Rolls-Royce Holdings Plc v Goodrich Corporation & Ors

The Claimant (RR Holdings) is the holding company of the Rolls-Royce group of companies (the Rolls-Royce Group), which group is engaged, inter alia, in the design, development, manufacture and sale of aircraft engines. The Third to Tenth Parties (the RR Entities) are other companies in the Rolls-Royce Group.

The Defendant (Goodrich) is part of a group of companies which supplies engine control products and aftermarket services in relation to Rolls-Royce aeroengines, and which is ultimately owned by Raytheon Technologies Corporation (RTC).

On 31 December 2008, the two groups of companies formed a joint venture company (the Tenth Party or JVC). Various documents were entered into in relation to the joint venture, including:

The issues which arise at this trial are:

The subject-matters of the PCOA and Exclusivity Obligation and Other ECSURS Issues are substantially different, and they are addressed separately and sequentially in this judgment.

On 31 December 2008, a number of agreements were entered into between RR Group, then the holding company of the Rolls-Royce Group, and Goodrich in connection with the formation of the JVC. These included a joint venture agreement (JVA) and the PCOA.

Clauses 6 to 9 of the JVA provided for circumstances in which Goodrich could require RR Group to buy its shares in the JVC, or RR Group could exercise a right to buy those shares. This included when there was a change of control in Goodrich (clauses 7.5 to 7.8).

Where RR Group exercised its right under the JVA to acquire Goodrich’s shares in the JVC, it also became entitled under the PCOA to require Goodrich to sell it the AM Package – the Call Option. The option was only exercisable by serving a Call Option Notice during the Call Option Period which was defined, in relevant respect, as the period “commencing on the date on which [RR Group] has served on [Goodrich] a Buyout Notice” exercising its right to acquire Goodrich’s shares in the JVC following a Change of Control of Goodrich, and terminating “on the earlier of: (i) the withdrawal or lapse of the Call Option in respect of such Buyout Notice; and (ii) the first anniversary of the date of such Buyout Notice”.

If the Call Option was exercised during the Call Option Period, then Goodrich was obliged to provide a package of information known as the Call Option Exercise Preliminary Information Documents to RR Group, who then had 40 days from the receipt of those documents to notify Goodrich whether it wished to withdraw the Call Option Notice. If it did not do so, then the parties were obliged to effect the sale of the AM Package from Goodrich to RR Group on a pre-agreed form of contract annexed to the PCOA, known as the Combined Asset and Share Purchase Agreement, or CASPA, for a price to be determined in accordance with a mechanism set out in the PCOA and the CASPA.

Clause 15 of the PCOA provided that neither party could “assign, transfer, charge or otherwise deal with all or any of its rights or obligations … under this Agreement …. without the prior written consent of the other party”, but that “if the proposed dealing is an assignment and if the proposed assignee is a party’s Affiliate, such prior written consent shall not be unreasonably withheld or delayed”. It also provided that “any such purported assignment, transfer, charge or other dealing without the prior written consent of the other party shall be null and void”.

Clause 16 provided that “any release, delay or waiver by any party in favour of the other party of any … [of] its rights under this Agreement shall only be binding if it is given in writing”.

Clause 17 provided that “a variation of this Agreement (or of any of the documents referred to in it) is valid only if it is in writing and signed on behalf of [RR Group] and [Goodrich]”.

On 23 May 2011, a scheme of arrangement came into effect which replaced RR Group with RR Holdings as the holding company of the RR Group. The scheme was effected by a share transfer by which shareholders in RR Group received one share in RR Holdings for each share they held in RR Group, with RR Group becoming a wholly owned subsidiary of RR Holdings. The existing directors of RR Group, with the exception of Sir John Rose who retired, became directors of RR Holdings.

On 30 November 2011, a draft of a proposed amendment to the JVC and the PCOA was sent by Adrian Thompson of Goodrich to Gareth Hopkinson of Rolls-Royce, which was intended to address an issue which had arisen as a result of a proposed relocation of the base of operations of the AM Package, Mr Thompson stating “we look forward to receiving your comments/confirmation that it is acceptable to R-R”. The draft made no reference to RR Holdings. Mr Hopkinson responded with a mark-up on 1 December which introduced references to RR Holdings, and described the amendments as having been made under clause 26 of the JVA and clause 17 of the PCOA. The mark-up described the parties to the proposed amendments as including RR Group and RR Holdings, and included a signature block for RR Holdings as well as RR Group. Mr Thompson replied stating “the proposed edits to the ‘parties’ clause and ‘whereas’ section are all fine (assuming that [RR Holdings] is a party to the JVA by the time this amendment is signed)”, and he made the same comments about the amendments to the signature block.

On 31 December 2011, a Deed of Adherence was signed referring to the scheme of arrangement, confirming that RR Holdings was entitled to the benefit of, and obliged to observe, covenant, and be bound by, the terms of the JVA “as if each reference to RR Group in the Agreement was a reference to RR Holdings”. On the same date, “Amendment No 1” was signed on terms which included the references to RR Holdings introduced by Mr Hopkinson by both RR Group and RR Holdings (the same individual signing for both companies).

In September 2011, United Technologies Group, now known as Raytheon Technologies Corporation or RTC, entered into an “Agreement and Plan of Merger” to acquire Goodrich. The proposed acquisition raised issues for the Rolls-Royce Group, both in relation to the JVC and in relation to the operation of the AM Package by Goodrich, and also competition issues in the USA because RTC owned Pratt & Whitney, a major manufacturer of aeroengines. The RR Group raised their concerns with the US Department of Justice (DOJ) who brought proceedings against Goodrich and RTC intended to address competition concerns. In the event, following negotiations between all interested parties, those concerns were addressed by twin-tracks of activity:

Thus, on 7 June 2012:

Against that background, the PFJ was issued on 26 July 2012 which stated in its recitals that:

Section VI of the PFJ addressed the JVC:

On 13 December 2012, in a letter to Rolls-Royce, Raytheon offered the Right to Purchase (the December 2012 RTP) provided for in the PFJ. On 29 May 2013, the terms of the PJF were made final (the Final Judgment).

A further Letter Agreement (the February 2014 Letter Agreement) was entered into between RR Holdings and Goodrich on 3 February 2014. This once again described the PCOA as an agreement between RR Holdings and Goodrich, and extended the Call Option Period to 26 July 2017. There was a further extension, in similar terms on 28 June 2017 (the June 2017 Letter Agreement), the Call Option Period being extended on this occasion to 31 December 2017. The June 2017 Letter Agreement required the parties (i.e., RR Holdings and Goodrich) to negotiate amendments to the PCOA and to the CASPA, so that it applied to the December 2012 RTP.

On 21 December 2017, RR Holdings and Goodrich entered into two further agreements:

The RTP Agreement stated in its recitals that RR Holdings had the right to acquire the AM Package under the PCOA (which was defined as an agreement between RR Holdings and Goodrich). It referred to the Call Option arising under the PCOA as the R-R Option, and provided that the parties “now wish the R-R Option to automatically expire” on the Unconditional Date such that:

The Unconditional Date was defined as “the first Business Day immediately following the satisfaction of the Condition or determination that no Condition is required” (clause 2.5). Clause 2.2 provided:

That right (the Modified RTP) was exercisable during a period (the RTP Period) which ran between 1 January 2020 and December 2023, and 12 months’ notice had to be given, such that the earliest the Modified RTP could take effect was at the start of 2021. If the Modified RTP was exercised, it was to be on the terms of the modified CASPA which accompanied the RTP Agreement. Clause 2.6 provided:

The timing position was, therefore:

Turning to the December 2017 Letter Agreement:

It is clear that the reference to “Rolls-Royce” in the December 2017 Letter Agreement was not intended to make Rolls-Royce plc a relevant party for the purposes of the PCOA. The December 2017 Letter Agreement clearly provides that the PCOA is an agreement between RR Holdings and Goodrich and, read in context, the subsequent references to “Rolls-Royce” are not intended to be read as references to Rolls-Royce plc, but to RR Holdings.

By further agreements dated 21 June and 19 September 2018, RR Holdings and Goodrich agreed to extend the Longstop Date, initially by 3 months and then to 21 October 2018.

On 9 October 2018, RR Holdings purported to serve a Call Option Notice (which it had issued the preceding day) on Goodrich. Goodrich disputes the validity of the Call Option Notice, and accordingly has not delivered the Call Option Exercise Preliminary Information Documents.

Goodrich challenges the validity of the Call Option Notice on three grounds.

First, Goodrich contends that, under the terms of the PCOA, only RR Group was entitled to serve the Call Option Notice, and that the Call Option Notice served by RR Holdings was invalid. In particular, it is said that the contractual requirements for transfer of the right of call from RR Group to RR Holdings were not satisfied.

RR Holdings has advanced two responses to that argument:

It is agreed that the court can conclude from the parties’ conduct that there has been a novation, the test being “whether that inference is necessary … to provide a lawful explanation or basis for the parties’ conduct” (Chitty on Contracts (34^th), [22-089]; Evans v SMG Television Limited [2003] EWHC 1423, [181]). The test is objective, and does not depend on the subjective intention of the parties (ibid, [186]). I was also referred to the decision of the Court of Appeal in Musst Holdings Limited v Astra Asset Management UK Limited [2023] EWCA Civ 128, [57] to similar effect.

In this case, it is perfectly clear that, from at least the time of the June 2012 Letter Agreement, RR Holdings and Goodrich have conducted themselves on the basis that it is RR Holdings, not RR Group, which is the party to the PCOA, and that the basis of their dealings cannot be otherwise explained. Not only have RR Holdings and Goodrich entered into agreements which describe the PCOA as an agreement between RR Holdings and Goodrich, but they have purported, by agreements between them to which RR Group is not expressly a party, to amend the terms of the PCOA, which amendments it is agreed were valid and effective. These steps were not taken by Goodrich in ignorance of the scheme of arrangement, because this was expressly referred to in the Deed of Adherence for the JVA of 31 December 2011. Further, if there was no novation, then Goodrich had granted rights to purchase to both RR Holdings (through the December 2012 RTP) and RR Group (through the PCOA), which would have placed it in a position of some difficulty had they both been exercised.

Goodrich submits, nonetheless, that there is no novation.

First, it says that there is no evidence that RR Group consented to the novation. As to this:

Finally, it is noteworthy that on Goodrich’s own case, the various Letter Agreements between RR Holdings and Goodrich were effective to vary the terms of the PCOA. If RR Holdings never became a party to the PCOA, and RR Group remained the only party entitled to exercise the Call Option, then this must be because RR Holdings was authorised to act on RR Group’s behalf. However, if RR Holdings had such authority, then its actions can be attributed to RR Group for the purposes of determining whether RR Group consented to the novation. I understood Mr Croall KC to accept this by the end of the closing submissions.

Second, Goodrich suggests, in reliance on MWB Business Exchange Centres Ltd v Rock Advertising [2019] AC 119, that there can be no novation otherwise than in writing because:

In support of its argument, Goodrich also relies on Kabab-Ji SAL v Kout Food Group [2021] UKSC 48. In that case, the claimant (C) alleged a form of novation by conduct in which the contract (the FDA) with the original counterparty, H, had been terminated, and replaced by a contract between the C, H and the defendant (D). Articles 3 and 19 of the FDA required H’s consent for any assignment or transfer, Article 24 provided that no termination or waiver would be valid unless signed in writing by both C and H, and Article 26 provided that the FDA could only be amended by a written document signed by both parties. The Supreme Court held that any termination of the original contract with H was ineffective because the requirement for writing signed by C and H was not satisfied ([64]), and that any new agreement between C and D would only be effective if C and H signed a document approving the variation of the FDA as required by clauses 17, 24 and 26 ([65]).

Some caution may be required in applying the reasoning in MWB as to the effect of a contractual formalities requirement as between the parties to a contract to a purported novation, because the formalities provision does not bind the proposed new party unless and until it has become a party (by which point the novation has taken place). In Kabab-Ji, the Supreme Court stated that D “would be entitled to assume that it could not become a party to [the FDA] with the claimant unless the requirements of the No Oral Modification clauses contained in the FDA were satisfied” ([68]). Kabab-Ji was a case in which the purported transferee disputed that it had become a party to the relevant agreement by novation, and the various No Oral Modification clauses were relied upon not as a legal bar to the conclusion of a contract between C and D through conduct, but as an evidential factor weighing against the conclusion of such a contract with the alleged new party.

