Lifestyle Equities CV v Hornby Street (MCR) Ltd

This is an appeal against a decision of HHJ Hacon (the “Judge”) who stayed the Appellants’ claim against the Respondents for trade mark infringement and passing off (the “Claim”) pursuant to the power contained in Section 9 of the Arbitration Act 1996 (“Section 9” and the “1996 Act”). The appeal raises a number of issues of conflict of laws in relation to arbitration agreements and the assignment of trade marks.

At least for the purposes of this appeal, the background facts are not in dispute and can be stated relatively shortly.

The Claim relates to the alleged infringement of a UK trade mark (the “UK Mark”) and various EU trade marks (the “EUTMs”) (together the “Trade Marks”) that protect a figurative mark with the following appearance (the “Beverly Hills Logo”) and a word mark for the words “Beverly Hills Polo Club”.

The First Appellant is the registered proprietor of the Trade Marks and the Second Appellant is the licensee of the First Appellant’s rights. Both companies are registered in the Netherlands. Neither has, or ever has had, any business in or connection with the United States of America.

The Beverly Hills Logo was originally used by a Californian entity called BHPC Marketing Inc. (“BHPC Marketing”) which was the owner of a number of US registered trade marks protecting the logo.

In the mid-1990s, a dispute arose between BHPC Marketing and the Eighth Respondent, the Santa Barbara Polo & Racquet Club (“SBPC”). SBPC is a sports and social club established in 1911 which used the following logo (the “Santa Barbara Logo”) in its business and was the owner of a number of US registered trade marks in respect of the same.

To resolve the dispute between BHPC and SBPC, a so-called “co-existence agreement” was entered into between those two parties on 31 June 1997 (the “1997 Agreement”). The 1997 Agreement recited the use by BHPC Marketing and SBPC of their respective logos and their ownership of a number of US trade marks.

Clause 1 of the 1997 Agreement provided as follows,

Clause 4 was in the following terms,

The 1997 Agreement also contained the following supplementary provisions,

In 2007 and 2008 the Trademarks were assigned from BHPC Marketing to another US corporation, BHPC Associates LLC, and then on to a further US corporation, BHPC International LLC. In 2009 the Trademarks were assigned to the First Appellant.

The application before the Judge and the appeal before this Court proceeded on the basis that neither of the Appellants knew of the existence of the 1997 Agreement when they took their assignments of the Trade Marks.

In April 2013 the Appellants became aware that SBPC had applied to register a Community trade mark in respect of the Santa Barbara Logo and stated that they had been informed of the potential existence of the 1997 Agreement. They were then supplied with a copy of the 1997 Agreement by SBPC.

Subsequently, in 2015, the First Appellant applied to the Mexican Trade Mark Office to register a trade mark and wished to show the office that the application had the support of SBPC. On 9 June 2015 a person claiming to act on behalf of the First Appellant sent an email to a representative of SBPC as follows,

A consent letter was then signed by SBPC. It stated that SBPC recognised the right of the First Appellant to use and register in Mexico the mark Beverly Hills Polo Club and the Beverly Hills Logo. The letter also said:

The Claim was issued in June 2020. It complained of infringement of the Trade Marks (and an additional EU trade mark) and passing off by the Respondents having used the Santa Barbara Logo on goods sold in the UK and the EU.

The Particulars of Claim asserted that the use by the Respondents of the Santa Barbara Logo had occurred without the authority of the Appellants. In that regard, and following the point having been raised in pre-action correspondence, the Appellants asserted (i) that they are not parties to the 1997 Agreement, (ii) that they did not know of the 1997 Agreement when they took their assignments of the Trade Marks, and (iii) that by reason of Article 27(1) of Regulation 2017/1001 on the EU Trade Mark and equivalent articles in predecessor regulations, and section 25(3)(a) of the Trade Marks Act 1994 (the “Trade Marks Act”), the 1997 Agreement was not binding on the Appellants when they took their assignment or licence of the Trademarks.

The Respondents applied for a stay of the Claim pursuant to Section 9. That section provides in material part,

Section 9 applies notwithstanding that the designated seat of an arbitration is outside England & Wales: see section 2(2) of the 1996 Act. An “arbitration agreement” is defined by section 6(1) of the 1996 Act as an agreement to submit present and future disputes to arbitration.

The Respondents advanced two grounds under Californian law in support of their application for a stay: (i) that under Californian law the Appellants are bound by the 1997 Agreement as assignees of the Trade Marks, and (ii) that by reason of the Appellants’ reliance on the 1997 Agreement to obtain the letter of consent to registration of trade marks in Mexico, the Appellants are bound by the 1997 Agreement by virtue of the operation of the doctrine of equitable estoppel under Californian law. The Judge had before him expert evidence on Californian law in relation to these two points from two Californian attorneys, Paul Llewellyn (for the Respondents) and Lukas Sosnicki (for the Appellants).

