National Bank of Kazakhstan the Republic of Kazakhstan v The Bank of New York Mellon SA/NV London & Ors

This judgment seeks to determine a question which has been “referred” to the English Court by the Belgian Court.

The immediate context in which that unusual, perhaps unprecedented, referral took place was a step taken in Belgium by “the Stati Parties”, the Second-Fifth Defendants who are individuals and companies from Moldova and Gibraltar, in the case of the Fifth Defendant, to enforce a Swedish arbitration award made in their favour against the Republic of Kazakhstan in the sum of about US$506 million. I have been told that the award was in respect of breaches by the Republic of its obligations under the Energy Charter Treaty which “led to the destruction of the Stati Parties’ investments”. The step taken to enforce the award was an attachment or garnishment order issued by the Belgian Court in October 2017 in respect of securities or cash held by The Bank of New

The jurisdiction of this Court was challenged by the Stati Parties but in December 2018 the challenge was dismissed; see [2018] EWHC 3282 (Comm), [2019] BLR 113.

In addition to enforcement proceedings in Belgium there are also enforcement proceedings in Sweden, the Netherlands, Luxembourg, Italy and the USA. There were enforcement proceedings in this jurisdiction which were challenged by the Republic of Kazakhstan but, following proceedings in this Court and in the Court of Appeal, they have been discontinued and the Stati Parties have undertaken not to enforce the award in this jurisdiction; see Stati and others v Republic of Kazakhstan [2018] EWCA Civ 1896, [2019] 1 WLR 897. Thus this Court’s involvement in this matter now arises solely because of the enforcement proceedings in Belgium and the “referral” of a particular matter by the Belgian Court to this Court.

In February 2019 directions for trial were given, including in particular the exchange of expert evidence on Kazakh and Belgian law.

The trial of the issue referred by the Belgian Court took place in late March 2020 in exceptional circumstances, namely, the outbreak of coronavirus. As a result of that outbreak and the restrictions advised or imposed on daily life by governments throughout the world the trial was conducted on a virtual or remote basis. The judge, counsel and solicitors participated from their homes by video link and the witnesses did so from their homes or offices abroad in Kazakhstan, Belgium and the USA. The proceedings could be watched on screen in Court 26 in the Rolls Building and they could also be viewed online (pursuant to the power granted in the Coronavirus Act 2020 to broadcast proceedings). The hearing was conducted without any technical hitch and all parties co-operated to ensure that the hearing took place efficiently and fairly. I am very grateful to the parties, their solicitors and counsel, the witnesses, transcribers, the suppliers of the necessary software and my clerk for enabling a case in the Commercial Court involving international parties and witnesses from several countries to take place notwithstanding the impediments caused by the outbreak of coronavirus.

The National Fund is a sovereign wealth fund whose purpose is “the stable socioeconomic development of Kazakhstan, accumulating financial resources for future generations and reducing the effects of unfavourable external factors on the economy”. It was created by Presidential Decree in August 2000. Its source of funding is tax revenue from the Kazakhstan oil industry and certain other revenues arising from, for example, the privatisation of state entities. There was evidence that it was modelled to some extent on the Norwegian sovereign wealth fund and the Alaska Permanent Fund both of which are funded by oil revenues for the benefit of future generations.

The National Bank of Kazakhstan (the “NBK”) is the central bank of Kazakhstan. It exists and operates under the Law on the National Bank of the Republic of Kazakhstan of March 1995. The experts on Kazakh law agree that the NBK is a legal person separate and distinct from the Republic of Kazakhstan. When performing public law functions (for example the development and implementation of monetary policy) it acts in the name of and as part of the Republic of Kazakhstan, but it acts in its own name when entering into commercial relations with other parties.

Pursuant to a Trust Management Agreement (the “TMA”) between the Republic of Kazakhstan and the NBK dated 11 June 2001 the National Fund is under “the trust management” of the NBK. “Trust management” is a Kazakh law concept. The NBK was entitled to possess, use and dispose of the National Fund for the benefit of the Republic of Kazakhstan. It is common ground that the assets in the National Fund “beneficially belong” to the Republic of Kazakhstan; see paragraph 16 of the Opening Submissions of counsel for the Republic of Kazakhstan. Consistently with this concession the Republic’s expert on Kazakh law, Professor Suleimenov, has stated that the Republic retains an economic and beneficial interest in the assets (see paragraph 131 of his first report). However, he also stated that all of the assets of the National Fund are owned by the NBK and not by the Republic (see paragraph 132 of his first report). That is not common ground. The case of the Statis Parties, supported by their expert on Kazakh law, Professor Maggs, is that the assets are owned by the Republic. It may be necessary to return to this topic later in this judgment.

BNYM holds the cash sums in question pursuant to a Global Custody Agreement (the “GCA”) dated 24 December 2001, the governing law of which is English. It is common ground that “the named contracting parties” to the GCA are the NBK and BNYM; see paragraph 82 of the Opening Submissions of counsel for the Stati Parties. (BNYM was not an original party to the GCA but became party to it by way of a novation on 23 January 2003.)

There is no dispute, I think, that the cash sums held by BNYM pursuant to the GCA are part of the National Fund. The cash sums are held in 15 accounts in US dollars and total US$530 million. In English law they are represented by the liability of BNYM to pay the sums on demand. The entity which holds the right to demand payment is either the NBK (the case of NBK and the Republic) or the Republic (the case of the Stati Parties). The case of the Stati Parties is that the Republic has the right to make the demand either because the NBK entered into the GCA as agent on behalf of the Republic or because, as beneficial owner of the National Fund (of which the cash forms part), it can require the NBK as trustee to make the demand or can do so itself by a “derivative action”.

On 29 September 2017 the Stati Parties applied to the Belgian Court for permission to levy an attachment or garnishment against the Republic of Kazakhstan, “including the National Fund of the Republic” on, inter alia, “all claims that Kazakhstan (including the National Fund) has against BNYM”. This was an attempt to enforce an arbitration award which the Stati Parties had obtained against the Republic in December 2013. The application stated at paragraph 19:

“The Bank of New York Mellon SA/NV (“BNY Mellon”) – garnishee according to the present request – acts as global custodian for the [National Fund] based on which Kazakhstan must have a claim against BNY Mellon relating to the assets in the [National Fund] that BNY Mellon holds for the [National Fund] as full part of Kazakhstan. ”

The Attachment Court in Brussels granted permission on 11 October 2017 and the court bailiff served the garnishment on BNYM on 13 October 2017.

The order was issued and served pursuant to article 1445 of the Belgian Judicial Code which provides:

“Any creditor can, on the basis of authentic or other instruments, through a bailiff, protectively garnish, in the hands of a third party, any amounts owed by such third party to its debtor.”

The scheme of the order is easily understood by an English lawyer familiar with an English garnishee or third party debt order. A leading authority on the Belgian law of execution has described it thus:

“The garnishment is therefore best defined as the attachment in the hands of the seized debtor’s debtor [the garnishee] on what the latter must pay or deliver to the seized debtor………

The debt claim between the garnished debtor and the garnishee is called the subject matter of the attachment.”

In response to the order BNYM declared on 30 October 2017 as follows:

“Although (legal predecessors of) BNYM entered into a global custody agreement dated 24 December 2001 (“Global Custody Agreement”) with the National Bank of Kazakhstan (the “NBK”), which is a “state entity” of the Republic of Kazakhstan …….. the Bank cannot fully exclude that the Republic of Kazakhstan (including the National Fund) has or will have claims on BNYM or that BNYM holds assets of or for the Republic of Kazakhstan (including the National Fund) which are the subject of garnishment in view of its contractual relationship with the NBK and the uncertainties of the legal relationship existing between the latter and the Republic of Kazakhstan.”

Accordingly BNYM froze the assets it held pursuant to the GCA.

On 20 November 2017 the Republic sought an order setting aside the attachment or garnishment order. All grounds advanced for setting aside the order were dismissed in a judgment of 25 May 2018. With regard to the contention by the Republic that the garnishee, BNYM, was not a debtor of the Republic the Belgian Court said:

“The argument that is raised by Kazakhstan is about the subjectmatter and the consequences of the attachment. Kazakhstan’s contention is actually that the garnishment could not have any subject-matter, and that the garnishee still wrongly froze the accounts.

The fact that the garnishee is not the debtor of the seized-debtor is not a ground for the withdrawal of the authorization nor for the lifting of the garnishment that has been authorized. The absence of a debt from the garnishee towards the seized-debtor only leads to the conclusion that the garnishment has no subjectmatter.

