The Czech Republic v Diag Human SE & Anor

The Czech Republic adduced oral evidence from two witnesses.

The first was Mr Jensen. Mr Jensen was a company lawyer at Novo Industries from 1986 and, following a 1989 merger with Nordisk Gentofte, with Novo Nordisk until January 2012. He made it clear that he was not involved in Novo Nordisk’s business nor how Mr Klaus Eldrup-Jørgensen (head of the blood plasma unit) ran that business. I accept that Mr Jensen was doing his best to assist the court. However, he had almost no involvement in the key events in 1990-1992, and his principal involvement came in 2004, when he was responsible for Novo Nordisk’s official interactions with a Czech Parliamentary Commission established to investigate the claim brought by Conneco against the Czech state. In that context, he was essentially dependent on the recollection of those ex-employees he spoke to, and those few documents to which Novo Nordisk still had access following the sale of the blood plasma business and the transfer of its archive in 1995. Even in relation to events in 2004, Mr Jensen’s recollection, some 20 years, on was clearly inaccurate in certain respects. For example, while he was adamant that he only attended one meeting with the Commission, an all-day meeting on a Sunday, the contemporaneous documents show that he attended two, both on Mondays: one on 31 May 2004 and another 28 June 2004. I have concluded that the contemporary documents and the inherent probabilities provide a better guide as to Conneco’s interactions with Novo Nordisk. I consider that material in some detail in Annex 2.

The Czech Republic’s second factual witness was Dr Petr Turek, a physician who specialised in haematology. He worked at the Institute of Haematology and Blood Transfusion in Prague between 1980 and 2005 and was a consultant to the Czech Ministry of Health. He explained why the committee established by the Ministry of Health (of which he was a member) did not select Conneco in the 1991 tender. I am satisfied that Dr Turek was doing his best to assist the court, albeit the events he was describing took place over 30 years ago. His evidence was of very limited relevance to the matters which remain in issue.

The Arbitral Claimants’ principal witness was Mr Josef Stava. He is 74 years old. Mr Stava founded both Conneco and its Swiss parent company, and is currently chair of Diag SE’s board of directors. He was clearly the moving spirit in establishing its business. Mr Stava is the only person involved in Conneco’s business who gave evidence in the BIT Arbitration and to the court. Perhaps for that reason, his arbitration statements went into a level of detail about that business which I doubt Mr Stava was familiar with contemporaneously, still less 30 years on. The reality is that Mr Stava was simply not in a position to speak to most of that detail, although, unfortunately, that did not prevent him from attempting to do so. Further, the more rigorous checking and testing of his evidence occasioned by the court process exposed a number of significant inaccuracies. At best, this revealed an unhealthy level of wishful thinking on Mr Stava’s part, and a wholesale failure to take the care necessary to ensure that his evidence was as accurate as it could be. The overall effect was to cause me to approach Mr Stava’s evidence with considerable caution.

The need for that caution was reinforced by an event which happened in the course of Mr Stava’s cross-examination. At each break in his evidence, Mr Stava was warned that he must not communicate with anyone about this case until his evidence had been completed. It was clear, however, that at the end of Day 3, Mr Stava was becoming frustrated at what he described as difficulties with his memory. He asked one of his counsel team (in open court, and hence on the transcript) if it was possible to finish his evidence then, so that he could talk to his legal team. Entirely properly, he was told that this was not possible, and the court reiterated that warning.

Most regrettably, Mr Stava ignored that warning and the court’s instruction. He contacted his Czech lawyer, Dr Kalvoda, to ask for a copy of a document from the Commercial Arbitration, and which he had previously sent to this London legal team in the days before he gave evidence. He was sent a further copy of the document, together with a covering email in Czech which contained commentary. Mr Stava asked his daughter Dagmar Stave to obtain a translation of the email, and then sent that translation along with the original email from Dr Kalvoda (but without any attachments) to his legal team early the following morning, suggesting certain questions relating to the material and one further topic which he could be asked in re-examination. His daughter Dagmar Stava was copied into that communication, and she also arranged for a hardcopy of the document to be printed and given to Mr Stava. The Arbitral Claimants’ legal team informed the court at the start of the hearing that morning that there had been an overnight development which might impact on Mr Stava’s continuing evidence, on which they wanted the opportunity to take instructions. I can now see that what was being sought was an opportunity to take instructions to provide the overnight communications to the court, albeit this was not clear to me at the time. The evidence proceeded, but it soon became apparent that Mr Stava was relying on a particular document (which he was holding in his hand) and questioning about the provenance of that document initiated a process which revealed the information set out in this paragraph. Dealing with these events occasioned a significant disruption to the trial and added considerably to the burdens on both legal teams, but most particularly on the Czech Republic’s legal team.

I should note the following facts:

I have little doubt that the long and frequently bitter legal battle which Mr Stava has fought with the Czech Republic has been a frustrating and draining experience. However, that does not excuse what I am satisfied was a clear act of non-compliance with the court’s instruction. The issue of what other consequences may flow from these events is for another occasion. I am concerned with their impact on the trial. I am satisfied that Mr Stava’s non-compliance provides a further reason to approach his evidence with caution, not least because it suggests that his dispute with the Czech Republic may have consumed him to a sufficient degree to override his better judgment. The net effect is that I have felt unable to rely on Mr Stava’s evidence save to the extent that it accords with the documentary record or the inherent probabilities.

The Arbitral Claimants called three other factual witnesses: Mr Stava’s three daughters, Ms Dagmar Stava, Ms Nicole Stava and Ms Silvia Stava. Ms Dagmar Stava gave evidence as to her perceptions of her father’s control of the Koruna Trust and Diag SE, and to the effect that in so far as she had rights of control over decisions, she would exercise them in accordance with her father’s wishes (Ms Nicole Stava and Ms Silvia Stava giving evidence to the same effect on that last issue). Ms Dagmar Stava was unwise in responding to her father’s request to be provided with documents during his cross-examination without checking matters with the English lawyers. I accept, however, that she was not involved in any attempt to influence the scope of her father’s re-examination and was essentially responding in an administrative way to what her father had asked her to do. A contemporaneous message shows her telling her father not to call anyone while in the course of giving evidence, which is to her credit.

The evidence as to how Mr Stava’s daughters would have responded if their consent had been required for some course of conduct to be pursued was wholly hypothetical in nature, and the factual position is better gauged by reference to the inherent probabilities rather than their assertions now in evidence served in support of their father’s case in this long running and bitter dispute. I accept that all three daughters felt a strong sense of loyalty to their father and would naturally be disposed to defer to him in matters of business. However, it is not difficult to envisage unhappy circumstances in which those loyalties might be tested to destruction (for example in the aftermath of a family row or following the introduction of a significant new element into the family dynamic).

In addition, the Arbitral Claimants adduced hearsay evidence from Colonel Miloš Bohonĕk MD, Ph D. From January 1991 to May 1995, Colonel Bohonĕk was Chief Doctor of the Transfusion and Plasmapheresis Centre of Hospital Mĕlník (“Hospital Mĕlník”), in which capacity he worked with Conneco and Mr Stava. His statement described the parlous state of Czech hospitals following the fall of the Communist state system, and in particular the deficiency in plasma derivatives. He referred to the arrangement which Hospital Mĕlník concluded with Conneco, and his understanding that there were similar arrangements between Conneco and Hospital Uherské Hradištĕ (“Hospital Hradištĕ”) and Hospital Frýdek Místek (“Hospital Místek”). He recollected about 30 centres or blood banks having entered into agreements of this kind with Conneco in the early 1990s. He stated that the Co-operation Agreement between Hospital Mĕlník and Conneco worked well and was mutually beneficial. Finally, he stated that he was surprised to receive the Bojar Letter, but for which he was “confident that the mutually beneficial relationship between Hospital Mĕlník and Conneco likely would have lasted for several more years.”

There is a dispute about the weight to be accorded to that evidence, which it is convenient to address now. I accept that the Czech Republic has not had an opportunity to test the evidence at this hearing which must affect the weight which I can accord to it. However, the effect of that factor is reduced by the fact that the Czech Republic had the opportunity to cross-examine Colonel Bohonĕk on the same evidence four years ago and chose not to do so. Nor am I persuaded that I should draw the inference that Colonel Bohonĕk is not willing to give evidence because he does not believe his evidence will withstand cross-examination. Colonel Bohonĕk has written to the Arbitral Claimants’ solicitors stating that he is applying for the post of Director of the Central Military Hospital Prague, and as a high-ranking professional soldier and civil servant, he believes it would not be appropriate for him to give evidence against the Czech Republic in public proceedings. I see no reason not to accept the broad terms of that evidence at face value. However, I feel able to place less weight on the detail of Colonel Bohonĕk’s evidence – for example the specific numbers given – because there is no explanation of how Colonel Bohonĕk knew of or arrived at those figures.

The Czech Republic has asked the court to draw an adverse inference from the Arbitral Claimants’ failure to call certain witnesses. In Efobi v Royal Mail Group Ltd [2021] UKSC 33, [41], Lord Leggatt stated:

The first individual whose absence is said to give rise to a positive inference against the Arbitral Claimants is Dr Rabanser, Mr Stava’s and the family’s personal Liechtenstein lawyer and a principal of the corporate trustee of the Koruna Trust. The absence of evidence from Dr Rabanser deprives the Arbitral Claimants of corroborative evidence for Mr Stava’s evidence as to the practical operation of the Koruna Trust. I am also satisfied that I can infer that, had he been called to give evidence, Dr Rabanser would not have supported any suggestion that he had reached a private agreement or understanding with Mr Stava to follow his instructions, to the extent this was not permitted by Liechtenstein law.

In addition, the Czech Republic invites me to draw an adverse inference from the absence of evidence from staff employed by Conneco who were “involved in the business the Defendants say was being conducted in the Czech Republic in the early 1990s”, with specific reference to Dr Gnädinger. Given the period of time which has since elapsed, I am not persuaded that I should attach positive significance to the lack of evidence from Dr Gnädinger. However, the fact remains that Mr Stava is the only live witness who has given evidence about the status and extent of Conneco’s business.

The court heard evidence from two experts in Liechtenstein law: Mr Nicholas Reithner for the Czech Republic and Dr Christian Batliner for the Arbitral Claimants. Both experts were knowledgeable, fair and sought at all times to assist the court by their evidence. For that reason, the matters in issue between them were very limited.

The Czech Republic made a number of criticisms of the Arbitral Claimants’ disclosure. Against a background in which orders for disclosure had been made in the BIT Arbitration, applications were made in these proceedings for targeted categories of disclosure, and two orders were made by me on 28 March 2023 and 21 March 2024.

The disclosure order made on 28 March 2023 required disclosure of documents evidencing whether the Lawbook Transaction was performed, and whether there was any transfer of shares. The Arbitral Claimants’ response to that order was tardy, and I have previously criticised the effort which was applied in complying with the order. However, a disclosure statement of 15 January 2024 describes the search process undertaken, which was supervised by a solicitor save for the search by Dr Rabanser, a qualified Liechtenstein lawyer, of his own files. These searches produced three new documents (i.e. documents not disclosed in the BIT Arbitration), one of which was withheld from inspection on the basis of a claim to privilege which has not been challenged. I am unable to conclude that there has been any deliberate non-compliance with that order.

The disclosure order made on 21 March 2024 required production of documents evidencing the existence of creditors of the Koruna Trust, and documents sufficient to evidence any instructions given by Mr Stava to Dr Rabanser in respect of the Koruna Trust, in each case limited to particular periods of time. The response to the first request was a statement that there were no creditors (rather than a search of documents). As to the second, a disclosure statement of 22 April 2024 described the searches made using keywords, the review of the returned documents by a solicitor, and the fact that there were no responsive documents.

In explaining the relative paucity of documents, the Arbitral Claimants rely on what Mr Stava alleges was a raid on Conneco’s offices in 1993 in which documents were taken by the Czech police. I am satisfied that that there was some form of criminal investigation by the Czechoslovak (“CSFR”) authorities into Conneco from March 1991, and some form of investigation into what the Czech Republic contended in the Commercial Arbitration was a different complaint in which Conneco featured. In the Interim Award in the Commercial Arbitration in 1997, the Commercial Arbitration tribunal referred to the criminal file number allocated to the investigation: CVS:MVV 65/91. At a later stage of the Commercial Arbitration, issues about the alleged insufficiency of contemporaneous documents came into sharper focus after Ernst & Young had produced a report on damages. On 19 September 2006, the Commercial Arbitration tribunal ordered the Czech Republic to produce the “commercial records and correspondence of Conneco, Novo Nordisk and Diag Human submitted and not returned from the closed investigation file of the Czech Police in 1993 and 1994” (referred to in the Final Award, [139]). The context in which those records were submitted is not clear. I am unable to conclude that the use of the word “submitted” is necessarily inconsistent with the police requiring documents to be handed over at Conneco’s premises after an unscheduled visit, but equally the Commercial Arbitration awards contain no reference to a “raid”.

It is not clear from the materials available to me how the Czech Republic responded to Conneco’s request for that order – and in particular, whether it denied that documents had been taken by legal compulsion, or that they had not been returned. The Final Award ([140]) records the Czech Republic expressing reservations about the Commercial Arbitration tribunal’s order but stating that it had nonetheless sought to comply with it to the greatest extent possible. Mr Bastin KC was not able to ascertain what the Czech Republic’s position had been in the Commercial Arbitration on the issue of whether documents had been obtained by the Czech police by legal compulsion and not returned.

At paragraph 190 of the Final Award, the Commercial Arbitration tribunal recorded Conneco’s submission that “the evidence of the plaintiff had been affected during the dispute by the unlawful practices by the defendant. Indications of this included the release of transaction and customs documentation of the plaintiff to the Police of the Czech Republic under the pretext of the criminal proceedings in 1993-1995 and the fact that after the postponement of the case they were not returned. The plaintiff said it had managed to obtain the documents and material unlawfully [retained] by the defendant and is now presenting the evidence from them. The commercial activity of the claimant prior to the relevant time is now shown in detail”. Once again, the position the Czech Republic took in response is unclear. While I accept that Conneco was submitting to the Commercial Arbitration tribunal that, in effect, its case was supported by comprehensive contemporaneous documentation and that any deficiency had been made good, I would have expected Conneco’s lawyers to put the best face on matters, and do not regard that submission as inconsistent with a state of affairs in which some relevant documents had been taken and not returned or replaced.

In an affidavit filed on Diag SE’s behalf in proceedings brought in the Commercial Court to enforce the Final Award in the Commercial Arbitration, sworn in November 2011, Mr Balboa referred to the fact that “all the business documentation and custom forms which the police confiscated as part of this [criminal] investigation were retained by the police and subsequently declared to be ‘lost’ causing considerable disruption and difficulties for Diag”. The Czech Republic’s response to that assertion is not apparent from the materials filed for this hearing.

The Arbitral Claimants raised the suggestion that relevant documents had been seized by the Czech police from Conneco’s offices and not returned in the BIT Arbitration. In its Counter-Memorial, the Czech Republic described the suggestion as implausible, but the only evidence adduced was from a Mr Petr Horacek (a former police officer, and a lawyer employed by the Czech Republic until 2015) who was not involved in any investigations of Mr Stava and Conneco. His evidence was that, had such a raid taken place and documents been seized, a protocol would have been issued recording what had been taken and Mr Stava would have been permitted to ask for the material back. In its submissions in the disclosure phase, the Czech Republic’s position in the BIT Arbitration was that no such raid took place, but no further evidence was adduced.

In correspondence in the BIT Arbitration, the lawyers from the Czech Republic stated that the file relating to one of the Czech police investigations had been destroyed in 2002. A police file in relation to a later and different police investigation was said to have been destroyed in 2018. In relation to the alleged seizure of documents in 1993, the letter denied that there had been a seizure but said that no protocol of seized documents had been found by the Czech police when asked to search.

The Tribunal was not persuaded that the existence of such a raid and/or removal of documents had been established, pointing to the absence of contemporaneous emails of complaint (Award, [383]).

I accept that documents are likely to have been taken from Conneco in the exercise of police powers. That appears probable against a background of at least one and possibly two ongoing criminal investigations, one directed to Conneco and the other involving it. This was not a complaint made for the first time in the BIT arbitration, and there is no evidence before me, at least, of the Czech Republic denying that documents had been taken in the course of the Commercial Arbitration. Nor do I feel able to place the weight which Mr Bastin KC asks me to place on Conneco’s understandable forensic assertion in the Commercial Arbitration that it had made any alleged documentary deficiency good. However, I do not feel able to make a positive conclusion as to what documents were taken and not returned or found elsewhere. I am not persuaded that Mr Stava would know the position in that level of detail.

It was common ground that in resolving the issues which arise as to the interpretation of the BIT, I should apply the “general rule of interpretation” in Article 31 of the Vienna Convention on the Law of Treaties (“VCLT”):

For the purposes of Article 31(2), the “context” includes the title and preamble of the Treaty. It has been noted that “contextually, the title may be the obvious starting point for identifying the ambit of a treaty or a section or provision in it”, although “titles are often too general to provide precise guidance” (Gardiner Treaty Interpretation 2^nd (2015), [4.2.2]). Context can also include “any structure or scheme underlying a provision or the treaty as a whole” (ibid, [4.2.3]).

Article 32 of the VCLT, “Supplementary means of interpretation”, provides:

In MSC Mediterranean Shipping Co SA v Stolt Tank Containers BV [2023] EWCA Civ 1007, [61], the Court of Appeal summarised the court’s interpretative task when faced with an international law treaty as follows:

Finally, I was referred to an important passage in the Supreme Court’s decision in Moohan v Lord Advocate [2014] UKSC 67, [64] when discussing Article 31 of the VCLT:

When the Czech Republic’s “No Investment” objection is unpacked, it can be seen to involve the cumulative effect of three different contentions:

The characterisation of the issue of whether there was an investment as jurisdictional has long been accepted by investment treaty arbitral tribunals (for the purposes of ascertaining their own jurisdiction), and this issue has also been classified as jurisdictional in English first instance court decisions (for the purposes of determining whether ss.67 or 103(2) of the 1996 Act are engaged): Gold Reserve Inc v. Bolivarian Republic of Venezuela [2016] EWHC 153 (Comm); PAO Tatneft v. Ukraine [2018] EWHC 1797 (Comm) and The Republic of Korea v. Dayyani and Others [2019] EWHC 3580 (Comm). It has also been accepted as a matter going to an arbitral tribunal’s jurisdiction by the Singapore Court of Appeal in Swissbourgh Diamond Mines (Pty) Ltd v Kingdom of Lesotho [2018] SGCA 81, [93].

This aspect of the jurisdictional challenges available in investment treaty arbitration merits further consideration. Jurisdictional challenges in commercial arbitration will generally offer a binary outcome – either there was an arbitration agreement, or there was not; either the dispute fell within it, or it did not; either the tribunal has been properly constituted, or it has not. It is certainly possible for issues to arise which may challenge the jurisdiction of the tribunal (and, hence, any award on the merits) only to some extent: for example if the tribunal makes a compendious award of damages by reference to two disputes, one of which is found by the court to fall within the arbitration agreement, and one which does not. A jurisdictional challenge in a commercial arbitration can sometimes involve detailed consideration of issues of fact – whether negotiations culminated in a concluded contract, whether a particular individual had authority to commit one party and so forth. But this is relatively rare, and, even when it does arise, it is generally possible to demarcate the jurisdictional and merits disputes in relatively clear terms.

In commercial arbitration, arguments that the jurisdiction of the tribunal can, in the alternative, be supported on a different basis to that found by the arbitral tribunal are almost unknown, which may explain why there is no provision in the s.67 context for the service of a respondent’s notice of the kind provided for in s.69 appeals (PD62 para. 12.6(2)).

However, these issues arise in much more acute form in investment treaty arbitration. As the arguments in this case show, determining whether there is a protected investment can involve a granular evidential analysis of a kind which a supervisory court would not ordinarily expect to be confronted with. Further, in finding that there is a protected investment, the tribunal will usually face a range of arguments as to what did, and did not, constitute a contribution; what assets were or were not owned or controlled by the investor and when they came into existence. While in some cases, they may involve a relatively simple task – was a factory acquired or established? – in others, the tribunal’s identification of the protected investment will be the culmination of a series of findings about particular items (shares, contracts, intellectual property rights, know-how, goodwill etc). To the extent that the tribunal concludes that it has jurisdiction, its findings on breach and relief will often reflect the particular investment it has found to exist.

