April 22, 2020

Can an “Episodic” Price-Fixing Conspiracy be Certified as a Class Action?

Many price-fixing class actions allege a reasonably uniform conspiracy. The stereotypical scenario alleged is that executives from different companies meet in a dark, smoke-filled room and agree to raise prices or restrain output in some uniform fashion. While that is an oversimplification, and reality is always much more complex, the basic core of most price-fixing allegations is that there was a uniform conspiracy that impacted all, or at least most, consumers in a broadly similar way. This is what has made so many price-fixing class actions amenable to certification.

The recent decision of the Ontario Superior Court of Justice in Mancinelli v Royal Bank of Canada is very different. In that case, the price-fixing conspiracy was accepted to be episodic. That type of case raises unique circumstances and challenges for all parties.


By way of background, the allegation in this case was that representatives of various financial institutions that participated in the foreign exchange (“FX”) market conspired from time to time to manipulate the exchange rates at which those transactions took place. The alleged effect of this manipulation was to increase the spread between the bid and the ask prices offered to customers looking to exchange one currency for another.

Contrary to most price-fixing class actions, the allegation was not that all or even most FX trades were affected by any type of systematic price-fixing conspiracy. Rather, the allegation was that representatives of the defendants communicated with each other through chat rooms to coordinate the prices offered to customers trading in the FX market to manipulate various FX benchmark rates. The allegation was that of an “episodic” conspiracy that impacted a meaningful number of customers, but by no means the entire market. As Justice Perell noted at para 74 of his certification decision, one of the plaintiff’s counsel stated in evidence filed over the court: “…based on Class Counsel’s investigation up to 5% of FX Instrument trades may have been impacted by the alleged wrongful conduct”.

The plaintiffs attempted to have a class action certified on behalf of three different groups of proposed class members:

  • Direct Purchaser from Defendant Class Members—these were entities or individuals who actually purchased foreign exchange through one of the defendants;
  • Direct Purchasers from Non-Defendant Class Members—these were entities or individuals who made foreign exchange transactions with banks, other than the defendants; and
  • Investor Class Members—these were individuals who did not directly make any foreign exchange transactions with any defendants, but rather, held funds where the fund managers would have made transactions with any of the defendants.

The defendants opposed certification of the class action. They took the position that none of the criteria for certification were met.

The Certification Decision

While Justice Perell certified the proceedings as a class action, he did so on the basis of a more limited class than the plaintiffs had sought. In particular, he held that the class could only be certified on the basis of the Direct Purchaser from Defendant Class Members, while the other two subclasses described above failed at the identifiable class criterion as well as at the common issues criterion.

With respect to the Direct Purchasers from Non-Defendant Class Members, Justice Perell rejected the submission that these class members were analogous to umbrella purchasers who were held to have a claim by the Supreme Court in Pioneer Corp v Godfrey:

[201] The case at bar is not like the price-fixing cases involving so-called umbrella purchasers who purchased from non-Defendants the goods whose sale prices were being fixed by the Defendants who controlled the market for those goods. In those cases, the Defendants who dominated the market by their misconduct effectively fixed the prices for the whole market in the goods. In contrast, in the immediate case, the Purchaser Class Members who purchased FX Instruments from non-Defendant banks entered into individually negotiated lawful transactions in which there is no commonality with the Purchaser Class Members who entered into FX transactions with the Defendant banks.

Justice Perell also held that there would be difficulties for class members to identify whether they were in fact part of the class, a problem which precluded certification in Sun‑Rype Products Ltd v Archer Daniels Midland Company:

[202] Further, given the episodic nature of the price-fixing perpetrated by the Defendant banks, it would be an impossible for the Direct Purchaser from non-Defendant Class Members to identify whether at the time they made their purchase of liquidity from a non-Defendant bank their transaction was affected by the unrelated illegal transaction. These customers cannot know whether their individually negotiated transaction with an innocent from collusion bank dealer was affected by the wrongdoing being perpetrated on the Direct Purchaser Class Members by the Defendants. 

[203]  The Direct Purchaser from non-Defendant banks are akin to - but even more remote to the wrongdoing – than the claimants in Sun Rype Products Ltd. v. Archer Daniels Midland Co., who were the potential victims of price-fixing but could not identify themselves as victims because it was impossible to know whether the sweeter purchased for their beverage was a sweeter whose price had been fixed by the Defendants. In Sun Rype Products Ltd., the Supreme Court of Canada decided that there was no identifiable class capable of being certified.

Justice Perell held that there were similar problems with the certification of claims by Investor Class Members.

With respect to the common issues criterion, Justice Perell held there was some basis in fact for issues regarding a conspiracy by the defendant banks to fix prices in the FX market.  He noted the fact that the conspiracy was episodic did not negate the possibility of a class action:

[228] I disagree with the Defendants’ essential argument against commonality. An episodic conspiracy can and does raise common issues including the common issue of whether the Defendants’ conspired to agreement to episodically price fix the Spread and the Fix.

