Al Mana Lifestyle Trading LLC & Ors v United Fidelity Insurance Company PSC & Ors

The issue on this appeal is whether the “Applicable Law and Jurisdiction” clause in a series of insurance policies issued by the appellant defendants contains an agreement which gives the English court jurisdiction over claims brought by the respondent claimants under the policies.

The claims are brought by the claimants under a suite of seventeen “Multi-Risks” insurance policies underwritten by the defendants. The claims are all for indemnities for business interruption losses, said by the claimants to arise from the Covid-19 pandemic and estimated by them to have a combined value of about US $40 million.

The clause in question provides as follows:

“APPLICABLE LAW AND JURISDICTION:

[1] In accordance with the jurisdiction, local laws and practices of the country in which the policy is issued. [2] Otherwise England and Wales UK Jurisdiction shall be applied,

[3] Under liability jurisdiction will be extended to worldwide excluding USA and Canada.”

The numbers in square brackets are not included in the clause, but were added by the parties for ease of exposition. The comma at the end of the second sentence is clearly a typographical error and should be a full stop. Nothing turns on this.

The defendants’ case is that, in each policy, the clause provides for the exclusive jurisdiction of the court of the country in which the policy was issued (for simplicity I will refer to this as “the local court”), with a fallback for English or Welsh jurisdiction in the event that the local court does not have or would not accept jurisdiction.

The claimants’ primary case, accepted by the judge, is that the clause gives whichever party wishes to bring a claim a free choice. It may bring proceedings either in the local court or in England. Alternatively, if that is wrong, the jurisdiction of the English court is available so long as the jurisdiction of the local court is not mandatory under the law of that country. Mr Gavin Kealey KC for the claimants made clear that this alternative arises only if, contrary to his primary case, there is what he described as an element of conditionality in the clause.

The claimants are 27 entities forming part of the Al Mana Group, carrying on businesses in the food, beverage and retail sectors, principally in the Middle East and Gulf region, but also with a small part of their operation in Ireland. The group does not carry on business in England or Wales.

The defendants are insurance companies operating within Gulf Cooperation Council countries, located respectively in the United Arab Emirates, Qatar and Kuwait, which is where the policies in question were issued: 15 policies issued by the first defendant in the UAE, some of which cover multiple insureds operating in different jurisdictions; one issued by the second defendant in Qatar; and one issued by the third defendant in Kuwait.

For present purposes, all of the policies are on materially identical terms. They contain five sections of cover or potential cover (because not all the claimants took out cover under each section), namely: (1) Property All Risks, which extends to include cover for certain business interruption losses, (2) Money All Risks, (3) Blanket Fidelity, (4) Workmen Compensation; and (5) Public & Product Liability.

Each policy contains a Schedule which includes, among other things, the “Applicable Law and Jurisdiction” clause which I have already set out. It includes also other clauses which were referred to in argument in the court below and are set out in the judgment, but which have not featured in the appeal.

The claim form was issued on 21^st May 2021. It was served on the defendants out of the jurisdiction in reliance on what the claimants contended was the agreement for English jurisdiction contained in the Applicable Law and Jurisdiction clause. The defendants’ challenge to the jurisdiction came before Mrs Justice Cockerill on 27^th July 2022. In her reserved judgment, handed down only two days later, the judge accepted the claimants’ primary case that the clause gives whichever party wishes to bring a claim a choice of bringing proceedings either in the local court or in England and Wales. In that sense the jurisdiction for which the clause provides is non-exclusive, although exclusive as against the rest of the world.

Applying the approach of the Supreme Court in Financial Conduct Authority v Arch Insurance (UK) Ltd [2021] UKSC 1, [2021] AC 649 at [77], the judge set herself the task of considering “how the words of the contract would be understood by a reasonable person to whom the document should be taken to be addressed”, that is to say “an ordinary policyholder” and not “a pedantic lawyer who will subject the entire policy wording to a minute textual analysis”.

The judge went some way towards accepting the main points made by the defendants. Thus she accepted that the clause provides for a single choice of the law of the country in which the policy was issued (“the local law”) as the governing law, and that this is so regardless of where proceedings are brought; that a choice of the local law is a factor in favour of the defendants’ construction; that the words “in accordance with” which provide for a mandatory choice of the local law were the same words as provide for the jurisdiction of the local court; that these words are capable, in an appropriate context, of being used in an “imperative and directory” sense; and that the formulation “in accordance with … otherwise … shall” might be understood as providing for an exclusive jurisdiction, together with a fallback. However, she considered that these points were not decisive, either individually or collectively.

