1 The appellant, Wong Mei Lee Millie (“Millie”), and the respondent, Jake Ngor Shing Rong (“Jake”), were in a romantic relationship. They jointly purchased a condominium unit in Hillcrest Arcadia (“Property”). Although Jake contributed more to the purchase price, the Property was registered in a 99:1 ratio in Millie’s favour. Their relationship did not last, and a dispute arose over the ownership of the Property.

2 A judge sitting in the General Division of the High Court (“Judge”) in HC/OC 657/2023 (“OC 657”) found that Millie held part of her registered interest in the Property on a resulting trust for Jake based on their respective financial contributions to its acquisition – Ngor Shing Rong Jake v Wong Mei Lee Millie [2025] SGHC 119 (“Judgment”). Millie brings this appeal against that decision.

3 Jake and Millie began their relationship in mid-2018. Sometime in December 2019 (“Purchase Day”), they exercised an option to purchase the Property for $1.865m before a representative from Advance Law LLC (“Advance Law”) named “James”. While the parties referred to him in their evidence as a “lawyer”, his name card exhibited in the evidence below described him as a “Senior Manager”. On the same day, the parties also signed a document prepared by Advance Law stating that they would hold the Property as tenants-in-common in the ratio of 99% (Millie) to 1% (Jake) (“99:1 Document”).

4 The purchase was completed on 20 March 2020 and the parties’ interest in the Property was registered in that 99:1 ratio. Jake and Millie paid the sums of $359,949.42 and $159,678 respectively to the purchase price. The remainder was financed with a mortgage loan. The mortgage payments were serviced primarily by the parties’ Central Provident Fund (“CPF”) moneys and the proceeds from the renting out of the Property.

5 The parties’ relationship subsequently deteriorated, and they separated in November 2020. While they continued to communicate in relation to the Property, including discussing its sale, a dispute eventually arose as to its ownership. Jake filed HC/OA 695/2023 (“OA 695”) seeking a declaration that he holds 71% of the beneficial interest in the Property. Millie disputed this, claiming that she holds her 99% registered interest in the Property both legally and beneficially. OA 695 was converted into OC 657 on Millie’s application in HC/SUM 2367/2023.

6 In the proceedings below, Jake’s case was as follows:

(a) In or around July 2018, Jake expressed to Millie his intention for them to own two properties, so that they could live in one property and rent out the other for passive income. In or around December 2018, Millie indicated that if they were to purchase a property together, the property should be registered legally in her sole name as a form of assurance and security of Jake’s fidelity, and that if he cheated on her, she would be entitled to the whole of the property. Out of love for Millie and to provide her with that assurance and security, Jake agreed to those terms.

(b) On Purchase Day, the parties initially agreed to purchase the Property in equal shares (“50:50 Agreement”). However, in further discussions that same day, Millie brought up Jake’s prior agreement that she would be registered as the legal owner of the Property and that she would be entitled to the whole of the Property if he cheated on her. In these further discussions, it was also agreed that if parties purchased a second property, Jake would transfer his “ownership” of the Property to Millie and have the second property registered solely in his name, so that they would not have to pay any Additional Buyer Stamp Duty (“ABSD”). Following these discussions, Jake agreed to Millie being registered as the 99% legal owner of the Property on the condition that she “shall be wholly entitled to the Property” if: (i) he “was disloyal to her during their relationship”; or (ii) he “intended to purchase a [second] property registered solely under his name and thus decided to transfer his interest in the Property to [Millie]”. Jake did not intend to give the Property to Millie save in these two circumstances. We refer to these conditions as the “Cheating Condition” and “2nd Property Condition” respectively.

(c) The parties did not know or understand the distinction between legal and beneficial interests when they purchased the Property. They had not discussed the proportion of their respective beneficial interests in the Property and had, in laymen terms, agreed to co-own the Property together.

(d) Since Jake’s financial contributions to the purchase price of the Property were significantly more than Millie’s, a presumed resulting trust arose in his favour. In the alternative, there was sufficient evidence to show that Jake did not intend to benefit Millie with his financial contributions to the Property, and thus an actual resulting trust arose in his favour. Conversely, there was no common intention between the parties for Millie to be the legal and beneficial owner of 99% of the Property.

(e) As neither the Cheating Condition nor the 2nd Property Condition were satisfied, Jake beneficially owned the Property in accordance with his financial contributions.

7 We pause to emphasise a couple of points about the Cheating Condition and how it featured in the case. As will be evident from our analysis, it was not consistently framed by Jake, and this has a bearing on our assessment of the evidence. More importantly, because Jake advanced his case based on a resulting trust, it was a key element of that case that Jake had to show he did not intend that the beneficial interest in the Property would vest in Millie upon the completion of the purchase. Hence, Jake’s case was advanced on the basis that Millie’s beneficial interest would only vest in the future upon an event occurring, namely if he was unfaithful to her. We emphasise this because it is quite a different situation from saying that that the beneficial interest in the Property did vest in Millie upon completion subject to Millie being obliged to transfer it to Jake in the event they separated on grounds other than Jake’s infidelity. The latter, which for convenience we refer to as the “possible alternative Cheating Condition”, was not how Jake advanced his case and is therefore not an issue that falls for our determination.

8 Millie’s case was as follows:

(a) When the parties discussed their plans to purchase two properties, it was on the basis that each of them would own one property. They never discussed or agreed that they would own parts of each property based on their financial contributions.

(b) Although the parties initially agreed that Millie would fully own their first property, when it came to purchasing the Property, Jake needed to own at least 1% of the Property to use his CPF moneys for the purchase, and Millie was comfortable with this arrangement.

(c) Millie was insecure about her relationship with Jake due to Jake’s history of infidelity. Jake promised he would give her 99% ownership of the Property to assure her that he wanted to settle down with her and would not cheat on her. Millie’s 99% ownership of the Property did not depend on whether Jake cheated on her. She would not have been assured if her 99% ownership was only in name as Jake could simply change his mind and take it away from her at his whim and fancy.

(d) On Purchase Day at Advance Law, Jake asked James about registering the Property in equal shares, but this was rejected by Millie. Jake also asked James what would happen if the Property was registered in a 99:1 ratio and the parties broke up. James told the parties to think carefully because ownership of the Property would follow the registered legal title. The parties then stepped out of Advance Law and had a short discussion. They confirmed that they were proceeding with their earlier agreement where they would own the Property in the 99:1 ratio.

(e) In the circumstances, when the Property was purchased in December 2019, the parties intended for the 99:1 ratio to reflect their legal and beneficial ownership. The 99:1 ratio was further intended to enable parties to save on ABSD when Jake transfers his 1% interest to Millie and purchases a second property. There was no agreement between the parties that Millie would only be entitled to the Property if Jake cheated on her during their relationship.

(f) The parties had a common intention for her to own 99% of the Property both legally and beneficially. The parties did not distinguish between legal and beneficial ownership and treated the registered interest as equivalent to ownership.

9 The Judge found that Millie held part of her registered interest in the Property on a resulting trust in favour of Jake based on their respective financial contributions towards the purchase price.

10 The Judge referred to the decision in Khoo Phaik Ean Patricia v Khoo Paik Eng Katherine [2025] 1 SLR 758 (“Patricia Khoo”) at [60], where this court set out four scenarios where “A” gratuitously transfers property into the joint names of A and “B” in law for no consideration, differentiated based on A’s intention. The Judge held that these four scenarios were broadly transferable to all cases of resulting trust claims (Judgment at [13]).

11 What is relevant for present purposes is the second scenario, which this court in Patricia Khoo described as follows:

12 However, the Judge described the second scenario in broader terms (Judgment at [13(b)]):

13 The Judge found that Jake did not intend to immediately and unconditionally benefit Millie with his financial contributions to the Property, but only if he cheated on her. Thus, a resulting trust over the Property subsisted in Jake’s favour (Judgment at [19] and [48]). While the Judge did not expressly state as such, it appears he was of the view that the resulting trust he found fell within the scope of the second scenario in Patricia Khoo as described by him.

14 Given his finding that Jake had no intention to benefit Millie with his financial contributions to the Property, the Judge rejected Millie’s contention that there was a common intention for her to own 99% of the Property (Judgment at [49]).

15 As to the parties’ financial contributions, the Judge rejected Jake’s contention that the parties had an agreement for the mortgage payments to be partially serviced with monthly payments of $1,350 from Jake’s CPF account and $250 from Millie’s CPF account. In the absence of a prior agreement on how liability for the mortgage loan would be split between the parties, each party was deemed to have contributed half the loan amount as the mortgage loan was taken out in their joint names. Combined with the parties’ upfront payments, their financial contributions were found to be in the ratio of 55.22% (Jake) to 44.78% (Millie). Since Jake held a 1% legal share of the Property, Millie held 54.22% of her legal share on a resulting trust for Jake (Judgment at [59]–[63]).

16 The Judge directed the parties to address the issue of illegality although this was not pleaded by Millie as a defence to Jake’s resulting trust claim (Judgment at [65]). The Judge was of the view that, as the registered 99:1 ratio did not reflect the parties’ beneficial ownership of the Property, illegality in the form of under-stamping or tax evasion could arise (Judgment at [76]). The Judge, applying the framework developed in Lau Sheng Jan Alistair v Lau Check Joo Richard [2023] 5 SLR 1703 (“Alistair Lau”) at [67]–[81] (“Alistair Lau framework”), found that the resulting trust over the Property arose as an incidental consequence of an illegal purpose because Jake contemplated under-stamping, but not tax evasion, at the time the Property was purchased (Judgment at [91]–[97]). He nonetheless determined that denying Jake’s claim would be a disproportionate response to the illegality (Judgment at [100]–[114]). Accordingly, the Judge declared that Millie held 54.22% of the Property on trust for Jake (Judgment at [115]).

17 Millie’s arguments on appeal are broadly as follows:

(a) In finding that a resulting trust arose over the Property, the Judge decided against the weight of the evidence. The Judge failed to consider the parties’ evidence on various points such as what transpired on Purchase Day. The Judge skipped over the issue of common intention with no analysis and erred in finding that Jake’s lack of intention to immediately benefit Millie with his financial contributions also meant that there was no evidence of a common intention.