When some but not all of the tribunal’s component findings on the issue of whether there was a protected investment are challenged under s.67 of the 1996 Act, what is to happen? Would the success of a challenge to a single component finding be sufficient to establish lack of jurisdiction for the award that was made, requiring the matter to go back to the arbitral tribunal? And equally, what of an investor? It can re-argue its case if it failed to establish a protected investment before the arbitral tribunal under s.67, but what is to happen if it succeeded, but not in respect of every component of its case? In particular if (as here) it is said that an “other investments” case has been argued before the court responsively, in the face of the state’s “no investment” challenge, then to the extent the investor succeeds on its alternative case, it might be said that the scope of the investment found by the court on the s.67 application is not that which was the subject of the breach and relief findings by the tribunal.

Whatever the position may be at the margins I am satisfied that the court should approach this issue by considering the substance of the investment established before the court. If that is the same as the substance of the investment found by the arbitral tribunal, then there can be no scope for setting aside the award or remitting it to the arbitral tribunal.

Finally, there is the issue of what is to happen when only some of the components of the investment found by the arbitral tribunal are challenged, or open to challenge (because they are not in dispute, or a challenge is precluded by s.73(1) of the 1996 Act). That issue arises in this case, where the Czech Republic’s “no investment case” was advanced on a wider front in court than before the Tribunal, and I held that aspects of it were not open. The Arbitral Claimants contend that the very fact that one component of the investment which the tribunal held had been established is not open to challenge is sufficient to preclude a s.67 “No Investment” challenge. That submission was only made after the s.73(1) hearing, so that this issue only became apparent after developments in the case had led to the decision to split what the court had previously ordered should be a single hearing. I cannot accept this submission in its full width. If, for example, the tribunal found that the investor had established two different factories at two separate points in time in the host state and awarded relief accordingly, I do not accept that if the Czech Republic had only challenged the investment to the extent of one of those factories, it would not be able to renew that challenge under s.67. There would be two distinct jurisdictional challenges in such a case, and relief awarded in relation to the challenged investment could not be said to be “in respect of” the unchallenged investment. By contrast, I can see that the position will be different where those matters not challenged or open to challenge are sufficient to establish the substance of the investment found by the arbitral tribunal and on the basis of which it awarded relief. Quite where, between those two situations, the line would fall to be drawn is, fortunately, not an issue which has to be resolved in this case.

I also reject the Arbitral Claimants’ contention that it is no longer open to the Czech Republic to advance the “No Investment” Objection because the issue of whether a successful attack on only a subset of the Tribunal’s investment finding might itself provide a basis for attacking the Award was a matter for the hearing which resulted in the March Judgment and any appeal from that decision. The Arbitral Claimants point to the Czech Republic’s ground of challenge in paragraph 3 of its Amended Particulars of Claim: “if the Court holds that the Tribunal was right that there was a protected investment, the Tribunal’s findings of quantum … are not with respect to that investment and accordingly should be set aside and/or declared to be of no effect”. However, that ground was directed to a different point – if the Tribunal’s findings as to the existence of investment were correct, whether there was a sufficient nexus between that investment and the damages awarded. There was no suggestion at that hearing that if no challenge could be brought to one of the components of the Tribunal’s investment finding, that was sufficient to take the entirety of the “No Investment” challenge off the table.

At this point, it may be helpful to summarise the Tribunal’s findings on the investment issue and their current status:

The arguments on whether there is any, and if so what, inherent meaning to the word “investment” in the BIT beyond establishing the existence of an asset was advanced by reference to three sources:

I consider these materials in turn, offering some preliminary observations as I do so, before reaching my conclusion in the light of all the material.

Article 1 of the BIT contains the following definitions:

It is the Czech Republic’s case that the word “investment” in the BIT has an inherent meaning which requires more than simply the investor owning or controlling an asset in the territory of the Czech Republic. It is not suggested that each of the factors in [39] above must be present for an investment to be established. That would, at least, involve a “bright line” analysis of the kind amenable to a jurisdictional enquiry. Rather, it is suggested that in determining whether there has been an investment, the court should consider the extent to which the factors in [39] above are satisfied, it being possible for the absence of one factor in an appropriate case to be outweighed by sufficiently strong presence of some or all of the others. That is a very evaluative exercise, and the prospect of it being undertaken de novo by the supervisory court, and (transnational issue-estoppel apart) by enforcement courts, is not an attractive one (cf. Butcher J’s observations in PAO Tatneft v Ukraine [2018] 1 WLR 5947, [101]).

The three factors which the Czech Republic contends are relevant to ascertaining the existence of an investment do not appear expressly on the face of the BIT. To that extent, the avowedly (but not exclusively) textual approach to interpretation of the VCLT does not immediately lend itself to this submission. However, Mr Bastin KC pointed to a number of indicia in the BIT which he said supported his argument that these factors formed part of the inherent meaning of the word “investment”.

First, there are statements as to the purpose of the Treaty in the Preamble, with its references to the parties desiring to “create and maintain favourable conditions for investments by investors of one party in the territory of the other Contracting Party”.

Second, there are a series of provisions in the BIT which refer to the “making” of investments. Prominent among these is Article 2(1) of the BIT which provides:

Mr Bastin KC also referred to various other provisions which he said showed an investment was more than an asset. I will not list all of them, but the following examples are representative: Article 3 (“investments by investors”; “admitted an investment on its territory”); Article 4 (“investments made … by investors”; “investments of the investors of the other Contracting Party”; “investments made within its territory” etc); Article 5 (“investments have been made”); and paragraph (1) of the Protocol (referring to “the territory of which the investment has been or is to be made”).

The use of the words “invest” and “investment” in investment treaties can address two concepts: both what the investor puts in (or contributes), and the assets within the host territory which are the manifestation or fruit of that contribution. Mr Justice Butcher in Republic of Korea v Dayyani [2019] EWHC 3580 (Comm), [45] drew a distinction between “property and assets put in by the investor” and “assets or property into which resources are put”, although it is clear that, on an appropriate treaty wording, the concept of assets put in “by the investor” can extend to property or assets in the host state acquired by the investor from a third party who made an investment: PAO Tatneft v Ukraine [2018] 1 WLR 5947, [64]-[67].

A similar distinction – in this context formulated by reference to the use of the word “investment” in BITs and the ICSID Convention – can be found in ITA jurisprudence. InMalicorp v Arab Republic of Egypt ICSID Case No ARB/08/18, Award of 7 February 2011 (Tercier, Baptista, Tschanz), it was noted at [108] that the BIT definition (an asset-based definition in the conventional form) “does not so much stress the contributions made by the party acting, asthe rights and assets that such contributions have generated for it”, in contrast to Article 25 of the ICSID Convention (albeit noting at [110] that “these two aspects are in realitycomplementary”):

These two senses of the word “invest” can be seen not simply when contrasting BITs and the ICSID Convention, but also within the terms of a BIT, and this can give rise to a degree of linguistic tension in the interpretation of BITs (as the Malicorp tribunal acknowledged). Focussing on the BIT in issue:

I accept that the provisions of the BIT (with expressions such as “investment in”, “investments made”, “promote investments” etc) and its preamble (with its reference to creating and maintaining “favourable conditions for investments by investors of one Contracting Party in the territory of the other Contracting Party”) provide support for the argument that the concept of an investment within that treaty is not established by the existence of an asset alone. I am conscious when doing so that arguments based on the use of the verb “invest” or “invested” have not appealed to some ITA tribunals who have concluded that these words are linguistic necessities to link the defined concepts of “invest” and “investor”, and are not intended to have significant legal content in their own right: e.g. Saluka Investments BV v Czech Republic (Watts, Fortier, Behrens) Award of 17 March 2006, [211]; Mytilineos Holdings SA v State Union of Serbia & Montenegro and Republic of Serbia (Reinisch, Koussitis, Mitrovic), Partial Award on Jurisdiction of 8 September 2006, [129]-[131]. However, that linguistic necessity may itself be informative as to the inherent meaning of the concept of “investment”.

There are two further provisions in the BIT on which Mr Bastin KC relied in support of his inherent meaning contention, neither of which I found particularly persuasive.

First, in reliance on Article 1(3) of the BIT and its definition of “returns”, he argues that an investment must be something inherently capable of generating returns. It is possible to find BIT provisions which define the concept of investment in this way: for example, Article 1 of the UK-Bolivia BIT defines “investment” as “every kind of asset which is capable of producing returns ….”. However, Article 1(3) does not form part of the definition of “investment” in this BIT, but defines the word “return” for the purpose of the free transfer of payments right conferred by Article 5. In my view, treating Article 1(3) as part of the definition of “investment” (which is the real effect of Mr Bastin KC’s argument) would fail to respect the structure and clear text of the BIT.

Second, he points to the fact that Article 9(3) of the BIT provides that:

Article 25 of the ICSID Convention defines the jurisdiction of the Centre as extending to “any legal dispute arising directly out of an investment between a Contracting State …. and a national of another Contracting State”, and there is a body of decisions which have defined the word “investment” in Article 25 in a particular way (applying the so-called Salini factors, first formulated in Salini Construction SpA v Kingdom of Morocco ICSID Case No ARB/00/4, Decision on Jurisdiction of 23 July 2001 (Briner, Fadlallah, Cremades) – although, as always in the investment treaty world, there is also a body of divergent opinion). Mr Bastin KC argues that the parties must have contemplated some similar definition applying to the BIT, given that they contemplated that a dispute under the BIT could be determined in ICSID arbitration.

Approached without regard to any authority, I found this submission unpersuasive:

It is right to note that an argument on the lines advanced by Mr Bastin KC has appealed to some ITA tribunals: e.g. Romak SA v The Republic of Uzbekistan PCA Case No AA 280, Award of 26 November 2009 (Mantilla-Serrano, Molfessis, Rubins), [193]-[195]; Alps Finance and Trade AG v Slovakia UNCITRAL arbitration, Award of 5 March 2011 (Crivellaro, Klein, Stuber), [239]; and Air Canada v Bolivarian Republic of Venezuela ICSID Case No ARB(AF)/17/1, Award of 13 September 2021 (Tercier, Poncet, Villanua), [293]. These awards offer two reasons in support of this conclusion – that it cannot have been expected that the substantive protection available to the investor would depend on which dispute resolution option was offered; and that the contrary conclusion would deprive the ICSID dispute resolution option of effet utile. I have not found either persuasive:

Further, a significant number of ITA Awards recognise that arbitrating a BIT claim using the ICSID dispute resolution option requires the claimant to establish an investment under both regimes (sometimes referred to as a “double-barrelled” or “double-keyhole” test): e.g. Salini Construttori SpA v Kingdom of Morocco ICSID Case No ARB/004, Decision on Jurisdiction of 23 July 2001 (Briner, Fadlallah, Cremades), [41]; Ceskoslovenska Obchodni Banka as v The Slovak Republic ICSID Case No ARB/97/4, Decision on Jurisdiction of 24 May 1999 (Buergenthal, Bernadini, Bucher), [68]; Mytilineos Holdings SA v State Union of Serbia & Montenegro and Republic of Serbia, Partial Award on Jurisdiction of 8 September 2006 (Reinisch, Koussoulis, Mitrovic), [114], [117]; Global Trading Resource Corp and Globex International Inc v Ukraine ISCISD Case No ARB/09/11, Award of 1 December 2010 (Berman, Gaillard, Thomas), [43]; and Malicorp v Arab Republic of Egypt ICSID Case No ARB/08/18, Award of 7 February 2011 (Tercier, Baptista, Tschanz), [107]. Professor Stephan Schill et al (eds), Schreuer’s Commentary on the ICSID Convention 3^rd (2022), [186] suggest that “this approach is followed by most arbitral tribunals”.

I was referred to in excess of 30 ITA Awards on this issue (it being made clear that the awards I was shown were far from a complete set of the available material). Both sides accept that there is a division of approach within the ITA jurisprudence, with each approach having a sufficient number of adherents that it cannot be marginalised as an outlier view. I have not sought to summarise this material or show its chronological progression – this is not a systematised body of jurisprudence of a kind which lends itself to a treatment of that kind. Rather, I have sought to identify some of the key themes underpinning the conclusions reached, to explore the ambit of the factors identified as relevant and to identify some difficult cases which reveal the tensions in both approaches.

First, there are a number of ITA Awards which suggest that a BIT definition that investment “means any asset, including” or similar language allows no room for introducing any requirement beyond the investor’s ownership or control of an asset – a fortiori, an asset of one of the listed kinds – for treaty protection to be available. I was referred to a number of awards to this effect, including Saluka Investments BV v Czech Republic, Award of 17 March 2006 (Watts, Fortier, Behrens), [211]: Mytilineos Holdings SA v State Union of Serbia & Montenegro and Republic of Serbia), Partial Award on Jurisdiction of 8 September 2006 (Reinisch, Koussoulis, Mitrovic), [117], [129], [135]; Casinos Austria International GmbH and Casinos Austria Aktiengesellschaft v Argentine Republic ICSID Case No ARB/14/32, Decision on Jurisdiction of 29 June 2018 (van Houtte, Schill, Torres Bernardez), [178]-[179]; Littop Enterprises v Ukraine SCC Arbitration Case No V2015.092, Final Award of 4 February 2021 (Lew, Fortier, Oreamuno), [330], [339] (a case under the Energy Charter Treaty – “the ECT”); and Michael Anthony Lee-Chin v Dominican Republic ICSID Case No UNCT/18/3, Final Award of 6 October 2023 (Fernandez Arroyo, Leathley, Kohen), [140], [156]. The reasoning in these cases is essentially textually driven.

Second, there is a substantial body of ITA Awards which suggest that the ordinary meaning of the word “investment” in this context embraces at least three criteria – a contribution to the host state, which is of some duration and which exposes the investor to investment or operational risk, as opposed to “ordinary commercial risk” or sovereign risk. A fourth factor – “contribution to the economic development of the State of the investment” – is sometimes suggested: see e.g. Salini Construttori SpA v Kingdom of Morocco ICSID Case No ARB/004, Decision on Jurisdiction of 23 July 2001 (Briner, Fadlallah, Cremades), [52]. However, that additional requirement has not found favour with most tribunals who embrace the “objective meaning” approach (Kriebaum, Schreuer and Dolzer, Principles of International Investment Law 3^rd (2022), pp.92, 95), and it is generally accepted that it is not necessary to show some positive net effect on the host economy, but at best some form of reasonable expectation of such an effect as a consequence of the investment.

This interpretation of the concept of “investment” can be traced back to at least the “state v state” ITA Award in Italy v Cuba, Interim Award of 15 March 2005 (Derains, Tanzi, Bravo), [81]. A particularly influential ITA Award is Romak SA v The Republic of Uzbekistan PCA Case No AA 280, Award of 26 November 2009 (Mantilla-Serrano, Molfessis, Rubins), [177], [181], [207]. Other “inherent meaning” ITA Awards I was referred to included:

The reasoning in these cases moves beyond a strict focus on the text of the definition of “investment”, with the following considerations being prominent:

Some ITA Awards have sought to reconcile the two approaches by suggesting that establishing the existence of an asset of an identified kind is sufficient to raise a presumption that there is an investment, capable of being rebutted in unusual circumstances (e.g. Komaksavia Airport Invest Ltd v Republic of Moldova SCC Case No 2020/074, Final Award of 3 August 2022 (Kalicki, Sands, Stern), [146]-[148]). However, the reference to non-exhaustive types of asset is intended to illustrate the width of the concept of asset, not to raise a presumption that requirements of contribution, duration and risk have been satisfied in the case of listed examples but not otherwise. Further, a presumption which operates when one of the illustrated types of asset is established, but not some other kind of asset, would not be principled.

Given the assets relied upon in this case, there are two particular types of assets – contractual rights and shares – which merit further discussion.

As I have mentioned, appeals to an “inherent” meaning of the word “investment” have been particularly prominent when the claim has been brought by reference to a contract for the supply of goods and services. The concepts of contribution, duration and risk serve to remove from the protection of a treaty cases such as the example given in Komaksavia Airport Invest Ltd v Republic of Moldova, [149]-[150] of someone in state A ordering a product over the internet from state B, with the price moving one way and the product the other, which the tribunal suggested could not credibly be described as an investment in either direction. An investment treaty claim in relation to a performance guarantee provided by the seller to the buyer of mining equipment was rejected on this basis in Joy Mining Machinery Ltd v Arab Republic of Egypt ICSID Case No ARB/03/11, Award on Jurisdiction of 6 August 2004 (Orrego Vicuna, Craig, Weeramantry). A claim relating to purchase and sale contracts concluded by US poultry interests failed for the same reason in Global Trading Resource Corp and Globex International Inc v Ukraine ICSID Case No ARB/09/11, Award of 1 December 2010 (Berman, Gaillard, Thomas), [55]. The tribunal in Alps Finance and Trade AG v Slovakia Award of 5 March 2011 (Crivellaro, Klein, Stuber) rejected the contention that what it described as a “mere one-off sale contract” amounted to an investment ([245]). The tribunal in Ipek v Turkey ICSID Case No ARB/18/19, Award of 8 December 2022 (McLachlan, Fortier, Levy), [292]-[293] distinguished between “purely contractual rights, which do not constitute an investment, and contractual rights constituting an ‘asset’”.

However, a more complex package of contracts, including a co-operation agreement intended to culminate in the claimant taking an equity stake in a local company was held to constitute an investment in Mytilineos Holdings SA v State Union of Serbia & Montenegro and Republic of Serbia, Partial Award on Jurisdiction of 8 September 2006 (Reinisch, Koussoulis, Mitrovic), [120]-[125], the tribunal concluding at [124]:

Similarly in GEA Group Aktiengesellshaft v Ukraine ICSID Case No ARB/08/16, Award of 31 March 2011 (van den Berg, Landau, Stern), a package of contracts which involved the supply by the claimants of naptha for conversion and the return of finished product was held to constitute an investment, on the basis that it “conveyed the right for GEA, through KCH, to exercise an economic activity in Ukraine at the relevant time” and was “more than just goods against a tolling fee – it established a relationship of ‘common interest’ whereby KCH (and, ultimately GEA) would, among other things, assist with delivery of logistics and pay for Ukrainian domestic freight, resolve customs issues, and supply the Oriana plant with necessary materials” ([149]). The fact that the identification of the alleged investment as a sale contract only takes the analysis so far is well-captured by an observation made by Professor Jan Paulsson in Pantechniki SA Contractors & Engineers v Republic of Albania ICSID Case No ARB/07/21, Award of 30 July 2009, [44]:

Shares in a company incorporated under the laws of the host state appear in most lists of assets in definitions of what constitutes an investment in BITs. There are ITA Awards which treat ownership of shares as sufficient of itself to establish an investment, without the need for further enquiry: e.g. RosInvest Co UK Ltd v The Russian Federation SCC Case No 079/2005, Award of 12 September 2010 (Bockstiegel, Berman, Steyn), [382] (a case in which shares had been purchased by the claimants from an existing owner, albeit that did not expressly form part of the tribunal’s analysis).

The issue of what happens when the claimant purchases the shares from their current owner who is not a national of the host state was expressly considered in Orascom TMT Investments Sarl v People’s Democratic Republic of Algeria ICSID Case No ARB/12/35, Award of 31 May 2017 (Kaufmann-Kohler, van den Berg, Stern), [382] in which the tribunal rejected the argument that there was no investment where shares in a host-state company were purchased from another foreign investor:

In KT Asia Investment Group BV v Republic of Kazakhstan ICSID Case No ARB/09/08, Award of 17 October 2013 (Kaufmann-Kohler, Glick, Thomas), the tribunal appeared to contemplate at [194] that a subsequent owner might only be able to rely upon a transferor’s “original contribution” where the shares were transferred in the context of an intra-group corporate restructuring. However, in that case the claimant was found to have made no payment itself, the share transfer being agreed at a notable undervalue, and the price not having been paid in any event ([206]).

That alternative analysis of KT Asia raises the further issue – must the purchaser of an existing foreign investment show that the investment was acquired for a contribution, and must the contribution come from its own funds? The ITA Awards show conflicting approaches:

In simplified terms, the acquisition of shares in an operating company involves the commitment of capital to the company’s operations, the recovery of which is subject to the risks of those operations. An exiting investor will realise its initial commitment, and cease to be exposed to those risks, with the capital thereafter being provided, and the operational risks born, by the incoming shareholder. Viewed in those terms, it is easy to understand why a number of ITA tribunals have been willing to treat the purchase of a shareholding in a host-country company from another foreign investor as the making of an investment. Cases in which shares are acquired without payment may raise more complex issues. However, it is not obviously clear why the application of an investment treaty should depend on whether the claimant was given the shares, or given funds which it then used to purchase shares, because in each case an asset of value is a source of the company’s capital and exposed to its risks. Further, to the extent that profits are generated which would be capable of being distributed to shareholders but are retained in the business, or the shares are not fully paid, the owner of shares may be said to have made a contribution to the company’s capital regardless of the amount paid for the shares.