In this case, there is a written agreement between the continuing party and purported transferee, but it is said that this cannot take effect because of No Oral Modification clauses which only have contractual effect as between the transferor and the continuing party. Assuming that it is tenable, in these circumstances, for the continuing party to invoke the formality requirement against the transferor but not the transferee, it would seem to be a better reason for denying the efficacy of any discharge of the transferor from its contractual obligations, rather than the agreement which the remaining party has entered into with the transferee in writing to accord it rights under the contract, in return for assuming obligations under it. These difficulties suggest that in a tripartite context such as the present, it may be easier to establish waiver of a formality requirement than in a purely bilateral context.

In this case, the statements in the various Letter Agreements to the effect that the PCOA is an agreement between RR Holdings and Goodrich meet any requirement for written consent by Goodrich to the transfer of rights under the PCOA to RR Holdings. In relation to the various technical objections to that contention:

Alternatively, any requirement for prior written consent has been retrospectively waived by Goodrich and RR Group. In Musst Holdings Limited v Astra Asset Management UK Limited [2023] EWCA Civ 128 the trial judge had found a novation by conduct notwithstanding clauses in the novated contract requiring prior written consent for any novation. At [83] and [86], Falk LJ in the Court of Appeal held that it was clearly open to the continuing party to waive the requirement for prior written consent, including retrospectively. I would note, given the focus in Goodrich’s submissions as to the position of RR Group, that the factual findings which led to that conclusion in Musst (and a similar conclusion on estoppel by convention) all concern dealings between the continuing party and the putative transferee, rather than the “exiting party”. In so far as clause 16.1 requires the consent of Goodrich and RR Group to be in writing, (a) the June 2012 Agreement provides such written consent for both (see [38] as to the position of RR Group) or (b) any requirement for the waiver to be in writing has itself been waived (RGI International Limited v Synergy Classic [2011] EWHC 3417 (Comm), [51(iii)]). As Mr Toledano KC put the matter:

It is also said that RR Group did not execute the Guarantee it had promised to provide under clause 15.3(B). Leaving aside the difficulty of invoking this point as an answer to the contract which Goodrich entered into with RR Holdings (the party seeking to exercise the Call Option), there was clearly a waiver of that requirement. To the extent that it is said that such a waiver is not valid because it was not in writing (a) clause 15.3(B) did not preclude RR Holdings and Goodrich entering into the June 2012 Letter Agreement, which is sufficient for RR Holdings’ purposes and (b) that formalities requirement has been waived in writing as between Goodrich and RR Holdings, through the June 2012 Letter Agreement (see [44] above).

In these circumstances, it is not strictly necessary to address RR Holdings’ alternative argument that Goodrich is contractually estopped from denying that it is entitled to serve the Call Option Notice, and I can therefore deal with the issue briefly.

The existence of a doctrine of “contractual estoppel” – by which one party makes a promise that the contract is to be approached on the basis that a particular state of affairs prevails, whether or not that is in fact the case – is well-established: see for example Peekay Intermark Ltd v Australia and New Zealand Banking Group Ltd [2006] 1 CLC 582 and Springwell Navigation Corp v JP Morgan Chase Bank [2010] 2 CLC 205. Notwithstanding its name, the doctrine is not a species of estoppel properly so-called, but simply a case of the court holding a party to its promise through a form of specific enforcement of the primary obligation assumed by requiring litigation to be conducted on the promised basis (rather than giving some other remedy for breach of that promise). Reflecting the fact that the doctrine is, at heart, about the enforcement of promises rather than protecting reasonable reliance, there is no requirement of unconscionability before the doctrine can be invoked (Springwell, [177]).

In this case, RR Holdings and Goodrich repeatedly contracted through the Letter Agreements on the basis that RR Holdings was a party to the PCOA and entitled to exercise the Call Option arising thereunder. Goodrich contends that the doctrine of contractual estoppel does not apply to statements made in recitals, which engage a different set of rules including the following:

Ultimately, the application of the doctrine of contractual estoppel will turn on whether the provision in issue is intended to constitute a promise by the parties that the particular state of affairs addressed by the provision is to be taken as established, whether that promise appears in a recital or some other provision. In Richards v Wood [2014] EWCA Civ 327, [15], Lewison LJ applied the doctrine to a statement in a declaration of trust as to the contributions made by the parties, stating “[a]ll parties are bound by [the] agreed statement, whether it represented the truth or not. This is the traditional function of recitals which in this respect are part of a sub-species of estoppel known as contractual estoppel”. In this case, the provisions relied upon by RR Holdings are essentially promissory, appearing in a contract whose operative provisions proceed on an assumed state of affairs which is the subject of the contractual estoppel.

However, assuming that the law is as Goodrich contends it to be:

That leads to the complication that the doctrine of contractual estoppel primarily operates in a bilateral context, as between the parties to the relevant contract, whereas the issue of whether an agreement originally entered into between A and B should be treated as an agreement between C and B has implications for all three parties. If Goodrich is contractually estopped as against RR Holdings from denying that RR Holdings rather than RR Group is the party entitled to exercise the Call Option, that might raise the issue of whether RR Group could continue to assert such a right, and, if so, how the conflict between the competing rights is to be resolved. On the facts of this case, that difficulty does not arise because RR Group has not obtained any extension of time for the exercise of any option it might have had. In any event, if Goodrich had placed itself in a position in which it might face conflicting claims to the same assets because it has made competing promises about them, that is its lookout, as would be the case if Goodrich had entered into two separate call options. No doubt, there might be circumstances in which Goodrich could invoke a “standing by” estoppel against RR Group under the principle recognised in Moorgate Mercantile Co Ltd v Twitchings [1976] QB 225.

Mr Croall KC alleged that there was some policy limit to the scope of the doctrine of contractual estoppel where it was capable of impacting on the rights of third parties (e.g., A and B contracting on the basis of a promise that A was the party with a call right in relation to certain property when it was in fact C). However, there is no such limit. As Lord Toulson noted in Prime Sight Limited v Lavarello [2013] UKPC 22, [47]:

Finally, Goodrich said that invoking contractual estoppel in this case involved using the doctrine as a sword rather than a shield. However, the language of estoppel in this context is apt to mislead. Goodrich and RR Holdings are parties to a legally binding contract in which a promise has been made as to the basis on which they are contracting. Enforcing the promise in that contract as between the contracting parties raises no difficulties – contractual promises, by their nature, forge swords (cf. Keiry CF Loi, “Contractual Estoppel and Non-reliance Clauses” [2015] LMCLQ 346, 349). Even if the sword/shield debate had been relevant, it is clear from Musst that a conventional estoppel – in that case estoppel by convention – can prevent a party from denying it owes obligations to the transferee in the context of a disputed novation, this being a context in which identifying consideration for the legal obligations asserted presents little difficulty (c.f. Tinkler v Revenue and Customs Commissioners [2022] AC 886, [76]).

Goodrich’s second defence concerns the effect of the December 2017 Letter Agreement, and whether it extended the period during which the Call Option could be exercised to the Unconditional Date or the Longstop Date, or whether it simply extended the Call Option Period, so as to continue obligations which only operated during that period, but on the basis that RR Holdings could not in fact exercise the Call Option.

There are points to be made on both sides of this argument, but I have concluded that the December 2017 Letter Agreement extended the period within which the Call Option could be exercised:

So far as the points made the other way are concerned:

Goodrich also argues that it and RR Holdings had determined prior to the date of the Call Option Notice that Governmental approval was required but could not be obtained before the Longstop Date so that the Call Option Period had expired at the date of that determination by virtue of clause 1(b) of the December 2017 Letter Agreement. By way of “recap”, this provided:

A threshold requirement for this argument to succeed is that clause 1(b) should be interpreted to have the effect that the December 2017 Letter Agreement will immediately lapse once (a) the parties determine that Governmental Approval is required and (b) the parties determine that Governmental Approval will not be obtained before the Longstop Date. On this construction, the December 2017 Letter Agreement and the Call Option might lapse significantly in advance of the Longstop Date.

I am satisfied that Goodrich is unable to overcome this threshold:

As to the points made the other way, Goodrich submits that the reference to the parties being unable to obtain Governmental Approval “on or before the Longstop Date” reflects the fact that the December 2017 Agreement could expire before the Longstop Date. However, those words simply reflect the fact that Governmental Approval could be obtained up to and including the Longstop Date. I would note that the expression “on or before” is one used very frequently in the parties’ contracts when imposing a deadline of some kind, and I am not persuaded that a “drafting tic” of this kind can bear the weight which Goodrich’s argument seeks to place on it.

If this issue had been resolved in Goodrich’s favour, then a further issue of construction would have arisen: what must happen for the parties to have “determine[d]” that Governmental Approval would not be obtained before the Longstop Date, and is it sufficient for each to have independently reached that view without the shared view “crossing the line”, in the sense that the parties manifest (coinciding) views to each other?

On the hypothesis under consideration, the parties must determine two things: that Governmental Approval is required, and that it will not be obtained by the Longstop Date. The question of what constitutes a determination must have the same answer in both contexts, and also for the purposes of clause 1(a) when the parties “determine no Governmental Approval is required” such that the Modified RTP becomes unconditional. In this last context, I am satisfied that much the more likely construction is that a determination requires a mutually shared position: given that the product of such a determination is that the Modified RTP immediately becomes unconditional, it is necessary that the parties know this is the case. Were matters otherwise, the Modified RTP might have in fact become unconditional without either party appreciating the position. In relation to clause 1(b), the effect of the two determinations will be that the December 2017 Letter Agreement expires automatically. Once again, it seems to me obvious that the parties must have contemplated that this would only happen when there was a mutual manifestation by the parties of the relevant determinations.

Further, clauses 2.3 and 2.4 of the RTP Agreement envisage that obtaining satisfaction of the condition will be a joint exercise, with an obligation on each party to share information with the other (clauses 2.4 and 2.5). That envisages a process in which the parties will make their positions known to each other.

That conclusion is also fatal to the No Governmental Approval Defence. Goodrich does not contend that any communication of its determination in relation to the issue of whether Governmental Approval would be obtained before the Longstop Date “crossed the line” prior to the exercise of the Call Option. Goodrich does suggest that RR Holdings’ position was manifest after the Call Option had been sent by RR Holdings on 8 October 2018, but before it had become effective through receipt at Goodrich on 10 October 2018 (the terms of the PCOA requiring service at Goodrich’s UK address, with the notice only being effective on receipt). While that would not be enough on my analysis, it is, in any event, a thoroughly unattractive argument, and one which is not consistent with the admission in Goodrich’s defence that the Call Option Notice was served on 9 October. In those circumstances, I make no apologies for looking at the terms and circumstances of the email of 9 October with some care. The email was not sent to Goodrich, but to Tim White of RTC, whose authority to receive communications on Goodrich’s part was not addressed at trial. It records the fact that Governmental Approval had not been obtained when the email was sent, and that RR Holdings had decided to call the business, but no more. It expresses no view on whether Governmental Approval will be obtained on or before 21 October 2018. Had a unilateral communication of its determination by RR Holdings been sufficient, I am not persuaded the 9 October email was sufficient to this end.

In relation to this issue, I am satisfied that each of RR Holdings and Goodrich had formed the view, with a sufficient level of confidence to constitute a determination for the purposes of clause 1(b), that Governmental Approval was required and that it would not be obtained before the Longstop Date prior to exercising the Call Option. This argument must be approached on the basis that Goodrich has succeeded in the two preceding construction arguments.

In my assessment, a decision by both parties that they would not proceed without Governmental Approval amounts, as a matter of practicality, to a determination that Governmental Approval was required for the RTP Agreement. I do not understand there to be any dispute that both parties had reached that view. In any event:

The evidence also establishes a decision by both parties that they should act on the basis that approval would not be forthcoming in time, because this was the likely outcome. I am satisfied that this amounts to a determination that Governmental Approval would not be obtained in time. That is how commercial parties would be expected to reach a determination as to future events.

As to the evidence:

I am satisfied, therefore, that prior to the exercise of the Call Option, RR Holdings had concluded that DOJ approval would not be forthcoming by the Longstop Date, and exercised the Call Option for that reason. However, that illustrates the lack of commercial sense in Goodrich’s construction – a determination to exercise the Call Option because DOJ approval was unlikely to be forthcoming before the Longstop Date would, on its case, extinguish the Call Option. As I suggested to Mr Croall KC in the course of argument, this was akin to a parachute which would self-destruct on pulling the ripcord.

It follows that RR Holdings are entitled to the declarations they seek, and an order requiring Goodrich to deliver the Call Option Exercise Preliminary Information Documents to RR Holdings.

This part of the case concerns alleged breaches of the ASA and ESCURS.

Taking the ASA first, by clause 2.1 (in combination with clauses 2.2, 2.3, 2.6 and 2.7) the RR Entities appointed Goodrich as “its Exclusive supplier of Aftermarket Services”.

Clause 2.2 of the ECSURS provided:

“Aftermarket Services” were defined for this purpose as:

As to the meta-definitions:

The ECSURS contained further relevant terms:

As the name suggests, Aftermarket Services are, broadly, services provided to customers in the period after the aircraft engine has been delivered to the customer.

Customers who acquire Rolls-Royce engines are not obliged to enter into aftermarket service contracts with Rolls-Royce, although many of them do, subscribing to one of the various levels of Rolls-Royce’s TotalCare aftermarket services offering. Those who do not contract to receive Aftermarket Services from the RR Entities may not enter into such contract at all, or they may do so with another service-provider, and both categories are referred to as “time and materials customers”.

The basic TotalCare package only covers repair and maintenance which is carried out after the engine has been removed to a workshop, and not repair and maintenance carried out “on wing”. It is possible to purchase additions to the package to cover “on wing” repairs, in the form of line replacement unit maintenance services (LRU Management Services) and maintenance of line replacement parts (LRP Management Services). There are two LRU Management Services options within the TotalCare offering. The first involves repair, or a replacement unit, within a specified turnaround time (Service Level 1) and the second an exchange service, under which an exchange unit is provided for the existing LRU within a shorter specified turnaround time (Service Level 2). Rolls-Royce in turn contracts with third parties to enable it to deliver LRP and LRU Management Services. The engine control system units supplied by Goodrich were line replacement units, and in return for supporting its LRU Management Services offerings, Rolls-Royce paid Goodrich a vendor support payment or VSP. Both the amount paid by the customer for LRU Management Services to Rolls-Royce, and the amount paid by Rolls-Royce to Goodrich by way of VSPs, were calculated by reference to the number of engine flying hours, referred to as an EFH basis.

It was a condition of acquiring LRU and LRP Management Services from Rolls-Royce that customers had access to spare units and parts, which could be used to repair or replace units or parts “on wing”. The recommended level of such units and parts was known as “Initial Provisioning”. The recommended level appears in a Recommended Spare Parts List. The recommended levels reflect various parameters, including customer-specific factors. However, on the evidence before me, it was not made a legal requirement of the provision of LRU and LRP Management Services that customers carried the recommended levels of Initial Provisioning. Customers can acquire their Initial Provisioning from Rolls-Royce or another supplier, although the effect of the evidence was that many customers chose, for commercial reasons, to acquire their Initial Provisioning from Rolls-Royce.

Customers proved increasingly resistant to carrying the level of Initial Provisioning recommended by Rolls-Royce, which could involve a significant upfront commitment. However, LRU and LRP Maintenance Services were an important revenue stream for Rolls-Royce. To avoid the costs of Initial Provisioning deterring customers from signing up to Rolls-Royce’s LRU and LRP Management Services packages, Rolls-Royce sales personnel offered significant discounts off the price of Initial Provisioning supplies as part of the TotalCare package negotiations.

In addition, a number of Rolls-Royce’s competitors in the aftermarket services market had introduced offerings which involved giving customers access to pooled LRUs but without requiring the same level of upfront purchases, including AJ Walters, Air France-KLM, SR Technics and Lufthansa Technik. These pooling services represented a significant commercial threat to Rolls-Royce’s aftermarket services business.

Against that background, Rolls-Royce had been considering the possibility of launching a parts availability service for some time. It is possible to find references to customer requests for a service of this kind in 2013 and 2014. In 2015, a limited number of customers – it would appear three operators – obtained access to such a pooling scheme. In 2016, plans to offer that scheme more widely came to fruition, Roll-Royce deciding to launch a new service offering known as the “Parts Availability Service” or PAS, with the service becoming generally available in 2017.

A number of internal presentations were delivered in 2016 which explained the rationale, and aspirations, for the PAS:

The PAS was made available to customers operating Trent 1000, Trent XWB or Trent 7000 engines and selected Service Level 2 customers paid for access to a pool of LRUs maintained for PAS customers. Rolls-Royce acquired the LRUs which were to be used within the PAS package, which were initially held in a Rolls-Royce storage location known as “1203”. From there, they could be drawn down to one of a network of Materials Service Centres (MSCs) which are operated on Rolls-Royce’s behalf by a third-party logistics service provider. Legal title in the LRUs in the MSCs is held by Rolls-Royce. There is no contractual or other legal limitation as to the uses to which LRUs held in MSCs can be put. I was pointed to one occasion in which LRUs from an MSC were used for two test flight engines, although it is not possible to determine whether this was in isolated example, or whether the use of LRUs in MSCs for purposes other than supporting the PAS was more widespread.

In addition to MSCs, PAS customers were required to have a certain quantity of LRUs stored in the customer’s onsite storage site (OSS), typically located at the customer’s main base. The customer managed its OSS and had legal title to the LRUs stored at its OSS. A proposal to retain title within a Rolls-Royce entity was rejected because of potential tax implications. However, the contractual terms relating to an operator’s use of the PAS were regarded by the RR Entities as having the effect that “in substance, Rolls-Royce is deemed to have economic control of the OSS and so should financially account for the OSS parts as inventory” (albeit the transfer of title was recognised to leave the RR Entities exposed in the event of the operator’s insolvency). As necessary, LRUs could be drawn down from the relevant MSC to “back-fill” an OSS when LRUs from the OSS had been deployed. I accept the evidence of Mr Loret that the initial stocking of OSS could come from Rolls-Royce directly rather than from an MSC, and, on the basis of the information assembled in Mr Bezant’s report, that 91% of the 128 AECSUs shipped to customers’ OSSs came from Rolls-Royce directly. In any event, the setting up of the PAS would inevitably have required the RR Entities to order AECSUs for the purpose of meeting the OSS requirement, whether delivered there directly (as appears to have been the case in the vast majority of instances) or via an MSC.

When an LRU on a PAS customer’s aircraft needed to be replaced, the customer used an equivalent unit in the OSS (which would then be “backfilled” by drawing an equivalent unit from an MSC), or if there was no such unit available, drawn directly from the MSC, with title in the replaced LRU passing to Rolls-Royce. Rolls-Royce then returned that unit to the relevant vendor – in the case of AECSUs, to Goodrich - who provided an exchange unit which was used to “back-fill” the MSC. Where it was economic to do so, the vendor repaired the returned unit.

Customers who subscribe to the PAS package pay an additional monthly fee, in addition to their EFH TotalCare rate and, in most cases, a commitment fee in respect of LRUs held at their OSS. The monthly fee is calculated at 0.95% of WLP/month for OSS units and 0.67% per month for MSC units, with discounts of up to 20% payable for some customers. The commitment fee appears to have been $250,000 in 2016 (representing 10% of the average value at WLP of the material at the OSS), to be escalated in accordance with an escalation formula in the TotalCare agreement. If the TotalCare package came to an end without the LRUs held at the OSS being used, those LRUs were returned to Rolls-Royce and the commitment fee returned.

By way of a reminder, the definition of Aftermarket Services is:

In what it confirmed in closing was its primary case, Goodrich contends that the PAS falls within the contractual definition of “Rework” and/or “the supply of Spare Parts for AECSUs”. As it was put in Goodrich’s written closing, “on any view the PAS involves ‘the supply of spare parts and initial provisioning’” (emphasis added). However, the effect of the definition cannot mean that any aftermarket service of which the supply of AECSU spare parts or rework on AECSUs forms part is exclusive to Goodrich. The PAS involved the supply of a range of spare part and rework services for the entire Rolls-Royce engine, albeit the evidence suggests that, by value, 47% of the parts in the OSSs and MSCs were sourced from Goodrich.

In considering this argument, it is important to note the following:

Further:

If it is not enough that the PAS “involved the supply of spare parts and initial provisioning”, then clause 2.1 provides an answer to this aspect of Goodrich’s exclusivity case. It provides:

I do not accept that Goodrich needs to supply the AECSUs direct to the customer, rather than to the RR Entities who supplies them to the customer, for this provision to apply. Provided that all Rework is done by Goodrich, or all AECSUs and Spare Parts supplied are sourced from Goodrich, the requirements of clause 2.1 are satisfied (the words “or otherwise” reflecting the fact that Goodrich need not assume the status of a sub-contractor for the clause to apply). In this regard, clauses 5.3, 5.5, 5.6 and 5.7 of the ECSURS envisage provision of Spare Parts “provided to R-R or a Customer” or collected “by R-R or the Customer”, with “R-R or the Customer” (as appropriate) acknowledging receipt.

In the alternative, Goodrich contends that the definition of Aftermarket Services should be interpreted as comprising three elements:

The R-R Entities, by contrast, argue that the definition has only two elements:

The considerations I have referred to at [91]-[92] above strongly support the RR Entities’ construction:

Further, the language of the definition lends some support to the R-R Entities’ construction:

The RR Entities’ construction also derives support from other terms of the ASA and the ECSURS:

Finally, the RR Entities’ construction is more consistent with the definition of Aftermarket Services in the PCOA, entered into as part of the same suite of agreements, which refers to:

This definition:

Goodrich’s arguments the other way largely assume what they seek to prove, in so far as they rely on the use of the term Aftermarket Services. As to the remaining points:

If the PAS (or at least those aspects of it relating to AECSUs) falls within the definition of Aftermarket Services, then the RR Entities contend that it is nonetheless permitted by clauses 11.10(c) and/or (d) of the ECSURS, the terms of which have been quoted at [76] above. The RR Entities say that this is the case because those supplies constituted Initial Provisioning. This argument is also crucial to the issue of what price was payable for those supplies under the ESCURS, and whether they attracted the pricing terms applicable to Initial Provisioning in clauses 11.10(c) and/or (d).

The contractual definition of Initial Provisioning is “Spares ordered by R-R or Operators in support of the operation of a particular Operator’s fleet”. The RR Entities’ primary argument is that it is enough that Spares are ordered which will “in certain circumstances” be transferred to an Operator as part of an EFH deal, such that all AECSUs acquired for the PAS constitute Initial Provisioning.

Both sides sought to support their interpretations of the concept of Initial Provisioning by reference to the evidence as to a particular witness’s understanding of the term – Mr Loret of the RR Entities and Mr Smith (a Business Development Director of Goodrich Control Systems) of Goodrich:

This material provides an unpromising basis for the suggestion that the term has a settled industry meaning, and I was not taken to any material other than the witness evidence in support of the suggestion that it did. I did not find the evidence of either Mr Loret or Mr Smith helpful when interpreting a defined term in a complex suite of legal agreements, and in answering the question of whether the PAS which RR Holdings introduced in 2016/2017 fell within the particular arrangements made for Initial Provisioning in the ASA and ECSURS in December 2008. In any event, the material relied upon by the RR Entities does not come close to establishing a generally understood meaning of the term initial provisioning to the effect contended for.

Further, documents internal to the RR Entities are not consistent with the alleged market understanding. While there are a few which refer to the pooled assets in the PAS as a form of initial provisioning, there are many more documents which contrast the PAS and initial provisioning:

Indeed, one of the documents I was taken to by the RR Entities which was said to be consistent with the alleged industry understanding (an email string of November 2016) referred to the fact that the PAS was being set-up, that “the plan is for R-R to build up various pools of parts to serve as Initial Provisioning” but presciently observed that:

The RR Entities’ response that “subjective understandings expressed [in] documents post-dating the contracts in issue are neither relevant or admissible” does not sit happily with its “industry understanding” case.

Further, the internal documents suggest that, at least as used within the Rolls-Royce Group, the concept of Initial Provisioning refers to Spare Parts acquisitions by customers at or around the time of entry into service (EIS) of an engine. That explains the term “initial” and is consistent with other material before me:

Mr Loret’s evidence was, in the main, to similar effect. He referred to the concept of initial provisioning when discussing the customer’s decision to purchase aftermarket services from the RR Entities, which are offered by the RR Entities “as part of its sale of new aircraft engines”, and he explained that the RR Entities calculate the amount of initial provisioning which the operator is recommended to carry “prior to entry into service of the fleet”. By way of a limited exception which acknowledges the general rule, he noted in footnote 16 that “customers may also be provided with an initial provisioning recommendation by Rolls-Royce, and customers may acquire further initial provisioning if they increase the number of engines within their fleet” (emphasis added), but he did not suggest that further spares acquired for the same engine over the life of the engine constituted initial provisioning. That also accords with that part of Mr Smith’s evidence, not the subject of cross-examination, which stated that initial provisioning is “first acquired at the time the operator’s fleet enters into service (or shortly prior to that time)”. It is also consistent with the definition of “Initial Provisioning” used in a different (and prior) agreement between the parties, the Product Support Package Agreement of October 2004, which defines “Initial Provisioning” as “the initial quantity of spare units that an Operator … requires in support of its fleet of engines”.

Consistent with that evidence as to timing, it is clear that there is a commercial link between the sale of aircraft engines or particular aftermarket service offerings by the RR Entities (and in particular LRU Management Services) and the sale of initial provisioning:

Against that background, I turn to the definition in the ESCURS:

Further, there are a number of places in the ASA and ECSURS which suggest that the term Initial Provisioning in the relevant transaction documentation contemplates something which has been or is being supplied to a particular Operator:

On RR Holdings’ construction, almost any Spare Parts supply will meet the definition of Initial Provisioning, because it will almost always be the case that it will be “transfer[red] in certain circumstances” to a particular Operator. However:

In response, the RR Entities point to clauses 8.5 of the ASA and 50.2 of the ECSURS.

Clause 8.5 of the ASA provides:

I am unable to see how this provision assists in the interpretation of the term Initial Provisioning, save that it is (yet) another provision recognising that Initial Provisioning “belongs” to a particular customer, who may (or may not) have been granted a contractual right to “put” surplus Initial Provisioning only back to its supplier, who may then re-sell it (by way of Initial Provisioning to someone else, or otherwise). Assuming in the RR Entities’ favour that AECSUs ordered to support a particular Operator’s fleet remain “Initial Provisioning” if re-supplied to another “particular Operator” for the purposes of supporting its fleet, that does not answer the question of whether AECSUs which are not ordered for that purpose can nonetheless constitute Initial Provisioning.

Clause 50.2 of the ESCURS provides:

I accept that clause 50.2 contemplates that Initial Provisioning might be held by the RR Entities, or the Customer, and I can see nothing in the definition of Initial Provisioning, or the concept of initial provisioning, which would be inconsistent with that state of affairs. However, that does not remove the requirement that Initial Provisioning has been ordered in support of a particular Operator’s fleet. Indeed, the definition of Exchange Service is “the service outlined in Section 3 of the ECSURS pursuant to which the stock of Initial Provisioning used by any Customer shall be replenished by [Goodrich] in accordance with the Programme Service Level”. These terms contemplate Goodrich replenishing a particular Operator’s Initial Provisioning following the Operator’s use of such Initial Provisioning.

I would also note that clause 50.2 (the only place in the ASA and ECSURS where language of this kind appears) would be an unlikely place to find a provision which had a significant impact on the scope of the concept of Initial Provisioning. It appears in Section 3 of the ECSURS which only applies “where R-R issues an instruction to [Goodrich] requiring Exchange Services”.

Further, clause 50.2 only applies to an Exchange Service operated by Goodrich, and would not apply to the PAS for that reason:

As to the suggestion that AECSUs exchanged under the PAS involve an Exchange Service operated by Goodrich:

Goodrich also argues that clause 50.2 can have no application because there has been no agreement between the RR Entities and Goodrich as to the level of Initial Provisioning which each PAS customer should hold. As to this:

Finally, the RR Entities argue that:

It is not clear to me whether this is an issue which actually arises. If it does, it is not a reason to interpret the term Initial Provisioning so that it extends to AECSUs acquired by the RR Entities for a pool from which they could be supplied to a number of competing airlines, but a reason to interpret the term Initial Provisioning so that it is satisfied when an order is placed by one economic unit which operates more than one airline for Initial Provisioning for both airlines. In that regard, I note that the definition of Operators is “any organisation or person operating an aircraft powered by a R-R Engines”, which might well be thought susceptible to an interpretation which accommodated the RR Entities IAG example. However, it is not necessary finally to resolve that question, because it does not assist in answering the issue which does arise in this case.

In these circumstances, it is not strictly necessary to consider whether the temporal limitation which I have found the concept of Initial Provisioning would ordinarily have (see [110]-[112]) applies under the ESCURS, an issue of construction on which both parties have forceful points to make. The definition does not include a temporal requirement, but the word “initial” in the defined term would suggest it is there, otherwise it would be a strange term to use. That interpretation derives some support from the following:

In the final analysis, I have concluded that the concept of Initial Provisioning as used in the ECSURS shares the same temporal limitation seen in uses of the term “initial provisioning” by the RR Entities in their documents, and which the evidence of both Mr Smith and Mr Loret seemed to acknowledge.

The RR Entities’ alternative argument is that AECSUs ordered for supply to the Operator’s OSS under the PAS constitutes Initial Provisioning. I accept that AECSUs ordered for the purpose of establishing a specific Operator’s initial OSS under the PAS have the requisite link with a particular Operator so as to constitute “Spares ordered by R-R or Operators in support of the operation of a particular Operator’s fleet”. Functionally, the OSS held by Operators who subscribe to the PAS serves the same purpose as Initial Provisioning ordered by Operators who do not, save that (presumably) the effect of the PAS is that PAS-subscribing Operators will be able to hold less stock, ceteris paribus, than a non-subscriber.

Goodrich has advanced a number of objections to this argument:

What, however, of the position when an AECSU sourced to form part of a pool is then used to “backfill” a PAS Operator’s OSS? I do not see how such an AECSU, which the RR Entities will have acquired from Goodrich at some anterior point in time when its eventual use would not be known or fixed, can subsequently become Initial Provisioning because the RR Entities later supply it to an Operator. Such AECSUs were not “ordered … in support of the operations of a particular Operator’s Fleet” and do not satisfy the requirement for Initial Provisioning at the point of acquisition from Goodrich, which is what clause 11.10(c) addresses.

On the evidence, I am satisfied that the 116 AECSUs which were supplied directly to the OSSs constitute Initial Provisioning. It seems to me more likely than not that the other 12 AECSUs were ordered for general pooling usage, and only subsequently shipped from an MSC to an OSS.

Finally, a separate issue arises as to AECSUs acquired for the Hong Kong MSC, on the basis that, at least initially, the only PAS subscriber supported from that MSC was Cathay Pacific. The RR Entities argue that all AECSUs acquired for the Hong Kong MSC constitute Initial Provisioning. As to this:

On the conclusions I have reached, the damages claim does not arise. However, having heard evidence and argument on these issues, and because in certain respects these issues are relevant to the quantification of the debt claim, I have set out my findings on most of the issues raised.

Goodrich understandably relied on a number of authorities which have stressed that, when the court is satisfied that a significant loss has been suffered, it must do its best on the material available to assess that loss. In particular:

It is also right to record the limits to the appeals of that kind, and their limited role when evidence as to the relevant matters is available or could have been produced (Marathon Asset Management Ltd v Seddon [2017] 2 CLC 182, [165]).

Goodrich also asked the court to grant it the benefit of any uncertainty as to the factual position on the basis that the RR Entities must have access to relevant evidence but have failed to disclose it. I am not persuaded that there has been any wrongful withholding of evidence by the RR Entities in this case, and the suggestions to that effect in Goodrich’s written closing was (rightly) not pressed orally. Model B disclosure was adopted by the RR Entities for quantum issues. Some 1.82 million documents were collected, 32,000 reviewed by Slaughter and May and 16,000 disclosed. In circumstances in which Practice Direction 57AD was introduced to streamline the disclosure process, and disclosure is now frequently conducted by reference to proportionate searches, often limited to particular custodians and which seek to locate documents which contain particular words, there will not be a complete documentary record at the trial. Where the course of proceedings suggest that documents of particular significance may not have been picked up, an application for supplemental disclosure can be brought, and, in some cases, the court may even ask for a further search at trial, or further searches may be undertaken by consent. That last course was adopted here, and it led to further material emerging, as summarised in Slaughter and May’s letter of 12 May 2023. That letter also convincingly answered the criticisms made of the RR Entities’ disclosure exercise in Goodrich’s written closing.

It is axiomatic that the benefits which Practice Direction 57AD is intended to bring will be largely neutralised if the absence, without more, of a document at trial which must once have existed is sufficient to justify an adverse inference. In this case, there is nothing to suggest that the RR Entities have not adopted an appropriate approach to record-keeping or complied with their obligations as litigating parties, and I am not willing to draw adverse inferences in relation to what other documents might have shown.

The ritual, respective, citations having been duly acknowledged, I turn to the principal legal issue on the quantum aspect of the case.

Goodrich’s case alleges the following:

While Goodrich’s Particulars of Claim plead the no-PAS Counterfactual as its primary case, at trial it was PAS Counterfactual which was in pole position. However, I am satisfied that the appropriate counterfactual is the no-PAS Counterfactual. Goodrich suggests that the relevant scenario is “what would in fact have happened if the RR Entities had complied with their contractual obligations and approached Goodrich under clause 4.2 of the ECSURS for a written quotation”.

As was made clear in Morris-Garner v One Step (Support) Ltd [2019] AC 649, [31]-[32], on the usual measure contractual damages “are intended to place the claimant in the same position as he would have been if the contract had been performed”, and, at [76], that “damages are based on the difference between the outcome of performance and non-performance”, and not the economic value of (or a substitute for) the right infringed. This is not necessarily the same position as if there had been no breach of contract. The position is put particularly clearly by Lord Reed at [91]:

In British Gas Trading Limited v Shell UK Limited [2020] EWCA Civ 2349, the seller of gas was contractually obliged to maintain a certain level of delivery capacity, which it was in fact impossible to maintain. It argued that damages should be assessed on the basis that, in order to avoid being in breach of contract, it would have served variation notices reducing the quantity of gas it was obliged to deliver. Males LJ rejected that contention, noting at [77]:

He continued:

In circumstances in which the breach of contract under consideration on the current hypothesis is undertaking activities in breach of a negative covenant not to do so (that being the essence of an exclusivity obligation), it is axiomatic that damages fall to be assessed by reference to the position as if that negative covenant had been performed, and the RR Entities had not undertaken those activities which breached Goodrich’s exclusivity right. The suggestion that damages should be assessed by reference to the price which the RR Entities would have paid had agreement been sought for a limited variation to those obligations involves an error of principle, because “damages for breach of contract depend on considering the outcome if the contract had been performed” not “the outcome if the contract had not been performed but had been replaced by a different contract” (Lord Reed, [142]), or on the basis that “that the party in breach would have taken steps to avoid being in breach of contract in the first place” (per Males LJ, [144]).

Goodrich seek to distinguish these authorities by relying on clause 4.2 of the ECSURS, which permitted the RR Entities (but did not oblige them) to seek a price from Goodrich for Goodrich to provide an Aftermarket Service for which the ECSURS did not provide a price. It is not clear, on Goodrich’s case, what Goodrich would have been asked or offered to do if such a request had been made – Goodrich’s argument was not really run on the basis that a price would have been provided for Goodrich to perform the service which the RR Entities wanted to offer their customers, but on the basis that the RR Entities would have agreed to pay Goodrich a premium for permitting the RR Entities do something they were not contractually entitled to do. While Goodrich did put forward a proposed amendment that it would have maintained the pools itself, and it has sought to keep this alive in a footnote in its opening, that amendment was objected to, and not pursued, nor were any of the practicalities of such a suggestion explored at trial. I am therefore satisfied that the only case open to Goodrich before me is one in which the PAS proceeded as implemented, save with a higher price being paid to Goodrich in return for Goodrich not objecting. I am not persuaded that, so formulated, Goodrich’s case involves a counterfactual in which clause 4.2 could have been invoked.

