Abstract
In Hong Kong, a transaction for the sale and purchase of land imposes a duty on the vendor to supply original title deeds upon completion. If this duty is breached, the purchaser would be entitled to refuse completion. Yet, much confusion has arisen over the precise requirements of this duty, given that it is thought to originate from two distinct sources: the common law, and s.13A of the Conveyancing and Property Ordinance (Cap.219). It is suggested that the current common law position is based on a misunderstanding of s.13A, and that the current statutory position under s.13A is not aligned with the normative rationale of the duty. As such, it is necessary to amend s.13A. When considering the substance of the amendment, the Hong Kong Legislative Council (“LegCo”) ought to ensure that the scope of the duty is aligned with its normative rationale, that the law is sufficiently certain, and that vendors who have relied on the current position would not be retrospectively prejudiced.
Introduction
In every sale and purchase of real property, vendors and purchasers have their respective rights and duties, which must be strictly observed. Yet, these rights and duties must be delineated in a clear and certain manner, as well as strike an appropriate balance between the interests of vendors and purchasers – this concern is especially relevant in the Hong Kong context, where property prices are the highest in the world. As Hong Kong law currently stands, the duty to supply original title deeds at the date of completion (on which this article will focus) has been the subject of much legal uncertainty, and strikes the incorrect normative balance between the interests of vendors and the interests of purchasers.
In short, this duty denotes an obligation on the vendor to supply certain types of title deeds over a certain period of time, so as to remove any real risk that the property in question is encumbered by an adverse interest that would bind the purchaser. The breach of this duty entitles the purchaser to refuse completion.
This duty is underpinned by the fundamental concern to ensure that the purchaser obtains clean title to the property at the date of completion. However, setting too high a standard for this duty would allow purchasers to refuse to complete, even if the ‘missing documents’ pose no real risk of the purchaser's title being affected by an adverse interest. Opportunistic purchasers may as such find themselves being able to refuse completion and escape from a bad bargain even where there is no real risk that their title would be affected. Conversely, setting too low a standard for this duty would lead to the purchaser's title being insufficiently protected, as he would not be entitled to refuse completion even if the missing documents do give rise to a real risk of an adverse interest affecting the property.
Therefore, it is imperative for the content of the duty to strike an appropriate balance between the interests of vendors and those of purchasers. In the final analysis, it is argued that Hong Kong law fails to strike this balance. There are areas in which the vendor is insufficiently protected, as well as areas in which the purchaser is insufficiently protected: as such, statutory reform is necessary.
With this backdrop in mind, this article will be divided into two substantive sections. Section 1 will examine what the current legal position is, and Section 2 will discuss how the law should be reformed.
1.The positive question: what is the law?
The law relating to the duty to supply original title deeds appears simple: the default position is governed by one provision, namely s.13A of the Hong Kong Conveyancing and Property Ordinance (Cap 219, Laws of Hong Kong;
A. Proving title, giving title and supplying original title deeds
In order to contextualise the duty to supply original title deeds, it is important to elucidate the distinction between the two duties of ‘giving’ title, and ‘proving’ title. These involve separate obligations arising from different points in time. The vendor is required to prove title at (a reasonable time before) completion, and this can be discharged by providing certified copies of the relevant documents under s.13 CPO. 1 Whereas for the duty to give title, this arises only on completion, and can, in general, only be discharged upon providing the original copies of the relevant documents under s.13A CPO. 2 Breach of each duty will constitute a repudiatory breach of contract, entitling the purchaser to refuse completion.
As a general rule, the vendor's duty to give good title entails the removal of any defect in title at the date of completion. This involves providing the requisite conveyancing evidence to establish that his title to the property to be sold is neither encumbered by an interest, nor rendered defeasible (void or voidable). The duty to supply original title deeds at the date of completion is, thus, a subset of the duty to give good title.
The twin duties of giving and proving title exist in Hong Kong because Hong Kong does not adopt a system of title registration (e.g. that in the United Kingdom), but instead adopts one of deeds registration. In a system of title registration, the register serves generally as conclusive proof of a vendor's title. However, a system of deeds registration differs in a crucial respect. The register does not serve as any conclusive proof of title, but rather, it serves as conclusive proof of the existence of deeds, and determines the priority between such deeds. As the existence of a deed per se does not indicate that the deed has been executed in a proper fashion, the determination of priority between competing instruments does not guarantee validity. Thus, if a purchaser of property wishes to ensure that he receives good title to such property, the register can only be of limited assistance. The purchaser will therefore need to turn to the vendor himself to show and provide good title, by examining the relevant title deeds dating back to the root of title to make sure that the chain of title is unbroken. This system is also adopted in the Canadian provinces of Quebec, Prince Edward Island, and Newfoundland and Labrador.
B. From Yiu Ping Fong to s.13A
The duty to provide original title deeds upon completion had been thought to originate from the common law, and the best starting point is the case of Yiu Ping Fong. 3 In that case, Yuen J noted that a vendor cannot provide good title merely by ‘handing over only certified true copies at completion without an adequate explanation as to why the originals cannot be handed over’. 4 The learned judge continued by noting that ‘the handing over of original title deeds and documents (or at least those which relate exclusively to the property being sold) is an important part of the vendor's obligation in a sale of land. That obligation is well established in the common law’. 5
The then-accepted understanding of Yuen J's remarks was that all title deeds, dating back to the ultimate root of title, 6 ought to be handed over upon completion, subject to the Re Halifax 7 rule. This finds support from Yuen J's emphasis on ‘[deeds] which relate exclusively to the property being sold’: the fact that Yuen J did not seek to limit the scope of the duty any further would corroborate the above conclusion. Indeed, this understanding was echoed in Loyal Hope Ltd v Leung Pui Ming 8 : ‘[T]his court is unlikely to entertain any argument against the subsistence of a purchaser's common law right to insist production of original title documents relating exclusively to the property on the date of completion of the land sale.’ 9
The rule in Re Halifax
10
provides that if the original title deeds are missing, the vendor can satisfy the duty to provide original title deeds by producing a statutory declaration as to the following three matters:
How the missing title document came to be lost or destroyed; and The contents of the document (by secondary evidence); and Proof of the document's due execution.