In the current case the attachment judge can only consider that the garnishment that has been authorized does indeed have a subject-matter. The subject-matter of the garnishment follows in fact from the declaration of the garnishee. According to the declaration …….[the declaration is then quoted]

The seized debtor is entitled to challenge the declaration from the garnishee before the attachment judge. However, this challenge relates to the debt of the third party and must be referred to the trial court in the proceedings on the merits under Article 1456 2^nd. para BJC.

The competent judge on the merits is, as stated by Kazakhstan itself, the English court who must apply its own national substantive law.”

The Belgian Court noted that the NBK also applied to set aside the order on the same grounds as the Republic and that BNYM sought an order that it was discharged towards the NBK and the Republic. The Belgian Court said:

“Both claims relate to the subject-matter of the attachment, notably whether or not a debt exists from BNYM towards Kazakhstan. Kazakhstan disputes the existence of such debt. The attachment judge cannot and may not settle such dispute, but only the judge on the merits. The judge on the merits is, as already mentioned above the English court who must apply its own national law.”

Article 1456 of the Belgian Judicial Code (“BJC”) to which the court referred provides as follows:

“If the garnishee disputes the debt of which the seizing creditor is seeking payment, the case is brought before the competent court or, as the case may be, referred to the competent court by the attachment judge.”

It is to be noted that article 1456 is premised upon the basis that the garnishee disputes the debt of which the creditor is seeking payment. In this case the garnishee did not do so but accepted that a debt owed to the Republic could not be excluded. It was the award debtor (the Republic) which disputed the debt. I point this out for completeness and because it was noted by one of the experts in Belgian law, Professor Storme. But no substantive point was taken as a result of it.

The scope of the referral by the Belgian Court to this court was debated on the Stati Parties’ challenge to the jurisdiction of this court. In my judgment, [2018] EWHC 3282

The precise ambit of the question “what assets, if any, does BNYM hold for the Republic” was not analysed in any detail either in my judgment or in, I think, counsel’s submissions. Perhaps I should have been clearer in my language but I believe that I had in mind that if matters such as a sham trust were established by whatever law was, in accordance with the English law of conflicts, the applicable law then it might be that the sums payable under the GCA were held by BNYM for the Republic. It was that sort of argument that I had in mind when approving Mr. Malek’s submission that the declarations sought were not limited to liability in contract. In that regard I accept that I could have expressed myself more clearly. For if BNYM holds the cash assets for the Republic the foundation of that liability must rest in contract, namely, the GCA. The cash assets are, as a matter of English law, a liability in debt.

When pleadings were exchanged it appeared that there was no common ground as to the law which was to be applied by the English Court. Thus in their Rejoinder the Stati Parties had not admitted that English law was the applicable law (see paragraph 8). It was said that the applicable law was Belgian and/or Kazakh (see paragraph 9). Reference was made to Belgian law concepts of “simulation” or “pretence” and to the use of a trust structure in an abusive manner (see paragraph 11). Considerable reference was made to Kazakh law and in particular to “the abuse of civil rights” under Kazakh law (see paragraph 21). It was alleged that the TMA was a sham or mock agreement under Kazakh and/or Belgian and/or English law (paragraph 27).

In consequence the List of Issues at paragraph 5(g) included the following:

“Does Kazakhstan have claims or rights against BNYM, or any capacity to enforce the GCA, arising out of any argument based on, or analogous to:

(i)

piercing legal personality;

(ii)

sham trust; or (iii) abuse of law?

under whichever law that governs that question.”

In preparation for the trial the parties exchanged evidence of Kazakh and Belgian law.

Similarly, one issue of Belgian law considered by the Belgian law experts was:

“In what if any circumstances could the Cash and/or the Securities held pursuant to the GCA or any claims or obligations in respect thereof fall within the scope of the Belgian Garnishment Order (a) having regard in particular to: any relevant feature of the legal personality of NBK and its potential equivalence to RoK, (b) having regard in particular to: any relevant features of the legal relationship between NBK and RoK and between each of them and the National Fund including the terms and effect of the TMA; and (c) having regard in particular to: any applicable rule or principle of law based on, or analogous to, piercing legal personality, simulation, sham trust, actio pauliana or abuse of law ?”

The declarations which are now sought by the NBK and the Republic in the proceedings before this court are these:

The fourth declaration was added by an amendment permitted on the first day of the trial. It did not appear to me to add to the issues to be debated. The addition of the declaration in place of an earlier declaration which appeared to criticise the conduct of BNYM also had the advantage that it meant that BNYM could take a lesser role in the proceedings. I was however concerned that the reference to “any system of law” required the court to examine whether under any system of law the Republic had a claim

When I gave a short ruling allowing the amendment I accepted this explanation:

“It seems to me that Mr Malek must be right when he says that the genesis of that phrase "under any system of law" lies in the circumstance that under the law of Belgium the attachment order will have subject matter if, under any system of law, Kazakhstan has a claim against BNYM in relation to the cash deposits. But the relevant system of law, because that is the system of law which governs the GCA, is English law, and that is why the Belgian court referred this question to the English court.”

At a later stage I was referred to the Belgian law experts’ joint memorandum from which it appeared that the phrase “under any system of law” echoed the agreement of the Belgian law experts that “if BNYM owes the Cash ….to Kazakhstan, the Cash ……….fall(s) within the scope of the Belgian Garnishment Order, irrespective of the law under which that claim arose.” It followed that the fact that the debt arose under English law rather than under Belgian law made no difference to the efficacy of the Belgian Garnishment Order. That was later stated in terms by Professor Allemeersch when cross examined at Day 2 p.13 line 9 – p.3 line 12. The question of the debt, being the suggested subject-matter of the Garnishment Order, was referred to the English Court because the governing law of the alleged debt was English.

However, at trial the Stati Parties did not seek to argue that the English law question which has been referred to this court, that is, whether BNYM owes the debt in question to the Republic of Kazakhstan, required any issues such as a sham trust to be addressed. Nor did they seek to establish the factual basis for such issues. They relied upon two arguments in support of their case that BNYM owes its debt to the Republic, namely, the law of agency and the law of trusts. They could have submitted, if they had wished, that in English law questions such as a sham trust were relevant to the question whether BNYM owed a debt to the Republic or that the foreign law on such topics was required to be assessed pursuant to the English law of conflicts but they chose not to do so. No such case was advanced either on the facts or in law. I cannot say why such a case was not advanced. The fact is that no such case was advanced either in fact or in law.

In the light of my ruling on the first day of the trial I was surprised to read in the Closing Note of counsel for BNYM (which was the first closing submission that I received) that it was essential that:

“the Court does not limit its consideration of the question of subject-matter to whether RoK has any (contractual) claim against BNYM under the GCA as a matter of English law. The Court should decide whether RoK has any claim against BNYM under any applicable system of law. The only potentially relevant laws identified by the Claimants and/or the Stati Parties are English, Belgian and Kazakh law. ”

In his oral submissions counsel explained, by reference to the pleadings and in particular paragraphs 6 and 11 of BNYM’s Defence, that this had always been BNYM’s position.

Counsel for the NBK and the Republic allied themselves with this approach and said that all issues pleaded in the Rejoinder were “in play”. In their Closing Submissions at paragraph 9 they said:

“The pleaded and agreed issues were not formulated in a way that restricted them to matters of English law. On the contrary, the Reply and the Rejoinder set out in detail the parties’ respective positions on the relevance and content of Belgian and Kazakh law on the central question of subject matter. The parties’ experts have given their opinions on an agreed list of questions of Belgian and Kazakh law, including (in the case of the Belgian experts) a broad formulation of the central question: “In what, if any, circumstances could the Cash and/or the Securities held pursuant to the GCA, or any claims or obligations in respect thereof, fall within the scope of the Belgian

Garnishment Order?”

Counsel for the Stati Parties, notwithstanding the scope of their Rejoinder, but consistently with my judgment on the jurisdiction challenge, stated their position as follows in their Closing Submissions at paragraph 9:

“The Stati Parties’ position is that the proper scope of this trial is whether RoK has a claim against BNYM under the GCA, as a matter of English law, taking into account foreign law only in so far as that is permissible and required by English conflict of law rules.”

