1 This is the appeal of the defendants in this action, KBP Biosciences Pte Ltd (“KBP”) and Dr Huang Zhenhua (“Dr Huang”), from the judgment dated 12 August 2025 (“Judgment”) of Philip Jeyaretnam J (“Judge”) sitting as a Judge of the Singapore International Commercial Court. By his Judgment, the Judge dismissed the appellants’ application to set aside a worldwide freezing order made by him against them on 14 February 2025, save for a minor variation to the disclosure provisions of the order from which there is no appeal. The proceedings arise out of the sale of the rights to a drug, Ocedurenone, which KBP was developing. The respondent, Novo Nordisk A/S (“Novo”), purchased these rights pursuant to an Asset Purchase Agreement executed on 11 October 2023 (“the APA”). The APA is subject to New York law and provides for arbitration in New York. Novo’s primary case arising out of this acquisition is that it purchased the rights to Ocedurenone as a result of fraudulent breaches of warranty and misrepresentations by KBP and Dr Huang. Pursuant to s 12A of the International Arbitration Act 1994 (2020 Rev Ed) (“IAA”), Novo sought a worldwide freezing order in support of a New York-seated arbitration which it intended to bring, and has now brought, against the appellants.

2 Ocedurenone was a potential new drug to control blood pressure and provide kidney protection. KBP, a Singaporean company, was founded by Dr Huang. Dr Huang, with Dr Yonghong Fred Yang as the Chief Development Officer, was responsible for the drug’s development.

3 The development of Ocedurenone involved clinical trials conducted in phases in various parts of the world. By the time of the acquisition, Phase 2 trials had been completed, and Phase 3 trials were about to begin.

4 The Phase 2 trials for Ocedurenone started in April 2018 and were managed by Worldwide Clinical Trials, Inc (“WCT”) as the contract research organisation. Patients were recruited from several countries, including the US, Chile, Georgia, Bulgaria, Hungary, Israel, and Australia. By 1 June 2020, all patients had been enrolled, with those from Bulgaria enrolled in March 2020. KBP was the sponsor. It appointed Dr George Bakris (“Dr Bakris”) and Dr Bertram Pitt (“Dr Pitt”) as lead principal investigators.

5 Phase 2 trials were completed on 5 August 2020 and the results demonstrated that Ocedurenone had a treatment effect. KBP’s position was that a Phase 2 clinical study report (issued after the Phase 2 trials and approved by the lead principal investigators) (“the Phase 2 CSR”) demonstrated efficacy in multiple regions. In 2021, articles were published by Dr Bakris, Dr Pitt, and others, claiming that Ocedurenone effectively lowered blood pressure and demonstrated clinical efficacy and safety.

6 Novo’s case in the dispute in the arbitration is that data from the Bulgarian site was solely responsible for Ocedurenone displaying a treatment effect in these Phase 2 trials. It alleges that material information regarding the Bulgarian site’s data was not disclosed prior to acquisition.

7 Moreover, Novo relies on the results of an unplanned administrative interim analysis (“Phase 2 Interim Analysis”) carried out by KBP in or around January 2020 to assist with manufacturing decisions for the Phase 3 study. On 13 April 2020, KBP and WCT approved a blinding plan (or strategy to minimise bias in treatment) for this Phase 2 Interim Analysis. KBP’s clinical team, including the lead principal investigators, were intended to have no knowledge of the Phase 2 Interim Analysis results due to this blinding plan. The Phase 2 Interim Analysis was made available to KBP in June 2020, based on data from the first 90 patients to complete week 12 of Phase 2. These results showed no clear treatment effect for Ocedurenone.

8 Novo became interested in acquiring Ocedurenone in early 2023 and conducted due diligence while the Phase 3 trials were ongoing.

9 KBP set up a data room comprising information concerning Ocedurenone so that Novo could investigate the drug. This included information as to the Phase 2 trials. As part of this data, KBP provided Novo with the Phase 2 patient data in Portable Document Format (“PDF”) format. However, KBP did not disclose its own analyses of that data or direct attention to the significance of the Bulgarian site results. Specifically, KBP uploaded (in PDF format) to the data room a listing containing data on the vital signs of every patient enrolled in Phase 2, referred to as the “Vital Signs Data Listing”.

10 KBP also uploaded a set of documents known as the “CTA Package”, previously submitted to the Chinese National Medical Products Administration, which included Phase 2 data.

11 It is noted that these documents enabled Novo to access patient-level data, including the site each patient was from and their systolic blood pressure (“SBP”).

12 However, while these documents contained the raw data, they were not provided in an analysable format that would have enabled Novo to readily identify anomalies, such as the significance of the results from the Bulgarian site. Analyses conducted by KBP of this data were not disclosed, nor (as appears below) were external concerns about the data’s anomalies.

13 Having conducted its investigations, Novo entered into the APA, with the acquisition closing on 29 November 2023. Novo paid US$700m to KBP and deposited US$100m in escrow.

14 Post-closing, in April 2024, PAREXEL International (IRL) Ltd, the organisation engaged to report on the Phase 3 trials released an interim analysis of the results of the Phase 3 trials. KBP contended that this contained errors. The overall result was that Ocedurenone lacked efficacy and fulfilled the criteria set for futility, leading to a public announcement by Novo in June 2024 of Phase 3’s failure and an impairment loss of over US$800m. Investigations by Novo revealed data anomalies, particularly at the Bulgarian site, leading to it submitting written complaints to regulatory agencies.

15 After the acquisition, and indeed after the ex parte hearing of its application for a worldwide freezing order, Novo also learnt that an earlier potential buyer of Ocedurenone, the pharmaceutical company, Otsuka, had decided to discontinue acquisition discussions because of doubts about consistency and data repeatability. Otsuka had reviewed the data room provided by KBP, including the data containing Phase 2 results, and had observed an “[i]mbalance in clinical data: Single site (Bulgaria) was responsible for bulk of efficacy, raising questions about repeatability in [Phase 3]”.

16 In response, KBP, in an email dated 25 August 2021, accepted that the SBP response at the Bulgarian site was superior to other countries and regions but explained this on the basis that the Bulgarian patients were enrolled later and were provided with additional guidance.

17 Otsuka was not persuaded by KBP’s explanation. An Otsuka representative, in an email dated 7 September 2021, indicated that they did not find the data compelling, saw no correlation between pre-dosing stability and change in SBP country by country, and hence were not convinced this explained Bulgaria’s superior results.

18 Otsuka decided not to proceed with the acquisition at that stage, primarily due to the inconsistent and anomalous results at the Bulgarian site. As explained below, the Judge considered that the evidence concerning the abortive negotiations with Otsuka indicated that the inconsistency in the results was a material issue for prospective buyers, and that senior management at KBP, including Dr Huang, were aware of its importance.

19 Novo had to show that there was a risk that the appellants would dissipate (ie, make unjustified disposals of) some US$330m being the balance of the proceeds of sale of Ocedurenone present in the jurisdiction. The Judge had evidence as to what happened to the proceeds of sale after payment by Novo. With respect to the risk of dissipation, he focused on the destination of approximately US$330m and, as explained below, he found that the transfer of this sum to Dr Huang personally was dissipative. His conclusion, which is challenged in this appeal, was based on a combination of factual findings and the application of established legal principles concerning the risk of dissipation. We deal separately with those legal principles below when we come to the parties’ submissions.

20 Of the total cash consideration of US$700m paid on acquisition, the Judge found that approximately US$330m was transferred from KBP to its ultimate beneficial owner, Dr Huang, and that such transfer was dissipative conduct. The Judge’s findings relevant to this issue may be summarised under the following subheadings.

(a) Speed and Sequence of Transfers: The Judge found that after the acquisition of Ocedurenone by Novo and the payment of US$700m to KBP, the majority of those proceeds were rapidly moved out of KBP. By June 2024, nearly all the purchase consideration had been transferred from KBP to its holding company in the Cayman Islands, KBP Biosciences Holdings Ltd (“KBP Cayman”). By the end of 2024, almost all of this was, in turn, transferred out by KBP Cayman. KBP held only US$435,000 as at 7 March 2025. Substantial sums were then routed to external lenders, consultants, and employees, but US$30m was paid directly to Dr Huang as bonus in tranches during 2024, and about US$300,965,696.69 to KBP Cayman’s parent company in the British Virgin Islands, KBP Holdings Ltd (wholly owned by Dr Huang), and (via Dr Huang’s company Panda Pharma Holdings Ltd (“Panda Pharma”)) ultimately to Dr Huang’s personal accounts in various tranches in June and August 2024 (Judgment at [50]).

(b) Anticipation of Claims by Novo: The Judge found that there was evidence that as early as January 2024, Dr Huang anticipated the possibility of claims from Novo related to the transaction, specifically if the drug failed its Phase 3 trials. Minutes from a KBP Cayman shareholder meeting recorded Dr Huang discussing the risk of Novo making damage claims against KBP if milestones or payments were not met (Judgment at [52]). The appellants rejected this interpretation of his evidence. Their submission was that the statement Dr Huang made related to the possibility of claims being made by Novo if KBP did not comply with its post-closing obligations under the APA. There was clearly a dispute about this, but it was not fundamental to the Judge’s overall view about the transfers of monies post-closing.

(c) Nature of Payments and Lack of Justification: The Judge found that, while large sums were paid out ostensibly to pay off obligations or creditors, the payment of approximately US$330m to Dr Huang was not shown to be one to which Dr Huang was contractually entitled. The appellants could not point to any contractual obligation for KBP Cayman to pay such sums to Dr Huang, only that Dr Huang felt it appropriate, given external investors had been repaid (Judgment at [54]).

21 Novo's position is that material information was withheld. In particular, it should have been told more about the Phase 2 trials than it was simply by the addition of certain data to the data room, as described above. That data was not true and complete because (among other matters) it did not in terms reveal that the result of the Phase 2 trials depended on the results of the Bulgarian site. The results of the Bulgarian site were out of line or inconsistent with the results of other sites.

22 Novo could have reached this conclusion by an analysis of the Vital Signs Data Listing, as indeed Otsuka had done.

23 However, the Judge held that there was a good arguable case that Novo was not obliged to carry out this analysis. Rather, KBP should have made a fair disclosure of that factor. The fact that Otsuka had withdrawn showed the materiality of information as to this inconsistency (Judgment at [44]–[45]). The APA required its disclosure. Section 4.8(h) of the APA stated that “Seller has made available to Buyer true, complete and accurate copies of … all material information in any KBP Group Entity’s possession or control concerning the safety, efficacy, side effects, toxicity, or manufacturing quality and controls of any Compound or Product” (Judgment at [33]).

24 There were also quality and compliance issues at the Bulgarian site which KBP had not disclosed (Judgment at [34]). The Judge considered that there was a good arguable case that KBP should have disclosed these matters. The appellants say that in error he stated that these issues were not “seriously contested” and point to the findings of the steering committee which oversaw the conduct of the Phase 3 trials (“Phase 3 Steering Committee”), but that error cannot undermine his conclusion that there was sufficient basis in this to find a good arguable case as it was a matter as to which the Court had itself to be satisfied. Likewise, the Phase 2 Interim Analysis was not disclosed even though the Phase 2 CSR released into the data room by KBP stated that “No interim analysis for efficacy was planned for this study” (Judgment at [38]). Nor was there disclosure of the negotiations with Otsuka (Judgment at [43]–[44]).

25 On 10 February 2025, Novo filed an originating application without notice, and succeeded in obtaining the grant of a worldwide freezing order by the Judge against KBP and Dr Huang on 14 February 2025. The APA stipulated arbitration under the auspices of the International Chamber of Commerce (“ICC”) in New York as the exclusive dispute resolution mechanism, and Novo filed for arbitration on 26 March 2025. On 14 April 2025, KBP initiated proceedings in Denmark to prohibit Novo from enforcing the worldwide freezing order. There were hearings in this application on May, June and July 2025, and on the last occasion there was a substantive hearing lasting three days. An anti-suit injunction sought by Novo against the Danish proceedings was granted on 19 June 2025.

26 Before the Judge on the application to set aside the worldwide freezing order, the fundamental issue was whether the conditions for a worldwide freezing order had been met. In particular, the appellants contended that there was no good arguable case and no real risk of asset dissipation. The appellants also claimed that Novo breached its duty of full and frank disclosure to the Court and that the legal requirements for interim relief under s 12A of the IAA were not satisfied (Judgment at [2]).

27 The Judge found that Novo had a good arguable case for fraud under New York law in that KBP failed to disclose material information about the drug Ocedurenone in breach of the APA. The Judge concluded that the information about the Bulgarian site was objectively material to a purchaser of the drug, as consistency of results was crucial for evaluating efficacy (Judgment at [36]). (The significance of it being “objectively” material was that it did not turn on Novo’s selection of this matter). Novo only obtained the Phase 2 Interim Analysis results after the APA was completed, and they showed no statistically significant response to treatment with Ocedurenone (Judgment at [39]). The evidence established a good arguable case that material information concerning efficacy was not disclosed, breaching the representation and warranty in the APA (Judgment at [40]).

28 The Judge then turned to the question of knowledge of the breach of warranty which would go to fraud and that, absent fraud, the APA’s US$100m escrow cap would apply (Judgment at [41]). There was evidence at the ex parte stage that Dr Huang knew the Phase 2 Interim Analysis showed no statistical efficacy without Bulgaria, as he received the results on 1 June 2020 (Judgment at [42]). The Judge found that the evidence established a good arguable case that KBP and Dr Huang deliberately withheld objectively material information about the Bulgaria site. They were aware from the prior negotiations with Otsuka that this information could determine the outcome of acquisition discussions (Judgment at [43]–[44]). KBP was under financial pressure.

29 KBP argued that it had disclosed all raw data in PDF format during due diligence, so Novo could have detected the Bulgaria site’s impact. The Judge found this unpersuasive because Novo would not have known to look for anomalies without being alerted, and simply supplying raw data in an unanalysed format, without highlighting the concerns, did not equate to full transparency. The manner of disclosure suggested an attempt by KBP to appear candid while withholding crucial information (Judgment at [45]).

30 In summary, in the Judge’s judgment, there was a good arguable case that KBP and Dr Huang deliberately withheld objectively material information about the Bulgaria site and that they intended to induce Novo to complete the acquisition it might otherwise have abandoned (Judgment at [45]–[46]).

31 The Judge found that there was a sufficient risk of dissipation of assets from Dr Huang’s and KBP’s pattern of transferring substantial funds, through various entities, into Dr Huang’s personal control, in circumstances indicating an awareness of possible claims and an intention to place those assets beyond the reach of enforcement, particularly given prior financial difficulties and allegations of deliberate non-disclosure (Judgment at [47]–[56]). The matters on which the Judge relied have been summarised at [20] above.

32 The Judge was accordingly satisfied that there was a real risk of dissipation of assets by KBP and Dr Huang. The overarching test is whether there was a real risk that judgment may not be satisfied due to unjustified dealings with assets (JTrust Asia Pte Ltd v Group Lease Holdings Pte Ltd [2018]2 SLR 159 (“JTrust”) at [64]). KBP transferred almost all of the US$700m purchase consideration to its holding company, KBP Cayman, which then transferred most of it out by the end of 2024. Substantial payments were made to external lenders, investors, and Dr Huang, who received US$30m as a bonus and approximately US$300m through his vehicle, Panda Pharma. The Judge found that Dr Huang anticipated a claim by Novo against KBP as early as January 2024, indicating a risk of dissipation. The Judge concluded that moving approximately US$330m from KBP to Dr Huang was unjustified and dissipative conduct. The Judge therefore relied on post-closing money movements as establishing the risk of dissipation, not the dishonesty of which he had found that there was a good arguable case. Contrary to the submission of Mr Cavinder Bull SC (“Mr Bull”), counsel for the appellants, the Judge therefore did undertake an analysis of how the appellants’ dishonesty in respect of the provision of information related to the risk of dissipation. He found essentially that that dishonesty was linked to the risk of dissipation of the remaining proceeds of sale through the evidence as to the movement of monies post-closing.