However, it is often the case that such statutory declaration requirement cannot be met, for instance in the case where the vendor simply did not know of the existence and contents of the document.
11
Also, a vendor may be unable to satisfy this requirement because of issues relating to mental incapacity.
12
As such, The Hong Kong Law Society viewed the duty to provide original title deeds as imposing an excessively onerous burden on prospective vendors. 13 Title documents may date back to the 19th century, and to insist on their production would not only entitle purchasers to escape bad bargains, but it would also stifle the efficiency of the Hong Kong conveyancing system. As a result, the Hong Kong Legislative Council decided to pay heed to the Law Society's suggestions, culminating in the enactment of s.13A of the Conveyancing and Property Ordinance in 2008.
Section 13A(1) clearly appears to make it easier for a vendor to give good title: most importantly, it restricts the time period over which title documents relating to the property
14
need to be produced.
15
Section 13A thus requires the following to be produced. First, the title deeds and documents relating exclusively to the property with respect to the period from the Intermediate Root to the current sale and purchase agreement.
16
The ‘Intermediate Root’ is the document referred to in s.13(1)(a)(ii), i.e. an assignment, mortgage by assignment, or legal charge, which deals with the whole interest in the land, and dates at least 15 years back from the current contract of sale. Second, any power of attorney executed within 15 years of the current sale and purchase agreement (“
The implication of this is that all original title deeds that pre-date the intermediate root period no longer need to be provided to the purchaser.
20
It can thus be seen that s.13A imposes a temporal qualification on the documents that need to be produced, in limiting the chain of title period to the period from the Intermediate Root onwards (
Nonetheless, s.13A(1) makes clear that the default position shown above is not to apply if “the contrary intention is expressed”. Thus, this preserves the common law threshold under Yiu Ping Fong as interpreted in the subsequent case of Loyal Hope, should parties elect to adopt this threshold by agreement. There are therefore two distinct legal sources of the duty to give good title, originating from common law and s.13A. 21
C. De Monsa and its interaction with s.13A
The enactment of s. 13A did not initially alter the Yiu Ping Fong understanding of the common law position (in requiring all title deeds relating exclusively to the property to be produced). This was the position adopted by the Hong Kong Court of Appeal (“
However, the Court of Appeal's judgment in De Monsa was overturned by a unanimous decision of the Hong Kong Court of Final Appeal (“
However, what is far from settled is the relationship between the common law position under De Monsa and the statutory position under s.13A. This has been thrown into sharp relief by a series of obiter dicta in De Monsa itself, 25 and the subsequent case of Chu Yin Fan. 26 These two positions (the Court of Final Appeal's position in De Monsa and Samson Hung DJ's position in Chu Yin Fan) will be analysed in turn.
I) The De Monsa interpretation of s.13A
Addressing the interaction between s.13A and the common law, Ribeiro PJ and Gleeson NPJ noted ‘for future reference that s.13A(2) allows recourse to the common law if the specified originals cannot be delivered’.
27
Meanwhile, Litton NPJ noted that “[w]here section 13A(1) imposes an obligation on the vendor, for the purpose of giving title, to deliver to the purchaser an original title deed, that requirement is subject to the common law rule: see subsection (2).”
28
This indicates that where the vendor is obliged to supply original title deeds under s.13A(1), i.e. where the document(s) in question fall within the chain of title period,
29
the document does not necessarily need to be produced. It only needs to be produced if its absence gives rise to a real risk of rights relating to the land being enforced against the current purchaser. In effect, the position adopted by Ribeiro PJ, Gleeson NPJ and Litton NPJ combines the temporal qualification and the purposive qualification: a vendor needs to fall outside both qualifications (15 years and ‘no real risk’)
30
before he breaches his duty to provide original title deeds (
On its face, s.13A(2) would seem to support the Court of Final Appeal's interpretation as s.13A(1) is expressed to be unaffected by any “rule of common law under which the vendor may discharge his obligation to give title to that land”. 31 Nonetheless, for such an interpretation to stand, it must be proven that there is legislative authorisation for imposing the additional purposive qualification on the duty to produce original title deeds (given that the temporal qualification is clearly authorised by statute). Thus, it is imperative to examine the legislative intention of s.13A as a whole.
As the Hansard debates show, the Legislative Council's aim was to correct the threshold required to satisfy the duty to give good title because the common law duty was viewed as being too harsh. Indeed, Legislative Councillor Miriam Lau noted the following during the 2008 Hansard debate: “[The effect of s.13A is that] a purchaser of land shall be entitled to…the delivery of the original of only the Government lease if it relates exclusively to that land and any document that relates exclusively to that land… This will overcome the problem of inability to produce the originals of many title deeds. Further, it would no longer be necessary to trace pre-intermediate root title documents for the purpose of proving and giving good title.”
32
Implicit in this passage is that s.13A is merely intended to rectify the situations involving the former need to ‘produce the originals of…title deeds’, and ‘trace pre-intermediate root title documents for the purpose of proving and giving good title’, clearly referring to a temporal qualification to the duty in Yiu Ping Fong. It does not follow from this that LegCo has authorised (through the enactment of s.13A) a further relaxation of the duty to that in De Monsa. The final threshold agreed to under s.13A was a considered choice, as a judgment of the correct balance to be struck between protecting the vendor and protecting the purchaser.
This understanding is corroborated by Sihombing and Wilkinson, who note that the effect of s.13A(2) is merely to ‘preserve…the common law rule in Re Halifax Banking Co and Wood, which was recognised as applying to missing original title deeds in Yiu Ping Fong’, 33 and thus not to authorise courts to modify the default standard to one lower than that mandated by s.13A(1).
Indeed, even from a linguistic perspective, s.13A(2) does not appear to authorise the discharge of the duty to produce missing originals from the intermediate root onwards through doing nothing (as would happen if one satisfies the ‘no real risk’ test under De Monsa). In using the words ‘discharge his obligation to give title to the land
As such, it is suggested that the Court of Final Appeal's reading of s.13A(2) in suggesting the cumulative ‘temporal and purposive’ qualifications pays insufficient regard to the purpose for which s.13A(1) was enacted. Section 13A(1) introduces the temporal qualification, but LegCo does not appear to authorise the additional purposive qualification through s.13A(2).