In my judgment foreign law issues arise where the English Court applies, as part of the English law of conflicts, a foreign law to a particular issue. For example the relationship between the NBK and the Republic is governed by Kazakh law. It was for this reason that, when I gave judgment on the jurisdiction challenge, I had referred (in the passage which I have already quoted above) to “issues analogous to the issues of piercing legal personality, sham trust and abuse of law which the Stati parties have raised in their written submissions in Belgium” which were to be “determined by this court, applying what it determines to be the applicable law.” Counsel for the NBK and the Republic and counsel for BNYM, in support of the submission that the court should consider the liability of BNYM to the Republic under Belgian or Kazakh law as well as English law, relied upon the penultimate sentence of paragraph 33 of my judgment on the jurisdiction challenge wherein I had approved Mr. Malek’s submission that the question before the court was “not limited to any liability of BNYM to RoK in contract: it includes any liability to RoK relating to the assets”. I have already accepted that I could have expressed myself more clearly in this regard but my approval of Mr. Malek’s submission was certainly not intended to say that the question before the court involved consideration of the question whether BNYM had any liability to the Republic relating to the assets under any system of law. That would have been a surprising question for this court to entertain. I did not understand Mr. Malek’s submission to have that effect. As I have explained above, I believe that I had in mind that if matters such as a sham trust were established by whatever law was, in accordance with the English law of conflicts, the applicable law then it might be that the sums payable under the GCA were held by BNYM for the Republic.

At the end of the trial it appeared that counsel for the NBK and the Republic accepted my understanding of the circumstances in which foreign law would be relevant:

“So your Lordship is right when he says that the court needs to decide foreign law issues to the extent that English conflict of law rules point to the foreign law, and that is clearly right.”

Counsel for the NBK and the Republic and counsel for BNYM said that this was an unsatisfactory state of affairs because it meant that the decision of this court might not bring finality. They suggested that the Stati Parties were intent on running further arguments in Belgium with regard to persuading the Belgian Court that even if this court decided that the Republic had no claim on BNYM to the debt the Belgian Court could properly be asked to consider whether there were other reasons why the garnishment order had subject-matter. It seems that that is indeed the intention of the Stati Parties. Counsel for the Stati Parties told me that further hearings in Belgium had

Counsel for the NBK and the Republic and counsel for BNYM invited this court to resolve, finally, the question whether the debt was owed to the Republic and, if the court held that the debt was owed not to the Republic but to the NBK, to give judgment on that debt claim to the NBK (or, as BNYM preferred, to give BNYM a period of time after the court handed down judgment in this matter to pay the debt) because the Belgian garnishment order would have been revealed to have no subject-matter and would be at an end. Counsel for the Stati Parties said, as I have already noted, that the final determination of the attachment or garnishment order was a matter for the Belgian courts.

The submission made by counsel for the NBK and the Republic that this court should finally determine whether or not the garnishment order had subject matter relied heavily on language used by me in my judgment on the jurisdiction challenge, specifically in paragraph 36 where I said that I was “unable to accept the submission that the Belgian Court has not in substance referred the question of the content of the attachment order to this court”. However, I was not addressing in that paragraph the question whether issues of Belgian law concerning the ultimate determination of the garnishment order should be determined by this court. I was addressing a submission “that there had been no referral of any question to this court and that the Belgian court had determined that the attachment order did have subject matter, on the basis of the BNYM declaration” (see paragraph 34 of my judgment). In the rest of paragraph 36 I made clear that this court was concerned with “the existence of a chose in action held by BNYM(L) for the RoK”. That was the issue which had been referred to this court and which will have a bearing on the question whether the garnishment order has subject-matter.

I have no doubt that this court cannot, as it were, purport to determine the outcome of the attachment or garnishment proceedings in Belgium. Once this court has determined the question referred to it, it must be for the Belgian court to determine the consequences of that decision in Belgium. Belgium is where the Stati Parties have taken steps to enforce the arbitration award by a garnishment order and Belgium must be the place where they are determined. This court’s role is to answer a question which the Belgian court has referred to this court. It is not to determine the outcome of the Belgian garnishee proceedings, notwithstanding that its determination of the issue referred to it may have a significant and possibly decisive role in bringing those proceedings to an end. It would be contrary to comity, that is, the respect which this court has for the procedures of another jurisdiction, for this court to assume responsibility for the determination of the garnishee proceedings in Belgium. In my judgment it is unthinkable that this court would presume to do that.

There are, as I understand the position, two broad matters in particular which the Stati Parties wish to argue before the Belgian Court.

The first is an argument developed by Professor Storme, the expert on Belgian law instructed by the Stati Parties, that the cash, being part of the National Fund, is within

I found Professor Storme’s opinion difficult to accept, for these reasons. First, article 1445 appears to focus on what is owed to the judgment debtor, rather than on what is owned by the judgment debtor. Second, if the garnishment had subject matter because the creditor had a right of recourse to all the assets of the debtor and, as noted in the

The second argument which the Stati Parties wish to advance in Belgium is that by reason of a sham trust or indeed fraud BNYM owes the debt to the Republic; see paragraph 21 of their Closing Submissions. In the event, although it had been anticipated by the NBK and the Republic (and by me) that these arguments would be advanced in this court, they were not advanced. The NBK adduced evidence from one of its deputy governors, Ms. Aliya Moldabekova, that she was unaware that the TMA was “just a sham or simulation or pretence” and that so far as she was aware it was no part of the purpose of the TMA to create a “mechanism for shielding RoK assets from creditors” and that the TMA had not been used for that purpose. But she was not crossexamined on that evidence. In circumstances where these allegations have not been advanced in this trial the court has necessarily considered the question as to whom BNYM owes the cash sums covered by the GCA on the factual basis that there has been no fraud, sham, simulation or pretence. The Stati Parties, if they consider that such matters are relevant to the question whether in English law a debt was owed by BNYM to the Republic, could have sought to establish the suggested sham but they chose not to do so. If the Stati Parties wish to allege in Belgium that there was fraud, sham, simulation or pretence and that on that basis in English law the cash was owed to the Republic it will, I think, be a matter for the Belgian Court to determine whether it is open to the Stati Parties to advance such arguments or whether they are estopped from advancing such arguments on the grounds of res judicata (as that principle is understood in Belgian law). It is the principle of res judicata which brings (desirable) finality to proceedings. But the application of that principle in the present case must be a matter for the Belgian Court to decide. It is improbable that the question whether under any other system of law BNYM may be regarded as owing a debt to the Republic would be regarded as relevant by the Belgian Court in circumstances where that court has referred the question of debt to this court. But again that must be a matter for the Belgian Court.

There is no dispute that in English law the relationship between a bank and its customer is that of debtor and creditor. The customer has a chose in action which entitles him to draw upon the credit balance on demand; see paragraph 75.1 of the Opening Submissions of counsel for the Stati Parties. Thus, in circumstances where it is common ground that the NBK is the named contracting party with BNYM the Stati Parties have to explain why in English law the Republic of Kazakhstan has a right to draw upon the credit balance held by BNYM. The Stati Parties seek to do so in two ways, first, by saying that the NBK entered into the GCA as the agent of the Republic of Kazakhstan and, second, by saying that the NBK holds the debt on trust and that the Republic of Kazakhstan can either compel performance of the obligation owed to the NBK or bring a derivative claim in its own name; see, again, paragraph 75.1 of the Opening Submissions of counsel for the Stati Parties.

The GCA states that the NBK is party to the GCA. The NBK is described as the Client who appointed BNYM as “banker to the Client” (clause 2(b)), instructions were “from the Client” (clause 4(a)), warranties were given by “the Client” (clause 13) and cash in the cash account was a debt owed by BNYM “to the Client” (paragraph 16(j)). Thus there was no suggestion that the Republic was in fact the Client or that the Client was acting on behalf of the Republic. Counsel for the Stati Parties pointed to recital A to the GCA which provided that the “Client is carrying out certain trust management services with respect to the certain securities of the Republic of Kazakhstan (“the National Fund”) in accordance with the Trust Management Agreement…”. But little reliance can be placed on this when recital B referred to the GCA as being between “Client” and BNYM and recital C stated that the GCA set out the terms upon which BNYM “will hold ….Cash of the Client as …banker”. Thus the recitals read together cannot fairly be read as evidence that the NBK was to enter the GCA as agent for the Republic.

However, English law permits a person not described as a party to adduce evidence that he was in fact the principal of the named party who made the contract on his behalf. This principle was not challenged in the present case and is well established, notwithstanding that it fits uneasily with the usual way in which contracts are objectively construed; see Teheran-Europe Co. Ltd v ST Belton (Tractors) Ltd. [1968] 2 QB 545 at p.552G (per Lord Denning MR who said the law relating to undisclosed principals was based on “business convenience”), The Magellan Spirit [2016] EWHC 454 (Comm), [2017] 1 A11 ER (Comm) 241, at paragraph 15 (Leggatt J.) and Filatona and Deripaska v Navifator and Chernukin [2020] EWCA Civ 109 at paragraphs 37 and 41 (Simon LJ).