(b) The Judge erred in expanding the concept of “conditional resulting trusts”, recognised in the second scenario discussed in Patricia Khoo in relation to resulting trusts which terminate upon a transferor’s death, to include resulting trusts which terminate on other conditions subsequent such as a person’s infidelity to his partner.

(c) In the alternative, the Judge erred in his characterisation of Jake’s illegality as one of under-stamping as Jake’s evidence referred to the avoidance of ABSD specifically. The Judge also erred in finding that it was disproportionate to deny Jake’s resulting trust claim.

18 Jake broadly aligns himself with the Judge’s analysis in the Judgment.

19 The issues to be determined are:

20 In determining the above issues, we are cognisant that an appellate court’s power of review with respect to findings of fact is limited because the trial judge is generally better placed to assess the veracity and credibility of witnesses, especially where oral evidence is concerned. However, where it can be established that the trial judge’s assessment is plainly wrong or against the weight of the evidence, the appellate court can and should overturn any such finding. Furthermore, where a particular finding of fact is not based on the veracity or credibility of the witness, but instead, is based on an inference drawn from the facts or the evaluation of primary facts, the appellate court is in as good a position as the trial judge to undertake that exercise. In so doing, the appellate court will evaluate the cogency of the evidence given by the witnesses by testing it against inherent probabilities or against uncontroverted facts (Tat Seng Machine Movers Pte Ltd v Orix Leasing Singapore Ltd [2009] 4 SLR(R) 1101 at [41]).

21 We begin by restating the four scenarios discussed in Patricia Khoo where A gratuitously transfers property into the joint names of A and B in law for no consideration, differentiated based on A’s intention:

(a) The first scenario is where the evidence establishes that A intended to benefit B immediately. Here, there is no occasion for a trust to arise and there is no separation of the equitable interest in the property from the legal interest (Patricia Khoo at [60(a)] and [64]).

(b) The second scenario is where the evidence establishes that A did not intend to benefit B immediately, but that B will inherit whatever is left in the estate when A dies. This court in Patricia Khoo noted that the juridical basis for such an arrangement had not been clearly articulated but observed that it could potentially be explained based on a resulting trust analysis (Patricia Khoo at [60(b)] and [66]–[76]).

(c) The third scenario is where the evidence establishes that A did not intend to benefit B at all. Here, an actual resulting trust arises which responds to A’s lack of intention to benefit B (Patricia Khoo at [60(c)] and [77]).

(d) The fourth scenario is where there is either no or insufficient evidence of A’s intention. Here, the parties’ respective interests in the property would turn on the application of the presumptions of resulting trust and advancement as may be applicable (Patricia Khoo at [60(d)] and [78]).

As we explain below, we are of the view that the present case falls within the first scenario.

22 Both parties referred to Chan Yuen Lan v See Fong Mun [2014] 3 SLR 1048 (“Chan Yuen Lan”) at [160], where this court set out a framework for analysing disputes over the beneficial ownership of a property. However, as this court cautioned in Patricia Khoo, the framework should not be applied mechanistically – depending on the facts of each case, the court does not have to go through the motion of applying the framework from start to finish and can instead simply zero in on the key issues in dispute (Patricia Khoo at [57]).

23 As is evident from the scenarios outlined in Patricia Khoo and the framework in Chan Yuen Lan, the most critical question where a resulting trust is alleged is the intention of the transferor at the time of the transfer of the property to, or when it was registered in the name of, the transferee. Where the transferor is the claimant, it is for him to discharge the legal burden of persuading the court that he did not intend to benefit the transferee at the material time, and if he fails to discharge that burden, he fails in establishing a resulting trust in his favour. It is in the context of determining the transferor’s intention that the presumptions of advancement or resulting trust may come into play. As this court noted in the fourth scenario in Patricia Khoo, the presumptions may apply where there is either no or insufficient evidence of the transferor’s intention. Consequently, recourse to presumptions is neither necessary nor permissible where the evidence before the court sufficiently establishes the transferor’s intention (Patricia Khoo at [78]).

24 The Judge described the presumption of resulting trust in the following terms (Judgment at [16]):

25 We agree that presumptions operate as evidentiary tools, but that label does not explain how they should be applied or how their application impacts the analysis of the evidence in relation to the transferor’s intention. In this regard, cases have not applied the presumptions clearly or consistently. Thus, for example, in Vandervell v Inland Revenue Commissioners [1967] 2 AC 291, Lord Upjohn said that “[a]ll the relevant facts and circumstances can be considered in order to ascertain [a transferor’s] intention with a view to rebutting this presumption [of resulting trust]” (at 312F). But he also said that “[i]n reality the so-called presumption of resulting trust is no more than a long stop to provide the answer when the relevant facts and circumstances fail to yield a solution” (at 313D). These are, on their face, two seemingly contradictory propositions – if the presumptions only apply where there is insufficient evidence to establish the transferor’s intention, then there is no question of analysing the evidence to see if the presumptions are rebutted, for that analysis has already been undertaken prior to the application of the presumption.

26 The historical origins of the presumption of resulting trust are explained in Lau Siew Kim v Yeo Guan Chye Terence [2008] 2 SLR(R) 108 (“Lau Siew Kim”) at [39]–[41] and we do not propose to repeat the same. The locus classicus for a modern restatement of the circumstances in which the presumption of resulting trust apply can be found in the judgment of Lord Browne-Wilkinson in Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669 at 708A–G, where he said:

27 It is true that the above passage suggests that the inquiry begins with the presumptions, and the evidence is later evaluated to determine if the presumptions are rebutted. However, it is not clear to us that this approach was the subject of a firm consensus among the earlier decisions of the chancery courts. We refer to a few cases to illustrate our point:

(a) A good number of cases involved a transferor who arranged for the transferee to have legal title to the property, and after the transferor’s passing, a dispute arose as to whether the transferee enjoyed the beneficial interest in the property or held it on trust for the estate (see, eg, Dyer v Dyer (1788) 30 ER 42, Rider v Kidder (1805) 32 ER 884 and Fowkes v Pascoe (1875) 10 LR 10 Ch App 343). But that is a paradigm situation (at least in the past) where direct evidence of the transferor’s intention is typically not available, and so the resort to the presumptions of resulting trust and advancement as the first port of call in these cases is explicable. We note that this was also how this court in Chan Yuen Lan rationalised the application of the presumptions in Lau Siew Kim, highlighting that the transferor in that case was no longer alive at the time of the proceedings and there was no direct evidence of his actual intentions (see Chan Yuen Lan at [51]).

(b) This is reinforced by the approach taken in Wheeler v Smith (1860) 65 ER 928 (“Wheeler”). In this case, the testator purchased bonds in joint names with his sister. After his death, a dispute arose as to whether the beneficial interest in the bonds resided with the sister or with the estate. The court ultimately found that the presumption of resulting trust was rebutted by evidence that the testator intended for the bonds to be gifts to the sister. There was in evidence various letters which the testator had written to the sister during his lifetime, wherein there was no mention of any trust, and wherein he informed the sister how he wished for her to distribute the interest but also said that he did not intend to bind her by way of legal document. On its face, Wheeler supports the view that the evidence is only evaluated after the presumption is applied, and this may be explicable on the basis that the testator was deceased. But on closer examination, the emphasis of the judgment was on the evaluation of the evidence of the testator’s intentions (particularly his letters) which led the court to find that “the case is wholly without evidence to prove that any valid trust was created or attached to the [sister’s] legal interest” (at 308). Notwithstanding the references to the presumption of resulting trust being rebutted, the substance of the court’s approach in Wheeler is consistent with our view that the first port of call must be the evidence as to the testator’s intention.

(c) Standing v Bowring (1885) 31 Ch D 282 (“Standing”) is a more difficult case. In this case, the plaintiff purchased stocks in joint names with the defendant but did not inform the defendant of this. Two years later, the plaintiff’s solicitors sent a letter to the defendant asking for the stocks to be re-transferred to the plaintiff’s name. This was the first time the defendant learned of the stocks. The plaintiff filed proceedings to compel the defendant to make the re-transfer. The court unanimously held that there was no basis for the re-transfer as the stocks were intended as gifts and the beneficial interest resided with the defendant. Lord Halsbury LC’s judgment was solely concerned with the issue of whether the gift was effective notwithstanding the defendant’s lack of knowledge of the stocks. Cotton LJ’s judgment began with the presumption of resulting trust, but he concluded that the presumption was rebutted by the evidence that the plaintiff intended to confer a benefit on the defendant. Lindley LJ’s judgment made no reference to the presumption of resulting trust; instead, his analysis went straight into the evidence of the plaintiff’s intentions, observing as follows (at 289):

We prefer the approach of Lindley LJ. Where evidence of the transferor’s intention is available, it is the evaluation of that evidence which ought to be the first analytical step.

28 In any event, regardless of what the approach was in the 18th and 19th centuries, the modern approach must foreground the analysis of the evidence of the transferor’s intention. This was explained by Lord Briggs in the Privy Council’s decision in Gany Holdings (PTC) SA v Khan (2018) 21 ITELR 310 at [17]:

29 After citing the above passage, the learned editors of Paul Matthews et al, Underhill and Hayton: Law of Trusts and Trustees (LexisNexis, 20th Ed, 2022) at paras 26.36–26.37 elaborated as follows:

30 We agree with the views stated above. They make the important points that: (a) the court first analyses the evidence from which the transferor’s intention may be inferred and recourse to presumptions may only be necessary in cases where there is no or insufficient evidence from which such an inference may properly be drawn; and (b) in modern times, it will be rare that such evidence on intention will be unavailable. If anything, the opposite is more likely true – with e-mails and text messages (among other technologies) creating a comprehensive and contemporaneous documentary record of parties’ interactions, there is nowadays an “information explosion” in most cases (see Chief Justice Sundaresh Menon, Supreme Court of Singapore, “The Complexification of Disputes in the Digital Age”, Goff Lecture 2021 (9 November 2021) at paras 17–18). Even in cases where the transferor is deceased or unable to give evidence, contemporaneous documentary evidence will likely exist to shed light on his intentions at the material time, and such evidence may be admissible under various exceptions to the hearsay rule, including ss 32(1)(c) or 32(1)(j)(i) of the Evidence Act 1893 (2020 Rev Ed) (“EA”).