Ultimately, the judge’s acceptance of the claimants’ approach was founded on three points. First, she considered that the word “otherwise” as used in the clause is most naturally considered as equivalent to “or”, so that the clause provides for a choice of jurisdiction. Second, she considered that the difficulty with regarding the first sentence as providing the primary rule, with the second sentence providing for a fallback, is that the clause does not identify the circumstances in which the fallback is triggered. Third, she regarded the claimants’ construction as supported by what she called “the commercialities or practicalities”: either there was a real prospect that the local court would not accept jurisdiction, in which case it was unattractive for a claimant to have to bring proceedings there in order to find out whether it would accept jurisdiction, only to start again in England if it would not; or there was no realistic prospect of jurisdiction being declined in which case the second sentence of the clause was otiose, a conclusion which the English court would not readily reach; and it made sense to provide for a single neutral venue, namely the English court, particularly as common issues could arise under policies issued in each of the three jurisdictions where the defendants are located.

The judge expressed her conclusion as follows:

“88.

… Both as a question of impression and on detailed analysis I consider that the better view is that the clause provides for non-exclusive jurisdiction – a true alternative. I accept that the clause is to some extent odd, but this result is less odd and creates fewer difficulties than the approach urged by the Defendants. A non-exclusive jurisdiction clause best harmonises the wording and the commercialities of the clause in the context of the wider factual matrix.”

Accordingly she did not need to consider the claimants’ alternative construction to the effect that the jurisdiction of the English court is available so long as the jurisdiction of the local court is not mandatory under the local law.

She did go on to consider and reject the defendants’ alternative case that jurisdiction should be declined on forum non conveniens grounds. That case has not been pursued on appeal.

For the defendants Mr John Lockey KC submitted that the clause, properly construed, provides for the exclusive jurisdiction of the local court. Sentence [1] contains a clear agreement on the applicability of the local law and local jurisdiction. Sentence [2] contains a fallback position, as regards jurisdiction only, which does not apply where (as here) the local court would accept jurisdiction; sentence [2] is incapable of transforming the agreement in sentence [1] into an agreement which treats the jurisdiction of the local court as no more than optional if one of the parties chooses to litigate in England or Wales. Sentence [3] is not concerned with jurisdiction at all, at least in the sense in which that word was used elsewhere in the clause, but with a separate question, namely which judgments of which courts are covered by the liability section of the policy (Section 5).

Mr Lockey submitted, in outline, that the opening words of sentence [1] are imperative or directory; that the choice of local law is a powerful factor in favour of construing the choice of local jurisdiction as mandatory; that this is in accordance with authorities such as Hin-Pro International Logistics Ltd v Compania Sud Americana de Vapores [2015] EWCA Civ 401, [2015] 1 CLC 901, Generali Italia SpA v Pelagic Fisheries Corporation [2020] EWHC 1228 (Comm), [2020] 1 WLR 4211 and AIG Europe SA v John Wood Group Plc [2022] EWCA Civ 781 at [62]; that no significance should attach to the fact that the word “exclusive” is not used in the clause; that sentence [2] catered for the possibility that the local court would not accept jurisdiction, even if this was a remote contingency, there being no reason why such provision should not be made; that the judge was wrong to place weight on the apparent desirability of a “single neutral forum” when that could be torpedoed by either party issuing a claim for a declaration of non-liability in the local court; and that the judge was wrong also to refer to the expertise of the English court and its experience of deciding Covid-19 business interruption disputes as a factor supporting the claimants’ construction of the clause.

For the claimants Mr Kealey, supporting the judge’s reasoning and conclusion, submitted that the construction of the clause is as much a matter of impression as of analytical interpretation and that decisions on different wordings in other cases were not helpful. Here, the simple point is that the clause provides an alternative of two jurisdictions, either the local court or, if not there (i.e. “otherwise”), the English court (or, for the liability cover in Section 5, anywhere except the USA or Canada). There is nothing in the clause to say that the English court is only available if the local court has declined or would decline jurisdiction – or in more technical language, there is no condition precedent to the operation of the second sentence. Moreover, it accords with commercial common sense to provide for a single neutral forum, particularly in a case where there is a suite of policies covering multiple insureds operating in a number of jurisdictions and common issues could arise under a number of policies. If the English court is only available if the local court has declined or would decline jurisdiction, the result is likely to be disagreement and delay.