In any event, I do not accept that the RR Entities’ option to ask for a price displaces the orthodox contractual principles I have summarised at [142]-[145] above. No doubt it might be said in any case in which a party undertakes an activity (or fails to so) in breach of contract that it could have approached the claimant and agreed a variation to the contract to allow it to undertake the same or similar conduct, or which relieved it from its obligations, and that the claimant would have agreed to accept an adjustment to the contractual consideration as the price of not standing on its rights. That does not have the effect that damages for the breach are to be assessed on the basis of what release fee would have paid had such a negotiation been undertaken.

Goodrich advanced two further arguments in response to the suggestion that the PAS Counterfactual was inappropriate as a matter of principle.

First, it suggests that the ECSURS gave the RR Entities a discretion as to how to perform their obligation not to undertake activities falling within the scope of Goodrich’s exclusivity right, and that in those circumstances the court must carry out a factual enquiry as to what the RR Entities would have done by way of performance (relying on Durham Tees Valley Airport Ltd v bmibaby Ltd [2011] 1 Lloyd’s Rep 68, [95]-[96], it being said that this was a category four case for the purposes of Toulson LJ’s categorisation). However, reaching an agreement with Goodrich which would relieve the RR Entities of the relevant contractual obligation to the extent necessary to undertake the PAS without breaching the ESCURS is, self-evidently, not a means of performing the RR Entities’ (assumed) obligation not to undertake the PAS. Adapting Males LJ’s observation in British Gas Trading, [81] to the negative covenant in this case, “this is not a category 4 case as described by Toulson LJ in Durham Tees Valley Airport. Rather, it is a single obligation case in which the defendants' obligation is simply [not] to do X”.

Second, reliance is placed on the recent decision of the Court of Appeal in Unicredit Bank AG v Euronav AV (The Sienna) [2023] EWCA Civ 471. That was a claim in contract by the holder of a bill of lading for breach of the contract of carriage evidenced by the bill, the breach arising from the owner delivering the cargo without production of the bill of lading. The owner argued that the breach had not caused the claimant (a bank which held the bill of lading by way of security) any loss, because, if the owner had refused to discharge the cargo without production of the bill, the bank would have consented to its discharge. At first instance ([2022] EWHC 975 (Comm)), Moulder J found that the bank had permitted discharge without production of the bill of lading, and alternatively that it would have done so if it had been asked. It appears that the case was argued at first instance on the basis that the issues of causation and loss depended on what the bank would have done if the owner had refused to discharge the cargo without production of the bill ([60]). The bank’s appeal on the quantum issue was premised on the factual assertion that, but for the breach, the cargo would have remained on the vessel, and the bank’s security interest retained and “that was as far as the counterfactual enquiry needed to go” ([101(2))].

At [103], Popplewell LJ held as follows:

At [108], he concluded:

The dispute in Unicredit has a number of interesting features. It arose in the context in which the breach of contract in question – delivery against a letter of indemnity rather than against a bill of lading – was the anticipated manner in which the contract would play out, both on the part of the original contracting parties who had provided for delivery without production of bills against an indemnity in their charterparty (the relevant contract) and on the part of the bank so far as the financing transaction was concerned. This reflected the difficulty of moving the bill through the sale and banking chain in time to meet the vessel’s arrival, exacerbated in that case by the effects of the Covid 19 pandemic, and the adverse financial consequences which would follow if the ship had to await the bill’s arrival at the discharge port. The effect of delivery against a letter of indemnity is that, in the usual course, the party entitled to enforce the contractual obligation to deliver only against the bill can look to the credit of the carrier, by a claim on the bill, who in turn will look to the credit of the party from whom it chose to accept the letter of indemnity in return for its agreement to deliver cargo otherwise than against the production of the bill. What, however, of the position where the party looking to enforce the contractual promise of the carrier only to deliver against the bill was party to the request that the carrier deliver without production of bills in the first place? Should they retain the right to look to the credit of the carrier, who would in turn be dependent on the credit of the party whose letter of indemnity they agreed to accept? Or should their conduct disable them in same way from setting up the breach they assented to and, on one analysis, were instrumental in bringing about? The answers to those questions are not obvious.

There is the further difficulty that answering that question posed by that scenario in what might be thought to be its more natural home – was there a breach at all? – is difficult if, as in The Sienna, (i) the bank has yet to become the holder of the bill at the time delivery is made without production of the bill, and therefore has no formal legal status to vary the carrier’s obligations under the contract, (ii) when there is a legal principle, at least when the alleged waiver takes place at a point when the bill of lading has the status of a mere receipt, that a prior waiver will not bind a subsequent holder of the bill (see [98]-[99] of The Sienna) and (iii) when, on the facts, the carrier was unaware of the dealings between shipper/seller, buyer and bank and so could not set up some species of estoppel ([39]).

It may also be important to note the manner in which the issues were pleaded and argued in that case. This was on the basis that, to establish loss, the bank had to show that it would have exercised its security in the goods ([39] and [103], and c.f. McGregor on Damages (21^st), [32-025] and, in relation to claims in conversion at least, The Jag Shakti [1986] AC 337), and, on the bank’s own case, it was necessary to show that but for the breach the owner would have contacted the bank who would not have permitted the ship-to-ship discharge of the cargo, something the bank described as “a necessary element” in its causation case ([104]).

Moving away from its specific commercial context, the decision might be said to raise the “clock-stopping” issue frequently encountered in private law damages assessments, and how far any counterfactual analysis should have regard to what would have happened subsequent to the date of breach in the “but for” world. English law offers different answers to that question in different contexts – e.g., the various debates about the “breach date” rule in contract, the debate in tort summarised in McGregor on Damages (21^st), [8-007]-[8-013] and the consideration of when the “clock stops” in cases of an alleged breach of trust, commercial and otherwise. What Unicredit emphatically does not support is the use of a counterfactual to measure damages in which the obligation not performed would never have arisen for performance. Rather, it is a case in which, on the facts, if the obligation had been performed as it should have been, that would have led to a further event “downstream” of the point when the breach occurred. For that reason, it provides no assistance to Goodrich in this case.

I heard factual evidence relating to this part of the case from:

I am satisfied that both witnesses were doing their best to assist the court, within the limits of their recollections, and the inherent difficulties in addressing counterfactual issues in a litigation context. I was grateful for Mr Kotrly’s well-judged submissions on the criticisms made of Mr Loret’s evidence.

I also heard expert evidence from:

Both experts sought to assist the court, and to narrow the issues, and I was materially assisted by their evidence. To a significant extent, the differences between them reflected different factual assumptions they had been asked to adopt, or their assessment of the factual evidence. I am satisfied, however, that certain of the assumptions which Mr Flower was asked to adopt were unrealistic, notwithstanding Mr Flower’s attempts to suggest otherwise.

The quantification of damages on this basis requires the court to determine a series of issues. Having determined those issues, the experts were able to arrive at an agreed damages figure.

The evidence on this question suffers from the difficulty that it is not clear what definition of Initial Provisioning the experts were using for the purposes of their calculations, and in particular whether it involved any chronological limitation or not.

Turning to the expert evidence:

I can deal with the first issue – whether I should have regard to recommended or actual levels of Initial Provisioning purchases – briefly. The evidence establishes that operators tended to acquire less by way of Initial Provisioning than the RR Entities had recommended. That was the evidence of Mr Loret. It was supported by the work of the experts which, looking at non-PAS operators during the relevant period, suggested 594 AECSUs were recommended, but only 354 units acquired. Mr Flower accepted that he had seen evidence that PAS Operators, on occasion bought less by way of Initial Provisioning than was recommended. The desire of Operators to minimise their outlay on Initial Provisioning was one of the commercial drivers of the PAS project. Accordingly, it is the figure for actual, rather than recommended, acquisitions which should be used.

The second issue is more complex: how many AECSUs, if any, would the PAS Operators have acquired above and beyond the level of Initial Provisioning (i.e., by way of replenishing Spare Parts) in the no-PAS counterfactual?

I do not believe I have got to the bottom of this point. However, on the material before me, I am unable to conclude that the PAS Operators would have acquired more than the range of 191 to 193 AECSUs put forward by the experts. I understand that, notwithstanding the proximity of these figures, complications would arise if I simply split the difference. Applying the broad axe in Goodrich’s favour, therefore, I find that 193 is the appropriate figure to use for AECSUs acquired by PAS Operators in the no-PAS Counterfactual, and that these would all have been acquired by way of Initial Provisioning.

The answer, on the evidence before me, is 193 units, for the reasons given at [165] to [166].

This issue raises two threshold questions of the construction of clause 11.10(c). This provides, in relevant respect:

The first question arises from the fact that there are different types of AECSU – the six principal types are engine electronic controllers, fuel pumps, hydromechanical metering units, fuel driven activators, data entry plugs and dedicated generators. The ECSURS defines an Engine Control System Unit as comprising 10 items, Appendix 4 sets out a non-exhaustive list of AECSUs

Goodrich argues that it is sufficient for the cumulative number of all AECSUs of whatever kind amount to more than 25% of the number of R-R Engines in the relevant engine programme contracted to be supported on an EFH basis. The RR Entities argue that the 25% threshold applies per type of AECSU.

The terms of the ECSURS do not offer a clear answer to this question, and a worked example in the agreement would have been helpful. In my view, the RR Entities’ construction is to be preferred:

Goodrich argues that, on the RR Entities’ construction, there is no prospect of the 25% Threshold ever being exceeded (inferentially for any type of AECSU). However, I do not have material before me which establishes that the parties would have held that view when the ASA and ECSURS were entered into, and in any event the threshold may, for all I know, have been included as a safety net to protect against an unanticipated event. Certainly, there is no sufficient material to persuade me that the outcome would have been sufficiently commercially absurd that the parties could not have intended it. I would note in this regard that, while I have not taken this into account, Mr Smith, from Goodrich, describes the threshold in his witness statement in the following terms:

Second, Goodrich argues that the 25% threshold should only be calculated by reference to R-R Engines which are contracted to be supported on an EFH basis for which Goodrich receives VSPs. This issue arises as follows:

On the basis of the first issue of construction alone, the 25% threshold was not exceeded.

If I had accepted Goodrich’s construction, then the calculation should have been performed by reference to the actual number of AECSUs that would have been acquired by PAS and non-PAS Operators (Mr Flower’s different treatment of those groups of Operators being unjustified, as he accepted in cross-examination).

It is accepted by the RR Entities that there is a real and substantial chance that up to 30% of PAS Operators would have become Time and Materials (T&M) Customers in the no-PAS Counterfactual. The issue which arises is whether calculations of loss based on that figure should be discounted on “loss of a chance” principles, and, if so, by what percentage.

I shall assume in the RR Entities’ favour that this issue (which arises at the stage of quantifying Goodrich’s assumed loss) is subject to loss of a chance principles (although cf. Chitty on Contracts (34^th) [29-089]). The evidence of both Goodrich and the RR Entities supported the figure of 30%. The actual figure might have been higher or lower than 30% (there are references in some documents to 35% or one-third), in contrast to the issue more frequently encountered in “loss of a chance” cases when the chance in question would play out in a binary fashion (would litigation have succeeded; would a contract have been concluded with a third party). Further, the evidence tends to suggest the figure of 30% may be conservative, with the result that a higher percentage is at least or more likely than a lower one. In those circumstances, I am satisfied that no discount is appropriate.

T&M Customers can acquire their Initial Provisioning from the RR Entities, or directly from Goodrich. Analysis performed by Mr Bezant showed that T&M Customers purchased a X% of their Initial Provisioning directly from Goodrich (the figure has been redacted for reasons of confidentiality). That figure reflected a calculation for (a) two T&M Customers who did not subscribe to any aftermarket services, who acquired Y% of their Initial Provisioning directly from Goodrich, and (b) a larger group of Operators, who did subscribe to TotalCare, who acquired Z% of their Initial Provisioning direct from Goodrich. There is a significant difference between the figures, and I would expect the conduct of operators who subscribed to TotalCare to be more representative of that of operators who subscribed to the PAS in the no-PAS counterfactual. Accordingly, I am satisfied that Z% is the appropriate percentage to use.

The issue which then arises is whether PAS Customers are, by virtue of their decision to subscribe to the PAS, operators who are likely to buy less Initial Provisioning directly from Goodrich than the whole pool of Operators. I am not persuaded that PAS Customers had a set of characteristics such that they were inherently less likely to purchase their Initial Provisioning directly from Goodrich in a no-PAS Counterfactual. In their opening, the RR Entities suggested that the particular characteristics of PAS Customers were such that not even 30% of them would have become T&M Customers and/or that they would not have purchased any Initial Provisioning directly from Goodrich. Both of these assertions had to be abandoned in closing in the face of the data, and they illustrate the difficulties of trying to attribute some special set of characteristics to the PAS Customers without a sufficient evidential base. The difficulty with the RR Entities’ submissions is that they can be equally applied to customers who never subscribed to the PAS, and, if they told the full picture, would mean that no T&M Customers would acquire Initial Provisioning directly from Goodrich, and yet we know that X% of them did, and must have had rational commercial reasons for doing so. Nor am I persuaded that it is appropriate to discount the Z% on loss of a chance principles, not least because, once again, on the material before me, it is as likely that the figure would have been higher as lower.

The experts are agreed that any direct purchases of Initial Provisioning from Goodrich need to factor in an agreed level of discount of D% from WLP (figure redacted for reasons of confidentiality).

PAS Operators were $/EFH customers who had opted for LRU Management Services and Service Level 2. Goodrich was receiving a VSP in respect of those customers. The 30% of these customers who would have become T&M Customers in the no-PAS Counterfactual would not have been $/EFH customers, and no VSP would have been payable. It follows that a credit must be given for the loss of the VSP.

However, it is common ground that Goodrich may have made additional revenue from those customers in the form of direct maintenance, repair and overhaul (MR&O) revenue which it would not have received from PAS Customers.

The total VSP received by Goodrich was $Am (figure redacted confidentiality), 30% of which is $Bm. As to MR&O services, assuming that Goodrich had provided the same level of services at the same cost, it would have been able to charge an additional $Cm from those services if provided to T&M Customers. Goodrich has suggested that, in due course over the life of the engine, the “lost” VSPs and the “gained” MR&O revenue would have balanced out. However, there was no expert evidence supporting this equivalence, and it might reasonably be thought that the VSPs would involve greater revenue than average MR&O revenue because Goodrich was assuming the risk of greater than expected levels of work.

Mr Flower suggested that the equivalence was inherent in the pricing model in the ECSURS, clause 47.3 of which, he contended, demonstrated that it was anticipated Goodrich would earn a mark-up of 4.3 over the life of an engine programme from Rework services. However, that provision did not apply to the Trent XWB and Trent 1000 engines, and it is not possible to determine whether clause 47.3 reflects a fair working hypothesis as to the link between VSPs and the value of MR&O work done, or simply a bespoke provision constituting one of a number of trade-offs in a very complex set of interlocking provisions. Indeed, when regard is had to the actual position, Goodrich’s profit margin on MR&O services provided to PAS Operators is much higher than in respect of MR&O services provided to T&M Customers. That might reflect the fact that MR&O work will become predominant in the later life of an engine. However, it might also reflect the fact that the VSP is not simply an attempt to capture the revenue payable in the T&M context, but reflects a significant risk premium or some other form of trade-off. I do not have the information to determine which, and in those circumstances the only “counter set-off” which has been established to be set against the $Bm loss of VSPs is $C.

Finally, it is common ground that the royalty provided for in clause 11.12 of the ECSURS would need to be factored in, which is a straightforward task which I will ask the parties to undertake.

On the basis of these findings, and in case this part of the claim becomes relevant, the experts are asked to agree the level of damages which would have been recoverable on this part of the claim had the relevant breach of contract been established in the no-PAS Counterfactual.

In contrast to a claim for negotiating damages under the principle recognised in Morris-Garner v One Step (Support) Ltd [2018] UKSC 20 (in particular at [1], [59] and [75]), this part of the Goodrich’s case does not involve the court assessing an objectively reasonable release fee. If the reality is that these parties would not in fact have arrived at an agreed outcome so as to permit the RR Entities to do what, on the current hypothesis, they were not permitted to do, then on Goodrich’s case, this damages claim must fail.

In this case, the parties’ pleaded positions as to what an acceptable price would have been are very far apart – the RR Entities say that they would not have been willing to pay much above IPC (which Goodrich says it would not have accepted), and Goodrich that it would not have been willing to accept much less than WLP (which the RR Entities say would have rendered the PAS uneconomic). Goodrich’s case at trial was rather more flexible.

I have found the process of making the findings of fact relevant to this issue difficult, because of the artificiality of the “jumping off” point. If the RR Entities had been under a primary contractual obligation to seek Goodrich’s consent to launch the PAS (as opposed to a primary contractual obligation not to operate the PAS, a breach of which could have been avoided by a variation), then it would be possible to determine when the obligation fell to be performed.

Goodrich’s case as to the point in time at which I should determine whether a negotiation would have been successful, and on what terms, remains obscure. If the operation of the PAS in breach of the exclusivity obligation is regarded as a continuing breach (which appears to me to be the correct characterisation), then I accept that it is open to Goodrich to advance a complaint in relation to the operation of the PAS when launched as a general offering in late 2016, even though there had been an earlier scheme which was also operated in breach of the ECSURS.

How negotiations would have proceeded, if they had been commenced, is very difficult to determine, because there were any number of ways of approaching the issue, including some form of profit/risk-sharing agreement, an increased price but reduced VSPs, a tiered approach depending on how successful the arrangement proved and so forth. However, the only variable explored at trial was a change to the price. If approached on the assumption that the request is to be made under clause 4.2 of the ECSURS (that being the only basis on which Goodrich suggested that this was the correct counterfactual) then I agree that Goodrich would initially have come back with a price, but of course that would not have precluded subsequent negotiation between the parties as to other terms in an attempt to arrive at a mutually acceptable deal.

When reaching its conclusion as to what the outcome of the assumed negotiation would have been, Goodrich asks the court to draw adverse inferences from the absence of a Rolls-Royce witness who would have been a decision-maker in any such discussion. However, Goodrich did not call someone in an equivalent position, and if it is suggested that Mr Smith met that requirement, he offered very little insight into Goodrich’s negotiating position. Further, the issue to be addressed is a hypothetical one, where the weight to be accorded to witness evidence is significantly less than in cases of direct observation of historical facts. Finally, Goodrich’s pleaded case as to the course of the hypothetical negotiation was sparse indeed, and in those circumstances I am not willing to draw an adverse inference from the RR Entities’ failure to call a witness simply to deny what is now accepted by Goodrich – that the RR Entities would not have paid WLP.

I accept that the RR Entities were strongly attracted to a parts availability service – that much is clear from the various internal business plans and proposals, identifying both the opportunities a parts availability service would bring in terms of revenue, and the threats presented by a failure to introduce such an offering to the RR Entities’ LRU Management Services business. However, I also accept Mr Loret’s evidence that the PAS needed to be able to “wash its own face”, and it is noteworthy that the internal assessments all appraise it on this basis, rather than quantifying benefits to the LRU business, and that the process of developing the offering and bringing it to market took a number of years. It is also clear that a parts availability service would only have been a viable offering if the AECSUs formed part of it, and the RR Entities did not suggest otherwise.

The following matters are also clear:

The question of how many AECSUs the RR Entities acquired for the PAS is a question of fact. The RR Entities say the correct number is 535 units, Goodrich that it is 572. As the party claiming that a debt is due or damages payable, the burden of proof lies on Goodrich, but, in determining whether that burden has been discharged, the court can have regard to the fact that the evidence relating to the supply of these AECSUs was exclusively within the control of the RR Entities.

The court has access to so-called “shipto” codes obtained from the RR Entities’ business information systems showing the shipment destination of various AECSUs, and evidence from Mr Loret (who I accept was involved in the design of those codes and therefore able to give evidence as to their significance) as to what they mean. The shipping codes cannot be conclusive, because there are occasions when it is clear from other evidence that the eventual destination of an AECSU is otherwise than as suggested by the “shipto” code. Nonetheless, as the primary business record of the RR Entities, I am satisfied that the “shipto” codes provide a prima facie basis for determining where AECSUs were shipped.

The first set of disputed AECSUs comprises 24 units:

There are then 85 units shipped to Rolls-Royce entities or joint venture companies:

Accordingly, I am satisfied that the number of PAS units was 538.

There is then a further issue as to the number of units which would have been supplied to the PAS in circumstances in which it is said the higher price payable by the RR Entities may have reduced demand. On the terms which I have determined would have been arrived at on the present counterfactual, I am not persuaded that the level of demand would have been meaningfully different to that which it was. I therefore find:

On the basis of these findings, the experts agree that the damages recoverable on the basis of the PAS Counterfactual would have been US$41,474,483.

Paragraph 36 of Goodrich’s Reamended Particulars of Claim provides:

Paragraph 36A provides:

Claims are asserted, therefore, by reference to clauses 46.1 and 46.2, 51.1 and 11.1 of the ECSURS.

Clause 46.1 appears in Section 2 of the ECSURS, entitled “Generic Rework Services $/EFH” and applies to all instructions placed with Goodrich. It provides:

Clause 46.2 provides:

No such information or summary was provided in respect of Initial Provisioning. The RR Entities say that this was not the subject of any complaint at the time, and that this gives rise to some species of estoppel, or perhaps waiver by conduct. Mr Smith, giving evidence for Goodrich, accepted that:

Against that background, I am persuaded that the parties, by their conduct, agreed to vary the requirement that RR Entities provide such information when placing orders, albeit it would have been open to Goodrich to request the information again had it wished to do so.

Further, no relief is claimed in respect of these breaches. Goodrich’s damages claim in the prayer is limited to “damages in respect of the RR Entities’ breaches of clause 11.1 of the ECSURS and the Exclusivity Provisions in the ASA and ECSURS”. Nor has there been any counterfactual evidence as to how the provision of this information would have placed Goodrich in a better position.

In these circumstances, this allegation takes Goodrich nowhere.

I have referred to clause 51.1 of the ECSURS above, I have already found that the PAS did not constitute the operation of an Exchange Service for the purpose of section 3 of the ECSURS. In these circumstances, I do not believe any issue as to clause 51.1 arises. If I am wrong, the position is as per clauses 46.1 and 46.2 above.

This aspect of Goodrich’s claim requires close attention, and undoubtedly loomed much larger by the end of the case than it had at any prior stage. By way of a summary, the pricing provisions in the ECSURS were as follows:

What is asserted is that the RR Entities breached clause 11.1 which provides:

It is alleged that clause 11.1 required the RR Entities “to specify in all Spares Orders the applicable price” (paragraph 22C), and that this obligation was breached by asserting that the Spare Orders placed asserted that the relevant orders “are for Initial Provisioning and that [IPC] is payable” (paragraph 36A). It is said that those assertions were wrong because “the correct price payable in respect of such units was World List Price pursuant to clause 9.2 of ECSURS” (paragraph 36A(1)). It is then said that if the RR Entities had complied with clause 11.1, “the RR Entities would have acquired the same number of [AECSUs] but would have specified in the relevant Spares Orders, and paid, the correct price for them” (paragraph 41D), it being alleged that the “correct price for all such units was [WLP]” (paragraph 41D(1)).

In opening, Goodrich defined its case under clause 11.1 as follows:

The alleged breach in failing to pay the correct price – as opposed to failing to specify it, even though it was payable – does not appear clearly in Goodrich’s Particulars of Additional Claim. However, paragraph 36A relies not simply on the initial failure to specify the correct price, but the subsequent conduct in asserting that the correct price is IPC when “the correct price payable in respect of such units was [WLP]”, and a generalised allegation of loss and damage was pleaded (paragraph 38). On that basis, I am satisfied that Goodrich’s case embraced a claim for damages for failure to pay “the correct price payable in respect of such units”, albeit no debt claim.

Reverting to clause 11.1:

In this regard, the agreed Case Memorandum describes Goodrich’s case as being “that in breach of clause 11.1 of the ECSURS, the RR Entities acquired the [AECSUs] for use in the PAS at a much lower cost than that to which they were entitled”, although it continues “the Defendant alleges that, had the RR Entities complied with clause 11.1 … the RR Entities would have acquired the same number of [AECSUs] but would have paid [WLP] … and claims the difference in price as damages”.

