Singapore International Commercial Court

1 The appellant, Senda International Capital Ltd ("Senda"), was ordered by a three-judge bench ("the Court") of the Singapore International Commercial Court ("SICC") to pay costs and disbursements in SIC/S 4/2017 ("SIC 4") in the sum of S$8.1m to the respondent, Kiri Industries Ltd ("Kiri"). SIC 4 was a suit commenced by Kiri, in which it succeeded in its claim of minority oppression against Senda and obtained a buyout order for its shares. This appeal, in which Senda urges us to interfere with the Court's decision on the costs of SIC 4, raises the question of the manner in which costs are to be assessed for proceedings in the SICC, and as to the principles by which such costs are to be assessed.

2 The DyStar group was a major player in the international dye industry and was based in Germany. It experienced financial difficulties in 2009 and insolvency administrators were appointed. Kiri wanted to acquire the DyStar group's business but could not raise the funds to do so on its own. Kiri accordingly turned to Zhejiang Longsheng Group Co Ltd ("Longsheng") and invited it to enter into a joint venture for this purpose. The relevant agreements were executed by Kiri and Longsheng in 2010 and DyStar Global Holdings (Singapore) Pte Ltd ("DyStar") was incorporated as the vehicle that would be used to acquire the DyStar group's business. Kiri was the majority shareholder in DyStar while Longsheng (through its wholly-owned subsidiary Well Prospering Ltd ("WPL")) held one share in DyStar and a €22m zero-coupon bond issued by DyStar, which could be converted at any time into ordinary shares at S$10 per share.

3 In July 2012, the convertible bond was transferred from WPL to Senda, which was another wholly-owned subsidiary of Longsheng. Then, in December 2012, which was also around the time DyStar turned profitable, Senda converted all of the debt under the convertible bond into equity. As a consequence of these developments, Senda became the majority shareholder of DyStar while Kiri became a minority shareholder. Respectively, they held 62.43% and 37.57% of the issued capital.

4 The relationship between Kiri and Senda as joint venture partners deteriorated following the emergence of Senda as a majority shareholder in DyStar. On 26 June 2015, Kiri commenced a suit in the High Court against Senda alleging minority oppression. On 11 May 2017, the suit was transferred to the SICC.

5 SIC 4 was tried before the Court in two tranches. The first tranche of the proceedings involved the determination of Senda's liability for minority oppression. The Court concluded that Senda had oppressed Kiri and ordered Senda to purchase Kiri's 37.57% shareholding in DyStar, which was to be valued as at the date of its judgment, 3 July 2018 (see DyStar Global Holdings (Singapore) Pte Ltd v Kiri Industries Ltd and others and another suit [2018] 5 SLR 1). That decision was upheld on appeal in May 2019 (see Senda International Capital Ltd v Kiri Industries Ltd and others and another appeal [2019] 2 SLR 1). After the conclusion of the first tranche, the Court held, by way of an oral judgment delivered on 8 January 2019, that Kiri was to be awarded "full costs", and that all such costs were to be "taxed if not agreed".

6 The second tranche involved the determination of the value of Kiri's shareholding for the purposes of the buyout order. The Court delivered its decision by way of three judgments, and in a final judgment dated 21 June 2021, adjudged the value of Kiri's shareholding to be US$481.6m (see Kiri Industries Ltd v Senda International Capital Ltd and another [2021] 3 SLR 215 (" First Valuation Judgment "), [2021] 5 SLR 1, [2021] 5 SLR 111). In July 2022, Kiri's appeal against the Court's decision on valuation was allowed in part while Senda's appeal was dismissed, though the details of that decision of the Court of Appeal are not relevant to the present appeal (see Kiri Industries Ltd v Senda International Capital Ltd and another and other appeals [2022] SGCA(I) 5).

7 For clarity, we explain that, in this judgment, we refer to the period beginning with Kiri's commencement of its suit against Senda in the High Court on 26 June 2015 until 3 July 2018, when the Court delivered its judgment on liability and made the buyout order, as "the Liability Tranche". The costs order made after the conclusion of the Liability Tranche in the oral judgment dated 8 January 2019 (see [5] above) is referred to as "the Liability Tranche Costs Order". The period commencing immediately after 3 July 2018 until 21 June 2021, when the Court adjudged the final valuation of Kiri's shares in DyStar, is referred to as "the Valuation Tranche".

8 After the conclusion of the Valuation Tranche, the Court directed the parties to file written submissions on costs. Two rounds of written submissions were filed, and on 8 December 2021, the Court issued its judgment ordering Senda to pay Kiri costs and disbursements of S$8,111,642.11 (see Kiri Industries Ltd v Senda International Capital Ltd and another [2022] 3 SLR 174 ("the Costs Judgment")). The Costs Judgment is the subject of this appeal.

9 We set out the procedural history leading to the Court's decision in the Costs Judgment and the arguments made below by Kiri and Senda in some detail because, as we will shortly explain, this has a bearing on some of the views that we have arrived at in this appeal.

10 On 21 June 2021, being the day on which the Court delivered its judgment as to the final valuation of Kiri's shareholding in DyStar, it also directed that the parties "file written submissions on costs, limited to 10 pages in standard font size, by 12 July 2021". On 22 June 2021, by a joint letter, counsel for Kiri and for Senda sought clarification in respect of that direction on, amongst other issues, whether the parties were to address the Court only on issues of their in-principle entitlement to costs for the Valuation Tranche, or if they were also to address the quantum of costs and disbursements that they should be entitled to for the Valuation Tranche, and whether their submissions should also address the quantum of costs for the Liability Tranche.

11 On 23 June 2021, the Court directed the parties to address the issues of the entitlement to costs for the Valuation Tranche and of the quantum of costs for the Liability Tranche and the Valuation Tranche.

12 On 16 July 2021, the parties filed their costs submissions as directed. Senda made the following arguments in its costs submissions:

(a) For the Liability Tranche, the order that Kiri's costs are to be "taxed if not agreed" (meaning the Liability Tranche Costs Order) should stand. Since Kiri had not made any costs proposal to Senda and no agreement had been reached by the parties on costs, Kiri's costs for the Liability Tranche should remain either to be agreed or, failing that, be taxed. The Court should therefore not fix the costs for the Liability Tranche. Alternatively, if the Court wished to fix costs, then Appendix G of the Supreme Court Practice Directions 2013 ("Appendix G") setting out "Guidelines for Party-and-Party Costs Awards" should apply, and Kiri should be awarded costs of no more than S$204,000 for the Liability Tranche.

(b) For the Valuation Tranche, the Court should make no order as to costs and the parties should bear their own costs. This was because, first, Senda had succeeded on some issues in the Valuation Tranche which entailed far greater expenditure in terms of time and costs than the issues on which Kiri had succeeded, and second, the final valuation which the Court arrived at was one falling between the respective valuations adduced by the parties, suggesting that the parties were essentially equally successful or unsuccessful in the valuation proceedings. In the alternative, if Kiri were to be awarded costs, a significant discount of at least 48% should be applied.

13 In its submissions, Kiri referred to the decision of the Court of Appeal in CBX and another v CBZ and others [2022] 1 SLR 88 (" CBX "). There, we held that, for a case (like SIC 4) that had been filed initially in the High Court and was then later transferred to the SICC Registry, the assessment of costs should distinguish between (a) costs for the period before the transfer to the SICC ("Pre-Transfer Costs") and (b) costs for the period after the transfer to the SICC ("Post-Transfer Costs").

14 In CBX , we held that, for the assessment of Pre-Transfer Costs, O 59 of the Rules of Court (2014 Rev Ed) ("the ROC 2014") and Appendix G would apply, unless the parties had at the time of the transfer consented to its disapplication, or the Registrar handling the transfer of the matter to the SICC had made an order to this effect (see CBX at [28]). This is because, in general, the policy underlying the adoption of Appendix G for cases filed in the High Court would continue to apply at least in respect of steps that had been taken in the High Court, even if it should later be considered appropriate to transfer the case to the SICC for adjudication (see CBX at [28]). However, in an appropriate case, the court may either depart altogether from Appendix G or apply an uplift from Appendix G in the assessment of Pre-Transfer Costs (see CBX at [34] and [40]). On the other hand, for the assessment of Post-Transfer Costs, O 110 r 46 of the ROC 2014 applies. Order 110 r 46(1) states:

15 Kiri accordingly drew a distinction between Pre-Transfer Costs and Post-Transfer Costs in its costs submissions. Kiri first submitted that even for Pre-Transfer Costs, Appendix G should be departed from, and it claimed a sum of S$500,000 by way of costs for the period up to the time of the transfer to the SICC, on 11 May 2017. Kiri said this was discounted from its actual time costs of S$790,325. As for Post-Transfer Costs, freed of the constraints of Appendix G, Kiri claimed a sum of S$7,297,718.50.

16 Kiri's claimed costs for the post-transfer period (as well as the actual time costs which Kiri said were incurred for the pre-transfer period) were based on calculations set out in a costs schedule ("Kiri's Costs Schedule"), which it provided to the Court as part of a bundle of documents filed together with its costs submissions. Kiri's Costs Schedule sets out for each procedural stage of the matter (namely, pleadings, discovery, interlocutory applications, affidavits of evidence-in-chief and affidavits, preparation for trial, attendance at trial, closing submissions, judgment, inter-solicitor correspondence and correspondence with court), the total time spent by each of its fee earners for that stage and their respective hourly rates, and correspondingly, the costs incurred at each stage. Kiri's Costs Schedule was based on Form 24 in Appendix B of the SICC Practice Directions (effective 2 June 2021) ("SICC PD 2021"). Paragraph 152(4) of the SICC PD 2021 describes Form 24 as a "sample costs schedule" which the parties may submit to the court for the purpose of costs assessments.

17 In its submissions, Kiri also claimed disbursements totalling S$5,944,073.44. For the purposes of this appeal, we need only concern ourselves with the part of Kiri's disbursements claim relating to the payments made to its valuation expert in SIC 4, Accuracy Singapore Corporate Advisory Pte Ltd ("Accuracy"), which amounted to S$3,336,800.42. Kiri sought to substantiate its claim for the payments made to Accuracy by producing invoices from Accuracy, each of which consisted of three components: "fees", "direct expenses" and "indirect expenses". Save for one invoice dated 4 April 2019, which was accompanied by a cover letter stating that the "direct expenses" in that invoice were the travel-related expenses incurred for a trip made by some of Accuracy's employees to Singapore, there was generally no explanation or description in Accuracy's invoices as to what such "direct expenses" and "indirect expenses" were, nor did Kiri provide any substantiation in its costs submissions for these "direct expenses" and "indirect expenses". As for "fees", which taken together across all the invoices amounted to S$3,036,504.00, Kiri did not provide any breakdown in terms of the hourly rates charged by Accuracy, the qualifications or seniority of Accuracy's fee earners that had worked on the matter, the number of hours spent by Accuracy's fee earners and the nature of the work for which those fees had been charged. For clarity, in this judgment, we refer to the sum total of the items represented as "fees" across all of Accuracy's invoices as "the Expert Fees" (S$3,036,504.00), which is to be distinguished from the sum total of the payments made by Kiri to Accuracy that Kiri sought to recover as disbursements (S$3,336,800.42).

18 On 28 July 2021, the Court issued the following direction, seeking that the parties address it by way of further written submissions also limited to ten pages on the following:

19 Pursuant to that direction, the parties filed further costs submissions on 18 August 2021. Senda contended as follows in its further costs submissions:

(a) Both Pre-Transfer Costs and Post-Transfer Costs for SIC 4 should be assessed "with primary reference to Appendix G". Senda submitted that the nature of the work done in the pre-transfer period did not justify a departure from the costs scales set out in Appendix G, and for the work done in the post-transfer period, Kiri had not shown how the nature of such work was so different from other similar proceedings in the High Court at or around that time that would warrant such a significant uplift from Appendix G. Relying on Appendix G, Senda argued that Kiri should only be entitled to S$204,000 for the Liability Tranche (with that sum apportioned equally as Pre-Transfer Costs and Post-Transfer Costs), and S$156,000 (representing 52% of S$300,000) for the Valuation Tranche.

(b) Further, Kiri's claimed costs for SIC 4 should be rejected, for two reasons.

(i) First, these were said to be "wholly disproportionate" to the previous costs awards made by the High Court in minority oppression proceedings, and thus they were "plainly unreasonable and excessive".

(ii) Second, although Kiri's Costs Schedule does set out a breakdown of time spent by Kiri's legal team, "there [was] no detailed breakdown or narration of what the time was precisely spent on" and it was therefore "difficult for Senda to assess whether the costs incurred [were] indeed justified and reasonable". Senda argued that it would suffer prejudice if Kiri's costs were "fixed instead of taxed", because Kiri has not been required to provide "the kinds of necessary details that are ordinarily expected to be provided in a taxation". Senda also pointed to five aspects of Kiri's Costs Schedule (the details of which we need not go into) and sought to show that these claims were "excessive and wholly disproportionate". Senda also emphasised that these objections were "merely illustrative" and that it would require a "more detailed breakdown and narration of the costs claimed by Kiri" to assess whether they were indeed justified and reasonable.

(c) As to the sum of S$3,336,800.42 claimed for payments made to Accuracy, this was said to be "exorbitant". There had been no means for Senda to assess and determine whether the amount claimed by Kiri was reasonable because, save for the invoices generally documenting the payments made to Accuracy, no breakdown of the fees incurred had been provided.

20 In the light of the parties' submissions, there were four main issues before the Court: (a) first, how the Liability Tranche Costs Order should be interpreted; (b) second, whether Kiri was entitled to costs for the Valuation Tranche; (c) third, what the quantum of costs to which Kiri is entitled should be; and (d) fourth, what the quantum of disbursements that Kiri could recover should be.

21 On the first issue, the Court held that the Liability Tranche Costs Order permitted the Court to fix costs in a manner more robust than a summary assessment, by reference to the bundle of documents and costs' breakdown which Kiri had provided (see the Costs Judgment at [19]). The Court reasoned that, in the context of proceedings in the SICC, "taxed" simply refers to costs it fixed following something more than a purely summary assessment; such an order did not require a court to conduct a taxation of costs using the procedure envisaged under O 59 r 20 of the ROC 2014 pursuant to which a detailed Bill of Costs would be produced by the claimant and be scrutinised by the Registrar (at [16] and [18]).

22 On the second issue, the Court held that Kiri should be entitled to costs for the Valuation Tranche because it had been successful in persuading the Court as to how its shares in DyStar ought to be valued (see the Costs Judgment at [26]). The Court rejected Senda's submission that each party should bear its own costs for the Valuation Tranche.

23 On the third issue, the Court awarded Kiri S$114,636.90 in Pre-Transfer Costs and S$4,846,178.36 in Post-Transfer Costs. As for Pre-Transfer Costs, the Court held that Appendix G was the starting point of the assessment, but it accepted that the complexities of SIC 4 warranted an uplift from the costs arrived at based on Appendix G (see the Costs Judgment at [39]-[41]). As for Post-Transfer Costs, the Court held that the "reasonable costs" to which Kiri was entitled under O 110 r 46(1) of the ROC 2014 referred to all costs that had been "sensibly and reasonably incurred" (at [77]). The Court used the claimed costs for the post-transfer period in Kiri's Costs Schedule as the starting point in the quantification of such "reasonable costs" and excluded some items on the basis of what it considered would constitute double recovery (at [95]). The Court held, however, that costs that were incurred in arguing matters that were doomed to fail or to oppress the other party would not have been reasonably incurred and therefore ought not be awarded as "reasonable costs" (at [80]).

24 The Court held that, in SIC 4, one such issue that had not been reasonably pursued by Kiri was the account of profits issue in the Valuation Tranche (see First Valuation Judgment ([6] above) at [179]-[189]) and accordingly the costs incurred on that issue were to be excluded (see the Costs Judgment at [90]). The Court explained that it would have preferred to account for such costs based on an assessment of the time spent by Kiri's lawyers on that issue and then reducing the hours claimed by Kiri accordingly, but this was not possible because Kiri had not presented its time costs based on the issues argued (at [91]). The Court held that, having regard to the information that was put before it, the next best available proxy in this regard was "Kiri's success on the value of DyStar's shares" (at [92]). The Court noted that, if Kiri had succeeded on the account of profits issue, its shareholding in DyStar would have been valued at US$888m, which was US$270m more than the final valuation that the Court arrived at (US$618m). The Court divided the figure of US$270m by the figure of US$888m, yielding a percentage of 30.41%. Following from this, the Court, determined that this percentage would "approximate the extent to which Kiri [had] been unsuccessful on [the account of profits] issue and which ought to be deducted as costs not reasonably incurred" and this represented the percentage reduction that was to be applied to the costs for the post-transfer period claimed by Kiri (at [93]). As a result, Kiri was only awarded 69.59% of the Post-Transfer Costs that it had claimed.

25 On the fourth issue, only the part of the Court's decision dealing with the payments made to Accuracy is relevant to this appeal. As mentioned earlier, Kiri substantiated this part of its disbursements claim with Accuracy's invoices, each of which consisted of three components: "fees" (the Expert Fees), "direct expenses" and "indirect expenses" (see [17] above). The Court allowed only the Expert Fees to be recovered as disbursements (see the Costs Judgment at [104]), and also applied a 30.41% discount to the Expert Fees, since the part of the Expert Fees relating to the account of profits issue would similarly not be recoverable. The Court quantified this part of Kiri's recoverable disbursements as S$2,113,103.15 (at [105]).

26 Senda applied for permission to appeal against the Costs Judgment. We granted Senda permission to appeal, save for the part of the Court's decision that Kiri was to be awarded the costs of the Valuation Tranche, as that fell within the Court's discretion in relation to costs after taking the relevant facts into account. In granting permission to appeal, we noted that the Court's decision on the meaning of "taxation" in the context of proceedings in the SICC and the guidelines for the assessment of "reasonable costs" under O 110 r 46 raised questions of general principle which fell to be decided for the first time, and upon which further argument and a decision of the Court of Appeal would be beneficial. Senda duly filed the present appeal.

27 Before us, Senda's arguments are directed to the Court's decision on Post-Transfer Costs and to allow Kiri to recover part of the Expert Fees that Kiri had paid to Accuracy; Senda does not take issue with the Court's decision on Pre-Transfer Costs. Senda also does not take issue with the Court's interpretation of what "reasonable costs" under O 110 r 46 entails. Senda also clarified that it is no longer arguing for the use of Appendix G as a starting point in the quantification of Post-Transfer Costs. The main point made by Senda in the appeal is that the Court erred as a matter of principle in (a) how it assessed "reasonable costs" and (b) in allowing Kiri to recover part of the Expert Fees as disbursements. On this basis, it contends that we should interfere with the Court's exercise of discretion on those matters and make an order for the costs of SIC 4 to be re-assessed. Senda made the following arguments.

28 First, the Liability Tranche Costs Order required Kiri to first attempt to come to an agreement on costs with Senda, and only failing that, was Kiri to resort to taxation, or in the alternative, to take up separate proceedings for taxation akin to that provided under O 59 of the ROC 2014. In the absence of Kiri doing either of the above, the Court should not have proceeded to fix costs for the Liability Tranche in the way it did.

29 Second, the burden of proving that claimed costs are "reasonable costs" under O 110 r 46 falls on the claiming party, in other words, Kiri. The information furnished in Kiri's Costs Schedule was deficient because it did not provide a sufficient breakdown that would enable Senda, or the Court, to assess the reasonableness of the costs incurred by Kiri. The Court therefore could not have undertaken any meaningful exercise of assessing "reasonable costs" based on Kiri's Costs Schedule. The same applied to Accuracy's fees because Kiri similarly provided no breakdown whatsoever for those fees. Senda also argues that the claimed amounts for various stages of work done by Kiri's lawyers, as reflected in Kiri's Costs Schedule, were "unreasonably high" and exorbitant.

30 In response, Kiri argues that the Liability Tranche Costs Order did not oblige the parties to first come to an agreement on costs, and the reference to the costs being "taxed" simply refers to the quantum of costs being judicially determined. The Court therefore did not err in proceeding to fix costs for the Liability Tranche in the way that it did. Kiri had provided sufficient information to enable the Court to assess "reasonable costs". The criticisms made by Senda of the individual stage items in Kiri's Costs Schedule amount to an abuse of process because they were not raised before the Court, and in any case, it does not lie in Senda's mouth to challenge the reasonableness of Kiri's costs when Senda did not provide any information as to the costs that it had incurred in SIC 4. Kiri accepts that there has been insufficient breakdown of Accuracy's fees in the proceedings before the Court, but it argues that the Court would nevertheless have been in a position to meaningfully assess the reasonableness of those fees, given the Court's familiarity with the matter and the nature of the expert evidence given by Accuracy.

31 The main issue dividing Kiri and Senda is how "reasonable costs" under O 110 r 46 are to be assessed. While they do not specifically, or at this stage, take issue with what "reasonable costs" under O 110 r 46 are, we do consider it necessary to express our views on the correctness of the Court's decision on this because that question is anterior to the question of how such costs are to be assessed. Accordingly, the issues arising for our determination in this appeal are as follows:

(a) First, what is the proper interpretation of "reasonable costs" under O 110 r 46?

(b) Second, what is the manner in which such "reasonable costs" are to be assessed?

(c) Third, in the light of our views on the first and second issues, whether the Court erred (i) in its assessment of Post-Transfer Costs, and (ii) in allowing Kiri to recover part of the Expert Fees as disbursements?

32 In our judgment, the resolution of this question turns on the distinction between the respective approaches to costs in O 59 and O 110 r 46 of the ROC 2014. The Court correctly characterised the entitlement to costs under O 110 r 46 as being whatever costs that had in fact been sensibly and reasonably incurred by the successful party, and in this connection, it correctly started with Kiri's claimed costs as the logical point from which to commence the inquiry into just what those costs should be (see the Costs Judgment at [81]). However, the Court did not seem to us to appreciate fully the distinction between O 59 and O 110 r 46 in their respective references to "reasonable" costs, and we think it would be beneficial for us to elucidate that distinction. We summarise the reasoning of the Court before explaining our views.

33 In its analysis of this issue, the Court considered the conceptual underpinnings of the costs regime under O 59 of the ROC 2014, which is applicable to proceedings in the High Court, as a "point of contrast" to costs awarded under O 110 r 46 (see the Costs Judgment at [69]). The Court noted that, under O 59 r 27, there are two methods to assess costs: the "standard" and "indemnity" bases (at [70]). We reproduce O 59 rr 27(2) and (3) of the ROC 2014 for ease of reference:

34 The Court, in apparent reference to O 59 rr 27(2) and (3) of the ROC 2014, noted that there are two ways in which the concept of "reasonableness" arises in O 59 - first, the costs incurred by a party must be "reasonably incurred", and second, the amount of the "reasonably incurred costs" that are to be awarded to the receiving party must be "reasonable". The Court described this as a "double attenuation" of costs based on the consideration of reasonableness (see the Costs Judgment at [72]). The idea behind "attenuation" is a reference to how the claimed costs should be moderated.

35 In contrast, there is only one mention of the concept of "reasonableness" in O 110 r 46 - "reasonable costs". The Court held that there is some ambiguity on a plain reading of O 110 r 46, because the single mention of "reasonableness" means either there is only a single attenuation that is applied under one broad inquiry that requires that costs awarded be "reasonable", or a double attenuation involving a narrower inquiry requiring that costs be reasonably incurred, and a reasonable amount of those reasonably incurred costs be awarded (see the Costs Judgment at [73]). The Court held that the former approach is the correct one (at [73]). It arrived at this view on the basis of the nature of the disputes and the typical or expected profile of the parties coming before the SICC, who it considered, would generally be better-resourced than the average litigant. Further, the importance of securing access to justice, which informs any inquiry into costs in proceedings in the High Court, would be superseded in proceedings before the SICC by the commercial consideration of ensuring that a successful litigant is not out of pocket for prosecuting its claim or defence sensibly (at [76]-[77]). Because concerns of access to justice are not at the forefront, there would be less of a need to attenuate the amount of costs awarded in the SICC (at [76]). Thus, given the commercial consideration underlying the SICC, "as long as the costs are sensibly and reasonably incurred, a party in the SICC ought to be able to claim them" (at [77]). The Court also held that, because there is only a "single attenuation" in O 110 r 46, the costs orders in the SICC will generally be higher than in the High Court (at [74]).

36 We have reservations over some aspects of the Court's reasoning, which could suggest a mistaken view of "reasonable costs" under O 110 r 46.

37 First and foremost, the Court's reasoning is built upon the premise that the costs regimes under O 59 and O 110 r 46 are comparable with and likely related to one another. This led the Court to conclude that, because there is only a single mention of "reasonable costs" in O 110 r 46, as compared to a double mention of the concept of reasonableness in O 59, costs awarded in the SICC would generally be higher than in the High Court because there would only have been a single attenuation (see the Costs Judgment at [74]). This, however, presupposes that the costs regimes under both provisions are capable of meaningful comparison in the sense that there is a relationship between them and that the references to "reasonable" or "reasonableness" in O 59 and O 110 r 46 are broadly consistent with one another. As we explain later, that is simply not the case. Were it so, one would expect that even if costs in the SICC were "generally" higher, they would not be so much higher . It is not controversial that the amounts claimed by Kiri as Post-Transfer Costs were several times what it might have gotten pursuant to Appendix G and O 59 (see [41] below) and it is not apparent to us how this could be justified or thought to be explicable based on the lifting of one element of attenuation otherwise applicable in O 59 from O 110 r 46.

38 In our judgment, the approaches to costs under O 59 and O 110 r 46 are fundamentally distinct (see [51] below). Importantly, the Court was aware of this distinction, as it observed that unlike the assessment of costs under O 59, there is no yardstick (such as Appendix G) against which the quantum of costs may be measured for the purposes of O 110 r 46 (see the Costs Judgment at [67]), a point of significance that we will come to later (see [53] below).

39 Second, it is also not clear to us how a broad inquiry under O 110 r 46 into whether the claimed costs are "reasonable costs" could ignore the question of whether such costs are reasonable in quantum in overall terms. We agree that, in proceedings in the SICC, access to justice considerations are superseded by the commercial consideration of ensuring that a successful litigant is not out of pocket. However, it does not follow from this alone that the inquiry into "reasonable costs" should exclude the inquiry into whether the costs incurred are, on the whole, reasonable in amount. The commercial consideration underlying the SICC is not a justification either for the parties to chalk up runaway costs or for the court to cease to scrutinise the overall quantum of the successful party's claimed costs. The assessment of "reasonable costs" by definition entails an inquiry into whether the claimed costs were reasonably incurred and are reasonable in amount. To use an example cited by the Court of Appeal in Lin Jian Wei v another v Lim Eng Hock Peter [2011] 3 SLR 1052 (at [47]), if the overall quantum of costs was outrageously disproportionate to the amount at stake in the litigation (and so necessarily was unreasonable in amount in overall terms), we do not see how it could be said that those costs were "reasonably incurred". In short, it is incorrect to hold that the inquiry under O 110 r 46 should only be one about whether the claimed costs were reasonably incurred and excluding the question of whether they are reasonable in amount, and there is nothing on the face of O 110 r 46(1) that justifies such an interpretation.

40 Third, as we have noted (see [37] above), the Court's reasoning suggests that there will be a discernible relationship between costs awarded under O 59 and costs awarded under O 110 r 46, explained by a single rather than double attenuation for reasonableness in the latter. We have some difficulty with this. The test of reasonableness is an objective yardstick and will therefore be context-specific (see also Then Khek Koon and another v Arjun Permanand Samtani and another and other suits [2014] 1 SLR 245 (" Then Khek Koon ") at [175]). There is no reason to assume that "reasonably incurred" costs, which is common to both O 59 and O 110 r 46, constitute the same thing in both provisions, especially where different considerations inform the costs assessments under each provision, as the Court recognised.

41 Indeed, the result in this case emphasises that there can be no discernible relationship between costs awarded under O 59 and costs awarded under O 110 r 46. As mentioned earlier, the Court assessed Pre-Transfer Costs using Appendix G but applied an uplift (see [23] above). In doing so, the Court first assessed the costs of the Liability Tranche on the footing that that it had been conducted entirely in the High Court by applying a tariff of S$35,000 per day, for all 12 days of the trial during the Liability Tranche yielding a total of S$420,000. It then applied a multiplier of 27.29%, which is the percentage of time spent by Kiri's lawyers in the pre-transfer period on work for the Liability Tranche (see the Costs Judgment at [43]-[49]). On this footing, if the costs for the post-transfer period of the Liability Tranche had been assessed under O 59, it would have been 72.71% of S$420,000, or S$305,382. That stands in stark contrast to the actual quantum of such costs that the Court awarded to Kiri, which was S$1,657,092.74. The vast disparity in the costs for the post-transfer period of the Liability Tranche, depending on whether they came to be assessed under O 59 or O 110 r 46, cannot be explained just by the absence of the second attenuation for reasonableness in overall quantum. This, in fact, is a consequence of the fact that there is no discernible relationship between the quantum of costs that may be awarded under the two bases.

42 We therefore consider that the Court erred in its understanding and explication of "reasonable costs" under O 110 r 46, and we proceed to consider this issue as a matter of first principle.

43 The starting point of our analysis is the "indemnity principle", which underlies the costs recovery scheme in the common law civil litigation system. The indemnity principle dictates that a successful litigant is to be indemnified by the unsuccessful party for the legal costs he has incurred (see Maryani Sadeli v Arjun Permanand Samtani and another and other appeals [2015] 1 SLR 496 (" Maryani Sadeli ") at [30]). It was described by Bramwell B in Harold v Smith (1860) 157 ER 1229 (at 1231) in the following terms:

44 The indemnity principle ensures that a successful party is not prejudiced by having to assert its rights or to defend itself against the unsuccessful party in court proceedings (see R v Lord Chancellor Ex parte Child Poverty Action Group [1999] 1 WLR 347 at 356). Its function is therefore a restorative or compensatory one (see Then Khek Koon ([40] above) at [156]; see also Adrian Zuckerman, "The costs indemnity principle - from restoration to blame" (2008) 27(3) CJQ 281). The indemnity principle also explains why a successful litigant is not allowed to recover more than what he has paid or is liable to pay to his solicitor (see, for example, s 112(2) of the Legal Profession Act 1966 (2020 Rev Ed)). Similarly, if a solicitor expressly or impliedly agrees that he will not in any circumstances charge his client, then no costs are recoverable from the other party (see Mohamed Amin bin Mohamed Taib and others v Lim Choon Thye and others [2011] 2 SLR 343 at [25]).

45 However, the entitlement of a party to costs recovery is not a substantive right ; it is an incident of the legal system's scheme for costs recovery, which in turn is driven by social policy (see Maryani Sadeli at [33]). Depending on what these policies are, limitations may be placed on the restorative or compensatory function of the indemnity principle, thereby limiting what it means for a successful litigant to receive an "indemnity" for his legal costs (or, as put by the High Court in Then Khek Koon (at [163]), "what the procedural law defines to be an indemnity" [emphasis in original]). This in turn shapes the principles on which such costs come to be assessed.

46 An example of a situation where limitations are placed on the restorative or compensatory function of the indemnity principle is the costs regime under O 59 of the ROC 2014, that applies to proceedings in the High Court. The civil litigation system in the High Court is underpinned by the policy of enhancing access to justice for all (see Maryani Sadeli ([43] above) at [34]). However, securing complete restoration or compensation to the full extent of the costs incurred by the successful party, will have consequences that may prove to be inconsistent with the policy of enhancing access to justice. A complete indemnity will mean that there is no ceiling placed on the unsuccessful party's liability to pay costs. This may deter litigants with a potentially meritorious case from commencing litigation for fear of their possible (unaffordable or otherwise prohibitive) liability for adverse costs if matters did not go as hoped (see Then Khek Koon ([40] above) at [160]). Also, the level of costs which a party actually incurs in litigation will generally depend on the extent to which that party is able or willing to dedicate financial resources to the effort. An uncompromising pursuit of a policy based on full indemnification may mean that the level of costs which a successful party will be awarded comes to be dependent on the means available to it to expend resources on securing the best possible assistance in prosecuting its claim or maintaining its defence. This has two consequences. First, it will leave the better-resourced litigant in a better position and over time, this will diminish rather than enhance access to justice. And second, such a policy will likely promote inequality of arms with the better-resourced able to secure more expensive representation, and if successful, visiting potentially ruinous consequences on those simply unable to afford those consequences. And, as we have already observed, the risk of such catastrophic consequences may deter the pursuit of some legitimate claims.

47 Therefore, costs awarded under O 59 are assessed at such a level as would enable a litigant with reasonable merits to pursue justice (see Maryani Sadeli at [32]; Then Khek Koon at [156]-[157]). This requires the application of an objective standard to determine the level of recoverable costs in each case, shaped by the normative question of what ought to be the amount of costs a successful party can recover for the particular work done in the context of the dispute in question, irrespective of the level of costs the successful party may have actually incurred in the legal proceedings. The use of such an objective standard manifests itself in costs assessments under O 59 in two ways.

48 The first is the use by the High Court of costs precedents, which means that courts tend to award the same levels of costs in what appears to be similar or comparable cases (see, for example, Lao Holdings NV v Government of the Lao People's Democratic Republic and another matter [2022] SGHC(I) 6 (" Lao Holdings NV ") at [44]). Where two cases are broadly similar, a party will not be able to claim a higher level of costs in its case, simply because it had been willing or able to incur more costs. To put it another way, access to justice considerations require that the level of recoverable costs be tailored to the attributes of the case itself and what is generally accepted to be recoverable in the usual run of similar cases, and not subjective factors such as how much an individual litigant might have been willing to spend.

49 The second is the use of Appendix G in the assessment of costs. Appendix G was prepared based on research and consultation with legal practitioners in Singapore. It infers a reasonable amount of fees that may be charged for various types of work, which are then expressed in ranges to account for the reality that there will be variances from one case to the next in terms of complexity (see Appendix G at para 2). The effect of Appendix G means that, even catering for the specificities of individual cases, the level of recoverable costs will generally remain within the ranges set out therein, which represent the level of fees which members of the public and the legal profession would generally accept as reasonable. It is consistent with access to justice considerations that costs are assessed by reference to these generally accepted levels and, again, are not dependent on subjective factors such as how much the individual litigant might have been willing to spend.

50 Costs under O 59 are therefore assessed on the basis of the tariffs or ranges set out in Appendix G, and in the precedents, these reflecting what is the generally accepted level of costs for a particular type of work done in a particular type of case and also representing the notional level of costs necessary to enable a meritorious litigant to pursue justice. The assessment is independent of subjective considerations such as how much costs the individual litigant might be willing to incur, thus helping to ensure that costs awards do not become excessive and run the risk of deterring meritorious litigants. It is in this light that the tests of reasonableness in O 59, namely, "whether the costs were reasonably incurred or were reasonable in amount " (see [33] above), are to be understood. Whether costs were "reasonably incurred" entails a consideration of whether the costs were incurred in such a way that it corresponds to the level of effort that is generally accepted as being likely to be expended for the particular type of work in question, and whether costs are "reasonable in amount" entails a consideration of whether the overall quantum of costs corresponds to the level of costs that is generally accepted as being likely to be incurred for a particular type of dispute.

51 Whether a similar limitation (or if any at all) is placed on the restorative or compensatory function of the indemnity principle in the costs regime under O 110 r 46 would depend on the policies or considerations underlying proceedings in the SICC. The "international" and "commercial" nature of the disputes that come to be litigated in the SICC (see O 110 r 1(2) of the ROC 2014 and O 2 r 1(3) of the Singapore International Commercial Court Rules 2021 ("the SICC Rules 2021")) means that the parties who come before the SICC will generally be better-resourced. Such parties may generally also be more willing to incur greater expense on litigation, in part because of the amounts at stake, and also because the cost of such litigation is generally seen as part of the necessary expense entailed in the pursuit of commercial objectives. The policy of enhancing access to justice is therefore less relevant in the SICC. The ordinary expectation stemming from the indemnity principle that one may vindicate its legal rights through litigation at the unsuccessful party's expense, is not in the SICC subject to the limitation that is imposed by access to justice considerations in the High Court. We therefore agree with the Court to this extent (see the Costs Judgment at [76]), and also endorse the views in Lao Holdings NV ([48] above) (at [64]-[67]), that in the SICC, the principal underlying consideration is a commercial one of ensuring that a successful litigant is not unfairly put out of pocket for sensibly prosecuting his claim or defence.

52 Given the commercial consideration underlying the costs regime in the SICC, an award of costs under O 110 r 46 is generally intended to restore or compensate the other party for the expense it had incurred in the legal proceedings as long as this been incurred in sensibly mounting his claim or defence. The determination of the level of recoverable costs in each case therefore involves, as a starting point, a subjective inquiry into just what costs were in fact incurred by the successful party in the particular case. The commercial consideration underlying the SICC, however, is not a reason for the successful party to recover whatever costs it had incurred. Even in the SICC where access to justice concerns are superseded, there remains an overarching interest in directing litigants to pursue their proceedings in a reasonable and sensible manner (see, for example, B2C2 Ltd v Quoine Pte Ltd [2019] 5 SLR 28 at [14]). Therefore, even in the context of O 110 r 46, the indemnity principle is not an unlimited one that entails full restoration or compensation. The successful party is only entitled to recover "reasonable costs" from the unsuccessful party, and not whatever costs it had incurred. However, given the subjective starting point from which costs are assessed under O 110 r 46, this test of reasonableness will be directed at the costs that had in fact been incurred in the particular case, and not at what an appropriate level of costs to be incurred might be in a generic sense for a type of case similar to the one at hand. This is self-evidently different from the tests of reasonableness applied in the context of O 59 (see [50] above).

53 Given the very different bases on which the level of recoverable costs is to be determined under O 59 and O 110 r 46 - the former being the level of costs which a successful party ought to be able to recover and the latter being the level of costs which a successful party has in fact reasonably incurred - O 59 and O 110 r 46 are not comparable in terms of their approaches to costs. The distinction drawn by the Court between O 59 and O 110 r 46 in terms of a double versus single attenuation of reasonableness (see [37] above) is therefore not a meaningful one and says nothing about how costs under O 110 r 46 are to be assessed. As we have explained earlier, there is no basis for the Court's view that there was only one inquiry into reasonableness under O 110 r 46, being an inquiry into whether costs had been reasonably incurred and excluding the inquiry as to whether the costs are, on the whole, reasonable in amount (see [39] above). Both these inquiries are inseparable aspects of an overall inquiry into what costs would be "reasonable" in the circumstances. We do not see how costs that are unreasonable in amount can be said to have been reasonably incurred; nor how the reasonable quantum of costs may be used to justify the recovery of costs for work that should not have been done in the first place and so which could not be said to have been reasonably incurred.

54 What then does the court look to in determining whether the claimed costs are "reasonable costs"? As we have explained earlier, this must entail the court looking both at whether costs were reasonably incurred and whether the overall quantum of costs is reasonable. The two inquiries are fundamentally inseparable from each other because the manner in which costs were incurred and whether they were reasonably incurred will almost inevitably bear on what the overall quantum of costs would be, and whether that overall quantum is or is not reasonable. The list of factors which the court may take into consideration in ordering "reasonable costs", as set out in para 152(3) of the SICC PD 2021, also shows that the inquiry into "reasonable costs" extends to both whether costs were reasonably incurred and whether the overall quantum of costs is reasonable:

55 The factor pertaining to the conduct of the parties in the proceedings under para 152(3)(a) directs the court to consider how the costs in the particular case came to be incurred, for example, the manner in which allegations or issues had been pursued in the proceeding and whether a party had acted reasonably in doing so. On the other hand, the factors set out under para 152(3)(b)-(f) direct the court to consider whether the claimed costs are "reasonable" in the light of specific features of the litigation, such as the amount at stake, the complexity of the issues involved, and the total time and effort expended in the proceeding. These factors together involve a holistic inquiry into whether the overall expense incurred by the litigant is reasonable in all the circumstances.

56 For completeness, we note that O 22 r 3(1) of the SICC Rules 2021 (which is the equivalent of O 110 r 46(1) of the ROC 2014 in the SICC Rules 2021) makes clear the subjective basis on which costs are assessed in proceedings in the SICC, as we have articulated earlier (see [52]-[53] above). Order 22 r 3(1) states:

57 The list of factors previously set out in para 152(3) of the SICC PD 2021 (see [54] above), which we have explained show that the inquiry into "reasonable costs" under O 110 r 46 of the ROC 2014 extends to both whether costs were reasonably incurred and whether the overall quantum of costs is reasonable, are now part of an illustrative list of "all relevant circumstances" under O 22 r 3(2) of the SICC Rules 2021 which a court may have regard to in considering "proportionality and reasonableness" for the purposes of O 22 r 3(1).

58 In the Costs Judgment (at [78]-[79]), the Court held:

59 We make two brief observations. First, as we have already noted, given the fundamentally different bases on which costs come to be assessed under O 59 and O 110 r 46 (see [47] and [52] above), there can be no discernible relationship between the two sets of costs. It is therefore incorrect as a matter of principle to infer a relationship between costs under O 59 and O 110 r 46 on the basis of the single versus double attenuation distinction that the Court has drawn. We accept that, in practice, costs awarded under O 110 r 46 are likely to be significantly higher than those awarded under O 59, but that is a consequence of the different bases on which costs are assessed under O 110 r 46 and O 59, and not because there is any inherent relationship between them. Second, in the context of O 59, the distinction between solicitor-and-client costs and recoverable costs (or party-and-party costs) exists because, given the objective basis on which recoverable costs are assessed (see [50] above), they will often fall short of and therefore be distinct from solicitor-and-client costs except in cases where the parties have contractually provided for the recovery of indemnity costs (see Then Khek Koon ([40] above) at [172]). Since the two bases for assessing costs in O 59 and O 110 r 46 are not related, then it must follow that there is also no basis for importing into O 110 r 46 the concept of solicitor-and-client costs in O 59.

60 We turn to the second issue, which raises two questions for our determination. The first is whether the Court's making of the Liability Tranche Costs Order required it to order a separate process for the assessment of costs, so that it erred as a matter of principle in proceeding to assess and thereby fix costs for the Liability Tranche in the way it did. This also raises the question of what the process for the assessment of costs in the SICC should generally be. The second question concerns the level of information that a successful party should provide in order to enable the Court to assess "reasonable costs" in a meaningful way, and on whom the burden lies to demonstrate that the claimed costs in question are reasonable or unreasonable, as the case may be.

61 As we indicated to counsel for Senda, Mr Toh Kian Sing SC ("Mr Toh") during the hearing before us, we are not persuaded by Senda's arguments as to the Liability Tranche Costs Order.

62 The interpretation of a court order, like any other exercise in interpretation, is one of giving to the words used a meaning which those words can legitimately bear (see Sembcorp Marine Ltd v PPL Holdings Pte Ltd and another and another appeal [2013] 4 SLR 193 at [27]-[28]).

63 Applying these principles, we find it difficult to agree with Senda that the words "taxed if not agreed" imposed on the parties an obligation to first attempt to make agreement on costs before they may resort to taxation. There is no ambiguity in the use of these words, the proper interpretation of which is simply that costs are to be "taxed" in the absence of an agreement on costs between the parties.

64 Such an order therefore does not oblige the parties to come to an agreement on matters of costs. However, that is not to say that the parties should not attempt to agree on matters of costs, which is an altogether different matter. It is generally desirable that the parties be encouraged to try to come to such agreement, or at the very least engage in conversations with each other about their respective positions on costs. The unsuccessful party should not wait for the other to initiate such conversations as it is in its interest to seek clarifications about the other party's claimed costs, which ultimately it is going to have to pay. The successful party too has as much an interest in initiating such conversations because it would want to settle the issue as far and as fast as possible so that it can be paid. This applies as much to proceedings in the SICC as it does to proceedings in the High Court.

65 We emphasise this because it appears that in this case, no sensible dialogue had taken place between Kiri and Senda about their respective costs positions, and it was only on filing the first round of costs submissions (see [12] above) that their costs positions, which were vastly apart, were communicated to each other. We do not find that satisfactory because it meant that the issues in the costs assessment process before the Court were unnecessarily broad, and the Court was essentially presented with virtually an all-or-nothing situation in terms of Kiri's high figure of claimed costs (amounting to close to S$8m) and Senda's position based on Appendix G (amounting to just S$360,000). Even after their respective positions were communicated to each after the first round of costs submissions were filed, the parties engaged in no dialogue, which meant that the vast divide between the parties simply persisted into the second round of costs submissions (see [19] above). At the hearing before us, Mr Toh accepted that the many challenges which he now sought to raise before us over the claimed items in Kiri's Costs Schedule had never been addressed to Kiri's counsel through correspondence or conversations between solicitors. Counsel for Kiri, Mr Dhillon Dinesh Singh ("Mr Dhillon"), also accepted that his team never spoke to Mr Toh's team about Senda's position on costs, and his reason was that, since the parties' positions were so far apart, any such conversation would have been difficult. We do not find these explanations satisfactory. We accept that in a protracted dispute like SIC 4, the relationship between the parties may become acrimonious but there is no need whatsoever for this to infect the solicitors in their dealings with one another. However adversarial their respective positions as counsel may be, as advocates and solicitors, counsel are bound by their shared commitments and duties as officers of the court, to advise their clients of the sensible position to take and which in their professional judgment would best assist the court. In our view, this case could have been far more satisfactorily resolved if the parties through counsel had tried to discuss and understand their respective positions on matters of costs and crystallised their differences before the costs submissions were filed.

66 Finally, there is nothing in the use of the word "taxed" or "taxation" that requires the court to order a separate process for the assessment of costs. The word "taxed" or "taxation" is no more than a description of any process by which the amount of recoverable costs and disbursements is judicially determined (see Gomba Holdings (UK) Ltd and others v Minories Finance Ltd and others (No 2) [1993] Ch 171 at 189). The word "taxation" does not import any specific requirement as to how that process of judicial determination should take place, which is ultimately dependent on the legal context in which that word is used. Exactly what the word "taxed" or "taxation" entails in the context of proceedings in the SICC depends on what is, as a matter of law, the process for the assessment of costs in the SICC. It is to this that we next turn.

67 In our view, consistent with the judge-led approach to case management for proceedings in the SICC, it is typically for the trial court that heard the matter to assess costs under O 110 r 46. This is consistent with and affirms the ideals of efficiency and procedural flexibility that the SICC espouses, since it is the judges who have heard the matter who will have the clearest sense of just how complex the matter was or not, as the case may be, and so what level of incurred costs may seem reasonable. That it is for the trial court which heard the matter to assess costs is now made express in the SICC Rules 2021. Order 22 r 2(3) states:

68 Unlike O 59, which provides for a process for the assessment of costs before a taxing registrar, O 110 r 46 is silent on whether "reasonable costs" are to be assessed by way of a similar separate proceeding before the trial court that heard the matter, or whether such costs may be assessed by the court at the conclusion of the matter by way of written submissions, as the Court did in this case, or whether such costs can simply be fixed summarily by the court. What is clear, however, is that the court enjoys a broad discretion as to how costs are to be assessed. This is stated at para 152(4) of the SICC PD 2021:

69 Thus, whether costs are to be fixed, assessed at the conclusion of the substantive proceeding as the Court had done in this case, or assessed by way of a separate process after the conclusion of the proceedings, is a matter for the trial court's discretion. It comports with the efficiency and flexibility inherent in the SICC regime that the trial court be left to decide which method of costs assessment to adopt, depending on the complexity of the matter at hand (see also the Costs Judgment at [23]). We note that this is also the position in the SICC Rules 2021, which affords wide discretion to the court in determining the procedure by which costs are to be assessed. Order 22 r 2(4) states:

70 In our judgment, the factors which the trial court should consider in the exercise of its discretion as to how to determine questions of costs include: (a) the complexity of the issues in the substantive proceeding; (b) the amount of costs claimed by the successful party; and (c) the nature and extent of the differences in the respective positions on costs taken by the parties. In its exercise of discretion, the trial court should be guided by the need to maintain a measure of proportionality between, on the one hand, the nature of the inquiry into "reasonable costs", the corresponding level of detail involved in such an inquiry, and the expense associated with such an inquiry, and on the other, the amount of costs claimed by the successful party. This stands to reason since a costs claim in the millions of dollars must warrant more scrutiny than one in the thousands of dollars, and depending on the circumstances, more may have to be done in order for the trial court to satisfy itself that such claimed costs are "reasonable costs".

71 In this case, the Court might have considered whether it was appropriate to impose such a tight page limit on the parties' costs submissions (see [10] above), given what was likely to have been the very substantial quantum of claimed costs in SIC 4. And it ought at least to have reconsidered the position after Kiri's first round of costs submissions, which made clear the large quantum of costs that it was claiming. It is obviously a matter for the trial court's discretion as to whether and what page limits to impose, but in doing so the court should be sensitive to the extent of likely dispute on matters of costs and the level of scrutiny which it ought to apply to the claimed costs in the case at hand. The parties are, of course, free to assist the trial court with their views on these matters.

72 Order 110 r 46(1) (and the corresponding provision in r 46(2) that applies to appeals from decisions of the SICC) provides for the entitlement of the successful party to reasonable costs from the unsuccessful party. The relevant "fact in dispute" (see Britestone Pte Ltd v Smith & Associates Far East, Ltd [2007] 4 SLR(R) 855 (" Britestone ") at [58]) that a trial court must be satisfied of, before it awards the successful party its claimed costs under O 110 r 46(1), is that the successful party's claimed costs are indeed "reasonable costs". Since it is the successful party who seeks to persuade the trial court that it is entitled to be paid costs under O 110 r 46(1), the legal burden must also be on that party to prove the fact in dispute, namely, that its claimed costs are "reasonable costs". This legal burden brings with it an evidential burden on the part of the successful party to adduce some evidence to propound the existence of the relevant fact in dispute (see Britestone at [58]).

73 Since the starting point for the assessment of costs under O 110 r 46 is just what costs had actually been incurred by the successful party (see [52] above), in order for the successful party to show that such costs are in fact "reasonable costs", it should adduce evidence of information on its incurred costs and include a sufficient breakdown of such costs. Such evidence would typically include: (a) a breakdown of the claimed costs in terms of the number of hours claimed; (b) information identifying by whom those hours were incurred, their levels of seniority and corresponding hourly rates; and (c) some explanation as to the types of work those hours were incurred for. This same level of information ought also apply to expert fees that the successful party seeks to claim from the unsuccessful party as disbursements, which are only recoverable where they have been "reasonably incurred" (see Centre for Laser and Aesthetic Medicine Pte Ltd v GPK Clinic (Orchard) Pte Ltd and others and another appeal [2018] 1 SLR 180 at [54]-[55]). In this regard, we endorse the following observations in Lao Holdings NV ([48] above) (at [113]):

74 We add that this is also the level of information that is contemplated in Form 24 of the SICC PD 2021, as well as Form 24 the SICC Practice Directions (effective 1 August 2022) that is presently in force.

75 Once the successful party has adduced the requisite level of information in support of the contention that its claimed costs are "reasonable costs", the evidential burden shifts to the unsuccessful party to adduce evidence to show that the claimed costs are not "reasonable costs". An inquiry into "reasonable costs" under O 110 r 46 is a subjective one concerning the question of what level of incurred costs can be said to be appropriate in the particular case, and against that yardstick, whether and to what extent the successful party's claimed costs can be said to be "reasonable costs" (see [52] above). Hence, the best evidence that the unsuccessful party can adduce to discharge its evidential burden will often be information as to the costs that it had correspondingly incurred for the matter, which might well be a sound proxy by which the trial court can determine what the appropriate level of costs in the particular case is. The court may of course do so in other or additional ways as well (see, for example, [79] below).

76 It will not suffice for the unsuccessful party simply to make unsubstantiated contentions that the successful party's claimed costs are disproportionate, exorbitant, or unreasonable. This is so for two reasons.

77 The first reason is a practical one. Unlike in the context of O 59, where the court has the benefit of the costs ranges in Appendix G or of previous costs awards in assessing the reasonableness of claimed costs, in the context of O 110 r 46, the trial court has no similar yardstick against which that exercise may be undertaken. Further, because the starting point for the assessment of costs under O 110 r 46 is just what costs were in fact incurred, the inquiry of reasonableness is specifically directed at what would be an appropriate level of costs to be incurred for the particular matter . However, the court will not always be in a position to make an assessment on its own of what firms should charge, the number of hours that should be spent on a particular matter, and the seniority of the lawyers that ought to be staffed on the file.

78 The second reason is one of principle. A plaintiff on whom the legal burden lies must adduce some evidence that is not inherently incredible proving the existence of the fact in dispute, and where no evidence in rebuttal is adduced by the defendant, the court may conclude from the plaintiff's evidence alone that it has discharged its legal burden and make a finding against the defendant (see Britestone ([72] above) at [60]). Therefore, even though the legal burden is on the successful party to establish that its claimed costs are "reasonable costs", in an adversarial system like ours, whether that burden ultimately is found to be discharged will depend on what evidence the unsuccessful party adduces in rebuttal. The fact that the legal burden is on the successful party does not mean it is open to the unsuccessful party to dispute the successful party's entitlement to claimed costs just by making empty and unsubstantiated contentions that the claimed costs are not "reasonable costs".

79 We accept that previous costs awards made by the SICC may be relevant in the assessment of "reasonable costs" under O 110 r 46 in a case sharing common features with those cases in which those awards were made (see CBX ([13] above) at [42]). However, the use of such previous costs awards does not mean that a tariff-based approach to the assessment of costs like that under O 59 is being used (see [48] above). The exercise of determining "reasonable costs" in such a case remains a subjective one because the starting point for the trial court is the level of incurred costs in the context of that specific case. However, to the extent that the case at hand shares common features with other cases, the costs awards made in those other cases might possibly inform the court of what is an appropriate level of costs to be incurred for the matter in question. Any reliance placed on previous costs awards, is not to determine the level of costs that should be awarded, but rather to provide a check as to whether the costs claimed by the successful party are reasonable or not.

80 In its Appellant's Case, Senda argued that, in the assessment of "reasonable costs" under O 110 r 46 for a case transferred from the High Court to the SICC, the court should consider whether the party liable to pay costs had objected to the transfer out of concerns over the SICC's costs policies; and, if that was so, any costs awarded should be adjusted to reflect its position as an unwilling participant to the transferred proceedings before the SICC. Senda contended that it had expressed its concerns over the SICC's costs rules at the pre-trial conference on 11 May 2017 where SIC 4 was ordered to be transferred to the SICC. Also, Senda argued that, to the extent that it could have anticipated the risk of higher costs in proceedings in the SICC, it would have been reasonable for it to anticipate that the costs awarded would not depart significantly from awards in the High Court and the costs ranges in Appendix G, given the quantum of the costs awards which the SICC had made up until that time. On this basis, the Post-Transfer Costs of S$4.8m that the Court awarded to Kiri far exceeded any amount that Senda might reasonably expect to have been awarded, and so cannot be said to be "reasonable costs". This argument was not pursued before us at the hearing by Mr Toh, but we address it for completeness.

81 We do not accept that a party's objection to the transfer of a matter from the High Court to the SICC that is maintained specifically over issues of costs will generally be a relevant consideration affecting how the court assesses "reasonable costs" under O 110 r 46. It may be noted that stringent requirements must be satisfied before a matter may be so transferred, whether on the court's own motion or on the application of the parties - the claims between the parties must be of an "international" and "commercial" nature (see [51] above) and the court must also consider that it is "more appropriate" for the SICC to hear the matter (see O 110 r 12(4) of the ROC 2014 and O 2 rr 4(1) and (4) of the SICC Rules 2021). When the court concludes that the requirements for transfer are met, it would already have been satisfied that the matter is of such a nature that in all the circumstances, the commercial consideration of ensuring that a party is not unfairly put out of pocket for sensibly prosecuting its claim or defence is the relevant one.

82 In any case, having examined the record, we are not satisfied that Senda had objected to the application of the SICC's costs regime under O 110 r 46 during the transfer of SIC 4. At the pre-trial conference where this issue of the transfer of SIC 4 was dealt with, the following exchange took place between the court and then-counsel for Senda:

83 From that exchange, it is clear that the mention of "SICC costs" was not a reference to recoverable costs that might come to be awarded to the successful party at the conclusion of SIC 4. Instead, it was a reference to court fees and filing fees. Counsel representing Senda at the time did make a reference to the possibility of "greater or higher legal fees", but that was not a reference to the legal costs that Senda might eventually have to pay should Kiri prevail in SIC 4, but to the fees that the lawyers would have to charge to their own client (Senda) for the greater amount of work that then-counsel for Senda perceived to be required if the matter was tried in the SICC. In other words, Senda did not object to the transfer of SIC 4 out of the primary concern that a different approach on costs might be taken under O 110 r 46. We add that, any expectation that the costs of SIC 4 would have been similar to those set out in the costs ranges in Appendix G could not have been reasonably held by Senda. The nature of the dispute in SIC 4 and the complexity of the issues raised therein meant that, even if SIC 4 had been heard in the High Court, a significant uplift would have been applied to the figures in Appendix G.

84 We come to the last issue in this appeal. We deal with the issue of Post-Transfer Costs and the Expert Fees separately. Since the Court's award of Post-Transfer Costs and part of the Expert Fees to Kiri is an exercise of its discretion, we can only interfere with the Court's decision if one of the following grounds is present (see Hong Leong Bank Bhd v Soh Seow Poh [2009] 4 SLR(R) 525 at [45]):

(a) the Court was misguided with regard to the principles under which its discretion was to be exercised;

(b) the Court took into account matters which it ought not to have or failed to take into account matters which it ought to have; or

(c) the Court's decision was plainly wrong.

85 As part of its case on appeal (see [29] above), Senda made extensive submissions in its Appellant's Case and Appellant's Reply on why various aspects of Kiri's Costs Schedule were deficient in terms of breakdown, and how the different components of claimed costs in that schedule were exorbitant and unreasonably high. Some of these arguments were raised below (see [19] above) but many are raised for the first time on appeal. At the hearing, we informed Mr Toh that it would be quite inconsistent with our appellate function to consider arguments that had not been made before the Court, and so we directed Senda to limit itself only to arguments and materials that had been brought to the Court's attention in the proceedings below. We also informed Mr Toh that it was not our function as an appellate court to examine, as though we were a taxing court, whether each individual item of Kiri's claimed costs had been correctly awarded. In any case, as it will soon be apparent, even if we were minded to consider Senda's arguments in full, these would have been futile anyway (see [88] below).

86 Senda agrees that many of the points which it now seeks to rely on could have been raised below, but it urges us to bear in mind that they could only have raised those points after the first round of costs submissions had been filed (which was when Kiri's costs position was first made known to them) and even then, for the second round of costs submissions, a ten-page page limit was imposed by the Court, though we note that each party nonetheless included a few pages of annexes. Significantly, however, before the Court, Senda did not seek to enlarge the imposed limits. Critically, in the direction given by the Court for the parties to file a further round of submissions after the first round of costs submissions had been filed, Senda was asked to provide its own calculations of Post-Transfer Costs, and it was specifically cautioned by the Court that, if no such calculations were provided, the Court would only have Kiri's calculations as a point of reference (see [18] above). While we consider that a trial court imposing page limits on written costs submissions should be sensitive to the quantum of the claimed costs in question (see [71] above), we are satisfied that in the context of this case, the Court had in its discretion set out what it thought was the most appropriate procedure for Senda to dispute Kiri's claimed costs. We therefore see no reason to interfere with that exercise of discretion and we are also satisfied that Senda had been afforded a proper opportunity by the Court to dispute Kiri's claimed costs in the proceedings below.

87 In any event, in its second round of costs submissions (see [19] above), Senda used a little more than three out of the ten pages to dispute Kiri's claimed costs and disbursements. Close to six pages largely reiterated or elaborated upon Senda's position in the first set of costs submissions. Despite the Court's invitation to Senda to provide its separate calculations for Pre-Transfer Costs and Post-Transfer Costs with reference to the Court of Appeal's guidance in CBX ([13] above) as a mirror to what Kiri had provided, Senda did no such thing, and its position in the second round of costs submissions on the Pre-Transfer Costs and Post-Transfer Costs that Kiri should be entitled to remained as it had been, namely that it should be based on Appendix G.

88 Aside from this, we do not accept that Kiri's Costs Schedule was defective because of insufficient breakdown, or Senda's argument that Kiri's claimed costs are unreasonably high and exorbitant.

89 A successful party substantiating its claimed costs is not required to provide an exhaustive line-by-line breakdown of each item of incurred costs. The inquiry into "reasonable costs" under O 110 r 46 is whether, in overall terms, the level of costs which the successful party says it had incurred can be said to be appropriate for the particular matter in question (see [52] and [75] above). A successful party need only provide information showing broadly how costs had been incurred at each stage of the proceeding. The question is whether the information provided is sufficient to enable the other party to compare the claimed costs against the costs which it had incurred for the corresponding stage of the proceeding, so that the unsuccessful party can assess whether the successful party's claimed costs are questionable or not. It will therefore be sufficient as a general position that the overall claimed costs are broken down in terms of the costs incurred at different broad stages of the proceedings, such as commencement of proceedings/pleadings, interlocutory hearings, affidavits, preparation for hearings and attending hearings, as is presently required by Form 24. It is then for the unsuccessful party to seek further information from the successful party and if necessary, to seek the assistance or guidance of the court. Kiri's Costs Schedule, which was based on Form 24, was therefore fit for purpose.

90 Once Kiri furnished such information in support of the contention that its claimed costs are "reasonable costs", the evidential burden was on Senda to adduce evidence in rebuttal, in other words, evidence showing that Kiri's claimed costs are not "reasonable costs". As we explained earlier, perhaps the best evidence that Senda could rely on for this purpose was information pertaining to its own incurred costs. However, Senda adduced no such evidence, whether before the Court or before us. It was insufficient for Senda to contend that individual items in Kiri's claimed costs are exorbitant or unreasonably high without explaining why and without pointing, for instance, to its own incurred costs. Nor did Senda raise before the Court, some of the points it raised before us suggesting that on the face of it, some items claimed by Kiri seemed questionable. They were: (a) that Kiri had claimed costs of S$61,740 (representing a total of 85 hours spent by its lawyers) for taking instructions on Senda's Defence (Amendment No 2) and for taking instructions and drafting Kiri's Reply to Defence following Senda's amended defence which, according to Mr Toh, pertained to only two issues; and (b) that a significant part of the work done in respect of the "preparation for trial and related hearings" stage during the Valuation Tranche related to the preparation of trial bundles. It is undisputed that these were not part of the arguments which Senda raised in its second round of costs submissions in challenging Kiri's claimed costs (see [19(b)(ii)] above), and so they were not raised before the Court. As we told Mr Toh, these should have been pursued below, and not before us.

91 Before us, Senda also submits that the Court did not independently address its mind to whether Kiri's claimed costs are reasonable, and instead resolved that issue by erroneously finding that Senda had not raised a dispute over Kiri's claimed costs. We do not agree with this submission.

92 The process for the assessment of "reasonable costs" is a dynamic one under which both the successful and unsuccessful party must adduce some evidence in support of their contentions that the claimed costs in question are reasonable or unreasonable, as the case may be (see [73]-[76] above). In this case, despite having been invited by the Court to provide its own calculations for Post-Transfer Costs mirroring those provided by Kiri after the first round of costs submissions were filed, Senda declined to do so. Instead, in the second round of costs submissions, Senda reiterated its position as set out in the first round of submissions and made some other arguments as to the allegedly insufficient breakdown of Kiri's claimed costs and as to these being unreasonably high. However, nothing objective was raised to show that there was anything untoward in the level of Kiri's claimed costs. The Court therefore did not err in finding that Senda had not raised a dispute over Kiri's claimed costs. In those circumstances, the Court did the best it could and quantified "reasonable costs" using a multiplier to discount some of Kiri's claimed costs in the final award of Post-Transfer Costs. That was an exercise of the Court's discretion and we are not persuaded we should interfere with it in all the circumstances.

93 We are therefore not persuaded that Senda's arguments in the appeal have raised any ground for us to interfere with the Court's discretion in its award of the Post-Transfer Costs of SIC 4, and we therefore dismiss this part of Senda's appeal.

94 As we indicated to both counsel at the hearing before us, the position in relation to Accuracy's fees is quite different from that in respect of Kiri's Costs Schedule. In the proceedings before the Court, Kiri provided no breakdown whatsoever in respect of the sums paid by Kiri to Accuracy and which it sought to recover from Senda as disbursements. There was no breakdown as to how the sums payable by Kiri to Accuracy had been incurred, in terms of the number of hours claimed, the relevant hourly rates of Accuracy's fee earners, and for what work those hours had been incurred. The claim for the Expert Fees (the sum total of the "fees" in each of the invoices) was not particularised at all. Also, save for a limited description of "direct expenses" in one invoice dated 4 April 2019 (see [17] above), the "direct expenses" and "indirect expenses" in each of Accuracy's invoices were also not itemised. In our view, the level of information provided by Kiri in respect of its claim to recover the payments made to Accuracy was clearly deficient and inadequate. As there was no breakdown as to how the Expert Fees, direct expenses and indirect expenses had been incurred, there would have been no basis on which Senda could contend that these sums had not been "reasonably incurred".

95 In the event, the Court allowed Kiri to recover only the Expert Fees, which was also subject to the 69.59% multiplier that it applied to Kiri's claimed costs (see [25] above). After Senda filed its Appellant's Case in this appeal, counsel for Kiri wrote to Senda with a breakdown of the sums that had been paid to Accuracy. This included a breakdown of the time costs incurred by Accuracy at the different stages of SIC 4, such as preparation of affidavits, joint statement of experts, and preparation for trial and related hearings, based on the number of hours claimed and the hourly rates of the relevant experts from Accuracy performing the work, as well as other miscellaneous expenses that had been incurred by Accuracy. In its Appellant's Reply, Senda argues that the breakdown shows that Accuracy's fees had not been reasonably incurred. Senda questioned the need for some 23 people from Accuracy to be involved in the preparation of the affidavits containing the expert reports used in the Valuation Tranche, and why a total of 733.5 hours had been claimed as part of the preparation for trial and related hearings by members of Accuracy's team who did not testify in the Valuation Tranche. However, Senda did not adduce in evidence, any information as to what costs had been incurred by its own experts in the Valuation Tranche.

96 Given that the level of information provided by Kiri in its claim for Accuracy's fees clearly fell short of the requisite standard, we were initially minded to remit this aspect of the matter to the Court for re-assessment. One factor that we considered relevant to whether we should exercise our power to remit was the quantum of the fees that Senda had paid its own valuation expert in the Valuation Tranche, because, if those fees far exceeded what Kiri claimed or had been allowed to recover, we would have been less inclined to conclude that Accuracy's fees were not "reasonably incurred". We therefore invited Senda to provide us with that information. Senda subsequently informed us that the total fees and disbursements charged by its valuation expert, PwC, from the start of PwC's engagement until end-December 2020 was S$2.82m. Senda also informed us that the sum of S$2.82m covered PwC's administrative meetings with Senda's representatives and various discussions with Senda's legal team from Rajah & Tann Singapore LLP.

97 As we noted earlier, the Court did not allow Kiri to recover the full quantum of the Expert Fees, but only 69.59% of the same. This was an exercise of the Court's discretion, which had been necessitated by the insufficient level of information provided by Kiri which prevented the Court from assessing if each item of Accuracy's fees (as well as the Expert Fees) had been reasonably incurred. In the event, Kiri was only allowed to recover S$2,113,103.15. The amount of expert fees that Senda itself incurred (S$2.82m) indicates that the level of expert fees which Kiri was ultimately allowed by the Court to recover was within an appropriate range. That being the case, we are not inclined to interfere with the Court's exercise of discretion and remit this aspect of the matter for re-assessment. In coming to that conclusion, we are also mindful that remitting the matter in this case will serve no purpose other than to protract the dispute between the parties.

98 For these reasons, while we agree with Senda that the level of information provided by Kiri in support of its claim for Accuracy's fees was insufficient, in the circumstances of this case, we are not persuaded that we should interfere with the Court's discretion in allowing Kiri to recover the sum of S$2,113,103.15 in respect of the Expert Fees it had paid Accuracy. We therefore also dismiss this part of Senda's appeal.

99 For these reasons, we dismiss Senda's appeal.

100 For the benefit of the parties and counsel, we summarise the points of principle that we have set out in this judgment pertaining to the assessment of "reasonable costs" under O 110 r 46 for proceedings in the SICC:

(a) The starting point for the assessment of "reasonable costs" is just what costs were in fact incurred by the successful party, to the extent that such costs are "reasonable". This is an open-ended inquiry to be undertaken with due regard to the specific facts of the case at hand.

(b) For proceedings in the SICC, it is for the trial court that heard the matter to assess costs and it is also within the trial court's discretion to determine the manner in which costs are to be assessed. In our view, the trial court should, before rendering a decision on the substantive matter, first consider directing the parties to file their respective costs schedules containing a suitable breakdown of their incurred costs based on the principles that we have set out in this judgment. The parties can then be invited to address the court on the appropriate costs orders after a decision on the merits has been rendered. Such an approach would encourage the parties to communicate with each other on matters of costs, and also help narrow the issues dividing them. It will also assist the court in crystalising the issues for resolution in the assessment of costs and in determining the level of detail that will be required in undertaking the assessment of costs and deciding whether a separate process of assessment is required.

(c) When costs come to be assessed by the trial court, the legal burden is on the successful party to establish that its claimed costs are indeed "reasonable costs", and it must provide information to show how the claimed costs had been incurred and thereby allow the unsuccessful party and/or the court to assess whether they are reasonable. This will typically include (i) a breakdown of the claimed costs in terms of the number of hours claimed; (ii) by whom those hours were incurred, their levels of seniority and corresponding hourly rates; and (iii) some explanation as to what work those hours were incurred for. This level of information will also apply to claims for expert fees.

(d) Upon the successful party providing a sufficient breakdown of its claimed costs, the evidential burden shifts to the unsuccessful party to adduce evidence in rebuttal. Since the inquiry on "reasonable costs" under O 110 r 46 is one about what is an appropriate level of costs to be incurred for the particular matter, the best evidence which the unsuccessful party can adduce to show that the claimed costs are not reasonable will often be evidence of its own incurred costs. However, it may also adduce other types of admissible evidence and point to flaws evident on the face of the costs claim.

101 In this appeal, Kiri seeks costs of S$123,250 and disbursements of S$7,928.61. That includes the costs of Senda's application for permission to appeal made by way of CA/OS 36/2021 (see [26] above), which had been reserved in the cause. Kiri has quantified its costs by reference to the hours expended by its solicitors working on the appeal and their respective hourly rates. A breakdown is also provided by Kiri for its claimed disbursements. On the other hand, Senda seeks costs of S$80,000 and disbursements of S$10,563.45. However, it does not appear that this sum is based on the level of costs that Senda itself had incurred for this appeal. Instead, Senda justifies the figure of S$80,000 by saying that it is comparable to the S$89,512.50 that Kiri had claimed in the proceedings on costs in SIC 4. We also note that Senda has not provided a breakdown of its claimed disbursements.

102 Since there is nothing to suggest that Kiri's claimed costs of S$123,250 are manifestly not "reasonable costs", and in the absence of any contention or rebuttal evidence by Senda to the contrary, we award Kiri S$123,250 as its "reasonable costs" of this appeal under O 110 r 46(2) of the ROC 2014. We also allow Kiri to recover disbursements of S$7,928.61, which we are satisfied had been "reasonably incurred" having regard to the higher level of disbursements that Senda says it had incurred in this appeal. The usual consequential orders will apply.