The relevant principles were summarised in Siu Yin Kwan (Administratrix of the estate of Chan Ying Lung deceased) and anor v. Eastern Insurance Co. Ltd [1994] 2 AC 199 by Lord Lloyd at 207D:

“For present purposes the law can be summarised shortly. (1) An undisclosed principal may sue and be sued on a contract made by an agent on his behalf, acting within the scope of his actual authority. (2) In entering into the contract, the agent must intend to act on the principal's behalf. (3) The agent of an undisclosed principal may also sue and be sued on the contract. (4) Any defence which the third party may have against the agent is available against his principal. (5) The terms of the contract may, expressly or by implication, exclude the principal's right to sue, and his liability to be sued. The contract itself, or the circumstances surrounding the contract, may show that the agent is the true and only principal.”

In order for a person to be bound by a contract as principal the person said to have made the contract on his behalf must have actual authority to do so. This is clear from the above summary and from the following quotations drawn from more recent authority to be found in Filatona and Deripaska v Navigator and Chernukhin [2019] EWHC 173 (Comm) at first instance:

In assessing whether actual authority has been conferred upon an agent by an owner of property to act in the name of the owner as principal it is necessary to distinguish between authority in the sense of permission to use the property and authority to bind as agent. That distinction explains why a ship owner permits a demise charterer (who has possession of the vessel) to repair the vessel, yet is not bound by the repair contract. Authority from an owner may however, in appropriate circumstances, encompass both senses of authority; see Turks Shipyard v The Owners of the Vessel NOVEMBER [2020] EWHC 661 (Admlty) at paragraphs 17-19.

Actual authority may be expressed in words or inferred from conduct. In The Magellan Spirit Leggatt J. explained at paragraph 29 what was required when there was no express agency relationship and conduct was relied upon from which to infer an agency relationship.

“In principle what must be shown is conduct from which (i) a reasonable person in the position of Mansel would have understood that it was authorised to enter into the charter as agent of VSA and (ii) a reasonable person in the position of VSA would have understood that Mansel was agreeing to do so. As in any case where an agreement is sought to be implied from conduct, it is not enough to point to conduct which was consistent with an agreement or mutual intention that Mansel would contract as agent of VSA. It is necessary to identify conduct which was only consistent with such an agreement or mutual intention and inconsistent with any other intended relationship between the two Vitol Group companies. Put another way, it must be fatal to the implication of an agency relationship if the parties would have or might have acted as they did in the absence of such a relationship: see, by analogy, cases such as The "Aramis" [1989] 1 Lloyd's Rep 213 and The

"Gudermes" [1993] 1 Lloyd's Rep 311. ”

Thus, in order for the NBK to have entered the GCA as agent for the Republic the NBK must have had actual authority to do so and must have intended to contract on behalf of the Republic. There has to have been an agency relationship between the Republic as principal and the NBK as agent with regard to the NBK entering the GCA on behalf of the Republic. What must be found is assent by the Republic to the NBK acting on behalf of the Republic so as to affect its legal relations with third parties, in this case BNYM, and assent by the NBK to act pursuant to that assent of the Republic; see Bowstead and Reynolds on Agency 21^st ed. at paragraph 1-001. Whether that relationship is established does not depend upon the parties themselves regarding it as an agency relationship or on the label chosen by the parties to describe their relationship but on what the parties have in substance agreed; see UBS AG v Kommunale Wasserwerke Leipzig GmbH

Where an English Court has to determine questions of authority in a case involving persons from countries other than England the applicable law must be determined in accordance with the English law of conflicts. In assessing what the parties have in substance agreed the applicable law, in my judgment, is that which governs the relationship between the suggested principal and agent. Although this proposition is not said to be certain, I accept the observation in Dicey, Morris and Collins on the Conflict of Laws 15^th ed. at paragraph 33-439 that “it accords with principle to say that whether A had actual authority and the scope thereof depends on his contract with P.” The relevant law in the present case is the law of Kazakhstan which is the law which governs the relationship between the Republic and the NBK and is the applicable law of the TMA. However the significance of the principal’s conferral of authority, what is meant by “actual authority”, will depend upon the law which governs the contract made by the agent with the third party; see Dicey et al at paragraph 33-440. That law in the present case is English law.

The Opening Submissions of counsel for the Stati Parties, at paragraph 83, rely upon a number of matters in support of the suggested agency relationship. The Stati Parties developed their case at paragraphs 30-58 of their Closing Submissions. It is necessary to consider the matters there relied upon but before doing so it is to be noted that it is difficult to find any passage in the opinion of Professor Maggs, the Stati Parties’ expert on Kazakh law, which supports their case on agency. This is surprising given that it is accepted that the law applicable to whether the NBK had authority to enter into the GCA on behalf of the Republic is Kazakh law; see paragraph 30 of the Stati Parties’ counsel’s Closing Submissions.

At section B of his report Professor Maggs, who gave his oral evidence clearly, concisely and with candour, said:

“Whether or not NBK had the “legal authority” from Kazakhstan to enter into this contract [the GCA] depends upon (1) whether or not the TMA was valid under the law of Kazakhstan and (2) if it is was valid, whether or not under the provisions of the law of Kazakhstan the TMA lawfully granted the authority to NBK to enter into the GCA.”

No submission was made to me that on the facts the TMA was not valid for the reason postulated in Section C of Professor Maggs’ report. In section E Professor Maggs considered the nature of the legal relationship between the NBK and Kazakhstan but did not in that section suggest an agency relationship. Instead he made clear his opinion that the transfer of the National Fund into the trust management of the NBK did not involve a transfer of title from the Republic to the NBK. In section G Professor Maggs considered the question whether the cash (held by BNYM) was the property of or held for Kazakhstan. He said that the cash was the property of Kazakhstan and was held by the NBK as entrusted manager under the TMA. No express reference was made to any relationship of agency when entering the GCA. In section I Professor Maggs addressed the question, in what if any circumstances could Kazakhstan have claims or rights against BNYM in relation to the cash, or any capacity to enforce the GCA? A number of matters were considered (abuse of law, sham trust and so forth) but no mention was made of an agency relationship. Since the relationship between the Republic and the NBK is governed by Kazakh law this is, it seems to me, a striking omission, especially so given that Professor Maggs was invited to consider in terms “any relevant features of the legal relationship between NBK and Kazakhstan”.

By contrast Professor Suleimenov, who also gave his oral evidence clearly, concisely and with candour, dealt with the matter expressly and concluded (at paragraph 85 of his report) that “when NBK entered into the GCA with BNYM it did so on its own behalf and not on behalf of RoK”. That opinion was based upon the view that by entering into a contract with BNYM for the provision of commercial services the NBK was not performing a public law function but was acting in its own name as a legal entity. Later (at paragraph 87) reference was also made to article 883 of the Civil Code which provided

“By establishing the trust management of property, the trust manager assumes an obligation to manage, in its own name, the property transferred for its possession, use and disposal in the interests of the beneficiary, unless otherwise specified in the agreement or in legislative acts.”

It is also to be noted that in the experts’ joint report it was agreed (at paragraph 8.6) that “in entering into commercial and other civil law relations, the NBK acts in its own name as a state institution”. The experts noted that they disagreed on the question of the independence of the NBK (see paragraphs 9-11) but Professor Maggs again made no express reference to any agency relationship.

In so far as an agency case was pleaded by the Stati Parties the only reference to agency was in paragraph 15 of the Rejoinder which responded to paragraph 17 of the Reply which in turn concerned the relevance of the issue of ownership of the National Fund. The Stati Parties pleaded that Kazakh public law concepts were relevant in the circumstances of the GCA and TMA. It was then pleaded:

“The TMA is a contract between a state entity, namely, the Ministry of Finance of the Republic of Kazakhstan on the one hand and on the other hand NBK as an agent, institution or organ of the Government of Kazakhstan.”

If this was intended to be an allegation that the NBK entered the GCA as the agent of and on behalf of the Republic so that the Republic could sue on the GCA it was a most oblique way of doing so.

With that introduction I return to the matters relied upon by counsel for the Stati Parties.

First, reliance is placed on provisions of Kazakh law and in particular upon article 26 of the Law on the National Bank of the Republic of Kazakhstan which provides as follows:

This article had a late appearance in the case. It was first mentioned by Professor Maggs in his supplementary report when he noted (at paragraph 9) that in an earlier case Professor Suleimenov had said, with regard to trust management by the NBK, that “it is unclear why we cannot apply Article 26 of the National Bank Law ………..to such relations.” Professor Maggs mentioned this, not in the context of a suggested agency relationship, but in the context of Professor Suleimenov’s opinion that the NBK was the owner of the National Fund. It was counsel for the Stati Parties who first relied upon the article in support of an agency relationship and noted (at paragraph 83.2 of his Opening Submissions) that in the present case Professor Suleimenov disputed that the NBK acts as an agent for the Republic “with no further explanation”.