Ii) The Chu Yin Fan interpretation of s.13A
In the subsequent case of Chu Yin Fan, 35 Deputy Judge Samson Hung inclined towards the view that “the obligations under Section 13A are the same as those at common law”, 36 meaning that as with De Monsa, s.13A imposes a purposive qualification, but not a temporal qualification. However, there are two reasons why this interpretation is untenable.
The first reason is that the Deputy Judge cites a passage in De Monsa
37
that merely supports the cumulative ‘temporal and purposive’ qualification, but does not go as far as to support the standalone ‘purposive’ qualification. The Deputy Judge noted the following: “I find support for this approach from the obiter dicta of Litton NPJ in De Monsa, supra, at paragraph 128 whereby he stated that:-
“Where section 13A(1) imposes an obligation on the vendor, for the purpose of giving title, to deliver to the purchaser an original title deed, that requirement is subject to the common law rule: see subsection (2). If an original in the chain of title is missing, and such loss gives rise to no risk that the title to be passed to the purchaser might be blemished, then the vendor would have fulfilled his obligation of giving good title under s.13A(1). “ (emphasis added)” 38
Litton NPJ's use of the words “[i]f an original
The second and more important reason is that if the effect of s.13A(2) were truly to preserve the De Monsa position in its entirety, then this would necessarily cut across the protections to the vendor offered by s.13A(1). As is apparent from the Hansard debates, s.13A(1) was intended to facilitate the discharge of the vendor's duty by limiting the necessary originals that need to be produced to those from the intermediate root onwards.
Indeed, Samson Hung DJ's view necessarily rests on the premise that s.13A(2) is not purely facilitative, but also contains a restrictive element, in that s.13A(2) can be interpreted so as to take away the rights conferred on the vendor by reason of s.13A(1). However, s.13A(2) is worded in a purely facilitative fashion: “Subsection (1) does not affect any rule of common law
Nonetheless, one may defend the Deputy Judge's view on the basis that although there is a conceptual difference between the cumulative ‘temporal and purposive’ qualification and the standalone ‘purposive’ qualification, its functional effect is effectively the same. 39 However, this is again not the case, as there are situations that satisfy s.13A(1) but do not satisfy De Monsa. The following example is one such situation. Purchaser P intends to purchase property X from vendor V. The last assignment was 15 years before the current purchase, and there is evidence that credibly suggests the possibility of an equitable mortgage by deposit of title deeds 18 years before the current purchase (which is evidenced, but not created, by a written memorandum). 40 When P asks V to produce original title deeds, V does so in relation to the assignment (which is the Intermediate Root) but cannot do so in relation to the equitable mortgage 18 years back. This situation does not satisfy the De Monsa threshold (i.e. the purposive qualification), because there is indeed a real risk of the mortgagee enforcing the mortgage against the current purchaser. 41 Yet, this situation satisfies s.13A(1), since the equitable mortgage was dated before the Intermediate Root.
As such, the view that s.13A imposes a standalone purposive qualification cannot be sustained.
Iii) The true position
From the above analysis, the true position is as follows:
The common law position under De Monsa imposes a purposive qualification and no corresponding temporal qualification on the duty to supply original title deeds. The position under s.13A imposes a temporal qualification but no purposive qualification on the duty to supply original title deeds.
The normative question: what should the law be?
There are two duties in play: the common law duty under De Monsa, and the statutory duty under s.13A. Yet, if both are to serve the same function, it makes little sense to retain the two distinct thresholds: doing so would merely cause further confusion. It is preferable to adopt one threshold for the sake of clarity.
This section will explore which approach is preferable from a normative standpoint. In sum, there are three main positions:
The ‘purposive qualification’ to the duty to supply original title deeds (i.e. the De Monsa threshold); The ‘temporal qualification’ to the duty to supply original title deeds (i.e. the s.13A threshold); and The ‘cumulative temporal and purposive qualification’ to the duty to supply original title deeds (i.e. the Court of Final Appeal's interpretation of the s.13A threshold).
It will be argued that in principle, the ‘purposive’ qualification is the most preferable. In practice however, the concern of retrospectivity means that the ‘purposive’ qualification needs to be modified.
A. The purposive qualification
I) Advantage: normative congruence
The main advantage of adopting the standalone purposive qualification is that it aligns the duty to supply original title deeds with its normative rationale, insofar as it allows the debate as to the ideal normative threshold to be reflected in the qualification itself.
In this regard therefore, it would be appropriate to examine the rationale for the duty to give title in more detail. It was established that the duty to give title is intended to (1) provide the purchaser with substantially what it had contracted to buy, and (2) furnish the means necessary for the purchaser to transfer title to a later purchaser. 42 Given these two objectives, what should the appropriate test be?
Lord Campbell LC in Moulton v Edmonds 43 stated the following: “Perhaps the test may be, whether the recited deeds not produced cast any reasonable suspicion upon the title shown by the deeds produced”. Indeed, it is suggested that this formulation upholds both objectives because if there is a realistic risk of enforceable proprietary rights being asserted against the purchaser, the purchaser would not be obtaining substantially what it had contracted to buy, and would not necessarily be able to transfer unencumbered title to a later purchaser. On the other hand, if there is no such risk, the purchaser would be obtaining substantially what it had contracted to buy, and would have the means of transferring title to a later purchaser without incurring legal liability. The test proposed by Lord Campbell would therefore achieve both objectives by protecting the purchaser against any real risk of future challenges to title. 44 Following this logic, it would be unfair on the vendor, and in any event superfluous, for any additional burden to be imposed in relation to such a duty. Indeed, Lord Campbell himself recognised the danger of imposing too high a standard on the vendor, namely that “the transfer of real property…would not only be difficult but would become often impossible, where there is an unwilling purchaser.” 45
More generally, is important to recognise that the duty to provide original title deeds (as part of the duty to give title) involves the classically unfortunate situation in which the court needs to allocate a necessary ‘shortfall’ to one of two innocent parties. 46 In this context, the shortfall needs to be allocated to either the vendor or the purchaser.
