Yukos Capital Sarl v OJSC Rosneft Oil Co
If from the imperfect evidence of foreign law produced before it, or its misapprehension of the effect of that evidence, a mistake is made by an English court, it is much to be lamented, but the tribunal is free from blame.Footnote 1 The mistake to be lamented presently is the High Court decision in Yukos Capital Sarl v OJSC Rosneft Oil Co Footnote 2 that a Dutch judgmentFootnote 3 gave rise to an issue estoppel in English proceedings, precluding a party from disputing as a fact the partiality and dependence of the Russian judiciary.Footnote 4 The decision was a mistake because on a proper construction of Dutch law the significance of the Dutch judgment was—if anything—evidential, not preclusive.Footnote 5 The outcome is lamentable, because a party was unduly shut out from litigation by the application of English preclusion law to a foreign judgment that was not preclusive in the jurisdiction where it was originally given.Footnote 6
The focus of this note is on the cause of the court's mistake. Why was it wrong to base an issue estoppel on the Dutch judgment? Owing to the particularities of the English approach to issues of (issue) preclusion arising in respect of foreign judgments, both English and Dutch law are potentially relevant for an answer to this question.Footnote 7 In this particular case, however, despite the court's unsatisfactory explanation of one of the core requirements for an issue estoppel under English law pertaining to what constitutes an ‘issue’,Footnote 8 the root of the error was the Dutch law evidence the court relied on as the basis for its decision.
In other respects, the decision is unremarkable. As most would have thought, Act of State or non-justiciability principles cannot prevent an English court from considering allegations of partiality and dependence of a foreign judiciary in the process of deciding whether or not to recognize foreign judgments in England and Wales.Footnote 9 Also, no lawyer familiar with English private international law will be surprised to hear that, in the interests of finality, foreign judgments can found an issue estoppel under English law, with the effect of precluding a party from disputing such serious allegations.Footnote 10 In both respects the judgment of Hamblen J is customarily well reasoned and irreproachable as a matter of English law.
1. The background of the decision
a) The dispute
The case related to a number of intra-group loans within the ‘Yukos Group’,Footnote 11 a well-known Russian group of companies engaged in oil production and trading, headed by OAO Yukos Oil Company (‘Yukos Oil’) until its insolvency, forced break-up and re-nationalization in December 2004.Footnote 12 Failing repayment of the loans by the borrowing party, OJSC Yuganskneftegaz (‘Yuganskneftegaz’), a Russian production subsidiary, the lender, Yukos Capital Sarl (‘Yukos Capital’), a Luxembourg company (still controlled by Yukos Oil and its shareholders), filed for arbitration in Russia.Footnote 13 In September 2006, the tribunal rendered four awardsFootnote 14 in favour of Yukos Capital for the equivalent of US$425 million exclusive of interest and costs (the ‘Awards’).
Having recovered the Awards, Yukos Capital pursued their enforcement through the 1958 New York Convention,Footnote 15 first in The Netherlands and after that in England and Wales. In the interim, however, the Awards were annulled by the Moscow Arbitrazh court in a series of judgments on 18 and 27 May 2007 (the ‘Annulment Judgments’),Footnote 16 on application of OJSC Rosneft Oil Company (‘Rosneft’), a Russian company controlled and majority owned by the Russian State, which in 2004 had acquired the majority of the assets of the Yukos Group, including Yuganskneftegaz.Footnote 17
b) The Dutch proceedings and the ‘Dutch judgment’
The enforcement proceedings in The Netherlands got under way in March 2007, when Yukos Capital filed an application for leave to enforce the Awards at the Amsterdam District Court.Footnote 18 Rosneft challenged the application on the basis of Article V(1)(e) of the 1958 New York Convention, by reference to the Annulment Judgments.Footnote 19 In reply, Yukos Capital alleged that these judgments were the product of a partial and dependent judicial process and, hence, unworthy of recognition in The Netherlands, in which case they lacked validity in the eyes of the Dutch legal order, so that they could not be taken into account for the purpose of determining the application. On 28 April 2008, the District Court rejected the application in accordance with the 1958 New York Convention, Article V(1)(e), after finding that Yukos Capital had failed to prove or even sufficiently allege partiality and dependence of the Russian judiciary.Footnote 20 Yukos Capital appealed.