However, Professor Suleimenov did explain his opinion by referring (i) to the fact that the NBK was not exercising a public function when entering into the GCA and (ii) to article 883 of the Civil Code.

Professor Suleimenov’s earlier comment was made in another case in this court when an attempt was made by claimants who had obtained an award against the Republic of Kazakhstan to obtain a Third Party Debt Order on cash sums held to the order of the NBK by BNYM; see AIG Capital Partners v Republic of Kazakhstan (National Bank of Kazakhstan intervening) [2006] 1 WLR 1420. Read literally article 26 appears to suggest an agency relationship between the Republic and the NBK. However, whether that was the meaning Professor Suleimenov intended to convey by his comment I do not know. His comment was made whilst addressing the opinion of another expert in the case, to which opinion I was not referred. It is therefore difficult to assess the true significance of Professor Suleimenov’s comment. Aikens J. held in the earlier case that the cash account in question was a debt owed by BNYM to the NBK; see paragraphs 30-32 of his judgment. It does not appear to have been argued that as a result of article 26 the NBK acted as agent for the Republic with regard to the GCA.

Although it can be said that article 26 of the Law on the National Bank appears to support the relationship of agency, it is necessary to view it in its context. It appears in a section of the Act entitled “Interaction of National Bank of Kazakhstan with State

Thus article 21 provides that the NBK is independent in its activities and article 22 provides that the Government “shall not be liable for the obligations of the National Bank of Kazakhstan, as well as the National Bank of Kazakhstan shall not be liable for the obligations of the Government, except when it takes on that responsibility.” Article 23 provides that the NBK may act as “as a bank, financial adviser and agent of the state bodies in agreement with them.” Article 24 deals with the NBK as a bank, article 25

Thus it is clear that the NBK may act as an agent (as accepted by counsel for the NBK and the Republic in their reply submissions) but article 26 also refers to the terms which have been agreed between the NBK and the Republic. That takes one back to the terms of the TMA. Article 22 provides that the Government shall not be liable for the obligations of the NBK “except where it takes on that responsibility”. Thus again one is directed back to the TMA, the agreement between the Republic and the NBK, to see whether there is anything in the TMA which provides that the Republic assumed liability for the obligations of the NBK. To similar effect is article 883(1) of the Civil

So it is necessary to consider the provisions of the TMA. The applicable law of the TMA is Kazakh law but there was no evidence that Kazakh law contracts should be construed in any different way than English law contracts.

Counsel for the Stati Parties submitted that the TMA confers “express authority” on the NBK to carry out trust management for the benefit of the Republic by way of investing financial assets by, amongst other things, entering into the GCA. If the authority was not express then it was, as it was put in oral closing submissions, “clearly implied”.

I have noted and read the clauses of the TMA on which reliance is placed at paragraph 34 of counsel’s Closing Submissions. Whilst they, or some of them, are consistent with the suggested agency relationship I am unable to accept that they manifest the grant to the NBK, expressly or by implication, of authority to enter into the GCA as agent for the Republic. Clause 1.1 makes clear that the NBK is to carry out trust management of the National Fund “for the benefit” of the Republic and to the same effect is clause 7.1 which states that “the beneficiary of this Agreement” is the Republic. However, such provisions, whilst perhaps consistent with an agency relationship, do not themselves

Reliance was placed on clause 2.1 which empowers the NBK to possess, use and dispose of the National Fund and to carry out investments of the National Fund. But this says nothing about acting as agent on behalf of the Republic or about binding the Republic to contracts made by the NBK with third parties. I should add that it refers to the NBK independently carrying out investments of the National Fund, a word omitted by counsel in their quotation from clause 2 and not commented upon. Of course, the TMA, and in particular clause 2, grants the NBK authority from the Republic in the sense of permission to deal with the National Fund and in so doing to enter contracts such as the GCA. But there is nothing in the TMA which authorises the NBK to enter such contracts as agent on behalf of the Republic so as to bind the Republic to the GCA. The word independently sits unhappily with authority in that sense. It is not suggestive of the required “closeness of identity”.

Reliance was placed on clauses 2.2.1 and 2.2.2 which oblige the NBK to carry out trust management of the National Fund “in accordance with the normative legal acts of the Republic” and to inform the Republic of the external managers or custodians of the

Reliance was placed on clause 2.2.5 which obliges the NBK to transfer money from the National Fund in accordance with the Republic’s instructions. This does not appear to me to say anything about whether the NBK enters into contracts with third parties as agent for the Republic.

Reliance was placed on clause 2.3 which provides for the NBK “to bear responsibility on its own behalf of the transaction obligations made by the Bank in excess of the powers conferred upon it or in violation of established restrictions.” It was not explained how this was said to advance the case of the Stati Parties. But it could be said that the (unstated) corollary of this provision is that the NBK has no responsibility on its own behalf in respect of transactions made within its powers because in such cases the Republic, as principal, was responsible. However, the clause as a whole is dealing with the liability of the NBK for losses in two eventualities. The first is where losses have been caused by “improper performance of its obligations”. The second is where the NBK acts “in excess of the power conferred upon it”, that is, without the permission of the Republic. I consider that read as a whole the clause is dealing with the liability of the NBK for losses in a manner which is consistent with the NBK entering into transactions regarding the National Fund as principal.

Reliance was placed on clause 3.1.2 which empowers the Republic to make decisions that regulate the activities of the National Fund in collaboration with the NBK. This clause appears to provide for the Republic to give instructions to the NBK concerning investments in or concerning the National Fund. Again, it is probably consistent with an agency relationship but I do not consider that it establishes such a relationship.

Finally, reliance was placed on clause 7.4 which provides for the TMA to be terminated by the President of the Republic. Even if this clause is to be construed independently of

In addition to article 26 and the provisions of the TMA counsel also relied upon article 889 of the Civil Code which concerns “entrusted management” and authorises the manager to instruct another person to take the action necessary to manage the property.

Indeed, Professor Maggs did not suggest an agency relationship. If it were the position that under Kazakh law the Republic as principal had authorised the NBK to enter agreements such as the GCA as agent on its behalf one would have expected Professor Maggs to have mentioned that when asked whether there were any circumstances in which the Republic could have a claim against BNYM in relation to the cash, “having regard to any relevant features of the relationship between the Republic and the NBK”. Instead of doing so he made reference to such matters as “abuse of right” over several pages. The relationship of principal and agent, if that were the position in Kazakh law, would have been a much simpler answer to the question.

Counsel further submitted that Professor Suleimenov had confirmed, when crossexamined, that the NBK entered into the GCA with “the authority” of the Republic. I have read the passages of cross-examination on which reliance is placed and which are set out in counsel’s Closing Submissions. I consider that counsel is reading too much into them. Of course the NBK had been granted, by means of the TMA, authority to manage the National Fund by entering into contracts such as the GCA. Thus the NBK had not, as it were, dealt with the National Fund in that way without the Republic’s permission. But that is not the same as conferring authority on the NBK to bind the Republic to the GCA as principal. It does not amount to an “assent that [the NBK] should act on [the Republic’s] behalf so as to affect his legal relations with third parties [such as BNYM]”.

The second matter on which counsel relied in support of their case on agency was the control which the Republic was able to exercise over the NBK in the performance of the GCA. Professor Suleimenov accepted that the Republic as “founder” of the trust management in relation to the National Fund could exercise influence over the nature of the relationships entered into by the NBK. Further, Professor Suleimenov did not dispute that the Republic could exert influence over the NBK by means of its “autocratic” power in Kazakhstan. The Republic also had powers to terminate the TMA; see clauses 7.3 and 7.4. There was a dispute as to their true construction but even if the power was limited in the manner suggested by counsel for the NBK and the Republic, the latter’s “autocratic” powers would probably suffice for it to achieve whatever the Republic wished. Ms. Moldabekova, a deputy governor of the NBK, accepted that the NBK required the approval of the Republic to investments. The Republic could also require sums to be paid out of the National Fund for the purposes of the state’s budgetary requirements. Finally, clause 2.2.7 of the TMA required the

The third matter relied upon was the payment of the NBK by way of a commission and reimbursement of expenses; see clause 2.1.3 of the TMA. This is again consistent with an agency relationship but not, I think, sufficient to demonstrate such a relationship.

The fourth matter relied upon was that the NBK, in circumstances where the trust management was for the benefit of the Republic (see clauses 1.1 and 7.1 of the TMA), owed fiduciary duties to the Republic. There was, I think, no evidence that in Kazakh law such duties were owed. But assuming that they were, they are consistent with an agency relationship but do not establish such a relationship.

