Case Details
- Citation: [2026] SGCA(I) 1
- Title: JONATHAN KUPETZ & 9 Ors v TERRAFORM LABS PTE. LTD. & 2 Ors
- Court: Court of Appeal of the Republic of Singapore
- Date: 6 March 2026 (judgment reserved after hearing on 19 January 2026)
- Judges: Steven Chong JCA, Belinda Ang Saw Ean JCA, Beverley McLachlin PC IJ
- Proceedings: Civil Appeal from the Singapore International Commercial Court Nos 2 and 4 of 2025; Originating Applications Nos 3 and 4 of 2025; Summons No 29 of 2025
- Appellants / Applicants: Jonathan Kupetz and nine others (representative appellants / applicants)
- Respondents: Terraform Labs Pte. Ltd. and two others (Terraform; Kwon Do Hyeong; Luna Foundation Guard Ltd.)
- Lower court reference: Beltran, Julian Moreno v Terraform Labs Pte Ltd [2025] SGHC(I) 17 (“Judgment”)
- Key procedural context: Representative action in SIC/OA 3/2024 (“OA 3”); bifurcated trial with a first tranche addressing common issues and disposing of individual claims of ten representative claimants
- Representative action scale: OA 3 involved 366 claimants
- Nature of claims: Misrepresentation (fraudulent, negligent, and innocent), breach of unilateral contract, inducing breach of contract, and unlawful means conspiracy
- Core tort issue on appeal: Measure of damages for the tort of deceit using a reliance measure; determination of the “cut-off” time for reliance
- Fresh evidence application: CA/SUM 29/2025 (“SUM 29”) to adduce further evidence on appeal
- Judgment length: 83 pages; 25,248 words
Summary
This Court of Appeal decision arises from a large representative action in the Singapore International Commercial Court (“SICC”) concerning the collapse of the Terra ecosystem, in particular the algorithmic stablecoin TerraUSD (“UST”) and its sister token Luna (“LUNA”). The investors alleged that they were induced to purchase UST by a set of pleaded representations about the stability of UST “by design” and the ability of the Terra protocol to maintain UST’s peg to US$1 through an arbitrage mechanism involving LUNA. The trial judge found that certain representations were false, and the central trial disputes became reliance and causation, followed by the quantification of reliance damages for fraudulent misrepresentation (deceit).
On appeal, the Court of Appeal addressed multiple issues, including (i) whether the respondents had sufficiently pleaded and/or given notice of their case on mitigation and the end of reliance; (ii) whether the trial judge erred by failing to consider the individual circumstances of each representative appellant; (iii) whether the trial judge erred in adopting a specific “cut-off” price/time (12 May 2022 at 12.01am UTC) as the end point for reliance; (iv) claims of certain individual representative appellants; and (v) costs. The Court also dealt with an application to adduce further evidence on appeal (SUM 29) and clarified the legal interplay between discovery of fraud, the end of reliance, and the duty to mitigate in deceit claims.
Ultimately, the Court of Appeal used the occasion to articulate a structured approach to the “cut-off” concept in reliance-based damages for deceit, emphasising that mitigation is not relevant until the misrepresentation is discovered, and that the end of reliance is closely tied to the time when the claimant could reasonably have sold the asset acquired under the influence of the misrepresentation. The Court’s reasoning provides important guidance for future crypto-related misrepresentation litigation and for representative proceedings where common findings may bind represented parties.
What Were the Facts of This Case?
The appellants were investors who purchased UST, an “algorithmic stablecoin” within the Terra ecosystem. UST was marketed as stable by design, pegged to a fiat currency (US$), and supported by a protocol intended to maintain price stability regardless of market size, volatility, or demand. The alleged mechanism depended on arbitrage between UST and LUNA: when UST traded above US$1, arbitrageurs would burn LUNA and mint UST; when UST traded below US$1, they would burn UST and mint LUNA. The intended market forces were said to keep UST at the US$1 peg.
