Much ink has been spilled on the 2014 Yukos arbitral awards, and rightfully so. They are notorious for collectively breaking the previous record for the largest arbitral award in history. Their magnitude (these were, in the tribunal’s words, “mammoth arbitrations”) also commands our attention, as do the issues at stake and the multiple companion arbitrations and enforcement proceedings around the globe that have sprung like mushrooms after the rain. One of such related cases is Luxtona Limited v. The Russian Federation, a PCA-administered, Toronto-seated UNCITRAL arbitration brought by a former Yukos shareholder. Though the tribunal has thus far only rendered an interim award on jurisdiction in 2017, the case is noteworthy as it shines the light on the Canadian judiciary’s take on the all-so-important issue of new evidence in the context of court proceedings under Art. 16(3) of the UNCITRAL Model Law (“ML”).

On June 30, 2021, the Ontario Superior Court of Justice (“Court”) allowed Russia to file new evidence because the Court was considering the issue of jurisdiction de novo, as opposed to reviewing the tribunal’s ruling. In reaching its conclusion, the Court drew inspiration from international authorities, including the UK Supreme Court judgment in Dallah v. Pakistan (“Dallah”), which was previously extensively discussed on this blog (for example, here and here).


The Evolving View of the Court

By way of reminder, Art. 16(3) provides that “if the arbitral tribunal rules as a preliminary question that it has jurisdiction, any party may request, within 30 days after having received notice of that ruling, the Court to decide the matter […].”

Interestingly, this was the third time the Court grappled with the issue of evidence in this case. Russia initially sought to set aside the tribunal’s award pursuant to Arts. 16(3) and 34(2) ML. In April 2018, Dunphy J. allowed Russia to file new evidence as of right because “[t]he court is directed to ‘decide the matter’ and not merely to review the decision of a tribunal whose very existence may or may not have been authorized” (2018 ONSC 2419, para. 33). Russia’s new evidence included two expert reports: one, which allegedly responded to the tribunal’s “incorrect findings”, and the other, which addressed statutory interpretation arguments that had evolved “following the hearing [before the tribunal] in response to positions asserted by Professor Stephan in reports submitted to the Hague Court of Appeal” in a different case (Professor Stephan is also Luxtona’s expert). In response, Luxtona filed additional expert evidence on Russian law.

Due to changes in judicial assignments, the application was reassigned to Penny J. who first accepted but then questioned Dunphy J.’s ruling and asked the parties “to reargue the narrow question of whether new evidence, which does not meet the test for new evidence under Ontario Law, is admissible on a court review of an arbitral tribunal’s jurisdiction under Art. 16(3)” (2019 ONSC 4503, para. 38). As Dunphy J.’s decision was merely an earlier evidentiary ruling, Penny J. could revisit and reverse it, as he did in December 2019 (2019 ONSC 7558). He determined that Russia could not file fresh evidence as of right. Rather, it had to show that “(1) the evidence could not have been obtained using reasonable diligence; (2) the evidence would probably have an important influence on the case; (3) the evidence must be apparently credible; and (4) the evidence must be such that if believed it could reasonably, when taken with the other evidence adduced at the hearing, be expected to have affected the result,” (para. 69) which Russia failed to demonstrate.

Penny J.’s decision was appealable with leave to the branch of the same Court, known as the Divisional Court, pursuant to s. 19(1)(b) of the Courts of Justice Act. Leave was granted in August 2020 (2020 ONSC 4668) and Penny J.’s decision was set aside in June 2021.

Why have there been so many different answers to a, seemingly, straightforward question within the same Court? Three main themes of divergence can be identified: (1) the relevance of authorities from a non-ML jurisdiction; (2) the significance and international acceptance of Dallah, and (3) the pertinence of United Mexican States v. Cargill, Inc. (“Cargill”) in the context of Art. 16(3) applications.

