1 This is a case about the consequences of one’s actions or, perhaps more accurately, one’s failure to act.

2 A peremptory, or “unless”, order has been described as one of the most powerful tools in the court’s arsenal to ensure compliance with its orders and maintain procedural discipline. Its premise is simple: a party is ordered to do an act, and “unless” the order is complied with and the act done as ordered, a prescribed adverse consequence is to automatically follow. It is no surprise, therefore, that the unless order has been referred to as an “order of last resort” and represents a party’s “last chance to put his case in order” (see Hytec Information Systems Ltd v Coventry City Council [1997] 1 WLR 1666 (“Hytec”) at 1674).

3 It should seem obvious, as night follows day, that if a party fails to make use of his “last chance”, the guillotine should drop and the prescribed consequence(s) ought to follow. For one, the court should not be seen as prone to making idle threats; an ultimatum is hardly an ultimatum if it is not ultimate. But it is contended in this appeal that this intuitive, if not reflexive, flow of events in fact does not flow.

4 Although the imposition and breach of unless orders are not uncommon occurrences per se in civil litigation, the present case is situated in the somewhat unusual context of curial proceedings to enforce a foreign arbitral award. A putative award creditor in foreign arbitral proceedings commences proceedings in Singapore to enforce the arbitral award it has obtained in its favour, intending to leverage on the policy of no-frills enforcement embodied by the Convention on the Recognition and Enforcement of Foreign Arbitral Awards Concluded at New York on 10th June 1958 (the “New York Convention”), which is given the force of law in Singapore as the Second Schedule of the International Arbitration Act 1994 (2020 Rev Ed) (the “IAA”). The award creditor obtains permission to enforce the award on an ex parte basis. The award debtor, however, claims to know nothing of the arbitration and intends to resist enforcement of the award. To this end, it takes out an application for specific production of documents, seeking various documents relating to the arbitration and the parties’ underlying dispute, pursuant to O 11 r 3 of the Rules of Court 2021 (“ROC 2021”). The court below grants the application and makes an order for the award creditor to produce specified categories of documents. After the award creditor fails to comply with its production obligations, an unless order is made against it, specifying, as the consequences of non-compliance, the dismissal of the enforcement proceedings and the rescission of the earlier grant of permission for the enforcement of the award. The award debtor alleges that the award creditor remains recalcitrant in its non-compliance. The broad issues that arose in the court below and in the appeal before me are twofold:

5 The award creditor prays that both questions be answered in the negative. It resists the award debtor’s attempt to enforce the unless order on various grounds, ranging from the unless order having (in actuality) been complied with to the court not having the power to enforce the unless order (even if it has been breached) in the face of the New York Convention. Above all, the catchword “proportionality” is the lynchpin of the award creditor’s position: it argues that it is disproportionate for the unless order to take effect, and for the enforcement proceedings to be stopped in its tracks.

6 In the court below, in his detailed grounds of decision reported at Wuhu Ruyi Xinbo Investment Partnership (Ltd Partnership) v Shandong Ruyi Technology Group Co Ltd and another [2024] SGHCR 7 (the “GD”), the learned assistant registrar Perry Peh (“the AR”) – who had also made the order for production of documents and imposed the unless order – found that the unless order had been breached and, more importantly, that it should take effect. The disgruntled award creditor appeals against the AR’s decision.

7 This appeal presents a useful opportunity to revisit and consider important aspects of the document production regime and the court’s jurisdiction on unless orders, particularly under the ROC 2021. Having considered the parties’ submissions carefully, I dismiss the appeal for the reasons explained in this judgment. My findings, in summary, are as follows:

(a) First, the unless order had been breached as the claimant did fail to comply with its production obligations.

(b) Second, it is not disproportionate for the stated consequences of the dismissal of the enforcement proceedings to kick in.

(c) Third, neither the New York Convention nor the policy of minimal curial intervention in arbitral proceedings deprives the court of the power to enforce the unless order in this case.

8 I start with the factual background underlying this appeal. The background to the dispute between the parties and the procedural history of this matter are somewhat complicated and warrants being set out in some detail.

9 The underlying dispute centres around the parties’ rights to shares in a company incorporated in Luxembourg, SMCP SA (“SMCP” and the “SMCP Shares”) held by the second respondent, European Topsoho S.àr.l. (“ETS”). The claimant, Wuhu Ruyi Xinbo Investment Partnership (Ltd Partnership) (“Xinbo”), claims that pursuant to a tripartite agreement (the “Guarantee”), ETS agreed to pledge some 40m shares in SMCP (comprising all of ETS’s shares in SMCP) to Xinbo as security for a debt owed by the first respondent, Shandong Ruyi Technology Group Co, Ltd (“Ruyi”), to Xinbo. Xinbo is the ultimate parent company of ETS. ¹ The Guarantee was dated 25 July 2018 and was signed on behalf of ETS by one Ms Qiu Chenran (“QCR”), who was, at that time, one of ETS’s managers. ²

10 The gravamen of Xinbo’s complaint is that, unbeknownst to it, ETS had double-pledged some 28m of the SMCP Shares (the “Pledged Shares”) by offering it as security to back an issuance of bonds (the “Bonds”) by ETS to bondholders. Subsequently, ETS defaulted on the Bonds, and the trustee for the bondholders, GLAS SAS (London Branch) (the “Trustee”), took possession of the Pledged Shares sometime on or around 27 October 2021. ³ After Xinbo learned of the Bonds and the Pledged Shares, it issued notices dated 8 and 9 October 2021 seeking the transfer of ETS’s remaining shares in SMCP – some 12m shares (the “Remaining Shares”) – to its nominee for the purpose of safeguarding Xinbo’s rights to the Remaining Shares. ⁴ The transfer of the SMCP Shares to Xinbo’s nominee was completed on or around 27 October 2021, ⁵ pursuant to an earlier share sale agreement dated 22 October 2021. ⁶

11 The Trustee subsequently commenced proceedings in the English courts against ETS and other persons. It obtained summary judgment on 17 October 2022 in respect of the debt owed by ETS under the Bonds, and thereby became a judgment creditor of ETS. ⁷ It has also challenged the transfer of the Remaining Shares ⁸ on the ground, inter alia, that it constitutes a transaction in fraud of ETS’s creditors under s 423 of the Insolvency Act 1986 (c 45) (UK). It appears that, at present, the Trustee had applied for and been refused summary judgment, and the matter is to proceed to trial before the English High Court (see GLAS SAS (London Branch) v European Topsoho SARL and others [2024] EWHC 83 (Comm)). Nonetheless, the previous and ongoing proceedings before the English courts are not of any particular relevance to the present matter.

12 As part of the Trustee’s enforcement action, it filed a bankruptcy petition against ETS in the Luxembourg courts in October 2021. A bankruptcy order was subsequently made against ETS in February 2023, the result of which was the appointment of a curator (the “Curator”) to take over control of and conduct investigations into ETS’s affairs. ⁹ I am given to understand that the functions of a Curator are similar to those of a trustee in bankruptcy or a liquidator under Singapore law. ¹⁰ In all proceedings before the Singapore courts, ETS has been acting under the control of the Curator.

13 On 18 November 2022, Xinbo commenced arbitration proceedings that resulted in the award that forms the subject-matter of its present enforcement proceedings in Singapore (the “Arbitration”). According to Xinbo, the Arbitration was commenced for the purpose of addressing its rights to the Remaining Shares under the Guarantee. ¹¹

14 The underlying arbitration agreement in the Guarantee had, initially, provided for the parties’ disputes to be resolved by arbitration under the auspices of the “Jining Arbitration Commission” (the “JAC”) in the People’s Republic of China. ¹² Article V of the Guarantee provided as follows: ¹³

15 However, the Arbitration was instead conducted under the auspices of the “Beihai Court of International Arbitration” (“BCIA”). As the AR noted, there is a curious inconsistency in Xinbo’s position as to when Art V of the Guarantee was varied to change the arbitral institution from the JAC to the BCIA (see GD at [5] and [107]).

16 Initially, in Xinbo’s supporting affidavit for its application for permission to enforce the award in HC/OA 222/2023 (“OA 222”), Xinbo had stated that the parties to the Guarantee – ie, Xinbo, Ruyi and ETS – had agreed to change the identified arbitral institution to the BCIA pursuant to a memorandum signed in or around June 2019 (the “Memorandum”), due to apparent concerns over the JAC’s ability to hear an arbitration over the parties’ dispute as it was a foreign-related arbitration. ¹⁴ Like the Guarantee, the Memorandum was signed on behalf of ETS by QCR. A copy of the Memorandum (undated) was exhibited to the supporting affidavit, and cl 1 of the Memorandum stipulated a variation to Art V of the Guarantee as follows: ¹⁵

17 However, in a subsequent affidavit filed by Xinbo’s representative, Ms Zhang Yu (“Zhang”), to comply with the Unless Order, Xinbo stated that the parties had agreed to the change of the arbitral institution from the JAC to the BCIA at an in-person meeting held on 9 April 2022 (the “9 April Meeting”). ¹⁶ According to Zhang, the attendees at the 9 April Meeting included:

At this juncture, it suffices for me to highlight that the fact that the variation was discussed and agreed on at an in-person meeting is an important part of Xinbo’s case that it has complied with the Production Order and the Unless Order.

18 Leaving aside the ambiguity as to the time when the agreement to change the arbitral institution occurred, the Arbitration was conducted at a private hearing on 30 December 2022, and the tribunal handed down its award on 10 January 2023 (the “Award”). It was recorded in the Award that the following persons had appeared before the tribunal at the hearing: ¹⁷

(a) Wei, as the lawyer acting for Xinbo;

(b) Ma, as the lawyer acting for Ruyi; and

(c) Mr He Hanchu (“HHC”), as the lawyer acting for ETS.

19 In the Award, the tribunal affirmed the validity of the debt owed by Ruyi to Xinbo as well as the Guarantee. It adjudged Xinbo to have rights to the Remaining Shares under a pledge of the SMCP Shares by ETS to Xinbo under the Guarantee, and held that Xinbo had a “priority right of compensation” out of the proceeds of sale of the Remaining Shares. ¹⁸ Notably, it was recorded at various parts of the Award that (despite being represented by HHC at the hearing) ETS had no objection to the evidence submitted by Xinbo, as well as the reliefs claimed by Xinbo. ¹⁹

20 In OA 222, Xinbo commenced proceedings in Singapore seeking permission to enforce the Award. The application was heard on an ex parte basis and permission was granted in HC/ORC 1189/2023 (“ORC 1189”).

21 Subsequently, ETS applied in HC/SUM 952/2023 (“SUM 952”) seeking to set aside ORC 1189. In an affidavit filed in support of SUM 952, the Curator alleged various irregularities relating to the Arbitration and the Award that supplied grounds for ORC 1189 to be set aside. Among other things, the Curator stated that:

(a) the arbitration agreement in Art V of the Guarantee was invalid and unenforceable as QCR did not have authority under ETS’s articles of association to enter into either the Guarantee or the Memorandum; ²⁰

(b) the Award was invalid as there was no dispute between the parties that the tribunal had jurisdiction to adjudicate upon; ²¹

(c) HHC had no authority to act on ETS’s behalf in the Arbitration; ²²

(d) the Award was procured by fraud as evidenced inter alia by the BCIA being (allegedly) a fictitious and non-existent arbitral institution; ²³ and

(e) enforcement of the Award would be contrary to the public policy of Singapore. ²⁴

22 In a nutshell, ETS’s case in SUM 952 was that the Arbitration and the Award had been concocted for the purpose of allowing Xinbo to steal a march ahead of ETS’s other creditors in ETS’s liquidation by giving Xinbo proprietary rights to the Remaining Shares. A common thread underlying the Curator’s assertions was that prior to being notified by Xinbo, the Curator had not had any knowledge of the existence of the Arbitration and the Award, and the Curator had also not been able to, in the course of its investigations into ETS’s affairs, uncover any internal documents relating to the parties’ alleged dispute under the Guarantee, the parties’ entry into the Memorandum, as well as the Arbitration itself (see GD at [6]). ²⁵

23 To fill the gaps in the Curator’s knowledge about the Arbitration and the Award, ETS applied in HC/SUM 2987/2023 (“SUM 2987”) for Xinbo to produce eight categories of documents which it claimed were material to the determination of SUM 952.

24 SUM 2987 was heard by the AR and he granted ETS’s application in respect of seven of the eight categories of documents requested (see DFD v DFE and another [2024] SGHCR 4 at [2]). Xinbo did not appeal against the AR’s decision. In the Production Order (ie, HC/ORC 5751/2023), Xinbo was ordered inter alia:

25 Generally, the Production Order required Xinbo to produce documents relating to (a) the parties’ dispute under the Guarantee; (b) the parties’ entry into the Memorandum; and (c) the Arbitration. More specifically, the following categories of documents that Xinbo was ordered to produce via the Production Order are relevant for present purposes (see GD at [12]):

(a) Category 1: Communications between Xinbo (which includes Xinbo’s representatives and lawyers) and ETS and/or any of its representatives (including QCR and QYF) regarding the need for and negotiation and execution of the Memorandum, including communications relating to the purported deficiency of the JAC and the circumstances which necessitated a variation of the agreed arbitral institution to the BCIA.

(b) Category 2: Communications from Xinbo (which includes Xinbo’s representatives and lawyers) to ETS and/or any of its representatives (including QCR, QYF, HHC or any other person from HHC’s law firm) in respect of each of the following time periods and subject matter:

(i) Category 2(a): prior to the commencement of the Arbitration, alleging a dispute and demanding reliefs from ETS and/or its representatives, whether in the form of a demand letter or otherwise.

(ii) Category 2(b): after the commencement of the Arbitration, notifying ETS and/or its purported representatives of the commencement of the Arbitration; and

(iii) Category 2(c): after the commencement of the Arbitration, in relation to any matter arising out of or in connection with the Arbitration, including but not limited to matters such as the terms of reference, list of issues, administrative matters and logistics of the hearing(s) for the Arbitration.

(c) Category 3: Communications in respect of any matter arising out of or in connection with the Arbitration from Xinbo to the BCIA.

26 On 11 December 2023, Xinbo filed a list of documents (the “LOD”), which ETS contended (in correspondence) was incomplete. Subsequently, on 30 January 2024, and without being prompted by any order or direction of the court, Xinbo filed a supplementary list of documents (the “SLOD”), which ETS similarly contended (again, in correspondence) was incomplete.

27 Dissatisfied with Xinbo’s attempts at complying with the Production Order, ETS filed HC/SUM 346/2024 (“SUM 346”) seeking an unless order to secure Xinbo’s full compliance with the Production Order, failing which ORC 1189 – ie, the order granting permission for the enforcement of the Award (see [20] above) – be set aside and OA 222 be dismissed.

28 SUM 346 was also heard by the AR, who granted the application and made the Unless Order (viz, HC/ORC 1035/2024). Xinbo did not appeal against the imposition of the Unless Order. Given that the present case centres around Xinbo’s alleged compliance (or non-compliance) with the Unless Order, it is useful to set out the relevant parts in full:

29 In apparent compliance with the Unless Order, Xinbo filed a second supplementary list of documents (the “2SLOD”), as well as an affidavit deposed by Zhang dated 14 March 2024 (the “14 March Affidavit”) in which she set out explanations for how Xinbo had complied with the Production Order and the Unless Order.

30 In the 14 March Affidavit, Zhang stated, inter alia, that two of the documents produced in the 2SLOD were responsive to Category 2(c) of the Production Order, whereas another document was responsive to Category 3 of the Production Order. The other documents produced in the 2SLOD did not come within any of the categories of the Production Order. ²⁶

31 Further, Zhang averred that, given the documents disclosed and produced in the LOD, SLOD and 2SLOD, “[Xinbo] has complied with the Production Order by producing all documents in its possession or control that are responsive to the categories of documents covered by the Production Order”. ²⁷ In other words, Xinbo took the position that it had fully complied with the Unless Order (insofar as it was imposed to secure full compliance with the Production Order).

32 Zhang then went on to elaborate why Xinbo had no further documents in its possession or control responsive to any of the categories in the Production Order. Her explanations have been helpfully set out in detail by the AR in the GD (see GD at [14]–[26]), but for ease of reference, I summarise them here.

33 Category 1 related to communications between Xinbo and ETS regarding the execution of the Memorandum. No documents responsive to Category 1 have been disclosed by Xinbo. ²⁸ Zhang explained that:

(a) The change of the arbitral institution had been discussed and agreed upon in-person at the 9 April Meeting. At that meeting, Ma (ie, Ruyi’s lawyer in the Arbitration) had raised the inadequacies of the JAC and suggested a change to the BCIA. The other attendees agreed. Following the meeting, Wang drafted the Memorandum, which was later executed in-person by the parties’ representatives on 11 April 2022. ²⁹

(b) The main mode of communication between the parties was through physical meetings, telephone calls or WeChat messages. Any WeChat messages that might have been responsive to Category 1 were no longer in Xinbo’s possession or control as “any relevant WeChat records of [Xinbo’s] representatives” – specifically, Wang – had been “deleted due to a change of mobile phones”. ³⁰

34 Category 2(a) concerned communications between Xinbo and ETS in respect of the parties’ dispute under the Guarantee before the commencement of the Arbitration. Although the AR recorded that Xinbo had not disclosed any documents responsive to Category 2(a) (see GD at [13]), I note that this may strictly be inaccurate as Zhang explained that the two notices sent by Xinbo to QCR and ETS on 8 and 9 October 2021 for the transfer of the Remaining Shares, which were disclosed in the LOD, were documents responsive to Category 2(a). ³¹ In fairness to the AR, his impression appears to have been engendered by a subsequent statement by Zhang – admittedly unequivocal on its face – that “[n]o documents have been disclosed by [Xinbo] under category 2(a) of the Production Order”. ³²

35 Be that as it may, Xinbo’s position seems to be that, those two notices aside, it had not disclosed any other documents responsive to Category 2(a) because no such documents remained in its possession or control. In this regard, Zhang explained that:

(a) The parties’ discussions on Xinbo’s interest in the Remaining Shares had taken place at in-person meetings, ³³ including at the 9 April Meeting. At the 9 April Meeting, the parties apparently had differing views on the effect of the terms of the Guarantee, specifically, on whether the Guarantee granted Xinbo any priority right to the Remaining Shares and/or their proceeds of sale. It was thus proposed by Wang that the Arbitration be initiated to determine Xinbo’s rights to the Remaining Shares under the Guarantee. ³⁴

(b) Any WeChat messages between Xinbo and ETS responsive to Category 2(a) had been lost, and were thus no longer in Xinbo’s possession or control, due to Xinbo’s representatives having changed their mobile phones. ³⁵

36 Category 2(b) encompassed communications between Xinbo and ETS after the commencement of the Arbitration, and specifically, those involving Xinbo notifying ETS of the commencement of the Arbitration. Xinbo took the position that it had complied with Category 2(b) by disclosing an e-mail from the BCIA to ETS communicating the commencement of the Arbitration in the SLOD. No further documents responsive to Category 2(b) were disclosed as there was no (and no need for) separate communication between Xinbo and ETS on this matter. ³⁶

37 Category 2(c) involved communications between Xinbo and ETS after the commencement of the Arbitration, and specifically, in connection with matters arising out of or in connection with the Arbitration. Xinbo’s position appears to be that it had complied with Category 2(c).

38 Zhang explained that Xinbo itself had no documents responsive to Category 2(c) in its possession or control given the parties’ main modes of communication (ie, physical meetings, telephone calls and WeChat messages) and the loss of the WeChat messages due to the change of mobile phones of Xinbo’s representatives. However, Xinbo had gone the extra mile by reaching out to one Mr Jing Lin (“Jing”), a secretary of the BCIA who handled administrative matters relating to the Arbitration, and who was able to provide some WeChat messages exchanged between (a) Jing and HHC (purportedly ETS’s lawyer in the Arbitration) from 7 December 2022 to 9 December 2022; and (b) Jing and one Zhang Tongtong (another secretary of the BCIA who handled the Arbitration). These messages that Xinbo obtained from Jing were disclosed in the 2SLOD. ³⁷

39 In SUM 346, ETS suggested that Xinbo had failed to comply with the Production Order by pointing to an apparent inconsistency between three documents in the SLOD: (a) an e-mail sent by the BCIA to ETS where the former had referred to ETS’s application for a simplified procedure; (b) a subsequent e-mail sent by the BCIA to all the parties referring to a joint application by the parties for a simplified procedure; and (c) an excerpt from the transcript of the hearing of the Arbitration which also referred to a “joint request” for a sole arbitrator. The Curator argued that the fact that a unilateral application had morphed into a joint application suggested that there had to be underlying communications between the parties preceding the conversion of the application into a joint one (see GD at [23]).

40 In her 14 March Affidavit, Zhang sought to explain away this inconsistency on the basis that the reference in the first e-mail to ETS’s application was infelicitous, as the application had all along been a joint one. Further, any discussions between the parties on the procedure for the Arbitration had been through their aforementioned main modes of communication, and any WeChat messages that existed had since been lost due to the change of mobile phones of Xinbo’s representatives. ³⁸

41 Category 3 concerned communications from Xinbo to the BCIA in respect of any matter arising out of or in connection with the Arbitration. Xinbo’s position appears to be that it had complied with Category 3.

42 Zhang averred that, in general, no documents responsive to Category 3 existed because, according to Wei (ie, Xinbo’s lawyer in the Arbitration), the parties had contacted the BCIA through its designated telephone line as this was the BCIA’s common practice. Wei no longer retained telephone records detailing those calls due to the passage of time. Further, Wei had advised that no written correspondence was exchanged between Xinbo and the BCIA because (a) the issues in the Arbitration were not particularly complex; and (b) when necessary, apart from discussions over the telephone, Wei had attended in person at the BCIA to deal with matters concerning the Arbitration, such as when he tendered evidence for the Arbitration. ³⁹

43 In SUM 346, ETS suggested that Xinbo failed to comply with the Production Order in respect of Category 3 by relying on an e-mail sent by the BCIA to Wang on 10 January 2023 seeking clarification on two questions relating to the Arbitration (the “10 January E-mail”). This e-mail was disclosed in the SLOD. Since the next e-mail in time was the e-mail from the BCIA notifying the parties of the outcome of the Arbitration, ETS argued that there must have been a prior e-mail to this where Xinbo had responded to the 10 January E-mail.

44 In her 14 March Affidavit, Zhang stated that Wei had advised that the 10 January E-mail had simply been intended to put on record matters that had previously been addressed at the hearing of the Arbitration. In any event, Wang had forwarded the e-mail to Wei, who subsequently contacted Zhang via the BCIA’s telephone line to confirm the same on 10 January 2023. Further, Jing had also liaised with Liang (Xinbo’s legal advisor) on WeChat on 10 January 2023 on the same issues in the 10 January E-mail, and Xinbo had disclosed these messages (which it obtained from Jing) in the 2SLOD. Apart from these messages, Xinbo did not have any further documents responsive to Category 3 in its possession or control. ⁴⁰

45 After the time for ETS to comply with the Unless Order had passed, various disagreements arose between the parties as to whether Xinbo had or had not complied with the Unless Order. Among other things, ETS took the position that the 14 March Affidavit – which was initially filed under cover of a solicitors’ affidavit – had been filed out of time, ⁴¹ which Xinbo disputed. ⁴² Other objections to the adequacy of Xinbo’s purported compliance with the Unless Order were also raised by ETS ⁴³ and refuted by Xinbo. ⁴⁴

46 In these circumstances, the AR gave directions for ETS to take up an application if it was of the view that Xinbo had breached the Unless Order. Pursuant to this direction, ETS commenced HC/SUM 643/2024 (“SUM 643”) in which it sought inter alia an order that the consequences in the Unless Order take effect. The appeal now before me in HC/RA 90/2024 is Xinbo’s appeal against the AR’s decision in SUM 643.

47 The AR held that Xinbo had breached the Unless Order and that it was appropriate for the stated sanctions in the Unless Order to take effect. Although the appeal against the AR’s decision is by way of rehearing based on the documents filed before the AR (see O 18 r 25(4) of the ROC 2021), I would record my appreciation for the assistance that I have derived from the AR’s detailed and extensive coverage of the issues in the GD.

48 The AR’s key findings on the issue of breach can be summarised as follows. First, he noted that it was not entirely clear that Xinbo’s position in relation to Categories 1, 2(a) and 2(c) was that the only form of communication between the parties had been through WeChat messages, such that no other forms of written communications (which the AR referred to collectively as “Other Written Communications”) existed (see GD at [43(a)]). However, it appeared that Xinbo’s position as gleaned from its affidavits was that the Other Written Communications did not exist given Zhang’s statements in her 14 March Affidavit that Xinbo had complied with the Production Order as apart from the documents disclosed in the LOD, SLOD and 2SLOD, no other documents responsive to the Production Order remained in Xinbo’s possession or control (see GD at [44]). Although the AR considered Xinbo’s failure to expressly state its position on the existence of the Other Written Communications to be odd, and also found some of Xinbo’s explanations questionable, he ultimately held that there was insufficient basis to conclude that there was either a reasonable suspicion, or that it was plain or obvious, that the Other Written Communications existed and were in Xinbo’s possession or control, such that Xinbo had breached the Production Order (and, in turn, the Unless Order) by failing to disclose them (see GD at [53]).

49 Second, the AR found that Xinbo had breached the Production Order in respect of Category 2(c). Although Xinbo had taken the position that it did not have possession or control of any WeChat messages between its representatives and ETS as its representatives had lost their WeChat messages after a change of mobile phones, the AR noted that Xinbo had been able to produce in the 2SLOD a screenshot of WeChat messages exchanged between HHC (purportedly ETS’s lawyer in the Arbitration) and Liang (Xinbo’s legal advisor) dated 30 December 2022 (the “30 December Message”), which Xinbo stated it had obtained from HHC and was responsive to Category 2(c) (see GD at [60]). Xinbo’s position was not that this 30 December Message was the only document that existed and which was responsive to Category 2(c), but that it was only able to produce the 30 December Message as Liang had changed his phone and thereby lost the other WeChat messages he had exchanged with HHC (see GD at [61]). As Xinbo did not explain how and why HHC had provided the 30 December Message, and also did not elaborate on whether the 30 December Message was the only message that HHC had been able to provide to it, the AR held that other WeChat messages between HHC and Liang existed and were in Xinbo’s control (as Xinbo could procure HHC’s assistance to provide them) (see GD at [63]). This either gave rise to a reasonable suspicion or made it plain and obvious that Xinbo had failed to produce these other WeChat messages responsive to Category 2(c), and had therefore failed to comply with the Production Order (see GD at [64]).

50 Third, the AR accepted Xinbo’s averment that it had no further documents responsive to Categories 2(b) and 3 on the basis that, respectively, Xinbo had not communicated separately with ETS about the commencement of the Arbitration and Zhang’s explanations in the 14 March Affidavit – as confirmed by Wei in a subsequent affidavit of his dated 26 March 2024 (the “26 March Affidavit”) – that no written communications were exchanged between Xinbo and the BCIA as Wei had either spoken to the BCIA over the telephone or attended in person when necessary (see GD at [67]–[68]).

51 Fourth, the AR held that Xinbo had breached the Production Order by failing to provide an adequate explanation of how it had lost possession or control of the WeChat messages. By failing to adequately explain its loss of possession or control of the WeChat messages, the court could not accept Xinbo’s bald assertions on affidavit that it did not have possession or control of the same, and in turn, Xinbo breached Categories 1, 2(a) and 2(c) insofar as the WeChat messages responsive to these categories were concerned (see GD at [81]). The Production Order required Xinbo to explain the circumstances in which it had lost possession or control as opposed to merely stating that it had lost possession or control (see GD at [73]). In this regard, Xinbo’s explanations were woefully inadequate:

(a) Although Zhang had stated that Xinbo’s representatives were unable to access their past WeChat messages after changing their mobile phones, Zhang had not stated that the past WeChat messages could no longer be accessed on these representatives’ old phones or that these representatives no longer had access to their old phones. Xinbo’s explanation was thus incomplete insofar as it only covered its representatives’ inability to access the WeChat messages on their current (new) phones (see GD at [75]).

(b) There appeared to be other persons from whom Xinbo could obtain the WeChat messages, but Xinbo had failed to account for these other persons and its attempts (if any) to obtain the WeChat messages from them. Instead, Xinbo had simply made a global statement that it had fully complied with the Production Order (see GD at [76]). For example, in respect of the parties’ dispute under the Guarantee and their discussion on the Memorandum, persons other than Wang and Liang had been in attendance at the 9 April Meeting from whom Xinbo could have obtained the WeChat messages exchanged between the parties. However, Xinbo had not stated that Wang and Liang were the only persons who had exchanged WeChat messages, and did not address their attempts to obtain the messages from these other persons (see GD at [77]–[79]).

(c) Although Xinbo had stated how Wang and Liang had lost possession of the WeChat messages (ie, through a change in their mobile phones), it had not stated when this had occurred. This was an important detail that Xinbo had to provide to comply with the Production Order as it corresponded to the time when Xinbo had lost possession or control of the WeChat messages (see GD at [80]).

52 Fifth, the AR held that, separate from the issue of production of documents, Xinbo had breached its obligation to allow ETS to inspect the documents that Xinbo had produced under the LOD, SLOD and 2SLOD. Instead of providing ETS the opportunity to inspect the originals of these documents, Xinbo had simply provided printed hardcopies of documents that it had already provided in soft copy to ETS (see GD at [85]).

53 The AR’s reasons for holding that the consequences in the Unless Order should take effect can be summarised thus. First, Xinbo’s breach of the Unless Order had clearly been intentional and contumelious. Since it took the position that it had complied with the Production Order and the Unless Order, Xinbo provided no meaningful explanation in SUM 643 as to why it had breached the Production Order and the Unless Order. Further, Xinbo had not made a real attempt at complying with the Production Order and the Unless Order, as evinced by both its woeful attempts at facilitating inspection and the inadequacy of its explanation on the loss of possession or control of the WeChat messages (see GD at [91]–[93]).

54 Second, it was neither contrary to the principle of minimal curial intervention nor the New York Convention to allow enforcement of the Unless Order to disrupt the enforcement of the Award. Minimal curial intervention did not mean that any decision that had the effect of disrupting enforcement of the Award would not be permitted (see GD at [95]). The coming into effect of the Unless Order and resulting disruption to the enforcement of the Award did not entail a challenge or review of the Award on the merits outside of the prescribed grounds for resisting enforcement of international arbitral awards under the New York Convention and s 31 of the IAA (see GD at [96]). The court, in enforcing the Unless Order, was not resisting enforcement of the Award (so as to be constrained by the exhaustive grounds set out in s 31 of the IAA and the New York Convention), and there was nothing in the ROC 2021 that indicated that the court’s case management powers were cut down in a case involving enforcement of an international arbitral award (see GD at [100]–[101]).

55 Third, the fact that enforcement of the Unless Order would give ETS the outcome that it was seeking in SUM 952 did not render the enforcement of the stated sanctions in the Unless Order disproportionate. The Production Order had been made for the purpose of ensuring that the court hearing SUM 952 would be able to properly consider whether enforcement of the Award should be allowed or refused. If it were disproportionate to enforce the Unless Order which was intended to secure compliance with the Production Order, there was a risk that parties seeking the enforcement of foreign arbitral awards would not take the court’s orders seriously (see GD at [102]–[104]).

56 Fourth, the AR did not agree with Xinbo’s argument that its breach of the Production Order did not affect the prospects of a fair hearing of SUM 952 as ETS was nonetheless able to mount its case based on the documents in its possession and those already disclosed. Among other issues in SUM 952, there had been a conspicuous change in Xinbo’s position on when the Memorandum had been signed to shift the arbitral institution from the JAC to the BCIA, as well as ETS’s contention that there had been no genuine dispute between the parties to found the tribunal’s jurisdiction in the Arbitration. In the face of Xinbo’s non-compliance with the Production Order, the court hearing SUM 952 would have no evidence before it other than the parties’ competing assertions. The AR was satisfied that the absence of documents due to Xinbo’s breach of the Production Order would frustrate a fair hearing and determination of SUM 952 (see GD at [105]–[109]).

57 Finally, the AR decided against giving Xinbo one final chance to comply with the Production Order and the Unless Order. Given Xinbo’s unequivocal position that it had fully complied with the Production Order and its failure to provide any explanation for its non-compliance, it appeared rather pointless to ask Xinbo to further comply (see GD at [117]).

58 I summarise the parties’ positions on appeal in broad strokes, which I shall address in greater detail in the course of my analysis below.

59 In the appeal before me, Xinbo noticeably placed much more stock in its arguments on why the stated sanctions in the Unless Order ought not take effect (on the premise that the court was satisfied that it had breached the Unless Order), as opposed to vigorously contesting the antecedent question of whether it had breached the Unless Order. Indeed, in oral submissions, counsel for Xinbo Mr Kelvin Poon SC (“Mr Poon”), spent little time addressing the issue of breach and was content to rest on his written submissions.

60 In any event, Xinbo’s written submissions contesting the AR’s finding of breach were rather sparse. In sum, Xinbo made the broad contention that the AR had misdirected himself in law by conflating the test of “reasonable suspicion” of Xinbo’s statements on affidavit professing compliance with the Production Order being false – “reasonable suspicion” being the applicable test under the former Rules of Court (2014 Rev Ed) (“ROC 2014”) – with the correct test (and the higher threshold) of “plain and obvious” applicable to the ROC 2021. It was argued that, had the AR applied the higher threshold correctly, he would have found that Xinbo had complied with the Production Order as it was not “plain and obvious” that Xinbo had failed to disclose documents responsive to the Production Order that remained in its possession or control. ⁴⁵

61 Turning to the consequences of breach, Xinbo emphasised that the New York Convention embodied a pro-enforcement regime, ⁴⁶ and framed the objectionability of enforcing the sanctions in the Unless Order – ie, the striking out of OA 222 and discharge of ORC 1189 – in two ways:

(a) First, the effect of the AR’s decision was to fashion a new ground for refusing enforcement of a foreign arbitral award beyond the enumerated grounds in the New York Convention and s 31 of the IAA. That, it was argued, was in contravention of the New York Convention. ⁴⁷ It was not open to the Singapore courts to make orders that had the practical effect of refusing enforcement of an award on a ground outside of these exhaustive bases, and in that respect, the New York Convention and IAA imposed constraints on the court’s procedural powers under the ROC 2021. ⁴⁸

(b) Second, it was contrary to the Singapore courts’ pro-enforcement policy in respect of arbitral awards for the court to strike out an enforcement action. Instead, the court should adopt other means to address a defaulting party’s non-compliance with an unless order, as opposed to using striking out as the first port of call. At the hearing of the appeal, in response to my query on what alternative sanctions might be meted out to deal with Xinbo’s non-compliance (on the assumption that I were to find that Xinbo had breached the Unless Order), Mr Poon suggested that the court hearing SUM 952 could draw adverse inferences against Xinbo ⁴⁹ or make appropriate costs orders. ⁵⁰

62 Further, Xinbo argued that allowing the sanctions in the Unless Order to take effect would be disproportionate as (a) it was still possible for SUM 952 to be heard and determined fairly; and (b) it granted ETS its desired outcome in SUM 952 through the backdoor. ⁵¹

63 ETS submitted, in the first instance, that Xinbo had failed to comply with the Unless Order. Among other things, ETS pointed to the inadequacies in Xinbo’s explanation as to how and when the WeChat messages had been lost from its possession or control, as well as the steps that it had taken to obtain and produce the WeChat messages. ETS also emphasised that Xinbo had completely failed to perform its obligation to permit inspection of the original documents. ⁵²

64 As for the consequences of Xinbo’s breach of the Unless Order, ETS aligned itself with the AR’s findings that (a) Xinbo’s breach of the Unless Order was intentional and contumelious; ⁵³ (b) there was no legal impediment against enforcing the sanctions stated in the Unless Order posed by the New York Convention, the IAA or the principle of minimal curial intervention; ⁵⁴ and (c) the enforcement of the sanctions was proportionate to Xinbo’s breach. ⁵⁵

65 The two main issues before me in this appeal are:

For Xinbo’s benefit, in the main, I shall focus on the findings that the AR made against it.

66 I start with the question of whether Xinbo breached the Unless Order. In my judgment, the answer is an unequivocal “yes”. I begin with a discussion on the applicable legal principles, before turning to elaborate on Xinbo’s non-compliance with the Unless Order.

67 The main legal issue in this appeal arises out of Xinbo’s contention that the AR had applied the wrong legal test when assessing Xinbo’s statements on affidavit that it had complied with the Production Order. To set the groundwork for resolving the issues arising in this case, including the question as to whether the AR had erred in law as Xinbo claims, it is necessary to first consider the applicable law relating to affidavits filed in document production applications and in response to document production orders.

68 It is trite that an affidavit relating to the discovery of documents is generally conclusive as to what documents are or were in a party's possession or control, responsive to a party’s discovery obligations, or protected by grounds such as legal professional privilege (see Jeffrey Pinsler SC, Singapore Civil Practice (LexisNexis, 2022) (“Singapore Civil Practice”) at para 30-76; Hodge M Malek & Paul Matthews, Disclosure (Sweet & Maxwell, 6th Ed, 2023) (“Disclosure”) at para 6-79). This principle has typically been articulated in caselaw (some of which I refer to in the following paragraphs) in the specific context of applications for discovery of documents under the ROC 2014 regime (or its English equivalent). Nevertheless, there is, in my view, no reason why the underlying principle as to the conclusiveness of affidavits should be any different under the current regime for production of documents under O 11 of the ROC 2021.

69 The principle as to the general conclusiveness of affidavits has been articulated in several authorities, including South Staffordshire Tramways Company v Ebbsmith (1895) 2 QB 669, where Lord Esher MR made the following reference to it (at 674):

70 However, it is equally well-established that the court is not invariably bound by a statement on affidavit, if it is satisfied that the statement is clearly false or deficient. The leading case for this proposition is the English Court of Appeal decision of Jones v The Monte Video Gas Company (1880) 5 QBD 556. All three members of the court were unanimous on this point. First, Brett LJ (as he then was), delivering the lead judgment, explained that (at 558):

In a similar vein, Cotton LJ framed the rule thus (at 559):

Although the last member of the court, Thesiger LJ, “confess[ed] that [he] felt somewhat apprehensive about drawing a hard and fast rule”, he nonetheless agreed that “unless the party seeking production can surmise from his adversary’s admissions that some documents are being withheld, he must rest content with the statements in the affidavit of discovery” (at 559).

71 The natural question that arises from this is: when can a court look behind the affidavit?

72 Under the ROC 2014, our courts proceeded on the basis of the well-known test of “reasonable suspicion”: the court could look behind the deponent’s affidavit as to the adequacy of its performance of its discovery obligations if the court entertained a “reasonable suspicion” that there were further documents to be discovered, based on the disclosed documents themselves or the surrounding circumstances (see Soh Lup Chee and others v Seow Boon Cheng and another [2002] 1 SLR(R) 604 at [9]; Natixis, Singapore Branch v Lim Oon Kuin and others [2024] 3 SLR 1502 (“Natixis”) at [27]).

73 But the position is different under the current ROC 2021. Earlier this year, in Lutfi Salim bin Talib and another v British and Malayan Trustees Ltd [2024] 5 SLR 86 (“Lutfi”), Chua Lee Ming J examined the authorities and held that the prevailing “reasonable suspicion” test was inapt having regard to the ethos of the ROC 2021. The learned judge considered a higher threshold of “plain and obvious” to be appropriate (at [32]):

While the appeal before me does not strictly concern a specific production application under O 11 r 3 of the ROC 2021, it is not disputed between the parties that Lutfi is applicable to the issue of whether Xinbo had complied with the Production Order (and the Unless Order) by virtue of Xinbo making statements on affidavit that it has complied with its document production obligations under the Production Order.

74 Although the view expressed in Lutfi is not strictly binding on me, I consider, on principle, that Chua J was correct in raising the bar in the context of the ROC 2021. As the learned judge alluded to, the ROC 2021 was intended to usher in a new approach to the disclosure and production of documents – specifically, a narrower regime – than the discovery regime under the former ROC 2014 (see Lutfi at [34], citing Civil Justice Commission, Civil Justice Commission Report (29 December 2017) (Chairperson: Justice Tay Yong Kwang) at p 19). Indeed, apart from the specific issue of the conclusiveness of affidavits filed in fulfilment of a party’s disclosure and production obligations, the paradigm shift is apparent from the move away from the previous test of “relevance-necessity” for specific discovery under O 24 of the ROC 2014 (see, for example, The Management Corporation Strata Title Plan No 689 v DTZ Debenham Tie Leung (SEA) Pte Ltd and Another [2008] SGHC 98), to the higher and stricter threshold of “materiality” under O 11 of the ROC 2021 (see Eng’s Wantan Noodle Pte Ltd and another v Eng’s Char Siew Wantan Mee Pte Ltd [2023] SGHCR 17 at [49(b)]; Cachet Multi Strategy Fund SPC on behalf of Cachet Special Opportunities SP v Feng Shi and others [2024] SGHCR 8 (“Cachet”) at [30]–[31] and [33]).

75 At first blush, the heightened threshold in Lutfi may raise concerns of the “plain and obvious” test becoming a charter for parties seeking to suppress or withhold material documents from production (see, for example, Jeffrey Pinsler, “Case Law Developments Concerning Production of Documents” (2024) CLU 5 at paras 7–8). In my view, any such concern is unfounded because the other side of the coin is that a party seeking to avail itself of this stricter standard cannot simply make bald and unsubstantiated assertions that it has complied with its production obligations. The court has the power under O 11 r 3(2) of the ROC 2021 to order a party who claims that certain documents are not in its possession or control to state this, as well as explain whether the documents were previously in its possession or control and if so, what has become of the documents. In the ordinary course, a court making an order for specific production under O 11 r 3 would include this proviso, such that a party either has to produce the requested documents or provide a satisfactory account of them if he does not do so. The effect of this, which is quite material to the present case, is that a party against whom a specific production order is made has two distinct and independent obligations: (a) to produce the document or state on affidavit that it is unable to do so for some reason; and (b) if it is unable to do so, explain why that is so. Both must be complied with (in letter and substance) before a party can be said to have complied with its obligations under the order for production.

76 It follows that the shift to the “plain and obvious” test does not allow a party to respond to a specific production order by simply making bald assertions on affidavit that it has no documents responsive to the order in its possession or control. It may well be that the court is unable to conclude on the evidence before it that it is “plain and obvious” that the assertion is untrue, but that in no way means that that party has complied with the production order, because it would have failed to discharge its secondary obligation to provide a satisfactory account of its inability to produce the document(s) ordered. In a particularly egregious case, the court may find the account given to be so lacking that it is satisfied that it is “plain and obvious” that the assertion cannot be taken as conclusive as it must be untrue. The point of the matter is that the assertion that one has complied with its production obligations by itself only satisfies one half of the party’s obligations. Put a different way, the separate obligation to provide a satisfactory account acts as a counterweight against attempts at abusing the conclusiveness of an affidavit filed in response to a document production application or order. In this way, a balance is struck between the interests of the party seeking the order (who either obtains production of the documents or an account of what has become of the documents) and the party subject to the order (who is not put in the invidious position of producing a document that it really does not have and having to prove a negative).

77 None of this is particularly novel or unique to the document production regime under the ROC 2021. The safeguard of requiring the party asserting that it is unable to produce a document that has been requested or ordered of it to adequately explain the circumstances of its inability has always been present within the court’s arsenal of powers, even under the ROC 2014 (see O 24 r 5(1) of the ROC 2014), and indeed, before it. The court has always reserved the right to satisfy itself as to the adequacy of a party’s explanation(s), independent of making a finding that the party is speaking untruths. An early example can be found in the English Court of Appeal decision of Gardner and another v Irvin and another (1878) 4 Ex D 49 (“Gardner”). The defendants in that case had resisted discovery of certain documents after a discovery order had been made by asserting privilege over the documents. The defendants’ affidavit stated that “one reason for objecting to produce the said documents set forth in the second part of the schedule is that the same are privileged”, before setting out a swath of different types of documents in the second part of the schedule. The plaintiffs contended that the affidavit was deficient as it stated merely that the documents were privileged, when it was prima facie clear that some of the documents in the second part of the schedule – such as the defendants’ cash-book, ledger and accounts – were not privileged. The plaintiffs objected and argued that the defendants had to set out the facts grounding the privilege so that the court could assess the legitimacy of the defendants’ claim to privilege (see Gardner at 50–51).

78 The court agreed with the plaintiffs that the defendants’ affidavit was insufficient. Brett LJ observed that “[t]he defendants ought to verify on oath the facts on which they claim the privilege” (see Gardner at 52). Cotton LJ considered that “[the defendants] ought to say not only that the documents are privileged, which is a statement of law, but they ought to set out the facts from which [the court] can see that [their] view of the law is right” (see Gardner at 53). His Lordship then went on to explain the reason as follows (see Gardner at 53):

79 Gardner is thus an example of a case where the court did not make a finding that the documents were not covered by privilege as the defendants claimed, but nonetheless found that the statements on affidavit were deficient in particularity. It illustrates that these are distinct obligations of a party who is subject to a production order. Indeed, Cotton LJ draws the perceptive link that it is precisely because the court accords a measure of deference to a party’s affidavit and generally treats it as conclusive that it is incumbent on that party to supply sufficient particulars.

80 Similar sentiments have been expressed in the local jurisprudence. For example, in Technigroup Far East Pte Ltd and another v Jaswinderpal Singh s/o Bachint Singh and others [2018] 3 SLR 1391, Steven Chong J (as he then was) said that it was not the position that (at [56]):

In the same vein, I made the observation in Natixis that (at [32]):

81 In my judgment, given the issues that have arisen in this case, it may be useful, without being unduly prescriptive, to lay down some general expectations as to affidavits filed in compliance with a party’s document production obligations.

82 First, it should go without saying that the affidavit should address all documents that the production order relates to. To use the present case as an illustration, the AR noted that Xinbo’s position in relation to the Other Written Communications was ambiguous as the existence and status of the Other Written Communications was not expressly addressed (see [48] above). In my view, it is a basic expectation from the court and the opposing party that no such ambiguities should be left. If a party takes the position that no documents responsive to a particular category that it has been ordered to produce are in its possession or control, it does not suffice to simply focus on one type of document falling within the relevant category while leaving the court and the other party guessing on other types of documents. The affidavit should address the category in its entirety and not merely a subset of it.

83 Second, and following from the first point, addressing a category as a whole does not mean that a party can get by with making a rolled-up or sweeping statement that “no documents responsive to Category X exist and/or are in our possession or control”. The court would expect some particularity in terms of a party’s explanation. If a class of documents comprises different types of documents, it would be prudent to address each type of document specifically (as appropriate) rather than simply referring to the class in its entirety. If necessary, the affidavit could also contain the broad statement as a “sweeping-up” provision to make clear a party’s position; but a blanket statement like that cannot be the only information or “explanation” that is presented to the other party and the court. The Gardner case illustrates this point.

84 Third, an explanation is hardly an explanation if it throws up as many, or worse, more, questions than it answers. Thus, for example, if it is asserted that certain documents do (or did) exist but are no longer in a party’s possession or control, it would be expected that the party sets out the steps that it has taken to verify this, and perhaps, even the steps that it has taken to make a good faith attempt at obtaining the documents. As a general litmus test, if there is an obvious follow-on question to what a party has put on affidavit, that is a fair indication that that question should be proactively addressed, so as to complete the explanation, as opposed to leaving the point hanging. To use another aspect of the present case as an illustration, if Xinbo was taking the position that it did not retain certain WeChat messages due to its representatives having changed their mobile phones, it would have been obvious to any bystander that the immediate follow-on questions to this would have been, amongst others, the individual identification of these representatives, the time(s) when the change(s) in mobile phones occurred, and what became of the old mobile phones.

85 Fourth, the affidavit should not be framed in equivocal terms that allow prevarication. If a party does not produce a document on the basis that it is not in the party’s possession or control, it should commit to that position and cannot ordinarily state that it is “unsure” or “not certain” if the document is in its possession or control. The case of Price v Price (1879) 48 LJ Ch 215 is illustrative of this proposition. After the plaintiff had obtained an order for discovery, the defendant filed an affidavit stating inter alia that “I am not able to say exhaustively that there may not be in my office, or may not have been in the possession of Ebenezer Roberts, deceased, my late clerk, documents relating to the matters in question in this suit, which I have been unable to find” (at 215). Bacon VC held that the affidavit was insufficient on various grounds, including that this clause was “evasive” (at 216). To be clear, the law does not require a party to be omniscient when making statements on affidavit in answer of its production obligations. But what is required is for a party to take all proper means to acquire the documents or information on them (Mertens v Haigh (1863) 46 ER 741 at 742), and to then give all the information in its power (Clinch v Financial Corporation (1866) LR 2 Eq 271 at 273).

86 These four points that I have identified are not matters that require the expenditure of significant mental bandwidth to conceive. On the contrary, these are matters that would (or should) occur to any party who approaches its document production obligations in good faith and a dose of common sense.

87 Indeed, while I have observed that the safeguard of requiring a proper explanation in the affidavit of a party who does not produce a document it has been ordered to produce is one of considerable vintage (see [77] above), there is a fair argument that the courts ought to apply greater scrutiny to explanations under the ROC 2021, as the “plain and obvious” test is more susceptible to abuse than the “reasonable suspicion” test.

88 With this overview of the applicable principles in mind, I turn to the facts of this case to consider the state of Xinbo’s compliance with the Production Order and the Unless Order.

89 As stated at [66] above, it is clear to me that Xinbo has failed to comply with the Production Order and the Unless Order.

90 I can quickly dispose of Xinbo’s suggestion that the AR conflated the tests of “reasonable suspicion” and “plain and obvious”. I flatly disagree. It is clear beyond peradventure that the AR was keenly aware of the difference between the two tests. Indeed, the AR expressly stated that he had initially made his decision based on the “reasonable suspicion” test as Chua J’s decision in Lutfi was handed down on the same day that SUM 643 had been heard and decided. The AR recognised explicitly that it was therefore necessary to consider “whether [his] decision [was] now indefensible given the principles set out by the court in Lutfi” (see GD at [32]). Xinbo’s contention is plainly unsustainable in the face of the AR’s awareness that he had to reassess whether his findings based on the “reasonable suspicion” test were maintainable based on the higher “plain and obvious” standard. Indeed, the AR immediately went on to state that “[he] would have reached the same conclusion in SUM 643 even if [he] had applied the ‘plain and obvious’ test in Lutfi” (see GD at [32]).

91 As I focus only on the findings of the AR which are adverse to Xinbo, my analysis will centre around the following three findings of breach of the Production Order and the Unless Order:

(a) Xinbo breached Categories 1, 2(a) and 2(c) of the Production Order by failing to provide an adequate explanation of how it had lost possession or control of the WeChat messages between Xinbo and ETS (see [51] above).

(b) Xinbo breached Category 2(c) of the Production Order by failing to disclose WeChat messages exchanged between Xinbo and ETS after the commencement of the Arbitration in relation to the matters in the Arbitration. The AR considered it to be “plain and obvious” that such documents exist but were not produced by Xinbo (see [49] above).

(c) Xinbo breached the Production Order by failing to allow ETS to inspect the originals of the documents it had produced (see [52] above).

92 I shall take each of these alleged breaches in turn.

93 Xinbo has taken the firm position that, apart from the WeChat messages that it has been able to obtain from Jing and HHC, it has no further WeChat messages in its possession or control that are responsive to the Production Order. Leaving aside the issue of whether it is “plain and obvious” that this position is untrue, Xinbo was nonetheless required to provide an adequate explanation of its inability to produce the WeChat messages.

94 I accept, in fairness to Xinbo, that the terms of the Production Order did not expressly state that Xinbo was required to provide an explanation for its inability to produce documents that it had been ordered to produce if it intended to assert such inability. The Production Order had only stated that Xinbo was to (a) “file a list of the documents in Categories 1 to 7 of the Schedule to SUM 2987”; (b) “provide [ETS] with copies of these documents”; and (c) allow ETS “to inspect the documents ordered to be produced … within 14 days from the time copies of those documents are produced to [ETS]” [original emphasis omitted].

95 However, I do not think that this carries Xinbo very far. The fact that Zhang had attempted to provide some explanation as to Xinbo’s loss of possession or control of the WeChat messages (by repeatedly pointing to its representatives’ change of their mobile phones) was not fortuitous. On the contrary, it indicates that Xinbo was clearly aware of the need for an explanation if it was not producing (or able to produce) the WeChat messages for whatever reason.

96 Mr Poon did not attempt to argue that Xinbo was in fact not required to provide an explanation, or that the explanations offered by Zhang in the 14 March Affidavit had been purely gratuitous. Such an argument would have been contrary to principle and authority. Although the court’s power to compel an explanation is acknowledged by O 11 r 3(2) of the ROC 2021, it does not follow that a party only has to provide an explanation when O 11 r 3(2) is expressly invoked. I do not understand Chua J’s comments at [19] of Lutfi, referencing the general possibility of specific production orders being made without an accompanying order under O 11 r 3(2) – to have been suggesting this – a party, especially a commercial entity acting under legal advice, should not require an express direction to act in accordance with common sense.

97 It is axiomatic that a party should obey both the letter and the spirit of a court order (see Lee Shieh-Peen Clement and another v Ho Chin Nguang and others [2010] 4 SLR 801 at [17]). Thus, in the context of contempt proceedings, it is well-established that the court focuses on whether a party has frustrated the order’s purpose. In Aurol Anthony Sabastian v Sembcorp Marine Ltd [2013] 2 SLR 246, Sundaresh Menon CJ explained that “the court will not adopt a myopic and blinkered view of the scope of an order”, as “[i]t is ultimately the purpose for which the order was granted that will be the lodestar in guiding the court’s determination as to the true effect of the order” [emphasis in original] (at [99]). Drawing on these observations, I noted in WestBridge Ventures II Investment Holdings v Anupam Mittal [2024] 3 SLR 332, when dealing with the interpretation of an anti-suit injunction, that the distinction between prohibitory and mandatory framing of anti-suit injunctions seemed to me rather facile as “one must not lose sight of their essential nature and purpose or be blinkered by form over substance” (at [71]).

98 All this goes to say that I would not have accepted any attempt at gamesmanship on Xinbo’s part by pleading (or feigning) ignorance as to the scope of its obligation to provide an adequate explanation in order to meaningfully comply with the Production Order. In any event, the position would have been crystal clear after the Production Order was supplemented with the Unless Order, as para 4 of the Unless Order spelt out that Xinbo was to file an affidavit stating:

99 It is against this backdrop that I assess the adequacy of Xinbo’s explanation. In my judgment, the AR was correct to find that Xinbo’s explanation was lacking.

100 Although Zhang made the general assertion at various points of the 14 March Affidavit that Xinbo’s representatives had changed their mobile phones, and some individual references were made to certain representatives specifically (viz, Wang and Liang), there was a general lack of particularisation in the affidavit as to: (a) the status of the WeChat messages in the possession or control of other representatives of Xinbo; (b) the time when Xinbo’s representatives had lost their phones; and (c) how exactly the change of mobile phones caused Xinbo to lose possession or control of the WeChat messages on the old mobile phones.

101 First, Xinbo did not account for all of its representatives that could reasonably be believed to have documents responsive to the Production Order. ⁵⁶

102 I use Xinbo’s explanation in response to Category 1 as an illustration. To recap, Category 1 concerned communications between Xinbo and ETS on the change of the arbitral institution from the JAC to the BCIA and the execution of the Memorandum. In the 14 March Affidavit, Zhang stated that the attendees of the 9 April Meeting where this issue had been discussed in-person by the parties’ representatives included QYF, QCR, Su, Zhang herself, Ma, Wei and Liang. ⁵⁷ Zhang also admitted that there were WeChat messages between the parties that corresponded to Category 1, as WeChat messages were one of the parties’ “main mode of communication” (the other two forms being physical meetings and telephone calls). ⁵⁸ The sum of these two facts, in my view, is that if Xinbo did not produce any WeChat messages responsive to Category 1, its explanation had to, at the very least, account for each of the attendees at the 9 April Meeting whom it had control over. Taking a broad view, this would have encompassed Su, Zhang, Ma, Wei, and Liang, all of whom were representatives of Xinbo in some capacity.

103 However, all that Zhang stated in the 14 March Affidavit was the following: ⁵⁹

104 I make two observations. The first is to emphasise the significance of the fact that the attendees at the 9 April Meeting included Zhang herself. Given this, it is curious that Zhang did not expressly address whether she ever had possession or control of the WeChat messages responsive to Category 1, and instead only stated what she had been “given to understand” from third parties. ⁶⁰

105 The second point is that, despite making the sweeping assertion that “any relevant WeChat records of [Xinbo’s] representatives … are no longer available as such WeChat records have been deleted due to a change of mobile phones”, Zhang went on to only address Wang’s situation specifically. In a different part of the 14 March Affidavit, Zhang also states that Liang had changed his mobile phone and thereby no longer has access to his old WeChat messages – presumably, although it is again not entirely clear, this includes any messages responsive to Category 1. Beyond the 14 March Affidavit, the lacuna was subsequently narrowed slightly when Wei stated his position – that he did not have possession or control of any written communications between Xinbo and ETS on the Arbitration – in the 26 March Affidavit. Even so, as ETS points out, ⁶¹ there is ambiguity as to whether Wei’s reference to the “Arbitration” encompasses the change of the arbitral institution and the Memorandum as opposed to the substantive arbitration, such that arguably even Wei cannot be said to have been accounted for. Although the AR took a rather generous view of the 26 March Affidavit by interpreting “Arbitration” broadly to include these matters, I consider that there is force in ETS’s submission that he may have been too generous. ⁶² But, even leaving that aside, what of the other representatives of Xinbo who are completely not addressed, including Zhang herself?

106 Although Zhang did say that Wang was “handling this change of arbitral institution issue as well as the matters concerning the arbitration”, this was equivocal as to whether Wang was the only representative of Xinbo who was ever in possession or control of communications responsive to Category 1 (see GD at [77]). The fact that Wang was “handling” or put in charge of addressing the change of arbitral institution and the Arbitration on Xinbo’s behalf did not necessarily mean that no other representative of Xinbo had any involvement. Indeed, the fact that four other representatives from Xinbo other than Wang attended the 9 April Meeting presumably means that they had some involvement in the change in arbitral institution and execution of the Memorandum. If Xinbo’s position was that only Wang and Wang alone had exchanged WeChat messages with ETS on these matters, it did not say so. Indeed, as the AR noted (see GD at [76]), Mr Poon conceded at the hearing before the AR that “there are these other individuals involved”, ⁶³ but that he was constrained by what Zhang had set out in the 14 March Affidavit.

107 Second, the details provided by Xinbo as to its loss of possession or control of the WeChat messages were deficient inasmuch as they did not state the time(s) when Xinbo supposedly lost possession or control. At various points of Zhang’s 14 March Affidavit, Xinbo simply asserted that its representatives had lost possession or control after changing their mobile phones, but it did not state when the change in mobile phones (which corresponded to Xinbo’s loss of possession or control) occurred.

108 The time at which a party supposedly lost possession or control of documents that it has been ordered to produce is clearly a material detail. This is acknowledged by O 11 r 3(2) of the ROC 2021, which provides that the court may order a party subject to a specific production order to file an affidavit stating inter alia “when that party parted with possession or control and what has become of the requested documents”. I wholly agree with the AR’s analysis on the need for Xinbo to particularise the time that it had lost possession or control of the WeChat messages, as there were several possibilities: (a) during the Arbitration; (b) before SUM 2987 was heard; (c) after SUM 2987 was heard and the Production Order was made; (d) after ETS commenced SUM 346 to obtain the Unless Order; and (e) after the Unless Order had been made (see GD at [80]). At the risk of repeating myself, an explanation in an affidavit cannot leave the other party and the court guessing.

109 Moreover, in the specific circumstances of this case, the time at which Xinbo lost possession or control was of especial importance for at least two reasons that the AR identified:

(a) First, as Xinbo had not taken the position that the WeChat messages were outside of its possession or control at the time when SUM 2987 was heard and the Production Order made, but only raised it when the time came for it to comply with the Unless Order, there was naturally going to be questions asked as to why the point was only being taken belatedly.

(b) Second, as a matter of common sense, faced with the coincidence in Xinbo’s position that, apparently, all of its representatives had lost possession or control of their WeChat messages for the same reason – ie, a change in mobile phones – it was predictable with almost certainty that ETS would raise doubt or even argue that the coincidence was not such but the product of a deliberate ploy on Xinbo’s part to suppress the messages. To be clear, I make no finding as to whether Xinbo had or had not suppressed the messages, and do not have to for the purposes of this appeal. The point is that it would have been obvious to any party in Xinbo’s position that questions would be asked of it, and it was incumbent on Xinbo to properly and pre-emptively attempt to answer those questions.

110 As I observed at [84] above, if an obvious follow-on question presents itself based on a party’s explanation, that would generally be a sign that the explanation should be firmed up to address the point pre-emptively. In my view, the position is a fortiori where, as in this case, the follow-on questions relate to or arise out of suspicions or doubts as to the veracity of a party’s explanation. I do not think it is an undue burden to require a party subject to an order for production of documents to anticipate readily foreseeable objections to its position. Document production should be approached with a lens of pragmatism. A party who has convinced the court to make a production order in its favour against another party would inevitably be suspicious if the latter provides a “nil” return; the former would hold the fair and reasonable expectation that documents responsive to the order do exist and are in the other party’s possession or control since the court would not have made the order in the first place if it had not been satisfied of those conditions. It is precisely because of this that the law demands an adequate explanation as a substitute when a party professes compliance with a production order without producing the document(s) ordered. It would make a mockery of the production regime, which is grounded in fair play in the conduct of litigation, if a party could simply hide behind the “plain and obvious” test and operate on a belief that the court cannot call its bluff because it holds all the cards.

111 Third, although Xinbo stated the event that had caused its loss of possession or control of the WeChat messages – ie, the change of mobile phones – its explanation was incomplete. I agree with the AR that Zhang’s explanation in the 14 March Affidavit “only supplies half of the story” as it states that Xinbo’s representatives are unable to access their old WeChat messages on their current (new) phones, but nothing is said about whether their old phones remain in their possession or control, as well as whether the WeChat messages can be accessed on their old phones (see GD at [74]–[75]). It could well be that the answers to either one of these questions is “no”. But the point is that the court does not know because Xinbo has not stated so, and it cannot assume a fact that Xinbo is not even required to prove but merely state. As I stated at [85] above, a party subject to a production order has to commit to a position, and it is antithetical to the policy underlying the document production regime to permit a party to hedge its position by making vague, general assertions in an affidavit filed in purported compliance with a production order.

112 In the final analysis, I consider that these inadequacies segue into a more general point. I have referred at [80] above to my observations in Natixis that a party has a duty to take reasonable steps to search for documents responsive to a production order made against it before the court will accept the explanation given on affidavit as conclusive of his professed inability to give production as ordered. The need to give a proper explanation that is sufficiently particularised on the material details of a party’s inability to give production is a corollary of that duty. A party who has taken the reasonable steps required of it under the law would have little difficulty providing an adequate account of those steps and what came of those steps. A party should have no difficulty committing to a position, backed by an adequate explanation, if it has made a good faith attempt at complying with the production order. Thus, if the court finds a party’s explanation lacking, it would likely also be justified in drawing the conclusion that that party has failed to make such an attempt.

113 The adequacy of a party’s explanations must also be viewed in context. I have already made the point above that it would have been patently clear to Xinbo that ETS would be suspicious of the position that Xinbo was taking in response to the Production Order, and the court should consider Xinbo’s omissions in that light. The failure to address a glaringly obvious objection is, self-evidently, far more indefensible than a failure to address an objection that is more subtle or, for want of a better word, of the nitpicking sort. If a party fails to address an obvious point, that will often be another indication to the court that the party has not undertaken reasonable endeavours to perform its production obligations in both letter and spirit.

114 There is a further piece of context. As Mr Tan pointed out in oral submissions, ⁶⁴ by the time Xinbo filed its affidavits to resist SUM 643, it was already known to Xinbo that ETS was taking issue with inter alia the adequacy of its explanations. ETS had raised these concerns in a letter to the court dated 5 March 2024 after Zhang’s 14 March Affidavit had initially been filed under cover of a solicitors’ affidavit on 4 March 2024. Even if I accept that Zhang’s 14 March Affidavit could not deviate from the draft earlier filed in court, Xinbo had every opportunity to respond meaningfully to ETS’s objections in its subsequent affidavits, namely, the affidavit deposed by Zhang on 21 March 2024 ⁶⁵ (the “21 March Affidavit”) as well as Wei’s 26 March Affidavit. However, neither of these affidavits shed further light or added details to Zhang’s 14 March Affidavit. In the 21 March Affidavit, Zhang largely offered bare denials of ETS’s allegations, and in relation to the WeChat messages specifically, she took the position that the 14 March Affidavit had been adequately comprehensive: ⁶⁶

115 Zhang’s position should be read against ETS’s objections in its letter dated 5 March 2024, which had explicitly referenced the lack of particulars as to the identities of Xinbo’s representatives and the time when they lost possession or control:

116 Thus, Xinbo’s defaults were directly and explicitly highlighted by ETS. Despite knowing that SUM 643 was an application that put the issue of its compliance with the Unless Order at the forefront, Xinbo chose to evade answering ETS’s questions, in both the 21 March Affidavit and the 26 March Affidavit. Indeed, even before me in this appeal, Xinbo has offered no justification or explanation as to why the details sought by ETS are unreasonable or unnecessary. It should therefore come as no surprise that I find Xinbo’s explanations woefully inadequate and that it has accordingly breached the Unless Order.

117 For the avoidance of doubt, my finding on the inadequacy of Xinbo’s explanation is an entirely separate matter from any finding that its unsubstantiated position that it has produced all the documents it can is false. As I have taken pains to emphasise, these are separate obligations imposed by the Production Order and the Unless Order. The court’s ability to look behind Xinbo’s asserted position is governed by the “plain and obvious” test, but the scrutiny over the adequacy of Xinbo’s explanation is not.

118 For this reason, I would respectfully not go as far as the AR to say that because Xinbo’s explanation is inadequate, the court cannot accept Xinbo’s position that it does not have possession or control over the WeChat messages as true, and that Xinbo has therefore breached Categories 1, 2(a) and 2(c) of the Production Order by failing to produce the WeChat messages (see GD at [81]). It appears that the AR’s reasoning was that the inadequacy of Xinbo’s explanation meant that it was “plain and obvious” that Xinbo had failed to produce the WeChat messages despite them being in its possession or control. In my view, although I accept that it is possible that an explanation may be so abjectly deficient that the court can see that it is “plain and obvious” that a party’s professed inability to produce the documents ordered is false, this chain of reasoning risks conflating the independence of the obligations imposed by a production order. As I explained at [76] above, a finding that it is not “plain and obvious” that a party has lied is not mutually exclusive with a finding that it is “plain and obvious” that a party’s explanation has been deficient. In this case, I am not entirely convinced that the deficiencies in Xinbo’s explanation provide a sufficient basis to meet the “plain and obvious” test so as to make a global finding that Xinbo has breached its obligation to produce documents (viz, the WeChat messages) responsive to all three of Categories 1, 2(a) and 2(c).

119 That said, it would be recalled that the AR held that Xinbo had breached Category 2(c) of the Production Order by failing to disclose certain WeChat messages, independently of the inadequacies in Xinbo’s account on its loss of possession or control of the WeChat messages. The AR’s analysis centred around the 30 December Message exchanged between HHC and Liang which, according to Zhang, Xinbo had been able to obtain from HHC despite Liang having lost his WeChat messages after changing his mobile phone (see [49] above).

120 I agree with the AR’s assessment that it is “plain and obvious” that there exist communications between HHC and Liang other than the 30 December Message. In the 14 March Affidavit, Zhang stated that there were “some communications” on WeChat between Liang and HHC “on matters relating to the Arbitration”, but Liang was unable to “retrieve the WeChat records … in the relevant periods of time” [emphasis added]. ⁶⁷ However, Xinbo had been able to obtain the 30 December Message from HHC which it disclosed in the 2SLOD. ⁶⁸ Although I am cognisant that words in an affidavit should not be analysed as if they are words in legislation, it is clear to my mind that plural references to “some” and “periods” at the very least suggest that there were more than a single chain of messages – ie, the 30 December Message – exchanged between Liang and HHC.

121 If that were the only basis for finding that WeChat messages other than the 30 December Message existed and were in Xinbo’s possession or control, I probably would not have been convinced that the high threshold of “plain and obvious” had been met. But that is not so. Zhang also stated in the 14 March Affidavit that Xinbo had been able to obtain another set of messages between Liang and HHC on 7 December 2022 from Jing. Curiously, Zhang did not say that this message was also obtained from HHC, which begs the question – why did HHC not produce this message, such that it had to be obtained from Jing? It is also not said by Zhang that this message and the 30 December Message are the only WeChat messages ever exchanged between Liang and HHC. In my judgment, this 7 December 2022 message is compelling proof that the 30 December Message was not the only message exchanged between Liang and HHC, and I am prepared to find based on this that it is “plain and obvious” that there remain WeChat messages other than the 30 December Message to be discovered from HHC.

122 The question that then arises is whether the WeChat messages in HHC’s possession can be said to be in Xinbo’s possession or control.

123 The phrase “possession or control” that is used in O 11 of the ROC 2021 is new; it replaces the former terminology of “possession, custody or power” in O 24 rr 1 and 5 of the ROC 2014. The change in nomenclature notwithstanding, it has been suggested in commentary that “[t]here is no substantive difference between the former and current wording as ‘custody’ or ‘power’ imply an element of control over the documents” (see Singapore Civil Practice at para 30-27) and “[t]he change in wording is not intended to narrow the scope of the provision” (see Singapore Civil Procedure 2024 vol 1 (Cavinder Bull SC gen ed) (Sweet & Maxwell, 2023) at para 11/2/3). Prior authorities addressing the definitional scope of “possession, custody or power” in the context of O 24 of the ROC 2014 are thus instructive in construing the scope of “possession or control” under O 11 of the ROC 2021.

124 Given that “possession” (a common concept across both O 24 of the ROC 2014 and O 11 of the ROC 2021) entails “a physical holding of the document coupled with a possessory right” (see Jeffrey Pinsler, Principles of Civil Procedure (Academy Publishing, 2013) (“Principles of Civil Procedure”) at para 17.005), it is clear that messages in HHC’s possession are not in the (physical) possession of Xinbo. It is not contended by ETS that there is any solicitor-client or other agency relationship between HHC (who acted for ETS, rather than Xinbo, in the Arbitration) and Xinbo that could, in principle, result in the attribution of HHC’s possession to Xinbo (see Disclosure at para 5-89).

125 So I am focused on the question of “control”. If the language or terminology in O 24 of the ROC 2014 is mapped onto O 11 of the ROC, “control” would correspond to “custody or power”. In this respect, “custody” can also be dismissed given that it entails physical possession simpliciter (see Principles of Civil Procedure at para 17.005; Disclosure at para 5-91). But “power” is much less straightforward. In the House of Lords decision of Lonrho Ltd and another v Shell Petroleum Co Ltd and another [1980] 1 WLR 627, Lord Diplock defined “power” as “a presently enforceable legal right to obtain from whoever actually holds the document inspection of it without the need to obtain the consent of anyone else” (at 635) [emphasis added].

126 If Lord Diplock’s definition of “power” is exhaustive of the scope of “control”, it would seem on the facts of this case that Xinbo cannot be said to have “power” over the WeChat messages in HHC’s possession as Xinbo does not have any strict in personam right against HHC to compel him to produce the messages to Xinbo or an in rem (or proprietary) right to the WeChat messages in HHC’s possession. However, subsequent cases have not read the requirement so narrowly. In Dirak Asia Pte Ltd and another v Chew Hua Kok and another [2013] SGHCR 1, the court observed that “a rigid adherence to Lord Diplock’s perceived definition of ‘power’ as an ‘enforceable legal right to obtain documents’ should give way to a more flexible contextual analysis” (at [24]), and that this “contextual approach” encompassed “not only the legal right to obtain documents, but also the practical ability to do so” [emphasis added] (at [28]). The court summarised the position thus (at [35]):

Subsequently, Kannan Ramesh J (as he then was) affirmed this approach in the Singapore International Commercial Court decision of Hai Jiao 1306 Ltd and others v Yaw Chee Siew [2020] 3 SLR 142 (at [46]–[47]), as did I in Natixis (at [35]). It can thus be seen that, even before the advent of the ROC 2021, our courts had in substance already been applying a test of “control”.

127 Some further insights into the concept of “control” may be gleaned from the English courts. English law shares a common experience with Singapore law as r 31.8(2) of the English Civil Procedure Rules 1998 (SI 1998 No 3132) (UK) (the “CPR”) also refers to “control”. In Various Airfinance Leasing Companies and another v Saudi Arabian Airlines Corpn [2022] 1 WLR 1027, Peter MacDonald Eggers QC, sitting as a Deputy High Court Judge of the English High Court, said this (at [21]):

128 The qualification at the end of this extract (which I have placed in emphasis) about the need for “specific and compelling evidence” going specifically to the issue of a party’s ability to access documents in another’s possession, as opposed to merely circumstantial evidence about the parties’ relationship generally, is significant. As Cockerill J subsequently opined in Loreley Financing (Jersey) No 30 Ltd v Credit Suisse Securities (Europe) Ltd and others [2023] EWHC 548 (Comm), “there is a degree of stringency required” (at [12]). In this connection, her Ladyship referred to the oft-cited decision of Ardila Investments NV v ECRC NV [2015] EWHC 3761 (Comm), where Males J (as he then was) cautioned that “practical control” in the sense that a third party could be expected to comply with the request of a party to the litigation did not suffice (at [11]). His Lordship preferred to state the test as one of “unfettered access” or “there [being] material from which the court can conclude that there is some understanding or arrangement” for the party to achieve such access (at [14]).

129 I agree with both Males J and Cockerill J that some circumspection is needed. An overbroad approach to “control” could place an undue burden on a party subject to a production order. But I would not go as far as to pitch the threshold at the level of “unfettered access” or to say that an expectation of compliance is invariably insufficient. In Natixis, I said that “[w]here the documents may lie with a third party, the duty extends to making reasonable efforts to request for the relevant documents” [emphasis in original] (at [32]). In my view, if there is clear evidence before the court that a party’s request for documents from a third party would be complied with, there is no reason why such documents should not be said to be in the former’s control. Indeed, this is illustrated by Phones 4U (in administration) v EE Ltd and others [2021] 1 WLR 3270, which I referred to in Natixis. In that case, the English Court of Appeal affirmed the making of an order against an employer that the employer should request that its present and former employees or agents make available their personal electronic devices for inspection.

130 Finally, in Berkeley Square Holdings Ltd and others v Lancer Property Asset Management Ltd and others [2021] EWHC 849 (Ch), Deputy High Court Judge Robin Vos identified the following general principles (at [46]):

Save for adopting the qualification to the first principle inserted by Jacobs J in the recent case of The Public Institution for Social Security v Al-Rajaan Al-Wazzan (in her capacity as representative of the estate of Mr Fahad Maziad Rajaan Al Rajaan (deceased)) and others [2024] EWHC 480 (Comm) that it goes too far to say that the relationship between the parties is “irrelevant”, and it is more accurate to say that the nature of the relationship is “not determinative” (at [28]), I generally accept these principles, subject of course to the caveat that they are signposts and should not be approached as if they were words of a statute or as entailing a box-ticking exercise.

131 In my judgment, applying the above principles, I am satisfied that it is “plain and obvious” based on the facts known to the court that Xinbo does have control over the WeChat messages in HHC’s possession. The fact that HHC has already voluntarily produced at least the 30 December Message to Xinbo on Xinbo’s request itself indicates that Xinbo can obtain possession of the WeChat messages from HHC. As I mentioned earlier at [121], it is ambiguous if the 30 December Message was the only message that Xinbo obtained from HHC or if HHC had in fact disclosed other messages to Xinbo as well. Moreover, I am buttressed in my conclusion by the peculiarity, highlighted by ETS, of HHC having cooperated with Xinbo’s requests whilst apparently rebuffing ETS’s own attempts at establishing contact with him – this is despite HHC having acted for ETS in the Arbitration. The unusual circumstances lead to the inference that Xinbo does have the practical ability to access, and thus has control of, the WeChat messages in HHC’s possession, if it does not already have them.

132 As I am satisfied that it is “plain and obvious” that WeChat messages between Liang and Shuang other than the 30 December Message (and the 7 December 2022 message obtained from Jing) exist and are in Xinbo’s control, I affirm the AR’s holding that Xinbo breached Category 2(c) by failing to produce these messages.

133 Lastly, I come to ETS’s complaint that Xinbo has breached the Production Order and the Unless Order by failing to produce the originals of the documents that it had disclosed in the LOD, SLOD and 2SLOD for ETS to inspect. According to ETS, when its solicitors attended at the offices of Xinbo’s solicitors on 6 March 2024 to inspect the documents, Xinbo had merely printed out documents that had already been sent in soft copy to ETS for ETS’s solicitors to inspect. ⁶⁹

134 I consider this to be straightforward. Although Xinbo contested this issue before the AR, Xinbo appears to have conceded (or at the least abandoned) the point before me, as there was complete silence in both its written and oral submission regarding this alleged breach of the Production Order and the Unless Order.

135 In any event, I agree with ETS and the AR that Xinbo’s non-compliance is clear. Before the AR, Xinbo argued that all it had were copies of the documents, it had prefaced its disclosure in the 2SLOD with the overarching description that the documents referred to therein were “copies and in Chinese unless otherwise stated”, and therefore, all that Xinbo could and had to give inspection of were copies of the documents and not originals. ⁷⁰ ETS’s frustration carries through in its written submissions where it animatedly refutes this argument as follows: ⁷¹

136 Although I would not go to the same lengths as ETS in its criticism, I accept the general thrust of its submission. In fairness to Mr Poon, his position before the AR was not only based on Xinbo having referred to “copies” in the 2SLOD, but also that Xinbo did not have the originals of at least some of the documents referred to therein such that it could only produce “copies”. But, in my view, Xinbo’s contention that its obligation to give inspection of “copies” entailed printing out hard copies of the soft copy documents that it had already sent to ETS was thoroughly misconceived.

137 It was already established under the ROC 2014 that a party’s obligation to give inspection relates to the originals of the documents in question (see Alliance Management SA v Pendleton Lane P and another and another suit [2008] 4 SLR(R) 1 at [35]; Fermin Aldabe v Standard Chartered Bank [2009] SGHC 194 (“Fermin Aldabe”) at [39]; Deutsche Bank AG v Chang Tse Wen and others [2010] SGHC 125 (“Deutsche Bank”) at [28]). Thus, in Fermin Aldabe, where the documents in question were 14 e-mail messages stored electronically on the defendant’s e-mail services, the court held that the defendant’s obligation to give inspection was not fulfilled by merely providing electronic copies of the discoverable e-mail messages; the defendant had to provide the technical means necessary for the plaintiff to inspect the e-mail messages which, at the minimum, entailed “provid[ing] a computer system from which the relevant e-mail mailboxes may be accessed and the 14 e-mail messages displayed on screen for the Plaintiff to view” (at [42]). Likewise, the court in Deutsche Bank held that the plaintiff’s assumption that providing printed copies of electronic documents for inspection discharged its obligations to give inspection of the electronic documents was wrong, as “where electronic documents are available, inspection should be given of the original and not merely a printed copy” (at [28]). Although the court noted that inspection of originals could be done for e-mails located outside of the jurisdiction by providing remote access to foreign e-mail services from the plaintiff’s offices in Singapore or having the devices containing such e-mails brought into Singapore, it considered that a reasonable alternative to these steps, in light of concerns as to costs and inconvenience, was to have the plaintiff provide electronic copies of the e-mails in their native format (at [29]).

138 The courts have arrived at the position above notwithstanding that O 24 of the ROC 2014 did not make any explicit reference to the “original” of a disclosed document (see Charles Hollander, Documentary Evidence (Sweet & Maxwell, 15th Ed, 2024) (“Documentary Evidence”) at para 9-25). But the point is put beyond doubt in the ROC 2021, as O 11 r 12 admits no such ambiguity because it expressly defines a party’s right to inspection as relating to the “original of any document produced”:

139 In the present case, even if all that Xinbo had were copies of the documents, its obligation to provide inspection of the copies of the documents related to the original or primary copy in its possession or control (see Documentary Evidence at para 9-25). In Rohan St George v 4Fingers Pte Ltd and another [2024] SGHCR 9, also a decision of the AR but in respect of a different case, the AR said that (at [75]):

140 I agree. A printed copy of an electronic document is not the “original”. Instead, the “original” of the electronic document would be the document as it exists on the device from which it was received or accessed by Xinbo. For example, if Xinbo received screenshots of text messages from a third party, its obligation to give inspection of the original screenshots required Xinbo to allow ETS to view the original screenshots on the phone or computer on which Xinbo had received the screenshot. Indeed, this mirrors what the court in Fermin Aldabe said in respect of e-mail messages (see [137] above). I therefore accept ETS’s submission, based on the authorities I have referenced above, that: ⁷²

141 The AR rightly observed that the illogicality in Xinbo’s position was apparent from the fact that the inspection of printed copies of the documents by ETS served no purpose (see GD at [84]); it added no value since ETS could have simply printed the copies it had received, and not have troubled itself with attending at the offices of Xinbo’s solicitors. I therefore have no hesitation in affirming the AR’s finding that Xinbo had breached the Production Order and the Unless Order insofar as the giving of inspection was concerned.

142 For the reasons I have explained in detail above, I hold that Xinbo did breach the Unless Order in a number of different aspects. It is thus necessary for me to go on to consider the consequences that should follow from Xinbo’s breach.

143 As a preliminary point, I consider it useful to make some observations about the nature and effect of unless orders. Specifically, I emphasise that, as a matter of law, an application like SUM 643 is not a necessary step for the prescribed consequence in an unless order to take effect. In this case, while ETS had commenced SUM 643 on the AR’s directions to confirm that the Unless Order had been breached and that the stated consequences therein had resultingly come into effect, it would be wrong to think of SUM 643 as the cause that triggered the consequences in the Unless Order. This is because an unless order is, by nature, “self-executing” as the consequence of non-compliance is embedded within the order as an automatic consequence upon breach (see DNG FZE v Paypal Pte Ltd [2024] SGHC 65 (“DNG”) at [100]). Indeed, this was especially clear in the present case as it had been specified in para 1 of the Unless Order that, if Xinbo failed to comply within the requisite time, ORC 1189 would be discharged and OA 222 would be dismissed “without any further order” (see [28] above). Any lingering misconception that an unless order does not automatically take effect should be dispelled.

144 I make two points that follow from this. The first is that it appears to have sometimes been overlooked that the breach of an unless order is a freestanding basis for the striking out of an action or defence. It is distinct from the court’s power to strike out an action or defence on the ground of a party’s breach of its production obligations under O 11 r 7(a) of the ROC 2021, which is a separate and independent regime that exists in parallel with unless orders (see DNG at [92]). Put differently, the court’s power under O 11 r 7(a) is exercisable based on a party’s non-compliance with its production obligations simpliciter, and the prior imposition or breach of an unless order is neither a necessary nor a sufficient condition for a party to bring an application and/or for the court to strike out a claim or defence under O 11 r 7(a).

145 The second point is that it is necessary to distinguish between: (a) the effect of the breach of an unless order, which is that the stated sanction in the order automatically takes effect; and (b) the court’s power to waive non-compliance either on request by the party in default or, more exceptionally, on its own motion (see Marcan Shipping (London) Ltd v Kefalas and another [2007] 1 WLR 1864 (“Marcan Shipping”) at [13]–[14]). The fact that the court has the power to forgive a breach of an unless order does not mean that the stated sanction in the unless order does not take effect until the court confirms that it has been breached and/or declines to grant relief to the party in default.

146 Earlier authorities had once taken so strict a view of the automatic effect of an unless order that it was thought that the court did not have jurisdiction to grant relief to a party in default after a stated sanction of termination or dismissal of a claim had been triggered, on the basis that the court could not resuscitate something that had ceased to exist (see Whistler v Hancock (1878) 3 QBD 83; Wallis v Hepburn (1878) 3 QBD 84; King v Davenport (1879) 4 QBD 402). This exceeding strictness was relaxed in Samuels v Linzi Dresses Ltd [1981] QB 115 (“Samuels”) on the basis that the court did have jurisdiction to extend the time for compliance with an unless order and thereby grant relief from the sanction(s) in the unless order. It was however recognised that “it is a power which should be exercised cautiously and with due regard to the necessity for maintaining the principle that orders are made to be complied with and not to be ignored” (at 126H). After Samuels, a practice developed, probably as a matter of caution, for a party seeking to take advantage of a breach of an unless order to apply to the court for a confirmation that the sanction had taken effect upon the other party’s breach. Nonetheless, and crucially, the development of this practice did not change the inherent nature of an unless order, viz, that the sanction imposed came into effect automatically upon breach (see Marcan Shipping at [13]–[14]). The position established by Samuels was subsequently endorsed by the Court of Appeal in Syed Mohamed Abdul Muthaliff and another v Arjan Bhisham Chotrani [1999] 1 SLR(R) 361 (“Syed Mohamed”) and, to my knowledge, is still the position today.

147 In Marcan Shipping, Moore-Bick LJ made the following observations on three consequences of the self-executing nature of an unless order, which are worth reproducing (see [34]–[36]):

148 Although the third consequence is also of much importance, and I will return to it in greater detail below, I focus for the present purposes on the first and second consequences which relate to the mechanics of an unless order. The first consequence affirms that the starting position upon the breach of an unless order is that the stated sanction(s) in the order automatically take effect. This is so regardless of whether the issue of breach or non-compliance is contested or not. If the issue is disputed, that may require an application to the court by the party seeking to rely on the unless order to confirm the other’s breach of the unless order, but the court’s decision would not be the trigger for the sanction, since the court would merely be declaring the status quo. The second consequence suggests that the sanction, once automatically triggered, would continue to operate unless the court grants the party in default relief from it (either on its application or, exceptionally, on the court’s own motion). In this regard, the court’s power to grant relief from the sanctions in an unless order can be located in O 3 rr 2(4) and 4(1) of the ROC 2021 which, respectively, refer to the power to (a) waive non-compliance with court orders; and (b) extend time for compliance with court orders.

149 In the present case, based on the second consequence identified by Moore-Bick LJ in Marcan Shipping, there appeared to me to be some substance in the argument that the court should only consider the issue of whether the Unless Order had been breached and thereby confirm if the sanctions did indeed automatically take effect. There was no application by Xinbo for relief from non-compliance (if the court did find such non-compliance) as SUM 643 was ETS’s application. In this regard, the leading scholar on English civil procedure law, Prof Adrian Zuckerman, expressed the following opinion in a case comment on Marcan Shipping (see Adrian Zuckerman, “How Seriously Should Unless Orders be Taken?” (2008) 27 CJQ 1 at 4):

150 Speaking for myself, I did see force in the view that the distinction between the operation of the sanction and the court’s power to grant relief means that the court should (unless it decides to act on its own motion to grant relief) confine itself in an application like SUM 643 to the issue of whether the unless order was breached or not. The court can act on its own motion, but it is not clear why this should be the general rule rather than the exception. If the defaulting party does not itself put in an application for relief, why should the court proactively act to protect a litigant that has repeatedly flouted its authority through disobedience of its orders from the consequences of its own actions? However, I am conscious that existing authority appears to have resulted in a general practice by the Singapore courts of considering, on its own motion, the appropriateness of the sanctions taking effect. I shall therefore also do so. In any event, ETS did not take the point that SUM 643 was not the proper forum to consider whether Xinbo should be granted relief from the consequences in the Unless Order.

151 The guideline judgment on the enforcement of unless orders is the Court of Appeal decision of Mitora Pte Ltd v Agritrade International (Pte) Ltd [2013] 3 SLR 1179 (“Mitora”). The following propositions can be distilled from this case:

(a) The starting position is that the breach of an unless order will automatically trigger its specified adverse consequences, and the onus is on the defaulting party to demonstrate that its breach had not been intentional and contumelious so as to avoid those consequences (see Mitora at [35], citing In re Jokai Tea Holdings Ltd [1992] 1 WLR 1196 at 1203B).

(b) A party may escape the characterisation of its breach being “intentional and contumelious” by demonstrating that it had made positive efforts to comply but was prevented from doing so by extraneous circumstances (see Mitora at [36], citing Syed Mohamed at [14]).

(c) But, even if a party’s conduct can be said to have been intentional and contumelious, that is not the end of the matter, as the court must nevertheless determine what sanction should be imposed as a result (see Mitora at [37]). In this respect, the court will be guided by considerations of proportionality (see Mitora at [39], citing Teeni Enterprise Pte Ltd v Singco Pte Ltd [2008] SGHC 115 at [64]).

152 Although it was decided just over a decade ago, Mitora is not a straightforward case insofar as it lies in the centre of an uneasy tension between the propositions that (a) a breach of an unless order would automatically trigger the stated sanctions; and (b) the court should nonetheless assess the sanction imposed having regard to proportionality. This difficulty was recently fleshed out with characteristic acuity by Goh Yihan J in DNG, who sought to rationalise the appropriate balance between these competing considerations in the following exposition that merits setting out in extenso (see DNG at [97]–[99]):

153 I respectfully agree with these observations. As Goh J points out, if the court were required to engage in a full reconsideration of the proportionality of the sanction in the unless order after the unless order has been breached, it would quite simply defeat the purpose of making the unless order in the first place. The court would essentially be conflating the distinct regimes of the enforcement of an unless order and its power to strike out a claim or defence based on a party’s non-compliance with its production obligations under O 11 r 7(1) of the ROC 2021, the exercise of which, as explained at [144] above, is not subject to a condition precedent of there having been a prior unless order in place. It must be, therefore, that if there is to be any consideration of the issue of proportionality at the stage of enforcement of an unless order, such an assessment must be on a light-touch basis so as not to render the earlier making of the unless order nugatory.

154 I would go further. In my judgment, the latitude for a defaulting party to make arguments on the (dis)proportionality of enforcing the stated sanctions in the unless order has to be considerably narrower than the status quo where no unless order has been made. I say this for the following reasons.

155 First, it is necessary to bear in mind the context in which unless orders are made. In the English Court of Appeal decision of British Gas Trading Ltd v Oak Cash & Carry Ltd [2016] 1 WLR 4530, Jackson LJ observed that (at [38]–[39] and [41]):

Since an unless order is only made on the premise of an antecedent breach of a court order (indeed, it is more likely than not that there has been more than one antecedent breach before an unless order is made), it is wrong to assume that a party who has breached an unless order is being called to account only for the breach of the unless order. Rather, the court should bear in mind that the defaulting party has breached (at least) “two successive obligations to do the same thing”. In the circumstances, there should be a strong prima facie assumption that enforcing the unless order – as well as the underlying breaches – would not be disproportionate.

156 Second, the court considering the enforcement of an unless order must bear in mind that the issue of the proportionality of the stated sanction in the unless order has already been considered by the court that granted the unless order in the first place. Thus, at least in a loose sense, the issue of the proportionality of the stated sanction is res judicata. In DNG, Goh J saw much force in the view that “if a court were to consider the proportionality of striking out following the intentional and contumelious breach of an unless order, that would amount to an attack on the earlier decision to impose the unless order” (at [100]).

157 I respectfully align myself with the learned judge’s view. Indeed, it also derives support from the UK Supreme Court decision of Thevarajah v Riordan and others [2016] 1 WLR 76 (“Thevarajah”). In that case, the defendants had failed to comply with an unless order and were resultingly debarred from defending the claim. In refusing the defendants’ (second) application for relief from sanctions under r 3.9 of the CPR (the first having been earlier refused), Lord Neuberger of Abbotsbury PSC observed that (at [18]):

158 This dovetails with Goh J’s observation in DNG: when it is established that a defaulting party has indeed breached the unless order, in asking the court not to enforce the automatic sanctions in the unless order, what the defaulting party is essentially seeking to do is for the unless order to be varied or rescinded. It is thus entirely logical that the court should generally refuse to do so, unless eminently good reasons are given. For the most part, any consideration that would have entered into the court’s mind when the unless order was made should not move the needle: ex hypothesi the earlier court would have considered the factor in question and assessed the sanctions stated in the unless order to nevertheless be proportionate. This is why, as Lord Neuberger noted, the court’s focus should be trained on a “material change in circumstances since the order was made” [emphasis added]: if such new factors exist at the time when the order is sought to be enforced, the enforcement court would not be undermining the earlier decision to grant the unless order since the judge making the order would not have factored in those factors into the calculus.

159 While the facts of Thevarajah were admittedly somewhat different insofar as they concerned a second (and not first) application for relief from sanctions, Lord Neuberger went on to say that (at [21]:

I do not think that these observations are confined solely to the specific context of Thevarajah. Thevarajah was an a fortiori case insofar as there had already been an earlier unsuccessful application for relief from sanctions. But the basic principle and thinking is the same: even in a first application for relief from the sanctions in the unless order, a decision granting such relief and refusing enforcement of the unless order’s automatic consequences would ex hypothesi be inconsistent with the reasoning of the court that made the unless order. It is doubtless that the court should not be seen as being too ready to contradict itself. As Prof Zuckerman cautions, “an automatic sanction set up by a rule or order might only be considered to be disproportionate ex post facto where circumstances have changed, or with the benefit of hindsight” and the court should not “lose sight of the difference between ex ante considerations of reasonableness and proportionality of process arrangements, and of ex post facto excuses for non-compliance” (see Adrian Zuckerman, Zuckerman on Civil Procedure: Principles of Practice (Sweet & Maxwell, 4th Ed, 2021) (“Zuckerman on Civil Procedure”) at para 11.157).

160 This segues into my third point. In my view, if a party considers the sanction in the unless order to be disproportionate, the proper forum for such complaints to be ventilated is not when the unless order has been breached and is sought to be enforced, but when the unless order is being sought. In other words, considerations of proportionality should be frontloaded to the time when the court is deciding (a) whether to impose an unless order; and (b) the terms of the unless order if one is to be made. If a party is disgruntled with the making of the unless order or the stated consequence(s) in the order, it is free to appeal against the making of the order and/or its terms. But once the unless order is made, and no appeal is filed (or, at any rate, successful), the ship to dispute proportionality further down the river ought to have sailed. I again refer to Prof Zuckerman’s observations in this regard (see Zuckerman on Civil Procedure at para 11.159):

It is, with respect, a distortion of the natural order of things for the court to be asked to consider the proportionality of an automatic consequence after the consequence has come to pass in law. The right balance, to my mind, is that closer scrutiny and circumspection may be warranted when deciding if an unless order is to be made (and on what terms), but if the court does make an unless order, the party in default should generally be accorded “no mercy” (see Hytec at 1675A). This would go some way towards mitigating the fundamental tension between the automatic nature of the unless order and the need to ensure proportionality and avoid overkill. This is referable to the third consequence identified by Moore-Bick LJ at [36] of Marcan Shipping in the extract I have earlier set out at [147] above – the draconian nature of the unless order may legitimately affect the court’s readiness to issue such an order, but it should not bear on the court’s resolve to enforce it once it is issued.

161 I return to Mitora for the fourth point. In my view, although the Court of Appeal’s observations on the importance of considering proportionality are no doubt of general application, they should be understood in the specific context in that case. In particular, the court expressly found that the appellant “did substantively comply with all its discovery obligations” (see Mitora at [21]). Any breach of the unless order by the appellant was so insignificant that the court considered that there had been a “false-positive diagnosis of evidential deficiency”, and even said that the courts below had erred by allowing the appellant’s statement of claim to be “struck out on a technicality” (see Mitora at [25]–[26]). The court’s comments on the importance of proportionality should be read in this context. It is therefore wrong, in my view, to cite Mitora as an authority for any wide-ranging proposition that, regardless of the nature and extent of the breach of an unless order, a prayer to proportionality may nonetheless save the day for the defaulting party. The Court of Appeal’s point in Mitora was really that striking out a claim for an arid or technical breach would be disproportionate where the court is satisfied that the unless order has been complied with as a matter of substance. And that must be unassailably correct. But in a case where the court does not find that the unless order has been substantively complied with, Mitora is, in my view, of little assistance to the party in default.

162 Indeed, the need to read Mitora in its context was subsequently emphasised by the Court of Appeal itself in Energy & Commodity Pte Ltd and others v BTS Tankers Pte Ltd [2021] 2 SLR 877. In that case, the appellants had failed to make a number of disclosures, resulting in an unless order being made against them which provided for their defences to be struck out and for judgment to be entered against them. After the unless order was breached, the aforesaid consequences followed. The Court of Appeal was unimpressed by the appellants’ reliance on Mitora in support of their argument that the unless order should not have been made. The following comments by Andrew Phang Boon Leong JCA (as he then was) on the factual setting of Mitora are apposite (at [21]–[22]):

163 Fifth, I would suggest that the principles in Mitora should be properly situated in the prevailing context of the ROC 2021, which is a different civil procedure regime that is underpinned by different values than that which persisted at the time when Mitora was decided. Put simply, what might have previously been considered disproportionate could be considered proportionate in the current legislative context. Amongst other things, the Ideals set out in O 3 r 1(2) of the ROC 2021 are of paramount importance, as the court is enjoined to “seek to achieve the Ideals in all its orders or directions” (see O 3 r 1(3) of the ROC 2021).

164 There have been practical consequences resulting from this change in philosophy. For example, Mr Poon suggested that one way to redress Xinbo’s breach of the Unless Order could be through a suitable costs order. ⁷³ This line of thinking appears to be influenced by the traditional assumption that, if prejudice occasioned by a procedural default was compensable by costs, the court would generally be content to grant its indulgence. In Affert Resources Pte Ltd (in compulsory winding up) v Industries Chimiques du Senegal and another [2024] 4 SLR 258 (“Affert Resources”), even in the context of a case governed by the ROC 2014, Goh J deprecated this assumption (at [3]):

165 If this assumption was already being sidelined in respect of the ROC 2014 regime, the position is now put beyond doubt in the ROC 2021. Order 3 r 2(4)(d) of the ROC 2021 expressly spells out that where there has been non-compliance with a court order, the court may “dismiss, stay or set aside any proceedings and give the appropriate judgment or order even though the non-compliance could be compensated by costs, if the non-compliance is inconsistent with any of the Ideals in a material way” [emphasis added]. In my view, this rule weakens the force of Mr Poon’s appeal to costs as an appropriate sanction as there is a clear intention in the ROC 2021 that the courts should move away from framing the question of prejudice from the perspective of costs being an adequate alternative recompense.

166 Moreover, as I intimated to Mr Poon at the hearing, I have some difficulty with the view that a costs order is a proportionate sanction for a breach of an unless order relating to a party’s breach of its document production obligations. ⁷⁴ A breach of production obligations bears on the substantive fairness of the proceedings as it can affect the other party’s ability to present its case. As Mr Tan rightly pointed out, there is a gradation of types of procedural defaults, and breaches of document production obligations are self-evidently among the most severe kind. ⁷⁵ To put the point somewhat bluntly, it seems to me perverse to suggest that a party can essentially “buy” the right to suppress or withhold material evidence from the other party or the court, by a payment of costs.

167 I would make a further point. Similar to how the court would have weighed the proportionality of the sanction in the unless order before making the order, the court dealing with a case under the ROC 2021 would have assessed the appropriateness of the unless order (and its embedded sanctions) against the Ideals. Thus, if the court refuses to enforce the unless order on a ground that was already factored into by the court (as opposed to an extraneous or unforeseen circumstance), the court would be refusing the enforcement of an order that it had already held to be upholding the Ideals of the ROC 2021. I am doubtful that such an approach or outcome can be correct.

168 Having regard to the above principles and, in particular, my observations on the limited role of proportionality, I consider that the present case is one where it is entirely appropriate for the sanctions in the Unless Order to be enforced. The onus is on Xinbo to demonstrate that the Unless Order should not be enforced, and I do not find that it has come close to discharging that burden.

169 For a start, I find that Xinbo’s breach of the Unless Order is intentional and contumelious. Indeed, I heard no argument to the contrary from Xinbo on this. Both its written submissions and Mr Poon’s oral submissions centred around the subsequent question of proportionality which, logically, means that Xinbo accepts (at the very least implicitly) that its breach was intentional and contumelious. Xinbo has chosen instead to hang its hat on the court’s discretion that is grounded on proportionality.

170 In any event, the point is straightforward. A breach of an unless order is by default intentional and contumelious, unless the party in default can show that it was prevented by outside circumstances from obeying the order (see DNG at [86]; Kraze Entertainment (S) Pte Ltd v Marina Bay Sands Pte Ltd [2014] 1 SLR 78 at [36]). There can be no suggestion here that Xinbo’s breaches which I have canvassed in detail above have been caused by anything but its own failings. As the AR observed, by taking the position that it had complied with the Unless Order, Xinbo logically disabled itself from providing any meaningful explanation for its non-compliance with the Unless Order (see GD at [91]).

171 Turning to proportionality, I am unconvinced by Xinbo’s arguments. As I have explained above, the starting point is that Xinbo did not appeal the making of the Unless Order or its terms. Therefore, the AR’s assessment on the proportionality of the sanctions in the Unless Order (and its consistency with the Ideals of the ROC 2021) should be given due respect. It should generally only be where the breach could either be properly characterised as trivial, or the result of circumstances that are extraneous (ie, outside of the defaulting party’s control) or unforeseen, that it would be disproportionate.

172 No such argument was advanced by Xinbo. Instead, the main point that Xinbo emphasised is that the impact of its breaches of the Unless Order on ETS’s ability to run its case in SUM 952 is not significant. Xinbo claims that since the AR agreed with it on this point, he was wrong to hold that enforcing the sanctions in the Unless Order was a proportionate response. ⁷⁶

173 Although I accept that this is, in principle, a relevant and important consideration, I disagree that it is of any assistance to Xinbo. In the first place, Xinbo has mischaracterised the AR’s decision. The AR did not express an unqualified agreement with Xinbo’s submission. In his words, he only agreed “[t]o some extent … with [Xinbo] that [ETS] appears to have been able to effectively mount its case in SUM 952 since day one” [emphasis added]. In fact, the AR then immediately went on to say that, in respect of the “critical question” of “whether there can nevertheless be a fair hearing”, he was “not satisfied that this [was] the case” (see GD at [105]).

174 Even if the AR did make a finding that ETS’ ability to run its case in SUM 952 was not affected by Xinbo’s breaches (which, to be clear, he did not), I would have disagreed with it. The link between Xinbo’s breaches of its production obligations under the Production Order and the Unless Order is manifest from the criteria that underpins the court’s jurisdiction to make specific production orders under O 11 r 3 of the ROC 2021. In particular, the touchstone of materiality means that the documents that the court has ordered to be produced are material to the adjudicative outcome; thus, the court in making a specific production order thereby decides that the documents have a real bearing on the way that the court would determine a pleaded issue in the parties’ dispute (see Cachet at [33]). In principle, therefore, there would be a clear enough nexus between a party’s breach of an unless order made to secure performance of a specific production order and the other party’s ability to present its case.

175 Further, Xinbo appears to conceive of the question of whether there can be a fair hearing between ETS and it in SUM 952 as a purely bilateral matter between the parties inter se. ⁷⁷ I question that premise. In my view, the court has a direct interest to ensure a measure of fair play in the conduct of litigation; so, even if it is not impossible per se for a party to present its case, the court should be entitled to take the view that the other party’s conduct has rendered the proceedings unfair. Put another way, it is wrong to assume that all there is to a “fair hearing” is the other party’s ability to present its case at a baseline level, such that one is then entitled to inflict disadvantage on the other so long as it does not go too far – that sets the standard too low, and the court should not countenance or permit any such manoeuvring that incentivises tactical non-compliance with court orders. I alluded to this concern in the course of Mr Poon’s oral arguments. ⁷⁸ In fairness to Mr Poon, he did candidly accept this point which, to my mind, was incontrovertible. Indeed, Mr Poon acknowledged that he could not run away from the point because it was on that basis that the Production Order was made. ⁷⁹

176 I have already expressed my reservations against substituting striking out for an adverse costs order at [166] above, and turn now to address Mr Poon’s alternative submission that the right (and proportionate) way forward is for the court hearing SUM 952 to draw adverse inferences against Xinbo. ⁸⁰

177 As a preliminary point, I note that the AR appears to have doubted the permissibility of adverse inferences being drawn against Xinbo if SUM 952 were to proceed without an application for the taking of oral evidence and cross-examination from ETS, on the basis that s 2(1) of the Evidence Act 1893 (2020 Rev Ed) (the “EA”) disapplies s 116 of the EA to proceedings decided solely based on affidavit evidence (see GD at [109]). If the AR was correct in proceeding on this premise, Xinbo’s reliance on adverse inferences as an alternative censure would be misconceived as its posited alternative would be no alternative at all.

178 In fairness to Xinbo, however, I respectfully doubt the correctness of that premise. In The Online Citizen Pte Ltd v Attorney-General [2023] 3 SLR 761, Aidan Xu J expressed the view that s 2(1) of the EA did not have the effect of preventing the court from drawing an adverse inference from affidavit evidence, whether on the basis that s 2(1) did not completely exclude the operation of s 116 of the EA or that an adverse inference was a general inference available in the common law that the EA did not displace (at [57]–[58]). Similarly, Prof Jeffrey Pinsler SC has opined, in relation to s 2(1) of the EA, that “[t]he non-application of the EA to interlocutory proceedings or other process involving affidavits does not prevent a court from applying basic evidential considerations in order to reach a just result” (see Jeffrey Pinsler SC, Evidence and the Litigation Process (LexisNexis, 7th Ed, 2020) at para 1.046). I am minded to agree with this perspective, and so I proceed on the basis – in Xinbo’s favour – that the court hearing SUM 952 could draw an adverse inference against Xinbo if it considers it fit, even if SUM 952 is a proceeding to be decided solely on affidavit evidence.

179 Nevertheless, the theoretical availability of an adverse inference does not advance Xinbo’s position. While the Court of Appeal in Mitora did list the drawing of adverse inferences against the defaulting party as one of various tools in the court’s arsenal other than the draconian sanction of striking out (at [45]), it is, in my view, insufficient for the defaulting party to point generally to the theoretical availability of a particular option, as the court has to assess the viability of any given option in the specific context of each case and its factual landscape.

180 In the present case, it is not at all clear to me how an adverse inference could satisfactorily address the prejudice occasioned by Xinbo’s deliberate breach of the Unless Order. Mr Tan referred me to Goh J’s decision in DNG, ⁸¹ where the learned judge had dismissed the viability of adverse inferences as “the extent of such inferences was difficult to measure because of the broad nature of the plaintiff’s breaches” and that “such inferences would overwhelm the plaintiff’s case almost entirely” (at [157]). I agree that the same can be said of this case. Xinbo’s breaches of the Unless Order are wide-ranging and affect a substantial, if not the entirety, of the issues raised in SUM 952. This is not a situation where there is a specific, discrete omission which the court could set right on a targeted basis; instead, the omissions infect the entirety of SUM 952. If Xinbo’s position is that the court in SUM 952 should draw adverse inference after adverse inference and adverse inference upon adverse inference, it is effectively suggesting that SUM 952 be decided entirely on a conjectural basis. I decline to proceed on such an unsatisfactory basis.

181 For all these reasons, I hold that the enforcement of the sanctions set out in the Unless Order is a proportionate response to the gravity and extent of Xinbo’s breaches of the Unless Order.

182 I come to the final arrow in Xinbo’s quiver. The argument is a legal one: it is said that the enforcement of a striking out sanction against an award creditor in proceedings to enforce a New York Convention award is either contrary to the New York Convention and/or the Singapore courts’ pro-arbitration policy undergirded by the principle of minimal curial intervention.

183 The argument is a creative one but there is no doubt in my mind that it cannot withstand scrutiny.

184 I start with the New York Convention. The gravamen of Xinbo’s complaint is that the striking out of an enforcement order for breach of an unless order is tantamount to fashioning a new ground for refusing enforcement of a foreign arbitral award by judicial fiat. It is argued by Xinbo that this contravenes Art V of the New York Convention and s 31(1) of the IAA which exhaustively prescribe the grounds on which a Contracting State to the New York Convention can resist enforcement of an award made in another Contracting State (which I shall refer to shorthand as a “Convention Award”).

185 In my judgment, there is no merit in this submission, for the following reasons.

186 First, it is normatively objectionable to say that any party that invokes the jurisdiction of the Singapore courts can claim to stand outside of our procedural rules or enjoy immunity from any part thereof.

187 Second, Art III of the New York Convention expressly puts paid to any such suggestion, as it expressly subjects the enforcement of a Convention Award to the procedural rules of each Contracting State:

188 It is thus explicitly clear that an award creditor must comply with the rules of a Contracting State if it chooses to invoke the jurisdiction of that Contracting State to enforce a Convention Award. The only constraint that is imposed on a Contracting State is the last sentence of Art III which sets out the principle of non-discrimination against Convention Awards: a Contracting State is not permitted to discriminate against a Convention Award by imposing more onerous procedural rules than those which apply to domestic arbitral awards. There is no suggestion from Xinbo that the striking out of an application to enforce an award is a sanction that somehow only applies to Convention Awards and not, for example, to domestic awards governed by the Arbitration Act 2001 (2020 Rev Ed).

189 The English High Court decision in Diag Human SE v Czech Republic [2014] 1 All ER (Comm) 605 (“Diag Human”) is particularly apposite. The court considered inter alia whether the New York Convention prevented the court from making an order for security for costs against an award creditor seeking to enforce a Convention Award. The award creditor made essentially the same argument as Xinbo in the present case that (see Diag Human at [12]):

Interestingly, the award debtor in fact appeared to concede this, as its counsel is recorded as having argued that the court had the option of imposing a stay of the enforcement proceedings rather than dismissal, which would not amount to a refusal of enforcement under Art V of the New York Convention (see Diag Human at [12]).

190 Burton J did not accept the award creditor’s argument. Indeed, his Lordship went even further than the position taken by the award debtor as he did not accept that dismissal of the enforcement proceedings for non-compliance with a security for costs order was impermissible under the New York Convention. After citing commentary that focussed on Art III’s prohibition on discriminatory treatment of Convention Awards, his Lordship said thus (see Diag Human at [15]–[16]):

Diag Human is thus authority that directly supports my conclusion that the Singapore courts can, in enforcement proceedings in respect of Convention Awards: (a) impose unless orders that prescribe the sanction of striking out of those enforcement proceedings; and (b) enforce that sanction in the event of non-compliance with the unless order.

191 Xinbo cites the English Court of Appeal decision of Gater Assets Ltd v Nak Naftogaz Ukrainiy [2008] Bus LR 388 (“Gater Assets”) as authority for the proposition that “English courts have declined to make certain orders if they would have the practical effect of refusing enforcement of an award”. ⁸² The court in Gater Assets considered the same issue in Diag Human as to whether it could make an order for security for costs against an award creditor seeking to enforce a Convention Award. Xinbo places particular reliance on the following comments in the judgment of Rix LJ, where his Lordship cast doubt on the permissibility of making such orders (see Gater Assets at [81]):

192 Although Gater Assets is a decision of a superior court in the English judicial hierarchy relative to Diag Human, I am not moved by Xinbo’s reliance on it. As neither decision is binding on me, I am free to prefer the reasoning in Diag Human if I am persuaded of its merits. In addition, the comments of Rix LJ in Gater Assets were strictly obiter and his Lordship himself made clear that he “[did] not make this a separate ground of decision” but was merely stating his opinion based on his understanding of the New York Convention’s structure (see Gater Assets at [82]).

193 In any event, I am persuaded by the merits of the approach taken in Diag Human. In Gater Assets, Rix LJ considered that the jurisdiction to make an order for security for costs went beyond merely Art III of the New York Convention because his focus was trained on the consequence of a Convention Award not being enforced if the order was not complied with. But, with respect, such an analysis seems to me to place the cart before the horse. This is because, as I explained at [188] above, Art III itself expressly subjects the enforcement of a Convention Award to the “rules of procedure of the territory where the award is relied upon”. Thus, a Convention Award would be enforced only if such enforcement would be in accordance with a Contracting State’s domestic rules of procedure. Put a different way, the pro-enforcement policy of the New York Convention bends to the domestic procedural rules of an individual Contracting State rather than the other way around. Indeed, the difficulty with Rix LJ’s approach was noted – correctly, in my view – in the separate judgment of Buxton LJ in Gater Assets who, in disagreeing with Rix LJ on this point, observed as follows (see Gater Assets at [116]):

I respectfully prefer Buxton LJ’s interpretation of the structure of the New York Convention, which was subsequently followed by Burton J in Diag Human.

194 Locally, there are other examples, albeit in different contexts or at a higher degree of abstraction, that demonstrate that there is no impediment to applying well-established domestic procedural rules to the context of enforcement of arbitral awards, even if the application of such rules would, as a practical matter, result in disruption to the enforcement of a Convention Award. For one, in The Republic of India v Deutsche Telekom AG [2024] 1 SLR 56, it was held that the doctrine of transnational issue estoppel from Singapore’s conflict of laws rules could be applied in the context of award enforcement proceedings. The Court of Appeal observed that “when dealing with the question of the enforcement of a foreign arbitral award, the New York Convention does not operate in isolation because the domestic law of the enforcement court also comes into play” [emphasis added] (at [97]). Like an unless order, the doctrine of transnational issue estoppel is a procedural tool that can result in the non-enforcement of an award.

195 In a similar vein is the duty to make full and frank disclosure of all material facts at the ex parte stage of an application to enforce an arbitral award under s 31 of the IAA. If Xinbo is right, it would mean that, even if Xinbo is found to have failed to make full and frank disclosure in its ex parte application for permission to enforce the Award, the court would be powerless to set aside ORC 1189 since discharging ORC 1189 would, in a sense, disrupt the enforcement of the Award on a ground not found in the New York Convention and s 31 of the IAA. But, as ETS rightly points out, ⁸³ there is local authority establishing to the contrary that the court can discharge its earlier grant of permission if the enforcement applicant is found to have breached its duty of full and frank disclosure, even if the court retains an overriding discretion not to do so in the circumstances of each case (see CZD v CZE [2023] 5 SLR 806 at [49], citing AUF v AUG and other matters [2016] 1 SLR 859 at [163]; National Oilwell Varco Norway AS (formerly known as Hydralift AS) v Keppel FELS Ltd (formerly known as Far East Levingston Shipbuilding Ltd) [2021] SGHC 124 at [176]–[184]). It is thus certainly not the case that permission to enforce a Convention Award, once given, can never be taken back or halted for non-compliance with domestic law or procedural rules.

196 Further, I note that the UNCITRAL Secretariat Guide on the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York, 1958) (United Nations, 2016) (“Guide on New York Convention”) also expressly recognises that the procedural rules of a Contracting State may result in the denial of the recognition and enforcement of a Convention Award (at p 86):

197 Third, the various examples I have raised to make the second point (at [194]–[195]) illustrate that Xinbo’s reliance on the exhaustive nature of Art V of the New York Convention and s 31 of the IAA rests on the false premise that anything that results in the practical consequence of a Convention Award not being enforced amounts to refusing enforcement of the Award in a way that falls foul of Art V and s 31.

198 Turning to Xinbo’s invocation of the well-established pro-enforcement policy of the Singapore courts and the principle of minimal curial intervention, I am also unconvinced that either of these support Xinbo’s attempt to fetter the Singapore courts’ jurisdiction to enforce unless orders made in proceedings to enforce a Convention Award.

199 An abstracted invocation of the Singapore courts’ pro-enforcement policy rings hollow. It cannot seriously be contended that a policy of pro-enforcement requires the court to ignore intentional and contumelious disregard for its authority in the form of breaches of its orders, still less an unless order. An arbitral award obtained by an award creditor in a Convention country is not a golden ticket to immunity from being visited with the natural consequences of the creditor disregarding orders made by this court in the course of enforcement proceedings that are before it. Xinbo’s argument attempts to set up a fence between the Singapore court’s pro-enforcement inclination and its protection of its authority and the due administration of justice as mutual exclusives, where the twain shall never meet. In my judgment, that is an ill-constructed fence. The court is entitled – and indeed, required – to strike the balance between these two imperatives as it sees fit.

200 Nor can Xinbo derive any assistance from the principle of minimal curial intervention. The principle of minimal curial intervention is premised on a proper ordering of roles between the court and the arbitral tribunal in the arbitration ecosystem (see Swire Shipping Pte Ltd v Ace Exim Pte Ltd [2024] 5 SLR 706 at [1]). In Soh Beng Tee & Co Pte Ltd v Fairmount Development Pte Ltd [2007] 3 SLR(R) 86, V K Rajah JA explained that (at [65(c)]):

It is plain that compliance with the court’s orders has nothing to do with the allocation of roles between the courts and the arbitral tribunal. It is exclusively a matter within the province of the courts. The court in meting out sanction for non-compliance with its orders in no way infringes or undermines the adjudicative jurisdiction of the tribunal or the tribunal’s mastery over its own procedure. I struggle to see how the principle of minimal curial intervention is of any relevance to this appeal or affords any way out for Xinbo.

201 For all of the foregoing reasons, I dismiss Xinbo’s appeal. In my judgment, Xinbo has undoubtedly breached the Unless Order, and the appropriate consequence flowing from the breach is for the specified consequences in the Unless Order to take effect. There is no impediment, whether in terms of legality or proportionality, that militates against Xinbo being visited with the natural consequences of its conduct.

202 Taking a step back and looking at things in the round, Xinbo has fear-mongered that enforcement of the Unless Order would set a “dangerous precedent”. I disagree. In my judgment, accepting Xinbo’s position would set a comparatively more dangerous precedent: it would give award creditors and enforcement applicants the impression that, with the award in hand, they have an open road to disobey the court’s orders with impunity. Anyone who invokes the Singapore court’s jurisdiction and powers for their own purpose – here, to enforce an award – necessarily subjects itself to the court’s authority and any orders made by it.

203 I will hear the parties on costs.