Marsh Limited & Anor v Greensill Bank AG & Anor

The first claimant, Marsh Limited (‘Marsh’) seeks the continuation until trial of an interim anti-suit injunction (‘ASI’) restraining the Defendants, Greensill Bank A.G. (in Insolvency Administration) and its Insolvency Administrator, Dr Michael Frege, to whom I shall refer compendiously as ‘GBAG’, from bringing proceedings against Marsh in Australia. GBAG seeks the setting aside of the interim ASI on the grounds of non-disclosure and, in any event, opposes its continuation.

Marsh is a UK company which provides insurance broking services. GBAG is a bank incorporated in Germany. GBAG is currently in Insolvency Administration, following the collapse of the Greensill group, of which it forms part, in March 2021. GBAG was once a subsidiary, and later a sister company, of Greensill Capital (UK) Ltd (‘GCUK’), which is itself now in administration.

The business of GCUK included both ‘Supply-Chain Financing’, namely the financing of a customer’s ‘payables’; and ‘Accounts Receivable Financing’ which involved receivables being purchased by GCUK. GBAG invested in or financed some of this business; finance was also raised from third-party investors, by GCUK selling (assigning) the benefit of the receivables it purchased or its customer’s obligations in respect of payables. In some cases, to make the obligations originated by GCUK more attractive, they were backed by trade credit insurance under policies issued to Greensill entities.

Where trade credit insurance was required, Greensill had a parallel policy structure, meaning that, in relation to each risk to be insured, two trade credit policies were put in place – one policy (the ‘master’ policy) naming GCUK as insured; the other (the ‘parallel’ policy) naming GBAG as insured. GCUK acted as ‘servicer’ for GBAG, and was responsible for originating financings, arranging and paying for insurance cover, and making disclosures and representations to insurers. GCUK also had responsibility for paying a sum equivalent to any deductible back to the insurer on a claim (there being no deductible on GBAG’s insurance) and the substantive premium (GBAG’s policy requiring payment of only a nominal $1).

From 2014 onwards, Marsh provided insurance broking and other services to GCUK and GBAG under a series of Letters of Engagement, addressed to GCUK and signed by an officer of GCUK, incorporating standard form Terms of Engagement (the ‘GCUK Letters of Engagement’). The overwhelming bulk of Marsh’s communications and instructions in relation to the insurance it brokered for GCUK and for GBAG, both the master and the parallel policies, were with GCUK rather than GBAG.

Under the Terms of Engagement, there was a limitation of liability clause, capping Marsh’s liability for insurance broking services at £7.5 million (later increased to £10 million). There was also a governing law and jurisdiction clause:

“The Engagement and any non-contractual obligations arising out of or in connection with it shall be governed by and construed in accordance with English law and any disputes related thereto shall be subject to the exclusive jurisdiction of the English courts.”

The Terms of Engagement contained a further clause (the ‘Affiliates Clause’) providing:

In 2018, in addition to a GCUK Letter of Engagement signed that year, a separate Letter of Engagement was concluded by GBAG with Marsh (the ‘2018 GBAG LOE’). The 2018 GBAG LOE was signed for a period of 12 months commencing on 1 March 2018. It was in materially identical terms to the GCUK Letters of Engagement, including the Affiliates Clause and the exclusive jurisdiction clause, save that as regards remuneration the 2018 GBAG LOE stated: “Our remuneration is covered in the separate engagement agreement with Greensill Capital (UK) Ltd”.

Since the collapse of the Greensill group, there has been a good deal of litigation commenced, both in England and elsewhere. In particular, there is heavy and complex litigation under way before the Federal Court of Australia (the ‘Australian Proceedings’). This involves ten proceedings, being three actions commenced by Credit Suisse entities (the ‘Credit Suisse Proceedings’); a further action commenced by an entity known as White Oak; and six actions commenced by GBAG against Insurance Australia Ltd (‘IAL’) (these six actions being the ‘GBAG Proceedings’). All of these cases are being case managed together, with a three-month trial scheduled to begin in March 2026. In England, there are proceedings by GBAG against Zurich Insurance, and a claim by White Oak against Marsh, both issued in 2023. Disputes concerning the authority of various individuals to act for GBAG arise in the first of those sets of proceedings, as they do in the Australian Proceedings.

In the GBAG Proceedings, under certain Australian statutes, GBAG brings claims for (alleged) misleading and deceptive conduct (‘MDC claims’). Such claims are apportionable, in that liability can be reduced if a concurrent wrongdoer is identified, whether or not the identified concurrent wrongdoer is party to the proceedings. I am told that MDC claims can be brought on the basis of incorrect statements of fact, even without alleging fault on the part of the representor; loss is recoverable on a statutory basis which may go wider than the scope of recoverable loss in tort; and Australian law is unsettled as to whether contractual limitation of liability clauses will be effective to cap liability under an MDC claim. I am also told that it is uncertain whether an Australian court would give effect to an exclusive jurisdiction clause pointing away from that forum, if jurisdiction to bring an MDC claim were otherwise established.

In August 2023, IAL responded to the claims against it in the GBAG Proceedings by pleading as one of its defences that, if it had done wrong, there were other concurrent wrongdoers and its liability should be reduced proportionately. Marsh was named as a concurrent wrongdoer. However, Marsh is not a party to the claims brought against IAL.

On 5 September 2023, GBAG and Marsh entered into a confidential Standstill Deed to toll limitation periods in respect of certain claims that GBAG might have against Marsh, including claims relating to the GBAG Proceedings (the ‘Standstill Deed’).

On 7 November 2023, Marsh was joined as a respondent in the Credit Suisse Proceedings. From that date, it has been party to the Australian Proceedings, and has played an active part in them, but there has been no claim directly by GBAG against Marsh.

On 2 July 2024, Ms Michelle Fox of GBAG’s Australian trial lawyers wrote to Mr Christopher Foster of Marsh’s English solicitors communicating GBAG’s intention to join Marsh to the GBAG Proceedings. That letter enclosed a termination notice issued under the Standstill Deed. This letter caused consternation for Marsh. Marsh considered that GBAG was bound by the exclusive jurisdiction clauses within the Letters of Engagement signed between Marsh and GCUK. Marsh feared that if GBAG went ahead and sued Marsh in Australia, Marsh would be exposed to an MDC claim with a headline value in the billions of Australian dollars, and Marsh would risk not having the benefit of the contractual limitation of liability contained in the Terms of Engagement.

Marsh decided that the right course for it to take was to obtain an interim ASI from the English court. Marsh was concerned, however, that if it gave full notice of such application to GBAG, GBAG might first obtain from the Australian court an anti-anti-suit injunction, preventing Marsh obtaining an ASI, based on arguments that the ASI would disrupt the Australian Proceedings, or enable Marsh to evade liability under MDC claims. As a result, Marsh decided to seek an ASI on very short notice to GBAG, after all necessary preparatory steps had been taken.

Marsh was aware that, since it would be giving GBAG only very short notice of the interim ASI application, Marsh would have the duties of full and frank disclosure and fair presentation which are incumbent on any applicant proceeding without notice. In order to comply with what were understood to be Marsh’s duties, in preparation for the interim ASI application, Marsh’s lawyers reviewed over 14,000 documents, including documents which had been disclosed to Marsh by GBAG in the Australian Proceedings.

The without notice application went ahead on 30 July 2024, before Cockerill J. Cockerill J granted the interim ASI in favour of Marsh, restraining GBAG from taking any steps to initiate or bring a claim against Marsh in Australia, in relation to the Engagements contained in or evidenced by any of the Letters of Engagement, and any non-contractual obligations arising out of or in connection with those Engagements. Cockerill J rejected an application for an interim ASI in favour of Marsh Pty Limited, the second claimant, and that application is not renewed before me, so I need say no more about it.

After the interim ASI had been granted, but in advance of the return date before me, GBAG brought an application in the Australian Proceedings, contending that, in obtaining the without notice interim ASI before Cockerill J, Marsh had breached the obligation not to make collateral use of documents disclosed in those proceedings. That obligation – familiar to English litigants as the ‘implied undertaking’ following Harman v Secretary of State for the Home Dept [1983] 1 AC 280 (HL), and now in CPR 31.22 – is referred to by the parties as the Harman obligation, and the documents subject to the obligation as Harman documents, which is a terminology I shall adopt.

In response to GBAG’s application in Australia, Marsh argued that obtaining the interim ASI from the English court was for the purposes of the Australian proceedings, and not for a collateral purpose. Alternatively, Marsh argued that the Harman obligation yielded to the obligation to make full and frank disclosure in the course of the without notice application. Following a hearing on 11 November 2024, Moshinsky J, sitting in the Federal Court of Australia, rejected these arguments in a judgment on 12 November 2024. He concluded, to the ‘beyond reasonable doubt’ standard, that Marsh had breached the Harman obligation, constituting a contempt of the Australian court, and declared that:

“Marsh Ltd and Marsh Pty Ltd relied upon documents, and information contained in documents, discovered in the [Australian] Proceedings in support of the Anti-suit Application… in breach of their obligation to this Court not to use discovered documents or discovered information for any purpose other than that for which it was given unless it is received into evidence, without leave of this Court.”

Moshinsky J went on to order, however, that, from the date of his order, Marsh be released from the Harman obligation, in relation to the documents disclosed by GBAG to Marsh in the Australian Proceedings, such that those documents may now be used for the purposes of the ASI application. Moshinsky J declined to make his order retroactive (‘nunc pro tunc’).

I respectfully express my thanks to Moshinsky J for the expeditious and clear way in which he determined the questions arising in relation to the Harman documents. His decisions enabled the issues before me to be significantly narrowed.

The court has power to grant an ASI under s.37(1) of the Senior Courts Act 1981. While the statutory language requires only that it be “just and convenient” to grant an injunction, it is well established that there are two main bases on which such an injunction might be granted – being (i) the ‘contractual basis’ where, normally by reason of a contractual jurisdiction clause, the applicant has a legal or equitable right not to be sued in a foreign court, or (ii) where the commencement or continuance of foreign proceedings would be vexatious or oppressive.

The parties are agreed that a convenient summary of the law can be found in the decision of Foxton J in QBE Europe SA/NV v Generali Espana de Seguros y Reaseguros [2022] EWHC 2062 (Comm) (‘QBE’). Foxton J held:

“10.

I was referred to Jacobs J's summary of the key principles which govern the grant of anti-suit relief in this wholly contractual context in AIG Europe SA v John Wood Group plc [2021] EWHC 2567 (Comm), [58] (which, to the extent it was in issue, was approved and further explained by Males LJ on appeal, [2022] EWCA Civ 781; [2022] 2 CLC 124, [10] ). The principles so summarised are as follows:

(i)

The court's power to grant an ASI to restrain foreign proceedings, when brought or threatened to be brought in breach of a binding agreement to refer disputes to arbitration, is derived from section 37(1) of the Senior Courts Act 1981, and it will do so when it is 'just and convenient'.

(ii)

The touchstone is what the ends of justice require.

(iii)

The jurisdiction to grant an ASI should be exercised with caution.

(iv)

The injunction applicant must establish with a 'high degree of probability' that there is an arbitration or jurisdiction agreement which governs the dispute in question.

(v)

The court will ordinarily exercise its discretion to restrain the pursuit of proceedings brought in breach of a forum clause unless the defendant can show strong reasons to refuse the relief (relying on Aggeliki Charis Compania Maritima SA v Pagnan SpA (The Angelic Grace) [1995] 1 Ll Rep 87).

(vi)

The defendant bears the burden of proving there are strong reasons.

11.

By way of further elaboration of those last two points:

(i)

It has been held that respect for comity is not a strong reason for the court not to give effect to a contractual choice of forum clause, and that comity requires that where there is an agreement for a sole forum for the resolution of disputes under a contract, that agreement is respected: Males LJ in AIG Europe, [8]. By way of parenthesis, in that context, comity is served by applying the same respect to choice of court or arbitration agreements in favour of other jurisdictions and arbitral seats.

(ii)

It has been held that the existence of a mandatory provision of foreign law applicable in the foreign court which overrides the contractual choice of jurisdiction is not a strong reason to refuse an ASI: Shipowners' Mutual Protection & Indemnity Association (Luxembourg) v Containerships Denizcilik Nakliyat ve Ticaret (The Yusuf Cepnioglou) [2016] EWCA Civ 386; [2016] 1 CLC 687, [34] –[37] and [57]–[58] and Thomas Raphael QC, The Anti-Suit Injunction (2nd edn) ( Raphael ), [8.31]–[8.44].”

GBAG opposes the continuation of the interim ASI on three grounds.

I will turn to each of these issues in turn.

It is now conceded by Marsh that there was a breach of the duty of full and frank disclosure.

The test whether that duty has been breached is an objective one and, in light of the decision of Moshinsky J of 12 November 2024 (see paragraph 19 above), it is accepted by Marsh that the use of Harman documents gave rise, at the least, to the real possibility that Marsh was acting in breach of its duties to the Australian court. That was a significant procedural matter which should have been brought to the attention of the court at the without notice hearing. Indeed, although Marsh has indicated an intention to appeal the ruling of Moshinsky J, Marsh accepts that for present purposes I should proceed on the basis that the use of Harman documents was an actual breach of such duties – i.e., not merely a possibility – which should have been disclosed to Cockerill J. Marsh submits, however, in light of expert evidence which has been served, that I should accept that there are respectable arguments to the contrary, which I do accept.

In advance of the hearing before me, Marsh filed two witness statements from Mr Foster, his second and his fourth statements, which dealt exclusively with the use of Harman documents and the argument that there had been a breach of Marsh’s duties on the without notice application. The facts stated in this evidence are not challenged by GBAG, and I accept the evidence as true.

Mr Foster is an English solicitor, and has no qualification under Australian law. Nonetheless, when the letter terminating the Standstill Deed came to his attention on 2 July 2024, he already had some familiarity with the Harman obligation under Australian law since Harman undertaking-related issues had arisen in the White Oak proceedings in England from around March 2024. In the context of making the without notice application for an interim ASI, Marsh also obtained further advice on the Harman obligation from Australian lawyers, but privilege in that advice has not been waived and so I do not know what that advice contained.

Mr Foster was well aware of the Harman obligation, but knew that there was a doctrine under which, in appropriate circumstances, such obligation would ‘yield’. Before instructing the review of Harman documents, Mr Foster considered and re-read relevant Harman-related passages from six Australian cases, most of which he had already read in full earlier in the year. In particular, he read paragraphs 71 to 74 of Hancock Prospecting Pty Ltd v DFD Rhodes Pty Ltd (No.2) [2023] WASCA 108, and considered that this provided a clear answer. This passage reads (with references omitted):

“Accordingly, the Harman obligation arises by operation of law. It is an obligation to the court. It is not absolute; nor is it unalterable. As to the first point (ie that the Harman obligation is not absolute), it is accepted that the Harman obligation is not unqualified. The obligation no longer applies where the material has been received into evidence. Similarly, the obligation does not apply to a document that has been read or referred to in open court in a way that discloses its contents. More generally, as was explained by Mason CJ in Esso Australia Resources Ltd v Plowman, the obligation may yield:

‘It would be inequitable if a party were compelled by court process to produce private documents for the purposes of the litigation yet be exposed to publication of them for other purposes. No doubt the implied obligation must yield to inconsistent statutory provisions and to the requirements of curial process in other litigation, eg discovery and inspection, but that circumstance is not a reason for denying the existence of the implied obligation. (emphasis added)’

In other words, the implied obligation is no answer to otherwise valid processes of law in other proceedings. Sometimes such processes will be compulsive. But they may also be permissive. Nothing in Mason CJ's statement confines 'the requirements of curial process' to mandatory orders that compel a litigant to take certain steps. The requirements of curial process are determined by what is required to do justice between parties in proceedings before the court. It will include, in an appropriate case, according litigants the ability to use documents and other materials that they are otherwise prohibited from using by reason of the Harman obligation.

