SS&C Technologies Canada Corp. v. Bank of New York Mellon Corp.

Joint Reasons for Judgment — Côté and Moreau JJ. (Wagner C.J. and Karakatsanis, Rowe, Martin, Kasirer, Jamal and O’Bonsawin JJ. concurring)

I. Overview

1Spoliation refers to the intentional destruction, alteration, mutilation, or concealment of evidence with a view to subverting the truth-finding process during litigation. It is a creature of abuse of process that comes in various degrees and forms (see Trillium Power Wind Corp. v. Ontario (Ministry of Natural Resources), 2023 ONCA 412, 167 O.R. (3d) 321, at para. 22). The common thread unifying all findings of spoliation is a manifest disrespect for the trial process, the truth-seeking function of the court, and the justice system generally. Parties who commit spoliation should not be rewarded for the evidentiary vacuums they create. Rather, as the authors of their own misfortune, spoliators must carry the presumption that the impugned evidence would be unfavourable to their case.

2This Court has not addressed the doctrine of spoliation since its seminal decision in St. Louis v. The Queen (1896), 25 S.C.R. 649. Since St. Louis was decided over a century ago, technology has become ubiquitous and demands on the civil justice system have evolved. Electronic document management is now the primary mechanism for record-keeping. Where once spoliation took the form of physical destruction of documents, in the digital age, spoliation is far more inconspicuous, occurring instantaneously at the click of a button (see G. Christian, “A ‘Century’ Overdue: Revisiting the Doctrine of Spoliation in the Age of Electronic Documents” (2022), 59 Alta. L. Rev. 901, at p. 902). This appeal invites the Court to clarify the doctrine of spoliation, from its principles to its practical function in modern civil litigation.

3The appellant, SS&C Technologies Canada Corporation (“SS&C”), is a Canadian data provider that acquires and processes specialized data used to generate pricing information for financial services companies to value securities. The respondent, the Bank of New York Mellon Corporation (“BNY”), is the world’s largest custodial bank, holding over US$35 trillion in assets under custody.

4The parties took diametrically opposed positions at every stage of the litigation. This appeal is no exception. At trial, while not making an express finding of spoliation, the trial judge drew adverse inferences against BNY based on its failure to produce and preserve relevant documents in the face of contemplated litigation. On appeal, the Court of Appeal for Ontario did not interfere with the adverse inferences drawn by the trial judge and made an express finding of spoliation against BNY. As these reasons explain, once a party proves spoliation, a mandatory presumption arises that the destroyed evidence would be unfavourable to the spoliator’s case. How this presumption is applied is discretionary and fact-specific. In this case, we would not disturb the Court of Appeal’s finding that BNY committed spoliation. However, the trial judge made errors of law and committed palpable and overriding errors in his assessment of damages that this Court cannot uphold.

5Assessing damages that properly compensate for a breach of contract can be a difficult and inexact art. Evidence is often lacking, and, when it is available, the nature of the loss itself may make damages hard to quantify. For this reason, damage awards need not be perfect. It is sufficient that courts do the best that they can with the rules governing evidence and with the evidence before them. However, this pragmatic standard does not allow judges to craft awards that are untethered from the evidence or loss in question. Damage awards must be connected to the case before the court. Without this connection, they cannot be justified.

6The difficulty of assessing damages is intensified when evidence that should be before the court is not, whether by simple oversight, non-production, or destruction. When the absence of evidence is the result of unjustified non-production or destruction — when there has been spoliation — courts must level the playing field. The absence of this evidence cannot accrue to the benefit of the party that flouts the administration of justice. Courts must be able to fill the evidentiary gap and adjudicate the matter before them.

7Both the issue of spoliation and the difficulties of crafting damage awards are at the heart of this case. The trial judge found that BNY redistributed proprietary market pricing data provided to it by SS&C, without SS&C’s authorization. This redistribution breached the clear terms of BNY’s contract with SS&C and, in effect, amounted to the conversion of SS&C’s proprietary data. It was not an isolated or one-time incident. Rather, BNY may have provided all of SS&C’s proprietary data to as many as 65 BNY-affiliated entities over the course of 17 years. It has been established that BNY not only redistributed this proprietary data, but also received payments from its affiliates for this redistribution. As a result, BNY repeatedly and knowingly breached the core of its contract with SS&C.

8When SS&C discovered this scheme, it explicitly requested that BNY preserve information with respect to how its data had been redistributed, to which affiliates its data had been redistributed, and how much BNY had earned from the redistribution. BNY refused to comply. It gave no justification for its non-production. Recognizing this failure to produce the data, the trial judge drew two inferences against BNY. First, he found “that the unaccounted for data was used by unauthorized entities within the BNY group other than CIBC Mellon” (2023 ONSC 4083 (“ruling on damages”), at para. 107).^[1] Second, he declined “to conclude that non-CIBC Mellon users made only de minimis use of the data” (para. 108). The Court of Appeal accepted these inferences and found that BNY intentionally refused to produce the requested data in order to suppress the truth in this litigation. It held that BNY engaged in spoliation and that its conduct “smack[ed] of contempt for the justice system” (2024 ONCA 675, 174 O.R. (3d) 410, at para. 165).

9In drawing these inferences, the trial judge failed to consider the scope and impact of the destroyed evidence. As a result, the two inferences he drew were an incomplete response to the spoliation that occurred. These inferences failed to level the evidentiary playing field and failed to provide a factual basis on which the trial judge could award damages. Instead, they simply affirmed what was already clear on the available evidence and argument. When spoliation has been established, there is no discretion — courts must draw an inference that is adverse to the spoliator and that is capable of filling the gap left in the evidence.

10The finding of BNY’s spoliation has not been appealed before our Court. What has been appealed is the award of damages made in favour of SS&C at first instance and affirmed by the Court of Appeal. Those damages are based on what the trial judge called a “rateable approach”, which both parties vociferously rejected. The trial judge found that BNY was able to account for less than half of the data that it received from SS&C. He inferred that the data BNY could not account for had been shared with unauthorized entities. Because he found that 44.6 percent of the data was accounted for, he was of the view that BNY had to pay for the remaining 55.4 percent of the data at what he understood to be the rate provided for in the agreement between SS&C and BNY. This led to a damage award of US$5,696,850.

11The trial judge justified his use of the rateable approach on the basis that a “third group of entities would have been prepared to pay a similar price for directionally similar data in directionally similar volumes” (ruling on damages, at para. 190). The rateable approach is meant to apply centralized pricing to the unaccounted-for data at the same rate that BNY was prepared to pay for the data. It is premised on the assumption that only the unaccounted-for data was shared and that the data was shared only once. As the Court of Appeal put it, the approach was based on “his conclusion that whatever data that was unaccounted for should be deemed to have been wrongfully shared” (para. 167).

12While we acknowledge that the trial judge was operating with poor evidence as a result of BNY’s spoliation, his damage award is undermined by the incomplete inferences he drew in response to BNY’s spoliating conduct and, as a result, relies on assumptions and calculations that are untethered from the evidence. First, the damage award had no connection with the contractual breaches found or with the incomplete inferences drawn by the trial judge. In particular, it failed to account for: (1) the number of entities that the data may have been shared with; and (2) the number of times that the data may have been shared. The trial judge awarded damages as if only one unauthorized entity accessed the data once — despite 65 unauthorized entities having had access to the data for years. Second, the trial judge grounded his award in the use of centralized pricing in a way that is not supported in the record. Third, despite his suggestions to the contrary, he created an entirely arbitrary rate for pricing the unaccounted-for data. Fourth, in crafting the award on this basis, the trial judge relied on a distinction between accounted-for and unaccounted-for data that bears little relationship to whether the data was improperly shared in breach of the contract.

13In light of the cumulative impact of the trial judge’s errors, we would allow the appeal. There is no doubt that SS&C should be compensated for the proprietary data that BNY wrongly shared with up to 65 unauthorized entities over 17 years, contrary to the express contractual agreement between the parties. Therefore, we are of the view that a new hearing is required on the specific issue of the quantification of the damages owed to SS&C based on adverse inferences that support concrete factual findings capable of setting the value of SS&C’s loss. The issue of damages is to be decided on the existing record, subject to the trial judge’s discretion to permit the parties to adduce additional evidence, particularly expert evidence.

II. Facts

A. The Parties

14The parties are both highly sophisticated and either affiliated with, or are, publicly traded entities.

15SS&C is the world’s largest hedge fund and private equity administrator and mutual fund transfer agency. SS&C has nearly US$45 trillion in managed assets.

16BNY is a global financial services firm. While it is headquartered in New York, its operations extend internationally through its many affiliates, subsidiaries, and joint ventures. BNY’s global custody business services its Canadian clients through CIBC Mellon Global Securities Services Company (“CIBC Mellon”), which is a Canadian joint venture that was previously a party to this litigation.

B. The Licensing Arrangement

17SS&C licenses its data to financial institutions like BNY through licensing agreements. These licensing agreements entitle the purchaser-client to order specialized data from SS&C; SS&C sends this data to allow the purchaser-client to value the accounts of its customers holding a specific security. The licensing agreement in this case was volume-based and operated on a per security basis — BNY would pay SS&C once for the data pertaining to a specific security, but the pricing information could be valuable to any number of customers holding accounts with BNY. In other words, the fee arrangement was contingent on the number of pricing data requests, not how many times the data was utilized to price individual client assets.

18Subject to the terms of individual licensing agreements, SS&C’s purchaser-clients are generally prohibited from sharing, redistributing, or selling SS&C-generated data to other parties, including purchaser-client subsidiaries and affiliates. This was the case as between SS&C and BNY. The proprietary nature of SS&C’s data, and the ease with which it can be shared, underscores the importance of the licensing scheme to SS&C’s business operations. Only the purchaser-client with which SS&C conducts its business is licensed to use the data it receives.

C. The Underlying Contractual Dispute

19The genesis of the parties’ relationship dates back to 1999, when SS&C’s predecessor corporation, Securities Valuation Company Inc. (“SVC”), entered into two separate data services agreements with “Mellon Trust” (“Mellon Trust Agreement”) and CIBC Mellon (“CIBC Mellon Agreement”). The trial judge found that the agreement with “Mellon Trust” actually bound Mellon Financial Corporation. Indeed, “Mellon Trust” was not a legal entity, but rather the brand name under which Mellon Financial Corporation operated its custodial business.

20In 2005, SS&C acquired SVC, thus absorbing SVC’s rights under both the Mellon Trust Agreement and the CIBC Mellon Agreement. In 2007, Mellon Financial Corporation and the Bank of New York merged to form BNY. Through these transactions, SS&C and BNY became the relevant parties to the Mellon Trust Agreement.

21The Mellon Trust Agreement remained in place for nearly 17 years, BNY having paid SS&C a total of approximately US$4.6 million over the life of the agreement.