The Amsterdam Court of Appeal reconsidered and granted the application by judgment of 28 April 2009 (the ‘Dutch Judgment’).Footnote 21 The court found that ‘it is so likely that the Russian civil court judgments setting aside the arbitral awards are the result of an administration of justice which is to be qualified as partial and dependent, that it is not possible to recognize those judgments in The Netherlands.’Footnote 22 The court therefore ignored the Annulment Judgments and, for that reason, it refused to apply Article V(1)(e) of the 1958 New York Convention. Its judgment became irreversible on 25 June 2010 when Rosneft's appeal in cassation failed on procedural grounds.Footnote 23
c) The English proceedings
Matters in England and Wales got under way in March 2010, when at the request of Yukos Capital and following an ex parte hearing, freezing orders were granted against Rosneft (and a number of associated companies) to secure funds for the enforcement of the Awards in England and Wales.Footnote 24 Specifically, Yukos Capital sought to recover the interest on the principal sum of the Awards (valued at over US$160 million at Russian Central Bank rate) it had lost owing to Rosneft's failure to satisfy the Awards between 2006, when they were issued, and 2010, when they were finally complied with following their enforcement in The Netherlands.
Up to a point the High Court proceedings mirrored the proceedings in The Netherlands. Like in the Dutch proceedings, Rosneft opposed enforcement by invoking the Annulment Judgments in conjunction with Article V(1)(e) of the 1958 New York Convention. In answer, as it had done in the Dutch courts, Yukos Capital alleged that the judgments invoked by Rosneft were in fact the product of a partial and dependent judicial process so offensive against English principles of substantial justice, that the resulting judgments could not be recognized in England and Wales.Footnote 25 But then proceedings took a markedly different course: when Rosneft freshly denied Yukos Capital's allegation, Yukos Capital replied that Rosneft was precluded—‘issue estopped’—from disputing its allegation, because the Amsterdam Court of Appeal, whose judgment English courts were bound to recognize, had previously determined the issue of partiality and dependence in its favour.Footnote 26 Yukos Capital further submitted that if Rosneft was indeed so precluded, it was essentially established that the Annulment Judgments should be refused recognition in England and Wales. This, in turn, would open the way to the granting of the claim for enforcement. Hamblen J clearly acknowledged the implications of Yukos Capital's plea of finality, or ‘res judicata’, as shown by his observation that: ‘If it [ie the court] decides not to do so [ie recognize the Annulment Judgments] that will be consistent with the decision reached by the Court of Appeal and it is in the interests of finality that the underlying factual issue decided by that court should not have to be relitigated.’Footnote 27
2. Evaluation—The decision to base an issue estoppel on the Dutch judgment
a) What is the nature of the question?
When a foreign judgment is invoked in an English court for preclusive purposes, two frequently associated but fundamentally distinct questions crop up: first, should the foreign judgment be recognized (as valid)Footnote 28 in England and Wales and, if so, what preclusive effects properly attach to the foreign judgment as a matter of English law? While recognition is a precondition for the attribution of preclusive effects, it is not equivalent to the attribution of preclusive effects. In other words, whereas every foreign judgment in order to have preclusive effect must first be recognized, not every recognized judgment is also preclusive. Hence, by asking the question whether the Dutch Judgment gave rise to an issue estoppel in the English proceedings, Hamblen J was concerned with the consequences of recognition, as opposed to the recognition itself. Still this is what Hamblen J appears to suggest, for instance,Footnote 29 where he notes that:
Considerations such as that it may have been impractical for a litigant to deploy his full case in an ‘earlier case of trivial character abroad’ are more likely to be relevant to whether there are ‘special circumstances’ which would make it unjust to recognise the decision.Footnote 30
But is application of the ‘special circumstances’ exception tantamount to refusing a foreign judgment recognition? It is respectfully suggested that this is conflating the issues. As the judge correctly explained, the exception applies in circumstances where the conditions for an issue estoppel are technically met, but where an issue estoppel would cause an injustice. In truth then, the exception is concerned with the legal consequences of a foreign judgment, as opposed to its recognition.
b) Does the common law control the issue?
