THE HONG KONG ELECTRIC CO LTD v. COMMISSIONER OF RATING AND VALUATION

CACV 27/2010

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF APPEAL

CIVIL APPEAL NO. 27 OF 2010

(ON APPEAL FROM LDGA NO. 224 OF 2004 AND
LDRA NO. 358 OF 2004 (CONSOLIDATED))

________________________

LDGA 224/2004

IN THE LANDS TRIBUNAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

GOVERNMENT RENT APPEAL NO. 224 OF 2004

________________________

BETWEEN

________________________

LDRA 358/2004

IN THE LANDS TRIBUNAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

RATING APPEAL NO. 358 OF 2004

BETWEEN

________________________

________________________

________________________

J U D G M E N T

________________________

Hon Rogers VP:

1.  I agree with the Judgment of Le Pichon JA and the orders she proposes.

Hon Le Pichon JA:

I.   Introduction

2.  This is an appeal by the Commissioner of Rating and Valuation (“the Commissioner”) from an order of the Lands Tribunal (“the tribunal”) dated 30 November 2009 allowing the appeals of the Hong Kong Electric Co Ltd (“HEC”) against the rateable value ascribed to its tenement for the year 2004/2005.  The original rateable value assessed by the Commissioner had been $6,294,000,000.  The tribunal had reduced the rateable value to $3,945,000,000 and made consequential orders.  At the conclusion of the appeal, judgment was reserved which we now give.

3.  I preface this appeal with the apposite remarks of Godfrey JA in China Light & Power Co Ltd v Commissioner of Rating and Valuation [1995] 2 HKC 42 at 43G-I:

“This court is concerned, in these appeals, with questions of rating. The world of rating appears, to one unfamiliar with the arcane, to be cloud-cuckoo land, a world of virtual unreality from which real cuckoos are excluded (although it seems that permission to land will be granted to a cuckoo flying in from the real world if it can demonstrate that its presence in cloud-cuckoo land is essential, not merely accidental: see Dawkins (V.O.) v. Ash Brothers & Heaton Ltd [1969] AC 366 per Lord Pearce at p. 382B-C.) A valuation for rating purposes must be based on hypothetical, not real, facts. Nevertheless, the authorities establish that such a valuation is itself to be treated as a matter of fact, impeachable only if the person responsible for the valuation has fallen into some error of law in arriving at it. In the absence of error of law, this court is not entitled to interfere with the valuation.”

4.  That passage underlines a very important fact: for rating purposes, there is the rating world, entirely hypothetical and imaginary, which Willmer LJ in Humber, Ltd v Jones (Valuation Officer) (1960) 6 RRC 161, 171 described as “a world of make-believe”, and there is the real world.  It is essential to distinguish the two and not lose sight of the ‘world’ that is relevant for the purposes of any particular aspect of the topic.

II.    Background

A.   Key features of the Rating Ordinance

5.  The “tenement” which these appeals concern comprises the land, buildings and structures occupied and used by HEC for the generation, transmission and distribution of electricity to its end users on Hong Kong Island, Ap Lei Chau and Lamma Island.  “Tenement” is defined in section 2 of the Rating Ordinance, Cap. 116 as

“any land (including land covered with water) or any building, structure, or part thereof which is held or occupied as a distinct or separate tenancy or holding or under any licence.”

The use of the words “held or occupied” presaged dual liability imposed by section 21(1) on both owner and occupier for the payment rates.

6.  For the purposes of ascertaining the rateable value, “tenement” has to be read together with sections 8 and 8A which deal with the treatment of machinery and plant.  In short, machinery used as “adjuncts” to the tenement is to be regarded as part of the tenement but no account is to be taken of the value of machinery required for its manufacturing operations or trade purposes.  “Cables, ducts, pipelines... oil tanks, settings and support for plant or machinery” fall within the definition of “plant” in section 8A(3) and are to be treated as part of the tenement.  As recorded in the judgment, HEC transmits electricity by means of 400 km of cables laid underground, undersea, in tunnels and overhead and distributes electricity via 5,000 km of cables.  The effect of section 8A is that such cables form part of the tenement.

7.  Further, any land, building or structure occupied by a person by means of any plant is deemed for rating purposes to be a separate tenement.  While HEC owns much of the tenement, it does not own the land occupied by cables, pylons and over 3,000 substations which are customer-owned.  Nevertheless, in the rating world, such land also forms part of the tenement because of the provisions of section 8A.

8.  In the present case, the Commissioner has exercised her power under section 10 of the Rating Ordinance to treat a property which otherwise might comprise several tenements as one single tenement.  This in cumulo assessment is not controversial.

9.  Turning to the rateable value of a tenement, it is to be noted that it governs both the amount of rates payable and Government rent.  The rates payable is a percentage of the tenement’s rateable value, currently set at 5%, while Government rent is calculated at 3% of the rateable value of the tenement.  Accordingly, the rateable value of the tenement and, hence, how it is arrived at, is of crucial importance to the parties.

10.  Section 7 of the Rating Ordinance provides that rateable value is the rent which the tenement might reasonably be expected to let from year to year on certain statutory assumptions.  Those are specified in subsection (2) of section 7.  While, in the real world, HEC does not rent the tenement from which it conducts its business of generating, transmitting and distributing electricity and, indeed, owns much of the tenement, in the rating world, one is not only required to proceed on the basis of there being a hypothetical landlord and a hypothetical tenant and the bargain on rent that they would strike, but also that part of HEC’s assets should be treated as belonging to the hypothetical landlord (thus forming part of the tenement and rateable) and the remaining part as a hypothetical tenant’s assets which are non-rateable.

11.  As regards the notional division of HEC’s assets between the hypothetical landlord and the hypothetical tenant, happily, the parties were able to reach agreement in the court below: of the total assets used by HEC to generate and supply electricity, approximately 52% by value is to be treated as belonging to the hypothetical landlord and rateable, and the remaining 48% by value as belonging to the hypothetical tenant and non-rateable.

12.  Further, the rateable value has to be assessed by reference to “the relevant date” which is the 1^st October prior to the coming into force of the valuation list on the 1^st April of the following year upon certain statutory assumptions set out in section 7A(2).  For present purposes, the assessment must be made on the basis that the tenement was in the same state as at the 1 April 2004 and any relevant factors affecting the mode or character of occupation were those subsisting as at 1 April 2004.

B.   The relevant rating assumptions

13.  Before turning to the rating assumptions, it is as well to bear in mind what rating is about.  As Lord Pearce observed in Dawkins (Valuation Officer) v Ash Brothers and Heaton Ltd [1969] 2 AC 366, 381H:

“Rating seeks a standard by which every hereditament in this country can be measured in relation to every other hereditament. It is not seeking to establish the true value of any particular hereditament, but rather its value in comparison with the respective values of the rest.”

Further, the hypothesis of the imaginary tenant is intended to assist and not hinder that process: Townley Mill Company (1919) Ltd v Oldham Assessment Committee [1936] 1 KB 585, 643.

14.  Rating law requires certain assumptions to be made.  Only those that are of particular relevance to these appeals will be mentioned.  An appropriate starting point is Lord Hailsham’s speech in The Railway Assessment Authority v The Southern Railway Company [1936] AC 266 which states some of the relevant rating assumptions and illustrates their practical application.  The case concerned a public utility undertaking.  Lord Hailsham observed (at 285) that:

“... when we come to a railway undertaking in which vast sums have been invested and in which the share of the tenant’s capital amounts to many millions it is apparent that the effort to visualize an actual tenant involves an almost impossible strain. In my opinion it would be a mistake to allow the percentage to be influenced by the largeness of the sum which the tenant is supposed to supply for capital, or by the difficulty of finding a tenant with such extensive means, or by the difficulty the tenant might find either in realizing his rolling stock and other chattels or in reinvesting so large an amount of capital on the expiry of his tenancy. Since the landlord is to be contemplated as a possible tenant, none of these considerations might be allowed to come in.”

Then (at 286-287) he added:

“I think it must be assumed that the hypothetical tenant would be able to realize his stock at the end of the year as easily as he could acquire it at the beginning of the year at the market value prevailing at the respective dates. There is a certain risk of fall in market value, just as there is the possibility of increase in market value, but he must not be given a higher percentage on the basis that he may be unable to sell the stock at all.