Before the exchange of written closing submissions, I raised the issue of whether Goodrich had advanced, or should be permitted to advance, a claim in debt with the parties before the service of the parties’ written closing submissions. Goodrich did not seek to advance such a claim in their written closing, albeit they did contend that their claim was “in substance for the failure of the RR Entities to pay the correct price”. Goodrich continued:

That might be thought to have suggested that a deliberate decision had been made not to plead a debt claim. The reason offered – the need to specify the details of each individual transaction – did not appear convincing, not least because it would seem to be equally true of the claim, which was brought, which involved an assertion that the correct price had not been paid. In the course of oral closings, Mr Croall KC confirmed that the issue of whether the claim could be advanced in debt had not come into focus for Goodrich until the close of the trial, as a result of the exchanges with the court.

The proposed amendments were not put forward until 5.30pm on Saturday 13 May (albeit I suspect the RR Entities were aware it could be coming from exchanges in court at the end of 11 May), with oral closing submissions in respect of this aspect of the case taking place on 15 and 16 May. The proposed amendments comprise:

On any view, this was, chronologically, a very late amendment indeed, and it comes against a background in which Goodrich had made a number of amendments to its statement of case, one set as recently as February 2023. The principles to be applied when considering whether to permit a very late amendment have been summarised by Carr J in Quah v Goldman Sachs International [2015] EWHC 759 (Comm) and more recently, the Court of Appeal in ABP Technology Ltd v Voyetra Turtle Beach Inc [2022] EWCA Civ 594, [23]. In ABP, the Court of Appeal approved the following summary of the principles given by Coulson J in CIP Properties (AIPT) Ltd v Galliford Try Infrastructure Ltd [2015] EWHC 1345 (TCC), [19]:

In this case, it was not in dispute that permission to amend should not be granted if it would occasion the adjournment of the trial or require the RR Entities to re-visit earlier steps such as disclosure, witness evidence or expert evidence. To the extent that it is viable at all, therefore, it will not be a “very late” amendment in the strict sense of that term. I accept that there is no good reason for the delay in bringing the amendment forward, and that it reflects the further analysis and clarity which the trial dialectic will sometimes bring.

The RR Entities identified certain reasons why it was said that the proposed debt claim was not realistically arguable. I am satisfied, however, that the issue met the (relatively low) threshold for arguability, and that these objections are better considered when addressing the amendments on their merits, rather than at the permission to amend stage.

As to the particularity and/or clarity of the proposed amendment, Mr Toledano KC criticised the amendment for not identifying the date each debt had accrued, and the specific debtor. However, that depends on the court’s findings as to which units constituted Initial Provisioning and which did not, and the issue which is raised by the debt claim in this regard is no different to that which arises on the clause 11.1 damages claim (which asserts a loss by reference to the difference between the amount paid and the amount which was the correct price, for each relevant order). The reality is that both parties have been content to argue the case on a basis which treated the RR Entities as a collective, a pragmatic decision, and one which may have reflected the fact that Rolls-Royce Plc undertook to procure the performance of the obligations of any Rolls-Royce affiliates (clause 41.1).

That leaves the key issue of prejudice. So far as the claim for damages to pay the correct price is concerned, I am satisfied that, albeit with much less clarity and insufficiency of focus, that is embraced within the existing pleading: see [216] above. So far as the debt claim is concerned (and also the damages for failure to pay the correct price, if I am wrong in my conclusion that this is already embraced within Goodrich’s Particulars of Claim), a key factor is that it was already a necessary ingredient of Goodrich’s case that the price payable for AECSUs provided for the PAS was WLP, and that this was the price which the RR Entities were obliged to pay:

In these circumstances, I have decided to permit Goodrich to amend their Particulars of Claim in the terms of the draft submitted on 13 May 2023. I am conscious of the risk that allowing the amendment at this late stage may leave the RR Entities with a feeling of dissatisfaction, particularly if the point succeeds, a sense of dissatisfaction which can only be enhanced by the fact that the point surfaced following exchanges with the court, and when Goodrich served amendments to its case as recently as December 2022, for which it wrongly initially asserted it did not need permission. However, in circumstances in which I am persuaded that the essentials of the point were already pleaded, and it raises only issues of construction which I am satisfied the RR Entities were able to deal with in closing, I do not accept the disappointed expectations as to the outcome of a case constitute a relevant head of prejudice, or being “mucked about” for the purposes of the application to amend. What significance these points may have when it comes to decisions about costs will have to be addressed in due course.

By contrast, not allowing Goodrich to advance the amendment might, depending on how the clause 11.1 damages issue is finally resolved, lead to the loss of the ability to advance a claim of considerable value. While that would not take Goodrich very far if there was credible evidence of prejudice the other way if the amendment was allowed, I am not persuaded that there is any such prejudice. It might also involve approaching the determination of the case on a very artificial basis.

I understood the RR Entities to advance three responses to the debt claim, which inevitably overlap or interrelate to some degree:

By clause 2.3, there was “incorporated into each Order … provided to [Goodrich] by the [RR Entities] in accordance with Clause 4 … the terms of this Agreement”. By clause 3.1 of the ECSURS:

Clause 3.3 provides that “Orders … (and any terms incorporated therein, in accordance with this Agreement), this Agreement …. and any other relevant provisions of the ASA shall be the only express terms which shall govern the provision of Aftermarket Services” and “any other terms that [the RR Entities] or [Goodrich] attempt to impose on any Orders shall be of no effect”.

Clause 4.1 provides that “neither Party shall be bound by and shall have no liability in respect of Orders …. which are not (b) consistent with the terms of this Agreement … unless the Parties expressly agree that a written term of such Order … shall prevail over the terms of this Agreement.”

Clause 5.7 provides that “title and risk of loss or damage to Spare Parts provided to [the RR Entities] or a Customer (as appropriate) by [Goodrich] shall pass on the delivery of the Spare Parts for the [AECSU] to the address specified on the Order ….”.

Finally, Clause 9.5 provides:

Against this background, I am satisfied that the specifying of the incorrect price in a Spares Order does not affect the price payable under the ECSURS. The effect of clauses 2.3 and 3.3 is to incorporate the pricing provisions of the ECSURS into any Spares Order, and only those provisions. I am not persuaded that Goodrich’s entitlement to refuse to accept Orders where the required information has not been provided (clauses 3.1 and 4.1) has the effect that the price payable for an Order accepted by Goodrich in respect of an Aftermarket Service can differ from the price which the ECSURS provides merely because the RR Entities have specified the wrong price on the Spares Order. Further, clause 4.1 does not, in my view, have the effect that the RR Entities’ failure to specify the correct price in the Spares Order relieves them of the obligation to pay the correct price. Rather, clause 4.1 is intended to ensure that the parties’ obligations in respect of an Order for which the ECSURS defines the relevant terms are as set out in the ECSURS, and not otherwise.

In oral closings, the RR Entities submitted that clause 9.5 of the ECSURS had the effect that unless Goodrich had invoiced the RR Entities for those AECSUs at the correct price, no debt in that amount accrued. As to this:

In my view, the better interpretation of clause 9.5 is that it does not affect the point in time at which the debt arises. The clause links the payment obligation not to the provision of the invoice as such, but to the time when the goods are despatched or the services completed. I understand the RR Entities to have accepted that this was the correct analysis in their supplemental closing submissions, but whether or not they did, that is the conclusion I have reached.

The Colledge line of cases is principally concerned with the issue of whether the submission of an invoice is necessary to start time running for limitation purposes, and a strong factor underlying the principle of law which evolved is avoiding the inherently unattractive position in which, by dint of choosing not to serve an invoice, a party with a claim can indefinitely postpone the running of the limitation period (c.f. Lord Esher MR in Colledge, 705). The question of what happens when a case reaches trial, and no invoice has been served in the appropriate amount, has received rather less attention. For example, it is not clear how far the making of the itemised claim in the litigation can itself be sufficient to render the debt payable so as to obtain judgment upon it.

Nor do the cases explore the legal consequences where, as is often the case and is the case here, an invoice has been submitted, but in the wrong amount. If an invoice has been submitted in too high an amount, it seems unlikely that the effect is to prevent any debt at all becoming payable, certainly to the extent that the party receiving the invoice is able to work out the amount due (see Lombard North Central Plc v European Skyjets Ltd [2022] EWHC 728 (QB), [137]-[138]). I would not regard clause 9.5 as having such an effect – the words “provided the invoice is accurate” – in my view are intended to do no more than make it clear that the RR Entities’ payment obligation only arises to the extent that the invoice is accurate.

If the invoice is lower than the figure actually due, what is the position?

I have observed that, clause 9.5 does not expressly link the payment obligation to the rendering of an accurate invoice, but to delivery or completion of the services, and I have concluded that the words “provided the invoice is accurate” are intended to do no more than make it clear that an inaccurate invoice does not itself create a payment obligation which does not otherwise exist. I am satisfied that is also the position where the invoice understates the amount of the debt. In these circumstances, I am satisfied that WLP was due and payable for Spares Orders under the ECSURS which did not constitute Initial Provisioning, even though these AECSUs were invoiced at IPC.

In any event, and as an alternative basis for my conclusion, I am satisfied that the parties to the ECSURS cannot have intended that the amount due for the supply of Spare Parts would not become payable unless invoiced in full if the reason for the inaccurate invoice was the RR Entities’ breach of contract in specifying the wrong price in the Spares Order. That is an unlikely construction, which would permit the RR Entities to take advantage of their (assumed) own wrong, contrary to the interpretative principle recognised in Alghussein Establishment v Eton College [1988] 1 WLR 587.

In the present context, in which the distinction between a claim in debt and damages is seen as highly significant because of the counterfactual arguments advanced in the latter context (see [259] below), it might be argued that any assumed breach of clause 11.1 should take effect only as an independent damages claim, rather than having the result that an invoice in an incorrect amount should be capable of triggering the obligation to pay. However, I am not persuaded that that is a reason for giving clause 9.5 a construction which I am not persuaded that sensible commercial parties would have attributed to it.

Further, there is a substantial body of authority to the effect that where one party is in breach of an obligation to satisfy a condition to the other party’s right of action, “the court will proceed as if the condition was satisfied (rather than engage in an assessment on conventional principles of the loss suffered by reason of the breach)” (Nautica Marine Ltd v Trafigura Trading LLC (The Leonidas) [2021] 1 CL 362, [95]). That doctrine is particularly likely to be engaged in relation to a condition to an obligation of payment (and, I would add, even more so to a condition to the obligation to pay an accrued debt). In Compagnie Noga d’Importation et d’Exportation SA v Abacha (No 3) [2002] EWCA Civ 1142, Rix LJ reviewed the relevant authorities at [95]-[108] and concluded at [106]-[107]:

This principle is criticised in Chitty on Contracts (34^th), [4-204], in terms unchanged from the 33^rd edition, and its scope may be a matter of debate (cf. The Leonidas, [107]-[110]). However, in its core field of application, where the breach prevents the satisfaction of a condition to the accrual of a debt, the principle has the approval of Lord Sumption in Société Générale (London Branch) v Geys [2012] UKSC 63, [131] as well as Rix LJ in Abacha. Its application in this case, to a procedural or pro forma pre-condition to the creditor’s ability to recover an accrued debt, presents an a fortiori case.

In order to determine whether a debt claim is due on that alternative basis identified at [242], it is therefore necessary to consider whether the RR Entities were in breach of clause 11.1, an issue to which I now turn.

Clause 11.1 is expressed in normative language (“shall be in the form of Spares Orders”), as is the definition of “Spares Order” (“an instruction … comprising the following requirements”, emphasis added). That does suggest it imposes an obligation, and does so as a matter of express language rather than implication.

Mr Toledano KC submitted that clauses 3.1 and 4.1 provide that Goodrich was not obliged to accept an order which contained inaccurate information, and that was its only remedy. If rejecting the order was Goodrich’s sole remedy, then it is not clear to me what the RR Entities say the consequence of accepting an Order which had used the wrong price would be. If it were to be suggested that, unless it chose to reject the Order, Goodrich was obliged to complete it for the price on the Spares Order, then I am unable to accept that contention, which is inconsistent with the terms of the ECSURS summarised at [230] to [235] above. Nor is there any language which suggests that clauses 3.1 and 4.1 contain an exclusive remedy for inaccurate information on a Spares Order.

Mr Toledano KC also pointed to the fact that, as originally drafted, it was envisaged that the RR Entities and Goodrich would reach agreement as to the level of Initial Provisioning which was to be recommended (clause 50.2) and that the RR Entities would provide Goodrich with information as to the level of Initial Provisioning being provided to each operator (clause 46.1), such that Goodrich would be able to work out for itself what the correct price was. As to this:

In circumstances in which the price specified by the RR Entities in the Spares Order was highly likely to be the price invoiced by Goodrich, and assuming that specification of the correct price in the invoice is necessary for that price to be recoverable as a debt, the argument for interpreting clause 11.1 as imposing an obligation to specify the correct price are particularly strong. As between the RR Entities and Goodrich, the RR Entities were far better placed to determine whether or not AECSUs were being ordered by way of Initial Provisioning, for Consignment Stores or otherwise.

Mr Toledano KC relied, in support of the contrary argument, on the decision in Verizon UK Limited v Swiftnet Limited [2008] EWHC 551 (Comm). In that case Gavin Kealey KC, sitting as a Deputy High Court Judge, held that a term that one party would render accurate invoices to the other was not to be implied into a contract, stating:

As Mr Kealey KC noted, there is authority establishing that demanding performance under a contract which is not due is not (or at least is not necessarily) a breach of contract. Many of these cases involve a party asserting a contractual right to more than is due (Lombard North Central Plc v European Skyjets Ltd [2022] EWHC 728 (QB), [170]). It might be said that asserting a contractual entitlement to a service or goods at a different price to that which the contract provides is no different.

In the Verizon case, there was no suggestion that the misstatement of the price by the invoicing party would impact on the other party’s ability to assert its contractual rights. Further, Mr Kealey KC was concerned with an attempt to imply a term whereas, as I have stated, Goodrich’s case is supported by the express terms of the ECSURS. Finally, the facts of Verizon were very different from those here. Swiftnet was exercising a contractual right to use MCI’s services. It was in a position to work out the amounts actually due, and its claim was, essentially, for the administrative costs of correcting MCI’s errors. In the present case, by contrast, the RR Entities were asserting a contractual right under the ECSURS to receive AECSUs at the price it had specified in the Spares Order, in circumstances in which it was highly likely that it would be invoiced by Goodrich at the price it had included in its Order. There are contexts in which a contractual demand for something which the other party is not in fact obliged to provide will constitute a breach of contract. This is the case of so-called “non-contractual orders” in both time charterparties (Temple Shipping Co Ltd v V/O Sovfracht (1945) 79 Ll L Rep 1) and voyage charterparties (Batis Maritime Corporation v Petroleos del Mediterraneo SA (The Batis) [1990] 1 Lloyd’s Rep 345, 349-50). In GW Grace & Co Ltd v General Steam Navigation Co Ltd [1950] 2 KB 383, 396, Devlin J noted:

A key feature of those cases, which is present here, is that the contract in question is one in which one party is permitted to give orders on a regular basis requiring the other party to take particular steps, and in which there is a high likelihood that any “orders” so delivered, given their peremptory character, will be complied with as a matter of course, without extensive consideration or discussion at senior management level, and where the person giving the order is in general best-placed to determine its legitimacy.

For these reasons, I am satisfied that clause 11.1 of the ECSURS obliged the RR Entities to specify the correct price in the Spares Order. I am also satisfied that if the Spares Orders had specified the correct price, Goodrich would have rendered invoices at that price (and I did not understand the contrary to be contended).

The RR Entities pleaded a limited estoppel case to address one aspect of Goodrich’s pleaded case – that in the absence of agreement between the RR Entities and Goodrich as to the recommended levels of Initial Provisioning for operators, the provisions of the ECSURS dealing with Initial Provisioning were not engaged and/or that the RR Entities were in breach of its obligation to notify Goodrich of Initial Provisioning requirements (paragraph 36 of the Particulars of Additional Claim). The RR Entities’ estoppel plea pointed to the fact that Goodrich had never complained about the absence of such information, but had supplied Initial Provisioning nonetheless and that:

Neither the type of estoppel nor its ingredients were set out. In the event, I have found that these provisions were varied by conduct (see [207]-[208]).

In the RR Entities’ oral closing, it was suggested that a similar estoppel argument arose in relation to the clause 11 breach. The argument was not further developed. However, I cannot see any basis on which Goodrich’s acquiescence in the RR Entities not providing information under clause 46.1 and/or (if relevant) the absence of attempts to agree recommended levels of provisioning in clause 50.2 could preclude Goodrich from complaining that the RR Entities had included an incorrect price in the Spares Orders and failed to pay the correct price.

Goodrich’s counterfactual case for the alleged breach of clause 11.1 is that “the RR Entities would have acquired the same number of [AECSUs] but would have specified in the relevant Spares Order and paid the correct price for them” (paragraph 41D of the Amended Particulars of Claim). The RR Entities’ response to that claim (pleaded at paragraph 28C(3) of the Amended Defence) is:

That paragraph is denied by Goodrich (paragraph 25C of the Amended Reply), and at paragraph 25B.3, Goodrich advances a positive case that RR Entities would have proceeded with the PAS.

A similar issue as to the correct counterfactual arose in Medsted Associates Ltd v Canccord Genuity Wealth (International) Ltd [2020] EWHC 2952 (Comm). In that case the claimant introduced clients to the defendant on terms that, if the defendant did business with them, the claimant would be paid commission. The defendant did business with those introduced clients without disclosing that to the claimant, and without making the promised payments. The defendant argued that damages should be assessed by reference to whatever level of trading would have taken place had it not concealed its activities from the claimant, on the basis that less trading would have taken place had it been necessary for the fees charged to take account of the payments due to the claimant. Mr Nicholas Vineall KC, sitting as a Deputy High Court judge, rejected that argument, stating at [39]:

I raised the Medsted and British Gas decisions ([143] above) with the parties prior to closing submissions. The RR Entities made a number of points in response.

First, they said that it was clear from Goodrich’s pleading that it was making a factual claim about what the RR Entities would have done. As to this:

Second, it is said that Goodrich’s case that, if the RR Entities had breached clause 11.1 of the ECSURS by stipulating the wrong price, it follows that the RR Entities would have purchased the same number of AECSUs at the same price is a non sequitur. However, the points made in the previous paragraph apply equally here.

Third, it was suggested in written closing submissions that the “least burdensome obligations” rule applied here, and required the court to assess damages on the basis that the relevant orders had not been placed. The argument was put as follows:

I am not persuaded that the clause 11.1 breach is susceptible to a “least burdensome obligations” analysis. As pleaded by Goodrich, the breach of contract alleged is a failure to identify the price payable “for the Spare Parts identified in (a)”, i.e., those which have been the subject of the Orders, which the RR Entities required Goodrich to deliver, and which Goodrich did in fact deliver. I am not persuaded that the obligation to specify the correct price for the AECSUs being ordered can be performed by not ordering AECSUs at all (in which scenario there would have been no delivery obligation at all). That is to confuse the position where there is more than one way of performing a contractual obligation to do a certain thing which obligation has fallen due for performance, and the position where different conduct by the Defendant would have prevented the obligation arising in the first place. In short, the argument directly cuts across the principle of law identified by Males LJ in British Gas Trading, [80]:

Fourth, it is suggested that the decision in Medsted is either distinguishable or wrong. The attempt to distinguish the decision on the facts because “what seems to have exercised the Judge in Medsted was the fact that [the defendant] had taken a deliberate decision to breach its contract”, whereas “in the present case … there is no suggestion that the RR Entities have deliberately breached clause 11.1 of the ESCURS” was rightly not pursued in oral closing. There is no relevant legal principle which was in play in that case which turned on the question of whether the breach in question was deliberate or not.

As to the suggestion that the decision is wrong:

Turning to Mr Kramer KC’s criticism of the decision:

In this case, I have concluded that, on its existing pleading, Goodrich can point to two types of breach – specifying the wrong amount in each Spares Orders, and not paying the amount payable on each Order. If the only breach pleaded had been the latter, then I can see no basis on which the RR Entities could argue that damages should be assessed on the basis that the later breaches would not have taken place had the correct price been paid on the first relevant Spares Order.

I am not persuaded that the existence of a breach of contract in relation to the specification of the price leads to a fundamentally different outcome, in which the RR Entities get to revisit their decision to order the AECSUs:

Mr Toledano KC submitted that this result was “entirely unfair and uncommercial” and “sits very uneasily with the overall aim of compensatory damages” because “Goodrich would get an enormous windfall”. I accept that, particularly through the judgments of Lord Sumption, English contract law has moved towards greater consistency between the rules which apply to the assessment of damages, and the compensatory rationale of those rules (see William Day and Sarah Worthington (Eds), Challenging Private Law: Lord Sumption on the Supreme Court (2020), xv-xvi and Adam Kramer, “Contract Damages”, chapter 5 of the same volume). However, the assessment of compensation is not undertaken on a holistic basis, but by reference to the breaches of contract pleaded, and what the consequences of performance of those obligations would have been.

Nor am I persuaded that arguments about windfalls are helpful in the current context, particularly when the “windfall” in question is the payment of the contractual price for the AECSUs ordered and used by the RR Entities. If the court is to have regard to windfalls, then it is instructive to consider the position of the parties if the PAS had been the success which the RR Entities had expected it to be. The RR Entities would have made a large profit, and developed a highly competitive service to the detriment of market competitors, and all because it had been acquiring AECSUs from Goodrich at IPC. As the RR Entities’ service design document of 6 October 2016 observed of the PAS:

Yet on the RR Entities’ case, their liability to Goodrich in damages would be assessed on the basis that, had the amount actually due been paid, the RR Entities would not have proceeded with the PAS, allowing them to retain the competitive benefits of their breaches of contract, but with damages being assessed by reference to a counterfactual in which those benefits would not have been obtained.

It should also be noted that this issue would only arise if the RR Entities had succeeded in their argument that the incorrect prices included in Goodrich’s invoices (which reflected the incorrect prices in the RR Entities’ Spares Orders) precluded what would otherwise be debt claims, to which counterfactual arguments of the type which I have just considered would be irrelevant. The wholly technical basis on which, on this assumption, no debt claims had arisen provide support for the approach to assessing damages which I have concluded is legally correct.

If I had proceeded on the basis that only one type of breach of contract was pleaded – the failure to include the correct price in each of the orders – then I do not believe this affects the analysis. Had each order included the appropriate price, Goodrich would have supplied the relevant goods and invoice and been paid the contractual price. A counterfactual which does not involve the RR Entities performing the relevant obligation (including the correct price in the orders placed), but the RR Entities never coming under such an obligation (because it did not place the orders) is, in my assessment, unsound and contrary to the decision of the Court of Appeal in British Gas Trading.

I have set out above my reasons for concluding that 538 AECSUs were supplied to the PAS at [196] to [198] above.

Of those 538 AECSUs, 101 were originally shipped to AOG stores and later re-purposed to the PAS. I accept that any debt claim in relation to these 101 AECSUs must fail – at the date of acquisition, the RR Entities were permitted to acquire AECSUs for AOG or consignment stores at IPC (Schedule 25C paragraph 1.2 of the ECSURS). While clause 4 imposed certain restrictions on the use of Consignment Stores, Goodrich has advanced no claim for damages in relation to any repurposed AECSUs, and the factual issues which such a claim might have raised were not explored. Indeed, paragraph 36A of the Particulars of Claim was limited to AECSUs “acquired for use in the PAS”. Goodrich’s attempt to secure the factual findings to support that unpleaded claim in their written closing submissions must be rejected.

That reduces the number of AECSUs for the purposes of the clause 11.1/debt claim to 437.

Of these 437, I am satisfied that 116 are supplies to OSSs and therefore did fall within the definition of Initial Provisioning (see [132]).

The result is that under clause 9.2 of the ECSURS, WLP was payable for 321 AECSUs, and the RR Entities are liable in debt for the difference between that figure, and the amount they have actually paid. That amount has been agreed by the experts at US$112,285,440.

While I have concluded that this is not a relevant enquiry, in case the matter should go further I have concluded that the RR Entities would not have placed the orders they placed to stock the PAS, to the extent that they did not constitute Initial Provisioning, had they appreciated that they would have to pay WLP for those units.

There was no argument at trial as to what, if anything, further would have happened in that scenario – e.g., as to whether there would then have been a negotiation of some kind and, if so, with what outcome – and therefore it is not necessary for me to make any further findings. Once again, Goodrich’s attempts to obtain the factual findings for such a claim (in this case in post-hearing correspondence) must be rejected.

This relief was claimed in support of Goodrich’s claim for breaches of the Exclusivity Provisions. As that part of Goodrich’s claim has failed, no orders are required.

For those reasons: