by way of decision in this judgment about the measure of damages essentially for the second and third of Mr Simmonds' reasons set out at paragraph 401 above. I would only add, by way of obiter observation, that I do not consider that the courts should adopt the same (or any similar) approach to equitable compensation as its approach to exemplary or punitive damages or anything like that approach.
410. I am not, therefore, prepared to give an answer to Issue 17.

Issue 18: In the light of the answers to Issues 13 to 17, should the Trustee administer the DB Plans on the footing that (subject to any events which would have caused a cessation of pensionable service apart from the service of the Exclusion Notices) the pensionable service of members under the DB Plans is (i) still continuing or (ii) has ceased as a result of the Exclusion Notices and if so on what date?

411. I do not consider that the answers to Issues 16 and 17 have any impact on Issue 18 (although they may influence the decision which affected members make in relation to their decision to set aside the Exclusion Notices vis à vis themselves). In the light of my answers to Issues 13 to 15, the Trustee should administer the DB Plans on the footing that affected members are entitled to elect to have the Exclusion Notices set aside vis à vis themselves. To the extent that they do so, this means that the DB Plans should be administered on the footing that pensionable service is still continuing; to the extent that they do not do so, the Exclusion Notices are to be treated as effective from 6 April 2011 with the treatment of affected members in other respects (eg Hybrid Deferred status and M Plan membership where appropriate) also being given effect.

Issue 19: If the Exclusion Notices are avoided and set aside,

19.1 do (i) the affected members and (ii) those Postponed Retirees who were expressed to be subject to the Exclusion Notices have a right, pursuant to the Pension Decision Tool or otherwise, individually in the future to elect:

(a) to be treated as a member of his/her Chosen DC Arrangement in respect of benefits earned thereunder on and from 6 April 2011;

(b) to retain the benefit of any Step Down Contributions and Matched Contributions paid by Holdings or by or on behalf of the member into the Chosen DC Arrangement; and

(c) to retain Hybrid Deferred status for his/her accrued benefits under the DB Plans in respect of service up to 6 April 2011;

instead of being treated as in pensionable service under the DB Plans on and from 6 April 2011?

19.2 If those members are entitled to be treated as in pensionable service under the DB Plans after 6 April 2011, and if they do not make the election mention in 19.1, are the payments already made by such members into their DC arrangement to be applied in the manner set out in paragraph 5.2 of the Pension Decision Tool?

Background

412. I have already mentioned the Pension Decision Tool and its relevant provisions: see paragraph 337ff above. It was issued in the knowledge of, and made express reference to, these proceedings. If the members elected to join the IBM DC arrangements, they would receive the benefit of enhanced employer contributions. These are known as "step down" contributions and "matched" contributions.
413. As can be seen from paragraph 5.2 of the Pension Decision Tool, provision is made for what is to happen to the contributions made by and on behalf of the member if, as a result of these proceedings, they are entitled to be treated as having continued in Pensionable Service under their DB Plan (unless the member chooses to remain in his or her chosen DC Plan).
414. The terms concluded by informing the member to select the "submit decision button" once he or she was ready to make the decision and, depending on the option selected, the member would be informed of one of the following, depending on his or her selection:

"A. I have read the important information above and I elect to join the M Plan on and from 6th April 2011 on the terms and conditions set out therein …

B. I have read the important information above and I elect to join the IBM UK Personal Pension Plan (the IBM UK PPP) on and from 6th April 2011 on the terms and conditions set out therein …

C. I have read the important information set out above and confirm that I elect NOT to join any IBM pension arrangement for service after 5th April 2011 consequently I will not be an active member of any IBM Pension Plan …"

415. Issue 19.1 asks whether, if the relevant members would otherwise be entitled to the benefit (if any) of the invalidity of the Exclusion Notices, they can choose instead to continue to benefit from their election under the Pension Decision Tool to be a member of their chosen DC arrangement (whether operated by IBM or by a different provider), with the benefit of the enhanced employer contributions and Hybrid Deferred status.
416. Issue 19.2 asks whether, if those members are entitled, and choose, to be treated as in pensionable service under the DB Plans on and after 6 April 2011, the payments made by and on behalf of such members into their DC arrangements should be applied in the manner set out in paragraph 5.2 of the Pension Decision Tool.
417. It is common ground that the answer to each of the questions asked in relation to Issue 19.1 is in the affirmative but subject to a disagreement about the extent to which the New ER Policy applies to Hybrid Deferred status. I have discussed this issue at some length above in relation to Issue 15.3. In my judgment, for the reasons there appearing, a member choosing to retain Hybrid Deferred status can only do so on the terms on which it was offered, terms which include the New ER Policy.
418. Issue 19.2 is concerned with the precise terms on which DC accrual is to be unwound. The Trustee now accepts that the payments already made by the relevant members into their DC arrangement are to be applied in the manner set out in paragraph 5.2 of the Pension Decision Tool. It may be that the details of how that provision is to be applied cannot be agreed, in which case I will need to make directions. I am not asked to do so at the present time; and so I will reserve the matter until later if such directions are necessary.
419. In slightly more detail, the Trustee has confirmed in correspondence that, as a matter of general principle, it agrees that paragraph 5.2 of the Pension Decision Tool has effect in accordance with its terms. Thus as a matter of general principle the Trustee accepts that if a member's DB accrual is reinstated or declared to have continued for any period after 6 April 2011, then (subject to his right to elect to remain a DC member as identified in Issue 19), the realisable value of the DC benefits he has earned during the period since 6 April 2011 ought to be transferred to the relevant DB Plan, insofar as possible. But the Trustee considers that paragraph 5.2 is not a comprehensive code and does not cover every case; numerous complicated issues could arise, just some of which were set out in Nabarro's letter to Bond Dickinson dated 11 July 2014.
420. There is one additional point which I need to mention in relation to Issue 19. The right of waiver found in paragraph 5.2 of the Pension Decision Tool would become exercisable only once the decision of the Court is known. Although this judgment provides that decision, the Trustee's view is that affected members should be allowed to wait until the final decision of the Court including after any appeal before they have to elect. Whether that is a sensible view depends on whether, pending an appeal, the Trustee is to administer the DB Plans in a way which reflects the result of this judgment or whether a stay is placed on the implementation of my decisions pending an appeal. I have not yet heard any submissions about permission to appeal against my conclusions in the Judgment let alone my conclusions in this judgment. I propose therefore, to address the question of the timing of any election at a further hearing after the hand-down of this judgment.

Issue 20 (the ER Window)

Issue 20.1: Are members who retired pursuant to the ER Window in principle entitled to damages or equitable compensation or some other remedy (and if so what) from Holdings if they can prove that, as a result of the ER Window and/or the impending introduction of the New ER Policy, they retired earlier than they would otherwise have done?

Issue 20.2 If members are entitled to damages or equitable compensation from Holdings, in principle what is the measure of such damages or compensation?

Issue 20.3: If members are entitled to damages or equitable compensation from Holdings, then further or alternatively, (if such a member so elects) should the Trustee administer the DB Plans on the basis that the member is entitled to any increased benefits under the DB Plan as a result of the implementation of the ER Window, and, if so, what increased benefits and subject to what (if any) conditions?

421. Holdings' submissions on these Issues are the same, mutatis mutandis, as those made in relation to Issues 16 and 17 concerning damages/equitable compensation in relation to service of the Exclusion Notices. The implementation of the ER Window was (subject to any appeal) a breach of the Imperial duty rather than a breach of the implied contractual duty. Breach of the Imperial duty does not give rise to any free-standing financial remedy on the part of individual members, whether in the form of damages, equitable compensation or otherwise.
422. The Trustee likewise relies on its arguments in relation to Issue 16 in relation to the principle of an award of damages and/or equitable compensation. There is some detail on the application of the principles to the facts. Mr Evans submits that it is possible that affected members may have suffered financial loss as a result of the shortness of the ER Window and the announcement of the pending introduction of the New ER Policy. In particular, there are a number of members who retired during the ER Window who contend that they would not have retired then, but rather would have continued in Holdings' service, had it not announced the impending change to its policy regarding early retirement from active status (ie the change from the Old ER Policy to the New ER Policy). The Trustee understands that the total number of these members is 861. A significant number of such members have already brought claims against IBM in the Employment Tribunal. These members contend that they have suffered financial loss as a result, for example lost earnings and other in-service benefits and lost DB accrual. They may also present claims for consequential losses, such as the reduced value of state benefits, share options and losses due to an inability to maximise the investment strategy and tax efficiency of their pension arrangements. In principle, and subject to proof that as a result of the existence and shortness of the ER Window they retired earlier than they would otherwise have done, such members should be compensated by a monetary award – either by way of damages or equitable compensation, as set out above.
423. On this issue of principle, my answer to Issue 20.1 is, mutatis mutandis, the same as the answer which I gave under Issue 16. In principle, an affected member is entitled to damages for breech of contract, or equitable compensation in respect of Holdings' breach of its Imperial duty. In relation to a person who retired during the ER Window, I do not consider that it is appropriate for me to give any further ruling. In the context of a damages claim, there are likely to be significant issues about causation, remoteness and mitigation; and similar issues may arise in relation to equitable compensation. I think that these issues would best be dealt with in proceedings (in which I include mediation) between Holdings and the affected members concerned. I am anxious, in any case, not to say anything which might have an impact on any damages claim made by such a member in respect of salary. If there is a contractual claim in respect of lost pension benefits, logically there ought to be a similar claim in respect of lost salary.
424. So far as concerns Issue 20.2, I decline to answer the question for the same reasons as I have declined to answer Issue 17.
425. The question raised by Issue 20.3 is whether the losses suffered by members who retired during the ER Window, but would not have done so but for IBM's breach of duty, can be compensated by means of an enhancement to their benefits under the DB Plans, rather than by a payment of money by IBM outside of the DB Plans. In effect, this would be to give them the additional DB accrual they lost as a result of their early retirement, plus compensation for other financial losses to the extent that those are attributable to pension provision (and not, for instance, for lost earnings which has nothing to do with the pension schemes or the Imperial duty).
426. The Trustee submits that there are two constraints on this possible approach to compensating such members.

i) First, it could not be forced on the member. Some members might prefer an award of cash outside the DB Plans; others might prefer enhanced pension benefits. It must be a matter for individual members to elect to seek compensation under the DB Plans.

ii) Secondly, it could not be forced on Holdings. The Trustee does not have a unilateral power to augment members' benefits under the DB Plans at Holdings' expense. Holdings might, or might not, prefer to compensate such members (assuming the obligation to do so is established) by enhancing their benefits under the DB Plans.

Accordingly, the Trustee accepts that this approach is only available if both the member concerned and Holdings agree on it.

427. As a matter of fact, Holdings does not agree to this proposal. It follows that the Trustee must administer the DB Plans on the footing that the relevant members retired (as in fact they did). If a compromise were to be reached between Holdings and a given member, it might be that part of such a compromise would include an agreement on the part of Holdings that the member's benefits should be augmented, in which case an agreement would also have to be made with the Trustee about the funding of such augmentation.
428. My answers are therefore as follows:

i) Issue 20.1: Yes: in principle, there is an entitlement to damages or equitable compensation.

ii) Issue 20.2: I decline to answer this Issue.

iii) Issue 20.3: the Trustee should administer the DB Plans on the basis that a member who retired during the ER Window is not entitled to any increased benefits under the DB Plans as a result of the implementation of the ER Window.

Longstop Date

429. I am conscious that I have, thus far, avoided committing myself, as I avoided in the Judgment too (but see below), from deciding when the Longstop Date actually is. In this judgment, I have already rejected Mr Simmonds' suggestion that the Longstop Date was 6 April 2011 (or at least a date not very long after that). In the light of my conclusions in relation to the Exclusion Notices, it is not in fact necessary to decide on the Longstop Date for the purpose of the remedies flowing from the invalidity of the Exclusion Notices since (subject to the Trustee's reservation of its position on appeal) it is accepted before me that it is no later than 31 March 2014. Nor is it necessary to decide the point for the purposes of remedies flowing from the invalidity of any of the NPAs (or rather of the non-pensionability terms resulting from them). However, if it were necessary for me to do so, I would adopt 31 March 2014 as the Longstop Date. The period ending on that date is the period relevant to the other Reasonable Expectation concerning continuation of the Old ER Policy. That policy could only apply to a member if he was an active member when he left service. But if a valid exclusion notice could be served taking effect before that date, the Old ER Policy would not apply to him. The result of that somewhat technical point is the same as the message that members might have taken from the combination of all of the many communications; it would be an unusually astute member, I suggest, who would have appreciated that, although he would be entitled to take early retirement on the Old ER terms, that entitlement could be abrogated by the simple expedient of excluding him from the Scheme.
430. Subject to one possible exception, this conclusion is not inconsistent with anything in the Judgment. The exception is that, in Judgment [865], I did say that a reasonable member remaining in the DB arrangements could not form an expectation that the DB Plans "would for the rest of his working life, or indeed until 2014" remain open to future accrual. On one reading, this could be said to be a finding that the Longstop Date was before 2014. But that would be contrary to what I said in Judgment [1532 i)] where I said that the Longstop Date "might be 2014 or perhaps even earlier". I could not have said that if I had previously decided that the Longstop Date was before 2014 unless I had overlooked that I had done so. I certainly did not intend to decide the point in the Judgment; Judgment [865] would more accurately have reflected my view if I had said "beyond 2014" rather than "until 2014". But that is only one reading. This paragraph must be read in its context in the Judgment as a whole. At this stage, I was considering Project Soto and the proposals announced and explained at the managers' meeting in January 2006. There were other communications after that (see Judgment [866] to [926]) which could have had an impact on how the earlier communications were to be seen. Judgment [1532 i)] reflects my ultimate view that I should not decide the point. I therefore propose to proceed on the basis that I did not decide the point in the Judgment or, if I did, that it is open to me to review that decision to eliminate the inconsistency with Judgment [1532 i)].
431. Mr Simmonds suggests that my reasoning in the Judgment has gone wrong when I said that the Reasonable Expectation following Soto lasted at least until 6 April 2011. He reaches that conclusion by placing, in my view, an incorrect interpretation on my use of the word "similar" in Judgment [1053]. By that I meant a Reasonable Expectation similar in content to that engendered by Ocean, not one which would, subject to a different starting point, have the same duration.

Issues 21 to 27: the New ER Policy

432. I should give an introduction to these Issues.
433. I have held that the introduction of the New ER Policy gave rise, even when viewed in isolation (by which I meant under Scenario A) to a breach of Holdings' Imperial duty. I did not mean to exclude the possibility that the same conduct gave rise to a breach of contract which in my view it certainly did.
434. There has been some debate about the basis on which Holdings' conduct gave rise to a breach of the Imperial duty and how that breach is to be characterised. One thing is clear which is that the Imperial duty is an equitable, not simply a contractual, duty; although Imperial Tobacco itself concerned the exercise of a power (just as the service of the Exclusion Notices concerned a power) I do not perceive any different approach being applied to the exercise by an employer of any other function under a pension scheme, in particular a change of policy which does not require any rule amendment or exercise of any other formal power. I use "the Imperial duty" as including the duty in respect of such functions. The remedies available for breach of duty in the exercise of an employer's functions under a pension scheme are no different, in principle, from the remedies available for breach of the Imperial duty in the exercise of a power, although the difference between the function in question and a power may require remedies to be tailored to suit the circumstances.
435. Although the introduction of the New ER Policy gave rise to a breach of Holdings' Imperial duty, it did not directly affect the members' rights: what it did was to indicate the circumstances in which a discretion given to Holdings under the Rules would be exercised. In that sense, there is a parallel with the NPAs which did not of themselves give rise to rights or obligations but indicated the basis on which future salary increases would be awarded. In my discussion under Issue 2, I expressed the view that it was inapposite to approach the NPAs by asking the question whether they were void, voidable, unenforceable or valid. Those questions were better directed at the contracts which came into being as a result of the NPAs having been entered into. The NPAs gave rise to breaches of duty and thus had an infecting impact on the contracts which were based on them.
436. Similarly, the change in ER Policy gave rise to a breach of the Imperial duty but it is inapposite to ask whether the New ER Policy itself is void, voidable, unenforceable or valid. What matters, in terms of remedy, is the action which was taken in the light of the New ER Policy which would not have been taken if the Old ER Policy had continued. There are two main classes of member to consider each of which comprises members who wished, at any time up to April 2014, to leave Holdings and who could have taken early retirement under the Old ER Policy but who would not be able to obtain (including those expressly refused) consent under the New ER Policy. The classes of such members are (i) members who nonetheless decided to leave service on the less favourable early leaver provisions and (ii) members who decided to remain in service rather than leave on the less favourable early leaver provisions. The first category is unlikely to include any member who could have taken advantage of the ER Window but there may be people who slipped through that net for one reason or another. And there is, of course, the category of member who did take advantage of the ER Window but who would have preferred to remain in service with the Old ER Policy applying: I have considered their position under Issue 20.3.
437. As to the first category, it is reasonable to assume that most members who left service voluntarily without consent would have left service if the Old ER Policy had continued to apply to them. It is not easy to imagine such a case, but I should not exclude the possibility of an exceptional case where a member would have decided to remain in service had the Old ER Policy continued to apply but who decided to leave service taking the less favourable early leaver benefits.
438. As to the second category, there were, no doubt, many members who remained in service who would have taken early retirement if the Old ER Policy had continued in operation. Members had a Reasonable Expectation that they would be able to take early retirement under the Old ER Policy, a Reasonable Expectation which would last until April 2014. Such a member, however, faces an evidential burden in satisfying one or more of Holdings, the Trustee and a court or tribunal that he would have taken early retirement had the Old ER Policy been available. If such a member cannot demonstrate that he would have taken early retirement, the change in ER policy will have had no impact on him and he is not entitled to a remedy.
439. In relation to many of the Issues concerning the change of ER Policy, the Trustee analyses the change of policy as the advance refusal of consent to early retirement. Under the Old ER Policy, a member could have asked for consent to take early retirement and would virtually automatically have been given consent if he fulfilled the relevant criteria. Once the New ER Policy was in place, it was still open to a member to seek consent but there would be little point in his doing so if he knew that he did not fulfil the new criteria. There would not, in his case, be an express refusal of consent.
440. As an overarching point, the Trustee contends, for the purpose of remedies, that Holdings should be treated as having given consent to retirement from active service to all eligible members (ie those who would have been entitled to take early retirement under the Old ER Policy) who in fact left service during the period for which the Old ER Policy should have continued in effect. The class under consideration here, of course, comprises members who did not have consent to early retirement (other than C Plan Over 60 Members who did not need consent) and either took immediate early leaver benefits or became deferred members. The Trustee's submission is that those who took immediate benefits should now receive (with retrospective effect) the benefits which they would have received under the Old ER Policy and that those who became deferred members should now be entitled (subject to an election to the contrary) to the same benefits (with retrospective effect) as if they had taken immediate benefits and not become deferred members.
441. It is said that this approach is consistent with the approach which I adopted in relation to the Rectification Action. I was there dealing with the position on the assumption that Holdings would be in breach of its Imperial duty in refusing to exercise the discretion then in issue (to consent to a rule amendment). I said that Holdings must consent, or be treated as consenting, if to refuse consent would be irrational or perverse – the test which I concluded was appropriate to the Imperial duty.
442. I will consider those general submissions on the part of the Trustee when considering the various sub-issues of Issue 21. Before I turn to those, there is one important point in relation to the New ER Policy which I need to address and which will feed into the answers to the Issues. It is the extent to which Holdings could have validly resolved on a new ER policy identical to and along the lines of the New ER Policy. In the Judgment, I decided that the Reasonable Expectation concerning ER policy was that the policy would not change until April 2014. Accordingly, Holdings would have been able to announce (subject to an appropriate, if any, period of notice) a change in that policy as from the end of March 2014 without thereby giving rise to any breach of duty. This area is explored further under Issue 22.2.
443. I mention, just to flag it, the question raised by Issue 21.2(a), namely whether there should be a different treatment in relation to pension earned before and after 6 April 2005 and to which I come in due course.
444. I turn to Issue 21 and its sub-issues.

Issue 21.1: In relation to the change from the Old ER Policy to the New ER Policy on 6 April 2010, and any refusals or deemed refusals of consent to early retirement pursuant to that change, were the change in policy or such refusals or deemed refusals

(a)

void and/or

unenforceable

(b)

voidable and liable to be set aside

(c)

valid?

445. The Trustee's primary position is that the change in policy and refusals were void. The same legal analysis is invoked as in the case of the exercise of the Exclusion Powers by the service of the Exclusion Notices, the submission being that the Exclusion Notices were void. I have rejected that submission and held that the Exclusion Notices were simply voidable. I rejected the notion that there was a half-way house, namely unenforceable, between void and voidable.
446. As I have already indicated, I do not consider it helpful to direct the question of void/unenforceable/voidable/valid to the New ER Policy as such. Nor do I find it particularly helpful to ask whether the "refusals or deemed refusals of consent" were void/unenforceable/voidable/valid. The more important issue is the effect of the New ER Policy on the eligible members. There is no doubt that, in the light of the Judgment, the change of ER Policy and its implementation gave rise to a breach of duty on the part of Holdings in relation to eligible members. The change of ER policy gave rise to a breach not only in the context of Project Waltz as a whole, but also viewed in isolation under Scenario A (ie taking account of the Reasonable Expectation relating to the continuance of the Old ER Policy). An eligible member who has left service is able correctly to say, in my judgment, that, had it not been for Holdings' breach of duty, he would have been treated differently; he would (subject to the answer to Issue 21.2) have become entitled, on leaving service, to benefits in accordance with the Old ER Policy (save to the extent that Holdings would have been entitled, in any particular case, to decline to apply that Policy to the member, for instance because of special operational requirements).
447. The appropriate – certainly the most straightforward – remedy for such a member is for the Trustee to provide him with the benefit (including retrospective adjustment) to which he would have been entitled if Holdings had given its consent to early retirement in accordance with the Old ER Policy.
448. Is there any reason in principle why the Trustee should not do so? I do not think so. The reason (the only reason) that it should not do so is that the Rules provide for the beneficial early retirement benefit to be paid only where Holdings has given consent to the early retirement. However, in a case where Holdings would be in breach of duty in refusing its consent (so that it would be acting irrationally and perversely), it would be appropriate, in my judgment, to direct the Trustee to administer the Plans as if consent had been given. This remedy, it is to be noted, is not one which requires any action on the part of Holdings. It is a remedy which goes only to the way in which the pension trusts are to be administered. This result can be reached in (at least) the following two ways, each of which is, in my judgment, valid:

i) First, as a matter of construction of the Rules, it is possible to imply a term to the effect that, where consent is withheld in breach of duty, it is not required. On any test for the implication of a term – whether the test articulated by Lord Hoffmann in the Belize Telecom case or the "goes without saying" test or "necessary to give business efficacy" approach – a term can be implied dispensing with the need for consent in such a case. In a different field, a covenant against assignment of a lease without the consent of the landlord, such consent not to be unreasonably withheld, allows an assignment without consent where consent is unreasonably withheld. There is nothing at all surprising – indeed it is to be implied – in the proposition that early retirement should only be with Holdings' consent but with a proviso that such consent should not be irrationally or perversely withheld.

ii) Secondly, an analogy with South West Trains provides another route to the same result. That case demonstrates how a Trustee is entitled to depart from the express trusts of a pension scheme (with injunctive relief being granted if necessary) where appropriate contractual arrangements have been entered into between an employer and affected employees. Both the Imperial duty and the contractual duty of trust and confidence were breached by Holdings' failure to consent to early retirements in the cases under consideration. The other side of that particular coin is that Holdings' duties required it oldings to give its consent to erarluy Hit to give its consent to early retirement. The Trustee is entitled, in my judgment, to recognise Holdings' duties in administering the Plans. It should do so by treating Holdings' consent as having been given (in contrast to the first route which dispenses with the need for consent at all).

449. Issue 21.1 has, however, been agreed by the parties in terms of void/unenforceable/voidable/valid. If, contrary to my approach, it is necessary to address the validity of the New ER Policy and the failure to consent to early retirement pursuant to the policy on the basis of the validity of the exercise of a power, then the discussion already carried out in relation to the exercise of the Exclusion Powers is relevant. The position needs to be addressed at two levels: first, the introduction of the New ER Policy; secondly the refusal (or deemed refusal) of consent.
450. At the level of the introduction of the New ER Policy, Holdings might be said to be acting in the exercise of a power. Consider, by way of analogy, the position which would have obtained if the policy which Holdings was permitted to adopt had to be set out in a Schedule to the Rules, Holdings having a power to amend the Schedule from time to time. A change of policy would then be effected by a change to the Schedule: effectively Holdings would be exercising a power given to it by the constitutional documents. The change of policy in the present case can be tested by the same criteria as the exercise of the hypothetical power of amendment in the example. I have examined the principles sufficiently in considering Issue 13. The result in relation to the change of policy is that the New ER Policy is not void, but is voidable and liable to be set aside in whole or in part at the election of the members. This would leave the Old ER Policy in force. As to the actual failure to consent, see the next following paragraph.
451. If it is inappropriate to treat the introduction of the New ER Policy in a way similar to the express power to amend the Schedule in the example, then the matter must be addressed at the second level, that is to say the refusal (or deemed refusal) of consent. And it is necessary to address matters at that level, in any case, following on from the setting aside of the New ER Policy. Once again, the analysis in terms of the exercise of a power is dealt with under Issue 13.
452. The difficulty at this stage of the analysis is that Holdings has simply failed to give consent. It has, albeit in breach of duty, failed to exercise a power (namely to consent to early retirement). To speak of a failure of that sort as being void, unenforceable or voidable and liable to be set aside does not make a great deal of sense unless there is a default position in such cases, that is to say a default position in which consent is to be treated as having been given. Suppose, for instance that an individual had asked for consent to early retirement and had been refused; and suppose that I were to hold that such refusal was void or was to be set side. That makes the refusal a non-decision but does not, of itself, substitute for it the granting by Holdings of consent.
453. It seems to me, therefore, that establishing the invalidity of the failure to give consent does not, without more, produce a result either way. The answer, I think, is to apply the approach which I have adopted, which makes the discussion in the preceding four paragraphs unnecessary and indeed irrelevant.
454. The answer to Issue 21.1 is that the New ER Policy cannot be relied on by Holdings (subject to the answer to issue 21.2) in relation to an Eligible ER Member who would, had the Old ER Policy remained in force, have been entitled to benefits in accordance with that Policy. A refusal or deemed refusal of consent in relation to such an eligible member who has left service was, or was part of, a breach by Holdings of both its Imperial duty and its contractual duty of trust and confidence. The Trustee should proceed on the basis that either consent was not required or that it is to be treated as having been given. A member may, however, elect to retain benefits as if the Old ER Policy did not apply to him.
455. These answers are subject to the impact of Issues 23 and 24 in relation to the various Separation Programmes to which those Issues relate.