Then again, it was said that the Court below had proceeded on the hypothesis that the hypothetical tenant would have to assemble together from all over the country an adequate staff to manage and run the railway undertaking and that this was wrong in law. The contention was based upon a passage in the judgment of Sir Francis Taylor, where, after citing Lord Herschell’s words in the Erith case, he goes on to say “vacant premises seem to me to be inconsistent with the presence of a highly trained staff and management.” My Lords, it is this expression which has given me most trouble in the present case. If it were to be construed as bearing the meaning which the appellants allege, if your Lordships were to assume that the Court based their allowance for tenant’s capital on the view that the tenant must be treated as somebody who came to take over this great railway system with no means of running it, with no directors or managers or staff, and that he had to be compensated for the risk which he ran of being unable to assemble such an organization in order to carry on his undertaking, then indeed, I think the appellants would be right in saying that the calculation had proceeded upon a wrong basis. It must be assumed that there is in existence the present staff and management of the railway and that the tenant can reasonably expect to avail himself of their services…”

So, the largeness of the sum to be invested by the hypothetical tenant does not affect his return on capital.  It is to be assumed that he can acquire the necessary plant, machinery and staff easily at the commencement of the tenancy and dispose of them as easily on termination.  Further, the landlord himself is to be contemplated as a possible tenant.

15.  Other relevant assumptions should also be mentioned:

(1)    It has to be assumed that the hypothetical landlord has already invested in the tenement and the hypothetical tenant has not when they negotiate for a tenancy (Re Railways (Valuation for Rating) Act 1930, And re Appeals by The Southern Railway Company, (Railway and Canal Commission Court) (1935) 33 LGR 101, 118).

(2)   The tenement is assumed to be “vacant and to let”, vacant in the physical sense and in the sense that the existing business has ended and any process machinery has been removed: Fir Mill, Ltd v Royton Urban District Council and Jones (Valuation Officer) (1960) 7 RRC 171, 185.  As HEC submitted in its opening below, this principle ensures that all possible tenants are considered on an equal footing and the actual occupier is not assumed to have an advantage in not having to incur the costs of moving in.  This assumption is also consistent with section 8(b) of the Rating Ordinance which requires that the value of machinery in or on the tenement for the purpose of manufacturing operations or trade processes be disregarded.

(3)    The tenancy is one from year to year, with a reasonable prospect that it would continue for an indefinite duration (Dawkins (Valuation Officer) v Ash Brothers and Heaton Ltd [1969] 2 AC 366, 387G-H) or, with a reasonable prospect that the tenancy would in fact continue for ‘x’ years: China Light and Power Company Ltd v Commissioner for Rating & Valuation [1995] 2 HKC 42, 45G-H (“the CLP case”).

(4)   The hypothetical tenant and the hypothetical landlord will be reasonable people and will act reasonably: see the Humber case (at 172).

C.   Valuation methodology

16.  It was common ground at the hearing below that the receipts and expenditure method is the appropriate method of valuation to arrive at a rent of the tenement. Accordingly, for present purposes, alternative methods of valuation are disregarded.

17.  The receipts and expenditure method of valuation has been in use for decades and used to be known as the ‘profits basis’ of valuation.  References in the older cases to the ‘profits basis’ or ‘profits method’ of valuation should be so understood.  Thus, inre Appeals of the Southern Railway Company when Mackinnon J stated (at 118) that “the profits basis” is best calculated “to secure a fair and just division of the net receipts between landlord and tenant”, he was referring to what is now termed the receipts and expenditure method of valuation which he proceeded to describe:

“The position of the landlord is entirely dissimilar to the position of the tenant; the landlord has invested his capital, but the tenant has not, and the tenant will only offer such a rent as he can afford to pay after satisfying himself that he has a prospect of getting a reasonable return upon his money, if he embarks it upon the somewhat hazardous business of operating a railway. The rent which has to be estimated is the highest rent a prospective tenant could reasonably be expected to offer, and it follows from this that, having ascertained the amount of capital which a tenant would require to invest in the business, and the lowest return upon that capital which he is entitled to expect, the balance of the net receipts, after satisfying the requirements of the tenant, represents such rent, and by this means a fair and just division of the net receipts is achieved.”

18.  The change in terminology came about after the Joint Professional Institutions’ Rating Valuation Forum published its Guidance Note on The Receipts and Expenditure Method of Valuation for Non-Domestic Rating in July 1997 where the inappropriateness of the term ‘profits basis’ to describe the method of valuation was explained.

19.  Thus, the receipts and expenditure method ascertains the rent by reference to the receipts and expenditure of the undertaking carried on at the tenement.  The difference between the gross receipts and the expenditure, referred to as the “divisible balance”, represents the profits of the undertaking.  That has to be divided between the hypothetical landlord and tenant.  The hypothetical tenant’s share represents the sum that provides him with a return on his capital and a reward for his efforts and risks.  The remainder would be the amount that should go to the hypothetical landlord as rent.

20.  The parties are agreed that the gross receipts and expenditure are to be taken from the audited accounts for the preceding year, subject to one caveat which arose from a new argument HEC raised during the hearing below.  As at the relevant date, there were various sites belonging to HEC that were under construction (“assets under construction”).  As I understand it, HEC’s stance is that if it were successful in its argument that assets under construction are non-rateable, some downward adjustment in the gross receipts would be required to reflect their non-rateability.

21.  So, subject to that caveat (which is a discrete issue), arriving at the ‘divisible balance’ is a simple matter given the ‘agreed’ gross receipts and expenditure.  However, it is the exercise of dividing the divisible balance between hypothetical landlord and tenant that has given rise to significant differences between the parties.  The principal issue on these appeals is how the hypothetical tenant’s share ought to be arrived at, which would determine the rent of the hypothetical tenancy.

22.  Before proceeding to consider that issue which lies at the heart of these appeals, it would be convenient to outline the scheme of control which governed HEC’s business and, as such, is an essential feature of the tenement.

D.   The scheme of control

23.  It is not uncommon for the business of public undertakings to be regulated.  That is achieved either by means of legislation (see, for example, the Public Bus Companies Ordinance, Cap. 230) or a scheme of control.  In 1993, HEC and HEC’s holding company entered into a Scheme of Control with the Government for a term of 15 years from 1994 to 2008.  It came into effect upon the expiration of an earlier scheme that had been in place since April 1980.  Its overall objectives can be discerned from recitals (D) and (E):

“ (D) HEC recognises its continuing obligation to contribute to the development of Hong Kong by providing sufficient facilities to meet the present and future demand for electricity and in pursuit of this objective, would construct additional generation, transmission and distribution facilities for sale of electricity to its consumers.

  (E)     The Government recognises that HEC and its shareholders are entitled to earn a return which is reasonable in relation to the risks involved and the capital invested in and retained in the businesses; in return the Government has to be assured that service to all consumers shall at all times be adequate to meet demand, will be efficient and of high quality, and is provided at the lowest possible cost which is reasonable in the light of financial and other considerations and is consistent with Government policy objectives on energy efficiency and conservation in the light of the need to protect the local and global environment and to meet international obligations.”

24.  The scheme of control in place for the rating year 2004/2005 monitored and controlled HEC’s electricity-related business and tariffs through the mechanism of providing a cap on HEC’s return.  It allowed HEC a “permitted return”, being the aggregate of two components:

(a)    13.5% of its average net fixed assets, and

(b)   1.5% of shareholders’ investments made after 31 December 1978 for financing the acquisition of assets.

“Net fixed assets” was defined to mean historic costs less depreciation calculated as provided in the scheme and “average net fixed assets” essentially corresponded to the net book value of the assets.  In broad terms, the maximum profit HEC could achieve under the scheme (the HEC’s permitted return) was capped at 15% of HEC’s average net fixed assets or, which comes to the same thing, 15% of the net book value of those assets.

25.  Since HEC’s revenue stream is dependent on the tariff it charges for electricity, the cap serves to control the tariff.  While HEC has the right to fix the tariff, the permitted return means that the tariff could be fixed with a view to generating receipts that would enable the permitted return to be achieved and not more.  Where the net revenue exceeds the permitted return in any year, the amount of the excess has to be transferred to the development fund HEC is required to set up under the scheme to fund the acquisition of fixed assets.  There is an upper limit for the development fund, any excess having to be returned to consumers in the following year in the form of a one-off rebate or tariff reduction.  For completeness, it should be mentioned that where in any year the net revenue is less than the permitted return, the deficiency is transferred from the development fund (but only up to and not exceeding its balance).  In other words, the deficiency is met from the development fund to enable HEC to achieve its permitted return provided there are sufficient monies in the fund to meet the shortfall.