[229] It is true that unlike a more run of the mill price-fixing conspiracy where the conspirators use their market position to control prices in the market, the Defendants in the immediate case are only alleged to have price fixed episodically, which is say that they agreed when opportunities arose to price-fix the Spread or the Fix. Should the Plaintiffs prove that allegations at a common issues trial, there would be a substantial advancement in the class proceedings.

Justice Perell acknowledged that the plaintiffs might not be able to establish causation or damages at the common issues trial. However, he held that this was not an impediment to certification of the action as a class proceeding. Rather, he held that, even in that case, a common issues trial “would be a launch pad for individual trials determined on an individual basis causation and quantification of damages”.

The Challenges with Certifying an “Episodic” Conspiracy

Justice Perell’s decision is seemingly an attempt to split the baby. The case was certified as a class action, but only a narrower subclass than what the plaintiffs sought to have certified. Yet, as is often the case, splitting that baby in this fashion created conceptual tensions that are difficult to resolve.

In Justice Perell’s analysis, it would seemingly have been sufficient for the class action to be certified that the only common issue resolved at a common issues trial is whether there was an episodic conspiracy by the defendants to fix FX prices. As he noted at paragraph 234:

[234] A workable methodology is not a sine qua non for the certification of every class action. The alleged price fixing conspiracy in the immediate case of a service (liquidity) has unique features and problems from a conspiracy to price fix a product like DRAM or corn syrup or lithium batteries. Further, it is no obstacle to a class proceeding that the common issues may not be dispositive of the Class Members’ causes of action.  In the immediate case, assuming success at the common issues trial, the Direct Purchaser from Defendant Class Members, unlike the Direct Purchaser from non-Defendant Class Members, will know at least that they entered into a FX Instrument transaction that may have been price fixed. Assuming success at the common issues trial and noting that the Defendants’ liability for conspiracy is a joint and several liability, the position of the Direct Purchaser from Defendant Class Members would be that they established potential liability for all those Defendants proven to have been co-conspirators and the Direct Purchaser Class Members would have proven general causation of harm.

Put simply, the fact that the conspiracy was episodic and that not all class members would have suffered the effects of that conspiracy would not, in Justice Perell’s view, have been a bar to certification. Rather, in his view, the conclusion that there was in fact a conspiracy with respect to FX exchange rates generally was a sufficient amount of commonality for the class to be certified.

This would be a slim basis for a class proceeding. In a typical price-fixing class action, the debate at certification is often whether the plaintiffs have a methodology to prove loss on a class-wide basis. In circumstances where there is a generalized conspiracy, and all class members have suffered loss, a class action will generally make sense. Justice Perell’s decision here is two steps removed from that archetype of a price-fixing class action. 

The implication of his decision is not just that a price-fixing class action can be certified without class-wide harm being suffered—a reasonable debate can be had about whether that makes sense. Rather, the implication of his decision is that a price-fixing class action can be certified where the vast majority of class members not only suffered no loss, but rather they were never even the subject of the conspiratorial conduct. The situation in this case is not the more familiar one where parties had agreed to have a common conspiracy, but certain class members suffered no loss because the conspirator “cheated” from the conspiracy and gave lower prices to those class members. Rather, this situation is one in which, based on the findings in Justice Perell’s decision, the overwhelming majority of the class engaged in trades that were subject to normal market dynamics without the defendants ever even considering conspiring with respect to those particular trades.  

While Justice Perell alludes to notions of general causation and specific causation that often come into play in tort cases (particularly in pharmaceutical and medical devices product liability cases), that analogy is unhelpful. In those cases, a conclusion on general causation may meaningfully advance a case, since 1) an affirmative answer on general causation will establish that all class members were at least exposed to a risk of harm, and 2) a quantitative conclusion on general causation can meaningfully help answer the question of specific causation.

The present circumstances, on Justice Perell’s description at paragraph 234, are very different. In this case, even if plaintiffs were able to prove at a common issues trial that there was a conspiracy that affected some trades, such a finding would do very little to advance class members’ positions. Those class members would still have to prove that there was a conspiracy with respect to their own trades. The fact that there might have been a conspiracy with respect to other class members’ trades would do nothing to advance their position. (I pause to note that in making these comments, I am reacting to Justice Perell’s characterization about what the value of certification would be even if there was no class-wide loss. Certainly, the decision at paragraph 230 seems to suggest that plaintiffs’ counsel did lead expert evidence of class-wide loss. To the extent that the evidence did provide a methodology for calculating class-wide loss or of easily identifying a sub-class that suffered loss, the criticisms above would not apply. The reactions in this post are not to that evidence, but rather to Justice Perell’s comments at paragraph 234 that imply that the existence of an episodic conspiracy would be sufficient to certify the class proceeding.)

This decision highlights again, as I have argued previously, that the bar for certification should not be too low. Even if certification is not the forum for litigating the merits of the case, it should be the venue for determining whether a case can meaningfully be heard as a class action and whether issues of real value to the class can be decided on a common basis. This is the very theory of the certification process. If the standard for certification is so low that virtually every issue of substance is either punted to the trial judge or left for individual issues trials, then there is seemingly little point in having a certification process. In that case, the certification process is not playing any role as a meaningful procedural screening mechanism.