The judge asked herself the right question, which is how the words of the contract would be understood by a reasonable policyholder. To some extent the answer to that question must depend upon the impression which the clause would convey to such a reader. For the reasons explained in Lewison, The Interpretation of Contracts (7^th Ed), paras 2.103 to 2.112, and as Mr Kealey’s skeleton argument puts it, “impressions (and first impressions and intuition and judgment) may be as powerful a tool as intricate linguistic and contextual analyses” when seeking to discern the true meaning of a contract. This is particularly the case when, as here, the clause in question is tersely expressed. Nevertheless, while such an impression is very often the starting point, it is necessary at least to attempt some further analysis.

My strong impression when I first saw this clause was that the first sentence contains the primary jurisdiction selected by the parties, with a fallback for English or Welsh jurisdiction in the second sentence. That impression has been confirmed rather than dispelled by the more analytical approach adopted in the parties’ submissions.

The fact that the first sentence deals not only with jurisdiction, but also with the governing law (i.e. the local law) and the need to apply local practices, while the second sentence is confined to jurisdiction, strongly suggests that the first sentence is intended to contain the primary rule, with the second sentence operating as a fallback. Thus, even when the second sentence applies and English jurisdiction is invoked, the English court would be required to apply the local law and practices. (Footnote: ¹) While it is possible, and not uncommon, for the English court to receive evidence of foreign law, and it would be equally possible for the English court to receive evidence about local practices, that is inevitably second best when compared with the application of local law by the local court, which can also be expected to be familiar with local practices. Just as the English courts have consistently held that “England is the best forum for the application of its own law” and that a choice of English law is itself a powerful factor showing that a choice of English jurisdiction is intended to be exclusive (e.g. Hin-Pro at [66] and [77], AIG v John Wood at [62]), so it must be accepted that a foreign court is the best forum for the application of its law and that a choice of foreign law is a powerful factor showing that a choice of foreign jurisdiction is intended to be mandatory.

The fact that the second sentence deals only with jurisdiction, so that even in English proceedings local law and practices must be applied, demonstrates also that, in the context of this clause, the words “in accordance with” are intended to be mandatory. In this regard I do not find it helpful to compare and contrast the words “in accordance with” with words such as “subject to” which have been used in clauses considered in other cases such as Hin-Pro and AIG v John Wood. It is sufficient to say that, in the context of this clause, the words “in accordance with” are imperative and mandatory. Prima facie, therefore, the choice of the jurisdiction of the local court is also mandatory.

All this, of course, is subject to the effect of the second sentence and, in particular, what is meant by the word “Otherwise”, with which that sentence begins. Obviously the clause must be considered as a whole. While it may be that in some contexts “otherwise” can be regarded as equivalent to nothing more than “or”, the context is important. If the parties had intended to provide for a free choice of jurisdiction for whichever party was to be the claimant (either the local court or England and Wales), this strikes me as an odd way of doing so. In the context of a jurisdiction clause such as this, the word “Otherwise” is more appropriate to introduce a fallback.

The question then arises, in what circumstances is the fallback available – and in particular, does the clause lack any indication of what these circumstances are? I see no real difficulty here. The fallback of England and Wales is available if the local court specified in the primary rule contained in the first sentence is not available – which in practice means, if the local court does not or would not accept jurisdiction. I do not regard this as introducing a condition precedent which does not exist in the clause, or as adding words which are not there, but rather as the natural meaning of the word “Otherwise” in the context of this jurisdiction clause.

Mr Kealey submitted that the word “Otherwise” could be regarded as equivalent to “if not”. I would be inclined to accept that submission, but it does not resolve the question, in what circumstances can the jurisdiction of the English court be invoked? It seems to me to make far better sense of the clause, and to be likely to be what the parties intended, that “Otherwise” or “if not” means “if not available”, rather than “if not fancied by whichever party is the claimant”.

In my judgment the objections to this construction are without substance.

I have already dealt with the principal objection, which is that it is necessary to read words into the clause which are not there in order to introduce an element of conditionality. For the reasons which I have given, I do not agree.

The second objection seeks to pose a dilemma: either there are no or very limited circumstances in which the local court would not accept jurisdiction, in which case the second sentence is otiose; or there are likely to be disputes about whether it would accept jurisdiction, leading to uncertainty for the claimant about where it should issue proceedings. In my judgment this is a false dilemma, and even something of an Aunt Sally.