22The CIBC Mellon Agreement remained in place until 2011, when CIBC Mellon terminated the agreement, having paid SS&C approximately C$3.1 million from 2000 to 2011. The reason given by CIBC Mellon for termination was that it “no longer require[d] the services provided under the Agreement” (2021 ONSC 2657, 18 B.L.R. (6th) 113 (“ruling on liability”), at para. 54). The trial judge did not accept this reason and found that although CIBC Mellon continued to require SS&C’s data services, it began receiving the data from BNY for free without disclosing this arrangement to SS&C. In doing so, CIBC Mellon circumvented SS&C’s licensing scheme.

23It was not until October 2016 that SS&C discovered CIBC Mellon’s use of the data under the Mellon Trust Agreement. A system glitch in the delivery of data from SS&C was the catalyst that led to this discovery. CIBC Mellon wrote to SS&C complaining of the glitch and inadvertently revealed that it had been receiving SS&C’s data from BNY. This led a CIBC Mellon employee to write to SS&C stating, “I just realized the point of your note being that we may be receiving information that we did not pay for” (ruling on liability, at para. 55). This discovery was the breaking point in the parties’ relationship, with SS&C terminating the Mellon Trust Agreement and commencing this litigation in 2017. SS&C sued BNY for breach of contract.

III. Decisions Below

A. Ontario Superior Court of Justice Decisions

24The procedural history of this lawsuit is protracted and complex; the two most relevant decisions from the trial judge are his separate rulings on liability and damages.

(1) Ontario Superior Court of Justice, 2021 ONSC 2657, 18 B.L.R. (6th) 113 (Koehnen J.) — Ruling on Liability

25At trial, SS&C argued that BNY breached its licensing agreement by sharing data with other affiliated entities within the larger corporate group. The interpretations of the licensing agreements were therefore a central issue at trial. BNY argued that the Mellon Trust Agreement was a “line of business agreement” and the CIBC Mellon Agreement was an “enterprise agreement” (ruling on liability, paras. 15-17). The difference between a line of business agreement and an enterprise agreement is that the former applies to an entire line of business within a corporate family, whereas the latter applies to an individual entity or entities referred to in the contract. BNY admitted to sharing data within its corporate family, claiming it was legally entitled to do so under the Mellon Trust Agreement. BNY maintained that by entering into a licensing agreement with the brand name “Mellon Trust”, SS&C intended that its data be shared across the entire corporate group under the Mellon Trust brand.

26The trial judge rejected BNY’s argument. Since BNY could not point to any language in the Mellon Trust Agreement to support the notion that it was a line of business agreement applicable to the Mellon Trust corporate family, the trial judge concluded that the language of the Mellon Trust Agreement, the factual matrix, and the conduct of the parties supported the interpretation that it was intended to apply to a single legal entity.

27The trial judge did, however, find that a subset of the Mellon Financial Corporation entities was captured under the Mellon Trust Agreement (see 2022 ONSC 1652, 31 B.L.R. (6th) 112 (“clarification reasons”), at para. 16). Since Mellon Financial Corporation was a holding company that would have no use for SS&C’s data, the trial judge held that a commercially reasonable interpretation of the contract was that the Mellon Trust Agreement contemplated data sharing amongst the custodial entities of Mellon Financial Corporation that were in existence at the time the predecessor parties entered into the contract. This conclusion did not apply to CIBC Mellon, however, as it entered into a separate agreement with SS&C that granted it access to different data.

28In sum, the trial judge concluded that the data provided by SS&C was meant for use only by the entity named in the relevant contract. In making SS&C’s data available to other affiliated entities of Mellon Financial Corporation that existed after 1999, BNY breached the Mellon Trust Agreement.

29The trial judge ordered a separate trial on damages.

(2) Ontario Superior Court of Justice, 2023 ONSC 4083 (Koehnen J.) — Ruling on Damages

30SS&C argued that its damages should equate to the monthly price under the Mellon Trust Agreement multiplied by the number of unauthorized entities that may have accessed its data through BNY. This approach assumed that each entity within the BNY family that was in the asset servicing business in a given month had access to SS&C’s data and would have paid the same monthly fee. This approach resulted in SS&C’s principal damages claim in the total amount of US$150,420,564. What raised this principal amount to SS&C’s global damages figure of US$889,752,087 is the late payment fee, equivalent to 1.5 percent compounded monthly, in lieu of pre-judgment interest. The late payment fee is expressly provided for in the Mellon Trust Agreement.

31BNY submitted that SS&C did not sustain any damages or, at most, nominal damages, because the cost of data had decreased since the predecessor parties entered into the Mellon Trust Agreement. BNY argued that SS&C’s theory of damages was unrealistic — if it was faced with paying an equivalent fee for all of the other affiliates with which it had shared data, it would have terminated the contract. In particular, BNY noted, without bringing the alleged contract into evidence, that when SS&C terminated the Mellon Trust Agreement, BNY entered into a subsequent data licensing arrangement at a 25% discount with another provider.

(a) Document Retention and Production

32Impropriety in document retention and production was a key issue at the trial for damages. The central allegation relevant to this appeal stems from BNY’s failure to preserve usage data showing the extent to which it shared SS&C’s data with unauthorized affiliates, despite SS&C having expressly requested a full accounting of this information.

33SS&C advised BNY of the need to preserve documents pertaining to data usage on December 23, 2016: “SS&C demands a full accounting by BNY of any and all affiliated entities and third parties with whom BNY shared Data. SS&C will seek to conduct a full audit of BNY’s use and unauthorized sharing of Data” (ruling on damages, at para. 95). On February 3, 2017, SS&C restated its demands to BNY:

34Even though SS&C learned of BNY’s breach in November 2016, it continued to supply market data until February 28, 2017. It was only when BNY refused SS&C’s request for a full accounting that SS&C terminated the Mellon Trust Agreement and filed its notice of application.

35BNY explained its limited data disclosure with reference to its data retention practices. SS&C data was uploaded to two BNY databases: InvestOne and the Institutional Accounting System (“IAS”). Data on InvestOne is removed after 30 days then destroyed and data on IAS is stored for two weeks before it is moved to backup storage. SS&C challenged BNY’s reliance on its retention policy, asserting that had BNY preserved the usage data as of December 23, 2016, when SS&C first advised it of the need to do so, the court would have had a full record of data usage between November 2016 and February 28, 2017. BNY offered no explanation for its failure to preserve the relevant documentation in the face of reasonably contemplated litigation. The trial judge was “satisfied that BNY knew or ought to have known that the evidence SS & C was asking it to preserve was required for litigation” (ruling on damages, at para. 99).

36SS&C submitted that BNY committed the “tort of spoliation” and asked the trial judge to draw an adverse inference “that each of BNY’s 65 asset servicing entities received SS & C data and should pay the same fee for that data as was paid under the Mellon Trust agreement” (ruling on damages, at para. 100). The trial judge declined to pronounce on whether spoliation as an independent tort applied in this case. He held that the court may draw adverse inferences from the failure to produce documents absent spoliation, as these types of adverse inferences “have nothing to do with spoliation but have to do with a party’s failure to produce evidence to support a proposition that it is advancing” (paras. 104-5).

37The trial judge drew two adverse inferences:

(b) The Trial Judge’s Assessment of Damages

38The trial judge rejected both theories of damages advanced by the parties, finding that neither party had provided the court with a sufficient record to assess damages. SS&C did not provide detailed information about the basis on which fees are typically calculated when dealing with a multi-enterprise entity like BNY. At the same time, BNY failed to provide a full picture about the use of SS&C’s data within the BNY family.

39In rejecting each party’s theory of damages, the trial judge determined that he was able to quantify damages based on the record. The trial judge invited submissions on his proposed approach. Though each party disagreed with the trial judge’s approach, he nevertheless concluded that when compared to the parties’ theories, it was the most commercially reasonable outcome.

40The trial judge divided the issue of damages into two categories: (1) for data BNY shared with CIBC Mellon; and (2) for data BNY shared with other affiliates. We discuss each of these categories below.

(i) Data Shared With CIBC Mellon

41From 1999 to 2011, CIBC Mellon had its own relationship and agreement with SS&C over the provision of specialized data. The trial judge assessed damages to SS&C from the time between the CIBC Mellon Agreement was terminated (April 1, 2011) to the time the Mellon Trust Agreement was terminated (February 28, 2017) to account for the unauthorized sharing of data between BNY and CIBC Mellon. He multiplied this timeframe by a monthly fee equal to the average monthly fee that CIBC Mellon paid during the last three months of its agreement. This damages figure was subject to two further adjustments. First, the trial judge added the average monthly fee from a contract CIBC Mellon negotiated to provide limited data access to an affiliate. Second, he made an upward adjustment for annual price increases that were contemplated in the CIBC Mellon Agreement at SS&C’s discretion. Using the price increase under the Mellon Trust Agreement as a proxy, the trial judge extrapolated a reasonable annual price increase for the CIBC Mellon Agreement. The trial judge arrived at a quantum of C$922,887 for this head of damages.

42The trial judge concluded that this approach was tethered to the record. Not only did it reflect commercial realities, but it extended the approach the parties had agreed to in having two separate agreements for 12 years and extended it to the time SS&C discovered the breach and terminated the Mellon Trust Agreement.

(ii) Data Shared With Affiliates Other Than CIBC Mellon

43BNY asserted that CIBC Mellon was the primary user of SS&C’s data and its use by other unauthorized affiliates was minimal.

44Due to a sparse evidentiary record, the trial judge was unable to accept BNY’s contention that CIBC Mellon was the overwhelming user of the unauthorized data. The trial judge provided several reasons for this conclusion. First, BNY relied heavily on an affidavit from one of its principal witnesses, Matthew McDonnell, who was BNY’s Managing Director and U.S. Group Head of Securities Data Management at the time of trial. Mr. McDonnell’s affidavit provided information about the use of SS&C’s securities quotations between 2007 and 2016, showing that approximately 95 percent of the quotations related to Canadian fixed income prices, of which CIBC Mellon was the primary user. Despite both Mr. McDonnell and BNY’s counsel having agreed that the SS&C securities quotations were uploaded onto two databases, IAS and InvestOne, the trial judge determined that Mr. McDonnell’s affidavit was inaccurate in that it only considered data from the IAS database, a fact that BNY admitted nearly three years after SS&C launched its application. The analysis in Mr. McDonnell’s affidavit demonstrated that it was missing a significant number of securities quotations that SS&C delivered.

45Second, when Mr. McDonnell encountered a security that was held in both a CIBC Mellon account and in another affiliate’s account, he categorized it as a CIBC Mellon security for the purposes of his affidavit, thereby skewing the proportion of CIBC Mellon’s usage.

46Third, BNY tried to explain this discrepancy in data by submitting that Mr. McDonnell’s affidavit did not reflect InvestOne data, which was where the missing data was stored. When BNY produced the InvestOne data, it did not resolve the discrepancy. BNY later revealed that it did not search the entire InvestOne platform, but instead, searched only one single user bank, user bank C5, which was reserved for CIBC Mellon accounts. BNY submitted that it limited its search because it assumed that only CIBC Mellon used the SS&C data. The trial judge labelled BNY’s circular analysis as “so replete with flaws as to be entirely unreliable” and “a bald assertion masquerading as an inquiry” (ruling on damages, at para. 50).