III.   This appeal

A.   Determination of the tenant’s share of the divisible balance

26.  This is the central issue.

27.  As stated in the Guidance Note (at 5.46), the tenant’s share has to be sufficient to induce the hypothetical tenant to take a tenancy of the property and to provide a proper reward to achieve profit, an allowance for risk and a return upon its capital.  HEC’s case below was that, first, the hypothetical tenant should be rewarded for the risk and enterprise in running the business by having a 25% share of the divisible balance and, secondly, there should be an apportionment to the hypothetical tenant of its share of the balance by reference to the permitted return multiplied by the value of the hypothetical tenant’s assets (“the permitted return approach”).  The Commissioner’s case is that in the circumstances of this case, the hypothetical tenant’s share should be arrived at by applying the weighted average cost of capital to the net book value of the hypothetical tenant’s assets (“the cost of capital approach”).  The weighted average cost of capital is the description of a mathematical formula that measures the minimum return that a company must earn on an existing asset base to satisfy its investors and is used to see if the investment is worthwhile to undertake.

28.  Although the tribunal rejected the contention that 25% of the divisible balance should be set aside for HEC to reward it for the risk and enterprise in running the business, it found in favour of HEC and held (at §359 (1)) that the tenant’s share be assessed on the basis of the permitted return multiplied by the hypothetical tenant’s average net fixed assets.  The tribunal’s order reflected that holding although earlier (at §159), it had expressed its conclusion somewhat differently:

“the tenant’s share under the divisible balance shall be the value of the Permitted Return x HT’s Average Fixed Assets, being a portion of the divisible balance in proportion to the asset split between the HL and HT.”

The meaning of the last clause in the passage quoted is somewhat opaque.  But, for present purposes, it is the correctness of the terms of the order that is material and, for this reason, any discrepancy between §§359 and 159 will be disregarded.

1.    Permitted return as basis for ascertaining the rent

29.  Mr Yu SC who appeared for the Commissioner accepts that the scheme of control is an essential feature of the tenement because it caps the profits of the undertaking, but he submitted that it does not follow that the permitted return should be used to arrive at the rent. Using it to arrive at the tenant’s share (which the tribunal did) amounts to the same thing in that determining the tenant’s share has the automatic effect of also determining the rent because under the receipts and expenditure valuation method what is left in the divisible balance after deducting the hypothetical tenant’s share represents the rent.  Mr Yu submitted that the scheme is not a proper basis for ascertaining the tenant’s share.  The points put forward can be considered conveniently under three main heads.

30.  The first is the general unsuitability of the scheme of control as a measure of the tenant’s share.  As earlier noted, the scheme of control is designed to regulate the profits of HEC, a public undertaking.  Its purpose is not to regulate rent, much less rent that is payable under the hypothetical tenancy.  It reflects a negotiated agreement arrived at between the Government and HEC’s holding company in 1993 to cap the level of profits HEC is permitted to achieve during over a 15-year period commencing in 1994.  Through that mechanism, the tariff chargeable to consumers is regulated.  The exercise required by the Rating Ordinance, on the other hand, is very different.  For one thing, the assessment of the rateable value i.e. the rent is an annual exercise whereas the permitted return is a long term arrangement.

31.  In the real world, economic factors do greatly influence rental negotiations of a tenement used for profitable purposes as was held in British Telecommunications PLC v Central Valuation Officer [1998] RVR 86, 95.  As a matter of common sense, economic conditions do not remain static. Adopting the permitted return approach would mean that, in practice, only economic circumstances prevailing at the time the scheme was negotiated (in 1993) will continue to influence the rent of the tenement years later (2004/2005) but the economic climate prevailing on the relevant date (1 October 2003) will not and cannot be taken into account in the rental negotiations under the hypothetical tenancy for 2004/2005.  By way of general observation, I would say that apart from the logic being difficult to fathom, it makes little sense.  Prima facie, it is far from apparent why the scheme of control is considered appropriate to determine the rent.

32.  Using the scheme to underpin the permitted return approach involves adopting a number of assumptions which, Mr Yu submitted, were erroneous.  When the tribunal (at §§97 and 100) accepted HEC’s submissions that it was reasonable for a hypothetical tenant to start by looking at its likely entitlement in terms of a return under the scheme, it appeared to have assumed that the hypothetical tenant would achieve the very “profit” provided under the scheme, namely, the permitted return.

33.  While, in the real world, it was open to HEC to set the tariff with a view to achieving the permitted return, there is no certainty that it would be achieved.  The permitted return merely represents the maximum amount HEC can expect to earn.  As a matter of construction of the scheme of control, plainly, that is correct.  In cross-examination, Mr McGee, the Group Finance Director of HEC, stated that for the year 2003, the board had decided “not to raise the tariff to earn its permitted return”.  In other words, HEC made a decision not to achieve the permitted return for that year.  In fact, that would appear to be the case for each of the years 2003 to 2006.  §33 of Mr McGee’s witness statement supports the view that the management of HEC decided not to set a tariff that would achieve the permitted return for those years which coincided with a period of economic downturn.

34.  The evidence further establishes that for the years 2003, 2004, 2005 and 2006, the permitted return was not achieved notwithstanding the existence of the development fund required by the scheme.  Although the scheme provides for transfers from the development fund to meet any deficiency in the permitted return, the development fund had a nil balance for the years 2003 to 2006 and so there was nothing available to meet the shortfall in the permitted return for those years: (McGee witness statement §§15, 31(v) and 33)  The tribunal’s reasoning in §106 (2) appears to have overlooked that evidence entirely which shows that in the real world, the permitted return is not a certainty.

35.  In my view, the tribunal should have had regard to that evidence as required by “the principle of reality” established in Hoare (VO) v National Trust [1998] RA 391.  That principle requires the valuer not

“to depart from the real world further than the hypothesis compels”

see per Schiemann LJ at 408 and where Peter Gibson LJ emphasised (at 415):

“... the necessity to adhere to reality subject only to giving full effect to the statutory hypothesis.”

Had the tribunal applied that principle, it would and could not have reached the conclusion that it did at §106.

36.  The permitted return approach is further premised on the hypothetical tenant achieving the permitted return based on the net book value of the hypothetical tenant’s assets.  As Mr Yu has convincingly demonstrated, that is a fallacy.  Reproduced below is Mr Yu’s written summary of his reasoning which speaks for itself:

“ 1. Under the [scheme of control], the operator of the integrated business gets a maximum profit of 15% x Net Book Value of HEC’s assets.

2. When the hypothetical tenant obtains the tenancy, his maximum profit will be governed by the scheme of control, i.e. 15% x Net Book Value of HEC’s assets.

3. The hypothetical tenant has to pay rent. Hence the hypothetical tenant’s maximum annual return is:

15% x Net Book Value of HEC’s assets - rent

4. There is no reason to believe that this is the same as what the Tribunal found, i.e.:

15% x Net Book Value of the hypothetical tenant’s assets

Unless one assumes that:

Rent= 15% x (Net Book Value of HEC’s assets – Net Book Value of the hypothetical tenant’s assets)

5. There is no warrant for this assumption to be made. Certainly nothing in the scheme of control requires it.”

37.  The nub of the submission is that the hypothetical tenant’s cash flow from the business cannot be assumed to be the permitted return multiplied by the hypothetical tenant’s assets because one cannot arrive at the hypothetical tenant’s cash flow or expected return without first knowing the rent that has to be paid.  This is what has been referred to as the circularity problem.  The only way out of the conundrum is to make an assumption as to the rent which is neither what the scheme of control provides nor what the Rating Ordinance requires.  The logic of Mr Yu’s submission is difficult to fault.