In most cases, the court is making the (often empirically uncertain) inference from a missing original title document as to three matters:
whether there is an interest created by the missing original; whether it is enforceable against the current purchaser; and whether the interest-holder will enforce the right against the current purchaser.
As the parties and the court are unlikely to know as to the existence or absence of one or more of the above three matters, it will have to make a predictive judgment as to elements (i)–(iii). Inevitably therefore, the court will not be accurate on every occasion. As such, a default rule is required in order to deal with the potential inaccuracy, i.e. the ‘shortfall’, in a just and equitable fashion. Indeed, on a more abstract level, distributing the shortfall equitably between the vendor and the purchaser is the normative function that the ‘duty to supply original title deeds’ has to perform.
The corollary question must therefore be this: how sure must the court be before deciding that the vendor should be given the benefit of the doubt, thus discharging his duty to provide the original title deeds? On one extreme, if we give no benefit of the doubt to the vendor, and presume there is no chance at all of (i)–(iii) in the absence of actual evidence of the same, the court would be foisting the shortfall entirely on P. On the other extreme, if we presume that every missing document gives rise to the existence of (i)–(iii), the court would be foisting the shortfall of the empirical uncertainty entirely on V. To ensure that a fair allocation of risk is achieved, both extremities must be avoided.
The ideal criterion of equitable distribution cannot possibly be settled by a short paper such as the present. Nonetheless, judges such as Lord Campbell LC in Moulton have attempted to formulate various tests to balance the shortfall equitably between the vendor and the purchaser. The ‘reasonable suspicion’ threshold proffered by Lord Campbell LC in Moulton is one such view as to how the shortfall should be distributed. One strong reason in support of this threshold is that it is closely aligned with the ‘real risk’ test for the duty to prove title under s.13 CPO 47 - a duty imposed on the vendor for the same purpose. 48
Admittedly, although Lord Campbell refers to ‘defect in title’ generally instead of (i)–(iii) above, it is suggested that Lord Campbell could not have meant ‘defect in title’ in the literal sense. If a ‘defect in title’ per se, i.e. a mere encumbrance irrespective of severity or likelihood of enforcement, is sufficient to entitle the purchaser to refuse completion, this would be inconsistent with Lord Campbell's own proffered threshold of ‘reasonable suspicion’. ‘Reasonable suspicion on title’ should refer only to all three elements of (i)–(iii) being satisfied. Thus, even if (i) and (ii) are satisfied, i.e. there is an encumbrance on the land giving rise to an enforceable interest, it is by no means a necessary inference that the purchaser needs to be concerned about the enforcement of such right. For example, if the interest-holder is a company that is dormant, there is no real risk of enforcement, and thus no ‘reasonable suspicion’ worthy of the purchaser's concern.
Lord Campbell is likely to have meant the following. If there is no reasonable suspicion (but merely a theoretical possibility) of the existence of (i)–(iii), the purchaser should bear the shortfall, with the vendor obtaining the benefit of the doubt. If however there is a reasonable suspicion as to the existence of (i)–(iii), the vendor should bear the shortfall.
At this juncture, one may object on the basis that the threshold for ‘equitable distribution’ is unsettled: why, for example, should reasonable suspicion be the touchstone for determining who should bear the shortfall? There is a straightforward response: although the precise threshold of this equitable distribution may be debated, adopting the purposive qualification allows the central normative debate to take place transparently. The precise threshold can, if necessary, be amended to reflect the ideal normative balance.
Ii) Disadvantage of uncertainty
Conversely, the principal weakness of having the De Monsa approach apply universally is its uncertainty: how is a vendor to know whether the missing title document gives rise to a risk of successful assertion of a proprietary interest that is superior to that of the purchaser? In particular, since it is an open-textured threshold making it difficult to produce clear-cut answers, vendors and purchasers alike may not be able to act swiftly and may engage in strategic behaviour. In contrast, s.13A carries the advantage that once the intermediate root of title is identified, a vendor would simply not need to worry about prior documents. 49
B. The temporal qualification
The temporal qualification represents the current position under s.13A. However, it will be argued (through an examination of the merits and demerits of adopting this threshold) that the current position is ultimately unjustifiable as it represents a significant departure from the normative rationale of the duty to supply original title deeds.
I) The advantage of certainty
In direct contrast with the purposive qualification, the temporal qualification carries the advantage of certainty. Where an original title deed is missing, the vendor will know whether he has discharged his duty to supply original title deeds simply by looking at the date of execution of such deed. If it pre-dates the Intermediate Root, it does not have to be produced. If it is within the chain of title period, it does have to be produced. No further inquiry arises. Parties do not need to assess whether there is a ‘real risk’ or surmise as to what this threshold requires.
Ii) Disadvantages
A) Not calibrated with normative rationale
However, the principal disadvantage of the standalone temporal qualification is that it falls out of kilter with the normative rationale of the duty to supply original title deeds. 50 This is for two reasons. First, where the missing original falls within the chain of title period, there would be a breach of duty even if its absence gives rise to no real risk of encumbrance and/or enforcement against the purchaser. This puts an undue burden on the vendor and allows the purchaser to escape a bad bargain without good reason: most notably where a purchaser wishes to take advantage of a falling market and escape its existing contractual obligations to the vendor. Second, where the missing original pre-dates the intermediate root, the duty to supply original title deeds is automatically satisfied even if the missing original gives rise to a real risk of enforcement of an interest against the purchaser. This unreasonably foists the risk of such enforcement on the purchaser.
B) Over-inclusiveness – post-intermediate root but no real risk
The over-inclusiveness of the standalone ‘temporal’ qualification is pointed up by a hypothetical scenario raised by Chan PJ in De Monsa:
51
“Take the common example of an owner who had for some reason failed to pay management fees and as a result, the owners incorporation imposed a charge on his property and registered it at the Land Office but the arrears were later paid and the charge was removed.”