Those ITA Awards which support the application of a tripartite test of contribution, duration and risk when considering the objective meaning of the word “investment” offer the following further guidance as to the application of that test.

First, it is necessary to consider the operation said to give rise to the investment globally, or as a whole, rather than considering its individual elements in isolation: Joy Mining Machinery Ltd v Arab Republic of Egypt ICSID Case No ARB/03/11, Award on Jurisdiction of 6 August 2004 (Orrego Vicuna, Craig, Weeramantry), [54]; Mytilineos Holdings SA v State Union of Serbia & Montenegro and Republic of Serbia, Partial Award on Jurisdiction of 8 September 2006 (Reinisch, Koussoulis, Mitrovic), [120], [125]. Kriebaum, Schreuer and Dolzer, Principles of International Investment Law 3^rd (2022), p.99 refer to this approach as the concept of “the unity of an investment”, noting:

Mr Bastin KC argued against this conclusion, referring to a passage in Teneris and Talta Trading e Marketing Sociedade Unipessoal Lda v Bolivarian Republic of Venezuela ICSID Case No ARB/11/26, Award of 29 January 2016 (Beechey, Kessler, Landau), [291], in which the tribunal held that the offtake contract in issue in that case was a contract for sale and delivery, and not an investment even if a holistic approach was adopted to the issue of whether there had been an investment. However, at [284]-[285] the tribunal expressly acknowledged the “combined effect” test of whether an investment had been established. I can see that there may be some activities – for example an internet order to pay money in exchange for goods – which may be so inherently lacking in the qualities of an investment that their character will not change merely because they are considered in a wider context. But generally, I am persuaded by the view that the issue of whether there is an investment should be looked at holistically rather than by considering different components of an integrated activity in isolation, at least where the claim relates to that holistic investment.

Second, the three factors are interdependent, and can counterbalance each other to some degree, rather than constituting three separate requirement which fall to be assessed and established independently: Salini Construttori SpA v Kingdom of Morocco ICSID Case No ARB/004, Decision on Jurisdiction of 23 July 2001 (Briner, Fadlallah, Cremades), [52]; Nova Scotia Power v Venezuela (II) ICSID Case No ARB(AF) 11/1, Award of 30 April 2014 (van Houtte, Williams, Vinuesa), [84]; Doutremepuich v Republic of Mauritius PCA Case No 2018-37, Award on Jurisdiction of 23 August 2019 (Scherer, Caprasse, Paulsson), [120].

Third, so far as any requirement of contribution is concerned:

So far as the issue of duration is concerned:

The concept of risk as applied in the ITA Awards is somewhat opaque, and attempts to formulate a satisfactory distinction between the “right” and “wrong” kinds of risk have been said not to “withstand closer scrutiny” (Markus Petsche, Investment Claims (2023), [4.36]). The distinction appears to be intended to differentiate the risks of investment (sometime referred to as “operational risk”) from the risk of counterparty non-performance inherent in any commercial transaction (Romak SA v The Republic of Uzbekistan PCA Case No AA 280, Award of 26 November 2009 (Mantilla-Serrano, Molfessis, Rubins), [229]-[230]). Some tribunals also distinguish operational risk from sovereign risk (e.g. the risk that a particular activity may be adversely affected by executive or legislative action, which arise both in trade contracts and in investments properly so-called: Postova Banka as v Greece ICSID Case No ARB/13/8, Award of 9 April 2015 (Zuleta, Stern, Townsend), [368]).

In Gold ReserveInc v. Venezuela [2016] EWHC 153 (Comm), the BIT defined a Canadian “investor” as a national of Canada who “makes the investment in the territory of Venezuela”. Venezuela argued that the claimant had not made an investment because it had, without any active contribution of capital or resources to the economy of Venezuela, acquired the indirect ownership of shares in a local company that held two mining concessions – as a result of a merger and share swap between its previous parent company and a subsidiary. At [32], Teare J noted that the term “investment” could have two possible meanings in ordinary language: “Investment can mean the contribution of resources, usually capital, to acquire an asset, as in ‘he made an investment of US$1m in acquiring a painting by Monet’. But an investment can also mean the asset which is acquired by the act of investing, as in ‘he exhibited his investment (the painting by Monet) at his stately home’.” The court found that the “expressed definition of ‘investment’ in the BIT used investment in this latter sense of asset”, and that “the ordinary meaning of ‘making’ an investment includes the exchange of resources, usually capital resources, in return for an interest in the asset” ([35]), or expending money on something it already owned ([37]), which Teare J contrasted with acquisition by inheritance ([35]). He concluded that “mere passive ownership of an asset is insufficient”, and that some action on the “investor’s” part was required ([37]). There was no finding on whether acquisition for value of an existing investment from another foreign investor would suffice, there being no evidence that the claimant had made any contribution for the shareholding which formed the basis of its claim, and the court being able to rest its conclusion on the $300m the claimant had subsequently invested into the assets acquired.

In PAO Tatneft v. Ukraine [2018] EWHC 1797 (Comm), [2018] 1 WLR 5947, [56]-[57], the issue arose as to whether the acquisition for value of a shareholding in an enterprise in the host state from another foreign investor amounted to an investment for the purposes of Article 1(1) of the Russia-Ukraine BIT. Butcher J answered that question in the affirmative. However, the basis of the court’s decision was not that the provision of value in the acquisition was enough, but that the language of the treaty did not require anything more than the ownership of the asset. Butcher J held that references in the BIT to assets “that are invested by an investor … in the territory of Ukraine”, to “making” investments or investments “made” within the territory of Ukraine were intended to provide “a link between the specification of the types of assets which are comprised within the definition of ‘investment’ and the person who owns or is otherwise interested in those assets”([68] and [70]-[71]). At [78]-[79], he distinguished Gold Reserve on the basis that that case was concerned with the meaning of “investor” and not directly with the meaning of “investment”, and because the terms of the BITs in the respective cases were “materially different”. I do not find those points of distinction wholly persuasive.

Finally, in The Republic of Korea v. Dayyani and Others [2019] EWHC 3580 (Comm), [2020] Bus LR 884, Butcher J was once again concerned with a BIT that similarly defined “investment” as “every kind of property or asset, and in particular, though not exclusively, including the following, invested by the investors of one Contracting Party in the territory of the other…”. He rejected the argument that, in order to constitute an investment under the BIT, the Salini criteria of contribution, operational risk and duration had to be satisfied ([58]-[61]), stating that he did not find the ITA Award in Romak SA v Republic of Uzbekistan PCA Case No AA280, Award of 26 November 2009persuasive.

I accept that, even outside the confines of Article 25 of the ICSID Conventions (where the absence of any definition of the word “investment” allows greater scope for ascribing an inherent meaning to the term), there will be cases in which a claimant under a BIT defining “investment” as “any asset” is required to do more than simply point to ownership or control of an asset to establish the existence of a protected investment. This much is necessary, if for no other reason than to address the hypotheticals often raised in this debate of the Paris Metro ticket, the weekend holiday let, the hiring of a car at an airport etc, and also to address the “pure” sale contract where the claimant sells goods into the territory or send a payment there for goods it has received. The concepts of contribution, duration and risk are of assistance in rationalising an intuitive sense that these instances of dealing with the host state do not amount to investing in it. Discussing the attempts to define the concept of investment in Article 25 of the ICSID Convention, Schreuer’s Commentary on the ICSID Convention 3^rd (2022) [275] suggests that its “purpose … is to filter out disputes arising out of activities or transactions that the Centre was not thought to address, in particular non-economic activities and commercial activities that do not involve investment specific risks”. The BIT ITA Awards which ascribe an inherent meaning to the word “investment” are often engaged in the same exercise. That was also the conclusion of the tribunal in Philip Morris Brands SARL v Uruguay ICSID Case No ARB/10/7, Award of 8 July 2016 (Bernardini, Born, Crawford), [206], when noting that the Salini criteria “may assist in identifying or excluding in extreme cases the presence of an investment.”

Outside of these examples, the extent of any additional requirement beyond ownership or control of the asset is far less clear. In particular:

It is not necessary in this case to try and map the full limits of any inherent meaning of the word investment, and I doubt the utility of seeking to do so given the essentially impressionistic nature of the issue. For reasons which I set out below, when regard is had to what the Tribunal described as the realities of Conneco’s business, I am satisfied that the Arbitral Claimants had made an investment in the CSFR which satisfied the three-limb test for which Mr Bastin KC contends.

Conneco was founded by, inter alia, Diag Human AG which originally held 49% of its shares. However, on 22 August 1990 one shareholder withdrew, as a result of which Diag Human AG had deposited 79% of its total share capital in the sum of CSK 2,370,000 (Articles of Association approved on 22 August 1990), later increased to CSK 2,400,000. The share capital was paid up (General Meeting Minutes of 14 May 1991 and 10 January 1992).

Conneco obtained various permits from the CSFR for the purposes of its activities:

I find that Conneco employed and trained employees, and acquired and refurbished premises in Prosek, all as part of or for the purpose of its business. Equipment acquired by Conneco included a walk-in refrigerator (as is apparent from its 1992 Budget of Expenditure). The Prosek premises were acquired for CSK 8 million by a contract dated 8 May 1991. Invoices and delivery notes indicate that Conneco operated from the Prosek premises. Prosek functioned as a distribution warehouse which was inspected by the State Institute for Drug Control on 6 August 1991 and found to “meet all the requirements for distribution of pharmaceuticals in accordance with the regulation No 284.90 Coll ‘On correct product, practice, quality management of human medicine and preparation of medical and packaging technology’”. On 27 August 1991, the Ministry of Health granted a certificate for distribution activities for the Prosek warehouse which “meets the requirements of Decree of the Ministry of Health No 284/1990.”

Correspondence refers to Conneco participating in training sessions with CSFR hospital staff (e.g. the letter from Conneco to University Hospital of 30 November 1990). The Czech Republic acknowledged in the BIT Arbitration that Conneco had been certified to train Czech Medical staff (Request for Bifurcation, para. 57). That is all consistent with the Tribunal’s (unchallengeable) finding that the Arbitral Claimants invested know-how and goodwill in the CSFR.

The formation of Conneco and the measures taken to put it in a place to conduct business are important when considering the issue of whether the Arbitral Claimants made an investment in the CSFR. This was not, as the Czech Republic’s submissions assumed, a case involving cross-border sales activity, capable of engaging the issues outlined at [75] and [92]-[93] above. Rather a functioning enterprise was established within the CSFR, which acquired physical infrastructure, employed staff and obtained legal and regulatory permits, to undertake business activities there.

The business model which Conneco sought to put into effect involved supplying hospital equipment to CSFR medical or pharmaceutical entities, in return for fresh frozen plasma (“FFP”) blood plasma, which would be sent to Novo Nordisk for processing, with fractionated products being re-imported to the Czech Republic.

I have set out my analysis of the evidence relating to Co-operation Agreements in Annex 1. In short, I am satisfied that four Co-operation Agreements were concluded, and that there were dealings with other health institutions in which equipment was exchanged for promises of FFP. I also accept Mr Stava’s evidence that the nature of the contractual arrangements incentivised the Czech health institutions to maintain the arrangements in existence, as the only means of (i) being able to acquire the equipment they needed and (ii) being able to pay for that equipment when they had no funds. Dr Bojar’s evidence in the BIT Arbitration was that the CSFR lacked the technology to process blood plasma and create blood derivatives and he confirmed Colonel Bohonĕk’s evidence that Czech hospitals “had limited financial resources, especially for the acquisition of modern equipment (e.g. for blood collection and blood processing)”. There is a letter from Dr Boris Bubenik of Hospital Frýdek Místek on 31 August 2003 to similar effect.

In those circumstances, I am satisfied that Conneco had created a business model and network which it had legitimate expectations would continue over a number of years and had established the necessary infrastructure to implement that model. It had also effected a significant amount of up-front supplies which equipped some Czech health institutions to participate in that business model, but also incentivised them to work with Conneco on a medium-term basis.

I have addressed this issue in detail in Annex 2. The effect of my conclusions is as follows:

The Czech Republic has raised a further issue as to whether this co-operation agreement between Conneco and Novo Nordisk was an asset in the territory of the Czech Republic.

As I explained at [89] of the March Judgment, in the BIT Arbitration, the Arbitral Claimants never clearly alleged that the Novo Nordisk agreement was an investment in its own right, and “the first occasion on which this agreement is specifically and unequivocally identified as an investment for jurisdiction purposes” was in the Award. In those circumstances, I found that, if necessary, the Czech Republic had acted with reasonable diligence by taking the point that there was no such agreement in its s.67 application ([89(vii)]).

The point initially taken in the Arbitration Claim Form served on 15 June 2022 was that “there was no investment of, or made by, Arbitration Claimant Diag Human SE in … the Czech Republic” (Details of Claim paragraph 2(a)) and that the “business co-operation agreement with Novo Nordisk” was not a “protected investment” (paragraph 2(b)). That was, perhaps, ambiguous as to whether the complaint was that there was no such agreement, or it was not an investment in the Czech Republic, or both. The Particulars of Claim served on 1 December 2023 repeated the assertions in paragraphs 2(a) and (b) of the Details of Claim. However, the Points of Reply served on 8 January 2024 unequivocally took a territorial point, as I explain below.

The hearing which culminated in the March Judgment was conducted on this point by considering whether a s.73(1) objection could be raised to the argument that there was no Novo Nordisk co-operation agreement. The question of whether the territorial objection was also open did not receive separate consideration. However, when asking the Arbitral Claimants to identify where they said that they had advanced a claim that the Novo Nordisk co-operation agreement constituted an investment, I addressed both the existence and territorial issues raised in the Czech Republic’s pleading as a whole. In finding that the Novo Nordisk agreement objection was open to the Czech Republic, I also referred to the territorial issue in [89] of the March Judgment, stating:

The Order made after the March Judgment described the issue open to the Czech Republic as an objection “so far as it concerns the existence and extent of the Novo Nordisk co-operation agreement, as identified in paragraph 89 of the Judgment”. As will be apparent, that paragraph acknowledges a territorial aspect of the objection.

Against that background, I am satisfied that it is open to the Czech Republic to take a territorial objection now, but the limits of that objection are those issues fairly put in play for this hearing in its Reply. That alleged:

The limits imposed by a requirement that an investment be made “in a territory” of a Contracting State are discussed in Schreuer’s Commentary on the ICSID Convention 3^rd (2022). The commentary reviews cases where the sole investment relied upon is the acquisition of a financial instrument issued by the state, noting that in those cases a number of ITA Awards had held that benefit to the host state was enough to satisfy the territorial nexus ([419]-[426]).

The issue has also been debated in cross-border contract cases which are also discussed in Schreuer, including a number involving pre-shipment inspection services provided by SGS. I was referred to two of these cases.

The first was SGS Société Générale de Surveillance SA v Republic of the Philippines ICSID Case No ARB/02/6, Decision of the Tribunal on Objections to Jurisdiction of 29 January 2004 (El-Kosheri, Crawford, Crivellaro). The BIT in issue extended protection to “investments in the territory of the one Contracting Party”. SGS had been retained to provide pre-shipment inspections of exports to be made to the Philippines, for which purpose SGS had also established a Manila liaison office. At [101], the tribunal rejected the attempt to subdivide the services which SGS was providing under its contract, the focal point of which was the provision in the Philippines of a reliable inspection certificate, which would enable goods to be exported to the Philippines and customs duties collected there. The tribunal also referred to the organisational role of the Manila office which was “a substantial office employing a significant number of people”. At [103], it concluded that “these elements taken together are sufficient to qualify the service as one provided in the Philippines”, and at [112] it noted that “there was no distinct or separate investment made elsewhere than in the territory of the Philippines, but a single integrated process of inspection arranged through the Manila Liaison Office, itself unquestionably an investment ‘in the territory of’ the Philippines”.

The second was SGS Société Générale de Surveillance SA v Paraguay ICSID Case No ARB/07/29, Decision on Jurisdiction of 12 February 2010 (Alexandrov, Donovan, Mexia). This was another case in which the BIT extended protection to “investments in the territory” of the host state. The services provided were similar to those which had been considered in the Philippines award. At [113], the tribunal said that an approach which rested “on a parsing of SGS’s investments and activities under the Contract is not sustainable”, and that it was not appropriate to attribute the BIT claim only to the activities abroad. The tribunal relied on the indivisibility of the claimant’s investment ([115]). At [117], the tribunal noted that the same analysis had been adopted on very similar facts in SGS Société Générale de Surveillance SA S v Pakistan ICSID Case No ARB/01/13, Decision of the Tribunal on Objections to Jurisdiction of 6 August 2003 (Feliciano, Faures, Thomas) and in BIVAC v Paraguay ICSID Case No ARB/07/08, Decision of the Tribunal on Objections to Jurisdiction of 29 May 2009 (Knieper, Fortier, Sands).

Having reviewed those authorities, the commentary concludes at [436]:

It will be apparent from these cases that, in considering whether any requirement of territoriality is met, it is necessary to consider the investment in its entirety, and not to take some part of an integrated whole and consider it in isolation. That approach, which I regard as intuitively correct, is supported by other ITA awards to which I was referred:

Approached from that perspective, I am satisfied that the co-operation agreement which I have found to exist between Conneco and Novo Nordisk was an integral part of the business established by the Arbitral Claimants in the Czech Republic. The raison d’etre of the agreement with Novo Nordisk was to enable Conneco to fractionate the FFP collected in the Czech Republic and deliver the fractionated blood products produced using that FFP to health institutions in the Czech Republic. For that purpose, Conneco was capitalised, acquired premises, established collection and storage facilities and concluded Co-operation Agreements with certain Czech health institutions, all within the Czech Republic. For that reason alone, the Czech Republic’s suggestion that the Novo Nordisk co-operation agreement does not meet the territoriality requirement of the BIT must fail.

In support of its contrary argument, the Czech Republic relied on the decision in Swissbourgh v Lesotho [2018] SGCA 81. In that case, the Singapore Court of Appeal emphasised the need for a “territorial link with the host state” ([99]), continuing:

Swissbourgh was not a case in which it was open to the claimant to present its claim as one based on an integrated investment with elements inside and outside of the host state. That is because the disputes concerning the lease agreements which it had entered into over mines in Lesotho had arisen before the relevant Investment Protocol relied upon came into force. The only claim open to the claimant concerned the decision of the member states of the Southern African Development Community (“SADC”), including Lesotho, to terminate the ability of the SADC tribunal to hear investor-state disputes, and the failure to provide an alternative means of dispute resolution, which were said to give rise to admissible claims under the Protocol. The “investment” in question, therefore, was said to be Swissbourgh’s right to bring an investor-state claim against Lesotho before the SADC tribunal and the claim it had brought before access to the SADC tribunal was closed off or “shuttered” by the SADC states.

The Singapore Court of Appeal noted that States cannot purport to protect rights or property located outside their borders ([102]), and that the existence of rights or property was a matter of domestic law and not international law ([103]). They held that the rights constituting an investment cannot be held “on the international law plane, unless that right is within the State’s sole control or the State has expressly undertaken to guarantee that right” ([137]). I note, by contrast, that the Singapore Court of Appeal was of the view that the right to arbitrate under a BIT against a particular state would pass the relevant territorial nexus because it was “a right that lies within the enforcement jurisdiction of the host state and can be guaranteed by that State acting unilaterally”. This is so even though the law governing that right would not be the law of the host State, but international law.

Even if the Novo Nordisk co-operation agreement had, contrary to my determination, fallen to be considered in isolation, I would still not have accepted that it did not satisfy the required territorial link. I do not regard Swissbourgh as holding that only contractual rights whose applicable law is that of the host state are capable of being an investment in that State. That would be inconsistent with its observations on a BIT arbitration offering and with a number of ITA awards (see Schreuer, [437]). In any event, no argument by reference to applicable law was raised by the Czech Republic until after the oral hearing, and I am not willing to allow argument on this point to surface at such a late stage. Nor is any argument open that if there was an agreement with Novo Nordisk creating rights which were enforceable under its applicable law (whatever that might be), those rights would not be capable of being enforced in the Czech Republic. The Czech Republic took its stand on the absence of any agreement and issues of governing law or non-recognition were not pleaded or advanced in oral argument.