Issue 21.2: This Issue is directed at whether the change in policy or such refusals or deemed refusals are severable and partly valid in relation to pension attributable to pensionable service from 6 April 2005.

456. I have put the Issue in this attenuated form because Issue 21.2 is formulated by reference to the answers to Issue 21.1 in terms of void, unenforceable or avoided and set aside. To my mind the real issue is whether Holdings can restrict the impact of my conclusions in relation to Issue 21.1 to that part of the pension benefits which is attributable to pensionable service prior to 6 April 2005 (ie the period prior to the Project Ocean changes) leaving the New ER Policy to take effect in relation to benefits attributable to pensionable service from 6 April 2005. In my judgment, it cannot do so.
457. In the Judgment, I decided (although Mr Evans made a valiant attempt to persuade me that I did not decide it at all) that the Reasonable Expectation concerning ER policy was that the Old ER Policy would apply until 2014 in relation to service up to the implementation of Project Ocean. In other words, members were entitled to hold a Reasonable Expectation that if they retired before 2014 (or rather before the end of March 2014) they would know that they would obtain an immediate pension which, in respect of that part which had accrued by 6 April 2005 (calculated taking account of the final salary link as Mr Simmonds accepts), would attract the favourable early retirement factors. They had no Reasonable Expectation that, if they retired early, that part of the pension attributable to service after 6 April 2005 would attract those favourable factors. It was therefore open to Holdings to implement changes to early retirement provided that the Reasonable Expectation was respected. It could, for instance, have introduced a rule amendment (subject to Trustee consent and to appropriate consultation) which altered the early retirement factors applicable on early retirement with consent so that different factors applied in respect of service before and after 6 April 2005. To the extent that it was possible to achieve the same objective by a change of ER Policy, there was nothing to prevent it doing so (at least viewing the change of ER policy ignoring the fact that it would form part of Project Waltz).
458. What it chose to do was to announce the New ER Policy. This change of policy was a change to the circumstances in which consent to early retirement would be given. It thought that this change would be effective in relation to the whole of a member's pension entitlement. But the result of the Judgment is that it could not be effective in relation to the part of the pension attributable to pre-6 April 2005 service. Faced with that fact, Mr Simmonds submits that the solution is simple: a member is to be entitled to retire early but the New ER Policy will apply only to that part of his pension accrued after 6 April 2005. It is a necessary element of this submission that, had the point been realised when the New ER Policy was introduced, it would have been possible, expressly, to adopt a new ER policy which said precisely that. I do not consider that this could have been done.
459. The principal reason why this could not be done is that it is not conceptually possible to divide the relevant consent into different components. The Rules permit a member to take early retirement with Holdings' consent: see Rule 2 Schedule D Main Plan DB Rules; Rule 6.6 Data Sciences Section Rules; and Rule 2 Schedule G I Plan Rules. If he does retire with consent, then the whole of his pension attracts the favourable early retirement factors. Either a member retires with consent or he does not retire with consent: I do not understand how, conceptually, it can be said that a member both (i) leaves service with consent and (ii) does not leave service with consent. As I have suggested, a rule amendment might have been possible to achieve Holdings' object. But that was not done and that is not what the Rules of the DB Plans in fact provide. It is not possible, in my judgment, to treat a consent to take early retirement under those Rules as operative in respect of one period of service but not another. The New ER Policy itself was simply not capable of effecting a change to the Rules which would have brought about this result. I can understand an argument that, had it been possible to implement a new policy which achieved Holdings' object, then the New ER Policy should be treated as doing so. But, for my part, I have not been able to think of such a policy and none was put forward by Mr Simmonds. Accordingly, I reject Holdings' case on this issue.
460. Further, the result of what I have said reflects, in my view, how the reasonable person would have viewed matters. Holdings having engendered a Reasonable Expectation that accrued benefits were protected, and thus that the benefits of the existing ER policy would continue to be available in respect of past service, the reasonable person would know, looking at the Rules and booklets, that he was entitled to early retirement with Holdings' consent. He would know of the practice and policy always to grant consent. It would not have occurred to the reasonable person, I venture to suggest, that by changing its policy and practice concerning the giving of consent Holdings would be able to treat different periods of employment differently. He would surely have understood that a Rule amendment would be needed to produce that result.
461. There is another reason why this severance of the consent requirement cannot take place. The adoption of the New ER Policy did not take place in isolation. It took place in the context of Project Waltz as a whole. It is the proposals as a whole which were flawed. Even if the New ER Policy had been reflected in a Rule amendment so that it applied only to future service, it would still form part of Project Waltz and thus part of a project the aim of which was to achieve an objective which could not be achieved without a breach of duty. One aspect of that, but not the entirety, is the lack of adequate consultation. It is far from clear that, if there had been a proper consultation and assuming that Holdings had acted entirely properly, any change at all would have been made to the ER policy. In these circumstances, I do not consider that the Court should countenance giving partial validity to the New ER Policy even if that were possible in principle.
462. Moreover, it seems to me that the conclusion for which Holdings argues is, in effect, asking me to treat the change of ER Policy as effecting a change in the Rules by allowing consent to be given for one period but not for another. Even assuming that such an amendment could be made, it could be made only in respect of future service. But no such amendment was made: it could only take effect from the date on which it is made and, no amendment having being made so far, the Old ER Policy must continue to apply at least up to April 2014.
463. The answer to Issue 21.2 is therefore that no distinction can be made between the periods before and after 6 April 2005 so far as the giving of consent to early retirement is concerned.

Issue 22: In the light of the answers to Issue 21 above:

22.1 should IBM have continued to apply the Old ER Policy on and after 6 April 2010 to Eligible ER Members (and if so, in relation to (a) all of their pension or (b) pension attributable to pensionable service prior to 6 April 2005)?

22.2 from what date may IBM properly apply the New ER Policy (or a similar replacement for the Old ER Policy) to Eligible ER Members, and what if any steps would IBM have to take (such as the giving of notice to members) to introduce such a policy?

22.3 until such time as the New ER Policy (or a similar replacement for the Old ER Policy) may properly be applied, does the Old ER Policy continue to apply to Eligible ER Members (and if so, in relation to (a) all of their pension or (b) pension attributable to pensionable service prior to 6 April 2005)?

464. As to Issue 22.1, Holdings should have continued to apply the Old ER Policy (including the right to refuse consent in some cases) on and after 6 April 2010 to eligible members in respect of all of their pension.
465. As to Issue 22.2, the Trustee's position is that Holdings must give reasonable notice of a change of policy and that a reasonable time would be 6 months from the date when notice is given. Holdings' position is that no further notice is required. The change of policy put in place in April 2010 should be allowed to take effect from 1 April 2014. This would not confound any Reasonable Expectation which came to an end at the end of March 2014.
466. I certainly agree with Mr Simmonds that there is no conceptual difficulty in Holdings' approach. The question is whether, when a member leaves service, he is to be treated as retiring with consent even though (i) no consent has been given and (ii) he had no Reasonable Expectation that the Old ER Policy would be in force after March 2014. There is no question of severance of the good from the bad. Further, even if that is wrong, there is no conceptual difficulty in severing the New ER Policy in respect of the periods before and after the end of March 2014: its operation can be ineffective so far as concerns members leaving service before that time but effective thereafter. I would add that the members have known since late April 2014 (by which time it was possible to have digested, so far as digestion was ever possible at all, the Judgment). It might be said therefore that the members have had plenty of time in which to make a decision whether to leave Holdings.
467. However, that is not necessarily to say that the New ER Policy can be treated as a valid policy known to the members and by which their early retirement requests would be assessed as from the end of March 2014. With the benefit of the Judgment (to the extent that it survives any appeal) it can be seen that the Reasonable Expectation did not survive beyond the end of March 2014. But nobody knew that until I had handed down the Judgment, with the parties taking radically different approaches to the duration of a Reasonable Expectation even assuming it was established. It is clear that, had anyone been asked to say how the good was to be severed from the bad, it would have been impossible to say. Holdings would have said the New ER Policy did not need to be severed: it was wholly good. The RBs would say that the New ER Policy was inherently flawed and should not be recognised at all. Even those who might have accepted that it could take effect from some future date would have been unable to say what that date was. Accordingly, were this a case where it was necessary to sever the good from the bad as can be done with the exercise of powers (as to which see the discussion of Bestrustees and Betafence at paragraphs 298ff above) I doubt very much that it would be appropriate to do so. It would have been impossible, when the policy was announced, to say when it would be effective from.
468. But that is not the real point which is, it seems to me, that by announcing the New ER Policy in 2010, Holdings made clear its intention to bring to an end the Old ER Policy. Holdings was constrained by the Reasonable Expectation which it had engendered from implementing that intention; but when the period for which the Reasonable Expectation endured came to an end, that constraint came to an end. Let me attempt to throw some light by looking at two counterfactual situations:

i) Situation 1: suppose that Holdings had served a notice in, say November 2013, saying that, although it expected to succeed in the litigation, it considered that on any view, the relevant Reasonable Expectation could not last beyond the end of March 2014 and it intended that the New ER Policy would operate from 6 April 2014 whether or not it succeeded on other parts of its case.

ii) Situation 2: suppose that Holdings had not announced or implemented any changes to the ER policy in 2010; suppose that it had told members in, say early 2013, that it was considering changes to its ER policy but, although it had not decided how to proceed, it was giving notice that no-one should assume that, after the beginning of 2014, the Old ER Policy would continue to apply. January 2014 comes and goes and then, on 30 April 2014 (more than 3 weeks after the Judgment), it announces that the New ER Policy will come into force the next day.

469. My question, in each scenario, is whether Holdings could validly implement the New ER Policy from 6 April 2014. It might put some members in a difficult position when making a choice about what to do because, until the Judgment was delivered, they would not know whether they would be entitled to the benefit of the Old ER Policy. What they would know is that it would not be available to them after 6 April 2014 if (as turned out from my Judgment) the Reasonable Expectation had expired by that date. Unless it can be said that to present members with that choice was itself another breach of duty, the difficult decision has to be made. Otherwise, I do not see why Holdings should not be able to implement the New ER Policy in situation 1.
470. The question whether Holdings should be able to implement the New ER Policy in situation 2 is more difficult. The members' Reasonable Expectations arising out of Ocean and Soto would not have been disappointed in the light of the Judgment; and the members would have known for months that a change was being considered and that they should not rely on the practice which had been adopted up until then. Unless it can be said that to present members with the change without giving them a last opportunity to take advantage of the Old ER Policy was another breach of duty, I do not, once again, see why Holdings should not be able to implement the New ER Policy. And if it did so the members would have no remedy.
471. The actual facts of the present case are, of course, different, but they reflect an aspect of each of situations 1 and 2. As in situation 1, Holdings has in fact indicated that the Old ER Policy is not one which it intends to continue (albeit it has to recognise that it is constrained by the Reasonable Expectation); and as in situation 2, its position is that further notice is not required. What both situations 1 and 2 have in common is that neither of them has anything to do with a breach of the members' Reasonable Expectation concerning continuation of the Old ER Policy. But that (subject to one point) is also the case on the actual facts. It is not the past breach of duty on the part of Holdings which gives rise to the position in which the members now find themselves. Rather, it is because the constraint arising out of the Reasonable Expectation has come to an end and because, having made clear that members should not harbour any expectation about the continuation of the Old ER Policy, members were faced with a difficult choice. In other words, the members' complaint will not be about disappointment of their Reasonable Expectation and will not, therefore, be about a remedy for the breach of Holdings' duty brought about as a result of that disappointment.
472. In this discussion, I have been considering the position in the context of the Reasonable Expectation (ie under Scenario A) but not in the context of Project Waltz as a whole. The fact is that the New ER Policy was developed as part of Project Waltz and was implemented as an element of Project Waltz. It might therefore be argued that the implementation of a policy identical to or along the lines of the New ER policy cannot now be implemented without a further announcement and without giving members the opportunity to take early retirement under the Old ER Policy; in other words, to implement the Policy now would simply be to implement a step in Project Waltz, a project which is fatally flawed.
473. It cannot be right, however, that Holdings is simply unable for ever to implement such a policy simply because, as part of Project Waltz, the New ER Policy and its implementation gave rise to a breach of duty. There must come a time after which Holdings is entitled to adopt such a policy free from the constraints of Project Waltz. Whether Holdings can implement such a policy once such a time has arrived without giving reasonable notice of the change of policy is not a question which arises in the context of remedies for the breaches of duty involved in Project Waltz. Rather, it is a question which would arise whenever Holdings decides to change its ER policy.
474. In my judgment, the time after which Holdings is entitled to adopt such a policy is 31 March 2014. Once the constraint imposed by the Reasonable Expectation had come to an end (which, as I found in the Judgment was 31 March 2014) Holdings was free to adopt a new ER Policy identical to or along the lines of the New ER Policy. If a change of policy could not, as a matter of general approach, be implemented without notice, then notice would be required for Holdings to adopt the new policy; but if notice was not required, then the fact that the new policy follows on from the end of Project Waltz and takes place in the context of the litigation leading to the Judgment and to this judgment does not result in notice being required, when otherwise it would not be required.
475. The Trustee's position is that 6 months' notice is required but, as I understand the submission, the requirement for notice arises because this is the only way in which a remedy for the breach of duty can be provided. I do not agree with that: if there is a requirement for notice it stems from a more general requirement for the giving of notice when a change of ER policy is to be adopted. The point has not been argued. If any party wants me to deal with the point, I will consider whether I should do so, taking account of any objections to that course. It may help if I provide some very provisional views which are not put forward even as preliminary decisions in principle:

i) Following the handing down of the Judgment, and allowing some time for it to be read and understood, it was known that the period during which members were entitled to hold the Reasonable Expectation concerning early retirement had already expired, on 31 March 2014.

ii) That date was the time from which Holdings was entitled to adopt a New ER policy free from the constraints of Reasonable Expectations. If notice of the adoption of a new policy was not required, then Holdings is entitled to apply the new policy to members leaving service after that date.

iii) If, in contrast, notice was required then, at the very latest, that notice came in the written opening submission for the remedies hearing dated 14 July 2014 and served, I believe, on 17 July. In that submission, Holdings made clear that the New ER Policy was to take effect. If a period of notice was required, the period from 17th July to date would be a reasonable one. If there is any doubt about Holdings' position at that time, there could not be any doubt by the end of the remedies hearing at the end of July.

iv) The Trustee suggests that a 6 month period of notice is required. I do not think that so long a period is required even if the Trustee's submission that notice is required is correct.olHo In my view, a period of 3 months is certainly enough. That takes one to mid or late October 2014. From that time, Holdings was entitled to adopt the new policy even if notice of the change of policy was required.

v) Holdings submits that no further notice is required because members have had some four years notice of introduction ie from the announcement in 2010. I do not accept that submission. The 2010 communication gave rise to a breach of duty: the announcement of an immediate and invalid change of policy cannot be taken, I think, as an announcement of a future and valid change of policy.

vi) I have formed no view about the need for notice of a policy change. I can see that there may be an argument that to give no notice at all and simply to announce an immediate change gives rise to a breach of one or both of the Imperial duty and the contractual duty of trust and confidence.

476. The answer to Issue 22.2 is therefore that Holdings was not, and is not, constrained at any time after 31 March 2014 by Project Waltz from adopting and implementing the New ER Policy or any other new policy. Whether it can in fact adopt such a policy without giving notice, and if so what duration of notice, is not a question which arises as part of the remedies for breach by Holdings of its Imperial duty or contractual duty of trust and confidence. I have given some provisional views about some of the issues and will consider giving further decisions if I am invited by any party to do so.
477. As to Issue 22.3, the Old ER Policy applies to all members who left service prior to 31 March 2014 and does so in relation to all of their pension. I have not made any decision about members leaving service after that date since their position depends not on a remedy for the breaches of duty dealt with in the Judgment, but on whether a more general requirement for notice of any change of policy is required. If such notice is necessary, I have expressed the provisional view that effective notice was given taking effect from, but not earlier than, the end of October 2014 so that members leaving service after that date would not be able to take advantage of the Old ER Policy.
478. The answers to Issues 21 and 22 are given subject to the answers to Issues 23 to 28 to which I now turn.

Issues 23, 24 and 28: Redundancy and Retirement, the Separation Programmes and the Pension Waivers

23. In respect of the termination of C Plan Over 60 Members' contracts of employment under the Separation Programmes (where those members were not expressly granted early retirement in accordance with Active ER Terms):

23.1 whether either or both of the methods of termination of the said contracts – namely voluntary redundancy or involuntary redundancy – constituted "retirement" in accordance with the Active ER Terms?

23.2 if either or both of the said methods did not constitute retirement in accordance with those terms, whether the Active ER Terms would not as a result apply in respect of such members? and

23.3 if either or both of the said methods did not constitute retirement in accordance with those terms and the Active ER Terms did not as a result apply in respect of such members, whether the members whose contracts were terminated are entitled to any other remedy and if so what remedy?

24. In respect of the termination of Eligible ER Members' contracts of employment under the Separation Programmes:

24.1 whether either or both of the methods of termination of the said contracts – namely voluntary redundancy or involuntary redundancy – constituted (i) a "retirement" and (ii) "with the consent of the Principal Employer" in accordance with Active ER Terms?

24.2 if either or both of the said methods did not constitute retirement with the consent of the Principal Employer in accordance with those terms, whether the Active ER Terms would not as a result apply in respect of such members? and

24.3 if either method did not constitute retirement with the consent of the Principal Employer in accordance with those terms and the Active ER Terms did not as a result apply in respect of such members, whether the members whose contracts were terminated are entitled to any other remedy and if so what remedy?

28. In relation to the C Plan Over 60 Members or Eligible ER Members whose employment was terminated under one of the Separation Programmes, do the waivers purportedly entered into by such members on termination of employment by IBM:

28.1 disentitle the members from any of the remedies identified in Issues 23 to 27?

28.2 have some other, and if so what, effect on any remedies available?

479. The Trustee's written closing helpfully sets out in summary the background to these Issues and the relevant provisions of the Main Plan which I adopt and set out in the following three paragraphs.
480. The questions arise from a series of Separation Programmes (Phoenix, Mercury and Apollo) implemented by IBM in 2012, 2013 and 2014 respectively. These programmes involved rounds of voluntary and compulsory redundancies. The underlying question is whether members who were made redundant in one or the other of these ways were entitled to early retirement benefits under Rule 2(1) Schedule D of the Main Plan DB Rules. That Rule (which I will refer to simply as "Rule 2(1)") provides as follows:

"Main Plan, Schedule D

2 Early Retirement

(1) On the retirement of a B Plan Member, C Plan Member, E Plan Member or an N Plan Member before his Normal Retirement Date with the consent of the Principal Employer and providing the Member is aged 50 or above (or at any age in the case of serious ill-health where the provisions of Rule 3 of this Schedule shall apply), he shall (subject to the other alternatives listed in Rule 1 of Schedule F) be paid a pension calculated as for normal retirement (see Rule 1 of this Schedule), by reference to his actual Final Pensionable Earnings and to his actual Pensionable Service completed to the date of retirement (unless the provisions of Rule 3 of Schedule F apply in the case of a B Plan Member or C Plan Member (as the case may be) having completed less than two years' Qualifying Service). In the case of a C Plan Member such pension shall at all times be subject to the C Plan Maximum."

It should be noted that the requirement for consent in this Rule does not apply to C Plan Members aged 60 or over as established in the Rectification Action.

481. The pension paid on early retirement under this Rule is reduced for early payment. The early retirement discount factors ("ERDFs") used under this Rule are more favourable to members than those applied in the event of early retirement from deferred status. Issues 23 and 24 are about access to these favourable ERDFs via Rule 2(1).
482. The same issues do not arise in relation to the I Plan or DSL Plan as the relevant ERDFs are no worse for members who are made redundant, however the redundancy is characterised. In fact, for certain categories of DSL Plan members, there are improved ERDFs where the member retires having been made redundant. The areas of common ground and dispute appear to be these:

i) Issues 23.1 and 24.1: It is common ground that voluntary redundancy is "retirement" for the purposes of Rule 2(1). The Trustee argues that compulsory redundancy is also "retirement"; Holdings takes the contrary view. The Trustee argues that members who (i) took voluntary redundancy or (ii) were made compulsorily redundant did so with the consent of the Principal Employer; Holdings takes the contrary view.

ii) Issues 23.2 and 24.2: If the answer to any of the disputes above is contrary to the Trustees position, the Trustee argues that Rule 2(1) would apply; Holdings takes a contrary view.

iii) Issues 22.3 and 24.3: The Trustee argues that the affected members are entitled to damages where they would have been entitled to the benefit of Rule 2(1) but for Holdings' withdrawal of the Old ER Policy in breach of contract; the Trustee takes a different view.

iv) Issue 28: the Trustee's position is that the waivers are ineffective. Holdings' position is that the members concerned are contractually bound not to take the benefit of Rule 2(1) to which they would have been entitled but for the waivers.

483. The parties have used different descriptions of their respective contentions. In the competition between wide/narrow and inclusive/selective I propose to adopt the former which is a conventional way of describing competing constructions of this sort.
484. In order to put Rule 2(1) in context, I need to refer to a few other provisions of the DB Section. The Trustee has referred, in its written opening submissions, to many other provisions as well: I will deal with those as necessary as I go along. The provisions which I need to refer to at this stage are these:

i) I mention, but do not set out, the definitions of "Retiree", "Deferred Retiree" and "Postponed Retiree". A Retiree is a former Member receiving a pension under the Plan. Deferred Retiree is a former Member (having terminated Pensionable Service before Normal Retirement Date ("NRD") who is entitled to pension from some future date. A Postponed Retiree is a former Member who remains in Service after NRD with the express consent of the Principal Employer and with the knowledge of the Trustee. When the (future) pension of a Deferred Retiree or Postponed Retiree becomes payable, the Member ceases to be a Deferred Retiree or Postponed Retiree, as the case may be, and becomes a Retiree.

ii) In the context of those definitions, Member is a person who became a Member, Service means Service with a participating company and Pensionable Service means Service while a Member. Membership ceases under Rule 2 Schedule B when (among other occasions) the Member reaches his NRD or ceases to be a regular employee or where the Member makes a written request that membership should cease. This last occasion reflects the statutory right of an employee to opt out of his employer's pension scheme.

iii) It can be seen that where a Member leaves Service before NRD, he ceases to be a Member and becomes a Deferred Retiree; and a Member who remains in Service after NRD also ceases to be a Member and becomes a Postponed Retiree. A Member who opts out of the Plan before NRD, ceases to be a Member and thus becomes a Deferred Retiree. He becomes entitled to a deferred pension under Rule 2 Schedule F. I do not think that, on any view, such a person can be said to have entered into "retirement" within Rule 2(1) let alone retirement with the consent of the Principal Employer.

iv) The definition of "Normal Retirement Date" varies between different sections of the Main Plan. For an N Plan Member, it varies according to the period of Pensionable Service under consideration. For men, it is 65 save in respect of Pensionable Service for the period from 6 May 1990 to 5 April 1997 where it is 60. For women it is 60 in respect of the period up to 5 April 1997 and thereafter it is 65.

v) That definition replaced the definition found in the 1995 TD&R of the N Plan where for men, NRD was 65 and for women it was 60 or such later age up to 65 as the woman Member might select.

vi) For the C Plan, NRD is age 63 as it was under the 1995 TD&R.

vii) Rule 1 Schedule D ("Normal Retirement"): "(1) On the retirement of a Member on his Normal Retirement Date…." the Member is to be paid the pension set out in the Rule;

viii) Rule 2 Schedule D ("Early Retirement"): "(1) On the retirement of [a relevant member, including C Plan and N Plan Members] before his Normal Retirement Date with the consent of the Principal Employer and providing the Member is aged 50 or above"….. [the Member is to] be paid a pension calculated as for normal retirement…" based on actual service to and salary at the date of retirement. The pension is actuarially reduced. The Member does not have to take the early pension: he has the choices set out in Schedule F;

ix) Rule 3 Schedule D: ("Ill-health Early Retirement"): "In the event of the retirement of [relevant member, again including C Plan and N Plan Members] due to serious ill-health (irrespective of age)….." the Member is to be granted (at the discretion of the Benefits Allocation Committee) an immediate pension;

x) Rule 4 (original 1994 version): ("Late Retirement"): "(1) On the retirement of a Postponed Retiree….he shall be paid a pension…". The amount of the pension is calculated as at normal retirement date but subject to actuarial increase (ie a late retirement factor);

xi) Rule 4 (1997 version) (Late Retirement): "(2) In respect of a female N Plan Postponed Retiree the pension at the level described in Rule 1(c) [this is an error: it should be Rule 1(d)] of this Schedule shall continue to be granted to the female until the age of 65 or actual date of retirement, if earlier.";

xii) Rule 4 Schedule D (1997 version): ("Late Retirement"): "(4) The Postponed Retiree's pension shall commence from such date as shall be agreed between the Trustee and the Principal Employer and the Postponed Retiree PROVIDED THAT such date shall not be later than the date of his actual retirement from Service nor, if he left Service prior to his Normal Retirement Date, later than his 70th birthday.",

Something has gone wrong with the drafting here since a person who has left Service prior to NRD will not be a Postponed Retiree, since a Postponed Retiree is defined as a Member who has remained in Service after NRD. I consider that one must simply excise the words ", if he left Service prior to his Normal Retirement Date".