52
In such a situation, it would flout common sense to contend that the purchaser can insist on delivery of such documents or a satisfactory explanation of their absence, in default of which he can refuse completion. Indeed, as Chan PJ notes, it would defeat the purpose of a sale and purchase transaction if the absence of the most trivial or insignificant title documents that do not give rise to any realistic possibility of affecting good title could allow a purchaser to refuse to complete. 53 For example, if the missing original is a mortgage deed, and the original of the discharge of such mortgage is provided, it would be unequivocally clear that the mortgage, which has been discharged, cannot affect the title of the purchaser. Thus, the missing original mortgage deed simply ought not to be a ground for the purchaser to refuse completion.
C) Under-inclusiveness: equitable mortgage by deposit of title deeds
The converse is also true. It would flout common sense if, given that a missing original title deed is thought to give rise to a realistic possibility of a prior interest encumbering or defeating the title of the current purchaser, the outcome of whether the vendor manages to discharge its duty to supply original title deeds depends on the arbitrary contingency of the time at which the interest was created.
In De Monsa, the Court of Final Appeal held that the main risk arising from missing original title deed(s) pre-dating the intermediate root is that of an equitable mortgage by deposit of title deeds, giving rise to currently enforceable rights.
54
Yet, it was noted that this risk is, in general, merely a fanciful possibility. Two lines of reasoning were adopted by the court: (i) in most instances, the mortgage is supported by a memorandum of deposit, and as such would be unenforceable against a subsequent purchaser under s.3(2) of the Land Registration Ordinance (Cap. 128, Laws of Hong Kong;
It is suggested however, that the Court of Final Appeal had underestimated the risk involved in relation to an equitable mortgage by deposit of title deeds encumbering or defeating the current purchaser's title, as a result of (erroneously) treating a memorandum of deposit necessarily as the instrument that creates the equitable mortgage by deposit of title deeds. The risk of such a mortgage encumbering or defeating the purchaser's title is ultimately a fact-sensitive question, and there are indeed situations in which the risk cannot be dismissed as merely fanciful. 55
I. Enforceability against subsequent purchasers
In Yiu Ping Fong, as well as the Court of Appeal in De Monsa, the judges seemed to take for granted that there could be a real risk of an equitable mortgage by deposit of title deeds giving rise to rights enforceable against the current purchaser. However, Litton NPJ in De Monsa took a different view. He relied on the reasoning that where there is an equitable mortgage by deposit of title deeds, 56 there is almost always an accompanying memorandum of deposit. As such, if the memorandum is not registered, then the equitable mortgage is (at the very least) unenforceable against subsequent purchasers of the property under s.3(2) of the LRO, even if the purchaser has actual knowledge of the mortgage. 57 Thus, he takes the view that in general, there is only a ‘real risk’ of the equitable mortgage by deposit of title deeds encumbering the current purchaser's title where the mortgage is purely created by parol, with no written memorandum of deposit. This view is consistent with Merry's remarks that the non-registration of a memorandum leads to automatic voidness against subsequent instruments. 58
The Court of Final Appeal continued on to note that it was difficult to imagine a situation where an individual would want to adopt an arrangement whereby “a person would lend a substantial sum of money with nothing more than an oral commitment and a deposit of title deeds, when he could have safeguarded his security by requiring a memorandum of deposit and having it registered”. 59
Ii. Memorandum versus underlying transaction
However, it is suggested that the Court of Final Appeal elides the crucial distinction between the memorandum and the underlying transaction. The two are not necessarily co-extensive. It is only where the memorandum creates the underlying transaction that the non-registration of such memorandum invalidates the underlying mortgage transaction. An equitable mortgage by deposit of title deeds may be created by parol and merely evidenced by a written memorandum. This may be the case especially in relation to family arrangements where there is likely to be less formality involved. In such a case, the non-registration of such memorandum does not invalidate the underlying transaction, given that the underlying transaction is created by parol. This is because “s 3(2) [of the LRO] only renders null and void a registrable, but unregistered instrument, but not the underlying transaction insofar as it can survive without the instrument”. 60
As such, where the mortgage is created by parol and merely evidenced by a separate written memorandum, it is the common law rules of priority, and not the LRO rules, that determine whether such a mortgage is enforceable against the current purchaser. 61 The common law rules provide that a prior equitable interest is enforceable against subsequent legal interest(s) provided that the holder of the subsequent interest is not a bona fide purchaser for value without notice. 62 Here, the very fact that the original title deeds are missing would put the subsequent purchaser on inquiry as to the validity of the vendor's title. The failure to answer such inquiry (i.e. disproving the existence of an equitable mortgage by deposit of title deeds) will render the purchaser fixed with notice if he carries through with the transaction, making him take subject to the mortgage. Indeed, in Oliver v Hinton, 63 quoted by the Court of Appeal in De Monsa, 64 it was held that a purchaser who had failed to ask for the production of title deeds for inspection took subject to an equitable mortgage by deposit of title deeds. From this, it can be extrapolated that all subsequent purchasers who know of the missing original title deeds would not be bona fide purchasers, giving rise to the risk of the earlier equitable mortgage encumbering the purchaser's title.
What, then, is the role of registering a memorandum of deposit where such memorandum merely evidences the existence of the oral mortgage? Its relevance lies in the fact that registration gives notice to the subsequent purchaser such that under the common law rules, he would not qualify as a bona fide purchaser for value without notice. It is not relevant to the determination of statutory priority between two instruments under the LRO.
III. A merely fanciful risk?
The foregoing analysis has established that there is the possibility of the purchaser's title being encumbered by an earlier equitable mortgage by deposit of title deeds. The corollary question thus needs to be tackled, namely: what is the risk of a mortgagee enforcing the terms of its mortgage against the land 15 + years later? The Court of Appeal in De Monsa, which did not mistakenly assume that the applicable rules were necessarily the LRO rules,
65
was unable to conclude that “any risk of an equitable mortgage [by deposit of title deeds] is so extremely remote that it should be [dis]regarded”.