33 The principles of full and frank disclosure require a claimant making an ex parte application to disclose all material facts to the court. The Judge directed himself as to the content of those principles by reference to [90] of the judgment of this Court in JTrust (Judgment at [57]):

34 The Judge also pointed out that the duty of full and frank disclosure is also reflected in O 18 r 1(7) of the Singapore International Commercial Court Rules 2021 (Judgment at [58]).

35 The appellants alleged before the Judge that Novo failed to disclose to the Court material facts at the ex parte hearing when it applied for the worldwide freezing order, including information as to the efficacy of Ocedurenone beyond Bulgaria and the Phase 3 Steering Committee’s findings. However, the Judge found that Novo’s case was based on KBP’s non-disclosure of information, not the inefficacy of the drug. The appellants raised some twelve heads of non-disclosure of relevant facts. The Judge went through each one and, applying the principles set out above, concluded that the non-disclosures were not material enough to warrant setting aside the worldwide freezing order (Judgment at [59]–[91]).

36 The Judge was satisfied that the Court could order relief under s 12A of the IAA in support of the New York arbitration. The requirements were fulfilled as the arbitral tribunal was unable to act effectively at the time of the ex parte application, the case was urgent, and it was appropriate for the Singapore court to make the order. The ICC emergency arbitrator could not grant the freezing order on an ex parte basis, and the risk of dissipation justified the urgency of the application (Judgment at [92]–[102]).

37 There is no dispute that the applicant for a freezing order must show a “good arguable case” on the merits of the claimant’s claims or as to the meaning of this requirement, so it suffices to set out the meaning given to this requirement in Singapore Civil Procedure 2026 vol 1 (Cavinder Bull gen ed) (Sweet & Maxwell, 2026)at para 13/1/63:

38 Mr Bull, for the appellants, submits that this requirement was not met and the Judge erred in concluding that it was met. All underlying data, including the Bulgarian site information, was provided through the Vital Signs Data Listing and Novo failed even to open or analyse it prior to completion of the transaction. The raw data would have enabled Novo to identify any issues with the Bulgarian site had it conducted a deeper analysis. There was no standard industry practice that required buyers to rely on interim analyses and such interim results were immaterial given the eventual positive topline findings and the involvement of external investors and independent scientists. The evidence before the Judge did not amount to a “good arguable case” because it was “shaken” by these points.

39 The appellants further submit that the Judge was not entitled to come to any conclusion as to dishonesty or deliberate conduct in relation to “good arguable case” because the question of fraud could only be determined at a trial. These matters were hotly contested. The Judge was wrong in law to make findings on these issues. There was evidence that placed a different interpretation on the Phase 2 CSR results and as to the practice about disclosure in these situations which would ultimately show that there was no wrongful withholding of information from Novo.

40 Mr Ong Tun Wei Danny SC (“Mr Ong”), for Novo, submits that the Judge correctly found a good arguable case of fraud, focusing on the deliberate non-disclosure of material information related to the efficacy of Ocedurenone specifically the significance of the Bulgarian site and related irregularities. Internal analyses showed the drug had no efficacy without the Bulgarian site; such analyses were not disclosed to Novo during due diligence. On Novo’s case, the Phase 2 trial’s positive efficacy was entirely driven by outlier results from the Bulgarian site, with no significant treatment effect at the other 61 global sites. On his submission, serious irregularities and compliance concerns existed at the Bulgarian site, including a remarkably low screen-failure rate, flagged issues by project auditors, and explicit breach letters; these were withheld from Novo. Otsuka had declined acquisition when satisfactory explanation for Bulgarian site data was not forthcoming. The APA’s representations and warranties required full disclosure of all material information; these obligations were not met.

41 Moreover, with regard to the appellants’ argument that it was sufficient for KBP to provide the raw data, Mr Ong submits that the APA imposed a higher obligation to affirmatively disclose materially anomalous results, rather than requiring Novo to discover issues in voluminous data. Novo had no reasonable basis to suspect the need for independent re-analysis.

42 In our judgment, all that had to be shown about the allegations made by Novo was that they were substantiated and had substance, not that they would necessarily succeed at trial. Moreover, Novo had the benefit of s 4.8(h) of the APA which required KBP to “ma[k]e available to Buyer true, complete and accurate copies of … all material information in any KBP Group Entity’s possession or control concerning the safety, efficacy, side-effects, toxicity, or manufacturing quality of controls of any Compound or Product.” That set the standard for disclosure of material information to Novo.

43 On the question whether there had been disclosure which satisfied s 4.8(h) of the APA, the Judge first rejected the argument that the comparison between the results of the Bulgarian site and the results of the other sites was not objectively material information. It was in his judgment material because it resulted in inconsistency between the results across all the sites and consistency was important to the purpose of evaluating the risk which the acquirer would undertake (Judgment at [36]). In our judgment that conclusion plainly establishes a good arguable case especially when coupled with the actions of Otsuka and the information provided by Novo that the results of the Bulgarian site drove the Phase 2 results. It does not cease to be a good arguable case because the appellants contend that the Bulgarian site did not in fact drive the Phase 2 results and that there was a reference in the data room to the existence of the Phase 2 Interim Analysis. Another factor noted by the Judge was that KBP referred to the importance of consistency in its Phase 2 CSR.

44 Moreover, KBP included in the data room a statement that no interim analysis was planned (Judgment at [38]). That statement was never corrected. Moreover, the Phase 2 Interim Analysis did not include the results of the Bulgarian site. The appellants did not dispute that the Phase 2 Interim Analysis showed no statistically significant result. Accordingly, that report must have been material. The Judge concluded that this evidence sufficiently established a good arguable case (Judgment at [40]).

45 The appellants’ primary case is that sufficient data had been uploaded in the data room for Novo to discover for itself that the Bulgarian site drove the results. The Judge correctly dismissed that case on the basis of s 4.8(h) of the APA. In any event, KBP had failed to disclose the Phase 2 Interim Analysis and had not disclosed the Otsuka matter. The Judge held that these matters suggested an attempt to “keep Novo in the dark about the problems with the Bulgarian site while offering a plausible story of ‘candour’ if Novo later discovered those problems” (Judgment at [45]). The Judge was not here making a finding as to motivation but pointing out that the evidence to which he referred was suggestive of the motivation stated. Whether it was so or not would be a matter for trial.

46 The Judge then had to consider the knowledge of Dr Huang because unless there was also fraud, there was a cap of US$100m on damages in s 11.2(b)(i) of the APA, which would be met by the funds held in escrow (Judgment at [41]). The Judge noted that Dr Huang would have been one of those who received the Phase 2 Interim Analysis and he must have known about the state of negotiations with Otsuka (Judgment at [42]–[44]). The Judge also bore in mind as possible motivation the fact that, at that time, KBP was “financially strapped”, having taken a high-interest emergency loan in 2023 (Judgment at [46]). This does not mean that he treated fraud as established evidentially but that there was material from which inferences could properly be drawn at trial.

47 There is no rule of law which prevents a Court from reaching a conclusion that there is a good arguable case as to fraud or deliberate conduct merely because it is contested. The Court must of course take into account the points which the defendants make in response to the allegation that they acted fraudulently or deliberately, and particularly where the allegation involves fraud, proceed with caution, as this Court advised in Bouvier, Yves Charles Edgar v Accent Delight International Ltd [2015] 5 SLR 558 (“Bouvier”) at [62] (which is considered in more detail below). In this case, the Judge heard the setting aside application over three days and so had ample opportunity to weigh up the arguments on each side. However, if the Judge nonetheless concludes that there is sufficient evidence to amount to a good arguable case on certain issues, it is his duty so to find.

48 It is clear that there is much detailed evidence about the information that was available to be disclosed and scope for argument as to what should have been disclosed. Nonetheless, in our judgment, there was sufficient evidence to support the Judge’s conclusion of a good arguable case against KBP and Dr Huang, and his conclusion that there was a good arguable case that the non-disclosure of information about the Bulgarian site was deliberate and dishonest for the reasons he gave (see [43]–[46] above). The case against them remained arguable on substantial grounds despite the other matters on which evidence was adduced. Ascertainment on the totality of the evidence of the standard of good arguable case was as far into the merits as the Judge needed to go on an application to set aside a worldwide freezing order. The merits of Novo’s claims and all of the appellants’ defences remain for final adjudication in the arbitration independent of and unaffected by any finding by the Judge in these matters on good arguable case.

49 Mr Bull submits that the Judge impermissibly inferred risk of dissipation based on “hotly contested” dishonesty allegations, contrary to the approach required by the decision of this Court in Bouvier. In Bouvier, this Court required not only a good arguable case of dishonesty, but also that the alleged dishonesty must be of a nature that has a real or material bearing on the risk of dissipation. As this Court put it at [94] of Bouvier:

50 On the appellants’ submission, being (for example) “dishonest in contract dealings” is not the same as evincing intent to dissipate assets: in assessing whether there is a real risk of dissipation, a mere allegation of dishonesty alone is insufficient because the dishonesty in question must have a real or material bearing on the risk of dissipation. The Judge had in error concluded, from his findings that Novo had a good arguable case against the appellants for fraud, that Dr Huang had as a result “exhibited dishonesty that bears on the risk of dissipation” (Judgment at [56]) without undertaking any analysis on how the alleged dishonesty bore on the risk of dissipation. In so doing, the Judge erred in law.

51 Mr Bull further submits that the Judge fell into error in his three main findings on risk of dissipation: by concluding from his finding of a good arguable case of dishonest non-disclosure that there was a risk of dissipation when the dishonesty was not linked to risk of dissipation, by failing to find that extensive data was transparently provided, and by anticipating a claim against KBP on the basis of statements made by Dr Huang as KBP had provided a different explanation. Those statements did not show that Dr Huang was aware of potential fraud claims from Novo and it was implausible he would be so open with external investors if cognisant of fraud risk.