38.  Insofar as the tribunal considered the circularity problem to be common to both the permitted return approach and the cost of capital approach and that therefore the circularity problem was not, of itself, a sufficient reason for not adopting the permitted return approach, it appeared to have overlooked another part of its judgment (§§256-269) which specifically rejected HEC’s submission that the circularity problem was an insurmountable problem in the cost of capital approach.  That aspect is addressed in §§53-56 below.  That error plainly must undermine the reason the tribunal gave (at §110(2)) for rejecting the Commissioner’s criticisms of the permitted return approach.  Further, it would appear that the cost of capital approach does not give rise to any circularity problem because it is directed to arriving at a return (net of rent) on the capital deployed and does not require any assumption as to rent to be made as what is left of the divisible balance after subtracting the return so ascertained represents the rent.

39.  Mr Yu’s second ground of challenge is that the permitted return approach involves giving an inflated value to the hypothetical tenant’s assets.  The focus of his criticisms was §142 of the judgment:

“142. In the premises, we accept HEC’s submissions ... For this purpose, we adopt HEC’s examples of calculation set out as follows:

“4. The following illustrative calculation shows how the evidence identified in paragraph 3 above would be applied:
The earnings stream that the tenant's assets bring under the SoC (ANFA of HT’s assets x PR)	HK$22404.1m x 17% = HK$3,808.7m
Present Value of that earnings stream (using WACC as discount rate) (income stream ¸ WACC)	HK$3,808.7m ¸ 8.31% = 45,832.7m
The amount that the HT would expect to receive to give him a return at WACC (WACC x capital value)	HK$45,832.7m x 8.31% = HK$3,808.7m ””

Pausing here, it should be noted that the passage quoted in §142 is taken from HEC’s note produced on Day 19 of the trial in answer to questions posed by the tribunal in relation to the evidence of Mr Jones, HEC’s expert, which only emerged in cross-examination.  The internal reference to “paragraph 3” is a reference to §3 of the note setting out Mr Jones’ view on the “economic value of the asset” which is reflected in §141 of the judgment:

“141. ... Mr Jones’ evidence should be understood in the context of his own views on what the economic value of the business’ assets is. His evidence on this respect is that the earning stream that the assets create for the enterprise represents the economic value of those assets.”

40.  Mr Jones’ view on the “economic value of the asset” first appeared in his second report (at §75).  In Mr Jones’ cross-examination, after Mr Yu had read §75 out in full, the following exchange took place:

“ A. That’s correct. Always assuming that in this you are pricing up to the permitted return. You have raised appropriately the caveat --

Q. Assuming that you can always achieve your permitted return.

A. For the record, the caveat that I just said was assuming that there is the ability in the market to price up to the permitted return, respondent’s counsel has properly pointed out that that’s not necessarily always the case.”

(Tr. Day 6, p. 86, ll. 16-24)

Clearly Mr Jones’ view was based on an assumption and should not be taken as unqualified.  Further, as has been shown in §§33-34 above, the assumption that the permitted return can always be achieved is not correct.

41.  I now turn to the 3 rows of calculations as set out in §142 of the judgment.  The first calculates the tenant’s income stream by multiplying the net book value of the tenant’s assets of 22 billion by the permitted return, using an average pre-tax percentage.  The fallacy underlying the formula used has already been explained (see §§36-37 above) and it appears immaterial whether the ‘tenant’ is the hypothetical tenant or the incumbent tenant.  The second calculates the present value of that income stream by using the weighted average cost of capital as the discount rate.  It will be seen that the present value of the incumbent tenant’s assets (reflecting Mr Jones’ view of the “economic value” of those assets) becomes more than twice its net book value.  When one comes to the third row, the hypothetical tenant’s capital investment becomes 45 billion while only achieving a return at the weighted average cost of capital rate.

42.  The analysis underlying those calculations which the tribunal accepted involves the assumption that the incoming (hypothetical) tenant will have to pay the incumbent tenant (i.e. the sitting tenant) the present capital value of his assets.  In short, what those calculations show is that the difference between 45 billion and 22 billion will go to the sitting tenant since only a sitting tenant is in a position to charge a premium over the net book value of its assets.  Mr Yu submitted (in my view correctly) that in rating law, there is no sitting tenant advantage or disadvantage.  This aspect is addressed under the ‘integration of assets’ heading below (§§76-83).  Finally and inexplicably, while the permitted return rate is used in the first row of calculations to illustrate the incumbent/hypothetical tenant’s income stream, the hypothetical tenant’s income stream is calculated differently – using the weighted average cost of capital rate.  Why that should be so is unexplained and is open to the surmise that the calculations were devised so as to present a high present value for the incumbent tenant’s assets.

43.  The table in §142 of the judgment is presented on the basis that it enables the hypothetical sitting tenant to achieve a return of 17% when his cost of capital is only 8.31%.  However, on that basis the hypothetical incoming tenant, if forced to invest capital of $45 billion, would then have to have an income stream of $7.65 billion if it were to achieve a return of 17%.  Followed to its logical conclusion there would be a year on year ever spiralling increasing value in the same assets.

44.  Mr Yu identified yet further problems with the permitted return approach which fall under his third ground.  It was said that no hypothetical landlord would have agreed a rent on the basis of (1) using the net book value (historic costs) of assets (that being prescribed by the scheme of control) as opposed to using the market value, (2) leaving certain assets out of account and (3) disregarding the value of monopoly of place and synergy.

45.  As to (1), it is self-evident and needs no further elaboration.  As to (2), the assets referred to comprise distribution pillars, poles and pylons and the land thereby occupied, “wayleaves” over land occupied by cables and some 3,500 customer-provided substations, all of which are not “assets” for the purposes of the scheme of control.  Yet, it is obvious that they are all of substantial value to HEC and essential to an integrated generation, transmission and distribution system.  At §305 of the judgment, the tribunal accepted the Commissioner’s submission that rateable value should be attributable to those assets to reflect the value and benefit of their occupation by the hypothetical tenant and, further, that whether or not HEC had to pay anything in the real world for the use of that land is irrelevant because in the rating world, it is well-established that private agreements and arrangements are ignored:

“to make the rent payable by the actual tenant the measure of the rateable value of premises ... would be an absolute innovation, in direct conflict with the principles of the law of rating as established for over a century.”

per Lord Atkinson in Popular Assessment Committee v Roberts [1922] 2 AC 93, 113. The tribunal’s finding at §305 is not challenged.

46.  As to (3), the tribunal found (at §181) which, again, is unchallenged, that:

“ (1) [The tenement] supports and enables a business of the generation and supply of an essential commodity, which produces steady profits.

(2) There is significant value in the synergy of the tenement: there is value added by reason of the way it is connected and configured.

(3) It enjoys a monopoly of place: the relevant business, that of generation and supply of electricity to Hong Kong and Lamma Islands, can only be viably carried out on this tenement, by virtue of its location and configuration.”

and, further, held as follows:

“186. Such a monopoly value should be reflected in the rateable value. As Lord Goddard CJ said in Amalgamated Relays Ltd v Burnley Rating Authority [1950] 2 KB 183 at 190:

“… the position at the present moment is that the company can say to any intending tenant, if it were minded to let at a rent, that it had a system which was working; that nobody else had it; that it was unlikely that any competitor would come into the market… The landlord would be in a position to say that if he let to the prospective tenant he would be putting him into a position in which he could make substantial profits and that he, the landlord, wanted a share of them. What share he would demand and what share the tenant would be willing to pay is a question of fact which the tribunal of fact must assess in the best way it can.” ”

47.  Given those findings, in my view, it is both illogical and irrational for the tribunal to adopt an approach that involves leaving out of account matters the tribunal has itself held should be reflected in the rateable value.  In any event, the rating assumption that the hypothetical landlord would act reasonably applies.  It takes little imagination to conclude that a hypothetical landlord, acting reasonably, would neither negotiate nor agree a rent on the basis of leaving assets of significant value out of account.

2.    Weighted average cost of capital as basis for ascertaining the rent

48.  It is not Mr Yu’s submission that the cost of capital approach is always appropriate in valuing a business but that it is the appropriate approach in the present case and that the permitted return approach is wrong in law.

49.  The cost of capital approach was the basis used in the CLP case, albeit by agreement between the parties.  It was considered by the Rating Forum who commented that the approach is only likely to be adopted in the receipts and expenditure method of valuation when considering properties occupied by major undertakings where there are sufficient tenant's assets to warrant that approach: Guidance Note, §5.54.

