PENNY'S BAY INVESTMENT CO LTD v. DIRECTOR OF LANDS

Press Summary (English)

Press Summary (Chinese)

FACV No. 1 of 2017

and FACV Nos. 2-9 of 2017

IN THE COURT OF FINAL APPEAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

FINAL APPEAL NO.1 OF 2017 (CIVIL)

(ON APPEAL FROM CACV NOS. 13, 14, 15, 16, 115,116, 119 & 120 OF 2015)

____________________

____________________

IN THE COURT OF FINAL APPEAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

FINAL APPEAL NOS. 2-9 OF 2017 (CIVIL)

(ON APPEAL FROM CACV NOS. 13, 14, 15, 16, 115,116, 119 & 120 OF 2015)

____________________

____________________

__________________________________

JUDGMENT

__________________________________

Mr Justice Ribeiro PJ:

1.  I agree with the judgment of Lord Neuberger of Abbotsbury NPJ.

Mr Justice Tang PJ:

2.  I agree with the judgment of Lord Neuberger of Abbotsbury NPJ.

Mr Justice Fok PJ:

3.  I agree with the judgment of Lord Neuberger of Abbotsbury NPJ.

Mr Justice Bokhary NPJ:

4.  I agree with the judgment of Lord Neuberger of Abbotsbury NPJ.

Lord Neuberger of Abbotsbury NPJ:

Introductory

5.  These two appeals arise out of a decision of the Lands Tribunal (Deputy Judge Kot and Mr K K Chiu) given on 15 October 2014 (LDMR No 23 of 1999). The decision concerned the compensation payable to Penny’s Bay Investment Company Limited (“PBIC”) for the extinguishment of marine rights enjoyed with certain land it owned at Penny’s Bay (“the Land”). The compensation was assessed at $10,952,500 (subsequently adjusted by agreement to $9,431,000), for reasons set out in a judgment which ran to 481 paragraphs, as there were a large number of legal and valuation disputes between the parties.

6.  Following that decision, PBIC and the Director of Lands (“the Director”) each appealed on a number of issues. In a judgment given on 16 May 2016, the Court of Appeal (Hon Lam VP and Barma and Poon JJA) resolved a total of eight issues, upholding the Lands Tribunal on some and reversing them on others (CACV 13-6, 115, 116, 119 and 120 of 2015).

7.  PBIC appeals to this Court on an issue concerning the interpretation of the restrictive user provisions of the grant under which PBIC held the Land, and the Director appeals to this Court on an issue as to whether the Land should be valued on the assumption that the proposed development giving rise to the loss of marine rights was certain to proceed.

8.  The Court of Final Appeal has already delivered a judgment (“the 2010 CFA judgment”) dealing with two questions, which came before it in 2010 as preliminary issues prior to the hearing before the Lands Tribunal – see Penny’s Bay Investment Co Ltd v Director of Lands (2010) 13 HKCFAR 287.

The factual and statutory background

The factual background

9.  The Land lies near the eastern end of Lantau Island. It is now included in the site of Hong Kong Disneyland, which opened in 2005 and was built on about 300 hectares of land reclaimed from the waters of the bay. It has its own MTR station and fast road connections to Hong Kong Island and the airport. In 1994, however, as the Court of Appeal said, Penny’s Bay was a remote corner of the coastline, inaccessible except from the sea or on foot.

10.  Before its development, the total area of the Land extended to about 2,010,000 square feet or 18.6 hectares.

11.  The Land was acquired by PBIC in 1970 pursuant to an Agreement and Conditions of Exchange dated 2 January 1970 and registered in Tsuen Wan Land Registry as New Grant No 4706 (“the Grant”). The initial term of the Grant was 99 years, less 3 days, from 1 July 1898, and it was subsequently extended to 30 June 2047 by the New Territories Leases (Extension) Ordinance, Cap 150.

12.  Para (a) of clause 3 of the Special Conditions of the Grant (“clause 3”) stated that:

“the [Land] shall be used for general industrial and/or godown purposes [excluding any offensive trade as defined] and shall not be used for any other purpose whatsoever. … [I]n the event of the [Land] or any part thereof being used for godown purposes or for the storage of petroleum products or dangerous goods … the method of storage … shall be subject to the approval of the Director of Fire Services. Not less than 285,000 sq ft of the [Land] shall be used for shipbuilding purposes only.”

Clause 3(b) stated, inter alia, that “no building shall be erected on the [Land] except a factory and/or warehouse, ancillary offices and … quarters for watchmen or caretakers”, although clause 3(c) permitted the erection of residential accommodation for staff working in any industrial activity pursued on the Land.

13.  The Land enjoyed limited rights of access to the sea, and, reflecting the nomenclature which has been adopted throughout these proceedings, I shall refer to those rights as “the Marine Rights”.

14.  By a tenancy agreement made in December 1975, PBIC let the Land to a subsidiary, Cheoy Lee Shipyards Limited, which carried on the business of shipbuilding and repairing. At all times thereafter until the surrender of the Grant to the Government in 2001 as described below, the Land was used for shipyard purposes, which of course involved making use of the Marine Rights.

15.  On 11 March 1994, pursuant to section 5 of the Foreshore and Sea-Bed (Reclamations) Ordinance, Cap 127 (“FSRO”), the Director of Lands published Gazette Notice No. GN 847 identifying a scheme which involved reclaiming the foreshore and sea-bed situated at Penny’s Bay in order to provide for, inter alia, land for the construction of two container terminals, land for industry, back-up areas, and associated infrastructure (“the CT Scheme”). The CT Scheme involved substantial land reclamation (“Reclamation”), which would inevitably involve depriving the Land of the Marine Rights.

16.  On 24 March 1995, a draft Outline Zoning Plan No S/I-NELP/1 (“the Plan”) outlining the development plan for the CT Scheme was published. Under the Plan, the area to be reclaimed at Penny’s Bay, together with part of the Land, was zoned for industrial use, container terminals, container back-up area, a business park, and other incidental uses. The majority of the Land was zoned under the Plan for industrial use, and there was to be a new access road which passed through the Land and linked the North Lantau Expressway with the container terminal sites.