47The trial judge found that BNY could not account for more than one-half of the data that SS&C delivered. SS&C invoiced about 16 million securities, however BNY could only account for 7.132 million quotations in its records. This left 8.875 million missing quotations, or 55.4 percent of the total quotations SS&C provided.

48The trial judge calculated damages at US$5,696,850 for BNY’s unauthorized data sharing with affiliated entities using a “rateable approach”. The rateable approach represents the crux of the disagreement on appeal.

(iii) The Rateable Approach

49The rateable approach fixes damages for sharing data with unauthorized entities other than CIBC Mellon at a rateable price for data that BNY could not account for by reference to the price BNY paid for data that it could account for. The trial judge found that the total fees paid by BNY under the Mellon Trust Agreement, for which BNY could only account for 44.6 percent of the data SS&C provided, amounted to US$4,586,273. He then deemed that the 44.6 percent of the data should be valued at the price of the full contract: US$4,586,273. The trial judge concluded that 100 percent of the data would attract a fee of US$10,283,123. The fee associated with the remaining 55.4 percent of data would therefore amount to the difference between US$10,283,123 and US$4,586,273, or US$5,696,850, which the trial judge fixed as the quantum of SS&C’s damages.

50BNY disputed the use of the rateable approach and proposed three alternative damages models. SS&C also disputed the use of the rateable approach, arguing that it was contrary to business convention for structuring licences, effectively awarding SS&C the same amount in damages whether BNY used market data to service one or multiple unauthorized affiliates. SS&C also argued that the rateable approach was a novel innovation that was contrary to the generally accepted valuation principle that uses existing agreements as a litmus test to measure damages. The trial judge disagreed, finding that “neither party led evidence about the conceptual basis on which data contracts are priced” (ruling on damages, at para. 188). The trial judge asserted this his approach operated on the basis “that a third group of entities would have been prepared to pay a similar price for directionally similar data in directionally similar volumes” and that it is “more consistent with the evidence about centralized pricing being a directional norm” (para. 190).

51The trial judge thus arrived at a quantum of US$5,696,850 to account for the unauthorized sharing of data between BNY and affiliated entities other than CIBC Mellon.

B. Court of Appeal for Ontario, 2024 ONCA 675, 174 O.R. (3d) 410 (Tulloch C.J.O. and Hourigan and Miller JJ.A.)

52On the issue of liability, the Court of Appeal dismissed BNY’s appeal and allowed SS&C’s cross-appeal. On the issue of damages, the Court of Appeal allowed BNY’s appeal in part. The Court of Appeal ultimately held that the Mellon Trust Agreement only authorized BNY to access SS&C’s data, as opposed to affiliates that BNY’s predecessor owned when it entered into the contract. The Court of Appeal also accepted the trial judge’s rateable approach to calculating damages, but set aside the portion awarding C$922,887 for CIBC Mellon’s data usage on the basis that it amounted to double-counting.

(1) Appeal on Liability

53BNY appealed the trial judge’s finding of liability on the grounds that he should have interpreted the Mellon Trust Agreement to authorize all BNY custodial affiliates, including CIBC Mellon, to access data under the agreement. SS&C cross-appealed on the ground that the trial judge should have held that the Mellon Trust Agreement only authorized BNY, not the custodial entities under Mellon Financial Corporation as of 1999, to access data.

54The Court of Appeal began its analysis by recognizing that the standard of review is highly deferential, as the agreement at issue was the result of negotiation between two sophisticated parties. Central to the liability appeal was the determination of which entities were considered the “Client” under the Mellon Trust Agreement, BNY or BNY and its affiliated entities. The Court of Appeal ultimately agreed with SS&C that the Mellon Trust Agreement only authorized BNY, not its affiliated entities that existed at the time of contracting, to access data under the agreement.

55The Court of Appeal determined that the trial judge made all the requisite findings supporting this conclusion: he found that the term “Client”, which the contract defined as “Mellon Trust”, referred to a single entity; the commercial purpose of the agreement was to ensure, as a data distributor, that SS&C was able to control a client’s distribution of data to protect its business model; the text of the contract reinforced the interpretation that a “Client” could not share SS&C’s data with any parties other than its customers; and the conduct of the parties pointed to BNY, instead of an entire corporate family, as the sole contracting party. Applying the principles of contractual interpretation should have led the trial judge to the conclusion that the contract only authorized BNY to receive SS&C’s data, not all of the Mellon Financial Corporation entities as they existed in 1999. This was precisely the trial judge’s extricable legal error, which was compounded by his factual error in finding that Mellon Financial Corporation itself did not require any data under the agreement. Simply because Mellon Financial Corporation was a holding company did not necessarily mean it was a shell company — it had its own market data employees and it provided services.

56The Court of Appeal therefore dismissed BNY’s appeal on liability, allowed SS&C’s corresponding cross-appeal, and varied the trial judgment by adding a declaration that no entities other than Mellon Financial Corporation and BNY were entitled to access the data SS&C provided under the Mellon Trust Agreement.

(2) Appeal on Damages

57SS&C took the position that it suffered US$889,752,087 in damages, while BNY asserted that SS&C was not entitled to any damages. Save for one aspect of the damages award, the Court of Appeal did not interfere with the trial judge’s rateable approach.

58The Court of Appeal concluded that the trial judge erred in awarding damages in relation to CIBC Mellon’s use of the unauthorized data as the damages award in addition to the contractual payments constituted double compensation. In other words, because the trial judge found that SS&C was already paid for CIBC Mellon’s use of 44.6 percent of the data under the Mellon Trust Agreement, SS&C was not entitled to damages under the CIBC Mellon Agreement as well. This finding is not under appeal to our Court.

59While not interfering with the damages awarded by the trial judge under the Mellon Trust Agreement, the Court of Appeal made a positive finding of spoliation by BNY. The Court of Appeal relied on the three-part test for spoliation set out in Stamatopoulos v. The Regional Municipality of Durham, 2019 ONSC 603, 85 M.P.L.R. (5th) 31, at para. 606, aff’d 2022 ONCA 179, 26 M.P.L.R. (6th) 1:

60The Court of Appeal noted that an intentional act consists of two elements: (1) the destruction of the document must be intentional and (2) the alleging party must prove a mala fides desire to prevent the use of the document in litigation, suppress the truth, and thus impact the outcome of the litigation (citing Christian, at pp. 911-12).

61Important to this case was the preservation notice SS&C delivered to BNY on December 23, 2016, after discovering the unauthorized usage of its data. The notice made SS&C’s position clear with respect to contemplated litigation and removed any doubt that BNY was under an obligation to preserve the data as relevant evidence. BNY’s former counsel responded to SS&C’s notice, stating that because CIBC Mellon was the assumed primary user of the data, “BNY Mellon categorically denie[d] that a breach of the Agreement ha[d] occurred” and declined to comply with the preservation notice (C.A. reasons, at para. 158).

62After a thorough review of the evidence, the trial judge determined that because the use by unauthorized entities other than CIBC Mellon was not de minimis, BNY’s rejection of SS&C’s preservation notice was not valid. More concerning was BNY’s disregard of its obligation to preserve relevant evidence in the face of litigation based on its own view of SS&C’s claim. The Court of Appeal noted: “It is not open to lawyers or parties to ignore their obligations under the Rules of Civil Procedure, R.R.O. 1990, Reg. 194, and at common law based on their opinion of the merits of a potential claim” (para. 159).

63The Court of Appeal found that to the extent that the trial judge held that the remedy for spoliation was limited to damages, and that adverse inferences were not available remedies where spoliation is proved, he erred in law. Regardless of whether spoliation is an evidentiary rule or an independent tort, an adverse inference is an available finding.

64The Court of Appeal had “no doubt” that spoliation had been proven (para. 163). BNY failed to preserve relevant data and it chose to ensure that the data would not be available in any subsequent legal proceeding. The trial judge was therefore justified in drawing adverse inferences, which the Court of Appeal noted to be a highly discretionary decision.

65The Court of Appeal did not interfere with the adverse inferences the trial judge drew. Despite SS&C reasserting its dispute with the rateable approach on appeal, the Court of Appeal found that the thrust of SS&C’s argument was to use a different damages calculation that imposed a series of hypothetical agreements with substantially the same terms as the Mellon Trust Agreement. The Court of Appeal held that the trial judge correctly rejected this approach, citing the lack of evidence supporting the proposition that a multi-enterprise entity would enter into these hypothetical agreements. Relatedly, the trial judge did not err in rejecting SS&C’s position that BNY would pay 65 times more for the supply of data by entering into these parallel agreements with the unauthorized entities, as the onus was on SS&C to produce evidence supporting this claim.

66Accordingly, the Court of Appeal dismissed the appeal and cross-appeal on damages, except with respect to the portion of the trial judgment fixing CIBC Mellon damages, which it reduced to $0.

IV. Issues

67This appeal raises the following issues:

V. Analysis

68The analysis will proceed in three parts. First, we will explain the historical underpinning of the doctrine of spoliation. Second, we will outline the proper test for proving spoliation, which necessitates a discussion of policy, scope, content, and corresponding remedies. Third, we will apply these principles to the facts of this case.

A. Historical Roots of the Doctrine of Spoliation

(1) Background

69The doctrine of spoliation dates back to the ancient Roman law of Justinian, when the Latin maxim omnia praesumuntur contra spoliatorem (“all things are presumed against the spoliator”) came to the fore (J. Cassels and C. Jones, The Law of Large-Scale Claims: Product Liability, Mass Torts, and Complex Litigation in Canada (2005), at p. 262; St. Louis, at p. 667). Spoliation originated in the context of loan transactions, as certain business records were considered prima facie proof of a loan (Cassels and Jones, at p. 262). The Romans applied the maxim with severity, as failing to produce such business records would not only cause claims to fail, but it would also amount to fraud upon the defendant (p. 262; see also St. Louis, at p. 667).

70The rigidity with which Roman law applied the maxim was not readily adopted by other legal systems (St. Louis, at p. 667). In the Canadian context, Girouard J. in St. Louis noted that spoliation, and illustrations of the maxim, were few and far between, which is still the case today (pp. 682-83; Cassels and Jones, at p. 262).

71From its origins, spoliation led to an evidentiary presumption: “When it was proved that a litigant had suppressed or destroyed evidence, the opposing litigant could rely on a presumption, generated by the fact of that destruction or suppression, that the spoiled evidence was harmful to the spoliating litigant’s case” (British Columbia Law Institute, BCLI Report No. 34, Report on Spoliation of Evidence (2004), at p. 3, citing Ward v. Apprice (1705), 6 Mod. 264, 87 E.R. 1011, at p. 1011; Armory v. Delamirie (1722), 1 Strange 505, 93 E.R. 664, at p. 664; and James v. Biou (1826), 2 Sim. & St. 600, 57 E.R. 475, at p. 477). The circumstances surrounding the destruction of documents had considerable influence on how the evidentiary presumption operated (BCLI Report, at p. 4).