In applying Carl Zeiss Footnote 31 as directing his basic approach to the problem of preclusion and foreign judgments and The Sennar (No 2) Footnote 32 as offering a further clarification of that approach specifically in relation to Dutch judgments, Hamblen J seems to take for granted that the common law controlled the consequences of the Dutch Judgment's recognition. This assumption is not as self-evident as it is correct; common law plainly governs if common law was the basis for recognition of the Dutch Judgment, but it was not. Admittedly, the Brussels I Regulation was entirely irrelevant in the circumstances.Footnote 33 This instrument could not, then, interfere with the judge's analysis, even if it were true that its reach would otherwise extend (implicitly) to the consequences of recognition.Footnote 34 However, of the two remaining options—the Foreign Judgments (Reciprocal Enforcement) Act 1933Footnote 35 and the common lawFootnote 36—the common law applies only residually and for the Dutch Judgment stood superseded by the 1933 Act.Footnote 37 It follows that Statute as opposed to common law was determinative of the consequences of recognition of the judgment,Footnote 38 in addition to defining the conditions forFootnote 39 and exceptions toFootnote 40 its recognition. Nevertheless, Hamblen J was right to revert back to the common law, since the House of Lords in Black Clawson Footnote 41 clarified that section 8(1) of the 1933 Act did not vary the common law for judgments recognized under the act as far as concerns the consequences of recognition.Footnote 42 In particular, the act did not affect the rule in Harris v Quine,Footnote 43 stated by Blackburn J, that ‘wherever it can be shewn that a court of competent jurisdiction has decided the matter, the plaintiff is estopped from disputing the decision, or litigating the matter in another court, while the decision of the first court remains unreversed.’Footnote 44
c) The cautious lex fori approach
English law invariably governs issues of preclusion arising in respect of foreign judgments recognized in England and Wales, not because this law is thought to be on balance the proper law of preclusion, but because issues of preclusion continue to be regarded as questions of procedure insusceptible to any choice of law analysis whatsoever.Footnote 45 At the same time, English courts are slow to impose English standards of finality of litigation if it is pleaded and, if contested, shown by evidence of foreign law that the issue in question can be (re)litigated in the jurisdiction where the foreign judgment was originally given which is invoked for purposes of preclusion. Thus there is an element of caution in the English approach; the imperative acknowledged by Lord Reid in Carl Zeiss, that English preclusion law be applied to foreign judgments in a manner consistent with ‘good sense’.Footnote 46 In keeping with this ‘cautious lex fori’ approach, Hamblen J applied English preclusion law to the Dutch Judgment and verified whether on that basis an issue estoppel arose with the effect of precluding Rosneft from disputing the fact of partiality and dependence of the Russian judiciary.Footnote 47 At the same time, he appropriately heard evidence on Dutch law when counsel for Rosneft alleged that this fact could be disputed afresh in The Netherlands. Logically this note should first assess the judge's application of English law and then the manner in which he took Dutch law into account. However, while the application of English law in the case is not entirely convincing,Footnote 48 it did not cause the eventual mistaken conclusion that the Dutch Judgment could properly found an issue estoppel. Accordingly, it is suggested to turn directly to the judge's assessment of Dutch law.
d) Taking account of Dutch law
When prompted to do so by counsel for Rosneft,Footnote 49 Hamblen J considered whether the Dutch Judgment would in The Netherlands have ‘force of res judicata’ (gezag van gewijsde), in the sense of being preclusive of litigation.Footnote 50 The judge's willingness to take Dutch law into account is illustrative of the good-sense approach of English courts to issues of preclusion arising in respect of foreign judgments, which mandates caution so as to avoid precluding issues that may be (re)litigated abroad. Accordingly, the judge heard expert evidence on Dutch law. In a joint memorandum,Footnote 51 the expertsFootnote 52 agreed that Dutch preclusion law, specifically Article 236 of the Code of Civil Procedure (‘Rv’),Footnote 53 attaches preclusive effect to the Amsterdam Court of Appeal's final decision (geschilbeslissing) on Yukos Capital's claim for enforcement of the Awards in The Netherlands.Footnote 54 However, they disagreed on whether the same was true for that court's findings which preceded its decision on the claim (voorbeslissingen), including the finding of partiality and dependence.Footnote 55 The expert for Rosneft denied the existence of such preclusive effect, albeit without proving his contention.Footnote 56 Conversely, the expert for Yukos Capital asserted that this effect does attach by virtue of Dutch law.Footnote 57 In support, he cited the Dutch Supreme Court decision in Siegers v Citco Bank Antilles NV. Footnote 58 Ultimately, following cross-examination, the expert for Rosneft was forced to concede that ‘res judicata extends to the establishment of a fact which ‘‘is part of a decision regarding the legal relationship in dispute’’.’Footnote 59 On that basis, Hamblen J concluded that it was ‘common ground on the expert evidence that the finding that the Annulment Decisions were the result of a partial and dependent legal process would be res judicata in the Netherlands’.Footnote 60