17.  On 25 April 1995, pursuant to section 8(2) of the Roads (Works, Use and Compensation) Ordinance, Cap 370, the Secretary for Transport published Gazette Notice No. GN 1450, which represented the proposed road scheme (“the road scheme”) associated with the CT Scheme. The notice described in general terms the nature of the proposed road works.

18.  On 5 May 1995, pursuant to section 8(1)(c) of FSRO, the then Governor in Council published Gazette Notice No. GN 1574 (“the 1995 Authorisation”), which formally authorised a proposal to reclaim about 1260 hectares of the foreshore and sea-bed at Penny’s Bay for the purpose of construction of two container terminals and associated infrastructure.

19.  In or about August 1999, the Government abandoned the CT Scheme and instead promoted the Disneyland Project. The road scheme, which was a necessary element of the CT Scheme and which had never been implemented, was also abandoned.

20.  Subsequently, by Gazette Notice No. GN 2230 dated 10 April 2000, the 1995 Authorisation was formally revoked. Four days later, on 14 April 2000, a new authorisation for the Disneyland reclamation in respect of a rather smaller area (about 330 hectares) was published as Gazette Notice No. GN 2231.

21.  The Land was required in order to implement the Disneyland project, and it was surrendered by PBIC to the Government by a Deed of Surrender executed by PBIC on 3 April 2001. Upon the execution of the Deed of Surrender, the Government paid PBIC a sum of $1,506,098,750, of which $1,483,380,000 was stated to be an ex gratia payment.

The statutory background

22.  The relevant provisions of FSRO are described in paras 16-22 of the 2010 CFA judgment, and I will limit myself to the provisions which are centrally relevant to this appeal.

23.  Sections 3-6 deal with preliminary steps to be taken before the Director can issue an authorisation for any reclamation (defined as “includ[ing] any work over and upon any foreshore and sea-bed”), and part of the reason for those steps is to enable objections to be made by anyone with “an interest, right or easement in or over the [relevant] foreshore” – ie any marine rights. If there are no objections or if any objections are overruled, the reclamation may be formally authorised under section 7 or section 8 respectively.

24.  Section 9 provides that notice of authorisation must be published, and that any such publication must state that anyone who considers that his marine rights “will be injuriously affected by the reclamation” may claim compensation. Section 10 provides, inter alia, that on publication of the notice all marine rights over the land designated for reclamation are extinguished, and section 11 states that the sole remedy of anyone injuriously affected is a claim for compensation under section 12.

25.  The legislative guidance on the question of compensation is sparse. It is contained in section 12 (“section 12”), subsection (1) of which provides as follows:

“Any person who claims that his [marine rights] will be injuriously affected by the reclamation may deliver to the Director a written claim stating the sum of money which he is willing to accept in full and final settlement of his claim …”.

Section 13 provides for either the claimant or the Director to refer the claim to the Lands Tribunal if the claim is rejected or the amount of compensation is not agreed within seven months of service of the claim.

26.  In these circumstances, it is unsurprisingly common ground that in this case publication of the 1995 Authorisation extinguished the Marine Rights, and that revocation of that Authorisation did not revive those rights. It is also agreed that PBIC had a vested right to compensation for loss of those rights as a result of the 1995 Authorisation. The question for the Lands Tribunal was how the compensation which had become payable to PBIC by virtue of the 1995 Authorisation should be calculated.

The 2010 decision of the Court of Final Appeal

27.  There were three preliminary issues before the Court of Final Appeal. They all arose from the question as to “how the compensation which had become payable by virtue of the 1995 Authorisation should be calculated and, in particular, the effect (if any) of the subsequent events” – see para 14 in the judgment of Lord Hoffmann NPJ, with whom the other members of the court (Bokhary PJ, Chan PJ, Ribeiro PJ, and Litton NPJ) agreed. As Lord Hoffmann explained in para 34, the injurious affection referred to in section 12 “means a diminution in the value of land caused by works authorised by statute which would otherwise have been tortious”.

28.  The first issue concerned the effect of events which occurred subsequent to the 1995 Authorisation on the valuation exercise required by section 12. As Lord Hoffmann explained in para 27, PBIC’s case was that compensation was to be assessed on the assumption that the 1995 Authorisation was effective, and that “the correct measure was the difference between the value on 5 May 1995 with marine rights and its value after their extinction” (para 28). Accordingly, PBIC argued, the valuation exercise required by section 12 was to be carried out without taking account of any event which occurred after that date. By contrast, the Director contended that, as the CT Scheme had been abandoned, compensation under section 12 should be limited to the loss “caused by the situation of the ship building business being precarious after the [1995] Authorisation and before the sale, together with the difference in the price for which the lease had been surrendered and what its value would have been with marine rights”. 

29.  On this first issue, after considering the wording of section 12 as well as the wording of the equivalent English legislation, Lord Hoffmann resolved the dispute in favour of PBIC, concluding at para 43 that “nothing which happened after [5 May 1995] can affect the valuation”.  This was based on his analysis of section 12, the effect of which he summarised in para 41, in a passage which I think should be quoted in full:

“Thus we have a statute which provides for compensation for injurious affection which will be caused by the reclamation but assumes that this has accrued and can be fully quantified before the reclamation has taken place – indeed, when it may be uncertain whether a reclamation will take place at all. In my opinion this can only mean that for the purposes of assessing the compensation, it must be assumed that on the date of authorisation it was certain that the reclamation would take place. This gives effect to the future tense in s.12: the owner claims that his land “will be” injuriously affected, not that it has been or may be so affected. The statute cannot have contemplated that the Government, having given notice that a reclamation was authorised, would be able to invite the Tribunal to speculate as to whether and, if so, when it might happen and have the compensation discounted accordingly.”