In a judgment delivered on 30 November 2020 (the “Judgment”), the Judge decided to stay the Claim. In paragraph 22 of his Judgment, he identified that the question he had to decide was whether the arbitration agreement in Clause 7 of the 1997 Agreement could be enforced against the Appellants. He then gave three reasons for concluding that the answer to that question was “Yes”. In summary these were,

that as a matter of English law, the Appellants had become parties to the 1997 Agreement as a result of their dealings with SBPC in 2015;

in the alternative, that the question whether the Appellants are bound by the arbitration clause in the 1997 Agreement should be determined by Californian law; and that on the basis of the expert evidence, under Californian law, a settlement or co-existence agreement is a burden attaching to a trade mark and passes with its assignment to become an obligation binding on the assignee, irrespective of whether the assignee knows about the agreement; and

in the further alternative, that the Appellants were precluded from denying that they were bound by the 1997 Agreement under Californian law principles of equitable estoppel.

The Appellants contend that all three reasons were wrong. I shall deal with these points in the same order as the Judge.

The Judge’s first reason was not based on any argument that had been raised by either side. The Judge concluded, however, that, applying English law,

The question of who is a party to an arbitration agreement is a substantive question which depends upon the underlying concept of contractual consensus. Under English conflict of laws principles, it is to be determined in accordance with the governing law of the (putative) agreement: see e.g. Peterson Farms Inc v C&M Farming Ltd [2004] 1 Lloyds Rep 603 at [45], and the more recent decision of the Supreme Court in Kabab-Ji SAL v Kout Food Group [2021] Bus LR 1717 (“Kabab-Ji”). That much was common ground between the parties in the instant case.

In the instant case, by Clause 9, the governing law of the 1997 Agreement was Californian law. However, because it had not even been suggested by the Respondents that the Appellants were parties to the 1997 Agreement, there was no expert evidence of Californian law on the question of whether, and if so, how, the Appellants had become parties to that agreement. For that reason alone, the Judge should not have chosen to express his own view of whether, if the matter had been governed by English law, the Appellants would have been parties to it.

Moreover, even if English law had been applicable, the Judge did not explain which principle of English law he had applied to reach his conclusion. In my judgment, his conclusion was plainly wrong. As a matter of English law, absent some express provisions for accession in the agreement, a person can only become a party to an existing agreement with the consent of all of the other parties, i.e. by novation or by the making of a new agreement: see e.g. Kabab-Ji at paragraph [60]. There was no such provision for accession in the 1997 Agreement and no suggestion, still less any evidence, that the other party to the 1997 Agreement, BHPC Marketing, had ever consented to the Appellants becoming parties to that agreement by novation or a new agreement.

There is a clear conceptual distinction between the question of whether a person is a party to an arbitration agreement, and the question of whether a non-party is bound in some other way by that agreement. The point was identified in general terms by Lord Collins in paragraphs [105]-[106] of his speech in Dallah Real Estate v Ministry of Religious Affairs, Government of Pakistan [2011] 1 AC 763 (“Dallah”),

The Appellants contended that the law to be applied to determine whether they were bound by the arbitration provision in the 1997 Agreement was English law and EU law. The Respondents contended that it was Californian law.

The Judge rejected the Appellants’ submissions which were based upon Rule 135 in Dicey & Morris, The Conflict of Laws (“Rule 135” and “Dicey”). Rule 135 is headed “Assignment of Intangible Things” and provides,

The Judge expressed the view that the relevant parts of Rule 135 were either concerned with the assignability of an intellectual property right (which is not in issue in this case) or that they did not apply to the effect of an arbitration agreement contained in a contract which was not a contract for an assignment of intellectual property rights.

Instead, the Judge accepted the submission on behalf of SBPC that the relevant rule in Dicey was Rule 64. That rule provides,

The Judge held, at paragraph 54 of his Judgment,

Having determined that Californian law should be applied to the issue, the Judge then went on to hold, on the basis of the expert evidence, that under Californian law an arbitration clause in a settlement or co-existence agreement was a burden that attached to a US or non-US trade mark, and that such burden passed with the assignment of the trade mark and was binding on the assignee irrespective of whether the assignee knew of the co-existence agreement. He thus held that the Appellants were bound by the arbitration clause in the 1997 Agreement.

Where there is a potential conflict of laws, the process of determining the law which applies to an issue in a case involves three stages: (i) characterisation of the issue; (ii) selection of conflict of laws rule which lays down a “connecting factor” for that issue; and (iii) identification of the system of law which is tied by that connecting factor to that issue: see Macmillan Inc v Bishopsgate Investment Trust (No.3) [1996] 1 WLR 387 at 391-392 per Staughton LJ, applied in Raiffesen Zentralbank v Five Star Trading LLC [2001] QB 825 at paragraphs [26] – [28] per Mance LJ. As Mance LJ emphasised, this analysis must be done according to English law, but not in an overly mechanistic way, or in a way constrained by particular notions of English domestic law.

As indicated above, the Judge took the view in paragraph 54 of his Judgment that the issue was “whether the effect of the arbitration provision in the 1997 Agreement is such that it binds the [Appellants] and that, “this falls under the heading of “interpretation” of the arbitration agreement”.