I would accept that it may be that the circumstances in which the local court would not accept jurisdiction are very limited or, perhaps, even non-existent. But I do not regard that as an objection. I see no reason why parties should not agree to confer jurisdiction on one court, with another as a fallback in case the primary court chosen is not available, without troubling to investigate whether or in what circumstances the primary court would decline jurisdiction. Such an agreement gives the parties the comfort of knowing that if, for any reason, their primary choice is not available, there is an alternative with which they are comfortable, and is a sensible agreement to make. The fact that, on investigation, it can be seen that the fallback is (or is likely to be) unnecessary, is nothing to the point.

Nor do I regard it as an objection that there may be uncertainty as to where proceedings should be commenced if there is a dispute about whether the local court would accept jurisdiction. In the first place, such a dispute is unlikely to arise. If the proposed defendant, which in most but not all cases will be the insurer, objects to being sued in the local court, which is itself unlikely, it will have no ground for complaint if the proposed claimant then sues in England. The picture painted, of a claimant trooping disconsolately back and forth from one court to another in order to find out where it is entitled to sue, seems to me to be a remote contingency which should not drive the construction of the clause. In the second place, on either construction, an element of uncertainty is inherent in the clause. On the claimants’ approach, either party may choose where to commence proceedings. So if the insurer is faced with a claim and wants to avoid English jurisdiction, it can commence an action for a declaration of non-liability in its local court. For jurisdiction to be established in one court rather than another depending on the chance (or sometimes the cunning) of whichever party gets in first seems to me to be most unlikely to have been what the parties intended.

That leaves the supposed desirability of the English court as a single neutral forum. However, this is of very limited if any significance in circumstances where (as the judge recorded at [91]) neither side has suggested that the claimants would not be able to obtain a fair trial in the UAE, Qatar or Kuwait or that those local courts would not be equipped to handle the claims in an efficient, cost-effective and timely manner, and where the claimants themselves are part of a group of companies operating in those jurisdictions. In those circumstances the need for a neutral venue does not arise: there is no challenge to the independence or neutrality of the local courts and both parties can fairly be taken to have appreciated this when making their contracts.

I can see that there may in some cases, where a common issue arises concerning policies issued in more than one country, be an advantage to both parties to have that issue decided in a single forum. However, where that is so, there is no reason why the parties should not agree on the jurisdiction to decide that issue. In any event, most of the risks giving rise to litigation would be likely to be of a local nature and the benefit, in a limited category of cases, of having a single forum cannot drive the construction of the law and jurisdiction clause, not least as (even on the claimants’ construction) it would always be open to a party to sue in the local court and thereby to negate the supposed advantage.

For these reasons I consider that the judge’s construction of the clause was mistaken. The second sentence applies only when the jurisdiction of the local court is not available. In this case it is common ground that each of the local courts would accept jurisdiction over the claimants’ claims. Accordingly the second sentence of the law and jurisdiction clause in the policies does not give the English court jurisdiction over those claims.

I should, however, mention two of the defendants’ criticisms of the judgment which I do not accept. First, I do not accept the criticism that the judge regarded the experience of the English court in dealing with Covid business interruption insurance as a relevant factor. As she said at [86], this “cannot be relevant to construction”. That was plainly right. Second, the judge rejected at [87] the defendant’s submission that the third sentence of the clause is not a true jurisdiction provision at all, but modifies the operative part of the liability cover in Section 5. She was plainly right to do so. The third sentence permits the claimant to sue in any jurisdiction other than the USA and Canada in the case of a claim under Section 5 of the policy. But the present case is not a claim under Section 5 and this point cannot affect the issue which we have to resolve concerning the construction of the first two sentences.

The claimants’ alternative case is that if, contrary to its submission, some form of conditionality is imposed on the jurisdiction of the English court, it is not that the English court has jurisdiction in circumstances where the local court has declined or would decline jurisdiction, but that the jurisdiction of the English court is available unless the jurisdiction of the insurer’s local court is mandatory as a matter of local law. As the judge made no findings whether such jurisdiction is mandatory under the law of the UAE, Qatar and Kuwait (on her approach to the clause, she did not need to), the claimants invited us to remit the case to her for such findings to be made.

I would reject this alternative case. While I see no difficulty, for the reasons which I have sought to explain, in reading the word “Otherwise” as referring to a situation where the local court is not available to the claimant, I can see no basis in the language of the clause for the claimants’ alternative construction. It is, moreover, most unlikely that the parties would have intended this construction, as it would require an investigation in every case, not whether the jurisdiction of the local court was available, a relatively straightforward point, but whether it was mandatory. Indeed, the parties served conflicting evidence as to the position under the three local laws. As that evidence demonstrates, to read the clause as allowing English jurisdiction unless the jurisdiction of the insurer’s local court is mandatory as a matter of local law would be likely to produce unnecessary and wasteful disputes which cannot sensibly have been intended.

Despite Mr Kealey’s powerful advocacy and the judge’s thoughtful judgment, I would allow the appeal and declare that the English court has no jurisdiction to try the claimants’ claims. Service of the Claim Form should therefore be set aside.

Much as I regret disagreeing with Lord Justice Males, I am firmly of the view that the judge was right in her construction for the reasons that she gave. That was my strong first impression, which was confirmed on further detailed analysis and after hearing the excellent oral submissions of both leading counsel.

Two matters on which I believe we are in agreement are that this clause is not well drafted, and that a key consideration is the meaning of the word “otherwise” at the start of the second sentence.

As Lord Clarke JSC said in Rainy Sky SA v Kookmin Bank [2011] UKSC 50, [2011] 1 WLR at [21]:

“ … the exercise of construction is essentially one unitary exercise in which the court must consider the language used and ascertain what a reasonable person, that is a person who has all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract, would have understood the parties to have meant. In doing so, the court must have regard to all the relevant surrounding circumstances. If there are two possible constructions, the court is entitled to prefer the construction which is consistent with business common sense and to reject the other.”

That means that the clause must be construed holistically, contrary to the approach advocated by Mr Lockey of looking at what the first sentence would mean, taken in isolation, and then considering how the second sentence qualifies or, as he contended, supplements it. It is nothing to the point that if it existed on its own, without the second and third sentences, the first sentence would probably be regarded as an exclusive jurisdiction clause. In that scenario there would be no alternative forum to consider. But that sentence does not stand alone. It is conditioned by the second sentence, which begins with the word “otherwise”; and as the judge says at paragraph 75 of her judgment, that provides for some form of alternative. The question is whether it is a true “either/or” alternative, as the claimants contended and the judge decided, or a primary/secondary alternative, as the defendants contended and Males LJ has decided.

In the context in which the word is used in this clause, “otherwise” naturally connotes a choice. As a matter of plain English, “otherwise” means “or”, “or else”, or “if not”. A common dictionary definition is: “in another manner, differently.” Collins English Dictionary pertinently states that: “you use otherwise to indicate that other ways of doing something are possible in addition to the way already mentioned.”

“Otherwise” may also be used to denote what will or may happen if something else does not happen, e.g.: “You must catch the last train, which leaves the station at 11pm. Otherwise, you will have to travel home by night bus.” However, even when it is used in that sense, it does not carry with it any implication of why the first scenario does not happen. In the example given, it does not matter why the person fails to catch the last train, it is the fact that they do not catch it which triggers the need to catch a bus instead. “Otherwise” does not carry with it any implication that the last train did not run, or that it did not stop at the station, or that it departed five minutes earlier than scheduled. Even when it is used in the sense of “if not”, “otherwise” does not mean “if that is not possible,” and that is not how it would be understood by the reasonable reader, particularly if they are not a lawyer.

The heading “Applicable Law and Jurisdiction” applies to the whole clause. All three sentences relate to jurisdiction. The first concerns the jurisdiction of the courts of the country where the policy is issued (to which I shall refer as “the local forum”), the second concerns the jurisdiction of the courts of England and Wales. The third sentence is of very limited assistance in construing the first two, because it expressly extends whatever jurisdiction has been agreed earlier to “worldwide, excluding the USA and Canada” in the case of liability claims. This has the practical effect of enabling an assured who has been sued in a jurisdiction other than the local forum to join the insurers as parties to that litigation, unless that underlying litigation is in the USA or Canada.

The extension of jurisdiction for liability claims does not directly assist in determining where non-liability claims are to be litigated. The most it does is indicate that the contracting parties envisaged circumstances in which a court other than the local forum would be applying the local law (as the chosen governing law) to a dispute between the insurer and the assured. That factor weighs against the express choice of the local law being a strong indicator of a mutual intention that the local forum should have exclusive jurisdiction. So too does the existence of the second sentence, because, as the judge pointed out, the parties envisaged that in certain circumstances the courts of England and Wales would have jurisdiction, and in those circumstances they would be applying the local law to the contract of insurance. The dispute is about when those circumstances would arise.