50.  The cost of capital approach is one of the recognised approaches used in calculating the return on the tenant’s capital.  §5.52 of the Guidance Note reads:

“5.52 Alternative approaches to determine the tenant’s share by means of a return on capital will include considering:

(a) an approach similar to the discount rate used for DCF valuations/appraisals;

(b) the Return On Capital Employed (ROCE) achieved by public companies for any particular industry from published accounts;

(c) the target ROCE for particular companies;

(d) the Weighted Averaged Cost of Capital (WACC).”

By way of contrast, the permitted return approach does not feature as a possible approach.

51.  After the CLP case was decided in the mid-1990s, the Commissioner applied the cost of capital approach in calculating the tenant’s return not only for CLP but for HEC as well.  There was no objection until the rateable value was ascribed for the year 2004/2005, giving rise to these proceedings.  It is pertinent to note that during the period prior to these proceedings, the permitted return yield was less than the cost of capital.

52.  Since the CLP case, the cost of capital approach has been applied in the British Telecommunications case, Dolgarrog Power Station, Interim Decision of the North Wales Valuation Tribunal (29 November 2007) and Southampton Container Terminal, Decision of the Hampshire South Valuation Tribunal (17 October 2008).  Although the tribunal paid little regard to these decisions on the basis that they are under appeal, I would be surprised if that were the case for British Telecommunications, a 1998 decision.

53.  HEC contended below that the cost of capital approach is unworkable.  This was based on the evidence of Mr Jones whose objection to the cost of capital approach appears from the following passage in the judgment:

“140. ... it has always been Mr Jones’ position that the owner occupier’s WACC (even assuming it to be same as the HEC’s WACC) is not a good proxy to the HT’s WACC. This is so because, where the HT would only have part of the necessary assets required for the business, and would have to pay rent, his WACC could not be determined until after knowing what rent it is required to pay, which is unknown. This creates a circular problem. In other words, it is Mr Jones’ view that the level of rent that the HT (or the potential purchaser) is required to pay represents another risk factor which the potential purchaser or HT would have to consider before deciding whether (and at who’s) WACC is sufficient to induce him to embark on the acquisition of the business.”

54.  This theme (a conceptual problem) is picked up later on in the judgment:

“256. HEC submits that there is an insurmountable problem in CRV’s adoption of HEC’s WACC as the HT’s WACC for the purpose of the R&E Method. It is said by Mr Jones that HEC’s WACC simply could not be taken as the HT’s WACC because of the different risk profile and factors, for example, the risk and need to pay rent. At the same time, this is also the circular problem identified by Mr Jones, which has been explained above in H H Judge Au’s judgment, that the HT’s WACC can only be determined after knowing the amount of the rent he has to pay, while under the R&E Method, the rent can only be determined after knowing the HT’s WACC. In other words, the rent is a prime determinant of the HT’s WACC, and not vice versa.”

55.  Shortly put, Mr Jones’ position was that the problem was insurmountable and that it is a virtual impossibility to estimate the hypothetical tenant’s weighted average cost of capital.  The tribunal held:

“266. We find Mr. Jones’ evidence and position difficult to accept. This is because: (1) it is agreed between the parties that the various methodologies for estimating WACC have been well established over the past years and under these methods, certain estimates have to be made out of available information from the business and the market, (2) estimates in the financial market by nature [are] bound to carry some inherent uncertainties, but these per se do not give rise to sufficient ground for not trying to make the best estimates out of available information from the business and the market, (3) if Mr. Jones were right, it would be very difficult, if not impossible to estimate the appropriate WACC for any business for the purpose of carrying out any R & E valuation in rating, and the R&E Method can never be properly used or relied upon, which is contrary to the practice.”

56.  The Commissioner’s position is that in the real world, HEC is an integrated business, being a combination of the hypothetical landlord and the hypothetical tenant.  So, unless there is some aspect of the hypothetical tenant’s business which would render it more risky than HEC or the hypothetical landlord, it would be fair to use HEC’s weighted average cost of capital as a proxy for the hypothetical tenant’s weighted average cost of capital.  The judgment records (at §259) that Mr Jones accepted that proposition.

57.  The Commissioner’s expert, Dr Lam, gave evidence (which the tribunal accepted) to the effect that the hypothetical landlord had greater risks and that HEC’s weighted average cost of capital could be used as a proxy for the hypothetical tenant’s weighted average cost of capital.  The tribunal’s acceptance of that evidence is the subject matter of the first contention raised in the respondent’s notice.

2.1  The respondent’s notice

58.  Mr Roots QC, leading counsel for HEC, criticised that conclusion as encapsulated in §267 of the judgment.  The gravamen of his criticism was that there was no or no sufficient evidence before the tribunal for it to have reached that conclusion.  It was said that the only evidence was that of Dr Lam, the Commissioner's expert.  Mr Roots questioned the quality of Dr Lam’s evidence, suggesting that it was unreliable.  He highlighted the following passage in Dr Lam’s cross-examination to make his point:

“ Q. It appears to be clear that your instructions required you to express opinions about the application of the law and practice of rating valuation to Hong Kong Electric’s property?

A. My work is related to the rating dispute, but I don’t think my task is to do the rating itself.

Q. But reading your report, we find frequent references to the “hypothetical landlord” and the “hypothetical tenant”, don’t we?

A. Yes.

Q. What previous experience do you have in connection with rating valuation?

A. No.

Q. So you don’t hold yourself out as an expert in rating?

A. You are right.

Q. So, what instructions were you given with regard to rating law and practice?

A. I have communications with the RDV people.  They provided me a lot of input about the concept of HT/HL.  They provided in the present case of CLP, involving CLP Power.  Whenever I have some uncertainty, I will call them and ask for further information.”

(Tr. Day 14, p. 103 l. 11- p. 104 l. 7)

59.  Dr Lam is an economist who, inter alia, conducts research on the regulation of public utilities.  There was no challenge to his expertise below and we have not been shown any evidence to the effect that Dr Lam had misunderstood rating concepts, or had confused them, and that this had been put to him in cross-examination.

60.  At the appeal hearing, Mr Roots was accorded the opportunity to go through all the evidence to show that no reasonable tribunal would have accepted Dr Lam’s evidence.  He declined that invitation.  The upshot was to request an appellate court to reverse the tribunal’s acceptance of Dr Lam’s evidence on the ground that it is unreliable, based on nothing more than cursory references to isolated passages from the transcript.  In my view, that approach was hardly realistic.  Nor do I consider it sufficient for Mr Roots to put his case simply on the basis that the tribunal failed to “get to grips” with the question whether it was right to rely on Dr Lam’s evidence at all because it was a matter for the tribunal whose evidence it preferred.  Further, it is not correct to say that the tribunal’s reference to “limited evidence” was only referable to Dr Lam’s evidence: there was evidence from Mr Jones and the tribunal had been provided with voluminous material by the parties.  In my view, the first contention raised in the respondents’ notice has no merit.

61.  The second contention raised in the respondent’s notice relates to the question whether the net book value or the depreciated replacement cost should be used to value the hypothetical tenant’s capital.  The topic is dealt with in §§343 to 350 of the judgment.  Neither party’s expert had done a valuation on that basis.  §346(2) records the Commissioner’s submission which the tribunal accepted:

“(2) Mr. Jones accepted that, in the long run, the summation of these figures over the years would be the same. The only difference is in profile. This is also the view of Dr Lam.”

Mr Roots submitted that §346(2) is inconsistent with the evidence and that “the proposition that net book value and depreciated replacement cost are effectively the same, subject only to inflation, only holds good if inflation tracks the change in the cost of the asset.”

62.  As to the correctness or otherwise of §346(2), Mr Jones’ evidence was as follows:

“ You have also agreed with me that there is a difference between a nominal WACC and a real WACC, if I could use that short form. A real WACC is a WACC calculated on the real economic value basis. It is not nominal.

A. Agreed.

Q. Would you agree that one way of arriving at the tenant’s share would be to use a real WACC multiplied by the depreciated replacement cost of the asset?

A. Insofar as you’re happy to accept depreciated replacement cost as the best available proxy for a market value that doesn’t make the exercise circular.

Q. Right. If you do that, you will not have a problem with double-counting?

A. No, you wouldn’t have a problem with double-counting.

Q. The difference between depreciated replacement cost and Net Book Value is that year after year, there would be an inflation element in the depreciated replacement cost; is that right?