As with all important findings of fact and especially where the establishment of an agency relationship depends upon what parties “in substance” have agreed it is necessary to stand back from the detail and look at the matter in the round. As stated by Allsop P. “…any arrangement must be understood and characterised by reference to its legal terms in context”. Although one necessarily considers the matters relied upon to establish agency separately when analysing them, it is necessary to view them collectively. The matters relied upon are (i) authority in the sense of permission by the Republic to the NBK that the latter may enter into the GCA, (ii) influence and control of the NBK by the Republic in matters concerning the GCA, (iii) payment of the NBK for its services and (iv) the (assumed) fiduciary duties owed by the NBK to the Republic. These matters are, as I have said, when considered individually, consistent with an agency relationship but they are also consistent with the NBK entering the GCA as principal. The question is whether, when considered collectively and in the light of the context in which they are found, they establish the suggested agency relationship in substance, if not in name.

The agreed position between the two experts on Kazakh law is that the NBK is the central bank of Kazakhstan, a legal person separate and distinct from the Republic, which, when entering commercial relations acts in its own name as a state institution. In that context, and looking at the matter objectively, the Republic would expect that the NBK, when entering into commercial transactions with financial institutions such as BNYM, would do so on its own behalf. The NBK was not a minor entity or functionary but an independent central bank. Article 21 of the Law on the National Bank states in terms that the NBK is “independent” in its activities and, consistently with that article, clause 2.1.2 of the TMA provides that the NBK has the right to

“independently carry out investments of the Fund”. I accept that the NBK could enter such transactions as agent for another state body or indeed for the Republic so as to bind them to the transactions. That appears to be provided for by articles 23 and 26 of the Law on the National Bank, and by article 833(1) of the Civil Code dealing with trust management. But whether such an arrangement was agreed with the Republic with regard to the GCA, with the result that the NBK had actual authority from the Republic to bind the Republic to the GCA, depends upon the terms of the TMA (as contemplated by articles 22 and 26 of the Law on the National Bank). Although the terms of the TMA reflect the control which the Republic has over the activities of the NBK (clause 3.1), the fee payable to the NBK (clause 4.1) and the fiduciary duties it is assumed to owe (clauses 1.1 and 7.1) I am unable to accept that read as a whole, and bearing in mind that the suggested agent is an independent central bank, that they manifest an assent by the Republic to the NBK acting on its behalf so as to affect its legal relations with third parties. The terms make it clear that the NBK acts with regard to the National Fund for the benefit of the Republic. But that is consistent with the NBK being liable on the GCA as principal. Ultimately, I was not persuaded that in the particular context with which this case is concerned there is the necessary “closeness of identity” that is required to establish an agency relationship.

It is, as I have said above, striking that Counsel for the Stati Parties have no support for their case from Professor Maggs. The only expert evidence on the subject came from Professor Suleimenov. Although he was taken to his earlier comment in the AIG case, his reasoning for his opinion as expressed in his report in this case (at paragraphs 8487), that there was no agency relationship, was not tested in cross-examination. In these circumstances, where Professor Maggs said nothing on the subject and where the reasoned opinion of Professor Suleimenov on this question was not tested it is tempting for this court simply to accept the evidence of Professor Suleimenov on this question. However, I am not convinced that Professor Suleimenov dealt fully with this issue. He did not note that articles 23 and 26 in fact contemplate that the NBK may act as agent. Whilst he may have regarded the TMA as not conferring authority on the NBK to bind the Republic to commercial contracts entered into by it he did not say so in terms.

I have to view the evidence as a whole. Having done so, my finding is that, having regard to the NBK being an independent central bank, to articles 21-26 of the Law on the National Bank, to article 883(1) of the Civil Code and to the terms of the TMA, the NBK did not have actual authority to enter the GCA as agent for the Republic, notwithstanding that some aspects of their relationship were consistent with an agency relationship.

Consistent with that conclusion is the absence of any suggestion in the GCA that the NBK entered into that contract as agent for the Republic. There is no requirement that there should be any such suggestion. A principal may be undisclosed. But if the Republic and the NBK had agreed that the NBK would enter into contracts such as the GCA as agent for the Republic it is difficult to see why the GCA, a contract governed by English law, would not have contained some clear reference to that agency.

I have also asked myself the questions identified by Leggatt J. in The Magellan Spirit as being appropriate to consider when there is no express authority to act as agent and reliance is placed on the conduct of the parties, in this case, the manner in which they deal with each other. Would a reasonable person in the position of the NBK have understood from the matters relied upon that it was authorised to enter into the GCA as agent for the Republic? Would a reasonable person in the position of the Republic have understood that the NBK was doing so? Counsel for the Stati Parties submitted that the answers to these questions was “plainly Yes.” I must disagree. Given the status of the NBK as the independent central bank of Kazakhstan, the provisions of Kazakh law to which I have referred, and the absence of any obvious or express grant of actual authority in the TMA I consider that the answer to both questions is No. Importantly, the Republic and the NBK might well have dealt with each other as they did in the absence of the suggested agency relationship. The four matters relied upon are consistent with the NBK alone being party to the GCA and having no actual authority from the Republic to bind the Republic to the GCA. That being so it is not possible to infer actual authority or an agency relationship from those matters, even if they are considered collectively.

In those circumstances, there being no express or implied agency relationship with regard to the NBK entering into the GCA as agent for the Republic the case of the Stati Parties on agency must be rejected. The Stati Parties have not established the necessary actual authority.

There being no relationship of agency with regard to the NBK entering into the GCA as agent for the Republic it is unnecessary to consider (a) whether, if there had been such a relationship, the Republic was the disclosed as opposed to an undisclosed principal and (b) whether the terms of the GCA excluded the Republic as principal in accordance with the principles discussed and applied in Filatona and Deripaska v Navigator and Chernukhin [2019] EWHC 173 (Comm) and [2020] EWCA Civ 109.

In essence the submission made on behalf of the Stati Parties is that the Republic is the beneficiary of a trust and that the NBK is the trustee. In those circumstances it is said that the Republic can enforce payment of BNYM’s debt. This submission emerged at trial. I do not think that it had been pleaded.

Reliance was placed on the principle of English equity whereby a beneficiary of a trust may institute proceedings to enforce payment of a debt owed to a trustee in an action to which the trustee is made party. Thus it was said that the Republic may sue BNYM, joining the NBK to the action. Reliance was also placed on the related principle of

English equity whereby the beneficiary may in certain special circumstances bring a

“derivative action” in its own name. These principles have been explored at length by the Supreme Court in Roberts v Gill [2011] 1 AC 240 but have been shortly summarised by Edward Murray (now Murray J.) sitting as a deputy judge of the Chancery Division in Seamus McEneaney v Craig Stevens [2017] EWHC 993 (Ch) in these terms:

“18.

In relation to a bare trust, such as that apparently asserted by the appellant in this case, the rule is that if the legal estate in the hands of the bare trustee is disturbed by a third party, the beneficiary may not institute legal proceedings in the name of the trustee without his authority, but may, on giving the trustee a proper indemnity, oblige the trustee to lend his name to assert his legal right: Lewin on Trusts (19^th.edition) at para 43-003. An alternative procedure is available if the trustee refuses to sue, namely, a derivative action. Under a derivative action, the beneficiary sues in his own name on behalf of the trust, joining the trustee as a defendant: Lewin on Trusts (19^th.edition) at paras 43-003 and 43-006.

……..

A beneficiary under a trust, however, may bring a derivative action only in special circumstances. There is no definitive list of what constitutes special circumstances, but the special circumstances must embrace a failure, excusable or inexcusable, by the trustees in the performance of the duty owed by the trustees to the beneficiary to protect the trust estate or to protect the interests of the beneficiary in the trust estate: Hayim v Citibank NA [1987] 1 AC 730 (PC) at 748F (per Lord Templeman).

The special circumstances are a necessary part of a beneficiary’s cause of action in bringing a claim as a derivative action and therefore must be pleaded: Roberts v Gill & Co at

The starting point of this submission is that the relationship between the NBK and the Republic is one of trustee and beneficiary. It was not explained to me how or why this was the case. The case of the Stati Parties is that the property in the National Fund remained at all times with the Republic. It was disputed that any property or title passed to the NBK. Professor Maggs said that the Republic had permitted the NBK to exercise “management rights” with regard to the National Fund. It was also, I think, common ground that Kazakh law does not recognise an ownership interest split into legal and beneficial ownership. Professor Suleimenov was asked to agree with that proposition in cross-examination and he did so. The submission therefore has a most unsure foundation. What can be said is that the rights to possess, use and dispose of the National Fund which were entrusted to the NBK by the TMA (and are said by article 188 of the Civil Code to be characteristic of ownership under Kazakh law) were to be used, pursuant to clause 1.1 of the TMA, for the benefit of the Republic, who was described in clause 7.1 of the TMA in terms as the beneficiary of the TMA.