66
The Court continued: “77. I believe it is invidious to have to decide such and such a chargee or assignee is more or less likely than others to be fraudulent. It is not a task on which a purchaser should be required to embark.
78. It lies ill in the mouth of a vendor to say that the risk is remote, when it is within its power, to remove the risk completely, for example, by properly accounting for the absence of the originals. I do not rule out the possibility that in a suitable case, a vendor, who, in spite of all reasonable efforts, is unable to explain a missing title deed, may be permitted to show that there was no reasonable doubt about the title.” 67
Indeed, it is difficult to disprove the risk of an equitable mortgage being enforced against the property, not only because the right subsists as against all subsequent purchasers, but also because the inherent nature of this inquiry entails the difficult exercise of proving a succession of negatives (that there was no equitable mortgage by deposit of title deeds, that such mortgage does not give rise to enforceable rights, and that such rights are not likely (or contemplated) to be enforced against the current purchaser).
The mere passage of 15 (or even 20) years per se is insufficient to render the risk of an equitable mortgage by deposit of title deeds so ‘fanciful’ as to be disregarded outright. Suppose that a mortgage was granted by a son in favour of his mother 18 years before the current purchase, and the son subsequently assigns the property to an intermediate purchaser 15 years before the current purchase, without producing the original title deeds (such assignment being the Intermediate Root). 68 If under this family arrangement the right(s) of action for repayment only begin to accrue (e.g.) 10 years after the creation of the mortgage, as may well be common in family arrangements, the mother may enforce such right against the current purchaser, as her action is not barred by the Limitation Ordinance. 69 This is not an inconceivable situation, as laypersons may not be aware of their legal rights until they are informed of them, which may be many years after the right of action accrues.
In addition, as Sihombing and Wilkinson note, 70 a lender may not wish to have his transaction made known to the public “for one reason or another”. This may well be the case in the mother-son example above, if other family members strenuously oppose such an arrangement. Creating an oral mortgage by deposit of title deeds (which may be accompanied by a supporting memorandum evidencing the terms of such mortgage) would allow the existence of the mortgage, as well as its terms, to be kept secret.
Equally, the risk of an equitable mortgage by deposit of title deeds in the commercial context cannot be described as ‘fanciful’ either. For example, up until 2012, it was common for homeowners under the Hong Kong Government's Home Ownership Scheme (“
These arrangements often involved homeowners depositing their title deeds with the lender, where the lender would take ‘custody’ of the deeds. 75 Where such an arrangement is created orally (with the terms evidenced in a written memorandum), it would prima facie take effect as an equitable mortgage by deposit of title deeds. 76 Nonetheless, because s.17B of the Housing Ordinance (Cap 283, Laws of Hong Kong) places restraints on the creation of a mortgage over a HOS flat, this raises the question of whether an equitable mortgage by deposit of title deeds would be prevented from arising.
Specifically, s.17B states that a purported mortgage in respect of a HOS flat would prima facie be ‘void’, meaning that the arrangement described above would not be effective to create an equitable mortgage by deposit of title deeds. However, s.17B also contains an exception, where the mortgage would be valid if government approval is obtained. 77 Thus, if the government approves a homeowner's plan to raise funds by way of mortgage, an equitable mortgage by deposit of title deeds 78 would not be struck down by the section. 79 Consistently, Merry notes that certain moneylenders become “approved replacement mortgagees”, since if a HOS homeowner succeeds in proving financial hardship, he or she can be permitted by the government to secure a larger loan through entering into a ‘substitute’ mortgage. 80 Additionally, such a mortgage may end up having legal effect in other ways. 81 It is clear therefore that the risk of an equitable mortgage by deposit of title deeds is not merely a ‘fanciful’ possibility, even in the commercial context.
The follow-up question arises as to whether such a mortgage would bind a current purchaser, and in particular, whether the restrictions on assignment under the Housing Ordinance make any difference to the analysis compared to the ‘family’ scenario previously explored. 82 It is suggested that there is no material difference, since homeowners under the Home Ownership Scheme are restricted by the government from assigning the property for just five years from the date of purchase. 83 Therefore, these owners would be free to assign the property after the five-year period. Where the property is so assigned, such assignment would constitute the Intermediate Root if the current purchase is (e.g.) 15 years after the assignment. Thus, the analysis is no different in this context: as in the ‘family’ situation described above, there may be rights enforceable against the current purchaser, since the right of action may have accrued (e.g.) 10 years after the creation of the mortgage, meaning that the mortgagee's right of action may not be barred by the Limitation Ordinance. 84 Thus, the risk of an equitable mortgage affecting the purchaser's title cannot be described as ‘fanciful’.
Iii) Conclusion: over-inclusiveness and under-inclusiveness
Taking into account the normative rationale of the duty to supply original title deeds, 85 it appears that placing a temporal qualification (as s.13A(1) does) is both over-inclusive (excessively favouring the purchaser) and under-inclusive (excessively favouring the vendor).
On the one hand, s.13A(1) requires the production of all relevant documents post-dating the intermediate root even if the absence of such documents is not referable to a suspected defect in title: this would include the obligation to produce the ‘original of a discharged mortgage’, even though its absence does not give rise to any risk of a defect in title. 86 If the absence of such a document entitles the purchaser to refuse completion, this imposes an inordinate burden on the vendor, since the duty should only require him to ensure there is no real risk that the purchaser's title would be adversely affected (which would be fulfilled if the purposive qualification were to be adopted). Any further obligation would be superfluous and draconian. It would also open the door for exploitation by the purchaser, since it would allow him to escape from a bad bargain without good reason, especially when the only reason that he wishes to avoid performance under the contract is because there is a falling market.
On the other hand, if there are missing documents pre-dating the intermediate root that are suspected to give rise to a defect in title, s.13A(1) exempts them from having to be produced. This overly favours the vendor, since it would be able to foist the risk of pre-intermediate root defects on the purchaser without incurring a breach of duty. The primary defect that could affect the purchaser's title is the equitable mortgage by deposit of title deeds, which (as we have established) cannot be dismissed as being a ‘fanciful’ possibility. In the event that such mortgage is enforced against the purchaser, this would cause substantial prejudice to him as he would be left with limited remedies. 87
C. Cumulative temporal and purposive qualification
The cumulative temporal and purposive qualification would appear to be the middle ground that mitigates the disadvantages of both the standalone purposive qualification and the standalone temporal qualification. At the same time however, the advantages of both standalone qualifications are significantly curtailed.