52 The Judge was wrong to describe the cash movements as “speedy” upstreaming of funds to Dr Huang. The movements took place months after alleged anticipation of claims and occurred in separate tranches. A person who wished to evade a claim would be more likely to dissipate assets quickly. Further, Dr Huang left US$295m in Singapore bank accounts, which was again inconsistent with an attempt to put assets beyond reach.

53 The upstreaming of funds to Dr Huang was justified as the proper return to a sole shareholder after redemption of external investors; the bulk of funds were not transferred rapidly, and further, significant assets remained in Singapore.

54 Mr Ong submits the Judge applied the correct approach in assessing risk of dissipation, requiring solid evidence of a real risk that any judgment will be unsatisfied due to unjustified dealings with assets. Dishonesty must “bear on” the risk of dissipation but need not be limited only to theft or classic misappropriation.

55 Novo relies on the prior transfers of money as evidence of risk of dissipation. KBP transferred US$339.1m to KBP Cayman within a month of receiving the proceeds, and a further US$300.9m to Dr Huang via corporate structures, without clear commercial rationale. The timing and explanation for the payments (eg, as “bonuses” to Dr Huang) were unconvincing and unexplained.

56 Mr Ong further argues that the dissipation was purposeful and timed in the context of anticipated claims from Novo, beginning soon after the closing and after knowledge of the drug’s failings and impending fraud revelation.

57 Mr Ong submits that the true holding in Bouvier and other authorities is that the alleged dishonesty must “bear on” risk of dissipation, especially since the conduct involved procurement of a US$700m up-front payment through non-disclosure.

58 Mr Ong submits that the appellants’ explanations for the movements of money were belated or based on technical distinctions that do not address the substance of their conduct. He further submits that assets have been reduced to very low levels within the company, inconsistent with purported ongoing operational needs. The retention of assets by Dr Huang in Singapore is not inconsistent with dissipation, as the risk assessment should focus on unjustified movement of funds and the potential for future dissipation, not whether all assets have exited the jurisdiction.

59 We now turn to our conclusions. The Judge dealt with the question of whether there was a real risk of dissipation in a structured manner.

60 The first question was whether Novo could establish a real risk of dissipation which was separate from the dishonesty which Novo alleged. It did not follow from the mere fact that KBP and Dr Huang were alleged to have been dishonest in relation to the non-disclosure of information to Novo that they intended to use their monies for unjustified purposes rather than pay any judgment that Novo might obtain. This point is established by Bouvier.

61 In our judgment, Bouvier is to be understood in the context of the facts of the case. An art dealer was alleged to have arranged for the purchase of artworks on the plaintiffs’ instructions not only for commission but also to have received a substantial undisclosed profit for his services from other parties. This Court was satisfied that there was a good arguable case against him for dishonest breach of fiduciary duty. The question then was whether there was a risk that he would dissipate his assets. On the facts, the art dealer was a person who was well-settled in Singapore and he informed the court that he had no intention to evade any judgment by removing assets. This was not apparently disputed. The plaintiffs did not seek to show what he had done with the proceeds for which they alleged the art dealer was accountable to them. It is thus reasonable to assume that in Bouvier there was no evidence of actual dissipation.

62 The question of law for this Court in Bouvier was the correctness of an earlier High Court decision, Spectramed Pte Ltd v Lek Puay Puay [2010] SGHC 112. This decision held that once there was an arguable case as to dishonesty no separate issue arose as to risk of dissipation.

63 In a case such as this, the facts relevant to dissipation are very different from those in Bouvier. The appellants have utilised some of the proceeds for proper corporate purposes. In addition, KBP as a trading company would continue to need money for operating its business. Furthermore, Dr Huang is an indirect shareholder, and in his affidavit evidence to the Judge he stated that he was entitled to the proceeds of sale not used for paying current commitments or other investors:

64 Dr Huang reached that view without there being any evidence of any agreement between the companies and himself to that effect.

65 In those circumstances, in our judgment, there are material differences between the facts of this case and those in Bouvier. In our judgment, the Judge was entitled to consider that there was evidence relevant to the risk of dissipation which was independent of the unproven allegation of dishonesty. Dr Huang believed he was entitled to US$300m without considering the company's liabilities or agreements. This was evidence that there was a risk he might unjustifiably transfer assets to prevent recovery of claims. In those circumstances, in our judgment, it cannot be said that the Judge failed to follow the decision of this Court in Bouvier.

66 Another part of Mr Bull’s case under this head is that the Judge went further than this and actually found dishonesty. Mr Bull founds this argument in particular on the Judge’s concluding paragraph under the topic of the risk of dissipation (Judgment at [56]):

67 In our judgment, Mr Bull’s argument fails to give weight to each sentence in this paragraph in its context. In the first sentence, the Judge addresses two matters: the good arguable case requirement and the requirement that the dishonesty bears on the risk of dissipation, which was the ratio of Bouvier. The Judge does not express the sentence in terms of a finding of dishonesty but simply as a statement that there is cogent evidence of dishonesty. This meets the good arguable case requirement for a worldwide freezing order.

68 When the second sentence refers to dishonesty, it is to dishonesty in the sense of dishonesty of which there is cogent evidence, no more. The second sentence connects the withholding of information with the speedy movement of the proceeds of sale.

69 The third and fourth sentences underscore that the Judge is not making a finding of dishonesty: that is a matter for the decision-maker at a trial. In the final sentence, the Judge distinguishes the situation after a trial. At this interlocutory stage, the Judge is assessing whether there is a good arguable case, rather than making a final determination. The Judge correctly holds that he must nonetheless decide at that stage whether there is a risk of dissipation, and do so on the basis of the evidence then available. In our judgment, the finding of a risk of dissipation at the interlocutory stage on the basis of the limited evidence then available is clearly not the same as a finding of fact at a trial.

70 We therefore also reject Mr Bull’s argument that the Judge made any finding of dishonesty which ought only to be made at trial. We further reject his argument that the Judge did not undertake any analysis to show how the appellants’ alleged dishonesty bore on the risk of dissipation.

71 As mentioned at [33] above, the Judge carefully explained the applicable principles by reference to the decision of this Court in JTrust. Those principles demonstrate that the failure to make full and frank disclosure does not automatically result in the termination of an order made by the Court. The Court has a discretion to continue the order or impose new terms, even if material non-disclosure is proven. This is because the Court considers matters such as the importance of the undisclosed fact, whether the non-disclosure was innocent, and the overall circumstances (see Tay Long Kee Impex Pte Ltd v Tan Beng Huwah [2000] 1 SLR(R) 786 at [27]). The claimant may still face penalties, such as costs, but the order may remain if the Court deems it appropriate. The Court may make an error in understanding the facts in issue and that error if material may lead to the setting aside of its conclusion as to whether there was a failure to make full and frank disclosure. Where, however, the Court directs itself in accordance with the applicable principles, takes account of (only) relevant matters (and does not omit to consider any relevant matter), and exercises a discretion available to it, and no procedural unfairness or irregularity is in issue, the Court’s assessment cannot be set aside on appeal except where the Court was plainly wrong, that is, exceeded the generous ambit within which reasonable disagreement is possible, and thus acted outside the ambit of its authority (see Nomura Regionalisation Venture Fund Ltd v Ethical Investments Ltd [2000] 2 SLR(R) 926 at [35]–[36]). The same applies where the Court relies on matters which require its evaluation, such as the materiality of a particular fact, or comes to its conclusion by balancing the factors. Where the decision of the lower Court is in accordance with these principles, the appellate Court will exercise restraint in setting aside the lower Court’s determination (see generally the discussion in Re Sprintroom Ltd [2019] 2 BCLC 617 at [72]–[78], a decision of the Court of Appeal of England and Wales which concerned the trial judge’s assessment of “unfair prejudice” at the trial of a minority shareholder’s action).

72 On this appeal, the appellants argue that Novo failed to make full and frank disclosure at the ex parte application under five heads, raising five of the matters which were raised before the Judge and which the Judge considered in detail.

73 The appellants rely first on Novo’s failure to point out that KBP disclosed the raw data in the data room. The essence of the Judge’s reasoning for rejecting this point was that simply providing raw data without explaining its significance did not meet disclosure obligations (Judgment at [71]). Novo was not required to re-analyse the data, and there was no material omission by Novo.

74 The second area of non-disclosure relied on by the appellants is Novo’s failure to point out KBP’s position that Ocedurenone displayed efficacy at sites other than Bulgaria, but the Judge did not consider that omission to be material to the case as Novo's case focused on undisclosed anomalies at the Bulgarian site, not on efficacy elsewhere (Judgment at [74]–[75]).

75 The third area was in relation to non-disclosures regarding findings of no fraud at the Bulgarian site. The Judge concluded that not disclosing findings of no fraud at the Bulgarian site was not a material omission as Novo's case was about concealing anomalies, not fraud at the site (Judgment at [79]).

76 The fourth area related to non-disclosures regarding errors in the futility determination after the Phase 3 trials. The Judge considered that these errors were immaterial, as Ocedurenone still failed Phase 3 (Judgment at [73]).

77 The fifth area concerned the question of misrepresentation of a charge granted by KBP in favour of DBS Bank Ltd over a US$218m fixed deposit (“DBS charge”). The Judge found that there was no material misrepresentation regarding the DBS charge, even though it subsequently turned out that the information provided by Novo was erroneous, because Novo's representation was based on documents available to it at the ex parte stage (Judgment at [89]).

78 We have summarised these matters briefly and record that we have had the benefit of extensive written submissions from the appellants, which we have carefully considered. We conclude, however, that the Judge was entitled to reach the conclusions that he did. We reject the submission that the Judge reached his conclusions on the basis that he had prematurely made findings of dishonesty: his Judgment does not support this submission. Furthermore, the appellants do not contend that the Judge misunderstood material facts or took account of irrelevant facts or left out of account relevant facts. It is not suggested that there was any procedural irregularity. In several respects their case is that they dispute the Judge’s assessment of what was material and had to be disclosed.

79 In those circumstances, for this appeal to succeed on the ground of non-disclosure, the appellants must show that the Judge’s conclusions were wrong or unreasonable. We do not consider that this has been shown. For example, the appellants (as the second area of non-disclosure summarised above) contend that on the ex parte application there was a failure to disclose their case that efficacy for Ocedurenone had been shown at other sites. The Judge did not consider this allegation to be material. As seen elsewhere in his Judgment, the Judge found that Novo’s case was about the non-disclosure of the outlier results at the Bulgarian site. In his judgment, a critical point was that the Bulgarian results suggested inconsistency in the global results and thus (as Otsuka had pointed out) raised questions of repeatability. In our judgment, his conclusion was open to him. Accordingly, the appeal on this ground must fail.

80 Section 12A of the IAA provides:

81 This key provision allows Singapore courts to offer a wide range of interim relief to support arbitrations, regardless of their location. Notably, Singapore courts can act without the arbitral tribunal's consent if an order is needed urgently to preserve assets or evidence. There are specific limitations. In particular, s 12A(6) states that the Courts may issue an order only if the arbitral tribunal is currently unable to act effectively, such as when the tribunal has not yet been properly established. That condition has of course to be determined immediately prior to making the order in question and when the Court’s decision is challenged on appeal, the question is whether the Court’s determination as at that time discloses an error of law or fact.

82 The first question is whether the case is one “of urgency” (see s 12A(4)). This is a jurisdictional requirement, and, in our judgment, it reflects the same requirement in domestic cases that applications for worldwide freezing orders should generally not be made except on notice to the defendant (see, eg, Bouvier at [115]). Even a jurisdictional requirement can in some circumstances involve matters of evaluation (as with forum non conveniens). The appellants argue that Novo could not show urgency because they had delayed some five to seven months before filing their application for a worldwide freezing order in February 2025. The appellants contend that Novo should have acted earlier, either in August 2024 after the Phase 2 Interim Analysis or in September 2024 after confronting KBP about anomalies. The Judge did not accept this argument. He held that a delay in filing does not necessarily negate urgency. The events of which, in his judgment, knowledge was critical to the risk of dissipation included knowledge gained around December 2024. In the Judge’s view, the urgency requirement was satisfied, especially as there was a risk of dissipation if prior notice of an application was given (Judgment at [100]).

83 We agree with the Judge’s analysis. Mr Bull’s arguments on appeal follow those made before the Judge. On his argument on delay, we agree with the Judge that delay could not be shown before Novo knew or should have known of the facts on which the risk of dissipation was based, which on his findings was not before December 2024. Delay, however, is not the critical requirement: the question is whether the urgency of the situation justified the application that was made. The decision on urgency in this case involved (among other matters) an evaluation by the Court on the information then available of the risk of dissipation, and in addition the potential limitations on the powers of the emergency arbitrator. In our judgment, for the reasons given below, these matters justified the finding of urgency for the purposes of s 12A(4) of the IAA. Accordingly, we dismiss the appeal on this point.

84 The further arguments on this appeal arise from the application of s 12A(6). As pointed out at [36] above, the Judge held that an emergency arbitrator could not grant the freezing order on an ex parte basis, and the risk of dissipation justified the urgency of the application in this case.

85 Mr Bull submits that an emergency arbitrator could have acted effectively, even if not on an ex parte basis. He further submits that the emergency arbitrator could have granted “initial orders” on short notice (eg, within 24 hours), citing ICC materials and expert evidence. He further submits that relief could have been obtained against Dr Huang as a non-signatory if he was ultimately found to be bound by the arbitration agreement.

86 Furthermore, on Mr Bull’s submission, the assets identified by Novo were not such that notice would risk dissipation. The appellants held property which could not be dissipated quickly, negating the need for urgent ex parte relief.

87 Mr Ong’s submission is that the Court had proper power as the tribunal and emergency arbitrator could not act “effectively” within the meaning of s 12A due to the inability of the emergency arbitrator to grant ex parte relief and bind third parties (eg, banks), and because Dr Huang was not formally a signatory to the arbitration agreement. The Judge acted properly in finding that obtaining effective interim relief from the emergency arbitrator was not feasible; this justified court involvement. He submits that the underlying procedural regime and rules (including the ICC rules and the APA’s contractual terms) do not preclude national courts from offering interim relief. Mr Ong relies on factors such as the appropriateness of the Singapore court as a forum and the availability and historical precedent for similar relief in comparable circumstances.

88 In our judgment, Mr Ong is correct in law in his submission that s 12A on its true construction may be invoked where a party to an arbitration agreement applies to the Singapore courts for interim relief against a non-party to the arbitration to preserve assets pending resolution of the disputes agreed to be submitted to arbitration and the non-party will not clearly be bound by any order made by the emergency arbitrator. That was the position in this case. Section 12A also applies where interim relief is sought against a party to the arbitration agreement without giving any prior notice to the party because that party may then act to defeat the purpose of the relief. That too was the position in this case with regard at least to the US$218m held on fixed deposit in relation to the DBS charge.

89 Accordingly, in our judgment, the Judge did not err in holding that he had jurisdiction to make orders under s 12A of the IAA in this case. It was additionally a matter for his discretion whether to make the orders sought. His exercise of discretion falls to be considered on the information available to him at the date of the relevant orders. The appellants have not shown that the Judge’s exercise of his discretion was plainly wrong and open to review on this appeal. So, their appeal on this ground must also be dismissed.

90 In their written submissions, the appellants sought a variation of the worldwide freezing order, but this point was not pursued before us. Therefore, no order is made on that application.

91 For the reasons given above, we dismiss this appeal on all grounds.