1 SIC/OA 3/2024 (“OA 3”) is the first representative action before the Singapore International Commercial Court (the “SICC”). It has been brought under O 10 r 19 of the Singapore International Commercial Court Rules 2021 (the “SICCR 2021”). In these bifurcated proceedings, ten representative claimants representing 356 other claimants are suing three defendants for breach of unilateral contract, misrepresentation (fraudulent, negligent, and innocent), inducing breach of contract, and unlawful means conspiracy. The claimants allege seven misrepresentations.

2 The dispute arises out of the Terra-Luna crash of 7 May 2022, a well-known episode in the crypto-world. An algorithmic “stablecoin”, TerraUSD or “UST”, pegged to the US dollar (“USD”) at a one-to-one exchange rate, lost its peg to the USD, and saw its exchange rate plunge against the USD, losing most of its market value. The claimants held UST tokens. They are aggrieved by the crash and seek compensation for their economic losses by invoking various causes of action. The first defendant has counterclaimed for a declaration that the claimants are bound by the terms and conditions hyperlinked on the first defendant’s websites.

3 Having considered the evidence and the parties’ submissions, I am of the view that: (1) the defendants made five fraudulent misrepresentations to the representative claimants which induced some of them to purchase UST in reliance; (2) the two remaining alleged representations are not actionable; (3) no unilateral contracts were formed between the defendants and any representative claimant; (4) the successful representative claimants are entitled to damages assessed on a reliance, as opposed to an expectation, basis; and (5) the claims for inducement of breach of contract and unlawful means conspiracy ought to be dismissed. I set out my reasons below, quantifying the damages due to successful representative claimants. I allow the first defendant’s counterclaim to the extent specified in this judgment.

4 The representative claimants, Mr Julian Moreno Beltran (“Mr Beltran”), Mr Douglas Gan Yi Dong (“Mr Gan”), Mr Daniel Davis (“Mr Davis”), Mr Ryan Connell Macquisten (“Mr Macquisten”), Ms Mariam binti Nawawi (“Ms Mariam”), Mr Xue Yao (“Mr Xue”), Mr Arun Kumar Sivaramavimalamoorthy (“Mr Arun”), Mr Andrew Joseph Che-Bin Lee (“Mr Lee”), Mr Phillip Robert Epstein (“Mr Epstein”), and Mr Tan Jian Loon (“Mr Tan”), have different nationalities and are based in different jurisdictions around the world. Their common characteristic is that they held UST tokens at the time of UST’s crash in May 2022. All ten representative claimants allege that they purchased or retained UST tokens in reliance on at least some of the alleged misrepresentations.

5 The 366 claimants can be categorised into ten classes, each class being constituted by claimants who allegedly relied on specific combinations of the seven pleaded representations. The representative claimants each represent one of the ten classes. The precise combination of representations which a representative claimant is said to have relied upon is unique to that representative claimant and the class that he or she represents (see the specific combinations reduced into tabular form at [38] below).

6 The first defendant, Terraform Labs Pte Ltd (“Terraform”), is a legal person incorporated in Singapore, with its principal business in software development and applications. Its majority shareholder is the second defendant, Mr Kwon Do Hyeong (“Mr Kwon”), a citizen of the Republic of Korea, who was the Director and Chief Executive Officer of Terraform. Mr Kwon was a member and a guarantor of the third defendant, Luna Foundation Guard Ltd (“LFG”), which was incorporated in Singapore on 29 December 2021 as a public company limited by guarantee.

7 Another guarantor of LFG was Mr Nikolaos Alexandros Platias (“Mr Platias”), Terraform’s Head of Research, a Greek citizen. He was previously a defendant in this representative action (see the General Division of the High Court’s decision in Beltran, Julian Moreno and another v Terraform Labs Pte Ltd and others [2024] 4 SLR 674 (“Beltran (Arbitration Stay)”) at [5]). Mr Platias and the claimants resolved their dispute. The claimants then applied to discontinue OA 3 against Mr Platias in SIC/SUM 9/2024, filed on 26 February 2024. I granted permission to discontinue on 4 March 2024 (see the court order numbered SIC/ORC 12/2024, extracted on 6 March 2024). On 1 April 2025, the claimants filed a statutory declaration (the “Platias SD”) made by Mr Platias on 27 March 2025, which they sought to rely on as evidence in these proceedings (see at [63] below).

8 The disputes between the representative claimants and the defendants stem from two cryptocurrencies – (1) UST, an algorithmic “stablecoin” that is part of the Terra blockchain protocol operated by Terraform and (2) UST’s associated sister cryptocurrency, LUNA tokens (“Luna”), which was not pegged to a fiat currency. Luna was intended to assist in maintaining UST’s peg to the USD by absorbing volatilities in UST’s market value via an arbitrage mechanism. The arbitrage mechanism was meant to incentivise arbitrageurs to take advantage of disparities between UST’s market value and the rate at which Luna could be exchanged for UST over the Terra protocol’s market module (see at [15]–[16] and [18] below for the operational details).

9 Another aspect of the “Terra Ecosystem” – a description used to refer to the decentralised network consisting of the Terra blockchain protocol, UST, and associated tokens and platforms – was the Anchor Protocol (“Anchor”), a lending and borrowing platform and application through which users could deposit or “stake” their USTs. Whenever a user staked USTs on Anchor, the transaction would be recorded as one whereby the user exchanged the USTs staked in exchange for an equivalent amount of “aUST”. The aUST reflected a right to claim a proportionate share of USTs within the pool of USTs deposited on Anchor at any one time. Mr Terence Lim Zheng Wei (“Mr Lim”), a former Terraform employee who gave evidence for Terraform and LFG, described aUST as “almost like an IOU”.

10 The defendants’ seven representations pertain to the functioning of UST, Luna, and Anchor within the framework of the Terra Ecosystem.

11 In April 2019, Mr Kwon and Mr Platias, along with two others, published a white paper entitled Terra Money: Stability and Adoption explaining the Terra Ecosystem (the “Terra White Paper”). The Terra White Paper was accessible from Terraform’s website at terra.money (the “Terra Website”). The Terra Website contained additional information about the Terra Ecosystem in a section under the heading “About the Terra Protocol”.

12 The Terra White Paper included the following section, the bolded and underlined words being the words relied on by the claimants to support the alleged First Representation:

13 The Terra Website contained the following statements, with the bold and underlined portions being the particulars highlighted in the claimants’ pleadings as constituting the alleged First Representation:

14 The pleaded meaning of the highlighted words is “that the UST tokens were stable by design, as they were pegged to a stable fiat currency”. This will be referred to as the “First Representation”.

15 The Terra White Paper contained the following wording, with the bolded and underlined portions being the particulars relied on by the claimants in their pleadings:

16 The Terra Website contained the following:

17 The claimants plead that the emphasised words meant “that the Terra protocol and its underlying token mechanics could maintain the price stability of UST regardless of market size, volatility or demand through an algorithm that enabled an arbitrage process to happen with UST’s sister token, LUNA, guaranteeing that UST would always return to the 1 USD peg”. This will be referred to as the “Second Representation”.

18 The Terra Website additionally contained the following information (see also the excerpt at [16] above):

19 The claimants say that the emphasised wording represented “that UST holders would always be able to exchange 1 UST for 1 USD’s worth of LUNA on the Terra protocol”. This will be referred to as the “Third Representation”. The Third Representation is pleaded as based only on information on the Terra Website. It is not pleaded as based on information in the Terra White Paper.

20 The First to Third Representations are collectively abbreviated as the “Terra Representations”. The claimants plead that the Terra Representations were made by Terraform and Mr Kwon (but not LFG). Additionally, the claimants plead that the Terra Representations formed a unilateral contractual offer extended by Terraform to the world at large, capable of acceptance by any user who purchased UST in reliance on the Terra Representations. The terms of the resulting contracts are said to track the pleaded meanings of the Terra Representations, reproduced at [14], [17], and [19] above.

21 In June 2020, Mr Platias and others published a white paper on the Anchor protocol entitled Anchor: Gold Standard for Passive Income on the Blockchain (the “Anchor White Paper”). The Anchor White Paper was accessible through a link on the Anchor protocol website at anchorprotocol.com (the “Anchor Website”).

22 The Anchor White Paper contained the following information, with the bolded and underlined portions being the particulars relied on by the claimants:

23 The claimants plead that the bolded and underlined words represented “that Anchor was a principal-guaranteed stablecoin savings product where UST holders could enjoy the stability of holding a stablecoin while earning passive income”. This will be referred to as the “Fourth Representation”.

24 A screenshot of a page on the Anchor Website included the following image:

25 The claimants plead that the image represented “that purchasers of UST would earn up to 20% Annualized Percentage Yield (“APY”) on Anchor if they staked their UST on the protocol” [emphasis in bold in original]. This is the “Fifth Representation”.

26 The Fourth and Fifth Representations are collectively abbreviated as the “Anchor Representations”, which the claimants plead were issued by Terraform and Mr Kwon (but not LFG). The claimants further plead that the Anchor Representations formed a unilateral contractual offer by Terraform to the world at large which was capable of acceptance by anyone who purchased UST in reliance. The terms of the resulting contract are said to have tracked the pleaded meanings of the Anchor Representations at [23] and [25] above.

27 On 22 February 2022, LFG issued a press release to the public which announced:

28 The claimants plead that the foregoing portions of the LFG press release meant that LFG “had successfully raised USD 1 billion through the sale of LUNA tokens as a reserve fund to maintain the UST peg to the US Dollar”. It is said that this caused several claimants to believe that LFG “would be able to maintain UST’s peg to the USD in any market situation due to its Bitcoin reserves”. LFG is pleaded to have made the Sixth Representation. It is further pleaded that Terraform made a unilateral offer to the world on the First to Sixth Representations. The contractual term offered by the Sixth Representation was supposedly “that [LFG’s] reserves would be able to protect the price stability of UST vis-à-vis the 1 USD peg”.

29 This contractual offer made by the “LFG Press Release” is alleged to have been capable of acceptance by anyone who purchased UST in reliance on the LFG Press Release. There is, however, a confusion in the claimants’ pleadings as to the identity of the contracting party or parties in relation to the LFG Press Release as well as the Terra and Anchor Representations. Earlier, I noted that the claimants seemed to be pleading that Terraform alone was the contracting party making unilateral offers on the Terra and Anchor Representations (see at [20] and [26] above). Elsewhere, however, the claimants plead that Terraform and LFG made unilateral contractual offers to the world via the LFG Press Release as well as the Terra and Anchor Representations. Thus, it is contended that “the 1st and 3rd Defendant made a unilateral offer to the world on the 1st to 6th Representations as the main terms of offer”. Terraform’s involvement in the LFG Press Release is not particularised. It is merely alleged that “the 3rd Defendant announced” the Sixth Representation. More pertinently, it is unclear how LFG was involved in the making of the Terra and Anchor Representations, since the claimants plead that LFG was only formed on 29 December 2021 (see at [6] above), and the Terra and Anchor White Papers predate LFG’s incorporation (see at [11] and [21] above).

30 Still elsewhere, the claimants suggest that all three defendants formed unilateral contracts with the claimants, stating under the sub-heading on breach of unilateral contract that “the 1st to 3rd Defendants had made the 1st to 7th Representations as set out above” and that the “representations were false and the Defendants were in breach of each of the express terms”. This would imply that even Mr Kwon was party to the alleged unilateral contracts, undertaking obligations thereunder and being capable of breaching their terms.

31 In contrast, when it comes to the relief claimed, there are only prayers to the effect that Terraform formed unilateral contracts with the claimants, not Mr Kwon or LFG. In closing oral submissions, the claimants’ counsel clarified in response to questions from me that only Terraform is alleged to have formed and subsequently breached unilateral contracts with the claimants. I will therefore approach the claimants’ case of breach of unilateral contract on this basis.

32 Following UST’s de-pegging from its one-to-one exchange rate with the USD, on 9 and 10 May 2022 (all dates and times for tweets are based on Universal Time Coordinated or “UTC”) Mr Kwon published the following tweets on the website now known as “X”, then known as “Twitter”, on his account (bearing the handle @stablekwon):

33 On 11 May 2022, all at 10.10am (forming a single “thread”), Mr Kwon tweeted:

34 “Lunatics” was common slang used to refer to avid supporters of the Terra Ecosystem. The claimants rely on the bolded and underlined excerpts of the 11 May tweets at [33] above, and all of the tweets at [32] above, in support of their pleaded meaning that “[LFG] were deploying more capital and had a plan to rescue the 1 USD peg that would be employed soon, that UST holders should not panic, and that UST and the Terra Ecosystem would ‘return to form’” [emphasis in original]. This will be referred to as the “Seventh Representation”.

35 In addition to the allegation that the Seventh Representation was a false representation of fact that is actionable in law, the claimants plead that it constituted a unilateral contractual offer by Terraform to the world. The offer was capable of acceptance by anyone who purchased UST on the strength of the tweets. A term of that unilateral contract is said to be a promise that Terraform and LFG “would come up with a rescue plan to restore the 1 USD peg for UST, and that the Terra Ecosystem would soon ‘return to form’” [emphasis in original]. I shall refer to this alleged unilateral offer as the “Return to Form Tweet”.

36 The next series of tweets came on 13 May 2022. These came after the function to exchange 1 UST for 1 USD’s worth of Luna tokens was disabled on 13 May 2022, as announced to the public through Terraform’s official Twitter account (with the handle @terra_money). Mr Kwon tweeted:

37 All tweets reproduced at [32]–[33] and [36] above remain publicly available on the website X (formerly Twitter) on Mr Kwon’s X account.

38 I summarise in tabular format which of the seven representations each representative claimant is relying upon as having induced him or her to purchase or retain USTs. I indicate each representation relied on with a cross (“x”) in the relevant box. The ordinal numbers reflecting each of the seven representations are shown on the horizontal heading (from “1st” to “7th”), and the ten representative claimants are indicated in the vertical boxes (abbreviated as “RC1” through to “RC10”, in the order listed at [4] above):

39 On 7 September 2022, the claimants filed their representative action (then HC/OC 247/2022 (“OC 247”)) in the General Division of the High Court prior to its transfer to the SICC as OA 3. Despite minor differences in the identities of the claimants and representations pleaded, OC 247 was a representative claim for (among others) fraudulent misrepresentation in respect of the Terra Ecosystem against the defendants and Mr Platias (see Beltran (Arbitration Stay) at [4]–[13]). The claimants were then represented only by Mr Beltran and Mr Gan (see Beltran (Arbitration Stay) at [6]).

40 On 1 March 2023, Terraform filed a summons for OC 247 to be stayed in favour of arbitration under s 6(1) of the International Arbitration Act 1994 (2020 Rev Ed). Mr Kwon and LFG sought case management stays dependent on the action against Terraform being stayed to arbitration (see Beltran (Arbitration Stay) at [1] and [29]).

41 The requested stay to arbitration was based on terms said to have been incorporated through notice by hyperlinks accessible from the Terra Website and Anchor Website (see Beltran (Arbitration Stay) at [14]–[16]). I shall employ the same abbreviations for the terms which Hri Kumar Nair J used in Beltran (Arbitration Stay), namely, “Terra Terms of Use” and “Anchor Terms of Service”. Nair J found that, although the terms contained arbitration clauses that formed prima facie arbitration agreements with the claimants (see Beltran (Arbitration Stay) at [151] and [161]), a stay should be refused as Terraform had voluntarily submitted to the court’s jurisdiction by taking numerous steps in OC 247 (see Beltran (Arbitration Stay) at [117]–[118]).

42 In relation to the Terra Terms of Use, Nair J analysed the claimants’ case thus in [142]–[144] of Beltran (Arbitration Stay):

43 Nair J observed on the Anchor Terms of Service and whether Mr Gan had constructive notice of them at [156]–[157] and [160]–[161] of Beltran (Arbitration Stay):

44 Mr Beltran was not relying at the time and does not now rely on the Anchor Representations (see at [38] above).

45 At a Case Management Conference of 6 January 2025, counsel for Terraform and LFG stated that they would amend their defences to concede to the First to Fourth and Sixth Representations. This was the result of a settlement reached between Terraform and Mr Kwon on the one hand and the United States Securities Exchange Commission on the other, over the latter’s complaint before the Federal Court of the Southern District of New York that the former had violated antifraud provisions in federal law. The settlement was arrived at on 12 June 2024 following a civil jury verdict rendered against Terraform and Mr Kwon on 5 April 2024.

46 Consequently, the defendants have conceded that the First to Fourth and Sixth Representations were misrepresentations of fact which had been made fraudulently (the “Conceded Representations”). There is no concession that the claimants were induced by the Conceded Representations, suffered loss in reliance thereon, or that the Fifth or the Seventh Representations are actionable misrepresentations (the “Disputed Representations”).

47 On 30 December 2024, the claimants took out a summons (SIC/SUM 67/2024 (“SUM 67”)) to bifurcate OA 3 into a stage for the determination of liability (to be heard on a representative basis) and a stage for the assessment of damages (to be heard on a non-representative or individualised basis). SUM 67 had been issued in response to Terraform’s and LFG’s summons (ie, SIC/SUM 56/2024 (“SUM 56”)) for OA 3 to be struck out on the ground that it was inappropriate for OA 3 to continue in a representative form. The defendants submitted that it was wrong to try a claim involving multiple claimants whose circumstances differed from one another as far as reliance, causation, and economic loss were concerned.

48 At a hearing on 13 February 2025, I dismissed the prayers for striking-out in SUM 56 and allowed the bifurcation sought in SUM 67 in a different form. I did not bifurcate OA 3 into one tranche for determining liability and one tranche for assessing damages. That would have been inefficient, since not all issues of liability and quantum were common among the claimants. For example, the question whether a claimant purchased or staked UST in reliance on a representation would depend on the circumstances of that claimant. The quantum of the claimant’s loss would then hinge on whether he or she was found to have relied on a misrepresentation.

49 I therefore bifurcated OA 3 into two tranches based on the more flexible provision in O 16 r 11 of the SICCR 202. That provides:

50 The result was that certain issues pertaining to liability could be heard in the first tranche, while other issues of liability were reserved for the second tranche. Such an approach addressed the defendants’ concern about the prejudice if their liabilities were solely determined on a representative basis, despite relevant factual circumstances differing among individual claimants.

51 It was clear, however, that OA 3 should proceed in a representative form in relation to those issues which were common to the claimants, such as whether the Disputed Representations are actionable misrepresentations. Trying these issues in relation to each claimant individually would not have been an efficient use of court resources.

52 I directed the parties to agree a list of the issues to be decided in the first tranche of the civil trial. The objective was to try the claims of all ten representative claimants in full in the first tranche, including all questions common to the individual claimants, with the remaining issues reserved for the second tranche.

53 The parties reverted with their lists of issues on 1 April 2025. The wording of all issues to be decided in the first tranche was agreed among them. However, the parties did not agree on whether three issues (namely, (1) did the defendants breach their duties of care in making the representations to the claimants, (2) did Mr Kwon cause Terraform to breach its unilateral contracts with the claimants through the disabling of the swap mechanism on 13 March 2022, and (3) the quantum of damages for which Mr Kwon would be liable in that scenario) were binding on all claimants or only the ten representative claimants. I allowed the parties’ agreed list of issues and directed that the three disputed issues would be decided in the first tranche. Whether a finding is binding on all claimants or only on the representative claimants is a question that will depend on the specific facts found. While the effect of a representative action is that this judgment will have a res judicata effect on all claimants, whether the judgment has the consequence of causing an individual claimant’s claim to succeed or fail will depend on the relevance of a factual finding here to the specific claim of that individual claimant.

54 I do not propose to set out the parties’ agreed issues in full here. As I told the parties at the trial, there was considerable overlap among the issues. In my view, the issues as encapsulated in [64] below capture the salient points raised in the parties’ lists.

55 On 3 April 2025, the claimants wrote to the Registry seeking orders for Mr Kwon and Mr Platias to attend court to give oral evidence per O 20 r 3(2) of the SICCR 2021.

56 In their letter to court dated 7 April 2025, Mr Kwon’s counsel objected to this request, citing futility and international comity concerns arising from the fact that Mr Kwon was in prison in America. Mr Kwon had been indicted for criminal fraud by a grand jury in the Federal Court of the Southern District of New York on 2 January 2025 and, according to his counsel, those proceedings remain ongoing.

57 On 15 April 2025 I refused the orders sought by the claimants, but on different grounds to those urged by Mr Kwon’s counsel. Since Mr Kwon remains in jail in the US and Mr Platias resides in Rodopoli, neither was compellable by a Singapore court per O 20 r 3(2) of the SICCR 2021, as neither was resident within Singapore (see JIO Minerals FZC and others v Mineral Enterprises Ltd [2011] 1 SLR 391 (“JIO”) at [71], which held that “a Singapore court cannot compel a foreign witness to testify in a Singapore court”.

58 I allowed Mr Kwon’s counsel to file and serve the latest version of Mr Kwon’s defence, despite the absence of a signed statement of truth. Mr Kwon’s counsel cited difficulties in procuring his client’s signature on account of the ongoing criminal proceedings mentioned at [56] above. On 2 May 2025, I directed the filing of the unsigned defence by 6 May 2025. Mr Kwon’s counsel did so on 5 May 2025. Although O 6 r 11(1) of the SICCR 2021 requires that pleadings be verified by a statement of truth, O 6 r 11(13) provides: “[u]nless the Court orders otherwise and subject to paragraph (14), a pleading which is not verified by a statement of truth remains effective but may not be relied on as evidence of any of the matters set out in it”. I was not minded to exercise my discretion to reject the filing of Mr Kwon’s unsigned defence under O 6 r 11(13). Nor was I minded to strike out the unsigned defence under O 6 r 11(14). This was due to the real difficulties faced by Mr Kwon’s counsel in procuring his signature on the statement. In my view, it would better serve the administration of justice and the fair conduct of OA 3 for Mr Kwon’s position to be reflected on the record (see O 1 r 3(1) of the SICCR 2021).

59 On 13 May 2025, at the start of trial, I disposed of further interlocutory matters. In particular, there was the representative claimants’ SIC/SUM 37/2025 (“SUM 37”) for permission to amend their originating claim, statement of claim, and further and better particulars. The amendments involved the removal of four represented claimants who were no longer party to OA 3, revisions to the damages sought by the claimants in their updated further and better particulars, and changes to the representations relied upon by specific represented claimants. These amendments were consequential upon the further and better particulars of damages which I ordered at the hearing of 13 February 2025 at [48] above in respect of all claimants, including the represented claimants. -

60 I did not think that allowing the amendments in SUM 37 would prejudice the defendants in the conduct of their defence. Hence, I allowed SUM 37 on 13 May 2025, with costs to the defendants. Terraform and LFG submitted that SUM 37 was a backdoor attempt to circumvent the deadline by which further and better particulars had to be produced. I disagreed. Their criticisms of the measures of damage pleaded in the latest round of amendments was a matter which could be taken up in closing submissions. It was conducive to a fair trial for the claimants’ pleadings to reflect their position on damages, enabling the real questions between the parties to be properly ventilated.

61 The latest statement of claim was filed on 14 May 2025. Terraform and LFG filed amended defences on 28 May 2025, after the start of trial on 13 May 2025. These were not wholly related to the claimants’ amendments of 14 May 2025, but instead introduced an unparticularised plea that the claimants were under a duty to mitigate their losses and failed to take all reasonable steps to do so. Although the claimants objected to the new plea, I allowed the same. But this was solely for the purpose of bringing to the fore an issue which, in the course of witness examination, I had directed the parties to consider. That issue was whether the buying or selling of UST by a claimant following the May 2022 de-peg should be treated as an attempt by that claimant to mitigate some of his or her losses. If the characterisation of such trading as “mitigation” was valid, did it follow that, when assessing damages, (a) the defendants should be credited with gains made by a claimant through such trading and, conversely, (b) the defendants should compensate a claimant for losses sustained as a result of unsuccessful attempts to salvage one’s position.

62 I raised a query on the law applicable to the pleaded misrepresentations. A question in the parties’ list of issues was whether the Disputed Representations satisfy the requirements of a claim for damages under the Misrepresentation Act 1967 (2020 Rev Ed) (the “MA”). This was potentially problematic as it was unclear that the parties accepted that Singapore law (including the MA) governed the alleged negligent misrepresentations. Case law holds that the MA does not apply to misrepresentations occurring outside of Singapore in relation to contracts not governed by Singapore law (see JIO at [104]). In some instances, the alleged misrepresentations would arguably have been made (and relied on) outside of Singapore. If so, the double actionability rule in private international law would apply. But none of the parties had referred to the actionability of the misrepresentations as civil wrongs in jurisdictions other than Singapore. This was presumably because the parties regarded Singapore law as the proper law of the misrepresentations. I asked whether that was actually the parties’ position? Having considered the foregoing conflict of law query at my invitation, the parties informed me that they were content to proceed on the basis that Singapore law (including the MA) governed the alleged misrepresentations and unilateral contracts.

63 Finally, there was the matter of the admissibility of the Platias SD (see at [7] above). I held that the Platias SD’s contents were admissible under s 32(1)(j)(iv) of the Evidence Act 1893 (2020 Rev Ed) (the “EA”). The section concerns witnesses who are competent but not compellable to testify and who refuse to give oral evidence. Mr Platias resides in Rodopoli and is non-compellable (see at [57] above). The claimants’ letter of 9 May 2025 noted that the claimants had contacted Mr Platias to seek his attendance at trial, but he was not agreeable. Despite admitting the Platias SD, I reminded the parties that the admissibility of the document was a separate question from the weight and value (if any) to be accorded to that evidence. I did not exercise my discretion to exclude the Platias SD under s 32(3) of the EA, after applying the factors for consideration in Gimpex Ltd v Unity Holdings Business Ltd and others and another appeal [2015] 2 SLR 686 at [103]–[109].

64 I proceed to consider the following topics based on the parties’ list of issues for the first tranche of the civil trial (see at [54] above):

65 The representative claimants submit that the Disputed Representations are actionable misrepresentations. The Fifth Representation on the Anchor Website represented that those who staked their USTs on the Anchor platform would earn up to 20% annualised percentage yield from such staking. That is not a statement of future, but of present, fact. It is (the claimants say) “a statement of fact about the returns that a user would obtain if you stake”.

66 Likewise, the Seventh Representation, especially the Return to Form Tweet (see at [33]–[35] above), was a representation of present fact that “a rescue plan existed which will result in it [UST] returning to form”. In other words, there was a plan which would restore UST to its previous form.

67 Terraform and LFG submit that the Fifth Representation was not a false representation of fact as the screenshot at [24] above did not guarantee that users who staked UST on Anchor would continually earn anything close to 20% of annualised percentage yield. There was therefore nothing false about the Fifth Representation.

68 The Seventh Representation did not state that UST would revert to a one-to-one peg with the USD in the tweeted “return to form”. All of the tweets of 11 May 2022 by Mr Kwon, read in context, made it plain that Mr Kwon was talking about the Terra Ecosystem as a whole (not UST’s peg to the USD specifically) returning to form. There was no representation as to the precise nature of that “return”. The Seventh Representation could not therefore have constituted an actionable misrepresentation of fact.

69 Mr Kwon argues that the Fifth Representation, at best, represented a present intent on Mr Kwon’s part as to the yield which users could earn from staking USTs on the Anchor platform. The representative claimants have failed to adduce any evidence to suggest that Mr Kwon lacked a genuine intention of enabling users to earn up to 20% annualised percentage yields from the Anchor platform. Accordingly, there cannot have been a misrepresentation. This was not a case where Mr Kwon made the Fifth Representation without a reasonable belief in its truth at the time when made.

70 On the Seventh Representation, Mr Kwon argues that there is no evidence to suggest the absence of any plan to bring about the recovery of UST or the Terra Ecosystem. Whether that recovery plan failed or succeeded in the end is irrelevant. The representative claimants have failed to establish that there was no plan.

71 An actionable misrepresentation is a false statement of fact, whether existing before or at the time the representation is made. It may include representations of a person’s present mental state or intentions. However, statements as to the future are not actionable as misrepresentations unless they are a valid contractual promise (see Tan Chin Seng and others v Raffles Town Club Pte Ltd [2003] 3 SLR(R) 307 (“Tan Chin Seng”) at [12]–[13] and [20]–[21]).

72 In determining the proper meaning of a representation, whether express or implied, the court’s approach is a subjective one, that is, based on the meaning that was intended by the representor and not the meaning that would be understood by a reasonable observer (see Wee Chiaw Sek Anna v Ng Li-Ann Genevieve (sole executrix of the estate of Ng Hock Seng, deceased) and another [2013] 3 SLR 801 (“Anna Wee”) at [37] and [41]).

73 Applying the principles at [71]–[72] above, I accept that the Fifth Representation was intended by Terraform and Mr Kwon to bear the pleaded meaning, namely, “that purchasers of UST would earn up to 20% Annualized Percentage Yield (“APY”) on Anchor if they staked their UST on the protocol” [emphasis in bold in original]. This is based on the screenshot of the Anchor Website at [24] above, showing the availability of an APY rate of 19.48% at 5 January 2022. I come to this conclusion in the context of the screenshot being promotional material on the Anchor Website intended to market the Anchor protocol to potential users.

74 But I disagree that the pleaded meaning can be read as a representation of present fact. Rather, it is a representation as to the future, namely, that users who staked their USTs on the Anchor protocol would earn up to 20% APY rates in respect of future deposits.

75 I accept that a representation as to a future fact could impliedly represent certain facts about a present state of affairs. For example, a representation that a future set of facts will inure from a given enterprise may impliedly represent something about that enterprise in the present that would make the predicted future state of affairs possible. Here, it could be argued that the Fifth Representation impliedly represented that the makers knew of no present facts that rendered it impossible for users staking their USTs on Anchor to earn any APY rates close to 20%.