Both the first and the second sentences use language which, taken in isolation, might be interpreted as mandatory. In the first sentence, the phrase is “in accordance with”, in the second the phrase is “shall be applied.” I agree with the judge that “in accordance with” is not the same as “subject to”, and that when “otherwise” is interposed between the two phrases:

“the reality is that this wording, with mandatory or quasi-mandatory elements in both parts, presents more naturally to a reader – and probably particularly to a non-legal reader - as an either/or.”

(Judgment paragraph 69).

It is important to bear in mind the nature of the reader from whose perspective this clause is to be interpreted. Whereas the defendants’ interpretation might commend itself to a commercial lawyer, I doubt whether it would even occur to the reasonable policyholder, apprised of all the relevant circumstances, that it could be understood as meaning that it was mandatory to bring proceedings in the local forum, and that they could not go to the English court unless they could establish that the local court had declined, or would decline jurisdiction. They would understand it to mean that if, for whatever reason, they did not bring proceedings in the local forum, they would have to do so in England and Wales.

I next turn to business common sense. Non-exclusive jurisdiction clauses are a known phenomenon. They may be less common than exclusive jurisdiction clauses, but they do exist. Giving the parties a choice between two competent courts which are each capable of determining the disputes fairly and in accordance with the chosen proper law appears to me to be a perfectly reasonable thing to have agreed. That does not involve casting any aspersions on the local forum. The conferring of a choice does not need to be predicated upon a perceived need for a neutral alternative venue to one of the selected jurisdictions. The fact that the local forum may be an excellent venue, or even the best venue for determining such disputes, does not mean that the parties wished to restrict themselves to it. On the claimants’ interpretation, the second sentence ensures that if the local forum is not chosen, the court which determines the dispute will be one with an international reputation, which is neutral, and which is familiar with insurance disputes and used to applying a foreign proper law. There is nothing uncommercial about that.

I am not impressed by the argument that an insurer could in practice negate the assured’s choice by commencing pre-emptive litigation in the local forum. That could happen in any case where there is a non-exclusive jurisdiction clause in a contract, but it does not affect the nature of the clause, let alone provide a justification for construing it as something else. Males LJ states that for jurisdiction to be established in one court rather than another depending on the chance (or sometimes the cunning) of whichever party gets in first seems unlikely to have been what the parties intended, but that is tantamount to saying that parties would be unlikely to agree to a non-exclusive jurisdiction clause, despite the fact that in practice they often do. I have seen numerous such clauses during my years in practice at the commercial bar, though they were generally better drafted. In any event, if on the face of it the parties agreed upon a choice of forum, why should the reasonable policyholder make an assumption when interpreting the policy that the insurer would want to preclude the assured from exercising that choice? Even if he did want to do so, he would be exercising a choice of forum that was conferred equally upon him, so the assured would be in no position to complain about it.

In the context of this case, it is possible that a common issue might arise which affects a policy issued in the UAE, a policy issued in Kuwait and a policy issued in Qatar. If the defendants’ construction is right, that issue would have to be determined separately in proceedings in each of those countries, with the concomitant risk of conflicting decisions. If the claimants’ construction is right, it would be possible for the issue to be resolved in the English court in one set of proceedings. Potentially that would be beneficial to all concerned.

Males LJ suggests that in that scenario, the parties (who could be as many as six different legal entities) could agree upon a single forum to determine the disputed issue. That is true, but it is equally true that in the far more unlikely event that a court of exclusive jurisdiction declined jurisdiction, the parties could agree that their dispute be referred to a different court (or to an arbitrator). A more important answer to that objection is that the fact that the parties could enter into a separate agreement to deal with the situation of potentially conflicting judgments as and when it arose, does not mean that it made no commercial sense for them to have made a bargain that would, among other matters, cater for that contingency in advance. It would provide them with the comfort of certainty that one court, with which they were all content, could resolve the common issue.

By contrast, the suggestion that the second sentence was intended as a “fallback” to cater for what everyone appears to accept is and was at all material times the highly unlikely prospect that the local forum would decline exclusive jurisdiction conferred upon it by the parties, makes far less sense. I have seen many exclusive jurisdiction clauses, but never a clause which provides what will happen if the chosen court declines jurisdiction. The hypothetical reasonable reader would think that it was unlikely to have occurred to the contracting parties that that might happen in practice, and even less likely that if it did occur to them, they would want to make provision for it. Of course, parties are generally free to make whatever bargain they choose, including making provision for remote or unrealistic contingencies, but I would expect them to do so in a manner which was less oblique and which identified the contingency in terms. It would not have been difficult to have started the second sentence with the words “If that court declines jurisdiction”, “If that court is unavailable” or “If that is not possible”. None of those phrases is synonymous with “otherwise”, which is the word that the parties used.

I accept Mr Lockey’s point that the fact that a hypothetical scenario may be remote does not necessarily mean that the parties did not wish to cater for it. But if at the time of the contract the parties would have appreciated that it was highly unlikely to happen, it is less likely that they intended to provide for it, and more likely that the alternative construction is correct.

One cannot use hindsight to justify a particular construction, so the fact that the evidence indicates that in practice the local courts of the UAE, Kuwait and Qatar would accept jurisdiction over the disputes only serves to confirm the truth of what the parties to these insurance contracts would have expected to be the case from the outset, had they addressed their minds to the point. This is not something which has only emerged upon investigation. A reasonable person, at the time the contract was made, would naturally expect that the courts of the country where the policy was issued and whose law governed the contract would have no reason to decline a jurisdiction expressly conferred upon them by agreement. A court, particularly a court which has real connections with the parties and the contract, will usually respect the freedom of parties to agree upon the venue for dispute resolution, subject to public policy considerations (which no-one has suggested would arise in this context). The insurer would naturally expect the courts of a country in which it is carrying on business to accept jurisdiction over it. The assured would expect to have an uphill, and probably futile struggle to persuade the local forum to decline jurisdiction in the face of an agreement that disputes should be resolved only by that court.

The more unrealistic the prospect that the chosen forum would decline jurisdiction, the less likely it is that the parties would have decided to cater for such a remote contingency. The lack of any realistic practical utility of a “fallback clause” compared with the potential benefits to the parties of agreeing on a limited choice of jurisdictions in which to resolve their disputes serves to reinforce the interpretation which is, to my mind, the natural and obvious reading of the clause.

Even when it means “if not”, “otherwise” does not mean “if not available” or “if not possible” and construing it in that way involves implying additional words (“available” or “possible”) which are not there and would not ordinarily be implied. It makes no commercial sense to me to strain the ordinary meaning of “otherwise” in this way in order to produce the result that the second sentence would serve no useful purpose, when the normal meaning of the word produces a construction which would give the parties a choice of appropriate jurisdictions which could be of real assistance if they wished to resolve a common dispute in a single venue.

This clause means that if, for whatever reason, the proceedings are not brought in the courts of the country where the policy was issued, they must be brought in England and Wales. I agree with the judge that interpreting it as a non-exclusive jurisdiction clause: “best harmonises the wording and the commercialities of the clause in the context of the wider factual matrix.”

I would therefore dismiss this appeal. I would add that I am entirely in accord with Males LJ’s rejection of the defendants’ further criticisms of the judgment for the reasons he states in paragraph 36 above, and with his rejection of the claimants’ alternative case as set out in the Respondents’ Notice, for the reasons that he gives in paragraphs 37 and 38. Neither of these matters affects the result.

I have had the great advantage of reading in draft the judgments of both Males LJ and Andrews LJ. For the second time in short succession, I find myself faced with persuasive but divergent judgments written by eminent judges with long experience of construing commercial documents. In the present case I agree with the judgment of Males LJ.

I can express my reasons for this conclusion quite shortly. The task is to identify how the words in question would be understood by the reasonable reader armed with the background knowledge reasonably available to the parties. For this purpose I find that it is usually helpful to start with the natural and ordinary meaning of the words that the parties have chosen.

It must be admitted however that it is not always easy to articulate with precision why one reading of a disputed provision seems more natural and ordinary than another, as the way in which language strikes a reader is an accumulation of experience of how language is ordinarily used. And, as the present case illustrates, the same words may strike different readers differently: thus Males LJ at paragraph 22 above has said that his strong impression when first reading the clause was that the first sentence contains the primary jurisdiction selected by the parties, with the second sentence a fallback, whereas Andrews LJ at paragraph 40 has said that her strong first impression was to the contrary.