The first step in assessing the accuracy of this conclusion is to note a simple error of translation of one of the key terms of Article 236 Rv.Footnote 61 The Dutch law experts tendered a literal version of ‘rechtsbetrekking in geschil’, which they interpreted as ‘legal relationship in dispute’. Accordingly, Hamblen J assumed that Article 236 Rv applies to a ‘decision regarding the relationship in dispute’, including any finding of fact forming part of such a decision.Footnote 62 But in reality, the provision no longer has its literal meaning, which in practice proved to be too vague, rendering it unworkable. Instead, the Dutch Supreme Court has interpreted ‘rechtsbetrekking in geschil’ as denoting the ‘rechtsvraag’; that is, literally, the ‘legal question’ under consideration, or more precisely, the claim (or issue) to be determined by the court.Footnote 63 In reference specifically to the issue to be determined, that Court has also applied the term ‘geschilpunt’.Footnote 64 Hence, an accurate statement of Dutch law to be offered to an English court having to take this law into account would have translated Article 236(1) Rv in light of its interpretation in practice, as follows: ‘Judicial findings regarding the claim or issue contained in an irreversible judgment, are conclusive between the same parties in another case.’Footnote 65
In light of this clarification, the relevant question to ask for the purpose of applying Article 236(1) Rv is whether the Amsterdam Court of Appeal's finding of fact (ie that the Annulment Judgments were the result of a partial and dependent legal process in Russia) was a ‘finding regarding an issue’. Answering this question involves an explanation of what exactly was the ‘issue’ to be determined by the court. There can be little doubt in this respect; the issue was whether recognition of the Annulment Judgments would be in breach of Dutch public order.Footnote 66 In relation to this issue, Yukos Capital made the point that the Annulment Judgments resulted from judicial partiality and dependency; a fact Rosneft denied. The court settled this question in favour of Yukos Capital. However, its finding that the Annulment Judgments were the result of a partial and dependent legal process in Russia was not a finding ‘on’ the issue in question, since the issue was whether the Annulment judgments violated Dutch public order, not whether those judgments were the product of a partial and dependent legal process. Nevertheless, as noted above, Article 236 Rv applies to findings ‘regarding’ the issue in question, a term broad enough to encompass findings on questions of fact and law made in the process of determining the issue. This broad interpretation was confirmed by the Dutch Supreme Court in Siegers/Citco Bank Antilles NV, the case cited by the Dutch law expert for Yukos Capital as authority for his assertion that Article 236 Rv applies to findings of fact. The Court held specifically that:
The grievance postulates that a judicial finding, despite concerning the issue in question, cannot have preclusive effect if that finding is factual in the sense that it, as the grievance puts it, establishes a fact. This starting point is incorrect. Such a finding is preclusive in another case between the same parties if that finding does not merely establish a fact (cf. HR 15 May 1987, NJ 1988, 164), but is part of the finding on the issue in question, meaning that the issue is (in some measure) determined by the fact's legal consequences which, according to that finding, apply between the parties.Footnote 67
Applying this reasoning to the finding of partiality and dependence by the Amsterdam Court of Appeal, it is clear that, while this is not a finding ‘on’ the issue in question, it nonetheless qualifies as a finding ‘regarding’ the issue in question covered by Article 236 Rv, because it forms part of the finding that Dutch public policy barred the recognition of the Annulment Judgments in The Netherlands—that is, the question of recognition was (in some measure) determined by the legal consequences (refusal of recognition) which, according to that finding (violation of public policy), apply between the parties on account of the fact established (partiality and dependence of the legal process in Russia).
Nevertheless, it is not sufficient for the applicability of Article 236 Rv that the finding of partiality and dependence by the Court of Appeal in the enforcement proceedings qualifies as a ‘finding regarding the issue in question’. In addition there needs to be identity between the issue determined in those proceedings and the issue arising in the succeeding case, for Article 236 Rv lacks application to a finding regarding the issue in question in one set of proceedings if the issue in question in the subsequent set of proceedings is different.Footnote 68 This requirement was clearly stated by the Dutch Supreme Court in Van Huffel/Van den Hoek,Footnote 69 where it ruled that ‘the finding of the [court whose judgment was relied upon as the basis for preclusion] related to a different issue—that is, the legal relationship in dispute—than that in question in the present case.’Footnote 70 Advocate General Asser in his opinion in Siegers/Citco Bank Antilles NV concluded similarly that: ‘The Supreme Court [in Van Huffel/Van den Hoek] accordingly expressed that it is inappropriate to invoke the preclusive effect of a finding of fact isolated from the finding on the issue in question in relation to which it was made.’Footnote 71
The requirement of identity of issues is fatal in the present case, because that identity is manifestly absent; the Amsterdam Court of Appeal determined the issue whether Dutch public policy barred the recognition of the Annulment Judgments in The Netherlands, whereas the High Court was called upon to determine whether English principles of substantial justice prevented the recognition of the same judgments in England and Wales.Footnote 72 If any authority is required for this conclusion, which is strictly logical, reference can be made once more to the opinion of Advocate General Asser in Siegers/Citco Bank Antilles NV, who made precisely the same point:
If a new case concerns another issue, a prior factual finding cannot be regarded as having preclusive effect by the court called upon to determine that issue in the new set of proceedings, notwithstanding that the factual finding is equally relevant for the finding on the new issue. This relevance relates to the evidential value of the prior judgment.Footnote 73
Accordingly, it is entirely irrelevant that both cases raised identical questions of fact, because the issue in question was different. This is not to say that the Amsterdam Court of Appeal judgment would be devoid of any relevance. But its proper significance would be evidential, not preclusive.