485. Schedule F deals generally with benefits following termination of pensionable service before NRD. Rule 1 provides that a member may receive his benefits in any one of the ways available to him which are listed. Where more than one of the alternatives is available, Rule 1(2) provides that the member may choose which to apply.
486. Schedule F is clearly intended to cover all situations of termination of pensionable service before NRD: thus the list includes receipt of an immediate early retirement pension and an immediate ill-health pension under Rules 2 and 3 Schedule D. Another listed item is a deferred pension under Rule 2 Schedule F itself. This provides that a member "who terminates Pensionable Service before his Normal Retirement Date….. shall (subject to the other alternatives listed in Rule 1 of this Section) become a Deferred Retiree….".
487. It is clear from the above that Schedules D and F are not mutually exclusive. Indeed, every member who falls within Rule 2 Schedule D will also fall within Rule 1 Schedule F and would be able, if he so wished, to take a deferred pension rather than an immediate pension. These Rules do not, like some other pension schemes, draw a sharp distinction between those who retire, on the one hand, and those who leave service early, on the other hand. Instead, the larger class – those who leave early – contains a sub-class – those who retire before NRD – with an immediate pension being payable only to those in the sub-class.
488. It is to be noted that Schedule F deals only with early leavers. Late leavers ("Postponed Retirees") are dealt within in Rule 4 Schedule D. If there is a category of late leaver (ie a person leaving service after NRD) in respect of whom leaving service does not amount to "retirement" within that Rule, then there does not appear to be any express provision in the Rules for him to receive any benefits at all. Mr Newman suggests that, whatever "retirement" might mean in Rule 2(1), it can include dismissal in Rule 4 so that the lacuna is filled. That is a point I will return to later.
489. The details of the three separation programmes are helpfully set out in Holdings' written opening submissions. It is useful to have that summary to hand: I set out the relevant paragraphs in Annex B to this judgment.
490. Before addressing the issues of construction of Rule 2(1), there are a few points to make.
491. The first is this. A number of cases have considered the meaning of the words "retire" and "retirement" in the context of pension scheme documentation. Some fine distinctions have been drawn. The message to be taken from the cases as a whole is that context is critical to ascertaining the meaning of the words used. Before turning to those cases, I remind myself of the proper approach to construction which I dealt with at Judgment [215] and [216] including reference to my earlier judgment in PNPF Trust Company Limited v Taylor [2010] EWHC 1573 [2010] PLR 261 at [127] to [145]. As before, I see no reason to depart from what I said there and I see no merit in repeating it. I add only that it would be hardly surprising, in the context of such a substantial document as the 1997 TD&R (and earlier iterations) to find gaps and minor inconsistencies. The court should not struggle to fill the gaps or avoid the inconsistencies if to do so would produce a result which, although literally full and consistent, fails to reflect what it can be said must have been the draftsman's intention.
492. The second point is that it is appropriate to construe the Rules in the context of the legislation (both tax legislation and social security/pensions legislation) and Inland Revenue/HMRC Practice Notes. This is admissible material: see for instance Mettoy Pension Trustees Ltd v Evans and others [1990] 1 WLR 1587 (see Warner J at p.1610H) and British Airways Pension Trustees Ltd v British Airways plc [2002] EWCA Civ 672, [2002] PLR 247 (see Arden LJ at [30]). However, care must be taken in carrying across to the rules of a pension scheme the meaning which particular words have in the legislation or Practice Notes. For instance, the word "retirement" may have one meaning in the Practice Notes. The rules should then be construed, if possible, consistently with the Practice Notes but that does not mean that the word "retirement" has to be given the same meaning in the rules. If, for instance, the rules spell out using different language the substance of what the Practice Notes provide, then there may be no need to give a particular word, such as "retirement" the same meaning in the rules as in the Practice Notes.
493. The third point is a longer one and relates to the various meanings which the words "retire" and "retirement" might have in different contexts. There is no doubt that the verb "retire" can be used in both a transitive ("X has been losing his edge recently and his employer has decided to retire him") and an intransitive sense ("X has had enough of his job and, now he is 64, has decided to retire"). Both are perfectly "natural" meanings although perhaps the intransitive sense is the more common. The noun "retirement" can be used to describe the result of either or both of a person retiring or being retired. It is all a question of context.
494. The same event may, at one and the same time, be seen as both a retirement and not a retirement. It depends on how the question is asked. Let me illustrate this by an example where 3 men, A, B and C, all aged 55, are carrying on the same job, let us say as senior assistant solicitors in the same firm, F LLP. They all decide to leave. A goes off to a better paid job as an assistant solicitor in another firm. B has decided on a change of career and has obtained a job as a forester. C, having inherited a large amount of money, has decided not to undertake any further paid work and has decided to devote his life to good causes and his vegetable garden.
495. If one asks the question in relation to each of them "Has he retired?" the answer might elicit another question which is the context in which the question is asked. Thus:

i) It can be said that each of A, B and C has retired from F LLP.

ii) It is clear that A has not retired in any other sense. He is still a practising solicitor and has not retired from the field of law.

iii) Although B can be said to have retired from practice of the law (as well as, like A, from F LLP), he has not retired from active employment altogether. If I were introduced to B over a pint of beer in the Woodman's Arms, I would not describe him as having retired (in abstract).

iv) C, in contrast, in his economically privileged position, has not only retired from F LLP and the practice of law, he has retired from active employment and can properly be described as having retired.

496. The fact that different answers can be given about retirement in relation to A, B and C, each of whom has left the service of F LLP indicates to my mind just how important the context of the use of the words "retire" and "retirement" is. It is the context which identifies what the question really is, when it is asked whether a person has retired or whether there is a retirement.
497. Before I leave the example, it will have been noticed that I have picked age 55 for the purposes of illustration. Most people – certainly the objective reasonable person – would see that as an age at which retirement, in the full sense of ceasing to be in paid employment, could take place. That, it seems to me, is why it is correct to say that A, B and C have all retired from F LLP. But if A, B and C had all been aged 30, it must be very doubtful that anyone would say that there had been a retirement of any sort. Save in exceptional cases, such as sports professionals, it would not be a normal to use the words "retire" or "retirement" in relation to a person who left his employment at that age. There is, perhaps, a more general exception: it is a perfectly normal use of the word to describe someone leaving service on grounds of ill-health as retiring even if that is at a comparatively young age.
498. In the right context, even dismissal can be described as retirement and similarly a person who is dismissed can be said to have been retired. In the present case, the Trustee's case is that the particular category of dismissal which is termed compulsory redundancy is said to be retirement within the meaning of Rule 2(1) although the Trustee accepts that the mere act of dismissing an employee by way of compulsory redundancy does not amount to the giving of consent for the purposes of that Rule adopting the approach to "consent" which I discuss in a moment.
499. One thing which is clear, in my view, is that the issue in the present case whether there has been a "retirement" within the meaning of Rule 2(1) has to be answered in the context of retirement from service with a participating employer. The issue is not whether the member has retired from all employment.
500. Both the Trustee and Holdings have addressed the issues treating "retirement" and "with consent" in a rather compartmentalised way. This is no doubt because the Issues themselves (in particular Issue 24) are compartmentalised in that way. I will adopt the same course. However, Rule 2(1) must be construed as a whole (and also in the context of the Rules as a whole) and the meaning of the word "retirement" is informed by the different situations in which retirement can occur. I must not lose sight of this in reaching my conclusions.
501. It is common ground that the consent referred to in Rule 2(1) relates to the taking of benefits rather than leaving service. Once the fact of retirement is established, the question is not whether that retirement was with consent; rather, the question is whether in relation to that retirement Holdings has consented to the payment of an immediate pension. This difference may, at first sight, seem immaterial. But it is in fact important because the different approaches lead to different arguments about the meaning of "retirement" in the first place. I shall refer to this as "the common ground approach" and to the approach that consent relates to leaving service as "the alternative approach".
502. Mr Short did not address why there was common ground on this point. He was content to adopt Mr Newman's reasons.
503. Mr Newman's contention was that the common ground approach is supported by four factors:

i) the archaeology of the DB Plans;

ii) the structure of Rule 2(1);

iii) a comparison between that Rule and Rule 1 Schedule F; and

iv) the purpose of the early retirement provisions.

504. As to the archaeology, it is potentially admissible: see most recently Sir Andrew Morritt V-C in Redrow plc v Pedley [2002] EWHC 983. In the present case, it might provide some assistance. The early retirement rule in the 1983 TD&R provided that an immediate pension could be paid on retirement "if the Principal Employer agrees that the Member may be offered an immediate pension". Clearly the focus was not on consent to retirement but on consent to payment of an immediate pension. If the member did not obtain that consent, he would become a deferred pensioner; the pension would become payable at the specified date (the later of leaving service and normal pension age under the preservation legislation). But if one goes back to the 1977 TD&R, one finds a rule which provided that an immediate pension could be paid "on retirement from service before the normal retirement date" where "such retirement is either … with the consent of the Principal Employer…" It is clear, in contrast with the 1983 TD&R, that the consent was to retirement not to payment of an immediate pension, although the latter would follow from the former. There had been a deliberate change of focus. The 1990 TD&R provided for an early pension "On the retirement of a [relevant member] before his normal retirement date with the agreement of the Principal Employer". It seems to me that the wording of Rule 2(1) is far closer to the 1977 TD&R than to the 1983 TD&R. I do not think that the archaeology provides a helpful pointer to the construction of Rule 2(1) since it might be said that the draftsman, far from wishing to continue the 1983 version, wanted to go back to the 1977 version.
505. Mr Newman draws attention to the provisions in the 1990 and 1991 TD&Rs which provide for deferred benefits to be payable from the NRD or any earlier date (within a 10 year limit before the normal date) subject to the prior agreement of the Principal Employer. The prior agreement is referable to the payment of the benefit, not the leaving service, and is, on his submission, entirely consistent with the corresponding consent requirement in the early retirement provision. I do not find that point even remotely persuasive. In each case – immediate or deferred pension – the trigger for the application of the relevant rule is leaving service. Bringing a deferred pension into immediate payment could not be linked to leaving service: the end of the member's employment would, by this time, be historic and the consent requirement could only relate to brining the benefit into payment. Indeed, in the 1977 TD&R, where the consent is to retirement rather than to the immediate payment of benefits, the deferred pension can be brought into payment "subject to the consent of the Trustees" and subject to the amount being determined by the Trustee which demonstrates the contrast between consent to retirement and consent to drawing a benefit.
506. I do, however, note at this point that the early retirement rule in the 1991 TD&R was, in all material respects, the same as Rule 2(1). The earlier rule (again Rule 2(1) – but then found in Schedule G (Retirement Benefits)) – falls to be taken into account when construing the new rule. The later provision does not necessarily have the same meaning as the earlier one since the 1995 TD&R was a replacement of the rules of the Plan and did not simply effect textual amendments. On a related point, Mr Newman submits that Rule 4(4) Schedule F cannot be relied on in the interpretation of the 1997 TD&R. That Rule was introduced after what is now Rule 2(1). That, as a matter of law, is wrong in my view. Rule 4(4) Schedule F forms part of the same TD&R as Rule 2(1) and must be taken account of in construing Rule 2(1). Again, the 1997 TD&R is a complete replacement of the pre-existing provisions; this is not a case of textual amendments to an existing set of provisions. This conclusion follows, in my view, from the Court of Appeal decision in Stena Line Ltd v MNRPF Trustees Ltd [2011] EWCA Civ 543 [2011] PLR 223 at [31] to [35], in particular the following passage from the judgment of Arden LJ:

"34. I accept Mr Spink's submission that, even though the very same clause is effectively readopted in the same form, its meaning may change on each re-introduction if the context in which it is re-adopted is materially different. Its meaning may be narrowed, or it may equally well have been widened, because of changes in the relevant background circumstances which fall to be taken into account in interpretation. Likewise I would also accept, as did the judge in paragraph 97 of his judgment, that it is possible that the meaning of a clause changes on re-adoption because there has been some material change in the scope or effect of some other clause in the period between its introduction and its reintroduction that has an impact on it……"

507. As to the structure of Rule 2(1), Mr Newman draws attention to what he calls the gateways to the provision of the deferred pension: (i) retirement (ii) before NRD (iii) with consent and (iv) after a minimum age (50) save in case of ill-health. The benefit "shall" then be paid (but this, I note, is subject to the other alternatives listed in Rule 1 Schedule F). He says that these four elements are not cumulative but independent. I do not understand that submission. There may be four gateways; but, on any view, the third of them is consent. That means consent to something, but, structurally, there is no reason why that should be consent to payment rather than consent to retirement.
508. As to the comparison with Rule 1 Schedule F, this is something I will look at further in a moment in addressing the submissions in relation to the meaning of "retirement". But so far as the meaning of "consent" in Rule 2(1) is concerned, I do not find the comparison between that Rule and Rule 1 Schedule F to be of any material assistance. Rule 1 Schedule F simply lists the range of alternatives which might be open to a member "who terminates Pensionable Service before his Normal Retirement Date". Where more than one alternative is open to a member, he may choose which is to apply to him. Again, I do not consider that the structure of Rule 1 Schedule F, whether read by itself or together with Rule 2(1), provides any pointer to whether the consent referred to in Rule 2(1) is focused on retirement or immediate receipt of benefits.
509. Mr Newman's fourth factor relates to the purpose of the early retirement provision with its favourable ERDFs. The requirement for consent can, no doubt, be partly explained from a funding perspective: the actuary would require a greater contribution than was actually required if there had been a right to take early retirement at, say, age 60 or even earlier. It can also be partly explained as an instrument for managing a suitable turnover in its workforce by providing it with a mechanism for encouraging employees to leave. But that cannot be the whole explanation since consent was not required in relation to C Plan Over 60 Members.
510. What can be said, I think, is that one purpose of the early retirement provision was to allow Holdings to decide, at least in the case of N Plan Members, whether to allow access to the early retirement pension and this is so whether consent is focused on retirement or on receipt of benefits. It seems to me that this is a pointer to a wider, rather than a narrower, construction of "retirement" in Rule 2(1) (although whether it goes as wide as the Trustee submits is a slightly different question). This is because it is reasonable to think that Holdings would have wanted to retain as wide a discretion as was reasonable and sensible so that the scope of "retirement" should be generously interpreted. As a good employer, Holdings would recognise that it needed to treat its workforce not simply in accordance with its implied duty of trust and confidence, but also fairly. It may for commercial reasons face the need on occasions to make individual employees or groups of employees compulsorily redundant but, at the same time, might want to provide them with early retirement pensions in accordance with the favourable ERDFs. It would be entirely unsurprising to find an early retirement provision which permitted Holdings to do precisely that; and so it would be unsurprising to find that Rule 2(1) itself could be construed to allow precisely that, without the need for Holdings to obtain an amendment to the TD&R to achieve that result. Rule 9 Schedule D provides for increases to pensions in payment, but I have been unable to find a provision providing for increases in other benefits (other than by amendment).
511. Mr Newman says that the Trustee's wide construction of "retirement" would be wide enough to cover dismissal for cause eg incompetence or misconduct. Relying on the (unidentified) purpose of the early retirement provision, he says that it cannot be correct that there should be an automatic entitlement to the advantageous early retirement terms in those cases. This point, he submits, falls away if consent is not consent to retirement but is rather consent to the immediate payment of benefits: even if dismissal is "retirement", it is not with consent unless consent is actually given, and nobody suggests that consent is an automatic consequence of dismissal.
512. I observe, however, that the point does not entirely fall away as Mr Newman suggests. Rule 2(1), as rectified following the Rectification Action, applies to the C Plan. C Plan Over 60 Members are able to retire without consent over the age of 60. Accordingly, if compulsory redundancy or dismissal for cause are "retirement" within Rule 2(1), a C Plan Over 60 Member would be entitled to an immediate pension applying the favourable discount factors. But if it is contrary to the purpose of Rule 2(1) to allow access to the favourable ERDFs in the case of Eligible ER Members, it is just as much contrary to that purpose in the case of C Plan Over 60 Members. And so, if effect is to be given to that purpose, then it is "retirement" which has to be given a narrow meaning since the problem is not solved by an appeal to the need for consent.
513. I propose to address the meaning of "retirement" in Rule 2(1) without at the moment resolving this issue of the focus of consent.
514. Issue 23.1 and 24.1 ask whether voluntary and compulsory redundancy constituted "retirement" for the purposes of Rule 2(1). It is conceded by Holdings that voluntary redundancy did so both in relation to Eligible ER Members and C Plan Over 60 Members. The outstanding questions arising out of Issue 23.1 and 24.1 are dealt with in the following paragraphs.

Issues 23.1 and 24.1 (paragraph (i)): Is compulsory redundancy "retirement"?

Holdings' submissions

515. Although Holdings accepts that voluntary redundancy is capable of being "retirement" within the meaning of Rule 2(1), it maintains that compulsory redundancy is not capable of being "retirement". Its primary approach is based on the common ground approach; its secondary approach is based on the alternative approach that the focus of consent is on retirement. The Trustee's case is that compulsory redundancy is "retirement": that is so whichever approach to consent is correct, although certain arguments which are open to it on the common ground approach are not open on the alternative approach.
516. One important case in this area relied on by Mr Newman is the decision of the Court of Appeal in AGCO Ltd v Massey Ferguson Works Pension Trust Ltd [2003] ECWA Civ 1044;[2003] PLR 241 ("AGCO"). This case considered whether an employee who had taken voluntary redundancy or who had been made compulsorily redundant "retires from service at the request of the employer". It was held by the majority that an employee who had been made compulsorily redundant did not come within that provision. The decision of Neuberger J to the contrary effect was reversed. Sedley LJ, siding with Neuberger J, dissented, deciding that, by parity of reasoning with the case of voluntary redundancy on the wording at issue, compulsory redundancy, too, led to the conclusion that an employee was retiring at the request of the employer.
517. Although AGCO was concerned with the meaning of a particular word in a particular context, the judgment of Rix LJ, in particular, is important in two respects. First, it distils the learning at that stage on the meaning of the word "retire" and demonstrates the proper approach to construction. Secondly, it removes the possibility of the temptation (which I would otherwise be unable to resist) of following the lead of Neuberger J in thinking that the authorities are of little assistance. That possibility is removed by the apparently disparaging words of Rix LJ at [63] of his judgment. After reviewing the authorities, he said this:

"I would therefore respectfully differ from the view of Neuberger J, who appears to have regarded the authorities as of little assistance, but also to have considered that such assistance as they gave supported the employee's primary contention of the possibility of a wider meaning for the word "retires"…"

518. I interpose here to say that, in reaching my conclusions, I bear in mind not only AGCO itself but also the authorities referred to in it, in particular Dorrell v May & Baker Ltd [1991] PLR 31 and Harris v Lord Shuttleworth [1994] ICR 991.
519. The relevant rule in AGCO appeared in Section A headed "Pensions on Retirement". Rule 13 started by providing that members should become entitled to a pension on the occurrence of various events: the introductory words did not include "retire" or "retirement". There followed three paragraphs ((a) to (c)) each of which contained a sub-heading starting with the words "Pension on retirement" followed by three different phrases: (i) "at the normal retirement date", (ii) "after the normal retirement date" and (iii) "before the normal retirement date". Under the relevant heading the paragraphs provided as follows: the first paragraph provided for a pension "on the retirement of a member from service….". The second paragraph provided for a pension where "by agreement with the employer a member remains in service and retires after the normal retirement date".
520. The third paragraph contained a number of sub-paragraphs providing for a pension:

i) where the member "retires from service…. on account of incapacity":

ii) where the member "retires from service …. of his own free will".

iii) where the member "retires from service at the request of the employer…" after his 50^th birthday. It was the meaning of this last paragraph which was in issue.

521. As to these, Rix LJ identified each of them as not being concerned with dismissal. Thus:

i) the first was not a retirement which depends on a resignation or a dismissal, but is in truth one of frustration and thus a real case of retirement from work: see at [56];

ii) the second was a pure case of voluntary, unpressured resignation, which was again prima facie incompatible with dismissal: see at [57].

522. As to the third, Rix LJ carried out a detailed analysis at [58] to [62] to show why "retires" did not, in the context of rule 13 as a whole, cover compulsory redundancy. In particular, he expressly referred to the use of the word "retires" in paragraph (iii) as plainly being used in an intransitive sense. The essence of his reasoning is found in [62].

"In sum therefore, as a matter of language, the words "retirement" and "retires" as found in rule 13 naturally cover situations distinct from dismissal. Dismissal does not naturally fall within the sense of those words. In particular, the rule 13(c)(iii) context of retirement "at the request of the employer" does not naturally embrace dismissal. Rule 23, on the other hand, with its broader and different language "leaves service", appears designed for situations going beyond that of rule 13's retirement. If "retires" in rule 13 means simply "gives up work" and otherwise takes its meaning entirely from its context, then it seems to me that there is no need for a separate expression relating to "retirement". Mr Furness suggests that "retirement" is a special word simply used for situations of various kinds where an employee is entitled to an immediate, as distinct from a deferred pension, and that those situations are all identified in rule 13. It is true that rule 13 is dealing with cases of immediate payment of a pension, but there is no reason why, if those cases might arise in any situation (within the stated parameters) where an employee leaves service, rule 13 would not itself have been written in terms of "leaves service". The fact remains that the basic rule is that an employee is entitled to a normal retirement pension at the normal retirement date and that the basic rule is that if he leaves service early then he is still entitled to that pension at that time. One comes back, therefore, to the fact that a normal retirement pension is measured against retirement in the ordinary way at the retirement age."

523. Mr Newman submits that Rix LJ's reasoning is precisely applicable to the Rules in question in the present case with the result that a compulsory redundancy is not an occasion of retirement. Thus:

i) "retirement" in the normal retirement rule found in Rule 2 Schedule D is to be seen as the entirely consensual event of ceasing work at the end of a working lifetime, and does not encompass dismissal;

ii) the ill-health early retirement rule in Rule 3 Schedule D is, as in AGCO, properly to be seen as the frustration of the employment contract rather than a dismissal;

iii) there is no other provision in the retirement benefits schedule (ie Schedule D) which uses the words "retires" or "retirement" to encompass enforced dismissal: for the early retirement provision to do so would, again as in AGCO, therefore be the first and only such occasion;

iv) in sum, as a matter of language, the word "retirement" as found within these provisions does not cover cases of dismissal;

v) this can be contrasted with the general deferred benefits provision Schedule F;

vi) the purpose of Rule 2(1) Schedule D is to enable Holdings to grant early retirement on preferential terms to employees over 50; and to extend that concept to enforced dismissals would undermine that concept. Dealing with this particular point in order to get it out of the way, it carries no weight, in my view. To categorise the purpose in that way is simply an attractive way of asserting the conclusion which Mr Newman wants to reach. Even if dismissal can amount to retirement, that retirement still has to be with Holdings' consent before entitlement to the beneficial terms of Rule 2(1) is available. The Trustee does not contend that dismissal, of itself, automatically provides the necessary consent. If there is a point to be made, it seems to me to be a point in favour of the Trustee rather than Holdings: it is that there may be occasions where Holdings wishes to dismiss an employee but also wishes the member to retire on the beneficial terms of Rule 2(1). The wide construction allows this beneficial treatment to be afforded through the simple mechanism of giving consent (although, no doubt, the object could be achieved by the use of other rules, including use of the amendment power if necessary).

524. Mr Newman finds further support for his submissions concerning the meaning of "retirement" in the case of Akester v Kingston Communications (Hull) plc [2005] PLR 153 ("Akester"). He also relies on it in support of the proposition that a compulsory redundancy cannot be seen as a situation in which an employer has given "consent" or "agreement" to a retirement.
525. In that case, the relevant provision applied where the member "retires from Service", in contrast with other provisions, where the right to benefits – including the right to immediate but actuarially reduced benefits - arose where the member "leaves Service": see at [9].
526. The Judge considered that the references in the rules to members "leaving Service" were of the widest generality, which would, on their face, apply to any of the various circumstances by which, as a matter of law, the contract of employment might come to an end: see at [101].
527. He contrasted this with the phrase "retires from Service", holding that the core meaning of the word "retires" involves a decision by an employee to bring his employment to an end on the grounds of increasing age or incapacity: see at [104]. That concept, as used in the relevant early retirement rule, did not apply to cases where the employer imposes its will upon an unwilling employee: see at [108].
528. The Judge applied AGCO, holding that it was entirely applicable to the circumstances of the case before him: see [166], [169]. He concluded that cases of compulsory redundancy could normally be expected to fall outside of the enhanced early retirement provision: see at [175].
529. The members had sought to argue that, even in cases of compulsory redundancy, the employees concerned would still come within the enhanced early retirement provision, provided that there was a sufficient degree of consent to the redundancy process, even if that consent was forthcoming only after the decision to make him redundant had been taken by the employer: see at [178]. That submission was rejected by the Judge, who held that, where the terms for dismissal were negotiated and agreed in circumstances where the employee was still being required to leave employment, that was not "retirement" for the purpose of the enhanced early retirement provision: see at [186] to [187] and [193] to [194].
530. Applying this approach to the present case, Mr Newman submits that there was nothing in the processes under the Separation Programmes which rendered the compulsory redundancies anything other than dismissals. Moreover, no employee who was made compulsorily redundant under the Mercury and Apollo programmes could reasonably have thought that, in so doing, they were "retiring", rather than being dismissed:

i) the employee information packs provided to such members in the Mercury programme stated that: "As an employee made redundant under this programme you will become a deferred member of any IBM pension plan; you cannot retire from service";

ii) the offer document provided to such members in the Apollo programme stated that: "As you are being dismissed on involuntary redundancy terms, you are not able to retire from Service".

531. And so Mr Newman therefore submits that the involuntary redundancies of employees under the Separation Programmes - even where those members signed settlement agreements in the course of being made so redundant – did not constitute "retirement" for the purposes of Rule 2(1), so that such employees are not entitled to the benefits ordinarily applicable to members who retired early from active service.

The Trustee's submissions

532. The Trustee's case in nutshell is this:

i) It turns partly on the proposition that the "retirement" and related terms in the governing documentation of the Main Plan DB Rules is a concept which is used to signify a particular state of affairs rather than to that state of affairs only if brought about in a particular way. Thus "retirement" does not exclude cases (i) where the member was retired (ii) where he was dismissed by reason of redundancy (iii) where he was dismissed by reason of compulsory redundancy; and it occurs when a member ceases to work for Holdings or, if some other factor is required, when the member takes or seeks to take benefits from the Plans after ceasing to work for Holdings.

ii) Various provisions concerning retirement after NRD do not work on the narrow construction. Further, the DB Rules would have been in breach of relevant Inland Revenue/HMRC requirements in 1997 if the narrow construction is correct.

iii) It is asking too much of the word "retirement" to distinguish between different types of dismissal and in particular between voluntary redundancy (which involved dismissal) and compulsory redundancy.