  1. After comparing the ML and UK legislation, Penny J. noted that the emphasis and scope for court intervention in the decisions of arbitral tribunals under the two statutory regimes differed significantly (so much that the UK approach undermined the competence-competence principle). Thus, English decisions were to be carefully scrutinized before being adopted in Ontario.1) 2019 ONSC 7558, paras. 54-55, 60-62. Conversely, the Divisional Court opined that there was no compelling reason to distinguish the nature of jurisdictional hearings under the two regimes.2) 2021 ONSC 4604, para. 33.
  2. With respect to Dallah, Dunphy J. observed that the international ML jurisprudence was quite unanimously in line with the approach suggested by Dallah and Cargill in the sense that a court need not defer to the decision of the tribunal on jurisdictional matters, nor is it explicitly confined to the record before such tribunal.3) 2018 ONSC2419, para. 28 Penny J. referred to Dallah briefly, noting its silence on the matter of new evidence.4) 2019 ONSC 7558, para. 46. Au contraire, the Divisional Court discussed Dallah extensively, concluding that it enjoyed “strong international consensus” and was cited by the Ontario Court of Appeal in Cargill with approval.5)2021 ONSC 4604, paras 30, 38.
  3. Cargill (previously discussed here) concerned Mexico’s application to set aside pursuant to Art. 34(2)(a)(iii) ML a NAFTA award that allegedly granted the investor losses in excess of the tribunal’s jurisdiction. Dunphy J. held that the “ratio” of Cargill was applicable to Art. 16(3) applications, adding that neither of these provisions constrained the Court to the four corners of the arbitration’s evidentiary record.6)2018 ONSC 2419, paras. 32-33. Penny J., while recognizing Cargill’s relevance for applications under both articles of the ML, highlighted that Cargill clearly described that “[o]n a true jurisdictional challenge, it is a review on correctness, without any deference, … but a ‘review’ nevertheless.”7)2019 ONSC 7558, paras. 57-58. Contrary to Dunphy and Penny JJ., the Divisional Court underscored that Cargill concerned Art. 34(2), which envisages a different test, i.e. a “review”. The Court of Appeal in Cargill “did not decide whether an application under Art. 16 [was] a ‘review’ or a hearing de novo,” nor did it comment on Dallah’s applicability under Art. 16(3).8)2021 ONSC 4604, paras. 23-24, 32.


A Closer Look at the Court’s Most Recent Decision in Light of Cargill and Dallah

By way of reminder, in Cargill the Court of Appeal discussed Dallah, albeit noting that the jurisdiction issue before it was “quite different under Art. 34(2)(a)(iii)” as it did not concern the ability of the tribunal to adjudicate altogether, but the content of the award itself. The Court of Appeal concluded that the standard of review was correctness, meaning that on a “true question of jurisdiction, the tribunal had to be correct in its assumption of jurisdiction to decide the particular question it accepted.”9)2011 ONCA 622, para. 53. The same court, however, cautioned that the standard of correctness does not presuppose “a broad scope for intervention in the decisions of international arbitral tribunals, [rather] only in rare circumstances where there is a true question of jurisdiction.”10)ibid., para. 44.

On the one hand, the Divisional Court says that Art. 34(2) which was at the center of Cargill, prescribes “the limited review” and provides for a different standard than Art. 16(3) (paras. 23-24). On the other, it upholds the de novo approach of Dallah, which concerned Art. V(1)(a) of the New York Convention which is much more similar to Art. 34(2) than Art. 16(3) ML. Can the two decisions be reconciled? Both concerned similar provisions, yet Dallah allowed a de novo hearing whereas Cargill provided for a review (on the standard of correctness).

If Cargill is not directly pertinent because it revolves around Art. 34, why is Dallah’s de novo approach applicable to an Art. 16 application? Is Dallah relevant whenever the existence of the arbitration agreement is in question, irrespective of which provision of the ML is invoked? This arguably leaves room for Dallah’s de novo approach in the context of an Art. 34(2) application, if the jurisdictional issue concerned the existence of the arbitration agreement (and not the award’s content, as in Cargill). If the de novo approach presupposes that a party can submit new evidence as of right, new evidence could then be introduced on an Art. 34(2) challenge to the final award. Would this be conducive to certainty, efficiency, fair play?

Conversely, if Dallah’s de novo approach is applicable only in the context of Art. 16(3) proceedings (because courts are invited to “decide the matter”), why should a party be able to file new evidence as of right just because the tribunal ruled on the jurisdictional issue as a preliminary question, and not have that same right if the tribunal did so in the final award?