Adopting the flexible approach referred to in Schreuer, [436] the Novo Nordisk co-operation agreement was an agreement involving a Czech party, to export Czech FFP collected by that Czech party, for fractionation in Denmark, with return of the fractionated products to the Czech Republic for supply by that Czech party to Czech healthcare institutions. That amply meets the requirement for a territorial link with the Czech Republic, even if viewed in isolation.

Finally, even if I had been persuaded that the Novo Nordisk co-operation agreement could not be considered as part of the Arbitral Claimants’ investment in the Czech Republic (or even that, however established the business relationship may have been, it was not a contract), I would still not have set aside the Award on that basis. As I observed in [89(i)] of the March Judgment, even if not an investment in itself, “it was central to the means by which the Claimants were intending to carry out the Czech business.” The termination of the Novo-Nordisk-Conneco relationship by the Bojar Letter was the mechanism by which the Czech business which benefited from that relationship in respect of a business which was established and operating in the Czech Republic was found to have suffered loss. It was the loss of that business – not any intrinsic value in the Novo Nordisk co-operation agreement as an asset – which the tribunal valued and compensated by means which I have held cannot be impugned in this court. Any complaints about those conclusions are matters of merits, not jurisdiction.

Section 67 provides the court with a discretion as to whether to award relief, and the court can refuse to grant relief if satisfied that the matters giving rise to the s.67 challenge would not have effected the outcome of the arbitration (Integral Petroleum SA v Melars Group Ltd [2015] EWHC 1893 (Comm), [26]; [2016] EWCA Civ 108, [26]-[28]). If, contrary to the conclusion I have reached, the Novo Nordisk co-operation agreement was not an investment, I would in any event have refused relief. On the basis of the reasoning of the Tribunal, I am satisfied that there would have been no difference in the amount of the award had the Tribunal proceeded on the basis that the Novo Nordisk co-operation agreement was not in itself an investment.

On the basis of these factual findings, I am amply satisfied that the Arbitral Claimants have established an “investment” for the purposes of the BIT. Looked at globally:

This state of affairs – not least in the significant presence the Conneco business established in the Czech Republic to enable it to conclude and perform its contracts – is very far removed indeed from a cross-border sale contract of the kind which has tested the limits of the concept of an investment in ITA jurisprudence:

Nor can I accept that the investment I have found can be described as lacking sufficient substance to merit protection under the BIT. While there is no minimal value which an investment must have to constitute a protected investment (see [87(ii)] above), the investment established on the evidence cannot conceivably be described as “minimal”, still less “de minimis”.

Even if I had been persuaded to apply some of the more marginal criteria seen in ITA Awards when seeking to ascertain if there has been an investment – the prospect of regularity of profit, and a contribution to the development of the host state – it is clear that the business established through Conneco was intended to generate profits over time through a period of continuous trading, and to do so by taking advantage of the economic opportunity presented by the desire of Czech health institutions for medical equipment and fractionated blood products.

Finally, I should address Mr Bastin KC’s submission that, at the time of entering into those Co-operation Agreements which subsisted at the time of the Bojar Letter, Conneco had no reasonable expectation of those agreements having any significant duration because of its failure in the two tender processes and the letters sent by Dr Bojar to Czech health institutions warning them off contracting with Conneco in 1991. I understand this argument to be essentially directed to the issue of duration. I am unable to accept it as a matter of law or fact.

So far as its validity as a matter of law is concerned, it seeks to take the 1992 Co-operation Agreements Conneco entered into with Polyclinic Uherské Hradištĕ and Frýdek-Místek(which I accept replaced the earlier Co-operation Agreements with those entities) and with Hospital Polyclinic Mĕlník (which may well have done so, although unlike the other contracts, it does not expressly say so) and examine them in isolation. However, they formed part of the overall investment which had begun in 1990, and continued relationships with those institutions which had resulted in contracts the year before. It would, in my assessment, be entirely artificial to treat the 1992 Co-operation Agreements as reflecting a new investment, whose duration fell to be tested on the date they were concluded. The suggestion that the Arbitral Claimants had a qualifying investment up to January 1992, but not thereafter, is equally unrealistic.

In any event, I am not persuaded on the facts that Conneco was or ought reasonably to have been aware that the three Co-operation Agreements entered into in January 1992 were “doomed from the start”. This issue overlaps with the objection ratione temporis which I address below. However, my conclusions at [155]-163] provide my answer to this part of the Czech Republic’s case on the facts.

I am satisfied that the investment I have found to exist is substantially that found by the Tribunal. The Tribunal sought to identify “the reality of Conneco’s business in Czechoslovakia”, and the “Claimants’ investment in Conneco” (Award, [384]), adopting the holistic approach which I have concluded is appropriate when seeking to identify the Arbitral Claimants’ investment. Within that framework, the Tribunal found an investment comprising the following:

Finally, as I have noted at [124] above, s.67 provides the court with a discretion as to whether to award relief, and the court can refuse to grant relief if satisfied that the matters giving rise to the s.67 challenge would not have effected the outcome of the arbitration. If, contrary to the conclusion I have reached, the investment I have found to be established differs to any meaningful extent from that found by the Tribunal, I would in any event have refused relief. On the basis of the reasoning of the Tribunal, I am satisfied that there would have been no difference in the amount of the award had the Tribunal proceeded on the basis of my conclusions.

The Tribunal also found that the Arbitration Agreement constituted part of the Arbitral Claimants’ investment (Award, [393]). I have dealt with this issue comparatively briefly. At the hearing, the Czech Republic had submitted that the issue of what the effect of a successful challenge to the Tribunal’s finding in relation to the Arbitration Agreement would be on the Award should be left until after judgment. The Arbitral Claimants did not expressly challenge that suggestion. In those circumstances, it may be necessary to hear further argument on this issue.

There is ITA jurisprudence addressing the issue of whether an arbitral award can constitute an investment. A majority of ITA awards have been willing to treat an arbitral award arising from a protected investment as representing the continuation of that investment, and crystallising the rights which arise from the original investment (see e.g. Saipem SpA v Bangladesh ICSID ARB/05/07, Decision on Jurisdiction of 21 March 2007 (Kaufmann-Kohler, Schreyer, Otton), [110], [127]; Frontier PS v Czech Republic UNCITRAL arbitration, Award of 12 November 2010 (Williams, Alvarez, Schreuer), [231]).

The tribunal in GEA Group Aktiengesellshaft v Ukraine ICSID Case No ARB/08/16, Award of 31 March 2011 (van den Berg, Landau, Stern) reached a contrary view ([158]-[162]), stating that it was not sufficient that an award arose out of investment because the two were analytically distinct. However, I am satisfied that this decision represents a minority view: see Schreuer’s Commentary on the ICSID Convention, 3^rd (2022), p.49; White Industries Australia v India UNCITRAL arbitration, Final Award of 30 November 2011 (Rowley, Brower and Lau), [7.6.3], [7.6.8] and 7.6.10]; Gavazzi v Romania ICSID Case No ARB/12/25, Decision on Jurisdiction, Admissibility and Liability of 21 April 2015 (van Houtte, Veeder, Rubino-Sammartano), [120]; Anglia Auto Accessories Ltd v Czech Republic SCC arbitration, Final Award of 10 March 2017 (Banifatemi, Reinisch, Sands), [151].

Consistent with that majority view, it has been held that if the arbitration award arises in relation to a transaction which is not an investment, the crystallisation of the parties’ rights by an award will not create an investment where none otherwise existed (Romak SA v Uzbekistan PCA Case No AA 280, Award of 26 November 2009 (Mantilla-Serrano, Rubins, Molfessis), [212]).

There is rather less in the way of ITA jurisprudence addressing the issue of whether an arbitration agreement can constitute an investment. The issue was left open in Saipem SpA v Bangladesh ICSID ARB/05/07, Decision on Jurisdiction of 21 March 2007 (Kaufmann-Kohler, Schreuer, Otton), [128] and in Anglia Auto Accessories Ltd v Czech Republic FSCC arbitration, Final Award of 10 March 2017 (Banifatemi, Reinisch, Sands), [154]. However, in ATA Construction v Jordan ICSID Case No ARB/08/02, Award of 18 May 2010 (Fortier, El-Kosheri, Reisman), [116]-[117] the tribunal held that a right to arbitration was a “distinct ‘investment’”.

Finally, the Singapore Court of Appeal in Swissbourgh v Lesotho [2018] SGCA 81 reviewed the ITA jurisprudence on this topic. The Court noted at [163]-[164] that “a part-heard arbitration claim …. could …. in principle be considered a stand-alone investment in it’s own right” or “a continuation or transformation of the original investment”, approaches which the Singapore Court of Appeal was prepared to accept “in principle” ([171]).

The issue of whether an arbitration agreement can constitute a “distinct investment”, if that means that it may satisfy the concept of an investment independently of any underlying investment to which it is linked, is controversial. However, I am satisfied that an arbitration agreement can be a crucial step in the continuity of investment between the original investment and any final arbitration award, and in an appropriate case can be regarded as a continuation of that investment, at least in conjunction with the claims to be arbitrated (see by analogy Chevron Corporation and Texaco Petroleum Company v The Republic of Ecuador UNCITRAL PCA Case No 348 77, Interim Award of 1 December 2008 (Bockstiegel, Brower, van den Berg), [180]) which treated legal claims relating to the liquidation of an original protected investment as part of that investment).

In this case, a bespoke arbitration agreement was entered into after Conneco had commenced court proceedings in respect of the effect of the Bojar Letter on what I have found to be a protected investment. It was entered into to address a particular dispute. It provided for arbitration in the Czech Republic under the provisions of Czech arbitration legislation. While the Czech Republic sought to suggest that the fact that this was a post-dispute arbitration agreement (rather than a general agreement to arbitrate future disputes forming part of a contract which was itself an investment) made it a less compelling candidate to constitute part of an investment, I am satisfied that the contrary is the case: the specific link between this agreement and the investment is clear from the circumstances in which it came into existence, its narrow scope and the fact that it was intended to provide a means for assessing compensation for the interference with what I have found to be a protected investment. The arbitration agreement also contained an express undertaking on the part of the Czech Republic to comply with any award. As the Singapore Court of Appeal noted in Swissbourgh v Lesotho [2018] SGCA 81, [128] when approving the submissions of the amici curiae, “a right associated with an investment need not be in existence at the time the investment was originally made in order for the right to be protected under the investment treaty, given that the ‘protection afforded by investment treaties is not static, it is dynamic’ ….”

Accordingly, I am satisfied that, in the circumstances of this case, the agreement to arbitrate the dispute concerning compensation payable for the loss caused by the Bojar Letter to Diag SE’s business (which I have found to constitute a protected investment) with the specific covenant on the part of the Czech Republic to comply with any award can properly be regarded as a continuation of that investment. It follows that I agree with the Tribunal’s decision on this issue as well.

In these circumstances, it is not necessary for me to address the additional assets relied upon by the Arbitral Claimants to establish the making of an investment. However, there is one issue which I should address. Mr Bastin KC submitted that where an arbitral tribunal does not find it necessary to rule on some of the assets which the claimant advances in a BIT arbitration as “investments”, the claimant can only rely on those investments in answer to the state’s s.67 challenge if the claimant brings its own s.67 challenge.

I would not have accepted that submission in this case, where the Arbitral Claimants were seeking, in broad terms, to establish the same general or overall investment as that for which they contended before the Tribunal and which the Tribunal upheld. I can see that there could be more complex cases in which an arbitral claimant identified two different investments before the tribunal which were said to lead to the same loss (e.g. an original investment and a subsequent lawsuit said to constitute the continuation of that investment), and the tribunal only ruled on one of them. Even in this case, if the arbitral claimant was not seeking different relief to that awarded by the tribunal, I am not persuaded that it would have to issue its own s.67 application to be able to raise the alternative route to the same destination before the court (although it would need to give fair notice of the point). Part 62 of the CPR only makes provision for the service of a formal document for an arbitrating party who, in response to a challenge by the other party, seeks to uphold the award before the court by alternative reasoning to that relied upon by the tribunal: in applications under s.69 of the 1996 Act (PD62 12.6(2)). There is no equivalent requirement for a party who seeks to resist a s.68 challenge on the basis that the tribunal would have reached the same result anyway (cf. RAV BahamasLtd v Therapy Beach Club Inc [2021] UKPC 8, [2021] AC 907, [36]).

For the reasons I have set out, I reject the Czech Republic’s No Investment Challenge to the Award.

I conclude on this issue merely by observing that there has always been a stark disconnect between the manner in which the Czech Republic treated Conneco’s business in the course of the underlying events, and the “No Investment” argument in the BIT Arbitration and before me. Whether in the sending of the Bojar Letter, or the subsequent meetings in 1992, or in entering into the agreement to the Commercial Arbitration, or in its conduct of that arbitration, the Czech Republic has always treated Conneco’s business as a significant state of affairs, which merited a significant (and at times excessive) state response. That provides some corroboration for my conclusion that the investment manifest in that business did not lack the substance necessary to constitute a protected investment under the BIT.

The Czech Republic argues that the Tribunal did not have jurisdiction to determine a dispute which was in existence before the BIT came into force on 7 August 1991. In the March Judgment, I upheld the Czech Republic’s submission that the objection they sought to raise was jurisdictional in nature: [157]. Before me, I did not understand the Arbitral Claimants to contend that if the dispute had arisen before the BIT came into force, this would not have deprived the Tribunal of jurisdiction (just as they had not challenged the Tribunal’s decision to the same effect). I am satisfied that that reflects the correct characterisation of the issue. The offer to arbitrate investment disputes contained in Article 9 of the BIT is to be interpreted in accordance with the principle of non-retrospectivity in Article 28 of the VCLT. That conclusion accords with the views expressed in a significant number of ITA Awards referred to in paragraph 102 of the Czech Republic’s skeleton argument, which it is not necessary to set out here.

The issue which arises on this challenge is whether the Arbitral Claimants’ claims relating to the sending of the Bojar Letter on 9 March 1992 was in reality the continuation of an existing dispute between the Arbitral Claimants and the Czech Republic, which began when Diag SE was notified that it had not succeeded in the first tender on 4 October 1990.

Both sides were content to adopt the definition of what constitutes a dispute applied in Empresas Lucchetti SA v Peru ICSID Case ARB/3/04, Award of 7 February 2005 (Buergenthal, Cremades, Paulsson), [48]: “a disagreement on a point of law or fact, a conflict of legal views or of interests between two persons” or “a situation in which two sides hold clearly opposite views concerning the question of the performance or non-performance of a legal obligation”.

In that case, the claimant sought to build a plant in a Peruvian municipality, its project being negatively affected by a series of administrative measures in 1997 and 1998 taken by the municipal authorities. It was common ground that this constituted a dispute. The claimant succeeded in a court challenge which ordered the municipality to provide the permit it had previously annulled, which the municipality did in notably grudging terms. However, in 2001, the municipality revoked the permits, in an order which recited the history of the matter, including the 1997 measures and the resultant litigation, suggesting in their decisions that the favourable court judgments had been obtained by corruption. The issue was whether this was a new dispute. In concluding (I would respectfully suggest, not surprisingly) that it was not, the tribunal considered “whether and to what extent the subject matter or facts that were the real cause of the disputes differ from or are identical to the other” ([50]). The dispute was the municipality’s desire to prevent the construction and operation of a factory which (both in 1997 and 2001) it regarded as contravening its environmental policies.

Other ITA Awards have referred to difficulties in applying the Lucchetti test, preferring the so-called “triple identity” test first espoused in this context in Railway Development Corporation v Guatemala ICSID Case No ARB/07/23, Second Decision on Objections to Jurisdiction of 10 May 2018 (Sureda, Eizenstat, Crawford), [131]: whether there was “identify of the parties, object and cause of action”. That was the view of the tribunal in Agility Public Warehousing Company KSC v Iraq ICSID Case No ARB/17/7, Decision on Jurisdiction of 9 July 2010 (Bull, Beechey, Murphy) [178]-[181].

A number of ITA Awards addressing this issue were considered by Kannan Ramesh J in Kingdom of Lesotho v Swissbourgh Diamond Mines (Pty) Limited [2017] SGHC 195, [121]- [168] (the issue was not considered on appeal). Kannan Ramesh J also criticised the test formulated in Lucchetti ([130]), suggesting that “the cause-of-action approach is a better way of ascertaining the real dispute than the subject-matter approach.” Based on his review of the ITA Awards, the Judge summarised the principles he derived from the ITA Awards at [171]-[176]:

I have had regard to these various formulations, but I have ultimately been guided by the principle in issue: is the tribunal, in substance, being asked to take jurisdiction over a dispute which came into existence before the treaty came into force? While the identity of the claimant or the cause of action may be of some assistance in answering that question, they cannot be determinative (not least because that would open the way for inventive lawyering to circumvent the non-retroactivity principle).

Mr Bastin KC invited me to test the issue by considering what would happen if there had been a pre-arbitration process in respect of the first set of facts, and the second set of facts had then occurred, and ask whether a second pre-arbitration process would be required. One might also raise a similar issue of whether a claim in relation to the second set of facts could be brought into a pending arbitration commenced by reference to the first. However, I do not think these analogies – in which considerations of procedural economy and expeditious dispute resolution are paramount – assist on the issue of the non-retrospective application of the BIT.

I have concluded that the answer to this question is relatively clear, and it is the answer the Tribunal gave.

I accept that the complaint relating to the 1990 tender was a dispute which pre-existed the coming into force of the BIT. Mr Stava raised this dispute with Dr Bojar on 4 October 1990, which I accept reflects knowledge on Mr Stava’s part that Conneco would not be declared the winner of the tender. In any event, a complaint about the 1990 tender process was made by Conneco to the Ministry of Health on 9 July 1991.

On 20 February 1991, Dr Bojar wrote to selected Czech hospitals requesting and insisting that they stop negotiations with foreign partners. I have seen no evidence that this directive was communicated to Conneco prior to 7 August 1991, still less that Conneco had made any complaint about it:

On 25 June 1991, the 1991 Selection Committee decided not to accept Conneco’s or Novo Nordisk’s proposals and recommended that three other companies move to the next round. I accept that Conneco was aware that it had not reached the second round by 9 July 1991 when it wrote complaining about this fact. Any dispute in relation to the second tender had crystallised at this point. The thrust of Conneco’s complaint was that the pricing comparison in the 1991 tender had not been performed appropriately. It is no answer that the tender had not yet been awarded: a party who fails to make it through the group stage cannot treat its non-appearance in the final as a separate dispute.

On 15 July 1991, Dr Turek sent a memorandum to Dr Bojar reporting on the outcome of the review of the tenders undertaken by the evaluation committee formed by the Ministry of Health. Dr Turek recommended that, with the launch of the new venture with the successful tenderers, it was necessary “to prohibit negotiations with processors other than the recommended ones” and that changes should be made in the competency of the Ministry of Health to permit this. Dr Turek confirmed that this was a recommendation to the Ministry of Health. There is no evidence as to when the recommendation was accepted.

In the period after 9 July 1991 (still less before 7 August 1991, when the BIT came into effect), there is nothing to indicate that Conneco perceived any threat to its existing business in the Czech Republic, although aware it had not won the tender, nor is there any evidence that it was aware of the recommendation made by Dr Turek’s committee on 15 July 1991 (still less that it had been accepted by the Ministry of Health, even assuming that was in fact the case). The contemporaneous documents establish the following:

The Czech Republic invited me to resolve this issue in the following way: “if one approaches this chronologically and asks how things looked shortly after the Bojar Letter was sent and whether the parties’ relationship was markedly different or simply a changed version?” The answer to that question is that the former is emphatically the case. The wholesale (and dramatic) changes which the Bojar Letter brought about are summarised in Annex 2. The contrast between the conduct of Conneco and Novo Nordisk after that date, and the position prior to 7 August 1991, is stark. The effect of the Bojar Letter was completely to cut-off the foundation of Conneco’s business in the Czech Republic and represented a dramatic departure from what had gone before, and a new dispute.

The terms of the Commercial Arbitration Agreement – stating “the parties have agreed that the dispute between them in respect of compensation for the loss allegedly caused in connection with [the Bojar Letter] should be arbitrated” – also confirms the fact that this represented a new and significant issue between the parties.

Reverting to the factors identified in the case law by way of further explanation of that conclusion, the Bojar Letter dispute:

For these reasons, the Czech Republic’s ratione temporae challenge is dismissed.

That brings me, finally, to what has long been the most important and most difficult issue in this arbitration challenge: did the placing of the shares in Diag SE in the Koruna Trust in June 2011 have the effect that it ceased to be a Swiss company for the purposes of the BIT, and therefore ineligible to accept the standing offer to arbitrate investment disputes contained in Article 9 of the BIT when the Request for Arbitration was served? That issue ultimately turns on the construction of Article 1(1)(c) of the BIT.

The issue arises as follows:

The Czech Republic contends that Diag SE does not fall within Article 1(1). It is common ground that it could only do so if it fell within Article 1(1)(c). The Czech Republic is not able to challenge the Tribunal’s conclusion that it did so up to June 2011 but contends that it ceased to do so at that point because its shares were transferred into a Liechtenstein discretionary trust called the Koruna Trust, the trustee of which is a legal person with its seat and real economic activity in Liechtenstein. I have held that the ambit of that challenge which is open before the court is limited to the period from the setting up of the Koruna Trust to the end of 2011.

For the purposes of this issue, I held that the scope of the objection open to the Czech Republic related to the effect of events occurring in the time period of the objection taken before the Tribunal, which related to the transfer of the shares in Diag SE into the Koruna Trust in June 2011 up to the end of the pledge arrangements put in place as an element of the Lawbook Transaction at the end of 2011. To the extent that some fresh event was relied upon after that date as altering the position – i.e. an argument to the effect that even if Diag SE was still controlled by Mr Stava in a relevant sense at the end of 2011, this was no longer the case thereafter – I held that this would not constitute the same “ground for objection” as that taken before the Tribunal. For that reason, the arguments before me and the factual findings made relate to the position as at the end of December 2011.

When I gave permission to appeal on the Stava June 2011 Objection, I noted the risk that, if my decision was overturned, a factual enquiry would be required in relation to Mr Stava’s legal rights concerning and factual control over Diag SE which the court would be considering as part of the Czech Republic’s nationality challenge to Diag SE’s claims. I made it clear that the court would expect the factual findings (including for this purpose findings of Liechtenstein law) made at this hearing to be binding on both the Czech Republic and Mr Stava in that eventuality. In the event, the parties were able to agree undertakings as follows:

I have set out my findings of fact, which include, for this purpose, my findings of Liechtenstein law, in Annex 3. My conclusions are as follows:

The effect of these factual findings is that if Article 1(1)(c) of the BIT requires control of a legal entity to arise from legal rights, or to derive to some extent from a proprietary interest in that legal entity (which I shall refer to by the shorthand “de jure control”), then Mr Stava did not have it. If, however, de facto control might in some circumstances be sufficient, then further analysis is required.

By way of a reminder, Article 1(1) provides:

The text of the BIT offers relatively little guidance as to what the concept of “control” involves:

Finally, the Protocol to the BIT provides:

The Czech Republic suggested that the reference to “proof” in Protocol (1) supported its case that control via legal rights or involving an ownership interest was required. However, de facto as well as de jure control can be “proved”, and I regard the reference as neutral. Protocol (2) appears to provide, inter alia, for a situation where a company incorporated in State X may have a claim under an investment treaty between State X and the Czech Republic, and also a claim under the BIT by virtue of Article 1(1)(c) control. However, it does not assist on the meaning of the concept of control.

I accept (as many ITA Awards have concluded) that the concept of control can, as a matter of its ordinary meaning in general usage and in the legal lexicon, embrace both de facto and de jure control. All will depend on the context in which it is used.

I was referred to a number of ITA Awards and passages from commentary addressing the meaning of “control” in investment treaties and the ICSID Convention. In considering that material, it is important to distinguish between the different contexts in which the issue of control of a corporate entity can arise in investor state dispute settlement.

The present context involves determining the nationality of a corporation, for the purposes of determining whether it has protected nationality:

Issues of corporate control can also arise when determining whether a claimant has protected nationality in the following contexts:

I was referred to a number of ITA Awards which had addressed the issue of control in these or related contexts.

Vacuum Salt Products Limited v Governor of the Republic of Ghana ICSID Case No ARB/92/1, Award of 16 February 1994 (Jennings, Brower, Hossain) had to address Article 25(2)(b) of the ICSID Convention. The claimant company was incorporated under the laws of Ghana, and in order to invoke the ICSID Convention for the purpose of its claim against the Republic of Ghana, the claimant had to bring itself within Article 25(2)(b). A Greek national, Mr Panagiotopulos, held 20% of the shares in the claimant, with the other 80% held by Ghanian entities. At [43]-[44], the tribunal held:

The tribunal considered the evidence intended to suggest that Mr Panagiotopulos had control over Vacuum Salt, concluding in the following terms at [53]:

Mr Dunning KC understandably placed reliance on the tribunal’s observation that “a total absence of foreign shareholding would virtually preclude the existence of such control.” However, I did not understand the tribunal to find that some shareholding was a legal pre-requisite to a finding of foreign control. It is difficult to see why the presence of a 1%, 2% or 5% shareholding which was not in itself a means of control should be determinative if the control was effected by other means. The overall effect of the decision is that the question of foreign control is a flexible one, to be determined in the light of all facts and circumstances, with material influence in a managerial rather than technical or supervisory way, and being “in charge”, being relevant factors.

In Plama Consortium Limited v Bulgaria ICSID Case No ARB/03/24, Award of 8 February 2005 (Salans, van den Berg, Veeder), when discussing the denial of benefits provision in Article 17 of the ECT, the tribunal stated at [170] that “control includes control in fact, including an ability to exercise substantial influence over the legal entity’s management and the selection of members of its board of directors or any other management body” (noting that that appeared to be common ground).

In Aguas del Tunari SA v Republic of Bolivia ICSD Case No ARB/02/3, Decision on Respondents Objections to Jurisdiction of 21 October 2005 (Caron, Alberro-Semerena, Alvarez), the issue of control arose in a very similar context to the present – whether a legal person which was not constituted under the laws of the Netherlands nonetheless constituted a Dutch national for the purpose of the Netherlands-Bolivia BIT because it was “controlled directly or indirectly by nationals of that Contracting Party …”. It is important to note the context in which the issue of “control” arose. A Dutch company owned 100% of another Dutch company, which owned 55% of a Liechtenstein company which owned 55% of the claimant. The claimant argued that this was sufficient to establish its Dutch nationality (it did not need to, and did not, argue that a certain level of shareholding was not simply sufficient, but also necessary). Bolivia argued that in addition to de jure control, it was also necessary to establish de facto control, and its case was that the two Dutch companies were mere shells, with actual control of the Bolivian corporate claimant being exercised further up the corporate chain by US companies. The majority of the tribunal accepted that the ordinary and legal meanings of the word “control” embraced both de jure and de facto control ([227, [231], [233]), although they did suggest at [242]:

The majority of the tribunal did, however, make a forensically powerful point against a requirement of factual control, which I accept is equally applicable to an alternative free-standing test of de facto control: that it was “sufficiently vague as to be unmanageable” ([246]). That said, the majority were clearly prepared to contemplate a wide variety of forms of legal control, referring at [264] to “the percentage of shares held, legal rights conveyed in instruments or agreements such as the articles of incorporation or shareholders’ agreements, or a combination of these”. In addition, they did not purport to define the concept exhaustively (“the tribunal observes that it is not charged with determining all forms which control may take”: [264]) and they appeared to accept that issues of factual control might be relevant in cases in which the alleged foreign controller had a minority shareholding ([288]).

In Yukos Universal Limited (Isle of Man) v The Russian Federation PCA Case No 2005-04/AA227, Interim Award on Jurisdiction and Admissibility of 30 November 2009 (Fortier, Poncet, Schwebel), the issue which arose was whether the “denial of benefit” provision in Article 17 of the ECT was engaged when the shares in the claimant company were held in Guernsey discretionary trusts, with most of the trusts having a Protector who was required to consent to important decisions. While it was not necessary to decide the issue to resolve the dispute, the suggestion by the Russian Federation that, in effect, the tribunal should “look behind” the trusteeship to those beneficially interested in the trusts was rejected, the tribunal holding at [517] that the trusts were “valid and effective as a matter of Guernsey trust law, and vest ownership and control to the Trustees within the framework of the trust instruments.” It is fair to say that the tribunal regarded the legal position under Guernsey trust law as effectively determinative of this question. However, this may have reflected the case management decision at [509] to defer the issue of whether the former shareholders “may have retained control over the shares transferred into trust as part of a structure they may have established or used at various points for fraudulent or criminal purposes”, addressing only “certain issues which arise within the more narrow framework” of Guernsey trust law ([510]-[511]).

In Bernhard von Pezold & Others v Zimbabwe ICSID Case No ARB/10/14, Award of 28 July 2015 (Fortier, Williams, Hwang), issues of control arose both to establish the Swiss nationality of locally incorporated companies for the purpose of Article 1(1)(c) of the Switzerland-Zimbabwe BIT and the availability of the ICSID Convention through the application of Article 25(2)(b). Article 1(1)(c) of the relevant BIT gave legal entities established under the law of the host state the nationality of the other state if they were “effectively controlled” by natural or legal persons who satisfied the terms of Article 1(1)(a) or (b). The chains of ownership of the locally incorporated companies were very complex, with members of the von Pezold family (comprising Rudiger who was German, Elisabeth, the Adult Children and Adam, who were Swiss and German) collectively having:

The Article 25(2)(b) issue in relation to the Border Companies was dealt with by the tribunal very briefly, and, on the facts, its conclusion is not surprising, and to that extent is compatible with the cases of both the Czech Republic and the Arbitral Claimants. However, the terms in which the tribunal resolved the issue was not to point to legal rights, but to the actual control exercised specifically by Elisabeth: “the evidence clearly demonstrates that Elisabeth exercises overall control of the Border Companies and that Rudiger, the Adult Children Claimants and Adam abide by Elisabeth’s exercise of ultimate control over those companies” ([215]). This appears to base the finding of control on Elisabeth’s de facto control. The tribunal found that the BIT nationality provision was satisfied by the Border Companies because they were “effectively controlled by Swiss nationals, and, in particular, by Elisabeth”, with “day-to-day management” by Heinrich, one of the Adult Children ([226]). Once again, the tribunal’s focus appeared to be the factual exercise of control. There is no focus on the word “effectively”, and I do not regard that word of itself as a strong pointer to one means of control rather than another – it is difficult to see how “ineffective control” could be control at all.

Having established jurisdiction ratione personae, the tribunal turned at [317] to what was described as “one last issue … to whom these investments belonged.” The issue of control here arises for a different purpose and in a different context, and I address this part of the von Pezold award at [204] below.

Guardian Fiduciary Trust Ltd v Macedonia ICSID Case No ARB/12/31, Award of 22 September 2015 (Heiskanen, Bucher, Stern) was another case concerned with a similar nationality provision in Article I(b)(III) of the Netherlands-Macedonia BIT. The claimant was a New Zealand company, owned by another New Zealand company CCT. Legal title in the shares in CCT was held by IN Asset Management, a New Zealand company owned by a Dutch Stiftung. However, IN Asset Management held its shares in CCT on trust for CCG, a Marshall Islands company. The tribunal observed that the issue of whether it was the Stiftung or CCG which had control of the claimant was not solely a question of New Zealand law but turned on the evidence ([134]). There was no evidence of actual control of the claimant being exercised by In Asset Management, whereas there was evidence of actual control being exercised by CCG ([137]). Accordingly, the claimant did not fall within Article 1(b)(III) of the BIT.

In United Utilities (Tallinn) BV v Estonia ICSID Case No ARB/14/24, Award of 21 June 2019 (Drymer, Stern, Williams), the issue of control arose in the same contexts: in the Netherlands-Estonia BIT, for the purpose of clothing a legal person constituted under the laws of a non-contracting state with protected nationality where it was “controlled, directly or indirectly by” natural or legal persons who satisfied the BIT’s nationality requirements in Articles 1(1)(a) and (b), and under Article 25(2)(b) of the ICSID Convention. In both contexts, the tribunal stated at [366] that “control is a flexible concept, which can only be determined on a case-by-case basis in the light of particular facts.” At [371], the tribunal cited with approval a passage from Professor Schreuer’s commentary on the concept in the ICSID Convention stating that “foreign control is a complex question requiring the examination of several factors such as equity participation, voting rights and management. In order to obtain a reliable picture, all these aspects must be looked at in conjunction”. The tribunal identified various factors – majority shareholdings, blocking shareholdings, voting rights, contractual arrangements and “the particular nature of the influence exercised by the foreign person over the domestic entity.”

In addition, two ITA Awards considering the effect of Article 1117 of the North American Free Trade Agreement (“NAFTA”) can usefully be considered in this context, because they raise the essentially similar issue of whether claims of a corporate claimant incorporated in the host state can nonetheless be advanced as investment treaty claims by a national of another signatory state because the foreign national “owns or controls directly or indirectly” the local legal entity.

So far as academic commentary on this issue is concerned, I was referred to Michael Schmid’s chapter in Professor Chester Brown’s edited volume, Commentaries on Selected Model Investment Treaties (2013), p.665 which refers to the extension of protection to investments of subsidiaries of the investor in Swiss BITs, and, when explaining why the extension should cover both owned subsidiaries as well as controlled subsidiaries, states “the control of a company requires at least a relative majority of shares”. I do not believe Professor Brown was seeking in this comment to address the complex issue of de facto and de jure concepts of control. He also notes (although I have not relied upon this as I am not persuaded it is admissible under the VCLT) that provisions of this kind reflect Switzerland’s “underlying policy stance for the purposes of investment protection … that it should not matter how an investor structures the investment as long as the investment he controls is located in the territory of the other BIT partner company.”

Schreuer’s Commentary on the ICSID Convention 3^rd (2022), discussing Article 25(2)(b) of the ICSID Convention states at [324]:

Issues of control of a corporation can also arise when determining whether the requisite link has been established between a protected national, and the investment said to give rise to the claim (i.e. as a means of identifying whether the claimant has a protected interest or standing to complain of a particular breach of the treaty).

Sometimes, the terms of the BIT will make it clear that an economic interest deriving from control of the investment, as well as one arising from ownership, will suffice. Thus, the USA Model BIT (2004) defined investment as “every asset that an investor owns or controls, directly or indirectly” and the NAFTA provides that “an investment of an investor of a Party” is “an investment owned or controlled directly or indirectly by an investor of such a Party”. Article 1(6) of the ECT defined an investment as “every kind of asset owned or controlled directly or indirectly by an Investor ….”

The Czech Republic relied on Professor Zachary Douglas’ commentary on these provisions in The International Law of Investment Claims (2009), where he argues both that, even where there is no express requirement that the investor control the investment in issue, such a requirement is implicit ([555]-[556]), and that the link between the investor and the asset must be a legal right of control stating:

At [560], Professor Douglas KC suggests that the words “controlled directly or indirectly” in this context do not correspond to the distinction between de jure and de facto control, with indirect ownership or control meaning cases where a claimant does not have direct ownership or control over the assets comprising the investment, but instead exercise such ownership or control indirectly by having direct ownership or control of the legal entity that does have direct ownership or control of the assets.

Before turning to the ITA Awards to which I was referred with reference to this aspect of control, it is important to note that the relevance of “control” in this context is different to its relevance in the nationality context. One obvious difference between the two contexts can be seen in the summary of the rules which Professor Douglas KC placed at the start of chapter 7 of The International Law of Investment Claims.

In the March Judgment, I held that issues of control addressing the subject of Douglas Rule 30 (does the claimant have protected nationality?) were jurisdictional for the purposes of s.30 of the 1996 Act, whereas at least part of those addressing the subject of Douglas Rule 31 (had the claimant with protected nationality disposed of the investment prior to one of the alleged breaches?) was not. As noted at [256]-[259] below, I have concluded that the same distinction provides the answer to the Arbitral Claimants’ issue estoppel case.

When the investor’s link to the asset is established by control, issues will arise not only as to the existence of the necessary link (the issue discussed by Professor Douglas KC) but also its extent, in particular when determining the extent of any compensable loss – for example, if the investor had minority (or no) ownership of the asset which gives rise to the claim, how is compensation to be assessed? Does control give the investor 100% of the value of the asset, in which case what of a co-investor without control? The ITA Awards I have been referred to do not address this issue.

I can see that considerations of this kind might be said to favour a requirement of control by legal rights as the link between the investor and the asset: legal rights are more readily susceptible to valuation than “de facto” control, not least because they are generally easier to sell. However, even in this context I should record that that has not been the approach in many of the ITA Awards I have been referred to and I am conscious that Professor Douglas KC’s views on this issue are controversial. For present purposes I would simply note that (i) the same issues do not arise in the nationality context, when the issue of control determines the nationality of the claimant, not the existence and extent of the protected investment; and (ii) in the latter context, the argument that the issue of control is jurisdictional becomes even less appealing.

In Bernhard von Pezold & Others v Zimbabwe. ICSID Case No ARB/10/14, Award of 28 July 2015 (Fortier, Williams, Hwang), having satisfied itself that it had jurisdiction ratione personae for the purposes of both the BIT and the ICSID Convention, the tribunal went on to consider in the ratione materiae context “one final issue for discussion: namely to whom these investments belong” ([317]). By way of scene-setting:

In Blue Bank International & Trust (Barbados) Ltd v Venezuela ICSID Case No ARB/12/20, Award of 26 April 2017 (Soderlund, Bermann, Malintoppi), a claim was brought by a Barbados company which managed trust assets under the Barbados-Venezuela BIT. Blue Bank managed the Qatar Trust, which had a Venezuelan settlor. The assets of the trust included two BVI companies which were indirect shareholders in two Venezuelan companies. It was accepted that Blue Bank satisfied the nationality requirement ([157]) but it was objected that it was not an investor because the Qatar Trust, rather that Blue Bank, owned the investment. The tribunal concluded that Blue Bank’s legal title did not answer the issue. It “cannot be considered as having committed any assets in its own right, as having incurred any risk, or as sharing the loss or profit resulting from an investment”, its reward coming in the form of fees. The arguments by reference to Barbados trust law were dismissed on the basis that “in actual fact and law, Blue Bank is not an owner in any relevant sense of the word” ([168]).

Italba Corporation v Oriental Republic of Uruguay ICSID Case ARB/16/9, Award of 22 March 2019 (Oreamuno, Beechey, Douglas) involved a claim under the US-Uruguary BIT, which defined an investment as “every asset that an investor owns or controls directly or indirectly”. The claimant, a US company, claimed to own a Uruguay company called Trigosul, but its claim to ownership was not established on the evidence because there had been no valid share transfer. It advanced an alternative claim that it controlled Trigosul. The tribunal found that this involved “factual situations that must be evaluated on a case-by-case basis” ([254]). The claim failed on the facts, there being no evidence that Italba took business decisions for Trigosul, ([259]), provided it with capital or funding for its operations ([261]-[262]) or controlled it generally ([263]). There was no discussion of the concept of control in principle ([264]).

In Agarwal v Uruguay PCA Case No 2018-04, Award of 6 August 2020 (Rigo Sureda, Johnson, Mayer), the claimants were three UK nationals who brought a claim under the UK-Bolivia BIT in relation to an investment in a Uruguay company, Minera Aratiri. On 1 April 2008, a discretionary trust had been established, which was the ultimate holder of the companies involved in the relevant investment. The trust had a trustee, a protector, the three claimants were in the class of beneficiaries, and the trustees could add additional persons to that class or, with the approval of the protector, exclude a current member. The discretionary trust was converted into a fixed interest trust on 1 August 2016. Against this background, Uruguay took two related jurisdictional or standing objections, although neither were ratione personae objections:

Uruguay succeeded on both points before the tribunal. The tribunal held that the BIT definition of an investment as “any kind of property” required “a possessory link” between the asset and the investor ([180]). It held that the rights of the claimants as discretionary beneficiaries did not constitute property under the BIT, the claimants merely having an expectancy of benefit if the trustee exercised its discretion in a certain way, and “a hope is not an asset” ([224]). As a result, the claimants lacked standing until the trust was converted into a fixed interest trust ([229]). By that date, the dispute had already arisen, which the tribunal held gave rise to a valid jurisdictional objection ratione temporis. The award was successfully challenged before the Paris Cour d’Appel (which included Judge Fabienne Schaller, President of the International Commercial Chamber of the Court) in a judgment of 21 February 2023. First, the Cour d’Appel rejected the argument that there was no jurisdiction because the dispute had arisen before the trust became a fixed trust ([70]-[71]), albeit that might provide a defence on the merits ([73]). Second, they rejected any requirement for a claimant to “actively make” an investment ([81]). Third, they rejected the contention that the BIT imposed a jurisdictional requirement that the claimant have protected nationality at the date of the alleged breach ([84]).