533. Mr Short refers to the definitions of Deferred Retiree, Postponed Retiree and Retiree. I do not find these of assistance save that the definition of Retiree means a person in receipt of a pension. It thus includes an early leaver whose benefit has since come into payment. It is said that the use of the language of retirement (ie "Retiree") in relation to such a person lends some support to the suggestion that the occasion on which he left service was a retirement. I do not agree with such a suggestion. Retiree is really being used as a synonym for pensioner: if one were to use the word "Pensioner" instead of "Retiree" in the definitions of Retiree, Deferred Retiree and Postponed Retiree, the meaning becomes clear.
534. Mr Short identifies a similar anomaly to that which was to be found in AGCO. On the wider view there is a retirement where a member is dismissed after NRD by reason of compulsory redundancy and even dismissal for cause. However, on the narrow view, there would be no "retirement" so that Rule 4(4) would not apply with the result that the Late Retirement factor would not be available. Indeed, it is not easy to see under what provision such a person would be entitled to any pension. He also says that this would breach the anti-forfeiture provisions of section 92 Pensions Act 1995: the court should therefore construe the provision in a way which avoids the section applying. He may be correct in saying that section 92 would be engaged, but that is not a killer point since, as Mr Newman points out, the effect of the section would be to override the Rule if it would otherwise give rise to an inadmissible forfeiture.
535. Further, this result is, he says, nonsensical and would leave the Plan in breach of HMRC requirements. On the narrow construction, Rule 4(4) does not require, and no other Rule requires, the pension to come into payment when the member leaves service since there will be no "actual retirement". This would not comply with the then Inland Revenue Practice Notes (IR12) which required pensions to come into payment by at latest age 75. This is a pointer to the rejection of the narrow construction and adoption of the wide construction under which this problem would not arise. Mr Short submits that "retirement" cannot have different meanings in Rule 4 and in Rule 2, to which Mr Newman asks "Why not?" and submits that the obvious answer to this problem is that it does have different meanings.
536. Rule 5 Schedule D relates to the lump sum option. Under Rule 5(5) a Postponed Retiree is able to receive a lump sum either on reaching NRD or on later retirement. On the narrow construction, compulsory redundancy will not give rise to a retirement so that a Postponed Retiree who did not elect to receive his lump sum at NRD is not entitled take one later. It seems to me that this point adds nothing to the one mentioned in the preceding paragraph of this judgment. In particular, if "retirement" in Rule 4 is given the wider meaning, then the reference in Rule 5(5) to "later retirement" would, in my view, mean "later retirement in accordance with Rule 4".
537. Mr Short also refers to various paragraphs of Schedule F. I have already adequately described its provisions. There is I think a slight indication in favour of the wide construction to be found in the words "A Member who terminates Pensionable Service" at the start of Rule 1. There can be no doubt, in my view, that Schedule F embraces all types of termination of Pensionable Service including dismissal and compulsory redundancy. A Member who is dismissed is nonetheless to be seen as one "who terminates Pensionable Service"; the draftsman did not see any need to use a different phrase such as "whose Pensionable Service terminates". If the verb terminate can be used in a transitive sense, why, it might be asked, should "retirement" not likewise be seen as encompassing a transitive use of the verb "retire". Mr Short makes the point, which I have made already in slightly different language, that "terminates Pensionable Service" and "on retirement" are not mutually exclusive given that Rule 1(1) encompasses members who are entitled to a pension on retirement under Rules 2 and 3 Schedule D.
538. In contrast, Mr Newman says that Schedule F supports his construction. There is a distinction between "retirement" in Rule 2(1) and "terminates Pensionable Service" in Rule 1 Schedule F. Mr Newman submits that, on the Trustee's wide construction of "retirement", there would be no need for two different concepts. I do not agree. There is clearly a need to provide for the early benefits of an early leaver who is not, on any footing, entitled to an immediate pension, for instance a person leaving service under the age of 50 or a person with less than 2 years' qualifying service. Rule 1 Schedule F is there to provide an exhaustive list of the available choices which might arise although in any particular case only some of the choices will be available. Thus "retirement" is a sub-set of termination of Pensionable Service, but that fact does not tell us anything about the contents of that sub-set.
539. Mr Short has included a long section in his written opening submissions dealing with the fiscal and legislative background against which the provisions of the Rules are to be construed. He submits that the word "retirement" should be construed consistently with the use of the same word in the tax legislation in the absence of a strong indication to the contrary. I do not propose to go through the many statutory provisions discussed in his historical exegesis. But there is one aspect I do need to deal with. It relates to the definition of "retirement or other benefit" which has continued in substance in the same form since section 23 (1) Finance Act 1947 right through to section 612 Income and Corporation Taxes Act 1988 (in force at the date of the 1997 TD&R). That definition makes no reference to benefits taken from deferral and so it is suggested that it is unclear whether they amount to benefits taken on retirement. This is important because, if they do, it will include benefits taken by those who have been dismissed. Further, although it is the case that a scheme could obtain discretionary approval for tax purposes even though it included provision of some benefits which did not fall within the definition of retirement or other benefits, deferred benefits taken are not mentioned in the list setting out the circumstances in which the discretion to approve a scheme could be given. The list was not exhaustive but it would be odd for such a common form of benefit not to be included in the list.
540. The answer to this, although it is not necessary to decide the point finally, appears to me to be found in the definition of "service" also originally found in section 23 FA 1947:

"service" means service as an employee or director of the body corporate in question, and "retirement" shall be construed accordingly"

541. Thus "retirement" is seen as the state of a person who was, but no longer is, in service by reference to a particular employer. Even a 30-year old who leaves service voluntarily would be seen as in a state of retirement. He would not, in accordance with the modern requirements for exempt approval, be able to receive his benefits until a much later age but when he does receive them, they will be benefits "in connection with past service, after retirement…" within the words of the definition. Similarly, when an employee is dismissed he ceases to be in service: like the 30 year old just mentioned, he may not become entitled to immediate benefits, but when he does become entitled to them, they again qualify, it seems to me, as benefits "in connection with past service, after retirement…". It can be seen, therefore, that the tax legislation contains a wide express definition of retirement. One question in the present caseis whether, and if so how, that definition and the definition of "retirement or other benefit" feeds into the interpretation of the Rules.
542. Mr Short has referred to Hoover v Hetherington [2000] PLR 297 (Pumfrey J) and to Venables v Hornby [2003] 1 WLR 3022 (CA) to demonstrate the impact of the statutory provisions on the interpretation of pension scheme rules. I do not find Hoover of any assistance in the interpretation of the particular provisions with which I have to deal.
543. As to Venables, although the actual decision was reversed on appeal to the House of Lords, much of what the Court of Appeal said withstood scrutiny by the higher court. Particular reliance is placed by the Trustee on this passage from the judgment of Chadwick LJ:

"With those provisions [relevant tax legislative provisions] in mind, it seems to me that - in the absence of some strong contra-indication – the parties to the trust deed must be taken to have intended that words and expressions in the trust deed should be construed in the same sense as that in which the same words and expressions would be understood in the context of the legislative provisions under which approval of the scheme established by the trust deed was to be sought. …

The Revenue, at least, could be expected to decide whether or not to approve the scheme on the basis that words and expressions in the trust deed were intended to be construed in the context of the legislative provisions unless it was made clear that that was not the parties' intention. If the parties intended otherwise, it was for them to say so."

544. Some care must be taken in applying that approach in any particular case. In particular, the word "retirement" is defined in the legislation by reference to "service" so that retirement and leaving service (for whatever reason) are reflective of each other. The second paragraph which I have quoted from Chadwick LJ is no doubt correct. The Revenue could be expected to satisfy themselves that the terms of the scheme in question provide benefits which are consistent with the grant of discretionary approval. It does not follow that the words "retire" or "retirement" have to be construed as having the wide meaning of the statutory definition. It would be a matter of indifference to the Revenue whether dismissal of a member over the age of 50 gave rise to a right to an immediate pension applying the generous ERDF or whether it gave rise only to a deferred pension. If Chadwick LJ's approach is applied in the present case, it means that every occasion of leaving service is a "retirement". But if that is right, it immediately gives rise to the question why Schedule F introduces a different concept when referring to "benefits following termination of pensionable service" (in the title to Schedule F) and to "terminates Pensionable Service" (in Rule 1). Why, I ask rhetorically, did not Rule 1 Schedule F simply refer to a Member who retires rather than one who terminates Pensionable Service if Chadwick LJ's approach is to be blindly applied?
545. Mr Short suggests that the narrow construction relies on a flawed premise, namely that retirement is consensual whereas dismissal is a unilateral decision by the employer. Mr Newman disavows that suggestion. Holdings accepts that voluntary redundancies fall within the concept of retirement; what Mr Newman says is that "retirement" is limited to resignations and voluntary dismissals but does not extend to involuntary dismissals such as compulsory redundancy.
546. Whether the flawed premise is there or not, I think I should address Mr Short's submissions on this topic. I agree that retirement is not always consensual: for instance, in the present case a male Eligible ER Member might resign at say, age 61 but without Holdings' consent: he would be a person who enters into retirement within the meaning of the Rules, it seems to me, notwithstanding that he does not become entitled to an immediate pension. Mr Short contends, in any case, that authority is against the proposition that retirement is always consensual. He relies on the decision of the House of Lords in Waite v CGHQ [1983] 2 AC 714. I do not perceive in that decision any generally applicable approach to the meaning of retirement. Lord Fraser referred to Mr Waite variously as having been dismissed, having been retired; and he used the phrase "compulsory retirement" in the course of his discussion. It is not contentious that "retire" and "retirement" can be used in a transitive sense if that is what the context indicates. It may be a contentious proposition – but it ought not to be and I think that Waite shows that it is correct – that the word "retirement" can include the state of affairs resulting from dismissal; moreover, the transitive use of the verb "retire" is entirely apposite to certain sorts of dismissal. Thus Mr Waite was dismissed on 30 April 1980 when he was aged 60½; he was described by Lord Fraser as compulsorily retired. The use of the word "retired" is entirely apposite in the context of the termination of Mr Waite's service; but there is no doubt that he was dismissed.
547. Other cases demonstrate the use of "retire" (in the transitive sense) being applied to dismissals, with judges referring to an individual in different places in a single judgment or decision as being dismissed and as being retired, in relation to which see two cases referred to by Mr Short, Gidella and Others v Wandsworth Borough Council and Another [2002] 3 CMLR 37 and Power v Regent Security Services [2007] UKEAT 0499/06.
548. Thus I agree with Mr Short that retirement and dismissal (and particularly compulsory redundancy) cannot, a priori, be treated as binary alternatives. In particular cases, they may, however, be mutually exclusive and that would appear to have been the result in AGCO.
549. As to the way matters operated on the ground with Holdings, Mr Short submits that the contracts of employment did not provide for automatic termination at retirement age. He submits accordingly that even ordinary retirement at the normal retirement age needs to be implemented by a formal termination of the contract of employment, in other words by dismissal.
550. For that proposition, he relies on the various iterations of the IBM Employee Handbook that deal with "Termination", which he says do not suggest otherwise. It should, however, be noted that the Handbooks do state that the normal retirement age is 63 and that late retirement is not permitted. Although, as Mr Short submits, a normal retirement age is not necessarily the contractual retirement age (he relies on Royal & Sun Alliance v Payne [2005] UKEAT 0122/05 at [28]), the question is one of construction of the employment contract. Under the Handbook which Mr Short gave as an example (which is all I have looked at in this context and is, I assume, typical), the "normal retirement age" for all permanent employees is stated to be 63. But "actual retirement age" is as defined in the pension plan to which the member belongs. Quite what the contractual retirement age is in the light of those provisions is not entirely clear. Further, although late retirement is stated to be impermissible, there would be nothing to prevent Holdings from extending a member's employment beyond the defined retirement age if it so agreed with the member in which case the contract would not come to an end. Again, by way of example, a female member of N Plan (having a normal retirement age of 60) was able to defer retirement until 65; if she elected to do so, the contract of employment would then be extended to that age but, absent any extension, would in my view come to an end at age 60.
551. I do not disagree with Mr Short's general proposition that a contractual retirement age will not necessarily bring the contract automatically to an end. It may instead be the date at which members can be obliged to retire. He relies on Wall v British Compressed Air Society [2004] ICR 408, where Simon Brown LJ said:

"it is, surely, normal not to be compulsorily retired until one has reached one's contractual retirement age, ie the age stipulated for automatic retirement, alternatively the minimum age at which one can be obliged to retire".

552. There are, in any case, various atypical employees whose contractual retiring age was later than age 63. This was the case with male N Plan members (who are described in the 1997 Handbook as having an "actual" retiring date of 65) In addition, some employees who transferred into IBM under the Transfer of Undertakings (Protection from Employment) Regulations 1981 may have (and there were many in 1996) had contractual retiring dates of over age 63. In those cases, it would have been necessary (in any case where the employee did not wish to leave) for Holdings to dismiss them at age 63 in order to enforce its then policy against late retirement, that is to say after age 63. As to that, it would be surprising to my mind if an employee who had a contractual retirement date of 65, but who was dismissed at age 63 in accordance with a policy concerning late retirement generally, was not then to be regarded as having entered into "retirement" within Rule 2(1).
553. I turn next to consider the position of a member who leaves service on account of serious ill health. Quite clearly, a member who is given notice on the basis of serious ill health is dismissed (not wrongfully, but dismissed nonetheless) and quite clearly he falls within Rule 3(1) (ie within the words "In the event of the retirement of a [member] due to serious ill-health (irrespective of age)….."). It cannot sensibly be argued, in my view, that his departure is on "retirement" if he leaves voluntarily but not "retirement" if he is dismissed. That conclusion is consistent with the approach in Harris v Lord Shuttleworth [1994] ICR 991 where Glidewell LJ said this:

"In my judgment, whether she gives notice of her intention to leave or the society gives notice dismissing her, the termination can still properly be described as "retirement from the service … by reason of incapacity"

554. In AGCO, Rix LJ did not suggest that this was wrong (stating: "all this court was saying in [Harris] was that the word "retirement" in context covered all circumstances in which an employee left service by reason of permanent incapacity"). It is relevant to note that Rix LJ saw retirement as a result of incapacity as being, in effect, a case of frustration. It is important to see what he meant by frustration in this context. He uses the word on a number of occasions including these:

i) In [46], when considering the decision in Dorrell v May & Baker (see above) he talks of an employee's service ending through frustration rather than dismissal.

ii) In [49], he observed that the issue of frustration did not arise in Brooks v National Westminster Bank Ltd (unreported: 8 November 1983, Court of Appeal) but went on to say that, if the contract is frustrated, then it cannot have been intended to make a critical difference whether the employee's departure is initiated by him or by the employer.

iii) In [56] he referred to rule 13(c)(i) which spoke of permanent incapacity stating that "Such a case is in truth one of frustration".

iv) In [60] he spoke of a retirement under that paragraph as "the frustration of his contract of employment".

v) In [79] he referred to "retirement" as a euphemism when applied to a person retiring through incapacity adding that the contract of employment "has come to an end by frustration".

555. It seems clear to me, from those references, that Rix LJ was using the word "frustration" as a term of art in the sense in which it is understood in contract law. He was not using it simply as a description of an event ie the termination of a contract of employment because the contracting party is unable to perform it. Although that would be a perfectly accurate description, it does nothing to illuminate the meaning of the word "retirement" in Rule 2(1).
556. In the present case, Mr Short submits that there is no question of frustration in the technical sense of that word in relation to contracts. He points out that the members' contracts of employment made express provision for serious ill health and permanent incapacity to be dealt with under the IBM Sickness and Accident Plan which is referred to in the Employee Handbooks and/or the ill health retirement provisions of the Plans, see Rule 3 Schedule D. Together, the Sickness and Accident Plan and the ill health retirement provisions make express and complete provision for cases of permanent incapacity.
557. In support of that contention, Mr Short relies on Villella v MFI Furniture Centres Ltd [1999] IRLR 468. In that case it was held that incapacity could not frustrate a contract of employment where the contract expressly foresaw and provided for long-term incapacity under a permanent health insurance (PHI) scheme. HHJ Green QC, sitting in the QBD, said (in a passage quoted by Mr Short in his written opening submissions):

"[48] … The contract expressly foresaw and provided for long-term incapacity due to illness. It follows inevitably that the later occurrence of such long-term incapacity cannot frustrate the contract. In Paal Wilson & Co AS v Partenreederei Hannah [1983] AC 855, at 909, Lord Brandon said this:

'Those pronouncements, which I do not consider that it is necessary for me to quote again myself, show that there are two essential factors which must be present in order to frustrate a contract. The first essential factor is that there must be some outside event or extraneous change of situation, not foreseen or provided for by the parties at the time of contracting, which either makes it impossible for the contract to be performed at all or at least renders its performance something radically different from what the parties contemplated when they entered into it.'"

In the light of that authority, I accept Mr Short's submission that the present case is not one of frustration in the technical sense in which that word is used in contract law.

558. The next class of leaver to consider consists of members who are dismissed for misconduct. Mr Short's submission is that their position does not point to the narrow meaning of "retirement". In this context, there is a potential anomaly (I will call it the "misconduct anomaly"). If "retirement" is given the wide meaning for which the Trustee contends, then a member who is dismissed for misconduct nonetheless enters into retirement (at least if he is over 50 being the age at which the Rules of the Plan at least envisage the possibility of retirement) and becomes entitled to the beneficial early retirement pension. Mr Short submits that there is, in fact, no anomaly because, under the common ground approach, access to the advantageous ERDFs is still controlled by the need for consent. There is, accordingly, no misconduct anomaly to point to Holdings' narrow construction. It is true, I agree, that there is no such anomaly so far as concerned the Eligible ER Members on that approach to consent.
559. However, as I pointed out in paragraph 509 above, it is not true so far as concerns the C Plan Over 60 Members who do not require consent to take early retirement. The misconduct anomaly can be wholly eliminated, if it needs to be eliminated, only if "retirement" is given a meaning which excludes dismissal for misconduct. But if it excludes that sort of compulsory dismissal, where is the line to be drawn between what is, and what is not, "retirement"? That is the real issue which I must answer.
560. I used, in the preceding paragraph, the words "if it needs to be eliminated". One view is that it does not need to be eliminated and that is simply something which has to be lived with just as was the case in AGCO. Restricted to C Plan Over 60 Members, it can be said that the anomaly does not amount to much and that to allow it to dictate the meaning the word "retirement" would be a classic case of dog-wagging. In any case if a member has an accrued right in the sense that he can access it immediately of his own volition (by resigning) it is not immediately obvious that he should forfeit the beneficial ERDF if he is dismissed for misconduct let alone if he is dismissed without misconduct.
561. Mr Short has, of course, addressed AGCO and Akester. He points out that what was at issue in AGCO was not the meaning of the noun "retirement" in the context of retirement with consent but the meaning of the verb "retire" in the context of a rule which applied where a member "retired at the request of the employer". He says that AGCO is distinguishable and does not require a distinction to be drawn in the present case between compulsory and voluntary redundancy. He perceives these distinctions on which he relies:

i) the wording of the early retirement provision was different from that in the present case. The particular formulation was crucial to the Court of Appeal's reasoning, see for example AGCO at [58] where Rix LJ's focus is on the phrase "at the request of the employer" which could not, in his view, cover cases where the employer not merely requests but successfully enforces retirement. This clearly coloured his interpretation of the meaning of the words "retirement" and "retire".

ii) the schemes as a whole are different. For example:

a) in the present case there is no 60/50 anomaly nor any misconduct anomaly as there was in AGCO: see at [28];

b) the "late retirement issue" in AGCO was answered by the precise wording of rule 13(b), see [54]. There is no wording anything like that in the present case;

iii) there are distinguishing features arising from the facts of the present case. In particular, the treatment of retirement and dismissal as binary alternatives cannot be maintained by reference to frustration given the IBM Sickness and Accident Plan; contrast AGCO at [56]; and

iv) the suggestion that retirement is always an entirely consensual event cannot stand in the light of Waite, see AGCO at [53].

562. As to Akester, Mr Short again says that it is distinguishable from the present case. In Akester:

i) the wording of the early retirement provision was different from that in the present case and was crucial to the approach to construction in that case: see Akester at [108] to [110];

ii) the schemes as a whole are different. For example:

a) there is no misconduct anomaly in the present case: contrast Akester at [111];

b) the reliance upon the use of the phrase "leaves service" in contradistinction to "retires" in relation to cases of ill health (see Akester at [102] and [103]) is not available in relation to the Main Plan given that the person who "terminates Pensionable Service before NRD" under Schedule F Rule 1 may still be entitled to an ill health pension under Schedule D Rule 2;

c) many of the features of the Main Plan identified above were not present in Akester. In particular, it was not necessary for retirement to include cases of compulsory redundancy in order to comply with HMRC rules: it is only the structure of the DB Rules in the present case which necessitate compliance with HMRC rules to be found in the meaning of "retirement";

d) there was no consideration of the fiscal framework: see Akester at [88], [95];

e) the suggestion that retirement is an entirely consensual event, at [112] of Akester, cannot stand in the light of Waite.

563. There is one other case I should mention which is relied on by Mr Newman. It is Young v Associated Newspapers Ltd (1971) 11 KIR 413. Mr Newman cites this authority to show that the court was not unduly concerned by an anomaly similar to the one arising in the present case if the narrow construction is correct in relation to Rule 2(1) and the same construction applies to Rule 4, namely that Postponed Retirees dismissed for misconduct would not be entitled to any pension at all. The Judge accepted that there might be an anomaly (although he left open the point of construction). But even if there was, he considered that it was one which the employer could eliminate by use of its amendment power. The relevant rule in that case referred to a member "retiring with the consent of the company". The Judge considered that the meaning of that provision was plain, namely that it did not include dismissal (in that case, a compulsory redundancy) and that its plain meaning could not be eliminated by reference to the anomaly.
564. As well as relying on Young, he also relies on AGCO as another case where an anomaly did not dictate the correct answer.
565. Mr Short submits that Young is of no relevance. His first reason for saying that is that Young was dealing with consent to termination ("retiring with the consent") whereas the common ground approach is that consent relates to the taking of benefits. I have not yet accepted that common ground. If I do so in the end, then his point is true in relation to Eligible ER Members: it will be possible (in contrast with Young) to construe "retirement" as including dismissal and yet to decline an early retirement pension because there is no consent. But even so, it does not meet the point in relation to C Plan Over 60 Members. If I do not accept the common ground, then the point cannot be made at all.
566. His second reason is that the context of the 1995 TD&R in terms of regulatory provisions is very different from the context of the rules in Young in 1953. In particular, it is no longer true, as it may have been in 1953 when the relevant trust deed was executed and as the Judge put it that "it would be natural for the company, when establishing a non-contributory pension fund, to exclude a dismissed person from a deferred pension". Put that way, I agree that it is no longer a sustainable proposition in the light of the preservation requirements. Discussion
567. The first point to make is that I do not regard myself as bound by the decisions in AGCO, Akester or Young to decide that compulsory redundancy does not give rise to "retirement" within the meaning of Rule 2(1). Those cases are distinguishable for the reasons discussed above. That is not to say that some of the arguments accepted in those cases, in particular by Rix LJ in AGCO, have no resonance in the present case. The question for me is, of course, one of interpretation of this particular Rule in this particular scheme.
568. The second point to make is that the reliance placed by Mr Short on the absence of the misconduct anomaly is not as strong as it might otherwise be because that anomaly does exist in relation to C Plan Over 60 Members.
569. In my judgment, the word "retirement" in Rule 2(1) is to be construed in such a way that it does not include dismissal for misconduct. It may be that the inclusion of such dismissal in the case of C Plan Over 60 Members could be seen as an anomaly which could be lived with, but it remains, nonetheless, an anomaly and is to be eliminated if possible. It is not, in my view, an ordinary or natural use of the word retirement that it should include dismissal for misconduct. I do not say that retirement can never include such dismissal, but the context must be such as to demand that construction. Even the transitive use of the verb "retire" would not ordinarily include a dismissal for misconduct. Whereas an employer might be said to "retire" an employee when he becomes less effective at his work than the employer would wish or expect, that cannot be said of a person dismissed for misconduct when that occurs before NRD.
570. I appreciate that this gives rise to another problem: if "retirement" in Rule 4 is to be given the same meaning as it has in Rule 2(1), then a Postponed Retiree who is subsequently dismissed does not become entitled to a pension under the express provisions of the Rules. This is not so much an anomaly as a lacuna. It seems to me that the lacuna can readily be filled in one of two ways:

i) First, by the implication of a term to the effect that any sort of termination of the Service of a Postponed Retiree is to be treated as a retirement for the purposes of Rule 4. There would be no inconsistency in such an implication with any express provision of the Main Plan. It is, if I may say so, obvious that there is a lacuna if retirement has the narrow meaning; and it is obvious – applying the "it goes without saying" reality check – what the implication should be.

ii) Secondly, it can be said that there is a material difference between the dismissal for misconduct of a member who has reached NRD (and could thus retire, even in the case of an Eligible ER Member, without consent) and such dismissal where the Member has not reached NRD. One objection to the use of "retirement" in relation to the member dismissed for misconduct before reaching NRD is that it simply cannot have been intended that such a member should obtain access to an immediate, and beneficial, pension. But that objection does not apply to a Postponed Retiree. Indeed, where a person leaves Service after his NRD, it seems a perfectly proper use of words to me to describe him as having entered into "retirement" in the context of Rule 4. The fact of the matter is that Rule 2(1) and Rule 4 serve very different purposes; and although the Rules appear close to each other in the same Schedule, their context is different. I see no reason why a different meaning should not be given to the word "retirement" in the two Rules.

571. It is clear that, in some circumstances other than in relation to a Postponed Retiree, dismissal can give rise to "retirement". It is a small point, but a person dismissed even by reason of misconduct on the very day of his NRD must surely enter into "retirement"; if not, there is nowhere else that his benefits can be found under any express Rule of the Main Plan. More significantly, I regard it as clear that a person who suffers serious ill heath (amounting to "Incapacity" as mentioned in Rule 3 Schedule D) enters into "retirement" within the meaning of Rule 3 whether he resigns or whether his contract is terminated by his employer because of the ill-health. For reasons given by Mr Short and which I have already addressed, this is not a case of frustration as described by Rix LJ in AGCO. It is simply inconceivable, to my mind, that a member who is dismissed in these circumstances should not, on the basis that this was not an occasion of retirement, be entitled to a pension in accordance with the discretion conferred by Rule 3. So here, as in Rules 1 and 4, one finds a different meaning of "retirement" from the narrow view espoused by Holdings in the context of Rule 2(1).
572. I do not propose to attempt to lay down criteria to distinguish which sorts of dismissal count as "retirement" for the purposes of Rule 2(1). Holdings accepts (rightly in my view) that the dismissal which results from voluntary redundancy gives rise to "retirement". My view is that dismissal for misconduct does not give rise to "retirement". Further, the termination of a member's contract of employment when he does retire may be brought about, in contractual terms, by a notice. It is perfectly proper for such a notice to be given, but it is, nonetheless, a dismissal.
573. In my judgment, compulsory redundancy falls to be treated, for the purposes of Rule 2(1), in the same way as a voluntary redundancy. The sort of objections which can be raised to dismissal for misconduct falling within the meaning of "retirement" do not apply to compulsory redundancy any more than they apply to voluntary redundancy. In contrast with the position in relation to dismissal for misconduct, it is a perfectly proper use of the verb "retire" in a transitive sense to describe a member made compulsorily redundant as having been retired. Indeed, that might be seen as a paradigm use of the verb in a transitive sense.
574. For a summary of the answers to the Issues just discussed, see paragraph 596 below.
575. The next issue relates to the consent requirement in relation to Eligible ER Members. This is the second part of Issue 24.1.

Issues 24.1 (paragraph (ii) – part): Is voluntary redundancy "retirement… with the consent of the Principal Employer"?