31 Further, in Tan Yok Koon v Tan Choo Suan [2017] 1 SLR 654, this court held that the rule in Shephard v Cartwright [1955] AC 431, that subsequent evidence of the transferor’s intention in the transferor’s favour is inadmissible, should be relaxed, such that the breadth of evidence which the court can now consider in determining the transferor’s intention is greater. As noted by Andrew Phang Boon Leong JA (as he then was) (at [110]):

32 The likely availability of admissible evidence as to the transferor’s intention largely explains why presumptions have been relegated to tools of last resort. In Teo Siew Har v Lee Kuan Yew [1999] 3 SLR(R) 410 at [29], Chao Hick Tin JA described the presumption of advancement as “an evidential instrument of last resort where there is no direct evidence as to the intention of the parties rather than as an oft-applied rule of thumb.” This was cited with approval by V K Rajah JA in Lau Siew Kim at [59] where he further observed that “the presumption of advancement will operate only where there is no direct evidence that may reveal the intention of the parties; only then will there be any necessity to infer or presume intention.” In Chan Yuen Lan at [49]–[52], Rajah JA referred to that passage in Lau Siew Kim in relation to both the presumption of advancement and the presumption of resulting trust, and reaffirmed the view that “there can generally be no justification to resort to presumptions, however much they may reflect the norms, customs and mores of society, if the court is faced with the actual intentions and desires of the transferor.”

33 In the circumstances, the first step must be to analyse the evidence of the transferor’s intention. The evidence may be biased, uncorroborated, contradictory or difficult to comprehend. But the relevance, reliability, credibility and persuasiveness of the evidence must be evaluated, no matter how challenging the exercise may be. Indeed, where the transferor is available to give direct evidence of his intention, it is difficult to conceive when the application of the presumptions will be necessary.

34 The recent decision of this court in Patricia Khoo exemplifies the correct approach. In that case, the testator executed a will stipulating that two bank accounts, which were then in his sole name, formed part of his residuary estate and should be distributed equally among his four children. He later arranged for the two accounts to be converted into joint accounts with his wife and eldest daughter, and for the execution of a codicil to amend the will in accordance with his discussions with his eldest daughter (which did not affect the way the joint accounts were to be distributed under the will). After the testator passed away, a dispute arose between the family members over the ownership of the joint accounts. The question before the court was whether the wife and the eldest daughter were beneficially entitled to the joint accounts upon the testator’s death, or whether they held them on a resulting trust for the estate. Even though there was (and could be) no direct evidence from the testator on his intention at the material time, there was extensive evidence in the form of, among others, his original will, his discussions with his eldest daughter, and his execution of the codicil. This court analysed the evidence and concluded that the testator’s intention was to retain the beneficial interest in the joint accounts, to distribute them equally among his four children in accordance with his will, and not to gift them to the wife and the eldest daughter. Crucially, this court did not rely on the presumption of resulting trust, nor did it analyse the evidence to determine if the presumption was rebutted.

35 It may be contended that so long as the evidence is properly evaluated, it matters not whether this is done before or after the presumptions are applied. But there is an important difference. Where the presumptions are applied as the first analytical step, the evidential burden shifts to the other party to rebut the presumption. Thus, for example, where a transferor-claimant asserts a resulting trust over a property and relies on the parties’ financial contributions to the purchase price to invoke the presumption of resulting trust, the evidential burden would shift to the transferee-defendant to prove that the intention of the transferor-claimant was to benefit the transferee-defendant with those financial contributions, notwithstanding that the legal burden is always with the transferor-claimant to make good his case of a resulting trust. In contrast, where the presumptions are applied as a tool of last resort, the transferor-claimant bears the evidential burden from the start to prove that he did not intend to benefit the transferee-defendant with his financial contributions. This will require the transferor-claimant to adduce sufficient evidence necessary to discharge or least shift that evidential burden. Where relevant evidence is notionally available but not adduced by the transferor-claimant, it is for the court to determine, in accordance with established principles, whether an adverse inference should be drawn against the transferor-claimant in respect of the failure to adduce that evidence.

36 The present case is such an instance where there should be no recourse to presumptions. Like the Judge, we are of the view that the evidence is sufficient to reveal the parties’ actual intentions as to the ownership of the Property. However, we disagree with the Judge’s evaluation of the evidence. In our view, the evidence shows that the parties intended for the 99:1 ratio to reflect their legal and beneficial ownership of the Property. It follows that Jake did intend to benefit Millie with his financial contributions, and a resulting trust did not arise over the Property. The Judge’s evaluation of the evidence was, with respect, incorrect because he did not properly address the numerous difficulties and inconsistencies in Jake’s case, and appears not to have considered or given weight to Millie’s evidence without explaining why that evidence was not credible. We turn to analyse these points in detail.

37 We begin with the parties’ evidence that they did not know or understand the distinction between legal and beneficial interests at the time the Property was purchased. In cross-examination, Jake admitted that he did not know or understand the difference between legal and beneficial interests, that he did not know he could hold a beneficial interest in a property based on his financial contributions to its acquisition (ie, a resulting trust), and this was a concept he learned only when the proceedings below began. Likewise, in her evidence, Millie attested that, prior to the proceedings below, she did not know that one’s beneficial ownership could be different from what was legally registered.

38 This suggests that when the parties signed the 99:1 Document on Purchase Day, they could not have contemplated or intended to own the Property in a manner different from the 99:1 ratio stated therein. In other words, their choice of the 99:1 ratio reflected their understanding of how the Property would be owned. This was even though they knew that Jake would be contributing most of the purchase price. The 99:1 Document is therefore objective evidence of a common intention between the parties to legally and beneficially own the Property in a 99:1 ratio, and of an intention by Jake to benefit Millie with his financial contributions.

39 To be clear, we are not saying that whenever laypersons sign a document specifying a specific ownership ratio, that specified ratio is necessarily conclusive as to their common intention on the ownership of the property. Nor are we suggesting that they must understand the concept of beneficial interest for a resulting trust to exist. What is critical is whether there is evidence to indicate that the registered interest is not reflective of the parties’ intentions as to ownership. This could, for example, be in the form of discussions referring to a different ownership ratio (with an explanation as to why the parties decided to register it differently), or the exercise of ownership in a manner different from the registered interest (such as control over the income earned from the property). As we will show, there is no evidence to suggest that Jake and Millie intended to own the Property in a manner different from their registered interests. They did not even discuss, much less agree to, ownership in any other manner, save for Jake’s assertion of the 50:50 Agreement which we deal with later. In the circumstances, the fact that the parties registered their legal interests in a 99:1 ratio in Millie’s favour, whilst not being aware that they could beneficially own the Property in a manner different from their registered interests, strongly supports Millie’s case that the parties commonly intended for her to legally and beneficially own 99% of the Property. This is an important point which the Judge failed to consider.

40 In this regard, Jake made an important concession, which the Judge did not refer to in the Judgment. In Jake’s second affidavit in OA 695 (“OA 695 Second Affidavit”), he stated that the 99:1 ratio was not a “sham”. In cross-examination, Jake similarly stated that the 99:1 arrangement was “genuine” and not a “sham”, and that he was not trying to trick the Inland Revenue Authority of Singapore into believing that he was only a 1% owner of the Property. Jake did not qualify or resile from these statements. In our view, the only logical interpretation of that evidence is that when the parties signed the 99:1 Document, they genuinely intended for the 99:1 ratio to reflect their legal and beneficial ownership of the Property. This further supports Millie’s case.

41 Before us, counsel for Jake, Mr Tito Isaac (“Mr Isaac”), submitted that when Jake signed the 99:1 Document, he did not intend for the 99:1 ratio to reflect the parties’ beneficial interests, but was content to have their beneficial interests remain unquantified and subsequently determined by reference to their financial contributions. We reject this argument as it is inconsistent with Jake’s own evidence that he did not know or understand the distinction between legal and beneficial interests, and was not aware that his beneficial interests could be determined by reference to his financial contributions.

42 Jake’s case was that the parties initially came to the 50:50 Agreement on Purchase Day (see [6(b)] above). The existence of the 50:50 Agreement was material to Jake’s case as it would undermine Millie’s position that the parties’ consistent understanding was that she would legally and beneficially own 99% of the Property. Having asserted it, the burden was on Jake to prove the 50:50 Agreement (s 105 of the EA). The Judge made no finding as to the existence of the 50:50 Agreement (Judgment at [20]). Save for Jake’s bare assertion, there is no evidence of the 50:50 Agreement. On the contrary, the evidence suggests that it was entirely contrived by Jake and this seriously undermined his case.

43 Jake’s first affidavit in OA 695 (“OA 695 First Affidavit”), statement of claim (“SOC”) and affidavit of evidence-in-chief (“AEIC”) merely contained a bare assertion that the parties initially came to the 50:50 Agreement on Purchase Day, without any further detail or explanation. At the trial, Jake explained how the 50:50 Agreement came about. He claimed that the issue of Millie’s insecurities only arose “from time to time”. On Purchase Day, the parties were excited to buy a house as a couple and the issue of Millie’s insecurities did not arise, and so they agreed to own the Property equal shares. This explanation is unconvincing. Jake’s own case is that Millie was insecure throughout their relationship and she had as early as December 2018 insisted that the property they purchased together should be registered in her sole name (see [6(a)] above). Millie’s insecurities were also documented in several text messages between the parties on 19 July 2018 and 3 December 2018, which we discuss below. This suggests that Millie considered her ownership of the property to be purchased to be especially important to her. It is therefore unbelievable that the issue of Millie’s insecurities did not initially arise on Purchase Day. Further, Jake’s evidence was also that the parties had no discussions about their percentage ownership of the Property prior to Purchase Day. If so, it is unbelievable that, without any discussions on the issue of ownership, Millie would have, contrary to her earlier text messages, simply agreed to the 50:50 Agreement on Purchase Day, only to raise her insecurities a few minutes later after leaving Advance Law’s offices.