MAKAE Europe v Saudi Arabia ICSID Case No ARB/17/42, Award of 30 August 2021 (Crook, Hafez, van Houtte) was a claim under the BIT, entered into between France and Saudi Arabia, in which “investment” was defined as “property owned or controlled” by the investor. The claimant advanced its case solely on the basis of control ([66]). It was common ground between the parties that de facto control would be sufficient ([117]) and the tribunal agreed ([118]). The tribunal rejected the contention that “de facto control requires a claimant to demonstrate some ownership or other form of economic interest in an investment” ([132]).

Finally in this context, Gramercy Funds Management LLC v Peru UNCITRAL Final Award of 6 December 2022 (Fernandez-Armesto, Drymer, Stern) concerned a claim under Article 10.28 of the US-Peru Free Trade Agreement which stated, “investment means every asset that an investor owns or controls directly or indirectly.” The tribunal derived the meaning of the word “control” as used in that treaty from other usages of the word in the same instrument ([626]) which it said referred to control over a corporation, and held that “control can only be exercised with regard to a corporation in which the investor already had an ownership interest” ([629]). The reasoning which led to that conclusion, and the extent to which it depended on other provisions in the FTA, is difficult to discern. However, the tribunal’s discussion of “ownership” in contrast to “administration” is of interest:

So far as commentary is concerned, I have already referred to the views of Professor Douglas KC at [198]. Schreuer’s Commentary on the ICSID Convention 3^rd (2022) discussed the definition of investment in Article 25 of the ICSID Convention as “every kind of asset, owned or controlled directly or indirectly”. At [322], the commentary states that “while control is often the consequence of sole or majority ownership of an investment, ownership is not the only determining factor for control. The ordinary meaning of the word encompasses other factors, such as special voting rights and powers of operation and management, often conferred through contractual arrangements.”

The attempt to draw these rather disparate strands together can no longer be postponed.

I accept that limiting Article 1(1)(c) to control exercised through legal rights would offer a simpler solution (cf. [186]). I also accept that it might be said that it would be rather unappealing if Mr Stava, having transferred the shares in Diag SE out of his ownership to achieve some legal or fiscal consequence, could when it suited him take the position that those shares were really “his”. That instinctive reaction is put in characteristically clear terms by the late Edmund King QC in his March 2018 lecture, “Is the Pugachev Ruling the End of Offshore Trusts and Should it Be?” He noted at [22]:

However, it is important in this respect to note that the context in which the “control” issue arises in Article 1(1)(c) is rather different. The claims of Diag SE remain the assets of Diag SE, any recoveries available for its creditors and shareholders. The issue which arises is the rather different one of whether Diag SE should benefit from the protections afforded by the Swiss-Czech Republic BIT because of control by a Swiss national. That is a context in which I have been persuaded that the word “control” in Article 1(1)(c) carries both of its ordinary meanings of legal and de facto control, such that, on appropriate facts, Article 1(1)(c) will be engaged where the control of the legal entity by nationals of a Contracting State falling within Article 1(1)(a) and/or (b) can be established by means other than the exercise of control through legal rights.

That is a conclusion which has been derived from the combined effect of a number of factors:

In this case, I am satisfied that Mr Stava had de facto control of Diag SE of a kind which satisfied Article 1(1)(c):

Finally, while I have found that Mr Stava’s influence over any distribution of the fruits of any investment in Diag SE, while significant, fell short of practical control, I do not regard that conclusion as telling against a finding that Mr Stava exercised control over Diag SE for Article 1(1)(c) purposes. Those with a majority ownership in or control of a legal entity may find their ability to distribute the proceeds of its activities curtailed for any number of reasons – contractual commitments, legal obligations arising from the interest of creditors, the presence of minority preference shares etc. To allow Mr Stava’s inability to control (rather than simply significantly influence) which of the members of the Stava family would get to enjoy the benefits of any recoveries by Diag SE if and when its arduous pursuit of its claims against the Czech Republic had a successful culmination to determine the issue of control altogether would be to allow the tail to wag the dog.

Finally, I have been asked by the Czech Republic to address an argument which, on my findings, I do not believe arises:

That is sufficient to reject the challenge to Diag SE’s status as a protected investor. However, in case my construction of Article 1(1)(c) of the BIT does not represent the final word on the subject, I should deal with the Arbitral Claimants’ application to amend its Points of Defence to contend that the Czech Republic was precluded from denying Mr Stava’s control of Diag SE for the purposes of establishing whether Article 1(1)(c) applied by the Tribunal’s finding (which I found is not open to challenge under s.67 of the 1996 Act) that Mr Stava had standing to bring a claim for breaches of the BIT in relation to his investment in Diag SE after May 2011. That preclusion was said to arise either by reason of the doctrine of issue estoppel or because it would be an abuse of process to permit the Czech Republic to advance the argument.

As it played out, it became apparent that the preclusion argument had a number of possible implications for other arguments in the challenges.

The application for permission to amend was opposed, both because it was said that the amendment should be refused as a matter of discretion, and because the point was said not to be arguable. I am satisfied that it is appropriate to grant permission to amend, but that the preclusion argument fails because the issue determined by the Tribunal vis-à-vis Mr Stava is not the same issue as that which arises before me in relation to the s.67 challenge to the Award in favour of Diag SE.

Taking the application for permission to amend first, I was referred to the well-known summary of the applicable principles in Quah Su-Ling v Goldman Sachs [2015] EWHC 759 (Comm), [38]. As to these:

I therefore turn to the argument on its merits.

The Arbitral Claimants relied on the four-fold test for issue estoppel summarised by Clarke LJ in Good Challenger Navegante SA v Metalexportimport SA (The Good Challenger) [2003] EWCA Civ 1668, [50]:

Mr Dunning KC took two preliminary objections, the first of which is that s.67 of the 1996 Act gives a de novo right to challenge a jurisdictional determination by an arbitral tribunal, and that the application of the doctrine of issue estoppel based on a determination in an arbitral award would be fundamentally inconsistent with that right. The first three elements in that argument are not controversial. I accept:

I accept that those provide unanswerable reasons why, to the extent that an arbitral award is subject to a timely and outstanding jurisdictional challenge in the law of the seat, it cannot have preclusive effect in the hearing of that challenge. However, the requisite element of finality would also be absent in those circumstances. That says nothing about those parts of an arbitral award which are not subject to a timely and outstanding jurisdictional challenge, and I am satisfied that there is no principle of law which prevents such a determination from giving rise to an issue estoppel in a s.67 challenge, even when it forms part of the award which is subject to the s.67 challenge (albeit not the part of the award which is the subject of that challenge).

There can be no dispute that a decision in an arbitration between A and B which is not challenged can give rise to an issue estoppel in a subsequent arbitration or court proceedings between A and B (AEGIS Ltd v European Reinsurance Co of Zurich [2003] 1 WLR 1041). The same result must follow where the relevant prior determination appears in an award which has been subject to a jurisdictional challenge in part, but the determination relied upon is not in the challenged part of the award. Does it make any difference if the determination of an issue in an unchallenged part of the award is relied upon not in a subsequent arbitration or hearing, but in a s.67 challenge to a different part of the same award? I can see no reason of principle why it should, and that instinctive reaction is supported by authority.

In Emirates Trading Agency LLC v Sociedade de Fomento Industrial Private Ltd [2015] EWHC 1452 (Comm), the arbitral tribunal had delivered an award on jurisdiction, to which there had been no challenge, and then went on to deliver an award on the merits, which was the subject of a s.67 challenge. Popplewell J held that the unchallenged jurisdiction award gave rise to an issue estoppel in relation to the challenge before him. At [29] he recorded and rejected a submission that “a decision of a tribunal as to its own jurisdiction could never give rise to an issue estoppel” advanced by reference to Dallah, stating:

In Port de Djibouti v DP World Djibouti FZCO [2023] EWHC 1189 (Comm), three claims were advanced in a London-seated arbitration: the JVA termination claim, the Share Transfer claim and the Breaches claim. A s.67 challenge was brought in relation to the arbitral tribunal’s determination of all three challenges, it being said that there was no jurisdiction because the arbitral claimant (“PDSA”) had ceased to be a shareholder in the joint venture company at a certain point. Henshaw J held that the tribunal had jurisdiction over the Share Transfer claim, in the course of which determination the arbitral tribunal had found on the merits that PDSA’s shareholder status was maintained. In relation to the Breaches claim, one of the grounds on which the jurisdictional challenge was rejected was because of an issue estoppel arising from the Share Transfer claim. Against that background, Henshaw J considered the issue of whether the determination of the shareholder issue in the Share Transfer claim gave rise to an issue estoppel for the purposes of the jurisdictional challenge to the Breaches claim.

Henshaw J referred to two authorities which supported the conclusion that an issue estoppel arose. The first was Westland Helicopters Ltd v Al Hejailan [2004] 2 Lloyd’s Rep 523 in which there had been no challenge to an arbitrator’s decision to resolve the dispute by reference to a notional annual retainer, but there had been a challenge to the arbitrator’s jurisdiction to award interest by reference to a principal sum quantified by reference to that notional annual retainer. At [34] Colman J stated:

At [37], he continued:

The second decision was C v D1 [2015] EWHC 2126 (Comm), in which the claimant sought to challenge an arbitral tribunal’s conclusion that it had jurisdiction under the arbitration clause in a contract (the SPA) to determine claims arising from alleged breaches by the claimant of an earlier contract (the PSC). The tribunal decided that clause 26 of the SPA did confer jurisdiction to decide claims arising from alleged breaches of the PSC, and on the merits that clause 11.1 of the SPA required the claimant to indemnify the defendant in respect of certain claims, including those in relation to the PSC. The claimant contended that it could challenge the second issue to the extent relevant to its s.67 challenge to the first issue, arguing that the effect of s.67 of the 1996 had the effect that “no relevant issue is or can be res judicata or the subject of an issue estoppel” ([81]-[82]), Carr J rejected that submission ([83]), noting that the findings on the second issue were not susceptible to a jurisdictional challenge, with the result that the determination of the issues within the second issue were final and binding on the parties.

At [98]-[99], Henshaw J concluded:

I agree with the conclusions expressed in these four cases, which to my mind provide a complete answer to this preliminary objection. I also agree with Henshaw J that the application of the doctrine of issue estoppel in a s.67 challenge cannot turn on the procedural happenstance of whether the same tribunal delivers more than one partial award in a single reference, or determines all of the issues in the same award.

The Czech Republic’s second preliminary objection is that the legal system which governs the question of issue estoppel in this case is not English law but international law, or at least that the Court should “take into account the approach taken by international law to the question of issue estoppel”. This argument was developed relatively lightly on both sides, and I was not persuaded by it:

The materials which Mr Dunning KC relied upon to establish “the approach taken by international law to the question of issue estoppel” were to the effect that international law recognised no binding system of precedent (which appears to me to raise a different issue) and cases in which it has been suggested that it is only the provisions of the dispositif which can have preclusive effect as a matter of international law, and not the reasons for them. These comprised:

In response, the Arbitral Claimants referred to a number of investment treaty awards which have applied a concept of issue estoppel extending beyond the dispositif: Apotex Holdings v USA ICSID Case No ARB(AF)/12/1 Award of 25 August 2014 (Veeder, Rowley, Crook), [7.18]; Grynberg v Grenada ICSID Case No ARB/10/6, Award of 10 December 2010 (Rowley, Nottingham, Tercier), [7.64]; and Ampal v Egypt ICSID Case No ARB/12/11, Decision on Liability and Heads of Loss of 21 February 2017 (Fortier, McLachlan, Orrego Vicuna), [258]-[274].

I am not persuaded that the Czech Republic’s submissions take account of the extent to which the ITA regime has moved disputes under particular international law treaties into the procedural framework of municipal law, particularly where the arbitration takes place under an ordinary private arbitral process rather than the ICSID Convention. However, in view of the decision I have reached that English law governs the issue which arises before me, I do not need to determine the extent to which any principles of issue estoppel as applied by ITA tribunals differ from those arising under English law.

The effect of my conclusions in this and the March Judgment rejecting all of the available jurisdictional challenges to the Award in favour of Mr Stava is that there is a final determination of a tribunal of competent jurisdiction.

The only point made by the Czech Republic to dispute that conclusion is that I have given permission to appeal on one of the jurisdiction challenges I rejected which was raised in relation to Mr Stava. It is common ground that if that appeal succeeds, the plea of issue estoppel will fall away. However, the mere fact that the March Judgment is subject to an appeal does not deprive it of its final and binding quality for the purposes of the doctrine of issue estoppel: Spencer Bower and Handley: Res Judicata 6^th [5.19] and AD Atelier International BV v Manès [2020] EWHC 1014 (Comm), [2020] QB 971, [48].

Self-evidently, a finding on an issue as between Mr Stava and the Czech Republic in the BIT Arbitration does not involve the same parties as the issue now before me between Diag SE and the Czech Republic. However, the doctrine of issue estoppel under English law extends not only to the parties to the original determination, but their privies. The test of what constitutes privity for this purpose has been the subject of controversy. It is an issue which I had cause to consider at some length in PJSC National Bank Trust v Mints [2022] EWHC 871 (Comm). In circumstances in which there was no real argument on the scope of the concept at this hearing, I hope I will be forgiven for adopting the conclusions I reached in that case for the reasons I gave in section C2 of that judgment:

Mr Dunning KC argued that Diag SE could not be a privy of Mr Stava’s in relation to the determination in arbitral proceedings to which Diag SE was itself a party. I accept that, ordinarily, issues of privity are unlikely to arise in this context, because issues will be usually decided as between all parties to the proceedings. However, that may not be the case if the tribunal only finds it necessary to decide an issue as between two of the parties to a case without referring to a third. In such circumstances, I cannot see why the putative “privy” should fall outside the scope of the principle merely because they were a party to the case. If anything, the fact of that participation and the ability to influence the content of the determination would appear to make this a stronger case for privity. In the present case, the premise on which the issue estoppel plea is being approached is that the Tribunal’s findings in relation to the alleged privy are not binding or determinative because it did not have jurisdiction over the issue. To treat the alleged privy’s participation in such a hearing as precluding the operation of the privity principle is particularly unpersuasive.

In this case, I am satisfied that Diag SE was Mr Stava’s privy in relation to the findings in the Award:

There has been a degree of position-changing on this issue, but the “law of the case” (and my finding as to the position in the BIT Arbitration) is as follows.

First, the Czech Republic relied upon the events forming part of the Lawbook Transaction as an answer to the claims of both Mr Stava and Diag SE. In its Rejoinder:

Second, the Arbitral Claimants responded to both contentions which they characterised as a single objection affecting both Mr Stava and Diag SE. For example, Section E of the Rejoinder on Jurisdiction was headed “the Claimants are both investors within the meaning of the BIT”, and referred to the Czech Republic’s objection that “neither of the Claimants can bring a claim in respect of conduct which occurred after June 2011” (at [74]). In their First Post-Hearing Brief, the Arbitral Claimants once again addressed the submission under the heading “the Claimants are investors protected by the BIT” ([254]).

Third, in its first PHB, the Czech Republic maintained the objection to the claims of both Mr Stava and Diag SE. At page 60, it alleged that in June 2011 “both Claimants lost any status as qualifying investors”.

Fourth, the Tribunal referred to and determined both objections not (as the Arbitral Claimants and the Czech Republic have at various times alleged) only the objection concerning Diag SE:

It is against that background that it is necessary to evaluate the intermittent submissions of the Arbitral Claimants and the Czech Republic that, having identified that both objections were being taken, summarised both objections, summarised the answer to both and noted no timeliness objections were taken, the Tribunal contrived to decide only one of them. At [398] of the Award, the Tribunal found that Diag SE was Swiss for the purpose of the BIT. At [403] the Tribunal referred to the submission that after the Lawbook Transaction, “Mr Stava no longer owned or controlled Diag Human SE” – the premise of both the Stava June 2011 Objection and the Diag SE June 2011 Objection – and rejected it. At [408], the Tribunal found (in effect) that Mr Stava controlled the Koruna Trust and through the Koruna Trust Diag SE, and at [410] that after June 2011, Mr Stava retained control of Diag SE. At [412]-[413] the Tribunal stated:

The effect of the finding in [1] is that the Stava June 2011 Objection failed. The effect of the finding in [2] is that the Diag SE June 2011 Objection failed. The effect of [3] is to record that Objection No 5 which embraced both of those objections was dismissed. I regard the finding as “crystal clear”.

I have addressed this (already determined) issue at some length because of a palpable sense in the Arbitral Claimants’ submissions that my refusal to grant permission to appeal on this issue after the March Judgment (an issue which I still struggle to discern was clearly argued at that hearing) stemmed from an insufficient understanding of the facts, and that a further two week hearing and Mr Dunning KC’s brief espousal of the point might reveal to me the error of my ways. While it has suited the Arbitral Claimants and the Czech Republic at different times to argue that only the Diag SE June 2011 Objection was determined by the Tribunal, and the Stava June 2011 Objection was left in some form of arbitral limbo, I have to confess I regard the argument as hopeless on the face of the Award, even without having regard to the benign approach the English court adopts to the interpretation of arbitral awards on curial challenges. It is utterly unarguable once the principle of “benevolent reading of awards” is brought into play. It was for those reasons I dismissed the Arbitral Claimants’ application for permission to appeal to argue the contrary. Had the Czech Republic maintained its argument that the Stava June 2011 Objection was never decided, I would have rejected it with an equal lack of hesitation.

I should also deal at this point with Mr Dunning KC’s suggestion that the March Judgment had only addressed the status of the Tribunal’s finding of Mr Stava’s ownership of Diag SE after June 2011, and not his control. I can deal with that briefly:

Issue estoppel applies to issues which are “an essential step in the reasoning” of the original tribunal, or a “fundamental issue …. which formed the basis of the [determination]” (Spencer Bower & Handley, 6^th, [8.01]). Is this requirement satisfied here?

I can deal with this briefly. The allegation Mr Stava had lost control of Diag SE in June 2011 was raised by the Czech Republic as a complete answer to his claim both in the BIT Arbitration and in the s.67 application which the Czech Republic issued. The Arbitral Claimants did not suggest in either context that this was not the case (merely in court contending that the issue was one of standing, and hence not reviewable on the s.67 challenge, a contention which I accepted). Nor did the Tribunal challenge the assertion that, if the point was good, it provided an answer to Mr Stava’s claim in the BIT Arbitration. Rather the Tribunal found that the premise of the contention – that Mr Stava had lost control of Diag SE in June 2011 – was not established and the objection failed for that reason.

That accords with the view expressed by Professor Douglas KC ([198]-[199]). And it is not realistically open to the Czech Republic to contend otherwise.

It is well-established that “an issue estoppel only arises if an issue in the second proceedings is the same as one decided or covered in the first” (Spencer Bower & Handley 6^th, [8.19]).

At this point, it is helpful to refer again to the contested issue of whether Mr Stava should be granted permission to appeal on the Stava June 2011 Objection, but on this occasion to the Czech Republic’s benefit. The Arbitral Claimants submitted the effect of the control issue in relation to Diag SE’s status as a Swiss investor was that:

The underlined words are correct. That is consistent with:

The fact that determining the Stava June 2011 Objection may involve traversing the same factual enquiry (and it is not necessary to decide whether it would) is not enough: that would be a matter of evidentiary fact, which does not provide the basis for an issue estoppel.

It has been noted that the special circumstances exception “should be kept within narrow limits to avoid undermining the general rule and provoking increased litigation and uncertainty” (PJSC National Bank Trust v Mints [2022] EWHC 871 (Comm), [59] citing from Spencer Bower & Handley on Res Judicata – see now the sixth edition, [8.32]).

Only one matter was put forward as constituting a “special circumstance” in this case: the statutory right to a de novo challenge to the jurisdiction of an arbitral tribunal under s.67. I am unable to accept that this constitutes a special circumstance, either as a matter of principle or given the authorities at [228]-[234] above. In this context, any issue estoppel deriving from a finding which an arbitral tribunal had jurisdiction to make is one of the “panoply of tools” available to the court as it is in other contexts in which the court has to make determinations of fact and law, be that in a s.67 hearing, when proceedings are brought to enforce an award or any other hearing in which issues arise for final determination (cf. Cockerill J’s comment on the objection to issue estoppel being relied upon when the court was asked to rule on an issue of state immunity under the State Immunity Act 1978 in Hulley Enterprises Ltd v Russia [2023] EWHC 2704 (Comm), [29]). As Males J noted in Central Trading & Exports Ltd v Fioralba Shipping Co (The Kalisti) [2014] EWHC 2397 (Comm), [50]:

Given my decision that Mr Stava’s control of Diag SE raises different ultimate issues as between Mr Stava and the Czech Republic and as between Diag SE and the Czech Republic, I am not persuaded that it is abuse of process for the Czech Republic to exercise its s.67 right to challenge the Tribunal’s jurisdiction over Diag SE’s claim on this ground. As Lord Lowry noted in Shaw v Sloane [1982] NI 393, 397, “the entire corpus of authority in issue estoppel is based on the theory that it is not an abuse of process to relitigate a point where any of the three requirements of the doctrine is missing.” Mr Goodman suggested that the abuse arose from the fact that that there would be a binding finding entitled to recognition in this jurisdiction that Mr Stava controlled Diag SE after June 2011, and it would bring the administration of justice into disrepute if the court found that he did not. However, once it is recognised that the issues are different (even leaving aside the possibility that those issues may involve different legal tests) I do not accept that any disrepute would follow from the fact that one tribunal, on the evidence and arguments before it, reached one conclusion on one issue, and a different tribunal on different evidence and legal argument reached a different conclusion on a different issue.