Though Canadian authorities dealing with Art. 16(3) ML are rare, they appear to demonstrate the expansion of the courts’ role in recent years. In 2005, the Alberta Court of Queen’s Bench held that, despite the appearance of “wide discretion”, Art. 16(3) did not go “so far as to allow a reviewing court to substitute its view simply because the court would not have reached the same conclusion,” and that the standard of review ought to be “one of reasonableness, deference and respect” (para. 53). In February of this year the Ontario Court of Appeal discussed Art. 16(3) ML in United Mexican States v. Burr, albeit focusing on other parts of the provision. Perhaps it is instructive that, when quashing Mexico’s appeal, the court underscored that the text of said provision prohibited an appeal from “the ruling of a Superior Court judge on the correctness of an arbitral tribunal’s ruling” (para. 26). Now, the Divisional Court ruled in favor of a de novo hearing. Courts in Québec also seem to lean towards the de novo approach (see, e.g., Groupe Dimension Multi Vétérinaire inc. c. Vaillancourt, para. 10).

Ultimately, whatever the nature of the court’s involvement or the standard of review, should a party be allowed to file new evidence as of right? While situations when it may be necessary to admit “new” evidence exist (for example, under the conditions enumerated by Penny J., or if a party did not participate in the arbitration and only decides to become involved after the tribunal it does not recognize rules that it has jurisdiction), what is the justification when a party participated in the arbitration all along? Despite the ML’s evasiveness on this point, the drafters’ intention to circumscribe court intervention under Art. 16(3) is evident from the 30-day deadline, absence of appeal and the tribunal’s discretion to proceed while the matter is pending before the court (para. 26).

The arbitral tribunal in the present case has reportedly suspended its proceedings while the Canadian challenge is pending. Four years after the tribunal’s interim award, the Court’s decision on jurisdiction is nowhere in sight (introducing new evidence will certainly not expedite the matter). Is this the “immediate court control” that the ML’s creators had in mind?


*The views expressed herein are those of the author and do not necessarily reflect the views of Woods LLP or its partners.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Profile Navigator and Relationship Indicator
Access 17,000+ data-driven profiles of arbitrators, expert witnesses, and counsels, derived from Kluwer Arbitration's comprehensive collection of international cases and awards and appointment data of leading arbitral institutions, to uncover potential conflicts of interest.

Learn how Kluwer Arbitration can support you.

Kluwer Arbitration


1 2019 ONSC 7558, paras. 54-55, 60-62.
2 2021 ONSC 4604, para. 33.
3 2018 ONSC2419, para. 28
4 2019 ONSC 7558, para. 46.
5 2021 ONSC 4604, paras 30, 38.
6 2018 ONSC 2419, paras. 32-33.
7 2019 ONSC 7558, paras. 57-58.
8 2021 ONSC 4604, paras. 23-24, 32.
9 2011 ONCA 622, para. 53.
10 ibid., para. 44.
This page as PDF

One comment

  1. One may wonder that even if “may request, within 30 days after having received notice of that ruling, the Court to decide the matter […]” in Art. 16(3) of the UNCITRAL Model Law means “de novo”, why that should mean “on a different record” as opposed to on the record established in the matter already (leaving aside the point about “new evidence” that meets the high test for new evidence noted by Justice Penny).

    The wording that the court is to “decide the matter” may well, and one might argue based on the overarching intent of the Model Law should mean, that the court steps into the shoes of the arbitral tribunal and says “here is the record, now let’s decide the matter”.

    Why should “decide the matter” mean “start from scratch”? or “add to the record”?

    Justice Corbett states in his conclusion for the Divisional Court “But because the court is hearing the jurisdictional issue de novo, the parties are entitled as of right to adduce evidence, including expert evidence, relevant to the jurisdictional issue.”

    Where does “entitled as of right” come from? Certainly not the Model Law.

    The Model Law does not say “hear the matter de novo”, it says “decide the matter de novo.”

    Nor does it say “decide the matter on a different — as opposed to the same/existing — record”.

    An overarching theme of the Model Law is the efficient and timely determination of disputes. That is why the tribunal can carry on.

    Based on the words of Article 16(3) and the overarching theme, there seems no reason to construe “decide the matter” to mean “rehear that matter afresh”, or “rehear the matter by permitting additional evidence from the party making the request.”

    Looking at the forest, rather than the trees, would limit the record the court needs to the existing record, “possibly” with the caveat of evidence that meets the “fresh evidence” test — a high hurdle.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.