576. It is clear that no express consent was given by Holdings to retirement in the case of voluntary redundancy. Its position in relation to Eligible ER Members is that consent cannot be deemed to have been given simply by virtue of the termination of pensionable service being a voluntary redundancy. In the present case, it is said that there is nothing to treat such consent as having been given so that there is no "retirement…. with the consent of the Principal Employer" and Issue 24.1 must be answered accordingly.
577. The written submissions of the parties addressed the issue of consent on the basis of the alternative approach that the focus is on retirement rather than the common ground approach where the focus is access to benefit. I propose to address, first, the arguments in the written submissions which now form Mr Newman's secondary case.
578. Mr Newman submits that the conclusion that there is no consent follows from Rix LJ's analysis in AGCO where he held that it would normally be possible for employees who apply for voluntary redundancy to "retire from Service at the request of the employer". He distinguished the position of voluntary redundancy from that of compulsory redundancy, because, although the former was formally completed by means of a dismissal, it occurred in reality as a result of a prior consent to such a dismissal (see at [65]), and it fitted better into the concept of retirement "at the request of the employer" (see at [69]): the practicalities are that, if the employer wants to have an early retirement programme, he has to take the first step and issue his invitation or make his request (see at [70]).
579. It follows from this that there is nothing inherent in the concept of "retirement" in an early retirement provision which precludes it from applying to the type of consensual dismissal involved in voluntary redundancy. Thus Holdings accepts that in the case of C Plan Over 60 Members, where there is no requirement for consent, the issue turns solely on the word "retirement" so that where those members accepted voluntary redundancy they fall within the scope of Rule 2 as it applies to them.
580. However, as regards the Eligible ER Members who accepted voluntary redundancy, the issue is whether there was "retirement … with the consent of the Principal Employer". In this situation, Mr Newman submits that the case-law supports the exclusion of those members from the scope of that phrase.
581. He relies on Minter v Julius Baer Investment Management Inc London [2004] EWHC 2472 (Ch); [2005] PLR 73 ("Minter"), in which the claimant, whose employment was terminated pursuant to a compromise agreement, claimed benefits under the early retirement provision of the employer's scheme, which required the member to have retired with the consent of the employer. Rimer J (at [84]) dismissed the claim on the basis that the compromise agreement did not entitle the claimant to any benefits under the pension scheme. The Judge also considered whether, if he did have such an entitlement, the claimant retired "with the consent of" the employer so as to qualify under the scheme's early retirement provision. The Judge (at [90] to [96]) considered that the claimant did not so retire, as the essential concept of consent in this context was that the initiative comes from the employee, and the employer in effect agrees to premature release from employment – the employer in that context cannot have consented to the termination as the employer cannot consent to his own act. The proposition that the employer does not, in these circumstances, consent to his own act is supported by Young. Thus, where the employee was dismissed from employment by reason of redundancy, even if he chose to be redundant, he did not retire "with the consent of" the employer.
582. That analysis, according to Mr Newman, applies equally in this case:

i) the Separation Programmes involved offers by Holdings to the relevant employees that their employment would be terminated on the ground of redundancy;

ii) the initiative for the termination therefore came from Holdings, not the employees;

iii) irrespective of the consensual nature of the redundancy, the employees' termination cannot have occurred "with the consent of" Holdings, as Holdings could not consent to its own act;

583. I do not accept those submissions. As Holdings accepts, a voluntary redundancy constitutes "retirement" notwithstanding that the process is initiated by the employer and notwithstanding that it takes effect, formally, by way of dismissal. According to Rix LJ, it qualifies because it is not in reality a dismissal but rather occurs as a result of prior consent to such dismissal (see at [65] of AGCO). The voluntary redundancy process commences, I accept, only because the employer has initiated it. But it is implicit, in my judgment, in the request from the employer for volunteers that any "retirement" which takes place (albeit by way of dismissal) will be one to which the employer has given its "consent" for the purposes of Rule 2(1). Indeed, in a voluntary redundancy programme, an employee may or may not have his application for redundancy accepted. If he is accepted for redundancy, the employer will clearly be consenting to his departure, a departure which is effected formally by a notice of dismissal. That notice does not vitiate or qualify the consent which is necessary for the process to continue. As Mr Short says, this is not a case of Holdings consenting to its own act (which is an argument which can be raised in relation to dismissal on compulsory redundancy); rather it is a case of Holdings consenting to the voluntary redundancy. Accordingly, even on the Minter analysis, the relevant initiative for the actual redundancy rather than for the commencement of the process comes from the member and it is Holdings which consents. The ensuing departure of the member is a retirement with consent. Access to the benefit then follows automatically because that is what Rule 2(1) provides. Holdings' policy (whether the Old ER Policy or the New ER Policy) is entirely irrelevant: the policy relates to when consent will be given whereas, on my analysis, consent is automatically given in a case of voluntary redundancy. It is not open to Holdings to say at one and the same time both (i) that consent is given to retirement but (ii) the member shall not be entitled to an early retirement pension.
584. That is, as I see it, the analysis, if the common ground approach is not correct. But if the common ground approach is correct, the issue of consent now needs to be addressed in that context. In that case Minter has no scope for application. Rimer J's reasoning, was that the employer could not "consent" to its own act in terminating the contract of employment. In the present case, in contrast on the basis of the common ground approach, Holdings does not need to provide "consent" to the termination of the contract of employment; there only needs to be consent to the access to benefits. Mr Short accepts that Holdings did not give express consent. Instead, he submits that it is the affected members that are to be treated as having consent in the light of the unlawful withdrawal of the Old ER Policy. When considering the position of whether consent was required to retirement rather than access to benefits, I said that it was necessarily implicit in the request from the employer for volunteers, that any "retirement" which takes place (albeit by way of dismissal), will be one to which the employer has given its "consent" for the purposes of Rule 2(1). On the common ground approach, it is not, however, necessarily implicit in the request from the employer for volunteers, that consent will be given to payment of an immediate pension with a favourable ERDF.
585. It is for the Trustee to demonstrate in the present case in relation to the voluntary elements of the Separation Programmes in question, circumstances which show that consent (ie under the common ground approach) has been given or must be treated as given. As Mr Short accepts, it has clearly not been given expressly. Instead, he submits that consent is to be treated as having been given as a result of Holdings' breaches of its contractual and Imperial duties.
586. I have discussed a very similar issue under Issue 21 above. In that context, I have decided that the Trustee is able to provide the benefit to which the member would have been entitled if Holdings had given its consent in accordance with the Old ER Policy. To repeat what I said, in a case where Holdings would be in breach of duty in refusing its consent (so that it would be acting irrationally and perversely), it would be appropriate to direct the Trustee to administer the Plans as if consent had been given. The same applies to voluntary redundancy once it is established, as I have decided, that voluntary redundancy is "retirement". This is so, in my judgment, even if the voluntary redundancy takes place in the context of a Separation Programme which is not itself part of Project Waltz (although the fact that it is not part of Project Waltz may be important when considering Issue 28, waivers). The question then is whether Holdings was in breach of duty in failing to provide its consent to an early retirement pension.
587. In accordance with the answer to Issue 21, a retirement before March 2014 is to be afforded the same treatment, so far as concerns consent, as if the Old ER Policy had continued in force. In the context of the common ground approach that consent is concerned with access to benefits, the question is then whether Holdings would have been bound by the Old ER Policy to give its consent to access to early retirement benefits.
588. I have no doubt that the Reasonable Expectation concerning the Old ER Policy can be invoked in relation to a voluntary redundancy as much as it can to a resignation. The material on which I decided, in the Judgment [680] to [687] that there was a Reasonable Expectation at all, largely related to voluntary separation programmes rather than individual requests to retire. And Mr Stallworthy and Mr Spink have referred me to a great deal of material which supports that conclusion. Accordingly, consent is to be treated as having been given.

Issues 24.1 (remainder): Is compulsory redundancy "retirement… with the consent of the Principal Employer"?

589. This Issue is to be answered on the assumption that there is a retirement; the question is whether the consent which is required by Rule 2(1) has been given or is to be treated as having been given.
590. On the basis of the common ground approach, the question, as in the case of voluntary redundancy, is whether the Reasonable Expectation concerning consent included an expectation that the Old ER Policy would be applied in cases of compulsory redundancy. This matter was not specifically addressed at the hearing. It is accepted by the Trustee that dismissal for misconduct is not "retirement with consent" and it is not suggested that there was a Reasonable Expectation that the Old ER Policy would be applied in the case of such dismissal.
591. The difficulty for the Trustee is that there were in fact no compulsory redundancy programmes prior to Project Waltz, let alone prior to the matters giving rise to the Reasonable Expectations. As Mr Stallworthy explains, there was no clear evidence before me at any of the hearings that Holdings ever had any compulsory separation programmes at all prior to Project Waltz. On the contrary, the witness evidence discussed the extent to which Holdings either redeployed staff to new roles or used incentives (including pension augmentations in conjunction with the Old ER Policy) to manage employees out of its employment.
592. There are, so far as I am aware, no statements made on behalf of Holdings which would have led to a Reasonable Expectation that the Old ER Policy would be applied in cases of compulsory redundancy. It is said against Holdings that there was no evidence from it that the Old ER Policy would not be applied in cases of compulsory redundancy. But it is not for Holdings to prove that negative, especially in the absence of any evidential support for the contrary proposition.
593. Although the evidence which I had contains comments which could be read as including compulsory redundancy, the comments are almost all based on the facts of actual situations other than compulsory redundancy where the Old ER Policy was applied. A number of witnesses expressed their concerns about the New ER Policy because they saw the Old ER Policy as a safety net if they were made redundant. But of course it was there as a safety net because a member would have been able to retire before being made compulsorily redundant and thus obtain an immediate pension. Whether a member could have formed a Reasonable Expectation that the Old ER Policy would apply if, contrary to the common practice, he was made compulsorily redundant, I doubt. And whether members could have formed a Reasonable Expectation that the Old ER Policy would apply in the event of an unprecedented compulsory separation programme I doubt even more. I do not decide either point. But what I do say is that I am not presently satisfied that the Old ER Policy would have applied in these situations. And if it would not have applied, then the compulsory phases of the Separation Programmes would not have given rise to retirements "with consent".
594. I turn then to the position on the basis that the common ground approach is incorrect, and that the focus of consent is on "retirement". Holdings' arguments under this scenario in relation to consent in the context of voluntary redundancy apply equally to compulsory redundancy. But the Trustee's principal answers to those submissions and the reasons which I have given above are not available. Holdings also argues that it is conceptually impossible for consent to retirement to be given in the context of compulsory redundancy.
595. The Trustee breaks the issue down into two components:

i) Whether the nature of compulsory redundancy is such that Holdings could never be said to have given consent for the purposes of Rule 2(1) in cases of compulsory redundancy. The Trustee says that there is no difficulty with the concept of Holdings giving consent to the retirement of a member made compulsorily redundant.

ii) If that is correct, whether Holdings is to be treated as giving consent by reason of its breach of the Imperial duty. As to that, it is not suggested, as already noted, that the act of dismissing a member by way of compulsory redundancy in itself comprises the giving of consent for the purposes of Rule 2(1): it is not inherent in compulsory redundancy (in contrast with voluntary redundancy as explained above) that consent is required. The source of the consent has to be found elsewhere if the Trustee is to succeed. The source can only be found, if it is to be found at all, in Holdings' withdrawing of the Old ER Policy in breach of its Imperial duty, so that it is to be treated as having given consent.

596. As to the conceptual possibility of consent being given to retirement in the context of compulsory redundancy (or indeed dismissal), I agree with the Trustee that this possibility does exist. I reach that conclusion for the following reasons. The purpose behind the requirement for consent under Rule 2(1) is to give Holdings a measure of control over the benefits which a retiring member is to receive: Holdings is concerned not so much with the fact of the retirement but with the consequences of the retirement. So long as Holdings has control over whether the (expensive) benefits of Rule 2(1) are available, the purpose of the consent requirement is acknowledged. In those circumstances, a purposive construction leads to the conclusion that Holdings is able to give a valid consent for the purposes of Rule 2(1) even in cases where the retirement occurs as the result of a compulsory redundancy (as it can, in accordance with my earlier decision). The important point here is that the member only becomes entitled to the benefits of Rule 2(1) if Holdings agrees that he should do so by giving its consent under Rule 2(1)). It would be open to Holdings, therefore, to make a group of members compulsorily redundant and (i) to agree expressly that such retirement (the hypothesis here is that there is "retirement") shall entitle the members concerned to the benefits provided under Rule 2 and (ii) to state expressly that consent was being given to such retirement. This would be effective, in my judgment, to entitle the members to benefits under Rule 2(1).
597. For completeness, I add that I see no reason why consent should not be given retrospectively. Suppose, for instance, that an Eligible ER Member leaves service aged 61 without consent. He is perfectly properly described as entering "retirement" but is not entitled to an immediate pension on the beneficial terms because Holdings has not consented. As a matter of ordinary language, Rule 2(1) can be read as including retrospective consent, a conclusion which is entirely consistent with my approach to consent in the context of dismissal by reason of compulsory redundancy.
598. As to the argument based on breach of the Imperial duty, the analysis is the same as that given in the context of the common ground approach at paragraphs 588ff above. I am unable to conclude on the evidence before me that the Reasonable Expectation concerning ER policy included an expectation that the Old ER Policy would apply in relation to compulsory redundancy.
599. The answer to Issues 23.1 and 24.1 are these:

i) voluntary redundancy is an occasion of "retirement";

ii) compulsory redundancy is an occasion of "retirement";

iii) voluntary redundancy is an occasion of retirement "with the consent of" Holdings whether "consent" relates to retirement or to access to benefits;

iv) compulsory redundancy under the compulsory elements of the Separation Programmes is an occasion of retirement "with the consent of" Holdings if, but only if, the Reasonable Expectation concerning the Old ER Policy applied to compulsory redundancies. I am unable to conclude on the evidence currently before me that the Old ER Policy did so apply.

600. That last conclusion is not an invitation to the Trustee and the RBs to produce further evidence. In these proceedings, at least, I must achieve some finality on the basis of the evidence and arguments already received. I say nothing about what might take place in other proceedings.
601. The answers which I have given make it unnecessary to decide for the purposes of the Issues which I have so far discussed whether the consent requirement relates to the retirement itself or the access to benefits. This aspect is considered in relation to Issue 28.

Issues 23.2 and 24.2: If either or both of the said methods did not constitute retirement (with the consent of the Principal Employer in the case of issue 24.2) in accordance with those terms, whether Rule 2(1) would not as a result apply in respect of such members?

602. In the light of my conclusion in relation to the meaning of "retirement" all that is left of these Issues is whether, in the case of Eligible ER Members, Rule 2(1) is to apply. For the reasons given under Issues 23.1 and 24.1, I am unable to conclude, on the evidence, that Rule 2(1) applies in the case of compulsory redundancy under the Separation Programmes.
603. I do not propose to address what the position would be if I am wrong in my conclusions about the meaning of "retirement".

Issues 23.3 and 24.3: If either method did not constitute retirement (or retirement with the consent of the Principal Employer in the case of Issue 24.3) in accordance with Rule 2(1) and that Rule did not as a result apply in respect of such members, whether the members whose contracts were terminated are entitled to any other remedy and if so what remedy?

604. The Trustee contends that damages or equitable compensation will be payable where members can establish, on an individual basis, that they would have retired before being made redundant and that, had they done so, the Old ER Policy would have applied to them.
605. I have already addressed, under Issue 16, the issue whether damages or equitable compensation can, in principle, be claimed for breach of the Imperial duty and have concluded that this is possible. In any case, Project Waltz gave rise to a breach of the contractual duty as well as a breach of the Imperial duty.
606. I agree with the Trustee's submission. If Holdings had not acted in breach of duty, the Old ER Policy would have remained in force. Thus, even assuming that the Separation Programmes did not, of themselves, involve any breach of any duty on the part of Holdings, the fact is that the Old ER Policy was not in force at the time of those Programmes when it should have been. If a member can demonstrate, on the facts, that he would in practice have been able to retire and take advantage of the Old ER Terms before he was made compulsorily redundant and would in fact have done so, then in principle he is entitled to compensation in respect of what he has lost by not having done so.

Issue 25 and 26:

607. The text of these Issues in the List of Issues is very long. All of the questions raised in Issue 25 relate to the period from 6 April 2010 during which Holdings should have applied the Old ER Policy and where:

i) an Eligible ER Member's employment has terminated;

ii) that termination is capable of being a "retirement with consent" and

iii) that Eligible ER Member did not receive a pension calculated under Rule 2(1).

608. The members covered by the list above include members taking voluntary redundancy. It also includes members taking compulsory redundancy if, but only if, the Reasonable Expectation concerning the Old ER Policy applied to compulsory redundancies. Since, as I have held, I am unable to conclude on the evidence currently before me that the Old ER Policy did so apply, they are not persons who are to be treated as actually entering into "retirement with consent" albeit that the termination of their employment was capable of being retirement with consent (in accordance with my analysis above). Accordingly, I exclude those members from the class to which Issues 25 and 26 are directed.
609. The Trustee's written opening states that the answers to Issues 25.1(a) and (b), Issue 25.2(a) and (b), Issues 26.1(a) and (b) and Issue 26.2(a) and (b) are common ground and that they are all to be answered in the affirmative. That common ground does not take account of the points made in the preceding paragraph. But subject to that, the common ground should stand and is, I think, correct. This is subject to the answer to Issue 28 (waivers). I answer the Issues accordingly.

Issue 27: Are any C Plan Over 60 Members or Eligible ER Members in principle entitled to damages or equitable compensation from IBM as a result of the application of the New ER Policy, or to some other remedy and if so what remedy?

In principle what is the measure of such damages or compensation?

610. The same legal analysis applies here as apply in relation to the Exclusion Notices discussed under Issue 16. In my judgment, a compensatory award of damages or equitable compensation is available in principle. I do not propose to say anything to add to what I said under Issues 16 and 17 about the measure of damage or equitable compensation.

Issue 28: the effect of the waivers

611. Members who left under the three Separation Programmes (Phoenix (2012), Mercury (2013) and Apollo (2014)) were required to enter into compromise agreements as a condition of receiving either voluntary retirement or enhanced compulsory retirement terms. Although there was a general carve-out of accrued pension rights from the settlement, the agreement did settle claims in respect of those rights referred to in the Pensions Waiver at Schedule 4 of the various agreements. IBM says that the members are barred from any remedy under Issues 23 to 27 above.
612. The waivers in question were signed under the Separation Programmes; they formed part of the compromise agreements pursuant to which the employment of the members was terminated by reason of redundancy. They all follow the same substantive form. They contain an explanation of the issue or issues relevant to each waiver (as to, for example, whether redundancy amounted to a retirement with consent including reference to AGCO) followed by a section written in the first person in the "voice" of the member, in which he or she agreed to waive particular rights. The waivers were signed by each member and the legal adviser to that member.
613. There are differences of detail between Phoenix waivers (which were given prior to the judgment in the Rectification Action) on the one hand and the Mercury and Apollo waivers (which were given after that judgment) on the other; and between voluntary and compulsory (referred to as "involuntary" in the documents themselves) redundancy.
614. I set out the substantive parts of the waivers below. But before I do that, I should summarise the salient parts of the introductory pages. As Mr Newman puts it in his written opening submissions, the voluntary redundancy waivers:

i) made it clear that, by agreeing to dismiss members on voluntary redundancy terms, Holdings was not giving its consent to early retirement from active service;

ii) referred to the AGCO and Minter cases and, in the case of the Phoenix waivers, the C Plan trial;

iii) recognised that, despite the Minter case which supported IBM's position that voluntary retirement did not amount to retirement with IBM's consent, the AGCO case may be said to give rise to a contrary argument;

iv) stated that IBM was not prepared to grant voluntary redundancy without a waiver of any claims to early retirement terms on that basis, and that the only basis it was prepared to do so was if the member agreed to the waiver by way of a bona fide compromise of a disputed or doubtful entitlement under the Plan.

615. And the compulsory redundancy waivers:

i) made it clear that dismissal on involuntary redundancy terms was not retirement from service;

ii) recognised that it may be arguable that, because the member has signed a compromise agreement in exchange for an improved financial settlement, the termination of employment was consensual and involved an element of choice, thereby introducing an element of doubt as to whether the redundancy was involuntary;

iii) stated that IBM was not prepared to offer the terms set out in the compromise agreement without a waiver of claims.

iv) further, even if there was a finding of voluntary redundancy, it repeated the argument arising from the AGCO and Minter cases as regards whether this could amount to the consent of IBM, and it stated that IBM was not prepared to grant the terms set out in the compromise agreement to a member without a waiver of claims in this regard, so that the only basis on which it was prepared to do so was if the member agreed to the waiver by way of a bona fide compromise of a disputed or doubtful entitlement under the Plan.

616. I now set out the substantive parts of the waivers; the passages giving rise to Issue 28 are highlighted. These passages are contained within, and at the end, of, the schedule which is referred to in the opening words.

Voluntary and Involuntary (Phoenix)

I have read the contents of this schedule and have had the benefit of legal advice thereon. I understand that in signing this waiver I am giving up claims which may be available to me and which may be valuable. By signing this document I agree to waive all and any claim that I may have, whether under the IBM Pension Plan, the limited exception to the July 1983 Members and Trial Members, the C Plan Trial or otherwise, to the effect that by granting me voluntary redundancy terms the Company is thereby consenting to my early retirement from active service, or that I am otherwise entitled to an immediate early retirement pension calculated under the beneficial terms of the IBM Pension Plan. I accept that I am being dismissed for redundancy and that I will, following redundancy, become a deferred member of the Plan.

Involuntary (Apollo)

I have read the contents of this schedule and have had the benefit of legal advice thereon. I understand that in signing this waiver I am giving up claims which may be available to me and which may be valuable. By signing this document I agree to waive all and any claim that I may have, whether under the IBM Pension Plan or otherwise, to the effect that by dismissing me on redundancy terms, whether on involuntary terms or otherwise, I am thereby retiring from active service or that the Company is thereby consenting to my early retirement from active service, or that I am otherwise entitled to an immediate pension calculated under the 'active member terms of the IBM Pension Plan. I accept that I am being dismissed for redundancy and that I will, following redundancy, become a deferred member of the IBM Pension Plan.

Voluntary (Mercury) (Apollo was in similar terms)

I have read the contents of this schedule and have had the benefit of legal advice thereon. I understand that in signing this waiver I am giving up claims which may be available to me and which may be valuable. By signing this document I agree to waive all and any claim that I may have, whether under the IBM Pension Plan or otherwise, to the effect that by granting me voluntary redundancy terms the Company is thereby consenting to my early retirement from active service, or that I am otherwise entitled to an immediate early retirement pension calculated under the beneficial terms of the IBM Pension Plan. I accept that I am being dismissed for redundancy and that I will, following redundancy, become a deferred member of the IBM Pension Plan.

617. There is nothing in the waivers or the compromise agreements expressly referring – or necessarily impliedly referring – to rights which members might have as a result of claims challenging Project Waltz or arising out of a breach of Holdings' Imperial duty or its contractual obligations of trust and confidence. The opening words of the substantive part of each of the waivers which I have set out above (and which I take as typical) expressly refers to the contents of the schedule and to having had the benefit of legal advice thereon (my emphasis). These opening words direct the focus of the waiver on the matters referred to in the schedule and clearly the legal advice referred to was directed at the issues disclosed in the schedule (including, I suppose, the effect of the waiver itself).

Holdings' submissions

618. Holdings' case as set out in its written opening is something I will come to in a moment. In his oral submission, Mr Newman presented a rather different argument. Putting the point in my own words, a member is given a choice under Rule 1 Schedule F of how to take his benefits. Even a member who is entitled to an early retirement benefit under Rule 2(1) does not have to take that benefit: he can choose to take a deferred pension, exercising his choice under Rule 1(1) Schedule F. Accordingly, the waivers can be seen as elections by the affected members to become deferred members (or Deferred Retirees) rather than to take immediate pensions (as Retirees). In making that submission, he relies on a number of communications with members which show, he says, what the members thought they were agreeing to. It is true that the documents show that Holdings regularly explained that, whether the redundancy was voluntary or compulsory, it was not consenting to early retirement: rather, the position was that there was a dismissal by way of redundancy which had the consequence that the member would become a deferred member. However I have not found any hint of election in these communications in the sense of the exercise of a choice under the Rules.
619. Nor is there any hint of election in the waivers themselves. The underlying basis of the waivers is that the member accepts (on Holdings' case) that he is not retiring with consent: in other words, he accepts that he does not fall within Rule 2(1) at all. He is not, therefore, within Rule 1(1)(b) Schedule F (entitlement to an immediate early retirement pension) in the first place.
620. I do not find this way of putting Holdings' case at all persuasive or attractive. I consider that Mr Newman is on firmer ground in his written opening submissions. The position taken there is that the intention behind the pension waivers was to render academic any arguments advanced on behalf of the affected members that the circumstances of their redundancies under the Separation Programmes entitled them to benefits under Rule 2(1).
621. Thus, if and to the extent that, at the time of the waivers, there was any dispute as to:

i) whether the authorities relied on by Holdings (AGCO and Minter) operated to take members outside Rule 2(1); or

ii) whether C Plan Over 60 Members were entitled to retire without consent,

the pension waivers are intended to bind the members who signed them as a matter of contract to the position for which Holdings was contending.

622. So far as the matters referred to in the immediately preceding paragraph is concerned, I agree with Mr Newman that this is clear from the explanation of the reasons for the pension waivers contained in the waiver documents themselves. Those matters are, however, concerned with the meaning of Rule 2(1). And when Holdings stated, as it did, that it was not prepared to grant voluntary redundancy without a release of any claims to early retirement terms on the basis of the waiver (and by way of a bona fide compromise of a disputed or doubtful entitlement under the Plan), it was referring principally to those matters.
623. Holdings submits that, applying ordinary contractual principles, the pension waivers constituted binding contracts by which the employees agreed not to claim benefits based on Rule 2(1). I do not disagree that the waivers constitute binding contracts: indeed the Trustee accepts that the waivers are effective but submits that, as a matter of construction, they are of more limited scope than that for which Holdings contends.
624. Mr Newman's submissions about the extent of the waivers are, in summary, these:

i) insofar as the waivers apply to compulsory redundancies, the member waives any claim that that redundancy constituted retirement or (if Holdings failed in the Rectification Action) that consent to early retirement was not required;

ii) the member waives any claim that the redundancy constituted Holdings' consent to early retirement from active service;

iii) the member waives any claim to immediate early retirement pension calculated on the terms of Rule 2(1);

iv) the member accepts that they were being dismissed for redundancy and that, following redundancy, they became deferred members of their Plan.

625. On the basis of those contractually binding terms, the waivers are effective, on Holdings' case, to preclude reliance on Rule 2(1) by the members concerned. They are thus unable to rely on Rule 2(1) directly to obtain benefits from the Trustee; and they are unable to rely on it to found a claim for damages or equitable compensation for breach of duty.

The Trustee's position

626. The Trustee says that the effect of any waiver is limited to the specific issues that were raised at that time (principally whether redundancy comprised retirement with consent under the Rules). The waivers do not also give up any other rights arising from the withdrawal of the Old ER Policy; this includes claims for damages even if the waivers are effective to waive any claim to actual entitlement to a benefit under Rule 2(1). If the waivers did have such a wide impact, the Trustee says that there is a claim for misrepresentation arising out of statements made at the time as to the effect of the waivers.
627. In more detail, the Trustee accepts that the waivers were effective to waive any claims that:

i) in cases of the voluntary Mercury/Apollo waivers, Holdings in fact consented to early retirement for the purposes of Rule 2(1) by granting voluntary redundancy (the Redundancy Equals Consent Assertion);

ii) in the case of the compulsory Mercury/Apollo waivers, that the Member retired from active service as a result of the dismissal (the Retirement Equals Dismissal Assertion) and/or that Holdings in fact consented to early retirement for the purposes of Rule 2(1) by dismissing the Member (the Redundancy Equals Consent Assertion);

iii) in the case of the Phoenix waivers, "that [Holdings] in fact consented to early retirement for the purposes of Rule 2(1) by dismissing the Member (the Redundancy Equals Consent Assertion) and/or arising from C Plan Trial". [I quote this passage from the written opening submission as corrected in oral submissions. I refer to it later.]

628. The Trustee contends that the waivers did not extend any further. They did not waive:

i) the rights of the C Plan Over 60 Members who were granted voluntary redundancy and whose entitlement is based upon the fact that they did not require consent rather than upon the argument that Holdings gave consent (save in respect of the Phoenix waivers);

ii) the rights of other Members who were granted voluntary redundancy and whose entitlement is based upon the argument that Holdings is to be treated as having given consent because of its unlawful withdrawal of the Old ER Policy rather than upon the argument that Holdings gave consent in agreeing to voluntary redundancy;

iii) the rights of Members to damages for breach of contract or the Imperial duty on the basis that they should have but did not become entitled to an immediate pension on advantageous terms because of the wrongful withdrawal of the Old ER Terms; and

iv) any other rights of Members based upon the wrongful withdrawal of the Old ER Terms.

629. Paragraph i) of the preceding paragraph, read with paragraph iii) of the paragraph before have caused me some confusion. These paragraphs were not enlarged on at the oral hearing and I had not picked up the point which now leads to confusion. It is that paragraph iii) appears to be saying that the waiver relates only to consent, in which case it would not apply where consent is not required (which, following the Rectification Action, is in fact the case for C Plan Over 60 Members). Paragraph i), however, says that a claim by a person who does not need consent (ie a C Plan Over 60 Member) is not waived "save in respect of the Phoenix waivers": in other words, it appears to be accepted that a C Plan Over 60 Member cannot take the point that consent was not required under the Phoenix Programme. I take the combination of the two paragraphs to be acceptance by the Trustee in relation to the Phoenix Programme that, in the case of a C Plan Over 60 Member, the waiver is effective to preclude him from arguing that consent is not required; but not as acceptance that he is unable to take the same points, such as they are, as are open to Eligible ER Members.

Discussion

630. In the light of my answer to Issue 24.1, members made compulsorily redundant did not enter into "retirement with consent" of Holdings. Accordingly, I address Issue 28 only in relation to voluntary redundancies. If I am wrong in my conclusion about compulsory redundancies, the position in relation to them is similar to that discussed below in relation to voluntary redundancies adopting the common ground approach to the focus of the consent requirement. However, there are differences in wording between the waivers for voluntary and compulsory redundancy which would be of relevance, but which I do not propose to address.
631. In addressing the question of construction, I make the point in relation to all three Separation Programmes that the Project Waltz proceedings were on foot before any of the Separation Programmes had been implemented or even planned. This is not a case where the waiver is relied on by Holdings as a waiver of claims which the person providing the waiver had no idea could be made where different considerations might apply. This is of relevance to the general, second, limb of each of the waivers which purports to waive all claims "that I am otherwise entitled to an immediate early retirement pension". The court is reluctant to extend waivers and compromises to matters about which the parties (or at least the party providing a waiver) could not possibly have had in mind. But that is not the position in the present case where the complaints about Project Waltz had been thoroughly aired and where it was known that a Reasonable Expectation about the duration of the Old ER Policy was in the frame.
632. Reliance has been placed on some FAQs produced in the course of the Mercury and Apollo Separation Programmes. They are not relevant to the construction of the Phoenix waivers. I deal with the Phoenix waivers first.

Phoenix

633. The waiver is effectively in two parts, the first limb relating to consent to retiring early (in the language of Rule 2(1), "retirement with consent") and the second limb relating to claims that the member is "otherwise entitled to an immediate early retirement pension…". I deal with the consent limb first.
634. It is clear that the first part of the waiver is aimed at uncertainties about the meaning of Rule 2(1). Holdings maintained that redundancy was not "retirement with consent" and was concerned to meet the argument that it might be. For Eligible ER Members who accepted the severance terms, Holdings required the Members to agree that the granting of voluntary redundancy was not to be taken as the giving of consent to retirement (or again in the language of Rule 2(1) that there was not a "retirement with consent"). Now, that was either correct (in which case the waiver was not necessary) or it was doubtful (in which case the waiver represented a compromise). If it had been clear to all concerned that voluntary redundancy did fall within Rule 2(1) then the language of the waiver would surely have been rather different. The members would not be waiving claims; they would be agreeing to a variation of their benefits (along the lines of a South West Trains agreement). But this was not the case.
635. And so, in relation to Eligible ER Members, the Phoenix waiver would clearly be effective to waive any claims to benefits under Rule 2(1) were it not for the argument which Mr Short raises in relation to Holdings' breach of duty in relation to the Old ER Policy.
636. The basis on which such Members can, apart from the waivers, obtain the benefit of Rule 2(1) has been discussed under Issues 21 and 22.
637. Under what I have called the alternative approach (that the focus of consent is on leaving service), the position is this. Voluntary redundancy is "retirement" and by offering voluntary redundancy, and then granting it to the applicant, Holdings is consenting to "retirement". The benefit then follows automatically by the application of Rule 2(1) and the Old ER Policy is of no relevance. If that were the only basis on which the Member would be entitled to benefit under Rule 2(1), it seems to me that the Phoenix waiver would take effect in relation to that entitlement. The entitlement would arise by the straightforward application of the Rules, but the entitlement arising from that consent has been waived.
638. But under the common ground approach the position is different. In that case, the mere fact of voluntary redundancy does not give rise to any entitlement under Rule 2(1). Instead, the member has to rely on Holdings' breach of duty in abrogating the Old ER Policy in order to demonstrate a deemed consent. This is not a straightforward application of the Rules since the result (ie that the member is entitled to an early retirement pension) can only be reached by treating Holdings as having given consent when it has not in fact done so.
639. Mr Short submits that this does not fall within the Phoenix waiver: he draws a distinction between an entitlement "based upon the argument that [Holdings] is to be treated as having given consent because of its unlawful withdrawing of the Old ER Policy" and "the argument that [Holdings] gave consent in agreeing voluntary redundancy". That, in my view, is a false dichotomy. The reason why, in either case, the Member is entitled to a benefit is because the relevant consent is established (on this hypothesis, consent to access benefits). The waiver relates to a claim under "the IBM Pension Plan…. or otherwise" to the effect that granting voluntary redundancy amounts to consent to early retirement. But this, it seems to me, precisely covers the consent which would, apart from the waivers, be treated as arising on voluntary redundancy as a result of the arguments based on the Old ER Policy. This, in my view, is so even if one ignores the words "or otherwise" where they appear in the first limb of the waiver. Taking those words into account, the case is only made stronger.
640. If that conclusion on the common ground approach is wrong, then Holdings relies on the second part of the waiver. On a literal reading of the words "or that I am otherwise entitled to an immediate early retirement pension…." are wide enough to cover any claims to a benefit under Rule 2(1). That does not, of course, result in the member having no benefit since the last sentence of the waiver quoted at paragraph 615 above expressly recognises that the member will become a deferred member (ie a Deferred Retiree to use the definition in the Plan).
641. Mr Short, correctly in my view, submits that the introductory pages (the essence of which is set out by Mr Newman as recorded at paragraph 613 above) form part of the waivers and are to be taken into account when construing them. The "or otherwise entitled" limb of the waiver is, he says, referable to the matters identified in the earlier parts of the waiver, that is to say, in essence, the points of construction of Rule 2(1).
642. I do not agree with that argument. It seems to me to be clear from the background and from the waivers themselves that Holdings was intending that those made voluntarily redundant should become entitled to deferred benefits only. The first limb focuses on the issue of construction of Rule 2(1), in particular the issue of consent. The second limb must have been intended to add something. On Mr Short's approach, it adds little, if anything. I see no reason to restrict its meaning by excluding claims to an entitlement which relies on Holdings' breach of duty in abrogating the Old ER Policy. I do, however, consider that the "or otherwise entitled" limb applies only to a claim arising out of the voluntary redundancy. In other words, the provision is to be read as "or that I am otherwise entitled thereby to an immediate early retirement pension…" where "thereby" refers to "by granting me voluntary redundancy terms".
643. In any case, the argument based on the Old ER Policy does not result in any claim other than one to a benefit under Rule 2(1): the early pension arises, and can only arise, under that Rule. But that is precisely what the waiver is directed at: by waiving the claim that he has retired with consent, the member accepts that he does not qualify for the benefit. The fact that the consent which triggers Rule 2(1) arises, under the common ground approach, as a result of the argument about the Old ER Policy rather than, under the alternative approach, as an inherent part of voluntary redundancy does not take the case out of the scope of the waiver.
644. In my view, therefore, the "or otherwise entitled" limb of the Phoenix waivers covers the claim to entitlement arising as a result of Holdings' breach of duty in abrogating the Old ER Policy.
645. In the result, it does not therefore matter which of the common ground approach is correct and alternative approach is correct. On both approaches, the result is that the Phoenix waivers are effective according to their terms to waive the entitlement to a benefit under Rule 2(1). It is not suggested by the Trustee – and is not open to them to do so, I think, on the materials placed before me – that the waivers themselves gave rise to any further breach of duty on the part of Holdings.

Mercury and Apollo

646. Turning to the Mercury and Apollo Programmes, the waivers, as has been seen, are in similar form to the Phoenix waivers, although there are a few differences, in particular reflecting the fact the result of the Rectification Action was known by the time of those waivers. In relation to the Mercury and Apollo Programmes, the Trustee relies, as part of its case on construction, on the FAQs produced during the course of those Programmes. These form part of the background against which members made their decision to sign the waivers. Mr Short says that they are relevant to construction of the Mercury and Apollo waivers, a proposition with which I agree. In each case, including both voluntary and compulsory redundancy, one finds this:

"What is the impact of the Good Faith/Project Waltz proceedings

As you may be aware, the High Court litigation to determine whether the pension changes made in 2009 to 2011 as a result of Project Waltz was heard in February and March 2013: the parties are still waiting for judgment. IBM confirms that, if the court finds those changes (or any of them) to have been invalid then you will not, by entering into a settlement agreement, have surrendered any rights you may have otherwise had."

647. The Trustee's arguments, in addition to those which I have addressed in relation to Phoenix, apply mutatis mutandis to the Mercury and Apollo Programmes. In addition, it is said that the wording of the consent limb of the waiver does not apply to C Plan Over 60 Members since, for them, consent was not needed and that was known at the time of the waivers, my judgment in the Rectification Action having been handed down. I will leave the position of those Members aside for the moment, and return to them at the conclusion of this section of my judgment.
648. Under the alternative approach to consent (ie the focus is on leaving service), I consider that the first limb of the waivers is effective in relation to the claims of Eligible ER Members. The reasons for these conclusions are the same as those which I have given in relation to Phoenix. I do not consider the FAQ statement points to a different result. The relevant invalidity to which the FAQ statement, we now know, applies, is the change of ER policy. The invalidity of that change, however, has no impact at all on the members' entitlement under the alternative approach. As I have explained, the Old ER Policy (or any other ER policy) is irrelevant. On the alternative approach, voluntary redundancy would give rise to "retirement with consent" and the member would be entitled to an early retirement pension. That right has not been affected at all by the invalidity of the changes to the Old ER Policy. The right to the early retirement pension was, it is true, eliminated on the voluntary redundancy, but this was all to do with the settlement agreement and nothing to do with the invalid change of policy. The FAQ statement is saying no more than that the members would not lose any rights which they would have had if the change of policy had not taken place.
649. However, if the common ground approach is correct, matters are, as with Phoenix, different. The same arguments apply as I have discussed in relation to Phoenix with the difference that the FAQ statement has more relevance. Taking that into account, Mr Short submits that it suggests strongly that only the matters expressly raised in the introductory section of the waivers are being compromised. Any claims based on the unlawful withdrawing of the Old ER Policy are not and the waivers should be construed accordingly.
650. It is quite difficult to see how the FAQ statement would operate even if it had contractual force; it is equally difficult to see what impact the statement would have as a non-contractual statement. It provides that a member will not, by entering into the settlement agreement, surrender any rights which he would otherwise have. On a literal reading, that means rights he would have if he had not entered into the settlement agreement. The settlement agreement is the separation agreement which itself contains the waivers found in a schedule. That makes perfectly good sense where those rights can be identified outside the settlement agreement itself: the member can enforce those rights. But where those rights would not exist apart from the settlement agreement, the position is rather more complex. If the member had not entered into the settlement agreement, he would not have been granted voluntary redundancy and there would have been no right which he would otherwise have had.
651. What the FAQ statement does not provide is that the members' rights in all respects should be treated as if the invalid changes had not been made. But even if it had done, that would not get the member far. Voluntary redundancy would then be an occasion when he would, absent any contrary agreement, be entitled to an early retirement pension. But that would be subject to a valid agreement to the contrary. Holdings would say the separation agreement was precisely such an agreement. Of itself, the separation agreements did not give rise to a breach of duty, so that it was open to Holdings to obtain such agreements with Eligible ER Members just as much as it was able to do so with a member where consent was not required at all.
652. What the Trustee cannot do, I think, is both to treat the separation agreements as having been made and then to excise the waivers on the basis that they amount to a surrender of rights which would otherwise have existed. It begs the question about what those rights would have been: there is no reason to think that the affected members would have been offered voluntary redundancy on terms which allowed for early retirement under Rule 2(1).
653. I return to the position of C Plan Over 60 Members. I accept Mr Short's submission in relation to the first limb of the Mercury and Apollo waivers that the waiver only goes to consent. Since consent was not required in the case of these Members, that part of the waiver does not affect their entitlement to an early retirement pension as a result of the voluntary redundancy. However, the second limb of the waiver does apply. These Members are not concerned with consent. Their benefits arise because voluntary redundancy is retirement. The only reason they might not have been entitled to early retirement benefits, in the absence of any waiver, was the argument that voluntary redundancy did not amount to "retirement". Quite clearly, the second limb of the waiver is wide enough to protect Holdings from the adverse (for it) result given by this judgment that voluntary redundancy does amount to retirement. This conclusion, indeed, lends further support to the conclusions I have reached in relation to the Eligible ER Members. It would be odd if those Members (in respect of whom consent was required) were in a better position, vis à vis the waivers, than the C Plan Over 60 Members (where no consent is required).
654. In my judgment, the position under Mercury and Apollo is the same as under Phoenix. Applying the common ground approach, the waivers are effective to waive all claims which the affected members have to benefits under Rule 2(1).
655. As with Phoenix, it is not necessary to decide whether the common ground approach or the alternative approach is correct. However, if it were necessary for me to decide the point, I would reject the common ground approach and adopt the alternative approach and I do not do so. My reason, in brief, is that the ordinary and natural meaning of Rule 2(1) is, in my judgment, unambiguous. It requires a very strained reading to relate the requirement for consent to access to benefits rather than to a change of status from active membership to retirement. I am wholly unpersuaded by Mr Newman's four arguments to the contrary.

Damages/compensation

656. The final issue in relation to the separation agreements relates to whether any of the affected members have a financial remedy arising out of the invalid change in ER policy. The Trustee's position is that the waivers do not affect any claim for damages for breach of contract or the Imperial duty on the basis that the affected members should have but did not become entitled to an immediate pension on advantageous terms because of the wrongful withdrawal of the Old ER Terms. It is said that the waivers deal only with an entitlement to an early retirement pension but do not purport to waive any entitlement to damages arising from the fact that a member was not entitled to an immediate pension because of Holdings' breach of duty.
657. Holdings' primary position is that the waivers apply to any financial claim just as much as they apply to a claim to be entitled to the early retirement pension. To the extent that a financial claim is for damages in respect of a failure to provide an early retirement pension as a result of voluntary redundancy under the Separation Programmes, I agree. I consider that, by waiving the claim to the entitlement which would otherwise have existed on his voluntary redundancy, the affected member necessarily waives any financial claim based on his inability to claim the entitlement which would otherwise have existed as a result of that redundancy.
658. What the waiver does not do, in my view, is waive any claim which an affected member has other than one arising out of the voluntary redundancy. The member does not waive any claim which he has in respect of any breach of duty on the part of Holdings arising out of Project Waltz. In practice, the important claim which is not waived arises if an affected member could show that he would not, had the Old ER Policy not been changed, have taken voluntary redundancy but would, instead, have retired. In those circumstances he may have a claim for financial compensation.
659. In relation to such a claim, I consider the position of a C Plan Over 60 Member first. I am told by Mr Simmonds that there were, as a matter of fact, no C Plan Over 60 Members who entered into waivers under the Mercury and Apollo Programmes.  It has not, at the time of the handing down of this judgment, been possible for the Trustee and the RBs to confirm the position.  If Mr Simmonds is correct, then there is no issue so far as concerns those Members.  But in case there were any such Members, I should deal with the issue. It is difficult to see how the invalid change in ER Policy has any impact on such Members taking voluntary redundancy under the Mercury and Apollo Programmes. They did so after my decision in the Rectification Action. Those members knew that consent was not needed. They nonetheless took advantage of the terms offered for voluntary redundancy, including provision of the waivers. They did so no doubt after taking, or at least after being invited to take, legal advice. I do not know why they would have taken that course. Whatever their reasons, Holdings' breach of duty in relation to Old ER Policy cannot have had anything to do with their choice. They have, in my view, no claim for damages arising as a result of the Separation Programmes. On the other hand, there may have been (I simply do not know on the evidence) C Plan Over 60 Members who decided not to take voluntary redundancy under the Separation Programmes but instead simply left service and claimed benefits under Rule 2(1). They clearly have no damages claim arising out of the Separation Programmes. So far as C Plan Over 60 Members taking voluntary redundancy under Phoenix, their position, in this respect, is the same as that of the Eligible ER Members discussed in the following paragraphs.
660. Turning to the position of Eligible ER Members, if the Old ER Policy had remained in force, it would have been proper for Holdings, nonetheless, to have offered voluntary redundancy under the Separation Programmes. A Member would then have had a choice to accept the offer (in which case he would have had to provide a waiver in respect of which no issue of breach of duty would arise since the hypothesis is that the Old ER Policy remains in force). Or the member could have refused the offer and instead resigned.
661. Accordingly, it is open to a Member to argue that his choice to take voluntary redundancy was affected by Holdings' breach of duty. Instead of having the choice of taking voluntary redundancy or resigning on the Old ER Terms (in which case he would become entitled to an early retirement pension), he only had the choice of taking voluntary redundancy or resigning on the New ER Terms (in which case he would not become entitled to an early retirement pension). He can argue that, but for the breach of duty, he would have resigned rather than taken voluntary redundancy and has suffered a loss as a result.
662. What he cannot argue is that because he was made redundant, it can properly be inferred that he would have resigned. Although Holdings has conceded that, for any period during which it ought to have applied the Old ER Policy, it may properly be inferred that a C Plan Over 60 Member or an Eligible ER Member who became a Deferred Retiree of the DB Plans would instead have taken early retirement from active status, that concession was expressly subject to its submissions in relation to Issue 28. The waivers prevent him, in my view, from running that particular argument (but do not prevent him from proving that he would in fact have resigned, see paragraph 655 above).
663. The Trustee raises an alternative case that the FAQ statement constituted a misrepresentation if, contrary to its primary case, the waivers are effective to preclude a damages claim. I doubt very much indeed that any claim in misrepresentation can be asserted given my analysis of the FAQ Statement. However, I decline to deal with this as a matter of decision. It is an aspect that should be dealt with in the context of an actual dispute in separate litigation between an affected member and Holdings.
664. The answer to Issue 28 is that the waivers do disentitle the affected members from the remedies identified in Issues 23 to 27 save that it remains open to such a member to argue (so far as permitted in accordance with the discussion above) that, but for Holdings' breach of duty, he would not have taken voluntary redundancy under the separation programmes.

Issue 29: consultation failures

29. Are any members in principle entitled to damages or equitable compensation from Holdings as a result of the manner in which the Consultation was conducted, or to some other and if so what remedy?

In principle, what is the measure of such damages or compensation?

665. Holdings accepts, subject to any appeal, that the result of the Judgment is that the manner in which Holdings consulted employees in relation to Project Waltz constituted a breach of its implied duty of trust and confidence. I shall refer to this as "the consultation breach". This was on the footing that, if a consultation was carried out at all (which was not mandatory), it had to be carried out properly. Although not every failure to carry out a consultation properly necessarily gives rise to a breach of the implied duty, I held that the failures in the present case were so serious as to give rise to a breach. The RBs put their claim on the basis that this breach was a breach of the implied contractual duty rather than a breach of the Imperial duty: see Judgment [1537]. I proceed on that basis. Holdings accepts that, in principle, this breach is capable of sounding in damages in the same way as any other breach of contract.
666. However, Holdings' primary position is that the Court cannot award damages or any other form of financial compensation because the remedies available are exhaustively stated in the Consultation Regulations which do not provide for a right to recover damages or other compensation. Mr Simmonds, perfectly correctly, points out that the complaint of each member in relation to the consultation is one, which if valid, would constitute a breach of the Consultation Regulations. He then submits that the Court cannot, in reliance upon a breach of the implied contractual term, supplement the remedies available to individual members since that would be to undermine the statutory scheme. He relies, in this context, by way of analogy on Johnson v Unisys Limited [2003] 1 AC 518 ("Johnson").
667. I remind myself of what I said in Judgment [1554] to [1556]. I there rejected Mr Simmonds' submission that the only remedies for a breach of the Consultation Regulations (by which I included conduct which not only breached the Consultation Regulations but which gave rise to some other breach of contract or other obligation) were those set out in the Consultation Regulations. It is right that the remedies for breach of the Consultation Regulations are restricted to those listed; but that is not to say that conduct which results in a breach of the Consultation Regulations and of some other right cannot be visited with some other remedy. I concluded that there was a breach of duty and that a remedy outside the Consultation Regulations was available.
668. Mr Simmonds submits that those paragraphs of the Judgment are not determinative because they go to remedies and the whole question of remedies was reserved to the later hearing. He says that I am not bound by what I said and that I should now say that the members have no damages claim.
669. In addressing Mr Simmonds' argument, it needs to be borne in mind that the consultation which Holdings actually conducted related to the whole of Project Waltz and not just the matters required by the Consultation Regulations: see Judgment [1541]. Even a consultation which is not mandatory must be carried out properly if it is carried out at all: this is clear from the decision in UK Coal Mining Ltd v National Union of Mineworkers (Northumberland Area) [2008] ICR 163 cited in Judgment [1543] and from a case in a public law context, R (Eisai Limited) v NICE [2008] EWCA Civ 438 at [24]. UK Coal was not concerned with a breach of the implied duty of trust and confidence; but what was said about the need for an actual consultation to be carried out properly does not depend on what, if any, duty is breached or on the remedies, if any, for breach of a particular duty.
670. In the Judgment, I held, in particular, (i) that there was a deliberate provision of misleading information (see Judgment [1573]) and that (ii) certain errors might be seen as deceptions (see Judgment [1583]) concluding that Holdings' behaviour was such as to cause members to question its integrity and good faith (see Judgment [1574]). I should add that the seriousness of the actual conduct remains, even if I was wrong (because of the analogy with Johnson) to have held that the implied duty of trust and confidence could apply to that aspect of the overall conduct, however serious, which gave rise to a breach of the Consultation Regulations.
671. In the course of the hearing, I was taken again to the Consultation Regulations and also to the enabling provisions found in section 259 Pensions Act 2004. Section 259(3) provides:

"The validity of any decision made in relation to an occupational pension scheme is not affected by any failure to comply with regulations under this section."

672. Mr Simmonds places considerable reliance on that provision in the context of his argument that the Consultation Regulations contain the exclusive remedy in relation to a breach of them notwithstanding that the breach might itself give rise to a breach of the contractual duty of trust and confidence or form an element of a wider course of conduct which gives rise to a breach of the implied duty.
673. The Trustee submits that Holdings' argument is to be rejected for three reasons (each of which I will look at in turn):

i) First, it is too late for the argument to be advanced.

ii) The Consultation Regulations do not cover the wrong for which the remedy is sought.

iii) The Consultation Regulations are not an exhaustive code.

674. The first point has two limbs. The first is that the matter is res judicata or it would be an abuse of process for it to be raised now. The second is that it is too late to take the Johnson point now.
675. As to the first limb there are tensions in the Judgment. Although Judgment [1554] to [1556] appears to resolve the point against Mr Simmonds, I did, at Judgment [1532] say that I was leaving the whole question of remedies to the further hearing. In the light of my answers to the second limb and to the second and third points, I do not propose to resolve the first limb.
676. As to the second limb, it is said that the argument is misconceived. Mr Spink submits that the conclusion of the House of Lords in Johnson was that the implied duty of trust and confidence did not apply at all to the manner of dismissal of the employee. Accordingly there was no breach of contract in the first place. There is no analogy to be drawn.
677. I accept that submission. It is to be noted that the statutory scheme in that case gave Mr Johnson a remedy for precisely the complaints which he made: see Lord Hoffmann at [56]. Lord Millett agreed with Lord Hoffmann's opinion, adding some observations of his own; and Lord Bingham agreed with both of them. Lord Steyn's opinion is consistent with the others. All of the judges considered that the statutory scheme contained the exclusive territory for the determination of Mr Johnson's rights and remedies. As Lord Nicholls succinctly put it:

"…I am persuaded that a common law right embracing the manner in which an employee is dismissed cannot satisfactorily coexist with the statutory right not be unfairly dismissed. A newly developed common law right of this nature, covering the same ground as the statutory right, would fly in the face of the limits which Parliament had already prescribed….."