30.  Lord Hoffmann went on in para 42 to say that:

“What this means in practice is that the reclamation is treated as having taken place on the date of publication of the authorisation and the compensation is the difference between the open market value of land in actual enjoyment of its marine rights and its open market value when deprived de facto as well as de jure of its access to the sea”.

31.  Accordingly, compensation was to be determined in accordance with Lord Hoffmann’s guidance, by calculating the difference between the value of the Land as at 5 May 1995 with and without the Marine Rights. In these proceedings, this has been described as the difference between “the before value” and “the after value”, and I shall adopt those expressions.

32.  The second issue, which only arose if PBIC succeeded on the first issue (which of course it did), was whether the Director was right in contending that, as Lord Hoffmann put it in para 28, because “the value [of the Land] on 5 May 1995 after … extinction [of the Marine Rights] … should also take into account that the [CT Scheme] was expected to bring road connections to [the Land]”, this “would increase its value, probably more than any diminution caused by the loss of marine rights [and so] no compensation would be payable”.

33.  In dealing with this second issue, Lord Hoffmann said at para 44 that “the parties to the hypothetical sale on 5 May 1995 would have had expectations about the future which, whether right or wrong, would have influenced the price at which they were willing to deal”. He went on to say that the published plans relating to the Land as at 5 May 1995 may have led to the owner having “lively expectations about how the value of [the Land] would be enhanced by the proposed works”. But he added that “[w]hether such expectations existed and the extent, if any, to which they would have affected the open market value of the land is a matter for evidence when the valuation comes to be done”.

34.  In para 45, Lord Hoffmann explained that the assessment of compensation was to be carried out on the basis that “the prospect of new uses for the land, and the increased value which that may have generated, would have entered into the valuations both with and without access to the sea” (emphasis supplied). However, he observed that this did not necessarily undermine the argument that the differences in such valuations could be “not … very substantial” (or, I would add, zero) if the prospect of a new access road would have led to the Land being seen as having value attributable to a use which primarily or exclusively relied on road, rather than marine, access. He concluded in para 46 that both “valuations must take into account all the information which was public knowledge at the time and (apart from the assumptions about marine rights) not be based on any artificial assumptions”.

35.  Thus, as Ribeiro PJ explained in a subsequent judgment on costs (FACV No 8 of 2009 dated 27 August 2010), the Court of Final Appeal plainly did not accept the Director’s contention that the after value would be greater than the before value as a result of the prospect of road access. Lord Hoffmann considered that the issue raised a matter of valuation, which could not properly be decided by the Court of Final Appeal as a preliminary point of law.

36.  Although there was a third issue, it was not the subject of any oral argument or any significant discussion in Lord Hoffmann’s judgment, and it is anyway of no relevance for present purposes.

The reasoning of the Lands Tribunal

37.  Thereafter, the valuation of PBIC’s claim for compensation for injurious affection was the subject of detailed expert and factual evidence and full argument before the Lands Tribunal. In accordance with the guidance given by the 2010 CFA judgment, the calculation of the compensation involved assessing the difference between the before value and the after value of the Land.

38.  As already mentioned, the Lands Tribunal gave a very long judgment because it had to resolve a plethora of different issues, both of law and of valuation, between the parties. Fortunately, only a few of those issues are relevant for the purposes of this appeal.

39.  While it involves a degree of oversimplification, the Lands Tribunal’s assessment of the before value of the Land as at 5 May 1995 can for present purposes be analysed as involving the following five steps:

 (a) Value of the Land assuming the Marine Rights continued until 2047, enabling shipbuilding use on 285,000 square feet and mid-stream use on the remainder, but with no industrial use (which was economically unfeasible without road access): $551,157,000;

 (b) Value of the Land assuming the Marine Rights continued (so shipyard use was still enabled on 285,000 square feet) but otherwise assuming that the CT Scheme went ahead (enabling industrial use to be carried out on the remainder): $841,828,000;

 (c) Additional value resulting from industrial development, as a result of the possibility of the CT Scheme proceeding was thus: $841,828,000 less $551,157,000 = $290,671,000;

 (d) Market assessment of the CT Scheme proceeding was assessed at 50%, so the additional value attributable to industrial development was accordingly half (c): $145,335,500;

 (e) Therefore, the before value was (a) plus (d): $696,492,500 (subsequently adjusted by agreement to $694,971,000). 

40.  Steps (c) and (e) of the before valuation were purely arithmetical and no more need be said about them. However, the other three steps involved evaluation or judgment, and four points should be made about them for present purposes.

41.  Steps (a) and (b) of the before valuation proceeded on the assumption that the shipyard use could continue for the duration of the Grant. This was because the before valuation assumed that the Marine Rights would not be lost (and that the other uses there described would also be able to benefit from the Marine Rights): the correctness of this assumption in relation to those steps is not in contention.

42.  Secondly, and this is contentious, the valuation exercise involved in step (a) included the assumption that “godown purposes” includes use for storage on open land, and more particularly, as the Lands Tribunal put it, for “mid-stream operation or open storage of containers”.

43.  As to steps (b) and (d) of the before valuation, there are two points to be noted, neither of which is contentious for present purposes, but both of which are relevant to the Director’s appeal before this Court. They each arise from the fact that, in order for the Land to be put to industrial use, the CT Scheme would have to be implemented (as the completion of that Scheme, and in particular the construction of road access to the Land, would have been seen as the only possible way by which industrial use could become economically viable).

44.  When assessing the value of the Land for industrial use as at 5 May 1995 at step (b), it was necessary to determine what the market would assess as the likely period before the CT Scheme was completed, and industrial use of the Land could accordingly begin. In the light of the evidence, the Lands Tribunal decided that eight years was the appropriate period, and accordingly the present day (or “dead ripe” as it is sometimes rather picturesquely called) industrial value of the Land as at 5 May 1995 had to be discounted to allow for that.