Such a 'yielding' (sometimes referred to as an 'over-riding' of the implied obligation) is consistent with the rationale for the Harman obligation. The yielding is explicable on the basis that the disclosure and use of the documents or information otherwise the subject of the implied obligation in the first proceedings is required in the interests of justice in the subsequent proceeding. In this respect the Harman obligation is necessarily subject to other requirements of the law.

Importantly, the implied obligation yields to the curial processes of courts other than the court to whom the obligation is owed. There are, as will be seen (see [80] below), numerous cases in which a second court has made an order for discovery which has the practical effect of over-riding a Harman obligation arising in a different court or tribunal. In this respect it is not necessary for the second court making the discovery order to await an order of the first court or tribunal releasing the obligation – or for the second court to itself make an order dispensing with or modifying the obligation.”

Mr Foster noted that none of the Australian cases had analysed exactly what “requirements of curial process” meant. However, since the rationale for the ‘yielding’ doctrine described in Hancock was to ensure justice was done, and this was the same rationale for the English requirements of full and frank disclosure, he thought he was “clearly required” to review the Harman documents and disclose to the English court any information materially relevant to the without notice application. As Mr Foster put it in his second witness statement:

“I did not believe there was any credible argument to the contrary. I remember feeling glad that the position on Harman was clear, but daunted by the work which would be necessary, as I viewed it as inevitable that the Defendants would be the first to challenge a failure to ensure that any prejudicial Australian documents had been disclosed and there was a large number of documents to review.”

It suited Marsh for Mr Foster to have reached this conclusion:

“I understood that an application could, in theory, be made to the original Australian court to release or modify the undertaking, and permit ‘use’ of the documents, but I understood that this was not obligatory. This was never a practical option for Marsh, partly due to timing issues, but also due to the risk of the Defendants becoming aware of Marsh’s intent to apply in England and Wales.”

Mr Foster has now apologised, on his own behalf and on behalf of Marsh, to the Australian court for breaching the Harman obligation. Mr Foster and Marsh, together with Counsel for Marsh, have also apologised to this court for failing to disclose the breach at the without notice application.

In Tugushev v Orlov [2019] EWHC 2031 (Comm) at [7], in a summary approved by the Court of Appeal in Derma Med Ltd v Ally [2024] EWCA Civ 175 at [29], Carr J distilled the following general principles from the relevant authorities:

“i)

The duty of an applicant for a without notice injunction is to make full and accurate disclosure of all material facts and to draw the court's attention to significant factual, legal and procedural aspects of the case;

ii)

It is a high duty and of the first importance to ensure the integrity of the court's process. It is the necessary corollary of the court being prepared to depart from the principle that it will hear both sides before reaching a decision, a basic principle of fairness. Derogation from that principle is an exceptional course adopted in cases of extreme urgency or the need for secrecy. The court must be able to rely on the party who appears alone to present the argument in a way which is not merely designed to promote its own interests but in a fair and even-handed manner, drawing attention to evidence and arguments which it can reasonably anticipate the absent party would wish to make;

iii)

Full disclosure must be linked with fair presentation. The judge must be able to have complete confidence in the thoroughness and objectivity of those presenting the case for the applicant. Thus, for example, it is not sufficient merely to exhibit numerous documents;

iv)

An applicant must make proper enquiries before making the application. He must investigate the cause of action asserted and the facts relied on before identifying and addressing any likely defences. The duty to disclose extends to matters of which the applicant would have been aware had reasonable enquiries been made. The urgency of a particular case may make it necessary for evidence to be in a less tidy or complete form than is desirable. But no amount of urgency or practical difficulty can justify a failure to identify the relevant cause of action and principal facts to be relied on;

v)

Material facts are those which it is material for the judge to know in dealing with the application as made. The duty requires an applicant to make the court aware of the issues likely to arise and the possible difficulties in the claim, but need not extend to a detailed analysis of every possible point which may arise. It extends to matters of intention and for example to disclosure of related proceedings in another jurisdiction;

vi)

Where facts are material in the broad sense, there will be degrees of relevance and a due sense of proportion must be kept. Sensible limits have to be drawn, particularly in more complex and heavy commercial cases where the opportunity to raise arguments about non-disclosure will be all the greater. The question is not whether the evidence in support could have been improved (or one to be approached with the benefit of hindsight). The primary question is whether in all the circumstances its effect was such as to mislead the court in any material respect;

vii)

A defendant must identify clearly the alleged failures, rather than adopt a scatter gun approach. A dispute about full and frank disclosure should not be allowed to turn into a mini-trial of the merits;

viii)

In general terms it is inappropriate to seek to set aside a freezing order for non-disclosure where proof of non-disclosure depends on proof of facts which are themselves in issue in the action, unless the facts are truly so plain that they can be readily and summarily established, otherwise the application to set aside the freezing order is liable to become a form of preliminary trial in which the judge is asked to make findings (albeit provisionally) on issues which should be more properly reserved for the trial itself;

ix)

If material non-disclosure is established, the court will be astute to ensure that a claimant who obtains injunctive relief without full disclosure is deprived of any advantage he may thereby have derived;

x)

Whether or not the non-disclosure was innocent is an important consideration, but not necessarily decisive. Immediate discharge (without renewal) is likely to be the court's starting point, at least when the failure is substantial or deliberate. It has been said on more than one occasion that it will only be in exceptional circumstances in cases of deliberate non-disclosure or misrepresentation that an order would not be discharged;

xi)

The court will discharge the order even if the order would still have been made had the relevant matter(s) been brought to its attention at the without notice hearing. This is a penal approach and intentionally so, by way of deterrent to ensure that applicants in future abide by their duties;

xii)

The court nevertheless has a discretion to continue the injunction (or impose a fresh injunction) despite a failure to disclose. Although the discretion should be exercised sparingly, the overriding consideration will always be the interests of justice. Such consideration will include examination of i) the importance of the facts not disclosed to the issues before the judge ii) the need to encourage proper compliance with the duty of full and frank disclosure and to deter non-compliance iii) whether or not and to what extent the failure was culpable iv) the injustice to a claimant which may occur if an order is discharged leaving a defendant free to dissipate assets, although a strong case on the merits will never be a good excuse for a failure to disclose material facts;

xiii)

The interests of justice may sometimes require that a freezing order be continued and that a failure of disclosure can be marked in some other way, for example by a suitable costs order. The court thus has at its disposal a range of options in the event of non-disclosure.”

Without challenging the truthfulness of Mr Foster’s evidence, Mr Fealy KC on behalf of GBAG made a number of criticisms of Mr Foster’s conduct. In summary, he submitted:

Mr Fealy KC submitted that the interim ASI should be set aside and not renewed. There had a been a breach of the duty of full and frank disclosure; the non-disclosure was both culpable and substantial; and setting aside the interim ASI was the appropriate response.

I find that the failure to disclose Marsh’s contempt of the Australian court in using Harman documents – both by reviewing those documents in advance of making the without notice application for an interim ASI, and by deploying those documents in the course of making the application – was a serious and substantial breach of Marsh’s duties to the English court.

In terms of culpability, the focus of GBAG’s arguments was, in reality, not so much on Mr Foster’s failure to disclose the breach of the Harman obligation to the English court, but on his prior failure to identify that there might be a breach at all. While Mr Foster characterised his conduct as showing “great care”, GBAG disputed that characterisation, noting that Mr Foster is not an Australian lawyer and that his research necessarily extended only to Australian cases which were not directly on point. GBAG submitted that since Marsh was not waiving privilege in the Australian legal advice it had taken, Marsh could not rely on this as showing care by Mr Foster.

