How Not to Draft Dispute Resolution Clauses in Standard-Form Contracts
By James Plotkin Q.Arb
This article explores the enforceability of dispute resolution clauses in standard-form contracts through the lens of the Ontario Court of Appeal’s recent decision in Binance Holdings v. Lochan.
Background
Binance Holdings operated the world’s largest cryptocurrency trading platform. Between 2019 and 2022, it sold cryptocurrency derivatives to Canadians through its website. That website boasted that users could create an account in “under 30 seconds”, even though Binance’s users could not proceed without accepting roughly 50 pages of terms, including an arbitration agreement.
The representative plaintiffs brought a proposed securities class action alleging that Binance failed to file or deliver a prospectus, contrary to its obligations under securities law. Binance moved to stay the proposed class action based on the arbitration agreement in its website terms.
Ontario courts refuse to refer the proposed class action in favour of arbitration
The motion judge dismissed the stay motion, finding that the arbitration agreement was unconscionable and contrary to public policy because:
- Binance had stated that a user can open an account in “under 30 seconds” while still expecting the user to be bound by a nearly 50-page contract.
- The arbitration agreement forced the user to agree in advance that Binance could change any part of the arbitration agreement without notice, with binding effect.
- During the proposed class period, Binance in fact changed the arbitration forum and governing law four times. The last of these changes provided for arbitration in Hong Kong, under Hong Kong law, administered by the Hong Kong International Arbitration Centre. The administrative cost was prohibitive—for disputes under $1 million USD, the median cost was $26,743 USD (approximately $36,000 CAD). This excluded other incidentals, such as travel and accommodation, legal fees and the like.
The Court of Appeal upheld the motion judge’s decision, rejecting Binance’s three arguments.
First, Binance argued that the motion judge erred in failing to apply the competence-competence principle—that the arbitral tribunal can and should generally assess challenges to its jurisdiction or the arbitration agreement’s validity before the matter is put to a court.
The Court found no error. The motion judge applied the Supreme Court of Canada’s framework in Dell Computer Corp. v. Union des consommateurs, 2007 SCC 34, and Uber Technologies Inc. v. Heller, 2020 SCC 16 applicable to assessing stay motions. Under that framework, the court should refer the challenge to the arbitral tribunal unless resolving it involves a pure question of law or a question of mixed fact and law requiring only a superficial assessment of the record.
Dell/Uber also permits a court to make the assessment where the party resisting a stay would, as a practical matter, have no opportunity to raise the challenge before the arbitral tribunal. This will be so if the cost and other burdens form a “brick wall” for the challenging party. The motion judge found that to be the case here. The Court of Appeal saw no basis to disturb that finding.
Second, Binance argued that the motion judge impermissibly went beyond assessing a question of law and considered the particular facts at issue.
Again, the Court rejected this argument, finding that the motion judge instructed himself properly on the Dell/Uber framework. It found that, to the extent the motion judge considered the record, he did not go beyond the superficial review permitted under Dell/Uber.
Third, Binance essentially repackaged the second argument, stating that the judge exceeded a superficial review of the record in finding that the arbitration agreement erected a “brick wall” preventing the proposed class members from accessing arbitration. Binance also argued that the motion judge should have considered the representative plaintiff’s specific circumstances rather than those generalizable to all class members in determining whether the clause foreclosed access to arbitration.
The Court rejected this argument on largely the same basis as the others. It found no palpable and overriding error in the motion judge’s decision to assess the arbitration agreement “on the basis of the typical cryptocurrency investor and the nature of disputes likely to arise under the arbitration clause.”
Key takeaways
In most respects, this case is a straightforward application of the Dell/Uber framework. However, it is interesting to consider the extent to which a finding of unconscionability can or should be made on a class-wide basis. On one hand, there does appear to be some force to Binance’s position. Unconscionability is an inherently fact-based analysis that accounts for the circumstances of the party resisting the stay motion (as well as the party seeking the stay and invoking the arbitration agreement).
At the same time, it is difficult to fault the Court. In Uber itself the Supreme Court of Canada assessed unconscionability based on the representative plaintiff’s circumstances, not the circumstances of each potential class member. The difference here is that the Court did not consider the proposed representative plaintiff’s specific circumstances as in Uber. Instead, it assessed the question “on the basis of the typical cryptocurrency investor and the nature of disputes likely to arise under the arbitration clause.”
This seems like a rough and ready practical solution to the problem of assessing unconscionability in a class action context, especially where class members may not all share the same material traits. Unlike Uber, which dealt with generally lower income rideshare drivers, the means and abilities of Binance’s clientele, crypto traders of different levels of wealth and sophistication, may vary to a greater degree. It would be untenable to assess them on an individual basis. Indeed, this would defeat the purpose of the class action certification process provided these individuals are all otherwise within an identifiable class.
This case also offers a lesson on how not to structure an enforceable arbitration agreement in a contract of adhesion. The terms here appear designed to erect the very brick wall Uber identified as an exception to the rule that the arbitral tribunal should assess a challenge to its jurisdiction or the arbitration agreement in the first instance. The post-Uber jurisprudence reveals ways that Binance could have mitigated the risk of the unconscionability/public policy hurdle.
For example, in Difederico v. Amazon.com, Inc., 2022 FC 1256 (aff’d 2023 FCA 165), the Court rejected an unconscionability argument. The arbitration agreement in that case was far less onerous. It provided for a modest $200 filing fee, which Amazon would reimburse for claims under $10,000. It expressly provided that the arbitration could be conducted in writing, by phone or at an agreed location. The clause also preserved the customer’s option to sue in small claims court in their home jurisdiction.
Counsel would be well-advised to consider Difederico and other similar cases (for example: Williams v. Amazon.com Inc., 2023 BCCA 314 and Petty v. Niantic Inc., 2023 BCCA 315) when crafting dispute resolution clauses in standard-form agreements.
James Plotkin is a partner at Gowling WLG LLP. His practice focuses on international and domestic arbitration, intellectual property litigation and administrative law. He has published broadly on arbitration-related topics and co-authors a commentary on Ontario’s international and domestic arbitration legislation.
