When material evidence tendered to an arbitral tribunal is fraudulent, we expect the court having jurisdiction to be very inclined to set the award aside. But as the recent decision of the English High Court in Chantiers de l’Atlantique S.A. v. Gaztransport & Technigaz S.A.S. demonstrates, “it ain’t necessarily so.” In that case the award was upheld despite the findings by the court that the arbitral tribunal heard material evidence which it held to be fraudulent. How did the court arrive at that conclusion?
The arbitration was an ICC arbitration held in Paris between French companies. The arbitration clause stated that the seat of the arbitration was in London, so the English Courts exercised supervisory jurisdiction over the arbitration and the application to set aside the award was made to the English courts.
The applicant to the arbitration, Chantiers de l’Atlantique (CAT), built LNG carriers. It licensed a technology from the respondent to the arbitration, Gaztransport & Technizaz (GTT) for the construction of the containment system in the tankers. During the sea trials of the tankers, it was discovered that nitrogen was passing through the barriers in the containment system, suggesting faults in that system. The faults could be due to GTT’s design or poor construction by CAT. Eventually, the ships were constructed with a barrier system that worked. However, CAT commenced an arbitration against GTT alleging that the initial failures in the barrier system were due to mis-design by GTT.
The arbitral tribunal dismissed the arbitral proceeding. It did so as it found that CAT could not establish “gross fault” on the part of GTT and thus could not meet the test imposed by French law for liability by a licensor to a licensee for a design fault or economic fault.
After the arbitral decision, CAT received a tip-off from a whistleblower raising questions about tests of the barrier system conducted by GTT and the alleged non-disclosure of those tests in the evidence that GTT had presented to the arbitral tribunal. CAT accordingly brought an application before the English courts for an order setting aside the arbitral award under section 68(2)(g) of the English Arbitration Act, 1996. That clause, like Section 46(1).9 of the Ontario Arbitration Act, 1991, authorizes the court to set aside the award if it is “obtained by fraud”.
After a very long review of the evidence led before the arbitral tribunal, the English Court concluded that fraud had been established for the purpose of section 68(2)(g) of the English Act. The misconduct arose from the non-disclosure of tests conducted by GTT. Yet the court declined to set aside the award. The reasoning of the judge is an important explanation and exploration of the factors which should guide a court in its consideration of the power to set aside an arbitral award for fraud.
First, the court set out the four principles:
- An arbitral award will only be set aside for fraud in extreme cases;
- Fraud is dishonest, reprehensible or unconscionable conduct. Fraud must be distinctly pleaded and proven to a heightened burden of proof.
- The Award must have been caused by the fraud. There must have been fraud “in the arbitration itself” and there must be “a causative link between the deliberate concealment of the document and the decision in the award”.
- The evidence of fraud must not be “of such as could have been obtained or produced at the arbitration hearing with reasonable diligence” and must be “so material’ it “would probably have affected the result of the arbitration.” The test does not require that the applicant show that the evidence would have affected the result, as such a test would usurp the function of the arbitral tribunal in the event that the matter is remitted to the tribunal, but the applicant must show that the evidence “would have had an important influence on the result.”
Except for the second test, these tests appear to be relevant for use by Canadian courts. But Canadian courts generally hold that there are only two standards of proof: balance of probabilities in civil cases, and beyond a reasonable doubt in criminal cases. So a Canadian Court would more likely apply a balance of probabilities to the second element of these principles.
Second, the judge went out of his way to point out that the arbitration had been conducted under the IBA rules. Under those rules, he said that “there was no duty to disclose relevant documents, akin to CPR Part 31, such as would be the case with London arbitration conducted in accordance with English procedure. In these circumstances, the court must be careful not to import into its assessment of GTT’s conduct….English concepts of the duty of disclosure.”
In the result, the judge found that, while some of GTT’s answers were misleading and inaccurate, there had been no fraud arising from the Requests that had been made during the arbitration. One is left to wonder whether the result would have been the same under common law duties of disclosure.
Third, the judge found that, while there had been fraud in the arbitration arising from the non-disclosure of testing results, the disclosure of the true position would not have affected the result of the arbitration. He gave seven reasons for arriving at that result. The most material and interesting are as follows:
- The other tests and evidence submitted by GTT to the arbitral tribunal demonstrated that the design of the barrier system was satisfactory. In light of all the evidence, the judge concluded that the impugned tests would not have affected the arbitral decision.
- Ultimately, the barrier system did work. In this context, a defect in design was not a likely explanation.
- If the impugned tests had been revealed, they would not have had a devastating impact on the conduct of the arbitration.
- The non-concealment of some tests was not the “tip of the iceberg.” It did not demonstrate that non-concealment was a wider issue.
- The arbitral tribunal had held that, even if design fault was proven, the standard of misconduct required by French law had not been established. The tribunal decided that, under French law, the design had to be “technically unusable or extremely difficult to use” to give rise to liability. Since the barrier technology had ultimately been implemented, that test could not be met.
These conclusions are noteworthy on a number of accounts:
First, the judge went into the technical evidence, and the evidence before the arbitral tribunal, to an extraordinary extent. His conclusions, and particularly as to whether the impugned evidence would probably have impacted the arbitral tribunal, came very close to a re-trial of the merits of the substantive issue between the parties, something which the judge warned himself that he should not undertake.
Second, the judge was largely unmoved by the impact on credibility that the disclosure of the impugned tests would have had on the arbitration process and tribunal. This reaction of the judge is somewhat surprising, having regard to the unpredictable impact which non-disclosure can have upon the credibility of parties and witnesses during any contested proceeding. Here, the issue might be what exactly is the question:
Is it: Would the disclosure of the true position probably have affected the result of the arbitration? (as stated by the judge). Or is it, or does it include: Did the non-disclosure of the true position probably affect the result of the arbitration? Those two questions do not necessarily lead to the same answer.
Third, the principles applied by English courts are based upon a strong disinclination to interfere with arbitration proceedings, even in the presence of fraud. Is this the right judicial attitude? Is the test used by the court too high a test? Is using this test the best way to ensure that both the stature and credibility, and the independence, of arbitral proceedings are protected? These questions will be further debated in the courts as arbitration proceedings, and in particular international commercial arbitrations, become even more common.
Fourth, the court’s decision seems to have been overwhelmingly driven by the very high test for liability under French law which the arbitrators applied, and the fact that the barrier system ultimately did work. The judge seems to have felt that, since the system did work, the applicant’s complaint was something of a tempest in a teapot, no matter how serious were his findings of misconduct.
These latter two factors may well not be present in another case. In North America, the test for negligence does not appear to be nearly as high as the test under French law used by the arbitrators. And if, in another case, the design or manufacture of the article is ultimately shown to be faulty, then the second factor will not exist.
In any event, this decision will be a useful precedent for future considerations of arbitral awards attacked on the ground of fraud.