(2) St. Louis v. The Queen (1896), 25 S.C.R. 649

72St. Louis was an appeal originating in Quebec and, as Girouard J. observed, Quebec law settled the point under dispute (p. 670). Because the Quebec law of evidence was understood to have mixed civil law and common law lineage, the Court considered the appropriate English sources including the treatment of the maxim omnia praesumuntur contra spoliatorem in the common law (p. 676). Justice Girouard’s review of sources beyond Quebec law included an extensive review of English and American law, a practice not uncommon in the Court’s jurisprudence at the time, and this led Girouard J. to conclude that “[t]aking this view of the law and the facts of this case, I have come to the conclusion that under both the Quebec Code and the English law the appellant cannot be regarded as a spoliator, and that even if he could he has fully rebutted the presumption” (pp. 689-90). This no doubt explains some of the currency of this Quebec case in the common law authorities, including recent ones, bearing on spoliation in common law jurisdictions.

73In St. Louis, the appellant contracted with the Department of Railways and Canals for the Dominion of Canada for stone and labour for the construction of railways and bridges (pp. 649-50). The appellant filed a petition of right against the Crown to recover the alleged balance owing on the contracts. The Crown argued spoliation, contending that the appellant fraudulently prepared payroll records that no longer existed, despite these records being the primary basis for his claim (pp. 650-51). The appellant admitted that he burned and destroyed all of his accounting records and time-books relating to these contracts such that none of the original records were available as evidence for trial. The Exchequer Court applied the presumption that the evidence wilfully destroyed by the appellant was harmful to his case. Accordingly, he was not entitled to any relief (p. 651).

74The Supreme Court allowed the appeal, concluding that the Exchequer Court had carried the omnia praesumuntur contra spoliatorem maxim too far (p. 652, per Taschereau J., concurring). Justice Taschereau noted that “[t]he destruction of evidence carries a presumption that the evidence destroyed would have been unfavourable to the party who destroyed it, but that presumption may be rebutted” (pp. 652-53).

75Justice Girouard, writing for the majority, agreed (p. 666). He found that because the operative statute did not create a duty to retain records and the appellant had destroyed his own evidence long before contemplating the court action, he “cannot be regarded as a spoliator, and that even if he could he has fully rebutted the presumption” by adducing “express and positive evidence to the contrary” (pp. 689-90; see also pp. 670 and 683-84).

B. The Modern Approach to Spoliation

(1) Policy Underpinning the Doctrine of Spoliation

76Engaging in spoliation runs contrary to two central tenets of the justice system: “. . . the establishment and maintenance of a fair trial process and the quest for the truth” (BCLI Report, at p. 1). It has been described as “a form of cheating” that “threatens to undermine the integrity of [the] civil trial process” (C. R. Nesson, “Incentives to Spoliate Evidence in Civil Litigation: The Need for Vigorous Judicial Action” (1991), 13 Cardozo L. Rev. 793, at p. 793).

77The sanctity of the civil justice system depends on the honesty of parties and witnesses (see Doust v. Schatz, 2002 SKCA 129, 32 R.F.L. (5th) 317, at para. 27). Parties are duty-bound, through statute and common law, to preserve, disclose, and produce documents relevant to litigation. Discovery in the civil process “is central to the conduct of a fair trial and the destruction of relevant documents undermines the prospect of a fair trial” by striking at the heart of the legal process (para. 27; see also Dreco Energy Services Ltd. v. Wenzel, 2006 ABQB 356, 399 A.R. 166, at para. 47). The central pillar of our legal system is built upon the examination of evidence in the pursuit of truth, and it is only effective if all parties respect the process and disclose all relevant evidence (Dreco, at para. 47).

78The principles underlying spoliation operate from two perspectives. As between the parties to litigation, the doctrine is concerned with ensuring an even playing field and trial fairness (G. Underwood and J. Penner, Electronic Evidence in Canada (loose-leaf), at § 8:1; BMW Canada Inc. v. Autoport Limited, 2021 ONCA 42, 456 D.L.R. (4th) 443, at para. 48). From the court’s perspective, there is concern for the viability and credibility of the fact-finding process (R. J. Sommers and A. G. Seibert, “Intentional Destruction of Evidence: Why Procedural Remedies Are Insufficient” (1999), 78 Can. Bar Rev. 38, at p. 42). As a form of abuse of process, spoliation amounts to egregious conduct that flies in the face of respect for the courts and the rule of law generally.

(2) The Test for Proving Spoliation

79In sum, spoliation refers to the intentional destruction, alteration, mutilation, or concealment of evidence with a view to subverting the truth-finding process during litigation. Once the court makes a finding of spoliation, it presumes that the destroyed evidence would have been unfavourable to the spoliator’s case. The mandatory nature of this presumption recognizes the gravity of spoliating conduct.

80To successfully prove spoliation, the alleging party must prove the following elements on a balance of probabilities: (1) the evidence was intentionally destroyed, altered, mutilated, or concealed; (2) at the time of destruction, litigation was ongoing or reasonably contemplated; (3) the evidence was relevant to said litigation; and (4) it is reasonable to infer that the evidence was destroyed to affect the litigation (see Nova Growth Corp. v. Andrzej Roman Kepinski, 2014 ONSC 2763, at para. 296; Catalyst Capital Group Inc. v. Moyse, 2016 ONSC 5271, 35 C.C.E.L. (4th) 242, at para. 136, aff’d 2018 ONCA 283, 46 C.C.E.L. (4th) 35; McDougall v. Black & Decker Canada Inc., 2008 ABCA 353, 440 A.R. 253, at para. 18; G. D. Cudmore, Civil Evidence Handbook (loose-leaf), at § C:1, § 1; K. A. McKague and E. Lin, “Why Would I Ever Plead Spoliation?” (2023), 54 Advocates’ Q. 93, at p. 94). While the burden of proof rests on the party alleging spoliation, if relevant evidence is unjustifiably destroyed, altered, mutilated, or concealed in the context of reasonably contemplated or ongoing litigation, there should be little difficulty in concluding that this was done to affect the litigation. This test highlights the notion, implicit in the doctrine of spoliation, that at the time the spoliator destroyed the evidence, they were under an existing duty to preserve it (Christian, at p. 907; see also St. Louis, at p. 670).

81The alleged spoliator may challenge any one or more of the above elements, as all four elements of the test are required to make out a finding of spoliation (see Cudmore, at § C:1, § 2.1, citing CMT v. Government of PEI, 2020 PECA 12). If the alleging party successfully establishes the four elements listed above, the court presumes that the destroyed evidence was unfavourable to the spoliator’s case. The burden then shifts to the spoliator to rebut this presumption. The spoliator can do so by showing that the destroyed evidence was not harmful to its case (as was the case in St. Louis). If the spoliator cannot rebut the presumption that the destroyed evidence was harmful to its case, then the presumption becomes mandatory and adverse inferences must be drawn. At this stage, the spoliator may bring evidence to potentially narrow the scope of any adverse inferences drawn against it pursuant to the presumption.

82The consequences of the mandatory presumption associated with a finding of spoliation will depend on the facts of the case and the extent of the prejudice to the non-spoliating party. Even where a party successfully proves spoliation, it does not necessarily follow that the party will succeed on its broader claim.

83Whether negligent destruction of evidence may qualify as spoliation, and whether spoliation should be recognized as an independent tort in Canada, are open questions in Canadian law. Without the benefit of full submissions, we leave these issues for another day.

(3) The Content of the Remedies Associated With a Finding of Spoliation Are Discretionary

84The nature and scope of the remedies associated with a finding of spoliation is the second issue at the heart of this appeal. This section will explain that, once the court makes a finding of spoliation, a rebuttable presumption arises that the destroyed evidence would be unfavourable to the case of the party that destroyed it. The mandatory presumption does not dictate how or to what extent the destroyed evidence would have been unfavourable to the spoliator’s case. This is a discretionary and context-specific determination, usually in the form of an adverse inference, that is “best left to the trial judge who can consider all of the surrounding facts” (McDougall, at para. 4).

85SS&C argues that the current law on spoliation is anachronistic. According to SS&C, once spoliation is made out, there should be no room for judicial discretion. In relying on judicial discretion, the adverse inferences drawn are usually weaker than the highest possible finding. In effect, SS&C seeks to bind trial judges to a presumption that the non-spoliating party has proven their maximum position on damages in every case where spoliation is found.

86BNY argues that SS&C’s appeal to trial fairness is a masked attempt to change the common law. The appropriate remedy for spoliation is discretionary. Because spoliation can occur in a variety of circumstances, the appropriate remedy should be addressed in a manner that is sensitive to the context.

87We agree with BNY. SS&C’s argument is based on an assumption unsupported by evidence that the highest possible remedy is the only effective means for a court to deter spoliators. What SS&C coins as the “uneven” application of remedies for spoliation is not a defect in the law, but rather an intentional choice to allow courts the flexibility to ensure remedies respond to the specific prejudice that arises on the facts.

88Spoliation, as it relates to the civil process, intersects with the rules of court and the court’s inherent power to control its processes, including preventing abuses of process. While the “principal remedy for spoliation is the imposition of a rebuttable presumption of fact that the lost or destroyed evidence would not assist the spoliator” (McDougall, at para. 29), this is usually not the end of the story.

(a) The Practical Operation of the Mandatory Presumption

89How the mandatory presumption practically applies depends on the facts of the case and the trial judge’s determination of how best to ensure fairness in the proceedings. After all, “it cannot be admitted that such a presumption should stand instead of all other evidence, and supply the total deficiency of proof” (St. Louis, at p. 685, citing W. D. Evans, A Treatise on the Law of Obligations, or Contracts, vol. II, Appendix (1806), at p. 169). On the contrary, in determining how best to give effect to the presumption that the destroyed evidence would be unfavourable to the spoliator’s case, trial judges should consider the context, scope, and impact of the destroyed evidence (see generally Endean v. Canadian Red Cross Society (1998), 157 D.L.R. (4th) 465 (B.C.C.A.), at para. 32).

90In so doing, the trial judge must draw an adverse inference against the spoliating party that is capable of filling the gap left by spoliation, though this does not foreclose the possibility that other remedies are available. In most cases, such an adverse inference works to correct the imbalance created by spoliation, to undo some of the prejudice caused, to ensure the proper functioning of the court’s fact-finding role, and to remedy the abuse of the court’s process that spoliation represents (see BMW Canada, at para. 48; Trillium Power Wind, at para. 22; Casbohm v. Winacott Spring Western Star Trucks, 2021 SKCA 21, [2021] 4 W.W.R. 506, at para. 36; Doust, at para. 29).