45.  Quite apart from delay, there would have been uncertainty in the market as at 5 May 1995 as to whether the CT Scheme would go ahead at all. In the light of the evidence they heard, the Lands Tribunal decided that the market as at that date would have assessed the prospects of the CT Scheme proceeding as being 50-50. Accordingly, the increase in the value of the Land attributable to its being able to be used for industrial purposes had to be adjusted downwards, and this was done at step (d) of the before valuation. 

46.  The Lands Tribunal’s after valuation of the Land as at 5 May 1995 was a shorter exercise. Again simplifying somewhat, it can be analysed as involving the three following steps:

 (a) Value of the Land for shipbuilding and godown use: $21,420,600;

 (b) Value of the Land for industrial use: $664,119,400;

 (c) Accordingly, the value of the Land is: $685,540,000.

47.  Step (c) of the after valuation simply involves aggregating the figures in steps (a) and (b) and nothing more need to be said about it.

48.  An important aspect of the assessments in both steps (a) and (b) of the after valuation is that the Lands Tribunal proceeded on the basis that, for the purposes of the after valuation, the CT Scheme was certain to go ahead.

49.  Step (a) of the after valuation, like step (a) of the before valuation, proceeded on the assumption that “godown purposes” included mid-stream storage use, and to that extent, but only to that extent, step (a) of the final valuation is controversial. Although step (a) of the after valuation involved taking the same value for shipbuilding use which was assessed as step (a) of the before valuation, it was then very substantially discounted by 90%. This was to allow for the fact that implementation of the CT Scheme would result in the loss of the Marine Rights, and would therefore put an end to shipbuilding use. However, rather than attributing no value to the shipbuilding use, the Lands Tribunal thought that some value should be attributed to it because of the prospect for a change of use.

50.  Step (b) of the after valuation ascribed an industrial value to the Land which was different from that contained in the before valuation. This was because the Lands Tribunal considered that, as they proceeded for the purposes of the after valuation on the basis that the CT Scheme was bound to go ahead, the delay factor should be six years, rather than eight years, as it had been in the before valuation. Furthermore, the after valuation included no discount for uncertainty when considering the effect of industrial use, as there was in step (d) of the before valuation. This was again because the Tribunal considered it right to carry out the after valuation on the basis that it was certain that the CT Scheme would proceed.

51.  Accordingly, the Lands Tribunal decided that the compensation payable to PBIC for the loss of the Marine Rights should be $694,971,000 less $685,540,000, namely $9,431,000.

The decision of the Court of Appeal and the issues on this appeal

52.  PBIC and the Director were dissatisfied with some aspects of the determination by the Lands Tribunal, and they each appealed to the Court of Appeal. As mentioned above, there was a total of eight issues raised in that Court, but for present purposes it is only necessary to identify four of them.

53.  The first relevant issue applied to step (a) of both the before valuation and the after valuation. It was whether mid-stream storage use was within the expression “godown purposes” in clause 3(a). In disagreement with the Lands Tribunal, the Court of Appeal did not consider that such a use was within the expression, and therefore it concluded that, subject to any further argument, mid-stream storage use was not permitted under the terms of the Grant. PBIC now appeals against the Court of Appeal’s conclusion on this issue.

54.  The second relevant issue arose from the rejection by the Lands Tribunal of an argument, to which I have not so far referred and which was raised by the Director, namely that, when carrying out the after valuation as at 5 May 1995, it should be assumed that the Reclamation, ie the reclamation works contemplated by the CT Scheme and the 1995 Authorisation, had actually been carried out. The Court of Appeal agreed with the Lands Tribunal on this issue and so rejected the Director’s appeal on the point. The Director was refused leave to appeal this aspect of the Court of Appeal’s decision by the Court of Appeal and the Court of Final Appeal.

55.  The third and fourth relevant issues can be taken together. They both arose from the fact that the Lands Tribunal carried out the after valuation on the basis of an assumption that as at 5 May 1995 the CT Scheme was bound to go ahead. As mentioned above, the Lands Tribunal therefore decided that the after valuation should be carried out on the assumptions that (i) the delay in being able to use the Land for industrial purposes should be six years rather than the eight years appropriate for the before valuation, and (ii) there should be no allowance for the prospect of the CT Scheme not proceeding and the Land therefore not being usable for industrial purposes, unlike the 50% allowance in the before valuation.  The Court of Appeal held that the Lands Tribunal were wrong on this issue, and that the after valuation should have been carried out on the basis of the same assumptions in relation to these two points as were made in the before valuation. In other words, step (b) of the after valuation should have involved the value for industrial use being adjusted to allow for (i) eight years, not six years of delay, and (ii) a 50% risk, as opposed to no risk, of industrial use not being available.

56.  PBIC now appeal against the Court of Appeal’s decision summarised in para 53 above. Contrary to the conclusion reached by the Court of Appeal, and in accordance with the reasoning of the Lands Tribunal, PBIC contends that “godown purposes” in clause 3(a) of the Grant included mid-stream storage use.

57.  And the Director appeals against the Court of Appeal’s decision summarised in para 55 above, namely that the Lands Tribunal were wrong, or not entitled, to assess the after valuation on the assumption that the CT Scheme was bound to be carried out.

58.  I will deal with those two issues in turn.

The first issue: the meaning of the user covenant

59.  The Lands Tribunal concluded that mid-stream storage use of the Land was permissible under the Grant because it constituted use of the Land “for godown purposes”, even though it involved storing goods on open land and a godown was, as they rightly accepted in their judgment, “a structure or building”. They justified this conclusion on the ground that the expression “godown purposes” was concerned with “the nature or purpose of the use instead of confining to whether a building was involved”, and they relied on the decision of Liu J in Cavendish Property Development Ltd v The Attorney General, HCMP 762/1987, 6 August 1988, as being “directly on point”.

60.  The Court of Appeal disagreed, stating, without giving much in the way of positively supportive reasons, that “a ‘godown’ must involve a building or part of a building. It does not cover storage in the open”.