I consider that this characterisation of the issue which the Judge had to decide was wrong. Interpretation is the ascertainment of the meaning of a contract or other document. Identifying what the terms of an agreement mean is (of course) important, but the process of interpretation cannot, of itself, answer the question of whether that agreement has any effect on someone who is not a party to the contract.

The point can be illustrated by a simple example. Assume that two parties to a contract (A and B) include in their contract a term in relation to C, who is not a party, which states, “C must pay £100 to each of A and B on 1 January 2022.” The meaning of that clause may be entirely clear, or may be made clear by a process of interpretation of the written document. However, unless there is something more, i.e. some further relevant principle of law which applies so as to impose the obligation that the contract purports to create on someone who is not a party to the contract, C will not be bound by it.

In the instant case, whilst it was therefore important to understand what the 1997 Agreement meant, the fact is that the Appellants were not original signatories and nor had they become contracting parties to that agreement. As such, no amount of interpretation of the terms of the 1997 Agreement could result in the Appellants being contractually bound by any of the obligations in it. Nor did the Judge suggest that any other part of Rule 64 might have been applicable.

Accordingly, even assuming that the provision in clause 12 of the 1997 Agreement that the agreement would be binding on “assigns” could be interpreted to mean that the agreement should be binding on transferees of trade marks covered by the agreement (rather than merely on assignees of the 1997 Agreement itself), without more, that provision in the agreement between SBPC and BHPC Marketing could not bind the Appellants.

Apart from estoppel (to which I shall return below) the only other factor alleged by SBPC to connect the Appellants to the arbitration clause in the 1997 Agreement is that they took assignments of the Trade Marks. As such, in my judgment, the correct characterisation of the issue was whether the assignments of the Trade Marks to the Appellants had the effect that the Appellants became bound by the arbitration clause in the 1997 Agreement.

The Appellants contended that when determining the appropriate English conflict of laws rule to apply to that issue, it is important to have in mind the underlying nature of a trade mark, the rights of action to which it gives rise, and how it can be assigned.

The general nature of a trade mark as an intellectual property right is described in Dicey at paragraph 22-051 in a paragraph that was, in an earlier edition, approved as an accurate statement of the law by Neuberger J in Peer International v Termidor Music Publishers [2004] RPC 22 at [22], (affirmed, [2004] Ch 212 (CA)),

It is therefore apparent that a trade mark is a right created and registered under a particular legal system which entitles the proprietor to make exclusive use of the mark, and to bring proceedings to protect its infringement in the defined territory in which that legal system applies.

So, under English law, a trade mark is an intangible property right obtained by registration of the mark with the registrar: see section 2 of the Trade Marks Act. The proprietor of the trade mark has exclusive rights in the trade mark which are infringed by use of the mark in the UK without the proprietor’s consent: see section 9 of the same Act. An assignment of a UK trade mark must be in writing, and until an application has been made to register the assignment, is ineffective against any person acquiring a conflicting interest in ignorance of it: see sections 24 and 25 of the Trade Marks Act.

The position in relation to an EU Community trade mark is similar. At the time of the assignments in this case the relevant EU law was the Community Trade Marks Regulation (EC) 40/94, which was then codified as Regulation (EC) 207/2009, and has now been amended and recodified in Regulation (EU) 2017/1001 on the European Union Trade Mark. A Community trade mark was (and is) created by registration and has effect throughout the EU. The registered proprietor of the mark is entitled to prevent all third parties not having his consent from (among other things) using in the course of trade within the EU any sign which is identical to the mark in relation to goods or services which are identical with those for which the mark is registered. An assignee of a Community trade mark can only invoke the rights arising from the mark when the transfer to him is registered.

It is against this background that the Appellants submitted that Rule 135 in Dicey was the appropriate choice of law rule to apply to all questions arising in relation to the effect of assignment of a trade mark. The Appellants submitted that becoming the proprietor of a registered trade mark by assignment gives the assignee a right to bring proceedings for infringement against anyone who makes use of the mark without consent within the relevant territory (referred to in Rule 135 as “the debtor”). They contended that all questions of the extent of the assignee’s right to rely upon the trade mark in such proceedings should therefore be governed by the law of the country under which the right was created, under which the transfer was registered, and in the territory of which the right applies (the lex loci protectionis): that is also the law with which the right, by its nature, has its most significant connection.

Mr. St Quintin illustrated his submission by an example using immovable property, which, as Neuberger J indicated in Peer International, has some of the same territorial characteristics as an intellectual property right. Assume that A buys land in England. After the acquisition, B enters on the land. When A brings proceedings against B for trespass, B claims that A’s predecessor in title (C) entered into an agreement with him which was expressly governed by Ruritanian law which gave him the right to enter and stay on the land, and which was expressed to bind purchasers of the land from C.

For the reasons that I have already given, I agree with Mr. St Quintin that the substantive issue in such a case should not be characterised as one of interpretation of the Ruritanian contract between B and C. The issue in the case would be whether, and to what extent A could assert his title as assignee and owner of the land so as to exclude B from it, and in particular whether the rights that he acquired when he bought the land were affected by the prior agreement between B and C. The applicable law to determine that issue would plainly be English law as the lex situs under which the system of registered title to land is created and under which ownership and transfers of ownership of such land are registered.