(b) Remedies Outside of the Evidentiary Presumption

91Some authors have characterized the remedies or sanctions flowing from a finding of spoliation as “almost unlimited” given the court’s broad discretion (see Underwood and Penner, at § 8:3). In addition to the mandatory presumption arising from a finding of spoliation, courts across Canada, exercising their inherent power to remedy abuses of process, have levied a variety of other sanctions to address the distinct prejudice created by spoliation.

92For instance, the “sanctions or remedies available to litigants who suffer due to spoliation include procedural remedies, evidentiary presumptions, contempt proceedings and costs orders. Preventive measures may also be taken through preservation orders” (Holland (Guardian ad Litem of) v. Marshall, 2008 BCCA 468, 301 D.L.R. (4th) 371, at para. 59). At the court’s disposal depending on the case before it is the ability to strike a claim or defence, make an adverse credibility finding against a party’s testimony, order substantial indemnity costs or punitive damages, exclude expert reports, issue an interlocutory injunction, or make a finding of contempt (see, e.g., Brandon Heating and Plumbing (1972) Ltd. v. Max Systems Inc., 2006 MBQB 90, 202 Man. R. (2d) 278; iTrade Finance Inc. v. Webworx Inc. (2005), 255 D.L.R. (4th) 748 (Ont. S.C.J.) (striking a pleading); Doust (adverse credibility finding); Chura v. Batten Industries Inc., 2023 BCSC 1708, 91 C.C.E.L. (4th) 38 (order for special costs); Cheung (Litigation Guardian of) v. Toyota Canada Inc. (2003), 29 C.P.C. (5th) 267 (Ont. S.C.J.) (excluding reliance on reports and other evidence); Western Tank & Lining Ltd. v. Skrobutan, 2006 MBQB 205, 207 Man. R. (2d) 176 (spoliation supporting remedy of interlocutory injunction); Fuller Western Rubber Linings Ltd. v. Spence Corrosion Services Ltd., 2012 ABQB 163, rev’d in part, but not on the finding of contempt, 2012 ABCA 137, 524 A.R. 246 (finding of contempt)).

93This list is not exhaustive.

94In crafting additional remedies, the court may find the following non-exhaustive list of considerations relevant: the level of culpability of the spoliator, the intention or reason behind the destruction of the evidence, the prejudice to the non-spoliating party, and the impact of the destruction of evidence on the court’s ability to fairly dispose of the issues (see generally The Sedona Conference, “The Sedona Canada Principles Addressing Electronic Discovery, Third Edition” (2022), 23 Sedona Conf. J. 161, at p. 313).

(c) The Remedy for a Finding of Spoliation Is Not Necessarily the Highest Possible Inference

95It is beyond doubt that spoliation is a flagrant attack on the civil justice system that flies in the face of its truth-seeking function and the fairness its processes aim to achieve. The penalty for this conduct should be severe, but it should not be blind. In the modern world, spoliation can occur in any number of ways and to varying degrees. To enforce a one-size-fits-all mandatory remedy that imposes the “maximum penalty” may create disproportionate consequences. Trial judges are well placed to determine the extent of the prejudice to a party harmed by spoliation, analyze the magnitude of the spoliating conduct, and fashion a proportionate remedy. Remedial discretion aligns with the jurisprudence, doctrine, provincial legislation, and other common law approaches. We discuss each of these areas below.

(i) Jurisprudence Addressing a Trial Judge’s Remedial Discretion

96The Latin maxim omnia praesumuntur contra spoliatorem (“all things are presumed against the spoliator”) is an organizing principle underpinning the doctrine of spoliation. While SS&C places significant stock in this maxim in arguing for maximum penalties, this Court in St. Louis tempered reliance on the maxim.

97Justice Taschereau, concurring, found that the Exchequer Court had carried the maxim too far (St. Louis, at p. 652). Justice Girouard, writing for the majority, agreed, and noted that most modern nations, including Canada, did not take the maxim as far as the Romans (pp. 666-67 and 682-83). Instead, Girouard J. endorsed the language of Campbell J. in Bott v. Wood, 56 Miss. 136 (1878), at pp. 140-41:

98SS&C cites Armory to buttress its claim that the maxim, when properly applied, ensures that the highest possible remedial value be awarded to the prejudiced party. In Armory, a chimney-sweep found a piece of jewellery and took it to a goldsmith. The goldsmith’s apprentice removed the stones and the goldsmith offered the chimney-sweep money, but the chimney-sweep declined and requested the return of the stones. The goldsmith’s apprentice returned only the empty socket for the stones. The chimney-sweep brought a claim against the goldsmith. Chief Justice Pratt directed the jury that unless the stones were produced, they should “presume the strongest against” the goldsmith and “make the value of the best jewels the measure of their damages” (p. 664; see also Cassels and Jones, at p. 262).

99In reviewing Armory, Girouard J. recognized that it was not directly on point, as the goldsmith was in the position of a thief. Justice Girouard went further in stating, at pp. 683-84, “[t]he appellant is his own master; he has taken nothing from the respondent; and I cannot understand how the maxim contra spoliatorem can generally be applied to a party who withholds or destroys his own papers.”

100In Lamb v. Kincaid (1907), 38 S.C.R. 516, a boundary dispute arose between the owners of placer claims in the Yukon. The appellants trespassed onto the disputed area and removed large amounts of auriferous material and mixed it with material taken from its own portion of the land (p. 526). From this mixture, the appellants extracted gold, prompting the respondents to bring a claim for invasion of their placer claim and damages equal to the value of the extracted gold. The appellants argued that they were entitled to deduct from the damages the expenses incurred in removing and washing the extractions, but did not maintain any account of this information. The Court declined to speculate for the benefit of the appellants, citing Armory in support of its position that “if a man by his deliberately tortious act destroys the evidence necessary to ascertain the extent of the injury he has inflicted, he must suffer all the inconvenience which is the result of his own wrong” (p. 540).

101SS&C also relies on Ponce v. Société d’investissements Rhéaume ltée, 2023 SCC 25, for the principle that evidentiary difficulties should be resolved against the wrongdoer where their actions prevent the other party from proving its damages. In Ponce, two presidents of a group of companies learned that another company was interested in acquiring the group of companies they ran. Instead of disclosing this information to their shareholders, they purchased the group of companies themselves and resold them at a substantial profit (para. 1). Though they sought compensation for the lost gain, the shareholders in Ponce contended that any difficulties in proving their damages was attributable to the presidents’ wrongdoing (para. 108). Ponce cited Kincaid for the principle that “where a fact cannot be proved because of a party’s dishonesty, that fact will be assumed to be true in the absence of evidence to the contrary” (Ponce, at para. 114; see also Kincaid, at paras. 539-40; Rainbow Industrial Caterers Ltd. v. Canadian National Railway Co., [1991] 3 S.C.R. 3, at pp. 14-16; Biotech Electronics Ltd. v. Baxter, [1998] R.J.Q. 430 (C.A.)).

102While Kincaid draws on Armory, and Ponce draws on Kincaid, these decisions do not squarely deal with spoliation. As BNY points out, Armory, Kincaid, and Ponce were not necessarily concerned with the intentional destruction of relevant documents in the face of reasonably contemplated litigation. Instead, these cases were primarily concerned with a party’s wrongs that preceded the litigation process and affected the party’s ability to prove their damages. As explained below, proving damages, while usually tied to the issue of spoliation, is conceptually distinct from proving spoliation generally.

103In sum, courts have moved away from a draconian approach to remedies for spoliation to more proportional remedies, as will be discussed below.

(ii) Broader Doctrinal Framework

104Spoliation poses a risk to the justice system because it “can mislead courts and can encourage parties to settle on unjust terms”, it “can result in people being denied the opportunity to obtain a legal remedy even when they have suffered egregious harm”, it can even cause judges to “render decisions that are based on inaccurate findings of fact, and as a result may be rendered wrong in law, based on flawed evidentiary records”, and it “can also frustrate litigants’ prosecution and defence of their actions” (Cudmore, at § C:1, § 1).

105Like with the doctrine of abuse of process, remedies for spoliation are inherently discretionary. In Saskatchewan (Environment) v. Métis Nation – Saskatchewan, 2025 SCC 4, Rowe J., writing for a unanimous Court, stated that “[w]here an abuse of process has been established, a subsequent question arises: What remedy is to be granted? That decision is discretionary” (para. 32 (emphasis added); see also Trillium Power Wind, at para. 22; Law Society of Saskatchewan v. Abrametz, 2022 SCC 29, [2022] 2 S.C.R. 220, at paras. 35-36; Behn v. Moulton Contracting Ltd., 2013 SCC 26, [2013] 2 S.C.R. 227, at para. 40; Toronto (City) v. C.U.P.E., Local 79, 2003 SCC 63, [2003] 3 S.C.R. 77, at paras. 37-38). Fixing a remedy for abuse of process is discretionary in order to allow judges to do justice in the array of circumstances in which abuses may arise. Spoliation requires similar flexibility.

106Given the jurisprudential developments in this area, it would be inconsistent for this Court to hold that a finding of spoliation attracts a bright-line, mandatory remedy, while emphasising that the related doctrine of abuse of process attracts contextual, flexible remedies.

(iii) Provincial Legislation

107In addition to the court’s inherent jurisdiction to control an abuse of its process, the court’s ability to grant remedies in response to a finding of spoliation stems from the relevant rules of civil procedure (McDougall, at para. 22; Trillium Power Wind, at para. 22).

108While it is most common for courts to invoke sanctions arising from its inherent jurisdiction, rules of the court contain remedial options for the destruction of evidence. For example, r. 30.08(1) of the Ontario Rules of Civil Procedure, R.R.O. 1990, Reg. 194, provides remedies for the failure to disclose or produce a document for inspection in compliance with the rules. If the document is not favourable to the impugned party’s case, the court “may make such order as is just” to remedy the prejudice arising from the lack of disclosure (r. 30.08(1)(b)).

109In stronger terms, the Nova Scotia Civil Procedure Rules provide that “[d]eliberate or reckless deletion of relevant electronic information, expunging deleted information, or destruction of anything containing relevant electronic information after a proceeding is started” may be dealt with under the “Abuse of Process” heading of the rules (r. 16.13). Under r. 88, a judge in Nova Scotia who is satisfied that a process of the court is being abused by a party may provide a remedy “that is likely to control the abuse” including: an order for dismissal or judgment; a permanent stay of proceedings; a conditional stay of proceedings; an order to indemnify each other party for losses resulting from the abusive conduct; an order striking or amending a pleading; an order expunging an affidavit or other court document or requiring it to be sealed; an injunction preventing a party from taking a step in a proceeding without permission of a judge; and/or any other injunction that tends to prevent further abuse (r. 88.02(1)).

110In Quebec, parties are duty-bound to cooperate, present evidence to the other parties on request, and ensure that relevant evidence is preserved until the end of trial as set out in arts. 20 and 251 of the Code of Civil Procedure, CQLR, c. C-25.01. The obligation of the parties to preserve evidence applies at all times, with the destruction of evidence typically resulting in an adverse inference against the party who destroyed it, or in some cases, a finding of contempt (Mag Energy Solutions inc. v. Falconer Cloutier, 2016 QCCS 2830, at paras. 26, 63 and 65).