Another component of the Terra ecosystem was the “Anchor Protocol” (“Anchor”), a lending and borrowing platform where users could deposit or stake UST in exchange for promised returns on an annualised yield basis, analogous to earning interest on a bank deposit. The investors’ case was that these ecosystem features and related communications induced them to purchase UST and to hold it until the collapse of the Terra ecosystem in May 2022.
In the SICC, the investors commenced OA 3 as a representative action. The trial was bifurcated: the first tranche dealt with issues common to all claimants and also disposed of the individual claims of ten representative claimants in full. Subsequent tranches would address the remaining represented claimants. The ten representative claimants (now the representative appellants on appeal) were therefore used as a vehicle to determine key issues that would have binding effect on the represented group, subject to the representative action framework and the scope of the trial judge’s findings.
Before the trial of the representative action, the defendants conceded that various representations—including those central to the investors’ pleaded case—were false. The trial therefore focused largely on reliance and causation, and, critically for damages, on when reliance would have ended for each claimant. The trial judge raised the issue of the appropriate “cut-off” date/time for reliance and invited submissions. The judge ultimately adopted a cut-off time of 12 May 2022 at 12.01am UTC (“Cut-Off Time”). The appellants challenged that choice on appeal, and they also argued that they were denied an opportunity to adduce further evidence because the mitigation issue was raised late, particularly after the defendants sought to amend their defence.
What Were the Key Legal Issues?
First, the Court of Appeal had to consider whether the respondents sufficiently pleaded or gave notice of their case on the end of reliance and mitigation, such that the trial judge could properly adopt a cut-off time for reliance damages. This issue was intertwined with the appellants’ complaint that the mitigation amendment was allowed at a late stage and that they were allegedly deprived of the opportunity to adduce further evidence.
Second, the Court had to address whether the trial judge erred by failing to consider the individual circumstances of each appellant. In representative proceedings, some findings may be common and binding, but the assessment of individual reliance and damages can still require attention to each claimant’s circumstances. The appellants argued that the judge’s approach effectively treated all claimants as if their reliance ended at the same time, without adequate consideration of individual facts.
Third, the Court considered whether the trial judge erred in adopting the specific Cut-Off Time (and associated cut-off price concept) of 12 May 2022 at 12.01am UTC as the end point for reliance. This required the Court to examine the legal principles governing the measure of damages for deceit, particularly the reliance measure, and how the “discovery” of fraud relates to the end of reliance and the claimant’s duty to mitigate.
How Did the Court Analyse the Issues?
The Court of Appeal began by framing the essence of a claim for fraudulent misrepresentation (deceit): it is premised on the representee’s reliance on the truth of a representation that is later discovered to be false. The Court emphasised that, implicitly, a claimant’s reliance must end at some point in time when the falsity is discovered and legal proceedings become possible or necessary. That end point is crucial because reliance damages are quantified by reference to the period during which the claimant continued to hold or act on the basis of the misrepresentation.
In this case, the Court noted that the trial judge was troubled by the cut-off issue and had raised it with the parties. The appellants’ arguments on appeal, however, were partly built on an incorrect premise: they assumed that the cut-off determination arose directly from the defendants’ late mitigation amendment. The Court rejected that premise. It explained that the cut-off analysis was not simply a by-product of the mitigation amendment, but a separate and necessary factual/legal determination tied to when reliance would have ended.
To clarify the legal interplay, the Court relied on its earlier decision in POP Holdings Pte Ltd v Teo Ban Lim [2025] 2 SLR 90 (“POP Holdings”). In POP Holdings, the Court had held that until the deceit or misrepresentation is discovered, there is no question of mitigation because the claimant is not yet in a position to take steps to reduce loss arising from the deceit. Once the misrepresentation is discovered—equated to the time when reliance would have ended and when the claimant could sell the property acquired under the influence of the misrepresentation—the law deems mitigation to have occurred at that time. The Court of Appeal in the present case used this to explain why the cut-off time is conceptually linked to discovery and the ability to sell, rather than to the timing of a pleading amendment.
On the representative action dimension, the Court addressed how common findings may bind represented parties, while individual circumstances may still matter for the assessment of each claimant’s reliance and damages. The Court’s analysis indicates that the representative framework does not eliminate the need for careful attention to each claimant’s position where reliance is inherently fact-sensitive. However, the Court also recognised that where the factual matrix relevant to the end of reliance is common—such as the point in time when information would have been discovered by a reasonable investor—then a common cut-off may be appropriate. This is consistent with the representative action’s efficiency rationale, provided the legal and factual basis for commonality is properly established.
Regarding the application to adduce further evidence (SUM 29), the Court considered the standards for admitting fresh evidence on appeal. The Court’s discussion (as reflected in the judgment outline) included the “locked UST evidence” and “liquidity analysis” and also “tweet link evidence” and “new PG evidence” that the appellants sought to rely on. The Court’s approach reflects the appellate principle that fresh evidence is not admitted merely because it could have been obtained earlier; rather, it must be relevant, credible, and capable of affecting the outcome, and it must satisfy the procedural requirements for adducing additional material at the appellate stage.
Finally, the Court addressed the specific cut-off time and the “reasonable investor” concept. The appellants argued that a reasonable investor would not have discovered the fraud by the cut-off time, and they also challenged the adoption of US$0.8011 as the cut-off price. The Court’s reasoning, as indicated by the issues framed, required it to evaluate the evidence and the factual/legal basis for when reliance ended. The Court’s analysis therefore combined (i) the objective “reasonable investor” lens for discovery and reliance cessation, and (ii) the practical ability to mitigate by selling the asset, which anchors the reliance damages calculation.
What Was the Outcome?
The Court of Appeal dismissed or resolved the appellants’ challenges to the trial judge’s cut-off approach and the related procedural complaints, while also clarifying the legal principles governing reliance damages for deceit and the duty to mitigate. The Court’s decision provides authoritative guidance on how to determine the end of reliance in deceit claims, particularly in cases where a representative action requires common determinations that affect multiple claimants.
In addition, the Court dealt with SUM 29 (the application to adduce further evidence) and the costs consequences of the trial and appeals. The practical effect is that the reliance damages assessment framework adopted at first instance—anchored to the Cut-Off Time—remains the governing approach, subject to any adjustments the Court made for particular representative appellants’ individual claims.
Why Does This Case Matter?
This decision matters because it addresses a recurring problem in deceit litigation: how to quantify reliance damages when the claimant’s reliance ends at some point after the misrepresentation is made. The Court’s articulation of the relationship between (i) discovery of fraud, (ii) the end of reliance, and (iii) the duty to mitigate provides a structured method for courts and parties. It also reduces uncertainty in cases where defendants concede falsity but dispute when claimants would have discovered it.
For practitioners, the case is particularly relevant to crypto and fintech disputes, where information dissemination, market dynamics, and the timing of “discovery” can be complex. The Court’s emphasis on an objective “reasonable investor” approach, coupled with a practical mitigation anchor (the ability to sell), offers a workable standard for future disputes involving algorithmic stablecoins and similar assets.
Finally, the decision is significant for representative proceedings. It illustrates how courts may treat certain findings as common and binding on represented parties, while still requiring careful consideration of individual circumstances where reliance is genuinely claimant-specific. Lawyers advising claimants or defendants in representative actions should therefore pay close attention to pleading, notice, and the scope of common findings, as well as to the procedural management of evidence and amendments.
Legislation Referenced
- Not specified in the provided extract. (The judgment likely references Singapore procedural and representative action provisions and general principles on damages and misrepresentation, but the specific statutory citations are not included in the supplied text.)
Cases Cited
- POP Holdings Pte Ltd v Teo Ban Lim [2025] 2 SLR 90
- Beltran, Julian Moreno v Terraform Labs Pte Ltd [2025] SGHC(I) 17
Source Documents
This article analyses [2026] SGCAI 1 for legal research and educational purposes. It does not constitute legal advice. Readers should consult the full judgment for the Court's complete reasoning.