In my judgment, Marsh is entitled to rely on the fact that Australian legal advice was taken, even without relying on the content of such advice. The fact of taking such advice shows that Mr Foster’s conclusion was not reached recklessly without caring whether he was right or wrong. Mr Foster read cases, he turned his mind to the issue, he had regard to relevant legal advice, and he reached a decision.

That is not to say, however, that I conclude that Mr Foster’s failure was entirely innocent. Of course, lawyers will on occasion turn their mind to a legal issue and innocently reach the wrong conclusion; but here Mr Foster’s failing lay in his unjustifiable belief that his conclusion was without “any credible argument to the contrary”. Mr Foster is not an Australian lawyer (and I cannot assume or infer that he received unqualified Australian legal advice to support his belief); the six cases he had reviewed were not directly on point; and Mr Foster had himself identified that none of the Australian cases had analysed exactly what “requirements of curial process” meant. Mr Foster’s conclusion was over-confident and, on the basis of the decision of Moshinsky J, wrong. In my judgment, Mr Foster should have been more aware of the risk of confirmation bias – that he was unwittingly being influenced by the knowledge that this conclusion suited his client: compare paragraph 32 above – and of his own limitations as an English solicitor when reaching a view as to the content of Australian law.

Moving on from the question of culpability, it is right (as held in Tugushev v Orlov at [29(ix)]) that Marsh should be deprived of any advantage it has achieved by its misconduct. This led to a debate between the parties as to whether any material advantage had been gained by Marsh.

I consider it is appropriate for me to have regard to the range of these counterfactuals as part of my evaluative judgment, rather than specifically deciding between them. In particular, I find GBAG’s proposed counterfactual somewhat artificial, especially bearing in mind that GBAG’s submissions on culpability focused primarily on a different aspect of Marsh’s conduct – i.e., the conclusion reached by Mr Foster, rather than the non-disclosure which resulted from that conclusion. In circumstances where Marsh’s breaches of duty (both to the Australian court and to the English court) were never deliberate, it is difficult to envisage a real counterfactual in which Marsh would have identified that using the Harman documents was a contempt of the Australian court, but pressed on regardless.

On the counterfactuals proposed by Marsh, I accept that the likelihood is that Marsh would still have obtained an interim ASI from Cockerill J on a without notice basis. There is no simple or straightforward answer to what a party is supposed to do when its duties of full and frank disclosure collide with other duties (compare, in the context of without prejudice privilege, Linsen International Ltd v Humpuss Sea Transport Pte Ltd [2010] EWHC 303 at [51]-[55]), but, where an applicant otherwise makes out a sufficient case for an interim ASI, a court is likely to be sympathetic where the applicant has acted responsibly in considering and appropriately managing any conflicting duties it may have.

On the other hand, I accept that on GBAG’s (somewhat artificial) counterfactual, Cockerill J would have had real concern at the prospects of granting an order based on an admitted and ongoing contempt of the Australian court. But, the mere fact (or hypothesis) that Marsh would not have obtained an interim ASI on 30 July 2024 does not, of itself, mean that GBAG would then be in a more advantageous position now. That submission depends on what would have happened next. Again – a range of possibilities arise, but two obvious ones may be considered:

Overall, I consider that the features of this case warrant the exercise of my discretion not to set aside the interim ASI, but rather to consider on their merits the substantive questions relating to its continuation. I recognise that such a discretion is to be exercised sparingly where there has been a significant breach, however, in my judgment, that is where the interests of justice lie in this case. While, as I have indicated, I find that the breach was not wholly innocent, and it may be that Marsh obtained an advantage by its breach, I have regard also to the following considerations:

It is common ground that an applicant for an interim ASI must establish with a “high degree of probability” that there is an arbitration or jurisdiction agreement which governs the dispute in question: see QBE at [10(iv)], cited at paragraph 23 above.

The rationale for this high threshold appears to be two-fold. In part, it is because interim ASIs tend to be determinative of the question in issue: here, for example, if I were to grant an interim ASI until trial of the claim for a permanent ASI, and the court at trial were to reject the claim, then it would likely be too late for GBAG to join Marsh into the Australian Proceedings. In part, the rationale is because an interim ASI does not merely interfere with the private rights of the respondent, but may interfere, albeit indirectly, with the working or output of a foreign court, raising issues of comity. See: Transfield Shipping Inc v Chiping Xinfa Huayu Alumina Co Ltd [2009] EWHC 3629 (QB) (Christopher Clarke J) (‘Transfield’) at [51]-[55]; Ecobank Transnational Inc v Tanoh [2015] EWCA Civ 1309, [2016] 1 Lloyd’s Rep 360 (‘Ecobank’) at [89]-[91] (Christopher Clarke LJ); The Anti-Suit Injunction (2^nd ed, Raphael 2019) at ¶13.37.

However, despite the ample authority in support of the ‘high degree of probability’ threshold, there appears to be scant authority as to what that threshold actually means. In Midgulf International Ltd v Groupe Chimiche Tunisien [2009] EWHC 963 (Comm), [2009] 2 Lloyd’s Rep 411 (‘Midgulf’), Teare J held that the claimant’s “strongly arguable” case was not sufficient to meet the threshold. In Transfield, Christopher Clarke J similarly held that a “good arguable case” was not sufficient.

Christopher Hancock KC, sitting as a Judge of the High Court in Tyson International Co Ltd v GIC Re, India, Corporate Member Ltd [2024] EWHC 236 (Comm) (‘Tyson’), considered this issue. At [97] he held:

“Both Counsel were agreed that TICL had to show a high probability that there was a binding jurisdiction clause, and both Counsel agreed that this meant something more than a good arguable case. However, beyond this, neither Counsel was able to be of much assistance as to what that phrase meant, and how it should be applied in a case such as the present where there was a competition between a jurisdiction and an arbitration clause. Although Mr MacDonald Eggers KC suggested that this meant more than establishing the existence of the clause on the balance of probabilities, and that I would have to be satisfied that there was no real prospect that there was a binding arbitration clause, he also accepted that he could cite no authority in support of this proposition, and for my part, in the absence of authority, I would not accept that submission.”

Where, in other contexts, interim relief is sought which is likely to be determinative, it is established that the court will need a “high degree of assurance” before granting such relief. In Koza Ltd v Koza Altin Isletmeleri AS [2020] EWCA Civ 1018 (‘Koza’), Popplewell LJ held at [77]:

“Nevertheless the inability to resolve the issue between these parties in this jurisdiction is not, to my mind, fatal to the existence or exercise of the ancillary jurisdiction. The court has developed principles catering for just such a situation when exercising its original jurisdiction to grant interim injunctions. Cases not infrequently arise of ‘interim’ injunction applications where the circumstances mean that the grant or refusal of relief will in practice be finally determinative. In such situations the court does not say that it has no power to restrain a threatened invasion of a disputed right simply because there will never be a final determination of that issue. Rather it recognises that the grant or refusal of the injunction will be a permanent and unjustified invasion of one party's rights, and so grants or refuses an injunction on the basis of the least irremediable prejudice, recognising that there is a heightened emphasis on the merits of the claim and that the court may need to have a high degree of assurance that the threatened conduct is an actionable invasion of the claimant's rights. It is not necessary to cite extensive authority for this well-known practice and the applicable principles: see, for example: NWL Ltd v Woods [1979] 1 WLR 1294; Lansing Linde Ltd v Kerr [1991] 1 WLR 251; and Forse v Secarma Ltd [2019] IRLR 587...”

I note that in NWL Ltd v Woods, the language of Lord Diplock was “high degree of probability” (at 1307H). This appears to have been the origin of the “high degree of probability” threshold in the context of interim ASIs. That formulation was first used by Colman J in Bankers Trust Co v PT Mayora Indah (unrep, 20 January 1999), and followed by Cresswell J in Bankers Trust Co v PT Jakarta International Hotels and Development [1999] 1 All ER (Comm) 785, 788h-789b, where Cresswell J also cited The Supreme Court Practice 1999, vol 1, para 29/L/15, which itself referred to NWL Ltd v Woods.

In Koza at [78], Popplewell LJ distinguished between cases where a court is “sure” and those where it has a “high degree of assurance”. In British Air Line Pilots Association v British Airways Cityflyer Ltd [2018] EWHC 1889 (QB) at [30], Butcher J said of the ‘high degree of assurance’ test, in the context of an application for an interim declaration as to the contractual rights of the parties which would be revisited at trial: “I consider that that test can hardly be different, or not markedly different, from the question of whether there should be summary judgment”. By contrast, in Lenovo Group Ltd v Telefonaktiebolaget LM Ericsson [2024] EWHC 2941 (Pat), Richards J held at [12]-[13] that where a particular declaration was sought, pending a FRAND trial, which would not be revisited, he was bound by Court of Appeal authority in Panasonic Holdings Corporation v Xiaomi Technology UK Ltd [2024] EWCA Civ 1143 to hold that the ‘high degree of assurance’ test did not require the applicant to establish its case to the summary judgment standard.

In addition to a lack of clarity as to the meaning of the threshold, there is also some uncertainty in the language of the authorities as to whether the court should ask itself whether, on the materials before the court at the hearing of the interim ASI application, there is (currently) a high degree of probability that there is an exclusive jurisdiction clause covering the dispute in question; or, whether the court should, in a process similar to that adopted in a summary judgment application under CPR 24, also have regard to evidence that can (or could) reasonably be expected to be available at trial. The difference may be substantial where, for example, there is known to be a corpus of documents that the parties have not yet had the opportunity to review. In Transfield at [52] Christopher Clarke J said the test was “whether or not the applicant has shown on the material adduced at the interlocutory hearing a high degree of probability that there was such an agreement”; by contrast, in Clearlake Shipping Pte Ltd v Xiang Da Marine Pte Ltd [2019] EWHC 2284 (Comm) at [19], Andrew Burrows QC, sitting as a Judge of the High Court, referring to Transfield, described that test as “a ‘high degree of probability’ that it would establish at trial that there was a binding jurisdiction agreement” – the ‘would’, it seems, signalling a recognition that the trial might well never happen.

Counsel for GBAG and for Marsh were agreed that the ‘high degree of probability’ test meant something beyond ‘more likely than not’. Beyond that, neither was willing to gloss the test, although Mr Edwards KC for Marsh said that it would be wrong to apply the summary judgment test – i.e., wrong to require the claimant to show there was ‘no real prospect’ of the claim failing at trial.

I do not find it easy to reconcile the authorities on ‘high degree of probability’ and ‘high degree of assurance’. However, I remind myself that the statutory requirement is that an ASI be just and convenient, and “the touchstone is what the ends of justice require”: QBE at [10(ii)]. I accept, accordingly, that it is not a hard and fast rule that the applicant must meet the summary judgment test. But, in my judgment, given the twin rationales for the test in this context, a ‘high degree of probability’ in the context of an application for an interim ASI will usually mean something close to that standard: the court should not usually proceed to determine the matter in the claimant’s favour unless the court considers that the defendant would have no real prospect of defeating the application at trial. In reaching that conclusion, the court should have regard to evidence that can (or could) reasonably be expected to be available at trial.

One way in which courts have dealt with applications where the case for an interim ASI meets the threshold of a ‘good arguable case’, but does not meet the ‘high degree of probability’ threshold, has been to give directions for the trial of the issue, and to continue the interim ASI to ‘hold the ring’ until that trial takes place. This was the course taken in Tyson, in Midgulf, in Impala Warehousing and Logistics (Shanghai) Co Ltd v Wanxiang Resources (Singapore) Pte Ltd (see [2015] EWHC 811 (Comm) at [9]), and in CNA Insurance Company Ltd v Office Depot International (UK) Ltd (see [2005] EWHC 456 (Comm) at [20]). While none of these authorities discuss the appropriate threshold to be met by the applicant in such a case, the learned author of The Anti-Suit Injunction observes at ¶13.48:

“Where the interim anti-suit injunction is sought only to hold the ring for a shorter period of time pending a further hearing or trial of the injunction, or where there will be sufficiently early trial of a final anti-suit injunction, so that the grant of an interim injunction will not be practically determinative of the question of forum, reasoning akin to Cyanamid is more appropriate. It may then be appropriate to grant relief on the basis of there being a sufficient probability of success, pending that further hearing or trial. In such a case, a trial of the final injunction may be accelerated.”

In the present case, Marsh did not in its application notice, nor in the evidence served in support of its application, seek an injunction to ‘hold the ring’ pending a fuller consideration of the issues at trial, nor any expedition of such trial. Nor did Marsh suggest that the threshold it would need to meet at this hearing was anything other than the ‘high degree of probability’ threshold. However, in footnote 34 to their skeleton, Counsel for Marsh observed:

“Current Commercial Court lead times indicate that a one-week trial could be fixed for May 2025, even without expedition. The Claimants would not object to an order for a speedy trial of the claim for a final anti-suit injunction, were the court to consider that appropriate. There is some precedent for that being done: see Impala Warehousing v Wanxiang Resources [2015] EWHC 811 (Comm) at [9] (Blair J).”

During his oral submissions, I explored with Mr Edwards KC whether Marsh was seeking an injunction to hold the ring pending trial, whether expedited or otherwise. Mr Edwards KC submitted that this was a course open to me.

This course was, however, opposed by Mr Fealy KC who drew to my attention the evidence served by his client, relying on the opinion of its Australian lawyer, Ms Fox, that “there is a real risk that the [Australian court] may not permit the joinder of Marsh Limited to the Australian Proceedings after January 2025 (with this risk naturally increasing with any further delay)”. This was said to be so even though Marsh has told the (English) court that it would not resist its joinder to a claim by GBAG in the Australian Proceedings on the grounds of lateness. This evidence is supported by an explanation of the pre-trial procedure in the Australian Proceedings. Mr Edwards KC submitted that these fears were overblown.

Overall, I do not consider that this is a case where I can or should grant an injunction to ‘hold the ring’ pending trial, applying some lower threshold than the ‘high degree of probability’ threshold. If the proposal had been made when the application was first issued, then the parties could have served evidence specifically addressing the procedure to be adopted in England, and the impact that might have on the Australian Proceedings. As it is, all I have is the evidence of Ms Fox’s opinion, which I have no reason to doubt. On that basis, even assuming that a trial of the final injunction could be expedited to, say, February or March 2025, I cannot be sure that a decision at that stage would not come too late to enable GBAG to join Marsh as a respondent in the GBAG Proceedings, or not without real inconvenience to the Australian court. It seems to me, accordingly, that the twin rationales underlying the ‘high degree of probability’ threshold apply here, and that this is the threshold I should apply.

That brings me to the substantive issue, which is whether, when GCUK signed various GCUK Letters of Engagement from 2014 onwards, it had actual or apparent authority to bind GBAG to those GCUK Letters of Engagement; or, whether GBAG ratified such agreements. (As noted above, there is no dispute that GBAG is bound by the 2018 GBAG LOE.)

The general principles on actual and ostensible (apparent) authority are not in dispute, having been helpfully summarised in The Law Debenture Trust Corpn plc v Ukraine [2023] UKSC 11 at [38]-[42].

As regards actual authority, Marsh accepts that it cannot point to a specific communication whereby a representative of GBAG conferred authority on GCUK to bind GBAG to broking contracts with Marsh. However, Marsh submits that I can infer, to a high degree of probability, that GBAG had conferred actual authority – whether by words or by conduct – by a combination of factors.

Marsh points out that in the Australian Proceedings it is accepted by GBAG, or positively averred, that GBAG was party to insurance policies which had been brokered by Marsh or an affiliate of Marsh. Marsh also says that, with GBAG’s knowledge, GCUK gave Marsh instructions to bind cover on behalf of GBAG. From this, Marsh argues that it can be inferred that GBAG also gave GCUK authority to bind GBAG in dealings with Marsh or, alternatively, that authority to bind GBAG in dealings with Marsh was a necessary incident of such authority. Marsh says that GBAG cannot have thought it was receiving the benefit of Marsh’s services on a non-contractual basis, and so must have authorised GCUK to contract with Marsh on its behalf.