111Accordingly, reliance on the rules of civil procedure to address the destruction of documents does not detract from the drawing of an adverse inference from a finding of spoliation.

(iv) Common Law Approaches to Remedies for Spoliation

112The consensus among other common law jurisdictions follows the flexible remedial approach. SS&C did not point to any other common law jurisdictions that have adopted a non-discretionary approach to assessing remedies for spoliation. This is likely because there is minimal support for the “maximum penalty” approach. Remedial discretion remains prevalent.

113For example, in England, where a court has found that a party suppressed documents, there are a number of remedies available:

114The starting point for the analysis is usually to determine whether a fair trial is possible given the destruction of evidence. If so, the court may determine whether the remedy of striking out a pleading would be proportionate and fair in all the circumstances, “or whether some other remedy will safeguard the position of the innocent party” (Active Media Services Inc. v. Burmester, Duncker & Joly GmbH & Co. Kg, [2021] EWHC 232 (Comm.), at para. 307).

115In Australia, “courts may remedy spoliation of evidence in a number of ways” (Mills & Anor v. Central Sydney Area Health Service & Anor, [2002] NSWSC 728, at para. 97). It is necessary “to weigh the effect of the alleged destruction of documents in respect of the issues in the proceeding, in order that the remedy be not out of proportion to the prejudice occasioned” (British American Tobacco Australia Services Ltd. v. Cowell (Representing the Estate of McCabe (deceased)), [2002] VSCA 197, 7 V.R. 524, at para. 190). A few years after the Supreme Court of Victoria (Court of Appeal) decided McCabe, the Evidence (Document Unavailability) Act 2006 (Vic.), was enacted to create a practical approach to dealing with spoliation by empowering judges to make a range of orders (see C. Kwan Lee, “Document Destruction in Victoria: the Correctness of Eames J’s Decision in McCabe” (2009), 6 Macquarie J.B.L. 37, at p. 47; Evidence (Miscellaneous Provisions) Act 1958 (Vic.)). Sections 89B and 89C of the Evidence (Miscellaneous Provisions) Act 1958 are instructive:

116In the United States, the remedial approach to spoliation is principally derived from the Federal Rules of Civil Procedure, 28 U.S.C. app. (2024), or other state procedural rules, and the court’s inherent powers (B. Finlay, M.-A. Vermette and M. Statham, Electronic Documents: Record Management, E-Discovery and Trial (loose-leaf), at § 5:35). In the context of electronically stored evidence, r. 37(e)(2) of the Federal Rules of Civil Procedure provides a uniform standard for U.S. federal courts in the imposition of sanctions after making a finding of spoliation (Christian, at p. 917). Rule 37(e) provides:

117Despite r. 37(e)(2) having standardized the remedial options for U.S. federal courts, a finding of spoliation does not require courts to apply the sanctions listed (as evidenced by the use of the word “may”) (Christian, at p. 917). Relatedly, the court’s “broad discretion in crafting an appropriate sanction for the destruction of evidence” through its inherent powers allows it to focus on restoring the prejudiced party to the same position it would have been in absent the spoliation (Finlay, Vermette and Statham, at § 5:37). Indeed, U.S. courts have imposed a variety of sanctions or remedies, including “fines, shifting costs and awarding attorney’s fees, the exclusion of evidence, adverse inference instructions to juries, the dismissal of claims, default judgments, and, where the spoliator is a lawyer, the recommendation of disciplinary action” (§ 5:35 (footnotes omitted)).

118In sum, several other common law jurisdictions retain the discretionary, flexible approach to remedying spoliation that supports rejecting the use of the mandatory “maximum penalty” approach advocated by SS&C. We therefore decline to adopt a change to the common law that would effectively amount to binding trial judges to a presumption that the non-spoliating party has proven their maximum position on damages in every case where spoliation is found.

VI. Application

119SS&C argues that the trial judge erred in drawing inferences that did not properly remedy BNY’s spoliation and thus tainted the damages awarded below. We agree. The inferences drawn by the trial judge did not meet the requirement that spoliation attract an adverse inference that usefully fills the evidentiary gap left by the spoliation. Nor was his damages award acceptable. Instead, the trial judge failed to draw inferences that usefully plugged the evidentiary gap left by the spoliation and awarded damages that are untethered from the facts of the case, unconnected to the incomplete inferences that he did draw, and inconsistent with his own rationale for the award. A reassessment of damages based on a finding of spoliation and the adverse inferences that must be drawn in response is necessary.

A. Standard of Review

120St. Louis indicates that when a presumption is drawn from a finding of spoliation, it is a matter of fact left to the determination of the trial judge. Drawing an adverse inference from the destruction of evidence is therefore reviewable for palpable and overriding error (see generally Benhaim v. St-Germain, 2016 SCC 48, [2016] 2 S.C.R. 352, at para. 52).

121Making a finding of spoliation, based on the four-part test described above, involves applying a legal standard to a set of facts (see generally Canada (Director of Investigation and Research) v. Southam Inc., [1997] 1 S.C.R. 748, at para. 35). Therefore, a finding of spoliation, which represents a question of mixed fact and law, is reviewable for palpable and overriding error unless the analysis is tainted by an identifiable legal error (Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, at para. 36; Southam Inc., at para. 35; H.L. v. Canada (Attorney General), 2005 SCC 25, [2005] 1 S.C.R. 401, at para. 55).

122The content and scope of the remedy associated with a finding of spoliation is discretionary and therefore attracts a high degree of deference. Absent an extricable legal error, a palpable or overriding factual error, or a failure to exercise discretion judicially, appellate courts will not intervene (see Canada (Transportation Safety Board) v. Carroll-Byrne, 2022 SCC 48, [2022] 3 S.C.R. 515, at para. 41, citing Canada (Attorney General) v. Fontaine, 2017 SCC 47, [2017] 2 S.C.R. 205, at para. 36; P. (W.) v. Alberta, 2014 ABCA 404, 378 D.L.R. (4th) 629, at para. 15; see also Canadian Imperial Bank of Commerce v. Green, 2015 SCC 60, [2015] 3 S.C.R. 801, at para. 95).

123The assessment of damages attracts considerable deference. When a damage award is appealed, it must be reviewed on a deferential standard that reflects its discretionary nature. As our Court stated in Naylor Group Inc. v. Ellis-Don Construction Ltd., 2001 SCC 58, [2001] 2 S.C.R. 943, appellate intervention is warranted only where the judge (1) made an error of principle in law; (2) misapprehended the evidence; (3) erred in reaching a conclusion for which there was no evidence; (4) failed to consider relevant factors, or considered irrelevant factors; or (5) made a palpably incorrect or wholly erroneous assessment of damages (para. 80; see also Housen; K.L.B. v. British Columbia, 2003 SCC 51, [2003] 2 S.C.R. 403, at para. 62; M.B. v. British Columbia, 2003 SCC 53, [2003] 2 S.C.R. 477, at para. 54).

B. The Trial Judge Drew Weak and Incomplete Adverse Inferences Against BNY

124When spoliation is found, the precise contours that the resulting adverse inference takes is subject to judicial discretion. This is essential to ensure that justice is done in the particular circumstances of each case. However, the requirements that the inference drawn be adverse to the spoliator and capable of filling the evidentiary gap left by the spoliation is not similarly discretionary. As stated above, once spoliation is found, a mandatory presumption arises that the destroyed evidence would be unfavourable to the spoliator’s case. The trial judge erred in drawing inferences that are little more than the arguments that BNY made in its own defence and simply reflected what the evidence before the court already showed. The result of drawing these weak and incomplete inferences was to improperly burden SS&C with the uncertainty and gaps in the evidence caused by BNY’s spoliation — rather than placing that burden on the spoliator.

(1) BNY Engaged in Spoliation

125The Court of Appeal found that BNY engaged in spoliation. Even so, for clarity, we briefly summarize and apply the test for spoliation in the circumstances. Establishing spoliation requires proof of the following elements on a balance of probabilities: (1) the evidence was intentionally destroyed, altered, mutilated, or concealed; (2) at the time of destruction, litigation was ongoing or reasonably contemplated; (3) the evidence was relevant to said litigation; and (4) it is reasonable to infer that the evidence was destroyed to affect the litigation.

126Each element is met in the instant case. The data usage evidence was clearly relevant, as three levels of court have now indicated. It was clearly destroyed or not produced, and this destruction or non-production followed the preservation notice sent by SS&C’s counsel. BNY was warned to preserve the data because of the pending litigation, and BNY therefore had full knowledge that litigation was going to be commenced. BNY refused to store the data on the basis that it rejected the allegations SS&C had made against it. In this context, and in the absence of argument to the contrary from BNY, the only possible inference is that the evidence was destroyed or not produced in order to affect this litigation. Spoliation was therefore clear. The fairness and truth-seeking functions of the trial were undermined by the destruction or non-production of records that would have shown the extent to which BNY had shared and sold the data.

(2) The Trial Judge Failed to Address the Scope and Impact of the Spoliation and Drew Inferences That Cannot Remedy the Spoliation

127Because spoliation was established, it was mandatory that the trial judge draw an inference, or inferences, that is adverse to BNY and that meaningfully filled the evidentiary gap left by the spoliation. Considered in the concrete reality of each case, the adverse inference drawn should have flowed from the context and scope of the spoliation and from the impact of the destroyed evidence.

128Here, millions of data points tracking BNY’s use of SS&C’s proprietary data were spoliated. That data was readily available to BNY and could have been preserved and produced. BNY was well aware of the request for preservation, but chose not to preserve the data because it disagreed with SS&C’s claim. This is never a justification for destroying or not producing evidence. It is particularly shocking conduct in litigation between exceptionally sophisticated and well-resourced parties. The result of the spoliation was the evisceration of the court’s ability to truly grasp the scale of BNY’s breaches of the contract. No reason for the destruction was given, the prejudice to SS&C was vast, and the court’s ability to dispose of the issues based on access to all of the relevant evidence was fundamentally undermined. In short, the spoliation and its result were severe.

129Despite this, the inferences drawn against BNY were weak and incomplete. At their highest, they simply accepted BNY’s own position at the hearing on liability or restated what the evidence before the court clearly demonstrated. At trial, BNY did not argue that it did not share the data — it argued that it was allowed to share the data with all of its affiliates. Indeed, BNY claimed that all of its affiliated entities were “entitled to receive” the SS&C data (ruling on liability, at paras. 6 and 58; C.A. reasons, at para. 19). Finding, therefore, that some unspecified number of entities beyond CIBC Mellon — itself an unauthorized entity — used the data to some unspecified degree renders the adverse inferences incomplete.