I) Certainty
In terms of certainty, the cumulative temporal and purposive qualification is an improvement from the pure purposive qualification. This is because a vendor need only worry about the question of ‘real risk’ in relation to documents within the chain of title period. However, in comparison to the pure temporal qualification, the cumulative temporal and purposive qualification falls drastically short. This is because vendors and purchasers alike need nonetheless to apply themselves to the question of ‘real risk’, which is a separate and additional inquiry. Also, this additional inquiry is a qualitative exercise in examining risk, as opposed to the quantitative exercise of ascertaining whether a document falls within the chain of title period under the standalone temporal qualification.
Ii) Calibration with normative rationale
In terms of whether the cumulative temporal and purposive qualification is calibrated with the normative rationale of the duty to supply original title deeds (i.e. whether or not it causes a lower risk of injustice), there is a much lower risk of injustice as compared to the pure temporal qualification. This is because the ‘real risk’ threshold goes towards filtering out unreasonable attempts on the purchaser's part to refuse to complete. For example, it prevents the ‘charge’ situation explained by Chan PJ in De Monsa, 88 or the ‘missing original of a discharged mortgage’ 89 situation from arising, meaning that purchasers are not given undue freedom to escape from a bad bargain, especially if the only reason why they wish to avoid the contract is because there is a falling market. This would run directly counter to the principle of pacta sunt servanda.
However, the cumulative qualification carries a higher risk of injustice as compared to the standalone purposive qualification, as the risk of an equitable mortgage by deposit of title deeds cannot be disregarded outright. Thus, restricting the chain of title period to that under s.13A(1) would mean that even if there are strong positive facts suggestive of there being an equitable mortgage by title deeds giving rise to a real risk of enforcement against the purchaser, the vendor would be able to turn a blind eye to this contingency and discharge its duty to the purchaser, leaving the latter to bear the risk of such enforcement without adequate protection.
D. Taking stock
Following from the above analysis, it is suggested that the purposive qualification represents the best position in principle. This is because it is the only position that conforms to the normative rationale of the duty to supply original title deeds upon completion, and because the potential concerns regarding ‘uncertainty’ can be allayed with relative ease.
I) Rejecting the temporal qualification
Certainty is the primary advantage of the temporal qualification. Nonetheless, certainty per se is insufficient to allow the incorrect normative balance to be struck, especially as it has been shown that there is a clear mismatch between the temporal qualification and the normative rationale of the duty to supply original title deeds. As has been shown, the temporal qualification is both over-inclusive and under-inclusive.
The weakness of the temporal qualification can be demonstrated through the example where a purchaser would be entitled to refuse completion in the event of a missing original mortgage deed, even if such mortgage has been discharged and there is a deed of discharge produced to the purchaser. As mentioned above, this confers undue freedom on the purchaser to escape from a bad bargain. 90
Ii) Rejecting the temporal and purposive qualification
Although the temporal and purposive qualification mitigates some of the disadvantages of both the standalone temporal qualification and the standalone purposive qualification, it is suggested that such a position represents an uncomfortable halfway house which neither reflects the normative rationale of the duty to provide original title deeds, nor fully achieves the goal of certainty.
Firstly, the cumulative temporal and purposive qualification applies the normative rationale of the duty to supply original title deeds in an incomplete fashion. As the purpose of the ‘real risk’ test is to distribute the shortfall in a just and equitable fashion as between vendor and purchaser, this should apply uniformly irrespective of the date of the missing original title deed. Thus, as the cumulative temporal and purposive qualification only adopts the ‘real risk’ test within the chain of title period, it does not reflect the normative rationale of the duty in its entirety.
Secondly, even if certainty is to be a serious concern, the cumulative temporal and purposive qualification is far less certain than the standalone temporal qualification: the question of ‘real risk’ is to be addressed for every missing original falling within the chain of title period.
In reflecting neither the normative rationale of the duty to provide original title deeds nor providing maximum certainty, it is suggested that this position attempts too much and achieves too little. The way forward should be for Hong Kong law to adopt the purposive qualification, as discussed below.
Iii) In favour of the purposive qualification
First, the concern for ‘certainty’ per se is insufficient to allow the courts to draw the incorrect normative balance. This is especially so since the ‘certainty’ problem can be mitigated by contracting out of the default rule, e.g. to restrict the chain of title period to 15 years by consensual agreement.
The prime objective of the default rule is to distribute the shortfall in a just and equitable manner, especially since parties who go by the default rule are unlikely to have applied their minds to the possible contingency of title deeds being missing. Thus, it is of paramount importance to prevent the default provision from drawing the normatively incorrect balance.
Also, the Court of Final Appeal in De Monsa did not appear to view the imposition of a purposive restriction such as ‘real risk’ as being too uncertain. Litton NPJ was perfectly content as a matter of principle to adopt the purposive qualification within the chain of title period, 91 with there being no indication whatsoever that such a qualification would render the law too uncertain. This, coupled with the court's conclusion that missing documents from the pre-intermediate root period is unlikely to give rise to any risk of enforceable rights encumbering or defeating the current purchaser's title, indicates that the Court does not view the adoption of a standalone purposive qualification as being too uncertain.
In any event, the ‘uncertainty’ problem can be mitigated by issuing general guidelines as to what constitutes ‘real risk’ or ‘reasonable suspicion’. This can be, for example, divided into three separate parts. First, the guidelines can refer to situations that involve no risk of affecting the value of the purchaser's title, such as a missing mortgage deed where the mortgage has been discharged. Secondly, the guidelines can also refer to situations where evidence additional to the missing deed (such as the surrounding circumstances or features of the transaction) needs to be adduced before ‘real risk’ or ‘reasonable suspicion’ can be established (e.g. where the missing deed is dated (say) 25 + years before the current purchase). 92 Finally, the guidelines can refer to situations where the fact that the deed(s) are missing per se would give rise to a real risk that the title of the purchaser will be substantively affected.