That hypothetical example does not, however, directly address the issue in the instant case. For that, one must also postulate that the Ruritanian law contract between B and C contained a clause providing for any disputes arising out of, or in relation to, the agreement to be arbitrated in Ruritania. What law would apply to determine whether B could enforce that arbitration agreement against A? Mr. St Quintin contended that it should be the same law as would apply to determine the other effects of the transfer of the property from B to A: Ms. Lane QC asserted that it should be the same law as the parties to the arbitration agreement (B and C) had chosen to govern their agreement.

In support of his argument that a law other than the governing law of the arbitration agreement should apply to the issue, Mr. St Quintin relied upon the decision of Burton J in Egiazaryan v OJSC OEK Finance [2017] 1 All ER (Comm) 207 (“Egiazaryan”). The case concerned a set of joint venture agreements between C and R1 in relation to a hotel development. R1 was a Russian company owned by another Russian company, R2, which was not itself a signatory to the agreements. The agreements were governed by English law and contained a provision for arbitration of any disputes in London. A dispute arose between the parties, and C began arbitration proceedings against R1 and R2. C contended that although R2 was not a party to the arbitration agreement, it could be joined to the arbitration pursuant to Russian law because under Russian law it would, as a parent company, be liable for the contractual obligations of R1.

It was common ground that R2 was not a signatory or party to the arbitration agreement which was governed by English law. However, Burton J held that Russian law was the applicable law to determine whether R2 could be joined to the arbitration. In doing so, he set out and agreed with the submissions of counsel for C (Mr. Smouha QC) as follows,

At paragraph 20, Burton J then summarised and accepted the further submissions of Mr. Smouha QC as to the distinction between the question of who is a party to an arbitration agreement and who can be joined to an arbitration. He explained his reasoning for not applying the governing law of the arbitration agreement with an example as follows,

Burton J then concluded, at paragraph 21,

The relevance of Egiazaryan is in the acceptance by Burton J that where the issue is whether a person who is not a signatory or party to the arbitration agreement is sought to be joined to an existing arbitration, the governing law of the arbitration agreement would not be the applicable law to determine whether they should be joined. In reaching that conclusion, Burton J accepted that it would not be appropriate for English conflict of laws rules to apply the governing law of the arbitration agreement. Instead it would apply another law to determine whether the non-party was bound on the basis of arguments such as agency, assignment or succession.

I consider that Burton J was right to identify the limits on the application of the governing law of the arbitration agreement. Where it is sought to treat a person who is not a party to an arbitration agreement as bound by it, the contractual consensus between the existing parties cannot provide the answer. There must be some other relevant factor to justify that conclusion. English conflict of laws rules would therefore look to a law connected to that other factor to determine the issue, and not to the law governing the arbitration agreement.

In the instant case, as I have indicated, the only additional factor that arguably connects the Appellants to the arbitration agreement in the 1997 Agreement is that they took assignments of the Trade Marks. Since it is plain that the law governing the Trade Marks (i.e. UK or EU law, as appropriate) would apply to the substantive underlying dispute as to whether the co-existence provisions of the 1997 Agreement were binding on the Appellants or not, it seems to me entirely logical that the same law should apply to the question of whether the Appellants are to be treated as bound by the arbitration clause in the same agreement.

For completeness, I should record that Ms. Lane also sought to rely in this respect upon the general principles of arbitration law (summarised in Dicey at paragraph 16-011) that an arbitration agreement contained in an arbitration clause within the body of a larger contract forms a separate and distinct agreement. A similar principle is encapsulated in section 7 of the 1996 Act which provides as follows,

I do not see that this principle or Section 7 assist in identifying the applicable law in the instant case. They are simply designed to save an arbitration agreement from the invalidity of any other terms of the larger agreement in which it is to be found. They do not purport to alter the terms of the arbitration agreement, or (more relevantly) expand the parties to it.

The same point can be made in relation to paragraphs 59-63 of the Judgment. In those paragraphs, the Judge held that even if English law (in particular Section 25 of the Trade Marks Act) might apply to render the 1997 Agreement ineffective against the Appellants for non-registration under the Act, the general principles of arbitration law would mean that the arbitration agreement “is distinct and remains unaffected”. That might be so, but it does not assist in identifying the applicable law at the earlier stage of the analysis, or provide any basis for treating the Appellants as bound by the arbitration agreement.

For those reasons I consider that the Judge was wrong to apply Californian law to the question of whether the Appellants were bound by the arbitration clause in the 1997 Agreement. In my view, the applicable law was English law in respect of the UK Mark and EU law in respects of the EUTMs.

On that basis, it would be unnecessary to consider whether the Judge was right to hold, on the basis of the expert evidence, that under Californian law an arbitration clause in a settlement or co-existence agreement was a burden that attached to a US or non-US trade mark, and that such burden passed with the assignment of the trade mark and was binding on the assignee. However, the point was argued on appeal, and as Lewison and Macur LJJ are of the opinion that Californian law is the applicable law on this second issue (see below), I should express my views on it.

Whilst the Judge’s conclusion was technically a finding of fact as to Californian law, findings of fact on foreign law are not subject to the same restrictions on appeal as ordinary findings of fact. It is open to this court to consider the evidence of foreign law that was before the Judge and decide whether it justified the conclusion to which he came: see Parkasho v Singh [1968] P 23 at 250, per Cairns J, referred to with approval by Megaw LJ in Dalmia Dairy Industries v National Bank of Pakistan [1978] 2 Lloyd’s Rep 223 at 286. That is certainly so where, as here, the Judge made his findings on the basis of written evidence alone, and did not hear any cross-examination of the experts on their reports.

The Appellants did not challenge the Judge’s conclusions that as a matter of Californian law, (i) in general, the terms of co-existence or settlement agreements under which each party agrees to the other’s use of trade marks worldwide could constitute a burden attaching to both US and non-US trade marks, and (ii) that such pre-existing agreements could bind an assignee of the trade marks, irrespective of any lack of knowledge of the agreement on the part of the assignee.

The Appellants did, however, contend that the Judge was wrong to find that these principles could apply where the obligation in question was to arbitrate disputes, as opposed to limitations on reliance on the trade mark. On that point, after setting out the relevant parts of the experts’ reports, the Judge summarised the position of the two experts as follows,

The Judge then summarised the decisions in Mag Instrument and Russell Road and concluded,

The Appellants contend that the Judge thereby wrongly placed the burden on the Appellants’ expert, Mr. Sosnicki, to show that the so-called “general rule” that applied to the terms of co-existence agreements could not apply to arbitration agreements. They contend that there was no such general rule and that Mr. Sosnicki had provided his answer by pointing out that there needed to be a “clear agreement” to waive a right to a jury trial in favour of arbitration.

Mr. Llewellyn had expressed the opinion in his first report that as a matter of Californian law, the Appellants would be treated as having stepped into the shoes of the assignors of the Trade Marks and be bound by burdens in the 1997 Agreement, which he asserted included the arbitration provisions. Mr. Sosnicki’s evidence in response on this point was contained in the following paragraphs of his report, (citations omitted),

I do not accept the Appellants’ criticisms of the Judge’s conclusions on this point. As the Judge correctly identified in paragraph 87 of his Judgment, the difference between the experts was whether taking assignment of a trade mark protected by a contract containing an arbitration clause could qualify as a “clear agreement” to arbitrate on the part of the assignee. However, neither expert was able to point to any Californian case which was precisely on the point. As such, it was entirely a matter of opinion for the experts as to how a Californian court might decide this novel question.

In that regard, Mr. Llewellyn at least expressed the opinion in his first report that the principles to which he referred would apply to the arbitration clause as much as to the other burdens in the 1997 Agreement. In contrast, in the extract that I have set out above, Mr. Sosnicki did not, in terms, disagree with that proposition or express his own opinion. Instead he simply drew attention to what Mr. Llewellyn “might have considered” or “did not analyse” in his earlier report. Given that rather unsatisfactory response from Mr. Sosnicki, I consider that the Judge was entitled to prefer the opinion actually expressed by Mr. Llewellyn.

As a third and final basis for his decision, the Judge held that the Appellants were prevented from contending that they were not bound by the arbitration agreement in the 1997 Agreement as a result of the operation of the doctrine of equitable estoppel under Californian law.

As a preliminary observation, the argument before the Judge, and on appeal, proceeded upon the basis that the issue of whether an estoppel operated to prevent the Appellants contending in these proceedings that they were not bound by the arbitration agreement in the 1997 Agreement should be determined in accordance with Californian law rather than in accordance with English law as the lex fori. I have some reservations about whether that is correct given the nature of the estoppel, but since the Appellants were content to deal with the Judgment on the basis that Californian law was applicable, I shall also deal with it on the same basis.

Commencing at paragraph 97 of his Judgment, the Judge summarised and quoted from the expert evidence on the Californian law of equitable estoppel. The broad thrust of Mr. Llewellyn’s evidence was that an assignee of a trade mark which is the subject of an agreement containing an arbitration clause might find itself estopped from denying that it was bound by that clause if it had sought to exploit other provisions of the agreement. The Judge then also referred to the evidence of Mr. Sosnicki, which was to the effect that the doctrine typically applied to preclude a litigant from pursing a claim that depended on the existence of a contract whilst at the same time seeking to disclaim portions of the contract (including an arbitration clause) that the litigant found unfavourable.

The Judge then considered the case of Namisnak v Uber Technologies Inc 971 F.3d 1088 (2020) upon which Mr. Sosnicki relied, and concluded, at paragraph 104,

The Judge then considered the meaning of “dependent upon or inextricably intertwined with”, before concluding, in paragraphs 106-107,

The Appellants contend that in reaching this conclusion in paragraphs 106-107, the Judge misapplied the test that he had earlier identified in paragraph 104 of his Judgment. I agree.

The Judge made two points in paragraph 106-107. The first was to refer to the Appellants’ conduct in 2015 in representing to the Mexican Trade Mark Office that they were parties to the 1997 Agreement. Given that the Judge had accepted in paragraph 104 that, “the equitable doctrine applies only where the plaintiff's claims are dependent upon or are inextricably intertwined with the obligations imposed by the contract containing the arbitration clause”, it is unclear what point he was making. The Appellants’ conduct in Mexico in 2015 has nothing to do with the claims which they make in the current litigation, and the Judge did not further explain its relevance.

Nor, if it were relevant, do I consider that the Appellants actually “exploited” the 1997 Agreement in their dealings with the Mexican Trade Mark Office. What the First Appellant actually did was to obtain and rely upon a separate letter of consent from SBPC. The First Appellant did not seek to rely upon the 1997 Agreement itself as evidence of SBPC’s consent, as it would have been entitled to do if it had been a party to the agreement: see clause 4 (cited in paragraph 9 above).

Moreover, the representations in the consent letter were actually made on behalf of SBPC alone, and were at very least ambiguous. The letter was on SBPC notepaper and only signed on behalf of SBPC. It also referred to the First Appellant being a party to the 1997 Agreement “through its predecessor entity”. Further, although the consent letter described the 1997 Agreement as having been executed between SBPC and the First Appellant, it would have been obvious to anyone reviewing the attached document that this was not the case.

The second point made by the Judge did address the test that he had identified in paragraph 104. However, although the Judge stated that “[the Appellants’] claims are likely to depend on the protection afforded to the First [Appellant’s] trade marks and the [Appellants’] goodwill by the 1997 Agreement”, that confused the basis upon which the Appellants made their claims with the defence asserted by the Respondents. The Appellants’ claims in these proceedings are not in any way dependent upon any provisions of the 1997 Agreement. The Appellants bring their claims simply as registered proprietors of the Trade Marks. It is only the Respondents who seek to advance a defence of consent and an obligation to arbitrate based upon the terms of the 1997 Agreement. Indeed, the Appellants maintain that their rights to bring their claims in an English court are unaffected by the 1997 Agreement to which they were not parties.

It was clear, moreover, on the expert evidence that the basis of equitable estoppel under Californian law focuses on the conduct of the party who is alleged to be estopped. Accordingly, as Mr. Sosnicki pointed out in paragraph 33 of his report, the Californian courts have held that a defendant’s reliance on a contract as a defence to a claim does not justify applying equitable estoppel as against the plaintiff: see Amergence Supply Chain Management, Inc. v. Changhong Trading Ltd., 2016 WL 8234652.

Nor was the Judge right to consider that the Appellants’ claims are “inextricably intertwined with the obligations imposed by” the 1997 Agreement. As I have already indicated, the Appellants place no reliance on the 1997 Agreement and are not otherwise subject to the obligations imposed by it.

For these reasons, I consider that the three reasons given by the Judge for granting a stay of the Claim were wrong. It follows that I would have allowed the appeal and discharged the stay on the proceedings ordered by the Judge.

However, I have the misfortune to be in the minority on the second issue. Although Lewison and Macur LJJ agree that the Appellants are not parties to the 1997 Agreement containing the arbitration clause, they consider that Californian law, as the governing law of that agreement, should apply to the question of whether the Appellants are bound by that clause, and that the Judge was entitled to reach the conclusions that he did on Californian law. The result is that the appeal will be dismissed.

The difference between us is that Lewison and Macur LJJ view the issue of whether the Appellants are bound by the arbitration clause as one which is essentially contractual in nature, and which, if it were necessary to do so, could be said to relate to the scope of the arbitration agreement within Rule 64 of Dicey: see per Lewison LJ at paragraphs 110-111 and 117 below. In essence Lewison and Macur LJJ consider that the same law should apply to the issue of who is bound by the arbitration agreement as applies to the question of who is a party to the arbitration agreement: see paragraph 112 below.

With respect, I do not see how that can be so. The question is whether a non-party is bound by an arbitration agreement between two other persons. The question is only being asked because the first person is not a party to the arbitration agreement as a matter of the law of contract. It is therefore difficult to see any logical reason, when addressing the issue as a matter of English conflict rules in the manner described in Raiffesen, to characterise it as a matter of contract.

The issue is also conceptually quite different from the issues addressed in Rule 64(1) of Dicey. Issues such as the interpretation and scope of an arbitration agreement all involve determining what effect to give to an agreement between the parties who have consented to it. It is thus entirely logical to apply the law that those parties have chosen to decide issues such as what the arbitration agreement means, or whether a particular dispute falls within its scope: see Enka Insaat ve Sanayi AS v OOO “Insurance Company Chubb” [2020] 1 WLR 4117 (SC) at [137]-[138].

It is also well established in private international law that where there is a dispute about whether an arbitration agreement exists at all or is valid, the general approach is to apply the law that would govern the agreement if it exists or is valid (the so-called “putative governing law”): see Kebab-Ji at [27]. Kebab-Ji decides that the same approach must apply where the question is whether a third person is or has become a contracting party to an existing arbitration agreement. That is logical because the essential issue in such a case is whether the contractual relationship between the existing parties either extends to, or has been modified by the introduction of, the third person by a contractual mechanism such as accession or novation. But the same logic cannot apply to the question of whether the third person is or has become bound by an arbitration agreement by a non-contractual mechanism.

I also do not think that Burton J’s decision in Egiazaryan can be explained and then distinguished on the basis that it raised the question of whether it was “permissible” to look beyond the governing law of the arbitration agreement, and that in the instant case there is “no need” to go that far to determine whether the Appellants are bound: see paragraph 114 below. In Egiazaryan there was an existing arbitration between C and R1, and Burton J applied a law other than the applicable law of that agreement to determine whether R2 should be joined to that arbitration. He did not do so out of necessity. The example which he gave of a Ruritanian arbitration (quoted at paragraph 52 above) made it clear that he did not think it would be appropriate to apply the governing law of the arbitration agreement, even though that would have permitted the joinder of a non-party to the arbitration.

I have had the advantage of reading in draft the judgments of my lords, Snowden and Lewison LJJ.

I agree with my lords that the appellant makes good and proper challenge to the judge’s finding that the claimants had become parties to the arbitration agreement or were estopped from denying the same by reason of communications to third parties following the discovery of the agreement.

However, I would dismiss this appeal since I conclude that the claimant is subject to the arbitration clause in the 1997 agreement which provides that any dispute shall be resolved by arbitration in Los Angeles.

I gratefully adopt Lewison LJ’s analysis of the issue in the appellant’s second ground of appeal and the entirety of his erudite reasoning in this regard.

In far less elegant a fashion than the judgment following, I merely summarise that I conclude that the judge was right to focus upon the arbitration agreement rather than upon the topic of the wider dispute afoot and therefore to identify Dicey, Morris & Collins on the Conflict of Laws 15^th Ed, Rule 64 as applicable to resolve the application before him. Thereafter, he was entitled to find from a reasonable evaluation of the expert evidence that in accordance with Californian state law which governed the question as to who may be bound by such an agreement, the claimants were bound by the arbitration process designated in the 1997 co-existence agreement.

I gratefully adopt Snowden LJ’s account of the underlying facts. The only issue for us to decide is whether the defendants are entitled to a stay of these proceedings because of the existence of an arbitration agreement.

Section 9 (1) of the Arbitration Act 1996 provides:

The defendants (or at least some of them) are party to an arbitration agreement. It is not suggested that we should distinguish between them. Those who are parties are therefore entitled to apply for a stay under section 9. But there is nothing in section 9 (1) (at least on the face of it) which says that the application for a stay may only be made against another party to the arbitration agreement (as opposed to a party to the proceedings). I agree with the judge, therefore, that the defendants are entitled to make the application against the claimants despite the fact that the claimants were not party to the arbitration agreement itself.

Section 9 (4) provides:

I am inclined to accept that if (for whatever reason) the arbitration agreement is not binding on the claimants it can, to that extent, be said to be “inoperative”, at least as against them.

I agree with Snowden LJ that the judge was wrong to hold that the claimants became parties to the arbitration agreement. I also agree that the judge misapplied the test he had formulated in holding that the claimants were estopped from denying that they were parties to it. I have nothing to add to what Snowden LJ has said on those issues. But I find myself unable to agree with him on the second question: are the claimants bound by the arbitration agreement, even though they are not parties to it?

Section 46 of the Act provides:

The effect of section 46 (1) (b) is that the parties are not restricted in their choice to the rules of law of a particular state. Dicey, Morris & Collins on Conflict of Laws (15^th ed para 16-053) explain:

The agreement containing the arbitration clause was made in the context of a trade mark dispute which arose in both the United States and Japan. The immediate purpose of the agreement was to compromise those disputes; but it also went on to make provision for the use of the marks “in any and all countries in the world.” It also provided that the agreement “may be used as evidence worldwide to assist with trademark and/or service mark registration, to show consent to the use of the parties’ respective above-named marks and/or registration of those marks.”

Clause 7 of the agreement provided for any dispute to be resolved by arbitration in Los Angeles, in accordance with the rules of the American Arbitration Association. Clause 9 provided that the agreement was to be “construed and governed in accordance with the applicable laws of California, without reference to its conflicts of law provisions.”

Thus the parties to the agreement expressly excluded the Californian conflicts of law provisions. Doubtless that was because the agreement was intended to have world-wide effect.

Clause 12 of the agreement provided that it should be binding on and inure to the benefit not only of the parties to it but also their “heirs, administrators, successors, assigns, licensees” and others.

The claimants are now the owners of some of the marks covered by the agreement, although they were not originally parties to it. It is accepted, for the purpose of this appeal, that when they acquired the marks they did not know of the existence of the agreement or the arbitration clause. They say that they are not bound by the arbitration agreement. That conclusion, they say, follows from the application of the English conflict of law principles which govern the validity and effect of an assignment of intangible property. Dicey, Morris and Collins summarise the law in rule 135:

It is rule 135 (2) on which they particularly rely.

The defendants, on the other hand, rely on the law applicable to arbitration agreements. That is summarised in Dicey, Morris & Collins in rule 64:

As can be seen from this summary, there is a potential distinction (reflected in rules 64 (1) and 64 (3)) between the law governing an arbitration clause and the law governing the substantive dispute. They are often the same, but need not be. The choice of law applicable to the arbitration agreement is relevant to the question: who decides the substantive dispute? The law applicable to the substantive dispute applies to the resolution of that dispute. We are only concerned with the first question. It is important to keep the two questions distinct. The example given by Mr St Quintin (summarised by Snowden LJ at [47]) relates to the substantive dispute, and not to the question: who decides the dispute?

In my judgment, rule 135 in Dicey, Morris & Collins may be relevant to the substantive dispute between the parties (although I should not be taken as saying that it is). But I do not consider that it governs the scope or effect of the arbitration agreement. In my judgment that is a contractual rather than a proprietary question: compare Mance LJ in Raffeisen at [34]. I consider that that question is governed by the proper law of the arbitration agreement (as stated in rule 64 (1)) which, in this case, is Californian law.

I agree with Snowden LJ that this question is not properly characterised as the interpretation of the arbitration agreement. The rule as stated in Dicey, Morris and Collins is not a statute, and it would be wrong to apply it as if it were. But if it were necessary to shoehorn this question into the rule as stated in Dicey, Morris and Collins, I would prefer to characterise it as an aspect of the scope of the agreement.

As a general proposition, the law governing the validity of the arbitration agreement also governs the question who becomes party to it: Kabab-Ji SAL (Lebanon) v Kout Food Group (Kuwait) [2021] UKSC 48 at [18] (i) and [53] (approving a decision of Sir Michael Burton GBE to that effect: J (Lebanon) v K (Kuwait) [2019] EWHC 899 (Comm) at [19]). Logically, the same principle must apply to the question who is bound by it.

Mr St Quintin relied on the earlier decision of Burton J in Egiazaryan v OJSC OEK Finance [2015] EWHC 3532 (Comm), [2017] 1 All ER (Comm) 207. The claimants in that case were parties to a shareholders’ agreement governed by English law. The subject matter of the agreement was the redevelopment of a hotel in Moscow. One of the parties to the agreement was a Russian corporation. Its parent company was not. The claimants alleged that both subsidiary and parent had engaged in a corporate raid designed to oust them from the project. They alleged that both the parent and the subsidiary were liable under Russian law. The question was whether the parent company (which was not a signatory to the arbitration agreement) could be made a party to the arbitration. The arbitrators held that it could not because it was not liable under the arbitration agreement as a matter of English law. Burton J allowed an appeal from that decision. The difficulty in that case stemmed from the divergence between English law, which was the law of the arbitration agreement and the shareholders’ agreement; and Russian law which made the parent liable for its subsidiary. Burton J held that despite the fact that the arbitration agreement was governed by English law, the parent could nevertheless be joined as a party to the arbitration. Burton J quoted the words of Lord Collins in Dallah Real Estate and Tourism Holding Co v Ministry of Religious Affairs of the Government of Pakistan [2010] UKSC 46, [2011] 1 AC 763:

The question in Egiazaryan was whether it was permissible to look outside the system of law chosen by the parties to govern their agreement. Burton J held that the “applicable law” was not restricted to the law of the arbitration agreement. Here there is no such need. The parties to the arbitration agreement chose Californian law (minus its conflicts rule) to govern their contract including the arbitration clause. Californian law is “the applicable law” at least as the starting point for deciding who may be party to the arbitration.

Having heard evidence from experts in Californian law, the judge held that in accordance with that law the claimants, as successors in title to the marks in issue, were bound by the arbitration agreement. I have no doubt that he was entitled to come to that conclusion on the evidence.

The relevant evidence, which the judge accepted, was Mr Llewellyn’s second statement (paragraph 10) in which he said that Californian law applied to the dispute in question “as a matter of contract”. The judge found at [78] that Mr Llewellyn had rowed back from his original reliance on US federal trade mark law; and that he relied on the application of “the Californian law of contract” to the 1997 agreement. The judge’s ultimate finding at [85] was:

This is not a question of whose law of trade marks applies. It is a question of whose law of contract applies to the separate agreement embodied in the arbitration agreement. As I see it, Snowden LJ approaches the question as a matter of the English law of contract, whereas the Californian law of contract (being the law of contract governing the arbitration agreement) takes a broader view of who is bound by the contract.

As far as the Ruritanian example that Burton J gave in Egiazaryan is concerned, I reserve my opinion on whether it is correct. But assuming that it is, the postulated example is that it is Ruritanian company law, rather than Ruritanian contract law, which deems the parent to be party to the subsidiary’s contract. I can well understand that a contract to arbitrate, governed by Ruritanian contract law, does not without more incorporate Ruritanian company law. But that is not this case.

Since, under Californian contract law, the claimants are bound by the arbitration agreement (even though they were not parties to it) the arbitration agreement is not inoperative. That means that the court is bound to grant the stay. I would dismiss the appeal.