On this I agree with Males LJ and I will try and explain why. A person reading the clause would probably start with the heading. This tells one that the clause will stipulate both what is agreed about the applicable law and what is agreed about jurisdiction. Then the reader will read the first sentence. This, read with the heading, tells one that these things are to be “in accordance with the jurisdiction, local laws and practices of” the place of issue. Pausing there, it seems to me that the natural way to understand the clause so far is that the applicable law is agreed to be “in accordance with” the local law of the place of issue, and that jurisdiction is agreed to be “in accordance with” the jurisdiction of the place of issue.

Now this of course has to be read with the second sentence, but before coming to that, it is worth teasing out what this means. First it means that the proper law of the policy is agreed to be the local law of the place of issue. That means that the words “in accordance with” are here being used to indicate what the proper law is. In this sense the words are mandatory. I do not myself think they are any different in effect from the words “subject to”: they are determinative of what the proper law is. The same applies to local practices: these are not optional but dictated. Second, one would expect the words “in accordance with” to mean exactly the same when it comes to jurisdiction. In other words, if this sentence stood alone, I would not myself have any real doubt that it would be equally determinative of which courts had jurisdiction, namely (only) the courts of the place of issue: see paragraph 24 of Males LJ’s judgment, with which I agree. And although of course one must be cautious about decisions on other words in other contracts, it is notable that similar provisions have consistently been held to provide for exclusive jurisdiction: see AIG Europe SA v John Wood Group plc [2022] EWCA Civ 781 at [59]-[63] per Males LJ.

The reader now comes to the second sentence. For my part, I do not think the punctuation can be ignored, and this is clearly a second and separate sentence. It is true that the comma at the end of the second sentence is obviously an error for a full stop (as the third sentence starts with a capital, and on a new line), but that does not mean that the full stop at the end of the first sentence can be ignored as if it were not there, and the capital at the beginning of the second sentence shows that it was intended. Punctuation admittedly often provides very slender, if any, support for a particular construction, but it can help to inform the question of how a provision would be read by the ordinary reader, and in the present case I think it does tend to suggest that the two sentences are independent rather than being read as a single composite provision. That means that when the reader comes to the second sentence he or she has already understood the first sentence to provide that the applicable law will be the local law of the place of issue, that local practices are to be applied, and that the courts of that place are to have jurisdiction, that is, for the reasons I have referred to, exclusive jurisdiction.

So how does the word “Otherwise” strike the reader in this context? I do not myself think it naturally reads as if it said “Alternatively”. I quite accept that there are contexts when “otherwise” can, as Andrews LJ says at paragraph 44 above, connote a choice. Mr Kealey gave an example in his written submissions (“You can sit here by the wall. Otherwise, you can sit by the window”). But that, as his example illustrates, is because the first sentence itself suggests an option; the second sentence is then naturally read as conferring a second or alternative option, where “Otherwise” in effect means “If you don’t want to sit by the wall”. To give it this meaning in the present context would require the reader to go back and revise their understanding of the first sentence so that rather than directing that jurisdiction was to be in accordance with the courts of the place of issue, it was instead to be understood as conferring a choice of suing in those courts (but still as mandatory in relation to local laws and practices). That seems to me an unnatural and rather jarring way to read the provision.

But if it does not mean “Alternatively”, what does “Otherwise” here mean? I agree that a useful paraphrase for “Otherwise” in this context is “If not”. As this shows, “Otherwise” here introduces a condition. But “If not” does not spell out precisely what the condition is. Does it mean “If you choose not to sue in the local court…”? Or does it mean “If the provision for the local court to have jurisdiction is ineffective…”?

Put like that, I think the latter is the more natural way to understand the provision. The problem with the reading “If you choose not to sue in the local court…” is that it sits uncomfortably with the first sentence which, for the reasons I have given, does not naturally read as if it conferred a choice. There is also the point that one would normally expect such a condition to be followed by another option, such as “…, you may alternatively sue in England”. But here the second part of the sentence is itself in mandatory form. That looks to me more like a mandatory requirement for suing in the local court, supplemented by another mandatory requirement if that provision is ineffective. Put another way, “Otherwise” here means “Failing that”, and that indicates that the second sentence applies where the stipulation in the first sentence fails. As Males LJ says (paragraph 26 above) that in practical terms means if the local court does not or would not accept jurisdiction.

For these reasons I think the natural and ordinary meaning of this not very well drafted provision is that preferred by Males LJ. Once one has reached that conclusion, the remaining question is whether there is any reason not to adopt the natural and ordinary meaning of the provision. On this, and on the other issues that were argued, I agree with Males LJ for the reasons that he gives.

I would therefore allow the appeal and make the other orders proposed by Males LJ.