44 When pressed in cross-examination, Jake claimed for the first time that the 50:50 Agreement was reflected in a document prepared by Advance Law which they initially signed on Purchase Day before James (“50:50 Document”). According to Jake, after Millie raised her insecurities and Jake’s previous promise to register 99% of the Property in her name, Jake called James to inform him that they wished to revise their ownership ratio, and they returned to Advance Law to sign the 99:1 Document. Jake claimed that he tried to obtain the 50:50 Document from Advance Law for the purposes of the proceedings below, but James told him that the firm no longer had it.

45 This belated reference to the 50:50 Document severely undermined the credibility of Jake’s case. Given that Jake referred to the 50:50 Agreement in both his SOC and AEIC, it would have been obvious that the 50:50 Document would be a key piece of evidence. His failure to mention it earlier is inexplicable and strongly suggests that it did not exist.

46 Even if Advance Law no longer possessed the 50:50 Document – and we note Jake did not procure even a letter from Advance Law confirming this – Jake could have called James to give evidence as to the existence of the 50:50 Document and the 50:50 Agreement. An adverse inference that the 50:50 Agreement did not exist should therefore have been drawn against Jake for his failure to call James as a witness (see Illustration (g) to s 116 of the EA and Sudha Natrajan v The Bank of East Asia Ltd [2017] 1 SLR 141 at [18]–[20]). At the trial, Mr Isaac highlighted that James was included in Millie’s witness list, but she chose not to call him as a witness. Counsel for Millie, Mr Terence Tan, explained that this was because James was not willing to testify. The fact that James was not called by Millie does not absolve or excuse Jake from doing so. While James’ evidence would have been relevant to both parties’ version of events, it was Jake who had the burden of proving the 50:50 Agreement.

47 In contrast, Millie’s evidence on what transpired on Purchase Day was clear and consistent. Her case was always that she wanted to own 99% of the Property, and Jake made a promise to her to that effect. Therefore, on her account of Purchase Day, when Jake brought up the possibility of registering the Property in equal shares, Millie rejected it, and the parties then agreed to proceed with their earlier agreement to own the Property in a 99:1 ratio (see [8(d)] above). On her case, there was no 50:50 Agreement or any document evidencing the same. In fact, if Jake’s testimony that he and Millie did not discuss the precise split of the Property prior to the meeting with James is to be believed, then Millie’s case that she rejected the possibility of registering the Property in equal shares is more logical, for that would have been a significant and abrupt departure from her position that the Property was to be registered in her sole name as assurance of Jake’s fidelity (see [43] above).

48 We accept that the unreliability of Jake’s case regarding the 50:50 Agreement is not in itself conclusive, as it is one-step removed from the question of the parties’ intentions when they executed the 99:1 Document and registered the Property in the 99:1 ratio. Nonetheless, this is a piece of evidence which the Judge ought to have considered, but failed to do so, when evaluating the overall credibility of Jake’s case.

49 Apart from the 2nd Property Condition which we address later, Jake’s case was that he only agreed to Millie beneficially owning her 99% registered interest in the Property if he cheated on her, ie, the Cheating Condition. This was a critical aspect of Jake’s case as it meant that he did not intend to benefit Millie with his financial contributions immediately (see [6(b)] above). The burden was therefore on Jake to prove the Cheating Condition as framed by him. We reject Jake’s case on the Cheating Condition for three reasons: (a) Jake’s explanation of the condition was inconsistent and unreliable; (b) the condition is illogical; and (c) the existence of the condition is not borne out by the evidence. We address each reason in turn.

50 Initially, Jake suggested that the Cheating Condition was that if he cheated on Millie, she would fully own the Property.

(a) In Jake’s OA 695 First Affidavit and OA 695 Second Affidavit, Jake said that if he cheated on Millie, she “could retain the [Property] in her sole name” [emphasis added].

(b) In fact, in the same paragraph in Jake’s OA 695 First Affidavit, he stated that if the parties were to purchase a second property, “I could transfer 1% of the [Property] held by me to Millie”, and also stated that if he cheated on Millie, he “would lose my share in the Property”. When these two statements are read in the context of the same paragraph, “my share” clearly meant his 1% registered interest.

(c) In Jake’s SOC, he said that Millie “shall be wholly entitled to the Property” if he “was disloyal to her during their relationship” [emphasis added].

(d) In Jake’s AEIC, Jake said that Millie “would be wholly entitled to the [Property] if I ever cheated on her” [emphasis added].

51 As the Property was already registered in a 99:1 ratio, the suggestion appears to be that Jake would discharge his obligations under the Cheating Condition by simply transferring his 1% registered interest to Millie if he cheated on her.

52 Jake’s case on the Cheating Condition morphed at trial. He suggested that the Cheating Condition was that if he cheated on Millie, he would not contest her 99% registered interest, and only then would she beneficially own 99% of the Property.

(a) When questioned on what was to happen if he cheated on Millie, Jake explained that “I would not claim back any cent … I will not contest. Basically … I would follow as such, that it is 99 per cent under her name.”

(b) Likewise, when questioned on what he meant by “wholly entitled” in his SOC, Jake explained that “[i]t means to say that I will not contest for the 99 per cent despite the differences of the financial contributions on my end.”

(c) He further testified that “if I ever cheat on her, then the 99 per cent will actually vest under her name.”

53 These statements represent a departure from his original version of the Cheating Condition that Millie would fully own the Property if he cheated on her. No explanation was given for this inconsistency and the Judge failed to consider this. Given that the Cheating Condition was central to Jake’s case, this inconsistency severely undermined that case.

54 Jake repeatedly stated that the Cheating Condition operated as a form of “assurance and security” to Millie of his fidelity (see [6(a)] above). But if Millie’s 99% registered interest was only nominal, and she held a sizeable share of it on trust for Jake, then she would have been afforded no meaningful “assurance and security” at all for Jake could always demand for his interest at any time and it would be for Millie to prove that Jake had cheated on her.

55 The Judge was nonetheless persuaded by Jake’s explanation that Millie’s 99% registered interest provided assurance and security by erecting “practical hurdles” against Jake recovering the Property, and this served as a deterrent against any infidelity on Jake’s part (Judgment at [31]). Before us, Mr Isaac pointed to these very proceedings as an example of the “practical hurdles” Jake would face. With respect, this does not make sense. On Jake’s own case, if he cheated on Millie, he would “not contest” Millie’s 99% registered interest, and so there can be no question of facing “practical hurdles” in recovering his share of the Property. If he did not cheat on Millie and they remained in a relationship, then (as Jake admitted in cross-examination) there would be no need for him to make a claim on the Property, and there would likewise be no question of facing “practical hurdles”. We therefore do not see how the existence of “practical hurdles” in commencing proceedings operates as a deterrent against Jake cheating.

56 In contrast, Millie’s case that her 99% beneficial ownership of the Property was not conditional on Jake cheating is more logical. Her position was that she wanted them to remain in a relationship. As she was admittedly insecure about Jake’s fidelity, she wanted 99% ownership of the Property as a demonstration of Jake’s commitment. If Jake cheated on Millie and they broke up, Jake would be left holding just 1% of the Property and would no longer be entitled to regard the Property as a joint investment. This deterrence was therefore precisely what Millie wanted – the Property would be held by her as ransom and Jake would only enjoy its benefits if he did not cheat and remained in a relationship with her.

57 What then happens if the parties were to split for reasons other than Jake’s infidelity? This is where the possible alternative Cheating Condition referred to at [7] may have been relevant. It appears that the parties might have briefly alluded to this scenario in their text messages reproduced at [61] and [64] below. In essence, there were discussions about how Millie would “give [Jake] back everything” in relation to the Property if they separated “on good terms”, as she had apparently done with her previous boyfriend. There were also discussions about how, if the parties married, Millie wanted Jake to fulfil his “promise about putting all [his] money with [her], and she was willing to “sign a contract” stating that she “will not take half [of Jake’s] assets if [they] split”, but if Jake cheated on her it would be “a whole diff[erent] story”. If these discussions amounted to a binding agreement (though we stress we make no finding on this), that may have provided Jake a potential remedy. However, we note that any notion that Millie would give assets back to Jake would be inconsistent with Jake’s resulting trust claim and would support the view that these assets belonged to Millie. In any event, that was not the case Jake advanced, and we say no more about this.

58 Importantly, the Cheating Condition is not borne out by the evidence, particularly the parties’ contemporaneous text messages. On the contrary, the text messages are more consistent with Millie’s case.

59 We begin with the parties’ text messages on 19 July 2018 (“19 Jul 2018 Messages”), shortly after their relationship began, where they discussed Millie’s insecurities and their plan to own two properties. We reproduce the salient portions of the 19 Jul 2018 Messages below.

60 When Millie asked if the two properties would be “u one I one”, Jake said “[y]es”. Later, Millie said “[i]f the hse isn’t mine I won’t bother / So put under my name at least I will bother”. This showed that her ownership of one of the properties was important to her. Nowhere in this or other exchanges did Jake suggest that Millie’s ownership would only be nominal. Crucially, despite Millie’s insecurities having already been raised, Jake did not suggest that Millie’s ownership was conditional on him cheating on her. While the Judge referred to the 19 Jul 2018 Messages (see Judgment at [22]), he did not appear to appreciate the relevance of the messages discussed above.

61 Later, on 3 December 2018, the parties again discussed the matter of Millie’s insecurities and their plan to purchase a property (“3 Dec 2018 Messages”). We reproduce the salient portions of the 3 Dec 2018 Messages below.

62 The only messages that could ostensibly be interpreted as evidence of the Cheating Condition as framed by Jake are Jake’s messages “[a]nd we said that if we do it under your name, any cheating on my side, all these go to you” and “[i]f I cheat, take it away from me by all means”. We make three observations. First, these messages are ambiguous. “[A]ll these go to you” and “take it away from me by all means” could refer to Jake transferring his 1% registered interest to Millie such that she would fully own the Property, as per the original conception of the Cheating Condition (see [50]–[51] above). Second, even if we interpreted these messages to mean that Millie would only beneficially own 99% of the Property if Jake cheated on her, it is important that nowhere in the 3 Dec 2018 Messages did Millie express her concurrence to this condition. If the Cheating Condition “was the agreement all along” as asserted by Jake in cross-examination, then it is surprising that Millie did not expressly affirm the condition but rather said “u think abt it first”. Third, these messages are to be seen in the light of the fact, as both parties accept, that they did not understand any difference between legal and beneficial ownership, and in that case, the messages are truly ambiguous as to whether they were referring to the Cheating Condition, as Jake contends, or the possible alternative Cheating Condition, which he did not advance and which would not aid his case for a resulting trust. The ambiguous nature of these messages and the lack of affirmation by Millie makes it doubtful whether the Cheating Condition ever existed.

63 We also highlight that the 3 Dec 2018 Messages were exchanged approximately one year before the Property was purchased. Significantly, Jake has not been able to point to any other text messages between 3 December 2018 and Purchase Day as evidence of the Cheating Condition even though, on Jake’s case, it was an important term of the purchase of their first property together.

64 Subsequently, on 20 and 21 July 2020, after the Property was purchased, the parties had an extended quarrel regarding Millie’s insecurities (“20–21 Jul 2020 Messages”). We reproduce the salient portions of the 20–21 Jul 2020 Messages below.

65 We discuss the 20–21 Jul 2020 Messages in greater detail at [82] below. For present purposes, we observe that while the parties spoke at length about Millie’s insecurities and her having a 99% registered interest in the Property, at no point did they mention the Cheating Condition or allude to the notion that Millie would only fully own or own 99% of the Property if Jake cheated on her. On Jake’s case, the Cheating Condition was an important facet of the parties’ ownership of the Property and an important source of assurance and security to Millie of his fidelity, so it is surprising that it was not mentioned at all in this extended exchange regarding Millie’s insecurities.

66 Overall, the text messages between the parties do not support the existence of the Cheating Condition. Save for two discrete references by Jake in the 3 Dec 2018 Messages, which were ambiguous in themselves, there was no reference to the Cheating Condition in the text messages. The evidence suggests that it did not exist.

67 Our conclusion is reinforced by the fact that the Cheating Condition was not mentioned in Jake’ letter of demand to Millie. The letter stated as follows:

68 While the Cheating Condition could theoretically be subsumed under the phrase “for a variety of reasons”, given its importance to Jake’s case, it is inconceivable that it would not be expressly mentioned in the letter of demand if it truly existed. Its absence is further evidence that it never existed.

69 Jake’s case was also that he only agreed to Millie beneficially owning her 99% registered interest in the Property if they purchased a second property together, ie, the 2nd Property Condition. This was also a critical aspect of Jake’s case as it meant that did not intend to benefit her with his financial contributions immediately (see [6(b)] above). The burden was therefore on Jake to prove the 2nd Property Condition. We reject Jake’s case on the 2nd Property Condition because: (a) Jake’s explanation of the condition was inconsistent and unreliable; and (b) the existence of the condition is not borne out by the evidence. We address each reason in turn.

70 Jake’s original account of the 2nd Property Condition was that if the parties purchased a second property, he would transfer his 1% registered interest to Millie such that she would fully own the Property, and he would then purchase the second property in his sole name without incurring ABSD. This was expressly stated in Jake’s letter of demand to Millie (quoted at [67] above) and in Jake’s OA 695 First Affidavit and OA 695 Second Affidavit. Importantly, there is nothing in this arrangement which suggests that the parties did not intend for Millie’s 99% registered interest to also reflect her beneficial ownership of the Property, or that such beneficial ownership would be conditional on a second property being purchased in Jake’s sole name. In fact, if the intention was for Millie to fully own the Property after Jake transfers his 1% registered interest to her, it was necessarily the case that she legally and beneficially owned her 99% interest from the very beginning.

71 Jake’s explanation of the 2nd Property Condition subsequently morphed, and his evidence became more obscure as to what exactly he would transfer to Millie if the parties were to purchase a second property. In his SOC, Jake stated that he would “transfer his interest in the Property to [Millie]” [emphasis added]. Similarly, in his AEIC, Jake stated that “I would transfer my ownership in the [Property] to [Millie]” [emphasis added].

72 In cross-examination, Jake’s testimony on the 2nd Property Condition followed a similar evolution. He was initially very specific, stating that he would “transfer my 1 per cent to [Millie’s] name, so that I can actually free up my side to actually buy the second property. And the second property would be in my name.” After giving this explanation, Jake affirmed that the 99:1 arrangement was genuine and not a sham. As mentioned at [40] above, the only coherent interpretation of these statements is that the parties genuinely intended for the 99:1 ratio to reflect their legal and beneficial ownership of the Property.

73 However, when questioned why he used the more ambiguous phrase “transfer my ownership” in his AEIC, Jake said his intention was to transfer his “entire beneficial interest” to Millie. But this explanation was obviously contrived given that Jake had no conception of beneficial interests, and did not know or understand the distinction between legal and beneficial interests, at the material time (see [37] above).

74 The way Jake’s case on the 2nd Property Condition transformed severely undermined its credibility. Jake’s original explanation that the parties would decouple their interest in the Property by him transferring his 1% registered interest to Millie coheres with Millie’s case, which she maintained consistently, that she held her 99% interest absolutely and this would enable the parties to save on ABSD (see [8(e)] above). Jake’s subsequent use of ambiguous terms such as “interest” and “ownership” appears to be an afterthought.

75 It is also noteworthy that in all the text messages discussed at [59]–[65] above, there was never any mention of the 2nd Property Condition as asserted by Jake. Even in the 19 Jul 2018 Messages, where the parties expressly discussed owning two properties, and Jake confirmed Millie’s understanding that the two properties would be “u one I one”, there was never any suggestion that Millie’s ownership of one of the properties would be in name only unless and until a second property was purchased and registered in Jake’s name. The lack of any reference to the 2nd Property Condition in the text messages suggests that it never existed.

76 We now turn to analyse the parties’ text messages in greater detail. These text messages were central to the Judge’s finding that a resulting trust arose over the Property. The Judge found that the text messages revealed a shared understanding that the Property was only nominally registered in Millie’s name to provide her with security, and that Jake never intended to benefit Millie with his financial contributions unless he cheated on her (Judgment at [21(a)]). We disagree with the Judge’s analysis of the text messages.

77 The Judge began by examining the 3 Dec 2018 Messages (set out at [61] above). He observed that the parties spoke of putting the property they purchased together “under [Millie’s] name”. He reasoned that one would not speak of placing property “under [a recipient’s] name” if the intention was to make an unconditional gift (Judgment at [23]–[26]). He then observed that the reference to placing assets “under [Millie’s] name” continued to play out in a separate conversation on 24 January 2019, and that these references to nominal registration formed the backdrop to the purchase of the Property (Judgment at [27]–[28]).

78 With respect, the Judge erred in ascribing such a meaning to the parties’ references to putting property “under [Millie’s] name”. First, there is no inconsistency between putting an asset in someone’s name and giving them the same. In fact, if the Judge was right that the parties had deliberately chosen their words to reflect that Millie only held nominal ownership of the Property, this suggests that the parties knew and understood the distinction between legal and beneficial interests, even if they did not use these specific labels. But the parties’ positions were that they did not (see [37] above).

79 Second, and importantly, the Judge’s interpretation is contradicted by other text messages between the parties:

(a) In the 19 Jul 2018 Messages (set out at [59] above), Millie said “[i]f the hse isn’t mine I won’t bother / So put under my name at least I will bother”. This message indicates that, from Millie’s perspective, putting property under her name was equivalent to the property being hers. There is nothing to suggest that Millie intended to mean nominal ownership when she said, “put under my name”.

(b) In the 20–21 Jul 2020 Messages (set out at [64] above), when Millie asked Jake if “[his] promise about putting all [his] money with [her] [will] still be kept”, Jake retorted by asking “i still havent gave you enough ah”. This suggests that, in Jake’s mind, he drew no distinction between putting money in Millie’s name and giving it to her. Later, Jake apparently transferred Millie $40,000, and on the next day, he described this as something which he had given her. Indeed, he went on to say, “[i]t was obvious you want the money in your own name / So I gave”, showing once again that he drew no distinction between putting money in Millie’s name and giving it to her. When Jake later sought to draw a comparison between the money he transferred to Millie and the registration of the Property in a 99:1 ratio, he did not attempt to distinguish the former as something which he had given her and the latter as something that was put in her name. The only distinction he drew was that the registration of the Property in a 99:1 ratio was done “willingly” whereas he felt pressured into giving Millie the money. Indeed, the fact that Jake drew this comparison shows that, from his perspective, the Property, much like the money, was something which he had given Millie and which Millie owned beneficially.

80 Therefore, the parties’ references to placing a property “under [Millie’s] name” is consistent with Millie’s case that she legally and beneficially owned 99% of the Property.

81 We next analyse other text messages which the Judge referred to in his analysis. We acknowledge that the Judge’s interpretation of some of these messages was plausible. However, the Judge erred in failing to recognise that all these messages could be reasonably interpreted in a way which supports Millie’s case.

82 The Judge examined the 20–21 Jul 2020 Messages (set out at [64] above).

(a) The Judge observed that Millie said that she wanted Jake to “think 3-4 times” before cheating on her, that she wanted something to “h[o]ld [him] back enough to reconsider n try many times till cannot”, that she “want the [heart] / [n]ot the [money]” and that “[she] only want assurance”. The Judge interpreted these messages to mean that Millie only wanted the Property and Jake’s money as “assurance” and a “security mechanism”, and not to benefit from them as gifts (Judgment at [33]–[34], [36] and [37(b)]). However, as we explained at [54]–[56] above, Millie would be afforded no meaningful assurance if her 99% registered interest was purely nominal, and the true deterrent against Jake cheating came in the form of Millie beneficially owning her 99% registered interest. Therefore, these messages support Millie’s case and not Jake’s.

(b) The Judge then observed that Jake said “[w]as I afraid that you will run away with the house? Who wouldn’t.” The Judge reasoned that such a fear was fundamentally incompatible with an intention to make an unconditional and immediate gift to Millie of his financial contributions to the Property (Judgment at [35]). However, if Millie held part of her 99% registered interest in the Property on trust for Jake, then she could not truly “run away” with it because Jake could always make a claim for that share, as he did in instituting the proceedings below. On the contrary, the risk that Millie would “run away” with the Property would only be real if she beneficially owned her registered interest in the Property and ended her relationship with Jake.

(c) The Judge also observed that neither party characterised Jake’s financial contributions to the Property as an outright and unconditional gift to Millie (Judgment at [37(b)]). The fact that Millie did not explicitly describe the Property (or Jake’s financial contributions thereto) as a “gift” to her in the text messages is neither here nor there. In the same vein, there is no explicit statement by Jake in these messages that he regarded himself as having an interest in the Property greater than his 1% registered interest.

83 There was one other text message which the Judge did not refer to in his analysis which further supports Millie’s case. In the 3 Dec 2018 Messages (set out at [61] above), Jake expressed his concern that “if I were to use my own money but yet have it under your name, if anything were to happen between us. All these money go down the drain”. Here, Jake explicitly acknowledged that he risked losing all his financial contributions to the Property should the parties break up. This is consistent with the understanding that Millie beneficially owned 99% of the Property.

84 The Judge also relied on other evidence to support his finding that Jake did not intend to immediately benefit Millie with his financial contributions to the Property. These were: (a) Jake’s involvement in selecting the property to purchase; (b) the parties’ discussions about the Property as an investment; (c) Jake’s continued financial contributions to the Property after breaking up; and (d) the parties’ discussions on how to split the Property after breaking up. But all of these were neutral.

85 First, the Judge found that Jake was significantly involved in selecting the property to purchase – attending viewings with Millie, and discussing with Millie which properties they liked and how much they were willing to pay. According to the Judge, this undermined the suggestion that the Property was intended as a gift from Jake to Millie (Judgment at [41]). We do not see how such an inference can be drawn. It must be remembered that, at the time, the parties were romantically involved and were making a joint investment on the basis that they would remain in a long-term relationship. It is not surprising that they would both be involved in all aspects of the purchase regardless of the ownership structure. The fact that Jake was actively involved in the process is equivocal.

86 Second, the Judge analysed the parties’ post-purchase conversations where they referred to the Property as an investment. For example, when the parties found out that the seller of the Property made a substantial profit from the sale, Millie expressed her disappointment and Jake comforted her by saying that they still bought the Property below its valuation price and that they too would make a profit from it. The parties also discussed how they intended to “milk” the Property for money. The Judge reasoned that an expectation of future profit was fundamentally incompatible with an intention to gift because if Jake truly intended to benefit Millie with his financial contributions, he would have no prospect of profit to speak of (Judgment at [42]–[43]). Again, this fails to acknowledge the fact that the Property was purchased as part of the parties’ plans for their future, and any return or income on the investment would have been enjoyed jointly as a matter of fact even if not as a matter of legal right, which again is consistent with the fact that they were both ignorant of the nuances of equity. There is nothing in these conversations to suggest that Jake intended to make a return on the investment in the Property individually.

87 Third, the Judge noted that Jake continued to service the mortgage payments even after breaking up with Millie. The Judge reasoned that if Jake’s financial contributions were intended as unconditional gifts to Millie, then his financial commitment to the Property would logically have ceased with the end of their relationship. The fact that he continued to make such financial contributions showed that there was never any intention to unconditionally benefit Millie with them (Judgment at [44]). We find Jake’s continued mortgage payments to be neutral. Jake and Millie were jointly and severally liable under the mortgage loan. Jake had every incentive to ensure that the mortgage payments were promptly serviced regardless of the parties’ precise ownership structure of the Property.

88 Fourth, the Judge noted that when the parties discussed selling the Property, Jake consistently proposed using the sale proceeds to reimburse their initial financial contributions and then splitting the remaining proceeds equally between them. The Judge reasoned that this expectation of Jake to recoup his initial investment contradicted any suggestion that his financial contributions were intended as gifts to Millie (Judgment at [45]–[46]). While Jake described his proposal as “what I feel should be fair for both parties”, it must be kept in mind that this was said after the parties had broken up. It was clearly in Jake’s interest to recoup as much of his financial contributions as possible given the failure of the relationship, and these messages have no bearing on the parties’ intentions when the Property was purchased. Like the earlier points, we find Jake’s proposal to be equivocal at best and not probative of Jake’s intention not to benefit Millie with his financial contributions.

89 The burden was on Jake to prove that he did not intend to benefit Millie with his financial contributions to the Property. The evidence, properly understood and seen in context, establishes that:

(a) The parties had extensive discussions about the Property being in Millie’s name because she was insecure about their relationship and Jake was willing to demonstrate, at his own risk and expense, that Millie’s fears were unfounded.

(b) In that regard, Jake was repeatedly making the point that he was trusting Millie by putting the Property and other assets in her name and at her disposal.

(c) Jake did not at any time claim that he had any interest in the Property beyond his registered interest.

(d) On the contrary, the evidence establishes that the parties agreed that Millie would own 99% of the Property. This was consistent with Jake’s intention as he expected them to purchase a second property which he would own; and they would save on ABSD by him then transferring his 1% registered interest to Millie before purchasing the second property.

90 In the circumstances, the clear intention of the parties was that they would legally and beneficially own the Property in accordance with their registered interests. Jake has failed to discharge his burden of proof that he did not intend to benefit Millie with his financial contributions. There is no room to apply the presumption of resulting trust as the evidence clearly establishes the intention of the parties.

91 On this ground alone, Jake’s claim for a resulting trust fails and Millie’s appeal must be allowed.

92 As we have found that that a resulting trust did not arise on the facts, it is unnecessary for us to consider whether the Judge erred in expanding the concept of “conditional resulting trusts”, and we express no view on this.

93 It is also strictly unnecessary for us to consider whether the resulting trust found by the Judge is unenforceable for illegality. Nonetheless, as this issue was dealt with extensively by the Judge and the parties (both in the proceedings below and in this appeal), we express some tentative observations. We stress that our comments below operate on the assumption that the Judge was right to find that a resulting trust arose over the Property in favour of Jake.

94 In this appeal, neither party takes issue with the Judge applying the Alistair Lau framework, which was adapted from the framework established in Ochroid Trading Ltd v Chua Siok Lui [2018] 1 SLR 363, to assess the impact of illegality on a trust claim. The following passage in Alistair Lau helpfully summarises the framework (Alistair Lau at [81]):

95 Millie is not claiming that the resulting trust is illegal in itself, but that it arose as an incidental consequence of an illegal purpose. Neither does Jake claim to have an alternative basis for enforcing a proprietary interest in the Property. Therefore, the key questions in the present case are: (a) whether the resulting trust arose as an incidental consequence of an illegal purpose; and (b) if so, whether denying Jake’s claim would be a disproportionate response to the illegality? We address them in turn.

96 Before we address the Judge’s analysis, we first explain how owning a first property in a 99:1 ratio impacts on the co-owners purchasing a second property in relation to ABSD, and how that potentially engages issues of tax avoidance, under-stamping and tax evasion under the Stamp Duties Act 1929 (2020 Rev Ed) (“SDA”).

97 In Singapore, a buyer must pay a stamp duty on his purchase of property, known as Buyer’s Stamp Duty. If the buyer is a Singapore citizen who already owns one or more residential properties, and seeks to purchase another residential property, he will have to pay an additional stamp duty on that purchase, known as ABSD. The current ABSD rates are set out in Art 3(bi) of the First Schedule of the SDA and range from 20% to 30% of the purchase price or the market value of the residential property. The purpose of the ABSD regime is to moderate demand in residential properties and promote a stable and sustainable property market. This is achieved by increasing the financial cost of owning two or more residential properties in Singapore.

98 The 99:1 arrangement is a method by which co-owners seek to purchase two residential properties without incurring ABSD. It operates in three stages. First, the co-owners purchase the first property in a 99:1 ratio. This allows them to obtain housing loans in both their names and to access both their CPF moneys to finance the purchase. Second, the co-owners decouple their interests in the first property. The 1% owner (“X”) transfers his 1% share to the 99% owner (“Y”), and the parties pay stamp duty on that 1% transfer. Third, X, who no longer retains an interest in the first property, is now able to purchase the second property in his name without incurring ABSD. These three stages minimise the total amount of stamp duty the parties must pay on the purchase of both properties. It should be noted that this method does not require the use of a trust structure in relation to either property purchased.

99 Section 33A of the SDA, which addresses the avoidance of stamp duty, states as follows:

100 Does the 99:1 arrangement constitute the avoidance of ABSD within the meaning of s 33A of the SDA? The Judge appeared to answer this question in the negative. He asserted that “[t]here is nothing inherently wrong about the decoupling arrangement – provided both parties genuinely intended to hold their first property in a 99:1 ratio, along with all attendant legal and beneficial rights that come with such a holding” (Judgment at [76]). But that does not address whether the 99:1 arrangement is captured by the wording of s 33A of the SDA, in particular, whether the “purpose or effect” of the 99:1 arrangement is to “relieve [X] from any liability to pay [ABSD]”. The Judge then stated that “there is nothing inherently objectionable about registering a property in a 99:1 ratio” and observed that “no tax liability was actually avoided at this stage” (Judgment at [88]–[89]). This misses the point. We have explained that the 99:1 arrangement involves three stages. The fact that the first stage does not amount to tax avoidance does not address whether the three stages collectively amount to tax avoidance. As this court observed in Comptroller of Income Tax v AQQ [2014] 2 SLR 847 at [43], a tax avoidance arrangement may constitute a combination of steps that may be individually unobjectionable.

101 We need not provide a definitive view of whether the 99:1 arrangement of the sort outlined at [98] above, where there is no question of a trust structure, amounts to the avoidance of ABSD. But, where the 99:1 arrangement involves a trust over the first property, which the Judge found to exist in this case, it is under-stamping and tax evasion which are potentially relevant.

102 In relation to under-stamping, s 4 the SDA provides as follows:

103 Article 3 of the First Schedule of the SDA prescribes the applicable stamp duty rates payable on various instruments. Where those duties are unpaid or underpaid in contravention of s 4(2) of the SDA, the instrument is said to be under-stamped. It should be noted that under-stamping is only concerned with whether the correct amount of stamp duty is paid. A person may commit under-stamping even if he did not intend to do so or did not know that an incorrect amount of stamp duty was paid. It is in this sense that the Judge described liability for under-stamping as “strict” (Judgment at [94]).

104 In relation to tax evasion, s 62 of the SDA states as follows:

105 Tax evasion within the meaning of s 62(a) of the SDA involves the non-disclosure of information with intent to evade the payment of stamp duty. This entails the deliberate and fraudulent intention to conceal information from the relevant authorities.

106 In what ways do under-stamping and/or tax evasion manifest when a 99:1 arrangement involves a trust over the first property? There are two different scenarios where the illegality may arise.

(a) First Scenario: The co-owners purchase the first property in a 99:1 ratio whilst holding their beneficial interests in a different ratio under a trust. When the co-owners decouple their interests in the first property, X transfers his entire beneficial interest to Y but represents to the authorities that only his 1% interest is being transferred, so that stamp duty is levied on only that 1% interest. In this scenario, it must be the case that X knew of his larger beneficial interest, otherwise he could not purport to transfer it to Y. Therefore, this is under-stamping (in contravention of s 4(2) of the SDA) and tax evasion (punishable under s 62(a) of the SDA) at the decoupling stage. X then proceeds to purchase a second property. Since he has disposed of his entire beneficial interest in the first property, he is legally entitled to purchase the second property without incurring ABSD. There is no under-stamping or tax evasion at this stage.

(b) Second Scenario: The co-owners purchase the first property in a 99:1 ratio whilst holding their beneficial interests in a different ratio under a trust. When the co-owners decouple their interests in the first property, X transfers only a 1% share of the Property to Y and pays stamp duty on the transfer of that 1% share. At this decoupling stage, the correct amount of stamp duty is paid, so there is no under-stamping or tax evasion. However, X still retains a beneficial interest in the first property (less 1% of his original beneficial interest which he has transferred to Y). X then proceeds to purchase a second property but falsely represents to the authorities that he does not have any interest in the first property, and so does not incur ABSD. If X was ignorant of his beneficial interest in the first property, then he may only have committed under-stamping in contravention of s 4(2) of the SDA (ie, ABSD which should have been paid was not paid). But if X knew of his beneficial interest in the first property and fraudulently concealed it from the authorities, he has committed tax evasion punishable under s 62(a) of the SDA (ie, evading ABSD with intent).

107 Having clarified how the illegality may materialise, we now examine the way the Judge understood how the 99:1 arrangement relates to under-stamping and tax evasion. In the Judgment at [76] and [90], the Judge said as follows:

108 According to the Judge, the illegality takes place at the decoupling stage when Jake transfers 1% of the Property to Millie without disclosing his remaining beneficial interest in the Property. This is incorrect. If Jake transfers 1% of the Property to Millie at the decoupling stage, there is no avenue for him to declare his remaining beneficial interest in the Property, and there is no need to pay stamp duty on that remaining beneficial interest. There is no under-stamping or tax evasion. As explained in the second scenario above, the illegality is engaged at the purchase of the second property stage, where Jake falsely represents to the authorities that he does not have any interest in the Property so that he can purchase the second property without incurring ABSD. As we explain below, based on the Judge’s evaluation of the evidence, the correct conclusion ought to be that Jake knew of his beneficial interest in the Property but intended to fraudulently conceal it at this stage. Therefore, the contemplated illegality would in fact have been tax evasion punishable under s 62(a) of the SDA (ie, evading ABSD with intent).

109 The Judge found that at the time the Property was purchased, the parties had intended to carry out under-stamping but not tax evasion. This was largely premised on the Judge’s finding that Jake did not know or understand the concept of beneficial interest at the time, and was therefore ignorant of his beneficial interest in the Property and could not have formed a fraudulent intention to hide it (Judgment at [92]–[93] and [97]).

110 In our view, this premise is flawed. If, as the Judge found, Jake did not intend to benefit Millie with his financial contributions or for her to own 99% of the Property, that same evidence necessarily indicates that Jake intended to have an interest in the Property larger than his 1% registered interest and knew that he had such a larger interest, regardless of whether he was aware of the specific labels of “legal” and “beneficial” interests. For example, if the Judge was right that the parties spoke of putting property “under [Millie’s] name” merely in the sense of nominal ownership and not as an outright gift (see [77] above), then this must mean that Jake appreciated an ownership interest different from the registered interest, and knew that his interest in the Property extended beyond his 1% registered interest. Similarly, if the Judge was right that Jake’s expectation to recoup his financial contributions after selling the Property showed that he did not intend for them to be gifts to Millie (see [88] above), then the expectation that he would be entitled to more than 1% of the sale proceeds also indicates that Jake knew that his interest in the Property extended beyond his 1% registered interest.

111 The upshot of this is that, assuming the Judge’s analysis of the evidence was correct, then the only logical corollary is that Jake must have known that his beneficial interest in the Property extended beyond his 1% registered interest, and that he would continue to retain an interest in the Property after he transfers his 1% registered interest to Millie.

112 It bears emphasis that the plan, as initially stated by Jake in his letter of demand, his two affidavits in OA 695 and whilst on the stand, and as corroborated by Millie, was always for Jake to transfer 1% of the Property to Millie and then purchase a second property without paying ABSD (see [70], [72] and [74] above). If Jake knew of his larger beneficial interest in the Property but intended not to declare it when purchasing the second property so that he would not incur ABSD, that would reflect an intention to evade ABSD within the meaning of s 62(a) of the SDA.

113 In these circumstances, the Judge would have been wrong to find that the illegality contemplated by Jake was merely under-stamping and not tax evasion.

114 For completeness, we are aware that Jake had asserted in cross-examination that what he intended to transfer to Millie was his “entire beneficial interest” (see [73] above). If true, then by the time Jake purchased the second property, he would not have any beneficial interest in the Property to declare, and he would not have committed tax evasion at that stage. But as explained at [74] above, we find this belated testimony unpersuasive and self-serving. Prior to this testimony, Jake had consistently stated that the intention was simply to transfer his 1% registered interest to Millie, and this was also Millie’s position which she maintained consistently.

115 Therefore, while we agree with the Judge that the resulting trust he found would have arisen as an incidental consequence of an illegal purpose, we disagree that the contemplated illegality was merely under-stamping. The evidence, as evaluated by the Judge, points to a paradigm case of tax evasion.

116 Given that the resulting trust over the Property arose as an incidental consequence of an illegal purpose, the next question is whether it would be disproportionate to deny Jake’s claim such that Millie would legally and beneficially own 99% of the Property despite contributing only 44.78% of the purchase price.

117 Where a contract is entered into with the object of committing an illegal act, the general approach is to examine the relevant policy considerations underlying the illegality principle so as to produce a proportionate response to the illegality in each case. The non-exhaustive factors for assessing proportionality include: (a) whether allowing the claim would undermine the purpose of the prohibiting rule; (b) the nature and gravity of the illegality; (c) the remoteness or centrality of the illegality to the contract; (d) the object, intent, and conduct of the parties; and (e) the consequences of denying the claim. The court should weigh and consider these factors in the context of the particular facts of each case (Ting Siew May v Boon Lay Choo [2014] 3 SLR 609 (“Ting Siew May at [66], [70] and [71]). These factors were expressly adapted into the trusts context and incorporated into the Alistair Lau framework as reproduced at [94] above.

118 The Judge’s conclusion that denying Jake’s claim was a disproportionate response to the illegality was largely premised on his anterior finding that the contemplated illegality was merely under-stamping. However, given our view that the illegal purpose would have been tax evasion, denying the resulting trust claim would not have been a disproportionate response. We elaborate below.

119 The first aspect of the Judge’s analysis was the fact that under-stamping is not a criminal offence and involved strict liability, and his finding that the parties had no nefarious intentions because they did not even know of Jake’s larger beneficial interest in the Property (Judgment at [100]–[101]). As discussed above, the only logical inference from the Judge’s evaluation of the evidence is that Jake knew of his larger beneficial interest in the Property but intended not to declare it when purchasing a second property so that he would not incur ABSD. In our view, the nature and gravity of the contemplated illegality was serious. Tax evasion involves deliberate and fraudulent concealment of information and is a serious criminal offence. Indeed, even the Judge noted that his views might differ in a case where the parties have knowledge of their beneficial interests and deliberately structured transactions to conceal them (Judgment at [111]).

120 The second aspect of the Judge’s analysis was his finding that the primary objective of the 99:1 ratio was to quell Millie’s insecurities, and the contemplation of purchasing a second property was a secondary consideration (Judgment at [102]). With respect, it is unclear to us how the Judge came to this finding. As noted at [59] above, there are text messages as early as 19 July 2018, close to one-and-a-half years before Purchase Day, where the parties expressly discussed their plans to purchase two properties. In Jake’s letter of demand to Millie (set out at [67] above), the parties’ plan to purchase a second property was explicitly mentioned, but the Cheating Condition was not. More importantly, as a matter of logic, Millie’s insecurity could have been quelled by any ratio which gave her a sizeable or majority share of the Property. On Jake’s case, the specific choice of the 99:1 ratio is only explicable because of the plan to purchase a second property, since this minimises the amount of stamp duty payable when Jake transfers his 1% registered interest to Millie.