I do not find this line of argument persuasive.

The Service Level Agreements (SLAs) between GCUK and GBAG contain no indication of GCUK being given such authority. On the contrary, the basic understanding of the relationship between these parties being at arm’s-length is set out in paragraph 1:

“This SLA governs the main feature of the working relationship between GCUK and [GBAG]. The aim is to have a clearly structured and long-term business partner relationship as well as an unambiguous basic understanding of the nature of the ‘SCF’ product. In the main, this concerns the refinancing of the SCF business acquired by GCUK with commercial customers. GCUK shall provide trust holdings to [GBAG]. The basic understanding of the working relationship features the ‘at-arms-length principle’, so that GCUK and [GBAG] may operate as separate businesses and are not a group of companies.”

As against this, Marsh points to the Affiliates Clause (see paragraph 7 above), arguing that it is inherently improbable that GCUK would sign up to Terms of Engagement including a contractual acceptance that it was accepting the Engagement on behalf of its affiliates, if it did not in fact have such authority. That argument does support Marsh’s case, but I do not place great weight on it in circumstances where is no evidence before me that anyone at GCUK actually turned their mind to this clause. The Affiliates Clause formed part of Marsh’s standard terms, and was not open to negotiation: the evidence of Mr Julian Macey-Dare, who was the main point of contact and relationship manager for Marsh on the Greensill account for much of the relevant period, was that, when negotiating the Letters of Engagement he did not have the ability to negotiate any of Marsh’s standard terms.

At the request of GBAG’s German lawyers, GBAG’s employees conducted a document review exercise over GBAG’s internal records covering a period from January 2014 to March 2021. From that search, according to the evidence of GBAG’s solicitor, Mr Edward Greeno, it appears that GBAG was never sent copies of the GCUK Letters of Engagement concluded in 2015, 2016, 2017, 2019 or 2020. The 2014 GCUK Letter of Engagement was different, in that it was circulated in draft to Mr Jason Austin who was a director of GCUK, but also on the supervisory board of GBAG. The earliest evidence in GBAG’s records of anyone at GBAG being conscious of the Affiliates Clause was only in October 2020. In my judgment, this evidence of GBAG’s very limited knowledge or understanding of the GCUK Letters of Engagement points away from GBAG having authorised GCUK to bind GBAG to those agreements.

As regards the 2018 GBAG LOE, the genesis of this is not clear. On 4 April 2018, Ms Birte Kuhlmann of GBAG emailed Ms Sophie Dyke of GCUK, asking: “I was asked for the insurance agents agreement signed between GCUK and Marsh and between GB and Marsh. Are those documents available to you and if so, would you be so kind to send them over to me, please?”. Later that day, Ms Dyke responded: “Please see the GCUK engagement letter attached. John Whelan has advised there may not be one for GB. Please check this with Danyon”. On 18 April 2018, Mr Macey-Dare sent an email, attaching a Letter of Engagement addressed to Mr Danyon Lloyd of GBAG, “As requested, Danyon. Please counter sign a copy for our records? Any questions, please let me know?”. Mr Lloyd proposed different language for the remuneration clause, to which Mr Macey-Dare replied “Of course sir”. The amendment was made as requested, and the 2018 GBAG LOE was then concluded on 18 April 2018, with effect as of 1 March 2018. Mr Macey-Dare has given evidence that he does not recall any details about these events. There is no evidence from anyone else involved.

Marsh says that the 2018 GBAG LOE demonstrates a willingness of GBAG to contract directly with Marsh. That is fair, but in my judgment that lends no support to – indeed, it undermines – Marsh’s case. In the absence of any other explanation, Marsh concluding an agreement directly with GBAG is inconsistent with Marsh’s argument that GBAG was already bound by the GCUK Letter of Engagement covering the same period. It is striking that, on the materials I have seen, neither Mr Macey-Dare nor anyone else said anything to the effect of there being no purpose to GBAG signing the 2018 GBAG LOE, since it was already bound into the GCUK Letters of Engagement. Moreover, the initial email in the chain – referring to agreements “signed between GCUK and Marsh and between GB and Marsh” and requesting “those documents”, in circumstances where, on any view, there had never been an agreement signed between Marsh and GBAG – suggests that, at least within GBAG, people had not previously turned their minds to the nature of the relationship with Marsh.

Marsh seeks to infer GBAG’s authorisation of GCUK from the knowledge and action, or inaction, of various individuals associated with GBAG. I find none of this persuasive. The evidence is that GBAG’s CEO in 2014, Mr Lloyd, was present at the first meeting between Greensill and Marsh, but I cannot infer from that bare fact that GBAG gave authority to GCUK to bind it to the contract GCUK concluded with Marsh some months later. Mr Jonathan Lane, who held the title of Group General Counsel, knew of the terms of the GCUK Letters of Engagement, and had some involvement in amending their terms, but Mr Lane was not an officer of GBAG. (Mr Lane did, at least once, sign a document on behalf of GBAG, but I cannot safely draw any general inference from that fact as to any ongoing role within GBAG.) Marsh relies on the knowledge and involvement of Mr Austin, described as a co-founder of the Greensill business, who in addition to being a director of GCUK was also on the supervisory board of GBAG. But, Mr Austin’s involvement is explained by his directorship of GCUK and, under German law, being a member of the supervisory board did not give Mr Austin any executive powers entitling him to bind GBAG.

Marsh relies on contemporaneous communications in which Marsh described itself as “broker” to GBAG, without demur from representatives of GBAG. However, a few isolated presentations or communications over the course of a seven year period are of limited assistance to the question I must decide, especially since the real issue here is not whether Marsh provided broking services to GBAG, but whether or not it did so under an agreement signed by GCUK with authority to bind GBAG. Moreover, there are other documents which point in the opposite direction, such as the ‘insurance disclosure pack’ provided by GCUK to Mr Macey-Dare and others in September 2020. That included a passage reading (with typographical errors corrected, and emphasis added):

“[GCUK] is predominantly the originator of assets that it wishes to be insured and is the party that has the relationship with Marsh and the insurers, and [GBAG] (and various other parties that take the benefit of [GBAG]’s insurances as loss payees, as described below) is an investor and is provided with the insurance which [GCUK] through Marsh arranges.”

Taking all the matters in the round, I am not satisfied to a high degree of probability that GCUK had actual authority to bind GBAG to the GCUK Letters of Engagement. On the contrary, on the materials I have seen, albeit without making any finding of fact at this interim stage, it seems to me more likely than not that GCUK did not have such authority.

The ostensible (or apparent) authority case was not abandoned by Marsh, but in his oral submissions Mr Edwards KC was content to rely on his skeleton argument. However, I find the ostensible authority argument to be even weaker than the actual authority argument.

GCUK relies on the same factual matters as those on which it relies to support its case on actual authority. But, in my judgment, there is no evidence there of a clear representation or ‘holding out’, express or implied, by someone on behalf of GBAG, that GCUK had authority to bind GBAG to the GCUK Letters of Engagement.

Moreover, even if some representation to that effect could be identified, the evidence that anyone at Marsh relied on any such representation is weak. Mr Macey-Dare’s evidence (untested, at this stage) is that he “understood GCUK and its employees were able to give instructions on behalf of both entities”, and “I never had any doubt that GCUK was speaking on behalf of both GCUK and GBAG”, but that appears to have been on the basis that “I cannot recall anyone associated with Greensill group ever suggesting that GCUK was not able to speak for GBAG in this way”. His rationale for believing that the GCUK Letters of Engagement were binding on GBAG seems to have been based only on an understanding of what the Affiliates Clause was designed to achieve, not any holding out by GBAG.