130Aside from BNY’s own arguments, the facts clearly established the inferences that the trial judge drew. The trial judge noted that in “2017 BNY initially indicated that eight entities including CIBC Mellon used SS & C’s market data to update their institutional accounts” (ruling on damages, at para. 120). He further noted that in 2019, BNY “provided additional names which raised the number of entities to 11” (para. 120). The trial judge found that when BNY disclosed the accounted-for data, “it indicated that SS & C data had been shared in accounts amongst 23 different strategic business units” (para. 121). BNY presented further evidence and argument to the effect that, as a result of various amalgamations, the court should reduce “the number of unauthorized users to 28” (para. 124).

131Similarly, the inferences drawn by the trial judge were not capable of filling the evidentiary gap left by the spoliation. He drew two inferences: (1) that the unaccounted-for data was used by unauthorized entities within the BNY group; and (2) that the unauthorized entities made more than de minimis use of the data. These inferences simply failed to fill the gap left by the spoliation, which required a determination on how many entities used the data and how much data they used. They did not establish, for example, how many unauthorized entities accessed the data or how often those entities accessed the available data. Instead, they established vague ranges of potential use: at least one unauthorized entity used SS&C data and that use was more than trifling. The practical impact of the spoliation remained unremedied.

132These weak and incomplete inferences reflect the trial judge’s error in failing to make an express finding of spoliation and failing to properly remedy the spoliation that took place. The trial judge, without the benefit the legal principles as now clarified by our Court, treated the missing evidence as an issue of a mere failure to produce (ruling on damages, at paras. 104-7). He proceeded to find that BNY could have produced other evidence showing minimal unauthorized use of the data and that it had not satisfactorily explained its failure to produce the missing evidence (paras. 108-11). His analysis did not engage with the scope or impact of the missing evidence itself.

133To be properly adverse in the circumstances of this case, the inferences drawn by the trial judge should have resulted in concrete findings of fact that were adverse to BNY and filled the evidentiary gap left by BNY’s spoliation. For example, in the absence of any explanation to the contrary for why it would be unreasonable or unjust, the trial judge could have inferred that the missing evidence was as damaging to BNY’s case as reasonably possible. That would mean that each of the 65 entities that could have accessed the data did so and that they each used all of the data that they could access. These findings would fill the factual gap as to data usage in a manner that is adverse to BNY, rather than providing, as the trial judge did, a vague range of potential unauthorized usage. Of course, if the trial judge wished to make a finding of fact that was less adverse to BNY, he was entitled to do so — though it would be expected that such a finding would be explained in his reasons.

134In this evidentiary and factual context, although the trial judge acknowledged that BNY had to bear the burden of an adverse inference (ruling on damages, at para. 187), the inferences he drew were not adequate to the task of assessing damages. They do not level the playing field, nor do they rectify the harm to the litigation process and the evidentiary gap that BNY’s spoliation created. By failing to turn his mind to the scope and impact of the destroyed evidence and by failing to draw inferences that were capable of filling the evidentiary gap in the concrete reality of the case, the trial judge erred.

C. The Rateable Approach Is Erroneous

135The trial judge also erred with respect to the methodology that he used to calculate damages. While they may at times involve more art than science, and even come close to approximations, damage awards must always be tethered to the evidence before the court. Trial judges must always do their best to ensure that the damages they award accord with the evidence before them and the breach or loss that occurred. In this case, the rateable approach failed to take into account the nature and scale of the breaches found, focused on an unsupported use of centralized pricing, relied on an arbitrary rate that is not connected to the evidence, and was premised on the flawed assumption that only the unaccounted-for data was shared and that it was shared only once. The trial judge did reflect upon some of the shortcomings of his approach, but this does not rectify the palpable and overriding errors inherent in the rateable approach.

(1) General Principles

136SS&C is owed damages for the breach of its contract. As the Court of Appeal rightly noted: “It was reasonable for SS&C to expect to be paid for the service it provided. That is its business. The business only works if the service has value” (para. 126). In crafting such awards, trial judges are not held to a standard of perfection. Appellate courts will not interfere with awards “where they have an evidentiary basis” because “damages cannot always be calculated with mathematical precision” (Extreme Venture Partners Fund I LP v. Varma, 2021 ONCA 853, 24 B.L.R. (6th) 38, at para. 53). While plaintiffs bear the burden of proving their damages on a balance of probabilities, courts must not let evidentiary difficulties get in the way of remedying a loss once that loss has been established (see Penvidic Contracting Co. Ltd. v. International Nickel Co. of Canada, [1976] 1 S.C.R. 267).

137Trial judges often lack certain, clear, or fulsome evidence regarding damages. In such evidentiary vacuums, they must do the best they can to quantify damages with the evidence available (Extreme Venture, at para. 53; Penvidic, at pp. 279‑80; Wood v. Grand Valley Rway. Co. (1915), 51 S.C.R. 283, at p. 289; Martin v. Goldfarb (1998), 41 O.R. (3d) 161 (C.A.), at p. 187; TMS Lighting Ltd. v. KJS Transport Inc., 2014 ONCA 1, 314 O.A.C. 133, at para. 61; Hollowcore Inc. v. Visocchi, 2016 ONCA 600, 351 O.A.C. 228, at paras. 53‑54; B.M.G. v. Nova Scotia (Attorney General), 2007 NSCA 120, 260 N.S.R. (2d) 257, at para. 172). These issues are often heightened where, as here, the case at hand is a complex commercial matter (see Gautam v. South Coast British Columbia Transportation Authority, 2020 BCCA 135, 37 B.C.L.R. (6th) 9, at para. 119, citing Mellco Developments Ltd. v. Portage la Prairie (City), 2002 MBCA 125, 166 Man. R. (2d) 285, at para. 107; see also Penvidic, at pp. 279‑80; Wood, at p. 289).

138In assessing damages, judges should feel neither disquiet nor hesitancy if the number arrived at is high (see, e.g., Nova Chemicals Corporation v. Dow Chemicals Company, 2020 FCA 141, [2021] 1 F.C.R. 551, at paras. 30‑31, aff’d 2022 SCC 43, [2022] 3 S.C.R. 352). Extensive and repeated contractual breaches can lead to a significant damage award. So long as the award emerges from the application of established doctrine and principles, there is no issue with a substantial award. Full compensation must be awarded — no more and no less. The plaintiff must be put in the position he or she would have been in had the contract been performed. Otherwise, the defendant will be rewarded for its wrongdoing and others might be encouraged to act similarly.

139While we do not doubt that this case involved significant issues with respect to assessing damages, the trial judge’s award was simply disconnected from its factual and evidentiary context. Faced with a critical gap in the evidentiary record caused by BNY’s spoliation, and having failed to draw adverse inferences that reflect the scope and impact of the destroyed evidence, the trial judge was left with a factual vacuum concerning the usage of the data. He crafted a damage award that is contrary to his own findings with respect to the unaccounted-for data and that is not rooted in the evidentiary record. These disconnects amount to palpable and overriding errors in addition to the legal errors discussed earlier. For the reasons explained below, the damage award cannot stand.

(2) The Rateable Approach Is Unconnected to the Case and the Adverse Inferences Drawn

140The methodology used in the rateable approach is fundamentally flawed because it bears no relation to BNY’s breaches. It gives no consideration to the number of entities that may have accessed SS&C’s data and to the number of times that each data point provided by SS&C could have been shared. It does not compensate SS&C for the data that was improperly appropriated and perhaps used by up to 65 entities. Instead, it is premised on a wholly arbitrary understanding of the “rate” in question and assesses damages on the unsupported basis that the unauthorized entities would have formed an enterprise agreement with SS&C to use the unaccounted-for data and that the usage of the unaccounted-for data was to be paid for on the same terms as in the Mellon Trust Agreement. With respect, the rateable approach is untethered from the facts as found and from the breaches at the heart of this case.

141The result of the rateable approach was that the trial judge failed to give any weight to his factual finding emerging from the inferences he drew that the unauthorized entities made use of SS&C’s proprietary data. What this means is that up to 65 entities repeatedly participated in breaches of the Mellon Trust Agreement. This is the factual foundation for the damage award. However, it played no role at all in the award granted. The number of breaches and the number of unauthorized entities that used SS&C’s data played no role in the rateable approach. Instead, the rateable approach fails to take into account the scope of the breaches in question, the number of entities that improperly accessed the data, and the extent of that improper access. This is not a peripheral issue; it goes to the heart of the harm suffered by SS&C — the repeated and extensive unauthorized use of its proprietary data. The trial judge’s rateable approach, therefore, was arbitrary in that it was entirely disconnected from the nature and scope of the breach, resulting in a reviewable error.

142The error is obvious when the trial judge’s and the Court of Appeal’s treatments of the breaches and damages are compared. The trial judge found that the entities that BNY’s predecessor owned when it entered into the Mellon Trust Agreement, with the exception of CIBC Mellon, were lawfully able to access the SS&C data. Sharing the data with those 21 entities was, therefore, not a breach of the contract (clarification reasons, at para. 43; see also ruling on damages, at paras. 19 and 122‑23). Accordingly, unlawful sharing occurred only with respect to 44 entities. This finding as to the breaches was overturned by the Court of Appeal, which rightly found that the entities that BNY’s predecessor owned when it entered into the Mellon Trust Agreement had no entitlement to the data, and so sharing the data with them was in fact a breach of the contract (paras. 4 and 38). As a result, on appeal, the number of unauthorized entities that could have used SS&C’s proprietary data increased from 44 to 65.

143Nonetheless, the damage award remained unchanged. A 50 percent increase in the scale of the breaches — in the number of unauthorized entities that used SS&C’s proprietary data — yielded no increase in the damages awarded. SS&C received no additional compensation despite the finding that 21 more entities could have benefitted from its data in breach of the Mellon Trust Agreement. What this reveals is that there is simply no relationship between the scope of the breaches and the damages awarded; between how many entities improperly received the data and how much SS&C was owed for that data. Given this disconnect, the damage award would remain the same regardless of whether all 65 entities made repeated use of each piece of data, only 44 entities made repeated use of the data, only one entity made only minor use of the data, or hundreds more entities used it extensively. The rateable approach is wholly inadequate because it is unresponsive to the full extent of the breaches that it is meant to remedy.

144Instead of this approach, the trial judge should have awarded damages that compensated SS&C for what BNY wrongfully acquired — the ability to repeatedly share SS&C’s proprietary information with up to 65 unauthorized entities for over 17 years. Each instance of data sharing, with each unauthorized entity, was a breach of a core provision of the Mellon Trust Agreement. SS&C is owed compensation for each such unauthorized use of its proprietary data, data that BNY had no right to share. While the number of breaches is unknown because of BNY’s spoliation, it is clear that the breaches were extensive and long‑lasting. The trial judge’s award takes no account of this finding.

(3) The Rateable Approach’s Reliance on Centralized Pricing Is Erroneous

145A principal reason that the trial judge used the rateable approach was his concern over the pricing of SS&C’s data and his focus on the notion of centralized pricing. Little evidence was before the court with respect to pricing beyond the contracts between the parties, but it was suggested that if BNY’s affiliates had been required to purchase the data, they would have purchased the data together and received the benefits of centralized pricing and economies of scale. In essence, had BNY’s affiliates been forced to pay for SS&C’s data, they would have negotiated together for a better rate. This was said to be in line with industry practice.

146This approach is untethered from the facts before the trial judge. First, while centralized pricing was found to be a “directional norm” (ruling on damages, at para. 190), there was no evidence or argument concerning how this industry standard would lead SS&C to consent to providing its data to all 65 unauthorized entities on the exact same basis that SS&C had sold its data to one entity, BNY. There is, for example, nothing to suggest that the combined amount of data used by the 65 entities and provided by SS&C would be similar to the amount of data used by BNY alone, rather than 65 times larger. Nor, as we discuss below, is there any basis for limiting the amount paid under this hypothetical centralized pricing model to the unaccounted-for data alone. Indeed, the trial judge appeared to recognize this problem, saying that the “unanswered question in the evidence” was “what the precise meaning of centralized pricing was. Did it mean that any contract would apply to all subsidiaries and affiliates of a contracting party or something else” (fn. 82).

147Second, while centralized pricing may occur as described by the trial judge in some circumstances, in the case at hand there is evidence that weighed against this premise. Indeed, this case arose in the context of distinct contracts between SS&C and two related entities that separately used SS&C’s proprietary data: BNY and CIBC Mellon. These separate agreements speak to the commercial reasonableness of BNY’s affiliated entities maintaining separate contracts with SS&C involving similar pricing. While enterprise agreements with economies of scale may be common practice in the industry, in this case the evidence demonstrates the commercial reasonableness of using separate contracts for related entities. We see no reason to ignore this context in favour of a hypothetical centralized enterprise agreement.

148Third, the result of focusing on economies of scale in the absence of strong evidence and to the exclusion of the nature of the breaches themselves was to discount the damage award because of the scale of BNY’s breaches. The fact that BNY’s breaches of the contract were significant enough that BNY could argue that it would have benefitted from economies of scale if it and its affiliates had chosen to properly pay for the data usage should not work in its favour. Had BNY shared the data with only one or two other entities, there would be little reason to reject the use of the pricing arrangement in the Mellon Trust Agreement. Indeed, this is what the trial judge did with respect to calculating the CIBC Mellon damages.

149Finally, the trial judge’s focus on centralized pricing also flowed from his understanding that when the Mellon Trust Agreement and the CIBC Mellon Agreement were amended to add new entities or groups of securities, the pricing for the supplemental information was lower than the unit pricing under the initial agreement. This understanding was erroneous. First, as counsel for SS&C noted before our Court, the amendments to the Mellon Trust Agreement and to the CIBC Mellon Agreement did not simply involve a decrease in price based on an increase in the amount of data bought. They involved negotiated prices for different sets of products. Also, any reduction was not to the unit price of the data as a whole, but rather only to the “supplemental information” that was part of the new agreement (see ruling on damages, at para. 28).

150Furthermore, even if the price did simply decrease as more data was provided, there is no logical connection between such a reduction in price based on a contractual renegotiation between two entities and a finding that SS&C would have agreed to provide up to 65 times the data without any increase in their remuneration — a 98.5 percent discount relative to what BNY paid SS&C for the exact same data. With respect, this finding does not flow logically from the evidence. Instead, this analysis simply assumes both that the 65 entities and SS&C would have agreed on an enterprise model and that they would have agreed on the price determined under the rateable approach. Noting an industry standard of centralized pricing is wholly insufficient to ground such reductions. In this context, we see no reason to allow BNY to benefit from the hypothetical situation in which, rather than breaching the contract individually as they did, all of the unauthorized entities would have come together to negotiate a preferable price for the data that they would collectively receive from SS&C. By failing to tether his damage award to the facts as found or to the nature of the breaches in question, the trial judge made palpable and overriding errors that justify our Court’s intervention.

151The trial judge justified his approach by noting that the pain point in the damage quantification was the absence of pricing information (ruling on damages, at para. 188). With respect, we disagree. First, the concern for pricing evidence speaks to a broader methodological issue. The trial judge’s task was not to reconstruct the kind of contract the parties would have negotiated had they bargained for BNY’s unauthorized distribution of the data. He was entitled, of course, to find that this was the most commercially realistic method by which to value the converted data. But having decided to pursue a different theory of damages, one not advocated for and indeed opposed by both parties, the trial judge cannot blame SS&C for failing to provide evidence that is relevant only because he adopted his rateable approach, rejected the contractual evidence of price contained in the Mellon Trust Agreement and in the CIBC Mellon Agreement, and refused to follow the ordinary method of valuing breaches of contract in these circumstances.

152Second, the issues of usage and price are related. Pricing depends in part on data usage — on how much data was used and what data was used. This is clear from the parties’ negotiated fee schedules, which set out the price per quote accessed and provided discounts based on the volume of usage (A.R., vol. II, at pp. 30-31). The relationship between data usage and price is further supported by the trial judge’s findings with respect to centralized pricing. While he erred in assuming that centralized pricing would lead to a 65-entity enterprise agreement on the same basis as the Mellon Trust Agreement, he did not err in noting a general inverse relationship between the number of quotes and the price per quote (ruling on damages, at paras. 31 and 33). Inferences that are adverse as to usage rates would, therefore, have been essential to understanding potential pricing and crafting an appropriate award. In their absence, the trial judge’s assumption that the data used by BNY and the data used by all of the unauthorized entities would be priced at the same rate is neither commercially reasonable nor supported by the evidence. This assumption further undermines his rateable approach.

(4) The Rate Used in the Rateable Approach Does Not Reflect the Trial Judge’s Own Rationale for His Methodology

153However, the trial judge’s approach would remain arbitrary even if these methodological and evidentiary errors could be accepted because it fails to accord with his own rationale. The trial judge said that he applied “centralized pricing to the unaccounted for data at the same rate as BNY was prepared to pay for the data for which it could account” (ruling on damages, at para. 190 (emphasis added)). But this is not what he actually did.

154If the trial judge meant that he applied the rate of the Mellon Trust Agreement to the unaccounted-for data, this was an error. The trial judge did not price the “missing” data at the same rate as the “accounted-for” data. Instead, he deemed BNY to have “in effect” paid the full US$4.6 million contract price “for 44.6% of the data it received” and then uses that assumption to calculate his award (ruling on damages, at para. 180). However, this assertion bears little relation to the established facts. BNY paid US$4.6 million for all of the data it received, not only the data that was accounted for. As a result of this assertion that 44.6 percent of the data should be valued at the full contract price, the rate applied to the remaining 55.4 percent of the unaccounted-for data bears little relationship to the rate agreed to by the parties under the Mellon Trust Agreement. Rather, the rate applied neither finds support in the evidence nor reflects what the rateable approach was intended to do. The error is palpable and overriding.

155Similarly, if the trial judge meant that he applied the rate that BNY would have been prepared to pay for the data, it is still an error. This could only be based on an understanding that BNY would actually have been willing to pay US$4.6 million for 44.6 percent of the data. However, the trial judge pointed to no evidence, argument, or explanation supporting this contention.

(5) The Rateable Approach Misuses Data That Was Found to Be Wholly Unreliable

156Moreover, even if the trial judge’s methodology was accepted, the rateable approach remains deeply flawed on the facts of this case. It misunderstands the nature of the data that could be accounted for, treating the accounted-for data as if it was not shared at all or was shared with only one entity. There is no evidence to support this interpretation of the accounted-for and unaccounted-for data. Indeed, the trial judge found just the opposite: (1) the analysis of the data was entirely unreliable; and (2) the data did not show that only CIBC Mellon had used the accounted-for data.

157As previously mentioned, the accounted-for data is derived from two different databases: InvestOne and IAS (ruling on damages, at para. 44). The trial judge found both databases to be deeply flawed and their accompanying analysis not useful for determining whether the data they contained had been used by unauthorized entities — in fact, the databases indicated that the accounted-for data had been used extensively by CIBC Mellon, an unauthorized entity.

158The trial judge made clear his concerns with the available data. He found that the analysis of InvestOne was “so replete with flaws as to be entirely unreliable” (ruling on damages, at para. 50). He found that the “InvestOne analysis is entirely circular” and “looks only at the CIBC Mellon user bank on InvestOne”, such that “it is no surprise that the overwhelming usage would be attributed to CIBC Mellon” (para. 50). He described the compilation of the InvestOne data not as “an investigation into the facts” but as a “bald assertion masquerading as an inquiry” (para. 50). Concerning IAS, the analysis of that database was strongly skewed because any use of SS&C’s data by other unauthorized entities was not recorded — it did not examine non-CIBC Mellon use, did not record non-CIBC Mellon use, and did not provide the percentage of non-CIBC Mellon use (para. 46). The trial judge also found that the affidavit confirming the usage of the accounted-for data was similarly weak (paras. 45‑46). The trial judge noted that key information in the affidavit “was never designed to determine where SS & C data was used within BNY”, and yet that was “precisely the purpose” for which BNY asked him to use it (para. 45).

159In this context, it was an error for the trial judge to assume that the accounted-for data was not shared with unauthorized entities. The distinction between accounted-for and unaccounted-for data does not speak to whether the data was shared with unauthorized entities; it concerns only whether BNY provided some information on that data’s usage. The databases did not show that 44.6 percent of the data was not shared — indeed, almost all of that data was said to have been shared with CIBC Mellon in breach of the Mellon Trust Agreement. There is simply no basis for the trial judge to have found that SS&C should only be paid damages for the unaccounted-for data. Because this assumption is the foundation of the rateable approach, its weakness collapses the whole structure of the damage award.

VII. Conclusion

160This case has been beset by both the absence of evidence and spoliation. Nonetheless, the trial judge was required to order BNY to properly and adequately compensate SS&C for the data that it improperly shared for its own profit. The trial judge failed to do so by drawing inferences that were weak and incomplete and by awarding damages that were unconnected to the facts of the case. These errors must be rectified.

161To do this, we would allow the appeal and remit the matter back to the Superior Court on only the issue of the quantification of SS&C’s damages. This matter should proceed on the basis of the existing record, but it will be open to the trial judge to accept evidence, expert or otherwise, with respect to pricing, based on the adverse inferences that must be drawn. The trial judge will be required to properly draw mandatory adverse inferences that remedy the effects of BNY’s spoliation. These adverse inferences will have to account for the errors identified and the context, scope, and impact of the destroyed evidence will have to be taken into consideration. In particular, the inferences should result in a concrete finding of fact relating to the unauthorized data usage (including both the number of entities that accessed the data and the number of times each entity accessed the data) and a concrete finding of fact about the value of that data. This usage and pricing information should then be used to arrive at the damages figure.

VIII. Disposition

162We would allow the appeal, set aside paras. 2, 3, 5 and 6 of the Court of Appeal’s order dated September 12, 2024, set aside the damage award made by the Superior Court in paras. 4 to 7 of its judgment dated July 11, 2023, remit the matter to the Superior Court on the issue of damages quantification only, and award costs to SS&C in this Court and in the Court of Appeal.