61.  So far as Cavendish is concerned, I agree with the Court of Appeal that it is of no assistance. The issue in that case was whether use of six floors of a building, let on terms that they could only be occupied “for industrial or godown purposes or both”, could be lawfully used for the purposes of a bank’s data processing centre. Unsurprisingly, Liu J answered that question in the negative. Much more importantly for present purposes, it is self-evident from my very brief summary of the facts that, as the Court of Appeal said, the issue in the present case, namely “whether a godown must involve a building was simply not an issue at all” in Cavendish. Although the relevant words of the user covenant in that case were almost identical to those in the present case, the issue in the present case, whether storage on open land is “for godown purposes” simply could not have arisen in that case, which involved activities on six floors of a building.

62.  In para 63 of their judgment, the Lands Tribunal referred to Liu J’s “findings” that “a godown is a place where at least movement of the stored items would be expected” and “storage without more is not user for godown or warehouse purposes”, and concluded they were “directly on the point”. Sentences extracted from a judgement in one case can often appear to support a proposition being advanced in another case, but once those sentences are read in their context, it can be quickly seen that they provide no support for the proposition that storage of goods on open land can be use of that land “for godown purposes”. Thus, Liu J’s quoted observations were not made in connection with the question whether godown purposes have to involve a building, and they do not give, and they did not purport to give, any guidance on that issue. Indeed, bearing in mind the issue in Cavendish, it would be surprising if they had done so.

63.  It is right to record that, on this appeal, as in the Lands Tribunal and in the Court of Appeal, both parties referred (albeit only in their written arguments) to other judicial decisions on the interpretation of user restrictions in statutes or contracts. As is so often the position in cases concerning issues of interpretation, earlier decisions involving different documents and different facts are of no assistance, and can sometimes be positively unhelpful as they take the eye off the ball.

64.  Turning then to the specific issue raised on PBIC’s appeal, I agree with the conclusion reached by the Court of Appeal. As in many cases involving a dispute as to the interpretation of a few words in a document, there is a limited amount which can usefully be said to explain how I arrive at this conclusion. In this case, as in many others, having identified the rival interpretations, the first stage is to decide what seems to be the natural meaning of the relevant words, and, the next stage is to consider whether there was anything in the document (ie the Grant), or in the surrounding circumstances when it was made, which called that interpretation into question. At any rate in this case, the first stage is almost a matter of impression and the second stage involves identifying a negative, so there is a limited amount that can be said by way of justification for this conclusion.

65.  It is clear that a godown is, in summary terms, a building in which storage takes place, and therefore a restriction to use “for … godown purposes” would naturally seem to imply, again in summary terms, use for a building in which storage takes place.

66.  However, it is said on behalf of PBIC that, if it had been intended to restrict the use of the land in that way, clause 3(a) would have been differently worded, restricting the use to, for instance, “as a godown or godowns”. When choosing between two rival interpretations of a document, a court should beware of placing weight on an argument that, if the parties had intended a particular meaning, they would have expressed themselves differently. Occasionally, such an argument has some force, but normally it is unhelpful, as both parties can normally play that game, not surprisingly as interpretation issues usually arise where the document concerned is not entirely clear. In this case, for example, the Director can say that, if the parties had meant what PBIC contends, they would have limited the use to “storage” rather than to “godown purposes”. Indeed, that is a point which has a little more force given the reference to storage later in clause 3(a).

67.  In any event, I consider that the natural meaning of “for … godown purposes” in the Grant is “for the purposes of a godown (or godowns)”. If it needs some sort of justification, the use of the word “purposes” is readily explicable both syntactically and substantively. Syntactically explicable, because of the inclusion of the words “industrial or”. Substantively explicable, because use “for … godown purposes” is potentially wider than use “as a godown or godowns”, as the former expression would clearly include uses which are ancillary to the godown use, whereas the latter might well not do so.

68.  It is also said on behalf of PBIC that “godown purposes” more naturally means “purposes (or uses) which would be carried out in a godown”. Even if that is right (and I am unconvinced that it is), it does not seem to me to point to a different conclusion in this case. The purposes (or uses) to which a godown is put is storage of goods (on a temporary basis) in a building. It would normally be surplusage to add “in a building” (as it was for Liu J in Cavendish), but it is appropriate to do so in this case. Analogies can be dangerous, but a user restriction to “house purposes” or “hospital purposes” would, subject to contextual considerations, involve a building: they are not the same as, respectively, “for residential purposes” (which could involve a tent or a caravan) or “for medical purposes” (which could involve an outdoor clinic).

69.  I see nothing in the documentary or factual context of the Grant to call this conclusion into question. As already mentioned, the reference to “storage” shortly after the words in issue could be said to support the conclusion on the ground that where the parties mean to refer to storage generally they say so, but that is a very slight point in my view. The reference in clause 3(b) to constructing a warehouse, which it was suggested also supported the conclusion, seems to me to be neutral.

70.  As to the surrounding facts, it was suggested by Mr Chang SC for PBIC that the sheer size of the Land called the conclusion into question. I am unimpressed with that point. It ignores the other uses permitted by the Grant, and in any event, there was no limitation on the number or size of godowns which could be erected on the Land, or to the ancillary uses to which neighbouring parts of the Land could be put.

71.  Having reached the same conclusion, the Court of Appeal pointed out that there had been no argument or ruling on the issue whether mid-stream storage use might nonetheless have been permitted under the Grant on the ground that it constituted use for “industrial … purposes”. As I understand it, the parties are agreed that the issue of whether the Land can be used for midstream storage use on the ground that it would be “for industrial … purposes” should be remitted to the Lands Tribunal together with certain other issues as a result of the unchallenged decisions of the Court of Appeal.

The second issue: the assumption about the CT Scheme in the after valuation

The approach of the Lands Tribunal and the Court of Appeal

72.  The Lands Tribunal decided that they should carry out the after valuation on the assumption that the market as at 5 May 1995 would assume that the CT Scheme would effectively be certain to proceed. This was inconsistent with their conclusion that, as at that date, the market would, as a matter of fact, have regarded the prospects of the Scheme proceeding as 50-50.

73.  The reasoning of the Tribunal is rather exiguous in this connection, but I think that it is tolerably clear from para 464 (where they said that “[w]ith the marine access being extinguished, one would expect that the Container Terminal Scheme would proceed as planned”) to para 477 (where they referred to it being “more certain that the Government has decided to proceed with the Container Terminal Scheme” because “marine access has been extinguished”) coupled with the after valuation itself. The reasoning effectively proceeded along the following lines:

 (a) For the purpose of the after valuation, it must be assumed that the notional open market sale as at 5 May 1995 was agreed on the basis that it was certain that the Reclamation would proceed, as it was the Reclamation that would deprive the Land of the Marine Rights;

 (b) If it is to be assumed that the Reclamation is bound to go ahead, then the Tribunal considered that the view in the market would be that the whole CT Scheme would be bound to go ahead;

 (c) If the market would have assumed that the CT Scheme was bound to go ahead then it would follow that there should be no discount for the uncertainty of being able to use the Land for industrial purposes;

 (d) There would, however, still have to be a discount for delay, but the Tribunal considered that it should be somewhat less than on the facts assumed for the before valuation, because of the confidence that the CT Scheme would proceed.

74.  If step (a) in that line of reasoning is correct, it seems to me to be hard to identify any convincing reason for quarrelling with the other three steps. As to step (b) the market assessment as to the likelihood of the CT Scheme proceeding in particular assumed circumstances, including where it was certain that the Reclamation would proceed, was a matter for the Tribunal. Having said that, it is not hard to understand why they concluded, on this assumption, that the market would have thought it very likely that the CT Scheme would proceed. Reclaiming well over 1,000 hectares of land would be a very expensive operation, and, if one is proceeding on the basis that the Reclamation was bound to happen, it would be unsurprising if people assumed that the subsequent works needed to complete the CT Scheme were therefore bound to happen as well. It is fair to add that the absence of any discount, so that it was assumed to be 100% likely, is rather more surprising. However, it was the Tribunal’s assessment of the market view, and it does not seem to me to be a view which could be characterised as unreasonable.

75.  Step (c) really follows as a matter of logic from step (b). And my observations on step (b) apply equally to step (d).

76.  It is step (a) in the Lands Tribunal’s reasoning which was successfully challenged by PBIC in the Court of Appeal. In that connection, the Court of Appeal said that the references to “before valuation” and “after valuation” “can be misleading”, because “the only difference which Lord Hoffmann postulated is that one valuation must proceed on the basis that the land has marine rights whilst the other valuation shall proceed on the basis that the land does not have any marine rights”. The Court of Appeal considered that it was illegitimate for the after valuation to proceed on any assumption different from the before valuation save in relation to the subsistence of the Marine Rights.

The three possible approaches

77.  Assessment of the before value and the after value each involved considering a hypothetical transaction, namely the sale of the Land on the open market as at 5 May 1995; in the case of the before value with the Marine Rights, and in the case of the after value without the Marine Rights. The question to be addressed is: when carrying out the after valuation, what, if any, further assumption is to be made?

78.  It seems to me that in the judgment of Lord Hoffmann in the CFA’s 2010 decision, one can find at least one sentence which appears to support each of three different possible (and mutually exclusive) answers to that question, namely:

 (a) It must be assumed that as at 5 May 1995 the Reclamation has actually been carried out. Support for that approach is to be found in para 42 (quoted in para 30 above) where Lord Hoffmann said that “the Reclamation is treated as having taken place on the date of publication of the Authorisation and the compensation” was to be assessed accordingly.

 (b) It must be assumed that as at 5 May 1995, the Reclamation has not been carried out, but that it is certain that it would be. Support for that approach is to be found in para 41 (quoted in para 29 above) where Lord Hoffmann said that “for the purposes of assessing the compensation, it must be assumed that on the date of authorisation it was certain that the reclamation would take place”.

 (c) No artificial assumption should be made as to whether, as at 5 May 1995, the Reclamation has been or will be carried out: one merely assumes that the Marine Rights have been lost. Support for that approach is in para 46 (quoted in para 34 above) where Lord Hoffmann said that the after valuation (and the before valuation) “must take into account all the information which was public knowledge at the time and (apart from the assumptions about marine rights) not be based on any artificial assumptions”.

79.  Approach (a) was rejected by the Tribunal and by the Court of Appeal, and both the Court of Appeal and this Court refused the Director permission to appeal on the point. Accordingly, the choice is between approach (b), which was adopted by the Lands Tribunal and is supported by the Director, and approach (c) which was preferred by the Court of Appeal and is supported by PBIC.  

The case in favour of the Court of Appeal’s approach

80.  There are undoubtedly strong arguments in favour of approach (c), the Court of Appeal’s view, as against approach (b), the Lands Tribunal’s view.

81.  First, approach (c) involves making no assumption which departs from reality, because it requires one to value on the basis that the Land, in terms of its physical state and its prospects, is as it was on the date of valuation without the Marine Rights - ie taking the actual market assessment (which the Lands Tribunal considered to have been 50-50) of the prospect of the Reclamation and the rest of the CT Scheme proceeding as it was as at May 1995. By contrast, approach (b) involves a departure from reality, in that it requires one to make the artificial assumption that, as at May 1995, the Reclamation will definitely proceed.

82.  Where a valuation is made on the basis of a notional sale in the open market then, as Lewison LJ put it in Harbinger Capital Partners v Caldwell (As the Independent Valuer of Northern Rock Plc) [2013] EWCA Civ 492, para 23, the hypothetical transaction “is only a mechanism for enabling one to arrive at a value of particular property for a particular purpose”. He added that “it does not entitle the valuer to depart from the real world further than the hypothesis compels. The various hypotheses must be taken no further than their terms make strictly necessary”.  A little later, he added this: “Where the hypothesis inevitably entails a particular consequence, the valuer must take that consequence into account” but “where the alleged consequence is not inevitable, but merely possible (or even probable), then the consequence cannot be assumed to have happened”. 

83.  It can fairly be said in this case that there is no logical reason why the assumption that the Marine Rights have been lost should require a person valuing on that assumption to search for or identify a reason why they have been lost, namely the Reclamation. Further, there is no practical reason requiring such a valuer to do so. Thus, it has not been suggested that it is impossible, or even more difficult, to carry out the after valuation without an explanation as to how or why the Marine Rights were lost; one simply values on the basis that they have been lost. Indeed, it is easier to value on that basis, as it involves making the assumptions as to delay and likelihood which had already been assessed and made in relation to the before value.

84.  Secondly, approach (b) involves effectively crediting the Government with the increase in value which the CT Scheme would confer on the Land through the provision of road access to and from the public highway. Yet, at least in England, the law has fairly authoritatively been stated to be that “[i]n the absence of any special provisions it appears that betterment arising from exercise of statutory powers cannot be set off against compensation for injurious affection” to quote from the current Encyclopedia on Compulsory Purchase, paras B-0061 and B-0482, citing Senior v The Metropolitan Railway Co (1863) 2 H&C 258 and Eagle v The Charing Cross Railway Company [1867] LR 2 CP 638.

85.  Thirdly, there is obvious force in the point that it could be unfair in practice on PBIC if its compensation for the loss of the Marine Rights enjoyed by the Land is reduced (or even extinguished) because of a potential improvement in the value of the Land attributable to a development which may not happen, particularly as it is largely up to the Government whether the development proceeds, and PBIC has no control and little, if any, influence over that decision. That point can be said to have particular force in this case, given that the development, the CT Scheme, did not in fact proceed.

The case in favour of the Lands Tribunal’s approach

86.  The presumption of reality, discussed in paras 81-83 above, undoubtedly should be borne strongly in mind by anyone undertaking a market valuation exercise - whatever its purpose, as Lewison LJ said in Harbinger at para 22. However, the ultimate question is what assumptions are required by the particular ordinance, contract or other document pursuant to which the valuation is being carried out, as Mummery LJ (with whom Beatson LJ agreed) effectively said in Harbinger at para 122. Indeed, that point is well illustrated by the fact that the majority of the Court of Appeal, differing from Lewison LJ, considered that the presumption did not apply in that very case.

87.  Turning to the present case, it seems to me that there is obvious force in the argument that, all other things being equal, if a development scheme on neighbouring property has both beneficial and detrimental effects on a person’s land, it would be wrong for that person to be able to claim compensation for the diminution in value attributable to the detrimental effect without having to take into account the increase in value attributable to the beneficial effect. It is worth bearing in mind in this context that as Lord Nicholls of Birkenhead said in Waters v Welsh Development Agency [2004] 1 WLR 1304, para 63, in relation to the application of the Pointe Gourde principle (namely that “compensation for compulsory acquisition of land cannot include an increase in value which is entirely due to the scheme underlying that acquisition” – per Lord McDermott in Pointe Gourde Quarrying & Transport Co Ltd v Sub-Intendent of Crown Lands (Trinidad) [1947] AC 565, 572) when assessing compensation for compulsory purchase, one should aim to “achieve … a fair and reasonable result”, and that plainly applies equally when assessing compensation for injurious affection.

88.  If one is carrying out a valuation assuming (as one must in the present case for the purposes of the after valuation) that any marine rights have been lost, it is, at the very least, not unreasonable, to consider why they have been lost, and then to take into account any valuation implications which naturally flow from that reason. It is true that in many valuations where one has to make an assumption that a right exists or does not exist, it would be illegitimate for the valuer to inquire why the right does or does not exist. However, while rules such as the presumption of reality have an important part to play, it seems to me that the courts have to do their best to find a valuation hypothesis which is consistent with the wording and aim of the legislation, is workable in practice, and which produces a fair and reasonable result.

89.  Given the paucity of statutory guidance, that is not an easy task in this case, as a reading of Lord Hoffmann’s judgment demonstrates. However, given that relevant marine rights are to be treated as irretrievably lost, de facto as well as de jure, on publication of the authorisation, it can be said to be logically consistent to make the assumption that the cause of the loss, namely the reclamation, will undoubtedly proceed. Further, the assumption is a relatively mild departure from reality - especially compared with that involved in approach (a), namely that the Reclamation has actually been completed as at May 1995 which does seem to me to be too extreme a departure from reality to be acceptable. A variant of the reality principle which seems to me to be applicable is that, if there is a departure from reality, it should be kept to a reasonable minimum. In this case, if a departure is appropriate, it is unnecessary to go further than assuming that the Reclamation will occur: it is unnecessary to go further and assume either that the Reclamation has occurred or that the remainder of the CT Scheme will occur (although that would not prevent the Lands Tribunal concluding as a matter of judgment that the market view would be that the remainder of the CT Scheme was bound to go ahead).

90.  As to the second point, which is discussed in para 84 above, assuming that it is indeed the law of England that a claim for injurious affection cannot be subjected to a set-off attributable to betterment, there is in my view no reason to apply that principle to a claim for compensation under section 12. As Lord Hoffmann said in para 36 of the CFA’s 2010 decision, the fact that the section refers to the English law concept of injurious affection “does not mandate the same method of assessing the compensation” as in English law. The very fact that section 12 provides that the loss is suffered and the compensation is assessed as soon as the reclamation is authorised suggests that a different approach to compensation may very well be required.

91.  In any event, I am unconvinced that the view expressed in the Encyclopedia is reliable: the two cases it cites do not, in my view, provide much if any real support for the general proposition it expresses. Of the four Judges who decided Senior, Pollock CB and Channell B expressed no view on the point (if anything they seem to have assumed that such a set-off might have been permitted), Bramwell B thought a set-off of the sort claimed in that case could not be claimed but made it clear that it was an obiter view, and Wilde J “doubt[ed] whether, in a case of compensation for injury to land, a Company can claim a set-off by reason of the land being subsequently benefited”. As for Eagle, I can find nothing in the judgments of Bovill CJ, Keating J or Montague Smith J which really touches on the point of set-off. The Court in that case upheld an umpire’s award of £656 for loss of light which rendered “the plaintiff’s premises less convenient and suitable for the requirements of his trade to the extent of £656” (per Keating J), despite the Umpire “having inserted an impertinent averment in his award” (per Montague Smith J). The issue in the case ultimately concerned the interpretation of a confused award.

92.  I turn to the third argument, which is discussed in para 85 above, namely that approach (b) would operate unfairly on a landowner in a case where the reclamation concerned never eventuates. It is equally true that approach (c) would operate unfairly on the Government in a case where the reclamation concerned does eventuate. At first sight, one might think that, as it is the Government which decides whether to publish a formal authorisation and it is the Government which normally decides whether the reclamation proceeds, the injustice involved in approach (c) is significantly less than that involved in approach (b).

93.  Closer consideration, however, seems to me point the other way. It appears unlikely that the Government would publish a formal authorisation for a particular reclamation if it did not intend pretty confidently to proceed with it. In the first place, it would mean that all the effort in getting to the point of publishing an authorization of a scheme would be wasted. Further, it appears to me that would be commercially disruptive, and administratively misleading to publish such an authorisation in the absence of a firm intention to proceed with the development so authorised, and therefore it is unlikely to happen.  Also, the very fact that the issue of an authorisation triggers a liability to pay compensation under section 12 is a strong disincentive to the Government not to publish an authorisation for reclamation with which it does not firmly intend to proceed. It is true that the facts of this case can be said to belie these points, but the circumstances in Hong Kong at the time of the 1995 Authorisation, which was during the period leading up to the hand-over in 1997, were exceptional.

94.  Quite apart from these considerations, the fact is that, so long as the reclamation does not go ahead, the landowner continues in practice to enjoy the marine rights; that is indeed what happened in this case (see para 14 above). Accordingly, so long as an authorised reclamation does not go ahead, a landowner will, it appears, continue to enjoy any marine rights as a matter of fact even though he may have been deprived of them as a matter of law. That is a practical pointer, in my view, in favour of approach (b) over approach (c).

95.  It also appears to me that the Lands Tribunal’s approach is more consistent with the guidance given in the CFA’s 2010 decision. It is fair to say that, as I have explained, different sentences in Lord Hoffmann’s judgment can be said to support different conclusions. However, Lord Hoffmann ended his judgment by saying that the compensation payable to PBIC should be “the difference between (a) the price which [the Land] would have fetched … on 5 May 1995 on the assumption that it enjoyed access to the sea as it had done up to that date and (b) the price which it would have fetched … on the assumption that access to the sea had been lawfully interrupted by the completion of the proposed reclamation” (emphasis supplied). This direction, like the statements in para 41 of Lord Hoffmann’s judgment, and (once one rejects approach (a) as one must) para 42, all support approach (b). While what Lord Hoffmann said in para 46 can be said to support approach (c), it is fair to say that it is not inconsistent with approach (b): it depends what is meant by “assumptions about marine rights”.

96.  It would be wrong to make too much of the 2010 judgment of this Court in the present context, because the point currently in issue was not before the Court. The two relevant points were simply those described in paras 28 and 32 above. The passages cited in paras 78(a) and (b) above are to be found in passages dealing not with a different valuation assumption from that currently in issue, namely what assumption is to be made as to the existence of the Marine Rights as at the valuation date, or, to be more precise, whether events after May 1995 are relevant to the valuation exercise. Further, because the judgment concerned preliminary issues, it could be dangerous to ascribe to statements in it wider implications than they were intended to have at the time. Lord Hoffmann himself presciently made that point at paras 9-13 and 30 of his judgment, which include a reference to “the danger inherent in the trial of preliminary issues”, the statement that it is “seldom if ever helpful to approach issues of law as an academic exercise”, and the warning that “an understanding of the practical consequences of applying a principle can sometimes demonstrate that it cannot possibly be right”.

97.  Nonetheless, in my view, it is relevant when deciding which of the various approaches to valuation to adopt that it would be inconsistent with the general thrust of a reasoned decision of this Court in a case between the same parties and relating to the same general issue, if we were to favour approach (c) over approach (b).

Conclusion on the second issue

98.  In these circumstances, while I cannot pretend to have found the point at all easy, I am of the view that the approach adopted by the Lands Tribunal to the after valuation, ie what I have called approach (b) as explained in para 78 above, was correct.

Disposition

99.  For the above reasons, I am of the view that PBIC’s appeal on the first issue, relating to the proper interpretation of the user covenant, should be dismissed, and that the Director’s appeal on the second issue, relating to the assumption to be made in the after valuation, should be allowed.

100.  I would also direct by way of order nisi that PBIC pay to the Director the costs of both appeals with liberty to the parties, if so advised, to lodge written submissions as to costs within 14 days of the date when this judgment is handed down, and in the absence of such submissions, that the order nisi stand as an order absolute without further direction.

101.  We were told that the parties have already agreed that the case should in any event be remitted to the Lands Tribunal as a result of the decision of the Court of Appeal. Accordingly, it would be inappropriate for me to say any more about the future conduct of these proceedings. However, the proceedings have been going on for an unconscionable period and have already taken up a substantial amount of court time. The parties and their legal advisers should do their best to achieve a final outcome as quickly and cheaply as is consistent with justice.

Mr Justice Ribeiro PJ:

102.  The Court unanimously dismisses PBIC’s appeal, allows the Director’s appeal and makes the orders as to costs set out in paragraph 100 of the judgment of Lord Neuberger of Abbotsbury NPJ.