Accordingly, I find that Marsh has not shown to a high degree of probability that GCUK had ostensible authority to bind GBAG to the GCUK Letters of Engagement.

Finally, turning to the question of ratification, the parties agreed that the relevant principles were to be found in the judgment of Waller J in Suncorp Insurance and Finance v Milano Assicurazioni SpA [1993] 2 Lloyd’s Rep 225, 234. So far as presently material, these provide:

“(i)

Where an act is done (as it was here) by a person in the name of another and that person does not have authority, then the person in whose name it was done may ratify by adopting the transaction (see Bowstead 15th ed. pp. 51 and 52).

(ii)

In order that a person may be held to have ratified an act done without his authority, it is necessary that at the time of the ratification, he should have full knowledge of all the material circumstances in which the act was done, unless he intended to ratify and take the risk whatever the circumstances may have been. (See Bowstead art. 16, p. 64.) The commentary makes clear however that this principle, designed to protect the principal from being held too readily to have ratified acts of his agent, is less strict in the contractual context than it is in a tort context.

(iii)

Ratification may be express or implied, and will be implied whenever the conduct of the person in whose name a transaction has been entered into is such as to show that he adopts the transaction in whole or in part; mere acquiescence or inactivity may be sufficient; (see Bowstead art. 17, p. 66.) The commentary has the following sentence:

‘Where the silence or inactivity is known to the third party, an estoppel may in appropriate cases arise against the principal, regardless of his actual intention (which alone is relevant to the ratification).’

An estoppel is not relied on by the plaintiffs in this case, but I am not sure that the position is quite as clear cut as that sentence would make out. If the principal is aware of all the material facts and appreciated that he was "being regarded as having accepted the position of principal, and took no steps to disown that character within a reasonable time, or adopts no means of asserting his rights at the earliest time possible" that can amount to "sufficient evidence of ratification". That is how it is put in par. 83 Halsbury’s Laws vol. 1(2) 4th ed. Reissue. If right, it comes very close to estoppel, but it is still ratification and in my view it is a correct formulation. It seems to me that it should not be open to a principal, who to the outside world by his conduct, or that of his duly authorized agents, appears to have adopted a transaction to be able to prove subjectively that in fact he had not, any more than such a principal would be able to prove subjectively that he did not intend to adopt a transaction when he does an act e.g. accepts the payment of money, which objectively adopts the transaction.

…

(v)

Ratification can also be by a duly authorized agent. … There is no doubt that ratification can be effected by an agent but that agent must be duly authorized either expressly or by virtue of being held out as having such authority. (See Bowstead p. 69)….”

Again, although the ratification case was not abandoned by Marsh, in his oral submissions Mr Edwards KC was content to rely on his skeleton argument. This argued that there was ratification (i) when Mr Lane signed the GCUK Letter of Engagement in October 2020, without GBAG then disclaiming that GCUK Letter of Engagement (or any of the earlier ones); and/or (ii) when GBAG pleaded its case in the Australian Proceedings (albeit, without Marsh identifying any particular aspect of the pleading).

I do not consider that these arguments have a high degree of probability of being established at trial. Indeed, I again consider – without making a finding – that they are more likely than not to fail. On the materials I have seen, Mr Lane was not an officer of GBAG. There is no evidence that Mr Lane knew or understood or intended when signing the GCUK Letter of Engagement in 2020 that he was binding GBAG, nor that he had authority to do so. Nor is there evidence that anyone on behalf of GBAG was aware of the signature of the 2020 GCUK Letter of Engagement, nor, when it was signed, that it purported to bind GBAG, such as might have required GBAG to disclaim it. Indeed, in my judgment, the signature of the 2018 GBAG LOE suggests the absence of any understanding that GBAG was bound by GCUK Letters of Engagement. As for the pleadings in the Australian Proceedings, while they do say that Marsh was GBAG’s insurance broker – which is common ground before me – they do not say that GBAG was party to any of the GCUK Letters of Engagement.

For these reasons, I cannot conclude to a high degree of probability that Marsh will (or would) succeed on the question of authority at trial. On the contrary, I consider Marsh would be more likely than not to fail to make good its case at trial.

In light of my decision on Issue 2, the only relevant Letter of Engagement in respect of which an interim ASI might be granted is the 2018 GBAG LOE. This raises the issue whether there are strong reasons for me not to grant an interim ASI to enforce the exclusive jurisdiction clause in that agreement.

In Catlin Syndicate Ltd v Amec Foster Wheeler USA Corp [2020] EWHC 2530 (Comm), Jacobs J explained the rationale for the ‘strong reasons’ requirement at [36]:

“I also consider that the approach is right. As Mr Stewart submitted (in the context of the defendants' cross-application for a mandatory injunction), the starting point is that the court will ordinarily act to protect the integrity of a contractual bargain reached between the parties. This is, in my view, one reason why 'strong reasons' are and should be required once the court is satisfied, to a high degree of probability, that there is a valid English jurisdiction clause to which the parties have agreed. Another reason is that where proceedings are started, in breach of contract, in a different jurisdiction to that which the parties have agreed, this will almost inevitably cause irremediable prejudice to the opposing party which cannot be satisfactorily compensated by damages. That party will be put to the expense, which can be considerable, of litigating a case, often over a lengthy period of time, in the different jurisdiction. There is always a serious risk that the result of the litigation will be different from that which would have resulted if the proceedings had been started in the correct forum, particularly so when – as is often the case and is the case here – the other forum is invited to apply a different law to that which would have been applied in the agreed forum. Even if the 'incorrect' forum were to be invited to apply correct law, it will often nevertheless be prejudicial to a party for this to happen in a case where the contractually agreed law and forum are the same. This is because it can reasonably be expected that the contractually agreed forum (i.e. England in the present case) will apply the contractually agreed law (English law in the present case) more reliably than the incorrect forum.”

Whether strong reasons are present in a particular case will “depend on all the facts and circumstances of the particular case”: per Lord Bingham in Donohue v Armco Inc [2001] UKHL 64 (‘Donohue’) at [24]. Lord Bingham went on to note at [25] that:

“Where the dispute is between two contracting parties, A and B, and A sues B in a non-contractual forum, and A’s claims fall within the scope of the exclusive jurisdiction clause in their contract, and the interests of other parties are not involved, effect will in all probability be given to the clause.”

The risk of parallel proceedings, in different jurisdictions, potentially leading to different results is a matter to which the courts will have regard. However, no weight is generally to be given to matters which were foreseeable to the parties at the time of contracting. As Coulson J held in Euromark Ltd v Smash Enterprises Pty Ltd [2013] EWHC 1627 (QB):

“14.

Where there is an exclusive jurisdiction clause, particularly if it selects the ‘home’ court of one of the contracting parties, foreseeable questions of convenience are irrelevant (see Beazley (on behalf of Lloyd's Marine Towage Insurance) v Horizon Offshore Contractors Inc [2004] EWHC 2555 (Comm). This principle was summarised by Gloster J, as she then was, in Antec International Limited v Biosafety USA Inc [2006] EWHC 47 (Comm) where she said:

‘Such overwhelming or very strong reasons do not include factors of convenience that were foreseeable at the time that the contract was entered into (save in exceptional circumstances involving the interests of justice); and it is not appropriate to embark upon a standard Spiliada balancing exercise. The defendant has to point to some factor which it could not have foreseen at the time the contract was concluded. Even if there is an unforeseeable factor or a party can point to some other reason which, in the interests of justice, points to another forum, this does not automatically lead to the conclusion that the court should exercise its discretion to release a party from its contractual bargain…’

15.

In essence, the party seeking to invoke the jurisdiction of the English court in the face of an exclusive jurisdiction clause, which provides for disputes to be determined in a foreign court, must point to a factor which could not have been foreseen when the contract was made. Moreover, what matters is whether it ought to have been foreseen, not whether it actually was (see by way of example the judgment of Moore-Bick J, as he then was, in Mercury Communications Ltd v Communications Telesystems International [1992] All ER (Comm) 33)).

…

18.

Mr Catherwood rightly drew my attention to the decision of Gross J, as he then was, in Import Export Metro Ltd and Another v Compania Sud Americana de Vapores SA [2003] EWHC 11 (Comm), reported at [2003] 1 Lloyd's Rep 405. In that case, at page 411 of the report, the learned judge said:

‘In the nature of things for the court to exercise its jurisdiction so as not to give effect to an EJC the strong reason relied on must ordinarily go beyond a mere matter of foreseeable convenience, and extend either to some unforeseeable matter of convenience, or enter into the interests of justice itself. Even then, it cannot simply be assumed that the court will automatically exercise its discretion so as to release one party from its contractual bargain. Once the interests of justice are engaged, then factors of convenience will be relevant to the exercise by the court, of its discretion.’”

In Donohue, Lord Bingham considered that, on the facts of that case, the ends of justice would be best served by a single composite trial in one forum, with the only possible forum being New York – which was not the contractual forum. However, this was subject to the qualification that Mr Donohue should be protected against liability under statutory RICO claims, to which he would not be exposed in England (i.e., the contractual forum), “because of the obvious injustice to him which such liability would in the circumstances involve”: at [36]. Lord Bingham therefore concluded that the interests of justice were best served if an anti-suit injunction were denied, but an undertaking proffered on behalf of the Armco companies not to enforce any award of multiple or punitive damages were accepted: at [39].

In Zephyrus Capital Aviation Partners 1D Ltd and ors v Fidelis Underwriting Ltd and ors [2024] EWHC 734 (Comm), Henshaw J reviewed a number of the authorities on ‘strong reasons’ before summarising applicable propositions at [124] (with references removed):

“(i)

The court is not bound to grant a stay but has discretion to do so.

(ii)

There can be no absolute or inflexible rule governing the exercise of the discretion.

(iii)

However, the English court will ordinarily exercise its discretion by granting a stay of proceedings unless the claimant can show strong reasons for suing in England.

(iv)

What constitutes a strong reason ‘will depend on all the facts and circumstances of the particular case’.

(v)

The burden of showing strong reason is on the claimant.

(vi)

Strong reasons are not shown merely by establishing factors that would make England the appropriate forum on a forum non conveniens analysis.

(vii)

Foreseeable factors of (mere) convenience should not be regarded as strong reasons to decline a stay.

(viii)

Regard can properly be had to whether the claimant would be prejudiced by having to sue in the foreign court because they would, for political, racial, religious or other reasons, be unlikely to get a fair trial.

(ix)

There are some judicial statements suggesting that even a matter pertaining to the interests of justice might not amount to a ‘strong reason’ if it was foreseeable and could be regarded as encompassed within the parties’ bargain in agreeing to the jurisdiction clause. However, the preponderance of the cases treat the interests of justice differently in that regard from factors of mere convenience.”

GBAG identifies five matters as, in combination, providing such ‘strong reasons’. These are, in summary:

I am unpersuaded by these arguments.

GBAG’s first argument (see paragraph 87(a) above) seems misconceived. Marsh has brought a claim for a permanent ASI in respect of all the GCUK Letters of Engagement, and there is no application before me to prevent Marsh pursuing that claim. So, whether I grant an interim ASI or not, the risk of inconsistent findings arises. In any event, since there is no dispute that GBAG is party to the 2018 GBAG LOE, and the only interim ASI I am now considering relates to the exclusive jurisdiction clause in the 2018 GBAG LOE, a (probably theoretical) trial of a claim for a final ASI arising out of this Letter of Engagement would not require findings regarding authority at all.

GBAG’s remaining arguments are primarily ones of convenience – namely the inconvenience of having to fight the same or similar disputes in different jurisdictions. However, even if the precise nature of what occurred was not predictable, inconvenience of this sort was nonetheless to a very great extent foreseeable when Marsh and GBAG agreed to an exclusive jurisdiction clause in 2018 in circumstances where (i) the insurance policies themselves contained exclusive jurisdiction clauses pointing to other jurisdictions, as is commonplace; and (ii) Marsh and GBAG had not previously concluded any such clause even while Marsh provided broking services to GBAG (or, at least, this is the assumption I make, which is GBAG’s own case). I also consider it relevant that there is already a multiplicity of proceedings in different jurisdictions, so there is no prospect of a single set of proceedings determining all issues.

GBAG does not contend that any third parties will be affected by an ASI being granted. Nor can GBAG identify any real injustice which it will suffer from having to bring its claim in England. True it is that GBAG may well be unable to recover the substantial sums from Marsh which it might recover under a statutory MDC claim in Australia, but protecting parties from exposure to substantial civil claims in foreign jurisdictions is a major reason why commercial parties, conducting business internationally, use exclusive jurisdiction clauses; it is not a matter of injustice for such a clause to be enforced.

As regards considerations of comity, it is helpful to recall the judgment of Christopher Clarke LJ in Ecobank at [132]:

“Comity has a warm ring. It is important to analyse what it means. We are not here concerned with judicial amour propre but with the operation of systems of law. Courts around the free world endeavour to do justice between citizens in accordance with applicable laws as expeditiously as they can with the resources available to them. This is an exercise in the fulfilment of which judges ought to be comrades in arms. The burdens imposed on courts are well known: long lists, size of cases, shortages of judges, expanding waiting times, and competing demands on resources. The administration of justice and the interests of litigants and of courts is usually prejudiced by late attempts to change course or to terminate the voyage. If successful they often mean that time, effort, and expense, often considerable, will have been wasted both by the parties and the courts and others. Comity between courts, and indeed considerations of public policy, require, where possible, the avoidance of such waste.”

Respect for comity is not, without more, a strong reason for the court not to give effect to an exclusive jurisdiction clause: see QBE at [11(i)], cited at paragraph 23 above. At this early stage, before Marsh has been joined as a respondent to a claim by GBAG in the Australian Proceedings, the burdens on the Australian court are not increased by an ASI. GBAG draws attention to Marsh’s breach of its Harman obligation as providing an exceptional feature, but I bear in mind that Moshinsky J in the Australian court took the decision that it was appropriate to release Marsh from the Harman obligation “such that those documents may be used for the purpose of the Anti-suit Application”. Moreover, essentially for the reasons I gave in respect of Issue 1 (see paragraph 45 above), I consider that it would be disproportionate to deny Marsh the benefit of the exclusive jurisdiction clause as a result of its breach of the Harman obligation.

I recognise that it may prove challenging for GBAG to formulate a claim against Marsh in the Australian Proceedings limited to alleged wrongdoing not covered by the exclusive jurisdiction clause in the 2018 GBAG LOE, while complying with an ASI securing compliance with that clause. However, that again seems to me to be a predictable inconvenience for GBAG, nothing more.

For these reasons, I conclude that there are not strong reasons for refusing to grant an interim ASI in respect of the exclusive jurisdiction clause in the 2018 GBAG LOE.

I have, at the invitation of Marsh, given consideration to whether I should make a conditional interim ASI – by ordering that unless GBAG undertakes not to pursue an MDC claim in Australia, I will make an interim ASI. However, this was not Marsh’s preferred outcome, and Mr Fealy KC told me he had no instructions to offer such an undertaking. Even if such an undertaking were given, in my judgment Marsh has a legitimate commercial interest in ensuring that non-MDC claims, which would be subject to English law, are tried in England. Accordingly, I see no reason to make my order conditional in this way.

At the end of the hearing I ordered that the existing interim ASI made by Cockerill J on 30 July 2024 would continue until the conclusion of the hearing of consequential matters arising from this reserved judgment. For the reasons set out in this judgment, I intend at that time to continue the interim ASI on a more limited basis, to restrain GBAG from taking any steps to initiate or bring any claim against Marsh in Australia, in relation to the Engagement contained in the 2018 GBAG LOE and any non-contractual obligations arising out of or in connection with that Engagement.