3. Is the case cause to reconsider the cautious lex fori approach?
The English cautious lex fori approach to issues of preclusion arising in respect of foreign judgments strikes a middle ground between dogmaFootnote 74 and good sense. The present case is no reason to reconsider it. While an alternative approach might be to apply the law of the judgment rendering jurisdiction, if anything, the mistake as to Dutch law in the present case would be cited in favour of excluding reference to foreign law altogether. Such a change should be undesirable, however, as it would give rise to complications like those presently inherent in section 34 of the Civil Jurisdiction and Judgments Act 1982Footnote 75 as construed by the House of Lords in The Indian Grace.Footnote 76 As currently understood, this claim-preclusive rule prevents the reassertion of a cause of action for which judgment has been recovered abroad without enabling English courts to take account of the fact that the cause of action can actually be reasserted in the judgment-rendering State. This imposition of English standards of finality of litigation without observing the need for conspicuous caution risks inflicting serious injustice to and between litigants.Footnote 77 Against this background, the inclination of English courts to take foreign law into account, as illustrated by the present case,Footnote 78 is only admirable.Footnote 79
At the same time, it should be noted that the cautious lex fori approach is clearly not the most efficient method, as it forces English courts and the parties to consider two legal systems cumulatively, while the application of a single law—the law of the judgment-rendering court—would (in theory) suffice to do justice. Moreover, it is unclear how the approach could work justice in circumstances where the scope of preclusion under the law of the judgment-rendering court is wider than in English law.Footnote 80 By mandating the application of English law, the approach sets a ceiling on the preclusive effect of a foreign judgment, whereas good sense would back preclusion beyond English standards of finality.Footnote 81 This is not the best place to undertake the considerable task of defining an alternative approach which is more efficient and just than the current approach. Nonetheless, two crucial questions which would arise in the process of devising such approach can be briefly identified and summarily addressed. First, should it matter that application of the law of the judgment-rendering court is tantamount to applying foreign law? Of course not—English courts are well able to make findings on the meaning of foreign law. It is something they do all the time.Footnote 82 The fact that they do so by reference to expert evidence and that this evidence may occasionally be wrong does not outweigh the benefits of international decisional harmony when it comes to the implementation of the principle of finality of litigation between legal systems.
Second, should it matter that application of foreign preclusion law involves applying foreign procedural law? Of course not, the belief that for reasons of convenience questions of procedure are always best governed by English law is mistaken, as demonstrated in the present context: whereas English courts theoretically apply only English preclusion law, in truth they have foregone the comfort of the lex fori by accommodating foreign law through the back door by taking it into account so as to avoid injustice. What then is left of the legitimacy of the procedural question exception? It is submitted that much is to be said for the approach recommended by Arden LJ in Harding v Wealands that ‘a reference to the law of the forum must be the exception, and it must be justified by some imperative which, relative to the imperative of applying the proper law, has priority. It may, for instance, be appropriate to apply the law of the forum where the court cannot put itself into the shoes of the foreign court’.Footnote 83 This approach was not accepted by the House of Lords in the context of international tort law when it comes to the quantification of damages.Footnote 84 However, in this context it may well be preferable. As noted, English courts have shown themselves more than capable to step into the shoes of the foreign judgment-rendering court. Accordingly, the application of English preclusion law as the law of the forum cannot be justified relative to the imperative of applying the proper law of preclusion. This dogmatic shift away from excluding any choice of law analysis in respect of issues of preclusion with a foreign element (here the judgment) need not imply that all issues are necessarily governed by foreign law; as Mance LJ (as he then was) in Raiffeisen Zentralbank said about the (three-stage)Footnote 85 process through which the common law identifies the appropriate law: ‘The overall aim is to identify the most appropriate law to govern a particular issue.’Footnote 86 Accordingly, the appropriate approach would be to look at the particular preclusion issue and to designate ‘in a broad internationalist spirit’Footnote 87 the preclusion law most appropriate for its resolution. This may be foreign law or the law of the forum as appropriate.