76 But there is no evidence establishing such impossibility. The representative claimants have argued that the Fifth Representation was false because the representors must have realised that APY rates of close to 20% were not sustainable in the long run. This was because the defendants are claimed to have established a yield reserve to buttress the APY. If there was a back-up yield reserve, the defendants must have known that there would not always be sufficient yields generated from interest earned on loans to borrowers or on collateralised staked assets to sustain APYs of close to 20%. The difficulty with this submission is that there is nothing to support the conclusion that the representors were asserting by the Fifth Representation that APY yields would not be subsidised by yield reserves.

77 Taking the screenshot at [24] above, which is the pleaded source of the Fifth Representation, there is nothing in the screenshot to suggest that the APY rates would only come from interest on loans to borrowers or collateralised staked assets to the exclusion of any yield reserve. Nothing in the screenshot suggests that the APY rates would (a) remain close to a 20% APY rate in the long run in a sustainable fashion or (b) APY rates may not materially deviate from 20%. The claimants’ pleaded meaning is careful to caveat the representation’s meaning with the words “earn up to 20%” [emphasis added]. To be false, it must be shown that the representors knew at the time that it was impossible for users to earn “up to 20%” APY rates from the Anchor protocol. This has not been demonstrated on the evidence.

78 A representation as to the future can also include an implied representation as to the representor’s present state of mind. But there is no evidence that Terraform or Mr Kwon lacked a true intention to pay yields of up to 20% APY to users who staked USTs on Anchor.

79 I conclude that the Fifth Representation was a representation of a future state of affairs which is not actionable in misrepresentation. I am unable to find any implied representation of past or present fact that has been shown to be false.

80 I find that the Seventh Representation did not bear the meaning pleaded by the claimants. There is no evidence that Mr Kwon subjectively meant to represent by his tweets that the defendants “had a plan to rescue the 1 USD peg that would be employed soon, that UST holders should not panic, and that UST and the Terra Ecosystem would ‘return to form’” [emphasis in original]. There is nothing in the Return to Form Tweet that supports an inference that this was the meaning subjectively intended by Mr Kwon at the time (see at [72] above).

81 The Return to Form Tweet has to be read in context (see Anna Wee at [36]), including not just the specific tweets highlighted by the claimants in their pleadings, but the entire series of tweets published by Mr Kwon on 11 May 2022 (see at [33] above). That is regardless of whether individual claimants saw all the tweets of 11 May 2022 or read only some, since the question of the meaning of a representation is not assessed from the perspective of the representee or even an objective reasonable reader, but the representor’s subjective meaning (see Anna Wee at [37] and [41]).

82 Reading Mr Kwon’s tweets of 11 May 2022 together, I do not accept that the Return to Form Tweet, the last in the series, can be understood to mean that there was a plan to restore the one-to-one UST-to-USD peg. The tweet instead only states that “Terra’s return to form will be a sight to behold. We’re here to stay. And we’re gonna keep making noise”. In Mr Kwon’s earlier tweets of 11 May 2022, he mentioned that the Terra Ecosystem was “one of the most vibrant in the crypto industry” and the rebuilding of UST would take the form of “adjust[ing] its mechanism to be collateralized”, implying changes to UST’s mode of operation.

83 Although by his 11 May 2022 tweets Mr Kwon conveyed that he and his team were in the midst of coming up with a recovery plan to improve the then-current state of the Terra Ecosystem, I cannot find anything to suggest that the specific recovery plan intended was one which would return UST to its original form of a one-to-one peg to the USD. In so far as the representative claimants fix on the phrase “return to form” to suggest that Mr Kwon meant that the UST would revert to its pre-depeg state without any changes, I cannot agree.

84 The representative claimants have failed to show that no recovery plan was in the works at the time of Mr Kwon’s tweets. On the contrary, the references in Mr Kwon’s tweets to (a) an endorsement of “community proposal 1164 to Increase basepool from 50M to 100M SDR” to “increase minting capacity” and “allow the system to absorb the UST more quickly” and (b) the discussion of further ideas in the community forums, support an inference that there was active exploration of avenues to improve the then condition of the Terra Ecosystem.

85 Consequently, the Seventh Representation is not an actionable misrepresentation. The Disputed Representations do not give rise to actionable claims, whether for deceit, negligent misstatement, innocent misrepresentation giving rise to rescission, or damages or other relief under the MA.

86 Terraform’s counterclaim seeks declarations that the Disputed Representations “w[ere] not false”. I decline to grant such wide declarations. I have found that the representative claimants have failed to establish that the Disputed Representations are actionable. It does not follow that I have found that the Disputed Representations were true.

87 The representative claimants emphasise that the defendants have accepted that the Conceded Representations were false statements of fact fraudulently made. Under the tort of deceit, reliance is made out so long as a fraudulent misrepresentation was a cause of a representative claimant purchasing UST. The fraudulent misrepresentation need not have been the sole or dominant cause of a representative claimant so doing.

88 It is not necessary for the representative claimants to have understood and acted on the representation in an objectively reasonable fashion. The test is not whether a person reasonably relied on a fraudulent misrepresentation. All that is required is that a representee was subjectively induced into acting based on the fraudulent misrepresentation. It is no defence to argue that a reasonable person in the representee’s shoes would have acted differently had due diligence or caution been exercised.

89 Terraform and LFG contend that the representative claimants did not rely on the Conceded Representations. First, they argue that the relevant test is one of “reasonable reliance”. Such criterion has not been satisfied because the representative claimants did not truly understand the meaning of the Conceded Representations in the Terra and Anchor White Papers and the Terra and Anchor Websites. On the contrary, the representative claimants were sophisticated or experienced crypto-investors who could not have believed that the Conceded Representations bore the pleaded meanings. Some representative claimants were instead engaged in opportunistic swing trading, buying USTs low to sell high once the de-peg of May 2022 presented a chance to make quick profits from UST’s volatile market movements. Such representative claimants did not act in reliance on the Conceded Representations.

90 Terraform and LFG raise several ancillary attacks bearing on the question of reliance, such as allegations that the representative claimants failed to adduce sufficient proof of having read the documents containing the Conceded Representations (including having accessed the Terra and Anchor Websites at relevant times) or failed to adduce credible evidence of their ownership and control of the private wallets through which UST-related transactions were conducted. Terraform and LFG observe that portions of the representative claimants’ witness statements concerning their alleged reliance on the Conceded Representations are nearly identical (“cut & paste”), supporting an inference that their evidence is contrived.

91 Mr Kwon emphasises that, even in cases of fraud and deceit, there is no presumption of reliance. The burden remains on the representative claimants to adduce evidence that they were induced into purchasing or retaining UST tokens in reliance on the Conceded Representations. The issue is whether reliance can be inferred from the evidence. Mr Kwon makes similar arguments to those advanced by the other defendants to argue that reliance should not be inferred.

92 The defendants have admitted to the Conceded Representations being fraudulent misrepresentations (see at [46] above). That having been accepted, the test is whether the Conceded Representations were a cause of loss. The Conceded Representations need not have been the sole or dominant cause of loss.

93  I accept that there is no presumption that a fraudulent misrepresentation means that representees relied on the same (see Anna Wee at [45]). The issue is whether an inference can be drawn that, per Anna Wee at [44], a misstatement “act[ed] upon the will of the representee” and “influences or leads the representee to change her behaviour in reliance on the misrepresentation” [emphasis in original]. However, where a representation is fraudulently made, in the sense of being intended to induce a course of action in reliance, reliance by a representee will often readily be inferred (see Edgington v Fitzmaurice (1884) 29 ChD 459 at 483–485). Nonetheless, it ultimately remains a question of fact (William Smith v David Chadwick, John Oldfield Chadwick, Ebenezer Adamson, and Edwin Collier (1884) 9 AppCas 187 (“Smith v Chadwick”) at 195–197).

94 The Conceded Representations were made in promotional materials to market the characteristics of USTs, Anchor, and the Terra Ecosystem. The tenor of the Terra White Paper, the Anchor White Paper, the Terra and Anchor Websites, and the LFG Press Release was to tout the positive features of the Terra Ecosystem. The fair inference to be drawn is that the representors were seeking to encourage representees to believe in the positive characteristics of UST and the Anchor platform and acquire USTs or retain their existing USTs.

95 In this light, the Conceded Representations must have been material to the decision to purchase or retain USTs or to stake USTs on the Anchor platform. The message was essentially that the UST’s peg to the USD meant that the UST was stable by design and the Anchor platform represented a means for users to earn passive income while enjoying the UST’s stability. In all likelihood, representative claimants who saw the Terra or Anchor White Papers and the Terra or Anchor Websites would have been influenced – even if only in part – by such promotional materials into purchasing or retaining USTs. Thus, the Conceded Representations, while possibly not the sole inducement underlying relevant UST transactions, would have “played a real and substantial part and operated in [the] minds” of those who bought or held onto USTs on the strength of what they read in the Terra or Anchor White Papers and the Terra or Anchor Websites (see Panatron Pte Ltd and another v Lee Cheow Lee and another [2001] 2 SLR(R) 435 (“Panatron”) at [23]).

96 A representee does not need to have acted in an objectively reasonable fashion in understanding a representation or entering into a transaction based on it. Terraform and LFG cited a number of cases for the proposition that “reasonable reliance” is an ingredient of the tort of deceit. But I am not persuaded that a representee must be shown to have acted reasonably in relying upon a representation. The Court of Appeal in Panatron (at [24]) held that it was no defence that the representees, as “knowledgeable and experienced businessmen”, ought to have acted differently with reasonable due diligence. Instead, “[a]ll that is required is reliance in the sense that the victims were induced by the representations. Once this is proved, it is no defence that they acted incautiously and failed to take those steps … which a prudent man would have taken”. Broadley Construction Pte Ltd v Alacran Design Pte Ltd [2018] 2 SLR 110 affirmed (at [36]) that “[i]t is still the law that representees are not obliged to test the accuracy of the representations made to them”. Thus, it is no answer for the defendants to say that the representees, as experienced or sophisticated crypto-investors, ought to have known better and should have realised that representations as to the stability of the UST-USD peg in all market conditions were too good to be true.

97 That is not to suggest that the subjective understanding of a representee will never be relevant. It is apparent from Anna Wee at [48]–[49] that the onus remains on a representee to show that he or she understood a representation in the sense pleaded and the “utterance was false in that sense”. But this is very different from a claim that there must be “reasonable reliance”. The question is whether the representative claimants saw, and understood from, the documents embodying the Conceded Representations that USTs were stable by design.

98 In summary, I approach the questions of reliance and causation as follows:

(a) It must be shown that a representative claimant read at least some of the pleaded sources of the Conceded Representations. It is for the claimants to decide what evidence to adduce in discharging that burden (for example, browsing histories or sworn averments in witness statements). There is no rule of law that an adverse inference is to be drawn merely because a claimant produced one form of evidence instead of another. Everything turns on the credibility of the evidence put forward, such evidence being assessed in the round on the balance of probabilities.

(b) It must be shown that a representative claimant subjectively understood the Conceded Representations to bear the pleaded meanings, which are here admitted to be false. That subjective meaning need not be something that a reasonable person would have understood. It is not relevant whether a representative claimant understood the exact operational mechanics of the UST-Luna arbitrage mechanism or how the stability of UST was to be maintained. The question is whether the representative claimant derived the pleaded meaning from a source document.

(c) It must be shown that the representative claimant transacted in UST in reliance on some combination of the Conceded Representations. Some or other of the Conceded Representations must have been “actively present to their mind” when representative claimants transacted in UST (see Axis Megalink Sdn Bhd and another v Far East Mining Pte Ltd [2024] 1 SLR 524 at [139]).

(d) The pleaded meanings of the Conceded Representations are material to an ordinary investor’s decision to transact in USTs (see at [93] above). Given the intention to induce UST transactions (see at [94] above), it would be a fair inference that the Conceded Representations induced a representative claimant to transact in UST (see Smith v Chadwick at 196). However, external circumstances may support a countervailing inference. Such external circumstances would include evidence that a representative claimant (a) did not really believe in a representation, (b) ceased to believe in a representation’s truth at a certain juncture, or (c) transacted in USTs for a different purpose. An example of a different purpose would be “swing trading” in which a claimant buys USTs when low and sells on rises in UST’s value with a view to speculating on short-term ups and downs in UST’s value in a volatile market.

(e) While the claimants bear the burden of proving that they own the cryptocurrency wallets through which relevant transactions were conducted, it is for them to decide how to discharge that burden (for example, by furnishing API keys, videos of them accessing their wallets, or something else). Whether that evidence is sufficient to discharge their burden is a question of the credibility of the evidence on the balance of probabilities. It is not a matter of drawing adverse inferences merely because one form of evidence was employed as opposed to another.

(f) The similarities among the claimants’ witness statements in relation to the Conceded Representations are a factor (by no means conclusive) to take into account in assessing the credibility of a witness’ evidence.

99 Applying the principles summarised at [98(a)]–[98(f)] above, I begin with the calculations in Mr Lim’s witness statement of the gains and losses of the representative claimants’ UST transactions. The representative claimants have each provided their own calculations in their witness statements. However, I prefer Mr Lim’s calculations as the better starting point for my quantification. Counsel for the representative claimants, in Mr Lim’s cross-examination, commented that the principal differences between Mr Lim and the representative claimants’ quantifications arose from (a) accounting (or not) for yields from depositing USTs on Anchor or other platforms and (b) differences in exchange rates used. The latter category accounted for minor differences between Mr Lim’s numbers and those of the representative claimants. The former category accounted for greater divergences between the two sets of figures.

100 Mr Lim’s calculations take account of yields generated from purchases of UST, whether staked on Anchor or another platform (such as “Gemini Earn” (in Mr Tan’s case) and “Binance Earn” (in Mr Davis’ case)). The yields are relevant to the reliance measure of damages and should be credited to the defendants. I reject the representative claimants’ distinction between intrinsic and non-intrinsic benefits. The latter category is alleged to include yields derived by staking on platforms such as Anchor, Binance Earn or Gemini Earn. Those yields are said to have stemmed from an “independent decision to stake or not to stake”. But such a distinction is irreconcilable with the reliance measure of damages, which is supposed to restore a representee to the position that he or she would have been in if the misrepresentation had not induced him or her to purchase or stake UST.

101 Profits which stem from the representee having been induced to acquire USTs – whether from a purchase and subsequent sale of UST tokens or from yields derived from depositing or staking USTs on a platform – would inevitably flow from the act induced by the representation. Yields must therefore be taken into account when restoring a representee to his or her original position. I do not think that yields can be disregarded as “non-intrinsic” or somehow the result of a wholly independent decision to stake UST tokens. Such an approach would place a representee into a better position than if the induced transaction had never occurred. It would constitute an unwarranted windfall to a claimant.

102 Mr Lim put forward several permutations of damage calculations. The calculations that I prefer are those which fall under the “Loss – Evidence Only (USD)” heading of the tables that he furnished. He describes that heading:

103 Mr Lim cross-checked transactions tied to the wallet addresses of representative claimants as disclosed within their further and better particulars, against accessible data on blockchain networks. He and his team: (1) extracted the transactions carried out on a wallet, (2) consolidated the inflows and outflows of USTs by taking UST’s hourly value at the time of a transaction based on Coingecko, a publicly available cryptocurrency data aggregator, and (3) used the value of total inflows and outflows to calculate the net loss sustained by a representative claimant. Mr Lim explained in cross-examination that differences between his calculations and those of the representative claimants were often attributable to his inclusion of yields from other platforms. For instance, Mr Lim accounted for Mr Tan’s staking of USTs on Gemini Earn as an outflow of USTs and did likewise with the yield which Mr Davis obtained from staking USTs on Binance Earn. In relation to a sum of about 31,000 USTs that Mr Gan claimed to have gifted to anonymous online friends and which Mr Gan excluded from his calculations of losses, Mr Lim treated the amount as an outflow of USTs. I agree with that treatment. While Mr Gan is entitled to deal with his USTs as he pleases, his net losses for the purposes of these proceedings cannot turn on the manner in which he disposes of his USTs by sale or gift. If Mr Gan had sold his UST tokens for USD and gifted the USDs to a friend, the USDs earned from the sale must be credited to the defendants as reducing Mr Gan’s net loss. The position should be the same if Mr Gan gifts the USTs instead.

104 Although I accept Mr Lim’s calculations as a starting point, I must have regard to potential attempts by the representative claimants at mitigation. Immediately after the de-peg, some representative claimants may have attempted to salvage their situation by swing trading. Given the volatility of UST, a representative claimant might have sought to buy when UST swung down and to sell shortly thereafter when UST swung up. Such endeavours may have enabled a representative claimant to recoup some loss. But the endeavour may have also caused a loss, if the representative claimant misjudged the market and, instead of swinging up, UST’s value went further down. I therefore invited submissions from parties as to an appropriate long-stop cut-off date beyond which a representee who continued to invest in USTs must be said to have been acting in a purely speculative manner. At such point, any gain or loss incurred thereafter would effectively be the result of a novus actus interveniens which should no longer be attributable to the original misrepresentations (see Doyle v Olby (Ironmongers) Ltd and others [1969] 2 QB 158 (“Doyle v Olby”) at 168–169 (per Winn LJ)). Any loss or gain from playing the UST market after the cut-off should be treated as purely the outcome of a claimant’s personal decision to undertake a risk. By similar token, if a representative claimant retained UST beyond the cut-off, that should be regarded as being for speculative purposes (that is, betting on an upturn). If the representative claimant makes a loss by eventually selling at less than the price on the cut-off date, he or she must bear the loss. On the other hand, if the representative claimant makes a gain by eventually selling at more than the price on the cut-off date, he or she gets to keep the gain as the fruit of his or her speculation or bet on an upturn.

105 Terraform and LFG proposed a number of alternative dates.

106 First, they suggested the start of the UST’s de-peg from the USD on 7 May 2022. In my view, this date would be too early. An investor in the representees’ shoes could reasonably have believed that the de-peg was only temporary on 7 May 2022. Mr Lim’s witness statement shows the fluctuations in the value of UST in the days immediately following 7 May 2022, particularly between 7 and 13 May 2022. The closing position for UST on 7 May 2022 was 0.9969, close to the one-to-one peg and higher than the lowest value (0.9927) at which UST traded throughout that day. It would be unrealistic to expect a claimant to have concluded on 7 May 2022 that the UST was unlikely to regain its original one-to-one peg with the USD.

107 Second, Terraform and LFG put forward the date of the Sixth Representation, namely, the LFG Press Release of 22 February 2022 (see at [27] above). This alternative rested on the premise that the Sixth Representation – regarding the building up of Bitcoin reserves as collateral to defend the value of UST – was inherently contradictory to the absolute stability of the UST’s peg to the USD and the immutability of the UST’s arbitrage mechanism. It was submitted that any representative claimant who read the LFG Press Release would have ceased to believe in the Terra Representations from then on, with the effect that later trades should be excluded.

108 I reject this approach. Whether the Terra Representations ceased to operate on the minds of a representative claimant is assessed on a subjective basis, that is, how the representee read those representations at the time. For the reasons I give in respect to individual representative claimants, the evidence does not support an inference that any of the representative claimants regarded the LFG Press Release as incompatible with the Terra Representations when they read the LFG Press Release. Nor am I persuaded that the Sixth Representation is incompatible with the Terra Representations. On the contrary, fairly read, the LFG Press Release supports an interpretation that the Bitcoin reserves were meant to buttress public confidence in UST’s peg in the USD and remove doubts among crypto-investors about the sustainability of the UST’s value. That is clear from Mr Platias’ statement in the LFG Press Release that the reserve means that “the primary counter-argument for the sustainability of algorithmic stablecoins is eliminated”. Mr Kanav Kariya’s statement in the LFG Press Release that the reserve “further strengthens confidence in the peg of the market’s leading decentralized stablecoin UST” is to similar effect. So is the sentence in the LFG Press Release that the “[r]eserve provides an additional avenue to maintain the stability of the peg in contractionary cycles that reduces the reflexivity of the system”. A reasonable person would not necessarily read the LFG Press Release to mean that the arbitrage mechanism was insufficient to defend the UST’s peg to the USD.

109 Third, Terraform and LFG proposed a cut-off of 12 May 2022 at 00.01 hours UTC. The proposal derives from Mr Kwon’s series of tweets on 11 May 2022 at [33] above, coupled with about 14-hours’ grace period for the representees to learn of the tweets, digest their information, and liquidate their UST holdings. I accept this approach. I have held that the Seventh Representation is not an actionable misrepresentation of fact (see at [85] above). On the contrary, it seems to me that a reasonable reader, considering the entirety of Mr Kwon’s tweets of 11 May 2022, would have appreciated that UST’s value was not stable and the mechanisms (algorithmic trading, Bitcoin reserves, etc) intended to guarantee its peg to the USD were incapable of safeguarding UST’s “stablecoin” characteristics or withstanding market volatility.

110 By this juncture, UST would have failed to re-peg to the USD over four days since the start of the de-peg on 7 May 2022. On 11 May 2022, UST’s value dipped to a low of 0.2998, opening at 0.7994 and closing at 0.8011. The instability of its value and the non-immutability of its peg to the USD would have been apparent to any reasonable investor by this point. The hopelessness of the peg would have been reinforced by Mr Kwon’s tweets, which highlighted the following:

(a) UST was then “trading at 50 cents, a significant deviation from its intended peg at $1”.

(b) The UST-Luna arbitrage mechanism was having difficulty restoring UST’s value and absorbing the fall in demand for UST, because “the cost of absorbing so much stablecoins at the same time has stretched out the on-chain swap spread to 40%, and Luna price has diminished dramatically absorbing the arbs [ie, the arbitrage trades]”.

(c) Any remedy to the “supply overhang of UST (i.e., bad debt)” would require solutions coming “at a high cost to UST and LUNA holders”, considering “the current on-chain spread, peg pressure, and UST burn rate”.

(d) UST needed to adjust its stability mechanisms to one where UST’s value was backed by assets serving as collateral. According to Mr Kwon, “we will adjust its mechanism to be collateralized”. This would have made it plain that UST’s arbitrage trading algorithm could not safeguard UST’s peg to the USD and LFG’s Bitcoin reserves would not suffice to restore the peg.

111 Within the 14-hour grace period posited by Terraform and LFG, a reasonable investor would have appreciated that all five of the Conceded Representations pertaining to the stability of UST’s value and its peg to the USD could not be true. Any further belief in the truth of the Conceded Representations after 00.01 hours UTC on 12 May 2022 would involve a representee failing to act with “reasonable prudence, reasonable common sense”, such that he or she can properly “be said to have been the author of his [or her] own misfortune” (per Winn LJ in Doyle v Olby at 168). Within the grace period, a reasonable investor, acting prudently to mitigate loss, would have ceased to acquire further UST tokens and liquidated his or her existing UST holdings to protect one’s portfolio against further falls in the UST’s value.

112 If then a representee voluntarily chose to trade in UST after 00.01 hours UTC on 12 May 2022 in the hopes of making a personal profit through swing trading, they would be doing so at their own risk. Any resulting losses would not be claimable against the defendants. But, equally, any profits from such trades should not be treated as reducing the net losses claimable against the defendants. If a representee chose to retain USTs in the hopes that the value of UST would somehow recover for whatever reason, that would be a legitimate choice. But the representee would so decide at one’s own risk. Any losses resulting cannot be said to flow from the fraud, once it was evident that the Conceded Representations were false.

113 I reject the alternative cut-off date of 14 May 2022 at 22.15 hours UTC suggested by the representative claimants. This timing takes, as a starting point, Mr Kwon’s tweets of 13 May 2022 at [36] above, which expressly stated that UST’s one-to-one peg to the USD would not be restored. In particular, Mr Kwon observed: “I still believe that decentralized economies deserve decentralized money – but it’s clear that $UST in its current form will not be that money”. Instead of the 14-hour grace period of Terraform and LFG, the representative claimants proposed a 24-hour grace period, leading to a cut-off of 14 May 2022 at 22.15 hours UTC. This strikes me as too late. A reasonable investor would long since have appreciated the unsustainability of UST’s one-to-one peg with the USD by this juncture. Mr Kwon’s 13 May tweets would only have confirmed what must have been obvious by then that the UST’s peg to the USD was unsustainable.

114 Consequently, in calculating the net loss of a representative claimant, any UST tokens purchased in reliance on the Conceded Representations which are still held by them at the cut-off date is taken to be sold at 00.01 hours UTC, 12 May 2022 (“the cut-off”), at the opening price of UST for that day – 0.8011. If their USTs were sold before the cut-off, the profits to be credited to the defendants are simply the actual amounts received from that sale. If their USTs were only sold after the cut-off (or never sold at all), the relevant outflow to be credited to the defendants is the value of their UST holdings at the cut-off, valued at 0.8011 per UST token.

115 I will now apply the principles of reliance, causation, and quantification of damages that I have articulated at [100]–[114] above to an examination of the circumstances of each representative claimant.

116 Mr Beltran estimated his loss at an amended figure of US$638,004.11 from trades in UST between 10 and 12 May 2022. Mr Lim’s calculations for Mr Beltran’s loss are slightly higher – US$641,056.55. Although a higher estimate, I shall take Mr Lim’s calculation as the starting point for an assessment of Mr Beltran’s position. Mr Lim accepted that the minor difference between the two figures is “[p]robably” attributable to the different exchange rates used.

117 Mr Beltran made many trades after the cut-off, selling Bitcoin for UST on Binance. These trades accounted for over US$100,000.00 of his losses. Applying the approach that I explained at [112] above, I exclude these trades from the losses claimable by Mr Beltran against the defendants.

118 But I have a more fundamental difficulty with Mr Beltran’s case. The difficulty is that I am unable to find that Mr Beltran’s trades from 10 to 12 May 2022 were transacted in reliance on the Conceded Representations. It follows that none of the losses flowing from those trades were caused by the defendants’ frauds.

119 Mr Beltran’s evidence was that he read the Terra Representations on the Terra Website and Terra White Paper in or around April 2022. He read the LFG Press Release on 5 May 2022. Then, after the de-peg began on 7 May 2022, and despite reading Mr Kwon’s tweets of 10 and 11 May 2022, he says that he formed the view “that it was a matter of time before the UST’s peg was restored”. According to him, he relied on the Terra Representations and LFG Press Release to purchase USTs as part of a “long-term view on investing in UST” as he “would effectively be getting a discount on a stablecoin which was certain to re-peg at some point in the future”.

120 Mr Beltran was thus a late adopter of UST. He only began investing after the de-peg had occurred and UST’s value had dropped by a considerable amount. His first purchases were on 10 May 2022 when UST’s opening position was 0.8059, its lowest position was 0.6841, and its closing position was 0.7999. These were well below UST’s one-to-one peg to the USD. For his first set of purchases on Gate.io on 10 May 2022, Mr Beltran bought UST at a price of 0.76. The inference that I draw in the circumstances is not that Mr Beltran believed in the truth of the representations and acquired UST from 10 May 2022 in reliance on the same. I infer instead that Mr Beltran was speculating that UST would increase in value after its dip from 7 May 2022 onwards and hoped to profit from a potential subsequent rise in UST’s value from 10 May 2022.

121 The inference that Mr Beltran was speculating on UST is supported by several matters. First, there is the absence of Mr Beltran’s browsing histories. He provided a screenshot of his browsing history on 30 April 2022, showing his accessing the Terra White Paper on that date. He explained his inability to retrieve browsing histories for the LFG Press Release due to the lapse of time. But, even accepting that he had seen the Terra White Paper in April 2022, why did he wait until 10 to 12 May 2022, well after the de-peg, before investing in UST? Mr Beltran’s conduct suggests that he was seeking to profit from the potential of a rise in UST’s market value, rather than acting in reliance on the Terra Representations and LFG Press Release. Mr Beltran’s explanations in cross-examination were unconvincing. At first, he denied that UST had de-pegged by the close of 9 May 2022, despite UST having dropped to a low of 0.7934. Later, he conceded, when confronted with UST’s value at the time, that by then he “would know that the depeg had happened”. He maintained that he believed in LFG’s ability to use its Bitcoin reserves to maintain UST’s peg to the USD when he first bought UST. But this was against the reality that UST’s value had dropped to 79 cents against the USD and had failed to re-peg since 7 May 2022. He asserted that his initial impetus for buying UST on 10 May 2022 was that he saw the “price recovering and that gave me hope”. But his witness statement averred that his primary motivation was to buy UST on the cheap as “I felt like I would effectively be getting a discount on a stablecoin which was certain to re-peg at some point”.

122 I find it difficult to reconcile Mr Beltran’s assertion of reliance with the fact that he apparently never had the impetus to acquire UST before the de-peg. This was despite supposedly having faith in UST’s absolute stability based on the Terra Representations and LFG Press Release. It would seem that he only had the impulse to buy UST on the faith of an alleged belief in its complete stability after UST’s value had declined steeply. That does not make sense. Therefore, I find that reliance and causation of loss are not made out in Mr Beltran’s case. I dismiss his claims for damages for misrepresentation.

123 Mr Gan’s claim comes to US$476,988.75. Mr Lim’s estimate is the same. I accept Mr Gan’s evidence of his ownership of various wallet addresses and accounts on cryptocurrency exchanges, as verified by way of screen recordings of him logging into his Kucoin account, an email from BTSE confirming a withdrawal of USTs purchased using his BTSE account, and his .csv file of the UST transactions on his Terra wallet address. I decline to draw an adverse inference from the mere fact that he has not disclosed his API keys due to privacy and hacking concerns.

124 However, in my view, the evidence adduced by Mr Gan falls short of establishing on the balance of probabilities that he accessed the Terra Website and read all of the Terra Representations. During the earlier stage of these proceedings, before Nair J (see at [40]–[43] above), Mr Gan filed an affidavit to resist Terraform’s application for a mandatory arbitration stay. The affidavit stated that Mr Gan:

(a) accessed the Terra Website and Terra White Paper sometime in December 2020;

(b) accessed both sources of representations “again in or around December 2021 and/or January 2022, in the course of my considerations on whether to purchase UST”;

(c) by reason of the accesses, saw all of the Terra Representations and the particulars thereof; and

(d) accessed the Anchor White Paper and Anchor Website in around February 2021.

125 In contrast, when Mr Gan first filed his further and better particulars in January 2023 to provide “[t]he precise date(s) on which the Schedule 3 Claimant accessed the said parts of the 1st Defendant’s website and the Terra Money White Paper referred to in paragraph 13 of the SOC Am1”, Mr Gan’s reply was:

126 In the affidavit at [124] above, Mr Gan acknowledged that, in his further and better particulars, he only mentioned having accessed the Terra Website and Terra White Paper in around December 2020. But he explained:

127 No explanation is given why Mr Gan only recalled this second access to the Terra Website when preparing his affidavit of May 2023, as opposed to when he was initially asked to furnish dates of access for the purpose of his further and better particulars of January 2023. In his witness statement, Mr Gan identified the different pieces of information which he obtained from the Terra Website during the two different points of access. In his first access on December 2020, Mr Gan only saw the information in the Terra White Paper and did not see the additional information entitled “About the Terra Protocol” on the Terra Website. That additional information formed part of the basis for the Terra Representations particularised in the pleadings. It is only during the second access sometime in December 2021 or January 2022 that Mr Gan says that he saw the additional information “About the Terra Protocol” which he had missed on his first visit. The latter additional information was in fact absent from the Terra Website in December 2020, as has been shown by archives extracted from the Wayback Machine.

128 I am unable accept that there was a mere lapse of memory on Mr Gan’s part. A more likely explanation for the change in Mr Gan’s version of events is that he only visited the Terra Website in December 2020. Upon the defendants pointing out, based on Wayback Machine archives, that all the alleged Terra Representations could not have been apparent on the Terra Website in December 2020, Mr Gan belatedly recalled the additional access in 2021 or 2022. In cross-examination, Mr Gan had no convincing explanation for why he suddenly recalled the dates of his supposed second point of access:

129 Having rejected Mr Gan’s evidence that he saw the “About the Terra Protocol” statements on the Terra Website, I turn to his claims of having relied on the Terra Representations in purchasing USTs. It is possible that Mr Gan saw the Terra White Paper in December 2020 as he claims. But the only evidence of that access rests upon his assertion that he did. The reliability of Mr Gan’s evidence as to when he accessed the Terra Website and what he gleaned from such access being in doubt, I approach his evidence on the matter with scepticism.

130 This was not the only unsatisfactory aspect of Mr Gan’s evidence. He claimed never to have seen a pop-up notice referring to the Anchor Terms of Service when connecting his wallet to the Anchor platform through the Dashboard on the Anchor Website. In his affidavit before Nair J, Mr Gan averred as follows:

131 I find Mr Gan’s equivocal statement of not “recalling” whether there was a link to the Anchor Terms of Service difficult to accept. Nair J remarked upon this in Beltran (Arbitration Stay) at [161] (see at [43] above):

132 When challenged on the Anchor Terms of Service in cross-examination, Mr Gan denied that such a “pop-up” would have appeared at all when he connected his wallet to the Anchor platform. He explained that he “already ha[d] a Terra Station on the Google Chrome extension”, so he could “just connect and then it will just automatically log you in” without having to see the pop-up notice with a link to the Anchor Terms of Service. According to him, one only sees the pop-up box with the words “By connecting, I accept Anchor’s Terms of Service” if one connects one’s wallet to Anchor without “hav[ing] the Terra Station extension installed on your Chrome browser”. But this assertion was falsified by Mr Lim’s evidence (supported by a video) that, even with the Terra Station extension installed, the pop-up box containing the words “By connecting, I accept Anchor’s Terms of Service” and a link to the Anchor Terms of Service, will still appear.

133 I am therefore unable to treat Mr Gan’s evidence as reliable. I consequently do not accept his evidence of having seen or relied upon the Terra Representations as found in the Terra White Paper or the Terra Website.

134 What of Mr Gan’s evidence in respect of the Fourth and Sixth Representations? He says that he relied on those representations when he purchased UST tokens in January and February 2022 and in May 2022. The problem is that the only evidence of the alleged reliance on the Fourth and Sixth Representations is a bare assertion in Mr Gan’s witness statement. Mr Gan averred that it was the totality of the first five representations which led him to be “convinced that UST was a stablecoin which could suit my investment needs”, since he “was on the lookout for a stablecoin which could also allow me to earn passive interest”. The totality of the first five representations are said to have led Mr Gan to believe that UST was stable because it was pegged to the USD and that he could earn passive income whilst enjoying the stability of a stablecoin, as “a good way to safely earn a good rate of return on an investment in UST stablecoins”. Given my rejection of Mr Gan’s evidence as to reliance on the Terra Representations, it is difficult to see how the totality of the first five representations, including the Fourth Representation, could have led Mr Gan to act as he did and to believe – as he averred – that “UST would remain stable and the Anchor protocol would function as represented”.

135 As for the Sixth Representation, Mr Gan said that the LFG Press Release reinforced his belief in the stability of UST’s value as derived from the Terra and Anchor Representations:

But again, having rejected Mr Gan’s evidence of having seen and relied on the Terra Representations, I am unable to accept his explanation of the role that the LFG Press Release played in reinforcing an already established belief in UST’s stability due to the Terra Representations.

136 For these reasons, I reject Mr Gan’s evidence that he purchased USTs in reliance on the Terra Representations, the Fourth Representation, and the Sixth Representation. Mr Gan fails to discharge the burden of proving reliance and showing causation of the losses from his UST investments. I dismiss Mr Gan’s claim for damages for misrepresentation.

137 Mr Davis’ claim is for US$28,129.98. Mr Lim’s calculation is similar – US$28,057.38. Mr Davis is in a different position from Mr Beltran in that all his purchases of UST tokens took place on 12 April 2022 and pre-date the de-peg of 7 May 2022.

138 The starting point is to take the net loss calculated by Mr Lim. Based on Mr Lim’s cross-examination by the representative claimants’ counsel, some of the difference between the parties’ figures – roughly US$87 – stems from Mr Lim having accounted for the yield obtained from Mr Davis staking USTs on Binance Earn. As I have explained, such yields are attributable to purchases of UST induced by the defendants’ misrepresentations and ought to be credited to the defendants in the net loss computations (see at [100]–[101] above). Other differences between the parties’ calculations are attributable to the different exchange rates used.

139 I reject the argument that adverse inferences should be drawn from Mr Davis’ failure to produce browsing histories of his access to the documents containing the representations. Mr Davis explained that Google Chrome’s browser only keeps 90 days’ worth of browsing history. I accept that Mr Davis was unaware of Chrome’s “My Activity” feature which could have been used to produce a more extensive history. I find that Mr Davis established ownership of his Binance account through a video of him logging into that account. Mr Davis met his evidential burden.

140 I also accept Mr Davis’ evidence of reliance. Mr Davis relied only on the three Terra Representations, which he gleaned from the Terra White Paper and Terra Website in December 2021. There is no evidence of speculation on UST’s sudden decline in value nor of circumstances undermining the credibility of Mr Davis’ evidence. I infer that Mr Davis was induced to purchase USTs based on the Terra Representations, since the Terra Representations about UST’s stability would be a material concern for any reasonable person deciding whether to buy UST.

141 Applying the cut-off (see at [111]–[112] above), while Mr Davis did not purchase UST after that date and time (see at [137] above), he sold UST after that time. On the general principles that I set out at [114] above, I should ignore his sales of UST for other cryptocurrencies on 14 May 2022. Any fall in value in Mr Davis’ USTs after the cut-off would not be attributable to the defendants’ fraud, but to an independent decision to bet on an upturn in UST’s value. Mr Davis bought 29,968.00 UST on 12 April 2022 for US$29,990.48. He sold 15,239.00 UST on 30 April and 14 May 2022. He therefore still holds 14,729.00 UST. Taking the approach that a reasonable investor would have sold all 29,968.00 UST by the cut-off (when the price was 0.8011), the sale of all Mr Davis’ UST tokens would have made US$24,007.3648. By 14 May 2022 when Mr Davis sold USTs, the value had decreased significantly to between 0.1483 and 0.255. Mr Davis cannot expect the defendants to indemnify him against the fall in UST’s value between 12 and 14 May 2022 or beyond (for the UST tokens he still holds). The retention of USTs beyond the cut-off would be attributable to Mr Davis’ independent decision to hold USTs.

142 Subtracting US$24,007.3648 from US$29,990.48 gives US$5,983.1152, say US$5,983.12. I therefore award US$5,983.12 to Mr Davis in damages for misrepresentation. Terraform and LFG used the average market price of 0.60485 USD for 1 UST on 12 May 2022 for the mitigation sale of Mr Davis’ UST holdings. But I held at [114] above that the relevant price should be the opening price of UST at the cut-off -- 0.8011.

143 Mr Macquisten claims damages of US$31,819.00 with no yields credited to the defendants, and US$27,053.88 if all yields are credited. Mr Lim’s calculation comes up with a similar figure of US$27,402.81, with UST-related yields included. I take Mr Lim’s calculations as the starting point for the reasons at [100]–[103] above.

144 I decline to draw an adverse inference based solely on the absence of a browsing history record showing access to the Terra Website in January 2021. Mr Macquisten attributed the absence to the deletion of files on his laptop in the course of time. In cross-examination, it was put to him that there was “no evidence of your browsing history in these proceedings”. He candidly conceded this. There was nothing to suggest that his explanation on affirmation for the absence of such evidence should be disbelieved.

145 I find that Mr Macquisten has proven his ownership of the relevant cryptocurrency wallets by tendering evidence of him logging into his Terra Station account and of his .csv files of his transactions on his Terra wallet. I decline to draw an adverse inference from his omission to provide relevant keys for (among others) his Kucoin account.

146 Among the Conceded Representations, Mr Macquisten’s evidence was that he relied on the Terra Representations. He was aware of, but did not rely on, the Fourth and Sixth Representations. He says that he also relied on the Seventh Representation (which I have found not to be actionable). Terraform and LFG submit that, Mr Macquisten having acknowledged that the Sixth Representation contradicted the Terra Representations, he ceased to rely on the Terra Representations from the day he read the LFG Press Release. The alleged contradiction was put to Mr Macquisten in cross-examination. Mr Macquisten accepted that there was a contradiction. But my impression, was that Mr Macquisten was simply accepting that counsel’s suggestion of a contradiction was tenable. I do not think Mr Macquisten was conceding that, when he saw the LFG Press Release, he realised that there was a contradiction. I therefore do not find that Mr Macquisten ceased to believe in the Terra Representations, merely because he saw the LFG Press Release.

147 Mr Macquisten saw the Terra White Paper, but did not see the additional details under the section heading “About the Terra Protocol” on the Terra Website. Those details were not present on the Terra Website in January 2021. The particulars of the First and Second Representations are a combination of citations from the Terra White Paper and the Terra Website. I am satisfied that Mr Macquisten derived the pleaded meanings of the First and Second Representations from the Terra White Paper.

148 For the Third Representation, Mr Macquisten relies on the following portion of the Terra White Paper:

The problem is that the passage cited is not among the pleaded particulars of the Third Representation. The Terra White Paper is not relied on as a source of the Third Representation. The pleaded source of the Third Representation is solely the information found in the “About the Terra Protocol” section of the Terra Website:

149 Consequently, I find that Mr Macquisten read and understood the First and Second Representations, but did not read the Third Representation.

150 On reliance, I accept as credible Mr Macquisten’s evidence that he was induced by his belief in UST’s stability of its peg to the USD based on the First and Second Representations. I find that Mr Macquisten’s purchases of UST starting on 9 February 2021 were motivated, at least in part, by that belief. Mr Macquisten purchased USTs on 11 and 12 May 2022. Although these were motivated, in part, by the non-actionable Return to Form Tweet, I cannot rule out the possibility that Mr Macquisten’s continued belief in the First and Second Representations led him to make further purchases of UST on 11 May. In my view, Mr Macquisten satisfies the burden of proving inducement and reliance and is entitled to compensation for losses flowing from his purchases of UST between 9 February 2021 and 11 May 2022 (that is, up to the cut-off).

151 According to Terraform and LFG, Mr Macquisten’s total purchases of UST less his total sales of UST as at 12 May 2022 amount to a net loss of US$28,141.33. In connection with that figure, Terraform and LFG state that Mr Macquisten did not hold UST tokens requiring a mitigation sale as at 12 May 2022. Mr Lim’s figure of US$27,402.81 is lower. But Mr Lim’s calculation takes account of Mr Macquisten’s UST purchases from 9 February 2021 to 12 May 2022 (that is, after the cut-off). The representative claimants’ estimate of Mr Macquisten’s net loss from UST purchases, inclusive of UST purchases on 12 May 2022 and accounting for yields, is similar -- US$27,053.88. Given the foregoing, it seems to me that the more reliable figure for Mr Macquisten’s loss is US$28,141.33, taking account of the submission by Terraform and LFG that such figure has been derived on the basis that no mitigation sale was required as of 12 May 2022.

152 I accept Ms Mariam’s evidence that she owned and controlled the Terra Station wallet and Binance account which were the subject-matter of a video that she tendered in court. She stated in her witness statement that she was unable to locate the browsing history of her accessing the Terra and Anchor Websites due to having accessed the sites on her work laptop in November 2021 and having since switched employers. I accept that explanation.

153 Ms Mariam’s evidence was that, of the Conceded Representations, she acted in reliance on the Terra Representations and the Fourth Representation to purchase UST tokens from 7 to 24 February 2022. Mr Lim’s estimate of her net losses is US$68,140.95. I accept Mr Lim’s estimate as an accurate starting point of the net losses arising from her disclosed UST trades, noting that it is not far from the representative claimants’ estimate of US$68,141.96.

154 I am, however, unable to conclude that Ms Mariam’s decision to purchase USTs was induced by her having read and relied on the Terra Representations and the Fourth Representation. On her evidence, she only read these representations on 1 November 2021. But the .csv file of UST transactions conducted via her Terra Station wallet shows that 498.879 UST was transferred to her wallet on 12 October 2021. Her witness statement denies that she conducted that transaction. She instead asserts that “I am unable to explain how this UST ended up in my Terra Station wallet” and “I do not recall purchasing UST at any time before February 2022”. She states:

155 The .csv file also shows that 495.181 UST had been staked on Ms Mariam’s Anchor account on 12 October 2021. But she denies having staked the USTs on her Anchor account, simply saying that she has “no recollection of” that transaction, and is “unable to explain” its occurrence.

156 I do not accept Ms Mariam’s explanation that she did not purchase and stake those UST tokens in October 2021 prior having seen the relevant representations in November 2021. In cross-examination, her oral evidence was contradictory. While claiming that she did not conduct these transactions, she was adamant that no one else could have had access to her wallet apart from herself. She denied that her wallet had been “compromised”. Her only explanation was that “[i]t’s a mystery to me”. The inescapable conclusion on the evidence (which I find) is that Ms Mariam had purchased 498.879 UST on 12 October 2021 and staked 495.181 UST on Anchor on the same day.

157  Ms Mariam strenuously maintained that, but for the Terra Representations and the Fourth Representation, she would not have invested in UST at all. She claimed to be looking for a safe and stable cryptocurrency in which to invest. The representations (she says) convinced her that UST was such a stable and secure cryptocurrency and she “would also be able to earn a good rate of return on this investment by staking the UST that I had purchased on Anchor”. She insisted in relation to the October 2021 transactions that “[i]t is highly unlikely that I would have purchased UST in October 2021, given that I had only learned about Terra and UST in November 2021”. In cross-examination, she maintained that she “only bought UST after [she] had seen the Terra website and the Terra white paper”, and that she “would not have used Anchor until [she had] read the Anchor white paper”. But, given my finding that she had carried out UST transactions in October 2021, Ms Mariam’s evidence on her reliance is not credible. In deciding to invest in UST, she could not have been relying on representations which she only saw in November 2021, if she was purchasing and staking USTs from October 2021.

158 Consequently, Ms Mariam has failed to make out her case in the tort of deceit. I dismiss her claim for damages for fraudulent misrepresentation.

159 I find that Mr Xue has established that he owned the relevant Binance, Kraken, OKX, and Kucoin accounts used to make the relevant UST transactions. Although some of his API keys apparently did not work, I decline to infer from such circumstance that Mr Xue did not make his claimed UST transactions. I accept Mr Xue’s explanation that he was unable to retrieve the browsing histories of his accessing the Terra Website and Anchor Website in around November 2021 on account of (a) the lapse of time since and (b) his practice of not retaining extensive browsing histories on his personal devices due to security and privacy concerns.

160 Mr Lim’s estimate of Mr Xue’s loss is US$22,917.94. This is similar to the representative claimants’ figure where credit is given for yields generated from Mr Xue’s UST holdings – US$22,917.78. I consider Mr Lim’s estimate as the starting point for quantifying Mr Xue’s losses.

161 Mr Xue’s evidence was that, of the Conceded Representations, he relied on the Terra Representations and the Fourth Representation in purchasing UST tokens between 19 February and 5 May 2022. According to him, he continued trading UST on 10 and 11 May 2022, after the de-peg of 7 May 2022, because he believed in those representations. He interpreted the Return to Form Tweet as a representation that UST’s peg to the USD would be restored.

162 In relation to the UST transactions pre-dating the de-peg on 7 May 2022, I accept Mr Xue’s evidence of reliance. Given the materiality of the Terra Representations and the Fourth Representation on the desirability of UST-related investments, Mr Xue’s evidence is credible. I accept that Mr Xue has proved a sufficient causal connection between the fraudulent misrepresentations and the economic loss flowing from his UST purchases up to 5 May 2022.

163 However, Mr Xue’s post-depeg UST purchases stand to be assessed differently. Mr Xue’s evidence was that he was influenced by the aforesaid representations as well as Mr Kwon’s Return to Form Tweet and other tweets on 10 and 11 May 2022 comprising the Seventh Representation. The former persuaded him that UST’s value was inherently stable because of its algorithmic features, while the May 2022 tweets reinforced that pre-existing impression and “further assured [him]” that UST’s peg to the USD would be “successfully restore[d]”. This, if accepted, would imply that the fraudulent misrepresentations continued to operate on Mr Xue’s mind in relation to his 10 and 11 May 2022 UST purchases. They would furnish a causal link to Mr Xue’s economic losses.

164 I reject this position in relation to the post-depeg trades. I find that the trades were instead Mr Xue’s attempt at pure swing trading, seeking to profit from UST’s volatility from 7 May 2022 onwards. This is because Mr Xue acknowledged that, by 12 May 2022, his faith in the stability of UST had wavered to the point that he began buying and selling UST in a panic rather than due to reliance on the representations. On the one hand, he claims to have been confident that the UST was algorithmically stable on 10 and 11 May 2022 despite the de-peg. On the other hand, from the next day and thereafter, he embarked on a series of panic sales and purchases of UST, ostensibly because he was becoming “increasingly fearful for [his] UST investment when the UST price continued to fall even more over the next few days”. I do not find that sudden transition from confidence to panic on Mr Xue’s part to be credible. The more likely explanation is that, the de-peg on 7 May 2022 having demonstrated that UST was not a stable currency, Mr Xue engaged in swing trading on UST’s volatility in a bid to make gains by speculating on the prospect of a rise in UST’s value, rather than out of continuing faith in the representations.

165 Mr Xue was selective as to which of his post-depeg transactions were (and which were not) made in reliance on the representations. Mr Xue was cross-examined on why, in relation to gains from sales of UST in the post-depeg period, he only credited the defendant with some (but not all) of the gains from his trades. For example, in respect of his sale over Binance of some 52,000 UST tokens on 12 May 2022, Mr Xue only credited the defendants with the gains from about 15,000 of the 52,000 USTs sold. He omitted around US$23,000 of his total gains of US$32,730.67 from the sale of the 52,000 USTs. Mr Xue explained that the gains from about 37,000 USTs were attributable to panic purchases on 12 May 2022 and were not made in reliance on the fraudulent misrepresentations. He identified two panic-buying transactions for a total of about 39,291 USTs made on 12 May 2022, which he said were excluded from the calculation of his losses on a “first-in first-out” basis, thus accounting for the USTs he excluded from his subsequent sale of 52,000 USTs that same day.

166 Mr Xue explained that his purchases on 10 and 11 May 2022 were premised on the defendants’ misrepresentations, such that the losses flowing therefrom remained claimable, while gains from purchases made on 12 May 2022 were independent panic purchases, credit for which should not be given to the defendants. The explanation strikes me as contrived. Mr Xue purchased over 98,000 UST tokens in ten tranches from 10 May 2022 06.09 hours to 11 May 2022 04.50 hours UTC, while having only purchased about 52,000 UST tokens over the longer period of 19 February to 5 May 2022. If Mr Xue’s account is to be believed, his faith in UST’s stability grew by a significant margin after UST had de-pegged from the USD on 7 May 2022 and failed to re-peg for 4 to 5 days afterwards. Moreover, while he staked the USTs purchased from 19 February to 5 May on the Anchor platform, with minimal withdrawals of USTs over that period, the same pattern does not hold true from 10 May 2022 onwards. Instead, he made three large withdrawals of over 40,000 USTs each on 10 and 11 May 2022. Mr Xue’s trading history on 12 May 2022 then has him buying and selling USTs within close intervals of one another. The picture that emerges is that, before the de-peg, Mr Xue believed in the representations and thought that UST was stable by design. He accordingly purchased USTs until 5 May 2022 in order to stake them on the Anchor platform and enjoy high yields. After the de-peg on 7 May, Mr Xue lost faith in the stability of UST and engaged in swing trading to profit from changes in UST’s market value, taking advantage of UST’s volatility. The former were transactions in reliance. The latter were speculative.

167 It follows that the proper assessment of damages for Mr Xue is to award his net losses flowing from pre-depeg UST purchases, while excluding the net losses flowing from post-depeg UST purchases. The representative claimants have estimated that sum at US$3,274.00 based on a scenario where post-depeg purchases of UST are not claimable and all yields are taken into account. Mr Lim’s estimate for the category “Loss – Nullified Post Depeg, Evidence Only (USD)” is only slightly different – US$3,378.81. For completeness, in contrast to the methodology for the category of “Loss – Evidence Only (USD)” (see at [102] above), this heading is described as follows by Mr Lim:

168 Accordingly, the sum of US$3,378.81 represents Mr Xue’s net losses by taking the sum he expended on the approximately 52,000 USTs purchased from 19 February 2022 to 5 May 2022 subtracting the gains accruing from those UST tokens being sold, whether before or after the de-peg. Following the approach at [114] above, if any of those USTs were only sold by Mr Xue after the cut-off, the gain to be credited to the defendants is the notional sale of those residual USTs at 0.8011 and not the actual gains from a sale occurring after the cut-off. In that event, there would have to be an adjustment to the figure of US$3,378.81.

169 Nonetheless, I am satisfied that the roughly 52,000 USTs purchased by Mr Xue by the close of 5 May 2022 (that is, pre-depeg) were sold off by the cut-off. Taking a “first-in, first-out” approach to his UST transactions, the sales of two tranches of over 40,000 USTs each on 10 May 2022 would have sufficed to dispose of Mr Xue’s pre-depeg USTs. Thus, adopting Mr Lim’s more generous figure at [167] above, I award Mr Xue US$3,378.81 in damages for misrepresentation.

170 Terraform and LFG point out that the Binance account which Mr Arun claims as his own is registered under the name of Mr Arun’s sister. Terraform and LFG accordingly urged me to infer that all transactions on that account were for Mr Arun’s sister, so that all damages alleged to have been sustained on that account should be excluded. But I accept Mr Arun’s explanation (supported by a video of him logging into the account with his email address) that he initially opened the account for his sister in an attempt to introduce her to the crypto-world. However, he now has full control over the account and carried out all relevant transactions there. Mr Lim confirmed in a supplementary witness statement that Binance does not allow a user to change their registered name once an account is created. A user can only change the account’s email address. While there is an absence of browsing histories in Mr Arun’s evidence, I accept as believable his explanation that his browsing history has been deleted, so he cannot retrieve evidence of his having accessed the Anchor Website and Anchor White Paper in December 2021. He recalled the approximate period of his access by reference to when he first purchased UST, since he bought USTs shortly after he read the Anchor White Paper and accessed the Anchor Website.

171 Terraform and LFG noted a discrepancy between two .csv files of Mr Arun’s Binance transactions exhibited in different witness statements, pointing out that Mr Arun was unable to explain the discrepancy. One record was the full .csv file as downloaded from Mr Arun’s Binance account and displayed in the video of him logging into that account. The other was an edited .csv file of relevant UST transactions in Mr Arun’s main witness statement. Contrary to what Terraform and LFG submit, the inference that I draw is that the discrepancy is attributable to an error in the course of deleting irrelevant non-UST transactions from a voluminous .csv file covering many transactions. I do not believe that there was a calculated attempt to suppress evidence.

172 Mr Arun merely relies on the Fourth Representation. He says that he read the Fourth Representation in the Anchor White Paper, and understood that UST holders could earn passive income whilst enjoying the stability of holding their principal savings in stablecoin. From 29 December 2021 to 7 May 2022, he claims to have purchased USTs in reliance on the Fourth Representation, believing that buying and staking USTs on the Anchor platform was a sound investment. I accept this evidence as credible. I find a sufficient link established between the Fourth Representation having been made and the economic losses flowing from Mr Arun’s December 2021 to May 2022 UST purchases.

173 Mr Arun purchased some USTs on 7 May 2022 at 17.04 hours UTC. I do not regard this as an attempt to profit from UST’s decline in value that started on 7 May 2022. Mr Arun’s purchases on 7 May 2022 do not appear out of place with his prior pattern of purchasing and staking USTs from December 2021 onwards. On 7 May 2022, UST closed at a value near its one-to-one peg (0.9969) and its lowest value for that day was 0.9927. I therefore accept Mr Arun’s evidence of reliance in respect of his 7 May 2022 UST purchases. Mr Arun is not claiming for post-depeg purchases of UST.

174 On the quantification of damage, I take as my starting point Mr Lim’s assessment of Mr Arun’s net loss – US$22,503.02. This is close to the representative claimants’ estimate of US$22,596.61, accounting for yields. But these do not account for UST tokens which remained unsold as at the cut-off at [114] above. These are merely estimates of the net losses from Mr Arun’s UST purchases, irrespective of when his USTs were sold. Mr Arun’s evidence is that “[b]etween 12 February 2022 and 3 June 2022, I sold all of the UST that I had purchased in reliance of [sic] the 4th and 5th Representations, as well as the UST that I had earned from staking the UST that I had purchased on the Anchor platform”. It is clear then that Mr Arun held onto some of his USTs following the cut-off. Selling such USTs after the cut-off would have increased Mr Arun’s losses to the extent those sales would have been at a significantly lower price than that of 0.8011 at the cut-off.

175 Of the estimates provided to me, the one that comes closest to what I should award to Mr Arun, namely, the amount spent on purchasing his USTs, minus the actual gains from selling his USTs and yields therefrom until the cut-off, and minus the notional sale of his remaining USTs at the cut-off, is the second permutation calculation by Terraform and LFG. They calculate that the “USD value of all purchases up to 12 May 2022 and deducting the USD value of all sales up to 12 May 2022” gives US$4,372.96, with 2,973.64 UST tokens still to be sold. Instead of the average price of 0.60485 USD used by Terraform and LFG for the notional sale of the remaining USTs at the cut-off, I take the opening price for UST at the cut-off of 0.8011 USD for the reasons stated at [114] above. Adopting Terraform and LFG’s methodology, the measure of Mr Arun’s loss would then be US$4,372.96 less the putative amount obtained from selling 2,973.64 USTs at 0.8011 USD (that is, US$2,382.183). This gives US$1,990.777, say US$1,990.78.

176 Accordingly, the quantum of damages due to Mr Arun for misrepresentation is US$1,990.78.

177 Mr Lee produced the browsing history of his accessing the Anchor Website in February 2022. I accept the videos of him accessing his various accounts as showing ownership of his Terra Station wallet and cryptocurrency accounts.

178 On reliance, acting on the Anchor Representations (of which the Fourth Representation is actionable and the Fifth Representation is not), he says that he purchased over 57,000 UST tokens from 2 March to 14 April 2022, and staked them on the Anchor platform. He believed that staking USTs on Anchor was a safe investment. I accept this evidence as credible.

179 Mr Lee’s evidence was that he made a series of UST purchases on 11 May 2022 solely on the faith of the Seventh Representation, thinking that UST’s value had only dropped temporary and would re-peg with a recovery plan. Since the Seventh Representation is not actionable, I find that Mr Lee is only entitled to consequential economic losses flowing from his UST purchases from 2 March to 14 April 2022, but not from his purchases on 11 May 2022.

180 I take Mr Lim’s estimate of Mr Lee’s net losses excluding all purchases after the de-peg (US$39,266.23) as the starting point. The methodology of Mr Lim’s estimate of loss from pre-depeg purchases only is that stated at [167] above. The representative claimants’ estimates for the scenario where credit is given for all yield accrued and post-depeg purchases of UST are excluded is slightly lower – US$38,702.00. However, if Mr Lee had sold any pre-depeg USTs after the cut-off at [114] above, these would still be included as profits to be subtracted from the net loss figure, following Mr Lim’s explanation at [167] above. Hence, it must be ascertained whether Mr Lee held onto his USTs beyond the cut-off. While Terraform and LFG submit that Mr Lee held 5,611.59 UST tokens as at 12 May 2022, these appear to be attributable to the USTs purchased by Mr Lee after the de-peg on 11 May 2022, when he bought on the basis of the Seventh Representation. On a “first-in, first-out” approach, Mr Lee purchased over 57,000 UST tokens before the de-peg, and the sales of USTs he conducted on 11 May 2022 from 7.14 to 22.02 hours – before the cut-off – exceed 200,000 USTs.

181 Hence, Mr Lee’s pre-depeg UST tokens would all have been sold off by the cut-off, and I award him damages in accordance with Mr Lim’s estimate of US$39,266.23.

182 I accept Mr Epstein’s explanation that his browsing histories were unavailable due to lapse of time. Mr Epstein’s witness statement exhibited screenshots of relevant UST purchases and sales, and the currency used for each transaction. Given the detailed trading logs disclosed in relation to Mr Epstein’s trades and his staking of USTs on Anchor, I am satisfied that Mr Epstein was the person who conducted the UST transactions for which he seeks compensation.

183 Mr Epstein says that he relied on all five Conceded Representations. His evidence is that, based on the Terra Representations which he saw on the Terra Website and Terra White Paper, he believed in the systemic stability of UST. Based on the Anchor Representations (inclusive of the Fourth Representation), he thought that the Anchor protocol was a safe product which would allow users to earn passive income in the form of a crypto-asset of stable value. He saw the LFG Press Release, making up the Sixth Representation in February 2022, and that reinforced his belief in the stability of UST. These representations appealed to him, since he wanted a place to park his savings to buy a house. I accept this account as credible.

184 I reject the suggestion by Terraform and LFG that Mr Epstein would have ceased to believe in the truth of the Terra and Anchor Representations once he read the LFG Press Release in February 2022. Mr Epstein stated that, in his view, the LFG Press Release reinforced the stability of UST’s market value. I accept that as a plausible understanding of the LFG Press Release.

185 From 13 March to 8 May 2022, Mr Epstein purchased more than 4.6 million UST tokens. The majority of them were bought after the de-peg on 7 May 2022. Mr Epstein purchased more than 3.3 million UST tokens on 8 May 2022 at or around 01.25 hours UTC. This constitutes more than 70% of the UST purchases for which he claims. The other UST purchases were more evenly spread out from 13 March to 1 May 2022. Mr Epstein also claims for purchases of UST made on 11 and 16 May 2022, comprising nearly 2 million UST tokens. These later purchases are alleged to have been made in reliance on all seven representations. But the timing of Mr Epstein’s later purchases suggests that these were purely influenced by the tweets comprising the Seventh Representation, which Mr Epstein understood to mean that the UST-USD peg would be restored and which I have held not to be actionable.

186 According to Mr Epstein, the Conceded Representations played a significant role in inducing all his UST purchases, even those made as late as 16 May 2022 when the price of UST had fallen to a low of 0.0944 (opening at 0.1766 and closing at 0.1264) well below the one-to-one USD-UST peg. I am unable to accept that as believable. Given how far the market value of UST had plunged by 11 May and afterwards and the non-recovery of the peg since 7 May 2022, I do not think that Mr Epstein could still have believed that UST was systemically stable. Mr Epstein may conceivably have taken Mr Kwon’s tweets of 10 and 11 May 2022 to mean that a recovery plan could restore the UST’s peg to the USD. But Mr Epstein’s assessment of the potential for UST’s recovery under a rescue plan is distinct from whether he continued to believe in the original Terra and Anchor Representations and LFG Press Release.

187 Mr Epstein’s stated motive for investing in UST was to yield passive income on a crypto-asset with a stable value in order to save for a house (see at [183] above). In stark contrast to that stated aim is the fact that he invested far more into UST tokens after the de-peg on 7 May 2022 would have cast serious doubt on UST’s professed stability. It seems to me that what Mr Epstein was doing after the de-peg was simply speculating on potential profits from any rise in the value of UST after it had de-pegged.

188 Hence, the correct measure of damages is to take the amount spent by Mr Epstein on the over 1.26 million UST tokens which he bought from 13 March 2022 to 1 May 2022 and subtract the amount which he would have earned from selling those tokens at 0.8011 on 12 May 2022 00.01 hours UTC, taking account of any yields from the UST tokens, including on the Anchor platform, and from sales of UST before the cut-off point.

189 The starting point for Mr Epstein’s loss is Mr Lim’s estimate of US$1,128,028.61, which uses the methodology at [167] above. But this sum does not account for any UST tokens held after 12 May 2022 at 00.01 hours UTC, and the depreciations in value post-dating the cut-off at [114] above. I am unable to rely on Terraform’s and LFG’s calculations in their submissions for the calculation at [188] above, because they calculate the net loss of Mr Epstein for all trades up to 12 May, which will include the trades of 8 May and 11 May 2022, which I have held at [185]–[187] above were not done in reliance on the defendants’ frauds and are not compensable. The claimants’ estimates where Mr Epstein’s post-depeg purchases are excluded nonetheless include his purchase of 8 May 2022 and do not account for a mitigation sale at the cut-off. I must therefore do what I can with the evidence before me to arrive at a correct calculation.

190 In Mr Lim’s “Summary of Analysis”, he puts the figure for Mr Epstein’s total pre-depeg purchases at US$1,261,300.26. Mr Epstein’s witness statement also has him purchasing a total of 1,261,300.26 UST tokens from 13 March 08.18 hours to 1 May 17.37 hours UTC. Mr Epstein’s tabular summary of his sales and purchases has him selling 223,854.94 UST tokens in three tranches from 29 March 00.21 hours to 11 May 10.32 hours for US$72,103.13. Taking a
“first-in, first-out” approach, all of these UST tokens must be attributable to the USTs which he had purchased by 1 May 2022. The remaining UST tokens would be the difference between 1,261,300.26 and 223,854.94, that is, 1,037,445.32 USTs. Applying the approach at [114] above, these USTs are taken to be sold at 0.8011 at the cut-off, yielding US$831,097.446. Hence, the amount to be awarded to Mr Epstein would be US$1,261,300.26 minus US$72,103.13 minus US$831,097.446. That comes to US$358,099.684, say US$358,099.68. This does not account for yields accruing to Mr Epstein, but in the absence of more detailed calculations on the measure at [188] from the parties, I take this as the closest calculation I can arrive at from the available evidence.

191 I award US$358,099.68 to Mr Epstein in respect of his claim for misrepresentation.

192 I find that Mr Tan has met the burden of showing ownership of relevant cryptocurrency wallets or accounts and accessing the representations upon which he claims. On ownership, Mr Tan adduced screenshots of his Gemini Earn account details and confirmed that the transaction details in his witness statement were derived from that Gemini account. He expressed his willingness to submit his phone for the defendants’ inspection to corroborate his access to the account.

193 I accept Mr Tan’s evidence that he attempted to produce his browsing history to confirm access to representations in the Terra Website in February 2022. But his browsing history had been deleted and was no longer available. He recalled the approximate time that he accessed the Terra Website because it was just before he began purchasing USTs in March 2022. I accept this explanation. In particular, I accept Mr Tan’s explanation that he was genuinely unaware of the option to retrieve a longer and more comprehensive browsing history via Google Chrome’s “My Activity” function.

194 Mr Tan claims to have relied on the Terra Representations and LFG Press Release, comprising the First to Third and Sixth Representations. He says that he understood the Terra Representations to mean that the price of UST would be pegged at a stable value to the USD through an arbitrage mechanism. He read the LFG Press Release to mean that a reserve fund had been amassed to defend the peg in volatile market conditions, reinforcing the stability of UST’s value. Terraform and LFG failed to show in cross-examination that Mr Tan subjectively ceased to believe in the truth of the Terra Representations upon his reading the LFG Press Release. Mr Tan did not consider the amassing of reserves as signifying that the arbitrage mechanism was insufficient on its own to maintain the stability of the UST-USD peg. I accept Mr Tan’s position as credible.

195 Mr Tan purchased over 64,200 UST tokens in March 2022, well before the de-peg of 7 May 2022. He is not claiming for losses from a purchase after the de-peg. As for when he sold his UST tokens, he only did so on 25 May 2022. That is after the cut-off at [114] above.

196 On Mr Lim’s calculations, Mr Tan spent US$66,059.50 on purchasing USTs in March 2022. He earned US$4,548.61 from selling USTs on 25 May 2022. That is how Mr Lim arrives at a “Loss – Evidence Only (USD)” figure of US$61,510.89. The correct measure of damages is to take Mr Tan as having sold off all his 64,246.40 UST tokens at the beginning of 12 May 2022 for 0.8011, yielding US$51,467.791. For the reasons at [114] above, I do not use the average value of USD 0.60485 for the mitigation sale, as Terraform and LFG have done.

197 Consequently, the damages to be awarded to Mr Tan would be US$66,059.50 less US$51,467.791. That comes to US$14,591.709, say US$14,591.71. I award Mr Tan US$14,591.71 as damages for misrepresentation.

198 I am unable to find that there were unilateral contracts as alleged by the representative claimants and dismiss their claims for breach of such. There are three broad grounds for my decision.

199 First, I do not think that there was an intention on the part of the defendants by any representations in the Terra and Anchor Websites (including in the Terra and Anchor White Papers) to extend unilateral offers to the world at large. Counsel for the representative claimants suggested that there was some sort of collateral contract whereby the defendants were promising to be bound by representations in consideration of a person purchasing UST from a third party. This seems to me contrived and artificial. I do not think that the Terra and Anchor Representations can by themselves be read as making any collateral contractual offer.

200 Second, any contract with Terraform would have incorporated the Terra Terms of Use and Anchor Terms of Service. The “No Warranty” clauses in those Terms make clear that any contract with Terraform would not include the Terra and Anchor Representations as contractual warranties.

201 Third, as for the LFG Press Release and the Return to Form Tweet, a reasonable reader would not have construed them as binding promises made to the world in respect of which Terraform would be contractually bound.

202 I elaborate further on my second and third grounds below.

203 Clause 10 of the Terra Terms of Use of 11 November 2021 provides:

204 Clause 3 of the Terra Terms of Use provides:

205 Similarly, clause 10 of the Anchor Terms of Service as at 12 October 2021 states:

206 Clause 2 of the Anchor Terms of Service states:

207 The foregoing terms are unambiguous. They say that no representations are intended to constitute contractual warranties or binding terms in any contract with Terraform.

208 The representative claimants make three broad arguments against the incorporation of the foregoing terms into their alleged contracts with Terraform.

209 First, they say that Terraform has not met the burden of showing that the Terra Terms of Use were on the Terra Website when accessed by certain representative claimants (Mr Macquisten and Ms Mariam) in 2021.

210 The earliest version of the Terra Terms of Use in the evidence before me is a version dated 11 November 2021. That version includes the “no warranty” provision in clause 10. Therefore, in all likelihood, clause 10 was accessible from the Terra Website as of 11 November 2021. As far as the Anchor Terms of Service are concerned, I accept Mr Lim’s evidence that those terms were accessible on the Anchor Website from 12 October 2021.

211 Ms Mariam claims to have accessed the Terra Website on about 1 November 2021. She says, based on a video showing the state of the website on 2 November 2021, that there was no link to the Terra Terms of Use on that date. On that basis, Ms Mariam argues that she would not be bound by the incorporation of terms which were inaccessible from the Terra Website when she visited. That may be true. However, given my finding that Ms Mariam had bought UST in October 2021 and that the Terra Representations could not have played a role in inducing her to purchase UST tokens (see at [157] above), I am unable to conclude that there was a unilateral contract between her and Terraform based on the Terra Representations. An offeree who knows of a contractual offer, but acts without any intention of accepting that offer, cannot be said to have accepted the offer, even if the offeree’s act happens to conform to the offeror’s requirements for acceptance. In such case, there is merely coincidence. What is needed is proof that an offeree acted with an intention of accepting a contractual offer. Here, Ms Mariam could not have accepted a contractual offer on the terms of the Terra Representations, as her decision to purchase UST in the first place was unrelated to those representations and could not constitute an acceptance of a contractual offer.

212 Mr Macquisten said that he only accessed the Terra Website in January 2021 and there was no link then to the Terra Terms of Use on the Terra Website. The difficulty, however, is that when Mr Macquisten accessed the Terra Website in January 2021, not all of the Terra Representations were on the website. There was only a link to the Terra White Paper on the Terra Website. The additional information in a section entitled “About the Terra Protocol” was not yet there. Not surprisingly, Mr Macquisten’s witness statement makes no mention of reliance upon any information in the “About the Terra Protocol” section of the website.

213 Terraform accepts that the Conceded Representations were false statements of fact fraudulently made. Terraform makes no concession as to the proper meaning of the particulars relied on by the claimants as giving rise to the Terra Representations, whether considered together or in isolation from each other, in relation to claims for breach of contract. I therefore have to assess whether the particulars mentioned in Mr Macquisten’s witness statement, which are solely to be derived from the Terra White Paper, are capable of being construed as a contractual offer of the nature pleaded by the representative claimants. In my view, the words relied on by Mr Macquisten cannot, on any reasonable construction, constitute the contractual offer alleged by him. The claimants’ pleadings posit that the contractual offer being conveyed via the Terra Representations was to the effect that (1) “UST was stable by design, and pegged to the US Dollar”, (2) “the Terra protocol and its underlying token mechanics would be able to maintain UST’s price stability regardless of market size, volatility and/or demand through its algorithm”, and (3) “UST [sic] would always be able to exchange 1 UST for 1 USD’s worth of LUNA on the Terra protocol”. None of those three meanings can be derived from the passages of the Terra White Paper on which Mr Macquisten relies.

214 The first meaning is said to be derived from the following passage in the Terra White Paper: “TerraSDR will be the flagship currency of [the Terra] family, given that it exhibits the lowest volatility against any one fiat currency”. The passage does not concern the volatility of UST but of TerraSDR.

215 The second meaning that the arbitrage mechanism would always maintain UST’s price stability “regardless of market size, volatility and/or demand through its algorithm” cannot reasonably be derived from any of the passages relied on by Mr Macquisten in his witness statement. Most of those passages merely explain how the arbitrage mechanism is supposed to work. The relevant passages describe how arbitrageurs will be incentivised to swap Luna for UST (or TerraSDR) or vice versa. The passages do not promise that UST’s value will be maintained regardless of how volatile market conditions may be.

216 The passages relied upon by Mr Macquisten from the Terra White Paper which come closest to the alleged contractual promise are:

217 However, the first of the above two paragraphs merely states that the arbitrage mechanism’s rewards will change based on prevailing market conditions. The passage does not promise that UST’s trading value will be maintained regardless of how severe or extreme the prevailing market conditions might be. The nature of an arbitrage mechanism is that the incentive to engage in arbitrage trading will grow as the disparity in the value of an asset in two markets increases. That does not mean that the incentive will be sufficient in every possible market scenario to overcome volatility and maintain a one-to-one peg between UST and USD. The second of the two paragraphs is merely a comparative statement which contrasts the volatility of Bitcoin with the volatility of UST and suggests that UST should be more robust. Contrary to what has been pleaded, the passage does not mean that UST’s price will always be stable “regardless of market size, volatility and/or demand”.

218 The third alleged unilateral contract term to the effect that users would always be able to swap 1 UST for 1 USD’s worth of Luna and vice versa is said to arise from the following passage in the Terra White Paper:

219 Fairly read, the passage does not guarantee that one will always be able to swap 1 UST for I USD worth of Luna. It simply explains how the system is supposed to function. The phrase “willingness to respect the target exchange rate irrespective of market conditions” cannot be read in isolation. In context, it explains how the swap mechanism allows arbitrage trading to take place when there is a disparity in values at which UST may be exchanged. There is no promise that the swap mechanism will never be terminated irrespective of how extreme market conditions became. In any case, the passage identified by Mr Macquisten is nowhere cited in the claimants’ pleadings. The Third Representation is instead pleaded to arise from a different passage in the section of the Terra Website entitled “About the Terra Protocol”:

220 On the basis of this different passage, it is pleaded that Terraform “made a unilateral offer to the world on the 1st to 3rd Representations as the main terms of offer as set out below”, including “that UST [sic] would always be able to exchange 1 UST for 1 USD’s worth of LUNA on the Terra protocol”. The claimants have not pleaded a case that, on the basis of the Terra White Paper, there was a unilateral offer being made to the world to the effect that users could always exchange 1 UST for 1 USD’s worth of Luna. The pleadings only rely on the “About the Terra Protocol” contents of the Terra Website for the Third Representation. Mr Macquisten’s difficulty is that he did not (and could not) have seen the section entitled “About the Terra Protocol”.

221 Consequently, I reject the contention that, in Mr Macquisten’s case, a unilateral contract was formed based on the terms of the Terra Representations. I note that there is no contradiction between this outcome and my upholding Mr Macquisten’s claim in the tort of deceit. The defendants conceded the claimants’ pleaded meanings regarding the Conceded Representations, inclusive of the Terra Representations. I accepted Mr Macquisten’s evidence of reliance thereon based on his subjective understanding of the cited excerpts of the Terra White Paper. But Terraform’s concession logically does not extend to Mr Macquisten’s claim for breach of unilateral contract. One considers the terms of a contract from an objective (rather than a subjective) standpoint. For the purposes of the claimants’ case on unilateral contract, I need to examine whether the invoked contents of the Terra White Paper can objectively be understood as a contractual offer on the pleaded terms and conditions. I have concluded that the contents cannot be so understood.

222 The representative claimants’ second argument is that the Anchor Terms of Service cannot bind the claimants who read the Anchor Representations, proceeded to purchase UST tokens on the open market, and then staked their USTs on the Anchor platform. The representative claimants contend that the Anchor Terms of Service only appear on the Dashboard of the Anchor Website when one attempts to connect one’s wallet to the Anchor platform.

223 This submission rests on my accepting that (1) the Anchor Representations constitute a unilateral offer to the world and (2) may be accepted by an offeree purchasing UST. I do not see how that can be a sensible understanding of the Anchor Representations, which are not aimed at promoting the stability of USTs, but at encouraging the adoption of the Anchor platform as a savings protocol through the staking of UST. In construing the Anchor Representations and how the putative offer would objectively be understood by reasonable persons, I accord weight to the fact that the Anchor Representations are meant to entice users to stake their UST tokens on the Anchor platform, not merely to purchase USTs and (say) deposit them on competitors’ exchanges. The logical construction of the Anchor Representations must be that the contractual offer (if any) being made is accepted by buying USTs and staking the same on Anchor. It follows that there is no need to consider the position of a claimant who purchases USTs on the strength of the Anchor Representations but does not stake the USTs purchased by connecting their wallet to the Anchor platform via the “Connect Wallet” function on the Anchor Website’s Dashboard.

224 The Terra Terms of Use and Anchor Terms of Service included provisions as follows:

(a) Clause 3 of the Terra Terms of Use stated that the information on the Terra Website “is made available solely for general information purposes” and Terraform does “not warrant the accuracy, completeness or usefulness of this information”.

(b) Clause 10 of the Terra Terms of Use stated that the contents of the Terra Website “are provided on an ‘as is’ and ‘as available’ basis, without any warranties of any kind, either express or implied”, and Terraform does not make “any warranty … with respect to the completeness, security, reliability, quality, accuracy or availability of the websites”, nor “warrants that the website, its content or any services or items obtained through the websites will be accurate, reliable, error-free”, and “disclaims all warranties of any kind, whether express or implied”.

(c) Clause 2 of the Anchor Terms of Service stated that “[a]ccess to the Interface is provided ‘as is’ and on an ‘as available’ basis only”, and that “[w]e do not guarantee that the Interface, or any content on it, will be free from errors or omissions”.

(d) Clause 10 of the Anchor Terms of Service stated that the “Interface is provided on an ‘AS IS’ and ‘AS AVAILABLE’ basis” [capitalisation in original], that “we disclaim any … warranties of any kind, whether express, implied, or statutory, including (but not limited to) the warranties of merchantability and fitness for a particular purpose”, and “do not … warrant that … the information contained in the Interface will be accurate, reliable, complete, or current; or that the Interface will be free from errors”, and that “[n]o advice, information, or statement that we make should be treated as creating any warranty concerning the Interface”.

225 I shall refer to the specific terms highlighted in the previous paragraph as the “No Warranty” clauses or conditions. The claimants allege that the “No Warranty” conditions were onerous and insufficiently signposted in the relevant websites. The claimants argue that, as a result, they cannot have been binding on the claimants to the extent that unilateral contracts were formed between Terraform and one or more of the claimants. I am unable to agree.

226 The “No Warranty” clauses do not exclude or limit liability. They instead stress that what is stated in the Terra and Anchor Websites should not be taken as giving rise to any contractual rights. There is nothing unusual, unreasonable or onerous in the “No Warranty” conditions. A person can hardly complain if a counterparty makes it clear that the counterparty is not intending to confer contractual rights on the person.

227 Sufficient notice was given of the existence of the Terra Terms of Use and Anchor Terms of Service so as to incorporate the “No Warranty” provisions. The Terra Terms of Use were accessible via a hyperlink at the bottom of the Terra Website which read “Terms of Use”. In my view, this arrangement was enough notice that the hyperlink contained terms that an ordinary user exercising reasonable and proper caution should consult before acting on matters stated in the Terra Website. The question of reasonable notice is not to be approached in the abstract. This must especially be the position in this case which involves representative claimants who took the time and made the effort to browse through the contents of the Terra Website and absorb its contents.

228 I do not accept the representative claimants’ argument that the hyperlinks to the Terra Representations were relatively more prominent than the link at the bottom of the page to the “Terms of Use”. The claimants submitted in opening their case at trial that:

229 Having examined the Terra Website, I cannot agree. Taking in the Terra Representations could not have been a matter of merely reading words that were prominently or conspicuously shown at the top of the Terra Website, without having to scroll to the “Terms of Use” hyperlink at the bottom of the page. On the contrary, a user would have had to go through several steps to click on hyperlinks with the specific intention of discovering more information about how the Terra Ecosystem worked, and how the value of UST was to be maintained. A user exercising ordinary care and prudence would not selectively click on some hyperlinks to bring up the Terra White Paper and the section entitled “About the Terra Protocol” on the Terra Website, while ignoring the “Terms of Use” at the bottom of the website. Nor would a reasonable person believe that he or she was forming a contract with Terraform solely based on a selective viewing of terms and conditions contained in some hyperlinks, while ignoring the hyperlink entitled “Terms of Use”.

230 I find therefore that the Terra Representations could not have formed part of a contractual bargain between Terraform and the representative claimants. The “No Warranty” conditions make it plain that Terraform was not making any representation or warranty by the Terra Website.

231 There was likewise reasonable notice of the Anchor Terms of Service. Before staking USTs on Anchor, one must connect one’s cryptocurrency wallet to the Anchor protocol. In the process of connecting a wallet through the Anchor Website, Mr Lim’s evidence (which I accept) is that a user would have come across the link to the Anchor Terms of Service through a drop-down box. Mr Gan claimed that when one uses the Terra Station extension on a Chrome browser, a wallet can be connected to Anchor without encountering a link to the Anchor Terms of Service. I have rejected this evidence at [132] above.

232 Since it is the act of staking which constitutes “acceptance” of any unilateral “offer” being made by Terraform, the “No Warranty” clauses within the Anchor Terms of Service would have been validly incorporated into any such transaction. As a result, Terraform cannot be regarded as having made any contractual warranties along the lines of the Anchor Representations. It is unnecessary for me to consider whether other terms (for example, exclusions or limitations of liability) in the Terra Terms of Use and Anchor Terms of Service were incorporated into the alleged unilateral contracts. In light of what I have held, there is no benefit to be gained, in so far as the claimants’ case is concerned, to considering whether terms other than the No Warranty provisions have been incorporated.

233 That leaves Terraform’s counterclaim which seeks declaratory relief as follows:

234 The declarations sought are too wide. As just observed, there is no need to make findings on whether all the terms and conditions in the Terra Terms of Use and Anchor Terms of Service were validly incorporated into the parties’ contracts, especially those terms relating to the limitation of liability and indemnification. I would also have to consider when particular terms came to be found on the relevant websites. On Mr Lim’s evidence, the earliest exhibited version of the Terra Terms of Use dates from 11 November 2021, while the earliest version of the Anchor Terms of Service dates from 12 October 2021. In those premises, I grant only limited versions of the declarations sought as follows:

235 There remain claims for breaches of unilateral contract based on the LFG Press Release (the Sixth Representation) and the Return to Form Tweet (the Seventh Representation). The claimants plead that the LFG Press Release amounted to an offer “that the 3rd Defendant’s reserves would be able to protect the price stability of UST vis-à-vis the 1 USD peg”. I am unable to agree. The LFG Press Release cannot reasonably be read as a contractual offer. It was not a proposal for members of the public to do anything to signify acceptance of the terms of an offer. It cannot objectively be read as making any promises or guarantees that the Bitcoin reserves amassed would maintain UST’s one-to-one peg to the USD under all circumstances and market conditions. The LFG Press Release merely informs readers that LFG would be amassing Bitcoin reserves to buttress the stability of the UST-USD peg. The LFG Press Release did not constitute a contractual offer to be bound by any terms and conditions.

236 Similarly, the Return to Form Tweet cannot reasonably be construed as an offer to be bound by any terms, let alone the pleaded term “that the 1st Defendant and 3rd Defendant would come up with a rescue plan to restore the 1 USD peg for UST, and that the Terra Ecosystem would soon ‘return to form’” [emphasis in original]. Nothing in Mr Kwon’s tweets of 9 to 11 May 2022 at [32]–[33] above, including the Return to Form Tweet, support such a reading. The tweets relied on for the Seventh Representation are:

237 Only the last tweet (the Return to Form Tweet) contains anything remotely close to a statement that the Terra Ecosystem would return to form. Even then, there is nothing to support an understanding that (a) the recovery plan was a plan to restore UST’s one-to-one peg to its original form prior to the de-peg or (b) Terraform was making a unilateral offer to the world at large, warranting that its rescue plan will be successfully implemented. There is nothing in the tweet which is capable of being accepted whether through the purchase of UST tokens or otherwise, so as to confer a right to sue if the rescue plan failed. In any event, it would be odd for Terraform to be making (or to be understood as making) contractual offers to members of the public through the personal social media account of an officer or shareholder. The social media platform X (formerly Twitter) is hardly the platform for companies to be extending binding contractual offers to potential customers and clients, especially through cryptic tweets.

238 I find that the Return to Form Tweet did not constitute a contractual offer to be bound by any terms and conditions. I dismiss all claims for alleged breach of unilateral contract.

239 That leaves only the claims of inducement of breach of contract (against Mr Kwon) and unlawful means conspiracy (against all three defendants). Since I have found that no unilateral contracts were formed between Terraform and the representative claimants, Mr Kwon could not have induced any breach of contract by Terraform. The claim for inducement of breach of contract fails. As for conspiracy to injure by the unlawful means of fraud and deceit, that requires proof of a combination among the defendants (a) to commit the relevant unlawful acts with (b) the intention of injuring the claimants (see Quah Kay Tee v Ong and Co Pte Ltd [1996] 3 SLR(R) 637 at [45]). There is not a shred of evidence of either ingredient. I therefore dismiss the claim for unlawful means conspiracy.

240 The final position is as follows:

241 Not all defendants were involved in the five Conceded Representations. For the avoidance of doubt, I set out the defendants which are jointly and severally liable in respect of the foregoing awards of damages. Of the five Conceded Representations, Terraform and Mr Kwon conceded the First, Second, Third, and Fourth Representations, while LFG admitted to making the Sixth Representation. Terraform and Mr Kwon only admit that the Sixth Representation “was made” albeit not by them. The claimants initially appear to treat LFG as the sole representor of the Sixth Representation. They plead that “the 3rd Defendant announced that it had successfully raised USD 1 billion … as a reserve fund”. But later they suggest that both Terraform and LFG made unilateral contractual offers to the world on the Terra Representations and Sixth Representation, before asserting that only Terraform made a unilateral contract on the terms of the Sixth Representation. Given the claimants’ confused position, I follow the approach in the defendants’ concessions and order as follows:

(a) Terraform and Mr Kwon are jointly and severally liable for the damages awarded to Mr Davis at [240(c)] above.

(b) Terraform and Mr Kwon are jointly and severally liable for the damages awarded to Mr Macquisten at [240(d)] above.

(c) Terraform and Mr Kwon are jointly and severally liable for the damages awarded to Mr Xue at [240(f)] above.

(d) Terraform and Mr Kwon are jointly and severally liable for the damages awarded to Mr Arun at [240(g)] above.

(e) Terraform and Mr Kwon are jointly and severally liable for the damages awarded to Mr Lee at [240(h)] above.

(f) Terraform, Mr Kwon, and LFG are jointly and severally liable for the damages awarded to Mr Epstein at [240(i)] above.

(g) Terraform, Mr Kwon, and LFG are jointly and severally liable for the damages awarded to Mr Tan at [240(j)] above.

242 All other claims in OA 3 are dismissed. I allow Terraform’s counterclaim for declaratory relief to the extent specified in this judgment at [234] above and dismiss the remainder of Terraform’s counterclaim.

243 None of the parties has fully prevailed. In those circumstances, it seems to me that on a rough and ready basis the appropriate costs order would be that each party is to bear its own costs. There will be an order nisi to that effect in OA 3 (inclusive of all interlocutory summonses for which costs were reserved in the cause of OA 3). The parties will have 14 days to file an application to vary the costs order nisi, accompanied by written submissions limited to 10 pages. In the event of such application being made, all other parties will have 14 days to file costs submissions in reply thereto, with the same limit of 10 pages applying. In the absence of any application to vary the costs order nisi, it will become absolute.

244 I awarded costs of SUM 37 to the defendants (see at [60] above), leaving the quantum to be determined at a subsequent date. The parties have since provided their costs submissions via correspondence with the court. Terraform and LFG seek costs of S$5,000.00, while the representative claimants submit that the quantum should only S$1,500.00. It seems to me that the former figure is reasonable and the defendants should be entitled to their costs of S$5,000.00 (all-in, inclusive of disbursements).