678. This is the same point which Lord Hoffmann makes at [58] of his opinion. Lord Nicholls himself explained the effect of Johnson in Eastwood v Magnox [2005] 1 AC 503 at [13] where he said:

"… it is not for the courts to extend further a common law implied term where this would depart significantly from the balance set by the legislature."

679. It is not open to Holdings to argue before me that there was no breach of duty in the first place. I have already decided, as part of my decision, that there was a breach of duty. Johnson is not an authority for the proposition that the court cannot give a remedy where there is a breach of duty. I do not, in any case, see Johnson as providing any significant support for an argument that the implied duty of trust and confidence does not extend to the statutory consultation required by the Consultation Regulations and which should lead me to question my decision. There was, in any case, a further reason why the implied term of trust and confidence did not apply to the circumstances of Mr Johnson's dismissal, a reason which is not applicable in the present case. It stemmed from the difficulty which Lord Hoffmann saw in extending an implied term, which was concerned with the continuing relationship of employer and employee, to the circumstances in which that relationship could be brought to an end. I agree with the submission on behalf of the Trustee that the whole point of the Johnson exclusion area is that the employee has already been dismissed and has an individual statutory remedy for unfair dismissal; in those circumstances, the implied duty of trust and confidence has no further part to play.
680. In the present case, however, Holdings' argument is to the effect that the Consultation Regulations have removed the contractual protection of the implied term in an ongoing employment relationship with no alternative statutory remedy being given to the employee. But that is inconsistent with my finding that Holdings was in breach of duty in relation to its conduct of the consultation and is one which it can raise only on appeal. But quite apart from that, the point which I wish to make at this stage is that Johnson does not, for the reasons given above, assist in the resolution of that issue one way or the other.
681. The Trustee's second reason for rejecting Holdings' submission is that the Consultation Regulations do not cover the wrong for which the contractual remedy is sought. It is the reason which I gave in Judgment [1555]. I stand by my conclusion and reasoning. The essential point is that the conduct of Holdings gave rise not only to a breach of the Consultation Regulations but also to a breach of the contractual duty. That contractual duty is not, in my view, circumscribed by the Consultation Regulations; there is, for reasons already given, no proper analogy with Johnson.
682. I would only add this. In Judgment [1555] I gave an example which illustrates why, in principle, there ought to be a remedy in a case such as the present. To build on that example (or perhaps merely to repeat it) suppose that an employer decides to consult on a matter which is not covered by the Consultation Regulations at all; and suppose that the manner of consultation is such as to give rise to a breach of the implied duty of trust and confidence. Clearly, in principle, the affected employees would have a claim for damages for breach of the implied duty. Why should it make any difference if the consultation includes, or even relates only to, matters which do fall within the Consultation Regulations? I consider that it should not do so. Section 259(3) Pensions Act 2004 does not require a different answer to be given: that section is concerned with the validity of the decision not with the consequences of that decision if it gives rise to a breach of some other duty.
683. The Trustee's third reason for rejecting Holdings' submissions is that the Consultation Regulations are not an exhaustive code and they do not exclude a claim to enforce contractual remedies. The Trustee has devoted 7 pages of submissions to this point including reference to Islington LBC v UCKAC [2006] EWCA Civ 340, Monro v HMRC [2009] Ch 69 (CA) ("Monro") (the submission referring to what Arden LJ said at [19] and her references to the authorities cited at first instance by Sir Andrew Morritt C), HMRC v Total Network SL [2008] 1 AC 1174 at [130] and R (Child Poverty Action Group) v Secretary of State for Work and Pensions [2011] 2 AC 15 ("CPAG"). I do not propose to refer to any of the authorities myself except for Monro and CPAG.
684. In Monro, at [22], Arden LJ stated the principle in this way:

"….if Parliament creates a right which is inconsistent with a right given by the common law, the latter is displaced. By "inconsistent" I mean that the statutory remedy has some restriction in it which reflects some policy rule of the statute which is a cardinal feature of the statute. In those circumstances the likely implication of the statute, in the absence of contrary provision, is that the statutory remedy is an exclusive one."

685. In CPAG, Lord Dyson JSC explained, at [32] to [33], that the test is whether in all the circumstances Parliament must have intended a common law remedy to coexist with the statutory remedy. "If the two remedies cover precisely the same ground and are inconsistent with each other, then the common law remedy will almost certainly have been excluded by necessary implication." He continued at [34]:

"The question is not whether there are any differences between the common law remedy and the statutory scheme. There may well be differences. The question is whether the differences are so substantial that they demonstrate that Parliament could not have intended the common law remedy to survive the introduction of the statutory scheme. The court should not be too ready to find that a common law remedy has been displaced by a statutory one, not least because it is always open to Parliament to make the position clear by stating explicitly whether the statute is intended to be exhaustive. … The question is whether, looked at as a whole, a common law remedy would be incompatible with the statutory scheme and therefore could not have been intended by coexist with it."

686. I propose to take the point very shortly bearing all of those submissions and authorities in mind. I pay particular attention to that citation.
687. The argument on behalf of Holdings depends critically on the proposition that the statutory scheme is the exhaustive source of any remedy for conduct which amounts to breach of the Consultation Regulations. In other words, the statutory scheme displaces the common law including what would otherwise be the members' contractual rights. In this context, the breach of the implied duty of trust and confidence is not that the consultation failed to comply with the Consultation Regulations; rather, there was a breach of the implied duty because of the way in which the consultation was carried out. Although, absent the Consultation Regulations, Holdings would not have been obliged to carry out a consultation at all, the fact is that it did so; and having done so, it was under an obligation to do it properly. If, as I have held to be the case, its failings were serious enough to have amounted to a breach of the implied duty, then Holdings' submissions, if correct, would result in the statutory scheme displacing the members' contractual rights.
688. The only basis on which it could be argued that these common law rights were displaced is, it seems to me, that those rights would (to use different possible formulations) undermine or circumvent or be inconsistent with the statutory scheme. In my judgment, this is not established: the statutory scheme is not undermined by giving a remedy to the members for breach of the implied duty of trust and confidence; no limitation on the members' contractual rights is circumvented because the statutory scheme provides no such rights.
689. There is not any inconsistency in the ordinary sense of the word in Holdings being subject to regulatory sanctions and the members having individual claims to damages or other civil remedies. Johnson provides no support to the contrary argument. Unlike in the case of the statutory scheme in Johnson, the Consultation Regulations give an individual member no remedy at all. The statutory sanctions are all regulatory and do not give the members any personal remedy, let alone any sort of compensatory remedy. In Johnson, it would have circumvented the statutory scheme if the court could give a financial remedy exceeding that which Parliament had laid down and which the employment tribunal could award. In the present case, I do not consider that the granting of a financial remedy to an individual member for breach of the implied duty would undermine the statutory regulatory scheme. I would see the argument that, if the Consultation Regulations had themselves provided an individual member with a financial remedy limited in amount, then the member would not be able to obtain a greater amount on the basis of a breach of the implied term. But that is not the present case.
690. Nor is there any inconsistency in the sense in which Arden LJ uses the word in Monro. It is not possible to detect in the Pensions Act 2004 (which is the statutory authority for the Consultation Regulations) any policy rule which is a cardinal feature of the statute which is found reflected in the restriction on remedies found in Regulation 18.
691. On the basis of the above discussion, I have no reason to doubt that the implied duty of trust and confidence was wide enough to cover the breaches which I held in the Judgment to exist. In my judgment, the members are entitled to a remedy for that breach. Their rights are not displaced by the statutory scheme. It follows in my view, as the Trustee submits, that the members are entitled to claim the ordinary remedies for breach of contract. These remedies are not, in principle, restricted and may therefore extend to injunctive relief (if appropriate in accordance with established principles) as well as damages.

Suspension of Project Waltz

692. The Trustee's contention is that the implementation of Project Waltz should be suspended pending a proper consultation. One aim of this submission is to ensure that Holdings should not be able now to take any step in Project Waltz which it could properly take assuming (i) that the consultation had originally been conducted properly and (ii) that Project Waltz would have been adopted following such a consultation. Another aim is to prevent Holdings from making any fresh decision to implement any of the same or similar proposals as were comprised in Project Waltz without a further consultation: this is particularly relevant to the possible service of new exclusion notices which, clearly, cannot form part of the original Project Waltz proposals (because they are new) but in respect of which the Trustee contends a further consultation must be conducted. I consider the most suitable way to address the Trustee's contention is to consider the impact on each of the 5 elements of Project Waltz, which I do in the following paragraphs. My discussion is without prejudice to the point (I will refer to it as "the overarching point") which I deal with after that discussion, namely the submission on the part of the Trustee that, in the light of announcements made to members and the setting up of the PCC, members were entitled to expect that Project Waltz would not be implemented until the consultation had taken place: the submission is that no such consultation has taken place and that the whole of Project Waltz must, for that reason alone, be postponed until a proper consultation has been carried out, and this is so even though the period during which any Reasonable Expectation could be held has expired.
693. Element 1: closure from 6 April 2011. I have already held that Holdings will need to serve new exclusion notices if it wishes to terminate further benefit accrual. Such notices can only have prospective effect. I have also held (see my answer to Issue 15.4), that a further consultation will be necessary in relation to that. Although the submission was not made in relation to Issues 13 to 15, that I should grant injunctive relief, it is made in relation to Issue 29. I consider that I should do so unless Holdings is willing to give certain commitments to the Trustee and members or undertakings to the court. The position is not entirely straightforward to deal with because of the possibility of appeal and also because it is part of the Trustee's case that the failures in the consultation process are, by themselves, sufficient to justify postponement of Project Waltz even if Holdings were to be successful in an appeal against my other conclusions in the Judgment.
694. Ignoring, for the moment, the impact of any appeal, it would not be necessary for me to grant injunctive relief in relation to the service of further exclusion notices if Holdings were to give a binding commitment to the Trustee and members and/or an undertaking to the court that it would not serve further exclusion notices without conducting a proper consultation as required by the Consultation Regulations. In saying that, I am simply indicating the sort of response from Holdings which would persuade me not to grant injunctive relief. Absent such a commitment and/or undertaking, I consider that it would be appropriate for me to grant injunctive relief to prevent a breach by Holdings of its statutory duty under the Consultation Regulations and to prevent a further breach of the implied duty of trust and confidence.
695. Now, taking into account the possibility of appeal, one outcome might be that the Exclusion Notices are held to be valid. In that case, an injunction or undertaking such as I have discussed above would not in fact bite: the injunction or undertaking would not prevent the Exclusion Notices taking effect according to their terms. But another outcome might be that the Exclusion Notices can be set aside and that new exclusion notices are required but without the need for any further consultation. In that case, to have granted the injunction or required the undertaking would have given the members more than they are entitled to: and so it can be said that there ought to be nothing to prevent Holdings from now serving new exclusion notices without suffering the delay of a further consultation. Accordingly, a mechanism needs to be found under which Holdings can serve an immediate exclusion notice which will have effect from the date of its service if the Judgment is overturned on appeal, but which will not have any effect if the Judgment is upheld (or not, in the event, challenged). And there will need to be a resolution of the basis on which Plans should be administered pending any appeal. I must leave matters there for the moment since such a mechanism has not, of course, been debated before me, and nor has the basis of interim administration; the parties must be given the opportunity to do so if a mechanism cannot be agreed.
696. Element 2: the NPAs. If my conclusions concerning the NPAs are correct, the failure to consult adds nothing to the RBs' case. In accordance with my answers to Issue 1 to 12, the affected members are not bound by the non-pensionability term (indeed for members sending protest emails, that term is not a term of the varied contracts of employment in the first place). It may be that fresh NPAs would be sought by Holdings. Even if new NPAs would form part of Project Waltz, and even if the overarching point is correct, there is nothing to preclude Holdings from seeking to obtain new NPAs which would, on an individual basis, take effect after the Longstop Date provided that such NPAs are not procured in a way which gives rise to a further breach of the implied duty of trust and confidence. I do not consider that I should accede to the Trustee's request that Project Waltz should be "suspended" in such a way as to preclude Holdings from seeking, in a proper way, further NPAs. In the light of that conclusion, the possibility of a successful appeal by Holdings in relation to the NPAs does not need to be considered.
697. As to "a proper way", I ought, I think, to mention Regulation 8(3)(e) of the Consultation Regulations which will need to be borne in mind. I do not say that it in fact applies: I have heard no submissions. Nor do I say that, even if it does, it actually has any real impact since it might be said that a member who agrees to a fresh NPA must, necessarily, have been properly consulted; I say only that it must be addressed.
698. Element 3: the early retirement window. This element of Project Waltz was fully implemented. The affected members have only financial claims. There is nothing to suspend.
699. Element 4: the New ER Policy. The period during which members were entitled to hold the Reasonable Expectation in relation to early retirement expired some months ago. There is therefore, prima facie, no reason why Holdings should not now be able to adopt the New ER Policy or something like it. I have discussed this aspect at paragraphs 468ff above. Subject to the overarching point, I do not consider that Holdings is precluded from adopting a new ER policy unless and until it has undertaken a further consultation. I should make it clear that I am not deciding that Holdings can properly introduce a new ER policy; there may be reasons why that would not be a proper exercise of its discretion although I do not know of any reasons why it would not be proper and it is not an issue before me. Whether Holdings can adopt a new policy without giving some further period of notice (during which members could make a final decision) is not a matter which has been debated before me.
700. Element 5: creation of Hybrid Deferred status. I have said a little about Hybrid Deferred status in addressing Issues 14.4 and 15.3. Further detailed questions arise under Issue 30. So far as I can tell, the only aspect of hybrid deferred status to which any question of "suspension" of Project Waltz might go is the extent to which affected members would be entitled to that status were Holdings to serve new exclusion notices. I have already dealt so far as necessary with the possibility of new exclusion notices under Element 1 above. There is nothing in Element 5 which calls for "suspension" of Project Waltz in any way.
701. My conclusion in relation to the Trustee's submission that Project Waltz should be suspended is (subject to the following paragraphs) that the only possible relief relates to the terms on which any new Exclusion Notices may be served, as to which see Element 1.

The overarching point

702. That leaves for consideration the overarching point (ie that Holdings cannot take any steps in relation to Project Waltz without conducting a full consultation). Not only would it be unable to serve new Exclusion Notices (which, as I have concluded, is the position quite apart from the overarching point), but it would not be able to adopt the New ER Policy or to procure new NPAs. This issue arises in two contexts: the first is that I am right in my conclusions in the Judgment about Holdings' breach of its implied duty of trust and confidence and its Imperial duty, and the second is that I am wrong. I propose to restrict myself to consideration of the first of those contexts. It is not possible sensibly to deal with the second context without knowing the extent to which and the basis on which the Court of Appeal disagrees with my conclusions. The consequences of the consultation breach in the second context should be dealt with by the Court of Appeal.
703. I turn, then, to the consultation breach on the basis that the Judgment is correct. I have found the question which the Trustee's submissions raise one ofthe most difficult questions to decide among the mass of issues which I am asked to decide.
704. It is relevant to note that the Trustee contends that Holdings should not implement any part of Project Waltz without further consultation. It is not suggested, and I do not consider that it could be suggested, that Holdings could not make any changes whatsoever to the Plans without prior consultation (save such consultation as is required by the Consultation Regulations). For instance, it would be open to Holdings to adopt a new ER policy different from the New ER Policy without consultation, although it would, of course, need to exercise its power to change that policy in a proper manner.
705. So, what does it mean to say that Project Waltz should not be implemented? Surely it was of the essence of Project Waltz that the 5 Elements should take effect as from the dates appropriate to each element: the Exclusion Notices were to take effect from April 2011 (it being thought that the final salary link would be broken), the NPAs were to achieve the non-pensionability of pension increases until then (and thereafter in case the Exclusion Notices did not break the final salary link) and the New ER Policy was to have immediate effect. The result of the Judgment and of this judgment is that nothing like those proposals can take effect. At best, from Holdings' point of view, new exclusion notices can be served only after a further consultation, new NPAs might be possible and the New ER Policy could be introduced (although perhaps only after a period of notice). It does not seem to me to be realistic to describe these possibilities as implementation of Project Waltz.
706. Rather, it will be the implementation of new proposals, an implementation which will require a fresh decision by Holdings. That decision will have to be made in the context of current circumstances. Any decision will need to take account of the different circumstances which exist today from those which existed in 2009. In particular, the financial and economic conditions are different, NPPC no longer features in IBM's EPS reporting and Holdings now has my decisions in the Rectification Action relating to the rights of C Plan members and in the Judgment relating to the preservation of the final salary link when the Exclusion Powers are exercised. Further, Holdings should now understand from the Judgment that it cannot simply act on the basis of threats of dire consequences from CHQ if it does not act in a particular way desired by CHQ; and CHQ should understand that it is no defence to an allegation of breach of duty by Holdings that it is simply implementing proposals which, for all practical purposes, have been imposed by CHQ.
707. I therefore reject the overarching point.
708. In case that is wrong, the position needs to be considered on the footing that Holdings decides to implement fresh proposals which can properly be described as being part of Project Waltz.
709. The consultation breach is particularly serious given the other breaches of duty established in the Judgment. Indeed, had the consultation been conducted properly, those other breaches might never have occurred. The consultation process was important as I explained in the Judgment. I held that the result of a further consultation is not a foregone conclusion; it would fulfil a useful purpose in informing Holdings in its decision making process. An obvious way of correcting that breach would be to require Holdings to hold a further consultation before deciding to implement Project Waltz.
710. Against that, it is to be remembered that Holdings is not, absent the overarching point, obliged to consult before attempting to persuade members to accept new NPAs or before changing its ER policy. Nor was it obliged to do so in 2009; its obligation was only that if it did consult (as it did) it should have done so properly. Of course, Holdings must act properly if it is to make a fresh decision to proceed with Project Waltz or so much of it as can now be implemented. And in doing so it must take account of the different circumstances which exist today from those which existed in 2009 which I have identified above. But that is not to say that a consultation is mandatory. Even if the Trustee is correct when it suggests that Holdings cannot, consistently with its implied duty of trust and confidence, introduce major pension changes such as Project Waltz, without consultation, I do not consider that Holdings would, simply by virtue of seeking new NPAs or changing its ER policy to accord with the New ER Policy, be putting itself in breach of duty unless it first consulted subject to the possible application of Regulation 8(3)(e) of the Consultation Regulations (as to which see paragraph 694 above).
711. However, that is not to say that the particular circumstances in which the original (non-mandatory) consultation took place might not lead to a different conclusion. Those circumstances include the following. In Judgment [1263] and [1304] I recorded that the consultation period ran until 20 October 2009. The employees were told that no final decisions would be made until the consultation was complete, that is to say, in the event, not before 20 October 2009. I also explained the process for the setting up of the PCC through which the consultation would be carried out. The Trustee's case is, that having announced the consultation and set up the PCC, it would be a breach of the implied duty of trust and confidence for Holdings to refuse to do what it promised, that is to say to carry out a genuine consultation before making its decisions on Project Waltz.
712. The Trustee's written closing submission puts its case succinctly in this way:

"Notwithstanding this, IBM – without having genuinely consulted – pressed ahead with making a final decision on Project Waltz in 2009, subsequently sought to implement Project Waltz, now seeks to have Project Waltz 'validated', and dismisses out of the hand any suggestion that it should consult again. All this amounts to not only a past breach of duty in 2009 but also a persistent and continuing breach of the implied duty of trust and confidence."

713. Holdings would say in relation to that submission that it is putting the Trustee's case far too high. The remedy for this breach of contract is damages. It would be to give the members too much to postpone the time at which Holdings can properly implement proposals to effect pension costs savings until after a further consultation. An award of damages would reflect the actual breach namely a failure to consult properly in 2009 and would reflect what would have happened at that time if Holdings had acted properly. Given the effect of my decision in relation to Issue 1 to 28, it may be that the members have suffered no loss at all as a result of the failure to consult. In contrast, to require a further consultation would be to give the members something which they might have come nowhere near obtaining if Holdings had acted properly in 2009: it is highly unlikely, Holdings would say, that it would simply have left pension provision as it stood.
714. As I have said, I have found this a very difficult point. In the final analysis, I am persuaded by the Trustee's argument. Once having decided to consult, Holdings was obliged to consult properly. In the context of the breaches to which Project Waltz gave rise, the failure to consult at the time was, as already stated, a serious matter. The affected members were entitled to have a properly conducted consultation carried out before any decision about Project Waltz was made which was not done. Before remaking any decision necessary to implement new exclusion notices, new NPAs or the New ER Policy (or anything resembling it) Holdings should if, but only if, contrary to my earlier conclusion, these changes would now form part of Project Waltz, carry out a further consultation. The affected members are entitled, in my judgment, to regard a failure by Holdings to do so as a further breach of the implied duty of trust and confidence.
715. Moving away from the suspension issue, the consultation breach gives rise to a claim for damages. In the light of my conclusions on Issues 1 to 28, I doubt that the consultation breach gives rise to any additional claim over and above the claims arising from the other breaches already discussed. I have considered the principles, to some extent, under Issue 5. I say that the consultation breach does not add to the affected members' claims because, as explained under Issue 5, the Court has to conduct an evaluation of what would have happened if Holdings had not acted in breach of its duties. It is not easy to see how that objective evaluation could depend on whether or not Holdings had carried out a proper consultation because the Court will be in possession of all the material which Holdings should have taken into account: all of the points which could have been made by the affected members, the PCC or the Trustee in an actual consultation can be made to the Court by way of submission.
716. The position may, I accept, be different if this matter goes to the Court of Appeal and if my conclusions in the Judgment or in the earlier parts of this judgment are, in whole or in part, reversed. Without knowing the extent of such reversal and the reasons for it, it is not possible for me to identify the correct principles to apply in relation to the assessment of damages for breach of the implied duty of trust and confidence in relation to the failure to consult properly. Contrast, for instance, these two possible outcomes on appeal so far as concerns the NPAs: (i) the procuring of the NPAs was a breach of the implied duty of trust and confidence but gives rise only to a claim in damages and (ii) the procuring of the NPAs was not a breach of duty at all (whether in isolation or as part of Project Waltz). The measure of damages for the breach of the implied duty in relation to the consultation breach would, I think, be entirely different in these two cases. Further, if I am wrong in my assessment that the conduct of Holdings amounted to a breach of the implied duty and the Imperial duty, it is possible that I am wrong, also, in my conclusion that the deficiencies in the consultation process amounted to a breach of either of those duties.
717. Accordingly, I do not propose to say anything more about damages, whether in relation to the NPAs or any other aspect of the case other than to say that any award of damages must take account of the other remedies which are granted to a member in the context of the breaches of duty brought about by Project Waltz overall. For instance, it may be that by setting aside the Exclusion Notices, a member will obtain a more valuable remedy than he would achieve by an award of damages in relation to the service of the Exclusion Notices if they were not set aside. That must be borne in mind when it is suggested that the consultation breach has resulted in actual damages. I consider that the matter of damages is best left there: it will be for the Court of Appeal to resolve issues of damages in the light of its conclusions should they differ from my own.
718. As previously, I am not willing to deal with the issue of exemplary damages.
719. The answer to Issue 29 is that members are in principle entitled to damages as a result of the manner in which the consultation was carried out. I am unwilling to say more about the principles to be adopted than as set out in my discussion under Issue 5. Holdings ought not to serve further exclusion notices without going through the consultation required under the Consultation Regulations. I would be prepared to consider granting injunctive relief to the limited extent discussed above if the parties are unable to agree a mechanism for preserving their respective interests as suggested above. Holdings is obliged to consult on any matters which form part of Project Waltz, but I am of the view that any new proposals are almost bound to be something very different from Project Waltz, not least because of the timing issues, so that any need for consultation on new proposals is not a consequence of the Judgment or this judgment. Since the possibility of injunctive relief along the lines I have discussed was not fully debated before me, I think it is appropriate to allow any party to make further submissions on the point.

Issue 30: the underpin applicable to Hybrid Deferred Members

720. By a letter dated 9 July 2014, Nabarro conceded on behalf of the Trustee (after consultation with, and with the agreement of, the RBs) that Holdings is correct in its contention that the basis for calculating the Hybrid Deferred underpin is that which is set out at issue 30.1 (rather than, as asserted in the Trustee's position paper, that which is set out at issue 30.2), ie that where a Hybrid Deferred member takes early retirement from active status, the underpin is calculated by applying to the member's accrued benefits the actuarial reduction provided for in Schedule F, rule 2(3)(b) of the Main Plan Defined Benefit Rules or Schedule I, rule 2(3)(b) of the I Plan Rules (as the case may be), in relation to the period between the member's retirement and his Normal Retirement Date. During the course of the hearing it was agreed between the Trustee and Holdings that the underpin would be calculated differently in the case of ill-health early retirements; the ERD (if any) to be applied in such a case is to be the factor which would apply to active members in the relevant Plan. I am prepared to make declarations to reflect these agreements.

Postscript

721. Following receipt of this judgment in draft, Mr Simmonds has raised a point in relation to paragraphs 378 and 430 above where I have referred to a contractual duty on the part of Holdings. He points out that, in the Judgment, I recorded that the employer of nearly all of the relevant individuals was IBM UKL (and he adds than none of the relevant individuals was employed by Holdings) and says that there cannot, therefore, have been a breach of any contractual duty on the part of Holdings. His point is relevant to other paragraphs too: there are many places in both the Judgment and this judgment where I have referred to Holdings' contractual duty.
722. In ordinary circumstances, I would have delayed the handing-down of this judgment to hear further argument about whether this point is something which should be reflected in this judgment and perhaps even in the Judgment as well. But the draft judgment having been sent to the parties and disclosed to IBM Corporation with my consent, there is now a commercial imperative for IBM Corporation in terms of its US reporting obligations for this judgment to be released.

a. The point now raised by Mr Simmonds was not taken on behalf of Holdings during the remedies hearing where references were, on the whole, to "IBM" without distinguishing which IBM entity was being referred to. The Judgment had not drawn the distinction and it was not suggested at the remedies hearing that my conclusion in the Judgment that Holdings had been in breach of its contractual duty was wrong for the reason that Holdings was never the employer of the members. If the point had been taken, I would have needed to analyse the positions of Holdings and IBM UKL separately, an analysis which would give rise to a number of questions which I do not propose to address at this stage.