I shall assume that the relationship between the NBK and the Republic is sufficiently analogous to that of trustee and beneficiary that English equity principles relating to trusts can be applied to it. Nevertheless, what is under consideration in the present case is a bank account which takes the form of a debt owed by BNYM to the NBK and which is governed by English law.

“Money paid into a bank account belongs legally and beneficially to the bank and not to the account holder”; per Lord Millett in Foskett v Mckeown [2001] 1 AC 102 at pp.128-129. As explained by Lord Cottenham in Foley v Hill (1848) 2 HLC 28 at pp.3637 the banker “has contracted, having received that money, to repay to the principal, when demanded, a sum equivalent to that paid into his hands”.

The person entitled to enforce that debt in this case is the NBK. That right forms part of the National Fund to which the Republic is, ultimately, beneficially entitled. The NBK would, it seems, be obliged to account to the Republic in respect of that debt and in respect of its proceeds as and when paid to the NBK. It would also seem, as acknowledged by counsel for the NBK and the Republic at paragraph 60 of their Closing Submissions, that, subject to matters of jurisdiction and state immunity, the debt due from the NBK to the Republic could be garnished from the NBK. But the Stati Parties have not sought to do that. They have sought to garnish the debt due from BNYM.

The question is whether that can be done by reference to the principles of English equity summarised above.

The claim which the Republic would be able to enforce pursuant to the equitable principles of English law is the claim which the NBK has against BNYM for the payment of the debt. If the Republic were to sue, joining the NBK as trustee, it would

Article 1445 of the BJC allows garnishment of “any amounts owed by such third party to its debtor.” BNYM is the third party and it owes sums to the NBK. The NBK is the creditor of BNYM. As a result of the Republic’s challenge to the garnishment order in this case the Belgian Court referred to this court the question “whether or not a debt exists from BNYM towards Kazakhstan.” In law the only debt owed by BNYM was to the NBK. Thus it is, I think, clear that equity principles of trust law do not assist the Stati Parties to say that the debt of BNYM is owed to the Republic so that it may be garnished.

Counsel for the Stati Parties offered no explanation as to why it would be in accordance with the applicable law of the debt, English law, to declare that in law BNYM owed sums to the Republic or to declare that a debt existed from BNYM to the Republic. I have asked myself whether the principle that equity treats as done that which ought to be done could assist the Stati Parties. But I have concluded that it cannot. Equity gives effect to that principle by permitting the beneficiary to enforce the debt which remains due to the trustee. Equity has not said that the debt may now be regarded as due to the beneficiary.

I accept that the Republic may ultimately be able to enforce the payment of the debt due from BNYM to the NBK but in law the debt remains a debt which is payable to the NBK and not to the Republic. I have therefore concluded that the principles of equity do not enable the court to declare that sums are owed by the garnishee, BNYM, to the award debtor, the Republic. This conclusion will of course be disappointing to the Stati Parties especially in circumstances where the Republic is ultimately beneficially entitled to the proceeds of the debt and has the right to require the NBK to make payments from the National Fund to enable demands on the Republic’s budget to be met, which could of course include payment of the award in favour of the Stati Parties if the Republic so wished. But these considerations do not enable the court to ignore ordinary contractual principles and to say that the debt which has been garnished is due from BNYM to the Republic when in law it is due from BNYM to the NBK.

Counsel for the NBK and the Republic considered that the Stati Parties were advancing a third argument that as “owner” of the cash the Republic had a claim to the cash in the hands of BNYM. However, counsel for the Stati Parties did not, I think, advance such an argument as a matter of English law. Nevertheless the ownership of the National Fund by the Republic was such a looming presence in the case that I ought to deal with the topic, albeit shortly.

The question depended upon whether in Kazakh law the property in or title to assets forming part of the National Fund passed to the NBK as trust manager (the view of Professor Suleimenov) or remained with the Republic (the view of Professor Maggs). I accept the opinion of Professor Maggs. His opinion fits naturally with an owner entrusting to an experienced or qualified manager the power to manage a property for the benefit of the owner. His opinion is supported by other academic writing on the subject. It was also, I think, common ground in the AIG case in which Professor Suleimenov did not disagree with the opinion of Professor Didenko that the trust manager did not become the owner of the property under trust; see [2006] 1 WLR 1420 at paragraphs 17 and 90. It was also common ground in the Belgian proceedings in May 2018 that the National Fund was the property of the Republic. Professor Suleimenov’s opinion in the present case (developed as a result of further thinking about the issue) appears to me to give too literal a meaning to the grant of the power “to possess, use and dispose” of the property to be trust managed (notwithstanding that such power is said by article 188 of the Civil Code to be characteristic of ownership under Kazakh law). His opinion seems to me to pay no, or at any rate insufficient, regard to the context of a property owner wishing to have his property managed by a qualified and trusted manager. Whilst there is every reason why such an owner would permit his manager to possess, use and dispose of the property it is very difficult to see why he would agree to transfer ownership of or title to the property or why Kazakh law should provide that such permission would have that effect. I was not persuaded that that was the effect of Kazakh law. Professor Maggs’ opinion, as expressed in cross-examination, was that the

However, the dispute in the present case concerns the right to demand payment from a bank of sums in a bank account. Professor Suleimenov suggested that there was a distinction between rights in rem (rights to property) and rights in personam (a right to make a claim). It seemed to me that there is such a distinction (for example article 115 sub-section 2.1 of the Civil Code draws such a distinction) but the consequences of that distinction were obscure (no article of any statute appeared to spell out the consequences). Nevertheless, the right to claim a debt appeared to be an example of a right in personam and both Professor Suleimenov and Professor Maggs appeared to be agreed that in Kazakh law, where a bank account is in the name of the trust manager, only the trust manager had the right to withdraw money from the account. Professor Suleimenov explained this in his Supplementary Report at paragraphs 37-39. Professor Maggs, when cross-examined, said “Well, if the Bank of New York Mellon founded a subsidiary as a Kazakh licensed bank, in that case it [the Republic] would have no right to take the money out of the Kazakh licensed bank……..Under Kazakh law on banking it would have no right because those rights had been assigned to the entrusted manager.” Furthermore, as submitted by counsel for the NBK and the Republic in reply submissions, the argument advanced by the Stati Parties based on the effect of transfer from the founder to the trust manager would appear to have no relevance to a US $ account which is likely to have been opened by the NBK.

The position in Kazakh law with regard to bank accounts is of interest but I am concerned with English law because the right to withdraw money from a bank account is a contractual right and the contract in question, the GCA, is governed by English law.

“Ownership” of the National Fund does not give the Republic a right to demand payment from BNYM or to say that the cash is held by BNYM for the Republic. The right to claim the cash is a contractual right and it lies with the NBK for the reasons I have endeavoured to express earlier in this judgment. Whatever ownership rights the Republic has in Kazakh law to the National Fund, whether as legal owner or beneficial owner or simply as owner, they do not in English law (just as in Kazakh law) confer on the Republic a right to claim the cash, the debt owed by BNYM, from BNYM. In the AIG case Aikens J. said much the same at [2006] 1 WLR 1420 at paragraphs 30-31. “There is no relationship of debtor and creditor between them. The fact that Kazakhstan may, ultimately, have a beneficial interest in the money represented in the cash accounts cannot, in my view, create such a relationship.”

For these reasons the Republic’s ownership of the National Fund does not enable it to claim from BNYM the cash sums owed by BNYM.

Whether or not a declaration is granted by the court as a remedy is a matter for the discretion of the court.

In the present case the English Court has, in effect, been asked by the Belgian Court to answer a particular question. The answer to that question is required in order to determine whether the garnishment order issued by the Belgian court has content or subject matter. That seems to me a good reason for this court to answer the question referred to it by means of a declaration.

It is therefore necessary to examine the reasons put forward by the Stati Parties for not granting the requested declarations.

Counsel for the Stati Parties submitted that there were three principal reasons for the

The first reason suggested was that the Republic of Kazakhstan has set itself “a very onerous burden, namely, to establish that there is no basis upon which BNYM could have concluded that Kazakhstan may have claims on BNYM or that BNYM may hold assets of or for Kazakhstan which were the subject matter of the Belgian Garnishment Order”. This argument was not specifically advanced in counsel’s Closing Submissions. But in case it is still pursued I ought to deal with it.

The submission appears to be based upon language used by me in my judgment on the jurisdiction challenge where I said that the referral from the Belgian Court involved “the correctness of the BNYM Declaration”.

The full passages in my judgment need to be quoted:

In the first extract I was responding to a submission by counsel for the Stati Parties that “there had been no "referral" of any question to this court, and that the Belgian Court had determined that the attachment order did have subject-matter, on the basis of the BNYM declaration” (see paragraph 34 of the judgment). I responded that “the Attachment Judge considered that the correctness of the BNYM declaration and the existence of a chose in action held by BNYM(L) for RoK to be questions for this court”. Reading the paragraph as a whole I was not saying that it was merely the “correctness of the BNYM declaration” that was to be considered by this Court. Rather, I was saying that what had been referred was the question whether the Republic of Kazakhstan had a chose in action against BNYM as suggested might be the case by BNYM in its declaration. The submission made by counsel for the Stati Parties requires the reader of my judgment to ignore the words “and the existence of a chose in action held by BNYM(L) for RoK”. In the second extract I was responding to a submission by counsel for the Stati Parties that a declaration was not the "most effective way" of resolving the issues raised, because the Belgian Court will ultimately rule on the validity of the executory attachment under Belgian law. I do not consider that anything in my response justifies the submission now made by counsel for the Stati Parties. On the contrary, I noted that the issue referred to the Court affects "the existence or extent of a legal right between" the Stati Parties and the other parties to these proceedings, namely, their right to attach the assets in question. At paragraph 33 of my judgment I described the central question as “what assets, if any, does BNYM hold for RoK?” That is a legal (and hard edged) question. It is not answered by examining whether there is a basis upon which BNYM could have concluded that Kazakhstan may have claims on BNYM. It is answered by examining whether Kazakhstan does have a claim on BNYM. Counsel for the Stati Parties now appear to accept that that is so; see paragraph 2.1 of their Closing Submissions where they say that “the relevant question under English law is whether [the Republic] has the right to enforce the debt owed by BNYM under the GCA.”

Unless that question is answered these proceedings would not advance the Belgian garnishee proceedings at all. It is worth repeating what the Belgian judge said of the claims by the NBK and BNYM:

“Both claims relate to the subject-matter of the attachment, notably whether or not a debt exists from BNYM towards Kazakhstan. Kazakhstan disputes the existence of such debt. The attachment judge cannot and may not settle such dispute, but only the judge on the merits. The judge on the merits is, as already mentioned above the English court who must apply its own national law.”

Thus the question for the English Court is not whether a debt may be owed to the Republic. The question is whether a debt is owed to the Republic. A declaration is the obvious way in which that question can be answered with clarity.

The second reason advanced for not making the declarations which have been sought was that, applying Rolls Royce PLC v Unite the Union [2010] 1 WLR 318, they were (i) overly broad and abstract; (ii) not satisfactorily dispositive or reflective of the issues which had been referred to this court and (iii) they raise issues which are inappropriate for determination by this court in view of what is to be determined by the Belgian Court.

I accept that, as stated by Aikens LJ at paragraph 120 of his judgment in Rolls Royce PLC v Unite the Union, there must be a real and present dispute between the parties, each party must be affected by the court’s determination of the issues concerning the legal right in question and that the grant of a declaration must be the most effective way of resolving the issues raised. In the present case there is, obviously, a real and present dispute between the parties. The Stati Parties say that BNYM’s debt is owed to the Republic. The NBK and the Republic say that it is not. The Court’s determination of that issue will affect the parties because it will be taken into account by the Belgian Court when determining the outcome of the garnishment proceedings in Belgium. If the debt is held to be owed to the Republic the Belgian Court will presumably say, subject to any further arguments of the Republic as to why the underlying arbitration award should not be enforced, that the Stati Parties are entitled to the garnished debt. If the debt is held not to be owed to the Republic the Belgian Court will presumably say, subject to such further arguments as the Stati Parties are entitled to advance in Belgium, that the Stati Parties are not entitled to the garnished debt. Finally, the making of a declaration is the obvious way of answering the question referred to this court by the Belgian Court. No other way of answering that question was suggested.

The first declaration which is sought is that the contracting parties to the GCA are BNYM London and the NBK (and not Kazakhstan). For the reasons which I have given the contracting parties to the GCA are indeed BNYM London and the NBK (and not Kazakhstan). The second declaration which is sought is that the obligations owed by BNYM London under the GCA are owed solely to the NBK (and not to Kazakhstan). For the reasons I have given the obligations under the GCA are indeed owed solely to the NBK. The third declaration which is sought is that BNYM London has no obligation to pay any debt due under the GCA to Kazakhstan. This declaration relates to the particular obligation owed under the GCA which is of immediate relevance to the parties. For the reasons I have given dismissing the arguments advanced by the Stati Parties under the headings of agency and trust BNYM London does indeed have no obligation to pay any debt due under the GCA to Kazakhstan. These three declarations are not overly broad and abstract. On the contrary they are clear and precise. They also reflect and seek to respond to the question referred to this court by the Belgian Court. Finally, they do not raise issues which are inappropriate for this court to determine.

The third reason advanced for not making the declarations was that it is inappropriate to grant declaratory relief in circumstances where it is required, as a matter of English private international law, to apply any substantive law other than English law to determine the merits of the declarations. It seems likely that this objection is no longer pursued because in their Closing Submissions counsel for the Stati Parties submitted that the proper scope of this trial is whether the Republic has a claim against BNYM under the GCA, as a matter of English law, “taking into account foreign law only insofar as that is permissible and required by English conflict of law rules.” In answering the question referred to this court by the Belgian Court I have taken into account Kazakh law because that is the law which applies to the relationship between the Republic and the NBK pursuant to English conflict of law rules. In the light of the position taken in the Closing Submissions I do not propose to discuss this third objection any further.

For the above reasons I consider that it is appropriate to make the first three declarations.

The fourth declaration is that Kazakhstan does not have any claims (under any system of law) against BNYM in relation to the cash deposits held by BNYM pursuant to the GCA which constitute a subject-matter falling within the scope of the Belgian Garnishment Order. This declaration was added by an amendment permitted on the first day of the trial.

Although I have explained earlier in this judgment the reasons for the phrase “under any system of law” (which I need not repeat here) it does not follow that it is appropriate to grant the declaration in the terms sought. To grant a declaration using the phrase “under any system of law” might cause uncertainty as to the scope of the matters addressed by this court. Such a declaration might suggest that this court had considered whether the Republic has a claim against BNYM under Kazakh or Belgian law when it has not considered that question. The question this court has considered is whether under English law the Republic has any claim against BNYM in relation to the cash deposits held by BNYM pursuant to the GCA. The Republic’s ownership of the National Fund does not give it a claim against BNYM in relation to the cash deposits.

I am also concerned at the inclusion of the phrase “which constitute a subject-matter falling within the scope of the Belgian Garnishment Order”. I have explained earlier in this judgment that there is, it appears, a dispute of Belgian law as to the scope or reach of the Order. Although the Belgian Court noted that the objection taken before it by the Republic was that the garnishment order had no subject-matter and referred that question to this court (as I had noted in my judgment on the jurisdiction challenge) that question depended upon whether BNYM owed a debt to the Republic. The court must decide that issue but if the question of subject-matter is wider under Belgian law (as it is said by the Stati Parties and Professor Storme to be) it is not appropriate for this court to decide that issue. It must be determined by the Belgian Court. For that reason any declaration by this court should avoid the appearance of deciding that issue.

I will therefore grant the fourth declaration but only in these terms:

“Kazakhstan does not have any claims against BNYM in relation to the cash deposits held by BNYM pursuant to the GCA.”

Whether the Stati Parties are able in Belgium to advance any argument that Kazakhstan does have claims against BNYM in relation to the cash deposits held by BNYM pursuant to the GCA based upon allegations of simulation, sham trust or abuse of right will be, as I have noted earlier in this judgment, a matter for the Belgian Court to consider having regard to the Belgian law of res judicata.

In the event that the court determined that the debt in question was owed by BNYM to the NBK, the NBK sought judgment upon its claim against BNYM. This assumes that the Belgian Court will decide, in the light of the decision of this court, to discharge the garnishment order. That may well happen but I cannot prejudge the decision of the Belgian Court. Accordingly, I do not consider that it is appropriate to give the judgment requested at this stage, but instead to defer further consideration of it, should that be necessary. In circumstances where BNYM is willing to pay the debt to the NBK as and when it is free to do so, BNYM (whose conduct in this matter is not criticised by the NBK or the Republic) should be afforded a reasonable time in which to pay after the Belgian Court has given its decision. Therefore, whether or not the Belgian Court decides to discharge the garnishment order I would not expect there to be any need for this matter to be brought back to this court.

For the reasons which I have endeavoured to express the court shall answer the issue referred to it by the Belgian Court by granting these declarations: