PH HYDRAULICS & ENGINEERING PTE LTD v AIRTRUST (HONG KONG) LTD

In the court OF APPEAL of the republic of singapore [2017] SGCA 26 Civil Appeal No 234 of 2015 Between PH HYDRAULICS & ENGINEERING PTE LTD … Appellant And AIRTRUST (HONG KONG) LTD … Respondent Civil Appeal No 96 of 2016 Between AIRTRUST (HONG KONG) LTD … Appellant And PH HYDRAULICS & ENGINEERING PTE LTD … Respondent In the matter of Suit No 219 of 2013 Between AIRTRUST (HONG KONG) LTD … Plaintiff And PH HYDRAULICS & ENGINEERING PTE LTD … Defendant JUDGMENT [Contract] — [Remedies] — [Damages] — [Punitive damages] [Contract] — [Contractual terms] — [Exclusion clauses] This judgment is subject to final editorial corrections approved by the court and/or redaction pursuant to the publisher’s duty in compliance with the law, for publication in LawNet and/or the Singapore Law Reports. PH Hydraulics & Engineering Pte Ltd v Airtrust (Hong Kong) Ltd and another appeal [2017] SGCA 26 Court of Appeal — Civil Appeals Nos 234 of 2015 and 96 of 2016 Sundaresh Menon CJ, Chao Hick Tin JA, Andrew Phang Boon Leong JA, Judith Prakash JA and Tay Yong Kwang JA 28 November 2016 11 April 2017 Judgment reserved. Andrew Phang Boon Leong JA (delivering the judgment of the court): Introduction 1 Is there any role for the concept of punishment in the common law of contract? Does the remedial response to a breach of contract lie only in compensation? Put simply, does the voluntary nature of an agreement in general and the recognition of the legitimate pursuit of self-interest in particular rule out − in the main, if not altogether − any possibility that egregious conduct by the party in breach might result in an award of punitive damages – that is, a monetary sanction in addition to the payment of compensation to the aggrieved party? To reiterate, is there any role for the concept of punishment in the common law of contract? And, if not, does contract law then necessarily take on a cloak of amorality? As we shall see, these deceptively simple questions belie deep (and oftimes complex) questions that bring us to the very heart of contract law. 2 Whether or not the law in Singapore should recognise the availability of punitive damages purely for breach of contract (that is to say, absent concurrent liability in tort) is the major issue in the present appeal (“Issue 2”); and to assist us in resolving it, we appointed Assoc Prof Lee Pey Woan (“Prof Lee”) as amicus curiae in this appeal. However, Issue 2 is sandwiched between two other issues. The first is whether the defendant below, and appellant in the first appeal, PH Hydraulics Pte Ltd (“PH”), had been reckless, dishonest and/or fraudulent in its design of a 300-ton Reel Drive Unit (“RDU”) for sale to the plaintiff below and the respondent on appeal, Airtrust (Hong Kong) Ltd (“Airtrust”) (“Issue 1”). In many ways, this particular issue is a threshold one, for if we answer it in the negative, then there would appear to be no legal basis to award punitive damages for breach of contract in the first place – although, for reasons we will come to, we do not think a negative finding on Issue 1 precludes us from expressing our views on Issue 2 (see below at [66]). The third issue (“Issue 3”) is whether a particular clause in the Sale and Purchase Agreement between the parties, Clause 25 (“Cl 25”), can be construed as limiting the extent of the damages payable by PH to Airtrust. 3 The High Court judge (“the Judge”) held, in Airtrust (Hong Kong) Ltd v PH Hydraulics & Engineering Pte Ltd [2015] SGHC 307 (reported in part in [2016] 1 SLR 1060) (“the Judgment”), in favour of Airtrust on Issue 1 and proceeded to find that PH’s conduct merited the award of punitive damages (thus also finding in favour of Airtrust with regard to Issue 2). Having found that PH had been fraudulent, he held that Cl 25 could not exclude such fraud. Issue 3 was not strictly speaking before the Judge but is an issue that has been raised on appeal. Issues 1 to 3 thus form the basis of PH’s appeal in Civil Appeal No 234 of 2015 (“CA 234”). 4 In a subsequent hearing, the Judge declined, in Airtrust (Hong Kong) Ltd v PH Hydraulics & Engineering Pte Ltd [2016] 5 SLR 103 (“the Costs Judgment”) to award Airtrust the costs of the action on an indemnity basis and ordered such costs to be awarded on the standard basis instead. Airtrust now appeals against this decision in Civil Appeal No 96 of 2016 (“CA 96”) and whether or not the Judge was right not to order indemnity costs is the final issue in the present appeals (“Issue 4”). 5 With this very brief overview, let us turn, first, to recount the facts as well as decision of the Judge in the court below. Facts 6 PH is incorporated in Singapore. It designs, manufactures, and supplies heavy machinery for offshore use in the marine and gas industry. Airtrust is incorporated in the Special Administrative Region of the People’s Republic of China, Hong Kong. 7 In 2007, an Australian company, Trident Offshore Services (“Trident”), approached PH with the intention of purchasing a 300-ton RDU. The RDU was to be mounted on board a vessel for the laying of undersea umbilical cables in the Bass Straits of Australia. Airtrust then entered into talks with Trident, and it was decided that it would buy the RDU from PH which was then to be leased to Trident. It is not disputed that all negotiations for the purchase of the RDU were conducted between one of Trident’s associate companies, on Airtrust’s behalf, and PH. 8 Airtrust and PH entered into a Sale and Purchase Agreement (“SPA”) on 7 September 2007. Under the SPA, PH was to design and supply the RDU. The purchase price was $895,000. Included in the purchase price was American Bureau of Shipping (ABS) “Full Certification” at an itemised cost of $20,000. The purpose of the ABS certification was to ensure that an independent entity would review the design, survey the manufacturing processes and witness the testing of the RDU to ensure conformity with the requisite design, construction, and structural codes and standards. For present purposes, “full certification” refers to a review of the RDU’s structural design, mechanical design and electrical components. 9 It transpired that ABS was not able to provide certification for machines like the RDU. PH thus suggested to Trident, in an email on 12 October 2007, that ABSG Consulting, Inc (“ABSG”), a subsidiary of ABS, provide the certification instead. The email correspondence on this issue was forwarded to Airtrust, who agreed. 10 As part of the process for obtaining certification, PH prepared a confidential bundle of documents containing design drawings and calculations and submitted it to ABSG (“Confidential Bundle”). The bundle was prepared by one of PH’s assistant managers, Mr Steven Gan (“Mr Gan”). It was submitted in December 2007. Airtrust was at no time aware of what was in the Confidential Bundle – it was not disputed that PH was solely responsible for designing the RDU and that it kept the design of the RDU confidential. 11 PH engaged a freelance structural engineer, Dr Liu Li (“Dr Liu”), to analyse its structural design using STAAD.Pro, a computer programme by which structures are modelled so that they could be tested against different loadings, or load cases. In applying load cases to a model, the software also processes what is termed a “unity check”. To pass the unity check, a value of less than “1.0” is required. 12 In Dr Liu’s preliminary report sent to PH on 26 October 2007, the STAAD.Pro analysis showed that the unity checks were not met. PH then decided to strengthen the structure and indicated so in the AutoCAD drawing that Mr Lei Chengyi, PH’s design engineer at the material time, sent to Dr Liu on 29 October 2007. On 7 November 2007, Dr Liu sent Mr Lei a final report indicating that the structural design passed the unity check. 13 On 4 February 2008, Ms Renuka Devi (“Ms Devi”), the lead structural engineer from ABSG, asked Mr Gan for a number of documents, including the STAAD.Pro input file. Ms Devi asked for the STAAD.Pro file again the next day, 5 February 2008, but this time from Ms Tan Sin Liu (“Ms Tan”), PH’s mechanical engineer who was in charge of the design of the RDU. On 15 February 2008, Dr Liu emailed the STAAD.Pro file to Mr Lei, who forwarded it on 18 February 2008 to Ms Tan; she sent it to Ms Devi the same day. 14 On 18 February 2008, Ms Devi emailed PH (her email was specifically addressed to Ms Tan) with, among others, the following query: “No wind load has been considered – Please clarify”. Ms Tan forwarded the email to Dr Liu for his advice on Ms Devi’s query. Dr Liu replied on the same day, “As per the requirement by [PH], wind load was not considered in the report. [PH] suggested that spooler located in the cabine, hence there is no wind load applied.” PH then forwarded that email to ABSG. 15 On 19 February 2008, ABSG issued three certificates, two of which are relevant to the present appeals: (a) a Structural Design Review certificate, and (b) a Mechanical and Electrical Design Review certificate. Both certificates stated that the “300-ton Wire Spooler Tower” was considered satisfactory provided that the calculations and drawings were adhered to and the workmanship was to the satisfaction of an ABS consulting surveyor. 16 On 10 April 2008, Trident took delivery of the RDU on behalf of Airtrust. The RDU was then mounted on the vessel Maersk Responder which thereafter set sail for Australian waters. On 20 May 2009, the hydraulic drive motor and gear assembly on Tower A of the RDU broke loose from its mounting and fell down mid-way during the umbilical laying operation involving the second reel. This was a catastrophic failure. In the course of repairing the RDU, and based on further investigations, Airtrust discovered a number of problems with its design and manufacture. It commenced a suit against PH for breach of the SPA. Airtrust averred that the RDU was not of merchantable quality, was not fit for the purpose which it was intended to be used, was not free from defects in design, manufacture, or workmanship, and did not meet the relevant industry standards and certifications. The decision in the court below 17 The trial before the Judge was bifurcated and dealt only with the issue of liability. The Judge found that PH had breached the SPA by not delivering an RDU that was of merchantable quality and fit for its purpose, and ordered damages to be assessed. There is no appeal against this finding. Fraud 18 The Judge also considered whether PH had been “reckless, dishonest or fraudulent” in the manner it had secured the ABSG certification for the RDU. 19 In this regard, Airtrust had pleaded in its statement of claim that: (a) PH did not obtain full and proper certification from ABSG because: (i) In its document submission to ABSG, PH did not provide ABSG with any calculations with respect to wind loading. Specifically, PH’s representatives instructed ABSG not to take wind loads into account in considering the adequacy of the design of the structure. (ii) The Mechanical and Electrical Design Review was not complete as PH did not instruct ABSG to conduct a mechanical design review, a hydraulic design review, a design review of the sub-frame assembly, or a design review of the components in the drive train. (b) The failure to obtain full certification was “fraudulent and/or deceptive conduct meant to mislead Airtrust”. (c) Further or alternatively, PH fraudulently, dishonestly, wilfully, and/or recklessly presented the certificates as amounting to full certification when they were not. (d) PH procured the ABSG certificates through misrepresentations. ABSG provided the Structural Design certificate based on the STAAD.Pro data provided by PH on 18 February 2008, but PH, in its data, falsely modelled that part of the unit which ABSG had been asked to certify. In particular: (i) the joints at the transverse and longitudinal struts of the towers, as well as the joints at the tower base clamps, were modelled as ball joints allowing for rotational movement, when they were in fact fixed joints; and (ii) the reel of the unit was modelled as being fixed in the X, Y, and Z directions and unable to rotate about the X axis when in fact it was able to slide along the X axis. (e) The purpose of the false modelling was to ensure that the unity checks for all beams would be less than 1.0. Airtrust would not have accepted the RDU had it known of the false modelling in the STAAD.Pro data; nor would ABSG have certified the RDU design. 20 We pause to observe that Airtrust did not bring a claim based on the tort of deceit. In its statement of claim, it only claimed, in its prayers for relief, damages to be assessed for breach of contract. Its case as developed in closing submissions, however, was that PH had been fraudulent in the way it had performed the SPA, and that this was an egregious breach of contract which justified an award of punitive damages in addition to damages to compensate it for breach of the SPA. 21 The Judge found that PH had, in four ways, deliberately and dishonestly misled ABSG into issuing its certification. (a) First, it had fraudulently stipulated the absence of wind load on the RDU in the submission of its STAAD.Pro programme to ABSG for the design review (see the Judgment at [246]). It had fraudulently misled ABSG into thinking that wind load need not be considered because the RDU would be located in an enclosed cabin in the ship (at [229]). (b) Second, in submitting its STAAD.Pro data to ABSG, it had fraudulently modelled the RDU reel as fixed in the X direction when it was not. (c) Third, it had, in the STAAD.Pro data, modelled the bolted joints of the RDU as ball joints. (d) Fourth, it had misled Airtrust into thinking it had obtained full ABSG certification when the certification was in fact only partial (at [248]). The Judge rejected PH’s defence that the certification it obtained from ABSG should be considered a full certification since it had included information relevant to mechanical design review in the Confidential Bundle submitted to ABSG (at [249]). The Judge was of the view that the “paucity” of the design information on the hydraulic system sent to ABSG demonstrated that PH had never intended to ask and never asked ABSG to perform a full mechanical review of the hydraulic system as part of its “full certification” for the RDU (at [253]). Punitive Damages 22 The Judge noted that this court had, in MFM Restaurants Pte Ltd and another v Fish & Co Restaurants Pte Ltd and another appeal [2011] 1 SLR 150 (“MFM Restaurants”) at [53], left open the possibility of punitive damages being awarded in contract law. Endorsing the approach in Canadian authorities cited by Airtrust, he held that a court could, in an “exceptional case”, award punitive damages for breach of contract if the defendant’s conduct in breaching the contract had been “so highly reprehensible, shocking or outrageous that the court finds it necessary to condemn and deter such conduct by imposing punitive damages” (see the Judgment at [264]). In this case, he judged that PH’s conduct had been “sufficiently outrageous and reprehensible” to call for the imposition of punitive damages; what was done was “blatantly irresponsible designing, engineering and manufacturing” (at [272]). Hence, the Judge awarded punitive damages to be assessed (at [275]). In reaching this conclusion, he pointed to six examples of how PH’s conduct had been outrageous and reprehensible: (a) It had asked a junior engineer to modify an existing RDU design without performing comprehensible engineering calculations to ensure that the design of the RDU would meet the requisite specifications and user requirements (at [266]). (b) It had acted “dishonestly and fraudulently” in omitting to model fatigue and wind load so that its false model would not fail ABSG’s unity checks (at [267]). (c) Despite knowing that only partial certification had been obtained from ABSG, it had misled Airtrust into believing at the time of delivery of the RDU that full certification had been obtained (at [268]). (d) It had installed gears, bearings, and bolts without having properly checked or tested them at the work sites. It had used a lower grade of bearings and bolts than that specified in the construction and engineering drawings respectively (at [269]). (e) It had demonstrated “blatant disregard” of its obligation to construct an RDU that was satisfactory in quality and fit for its purpose, as could be seen from its fraudulent input into the STAAD.Pro data, deliberate deviations from design drawings during construction, and installation of components without checking on quality (at [270]). (f) Its workmanship and manufacturing quality had been poor (at [271]). It can be seen that points (b), (c) and (e) above are based on the findings of fraud that the Judge had earlier made. In other words, it was partly on account of PH’s fraudulent conduct, and partly on account of instances of “irresponsible” designing, engineering, and manufacturing, that the Judge imposed an award of punitive damages Limitation of liability 23 The parties agreed that, were the Judge to find fraudulent conduct on PH’s part, the limitation of liability clause in the SPA (ie, Cl 25) would not apply. Thus, the Judge did not need to consider the scope of Cl 25, which excludes PH’s liability for “consequential or indirect losses” which include, but are not limited to, loss of profits. 24 Whether or not Cl 25 can be construed to limit the extent of damages payable by PH to Airtrust is, however, a point which was included in PH’s Notice of Appeal. Specifically, the question we are asked to address is whether it excludes the following two heads of loss and damage which Airtrust had pleaded in its statement of claim: (b) Damages for loss of profits from the rent which the RDU would have made over the period of normal utility of such a unit if it had been properly designed and manufactured. For instance, [Airtrust] would have been able to take up the Subsea 7 Job and jobs offering equivalent rental rates. … (f) [Airtrust] would have used the profits from the rental of a properly designed and manufactured RDU to invest in other RDUs and lost the opportunity of earning profits from the rental of RDUs acquired from the earnings of the first and other RDUs. Indemnity costs 25 In the hearing on costs, Airtrust sought an order of indemnity costs on the basis that PH: (a) had failed to give full and proper disclosure of relevant documents; (b) had suppressed important facts which came to light only in the midst of trial; (c) had sought to keep away from trial individuals connected with the design of the RDU; and (d) had consistently maintained plainly unsustainable arguments despite contradictory evidence. 26 In the Costs Judgment, the Judge held that none of these four factors justified an award of indemnity costs against PH. He found (at [54] of the Costs Judgment) that PH did not “display such a degree of unreasonable or improper conduct” as to warrant a departure from the usual basis for awarding costs. The parties’ submissions PH’s submissions 27 We will first set out PH’s arguments in CA 234 with regard to Issues 1 to 3 as well as its response to the arguments with regard to Issue 4 raised by Airtrust in CA 96, before setting out Airtrust’s response to PH’s arguments with regard to Issues 1 to 3 in CA 234 and its arguments with regard to Issue 4 in CA 96. 28 PH’s contentions, in brief, are as follows: (a) The Judge made erroneous findings on fraud because he had (i) drawn inferences contrary to the primary and objective evidence; and (ii) made findings on allegations of fraud which Airtrust had belatedly included in the third amended statement of claim after the trial had concluded. (b) Punitive damages should not be allowed in contractual claims as there is a greater need for certainty in commercial contracts; in any event, such damages should not be awarded in the present case because there is no objective evidence to indicate that PH had acted in bad faith, or conducted itself in a manner that was especially profit seeking or calculated to injure Airtrust. (c) Cl 25 applies in this case to exclude two heads of claim pleaded by Airtrust (reproduced above at [24] above). (d) On costs, Airtrust is rehashing its submissions in the court below, which were duly considered and rejected by the Judge; it has not demonstrated how the Judge had misdirected himself with regard to legal principle or had made a decision which was plainly wrong. Airtrust’s submissions 29 Airtrust’s response to the submissions made by PH above is as follows: (a) Fraud has always been in issue in these proceedings from its first amendment to its statement of claim, and the Judge’s findings of fraud were amply supported by the evidence. (b) There is a cogent case for recognising the availability of punitive damages in contract. Since punitive damages can be awarded in tort claims, it would be “inconsistent to exclude them from contract claims”, especially since torts and breaches of contracts are both civil wrongs requiring the defendant to pay damages for infringing the plaintiff’s rights, the only difference being the source of those rights. (c) Airtrust’s claims based on its inability to earn profits for renting out the RDU were direct losses, not indirect and consequential losses, and are therefore not excluded by Cl 25. Read as a whole, Cl 25 is only intended to qualify the extent to which the parties must indemnify each other against third-party claims; it does not cover losses arising from a breach of contract. (d) There was no doubt that PH’s unreasonable conduct protracted the trial, making this an appropriate case for indemnity costs to be awarded. The issues 30 The issues have been set out at the outset of this judgment. To recapitulate, Issue 1 is whether PH was guilty of reckless, dishonest and/or fraudulent conduct. Issue 2 is whether or not this court will recognise the award of punitive damages for a breach of contract per se (ie, a breach of contract without any other concurrent liability, for example, in tort). Issue 3 is whether Cl 25 in the Sale and Purchase Agreement between the parties could be construed as limiting the extent of the damages payable by PH to Airtrust. Finally, Issue 4 is whether Airtrust is to be awarded the costs of the action on a standard basis or (as it sought) on an indemnity basis instead. 31 Let us now turn to consider each of these issues. Our decision Issue 1 – Fraud 32 It is undisputed that there is fraud when one makes a false representation (a) knowingly, (b) without belief in its truth, or (c) recklessly, careless whether it be true or false (see the classic House of Lords decision in Derry v Peek (1889) 14 App Cas 337 at 374, cited by this court in Wee Chiaw Sek Anna v Ng Li-Ann Genevieve (sole executrix of the estate of Ng Hock Seng, deceased) and another [2013] 3 SLR 801 (“Anna Wee”) at [32]). Negligence, however gross, is not fraud (see Anna Wee at [35]). In this regard, the representor’s subjective belief is crucial. The court cannot substitute its own view. Even if the representor’s statement was objectively unreasonable, that would not make it fraudulent if the representor honestly believed what it was representing (see Anna Wee at [37]). To that end, irrational belief is not fraud unless it is so incredible or unreasonable as to lead to an inference of an absence of honest belief (see Anna Wee at [40]). In addition, it must be kept in mind that the burden lies on the party asserting fraud to prove it on the balance of probabilities. Either there is fraud or there is not; there is no room for a finding that it might have happened. Because of the serious implications of fraud, cogent evidence is required before a court will be satisfied that the allegation of fraud is established (see the decision of this court in Alwie Handoyo v Tjong Very Sumito and another and another appeal [2013] 4 SLR 308 at [161]). 33 We deal first with PH’s preliminary objection that Airtrust’s allegations of fraud had been introduced “belatedly” by way of an amendment to the pleadings after the conclusion of the trial, which caused it prejudice. 34 We are of the view that this objection is unfounded. Airtrust had pleaded in its first amended statement of claim that PH’s failure to obtain a full ABSG certification amounted to fraudulent and/or deceptive conduct. Therefore, that part of Airtrust’s case on fraudulent conduct was always in issue. It is true that, over the course of the trial, its case expanded to include Dr Liu’s alleged misrepresentations as well as PH’s failure to undertake a hydraulic design review, but that was because these matters only came to light during the trial itself. In particular, Airtrust was not even aware of the existence of Dr Liu until the cross-examination of one of PH’s witnesses. As the Judge noted, with “fresh evidence emerging” as the trial progressed, the parties felt that amendments to their pleadings would be necessary, but decided to wait till the end of trial to consolidate their intended amendments (see the Judgment at [7]). The Judge was satisfied that no prejudice would be occasioned to either party by allowing the amendments. Furthermore, PH did not object to the allegations of fraudulent conduct in the amended statement of claim which Airtrust filed after the trial; it in fact filed its own amended defence to deal with these allegations. Thus, we can see no basis for PH’s objection that it was prejudiced by the allegations of fraud made belatedly. 35 We will now address the four instances of fraudulent conduct which the Judge had found on PH’s part (see above at [21]) and which justified, in his view, the imposition of punitive damages. In our view, and with respect to the Judge, we do not think that any of these findings were justified by the evidence. Let us elaborate. Misrepresentation that no wind load was to be considered 36 The Judge found that both Dr Liu and PH had deliberately and dishonestly misrepresented to Ms Devi, who had conducted the structural design review on behalf of ABSG, that it was unnecessary to consider wind load, and that there was dishonesty because Dr Liu and PH had represented that the reel (or spooler) was to be housed within an enclosed cabin in the ship despite knowing it was not (see the Judgment at [229]). The primary evidence on which the Judge based this conclusion was the email from Dr Liu to PH on 18 February 2008 which was eventually forwarded to Ms Devi. The Judge rejected PH’s explanation that the omission to consider the wind load was due to a “miscommunication” between Mr Lei and Dr Liu (see the Judgment at [228]). He found that they were both aware that wind loads had to be omitted in order to avoid a failure of the unity checks. 37 The evidence on this point is not entirely satisfactory but three things are clear: (a) PH does not deny that it did not consider wind load in the design process; (b) PH agrees that it should have checked for wind load, as its general manager, who was one of the witnesses, admitted to the Judge; and (c) PH accepts that Dr Liu’s statement about the RDU being located in the cabin was “obviously erroneous”. Mr Lei himself acknowledged that it was “not possible” for the spooler to be in the cabin. He could only say that PH telling ABS that the spooler would be in the cabin arose from a “misunderstanding”, and had no answer when asked by the Judge whether, in his professional opinion, wind load needed to be considered given that the RDU would be on the deck of a ship at open sea. 38 We do not think, however, that these problems with PH’s evidence necessarily lead to the inference that it was fraudulent. With respect, there are a number of difficulties with the Judge’s finding. 39 First, it is not clear if there was a false representation to begin with. The false representation, as pleaded, was that PH had “instructed [ABS] not to take wind loads into account in considering the adequacy of the structure” [emphasis added]. But Dr Liu’s response merely explains why he did not consider wind load; it does not purport to instruct Ms Devi to similarly disregard this factor. 40 Second, it is not clear if ABSG was misled by Dr Liu’s email into not considering wind load. Ms Devi explained to the Judge that environmental conditions, in particular inertia loads generated by wind, were already considered in the certification process. She added that, in asking Dr Liu to “clarify”, what she had meant to ask was whether Dr Liu specifically wanted her to also consider the wind velocity during transit. Certain clients would ask specifically that this factor be taken into account. Though there is no dispute that Dr Liu’s email was indeed sent to ABSG and Ms Devi, there is no evidence of the actual email from PH forwarding Dr Liu’s response to ABSG or the subsequent correspondence between the parties. Hence, it is not clear if Ms Devi noticed or had any questions about Dr Liu’s statement that the spooler was to be located in the cabin. 41 Third, there is insufficient evidence to support the finding that PH had a motive for fraud, and, consequently, insufficient evidence to support the Judge’s inference that Dr Liu and PH had knowingly made a false representation. It was clear in the Judge’s mind that wind load had to be omitted in order to avoid a failure of the unity checks. That was ostensibly the motive for Dr Liu and PH to deliberately misrepresent to Ms Devi that wind load need not be considered because the spooler would be located in a cabin. However, no evidence has been brought to our attention on whether it was technically true that the wind load, if considered, would have caused the unity checks to fail. From the Judgment, it does not appear that any such evidence was produced to the Judge either. In other words, there is no evidence supporting the factual foundation on which the inference of dishonesty rests. It is therefore difficult, with respect, to see how the Judge then drew the inference that Dr Liu and PH knew that the inclusion of wind load would harm their prospects of obtaining the ABS certification, and, consequently, that the misrepresentation was for the express and dishonest purpose of achieving a pass on the unity checks. Inaccurate modelling of fixed joints as ball joints 42 We now turn to the inaccurate modelling of the joints in the STAAD.Pro files. 43 As mentioned, it was not disputed that Dr Liu’s preliminary analysis on 26 October 2007 demonstrated that some beams did not pass the unity check. The Final Report he sent Mr Lei on 7 November 2007 indicated that the structural design passed the unity check. Eight beams were modelled as fixed joints instead of ball joints – it is accepted that the software only allows one to model a beam as either one of these two types of joints. The question is whether these changes meant that Dr Liu, and therefore PH, had made a false representation. The Judge thought so and arrived at his conclusion in the following way: (a) Dr Liu and PH “must have discussed at some length on how to get the RDU structure to pass the STAAD.Pro analysis and the unity checks with minimal change to the actual structural design, as any major changes would probably involve additional cost and delay for [PH]” (see the Judgment at [233]). He did acknowledge however that that there were no records of such discussions (at [231]). (b) The Judge was unconvinced by Dr Liu’s explanations for modelling the RDU with ball rather than fixed joints. In the Judge’s view, there was “clearly no technical justification to treat them as completely ball joints” (at [236]). Given that a structural engineer of Dr Liu’s experience was making such an “absurd engineering assumption”, the Judge was “driven to conclude” that this shift in analysis was due to “a deliberate and dishonest effort to try and get a pass for the unity checks” without considering the true nature of the relevant joints (at [238]). (c) The Judge proceeded to find that “[PH] and [Dr Liu] had a clear motive to achieve the unity check of 1 for every beam in the RDU structure”. They knew the RDU was to be constructed with joints that were nothing like ball joints. PH never had any intention to construct ball joints for the RDU and Dr Liu knew that. Therefore, both of them had been “less than honest” in “knowingly setting the wrong boundary conditions” so that the RDU could pass the unity checks (at [239]). (d) ABSG was misled into issuing the certification based on the inaccurate STAAD.Pro data submitted by PH. The ABSG certificate issued for an RDU modelled as having ball joints was inaccurate because the RDU was built without any ball joints (at [240]). 44 As a preliminary matter, we agree with Airtrust that PH cannot rely on the assertion that the fraud was by Dr Liu to excuse itself from liability. As a matter of law, a principal is liable for the false representation made by its agent even if it reaches an innocent party through the innocent principal (see the House of Lords decision of S Pearson & Son Ltd v Lord Mayor of Dublin [1907] AC 351). Having outsourced the STAAD.Pro analysis to Dr Liu, PH would be liable for any false representation that he had made – the false representations here being the changes in the STAAD.Pro data. 45 The Judge was, in our view, correct on the first point: given that Dr Liu was engaged specifically for the purpose of doing the STAAD.Pro analysis, it was not unreasonable to assume that there must have been discussions between Dr Liu and PH (in particular, Mr Lei) about how to get the RDU structure to pass the unity checks in the STAAD.Pro analysis. 46 However, with respect, we disagree with the second point made by the Judge (and thus with the third and fourth points, which flow from the second). There is no cogent evidence that the shift in the STAAD.Pro analysis was due to a deliberate and dishonest effort to get a pass for the unity checks. The Judge found that Dr Liu had no technical justification for using a ball joint to model the RDU’s bolt joints. There is no need to question whether that finding was technically correct since PH did not attempt to defend Dr Liu’s approach of modelling the joints as ball joints. However, although the Judge disagreed with Dr Liu’s technical assumptions, it was not appropriate in our view to conclude that Dr Liu was dishonest. Even if Dr Liu had modelled the bolt joints as fixed joints (as the Judge thought should have been done), this would have been, at best, a closer approximation of the RDU’s bolt joints. It would still not have been a fully accurate model. Dr Liu testified that he had to “simplify” the RDU’s bolts in STAAD.Pro because he could not “model the bolt in the STAAD.Pro programme”. He could only choose to model each joint as either a fixed or a ball joint. The Judge accepted this, acknowledging that the RDU’s bolt joints were “somewhere between a ball joint and a fixed joint”. The Judge in fact agreed with Dr Liu that it was difficult to get an “exact simulation” of a bolt joint. The Judge even appreciated Dr Liu’s explanation that he believed that the tower was a “weak structure” which meant that the joint does not function “like a bolt; it functions … like a more flexible thing” and is therefore closer to a ball joint. The Judge may have found that Dr Liu’s assumptions were not borne out by the literature he cited. But that does not necessarily mean that Dr Liu knowingly made a false statement. It must be borne in mind that Dr Liu’s STAAD.Pro data would not have been entirely accurate regardless of whether he had chosen to model the RDU’s bolt joints as either fixed or ball joints. 47 Nor do we find that Dr Liu was reckless as to the truth of the representations he made through the STAAD.Pro data. Such recklessness is found if a person “makes a representation, conscious of a risk that it may be false, but who is indifferent to that risk” [emphasis added] (see the Singapore High Court decision of Chu Said Thong and another v Vision Law LLC [2014] 4 SLR 375 at [124]). We do not think Dr Liu was conscious of such a risk. He had had his reasons for choosing the ball joint model. As mentioned above, he thought that the RDU tower was a weak structure and that the joints were therefore closer to ball rather than fixed joints. He also believed that his conclusions were supported by research. He relied on “some reports” to arrive at his decision on how to model the bolt joints. He told the Judge, “[i]n many reports, they assume [the bolt joint] just as the [fixed] joint”, and that this was “accepted in the industry field”. He candidly admitted that he did not do any calculations to confirm that it was appropriate to model the RDU’s joints as ball joints, contrary to what the Judge would have expected. However, in our view, that was at most negligent behaviour, or even grossly negligent behaviour, which did not amount to fraud. 48 In the circumstances, we do not think that Dr Liu’s change in analysis effected by modelling the bolted joints of the RDU as ball joints was a deliberate and dishonest effort to get a pass for the unity checks. Inaccurate modelling of RDU reel as fixed in the X Direction 49 In so far as the inaccurate modelling of the RDU reel as being unable to slide in the X Direction is concerned, the Judge said that had this been the only mistake (ie, if PH had not also intentionally excluded wind load or intentionally modelled fixed joints as ball joints) he might have given Dr Liu the benefit of the doubt, and found this mistake to be attributable to gross negligence (see the Judgment at [245]). In other words, the Judge reasoned that the fact of the other two mistakes being intentional lent support to the mistake made in modelling the RDU reel being intentional and, in addition, fraudulent (at [246]). 50 If it is accepted that the exclusion of wind load from the design process and the modelling of the RDU’s joints as ball joints were not fraudulent acts, then the basis for the Judge’s finding on this point is substantially diminished. As the Judge expressly acknowledged (see the Judgment at [245]), the mistake made in the modelling of the RDU reel, when taken in isolation, might have been due to gross negligence on the part of Dr Liu. It does not necessarily lead to the inference that there was fraud. Failure to obtain full certification of the RDU 51 Finally, we turn to the issue of whether PH fraudulently failed to obtained full certification of the RDU. The Judge found that: (a) PH had not instructed ABSG to carry out a mechanical design review (see the Judgment at [248]). (b) PH had never intended to ask, and never asked, ABSG to perform a full mechanical review of the RDU’s hydraulic system. PH had therefore misled Airtrust into believing that it had obtained full certification from ABSG, when it knew as a fact that ABSG had not provided full certification for the RDU (at [253]). 52 In so far as the mechanical design review was concerned, the Judge based his finding on the totality of the following circumstances (see the Judgment at [248]): (a) PH did not instruct ABS to carry out a mechanical design review, (b) mechanical design review was not included in scope of work in ABSG’s quote, (c) PH should have realized that Mr Ravi Chandran, the lead electrical engineer of ABSG, with whom they had worked for many years, was not qualified to carry out a mechanical design review, and (d) there should have been a separate invoice for mechanical design but no such invoice was produced. In our view, these points do not, with respect, unequivocally support a finding of fraud. 53 First, it is not entirely true that PH did not instruct ABS to carry out a mechanical design review. When Mr Gan, the project engineer from PH at the material time, first emailed ABS (before ABSG had been engaged) about the certification process, Mr Gan specifically stated that PH had received a project from Airtrust requiring “ABS full classification on design review and approval” [emphasis added]. The subject title of the email also reads “ABS Full Classification pertaining to Design Review & Inspection – Trident Australasia 300 reel drive unit” [emphasis added]. 54 Second, although the quote from ABS to Mr Gan on 2 January 2008 only demonstrated that ABSG’s scope of work related to design review of the spooler tower and power pack (at a lump sum cost of US$2,400), and design review of the submitted electrical items (at a lump sum cost of US$1,500), and seems to make no mention of mechanical design review, this is contradicted by the wording of the invoice dated on 20 March 2008. The invoice listed a charge of US$2,400 for “Structural & Mechanical Design Review” and US$1,500 for electrical design review. This gives cause to believe that ABSG may also have contemplated that the “design review of the spooler tower and power pack” described in the quote from Mr Gan would include both structural and mechanical design review. It is worth noting that the invoice came after Mr Chandran’s letter dated 19 February 2008 to Mr Gan, captioned “Design Review for 300T on Wire Spooler Towers – Mechanical and Electrical Design review” [emphasis added]. 55 Third, the Judge’s point was that PH must have known that Mr Chandran could not sign off, on the certification letter he sent, as having performed the mechanical review. That, however, does not necessarily mean PH knew that ABSG had not conducted the mechanical review. It is possible that ABSG had authorized Mr Chandran, for the sake of convenience, to sign off on the certification letter on behalf of the mechanical engineer, as was previously done with an RDU made for a previous client, Acergy. In any event, there was no exploration at trial of whether PH knew that Mr Chandran could not sign off on the mechanical review. 56 Fourth, the Judge’s assumption that there must have been a separate invoice for mechanical review was presumably premised on Mr Chandran’s explanation at trial that there should be a separate letter from a mechanical engineer to certify that the mechanical review has been done, and therefore a different “costing” (presumably he meant ”invoice”). Mr Chandran then explained that their invoice was only for structural and electrical design review. But the invoice ABSG issued was for structural and mechanical review, as highlighted above at [54]. No evidence was put before the Judge to suggest that it was not possible for electrical and mechanical review to be invoiced together. 57 In so far as a review of the hydraulic system was concerned, the Judge inferred from the paucity of the design information on the hydraulic system sent to ABSG that PH had never intended to ask and never asked ABSG to perform a full mechanical review of the RDU’s hydraulic system (see the Judgment at [252]). However, this was not a situation in which the Confidential Bundle did not contain any documents pertaining to the hydraulic system design; it did contain documents such as the hydraulic schematic drawing, the 430 Hp Engine Hydraulic Power Unit General Arrangement drawing, and the system calculation and hydraulic motors specification sheet. The Judge formed his own view of what documents should have been included but were in fact absent from the bundle (see the Judgment at [251] and [252]). Although he was entitled to do that, it is a much further step to draw the inference that PH had no intention for ABSG to perform a hydraulic system review, or that they had fraudulently chosen to omit documents in the Confidential Bundle to ABSG in order to prevent the hydraulic system from being mechanically certified. There is no evidence to suggest that these omissions were deliberate, let alone fraudulent. 58 In any event, the lack of full documentation can be explained by more innocuous reasons. In PH’s submission, it had believed that it had given sufficient detail and that, if it were not so, this would have been raised by ABSG. The classification society was best placed to know what documents were necessary to perform the reviews. In other words, PH was prepared to give more documents if ABSG required additional documents for the purpose of carrying out a full certification. This was in fact done on one occasion, albeit in respect of a different review pertaining to structural design. On that occasion, Ms Devi requested further documents such as those relating to skidding assembly and jacking frame assembly, and PH duly complied with the request. It is therefore insufficient to rely on the mere fact that documents were missing from the Confidential Bundle to conclude that their omission was fraudulent. 59 In the circumstances, we do not think that the Judge was correct in concluding that PH must have known that ABSG had not in fact performed any mechanical design review or hydraulic system review. With respect, given the number of references in the invoices and emails to “full certification” and “mechanical review”, and given the inclusion of some documents on hydraulic review in the Confidential Bundle, there was a possibility that PH had indeed been mistaken. It is possible, in our view, that PH assumed that ABSG would perform these reviews without actually verifying that they had been done. However, this lapse appears to us to be, at most, negligent or grossly negligent behaviour. Certainly, there was no cogent evidence that PH fraudulently failed to obtained full certification of the RDU. Conclusion on the findings of fraud 60 In our view, the Judge’s findings of fraud should, with respect, be set aside because there is no cogent evidence that PH had acted fraudulently. Looking at the evidence in the round, the most that could be said was that PH had been grossly negligent. It is clear that PH was supposed to obtain full certification from ABSG and that the certification which was issued was not a “full” one. We do not wish to speculate unnecessarily about what went wrong, save to say that it appears to us that the mistakes in the data or documents PH submitted to ABSG for the certification process, which the Judge found to have been motivated by fraud, could simply have been due to oversight. 61 We find in favour of PH on this ground of appeal. We now turn to consider Issue 2. Issue 2 – Punitive Damages Introduction 62 We take as our starting point the well-accepted proposition that “the purpose of the law of contract is not to punish wrongdoing but to satisfy the expectations of the party entitled to performance” (see the House of Lords decision of Co-operative Insurance Society v Argyll Stores (Holdings) Ltd [1998] 1 AC 1 at 15, per Lord Hoffmann). That is why the general aim of damages for breach of contract is to compensate: the plaintiff is to be placed, as far as a payment of money allows, in the same position as if the contract had been performed (see the oft-cited English decision of Robinson v Harman (1849) 1 Exch 850 at 855). Besides orthodox compensatory damages, which are assessed by reference to the plaintiff’s pecuniary loss, there are other measures by which contractual damages can be assessed, which we will return to later. But it suffices to note that these other remedial options serve also to protect the plaintiff’s interest in contractual performance and remain primarily compensatory in purpose (see below at [79]). 63 By contrast, punitive damages, also referred to as exemplary damages, are awarded against a wrongdoer as a form of punishment (see Chitty on Contracts vol 1 (Hugh Beale gen ed) (Sweet & Maxwell, 32nd Ed, 2015) (“Chitty on Contracts”) at para 26-044). The primary aim of such damages is to punish the wrongdoer and deter similar behaviour by the wrongdoer and others in the future. Being a departure from the general compensatory function of damages for breach of contract, the recognition of the availability of punitive damages in contract law is not a step to be taken lightly. As alluded to at the outset of this judgment, whether or not a court should award punitive damages is an issue which involves a number of complex theoretical and practical considerations even though it appears to be simple. Before proceeding to consider the issue proper, a few preliminary (albeit not unimportant) observations are in order. Preliminary observations 64 The first observation is that the availability of punitive damages is recognised in tort; therefore, where there is concurrent liability in both contract and tort, the court may award punitive damages in appropriate circumstances. It is well-established that “where the claimant has a cause of action both in tort and for breach of contract, he may be able to recover punitive damages by framing the claim in tort” (see Edwin Peel, Treitel: The Law of Contract (Sweet & Maxwell, 14th Ed, 2015) (“Treitel”) at para 20-019). It is also now established, following our recent decision in ACB v Thomson Medical Pte Ltd and others [2017] SGCA 20 at [176], that the availability of punitive damages in the law of tort is no longer restricted to the three categories set out in Rookes v Barnard [1964] AC 1129 – that is, where there exists oppressive, arbitrary or unconstitutional action by the servants of the government, where the defendant’s conduct was calculated to make a profit for himself which may well exceed the compensation payable to the plaintiff, and where punitive damages are expressly authorised by legislation. In this judgment, therefore, we are concerned only with whether punitive damages can be awarded for breach of contract where the plaintiff has not chosen (or is not entitled) to bring a concurrent claim in tort (and it is in this sense that we will hereafter refer to punitive damages “for breach of contract”). 65 The second observation, which has already been hinted at, is that even if punitive damages cannot be awarded for breach of contract, there are alternative bases for the award of damages which are, to some extent, deterrent in function; these can fill the remedial gap which, it has been argued, justifies the award of punitive damages. We deal with this point below from [77] onwards. 66 The third observation – which is, in fact, closely related to the outcome of the present appeal – is whether, given our finding in relation to Issue 1 that the Defendant was not guilty of any fraud but was, taking Airtrust’s case at its highest, only guilty of gross negligence, there is a need to consider Issue 2. In our view, there is still a need to do so because, assuming for the moment that punitive damages can be awarded for breach of contract (such as the present), it does not necessarily follow that such damages will not be awarded on the facts of this particular case simply because there had not been any fraud as such. Put another way, this court would, assuming it was prepared in principle to award punitive damages for breach of contract, have to ascertain under what circumstances such an award might be made. It therefore follows that the prior question as to whether this court would, in principle, be prepared to award punitive damages would need to be decided first. 67 We turn now to consider the substantive issue – commencing first with the arguments against recognising punitive damages for breach of contract, before turning to examine the arguments in favour of doing so. Arguments against an award of punitive damages in a purely contractual context (1) The distinction between contract and tort 68 The first argument against the availability of punitive damages is that allowing the courts to punish a party who has breached a contract sits uneasily with the concept of a contract as an obligation voluntarily undertaken. By contrast, punitive damages are less objectionable in tort. To understand why, it is necessary to start with a basic distinction between contract and tort, which are different sources of private law obligations. At the broadest level of generality, it can be said that obligations in contract arise from a voluntary and binding agreement between the parties whilst obligations in tort generally do not (absent concurrent liability in both contract and tort). Putting the point in another way, “contractual obligations are voluntary and particular to the parties, whereas liability in tort is imposed by law as a matter of policy and affects persons generally” (see Chitty on Contracts at para 1-145). This distinction has many significant implications where remedies are concerned. 69 First, and most importantly, although damages for the breach of a contract and for the commission of a tort are both meant to compensate the plaintiff, they nevertheless serve different remedial purposes. Compensation for a breach of contract is meant to put the aggrieved party in the same situation as if the contract had been performed – what is being protected is usually referred to as the “expectation interest” or the “loss of a bargain”. On the other hand, compensation in tort is meant to restore the aggrieved party to the position it would have been in had the tort not been committed. 70 Second, in so far as remoteness of damage is concerned, the relevant rules and principles in contract are quite different from those in tort. The rules and principles in contract centre on the concept of “reasonable contemplation” whereas those in tort centre, instead, on the concept of “reasonable foreseeability”, the latter being broader and more generous than the former (see the decision of this court in Robertson Quay Investment Pte Ltd v Steen Consultants Pte Ltd [2008] 2 SLR(R) 623 (“Robertson Quay”) at [71]−[76]). In Robertson Quay, we explained the contract-tort distinction in the following way: 75 The law of contract, put simply, is about agreement. It is true that there can be concurrent liability in contract and in tort, but, where there is no such concurrent liability, the law of tort clearly relates to civil wrongs that occur not as a result of a contractual relationship between the party that has suffered damage and the party that committed the tort(s) in question as such but, rather, despite the fact that both have hitherto been strangers to each other. This is a simple – yet profoundly important – starting point, which is also captured by the observations of both Lord Reid and Lord Upjohn in The Heron II at 386 and 422, respectively (reproduced above at [71]). 76 If there has, ex hypothesi, been agreement between the parties to a contract, it follows that they have already been afforded the opportunity to consider various matters thought to be relevant to their contractual relationship itself. This would, of course, include any matters relating to remedies in general and damages in particular. Indeed, the quintessential illustration of contractual provision in this regard is the liquidated damages clause, which purports to constitute a genuine pre-estimate of loss that might result from a breach of contract … [emphasis in original] 71 This (key) element of agreement in the contractual context militates against the availability of punitive damages. Since the parties to a contract have been afforded the opportunity to consider various matters that are relevant to the bargain that they entered into, and since they have decided between themselves what the terms of the contract should be, the courts ought to have but a minimal role in regulating their conduct without regard to their agreement. This is why, absent vitiating factors, courts must give effect to the bargain entered into between the parties – this is embodied in the concept of “sanctity of contract”. Indeed, even vitiating factors cannot – for the reason just mentioned – operate beyond the strict parameters which surround each of them and certainly cannot be applied in a liberal fashion lest the very raison d’être of the common law of contract be undermined. And where one contracting party fails to keep its end of the bargain, the court’s task is to ensure that the other contracting party is compensated for the loss of that bargain through an award of damages. That award of damages gives effect to the bargain struck by both parties and generally carries with it no disapproval (and, equally, no approval) of what the contract-breaker has done, regardless of whatever may have been the motive for breaching the contract. Indeed, the common law of contract has largely been “tolerant of a certain amount of self-interested behaviour” (see Sèlene Rowan, “Reflections on the Introduction of Punitive Damages for Breach of Contract” (2010) 30 OJLS 495 (“Rowan”) at p 504). 72 Looked at in this light, it would, in our view, be anomalous or even inappropriate for the court to regulate the contracting parties’ conduct by imposing an award of punitive damages on the party in breach by way of what is in effect an external standard. The standard is an external one because, with the award of such damages, the court goes further to signify its own outrage at the contract-breaker’s conduct, and to communicate its own view of what proper commercial behaviour should be. The court is no longer giving effect to the standard set by the contracting parties. Such an external standard may be said to be antithetical to the very nature and function of the law of contract in general and its remedial structure in particular (which is, in the main at least, to compensate, and not to punish). The concept of punishment connotes the (related) concept of deterrence and, looked at in this light, sits very uneasily with the concept of a contract which, as we have just noted, is a voluntary agreement entered into between willing parties who, ex hypothesi, would regulate their legal relationship themselves. Punishment and deterrence are quintessentially part of the legal landscape of the criminal law. 73 The strength of contract law’s commitment to the compensatory principle, and its aversion to concepts of punishment and deterrence, can also be seen in the rule against contractual penalties. As we mentioned in Robertson Quay (see the extract at [70] above), the liquidated damages clause is perhaps the quintessential example of how the contracting parties themselves regulate the consequences of their own default. Yet the sum of liquidated damages provided for must be a genuine pre-estimate of loss; the law does not allow them to impose a sum so extortionate as to threaten them to perform the contract (and thus discourage or deter them from breaking their promise). This is commonly referred to as the rule against penalty clauses. There is considerable force in the observations that this rule expresses the law’s “distaste for deterrent damages in contract” (see Neil Andrews, Malcolm Clarke, Andrew Tettenborn and Graham Virgo, Contractual Duties: Performance Breach, Termination and Remedies (Sweet & Maxwell, 2011) at para 20-024)) and “constitutes a resounding rejection of deterrence and punishment as acceptable aims in the law of contract” (see Rowan at p 508). And if the law does not allow the contracting parties themselves to stipulate such penalties or sanctions in advance, it seems anomalous that the law would then allow a judge to impose unforeseen penalties later on (see Matthew P Harrington, “‘A Lawless Science’: Punitive Damages for Breach of Contract in Canada” (2015) 69 Supreme Court Law Review 185 (“Harrington”) at p 210). The law would not only be imposing its own external standard on the contracting parties, as we have mentioned, but would be doing so even though this were not anticipated by either contracting party (and this does not cohere with the rule of remoteness in contract, which is that damages must be reasonably contemplated). 74 In contrast (and in so far as the private law of obligations is concerned), the law of tort affords far more latitude to the courts in regulating conduct between the parties (who could be but are not generally in a contractual relationship). And unlike the law of contract, the law of tort also imposes standards of normative behaviour between complete strangers. In the English Court of Appeal decision of Robinson v PE Jones (Contractors) Ltd [2011] 3 WLR 815, Jackson LJ observed as follows (at [79]): Contractual obligations are negotiated by the parties and then enforced by law because the performance of contracts is vital to the functioning of society. Tortious duties are imposed by law (without any need for agreement by the parties) because society demands certain standards of conduct. In that sense, there is greater room for the court to impose an external standard – its own view of how parties should conduct themselves towards each other. It is not surprising, therefore, that the availability of punitive damages is well-established within the law of tort because conduct which falls far short of the standards expected by society may in some situations need to be met with sanctions to signify that such conduct should not be tolerated. (Even then, it must be remembered that punitive damages in tort are not awarded except in exceptional circumstances.) The point is put by Prof Andrew Burrows in this way (see Andrew Burrows, “Dividing the Law of Obligations” in ch 1 of Understanding the Law of Obligations: Essays on Contract, Tort and Restitution (Hart Publishing, 1998) at p 13): … [B]eing based on a voluntary undertaking, the courts ought to tailor the remedy in contract to what was voluntarily undertaken and should therefore be reluctant to invoke non-compensatory remedies, such as punitive and restitutionary damages. In contrast, where the liability is purely imposed, as in tort, there need be no such reluctance. 75 By way of comparison, the law of contract stands on one end of the spectrum whilst the criminal law stands on the other – with the law of tort standing somewhere in between. To reject the availability of punitive damages in contract law while recognising it in tort law is not to concede that contract law pays no regard to morality or turns a blind eye to the behaviour of the contracting parties. It is, rather, to recognise that contract and tort encompass different causes of action which have different remedial goals. The plaintiff remains at liberty to plead the cause of action which it thinks most advantageous to it and most likely to return the remedial outcome it seeks (save for when a crime is committed, in which case the law will certainly intervene). 76 As we have seen, the distinction between contract and tort is not an arbitrary exercise in classification. On the contrary, it is one that has substantive implications in so far as the present appeal is concerned. In particular, this distinction underscores the reason why the concept of punishment is inapposite in the contractual context (compared to that in tort and, a fortiori, the criminal law). However, could it be argued that the residual power to award punitive damages for breach of contract is nevertheless necessary in light of the fact that there would otherwise exist a remedial gap? This was a point which was also dealt with by the learned Prof Lee in both her written as well as oral submissions to this court. It is to this particular issue that our attention now turns. It will suffice for the moment to say that this particular argument is, at best, neutral. Indeed, as we shall see, whilst the available alternatives might be qualitatively different from the concept of punitive damages, they more than fill any possible remedial gap. Thus, this argument becomes unpersuasive. (2) No remedial gap which only an award of punitive damages can fill 77 The argument based on a “remedial gap” is that it is necessary to have a residual discretion to award punitive damages in contract law because existing remedies are inadequate to punish and deter outrageous behaviour. 78 The argument appears to us an odd one because if we accept (as has been elaborated upon in the preceding part of this judgment) that the concept of punishment (and its related concept of deterrence) is inapposite in the context of the common law of contract, then there is, ex hypothesi, no gap to fill in the first place. 79 However, putting this difficulty to one side for the sake of argument, our view is that, as Prof Lee has perceptively pointed out, the court has a number of remedial options with punitive or deterrent effects that it could utilise (either as an alternative or in addition to the ‘traditional’ award of compensatory damages). Indeed, as we shall see in a moment, these remedial options serve also to protect the plaintiff’s interest in contractual performance and remain primarily compensatory in purpose, even if they may incidentally have a punitive or deterrent effect. 80 One alternative referred to by Prof Lee is what has now come to be known as “Wrotham Park damages” (based on the seminal decision by Brightman J in the English High Court in Wrotham Park Estate Co Ltd v Parkside Homes Ltd [1974] 1 WLR 798 (“Wrotham Park”)). Such damages are quantified not by reference to a plaintiff’s pecuniary loss (which may be nominal or difficult to quantify) but by reference to the sum of money which the plaintiff could have reasonably demanded in return for permitting the defendant to breach a restrictive covenant or other legal restriction. The legal requirements for the award of such damages – the defendant’s deliberate breach of contract for its own reward, the plaintiff’s difficulty in establishing financial loss, and the plaintiff’s interest in preventing the defendant’s profiting from a breach of contract (see the English Court of Appeal decision of One Step (Support) Ltd v Morris-Garner and another [2016] 3 WLR 1281 at [147]) – suggest that it has the effect of deterring deliberate breaches of contract. Ultimately, however, Wrotham Park damages are largely accepted to be compensatory in nature, although they are different, in substance, from a traditional award of compensatory damages (see also the justly famous, as well as widely cited, article by Robert J Sharpe and S M Waddams, “Damages for lost opportunity to bargain” (1982) 2 OJLS 290). 81 Another alternative is an award for account of profits for breach of contract, the genesis of which is the seminal House of Lords decision of Attorney General v Blake (Jonathan Cape Ltd, Third Party) [2001] AC 268 (“Blake”). The precise nature of the award of damages in Blake is, however, less clear. In our decision in MFM Restaurants, both Wrotham Park and Blake were referred to briefly as follows (at [54]−[55]; reference may also be made to the decision of this court in ACES System Development Pte Ltd v Yenty Lily (trading as Access International Services) [2013] 4 SLR 1317 at [53]): 54 The [decision in Blake] is less controversial [compared to whether punitive damages can be awarded in a purely contractual context], but no less problematic. Indeed, Blake itself generated – not surprisingly, perhaps – a learned body of legal literature of a magnitude that has been witnessed only occasionally (in this regard, the account by Prof Graham Virgo in his book, The Principles of the Law of Restitution (Oxford University Press, 2nd Ed, 2006) (“Virgo”) ch 17 furnishes an exceptionally clear and succinct overview of both the relevant case law as well as legal literature). Put very simply, the principles set out in Blake permit the court to award damages to the plaintiff (in a situation relating to the breach of a contract) on the basis of the gains or profits made by the defendant even though the plaintiff could not otherwise be awarded any damages based on traditional contractual principles (for example, because there has been no difference in value of the contractual subject matter and, hence, no justification for the award of expectation loss). Such damages would, however, be awarded only in exceptional cases. This category of damages has sometimes been termed as “restitutionary damages”, although the House in Blake preferred to classify such an award on the basis of an account of profits. 55 Apart from the various conceptual as well as (as referred to at the end of the preceding paragraph) terminological difficulties, there are (concurrent) practical difficulties as well in so far as the award of damages under the principles set out in Blake are concerned. In Blake, for example, in the leading judgment of Lord Nicholls of Birkenhead, the statement of principle (at 284–285) does not really furnish concrete guidance as to when the power to award such damages will arise. That the award of such damages is (as already noted in the preceding paragraph) exceptional still leaves the (very practical) issue as to the criteria which will enable the court to ascertain whether or not a given fact situation is indeed exceptional. Case law developments since Blake (see, for example the English Court of Appeal decision of Experience Hendrix LLC v PPX Enterprises Inc [2003] 1 All ER (Comm) 830 (noted in Pey-Woan Lee, “Responses to A Breach of Contract” [2003] LMCLQ 301 (“Lee”); Martin Graham, “Restitutionary Damages: The Anvil Struck” (2004) 120 LQR 26 (“Graham”); and David Campbell and Philip Wylie, “Ain’t No Telling (Which Circumstances are Exceptional)” (2003) 62 CLJ 605 (“Campbell and Wylie”)) have also suggested a possible alternative approach that is premised not on a restitutionary basis as such but, rather, on a compensatory one (centring on the much discussed decision by Brightman J in the English High Court decision of Wrotham Park Estate Co Ltd v Parkside Homes Ltd [1974] 1 WLR 798 (“Wrotham Park”) (see also Robert J Sharpe and S M Waddams, “Damages for lost opportunity to bargain” (1982) 2 OJLS 290); reference may also be made to the recent Privy Council decision (on appeal from the Court of Appeal of Jersey) of Pell Frischmann Engineering Limited v Bow Valley Iran Limited [2010] BLR 73 (especially at [46]–[54])). However, it would appear that the precise contours of this particular head of compensatory loss (as well as its relationship with the established head of expectation loss) have yet to be worked out fully (see, for example, Virgo (at pp 490–492) as well as Lee; Graham; and Campbell and Wylie). It might well be a difference of degree rather than kind, at least in so far as the former consists in an attempt to quantify compensatory damages by reference to the defendant’s gain (see Lee (at 302–303) and, by the same author, “A New Model of Contractual Compensation” [2006] LMCLQ 452 at 453; cf also Samuel Stoljar, “Restitutionary Relief for Breach of Contract” (1989) 2 JCL 1 at 3–4). To the extent, of course, that both heads are coincident, the head of damages will, in substance, be based on orthodox principles of the common law of contract (which centre on expectation loss). [emphasis in original] 82 To be sure, the two heads of damages referred to in the preceding paragraph are not of precisely the same nature as punitive damages. Although they are a departure from the traditional loss-based measure of damages, their primary purpose can still be said to be compensatory, in that they protect a plaintiff’s interest in contractual performance. Any punitive or deterrent effect they may have is arguably incidental. However, there is no gainsaying that they do comprise more than nominal damages and go beyond the orthodox compensatory measure assessed by reference to the plaintiff’s pecuniary loss. They can thus serve as remedial alternatives to punitive damages in ensuring that deliberate breaches of contract are met with just remedial responses. The argument that punitive damages are necessary to fill a remedial gap is, looked at in this light, neutral at best. 83 In this regard, there is a further – and third – possible head of damages that might serve as another alternative to an award of punitive damages. This is an award of damages to compensate the aggrieved party for mental distress resulting from the breach of contract. As the law now stands, it would appear that such damages can be recovered in situations where the purpose of the contract itself was to provide peace of mind or freedom from distress (see, for example, the Singapore High Court decision of Kay Swee Pin v Singapore Island Country Club [2008] SGHC 143 (“Kay Swee Pin”) – with the possibility that the scope of such an award might possibly be broadened in the future (see Kay Swee Pin, especially at [79]−[80]; and cf Andrew B L Phang and Goh Yihan, Contract Law in Singapore (Wolters Kluwer Law & Business, 2012), especially at paras 1539−1540). We think that, even on the narrower basis just referred to, awarding a plaintiff damages for mental distress caused by the defendant’s egregious conduct in breaching the contract would be more principled than awarding it punitive damages. Such an award would still, as a matter of principle, be consistent with the compensatory rationale of damages for breach of contract. As with Wrotham Park damages and the account of profits remedy, such damages may also have a punitive effect (which, as has been aptly observed, makes the distinction between compensation of the plaintiff and punishment of the defendant difficult to draw in practice: see Treitel at para 20-019). 84 In summary, there does not appear to be any remedial gap that justifies recognising the availability of punitive damages for breach of contract. The argument is that there would be a “gap” if the court did not have at its disposal a remedy to punish and deter those who breach their contracts in a deliberate or outrageous way. Our response is that there is no gap to begin with because the purpose of damages for breach of contract is not to punish. Even if there is such a gap, it can be filled by the alternative remedies mentioned above which target deliberate breaches of contract and arguably have punitive or deterrent effects. Given their punitive or deterrent effects, such remedies may already seem anomalous when compared with a “traditional” award of damages for breach of contract. However, being ultimately compensatory in nature, they can at least be measured by a clear and definable standard – that is, the value of the plaintiff’s contractual bargain. On the other hand, an award for punitive damages would appear to be measured (and thus constrained only) by a court’s sense of what conduct merits punishment. This would arguably be a difficult standard to define in advance. 85 This is an appropriate juncture to turn to the third argument against recognising punitive damages for breach of contract: the absence of clear criteria by which to determine when punitive damages should be awarded, and the consequent uncertainty this would lead to. (3) The absence of criteria and consequent uncertainty 86 Prof Lee argues that punitive damages ought not to be awarded in a purely commercial context because the concept of an “outrageous” breach – to take but one of many epithets used to express the kind of conduct deserving of punishment – is particularly elusive in this context where self-serving behaviour is an accepted facet of contracting norms. She argues therefore that extending the reach of punitive damages (to a purely contractual context) on the basis of such an uncertain concept would likely harm commercial stability. 87 Whilst not conclusive, this is, in our view, a powerful argument. Any argument to the contrary based on the fact that punitive damages are already awarded for “outrageous” conduct in the context of the law of tort loses its force in light of the fact that there are separate and distinct reasons why punitive damages ought to be awarded in tort but not in contract. These reasons have been canvassed earlier in this judgment (see above at [68]−[76]; see also below at [110]−[111]). 88 Turning to the argument proper, it is, as just mentioned, a powerful one because most parties enter into a contract with self or vested interest as a main (even paramount) consideration. Indeed, this is the basis for the theory of efficient breach – ie, that a contractual breach should be permitted if it would mean that limited resources can be more efficiently allocated; that a contractual remedy serves only to incentivise the efficient allocation of resources; and that an award of punitive damages is an economically inefficient response to a contractual breach. This was a point Prof Lee also dealt with ably in her written submissions (but which, given the inherent controversy surrounding it and the fact that its resolution is not necessary for the purposes of the present appeal, we will put to one side). 89 Even leaving aside the controversial theory of efficient breach, it is clear that contracting parties are – save in rare and truly exceptional circumstances (such as when there is a duty of good faith, with which we will deal with below at [131]) under no duty of altruism or, indeed, any duty to even have regard to the interests of the other contracting party in the first place. This is especially the case in the commercial context where profit is often the dominant motive. If this analysis, however, is accepted, it becomes very difficult, in our view, to identify specific (as well as workable) criteria for ascertaining when a contracting party’s conduct has crossed the line from self or vested interest into the realm of the “outrageous”. Indeed, the court must be especially careful in this particular regard simply because what is “outrageous” will vary from person to person. However, this is not to state that the court will countenance all improper conduct. Control mechanisms which regulate improper conduct by contracting parties are not unusual and are often embodied within the internal legal framework of the common law of contract itself. In particular (and as already mentioned above at [71]), there exists (by analogy) an established body of vitiating factors that operate in situations where injustice might otherwise occur as a result of specific conduct on the part of a contracting party. As also noted above (at [79]), there are (in the contractual sphere) alternative remedies that are more principled substitutes compared to the possible award of punitive damages. What is clear is that, taking all these factors into account, the argument presently considered is a powerful one, with the result that the case against the award of punitive damages in a purely contractual context is buttressed further. 90 This would be an appropriate juncture to turn from considerations of general principle to the specific case law itself. (4) The relevant case law 91 It is clear – even from a cursory survey of the relevant case law – that the authorities lean heavily against the award of punitive damages in a purely contractual context. In our view, this is not surprising (and might even be expected) given our analysis on general principles as set out above. The position in Singapore 92 The position in Singapore is – as yet – unclear. Not until the present decision has there been a need to take a firm view on whether punitive damages can be awarded. This court did consider the point in MFM Restaurants, where it observed in passing that whether punitive damages ought to be awarded in a contractual context was an “especially controversial” issue (at [53]). It did however suggest that there was an arguable case for the award of punitive damages: … Although there is an arguable case, in principle, for the award of punitive damages in contract law (see, for example, Ralph Cunnington, “Should punitive damages be part of the judicial arsenal in contract cases?” (2006) 26 Legal Studies 369 (“Cunnington”) and Pey-Woan Lee, “Contract Damages, Corrective Justice and Punishment” (2007) 70 MLR 887; though cf Solène Rowan, “Reflections on the Introduction of Punitive Damages for Breach of Contract” (2010) 30 OJLS 495), the case law is itself inconclusive (see, for example, Andrew Phang and Pey-Woan Lee, “Restitutionary and Exemplary Damages Revisited” (2003) 19 JCL 1 at 21–31 and Cunnington at 389−393). This is not surprising in view of the radical as well as draconian nature of punitive damages themselves. The position in Singapore is, not surprisingly, still an open one. In the Singapore High Court decision of CHS CPO GmbH v Vikas Goel [2005] 3 SLR(R) 202, it was observed thus (at [65]–[66]): 65 … Indeed, the rather limited circumstances under which exemplary damages will be granted under English (and, presumably, Singapore) law appears to be in a state of transition, even flux (compare, for example, the House of Lords decisions of Rookes v Barnard [1964] AC 1129 and Cassell & Co Ltd v Broome [1972] AC 1027 on the one hand with both the House of Lords decision of Kuddus v Chief Constable of Leicestershire Constabulary [2002] 2 AC 122 and the New Zealand Privy Council decision of A v Bottrill [2003] 1 AC 449 on the other; further reference may be made to the English Law Commission’s Report on Aggravated, Exemplary and Restitutionary Damages (Law Com No 247, 1997)). 66 There is the yet further issue as to whether or not exemplary damages could be awarded for cynical breaches of contract (see generally, for example, the Canadian Supreme Court decision of Whiten v Pilot Insurance Company (2002) 209 DLR (4th) 257). All these issues raise important questions of great import but fall outside the purview of the present decision. 93 A seemingly different position was taken in Sim Yong Teng v Singapore Swimming Club [2016] 2 SLR 489 (“Sim Yong Teng”), a subsequent decision of this court (released after the Judge’s decision in the present case), where the view was expressed (at [102]), arguably by way of obiter dicta) that punitive damages would not be awarded for breach of contract. In arriving at this view, this court endorsed the analysis of the law in Kay Swee Pin. We note however that the Assistant Registrar in that case only took the tentative view, after a survey of a number of decisions as well as arguments from principle, that the “weight of authority” was against the recognition of punitive damages, and that “no argument in principle or policy had been put forth to show why the position should be otherwise” (at [102]). It thus appears to us that Sim Yong Teng did not reject outright the availability of punitive damages in a purely contractual context. 94 However, Prof Lee also refers to decisions from the High Court which appeared to have accepted that in principle punitive damages could be awarded for a breach of contract (Scott Latham v Credit Suisse First Boston [1999] SGHC 302 at [43]; JES International v Yang Shushan [2016] 3 SLR 193 at [189]). We should also mention a decision of the Singapore International Commercial Court which was released after this appeal was heard. In Telemedia Pacific Group Ltd and another v Yuanta Asset Management International Ltd and another [2017] 3 SLR 47, Patricia Bergin IJ held that it was not necessary to decide whether the court had the power to award punitive damages, but added that, in any event, punitive damages should not be awarded because that was not an “exceptional case of high-handed and reprehensible conduct” that required punishment (at [35]). It will suffice for the moment to reiterate that the law in Singapore is – in these circumstances – not yet settled and still in a state of flux. 95 More generally (and importantly), Prof Lee’s submissions to this court contain a very comprehensive survey in comparative context of the relevant case law (and even includes references to the relevant position in the United States of America). We are greatly indebted to her industry and scholarship, and gratefully adopt her survey and analysis. The position in England and Wales 96 Turning, first, to England and Wales, it is clear that punitive damages cannot, under English law, be awarded for a breach of contract. This position is commonly taken to have been established by Addis v Gramophone Co Ltd [1909] 1 AC 488 (“Addis”), in which the House of Lords held that, in an action for wrongful dismissal, a claimant could not be awarded damages as compensation for the manner of the dismissal. Lord Atkinson observed (at 494) that “damages for breach of contract were in the nature of compensation, not punishment”, although it is worth noting that there was a dissenting opinion by Lord Collins, who preferred retaining the discretion to award such damages so as to allow redress of wrongs for which there might be no remedy (at 500). Notwithstanding this difference of views, the bar against punitive damages took root in English law. In Johnson v Unisys Ltd [2003] 1 AC 518, decided close to a century after Addis, Lord Steyn, having noted the difficulty in discerning the ratio decidendi of that case, nonetheless stated in no uncertain terms that, as a matter of English law, punitive damages “have never and cannot be awarded for breach of any contract” (at [15]). 97 Furthermore, in a report issued in 1997, the UK Law Commission recommended that the state of the law – which was that punitive damages could not be awarded for breach of contract – should not be changed (Report on Aggravated, Exemplary and Restitutionary Damages (Law Com No 247, 1997)). This was for the following five reasons (at para 1.72): First, exemplary damages have never been awarded for breach of contract. Second, contract primarily involves pecuniary, rather than non-pecuniary, losses; in contrast, the torts for which exemplary damages are most commonly awarded, and are likely to continue to be most commonly awarded, usually give rise to claims for non-pecuniary losses. Thirdly, the need for certainty is perceived to be greater in relation to contract than tort and, arguably, there is therefore less scope for the sort of discretion which the courts must have in determining the availability and quantum of exemplary damages. Fourthly, a contract is a private arrangement in which parties negotiate rights and duties, whereas the duties which obtain under the law of tort are imposed by law; it can accordingly be argued that the notion of state punishment is more readily applicable to the latter than to the former. Fifthly, the doctrine of efficient breach dictates that contracting parties should have available the option of breaking the contract and paying compensatory damages, if they are able to find a more remunerative use for the subject matter of the promise. To award exemplary damages would tend to discourage efficient breach. We naturally find the fourth reason mentioned by the Law Commission particularly persuasive given that it accords with what we have said about the contract-tort distinction above. 98 Prof Lee observes (correctly, in our view) that although there have been developments in the law of damages in England since the publication of this report, in particular the recognition of the account of profits as an arguably non-compensatory remedy for breach of contract, “these developments have not culminated in any observable traction for reform of the rule against punitive damages in contract law” and that “[o]n the contrary, recent authorities continue to affirm the orthodoxy that it is not the purpose of contract damages to punish”. The position in Australia and New Zealand 99 The legal position appears to be the same in Australia (see, for example, the High Court of Australia decisions of Butler v Fairclough (1917) 23 CLR 78 (at 89) and Gray v Motor Accident Commission (1998) 196 CLR 1 (at 6–7); the Federal Court of Australia decision of Hospitality Group Pty Ltd v Australian Rugby Union Ltd (2001) 110 FCR 157 (at [142]); and the New South Wales Court of Appeal decision of Harris v Digital Pulse Pty Ltd (2003) 44 ACSR 390). In the last-mentioned case, Spigelman CJ noted that “in Australian law [punitive] damages are not generally available for breach of contract” and that this was similar to the position in England as established by Addis (at [294]). 100 The legal position is also similar in New Zealand following the decision of the New Zealand Court of Appeal decision in Paper Reclaim Ltd v Aotearoa International Ltd [2006] 3 NZLR 188 (“Paper Reclaim”). In Paper Reclaim, the New Zealand Court of Appeal surveyed the position in English, Australian, Irish, American and Canadian law and concluded that the position in New Zealand should conform with the “clear trend of … authority” which was against the possibility that punitive damages should be awarded in breach of contract cases (at [180]). It observed that there was “no need for [punitive damages] to fill any hole in the range of compensatory damages in the contract field” given the availability of a wide range of contractual remedies such as damages for mental distress and an account of profits (at [182]) – this accords with our view that the argument for punitive damages based on a “remedial gap” is unpersuasive. 101 We pause to note that the sheer weight of case law alone is not necessarily conclusive of the issue although it cannot, of course, be ignored. The more important point for the purposes of the present appeal is that, in addition to the weight of case law authority, the arguments from general principle also support the case against awarding punitive damages in a purely contractual context. Many of these arguments of principle were also relied on in the cases and law reports from other jurisdictions. However, this is not the end of the matter because there is strong case law authority in the Canadian context which supports the award of punitive damages in a purely commercial context. We deal with this line of cases from [112] onwards. (5) Policy considerations 102 Whilst we would not over-emphasise this particular point, there also appear to be policy considerations that militate against the award of punitive damages in a purely contractual context. 103 It would be apposite to note that the distinction between considerations of legal principle on the one hand and those of policy on the other is often a fine one. For example, the distinction between contract and tort (see above at [64]−[76]) could, on one view, be said to be one of policy. However, on balance, it is our view that that is a consideration which relates more to legal principle than to policy. That leaves two other considerations that were, in fact, raised by Prof Lee in her written as well as oral submissions to this court which would fall, more appropriately, within the sphere of policy. Let us elaborate. 104 The first policy argument is that awarding punitive damages in a purely contractual context may adversely affect the manner in which litigation is conducted inasmuch as it may add to its length, complexity and costs and confer upon plaintiffs an undue advantage in forcing large (or larger) settlements. Whilst there may be, by its very nature, no easy way of ascertaining how persuasive this argument is, it cannot, in our view, be ignored. It could – at the very least – be the case that contracting parties will simply tag on a claim for punitive damages for breach of contract as a matter of course. And this would entail, in turn, time and effort on the part of the court in having to deal as a matter of routine with claims that would very rarely succeed. 105 Indeed, the New Zealand Court of Appeal relied on this policy argument in Paper Reclaim as a reason to rule out entirely the availability of punitive damages for breach of contract rather than allow that it could be awarded in exceptional circumstances. In its view, leaving the position open would mean that any plaintiff could “blithely” plead a claim for punitive damages, asserting that his or her case was exceptional; this might cause claims to go to trial unnecessarily, and consequently, plaintiffs would have a “powerful weapon with which they [could] harass defendants and, perhaps, extract large settlements because the costs of defending even an unmeritorious claim [might] be huge” (at [181]). The difficulty of devising clear criteria for determining the quantum of such awards adds to the risk that defendants may be subjected to unjustifiably high awards. 106 The second policy argument is that punitive damages in a purely contractual context are most commonly awarded in circumstances where there is a heightened risk of recurrent reprehensible conduct, especially where the parties are of unequal bargaining power, as is the case with insurance, employment and consumer transactions. In Prof Lee’s view, however, where such a risk is confined to particular types of contracts or specific industries, it would be more appropriately managed by regulation rather than by judicial remedies such as an award of punitive damages. This is because the regulator would generally be in a better position to assess the wider social as well as economic impact of a particular sanction and would therefore be better able to devise a suite of remedial responses (including, but not limited to, penal sanctions). We see much force in this argument which also buttresses the case against the award of punitive damages in a purely contractual context. 107 Let us turn now to consider the opposite side of the coin, so to speak, viz, arguments in favour of an award of punitive damages in a purely contractual context. Arguments in favour of an award of punitive damages in a purely contractual context (1) Introduction 108 In our view, the general arguments as well as the case law against the award of punitive damages in a purely contractual context are extremely powerful. In fairness to Airtrust, however, we will now examine the arguments in favour of the award of such damages. In this regard, there appear to be three main arguments. The first is an argument from uniformity. Put simply, if outrageous conduct might lead to the award of punitive damages in tort, why should such damages not similarly be awarded for outrageous conduct in a purely contractual context? The second centres on the Canadian case law and the third is whether there is a case for the award of punitive damages in a purely contractual context where there exists either an express or implied duty of good faith on the part of the party in breach of contract. The third argument is in fact closely related to the second inasmuch as the leading Canadian case did (as we shall see) premise its decision to award punitive damages in a purely contractual context, in part, on the fact that a (separate and distinct) contractual duty of good faith had been breached. 109 We turn now to consider these arguments. (2) The argument from uniformity 110 We note that Airtrust places much emphasis on this argument (see above at [29(b)]). It submits that there is no reason to suppose that morally outrageous behaviour deserving of punishment can only be found in tort cases and not contract cases. A similar sentiment was expressed by the editors of McGregor on Damages (Sweet & Maxwell, 19th Ed, 2014) at para 13-016): as they ask, since punitive damages can be awarded in tort if there has been, for example, high-handed public conduct or profit-motivated private conduct, “may not such conduct deserve the same sanction whatever the cause of action?” 111 The answer to the question posed in the preceding paragraph is straightforward: to the extent that the law of tort is qualitatively different from the law of contract, the argument from uniformity is, in the final analysis, an unpersuasive one. Indeed, the qualitative difference just mentioned has been dealt with earlier in this judgment in some detail (see above at [68]−[76]). (3) The Canadian case law 112 By far the strongest arguments in favour of an award of punitive damages in a purely contractual context are (as alluded to above) to be found in the Canadian cases, in particular, the Supreme Court of Canada’s decision in Whiten v Pilot Insurance Company (2002) 209 DLR (4th) 257 (“Whiten”) where the court did endorse for the first time the availability of punitive damages in a purely contractual context. It is necessary to analyse Whiten in some detail as it was the key authority relied on by the Judge. 113 Before turning to Whiten, it is important to mention a case from the same court which foreshadowed it. In Vorvis v Insurance Corporation of British Columbia [1989] 1 SCR 1085 (“Vorvis”), it was held that punitive damages would only be awarded for breach of contract where the conduct complained of simultaneously constituted an “independently actionable” wrong which caused the injury complained of by the plaintiff (at [22] and [25]). The court in Vorvis held that an award of punitive damages was not warranted on the particular facts of that case, but the decision left many in doubt about what the requirement that the wrong had to be “independently actionable” meant. The assumption of most courts and commentators, at least until Whiten, was that the court was referring to concurrent liability in tort (see, for example, Yehuda Adar, “Whiten v Pilot Insurance Co: The Unofficial Death of the Independent Wrong Requirement and Official Birth of Punitive Damages in Contract” (2005) 41 Can Bus LJ 247 (“Adar”) at p 254). Looked at in that light, the position in Vorvis was less of a departure from orthodoxy than it seemed to be. 114 In Whiten itself, the majority of the seven-member court (comprising all but LeBel J, who dissented) held that the requirement of an independent actionable wrong was satisfied by the breach of a separate contractual obligation – specifically, the breach of a duty of good faith. As already alluded to above, we will deal with the effect (if any) of a duty of good faith in the next part of this judgment. However, on a more general level, there are conceptual difficulties with the court’s identification of a breach of a duty of good faith as satisfying the “independent actionable wrong” requirement. As Prof Lee persuasively argues (citing John D McCamus, “Prometheus Bound or Loose Cannon? Punitive Damages for Pure Breach of Contract in Canada” (2004) 41 San Diego L Rev 1491 (“McCamus”) at p 1504), it is not clear why punitive damages may be awarded in tort for a single breach of duty but should be contingent upon proof of an independent actionable wrong in contract. She further observes that “[t]he perplexity is compounded by the Whiten court’s acceptance that the requirement may be satisfied by a separate contractual breach, thus suggesting that the award may be made for two or more breaches but not a single egregious breach”. And in a footnote to this observation, Prof Lee elaborates by noting that “[t]he obvious unsoundness of such a proposition must inevitably lead one to conclude that a single breach of contract could, in appropriate circumstances, also justify an award of punitive damages” (relying on McCamus at p 1504 and Adar at p 263). Finally, she concludes that “[o]n this interpretation, the requirement [of an independent actionable wrong] is effectively denuded of any practical utility as a means of limiting the availability of the award, for it can easily be fulfilled by ‘dividing up the defendant’s contractual obligations into little pieces’” (relying on John Swan, “Punitive Damages for Breach of Contract: A Remedy in Search of A Justification” (2004) 29 Queen’s LJ 596 (“Swan”) at p 616). 115 Quite apart from the artificiality that results from this particular requirement (as we have mentioned), it is also, with respect, flawed simply because it does not meet the argument (also just set out) that there is no reason, in principle, why a single breach of contract ought not also to justify an award of punitive damages. After all, if it is the egregious nature of the conduct of the party in breach that is being punished, it ought not to matter whether that conduct is the result of a single breach or more than a single breach. If, of course, the independent actionable legal wrong is a tort, there would be no difficulties to begin with simply because there would then be concurrent liability in both contract and tort, thus paving the way for an award of punitive damages in any event (see also above at [64]). 116 To turn to the facts and holding in Whiten: an insurance company had engaged in “stonewall tactics” calculated to deny the legitimate claim of a policy holder (“the insured”) whose home had been destroyed by an accidental fire. It appeared that the insurance company was engaging in tactics which would ultimately result in the insured capitulating to a cheap settlement. In particular, the insurance company accused the insured of arson, notwithstanding the fact that various investigations had revealed no evidence of such misconduct. As a result, the insured, who was already in poor financial shape, was deprived of the means of securing a permanent (and replacement) home and was forced to embark on costly litigation to recover what was lawfully hers pursuant to the insurance policy concerned. At trial, the jury was satisfied that the insurance company had behaved in a manner that was malicious, high-handed, arbitrary as well as capricious; it imposed upon the insurance company punitive damages of C$1m in addition to compensatory damages of C$318,252.32. On appeal, a majority of the Ontario Court of Appeal, whilst agreeing that punitive damages ought to be imposed on the facts, nevertheless reduced the quantum of punitive damages awarded to C$100,000. On a further appeal to the Supreme Court of Canada, however, the original award of punitive damages imposed by the jury was restored. Whilst Binnie J, who delivered the judgment of the majority of the court, conceded that the award of C$1m in punitive damages was high, he was of the view that such an award was nevertheless “not so disproportionate as to exceed the bounds of rationality” (at [128]). 117 As Prof Lee points out, Whiten has been subject to no small measure of criticism (significantly, in our view, from a number of eminent academic experts in the field of Canadian contract law). We have already considered one of the criticisms directed at the court’s application of the independent actionable wrong requirement. Prof Lee also points out, inter alia (and citing Swan), that the court in Whiten could have awarded damages for mental distress instead (see also [83] above) and/or penalised the insurance company by way of indemnity costs (although the latter suggestion may not yield as substantive an award of damages as the other alternatives considered earlier in this judgment). She also raises points that we have considered earlier – for example, the fact that an award of punitive damages in a purely contractual context violates the compensation principle that arises from the agreed allocation of expectations by the contracting parties as well as the fact that there are no clear criteria for ascertaining when precisely such damages ought to be awarded. 118 Finally, Prof Lee also notes Prof S M Waddams’s view (S M Waddams, The Law of Damages (Canada Law Book, 5th Ed, 2012) at para 11.257) that Whiten may not be a true authority for the award of punitive damages in a purely contractual context as “there were features of the case not common to ordinary commercial contracts, notably a public, quasi-regulatory, interest in inducing insurers to investigate claims fairly and meet their obligations” (see also [124] below). It was also a case involving tortious defamatory conduct, given that the insurance company had accused the insured of having committed arson. (A) The Whiten principles 119 The judgment by Binnie J on behalf of the court in Whiten was, nevertheless, a comprehensive and wide-ranging one. The key principles are succinctly captured by Prof Lee in the following helpful summary. (a) The attempt to limit punitive damages by “categories” [a reference to the test propounded by Lord Devlin in the House of Lords decision of Rookes v Barnard (see [64] above), in relation to the award of punitive damages in tort law] is unhelpful. Canadian cases have rightly rejected this limitation and awarded punitive damages whenever the defendant’s conduct was such as to merit judicial condemnation (see Whiten at [67]). (b) The general objectives of punitive damages are punishment (in the sense of retribution), deterrence and denunciation of egregious conduct (see Whiten at [68]). (c) Punitive damages should only be resorted to “in exceptional cases and with restraint” because criminal law remains the primary vehicle for punishment (see Whiten at [69]). (d) In setting the quantum, the court should be guided by the principle of rationality. An award is rational if it is no larger than the minimum necessary to achieve one or more objectives (punishment, deterrence and denunciation) of awarding punitive damages (see Whiten at [71]). (e) It is rational “to relieve a wrongdoer of its profit where the compensatory damages would amount to nothing more than a licence fee to earn greater profits through outrageous disregard of the legal or equitable rights of others” (see Whiten at [72]). (f) A rational award is not one constrained by fixed cap or ratio (see Whiten at [73]) but one that accords with the principle of proportionality. This means that the overall award (compensatory plus punitive damages and other awards relating to the same misconduct) should be “rationally related to the objectives for which the punitive damages are awarded (retribution, deterrence and denunciation)” (at [74]). Or, to put it in more familiar language, punitive damages should only be awarded “if, but only if” compensatory damages are insufficient to punish the offensive conduct. (g) Specifically, “proportionality” is determined by reference to several dimensions including: (i) the blameworthiness of the defendant’s conduct; (ii) the vulnerability of the plaintiff; (iii) the harm or specific harm directed specifically at the plaintiff; (iv) the need for deterrence; (v) other penalties, both civil and criminal, which have been or are likely to be inflicted on the defendant for the same misconduct; and (vi) the advantage wrongfully gained by the defendant from the misconduct. 120 The Whiten principles are, in the final analysis, unpersuasive. Let us elaborate, considering each principle in turn (we will refer to each of the principles above as Whiten principle (a) to (g)). (B) The Whiten principles are unpersuasive 121 Whiten principle (a) is not persuasive because the court assumed that there was no difference in principle between awarding punitive damages in tort and in a purely contractual context. The court surveyed the cases from various jurisdictions in which the availability of punitive damages had been extended to various torts, without considering that the position in those same jurisdictions was against punitive damages for breach of contract. For instance, the court cited (at [49]) the UK Law Commission’s support of punitive damages in tort, but did not refer to the Commission’s disapproval, in the same report, of punitive damages in contract law (as we have noted at [97] above). The critique of the limitation of the award of punitive damages to “categories” was a reference to the narrow position set out in the law of tort and does not, with respect, really address the issues of principle relating to the justification for awarding punitive damages in a purely contractual context. 122 With respect, Whiten principle (b) is susceptible to the same critique as Whiten principle (a). The three objectives of retribution, deterrence, and denunciation were based on a statement in the English decision of Wilkes v Wood (1763) Lofft 1. In that case, government agents had trespassed on the premises of a politician to seize certain political publications, and it was held that a jury had the power not just to give damages for the injury received but also to punish (see Whiten at [40]). It was a case in tort. Again, the significant differences between tort and contract were overlooked and the broader issue of principle – why punishment is a legitimate remedial response to a breach of contract – was not really dealt with. 123 The same difficulty applies to Whiten principle (c). Given the court’s reliance on a string of authorities on punitive damages in tort, the underlying premise that punishment is a legitimate or necessary response to a breach of contract was assumed without argument. The assumption has been criticised by one Canadian commentator on the basis that damages for breach of contract are only meant to compensate and that considerations of punishment only “distort the contractual relation” between the parties (see Swan at p 635). This view echoes the concerns we have expressed above (at [71]–[73]) on the uneasy relation between the law’s tolerance of self-interested behaviour in the contractual context and any perceived need to punish those who breach their contracts. 124 Another difficulty we have is with the court’s explanation that punitive damages should be exceptional because criminal law is the main vehicle of punishment. This suggests that an award of punitive damages is meant to fulfil a quasi-regulatory function. At [123] of Whiten, Binnie J explained the relevance of taking into account other civil or criminal penalties in determining what would be a proportionate sum of damages to inflict on the defendant for the same misconduct: Compensatory damages also punish. In many cases they will be all the “punishment” required. To the extent a defendant has suffered other retribution, denunciation or deterrence, either civil or criminal, for the misconduct in question, the need for additional punishment in the case before the court is lessened and may be eliminated. In Canada, unlike some other common law jurisdictions, such “other” punishment is relevant but it is not necessarily a bar to the award of punitive damages. The prescribed fine, for example, may be disproportionately small to the level of outrage the jury wishes to express. The misconduct in question may be broader than the misconduct proven in evidence in the criminal or regulatory proceeding. The legislative judgment fixing the amount of the potential fine may be based on policy considerations other than pure punishment. The key point is that punitive damages are awarded “if, but only if” all other penalties have been taken into account and found to be inadequate to accomplish the objectives of retribution, deterrence, and denunciation. The intervener, the Insurance Council of Canada, argues that the discipline of insurance companies should be left to the regulator. Nothing in the appeal record indicates that the Registrar of Insurance (now the Superintendent of Financial Services) took an interest in this case prior to the jury's unexpectedly high award of punitive damages. [emphasis added] This passage suggests that punitive damages should be imposed where misconduct by the defendant has gone unchecked by the relevant regulatory authority (in that case, the Registrar of Insurance), or to put it bluntly, that courts should take it upon themselves to punish where the state has failed to. We would have serious reservations accepting that as the proper function of a court in any civil proceeding, whether in tort, contract, or otherwise. As we have observed above at [106], such regulatory interests are best taken care of by the regulator. 125 The strength of Whiten principle (e) is diminished by the recognition that an account of profits can be awarded as a remedy for breach of contract (see [81] above). 126 Finally, Whiten principles (d), (f) and (g) attempt to impose a rational limit on punitive damages awards. However, it appears that the principle of proportionality does not, despite the “several dimensions” referred to in Whiten principle (g), furnish sufficient guidance, and that this leads, in turn, to uncertainty. It appears that even courts at different levels of the judicial hierarchy, looking at the same set of facts, may come to very different conclusions about the appropriate quantum of punitive damages. In Whiten itself, the award of C$1m in punitive damages by the trial court was reduced to a tenth of that sum by the appellate court and then reinstated by the Supreme Court (and even then not unanimously, since LeBel J dissented, finding that the award of C$1m went “well beyond a rational and appropriate use of this kind of remedy” (at [143]). To take another example, in Honda Canada Inc v Keays [2008] 2 SCR 362 (“Honda Canada”) the trial court’s award of C$500,000 was reduced by the Ontario Court of Appeal to C$100,000 and then vacated by the Supreme Court. This has led to awards of punitive damages being described as arbitrary and incoherent (see Harrington at p 219). 127 There is a further reason to be cautious about adopting the Whiten principles as part of Singapore law: it is not clear whether Whiten is to be read as recognising the availability of punitive damages for all contract breaches or only certain types of contracts where there is, in Prof Lee’s words, a “material power imbalance”. A useful contrast may be drawn between Whiten and its companion case, Performance Industries Ltd v Sylvan Lake Golf & Tennis Club Ltd [2002] 1 SCR 678 (“Sylvan”), which was heard on the same day as Whiten and which adopted entirely its principles on the issue of punitive damages. In Sylvan, two businessmen had an oral agreement. One of them (the defendant) reduced the agreement into writing but deliberately changed one of the terms to his advantage. He then insisted on the written terms when the other (the plaintiff) tried to enforce the agreement despite knowing that the written terms did not accurately reflect the oral agreement. The plaintiff brought a claim for rectification and damages. One issue was whether punitive damages should be awarded against the defendant for his fraudulent misrepresentation that the written document accurately reflected the terms of the oral agreement. The trial court awarded punitive damages of C$200,000; the Alberta Court of Appeal set the award aside and its decision was affirmed by the Supreme Court. The judgment was also delivered by Binnie J, who noted that “fraud is generally reprehensible, but only in exceptional cases does it attract punitive damages” (at [87]). He also suggested that a falling out between two business partners would not attract an award of punitive damages if there was no “abuse of a dominant position”, notwithstanding the fact that the defendant’s misconduct was “planned and deliberate” (at [29] and [88]). 128 Read together, Sylvan and Whiten suggest that the Supreme Court of Canada intended only to allow punitive damages for breaches of certain exceptional types of contracts. There is no agreement on what such contracts should be. Some commentators suggest that contracts of a “personal nature” which are of “obvious importance to a plaintiff’s well-being” would fall in this category (see James Cassels & Elizabeth Adjin-Tetty, Remedies: The Law of Damages (Irwin Law, 3rd Ed, 2014) at p 353); others suggest insurance contracts alone, as in Whiten, or more generally contracts where there is a duty of good faith can be the subject of an award of punitive damages (see Adar at p 272). This last-mentioned interpretation has some force given, first, that in Whiten, the independent actionable wrong requirement was satisfied by the finding that the insurer had breached a duty of good faith, and, second, that subsequent decisions of the Supreme Court of Canada have rejected awards of punitive damages on the basis that the defendants had not acted in bad faith (see Fidler v Sun Life Assurance of Canada [2006] 2 SCR 3 and Honda Canada). 129 We shall therefore consider, notwithstanding our doubts about Whiten, whether or not an award of punitive damages in a purely contractual context can nevertheless be justified where there has been a breach of an express or implied duty of good faith. (4) What if there exists an express or implied duty of good faith 130 It is important, at the outset, to note what the Singapore position is insofar as a duty of good faith in the law of contract is concerned (and, for an excellent as well as nuanced account of the respective legal positions in (especially) England and Canada, see Tan Zhong Xing, “Keeping Faith with Good Faith? The Evolving Trajectory Post-Yam Seng and Bhasin” [2016] JBL 420). 131 The legal position in Singapore with regard to a duty of good faith in the law of contract is summarised in the following observation of this court in The One Suites Pte Ltd v Pacific Motor Credit (Pte) Ltd [2015] 3 SLR 695 (at [44]): … [T]he present position in Singapore as set out in the decision of this court in Ng Giap Hon v Westcomb Securities Pte Ltd [2009] 3 SLR(R) 518 (“Ng Giap Hon”) is that there is no implied duty of good faith based on a “term implied in law”. However, this court in Ng Giap Hon was prepared to leave open the possibility that such a duty could be implied by way of the narrower category of “terms implied in fact”. Whilst it is thought in some quarters that the English High Court decision in Yam Seng Pte Ltd v International Trade Corporation Ltd [2013] 1 Lloyd’s Rep 526 heralded a more expansive approach towards the doctrine of good faith, a close analysis of the judgment itself (especially at [131]) suggests that the English position is not that much different from the existing Singapore position in Ng Giap Hon (reference may also be made to the English Court of Appeal decision of Mid Essex Hospital Services NHS Trust v Compass Group UK and Ireland Ltd (trading as Medirest) [2013] EWCA Civ 200, especially at [105]). It might also be usefully noted that this court in HSBC Institutional Trust Services (Singapore) Ltd v Toshin Development Singapore Pte Ltd [2012] 4 SLR 738 held that where there is an express term in the contract that the parties would negotiate in good faith, this would be given effect to. However, the law in this particular sphere (viz, good faith) continues to be in a state of flux. For example, the Supreme Court of Canada in Harish Bhasin v Larry Hrynew and Heritage Education Funds Inc [2014] SCC 71 recently held that there is a duty of honest performance which is, in turn, a manifestation of the general organising principle of good faith. Whether or not the formulation in this last-mentioned decision is too vague and general is a question which is (fortunately) outside the purview of the present appeal – as is the (more specific) issue as to whether or not the implied term of mutual trust and confidence ought to apply in the employment context such that a term ought to be implied in law that neither party will, without reasonable cause, conduct itself in a manner that is likely to destroy or seriously damage or undermine the relationship of trust and confidence between employer and employee (this principle being first established in the leading House of Lords decision of Malik v Bank of Credit and Commerce International SA (in compulsory liquidation) [1998] AC 20, which was, however, not followed in the recent High Court of Australia decision in Commonwealth Bank of Australia v Barker (2014) 88 ALJR 814 (with the position in Singapore still left open for decision in a future case (see the decision of this court in Wee Kim San Lawrence Bernard v Robinson & Co (Singapore) Pte Ltd [2014] 4 SLR 357))). It will be immediately seen from this very short synopsis of the various issues why the question as to whether or not there is a duty on the part of the parties to cooperate and, if so, what its scope is and what its relationship is to doctrines such as good faith, are all matters that ought best to be decided in a definitive fashion only when they next come directly for decision before the courts. [emphasis in original] 132 It will be seen that, apart from an express (contractual) duty of good faith, an implied duty of good faith would be rare (being premised, if at all, not on a “term implied in law” but, instead, on the much narrower basis of a “term implied in fact”). As an aside, it is apposite at the present juncture to note that the position in Canada (as set out in the Supreme Court of Canada decision of Harish Bhasin v Larry Hrynew and Heritage Education Funds Inc [2014] 3 SCR 494 (“Harish Bhasin”)) on the duty of good faith in contract law is quite different from ours and indicates that Whiten’s ostensibly liberal approach to punitive damages may come to be more firmly entrenched in that jurisdiction. Harish Bhasin established that there is an “organizing principle of good faith” in the (Canadian) common law of contract that underlies and manifests itself in various more specific doctrines, and that, as a manifestation of this organising principle, “a general duty of honesty in contractual performance” is to be recognised in all contracts (at [63] and [73]). It appears to us that if, according to Whiten, the independent actionable wrong requisite for an award of punitive damages is satisfied by the breach of a good faith obligation, and if there is now a new duty of honest performance applicable to all contracts (as a particular manifestation of the general organising principle of good faith) after Harish Bhasin, it follows that the Canadian courts may be able to award punitive damages in response to a greater variety of contractual breaches. This is especially so given that the court in Harish Bhasin suggested that the list of specific doctrines that may be subsumed under the organising principle of “good faith” is not closed and can be developed incrementally (at [66]). 133 Although no similar organising principle of good faith has been recognised in Singapore law, it is still possible to argue, pursuant to Singapore law, for the implication of a duty of good faith. If so (and given the rarity with which such a duty would be found in the Singapore context), could it be argued that, where such a duty in fact exists, it should also be possible for our courts to award punitive damages? 134 We do not think the existence of a duty of good faith would, on its own, justify an award of punitive damages in the event of its breach. The existence of such a duty may help overcome one of the objections in principle to punitive damages for breach of contract: the undesirability of a court imposing on the parties its own normative standard of contractual performance. If there is either an express or implied duty of good faith, the argument that the contracting parties are justified in performing the contract in a way that prioritises their self-interest falls away. However, it does not necessarily follow that punitive damages ought to be awarded. This is because if a contracting party falls short of an obligation of good faith, any damages awarded could arguably be viewed as compensation to the aggrieved party for a breach of such an obligation. The aggrieved party could perhaps argue that, in addition to such compensatory damages, the court should make a further award of punitive damages for breach of an obligation of good faith. However, that merely brings us back full circle to the justification for and/or criteria by which punitive damages ought to be awarded in a purely contractual context in the first place. Put simply, the existence of an express or implied duty of good faith is a neutral factor. Conclusion 135 We have set out the arguments both for as well as against the award of punitive damages in a purely contractual context. It is clear, in our view, that the arguments against the award of such damages far outweigh the arguments in favour of such an award. Indeed, the case authority which most strongly supports the making such an award is itself afflicted with no small measure of difficulties. Not only is Whiten the only case in the sea of Commonwealth authorities that supports such an award, it is also subject to a number of very persuasive criticisms. In the circumstances, we are of the view that there ought to be a general rule that punitive damages cannot be awarded for breach of contract. 136 Given, however, that the instances in which a breach of contract can occur are manifold, we would not rule out entirely the possibility that a case may one day come before this court, involving a particularly outrageous type of breach, which necessitates a departure from the general rule. As was said in a case concerning punitive damages for negligence, “‘never say never’ is a sound judicial admonition” (see the Privy Council decision of A v Bottrill [2003] 1 AC 449 at [26] per Lord Nicholls of Birkenhead). That said, any argument for the award of such damages would need to surmount the many reasons of principle and policy set out in this judgment against doing so. It would therefore take a truly exceptional case to persuade this court that punitive damages should be awarded for breach of contract. Indeed, we are not sure that we would award punitive damages even if we were faced with the facts of a case as extreme as Whiten. As we have noted, there are a number of other possible alternative remedies (including the award of damages for mental distress for breach of contract) that could also be invoked by the court to do practical justice while respecting the compensatory function of damages for breach of contract. The phrase, “hardly ever”, used by Lord Hailsham of St Marylebone LC in the House of Lords decision in National Carriers Ltd v Panalpina (Northern) Ltd [1981] AC 675 (at 692) (albeit in relation to the possibility of frustration of leases) more accurately expresses our present sentiments on the availability of punitive damages for breach of contract. Whether we would have awarded punitive damages in this case 137 Given that we have not ruled out wholly the possibility of the award of punitive damages in a purely contractual context, can a case be made on the facts of this case that such damages ought to be awarded? It is important, in our view, to reiterate the general rule that punitive damages are not awardable in a purely contractual context. 138 Turning to the specific facts of the present case, we note that we have found that PH had not been fraudulent; there was, at best, only gross negligence on its part. That would fall short of the kind of egregious conduct exhibited by the defendant in Whiten which was found to be deserving of punishment. To be clear, our assessment would not be affected even if these purported acts of fraud are taken together with the other instances of irresponsible designing, engineering and manufacturing on PH’s part identified by the Judge (see above at [22]). In our view, this was clearly not a situation which merits an award of punitive damages. Put simply, the general rule referred to in the last paragraph applies. 139 We hasten to add that even if fraud had in fact been established (which was not the case here), it does not follow inexorably that punitive damages should be awarded. The Judge relied chiefly on Whiten but, as we have explained, even in Canada, where punitive damages are an available remedy for breach of contract, it was recognised in the subsequent case of Sylvan that fraud in and of itself, is not sufficient to justify an award of punitive damages particularly in the commercial context (see our discussion of Sylvan at [127]–[128] above). We would thus have been of the view, even if fraud had been established, and even if such fraud had been planned and deliberate, that this was hardly an exceptional case as to bring us anywhere close to departing from the general rule. Given that the same court that awarded punitive damages in Whiten declined to do so in Sylvan, and given that we are not even sure that we would award punitive damages on the facts of a case as extreme as Whiten, it follows, unquestionably, that on the facts of a case such as the present (which are more similar to those in Sylvan), an award of punitive damages would not, in our view, be even remotely possible. 140 Finally, we note that, had PH made a fraudulent misrepresentation, Airtrust would have been entitled (provided the requisite evidence was forthcoming) to a generous measure of damages pursuant to the common law and/or s 2(1) of the Misrepresentation Act (Cap 390, 1994 Rev Ed) (see, generally, the decision of this court in RBC Properties Pte Ltd v Defu Furniture Pte Ltd [2015] 1 SLR 977, although a provisional view was expressed in that case that perhaps the latter (statutory) measure of damages might not be the same as that obtainable pursuant to a common law action for fraudulent misrepresentation or deceit). The generous measure of damages for fraudulent misrepresentation at common law includes all loss which flows directly as result of the entry by the representee into the contract concerned, regardless of whether or not such loss was foreseeable, and all consequential loss as well. As Lord Steyn observed in the leading House of Lords decision in Smith New Court Securities Ltd v Citibank NA [1997] AC 254, this generous measure is justified and a stricter approach towards the fraudster is adopted because “[f]irst it serves a deterrent purpose in discouraging fraud” (at 279) and “[s]econdly, as between the fraudster and the innocent party, moral considerations militate in favour of requiring the fraudster to bear the risk of misfortunes directly caused by his fraud” (at 280). In those circumstances, and assuming that such a generous award could be made, we would have found it difficult to justify imposing a further award to punish PH in this case. 141 In the circumstances, it is clear that Issue 2 must be resolved in favour of PH. We now turn to Issue 3. Issue 3 – Interpretation of cl 25 142 We commence our analysis of Issue 3 by setting out cl 25 in full: 25) Indemnity Company, Contractor and Supplier hereby indemnifies and holds harmless the other from and against all claims, costs, damages, expenses, and liabilities arising out of personal injury or death to its or its other contractors, officers or employees and any loss or damage to its or its other contractors properly arising out of or as a result of the performance of the work associated with this Purchase Order and regardless of the cause or reason for the said loss or damage and regardless of whether the same may arise of, or as a result of the negligence of the other Party. Neither party shall be liable to the other for any consequential or indirect losses (whether or not foreseeable by either Party at the date hereof) including but not limited to loss of profits, products, and business interruption or economic losses arising out of or as a result of the performance of the work and regardless of the cause or reason for the said loss or damage and regardless of whether the same may arise as a result of the negligence of the other. Each party shall assume its legal liability towards the Third Parties and shall indemnify and hold harmless the other Party accordingly. For the purpose of this Article each Party shall be deemed to include its parent, affiliate, subsidiary companies, their respective officers, directors, employees and agents. [emphasis added] Whether Cl 25 applies only to third-party claims 143 Airtrust’s primary contention is that Cl 25 is wholly inapplicable. It reads Cl 25 in the following way: the first paragraph obliges the contracting parties to indemnify each other against claims by third parties; the second paragraph qualifies the first paragraph by limiting the indemnity to the third parties’ direct losses and excluding the third parties’ consequential or indirect losses. 144 We do not agree with Airtrust’s reading of Cl 25’s second paragraph. On its terms, it refers to “neither party [having any liability] to the other for any consequential or indirect losses”. The premise must be that the consequential or indirect losses are suffered by the other contracting party. It is significant that the phrase “[n]either party shall be liable to the other” is used, rather than “[n]either party shall indemnify and hold harmless the other”. If paragraph 2 were only meant to qualify paragraph 1, one would have expected the same phrase “indemnify and hold harmless” to be used instead. 145 Airtrust’s argument that paragraph 2 limits the extent of the indemnity under paragraph 1 assumes that the “consequential or indirect losses” are suffered by the “third parties”, such as employees or contractors; and that it is the third parties’ consequential or indirect losses which do not have to be indemnified. Hence, Airtrust provides this illustration of how Cl 25 operates: if the poor design of the RDU had resulted in damage, Cl 25 “would have required [PH] to indemnify [Airtrust] against any personal injury or property claims against [Airtrust], but not in respect of consequential or indirect losses claimed against [Airtrust]” [emphasis added]. 146 That does not appear to us to be the most natural interpretation of the second paragraph, or at any rate the only one. At its highest, the argument suggests that paragraph 2 limits both the liability between the contracting parties and the extent of the indemnity for third-party claims in paragraph 1. In other words, (a) PH is not liable to Airtrust for any consequential or indirect losses arising out of the performance of the work, and (b) PH does not need to indemnify and hold Airtrust harmless against consequential or indirect losses suffered by its personal employees or other contractors. But that still means that Cl 25 is applicable to the present case to limit PH’s liability to Airtrust for its breach of contract to those losses which were not consequential or indirect losses, ie, direct losses. 147 In support of its interpretation, Airtrust relied on the interpretation of what it contends is a similar clause in the English Court of Appeal decision of Kudos Catering (UK) Ltd v Manchester Central Convention Complex Ltd [2013] EWCA Civ 38 (“Kudos Catering”). However, that case does not assist Airtrust but in fact fortifies our interpretation of Cl 25. 148 In Kudos Catering, the relevant clauses of a contract in a section headed “Indemnity and Insurance” read as follows: 18.4 The Company shall indemnify and keep indemnified the Contractor against all actions, claims, demands, proceedings, damages, costs, charges and expenses whatsoever in respect of or in any way arising out of the provision of, or damage to, any property including property belonging to the Contractor to the extent that it may arise out of the negligence of the Company, its employees or agents. … 18.6 The Contractor hereby acknowledges and agrees that the company shall have no liability whatsoever in contract, tort (including negligence) or otherwise for any loss of goodwill, business, revenue or profits, anticipated savings or wasted expenditure (whether reasonably foreseeable or not) or indirect or consequential loss suffered by the Contractor or any third party in relation to this Agreement and the limitations set out in this condition 18.5 shall be read and construed and shall have effect subject to any limitation imposed by any applicable law, including without limitation that this Condition shall not apply to personal injury or death due to the negligence of the Company. [emphasis added in italics and bold italics] 149 Tomlinson LJ (with whom McCombe and Laws LJJ agreed) held (at [26]) that “[b]oth its position and its content” showed that cl 18.6 “intended to qualify the wide ambit of the indemnity afforded by Clause 18.4”. It is, however, explicit in cl 18.6 that the consequential and indirect losses excluded are those suffered, not just by the other contracting party, but also those suffered by “any third party”. Cl 25 in our case refers only to consequential and indirect losses without specifying whom those losses are suffered by. As we have observed, Airtrust has attempted to confine the consequential or indirect losses which are excluded to those suffered only by a third party. But the very case it has cited demonstrates that it is equally likely that a clause such as Cl 25 may exclude consequential losses suffered both by the other contracting party and third parties. That is, in fact, the interpretation we have preferred. 150 The interpretation which we have endorsed is neither too generous nor uncommercial. The consequential and indirect losses excluded are only those which arise out of performance (or, more accurately, defective performance) of the work. Paragraph 2 does not exclude consequential and indirect losses arising from other breaches of contract. Indeed, Tomlinson LJ arrived at the same view in Kudos Catering. At [27], he found that, given the reference to third-party losses, the phrase “in relation to this Agreement” in cl 18.6 referred to losses arising from the performance of the agreement, and not from non-performance or refusal to perform; if there was no performance, no third-party losses could arise. In form, cl 18.6 is clearer than Cl 25 here in one respect, but less clear in another. Clause 18.6 is clearer in that it explicitly states by whom the “consequential and indirect losses” are suffered though Cl 25 does not. It is less clear in that it uses the broadly-worded phrase “in relation to this Agreement” though Cl 25 limits it to performance of the agreement. In substance, however, it appears that cl 18.6 in Kudos Catering and Cl 25 in the present case are the same. 151 Therefore, properly interpreted, Cl 25 excludes consequential and indirect losses suffered by the contracting parties as well as third parties arising out of the (defective) performance of the SPA in question, but not consequential or indirect losses arising from a refusal to perform the contract or to be bound by it. Airtrust’s claim against PH is for consequential losses suffered by it arising out of the (defective) performance of the SPA. On its face, therefore, Cl 25 applies to exclude such losses. 152 We therefore turn to examine whether or not Airtrust’s claims for lost rental profits and lost opportunities to earn further profits are consequential and indirect losses, which would be excluded, or direct losses, which would not. To be precise, the issue is whether the phrase “loss of profits”, which is said in Cl 25 to be an example of consequential or indirect losses, covers Airtrust’s claims. Whether loss of profits is excluded 153 Both parties accept that the phrase “consequential and indirect losses” in Cl 25 refers to those claims falling under the second limb in Hadley v Baxendale (1854) 9 Exch 341, ie, damage which does not arise directly, naturally and in the ordinary course of events from a breach of contract (see Robertson Quay at [82]). Direct losses, on the other hand, are those which fall within the first limb of Hadley v Baxendale. Such direct losses ought to be well within the reasonable contemplation of all the contracting parties concerned such that it is neither unjust nor unfair to impute knowledge of such damage to them (Robertson Quay at [81]). Airtrust argues that its “inability to earn revenue from renting out the RDU is a direct loss, not an indirect or consequential one”. 154 It is important to note that the claim pleaded by Airtrust was for loss of profits which the RDU would have made over a period of normal utility if it had been properly designed and manufactured. Airtrust has now argued that it is the loss of revenue which is a direct loss. 155 Regardless, we are of the view that the loss of revenue or profit could not be a direct loss which was within the reasonable contemplation of both parties. All that Airtrust points to as indicating that loss of revenue would have been within the reasonable contemplation of the parties is that PH had agreed to design and build the RDU for Airtrust knowing that it would be rented out to Trident and have a working life of at least 15 years, and that PH had also built RDUs which it rented out itself for income. However, it is not entirely clear that the parties knew that the RDU would be rented out for the full 15 years. It may have been necessary to show evidence that the parties knew the length of the rental contract to Trident. We were not referred to any such evidence. All that has been shown is that Trident paid a daily rate of US$5,050 per day for use of the RDU; there is no evidence that Trident was obliged to rent the RDU for any fixed period or minimum period. It is also not clear if the evidence supports the usability of the RDU, as constructed, in any other area other than that in which Trident was intending to use it, such that the inference could be drawn that both parties would have reasonably contemplated (and expected) Airtrust to rent out the RDU to other users within that 15-year period even if Trident could not use it. In the circumstances, we are not persuaded that the loss of revenue or profit which the RDU would have made over the period of 15 years of working utility was a direct loss. 156 The second head of loss is for loss of opportunity to earn profits from the rental of RDUs acquired from the earnings of the first and other RDUs. We do not think that this could be a direct loss. There is no evidence indicating that the parties knew that any profits Airtrust made from the RDU in question would be used to purchase other RDUs, which would in turn be rented to other parties. This head of loss would clearly be excluded by Cl 25 since it is a claim for loss of profits. 157 In the circumstances, we find in favour of PH on Issue 3. The two heads of claim Airtrust has pleaded based on loss of profits (see [24] above) are excluded by Cl 25. Issue 4 – Indemnity Costs 158 We turn finally to deal with CA 96. We do not think there are any substantive grounds for interfering with the Judge’s discretion on this particular issue. Airtrust sought indemnity costs in respect of the entire trial, which took place over three tranches and had 24 witnesses. The specific instances of unreasonable conduct it has alleged, even if made out, only affect a few of the issues in dispute and a few witnesses. In the circumstances, we see no reason to disturb the Judge’s decision not to order an award of indemnity costs against PH. We find in favour of PH on Issue 4. Conclusion 159 For the reasons set out above, we allow CA 234 with costs and dismiss CA 96 with costs. The usual consequential orders will apply. 160 We would also like to express our deep appreciation to Prof Lee for her outstanding scholarship as well as cogent oral submissions that aided this court greatly in arriving at its decision on a particularly thorny area of the common law of contract. Sundaresh Menon  Chao Hick Tin  Andrew Phang Boon Leong Chief Justice   Judge of Appeal  Judge of Appeal Judith Prakash   Tay Yong Kwang Judge of Appeal  Judge of Appeal Tan Chee Meng SC, Josephine Choo and Jeffrey Koh (WongPartnership LLP) for the appellant in Civil Appeal No 234 of 2015 and the respondent in Civil Appeal No 96 of 2016; Tan Chuan Thye SC, Avinash Pradhan, Daniel Gaw, Alyssa Leong and Arthi Anbalagan (Rajah & Tann Singapore LLP) for the respondent in Civil Appeal No 234 of 2015 and the appellant in Civil Appeal No 96 of 2016; Assoc Prof Lee Pey Woan (School of Law, Singapore Management University) as amicus curiae.