723. Accordingly, this judgment must be taken as provisional so far as concerns the answers which I given to some of the questions concerning contractual liability. But the answers which I have given to many of the questions will be unaffected. Thus the answers relating to Holdings' breach of its Imperial duty and the remedies for breach of that duty would remain unaffected. Further, the contractual analysis in relation to the NPAs and Consultation would be the same, substituting IBM UKL for Holdings. This would be so, at least, unless it is sought to argue that IBM UKL was not in breach of its contractual duty of trust and confidence at all; it can be said that there is no finding in the Judgment that IBM UKL was in breach of duty since my express finding was only that Holdings was in breach of its (non-existent according to Mr Simmonds' point) duty.
724. I must leave matters there for the moment. A further hearing will need to be fixed to decide whether Mr Simmonds' point should be taken account of when determining remedies or whether it should left to the Court of Appeal to correct any error which I have made. Even if the point cannot be taken, I would want to say something about the appropriate remedies assuming that Holdings itself was not under any contractual duty. Given my conclusions in relation to the remedies for breach of the Imperial duty, it may be thought that little, if anything, turns on the point other than that certain money remedies will be available only against IBM UKL rather than Holdings.

Concluding remarks

725. I adopt the remarks made in Judgment [1596] and thank all of the teams involved once again. Although the task of writing this judgment has been less burdensome in some respects than the task of writing the Judgment, it has, nonetheless, been a mammoth undertaking. I make no apology for its length for, in reality, each of the first 29 Issues have demanded what is in effect a judgment of its own. The length of the written submissions which I received (around 400 pages), with reference to many authorities and documents, indicates, what I have been faced with. It is a matter of regret that, once again, there has been a significant period between the end of the hearing and the delivery of this judgment. I am not, myself, willing to call it delay: it is the period which the task took to complete given all the exigencies of life.
ANNEX A

List of Issues

In this List of Issues, the following definitions apply:

the "2006 Partial NPAs" means the elections made by individual members following Project Soto to remain in the DB Plans on terms as to partial non-pensionability

the "2009 NPAs" means the Initial 2009 NPAs and the Later 2009 NPAs or any of them

the "2010 NPAs" means the non-pensionability agreements purportedly entered into by members after 5pm on 16 November 2009 and before 3 March 2011

the "2011 NPAs" means the non-pensionability agreements purportedly entered into by members in connection with the 2011 salary cycles and all subsequent salary cycles (and any later versions of such agreements entered into after the 2011 salary cycle in connection with all subsequent salary cycles) after 3 March 2011 down to the current time

"Active ER Terms" means terms under Rule 2(1) of Schedule D to the Main Plan DB Rules and the first two paragraphs of Rule 6.6 of the DSL Plan Rules, including the said terms insofar as they are applicable to Enhanced Deferred Members and Hybrid Deferred Members [We are concerned in practice only with Rule 2(1) Schedule D]

"Affected Members" means the members of the DB Plans expressed to be the subject of the Exclusion Notices other than Postponed Retirees

"C Plan Over 60 Member" means a member aged at least 60 on the date his IBM employment ceased with access to Active ER Terms of the C Plan

"Chosen DC Arrangement" means the replacement pension arrangement chosen by members pursuant to the Pension Decision Tool, namely (Option A) the Hybrid M Plan, (Option B) the IBM UK PPP or (Option C) to opt out of IBM pension provision and (if chosen) to join an alternative pension arrangement provided outside IBM

the "Consultation" means the consultation exercise carried out by IBM in relation to the implementation of Project Waltz which purported to be but which was not in fact conducted in accordance with the Occupational and Personal Pension Schemes (Consultation by Employers and Miscellaneous Amendment) Regulations 2006, SI

the "DB Plans" means the defined benefit sections of the Main Plan and the I Plan

"Eligible ER Members" means those members in the C, N and (except for the purposes of Issues 23, 24 and 28, which concern the C and N Plans only) DSL Plan (including, for the avoidance of doubt, Enhanced Deferred Members and Hybrid Deferred C, N and DSL Plan Members (if created) and any purportedly excluded members who are treated as being in pensionable service on or after 6 April 2011 pursuant to Issues 13 to 18 below) who are under their Normal Retirement Date and who have a right to be considered under Active ER Terms but who need IBM's consent to retire

the "ER Window" means the early retirement window implemented in relation to the DB Plans in November-December 2009

the "Exclusion Notices" means the Exclusion Notices issued by IBM dated 28 May 2010 and the Supplemental Exclusion Notices issued by IBM dated 28 March 2011

"IBM" means the First and/or Second Claimants

the "Initial 2009 NPAs" means the non-pensionability agreements purportedly entered into by members in connection with the 2009 and all subsequent salary cycles prior to receipt of IBM's email of 9 November 2009 at 16:24

the "Judgment" means the judgment handed down on 4 April 2014 with neutral citation number [2014] EWHC 980 (Ch)

the "Later 2009 NPAs" means the non-pensionability agreements purportedly entered into by members in connection with the 2009 and 2010 salary cycles after receipt of IBM's email of 9 November 2009 at 16:24 and before 5pm on 16 November 2009

"members" (without a capital M) means the persons expressed to be the subject of the Exclusion Notices (including for the avoidance of doubt Postponed Retirees of the DB Plans) and (where the context admits) any other IBM employees with benefits under the DB Plans who were purportedly affected by the New ER Policy such as Enhanced Deferred Members of the DB Plans

the "New ER Policy" means the new early retirement policy purportedly introduced by IBM with effect from 6 April 2010

"NPAs" means the 2009 NPAs, the 2010 NPAs and the 2011 NPAs or any of them (but, for the avoidance of doubt, does not include the 2006 Partial NPAs)

the "Old ER Policy" means IBM's pre-Project Waltz policy of invariably granting consent to early retirement for members of the DB Plans who were eligible for non-ill health early retirement under the rules of the DB Plans applicable to active "Members" (as defined therein)

"Pension Decision Tool" means the announcements emailed to members by IBM on

December 2010 which included an online tool whereby members could select an option for pension provision from 6 April 2011 subject to terms entitled "Important Information"

"Separation Programmes" means the programmes operated by IBM, known as Phoenix, Mercury and Apollo, under which the employment of employees with IBM was terminated.

Other terms not defined above have the meanings ascribed to them in the trust deeds and rules of the DB Plans or the Judgment.

1. Are the Initial 2009 NPAs:

1.1 void and/or unenforceable;

1.2 voidable and liable to be set aside,

and if so:

1.2(a) can they only be avoided and set aside by the Court in relation to all or

a group of members collectively, and if so will the Court now avoid them and set them aside;

1.2(b) can they be avoided and set aside at each member's individual election, and if so can members now avoid them and set them aside, and what would they have to do or prove in order to do so; or

1.2(c) are they liable to be avoided and set aside in some other manner, and if so what manner;

1.2(d) with effect from what date (in particular from the date on which they were purportedly entered into or from the date of election to avoid and set aside);

1.3 valid?

2. Are the Later 2009 NPAs:

2.1 void and/or

unenforceable

2.2 voidable and liable to be set aside,

and if so:

2.2(a) can they only be avoided and set aside by the Court in relation to all or

a group of members collectively,

and if so will the Court now avoid them and set them aside;

2.2(b) can they be avoided and set aside at each member's individual election,

and if so can members now avoid them and set them aside, and what

would they have to do or prove in order to do so; or

2.2(c) are they liable to be avoided and set aside in some other manner, and if

so what manner;

2.2(d) with effect from what date (in particular from the date on which they were purportedly entered into or from the date of election to avoid and set aside);

2.3 valid?

3. As regards any 2009 NPAs that are void and/or unenforceable, or are avoided

and set aside:

3.1 are members entitled to keep all or part (and if so, which part) of the salary increases they were awarded subsequent to entering the agreements, and are they entitled in the future to continue to be paid salary incorporating such increases;

3.2 are all or part of such salary increases (and if so, which part) to be treated as

pensionable for the purposes of the DB Plans?

4. Having regard to the answers to Issues 1 to 3, are members in principle entitled to damages or equitable compensation from IBM for having purportedly entered into the 2009 NPAs, or to some other remedy and if so what remedy?

5. If members are entitled to damages or equitable compensation from IBM, in principle what is the measure of such damages or compensation?

6. Are the answers to Issues 1 to 5 any different for those members who purportedly entered into the 2009 NPAs having sent emails of protest substantially in the form of the 582 emails referred to at paragraph 1325 of the Judgment or other protests to similar effect?

7. To what, if any, remedy are the members who did not enter the 2009 NPAs entitled?

8. In the light of the answers to Issues 1 to 7, should the Trustee administer the DB Plans on the footing that salary increases awarded following the 2009 NPAs (and/or any salary increases that members who did not enter the 2009 NPAs are entitled to be treated as having) are:

8.1 non-pensionable or are to be treated as not having been awarded;

8.2 pensionable;

8.3 to be treated in some other way and if so what way?

9. Was it a breach of the Imperial duty or the contractual duty of trust and confidence for IBM to procure that members enter into the 2010 NPAs?

10. If there was such a breach of duty, Issues 1 to 8 are repeated in respect of the 2010 NPAs with all necessary changes.

11. Was it a breach of the Imperial duty or the contractual duty of trust and confidence for IBM to procure that members enter into the 2011 NPAs?

12. If there was such a breach of duty, Issues 1 to 8 are repeated in respect of the 2011 NPAs with all necessary changes.

13. Were the Exclusion Notices:

13.1 void and/or unenforceable

13.2 voidable and liable to be set aside,

and if so:

13.2(a) can they only be avoided and set aside by the Court in relation to all

or a group of the Affected Members collectively, and if so will the Court now avoid them and set them aside;

13.2(b) can they be avoided and set aside at each Affected Member's individual election, and if so can Affected Members now avoid them and set them aside, and what would they have to do or prove in order to do so;

13.2(c) are they liable to be avoided and set aside in some other manner, and if so in what manner;

13.2(d) may any power to avoid or set aside the Exclusion Notices be exercised independently of any such power in relation to the NPAs (so that, for example, the NPAs are avoided but the exclusion is affirmed);

13.3 valid?

14. If the Exclusion Notices are void and/or unenforceable, or are avoided and set aside:

14.1 If the Exclusion Notices are void:

(a) are they severable and partly valid

(b) if severable and partly valid, to what extent are the Exclusion

Notices valid;

14.2 If the Exclusion Notices are unenforceable:

(a) are they partly enforceable;

(b) if partly enforceable, to what extent are the Exclusion Notices

enforceable

14.3 If the Exclusion Notices are avoided or set aside:

(a) can and should the Court grant relief on terms, so that the Exclusion Notices are treated as taking effect at some date after 6 April 2011 and as terminating such pensionable service at that date,

(b) and if so what is that date;

(c) what, if any, other terms should apply to the setting aside;

14.4 In the light of the answers to 14.1 to 14.3 above, if the Exclusion Notices are

treated as so taking effect, will the excluded members be entitled to Hybrid

M Plan membership and Hybrid Deferred status as from the date of termination of such pensionable service (subject to any necessary consents being given by the Trustee and subject to any subsequent events which would have caused cessation of Hybrid Deferred status)?

14.5 If the Exclusion Notices are wholly void, wholly unenforceable or are set

aside in their entirety (and not given effect on terms), are Affected Members to be treated

(a) as having continued in pensionable service under the DB Plans

after 6 April 2011 and

(b) as still being in such pensionable service at present (subject to anyevents which would have caused a cessation of pensionable service apart from service of the Exclusion Notices);

15. If the Exclusion Notices have not yet effectively terminated pensionable service under the DB Plans or the Exclusion Notices are in the future avoided or set aside:

15.1 is IBM's power to terminate such future pensionable service now inhibited

by the Reasonable Expectations engendered by Project Ocean and Project Soto (and if so to what date);

15.2 if IBM wishes to effect such termination by excluding members from the DB

Plans, would it have to serve new exclusion notices;

15.3 would the members be entitled to Hybrid M Plan membership and Hybrid

Deferred status as from the date of such termination (subject to any necessary

consents being given by the Trustee); and in particular, if IBM were required

to and did serve new exclusion notices in order to terminate pensionable service under the DB Plans, would it be obliged to offer all (or any, and if so what) elements of Hybrid Deferred status to the members in question (subject to any necessary consents being given by the Trustee);

15.4 must IBM carry out a further consultation before such termination?

16. Having regard to the answers to Issues 13 to 15, are members in principle entitled to damages or equitable compensation from IBM as a result of the service of the Exclusion Notices, or to some other remedy and if so what remedy?

17. If members are entitled to damages or equitable compensation from IBM, in principle what is the measure of such damages or compensation?

18. In the light of the answers to Issues 13 to 17, should the Trustee administer the DB Plans on the footing that (subject to any events which would have caused a cessation of pensionable service apart from service of the Exclusion Notices) the pensionable service of members under the DB Plans:

18.1 is still continuing;

18.2 has ceased as a result of service of the Exclusion Notices,

and if so on what date?

19 If the Exclusion Notices are void and/or unenforceable, or are avoided and set aside,

19.1 do (i) the Affected Members and (ii) those Postponed Retirees who were expressed to be subject to the Exclusion Notices have a right, pursuant to the Pension Decision Tool or otherwise, individually in the future to elect:

(a) to be treated as a member of his/her Chosen DC Arrangement in respect of benefits earned there under on and from 6 April 2011;

(b) to retain the benefit of any Step Down Contributions and Matched Contributions paid by IBM or by or on behalf of the member into the Chosen DC Arrangement; and

(c) to retain Hybrid Deferred status for his/her accrued benefits under the DB

Plans in respect of service up to 6 April 2011; instead of being treated as in pensionable service under the DB Plans on and from 6 April 2011?

19.2 [If those members are entitled to be treated as in pensionable service under

the DB Plans after 6 April 2011, and if they do not make the election mentioned in 19.1 (or if such right of election does not exist), are the payments already made by such members into their DC arrangements to be applied in the manner set out in paragraph 5.2 of The Pension Decision Tool.]

726. 20.1 are members who retired pursuant to the ER Window in principle entitled to

damages or equitable compensation or some other remedy (and if so what) from IBM if they can prove that, as a result of the ER Window and/or the impending introduction of the New ER Policy, they retired earlier than they would otherwise have done;

20.2 if so, in principle what is the measure of such damages or compensation;

20.3 if so, further or alternatively, (if such a member so elects) should the Trustee

administer the DB Plans on the basis that the member is entitled to any increased benefits under the DB Plan as a result of the implementation of the ER Window, and, if so, what increased benefits and subject to what (if any) conditions?

21. In relation to the change from the Old ER Policy to the New ER Policy on 6 April 2010, and any refusals or deemed refusals of consent to early retirement pursuant to that change:

21.1 Were the change in policy or such refusals or deemed refusals

21.1(a) void and/or

unenforceable

21.1(b) voidable and liable to be set aside;

21.1(c) valid?

21.2 If the change in policy or such refusals or deemed refusals are void and/or unenforceable, or are avoided and set aside:

21.2(a) If void, are they severable and partly valid in relation to pension attributable to pensionable service from 6 April 2005

21.2(b) If unenforceable, are they partly enforceable in relation to pension attributable to pensionable service from 6 April 2005

21.2(c) If avoided and set aside:

(i) can and should the Court grant relief on terms, so that they

are treated as taking effect in relation to pension attributable to

pensionable service from 6 April 2005;

(ii) if so, what, if any, other terms should apply to the setting

aside;

22. In the light of the answers to Issue 21 above:

22.1 should IBM have continued to apply the Old ER Policy on and after 6 April

2010 to Eligible ER Members (and if so, in relation to (a) all of their pension or (b) pension attributable to pensionable service prior to 6 April 2005)

22.2 from what date may IBM properly apply the New ER Policy (or a similar

replacement for the Old ER Policy) to Eligible ER Members,and what if any steps would IBM have to take (such as the giving of notice to members) to introduce such a policy;

22.3 until such time as the New ER Policy (or a similar replacement for the Old ER

Policy) may properly be applied, does the Old ER Policy continue to apply to Eligible ER Members (and if so, in relation to (a) all of their pension or (b) pension attributable to pensionable service prior to 6 April 2005)?

23. In respect of the termination of C Plan Over 60 Members' contracts of employment under the Separation Programmes (where those members were not expressly granted early retirement in accordance with Active ER Terms):

23.1 whether either or both of the methods of termination of the said contracts –

namely voluntary redundancy or involuntary redundancy – constituted "retirement" in accordance with the Active ER Terms?

23.2 if either or both of the said methods did not constitute retirement in accordance

with those terms, whether the Active ER Terms would not as a result apply in respect of such members? and

23.3 if either method did not constitute retirement in accordance with those terms

and the Active ER Terms did not as a result apply in respect of such members, whether the members whose contracts were terminated are entitled to any other remedy and if so what remedy?

24. In respect of the termination of Eligible ER Members' contracts of employment under the Separation Programmes:

24.1 whether either or both of the methods of termination of the said contracts –

namely voluntary redundancy or involuntary redundancy – constituted (i) a"retirement" and (ii) "with the consent of the Principal Employer" in accordance with Active ER Terms?

24.2 if either or both of the said methods did not constitute retirement with the consent of the Principal Employer in accordance with those terms, whether the Active ER Terms would not as a result apply in respect of such members? and

24.3 if either method did not constitute retirement with the consent of the Principal

Employer in accordance with those terms and the Active ER Terms did not as

a result apply in respect of such members, whether the members whose

contracts were terminated are entitled to any other remedy and if so what

remedy?

25. For any period from 6 April 2010 during which IBM should have applied the Old ER Policy and where:

an Eligible ER Member's employment has terminated; that termination is capable of being a "retirement with the consent of the Principal Employer" in accordance with Active ER Terms; and that Eligible ER Member did not receive a pension calculated on Active ER Terms

25.1 Where such Eligible ER Member became a Retiree of the DB Plans during this

period:

25.1(a) Is such person entitled to be treated as having received consent to early

retirement in respect of such part of his pension as was subject to the Old ER Policy?

25.1(b) If so, with effect from the date on which he became a Retiree, is such person entitled to an early retirement pension calculated in accordance with Active ER Terms, and to back-payments (with interest) in respectof any underpayment of pension since retirement?

(This is subject to Issue 28.)

25.2 Where such Eligible ER Member became a Deferred Retiree of the DB Plans

during this period:

25.2(a) Is it necessary for him to show that he would have taken an immediate early

retirement pension rather than a deferred pension had the Old ER Policy still been in force, or can this be inferred without further proof?

25.2(b) Subject to Issue 25.2(a), is such person entitled (if he so elects) to be treated as having received consent to early retirement in respect of such part of his pension as was subject to the Old ER Policy? If so, with effect from the date on which he became a Deferred Retiree, is such person entitled (if he so elects) to an early retirement pension calculated in accordance with Active ER Terms, and to back-payments

(with interest) in respect of any non-payment of pension since the date on which he became a Deferred Retiree.

(This is subject to Issue 28.)

26. For any period from 6 April 2010 during which IBM should have applied the Old ER Policy and/or during which the New ER Policy was purportedly applied on the (incorrect) basis that C Plan Over 60 Members required IBM's consent to retire on Active ER Terms, and where:

a C Plan Over 60 Member's employment has terminated;

that termination is capable of being a "retirement" in accordance with

Active ER Terms; and

that C Plan Over 60 Member did not receive a pension calculated on

Active ER Terms:

26.1 Where such C Plan Over 60 Member became a Retiree of the DB Plans during this period:

26.1(a) Is such person entitled to be treated as having exercised his right to take early retirement from active status as established in the Rectification Action?

26.1(b) If so (and/or by virtue of his Reasonable Expectation as to the continued application of the Old ER Policy as found in the Judgment), with effect from the date on which he became a Retiree, is such person entitled to an early retirement pension calculated in accordance with Active ER Terms, and to back-payments (with interest) in respect of any underpayment of pension since retirement?

(This is subject to Issue 28.)

26.2 Where such C Plan Over 60 Member became a Deferred Retiree of the DB Plans during this period:

26.2(a) Is it necessary for him to show that he would have taken an immediate early

retirement pension rather than a deferred pension had the Old ER Policy still been in force and/or had he known of his rights as established in the Rectification Action, or can this be inferred without further proof?

26.2(b) Subject to Issue 26.2(a), is such person entitled (if he so elects) to be treated as having exercised his right to take early retirement from active status as established in the Rectification Action? If so (and/or by virtue of his Reasonable Expectation as to the continued application of the Old ER Policy as found in the Judgment), with effect from the date on which he became a Deferred Retiree, is such person entitled (if he so elects) to an early retirement pension calculated in accordance with Active ER Terms, and to back-payments (with interest) in respect of any non-payment of pension since the date on which he became a Deferred Retiree.

(This is subject to Issue 28.)

27. Are any C Plan Over 60 Members or Eligible ER Members in principle entitled to damages or equitable compensation from IBM as a result of the application of the New ER Policy, or to some other remedy and if so what remedy?

In principle what is the measure of such damages or compensation?

(This is subject to Issue 28.)

28. In relation to the C Plan Over 60 Members or Eligible ER Members whose employment was terminated under one of the Separation Programmes, do the waivers purportedly entered into by such members on termination of employment by IBM:

28.1 disentitle the members from any of the remedies identified in Issues 23 to 27?

28.2 have some other, and if so what, effect on any remedies available?

29. Are any members in principle entitled to damages or equitable compensation from IBM as a result of the manner in which the Consultation was conducted, or to some other remedy and if so what remedy?

In principle what is the measure of such damages or compensation?

30. Where a Hybrid Deferred Member takes early retirement from active status, and it becomes necessary to calculate the amount of the underpin referred to in paragraph 223.3.3 of James Lamb's first witness statement dated 28 July 2010, is the underpin calculated by applying to the revalued amount of the member's accrued benefits either:

30.1 the actuarial reduction provided for in Schedule F rule 2(3)(b) of the Main Plan

Defined Benefit Rules, or Schedule I rule 2(3)(b) of the I Plan Rules, in relation to the period between the member's retirement and his/her Normal Retirement Date; or

30.2 the early retirement discount factors prescribed in Schedule D rule 2 or (if applicable) rule 3 of the Main Plan Defined Benefit Rules, or Schedule G rule 2 or (if applicable) rule 3 of the I Plan Rules?

ANNEX B – a summary of the Separation Programmes

Phoenix

The Phoenix programme was carried out between July and August 2012 (ie prior to the judgment in the Rectification Action being issued). It provided in summary as follows:

(i) No employees under the NRA of their respective DB plans whose employment terminated via redundancy (both voluntary and involuntary) were treated as having retired from active service; they were all treated as deferred members of their plans and, if over minimum pension age, were able to draw their pension from deferred status. At this stage, IBM proceeded on the basis that:

(a) no such redundancy could constitute "retirement" within the early retirement terms of the DB Plans; and in any event;

(b) consent was not granted to retire early from active service.

(ii) As regards C Plan Over 60 Members and Eligible ER Members aged between 50 and the relevant NRA of their respective plans:

(a) those who accepted an offer of voluntary redundancy were treated as deferred members of their DB Plan and entered into a pension waiver signed by the member, and countersigned by his or her legal adviser;

(b) those who were made redundant under the compulsory redundancy programme all signed a settlement agreement on the terms set out in an offer letter, settlement agreement and pension waiver.

Mercury

The Mercury programme was carried out between May and July 2013, after judgment

in the Rectification Action had been issued. For the purposes of this programme (and the subsequent Apollo programme), Holdings proceeded on the basis that:

(i) C Plan Over 60 Members did not require the consent of Holdings to retire early (following the decision in the Rectification Action);

(ii) all affected members who were made voluntarily redundant could be treated as having "retired" under the terms of the DB Plans;

(iii) Eligible ER Members who were made voluntarily redundant would not be treated as having retired "with the consent of the Principal Employer" for the purposes of the early retirement terms of the DB Plans;

(iv) any members who were made compulsorily redundant would not be treated as having "retired" under the terms of the DB Plans;

(v) in any event, no consent was given to Eligible ER Members to retire early from active status where they were made compulsorily redundant.

Thus, under the Mercury programme:

(i) C Plan Over 60 Members who took voluntary redundancy were entitled to receive an immediate unreduced pension and did not enter into a pension waiver;

(ii) Eligible ER Members who took voluntary redundancy were treated as deferred members of their relevant DB Plan and entered into a pension waiver signed by the members, and countersigned by their legal advisers, in terms which are identical in all material respects to the waiver for the Phoenix.

(iii) all members who were made compulsorily redundant were treated as deferred members of their DB Plans; if such members wished to draw their pensions before their NRA, they would have had their benefits reduced accordingly;

(iv) the members who were made involuntarily redundant and who signed compromise agreements, also signed pension waivers which were tailored to their specific pension circumstances:

(a) C Plan and N Plan members who were aged between 50 and 59 entered into a pension waiver signed by the members, and countersigned by their legal advisers, in the similar terms to the waivers already mentioned;

(b) N Plan members who, by reason of equalisation requirements, had the right to benefits based on an NRA of 60 in respect of part of their service and the right to benefits based on an NRA of 65 in respect of the remainder of their service ("Split NRA Members"), entered into a pension waiver signed by the members, and countersigned by their legal advisers, in identical terms to the previous waiver. The waiver document explained that an early retirement pension taken between 60 and 65 would be paid without reduction in respect of the NRA 60 element of service, whereas the benefit in respect of the NRA 65 element would be reduced.

(c) C Plan Over 60 Members entered into pension waivers signed by the members, and countersigned by their legal advisers, gain in similar terms save that there is no waiver of any claim that IBM is thereby consenting to early retirement from active service, it being clear by this time that, following the Rectification Action, there was no such requirement for C Plan Over 60 Members.

Further, the relevant FAQ document {C2/11.2.49/1224} explained that those members had the choice to retire under the active early retirement terms and receive an immediate unreduced pension, or leave under the involuntary programme and become a deferred member of the plan, which would attract reductions to any early retirement pension paid before reaching age 63.

(v) Pension waivers were in all cases signed only in conjunction with a compromise agreement.

Apollo

The Apollo programme was carried out between January and March 2014. This programme essentially mirrored the Mercury programme, and in particular it proceeded on the same bases as the Mercury programme as set out above.

(i) C Plan Over 60 Members who took voluntary redundancy could retire on an unreduced pension; and did not sign a pension waiver;

(ii) Eligible ER Members became deferred members of their Plan on leaving service under the voluntary programme, and were required to enter into pension waivers in the same terms as under the Mercury programme; thus:

(a) Split NRA Members who were over age 60 entered into pension waivers signed by the Members, and countersigned by their legal advisers.

(b) C Plan and N Plan members who were aged between 50 and 59 entered into pension waivers {C2/11.3.49/1537} signed by the members, and countersigned by their legal advisers.

(iii) all members who were made involuntarily redundant were treated as deferred members of their DB Plans. As a consequence all such Eligible ER Members and C Plan Over 60 Members who wished to draw their deferred benefit before reaching their NRA would receive reduced benefits accordingly;

(iv) the different categories of members who were made involuntarily redundant and who signed compromise agreements, entered into pension waivers in the same terms as the corresponding members under the Mercury programme.