A. Correct.

Q. And would you agree with me that if you use a real WACC to multiply by the depreciated replacement cost, compare that with a nominal WACC multiplied by the Net Book Value, you probably arrive at pretty much the same kind of figure?

A. Well --

Q. You should.

A. In any given year, no.

…

Q. They are different in profile, but in terms of the economic value, if you have to estimate the value of the earnings, would it pretty much come down to the same thing?

A. If you summate them, if you discount them over time, then they automatically come to the same number which is the original investment. I mean, it doesn’t -- it is not a --

Q. So the difference is only the profile? In other words, the money you get every year is different?

A.That’s correct.”

(Tr. Day 6, p. 116, l. 15 – p. 118,l. 10.)

In view of that evidence, I cannot accept that §346(2) is incorrect.

63.  Mr Roots then criticized the tribunal for overlooking Dr Lam’s evidence where he agreed that there would be a mismatch between the net book value x nominal weighted average cost of capital and depreciated replacement cost x real weighted average cost of capital if the price index for the assets and general inflation were different.  That passage was said to be contrary to the Commissioner’s position and to support Mr Roots’ submission.  As was the case with the first contention, Mr Roots did not go into the issue in any meaningful way to enable an appellate court to develop a sufficient grasp of the underlying economic theories and the expert evidence involved.  The court was not referred to evidence showing and analysing the differences between the relevant price indices and inflation and how they affect the rateable value.  In those circumstances, the court simply is not equipped to form a view of the matter, much less reverse the tribunal’s conclusion based on nothing more than a short passage from Dr Lam’s evidence.

64.  Further, I agree with Mr Yu that the conclusion that in a valuation exercise the net book value is a fair approximation of value and an acceptable proxy for the value of the hypothetical tenant’s assets is a conclusion of valuation which is a matter for the tribunal, rather than one involving a question of law.  Accordingly, the second contention also has no merit.

65.  I would add that in the course of arguing the second ground in the respondent’s notice Mr Roots raised a point, based on §§4 and 5 of the supplementary report of Mr Jupp that it would be impossible to fit machinery such as turbines, generators, transformers and the like, that are currently manufactured into the existing power stations.  Quite simply, under the Southern Railway principle, the position must be that the case has to be decided on the basis that the hypothetical tenant could acquire the existing plant and machinery easily, presumably from the existing tenant.

3. Conclusion

66.  For the reasons stated, I have no hesitation in concluding that:

(1)   the tribunal erred in law in accepting the permitted return approach as the proper basis for assessing the tenant’s share of the divisible balance;

(2)   in the circumstances of this case, the cost of capital approach is appropriate.

B.    Integration of assets

67.  HEC’s contention is that while the hypothetical landlord has a unique tenement, there is a unique set of assets (turbines, boilers and other apparatus) which an hypothetical tenant must acquire in order to be able to conduct the business of generating, transmitting, distributing and supplying electricity.  This ‘uniqueness of asset’ argument was deployed in a number of ways.

68.  It was said to give the hypothetical tenant a strong bargaining position which must be taken into account.  Mr Roots referred to the judgment of Pearce LJ in Tomlinson (Valuation Officer) v Plymouth Argyle Football Co Ltd and another (1960) 6 RRC 173, 178-179 as an illustration of the strength of the hypothetical tenant’s bargaining power:

“... it was important for the tribunal to try to decide what in “the higgling of the market” would be the resulting rent of this special hereditament as a result of the probable negotiations between the lessors and the ratepayer tenants.  Each had powerful bargaining arguments.  The landlord needed a tenant for a valuable hereditament which demanded some thousands of pounds to be spent on it manually by way of repairs.  On the other hand, the ratepayers’ existence depended on their taking the hereditament.  I cannot but think that they would have arrived at some reasonable compromise and that compromise would represent the rent.  Nowhere in the case is there any reference to the bargaining power which the ratepayer tenants had against the lessors by virtue of their being the only bidders for the hereditament and, indeed, the case was decided on the basis that they had no such power.  That I think is bound to give an unreal picture of the rent which was likely to result and in my view that defect springs from the fact that the tribunal assumed more than one hypothetical tenant, where in fact it was clear that there would be only one.”

69.  In Tomlinson, the ratepayer tenants were able to establish that there would be no competition because they would be the only bidders.  They were able to rule out everyone else including the owner as a potential hypothetical tenant.  Their strong bargaining power was due to the absence of competition.  HEC does not put its case on the basis that there would be only one hypothetical tenant.  Whether or not interest would be ‘limited’ as HEC sought to suggest is irrelevant: it would not detract from the fact that there will be more than one potential hypothetical tenant.  It must follow that there will be competition and Tomlinson is of no assistance.

70.  I digress to address the issue of competition.  As to potential hypothetical tenants who might be interested in acquiring HEC’s business, Dr Lam’s evidence was to the effect that apart from CLP (the other regional monopolist) which has an obvious interest if horizontal integration were acceptable to the Government, entities such as ExxonMobil, Singapore Power, China Hong Kong Power Development Co Ltd, various Chinese electricity companies and investors, the Macquarrie Group and sovereign wealth funds have the financial capability and commercial interest to acquire HEC’s assets and to rent the tenement.  The attraction appears obvious: the absence of competition for the supply of electricity makes the risk of operation low and the return stable and so creates a favourable operating environment for the hypothetical tenant.  Further, prior experience in the electricity business would not appear to be necessary bearing in mind that when the Cheung Kong Group acquired HEC from Hong Kong Land, it had no prior experience in running a utility business.

71.  Mr Jones, HEC’s expert, accepted that there would be interest in the tenement and it was only a matter of price.  §137 of the judgment records Mr Jones’ evidence on the topic of whether there will be potential candidates who would be willing to take over HEC’s business.  He opined that if the return were to be measured by the weighted average cost of capital, interest from potential investors might be low, but if the return was a bit more (i.e. an additional 1, 2 or 3%) the level of interest would be greater.

72.  The tribunal commented:

“139. ... Mr Jones was addressing the potential sale of the integrated business of HEC as a whole, but not the HT’s undertaking. Mr Yu for the CRV is also premising his submissions on the proposition of the sale of the entire integrated business.

140.This puts a question mark as to whether Mr Jones was considering the scenario of the hypothetical rating world, where rent had to be paid, and thus be factored in, by the potential purchasers in acquiring the HT’s business.  As such, we have doubts as to whether these parts of Mr Jones’ evidence could be treated as clear and sufficient evidence to deal with the issue on whether, if it is only the HT’s business that is on sale, there would be interested potential HT who would be prepared to purchase it with a return expected at WACC level...”

The tribunal appeared to doubt the relevance of the evidence adduced on the issue of competition (in the sense that there would be more than one potential hypothetical tenant) because the questions had been directed at the sale of the integrated business and not at the hypothetical tenant’s undertaking.  That distinction and its practical effect require consideration.

73.  In the rating world, the statutory hypothesis is a tenancy from year to year.  As Willmer LJ observed in the Humber case (at p. 171):

“... in the world as it is, no sane manufacturer would take a tenancy from year to year of a factory in which he is going to install valuable machinery and for which he is going to assemble a skilled labour force to work for him. It seems to me that, if one had to value such a hereditament in the conditions of the world as they are, the only result would be a nil valuation, because there would be no possibility in practice of finding a tenant for such premises at all. That seems to me to be the reductio ad absurdum of the argument presented on behalf of the ratepayers.

The fact is that it is impossible to get away from the situation that the statute postulates not only a hypothetical tenant but also a hypothetical landlord, and … in the context of a hypothetical world in which the hypothetical tenant cannot become the owner of the premises and cannot get a lease for a term of years.  Moreover, one has to postulate a world in which not only this hypothetical tenant is in that position, but everybody else is in the same position.”

74.  Therefore when considering whether there is competition, one would be constrained to consider that question by reference to the real world rather than to the hypothetical tenancy because no one in the real world having to invest so much in plant and staff would take on a tenancy from year to year.  The evidence is clear that in the real world there will be persons interested.  If it had to be considered in the context of the rating world, then as explained in the Humber case, one has to postulate a hypothetical world in which everyone including the hypothetical tenant will be in the same position.  In that scenario, leaving aside, for the moment, Mr Jones’ criticisms of the appropriateness of the cost of capital approach which has been dealt with above (see §§53-64), it is far from clear quite why the tribunal cast doubt on the relevance of that evidence.

75.  Mr Yu submitted that Mr Jones’ evidence (to the effect that there would be persons interested and the level of interest would rise with an increase in the rate of return) applies equally, whether in the context of an integrated business or in the context of a hypothetical tenancy.  I agree.  I accept that where the price is set by reference to the cost of capital, common sense would suggest that in the commercial world an increase in the rate of return by adding one or more percentage points to the cost of capital rate could only have a positive effect on the level of interest that would be generated.

76.  Another aspect of the ‘uniqueness of asset’ argument is the supposed ‘symmetry’ between the respective assets of the hypothetical landlord and the hypothetical tenant, such as to give them equal bargaining power.  Mr Roots submitted that it must be taken into account in the assessment of the rent and that neither the vacant and to let principle nor the Southern Railway principle requires otherwise.  He stressed that the assets of the hypothetical tenant are of great value as Mr Jones’ calculations set out in §142 of the judgment show and that that must be recognized in assessing the tenant’s share.

77.  The Commissioner disagrees on the basis that, on analysis, the vacant and to let principle and the Southern Railway principle go hand-in-hand for the purpose of assessing the value of the occupation of the tenement and their joint effect is to negate any sitting tenant advantage or disadvantage.  Further, if the tenant’s assets are of great value, that is not because of their uniqueness; rather, it is attributable to the sitting tenant advantage.

78.  The principle enunciated by Lord Hailsham in Southern Railway (see §14 above) requires one to disregard the difficulties a tenant might face in the real world in acquiring the rolling stock required for operating the railway, recruiting adequate staff to operate and manage it and in disposing of the rolling stock at the end of the tenancy.  Similarly, in Humber, the largeness of the sum required for tooling up a large car assembly works and length of time (5 years) it would take to assemble the necessary labour force before achieving full production are disregarded.  Conceptually, the need to acquire machines and apparatus that fit the tenement is no different from having to tool up a factory or to assemble the necessary staff. Indeed, Southern Railway involved acquiring rolling stock that plainly had to fit the railway tracks.  In substance, that could be said to involve the integration of the hypothetical tenant’s assets with the hypothetical landlord’s assets.  Those matters do not give the hypothetical tenant any special bargaining power or entitle him to compensation for the risks involved and must be disregarded under the Southern Railway and Humber principles.

79.  The vacant and to let principle requires an assumption that the premises are vacant and all the process plant and machinery are removed.  Thus, in Edmondson (Valuation Officer) v Teesside Textiles Ltd [1984] RA 247 (CA), in situ process plant and machinery occupying two-thirds of the factory was to be ignored and treated as if it were not there.  But vacant and to let must not be taken literally or “carried too far” for otherwise it would mean that it would to take years before the hypothetical tenant could acquire the necessary plant and assemble the requisite staff and managers to run the operation.

80.  Amies, Law of Rating, pp. 417-418, notes that the assumption would not seem to apply to hereditaments of which continuity of operation is essential for the maintenance of their value in that the whole concept of the valuation of a public utility undertaking was based in effect on the assumption that it was taken as a going concern and that the actual occupier handed over to the new tenant (or took a new tenancy himself, since he was conceived as a possible tenant) and received (or credited himself with) the value of his tenant’s capital.  After considering Southern Railway and Humber, the passage concludes:

“Existing non-rateable process plant suitable only for use in the large steelworks in which they are used must be assumed to be available to the hypothetical tenant, whether the actual owner or other person acquiring it from him as the outgoing tenant, and it is wrong to assume that the factory is stripped of all its plant other than motive plant and that the hypothetical tenant is exposed to all the hazards of going into the market to look for new plant.”

It can be seen that, in the world of make-believe which the statute imposes, the vacant and to let principle operates and takes effect in tandem with the Southern Railway and Humber principles.

81.  As to the “value” of the tenant’s assets, Mackinnon J in his judgment in the lower court (the Railway and Canal Commission Court) in Southern Railway, stated (at 109-110):

“… the company seeks also to inflate the manufacturing cost in their own workshops by a further addition based upon the reduction of that cost that they have effected by what is called mass production. The hypothetical manufacturer, it is said, must not be supposed to possess such methods of economy; his cost price will, therefore, be higher, and it is to such a higher cost price that he will seek to add his profit. We think the principle on which this sort of addition to the cost price to the railway company is claimed, and apparently has often been allowed, is unsound. The hypothesis which divides the possessions of the railway company into two parts, belonging to it on the one hand as landlord and on the other as tenant, is necessary if the rent payable by the one entity to the other has to be calculated. But the hypothesis need not be, and in my view ought not to be, carried further than is necessary for that end. The railway company, qua tenant, is the only conceivable tenant of the railway company qua landlord, and in valuing the tenant’s capital of the railway company qua tenant I think account should be taken only of that which it has in fact cost the company.”

On appeal, Lord Hailsham spoke in terms of “market value”: see the first of the passages from his speech set out in §14 above.

82.  The Guidance Note (at 5.50 (e)) is of some assistance on this issue:

“In considering the capital sum to be attributed to items of tenant’s capital, it ... is usual in the transfer of going concerns for assets to be transferred at their market value (i.e. for this purpose, their value to an incoming tenant). Where it is not practical to determine the market value - for example, because of the unique nature of the capital item - then the replacement cost of the item should be established and a deduction made to allow for any obsolescence.”

While no valuation had been made on the basis of depreciated replacement cost, as we have seen (§§61-64), the tribunal accepted the net book value as an acceptable proxy for the value of the tenant’s assets.

83.  In conclusion, upon analysis, the integration of assets point on which HEC placed considerable reliance provides no sound basis for entitling the sitting tenant to charge a premium over the net book value of its assets.

C.   Assets under construction

84.  The issue is whether such assets are rateable under the Rating Ordinance.  The tribunal held that they are non-rateable.  Further, it accepted HEC’s contention that, as a consequence, a downward adjustment in the gross receipts was warranted.  That meant a re-opening of the ‘agreed’ gross receipts from the preceding year’s audited account.  Broadly speaking, the contention was based on the fact that assets under construction are “fixed assets” under the scheme of control and, as such, are taken into account in the calculation of the permitted return.  At the same time, the tariff (which directly affects the gross receipts to be generated) is set with a view to achieving the permitted return although, as we have seen, that is not always achieved.  So the argument goes that assets under construction contribute to the generation of gross receipts and, if non-rateable, require the re-opening of the accounts.  For present purposes, we need only be concerned with the correctness of the tribunal’s conclusion that assets under construction are non-rateable and not with the correctness of HEC’s arguments for a downward adjustment.

85.  As earlier noted, under the Rating Ordinance, “tenement” includes land that is “held or occupied”.  Section 21 imposes liability for rates on both owners and occupiers although the primary liability is on the occupier who is deemed to be liable in the absence of any agreement to the contrary.  Under section 18, the liability to pay rates attaches to every tenement whether or not it is occupied, subject to refunds where the provisions of section 30 apply.  Applying those statutory provisions, it is Mr Yu’s submission that building sites (which unarguably fall within the definition of “tenement”) are rateable.

86.  The tribunal accepted (at §165) that a tenement is rateable whether or not it is factually occupied.  It had understood Mr Roots as agreeing with that proposition.  Nevertheless, the tribunal accepted Mr Roots’ submission that a tenement must be capable of being occupied before it can become rateable and, as assets under construction are not capable of being occupied at the relevant date, as a matter of law they are not rateable.

87.  On these appeals, Mr Yu’s position is that the concept of “capable of being occupied” is an entirely new concept created for the purpose of these proceedings that has no basis in either UK or Hong Kong case law on rating.  For his part, Mr Roots submitted that Mr Yu’s point is unarguable in that the decision of the Court of Final Appeal in Commissioner of Rating & Valuation v Agrila Ltd & Others (2001) 4 HKCFAR 83 is dispositive of the matter, citing the following passages:

“... as is agreed by all, under rating law “a house in course of construction cannot be rated” (as per Lord Wilberforce in Dawkins (Valuation Officer) v. Ash Brothers and Heaton Ltd [1969] 2 AC 366 at 385-H)”

per Litton NPJ at 91D

“Another fundamental proposition of rating law, as stated by Lord Radcliffe, is that :

“Building sites themselves are not treated as rateable hereditaments [the English rating equivalent of tenements] while the work of building is in progress.”

(London County Council v. Wilkins at 380). What is important for present purposes is that the statement expresses the proposition in terms of rateability of the property, not in terms of rateable value. The proposition explains why development sites in Hong Kong have not been rated.

Viewed in the light of these well-established principles of rating law, the purpose of regulation 2 seems to be reasonably clear.  It is to overcome the problem that building sites are not rateable tenements for the purposes of the Rating Ordinance.”

per Mason NPJ at 99C-E.

88.  I turn to consider the basis for the observations set out above and whether Agrila is a bar to Mr Yu’s submissions.

89.  The first matter to note is that the issue in Agrila was whether vacant tenements such as building sites are liable for Government rent which involved the interpretation of regulation 2 of the Rent Regulations.  Regulation 2 is a deeming provision - that the rateable value of the land before development shall be ascertained “as if [it] were a tenement liable for assessment to rates under the Rating Ordinance.”  The proper interpretation of regulation 2 therefore did not require the determination of whether building sites are rateable under the Rating Ordinance and the Court of Final Appeal did not engage in that exercise.  Second, as recorded in the judgment of Mason NPJ (at 94D-E), the parties in Agrila agreed that

“the relevant sites were not rateable under the Rating Ordinance. While under construction, the sites were not regarded, for rating law purposes, as being in rateable occupation.”

The decision proceeded on that basis. As regards the parties’ agreement, it is perhaps explicable by the fact that both parties were represented by Queen’s Counsel from London who may have assumed a common historic origin for Hong Kong and English rating law when, in fact, the historic position is very different.

90.  Historically, liability for rates in England originated with the Poor Relief Act 1601.  Under that Act, the persons made chargeable were ‘every inhabitant, parson, vicar, and other, and every occupier of lands’.  In 1840, the liability of an inhabitant was abolished but the 1601 Act remained in force until the General Rate Act of 1967.  Even after 1967, the only person chargeable with the payment of rates remained the “occupier” of the land and that continued to be the case until 1992.

91.  Without a doubt, liability for rates under English common law was firmly grounded on the concept of ‘occupation’, that being the sole test of rateability.  This can be seen from John Laing & Son v Kingswood Assessment Committee [1949] 1 KB 344 and London County Council v Wilkins (Valuation Officer) [1957] AC 362 (to which Mason NPJ referred) which establish that the four ingredients necessary for rateable occupation in English law are (1) actual occupation or possession; (2) which is exclusive for the particular purposes of the occupier; and (3) of some value or benefit to the occupier; and (4) not for too transient a period.

92.  The central importance of ‘occupation’ in English rating law can be seen from John Laing where building contractors to whom a building site had been handed over were held to be in rateable occupation of various hereditaments consisting of offices and other structures they had erected on the site for the purpose of carrying out the contract.  The basis of liability was that they were in actual and exclusive occupation of those hereditaments.  The building site as such was not rateable because construction was taking place, but the structures erected thereon, albeit impermanent, were rateable.

93.  While the fundamental proposition cited by Mason NPJ from the speech of Lord Radcliffe in the London County Council case is correct as a matter of English law, it should be noted that the phrase “while the work of building is in progress” was an integral part of Lord Radcliffe’s proposition.  Insofar as Mason NPJ’s statement (at 99E) that

“building sites are not rateable tenements for the purposes of the Rating Ordinance”

is to be read as a shorthand reference to Lord Radcliffe’s proposition, that issue was not argued because of the common ground between the parties.  Notably, the Court of Final Appeal did not give consideration to the separate issue of non-occupation liability for rates under the Rating Ordinance.  Nor did it consider the issue of liability to pay rates for unoccupied tenements subject to the right to recover rates payable under section 30.

94.  In Agrila, Mason NPJ (at 98J) read the decision in Yiu Lian Machinery Repairing Works v Commissioner of Rating and Valuation [1982] HKDCLR 32 at 39 as establishing that in Hong Kong a tenement is not in rateable occupation unless the four requirements for rateable occupation in English law are satisfied.  The underlying assumption was that there was no difference between Hong Kong rating law and English rating law.

95.  The relevant passage in Yiu Lian is to be found at 41G- I (and not at 39).  After first observing:

“… that those four criteria relate neatly to English rating concepts but can only be applied in Hong Kong subject to several important qualifications” (emphasis added),

Judge Cruden proceeded to identify the differences:

“For in England occupation per se gives rise to liability for rates. It matters not whether the land occupied is freehold or leasehold or whether the occupier is in occupation as owner, lessee, licencee or otherwise. In Hong Kong under our Section 2 actual occupation by itself is not enough. Before land or any building or structure can become rateable it must in addition be held or occupied as a distinct or separate tenancy or holding or under licence. So there must be either ownership or occupation under one of those three kinds of limited title. In England if the four tests are satisfied the occupation becomes rateable. But in Hong Kong the satisfaction of the four English tests would only give rise to liability for rates if, in addition, the land was occupied under one of the three limited forms of tenure set out in Section 2. In England the four tests are complete and decisive criteria whether liability for rates arises. In Hong Kong they are, at most, merely an indicator that there may be a liability for rates.”

In other words, while those tests are applicable in Hong Kong, they are not decisive tests for rateability.  In my view, Yiu Lian provides no support for the view that the rateability of building sites in Hong Kong is the same as in England.

96.  In fact, significant differences exist between Hong Kong and English rating law.  In Hong Kong, there is dual liability of the owner and occupier stemming from the definition of “tenement”, whereas in England liability was limited to occupiers.  That was the position from 1601 until 1992 when the Local Government Finance Act 1992 had to be enacted to extend liability to owners for certain limited classes of vacant or unoccupied hereditaments whereas in Hong Kong, what constitutes a “tenement” does not hinge on whether or not it is occupied.  As earlier noted, the tribunal had accepted that proposition.

97.  Section 30, which deals with “refunds in respect of unoccupied land”, supports the view that an unoccupied tenement is rateable: that section would be otiose if rateability of a tenement depended on the fact of occupation.  Moreover, the fact that section 30 applies and a refund is made cannot alter the fact that the tenement is rateable in the first place.  It may be that from the perspective of the party liable for the payment of rates, whether the rateable value is assessed at nil or a refund is made under the provisions of section 30, the end result is the same.  Again, that cannot alter the fact that an unoccupied tenement is nonetheless rateable.

98.  Once there is a tenement, its rateable value must be ascertained in accordance with section 7.  Nothing in that section precludes a rateable value of nil being ascribed where circumstances warrant it.  It has been suggested that where, before issuing a rate demand, the Commissioner is aware that a tenement is unoccupied, a discretion may be exercised merely to issue a nil demand or a demand for a nominal sum.  The commentator noted that in clear cases of unoccupied tenements, this would avoid the cost and delay of issuing a demand and then later processing a refund.  See Cruden, Land Compensation and Valuation Law in Hong Kong, 3^rd edition (2009), p. 466, fn. 113.

99.  In my view, it follows from the analysis above that unoccupied tenements are rateable under the Rating Ordinance.  For the reasons explained at length, Agrila is no bar to this court to so hold because this very issue was not the issue the Court of Final Appeal had to decide or did decide after full consideration.

100.  As to the tribunal’s holding that in order to be rateable, the tenement must be of a state capable of being occupied, I agree with Mr Yu that not only did the proposition go beyond Agrila, it has no valid legal basis and must be rejected.  I would add this.  A tenement that is “incapable of occupation” is specifically addressed in section 31(ba) of the Rating Ordinance.  That provision mandates a refund where the Commissioner is satisfied that a tenement has become “incapable of occupation, as a result of any order made by a court on the application of Government”.  The fact that a specific refund provision was considered necessary but only in strictly circumscribed circumstances puts paid to the notion that, as a matter of general rating law, no liability to rates can arise where a tenement is incapable of occupation.

D.   Order

101.  I would allow the appeal and set aside the orders below dated 30 November 2009 and 12 April 2010.  I would also order that there be an order nisi of costs in favour of the Commissioner.

Hon Stone J:

102.  I agree with the judgment of Le Pichon JA and the orders she proposes.

Hon Rogers VP:

103.  Accordingly, there will be an order in terms of §101.