In the March Judgement, I left open the issue of whether the Third Section 68(2) Challenge met the requirement of substantial injustice imposed by s.68(2) of the 1996 Act. That question arose from the fact that at the time of the Bojar Letter breach, Mr Stava did not personally own 100% of Conneco, such that if he was claiming for his own loss at that date, the issue arose (but was not determined) as to whether the award of damages to him should have reflected the extent of his interest. However, the Tribunal also found that there had been a breach – the 2014 Review breach – in 2014 when there was no suggestion that Mr Stava did not own 100% of Diag SE, and for which the Tribunal awarded the same loss, via different reasoning.

On the basis of the findings I have made, I am satisfied that the Czech Republic has suffered no substantial injustice from the Third Section 68(2) Challenge. However, it is necessary to allow for the fact that if the appeal against the March Judgment succeeds, or other appeals are brought and succeed, Mr Stava’s ability to bring a claim in respect of the 2014 Review breach might come into issue. The order will need to preserve the Czech Republic’s ability to resurrect the Third Section 68(2) Challenge if successful appeals have the effect that the Czech Republic is able successfully to challenge any award in Mr Stava’s favour in respect of the 2014 Review breach under s.67. I will ask the parties to consider the terms of an order to address this issue.

For these reasons, the Czech Republic’s surviving challenges to the Award under s.67 of the 1996 Act are dismissed, and, subject to [263] above, the third Section 68(2) Challenge fails.

I would like to conclude this judgment by thanking both legal teams for their Herculean efforts in what has been a very hard-fought dispute. The court has benefited from thoroughly researched, cleverly conceived and highly skilled advocacy on both sides, which did full justice to the interests of the protagonists.

On 13 May 1991, Conneco entered a Co-operation Agreement dated 10 May 1991 with the District Institute of National Health (District Hospital) – Transfusion Centre Mĕlník (“Transfusion Service Mĕlník”). The agreement provided for a minimum supply of 800 litres of FFP in each 12-month period in return for blood derivative products or consumables and equipment. The schedule included a range of equipment which the Transfusion Centre could order at agreed prices. The agreement had an initial duration of two years but could be extended by the agreement of the parties “no later than within a period of 6 months prior to the expiry of the term.” I accept that the expectation at the time of entry was that this would not be a one-off transaction, but a rolling contract which would be renewed if performance proved satisfactory for both parties.

An addendum to that agreement was signed on 7 October 1991 identifying equipment which Conneco would supply up front, at a value of CZK 303,000, and recording Conneco’s agreement to supply further equipment up to CZK 1,300,000 in value. The supplied equipment was to be paid for “over the first two years of the validity” of the Co-operation Agreement through supplies of FFP, or, in default, in cash.

On 21 January 1992, Addendum No 2 to the Co-operation Agreement was signed providing for interest to be paid to Conneco at 18% on goods supplied other than by Conneco.

On 9 March 1992, Conneco entered into “Agreement No 3/92” with “Hospital Polyclinic Mĕlník”. This had the same address as the entity which entered into the first two agreements but a different company number and signatory. Its relationship with the earlier agreements is not clear, although it is fair to say that the Arbitral Claimants’ position at the opening of this hearing was that it did replace the 1991 agreement. It involved the same basic structure but provided for a minimum total supply of FFP of 1,200 litres over 12 months. The agreement had a 12-month duration, with extension being subject to the parties’ agreement.

On 1 November 1992, Conneco and Hospital Polyclinic Mĕlník entered into a termination agreement. This records that various items of equipment were supplied by Conneco at a total price of CKS 1,335,252, and 415 litres of FFP supplied to Conneco at a price of CKS 747,630. It was noted that “the Hospital with Polyclinic does not currently have the alternative to settle the deliveries of the equipment from Conneco as other than by the FFP supply”. I am satisfied that this reflects the position of recipients of equipment from Conneco generally, and that this was a reason why Conneco had a legitimate expectation that the co-operation agreements would be renewed by its counterparties. That assessment is consistent with the evidence of Colonel Milos Bohonĕk as to his expectation that the commercial relationship would have continued but for the Bojar Letter.

On 13 May 1991, Conneco entered into a Co-operation Agreement with Frýdek Místek Hospital adopting essentially the same structure. The minimum total quantity of FFP to be supplied was 1,000 litres in each 12-month period, and the agreement had a 2-year duration and provision for a negotiated extension.

Addendum No 1, signed on 12 August 1991, showed extensive supplies of equipment being made to Frýdek Místek Hospital before FFP supplies would begin. As with the Transfusion Service Mĕlník Co-operation Agreement, the recipients needed to receive equipment from Conneco before being able to deliver the FFP to pay for it.

On 1 January 1992, Conneco entered into a replacement agreement with Frýdek Místek Hospital dated 1 January 1992 . This revised the minimum supply to 2,000 litres over the 12-month term of the agreement.

A letter from Dr Miroslav Malchar, Director of the Hospital, to the Czech Ministry of Health dated 24 March 1992 stated that under the agreement, Conneco selected from frozen plasma directly in the Hospital, referring to selections being made on 4 November 1991, 28 November 1991, 5 January 1992, 10 February 1992 and 4 March 1992. Sixteen crates had been removed, storage being Conneco’s responsibility.

A letter of 7 June 1992 records an agreement to withdraw from the contract, with an unpaid amount of CKS 185,668.60 still due to Conneco.

On 30 May 1991, Conneco entered into a contract with Hospital and Polyclinic Uherské Hradištĕ in similar form to these entered into with Transfusion Centre Mĕlník and Hospital Frýdek Místek. The minimum supply was to be 800 litres of FFP per 12 months, with the contract having a 2-year duration, and the parties being able to extend it with 6 months to go.

On 2 January 1992, this was replaced by a one-year agreement with a minimum quantity of 2,000 litres of FFP for each year the agreement was in operation. The agreement had a minimum duration of 1 year. There are invoices showing equipment ordered by Conneco for delivery to the Polyclinic (5 February 1992). An invoice of 5 March 1992 shows delivery of blood derivatives to the Polyclinic. There was some delivery of FFP by the Polyclinic to Conneco – for example a test on FFP is dated 10 March 1992 countersigned by a doctor at the Polyclinic.

There is an unsigned Co-operation Agreement between Conneco and Olomouc Hospital in similar terms to the other 1991 Co-operation Agreements. There is also a signed document described as Amendment No 1 to a Co-operation Agreement identifying equipment which Conneco was to provide, in very similar terms to the amendments to the 1991 Co-operation Agreements with Frýdek Místek and Transfusion Service Mĕlník. There are also invoices for the supply of additional equipment to Hospital Olomouc beyond that appearing in Addendum No 1. I am satisfied that it is more likely than not that a Co-operation Agreement was concluded between Conneco and Olomouc Hospital.

A document dated December 1991 records the collection of 198.564 litres of FPP from “Ostrava – Blood Donation Centre, Poruba”, with a welding clamp of CZK 67,000 still to be paid for. This was clearly an “FFP for medical equipment” arrangement.

There is an unsigned Co-operation Agreement between Conneco and Hospital Polyclinic Kolin providing for a minimum delivery of 2,300 litres of FFP during each 12-month period of a renewable 2-year agreement. Given the letter from the identified signatory of that draft agreement, Dr Ludek Rubas, to Dr Bojar of 8 May 1991 stating that they had interrupted their negotiations with Conneco, I am not persuaded that a Co-operation Agreement was concluded with Hospital Polyclinic Kolin.

In addition, there is evidence of supplies of equipment by Conneco to other CFSR health institutions at Cheb, Prerov, Kladno, Caslav, Havlickuv Brod, Ceske Budejovice, Boleslav and Benesov u Prahy. I do not have sufficient information to know the arrangements under which these supplies were made. However, given the evidence as to the basis on which Conneco dealt with CFSR health institutions generally, and the evidence that the only means by which the institutions could have funded these acquisitions was through the supply of FFP, I am satisfied that a substantial part of these supplies would have been under similar arrangements to those in the Co-operation Agreements, albeit without any evidence of a binding framework contract of a specified duration.

The evidence establishes that 2,500 litres of FFP was shipped to Novo Nordisk (a figure confirmed by contemporary correspondence from Novo Nordisk and a Czech secret service report). There is no information as to whether there had been FFP deliveries from hospitals to Conneco which had yet to be exported. However, this seems likely:

Further, the very fact that FFP had been collected from Czech hospitals by Conneco, that FFP had been shipped to Denmark and that at least some blood derivatives were supplied by Conneco to a Czech health institution shows that Conneco had been able to equip the hospitals to collect and store FFP, and that it had the infrastructure to collect FFP from contributing hospitals, store it and ship it to Denmark and the ability to delivery blood derivatives to Czech health institutions.

The point in time at which some form of co-operation began between Diag Human AG and Novo Nordisk (or one of its constituent companies) is not clear. However, on 3 October 1988 an agreement was signed for the supply by Diag Human AG to Novo Nordisk of 48,000 litres of FFP from the Bavarian Red Cross in 1990.

From 1990, there are minutes of regular meetings in which Diag Human AG and Novo Nordisk discuss business in Eastern Europe, and the terms of a distribution agreement between them in relation to that area. Frequent reference was made to business in Poland, the CFSR and Yugoslavia, as well as the ongoing Bavarian Red Cross arrangement, and there are also occasional references to business opportunities in Switzerland and Italy. Drafts of distributions agreement were discussed and exchanged in 1990.

It is clear that there was performance under the Bavarian Red Cross arrangement in 1990 – for example a meeting note of 9 July 1990 refers to 20,000 bottles of albumin being ready for delivery under that contract. The same meeting also refers to the CFSR, which it is anticipated would be the source of 12,000 litres of FFP, and records “the condition for the cooperation are virtually agreed”.

Regular meetings continued in 1991, with discussion of a written co-operation agreement continuing (e.g. at a meeting on 14 May 1991 when the contents and duration of the proposed co-operation agreement were outlined). At the same meeting, there was specific discussion of the possibility of procurement of FFP from CFSR, in an estimated quantity of up to 24,000 litres a year, and a statement that a draft co-operation agreement between Conneco and Novo Nordisk would be worked out.

In June 1991, Novo Nordisk submitted a bid in the tender process run by the Ministry of Health in which it referred to Conneco as “our partner in Czechoslovakia”. That does suggest that there was some form of business co-operation relationship between them. The bid answered a question as to the processing capacity of Novo Nordisk by stating that “Novo Nordisk has capacity to fractionate from 10 tons to maximum 100 tons supplied from Czech Republic” and proposed a volume of 5000 litres of FFP.

At a meeting on 14 August 1991, it was noted that the contractual proposal for a distribution agreement covering Poland, CFSR and Yugoslavia had been accepted by both parties, but that the contract would only be signed when outstanding amounts arising from supplies to Yugoslavia had been resolved. It was stated that “day-to-day cooperation shall take place in accordance with the draft contract”. It was noted that Conneco had not succeeded in the CFSR tender process, but this was not regarded as bringing an end to business there. There was reference to the fact that “Diag Human and Conneco are now concentrating on the effective utilisation on the newly founded plasmapheresis facility in Brno”, with it being anticipated that 3,000 litres of CFSR plasma would be delivered to Novo Nordisk in 1991.

On 14 October 1991, Conneco applied to the Customs Office at Prague for a permit to export FFP from the CFSR to Novo Nordisk for processing there and to re-import derivatives into the CFSR. The terms on which this could be done were approved by the Customs Office on 30 October 1991. The permit referred to an anticipated value of 14 million Czech crowns per year. At this stage, therefore, some form of business co-operation relationship was continuing between Conneco (as the Czech subsidiary of Diag Human AG) and Novo Nordisk in relation to the processing of Czech FFP. An invoice for a shipment of FFP to Novo Nordisk was issued by Conneco on 5 March 1992, and a quality certificate for that shipment on 10 March 1992.

On 9 March 1992, the Bojar Letter was sent by the Czech Minister of Health to Novo Nordisk, stating that one of the reasons its participation in the 1991 tender had failed was “a doubt about respectability of the Conneco”. The letter concluded “we expect that you will draw appropriate conclusions from the problem mentioned above …” The implications of that last sentence were not lost on Novo Nordisk. On 18 March 1992, Mr Knud Daugaard of Novo Nordisk wrote to Mr Kubálek of Conneco stating:

Once again, the letter acknowledges the existence of a relationship of business co-operation between Novo Nordisk and Conneco at that point, which the Bojar Letter had caused Novo Nordisk to put on hold.

On 23 March 1992, Mr Klaus Eldrup-Jørgensen of Novo Nordisk wrote to Dr Bojar stating that the letter had “received our careful and urgent attention” (something readily understandable given the content and source of the Bojar Letter). He continued:

It is, in my view, significant that in a context in which it would have been in Novo Nordisk’s interests to downplay the nature and extent of any dealings with Conneco, or even to deny any relationship with them, the letter did no such thing. It referred to Novo Nordisk being “active” in the CFSR “through” Conneco.

On 8 May 1992, Dr Gnadinger of Diag Human AG and Mr Eldrup-Jørgensen, Mr Thaning Larsen and Mr Habo of Novo Nordisk met in Gentofte in Denmark. Novo Nordisk was told that the Ministry of Health had agreed to send a corrective letter, but none had been received. The meeting notes that Novo Nordisk had set up a meeting with the Minister of Health which was scheduled for 18 May. They also state:

On 18 May 1992, Mr Eldrup-Jørgensen and Mr Larsen had a meeting in Prague with the CFSR Minister and Deputy Minister of Health and two others government officials. The official minutes are singularly uninformative as to what as discussed. However, Mr Eldrup-Jørgensen wrote to Dr Gnadinger on 26 May stating that Novo Nordisk had decided not to accept supplies of plasma from the CFSR until the position had been clarified, stating “Novo Nordisk has not and do not want to be involved in matters between Conneco and the CFSR Ministry of Heath” and “we must ask you to abstain from using the name of Novo Nordisk in your dealings with the CFSR Ministry of Health”. It is entirely understandable that Novo Nordisk, faced with a direct communication from a Minister of the CFSR with a barely veiled threat, reacted in a defensive manner.

On 18 August 1992, Mr Eldrup-Jørgensen wrote to Conneco stating:

On 16 September 1992, with (on his evidence) the input of Mr Jensen of Novo Nordisk’s legal department, an agreement was signed agreeing the outstanding balance as between Diag Human AG and Novo Nordisk. The agreement states that it was reached “following discussions in Copenhagen between Novo Nordisk being represented by Klaus Eldrup-Jørgensen, Torben Thaning Larsen and Diag Human represented by M Gnadinger”. The agreement also stated:

The communication clearly contemplated that Diag Human AG and its trading partners (including Conneco) had had some responsibility for marketing Novo Nordisk’s products in Eastern Europe (including the CFSR) and that this relationship was being brought to an end.

I have found the exchanges up to 16 September 1992 the best guide to the nature of Conneco’s relationship with Novo Nordisk at the time of the Bojar Letter:

In 1995, Novo Nordisk sold its plasma business to the US company HemaSure. Evidence given by Mr Jensen to the Czech Parliamentary Investigative Committee in June 2004 suggested that HemaSure took over all documents relating to the production of blood derivatives, and that the majority of the employees of this part of Novo Nordisk’s business was transferred across. HemaSure itself appears to have become insolvent a relatively short period thereafter.

Novo Nordisk no doubt hoped that the September 1992 agreement marked the end of any involvement in the dispute between Mr Stava and the Czech Republic. However, in in the early 2000s, two former employees of HemaSure provided witness statements to Conneco for the purpose of the Commercial Arbitration against the Czech Republic. Two statements were provided by Soren Bogo Jørgensen (who had been director of Q&A in the biopharmaceuticals division at Novo Nordisk and then HemaSure). In his second statement, he stated that “the sole reason why Novo Nordisk A/S in 1992 stopped to accept plasma from CFSR” was the Bojar Letter and the subsequent meeting between Novo Nordisk and the CFSR officials. The other statement was from Claus Bildsoe Astrup (formerly Jensen) who had joined Novo Nordisk in 1995 as a marketing manager in the plasma product unit, and who gave hearsay evidence of statements by various colleagues.

On 10 June 2003, the Czech Chamber of Deputies resolved to set up an inquiry into the settlement of the dispute between Diag Human and the Czech Republic. On 9 April 2004, the secretary of the Investigative Commission met Mr Binter of Novo Nordisk in Prague to seek the company’s assistance. On 16 April 2004, the Commission sent a set of 8 questions through to Novo Nordisk. On 11 May 2004, Novo Nordisk agreed to co-operate and promised to contact former employees and obtain the required information.

It is likely that the emergence of this ghost from the past was unwelcome news for Novo Nordisk, who put the matter in the hands of Mr Anders Jensen, its inhouse lawyer. The request came over 10 years after the relevant events and 9 years after the sale of the relevant business unit. I am also satisfied that Novo Nordisk was keen, so far as possible, to distance itself from the ongoing dispute, and to minimise its role.

Mr Anders Jensen attended a meeting with two members of the Commission in the Czech Embassy in Copenhagen on 31 May 2004. In the course of that meeting, Mr Jensen appears to have handed over a document with answers to the 8 questions which he had signed. While Mr Jensen thought at trial that the document had been prepared and signed at the 31 May 2004 meeting, I am satisfied that it was prepared either by Mr Eldrup-Jørgensen or Mr Thaning Larsen before the meeting:

Mr Jensen attended a further meeting with the Commission at the Czech Embassy on 28 June 2004. The minutes on this occasion record the presence of an interpreter, and the written minutes in English were not produced until 15 December 2004. It is not clear how they came to be sent to Mr Jennsen for him to sign, but it must have occurred some time after the meeting.

Mr Jensen began by seeking to distance Novo Nordisk from the statements made by Soren Bogo Jørgensen and Claus Bildsoe Astrup (formerly Jensen), stating:

Mr Jensen confirmed that he had searched those records which Novo Nordisk still had after the sale of the business and produced three agreements dating back to 1991. He was careful to confirm that he could not say that these were the only agreements ever concluded. Mr Jensen had no direct involvement in most of the events: he was very clear in his evidence to the court that he was “not involved in business” and, beyond helping Mr Eldrup-Jørgensen to buy his furniture when he became head of the blood plasma unit, he was “not involved in how he ran the business”. His statements to the Commission must have reflected what he had been told or concluded, and the underlying basis for his statements is not clear. His statements of the extent of Novo Nordisk’s dealings with Diag Human understated the position, probably because the true position could no longer be ascertained from the former employees he had spoken to:

In his 2004 interview, and in his evidence at this trial, Mr Jensen was rather dismissive of the aspirations which Mr Eldrup-Jørgensen had to expand the blood plasma business. However, I am not persuaded that he was privy to the plans of the blood plasma division in Novo Nordisk in 1992, and his statements in 2004 are likely to have been coloured by hindsight. Nor am I persuaded that he was privy to the production capacity of the blood plasma unit, and I am satisfied that Mr Eldrup-Jørgensen (who Mr Jensen confirmed was a serious businessman) would not have made statements about Novo Nordisk’s capacity in the June 1991 tender documents unless they were accurate. In any event, even if Mr Jensen is right that Novo Nordisk was looking to dispose of the blood plasma unit from the early 1990s, that does not mean that it wanted to prevent the blood plasma unit developing its business (something which may well have facilitated such a disposal). The participation of the Novo Nordisk blood plasma unit in the 1991 CFSR tender process is significant in this respect.

In short, none of the evidence given once the Commercial Arbitration began – whether of ex-Novo Nordisk employees in support of Conneco at the arbitration, or the statements made by Mr Jensen either in 2004 or at this trial – are of real assistance. The best evidence of the position in 1992 is to be bound in the documented dealings in and around that period, the effect of which I have summarised above.