121 In any event, the primary objective behind the selection of the 99:1 ratio is beside the point. Jake’s intention behind the selection of the 99:1 ratio and his intention not to benefit Millie with his financial contributions (which gives rise to the resulting trust) are two distinct intentions. Here, we are concerned with the remoteness or centrality of the illegal purpose of tax evasion to the latter intention. Regardless of whether it was Millie’s insecurities or the purchase of a second property which was the primary reason for the selection of the 99:1 ratio, that does not speak to whether the illegal purpose of tax evasion was central to Jake’s intention not to benefit Millie with his financial contributions. In our view, the Judge wrongly conflated these two intentions.

122 This distinction is illustrated by the decision of Cheong Kok Leong v Cheong Woon Weng [2017] SGCA 47 (“Cheong Kok Leong”), which Jake cites in support of his case, but which does not assist him. In that case, there was an oral agreement for the appellant to hold a private property on trust for himself and the respondent in equal shares. The appellant argued that the oral agreement was void for illegality because the respondent owned a Housing and Development Board (“HDB”) property when he entered into the oral agreement, and thus the oral agreement was made to circumvent the prohibition against the concurrent ownership of private property and HDB property. In rejecting this argument, this court observed that the illegality in concurrent ownership under the Housing and Development Act (Cap 129, 2004 Rev Ed) did not attach to the ownership of the private property but to the acquisition or retention of the interest in the HDB property. Even if it was illegal for the respondent to retain his HDB property when he acquired his interest in the private property, there was no illegality with respect to the respondent’s ownership of the private property (at [19]–[20]). In addition, it was the parties’ common intention for the respondent to have a half-share of the private property which gave rise to the trust, and this common intention was distinct from the respondent’s purpose in registering the appellant as the sole owner of the private property. Thus, the trust arising from the common intention was untainted by the illegality (at [21]). Cheong Kok Leong was therefore a case where the illegality in respect of the respondent’s ownership of the HDB property was unrelated to the trust over the private property. The same cannot be said of the present case. The resulting trust over the Property (as asserted by Jake and found by the Judge) is the precise legal arrangement which, if undisclosed to the authorities, would have allowed Jake to retain a beneficial interest in the Property whilst purchasing a second property without incurring ABSD, thereby amounting to tax evasion.

123 The third aspect of the Judge’s analysis was that the under-stamping was only contemplated and never executed (Judgment at [103]). We accept this is equally true even when the illegality is properly characterised as tax evasion. The Judge relied on two cases – the decisions of the UK House of Lords in Tinsley v Milligan [1994] 1 AC 340 and the High Court of Australia in Nelson v Nelson [1995] 132 ALR 133 – which allowed the claims in resulting trusts despite completed illegal acts, and found that the case for allowing Jake’s resulting trust claim more compelling given that the illegal purpose was only contemplated (Judgment at [104]–[106]). However, the issue of whether the illegality was contemplated or completed was not squarely before the courts in those cases, and the courts did not make that distinction. The inference which the Judge sought to draw from the outcomes of those cases is unpersuasive because it was not rooted in the reasoning of those cases.

124 In any event, we disagree with the Judge’s view that there is a stronger case for not denying a claim when the illegal purpose is merely contemplated but not completed. In Ting Siew May at [98], this court observed that “[o]nce an illegal object of the contract has been established, that object taints the contract itself and it is no answer to say that the illegal object has not been carried out” [emphasis in original omitted]. This court also specifically highlighted at [94] that “[t]here is no authority for the proposition that a contract entered into with the intention of committing an illegal act will no longer be considered illegal and unenforceable if the party with that original intention subsequently decides not to carry out that intention.” Applying this to the present context, it is irrelevant that no tax was evaded and the parties will no longer pursue this illegal purpose as they have broken up; Jake’s purpose of tax evasion tainted the resulting trust when it first crystallised.

125 The Judge however reasoned that the present case fell within the “exception” recognised in Ting Siew May at [96] where “any abandonment of an original unlawful intention can be taken into account only if there was an ignorance of the unlawfulness of the intention in the first place”. This was because of his finding that the parties here did not set out to deliberately engage in an illegal act and did not even know that their actions would constitute unlawful under-stamping (Judgment at [107]–[109]).

126 With respect, the Judge may not have fully appreciated this court’s comment in Ting Siew May at [96]. In that case, the respondents argued that they had abandoned their original unlawful intention and referred to the English case of Waugh v Morris (1873) LR 8 QB 202 (“Waugh”) as authority. In Waugh, the English court held that to avoid a contract which could be legally performed, on the ground that there was an intention to perform it in an illegal manner, it was necessary to show that there was the wicked intention to break the law, and in this respect knowledge of the law was of great importance. On the facts, the English court found that the contract was not void because it was not made knowingly with the intention to violate the law. However, this court in Ting Siew May found that Waugh did not assist the respondents’ case because it at most stood for the proposition that “any abandonment of an original unlawful intention can be taken into account only if there was an ignorance of the unlawfulness of the intention in the first place”. This however did not apply to the respondents who knew that their original intention was unlawful and therefore had the wicked intention to break the law (at [94]–[96]). While refuting the respondents’ reliance on Waugh, this court did not endorse any “exception” as understood by the Judge.

127 Similar to the Judge’s line of reasoning, Jake cites Kliers v Schmerler [2018] EWHC 1350 (Ch) (“Kliers”) as an example of a case where the court allowed a trust claim despite the illegality having been completed. To the extent Jake relies on this case as authority for the argument that the case for allowing a trust claim must be even stronger when the illegality is merely contemplated, that simply does not follow for the reasons already provided. In any event, the facts of Kliers are so vastly different that it is not remotely useful as a comparator to the present case. In that case, the claimant and her husband purchased a matrimonial property where she contributed most of the purchase price. However, the property was registered in the name of the claimant’s brother. This was so that an interest-only mortgage could be obtained by falsely declaring that the property would be for the brother’s sole occupation, and so that the claimant could obtain housing benefits by entering a shorthold tenancy of the property. While the court allowed the claimant’s trust claim, this was largely influenced by the court’s finding that the claimant was unduly influenced into registering the property in her brother’s name by, among others, her father and religious leaders in her community (at [91]–[93]). The court therefore was of the view that this was a “strong case where our principles of illegality do not and should not bar the remedies” sought by the claimant (at [90]). The decision in Kliers therefore turned on the unique facts of the case. In contrast, there are no special facts or considerations in this case which warrant looking at Jake’s position through a more generous lens.

128 In our view, the illegal purpose of evading ABSD tainted the resulting trust the moment it crystallised. That was, on Jake’s own case, his intention from the outset, and it did not come to pass because the parties ended their relationship. Recognising a resulting trust would effectively be endorsing that intention, and it is no answer to say that no ABSD was in fact evaded or that the parties will no longer pursue that intention.

129 The fourth aspect of the Judge’s analysis was that there exists a statutory mechanism under s 46 of the SDA to correct under-stamped instruments and impose penalties, and therefore allowing the resulting trust claim would not undermine the prohibition against under-stamping (Judgment at [110]). This statutory mechanism applies to evaded tax as well. However, in our view, the mere existence of a statutory mechanism to retrospectively correct an illegality does not detract from the fact that it is impermissible and unacceptable for parties to deliberately structure their transactions in a way which evades their tax obligations.

130 As observed at [97] above, the ABSD regime plays a crucial role in promoting a stable and sustainable property market. Further, the Government has recognised that trust structures may be exploited to circumvent ABSD rules, which it is why it introduced a new category of ABSD for residential properties transferred into a living trust, commonly referred to as ABSD (Trust) (see Ministry of Finance, “Additional Buyer’s Stamp Duty (ABSD) For Residential Properties Transferred Into A Living Trust”, press release (8 May 2022)). Jake’s contemplated tax evasion is another means by which trust structures can be used to circumvent the ABSD rules. Even though there exists a statutory mechanism to correct the illegality had ABSD been evaded, allowing Jake’s resulting trust claim would amount to condoning an illegal purpose which, in the present case, would undermine the ABSD regime.

131 The last aspect of the Judge’s analysis was his view that to strip Jake, who contributed most of the purchase price, of a substantial beneficial interest in the Property would be an excessive penalty that outweighed the gravity of the intended illegality (Judgment at [113]). We accept that there are significant financial consequences to Jake if we find that the resulting trust is unenforceable for illegality. However, given our view that the contemplated illegality would have been tax evasion and not under-stamping, the gravity of the intended illegality was far more serious than what the Judge had in mind. In our view, where the illegal purpose in question involves dishonesty, it would be extremely rare that the financial consequences of denying a claim would outweigh the gravity of the illegality.

132 In sum, if the Judge was right to find that a resulting trust arose over the Property, then the only logical inference is that Jake knew of his larger beneficial interest in the Property and deliberately intended to conceal it when purchasing a second property so that he would not incur ABSD. It follows that the illegality contemplated by Jake at the time the Property was acquired would have been tax evasion and not simply under-stamping. Given the serious nature and gravity of tax evasion and the importance of maintaining compliance with the ABSD regime, we are of the view that it would not have been a disproportionate response to the illegality to deny Jake’s claim. Therefore, even if we had not found that the parties’ common intention was for the 99:1 ratio to reflect their legal and beneficial ownership, the issue of illegality would have been an independent basis for denying Jake’s claim.

133 For the foregoing reasons, we allow the appeal. Both Millie and Jake submit that costs should be fixed at $50,000 in favour of Millie if the appeal is allowed, although they appear to diverge on whether this sum ought to be inclusive of disbursements. Having regard to Appendix G of the Supreme Court Practice Directions 2021 which stipulate costs in the range of $30,000 to $150,000 for appeals before the Court of Appeal against a judgment obtained following a trial, we fix costs at $50,000 (all-in) payable by Jake to Millie.