E. The best position in practice
Though the best position in principle would be to adopt the standalone purposive qualification, the practical reality is that in order to achieve this, s.13A needs to be repealed.
It is suggested that s.13A cannot be repealed without causing significant upheaval. This is because s.13A has been enacted for approximately 10 years, and thus vendors have relied on the fact that under this regime, the production of title deeds that include and post-date the intermediate root is sufficient to discharge the duty to give title. Yet, if s.13A were to be repealed, vendors would be required to produce pre-intermediate root documents as long as the ‘real risk’ threshold is satisfied. It may well be difficult for vendors now to dredge up any pre-intermediate root documents (should they give rise to a realistic possibility of an encumbered or defeasible title being enforced against the purchaser). A vendor may have to contact one or more predecessors in title for such documents, which may be unrealistic given the speed at which land transactions are conveyed in Hong Kong. In addition (and more importantly), even if such predecessors in title can be contacted, there is no assurance that such predecessors have not lost or thrown away the originals of the missing pre-intermediate root title deed(s).
This affects the saleability of the property, as vendors may have to negotiate to contract out of the default rule (i.e. the standalone ‘purposive’ qualification) in order to shorten the chain of title period: this may well cost an extra fee for the vendor, and/or affect the saleability of the property outright.
In response to this concern, it is suggested that s.13A should instead be amended to reflect the spirit of the purposive qualification as much as possible and prevent vendors from being retrospectively prejudiced. The latter concern is to be addressed by exempting the vendor from needing to produce any title deeds before a particular date. This date should be set at 15 years prior to the date on which the new provision comes into effect: thus, if the new provision comes into effect on 1 January 2022, any title deeds before 1 January 2007 would not need to be produced. This ensures that current property owners would not need to produce any deeds additional to those received from their vendors. 93
In turn, in order to reflect the purposive qualification, all title documents from such a date onwards need to be produced by the vendor, except if the absence of any title document(s) does not give rise to a 'real risk' of rights enforceable against the purchaser. Under the amended section 13A therefore, there will be three additions to subsection (1), namely s.13A(1)(c), (d) and a ‘proviso’ at the end. The other subsections will remain unaltered.
The new s.13A(1) should read as follows: “(1) Unless the contrary intention is expressed, a purchaser of land shall be entitled to require the vendor to deliver to him, for the purpose of giving title to that land, the original of both of the following only—
(a) if there is a Government lease that relates exclusively to the land, the lease; and
(b) any document that relates exclusively to the land and is required to be produced by the vendor as proof of title to that land under
section 13
(1)(a) and (c); and
(c) all title deeds dating back to [15 years prior to the date on which this provision comes into effect] that relate exclusively to the land.
Provided that, if any missing document(s) under (a)–(c) above do not give rise to any real risk of the purchaser's title being affected, they do not need to be produced.”
Under the new amended section, the prima facie position is that there is no ‘time limit’ on the chain of title period, given we have established that the standalone temporal qualification should not be adopted.
At the same time however, there needs to be a ‘longstop date’ to prevent current property owners from being retrospectively prejudiced by the current regime. The ‘longstop date’ in s.13(A(1)(c) of ‘15 years prior to the date on which [the amended] provision comes into effect’ (i.e.1 January 2007 under the previous example) 94 ensures that current property owners who have received documents that are sufficient to comply with the current s.13A regime would not be prevented from discharging the duty to give title vis-a-vis subsequent purchaser(s).
It is important to note that this ‘longstop date’ is very different to the ‘time limit’ in respect of the current s.13A (namely the ‘Intermediate Root’), since the former is absolute whereas the latter is relative. With the Intermediate Root, the time period of ‘15 + years’ is assessed relative to the time of the purchase, meaning that the ‘minimum date’ changes over time: property purchased on 1 January 2022 instead of 1 January 2012 would produce a different minimum ‘chain of title period’ (namely 1 January 2007 instead of 1 January 1997). In contrast, the ‘longstop date’ under the amended section is a fixed date – for example, if the amended section takes effect from 1 January 2022, the ‘longstop date’ means that documents before 1 January 2007 would not need to be produced. This does not vary according to the date of a particular purchase: thus, for example, if a property is sold on 1 January 2030, the ‘longstop date’ remains 1 January 2007, as opposed to 1 January 2015 (which would be the minimum ‘chain of title period’ under the current regime). It is in this crucial respect that the amended provision departs from the standalone temporal qualification.
In turn, the amended section gives effect to the standalone purposive qualification through the ‘proviso’, as a document would not need to be produced if it does not ‘give rise to any real risk of the purchaser's title being affected’. This can then be supplemented by guidelines as to what would violate the ‘real risk’ threshold, as stated previously. 95
In sum therefore, it is suggested that this amended provision strikes the best balance between upholding the normative rationale of the duty to provide original title deeds and preserving the ability of current property owners to transfer good title to a subsequent purchaser.
Conclusion
It is clear that the current law relating to the duty to provide original title deeds at completion is not aligned with its normative rationale, as it is both over-inclusive (in that it excessively favours the purchaser) and under-inclusive (in that it excessively favours the vendor). As a result, this article has proposed that the way forward would be to amend s.13A of the CPO in a way that (1) recalibrates the scope of the duty with its normative rationale, and (2) ensures that current landowners are not prejudiced by the proposed amendment. The amended section should therefore adopt the standalone purposive qualification, subject to a ‘longstop date’ that prevents current landowners who have relied on s.13A from being retrospectively affected. If the amendments suggested in this article are taken on board, the law on the duty to supply original title deeds would strike the right balance between vendors and purchasers, such that it curtails opportunistic attempts by purchasers to refuse to complete, and protects purchasers who are subjected to a real risk that the property in question is encumbered by an adverse interest.
Footnotes
Acknowledgement
This article is dedicated to the late Professor Michael Wilkinson, who inspired me to write this piece.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship and/or publication of this article.