In the course of 2011, two connected events took place:

Lawbook Limited has its registered seat in the Irish township of Ballylickey in Bantry, Co Cork. ARAM Global describes itself as a company founded in 2009 by Raja Visweswaran and Michael Balboa, based around a group of 15 experienced financiers and investment managers who specialise in debt recovery.

The Lawbook Transaction comprised:

On 27 May 2011, Diag SE retained ARAM International Partners to represent it in negotiations relating to the Final Award in the Commercial Arbitration in return for 12% of the cash proceeds, with Diag SE meeting recovery costs (which made the Lawbook Transaction an even less rational business proposition).

On 1 June 2011, Diag SE issued a press release reporting that Mr Stava had transferred all decision-making power in Diag SE with immediate effect, as “part of an agreement through which Mr Stava has sold Diag Human to an international consortium of institutional investors”. On 7 June 2011, Mr Stava gave an interview repeating the suggestion he had sold Diag SE, and that “the Czech government already knows perfectly well who the new owner of Diag Human is”. On 10 June, another press report appeared referring to “the anonymous new owner of Diag Human” who had hired ARAM Global, and to a press statement by Mr Stava that he had sold Diag SE to a consortium of international investors. The report stated that neither Mr Stava nor ARAM Global would reveal the identity of the buyer.

On 8 June 2011, Mr Visweswaran of ARAM Global wrote to the Prime Minister of the Czech Republic referring to “the recent change of ownership” and stating ARAM had been authorised to negotiate on the new owner’s behalf.

On either 10 or 20 June 2011, the Ministry of Health replied to this letter, indicating that the legal dispute between the Czech Republic and Diag SE was ongoing, and they could see no “evident reason” for talks at that time. It also raised Diag SE’s ongoing enforcement efforts in relation to the Commercial Arbitration Award. At no stage did the Ministry of Health ask any questions about the new owner of Diag SE, even though that would have been an obvious point of interest.

By 21 June 2011, Mr Balboa of ARAM Global had been appointed to the position of Diag SE’s President and to its board of directors, as recorded in the Liechtenstein Public Register.

On 20 July 2011, ARAM Global sent a further letter enclosing an authorisation to act for Diag SE signed by Mr Balboa and Dr Rabanser, and a copy of the Liechtenstein Public Register. The Ministry of Health responded in negative terms to the request for negotiations.

On 12 October 2011, the Koruna Trust and the Trustee’s trusteeship were formally registered in Liechtenstein.

On 2 November 2011, Mr Balboa swore an affidavit in English enforcement proceedings commenced by Diag SE, stating ARAM Global had been retained by Diag SE to provide recovery advice.

On 27 November 2011, Mr Stava accepted appointment as the Protector of the Koruna Trust.

On 1 December 2011, the United States Attorney’s Office for the Southern District of New York announced that Mr Balboa had been charged with fraudulently overvaluing hedge fund assets.

On 2 December 2011, minutes of a meeting of Diag SE’s shareholders record Mr Stava holding 25 bearer shares in the company, and voting to remove Mr Balboa from the position of President and Director of Diag SE.

On 6 December 2011, Mr Stava wrote to Lawbook Ltd referring to the non-payment of the first two instalments of the price and stating that the 31 December deadline for payment of the full price would not be extended.

On 15 December 2011, Diag SE terminated its engagement letter with ARAM SE for material breach and revoked the authority of ARAM Global.

The Lawbook Transaction expired on 31 December 2011.

It was Mr Stava’s evidence in the BIT Arbitration and before me that the Lawbook Transaction was entered into because in early 2011, the Czech Minister of Finance, with the knowledge of the Czech Prime Minister, suggested that if Mr Stava distanced himself from Diag SE, it would facilitate the settlement of their longstanding dispute. In effect, it was intended to create the illusion that he had severed his connection with Diag SE to facilitate a settlement. That account was challenged by the Czech Republic in the BIT Arbitration and, with less enthusiasm, before me as an untruthful fabrication on Mr Stava’s part. The submission made to me was that:

I am satisfied that Mr Stava’s has been substantially truthful in his evidence that the Lawbook Transaction represented an attempt to create apparent distance between himself and Diag SE following indications from high levels within the Czech Government that this would be helpful to any attempt to resolve the dispute. I have reached this conclusion in reliance on the following matters:

By contrast, there is no dispute that the Koruna Trust created a valid trust which was intended to and did take effect in accordance with its terms and the applicable principles of Liechtenstein law. There was some debate about why the Koruna Trust was established. The one consistent – but not at all times the only – explanation offered by or on behalf of Mr Stava was “succession planning”. I did not find that a particularly informative explanation of why a discretionary trust was set up whose sole asset was shares in a non-trading company whose sole business was the attempt to enforce a commercial arbitration award against the Czech Republic. I accept, however, that Mr Stava had benefiting his family in mind when he established the trust, as evidenced by the Class of Beneficiaries (see [23(iii)] below). I also accept that “asset protection” may have been a reason why this asset – which by this point had had a controversial and troubled history, in which the resources of the Czech state had been deployed to a striking degree in an attempt to prevent the realisation of any value – appears in a discretionary trust essentially on its own. In neither respect am I persuaded that the purpose of the trust imposed a meaningful limit on the Trustee’s responsibilities outside the context of fixing the members of the Class of Beneficiaries and in the distribution of Trust assets.

Finally, I accept that it is likely that there is some written record of Mr Stava’s intentions in relation to the distribution of the Koruna Trust, and that this has not been produced despite disclosure orders by this court. However, I am not persuaded this would have added meaningfully to the obvious fact that Mr Stava intended the Stava family to benefit.

The terms of the Koruna Trust were set out in the Trust Deed, which was subject to an express choice of Liechtenstein law and the jurisdiction of the Liechtenstein courts (clause 2). The Koruna Trust was declared to be irrevocable (clause 30).

The Trust Deed identified:

The Trust Deed relieved the Trustee of various default obligations arising under Liechtenstein trust law (including keeping records) (clause 2.2) and accorded the Trustee wide powers, including:

Clause 24 provided that “the Trustee shall not be required to interfere in the management or conduct of the business of any Company, securities of which comprise the whole or part of the Trust Fund”.

So far as the members of the Class of Beneficiaries are concerned:

Clause 26 permits any Trustee to acquire assets from the Koruna Trust or dispose of assets to the Koruna Trust “without being liable to account for any profit and without the transaction being void provided any transaction is effected at a market price or a price certified as fully commercial” by qualified professionals.

Liechtenstein is a classic, continental European civil jurisdiction in which laws enacted by Parliament are the central sources of law. The principal statute relevant to issues in this case is the Persons and Companies Law (“the PGR”) which incorporates the Trust Enterprise Law (“TrUG”). The General Civil Code (“ABGB”) and the Property Law (“SR”) are also relevant to some extent.

The following matters of Liechtenstein trust law were common ground between the Liechtenstein law experts:

I will now briefly consider the matters on which the experts were not agreed.

First, there is a dispute as to what happens if a trustee enters into a mandate with the settlor and binds himself to follow the settlor’s instructions: Mr Reithner saying there is no trust in such a scenario, and Dr Batliner saying the trustee could bind himself in this way, although this would not relieve him from the obligation to comply with his duties as trustee. I am not persuaded on the evidence that any binding agreement of this kind was entered into – there is no documentary evidence to support it, Mr Stava’s evidence was at best equivocal as to whether there was a mutual understanding to this effect with Dr Rabanser or that was merely his own understanding, and Dr Rabanser did not give evidence, from which I have inferred that he would not have supported the suggestion he had bound the Trustee in this way. Accordingly, it is not necessary to decide this point.

Second, there is an issue as to whether a protector who is also a beneficiary could exercise the power to remove the existing trustee and/or appoint himself as trustee and distribute trust property to himself. As to this:

Third, whether clause 26.1 of the Koruna Trust Deed and Article 925 of the PGR would prevent a trustee who is also a beneficiary from making any distribution to himself. As to this:

There is one further submission which I should address, namely that Mr Stava owned the shares in Diag SE because he held the bearer shares at the end of 2011. In closing, Mr Riches KC accepted that “it’s not how the case has been put” but said that the submission was “based on the fact that Mr Reithner accepted in his evidence that if you hold the bearer shares as a matter of Liechtenstein law, that makes you the owner. Now I accept that’s not the way we pleaded it”. The background to this issue is now set out.

The Arbitral Claimants in their pleading pleaded that Mr Stava was the legal and beneficial owner of the entire share capital in Diag SE until 27 May 2011 when legal title in the shares was transferred to Lawbook Limited, but that Mr Stava continued to control Diag SE, doing so “as holder of all of the bearer shares.” It was expressly accepted, therefore, that Mr Stava’s ownership of the shares in Diag SE ended on 27 May 2011.

The Arbitral Claimants also served a pleaded case on Liechtenstein law in which they made the following admissions (which they have not sought or obtained the court’s permission to be released from):

In support of the argument the Arbitral Claimants sought to advance in closing that ownership of the shares was in fact vested in Mr Stava, reliance was placed on Mr Reithner’s first report. This had stated “when the shares became trust property, the Trustees were obliged to take control of them … They were (and are) the owners of the shares”. Mr Reithner also referred to Mr Stava taking “possession of the share certificates on behalf of the Trustee” and acting “as an agent of the Trustee”. However, in a section addressing events in 2014, and the effect of a 2014 document which, on one view, purported to transfer voting rights separate from the shares, Mr Reithner stated:

Dr Batliner’s report stated that the Koruna Trust became the owner of the shares as from 1 January 2012. He noted the evidence that the bearer shares were in the physical possession of Mr Stava from 27 May 2011 to 28 November 2013 and that there was no evidence as to what agreements were in place to deal with this state of affairs. However, he stated that Mr Stava was only the shareholder until 27 May 2011, with the Koruna Trust “holding the share certificates as pledgee from 27 May until 31 December 2011” and becoming “the owner of the shares as of 1 January 2012”, suggesting he had to assume that Mr Stava held the certificates “by mutual agreement”, on the basis of “an unlimited power of attorney”. He expressly stated that this did not involve a transfer of the shares out of the Koruna Trust.

The experts’ Joint Memorandum referred to Mr Stava acting as an agent of or custodian for the Koruna Trust when holding and voting the shares.

In the Arbitral Claimants’ opening, reference was made to the quoted paragraph of Mr Reithner’s report in support of the submission that Mr Stava “was the only person entitled to exercise the voting rights in respect of the … shares”, but “even if Mr Reithner was wrong about this and the bearer shares were owned by the Koruna Trust while held by Mr Stava”, Mr Stava had an unlimited power of attorney. A later paragraph stated, “as noted above, at least when voting with possession of the bearer shares, the Claimants’ expert evidence is that Mr Stava was as a matter of Liechtenstein law treated as owner” (emphasis added). Finally, in opening submissions, Mr Riches KC finished with a “final general observation” in which he stated, “it is not disputed that the trustee had some form of de jure control, subject to the undisputed Liechtenstein law evidence that the holder of the bearer shares is the owner”.

With the benefit of hindsight, it is possible to see in these fleeting references an early attempt subtly to embrace the new ownership case, but there was nothing to suggest that the Arbitral Claimants were seeking to depart from the extensive admissions they had made, still less seeking the requisite permission to do so. As will be apparent from my summary, there was no “undisputed Liechtenstein law evidence” that Mr Stava’s physical possession of the shares made him the owner to the extent that he was holding the shares (as both experts accepted he was) as agent or custodian for the Koruna Trust.

The order of the expert witnesses was reversed because of availability issues, Dr Batliner going first. In the course of his cross-examination Dr Batliner confirmed the following:

Mr Riches KC raised the point very briefly as his sole point in re-examination (and he did not cross-examine Mr Reithner about this topic):

The point made by Dr Batliner is one which any English lawyer will immediately recognise. The postman or courier delivering bearer shares does not, by virtue of their possession, became the share owner, however briefly, nor the safety deposit box company where bearer share certificates are stored. Nor, closer to home, did the custodians with whom bearer shares in Liechtenstein companies have had to be deposited since 1 March 2013 and who are obliged to maintain a register, giving the names and details of the shareholders. This is because there is no basis for the transfer of the shares, even if there has been a handing over (with the result that the custodian is not holding them in their own right). So too of Mr Stava from 27 May 2011 who held the shares as agent or custodian for their owner, the Trustee on behalf of the Koruna Trust. That is the consistent effect of the evidence of Mr Reithner and Dr Batliner, as I have sought to show, and it is fatal to the Arbitral Claimants’ attempts to depart from their pleaded case and argue that Mr Stava was the owner of the shares on 1 January 2012. That is sufficient to dispose of the point, but I would in any event have held that it was not open to the Arbitral Claimants to advance it, against the procedural background I have outlined.

It is important to note the following matters:

When settled into the Koruna Trust, therefore, as the Czech Republic stated, Diag SE “had no business other than seeking to recover under its commercial arbitration award.” I accept that it was Mr Stava who was in de facto control of that process – he was the only person with first hand knowledge of the relevant events, not LNR Trust Reg nor Dr Rabanser. It was Mr Stava who was the sole source of funding for the litigation: this was stated in Mr Stava’s witness statement in the English enforcement proceedings in response to the Czech Republic’s application for security for costs; and it was noted by Mr Justice Burton in [2013] EWHC 3190 (Comm), at [42(i)]. When the Czech Republic later sought a third-party costs order against Mr Stava on the basis that he was “Diag SE’s owner and controller” – and I should interject there that I place no reliance on that particular statement – it stated that Mr Stava “was directly responsible for the instigation of this litigation. He was the sole funder of the litigation for [Diag SE]” (Mr Bridson’s eleventh witness statement of 6 November 2015). Mr Bridson also referred to confirmation from two firms of solicitors who had acted for Diag SE in English proceedings that Mr Stava was the sole funder and asserted it was clear that those solicitors received instructions from Mr Stava. Mr Bridson’s description of Mr Stava as the guiding hand and financial force behind Diag SE was accurate.

I accept, therefore, that to all intents and purposes Mr Stava had practical control of the pursuit of Diag SE’s only assets, whose pursuit was dependent on funding by him. In seeking to pursue and realise those assets, there was no scope for any conflict between Mr Stava and the Trustee (the interests of Mr Stava and the Koruna Trust which he had had established, of which he, his daughters and their children were in the Class of Beneficiaries, and into which he had settled Diag SE were perfectly aligned in that respect). Nor was there any practical scope for the Trustee to second-guess Mr Stava’s decisions in the attempt to realise that asset. These matters were peculiarly within Mr Stava’s knowledge and were being funded by assets outside the Koruna Trust.

Clauses 19.2, 19.5 and 24 of the Trust Deed permitted the Trustee to act in a way which reflected that alignment of interest, and Mr Stava’s greater knowledge of and ability to pursue the exploitation of Diag SE’s only asset. The Trustee was also entitled (but not obliged) to have regard to the wishes expressed by Mr Stava as settlor, protector and a member of the Class of Beneficiaries and on issues relating to the pursuit of legal claims open to Diag SE, I am sure the Trustee would have done so. Deferring to Mr Stava on matters relating the pursuit and control of Diag SE’s claim would inevitably comply with the “business judgment rule” and, as I have stated, there was no meaningful possibility of the Trustee forming and seeking to act on a different view to Mr Stava’s as to how to enforce Diag SE’s claims.

However, Mr Stava was not exercising legal powers of control held by him in his own right and those powers could not be exercised exclusively in his own interests. The immediate source of those powers was Diag SE itself.

Mr Stava was a member of the board and President of Diag SE and held sole signatory rights. Those rights were held on behalf of Diag SE, and fell to be exercised in the best interests of the company, and in particular its sole shareholder, the Koruna Trust.

As the holder of the bearer share certificates in Diag SE at the end of January 2012, Mr Stava had the legal right to exercise the voting rights which attached to those shares in what he believed to be the best interests of the Koruna Trust. The exercise of those voting rights was determinative of decision-making within Diag SE. The Czech Republic also accepted that, independently of the voting rights, Mr Stava was able to influence Ms Dagmar Stava and Dr Rabanser (two other board members).

I accept that the combined effect of the very limited nature of Diag SE’s business, the particular dependence of Diag SE on Mr Stava if it was to seek pursue those rights, and the absence of any realistic scope for a conflict between Mr Stava’s personal interests and those of Diag SE and the Koruna Trust in seeking to enforce Diag SE’s rights, was that Mr Stava was the determinant voice in all decisions taken by Diag SE.

For the reasons I have set out, I am satisfied that the Trustee had effectively delegated the conduct of Diag SE’s sole business – the exploitation of its rights under the Commercial Arbitration award and any associated rights – to Mr Stava, and that this was a perfectly proper course.

The Trustee of the Koruna Trust had the legal right to sell the shares in Diag SE, where the Trustee formed the good faith view that such a sale was in the business interests of the Koruna Trust and represented a prudent business judgement. However, it is virtually inconceivable that the Trustee would have concluded that the best interests of the Koruna Trust, and prudent business judgement, justified such a sale if Mr Stava was opposed to it. That reflected the fact that Mr Stava was uniquely placed to assess the value of Diag SE and how best to realise that value. In addition, it is difficult to see how such a sale would have been a viable business proposition without Mr Stava’s support.

Had the Trustee been intent on pursuing such a sale in defiance of Mr Stava’s wishes, I am satisfied that Mr Stava’s daughters would have supported Mr Stava in removing the Trustee in the exercise of his legal power as Protector to take that course with the requisite beneficiary consent. This would have been a means of giving effect to a straightforward business judgement, on an issue where Mr Stava’s daughters would have deferred to Mr Stava’s judgement as to the best means of securing value from the company he had created and the claims he had pursued for so long and with such effort. I am also satisfied that, in such an eventuality, Mr Stava would have been entitled properly to exercise his powers to remove the Trustee seeking to effect a sale and to appoint himself as Trustee, together with a Liechtenstein domiciled trustee. He was uniquely placed to administer the Koruna Trust given its very limited and particular asset base. Mr Stava would have been supported by his daughters in doing so. That situation never came to pass – nor was it ever remotely likely that it would be necessary.

In his capacity as Protector, Mr Stava had the legal power to prevent the Trustee adding an individual to or excluding an individual from the Class of Beneficiaries (by withholding the required consent). That power was one which Mr Stava was obliged to exercise in the interests of the Koruna Trust. Equally, the Trustee was not obliged to follow Mr Stava’s instructions to add or exclude individuals from the Class of Beneficiaries.

I accept that the Trustee would not automatically have followed Mr Stava’s wishes in relation to the addition or exclusion of potential Beneficiaries or in making distributions from the Koruna Trust. These were very different contexts to the management of Diag SE’s business. While I accept that as the Settlor and a member (albeit one of ten) of the Class of Beneficiaries, Mr Stava would have had a very important voice in these decisions, the Trustee would have had regard to all relevant circumstances (including Mr Stava’s wishes as Settlor and the source of the assets, the views of members of the Class of Beneficiaries, the purpose of the Koruna Trust of settlement planning, tax consequences, whether there were any creditors etc).

In January 2012, had Mr Stava’s three daughters (as members of the Class of Beneficiaries themselves and as parents of the other members of the Class of Beneficiaries save for Mr Stava) supported Mr Stava’s wishes as to membership of the Class of Beneficiaries or the distribution of assets from the Koruna Trust, it is virtually certain that the Trustee would have formed the perfectly proper professional judgement that it should act in the manner supported by the Stava family.

I accept the evidence of Mr Stava’s three daughters that they found it difficult presently to foresee circumstances in which they would have opposed their father’s wishes in relation to the identity of eligible Beneficiaries or the distribution of assets from the Koruna Trust. In my assessment, it is highly unlikely that there would have been any disagreement between Mr Stava and his daughters had these issues arisen as at January 2012. Had it been necessary to do so, I am satisfied that there is a high likelihood that Mr Stava’s daughters would have supported him if he had decided to exercise his legal power to remove the Trustee if it had refused to give effect to his wishes on these issues, and to appoint himself as Trustee together with a Liechtenstein-domiciled trustee.

However, I accept Mr Dunning KC’s submission that it is not fanciful to suppose that there could be circumstances in which there would be a disagreement within the Class of Beneficiaries as to who the members of the Class of Beneficiaries should be or how assets should be distributed from the Koruna Trust. In that eventuality, it is not possible to determine how the Trustee would have acted or whether the Trustee could have been removed. Everything would depend on the particular circumstances.

My conclusions are as follows: