Shun Fung Ironworks Ltd v. Director of Buildings and Lands (Volume Iii)
Read the full judgment text of on BabelCite..
1. Assuming that SFI's business at Junk Bay was extinguished by the resumption of its land there, section 10(2)(d) of the Ordinance requires this Tribunal to determine the loss or damage to SFI's business as a result of that.
|
LDLR000018B/1987 Crown Lands Reference No. 18/87 IN THE HONG KONG LANDS TRIBUNAL ----------------- Between
----------------- VOLUME III ----------------- RAPT 2 PARTICULAR ISSUES - continued
PART 2 PARTICULAR ISSUES - continued
PART 2 PARTICULAR ISSUES - continued
PART 2 PARTICULAR ISSUES - continued
PART 3 ACCOUNTS
PART 4 CONCLUSION Section IV CAPITALISATION RATE 1. Assuming that SFI's business at Junk Bay was extinguished by the resumption of its land there, section 10(2)(d) of the Ordinance requires this Tribunal to determine the loss or damage to SFI's business as a result of that. 2. Both sides agree that, in practice, in the present instance, the approach to be adopted is the same as if the Tribunal were trying to ascertain the value of SFI's goodwill. 3. On that approach, the Tribunal needs to ascertain the value of SFI's business at Junk Bay as a whole by (i) finding the stream of profits which SFI could reasonably have been expected to earn at Junk Bay in the No-Scheme-World, and (ii) finding the capital value of that stream of profits as at 19th January 1987, ("the valuation date"), which is when SFI physically vacated the Junk Bay site. 4. The stream of profits is readily ascertainable from the Profit and Loss Accounts we have caused to be prepared for Financial Years 1986/7 until 1998/9, inclusive, for the No-Scheme-World, embodying our findings on SFI's rebar-making capacity, production and sales volumes, delivery prices, and scrap prices. The final version of those Profit and Loss Accounts is to be found in D.H.S.'s Bundle dated May 1992, Revised Appendix V at pages 0019 and 0020. 5. According to SFI, its future profit streams have been projected on such a conservative basis that they already reflect future risk. That conservative approach, so SFI urges, is to be seen, for example, in its profit projections for the No-Scheme-World not taking account of :-
6. Assuming such increased production and sales on the one hand and reduced costs on the other, SFI went on to contend that, whichever method or methods for capitalization of future profits the Tribunal selected, the Tribunal should pay heed to risk already having been taken into account in the conservative profit forecasts and should guard against the double counting which would occur if the same risk was again allowed for in the discount rate arising from the capitalisation method chosen. 7. Obviously, the same risk should not be counted twice against SFI. SFI's fears of that happening, however, are groundless, we feel. The possibility of such double counting is predicated on the assumption that the profit projections for the No-Scheme-World are conservative. We do not, however, share SFI's view on that. 8. In fact, we feel that the profit projections, if anything, err on the side of being favourable to SFI, because of their major assumption that the sale price of rebar will keep up with future inflation. One has only to see how SFI's sale prices fared against inflation in the Scheme-World to realize that inflation-proofing might well prove to SFI's advantage in the long run. 9. By way of general comment on the improvements in the form of increased production, greater sales and lower costs SFI contemplated for itself in the future, we find ourselves in agreement with Mr Li that such a picture is too one-sided, as it leaves out the possibility of unfavourable developments such as, for example, wage inflation or increases in the price of electricity of such a degree as to impair SFI's profit projections significantly. 10. As SFI's history shows, many things can go wrong in business, and we do not regard it as reasonable to suppose that SFI's earnings projections for the No-Scheme-World are anything like risk-free. Later in this Section, we will see, in the context of what is known as "The Capital Asset Pricing Model", that the parties have agreed that a suitable risk-free rate for Hong Kong is 8.76%. If, as SFI seeks to contend, all the risk relating to its future earnings has really been embodied in its cash flow projections, one would expect, as a corrolary, that SFI would have contended its cash flows should be discounted at 8.76% rather than the 12% to 13% SFI has in fact plumbed for. None of the textbooks or articles drawn to our attention appears to contemplate the possibility that in discounting a company's cash flow all the risk can somehow be wrung out of the discount rate, and then be carried exclusively by means of the cash-flow projection. 11. As a matter of reality and common sense in valuations of goodwill, some risk is reflected in the capitalisation rate, some in the projected cash flow. How all round risk is apportioned between the discount rate and the cash flow is a matter of impression and judgment. Sensibly, Mr Best, SFI's accounting expert, has neither explicitly nor implicitly suggested we should come up with a risk free discount rate. 12. On the specific points about SFI's already-planned improvements to its plant and machinery, and the possibility of increasing production in the future, we do not feel it necessary to make any further comment beyond what we have already said in our Section I : SFI's REBAR-MAKING CAPACITY. 13. In relation to the possibility of SFI' expanding its trading activities in the No-Scheme-World, we are far from persuaded that SFI was likely to do that on a significant scale. Even in the Scheme-World, there was nothing to stop SFI becoming a stockist, trading in rebars, but it only did that on a minor scale. The evidence does not point towards SFI possessing any particular aptitude for such trading. 14. SFI pointed out that its former neighbour, SWS, both manufactures and trades, but we do not think that necessarily suggests SFI would have chosen to adopt similar business policies. 15. Concerning SFI's contention that its capacity was in fact 123,000 M/T's rather than 110,000 M/T's, we refer to our finding in our Section I : SFI's REBAR-MAKING CAPACITY to the effect that, for the years 1985/6 and onwards, SFI would, in the No-Scheme-World, have produced at least 110,000 M/T's of rebar comfortably, but the evidence is too vague and there are too many imponderables for us to say whether SFI's capacity would have been more or less than 123,000 M/T's 16. Attention can now be turned to ascertaining a discount rate to convert the profits we have found SFI would have made in the No-Scheme-World into a capital sum as at the valuation date (19th January 1987). Before that, though, we wish to say something on the related topic of goodwill. 17. Goodwill existed only to the extent that the combined value of SFI's tangible assets, namely, its Junk Bay site ($23.750 million), the buildings there ($26.000 million), its plant and machinery ($60 million), and its working capital, are exceeded by the value which we determine for the business as a whole. 18. A somewhat exotic variant of what we have just described is to be found in Mr Best's Appendices XXXIV and XXXIV.1 where he sets out the way to find goodwill by what we were given to understand is known as an "iterative" process. (Mr Li duplicated the process with the government's numbers in 42/05E (II) pages 12 to 16). For the sake of completeness, we have had the expert accountants for both sides repeat that process with the figures we have found, and the final version is to be found in D.H.S.'s Bundle, dated May 1992 at pages 0025 and 0026. 19. Before moving on to consider the various methods drawn to our attention as possible approaches for ascertaining the discount rate, there is a further general observation we feel we should make on the topic of risk - one of the critical factors, either explicit or implicit, underlying any discount rate. 20. On the view we take, risk, for present purposes, should be perceived as it would have been by someone standing in SFI's shoes on the valuation date (19th January 1987). That does not mean the Tribunal has to disregard all evidence arising, or events occurring, after the valuation date. Heed can be paid to such evidence or events, but, only in so far as they throw light on how risk might have been perceived as at the valuation date (19th January 1987). 21. As a matter of law under what is known as "the Bwllfa principle" (Bwllfa v. Pontypridd Waterworks Co. [1903] A.C. 426), the Tribunal is required to make its findings on the basis of the actual loss or damage the evidence shows the litigant to have suffered after the valuation date. Courts prefer not to guess on an issue where actual data are available. Thus, in personal injury cases, the courts do not speculate on what, say, an injured workman as at the date of his injury might reasonably have expected his future wages to be. Instead, in assessing damages, the courts sensibly use hindsight, and base themselves in that type of case on what actual wage rates turned out to be down to the time of assessment. 22. Similarly, in the case before us, we will assess SFI's loss or damage on the basis of the evidence showing the actual delivery prices for rebars and the price of scrap down to and during the trial. 23. To the extent just shown, hindsight is permissible. 24. Unless, however, we have misunderstood the submissions made on SFI's behalf, SFI considers the use of hindsight should be extended to the point that all happenings down to the time of trial should be taken into account as part of the exercise for determining the discount rate. 25. SFI states its position explicitly in the following passage from CCS, Section 9, Vol 1, page 16, para.3.9 :
26. A somewhat ambiguous passage from Mr Best's Report 33/08, page 3, para.17, implicitly, on one reading, makes the same point. That paragraph is as follows :
27. If Mr Best's paragraph purports to mean that it is the rate of return which should be estimated with the benefit of hindsight, then we beg to disagree. On the other hand, if it intends to convey no more than that SFI'S earnings in the No-Scheme-World can be calculated with the benefit of hindsight then, on that, we agree wholeheartedly with Mr Best. 28. As SFI would have it, the Tribunal is presumed to know everything happening uptil the conclusion of the hearing, and should act upon it In particular, SFI wants the Tribunal to take cognizance of announcements made by government in mid-1989 of vast infrastructure projects for building a new airport and port development works at the western end of Hong Kong's harbour. The prospect of such works reduces the risk of SFI not selling its product, according to SFI. 29. Both as a matter of law and for practical reasons, we do not think the Tribunal is entitled to use hindsight in the way SFI suggests. 30. An illustration of how hindsight can be employed by virtue of the Bwllfa principle is afforded by Emerald Quarry Industries Pty Ltd v. Commissioner of Highwilys [1976] 18 S.A.S.R. 438, 504. 31. The question which arose in the Bw1lfa case (called by some the "Pontypridd" case) was how to compensate a coal-mining company for coal it could have expected to win from its seam during the five year period, commencing October 1898, it would have taken to exhaust the seam. In that case, the valuation date was October 1898 when the coal-mining company received the equivalent of a notice to treat. 32. In proceedings taken by the coal-mining company in 1900, it sought to be compensated for the coal it could have expected to extract uptil the time of the hearing on the basis of actual coal prices during that time, whereas the respondents, who had served the notice to treat, contended compensation should be based on the lower price obtaining at the time of the notice to treat. The court awarded compensation on the basis put forward by the coal-mining company. 33. Commenting on that case, Wells, J. in Emerald Quarry at page 504 said :
34. For the purposes of our case rebar should, mutatis mutandis, be substituted for coal in that passage. 35. In the same way that, in Bwllfa, compensation was to be assessed at the date of the notice to treat, it should in our case be assessed on the equivalent valuation date, which is 19th January 1987. 36. Moreover, in our case, while it has been perfectly in order to have regard to the actual price of rebar and scrap until about the end of 1989 (particularly for the purpose of determining trend prices) that in Wells, J.'s words :
37. Insofar as the discount rate to be applied at the valuation date is concerned, we are firmly of the view it should be based only on information ascertainable at the valuation date, i.e. 19th January 1987. 38. About the only advantage we can see for a trial lasting as long as this one is the opportunity it provided for observing the ebb and flow of public events having implications for Hong Kong's economy and for companies operating within that economy. It demonstrated to us the wisdom of not getting too euphoric over good news and not getting too disheartened by bad news, since the situation keeps changing. 39. For example, there was the massive blow to business confidence in Hong Kong associated with events in Tienanmen Square in June 1989. On the other hand, there was the initial boost to business confidence from the new airport and port announcement, but, for a long time after that, confidence fluctuated with "on again", "off again" pronouncements. 40. We imagine that anyone in SFI's position would feel greatly aggrieved if told that compensation for a business extinguished on 19th January 1987 was going to be reduced on account of unfavourable economic developments two and a half years later in June 1989. Such a person would, we do not doubt, want the value of his business at the time it was extinguished. On the basis of what is sauce for the goose is sauce for the gander, he could not expect to be paid more because of favourable economic developments years after the valuation date. Conversely, neither should the governemnt expect to pay out less in a situation where economic developments after the valuation date had brought about a deterioration in a claimant's economic prospects. 41. At the practical level, it is difficult enough to focus on one date such as 19th January 1987 for the purpose of ascertaining events at that time, with a view to fixing an appropriate discount rate in the light of the risk/reward situation as perceived then. 42. If, instead of trying to tie down an appropriate discount rate to a particular date, the Tribunal were supposed to juggle with a discount rate which keeps changing with every political and economic whirl and eddy manifesting itself as the trial proceeds, then the Tribunal would find itself confronted with a far less workable approach, and one which, in our view, would tend to a less certain and less just result between the parties. 43. Having made our point that the capitalisation rate is to be arrived at with perceptions existing at the valuation date (subject only to the Bwllfa principle), we turn our attention now to the various methods advocated by the parties for determining that rate. 44. Those methods can be conveniently grouped under three major headings, as follows :-
45. By way of a prefatory general comment, we do not regard the above classifications as mutually exclusive, and, in particular, resort to judgment must always be made to ensure the mechanical application of stockmarket-based or rule-of-thumb techniques does not lead to a result in any way unreasonable. I. STOCKMARKET-BASED APPROACHES 46. Whether on a P/E approach, or for purposes of a D.C.F. calculation, we see no advantage in having recourse to non-Hong Kong public companies in the way D.H.S. has done, since none of those companies D.H.S. has chosen is in any way comparable to SFI, except on the single point of coming from the mini-mill sector. 47. Initially, D.H.S. put together a basket of publicly-quoted companies, operating mini-mills, from stock-exchanges in the United States, Malaysia, and Singapore. 48. The United States companies were :-
49. Singapore was represented by National Iron and Steel Mills Ltd, while from Malaysia came Amalgamated Steel Mills Berhad. 50. All of those American mini-mill companies, except New Jersey Steel, operated on a vastly greater scale than SFI, and have diversified into making other types of steel product in addition to rebars. 51. Even New Jersey Steel Inc, the smallest and only undiversified U.S. mini-mill selected, has an annual capacity of 350,000 tons, more than three times that of SFI. 52. The four other U.S. companies have a production range of 2.1 million tons annually for Nucor at the top, down to Birmingham Steel, 650,000 tons, at the bottom. 53. Singapore's National Iron and Steel Mills Ltd has an annual capacity exceeding 500,000 tons of steel, and makes wire rod as well as steel bars, whilst Malaysia's Amalgamated Steel has a capacity of 400,000 tons per annum, the break-down of which into different products we do not know, but, anyway, it makes no difference in the light of our general view on the irrelevance of this company (or any of the other foreign companies), for present purposes. 54. Hong Kong is unique in the world as a country which not only preaches, but actually practices free trade. There is no feather-bedding for any Hong Kong companies, public or otherwise. 55. SFI, like any other Hong Kong company, stands to survive or perish by its own efforts. Apart from trying to keep a more-or-less level playing field, Hong Kong's government does nothing to protect any sector of the economy from outside competition. Imports of all sorts flood in from every other part of the world, without let or hindrance. Rebar is no exception. It was SFI's steel-marketing expert, Mr Gillett, who explained how the rebar-market in Hong Kong is dominated by imports. Manufacturers producing rebar elsewhere in the world, subsidised or unsubsidised, are free to dump their products in Hong Kong below cost, so, to remain competitive, SFI would have to adjust its prices downwards, SFI's only commercial advantage being the extra it could command for special lengths and its general attractiveness as a local producer. 56. Hong Kong's trading regime immediately distinguishes a Hong Kong mini-mill from one in the United States, operating behind a tariff-wall. Whether Singapore and/or Malaysia shield their steel producers from international competition we do not know. D.H.S., who selected them, did not tell us. In fact, we were told precious little about them - five or six lines on each (see Mr Best's Report, 33/08 pages 107 and 108). 57. The type of information the Tribunal needs to know for deciding whether a particular public company can be used as an indicator to assist in the valuation of a private company is described in Chilvers and Lemar's Litigation Support at page 40 para.320. We set that out:-
58. That list is by no means exhaustive. Mr Best, for example, stated in his report 33/02, para 260 :-
(It is convenient to note at this point that we prefer the views of Glover that business risk is not affected by whether a company happens to be private rather than public (op.cit. p.395), and that if the entire share capital of a private company is being valued, there is no more justification for a discount for lack of marketability than in the case of a public company (op.cit., p.212), our parentheses.) 59. For the Singaporean, and Malaysian companies, information of the type pinpointed by Chilvers and Lemar is totally lacking. 60. For four of the five United States publicly quoted mini-mills, there are Standard and Poor's reports (33/08 pages 213A - H), which are full of information, showing just how dissimilar they are from SFI. 61. The one of the five U.S. companies without a Standard and Poor's report is New Jersey Steel, about which we know as little as the Malaysian and Singaporean companies. 62. So different in scale, financial structure, (insofar as we know it), products (except for New Jersey Steel), and operating history (insofar as known), are these foreign mini-mills that, it is as inappropriate to try to pray them in aid as near-comparables to SFI on the basis of sharing the one characteristic of being mini-mills, as it would be to select Maxims of Paris as some sort of comparable for a Hong Kong cooked-food-stall on the basis that both come from the catering sector. 63. Individually, each of the foreign companies has to be rejected, on our view, for utter lack of comparability with SFI. Collectively, they are equally useless. 64. Besides, firstly, operating under trading regimes markedly different from Hong Kong's, and, secondly, either viewed singly or collectively, having virtually nothing in common with SFI beyond operating mini-mills, there is a third major ground militating against the foreign mini-mill companies, namely, that they are quoted on stock exchanges uninfluenced by events which make Hong Kong uniquely risky. 65. Throughout the case, the Hang Seng Index was treated as a barometer for Hong Kong's business risk. From an article entitled "The Risk and Return of Investing in the Far East Emerging Markets" by Dr Ho Yan Ki (Mr Li's Report, 42/05A p.66), it was clear that investing in Hong Kong is far riskier than investing in the USA, and considerably riskier than investing in either Singapore or Malaysia. It has to be borne in mind, too, that in 1986 - 87 (which straddles the valuation date) "... an extreme speculative mood prevailed", according to Dr Ho. (42/05A, p.67) 66. A further significant point arising from Dr Ho's article is how little correlation exists between stock markets in Hong Kong and in the USA, and, to a lesser extent between Hong Kong, and either Malaysia or Singapore. 67. Thus, not only is Hong Kong's stock market far more volatile (and, hence, risky) than that of the USA, Malaysia or Singapore, but, also, those foreign stock markets are moved by currents different from the ones affecting Hong Kong. 68. In sum, the foreign stock markets to which Mr Best took us are poor surrogates for Hong Kong's. That is particularly so in the instance of the United States, the main foreign market with which Mr Best's strove to make comparisons in the search for a discount rate. 69. True, there has been virtually no currency risk since 1983 for Hong Kong investors in the USA, because of the interbank interest rate mechanism introduced in Hong Kong that year to prevent the HK$ fluctuating against the US$ by more than a tiny margin (US$1 to equal HK$7.78 to 7.82), but that has not prevented the Hong Kong stockmarket being almost twice as volatile as that of the U.S.A. In this context, it is also enlightening to read the extract from Investments by Jacob and Pettit in 42/05D, pages 16 to 18, which shows that investments in Asian stock markets are the riskiest in the world. Then one has to bear in mind Dr Ho's article showing that Hong Kong is either the riskiest, or at best, the second riskiest after Philippines, of the Asian stock markets. 70. While far from ideal because of the lack of any publicly quoted mini-mills, we felt, nonetheless, that the Hong Kong stock market was infinitely preferable to any of the overseas stock markets as a basis for valuing SFI. 71. Resort to foreign companies was made by D.H.S. both for the purpose of determining a suitable Price/Earnings multiple, and to find a discount rate via the Capital Asset Pricing Model, of which more anon. 72. Having made those prefatory remarks, in relation to both stock market-based methods for deriving a capitalisation rate, namely, P/E multiples, and DCF, we will now move on to consider each of those in turn. (a) PRICE/EARNINGS ("P/ E") MULTIPLES 73. The theory behind a P/E - based approach to finding a capitalisation rate for a private company is that, from a public company with risks and prospects, at or around the valuation date (19th January 1987), comparable to the subject's, a P/E ratio (based on the most recent financial year's earnings in the case of a company already listed, but on next year's earning, as per the prospectus, in a company about to go public) should be deduced for application to the subject's earnings. 74. Superficially, such an approach might appear somewhat mechanical and objective, but, unfortunately, that is not even generally the case, and, more particularly, special problems have emerged in seeking to apply this technique to SFI in the Hong Kong context. 75. At the best of times, and even in the most sophisticated markets like London or New York, a P/E ratio from a publicly - quoted company is a crude analogue for trying to value a private company. 76. The P/E ratio implicitly embodies two interrelated elements : (i) the discount rate and (ii) the growth rate for earnings, also known as "growth opportunities" : See Brealey & Myers (Principles of Corporate Finance, 3rd Ed., p.59 et.seq.) 77. Mr Li, at page 1667, line 16 of his Transcript, described the existence of those two elements. We quote : "Yes. In a perfect world the two (P/E ratio and D.C.F., our parenthesis) should give us the same result, and under the D.C.F. approach we have to do two things. One is to project the company's cash flow, in other words, what is the company's prospects in the future. The other is to determine the discount rate for the company. In other words, for the level of risk that is inherent in this company what discount rate should be used. The P/E ratio is a combination of both which would mean that in order for the P/E ratio of these companies to be applied to Shun Fung we would have to know the prospects of each one of these companies. In other words, is their cash flow the same as Shun Fung, are their discount rates the same as Shun Fung.", and Mr Best made a similar point in his Report 33/02, para.260, already cited, and replicated in his later Report 33/08, para.21. 78. The textbooks say the same thing. Besides Brealey & Myers, see also Chilvers and Lemar, (op.cit. p.39 (para.317) and 191); Glover, Valuation of Unquoted Securities p.101, 194, 197-201, 215, 227, 231, 234, 254-257, and 401; Dixon, A Practical Approach to Financial Management p.176; and one should also see the article by Schilt, "A Rational Approach to Capitalization Rates", in 42/05D, p.20, where he says "A major problem encountered with stock market valuation is that there is no explicit reference to longer term discount rates or growth expectations, even though the price which is paid for a stock is generally related to expected returns rather than current returns". 79. Whether a minority interest in a company is being valued by reference to dividends and their prospects for growth, or whether the valuation being done is of a controlling interest by reference to the company's earnings and their growth prospects, the approach is basically the same. Provided one is satisfied such growth prospects do in fact exist, one needs, in practice, to make the simplifying assumption of a normal long term growth rate. 80. The formula expressing the Present Value of a controlling interest through the medium of the future earnings stream of such a company is :-
81. The formula for valuing minority interests by reference to dividends is :-
82. Those formulae explicitly consider the growth rate - g - in earnings or dividends, as the case may be, and the expected rate of return. That basically is what Mr Li was describing in the passage from his Transcript at page 1667 relating to a company's cash flow in terms of its future prospects and the discount rate. 83. Unless one can isolate the growth element implicit in the P/E ratio of a particular quoted company, that P/E ratio will be so ambiguous as to be virtually worthless as the basis for valuing another company. Unanalysed, a high P/E ratio can mean the company is perceived by investors as having above-average growth expectations for the future, or that it is expected to be about to pay out above average dividends, or, to ring all the changes, some combination of the two. 84. As Glover (op.cit. p.283) points out, a P/E ratio is merely the ratio of price to a single year's earnings and is not a discounting mechanism. 85. It is only if the subject private company has similar growth prospects to the quoted company that the quoted company will have the necessary characteristics of comparability. 86. There were seven Hong Kong publicly quoted companies selected by Mr Best to be what he variously described as "near comparables" or "indirect comparables" for his P/E multiple and D.C.F Approaches. 87. Although Mr Best and Mr Li both agreed that "normal potential earnings" should be the multiplicand, to which an appropriate P/E ratio should be applied, it turned out that each had very different notions of the concept of "normal potential earnings". 88. To Mr Best, it meant a figure based on his projections of SFT's earnings in the No-Scheme-World, from the valuation date (19 January 1987), onwards (i.e. till 1998/99). 89. For Mr Li, "normal potential earnings" should be based principally on SFI's earnings in the No-Scheme-World for the Financial Year 1986/7, during which resumption occurred. (See Exh SFI 219, page 3) 90. Both Mr Li and Mr Best agreed that the market only looks at earnings' prospects a relatively short time ahead in the context of factoring in the sort of information which gives rise to a particular company's P/E multiple, Mr Best suggesting a period of about a year. (Mr Best's Transcript, p.1522, lines 3 to 9; p.1523, lines 34, 35) 91. As the reason urged upon the Tribunal for adopting the P/E multiple approach to valuation is its being market-based, we are of the view that Mr Li's interpretation of the concept of "normal earnings potential" or "normal potential earnings" is to be preferred to Mr Best's, since the former's is a closer simulation of how the market in fact behaves. 92. Mr Best's seven publicly quoted nearest-comparable companies were :-
93. They were choosen by Mr Best, in the main reasonably, we think, by virtue of their being either suppliers (i.e. K. Wah Stones, Tung Wing Steel and Green Island Cement), to the construction industry in the same way as SFI, or, at least, construction companies (Kumagai Gumi, Sung Foo Kee, Paul Y Holding and Kier Kin Sun (also known as Beazer), which would be expected to follow the same economic cycle as SFI. 94. Rightly, in his evidence, Mr Best emphasised SFI's growth prospects, but, wrongly, failed to analyse how they compared with the growth prospects of his closest comparable companies. 95. K Wah Stones, Kumagai Gumi, Tung Wing Steel and Sung Foo Kee all issued prospectuses to go public on the offer dates listed below, on the prospective P/E ratios shown :
96. Although Tung Wing Steel had to abandon its attempt at listing on the date shown because of legal problems, we do not see that as a reason to exclude it from the present exercise. (As a matter of interest, it was in the end privatized in September 1989 on a prospective P/E ratio of 5.7). 97. We fail to see how SFI is in any way helped by bringing any of those four raw, prospective P/E ratios to the Tribunal's attention. The last thing SFI wants is to follow those four companies with their prospective P/E ratios of, on average, about 7.5, and SFI's No-Scheme-World prospective earnings of $6.506 million for its Financial Year 1986/7, which corresponds with the end of the next Financial Year for each of those four about-to-be listed companies. What SFI, in fact, has in mind is to borrow their average P/E ratio of, say, 7 to 8, but then to apply that to a well-massaged "future potential earnings" figure, rather than to its prospective earnings for the Financial Year 1986/7. SFI's prospective earnings for the Financial Year 1986/7 are, we think, the closest and most natural analogy to the next-to-expire Financial Years of those four companies when they went public. 98. As we know nothing about the expected growth rates of any of those four companies, we fail to see how knowledge of their prospective P/E ratios can assist the Tribunal in valuing SFI. To us, it is a travesty to describe D.H.S.'s prospective P/E approach as "market-based", since D.H.S. only wishes to avail itself of the advantages, such as the relatively high P/E ratio of 7 or 8, but wants to discard the part tieing that multiplier to the forecast earnings for the Financial Year next due to expire. 99. Chilvers and Lemar (op.cit., p.39, para.317 and 318) described the conventional price/earnings method for arriving at a capitalization rate :-
100. There is not a whisper of a suggestion in either of those paragraphs of it being legitimate to take account of growth expectations for the subject not only by borrowing the Price/Earnings Ratio of the comparable but also by adjusting the earning of the subject. Mr Best's hybrid seeks to have it both ways. As Mr Best correctly observed, (Transcript 1524), "It's not the conventional P/E approach." 101. For reasons unfathomable to us, Mr Best seemed to think the paragraphs we have just cited from Chilvers and Lemar had something to do with the difference between a company's historic and prospective earnings. 102. Mr Best struck us as suffering from something of a blind spot in relation to the significance of historic and prospective P/E ratios. (See the Transcript of his evidence, pages 639 to 683 passim, and page 1530). 103. Irrespective of whether the prospective P/E ratio arises in a prospectus situation or a journalist's or company chairman's forecast, we see no merit in a suggestion made by Mr Best that a prospective P/E ratio should be treated as half an historic P/E ratio. The reality is a prospective P/E ratio can just as easily be more than the historic, as it can be less, or the same. There is no rule. Everything depends on the particular company's earnings trend as perceived by the market. 104. In the context of prospective P/E ratios, there was a further strange feature in Mr Best's approach to his Hong Kong near-comparable companies. Three of them were already publicly quoted at the valuation date (19th January 1987). They were Green Island Cement, Paul Y Holdings and Kier Kin Sun (also known as Beazer). 105. He came up with what he described as a "prospective Price/Earnings ratio" for them by a method which has never ceased to astonish us. This involved taking the price of each of the three companies on the valuation date (19th January 1987), and the earnings of each of them at the end of its Financial Year then running. The earnings at the end of that Financial Year were then divided into the company's price on 19th January 1987, and the result described as "the Price/Earnings ratio". 106. On that hypothesis, the market had not only perfect hindsight, but also perfect foresight. Nothing could be less like the real-world market. Brealey and Myers (op.cit, p.60) make the observation in relation to price earnings ratios that, some financial analysts, "... often use the ratios in odd ways". 107. The results Mr Best got from his exercise are as follows:
108. At Mr Best's request, Paul Y Holding's 35.22 gets discarded as being too far out of line with the other prospective P/E ratios. 109. Even if it had not been reached in this extraordinary way, Kier Kin Sun's (also known as Beazer) P/E ratio should in any event be excluded, in our view, because its shares saw so little trading. A share traded so infrequently cannot be safely treated as a basis for drawing inferences about public companies. Some weeks it did not experience a single trade : See Exh R146 and 150; and Exh SF256. 110. Thus, on the view we take, Green Island Cement is the only one of Mr Best's three Hong Kong companies actually listed at the valuation date (19th January 1987), whose strangely derived "prospective Price/Earnings ratio" needs any further mention. What further we have to say about it is that the Tribunal can set no more store by it than it does by the four companies each with a prospectus, namely, K Wah Stones, Kumagai Gumi, Tung Wing Steel and Sung Foo Kee, since, as with them, no analysis has been done to identify growth rates. Hence, the Tribunal lacks a rational basis for comparison with SFI. 111. Albeit by very different routes, Mr Best came up with the same P/E ratio - approximately 8 - for the American, Singaporean, Malaysian mixture (historic), as the Hong Kong publicly quoted companies (prospective). 112. SFI's whole P/E multiple exercise should have been abandoned in our view. Not only did Mr Best acknowledge that the Discounted Cash Flow approach was theoretically superior, but, also, in our view, the P/E approach was redundant in the light of SFI's flat, inflation - proofed No-Scheme-World earnings from 1988/9 anal onwards. That inevitably meant that SFI's P/E ratio was the reciprocal of its Discount Rate, (See : Exh R121 and 122, and Brealey & Myers, (op.cit.), pages 56 and 60), so, if you knew the one, why bother with the other? 113. As the Concise Oxford Dictionary defines it, "reciprocal" is the, "function or expression (of one number, our parenthesis) so related to another that their product is unity". That dictionary gives the example of 1/5th being the reciprocal of 5. Another example in the present case was 8 being the reciprocal ofl2<<%,. i.e. 100% : in other words, unity. 114. One practical reason why valuers have recourse to the P/E multiple approach to valuing a private company rather than the D.C.F. method is the problem of working out mutually acceptable projections of future earnings for the private company to be valued. Without such mutually acceptable projections, negotiations on the D.C.F. basis are likely to founder. 115. In the particular circumstances of the present case, however, the methodology for estimating SFI's future cash-flow has been agreed, and, even though the parties might not agree on the cash-flow's quantum, that does not present any obstacle to our following the D.C.F. approach. Both sides agree it can be used. The only question has been whether the P/E multiple method can be used in addition. 116. Assuming that a publicly quoted company could be found with sufficient points of comparability qualifying it to become a surrogate for SFI, the Tribunal would then need to bear in mind that the price of the public company on the stock-market represented minority interests, which, by definition, lacked control of the public company, whereas our hypothetical purchaser would be buying SFI as a whole, meaning, with full control. We find ourselves in agreement with the view put forward by Glover (op.cit. p.229 and 236) that, for P/E purposes, only recent take-over bids of comparable quoted companies are of any real help when it comes to ascertaining an appropriate P/E multiple for a controlling interest in a private company. P/E ratios evident from share, prices generally relate to trades of small parcels of shares and there is the obstacle of finding a rational basis for applying that type of minority P/E ratio to a controlling interest. 117. Experience from all stock markets drawn to our attention shows that, a bidder for sufficient shares in a public company to give him control, will, in practice, have to pay a premium above the price being quoted before news of the proposed take-over broke. 118. To ask how much, as a rule, that premium is likely to be, is akin to asking how long, as a rule, is a piece of string : the answer depends on the circumstances. Sometimes such a premium will be 50% or more (see Chilvers & Lemar : Litigation Support, para.321), whereas, at other times, it might be as little as 7.8% (e.g. the premium paid when Green Island Cement went private in 1989), or it could be even less. 119. One of the factors which frequently plays a part in the size of the premium for a public company in Hong Kong is the shares being traded at a discount to the company's tangible assets. The take-over or privatisation offer-price is likely to be pitched at a level eliminating, or at least reducing, that discount. One of the attractions of the D.C.F. approach is the explicit recognition of the full value of the tangible assets of the subject. 120. Another factor which can lead to a premium for control is that the purchasing company might be willing to bid up the price because of special benefits it discerns for itself such as increasing profits by changing the managemnt policies of the target company or from synergy, meaning the additional benefits to both the bidder and the target company from integrating their businesses together. Put another way, the combined profits of the two companies once integrated will be expected to exceed the sum of the profits of the two companies if there is no take-over. That type of consideration has no part to play in setting a capitalisation rate for SFI since both sides proceed on the assumption that SFI's managment policies will be the same in the No-Scheme-World as in the Scheme-World, and the operations of New World and its subsidiaries were already integrated with SFI's. 121. Another reason, according to SFI, the P/E ratio selected for it deserves a lift is recognition in any hypothetical sale of its business being debt free, and the value to SFI of its relationship with its parent, New World, and the other companies in the New World Group, which assured SFI's not only being able to sell all its product, but also made available to SFI a reliable and cheap source of finance. 122. The elements of being sold debt free, and being able to sell all its production certainly would count in SFI's favour. Being able to sell all its product is, in our view, reflected in the projections of SFI's sales in the No-Scheme-World. 123. We do not think too much can be made by SFI of the point about cheap and reliable finance from the New World Group. 124. Cheapness is relative. Because SFI was already so loaded down with debt, and had been consistently making losses after 1974/75, the finance made available by New World to SFI from 1978 onwards, when New World took over SFI's funding, was, at 1% over Prime, cheap for SFI. As we gathered from Mr Best, no bank would have entertained making further loans to SFI in the Scheme-World from 1978 onwards, the risk being too great. In coming up with a P/E multiple for SFI, SFI's chronic liquidity problem over the years goes on one side of the scale along with SFI's other disadvantages, whilst the relative cheapness of the funds flowing to it from its parent, joins with SFI's other advantages on the other side of the scale. 125. To what extent New World's financing of SFI deserves the description "reliable" is open to question. 126. Clearly, SFI, in the Scheme-World, from at least 1978/9 onwards, could not, in practice, rely on getting adequate financing from New World to buy reasonably priced scrap, either from ships SFI might have broken if it had had the money to buy them in the first place, or from dealers. New world, in effect, stood by while SFI suffered for years from a chronic cash shortage. 127. Money was, however, made available by New World to SFI for capital investment in new plant and machinery, and, in the end, for SFI's operations. 128. On the evidence, we accept that SFI could rely on getting funds, ultimately, from New World for new plant and machinery, but whether SFI could rely on getting such funds within a reasonable time of needing them was far from clear on the evidence. Either SFI's management and work-force carried out the Second Phase of Expansion at a glacial pace from 1973 to 1981 because it lacked the necessary skill and experience to move faster, or else they were held up because they could not rely on getting the necessary funds as and when they needed them. Whichever alternative is correct, SFI does not emerge with any unalloyed advantage. 129. In the quest to extract a P/E ratio for SFI, no public company, local or foreign, measures up to being a comparable or near-comparable. Even if there were, the problem then arises of the absence of take-over bids to provide the foundation for a P/E ratio suitable for a controlling interest. Added to that, there is the whole host of factors, already identified by us, which need weighing in the balance to make allowance for SFI's unique circumstances. 130. In such a situation, there is no practical hope of reaching an appropriate P/E ratio for SFI by any properly reasoned approach. The Tribunal is effectively thrown back on the essentially subjective judgment of the parties' accounting experts, or the Tribunal's own robust impression. 131. Because of the inadequacies of the quoted-company-comparable, P/E approach, at the best of times, particularly the problem of having to try and tease out the growth rate - g - for the publicly-quoted company's earnings, and because of the particular futility of going through the motions of a P/E approach for SFI, the inflation-proofed, flat earnings of which, in the No-Scheme-World, inevitably lead to SFI's P/E ratio being the reciprocal of its discount rate, we see no point in going through the motions, of a P/E exercise any further. (b) DISCOUNTED CASH FLOW (i) CAPITAL ASSET PRICING MODEL 132. The Capital Asset Pricing Model (C. A.P.M.) is one of a series of interrelated theoretical concepts, including the Efficient Market Hypothesis (E.M.H.) and Modern Portfolio Theory (MPT), developed by academic economists over the past 20 years or so, in an endeavour to explain stock-market behaviour. 133. Using C.A.P.M. as a tool, Mr Best and Mr Li have come up with "real" i.e. inflation-proofed, discount rates to be applied to SFI's post-resumption real cash-flows in the No-Scheme-World, Mr Best's real discount rate being 12% to 13%, and Mr Li's, 28%. 134. The element of C.A.P.M. winning universal approval is its recognition that the rate of return required by reasonable investors from a share will be the risk free rate represented by wholly safe government securities (e.g. Treasury bonds or bills in the United States or "gilts" in the United Kingdom), plus a premium (known as the "Market Risk Premium" or "Expected Risk Premium") for the additional risk inherent in any investment in shares, whether in a single company, like, say, Hong Kong Bank, or, in a basket of shares, represented by an index, like the Hang Seng Index. (H.S.I.) 135. When one speaks of the "rate of return" in this context, it means the discount rate (of the type to be applied to future cash flows to transform them into net present value), and does not mean the compounded yield a share or an index has earned in the past. 136. For present purposes, the object in ascertaining these discount rates is so that, if a particular public company, or perhaps a sector of an index, or even a whole index, can be regarded as a comparable, or at least as a reference point for use in relation to SFI in the No-Scheme-World, the rate of return (i.e. discount rate for that quoted company, sector, or index) will be used as the discount rate (or at least as a reference point) to be applied in the Discounted Cash Flow (D.C.F.) exercise for S.F.I. 137. By way of an example, the nominal rate of return (discount rate) calculated for the Hang Seng Index (H.S.I.), using data for the twenty-five year period from 1964 to 1988, was 28.59%, whereas the total compound yield, (by way of capital gains with dividends reinvested), receivable by an investor who had invested in the 33 constituent stocks of the H.S.I. for that same period (with the same weighting) would be 18.33% (See SF 231). Assuming that we were to regard SFI as being about as risky as the H.S.I., then the nominal discount rate for our DCF exercise on SFI would be 28.59%. 138. The "rate of return", i.e. discount rate, calculated from the H.S.I., is the arithmetic average over the period 1964 to 1988 of the annual changes (i.e. plus or minus any capital gain or loss) in that index, together with each year's dividend yield, giving the result 28.59%, whereas the 18.33% compounding yield is the geometric average of capital gains and dividends over that same period. 139. Probably the most effective way of showing how a rate of return/discount rate comes about in the C.A.P.M. situation is to present the table illustrating Mr Li's workings for the H.S.I. (extracted from his Report 42/05A page 50) :-
SHUN FUNG IRONWORKS, LIMITED
NOTES 1. Being the percentage change in the Hang Seng Stock Index from the first trading day to the last of the year. The change of the year 1964 has been annualized. 2. Index and dividend yield information are extracted from 'Hong Kong Quarterly Review', January 1989 by W.I. Carr. An extract is provided in Doc 2." 140. Much confusion arose during the hearing as the result of the accountants, counsel, and, no doubt, the Bench loosely saying something like, "If you had invested in the HSI in 1964, by the year 1988 you would have got a return of 28.59%," when, obviously, in context, and against the background of the case, the discount rate of 28.59% was meant. 141. An example of the type of confusion we are describing occurs in CCS, Section 9, Part I, blue page 113, para.18.8. 142. That SFI know and accept that it is the arithmetic, and not the geometric return which needs to be applied when ascertaining rates of return/discount rates can be seen from, inter alia the following sources:-
143. Moreover, there are countless occasions to be found in the Transcript when we corrected counsel on this point. 144. If any more authority is required for the proposition that it is the arithmetic average return which is to be used for the purpose of estimating discount rates, we refer in particular to foot-note 2 on page 126 of Brealey & Myers (op.cit), and there are further references elsewhere in that work along the same lines. Glover (op.cit) is peppered with passages to the same effect. 145. Inconvenient though it may be to SFI to have to accept this, there can be no doubt that the nominal discount rate, derived from the H.S.I. as a whole, is 28.59%. 146. As it is agreed that Hong Kong's historic risk free rate is 8.76%, a process of simple arithmetical deduction shows the Market Risk Premium for investing in the Hang Seng Index is 19.83%. 147. Anyone contemplating an investment in the 33 constituent stocks of the HSI (with the same weighting) would, on the strength of C.A.P.M., know that the required rate of return was the Market Risk Premium of 19.83% plus whatever the future risk free rate happened to be at the time of the investment. An important feature of C.A.P.M. is the assumption that the Market Risk Premium, which we have already explained has been derived from historic market data, represents, in effect, a "normal" risk premium which is also made to serve as the future risk premium. (See Brealey & Myers, op.cit. pages 126 and 127). 148. For the purposes of the present case, it was agreed that 8.76% should also be deemed the future risk-free rate. That was a simplifying assumption. In sophisticated markets like London, say, or New York, the future risk-free rate will be based on yields from government bonds maturing at some date in the future. As there is agreement between the parties on a future risk-free rate for Hong Kong, we are spared the problem of how, if an issue were made of it, a risk-free rate would be determined for a place like Hong Kong where the government is not generally funded by debt, with the result that the substratum of a market in government bonds hardly exists. 149. With this agreed future risk-free rate of 8.76%, to which must be added the Market Risk Premium of 19.83%, the required nominal rate of return on the Hang Seng Index as a whole becomes 28.59%. 150. From that, it is an easy step to say that the required nominal rate of return (i.e. discount rate) for projects attended by the same degree of risk as investing in the Hang Seng Index as a whole should be 28.59%. This is uncontroversial. 151. It was Mr Best in his Report 33/02 who first introduced C.A.P.M. as a basis for ascertaining a discount rate for SFI. He used American material based on sixty-odd years' data on shares making up the Standard and Poor's 500 - stock index, as well as other indices. Mr Li then did his equivalent exercise for Hong Kong, already-referred-to, using the H.S.I. since its inception in 1964. 152. Now that the spade-work has been done for a Hong Kong C.A.P.M., we regard it as bordering on the perverse for SFI still to try to cling to the United States equivalent. It is possible to contrive an appearance of concordance between the U.S. Standard & Poor's 500 - stock index and the H.S.I. when applying C.A.P.M. for the purpose of arriving at a discount rate for use with SFI's cash flow. With creative massaging of the data input, neither Mr Li, nor Mr Best had any difficulty in arriving at a preconceived rate. By adding on, or taking off a few percentage points here or there for a multiplicity of reasons, such as a special Small Stock Premium, or a change in the historic risk-free rate culled from some recently published new statistics, no doubt the accountants would be able, say, to reconcile data from the Helsinki or Buenos Aires market indices and the H.S.I. with a similar approach, but we ask, "What is the point?" In our view, nothing is to be gained from resorting to foreign indices. 153. Already, in relation to P/E multiples, we have drawn attention to how volatile the Hong Kong stock market is when compared to markets in the U.S.A. As we mentioned previously, the Hong Kong stock market is almost twice as volatile as the U.S. market, according to the article by Dr Ho "Investing in Far Eastern Emerging Markets", and Jacob and Pettit on "Investments". 154. Because of Hong Kong's high volatility, the H.S.I. and publicly-quoted Hong Kong comparables, or near - comparables, (if there are any), should be employed as surrogates for SFI in any C.A.P.M. exercise. 155. It will be recalled that in relation to P/E Multiples of quoted closest comparable companies, an issue arose on whether the circumstance that such multiples represented minority interests made them unsuitable without adjustment for the purpose of valuing a controlling interest in a private company such as SFI. A question arose during the hearing whether that problem with P/E Multiples is paralleled in C.A.P.M., which is based on a share price index, (the H.S.I.), reflecting trades of minority interests only. 156. According to Mr Best, a discount rate for SFI derived from applying C.A.P.M. to quoted comparables should be adjusted downwards, since C.A.P,M. is based on public trading of small parcels of minority shares, whereas, it was common ground, the valuation exercise for SFI was to be done on the basis of a 100% controlling interest. 157. As we will in due course more fully explain, C.A..P.M. is concerned solely with measuring the volatility of quoted shares relative to movements of the market as a whole, such market as a whole in Hong Kong's case being represented by the Hang Seng Index. Via C.A.P.M., a quoted share with low volatility has a lower capitalisation rate conferred on it than one with relatively higher volatility. It would only be if bids for controlling interests in quoted companies made the shares of such companies less volatile that a lower capitalisation rate would ensue from recourse to C.A.P.M. 158. As is common knowledge, take-over bids for public companies almost invariably make their shares more, not less, volatile. Hence under C.A.P.M., a take-over bid for a company will result in a higher discount rate for it. 159. Such a topsy-turvey consequence flowing from the C.A.P.M. concept calls C.A.P.M.'s credibility into question. That is not C.A.P.M.'s sole outlandish consequence, of which more anon. 160. Conceptually, then, where resort is made to C.A.P.M., there is no justification for adding a premium for control (i.e. adopting a lower discount rate) when valuing a controlling interest in a private company by means of a quoted comparable, for which a discount rate has been calculated using C.A.P.M. 161. Away from C.A.P.M.'s Wonderland, the position in the real world is that premiums for control of public companies can often be substantial, but that, frequently, means no more than that the price for the target comany is being raised to a level closer to net asset value. See Chilvers and Lemar, op.cit, p.40, para.421. 162. We accepted what Mr Li said about such a phenomenon occuring in Hong Kong of quoted shares frequently trading at a discount to their net asset value. From there, he went on to argue that, in SFI's situation, there was no danger of reliance on C.A.P.M. leading to a valuation for SFI less than its net asset value (or, at least, the net asset value of its tangibles), since the tangible assets have been separately valued. 163. The Hong Kong public companies already accepted by us as surrogates for SFI on a P/E approach will also be the most suitable for use with C.A.P.M., subject only to those being excluded which have not been publicly quoted for a sufficiently long time - a minimum of two years - to render their use with C.A.P.M. satisfactory. Thus excluded will be Sung Foo Kee and Tung Wing. 164. Kier Kin Sun, (also known as Beazer), the quoted company we eliminiated as inappropriate for the P/E approach on account of the infrequency of dealings, will likewise be unsuitable for C.A.P.M. 165. If a C.A.P.M. exercise is to be performed for SFI, Hong Kong's nearest-comparables, in our view, are the suppliers of construction materials, Green Island Cement and K. Wah Stones, and the construction companies, Paul Y, and Kumagai Gumi. 166. We will not adopt the property development companies (Cheung Kong, Henderson Land, Hong Kong Land, New World and Sun Hung Kai) put forward by Mr Li, since property development does not necessarily follow the same business cycle as supplying construction material or construction itself, which, in our view, are the sectors closest, cyclically, to SFI. However, in practice, the rates of return obtained by applying C.A.P.M. to Mr Li's property development companies fall within the same range as Mr Best's construction and construction material companies. 167. A particular candidate put forward by Mr Li as a near-comparable was New World Development, but, because of the diversity of its activities, which range from shipping to operating a television station, we reject it as a near-comparable. 168. However, in our view, New World Development, and a selection of other companies we shall mention, do have a role to play when it comes to standing back, and using common sense to see whether the discount rate arrived at by C.A.P.M. for SFI is realistic. 169. In due course, we will explain how Mr Best used C.A.P.M. to arrive at an unadjusted real discount rate of 16.4% for SFI, whilst the equivalent rate for New World Development itself was 25.9%, and the Hang Seng Index as a whole was 20.1%. 170. Such a result is sufficient to make one pause and ask whether there might be something faulty about C.A.P.M. Via C.A.P.M., one is invited to accept that SFI, a cyclical, single-product company, which, even in the No-Scheme-World, never made a profit and suffered chronic liquidity problems between 1975/6 and the valuation date (19th January 1987), is considerably less risky than its parent, New World Development, which is highly diversified and has always been healthily profitable at all material times. C.A.P.M., as applied by Mr Best, also tells us that it should be far less risky to invest in SFI than to buy the full spectrum of stocks making up the H.S.I., with the same weighting. 171. Mr Best adjusted that 16.4% for SFI down to 13% on the basis that the 16% had been derived from the H.S.I., which is based on trades of minority interests, whereas SFI's discount rate had to be on the basis of a 100% controlling interest, and also taken into account for the purpose of reducing that 16.4% to 13% was the circumstance of the advantages SFI was said to enjoy by virtue of being a subsidiary of New World Development. 172. If the C.A.P.M. - derived real discount rates of 25.9% for New World Development itself and 20.1% for the Hang Seng Index as a whole undergo a similar adjustment by scaling them down on the basis of a controlling interest in the way Mr Best has done for SFI, they will both still finish up with a far higher real discount rate than SFI. 173. With our feet planted firmly on the ground, those were the sort of implications we needed to ponder in our minds before deciding whether to become converts to C.A.P.M. 174. Because C.A.P.M. had us so perplexed, we cast around for a source of enlightenment. It was not until we had the opportunity to consult Glover's "Valuation of Unquoted Securities", which is recommended as a reference book by Chilvers & Lemar (op.cit, 195), that the scales started to fall from our eyes. 175. Glover certainly struck a sympathetic chord with us when he wrote (op.cit., 167), "MPT ("Modern Portfolio Theory", our parenthesis) is unintelligible to most practitioners. It is neither simple nor elegant but labyrinthine and abstruse". 176. A chartered accountant himself, Glover spent eight years in the U.K. specialising exclusively in the valuation of unquoted shares with the international accounting firm of Ernst & Whinney. From his book written in 1986, he comes across as a practitioner who does, in fact, understand MPT, including C.A.P.M., and, having weighed it in the balance, he finds it wanting. 177. As part of their training, accountants obviously encounter MPT including C.A.P.M. in their text-books. It gets a mention in the various accounting or corporate finance books brought to our attention in evidence. (e.g., Brealey & Myers (op.cit.); Dixon (op.cit.); Chilvers Lemar (op.cit.); Jacob and Pettit (op.cit.); Franks & Scholefield, Corporate Financial Management (2nd Ed.) and the various works referred to in Doc. Nos.27/18, 27/19 and 27/20. 178. Both Mr Best and Mr Li knew about it, but there was not the hint of a suggestion that either of them had used it previously in practice. 179. We now have the feeling of having been used as guinea pigs in an experiment which failed. 180. Five years ago, Glover (op.cit. 169) wrote as follows :
181. Based on our own numbing experience in the present case, we think, with no small measure of confidence, we can add the judicial to any other worlds where C.A.P.M. is not likely to gain widespread acceptance. 182. SFI's submissions (CCS Section 9, Vol I, page 12, para.21), no doubt seeking to allay the sort of doubts which beset us, raised the rhetorical question what possible risk could there be in investing in SFI in the No-Scheme-World. 183. To that, we give the considered reply that the risk of events similar to those of 1976 onwards, repeating themselves, has not gone away. During, say, the two or three years running up to 1976, SFI's prospects must, no doubt, at times, have looked almost equally roseate, as those acting for SFI say they were in January 1987. Who is to say there will not again be technological advances in steel-making which cause SFI to lag behind, or another oil-crisis which greatly increases the price of electricity, or another world-wide glut of rebars, leading to below-cost dumping in Hong Kong? 184. As we have already indicated, we regard C.A.P.M. as having made a useful contribution to peoples' understanding of the implications of investing in the stock market by pointing out how investors require a premium above the risk-free rate to compensate for the greater risk of equity investment. Applied to a whole stock index, such as the H.S.I., C.A.P.M. gives a sensible result by way of a rate of return (discount rate), which needs no massaging. 185. It is when attempts are made to apply C.A.P.M. for the purpose of finding a rate of return for individual stocks that we find ourselves beset by serious misgivings. 186. We confess to having been initially attracted to C.A.P.M. by its aura of scientific method. Instead of the old-fashioned method of intuitively plucking discount rates from the air or using crude rule-of-thumb approaches like the pay-back period, here was an elegant mathematical formula :-
"r" being the required rate of return, "rf" the risk-free rate, which, as we have shown in the Hong Kong context, is agreed at 8.76%, "B", the Beta coefficient, measuring a particular stock's sensitivity in relation to the movements of the market as a whole, and "rm", the market's historic rate of return. The "market" in our case is, of course, the Hong Kong stock-market, as represented by the 33 stocks in the H.S.I., which, as we have earlier said, has a nominal required rate of return of 28.59%, meaning that a business venture, with risks of the same magnitude as the H.S.I. as a whole, should have its estimated nominal future cash-flows discounted at 28.59%. 187. We will now explain the function of the Beta Co-efficient (usually shortened simply to "Beta" or, in the formula, just "B"), in the context of C.A.P.M. 188. Firstly, one needs to understand C.A.P.M.'s approach to the measurement of risk. 189. By looking at Mr Li's table (set out earlier from his Report 42/05A, at page 50), showing the arithmetic average returns on the Hang Seng Index over the 25 year period from 1964 to 1988, one can see how there was a wide spread of outcomes in the sense of fluctuations in returns from year to year. To take extreme examples, the highest of those annual returns was a gain of 148.47% in the overheated market of 1972, and the worst loss was 53.74% in 1974. 190. For ease of understanding, that spread of outcomes of the H.S.I. could have been presented in the form of a histogram of frequency distributions, as has been done by Brealey & Myers in their Figure 7-2 (op.cit., page 129) for the Standard & Poor's 500-stock index. 191. Statistical measures of spread of outcome are variance, and variability, this latter also being known as "standard deviation". Variance and standard deviation, in any particular instance, are different facets of the same data, variance being the expected squared deviation from the expected return, while the standard deviation is the square root of the variance. 192. We will confine ourselves as far as possible to standard deviation, since it is expressed in the same units as the required rate of return. To give an example of that from Brealey & Myers (op.cit. page 131), the standard deviation on the Standard and Poor's 500-stock index, measured over the years 1926 to 1985, was 21.2%, while the rate of return (arithmetic) for the same period was 12% (nominal). 193. Like the Market Risk Premium to which we have previously referred, the standard deviation of the market, too, is calculated from historical material relating to the market. For present purposes, the market is represented by the H.S.I. The spread of outcomes of the H.S.I. from 1964 to 1988 comprise the historical data giving rise to the probabilities incorporated in the calculation of the H.S.I.'s standard deviation for that period. 194. Although we know that the rate of return for the H.S.I. for the period 1964 to 1988, was 28.59%, we were never informed of the H.S.I.'s standard deviation for that period. 195. The only information presented to us on the standard deviation of the H.S.I. is that to be found in Mr Best's exhibit SF219 from page 97 to 100 for the period 1978/9 to 1988/9. Unfortunately, that is based on S.F.I.'s Financial Years, whereas Mr Li's exercise for the rate of return on the H.S.I. was on the basis of calendar years. A further problem in reconciling Mr Best's tables with Mr Li's is that Mr Best has not added the dividend yield, which means he has not shown annual total returns in the way Mr Li did. 196. In addition to the presentational problems just mentioned, we strongly suspect that Mr Best has used a different approach from Brealey & Myers for calculating standard deviation. Mr Best's answer of 0.1 (at page 100) strikes us as very odd for the standard deviation of one of the world's most volatile stock markets in respect of an 11 year period of high volatility. We did not query that standard deviation result with Mr Best during the trial, since it was only subsequently that the full significance of standard deviation to C.A.P.M dawned on us. 197. Not knowing the actual standard deviation for the H.S.I. calculated on the basis of annual total returns for Mr Li's period of the twenty-five calendar years 1968 to 1988 is not an impediment to understanding the implications of C.A.P.M. for SFI : what really matters is being aware that risk, for the purposes of C.A.P.M., has to be considered in terms of standard deviation. 198. The importance of standard deviation in our context is emphasised by Glover (op.cit., page 165) where he says, "At the heart of MPT Lies the belief that the risk of a share can be measured by the standard deviation of 'expected' future rates of return. The beta co-efficient, which measures a stock's systematic or market risk, is intimately linked to the standard deviation of returns." In due course, we will enlarge upon "systematic or market risk." Brealey & Meyer's, too, (op.cit., pages 127 to 132) give prominence to the concept of standard deviation (which they, more often than not, call "variability) in their section, "Measuring Portfolio Risk." 199. One of Glover's many criticisms of C.A.P.M. theory is its obsession with standard deviation. According to C.A.P.M. logic, the lower the standard deviation of a portfolio, the more successful that portfolio in terms of risk avoidance. Rather than measuring the success or otherwise of a portfolio in terms of risk adjustment, Glover asserts (op.cit. 140), with much good horse-sense, we think, that the only criterion which really matters when judging stock market investments is how they have performed in terms of terminal wealth, whereas, for C.A.P.M. purists, the realized outcome for the investments making up a portfolio is of no significance. 200. So far, we have touched upon the importance to C.A.P.M. theory of the standard deviation of an index, such as the Standard & Poor's 500-stock index in the U.S.A. or the H.S.I. for Hong Kong, as an indicator of risk. We explained how the standard deviation of such an index is constructed from the historic spread of outcomes over a period. As we were not told the standard deviation of the H.S.I., we had to make do with an illustration from the Standard & Poor's 500-stock index which, over a sixty year period, had a standard deviation of 21.2%. 201. Besides a whole index having a standard deviation, the constituent stocks of that index will each have its own standard deviation, calculable from its historic performance over a period. 202. Yet again, we must have recourse to the United States for illustration purposes, since we have not been provided with the necessary calculations based on Hong Kong stocks. 203. From the United States material to be found in table 7-3 at page 132 of Brealey & Myers (op.cit), we see that, from the ten United States' stocks selected, only three have a standard deviation lower than the 21.2% standard deviation of the Standard & Poor's 500-stock index, and we accept what Brealey & Myers say about most stocks having a higher standard deviation than the index. 204. We now set out that Table 7-3, not overlooking that it covers only the period 1981 to 1986, whereas the Standard and Poor's 500-stock index spanned a 60 year period. The difference in periods is of no materiality, so we gather, on the point that the standard deviation for most stocks is higher than that of their home index.
TABLE 7-3
205. We presume that, in Hong Kong, too, most companies constituting the H.S.I. must have a higher standard deviation than that for the H.S.I. as a whole. Were it not so, C.A.P.M. theory would be robbed of its most vital ingredient, since the central tenet of C.A.P.M., and the allied concept of Modern Portfolio Theory (MPT), is that the standard deviation of a portfolio of stocks will be less than the weighted average of the standard deviations of the individual stocks comprising the portfolio. At the same time, though, according to MPT, the rate of return on that portfolio will be a weighted average of the rates of return of the constituent stocks of the portfolio. MPT and C.A.P.M. thus hold out the promise of maximum return for minimum risk (See Glover, op.cit., pages 108 & 109). 206. On the point of diversification reducing variability, it is, in fact, a mathematical necessity that this should be so, unless the stocks making up the portfolio happen to be perfectly correlated. 207. Although combining stocks in a portfolio reduces risk, as measured by standard deviation, the element of risk which still remains cannot be ignored. That risk which remains, as reflected by the portfolio's standard deviation, is known by a variety of synonyms such as "undiversifiable", or "market", or "systematic" risk. We think that the best way of conveying the sense of this concept is to describe such risk as, "market/undiversifiable". 208. The risk which does get eliminated by combining shares in a portfolio is known as "diversifiable", "unique", "unsystematic", "residual" or "specific" risk. From that selection, we regard the words "unique risk" or "diversifiable risk" as equally suitable. We will adopt the description, "unique/diversifiable". 209. Market/undiversifiable risk relates to economy-wide perils which will make the market as a whole rise or fall, e.g. a recession or an "Oil Crisis", and is to be distinguished from unique/diversifiable risk which is confined to a particular stock. Unique/diversifiable risk could, adversely, include, say, a strike at one of such a particular company's plants, or, favourably, its winning a large contract. 210. In the case of a well-diversified portfolio, by which we gather inferentially from Brealey and Myers (op.cit., p.132), is meant ten to fifteen or more stocks combined, unique/diversifiable risk will have been more or less eliminated, the only risk remaining being of the market/undiversifiable variety. 211. The contribution from each individual stock in the portfolio to the portfolio's overall market/undiversifiable risk depends on each stock's sensitivity to market movements. That sensitivity to market movements by a stock is known as its Beta. 212. The index taken as representative of the market-in our case the HSI - is the reference point, and, by definition, its Beta will be 1.0. 213. One cannot stress too strongly how important as a benchmark is the Beta of 1.0 for a whole index. HSI, comprising as it does thirty-three stocks, is about as diversified a portfolio as one could reasonably hope for. For all practical purposes, there should be no unique/diversifiable risk remaining, so that the 28.59% nominal rate of return calculable for the HSI effectively represents market/undiversifiable risk only. When we come to look at the Beta for individual stocks viewed in isolation such as, for example, the Beta of 1.313 for New World Development giving rise to a nominal rate of return of 34.80%, it has to be borne in mind that the rate of return in such circumstances represents market/undiversifiable risk only, and that if the unique/diversifiable risk is taken into account, too, the overall rate of return for that single stock will be considerably higher. 214. For a stock in a portfolio, its Beta is measured on the basis of how the price of the stock responded to past market movements. Again, resort is had to historic data. 215. Beta is usually measured over a relatively short period, the usual period in the USA, apparently, being 5 years, judging from the examples in Brealey & Myers, whereas in our case, the calculations, done from data on Hong Kong quoted-companies, have been over a period of at least two years. 216. The mathematical technique employed to ascertain Beta for an individual stock involves regressing that stock's monthly price changes against the market's (i.e. the H.S.I.'s) returns over the same months during a chosen period. In our case, that period was a period of at least two years, terminating on 29th may 1989. Having plotted those monthly rates of return for the stock against the market's returns for the same months, one fits a line through the points (the "line of best fit"), its Beta being the slope. The slope of the regression line (the "line of best fit") correlates the rate of return for the stock with the market, and gives the regression coefficient, i.e. the Beta. Examples of how this is done are to be found in Brealey & Myers (op.cit. pages 175 et seq), and Glover, (op.cit, page 163). 217. Beta measures the amount by which investors expect a stock's price to change for every 1% change in the market, represented in Hong Kong by the H.S.I. 218. If a 1% rise in the market has resulted in a 2% rise in a stock, that stock will have a Beta of 2. Another way of putting that is to say Beta measures the amount a stock's price is expected to change for each additional 1% change in the market. 219. The Beta of a portfolio will be the weighted average of the Betas of its constituent shares. 220. A portfolio with a Beta of 2 will have twice the variability (i.e. standard deviation) of the market, as represented by the H.S.I. 221. Besides measuring market/undiversifiable risk, Beta performs another crucial function, namely, determining the Expected Risk Premium on a stock. As the Expected Risk Premium on a stock is in direct proportion to its Beta, it necessarily follows that a stock with a Beta, say, of 0.5 will have half the Expected Risk Premium of the market. Thus, in Hong Kong, using the H.S.I. as the base index, with a Beta of 1.0, and, an Expected Risk Premium of 19.83% for that index, a locally listed stock with a Beta of 0.5 will have an Expected Risk Premium of 9.915% (19.83% divided by 2) 222. Thus, according to C.A.P.M. theory, both of the following are proportional to Beta :-
223. The circumstance that both the required rate of return for a stock and its standard deviation are derived from that stock's price is seized upon by Glover to mount a powerful argument against the soundness of C.A.P.M. in theory, and its usefulness in practice. 224. In essence, Glover's attack on C.A.P.M. takes the form of pointing out how C.A.P.M. suffers from inherent circularity of argument, its two parameters - risk and rate of return - being defined, basically, in terms of the same phenomenon : a stock's price. As the risk measure, Beta, is a function of the share price, which is itself a function of risk, risk finishes up getting measured in terms of itself. 225. Glover's extended arguments exposing not only the fallacy underlying C.A.P.M.'s theory, but, also, its inadequacy, in practice, for finding a rate of return, can be found at pages 109, 110, 117, 130 et seq., 138 to 141, 146 et seq., 197 and 198 of his already-cited book. 226. We strongly suspect that Glover has done for C.A.P.M. what Copernicus did for those who believed the sun moved around the earth. 227. A further general point we feel we should explain at this point about C.A.P.M. is how the nominal discount rate derived from it can be turned into a real (inflation-free) rate by deducting the historic average inflation rate, such rate for Hong Kong being agreed between the parties in the present case as 7.1%. The method of deduction is agreed to be what is known as "geometric", employing the formula :-
228. There is a worked example of this in Mr Li's Report 42/05A, page 42, on the basis of Mr Li's nominal discount rate of 36.73% for SFI, and the agreed inflation rate of 7.1%. The real discount rate then becomes 27.67%, as follows :
229. For ease of understanding, we set out, by way of example, C.A.P.M. calculations on the alternative bases of Mr Li's Beta coefficient of 1.209, derived from the average Beta coefficients of five publicity-quoted Hong Kong property development companies, as a surrogate for SFI, and Mr Best's equivalent Beta coefficient of 0.8, based originally on the average of the Beta coefficients of the listed companies of the Industrial Sector of the Hong Kong Index, measured in terms of the H.S.I. (Although, ultimately, the Beta of 0.8 which Mr Best reached had the same workings and gave the same answer of 16.36% for the real discount rate, as shown below, his later 0.8, however, purported to be what is known as an "Asset Beta", whereas the earlier was of the "Equity" variety, but, for the time being, we will defer consideration of the "Equity/Asset" dichotomy, and will, meanwhile, simply refer, generally, to "Beta coefficients" or "Beta". 230. That example of C.A.P.M. calculations is as follows :-
231. A description of the relationships exemplified by the C.A.P.M. calculations, just shown, is as follows :-
232. Assuming for the purposes of illustration that 0.8 is the correct Beta corefficient for SFI, appropriate substitutions in the formula r = rf + B (rm - rf) are as follows :-
233. Geometric deduction from that of the historic inflation rate of 7.1% results in a real discount rate of 16.36%. 234. By various subjective adjustments, Mr Best whittled that down to 13%. That 13% was the discount rate Mr Best contended should be applied to SFI's cash-flows in the No-Scheme-World for the purpose of ascertaining goodwill. 235. As we have already noted; C.A.P.M. links up with modern Portfolio Theory (MPT) by contending that unique/diversifiable risk can be diversified away by holding a sufficiently wide variety of different stocks, say, ten to fifteen or more, according to Brealey and Myers, (op.cit. at page 132), in the one portfolio. 236. Because unique/diversifiable risk can be diversified away, it is. unimportant, according to these theories. All that is said to matter for portfolio purposes is market/undiversifiable risk, which cannot be diversified away. The market/undiversifiable risk of a portfolio will be a weighted average of the Betas of the stocks in that portfolio. 237. Even assuming C.A.P.M. were an unassailable theory, it was, nonetheless, never satisfactorily explained to us during the hearing how C.A.P.M., which claims only to concern itself with market/undiversifiable risk in the context of shares held in a portfolio, could be adapted for the purpose of coming up with a discount rate for SFI on its own, with its unique/diversifiable risk, and without any pretence it should be viewed as part of a portfolio. As pointed out by Brealey and Myers, (op.cit.132), "If you only have a single stock, unique risk is very important, ..." and at page 143 "... the effective risk of any security cannot be judged by an examination of that security alone". 238. Glover, too, (op. cit. 161) describes the position regarding a single stock in similar terms : "Thus, by holding a portfolio of stocks, some of the risk of the constituent stocks is diversified away and the investor can obtain a level of risk which is lower than the risk of a security with the same rate of return but held in isolation". 239. Such a lone stock will be exposed to the total risk, made up of the unique/diversifiable and the market/undiversifiable elements, as we have already noted in relation to the contrasting position of a whole index with a Beta of 1.0 representing market/undiversifiable risk only, since the unique/diversifiable risk will have been diversified away. 240. The topic of allowing for an element of unique risk in the rate of return required for a company isolated from a portfolio is touched upon in Exh R140, an article from the Investment Analyst of January 1989 by Dimson and Marsh, entitled "The Smaller Companies Puzzle", which explains that the rate of return on a publicly-quoted share is made up of three components, the first being the interest rate, which is the reward for the time value of money, secondly, a risk premium proportional to the market/undiversifiable risk (Beta) of the stock, and, thirdly, an unpredictable element, (the unique/diversifiable risk), which reflects factors such as the performance of the individual company relative to its industry's performance or the performance of companies of similar size. 241. As Beta measures only market/undiversifiable risk, we find ourselves in agreement with the government's submission that it was generous on the government's part to allow SFI to avail itself of C.A.P.M. which ignores unique/diversifiable risk. In the United States, so Brealey and Myers (op.cit.132) point out, most stocks are more variable (i.e. have a higher standard deviation) than the Market Portfolio (Standard and Poor's 500-stock index), and we know from Glover (op.cit.165 and 171) that, as a matter of mathematical necessity, the same must apply in Hong Kong in relation to the H.S.I., unless its constituent stocks are perfectly correlated. 242. In relation to the seldom-traded Beazer, the so-called comparable to which SFI regards itself as closest, such data as exist (Exh R146) shows that only 32.7% of Beazer's total risk is market related, the remaining 67.3% of the risk being unique/diversifiable. Beazer's Beta is just over 1 - 1.063. For other close comparables the position is shown by Exh R146, as follows :-
243. For New World Development, with its Equity Beta of 1.313, 76.8% of its total risk is market-related. 244. Where, as C.A.P.M. contemplates, unique/diversifiable risk is diversified away in a portfolio, such unique/diversifiable risk is of no signficance, but the same cannot be said of just one company's stock standing on its own, as happens with S.F.I. 245. As C.A.P.M. does not concern itself with stock volatility not related to the volatility of the stock market as a whole, it is curious that C.A.P.M. has been prayed in aid at all for present purposes. 246. In coming up with a discount rate via C.A.P.M. for SFI divorced from a portfolio setting, both accountants have found what is known in C.A.P.Mese as "the company cost of capital" (see Brealey & Myers, op.cit., p.173, 182-187), a concept which we now regard as misguided. Bascially, "the company cost of capital" is the rate of return supposedly required by investors in a company. C.A.P.M. gets prayed in aid for the necessary calculation. As we have already seen, the rate of return required by investors in, for example, New World Development, through the application of C.A.P.M. is 34% nominal. That 34% nominal, according to those prepared to swallow the rest of C.A.P.M., should be used by New World as the rate for discounting cash flows for new projects or for company acquisitions. Thus used, that 34% is characterized as New World Development's "company cost of capital". 247. We will encounter the "company cost of capital" again when we touch upon the related topic of "Asset Betas". 248. As we have already indicated, there are four Hong Kong listed-companies upon which the parties are agreed as the nearest-comparables to S.F.I. We now set out those four companies, with their agreed Equity Betas :-
249. SFI had second thoughts about the suitability of K. Wah for C.A.P.M. purposes, because, during the period over which Beta was measured, K. Wah's share price was affected by a host of factors unrelated to K. Wah's core business as a supplier to the construction industry (see Exh SF 219, pages 148-149, para. 42). 250. As a matter of degree, K Wah is even less comparable than the other closest comparables, and we are prepared to exclude it from our Beta averaging exercise. 251. Green Island Cement, too, attracted objections from SFI, too, on the basis that it had suffered from unusual influences, likely to affect its Beta such as shipping losses prior to 1986, it had merged with China Cement, and, in early 1989, had been privatized. As SFI request, we will eliminate that, too, although not much is then left to assist in our quest for a Beta. 252. The average of the Equity Betas of the two remaining closest comparables, Paul Y and Kumagai Gumi, works out at 1.2 (rounded). Without any adjustments, that, via C.A.P.M., gives a nominal discount rate of 32.56%, and a real rate of 23.77%. 253. Neither side adopted a Beta based on the average Betas of those two agreed closest-comparables. 254. Instead, based on Hong Kong listed companies, Mr Best consistently contended an appropriate Beta for SFI was 0.8, although he used two different routes to reach that conclusion. (See Exh SFI219, pages 52 and 149.) Through C.A.P.M. applied to the H.S.I., a Beta of 0.8 translates into a nominal discount rate of 24.62%, and real 16.36 (say, 16%). 255. With subjective downwards adjustments for a range of factors such as SFI's supposed advantages flowing from its relationship with the New World Group (e.g. a captive market, "cheap and reliable finance"), and the circumstance that the raw data of stock prices fed into C.A.P.M. represent minority interests, whereas the hypothetical sale of SFI was on the basis of full control, Mr Best consistently concluded 13% was the correct real discount rate derivable from C.A.P.M. 256. Mr Li, on the other hand, invariably adhered to the view that 28% real was the correct discount rate for SFI, after subjective upwards adjustments of a Beta of 1.209, based on five listed Hong Kong property companies (see Mr Li's Report 42/05F page 19), and 1.1, arrived at in the light of the Betas ("Asset", of which more anon) of the four comparables, Paul Y, Kumagai Gumi, Green Island, and K Wah Stones, plus New World (see 42/05F page 14 and 42/05F1, pages 1 and 2). With K Wah, Green Island and New world excluded, Mr Li's Asset Beta averages out at approximately 1.0. 257. Even when applied in the most favourable circumstances, we seriously doubt that C.A.P.M. has the power to give a sensible answer when resort is made to it with a view to finding a capitalisation rate for a single stock, and we have even less faith in it in the highly unfavourable circumstances of the present case where none of the publicly-quoted so-called closest comparables, either singly or collectively, has anything but the most tenuous similarity with the subject, SFI. 258. The "closest", Beazer Asia, according to SFI's submissions, has for its principal activities (see Exh SFI219 page 158) "design and installation of piling and foundation work, heavy civil engineering, marine work, prestressed concrete construction, geotechnical works and supply, and installation of specialist civil engineering materials." 259. During the period in respect of which Beazer's Beta was calculated, it acquired a 50% interest in a development site for a 16-storey residential building. Six months later, (but still within the Beta period), that was sold. It also had a 10% interest in a commercial development in Sydney. Its average gearing ratio over the period 1985 to 1988 was 23%. 260. That, the closest comparable, is so utterly different from SFI that we fail to see how anything would be gained from C.A.P.M.'s answer on Beazer, even if C.A.P.M. embodied reality. A fortiori, we despair of the even less comparable companies we are still considering, namely, Paul Y, and Kumagai Gumi, telling us anything useful about SFI as a result of giving them the C.A.P.M. treatment. The activities of that last mentioned pair of companies can be found usefully summarised in Mr Li's report, 42/05F, page 21, such information in its turn having been gleaned from Exh SF 219, page 153. 261. We now present extracts from that Report of Mr Li's, pages 21 and 22, showing the principal activities of the two companies just named, together with information on their market capitalization, earnings trend, and factors to be considered in determining SFI's Beta. As in Mr Li's Report, we have also included a column with information on SFI for comparison. We have supplemented the information in the column for SFI by including material from the submissions made on SFI's behalf regarding factors alleged to decrease SFI's risk in the No-Scheme-World for Bate purposes.
262. Based on the factors alleged to increase or decrease risk as between SFI and the two closest comparables, Paul Y and Kumagai Gumi, those acting on SFI'S behalf have sought to engage in an exercise adjusting SFI's Beta for differences between SFI and those two closest-comparables by a process analagous to the one used in land valuation, as in the present case, for example, where additions to, or deductions from the value of the subject land are made in the light of a comparable's characteristics. 263. While, for the purpose of land valuation, the technique of adjusting for the differing features of the subject land and comparables is established beyond all question, SFI's attempt to do something similar with the Betas of shares is totally novel. 264. Novelty by itself is no reason for disallowing SFI's attempt. The reason why, in fact, we consider this type of adjustment misconceived rests with the impossibility of reconciling adjustments of that nature with the most fundamental concept of C.A.P.M., namely, that C.A.P.M., and the Beta derived from it, purport to measure market/undiversifiable risk only, and not any unique/diversifiable risk. However, not even Brealey and Myers (op.cit., p.189), who are obviously state-of-the-art on Betas, know which characteristics of a share determine Beta. Bearing in mind that Beta measures market/undiversifiable risk only, it follows as a matter of logic that Brealey and Myers do not know which characteristics of a share determine its market/undiversifiable risk. Not knowing which characteristics determine the market/undiversifiable risk, neither can they know what determines the unique/diversifiable risk. If they knew which characteristics of a share determined the unique/diversifiable risk, presumably the remaining characteristics would be the ones which determined the market/undiversifiable risk. 265. Accepting for the sake of argument the correctness of the assertions tabulated for SFI as "Factors which increase risk", and "Factors which decrease risk", which of those factors would go towards determining SFI's market/undiversifiable risk, and which, unique/diversifiable risk? 266. Take, for example, the first two factors tabulated as increasing risk - "Fluctuations in steel bar prices relative to inflation" and "Increasing competition in scrap steel market" - and the first two tabulated as decreasing risk - "Local manufacturing ability" and "Niche for Special Lengths". Which of those relate to market/undiversifiable risk and to what extent, and which to unique/diversifiable risk? 267. As not even Brealey and Myers could tell us with any confidence, it does not strike us as unreasonable to suppose that the accounting experts giving evidence before us will not be able to help us either. 268. Besides the theoretical objections to trying to adjust the Beta of a stock without having any means of knowing which type of risks for it fall into the market/undiversifiable category, and which into unique/diversifiable, there was also a solid practical reason which made Mr Best unsuited to the task he attempted of trying to make adjustments for each separate close-comparable company, rather than performing a broad-brush averaging approach. As he was neither an expert on the steel industry, nor the construction/construction materials industry in Hong Kong, he was in no position to weigh their relative risks with a view to adjusting the Betas of companies in those industries. For example, with Mr Pong, a director of Shiu Wing Steel, saying in 1988 there was fierce competition in the trading of rebars in Hong Kong, Mr Best was simply not qualified to gainsay him, or to weigh the relative risks of the steel and construction/construction materials sectors. 269. A further difficulty in the way of an approach, such as Mr Best's, of trying to adjust the Betas of individual companies is that Betas are imperfect guides at the best of times, since they are based on a limited number of observations and are prone to large estimate errors (see Brealey and Myers, op.cit. p.178, 181, and 197), with the result that too much reliance should not be placed on such individual Betas. 270. Before the present case, we doubt that anyone has ever suggested C.A.P.M. and Betas were designed for the type of comparables - adjusting exercise SFI contemplates, and we see no sensible way of adapting C.A.P.M. and Beta for the purpose SFI has in mind. 271. Again, bearing in mind that Beta concerns itself with market/undiversifiable risk only, it is not likely to be helpful when what we want to know is SFI's unique/diversifiable risk. 272. C.A.P.M., Betas, market/undiversifiable risk and unique/diversifiable risk are not of interest in themselves, of course, and are only useful in so far as they help us ascertain SFI's "loss or damage" due to the resumption. The C.A.P.M. approach, which requires us to pretend that SFI was part of a diversified portfolio of shares where unique/diversifiable risk but not market/undiversifiable risk would be diversified away, to us does seem far removed from the issue of SFI's loss or damage on which we have to focus. 273. As we have already indicated, we do not think the Betas of the two remaining closest comparables, Paul Y, and Kumagai Gumi, are individually helpful in finding an appropriate discount rate for SFI. The best that can be done with them is to use them collectively for such guidance as they can give on an appropriate range for SFI. Their average Beta works out at 1.2. Applying C.A.P.M., that gives a nominal discount rate of 32.566% and a real rate of 23.77%. 274. When in the closing submissions on behalf of the government, it was conceded that Equity Betas have a track record, and are empirically based, we gathered from the context and the submission as a whole, the government was referring, firstly, to how the Beta of a whole index such as the HSI can provide a benchmark against which a rate of return on the shares of an individual unlisted company can be measured in a common sense sort of way, and, secondly, how the Equity Betas of the closest comparable companies might be used by way of collective guidance to help come up with a sensible answer on the rate of return for the subject. 275. There is certainly no empirical evidence of which we are aware supporting the notion of "the company cost of capital" for a single company outside a portfolio, and we will be greatly surprised if the empirical approach of observation and experiment ever shows that, in terms of investment outcome, "the company cost of capital "derived via C.A.P.M. has any validity. As "the company cost of capital" is conceptually so flawed, being in conflict with C.A.P.M.'s central idea of holding stocks in a portfolio with a view to eliminating unique/diversifiable risk, we dismiss it on theoretical grounds, too. 276. It is common ground that an answer produced by applying C.A.P.M. is open to modification on grounds of business judgment. 277. Long before acquaintance with the contents of Glover's Valuation of Unquoted Shares had opened our eyes to the theoretical objections to C.A.P.M., we had undergone total disenchantment with its practical value for our case. C.A.P.M. turned out to be no better than the occasion for a charade. Neither accountant impressed us as paying anything more than lip-service to C.A.P.M., which proved powerless to inhibit unlimited subjective manipulation of the numbers. 278. Perhaps, initially, we expected too much of C.A.P.M. We were warned by Mr Best, (in his Report 33/02, para.274), echoing Brealey and Myers' view (op. cit. page 175), that C.A.P.M.'s results stood to be overruled by business judgment. 279. What we had not expected was how generally - unhelpful C.A.P.M. would turn out to be for the circumstances of the present case. Invoking C.A.P.M. with Hong Kong data, Mr Best produced the answer 13% (real) for the capitalisation rate; using the same method and the same data, Mr Li's answer is more than twice that rate - 28% (real). 280. C.A.P.M., which beckoned initially as a possible deus ex machina for scientifically determining a discount rate for SFI, disappointed us by turning out to be yet another god with feet of clay. 281. As C.A.P.M. is, in our view, an unsustainable theory, attempts by Mr Best to play variations on the C.A.P.M. theme were inevitably doomed for want of a sufficient foundation. 282. In the context of the significance of debt to the rate of return for the two publicly-quoted Hong Kong closest-comparables, which the Tribunal has accepted (i.e. Paul Y and Kumagai Gumi), Mr Best introduced a different type of Beta : the Asset Beta. 283. Hitherto, the Beta of which we have spoken, has been of the Equity variety which measures the equity risk of a share. 284. According to Mr Best (SF219, page 145), citing Brealey & Myers (op. cit., Chapter 9, page 184 et seq.), shareholders bear not only the business risk of a company's real assets (i.e. tangible assets such as machinery, and intangible assets such as technical expertise, or trademarks; see Brealey & Myers op.cit, page 3), but, also, financial risk, to the extent that the company has incurred interest-bearing debt to finance those real assets. The more a firm relies on debt-financing, the riskier its common stock. We do not think there is anything controversial in the present paragraph. 285. At this point, we will look at the level of debt of each of the publicly quoted Hong Kong closest comparables. 286. A useful concept in the context of considering a company's debt is that of gearing, (also known as financial leverage), which is a measure of the extent to which the company's business is financed by interest-bearing debt. This measure is calculated for a public company by dividing the value of its interest bearing liabilities by the sum of the interest bearing liabilities and the company's market capitalisation. 287. The average gearing ratio of each of the Hong Kong closest-comparables was as follows :
288. The gearing ratio for Paul Y was agreed (see Exh R153), but not that of Kumagai Gumi. 289. The description of those companies by SFI in Exh SF219, page 145, para.24 is "highly geared". There is a degree of irony in such an observation coming from the SFI camp, bearing in mind SFI's own leverage in the period before the shadow took effect, its Total Liabilities/Total Assets Ratio being 1.47 for Financial Year 1980/1 and 2.52 in 1981/2. (See Mr Li's Report 42/05B at page 123). That ratio measures the portion of assets financed by creditors, a low ratio being usually desirable because it indicates a low fixed cost burden of interest on debt. 290. To us, the gearing of the pair of closest-comparables, as set out above, appears moderate. 291. There is no dispute that a factor tending towards a lower discount rate for the reincarnated SFI of the No-Scheme-World is the circumstance that it is to be regarded as debt-free for purposes of the No-Scheme-World D.C.F. valuation on an extinguishment basis. Consistently, from at least Exh.R78 onwards, the government has conceded that. 292. Controversy enters the picture from Mr Best's contention that the risk from a company's debt can be measured, more or less mechanically, by applying the following formula, extracted from Brealey & Myers (op. cit page 185).
293. As the theory would have it, the Asset Beta thus derived, will measure purely the business risk of a company, its financial risk having been separated out. 294. Applying this formula to data relating to the Hong Kong closest-comparable companies, Mr Best's idea was to use their resultant Asset Betas as indicators of an Asset Beta for debt-free SFI. 295. An illustration of the working of the formula, using Hong Kong data, was provided by Mr Best in Exh SF250B, as follows : "STEPS IN DETERMINING THE REAL DISCOUNT RATE 1. Determine equity beta coefficients (calculated by Compuserve). 2. Determine gearing. 3. Calculate asset beta coefficients by applying the following formula :
4. Apply asset beta coefficient to CAPM
296. At first blush, the formula might appear an attractive alternative to the sort of subjectivity inherent in Mr Li's opinion in Exh R78 that, the combined adverse effects of SFI's allegedly high operating leverage, and small size when weighed against the favourable effect of no debt, resulted in SFI having a nominal discount rate of 36.73%, which was 4% higher than the nominal discount rate (32.73%) for five of Hong Kong's largest publicly-listed property developers, (Cheung Kong, Henderson Land, Hong Kong Land, New World and Sun Hung Kai), worked out via C.A.P.M., on the basis of an average Equity Beta of 1.209. 297. In relation to "high operating leverage" this normally means that a company has a high ratio of fixed costs to variable costs. That is certainly not the case with SFI which has low fixed costs, far and away the greater part of its costs being variable in respect of such items as scrap and electricity. (See Exh SF219 at p.59). 298. Mr Li said he was using "high operating leverage" in the sense of not being able to change direction quickly. He pointed out that an industrial undertaking such as SFI, with its heavy investment in plant and machinery, and specialised workforce, cannot close down quickly or switch easily to another line of business. Such inability to change quickly is perceived by the market as adding to the risk of a company, according to Mr Li, and, hence, merits a higher capitalisation rate. We are not persuaded that Mr Li is right on that, and we see no justification for any increase in the capitalisation rate for SFI on account of operating leverage. 299. If, as we suspect, Mr Li was not right about operating leverage, it is easy to understand how he might have gone wrong on it in view of what Mr Roy Leung himself said on this topic in his First Affidavit, para.16.34. We quote :
300. We also draw attention to what Brealey & Mevers (op.cit., 190) say on this topic :-
301. Continuing our commentary on Asset Betas, we know that at least once resort has been made to them in the United States from the material at page 182 of Brealey and Myers (op.cit.) which shows that one Gerald A Pogue gave evidence to the U.S. Federal Energy Regulatory Commission (F.E.R.C.) on the Asset Betas of various industries including steel. For what purpose his evidence was used and with what effect, nothing is said. 302. One possibility is that Mr Pogue was giving evidence in relation to one of the applications described by Brealey and Myers (op.cit. page 53 and 54), where an electricity utility in the United States seeks to justify its price on interstate sales. We quote from Brealey and Myers :
303. Even if F.E.R.C. accepted Mr Pogue's evidence on Asset Betas on the one occasion of which we know, and even if it has accepted similar evidence on countless other occasions, we remain less than satisfied on the conceptual soundness of Asset Betas, and of C.A.P.M. generally, when it comes to the task of findinag a discount rate for a lone company. 304. At a practical level, we can appreciate the attraction to a body like F.E.R.C. of an apparently cut-and-dried formula for compensating electricity utilities. 305. So long as no one questions such a formula, it can be a convenient way of resolving what might otherwise be a complex dispute. Already in this case, we have encountered the not dissimilar situation of the Hong Kong government going along with the Average Wholesale Price Index ("A.W.P.I.") for the purpose of operating escalation clauses in its building contracts, but not being prepared to do so in the present case on the issue of the price of rebars before us. 306. After much probing, we were satisfied that the A.W.P.I. was not a satisfactory indicator for the prices SFI might reasonably have expected for its product in the No-Scheme-World. 307. Likewise, after prolonged investigation of C.A.P.M., and, inter alia, its suitability for fixing the discount rate for a single stock, we have been made aware of C.A.P.M.'s short-comings. 308. According to Brealey & Myers (op.cit. p.173, 175, 181-187), whom Mr Best appears to have followed, a company discounting its forecast future cash flows should do so by incorporating the company's Asset (rather than Equity) Beta when arriving at a discount rate (i.e. expected rate of return) through the medium of C.A.P.M. A discount rate derived in that way becomes the so-called "company cost of capital", to which we have previously made reference. 309. Following Brealey & Myers (op.cit. p.186), Mr Best adopts their position that a company's business risk, as reflected by its Asset Beta, remains unaffected, regardless of the amount of debt the company takes on (See C.C.S., Sect.9, App 12 : "Asset Betas"). As the company cost of capital is a function of its Asset Beta, it follows that the company cost of capital remains unaffected by the extent of the company's debt. 310. During the final submissions on SFI's behalf, we raised a query, (Transcript, p.3086), on how it could be, under the Brealey & Myers formula (op.cit.185) for finding an Asset Beta, that the Asset Beta for a company did not change, when the proportion of the company's debt to equity changed. 311. In the example given by Brealey & Myers (op.cit.185), Philadelphia Electric's mix is given as 46% equity and 54% debt. The Equity Beta is given as .51, and the Debt Beta zero. 312. The calculation is shown as follows :
313. In our query, we wanted to know why, if the mix became 36% equity and 64% d0bt, with the same Equity Beta of .51 and Debt Beta of zero, the Asset Beta did not become .18 (rounded), based on Brealey & Myers formula as follows :
314. Philadelphia Electric's company cost of capital (i.e. expected rate of return) is to be found by applying the following formula :
315. With the two different Asset Betas .235 and .184, but the same assumed risk free rate of 5.6% and market risk premium of 8.4%, the calculations are as follows :-
316. Thus, on the occasion Philadelphia Electric has the more debt (64%), the discount rate at 7.15% is lower than when it has less debt (54%), the discount rate then being 7.6%. 317. The corollary of such a state of affairs is the more debt a company takes on, the lower its cost of capital. If Asset Betas can give rise to such bizarre consequences, one is impelled to question whether they can represent reality. We ourselves, having sensed there was something rather peculiar about Asset Betas, were not surprised to find them coming under attack from Glover, (op.cit. 253 to 257), in the context of the so-called "company cost of capital". As he points out (op.cit.255), if a company were to keep on increasing its debt with a view to taking advantage of the correspondingly lower cost of capital which Asset Betas in conjunction with C.A.P.M. appear to promise, the result, ultimately, would be bankruptcy for the company. 318. One need look no further than SFI's history to be made aware of the crippling effect debt can have on a company, and of the absurdity inherent in any theory which embraces the notion more debt will mean a lower cost of capital. SFI's not inconsiderable debts from 1976 onwards brought about a situation where, ultimately, no one but its parent, New World, was prepared to make any further advances. 319. Far from answering the Tribunal's query, CCS, Sect.9, App 12, under the heading "Asset Betas", has done an exercise on the same basis as Brealey and Myers (op.cit.p.186), assuming Philadelphia Electric's Asset Beta remained a constant at 0.235. As a matter of simple algebraic substitution, in such circumstances the Asset Beta remains the same, regardless of the debt/equity mix. 320. In our view, it is artificial and contrived to assume the Asset Beta would remain constant. 321. As we have already explained, it is from the following formula (Brealey & Myers, op.cit 185) that the Asset Beta is to be calculated :
322. Applying that formula, as already described, with a mix of 54% debt and 46% equity, Philadelphia Electric's Asset Beta worked out at .235, and when the proportions are changed to 64% debt and 36% equity, the Asset Beta becomes .184. 323. [A minor point we mention, in passing, in relation to the CCS on this topic is the apparent error in para.13 in referring to "... 36% debt/64% equity ..." when what was meant, we presume, was "64% debt/36% equity".] 324. In other ways, too, Asset Betas turn out to create as many problems as they solve. 325. Far from proving the welcome equivalent of "bringing charcoal on a wintry day", Asset Beta's turned out to be more in the unhelpful nature of "embroidery on flowers", to quote a local proverb. 326. The following (from 42/05F, page 2) is what Mr Li had to say generally about Asset Betas :-
327. Mr Li's Report 42/04 subsequently became 42/05A, the relevant part of which, at page 36 et seq. gives Mr Li's general views on C.A.P.M. before the Asset Beta refinement was introduced by Mr Best. 328. To counter Mr Best's Asset Betas, Mr Li now played Debt Betas. It is only if Debt Betas can be taken as zero (which is what Mr Best has done for all three closest-comparables, including Paul Y, used by Mr Best for illustration purposes in SF250B), that the Asset Beta formula will get the opportunity mechanically to grind out an answer. 329. In Mr Li's opinion, the Debt Betas of the Hong Kong closest comparables were 0.3 or 0.4 (See Mr Li's Report 42/05F, page 37). 330. For Paul Y, for example, he thought its Debt Beta should be 0.4. With that one change, Mr Best's calculation from Exh SF250B gets re-worked as follows :
Apply Asset Beta coefficient to C.A.P.M.
331. Two principal grounds were advanced by Mr Li in support of his Debt Betas of 0.3 and 0.4 for the closest comparables, the first being the risks posed to lenders by interest rate volatility, and the second, the risk of default. (See Mr Li's Report 42/05F, page 12). 332. A factor likely to be linked with the risk of default, according to Mr Li, was the relatively small size of the closest-comparables. We think he was correct on that, and we agree also with the other factors he identified in his Report 42/05F, page 37 as relevant to the assessment of Debt Betas, namely, whether a company has a history of losses, the extent of gearing, and the specific risks of each of the companies. 333. Those specific risks for the two closest-comparables accepted by the Tribunal for the C.A.P.M. exercise, are correctly identified in Mr Li's Report, 42/05F, pages 37 and 38. For Paul Y, its relative smallness and history of fluctuating earnings increased risk; whereas for Kumagai Gumi, risk was lessened by its relatively large market capitalization and sound earnings history. 334. That Debt Betas can exist we know from American experience, as exclusively reported by Brealey and Myers who tell of Debt Betas of 0.2 to 0.4 in times of interest rate volatility, such as occurred in the U.S.A. in the early 1980's. 335. We greatly doubt whether anyone had ever thought Asset Betas, let alone Debt Betas, had any relevance to Hong Kong before Mr Best let the C.A.P.M. genie out of the bottle. 336. There has not been the faintest whisper of a suggestion of any data, hard or soft, existing about Asset or Debt Betas in Hong Kong. 337. Not only is there a total dearth of practical information about the role of these more exotic types of Beta in Hong Kong, but, also, no satisfactory theoretical underpinning was forthcoming. That can be gathered from a perusal of Mr Best's Transcript, from pages 1418 to 1424. 338. What was never adequately explained for us was why we were supposed to concern ourselves with risks to lenders, when our actual concern, presumably, should be with the risks to the business of the borrowing listed-company, which was only of interest as a surrogate. 339. Mr Best frankly admitted that he could not justify what he described as the "conceptual accuracy" of the Asset Beta formula, and he made it clear he was relying on the authority of Brealey and Myers. 340. In view of Mr Best's confessed lack of acquaintance with Debt Betas, beyond what appears in the pages of Brealey and Myers, we approached his evidence on the topic with some scepticism. 341. What he told us about U.S. corporate debt frequently taking the form of fixed interest bonds, whilst the borrowings of Hong Kong companies are almost invariably at floating interest rates from banks, struck us as probably correct. We think Mr Best is probably right, too, that, from a lender's point of view, fixed interest loans are more risky, and, hence, volatile than floating rate loans. Why that fact should set off a chain of cause and effect, starting off with Debt Betas, then on to Asset Betas, whence to C.A.P.M., and a rate of return for a public company serving as surrogate for the subject, we found elusive. We were left unpersuaded that the difference between the fixed form of U.S. corporate debt and Hong Kong's floating variety meant that Hong Kong Debt Betas should be zero. 342. On the point about loan default, we did not agree that the level of security for the two Hong Kong closest-comparables was such that any of the Debt Betas should be zero. The details of each of the pair's interest-bearing debt and property in Hong Kong are to be found in SF240. Bearing in mind the gearing of each, and the various types of property listed for each, we do not regard a zero Debt Beta for either of them as more likely than the Debt Betas assessed by Mr Li in his Report 42/05F, pages 37 and 38. 343. On Kumagai Gumi, we are satisfied that Mr Best was right in finding its gearing 26%, and Mr Li wrong contending it should be 21%. 344. The point turned on the level of Kumagai Gumi's interest bearing debt over the 1987 - 1988 period. Mr Li argued that "Cash at bank and in hand", as shown in his report 42/05F, page 30, should be treated as free cash, and used for the purpose of reducing the level of interest bearing debt and hence the gearing (see R147). The lower the gearing, the higher the Asset Beta. 345. Mr Best argued that, far from being free cash available generally for reducing debt, the sums in issue were used to conduct a separate part of Kumagai's business, namely, the business of investment. According to Mr Best, this "Cash at bank and in hand" was, in effect, part of Kumagai Gumi's investment portfolio, and was not available for reducing interest bearing debt. The argument can be followed in extenso in Mr Best's Transcript from page 1342 to 1354, and we find ourselves in agreement with the conclusion he reaches. 346. Having immersed ourselves in the arcana of Asset Betas and Debt Betas, it was somewhat dispiriting to learn from Mr Best, who triggered it all off with his late foray into Asset Betas, that probably only one of the originally proposed closest comparable companies - Green Island Cement - was affected, and that only by 0.1% of a Beta (Mr Best's Transcript, page 1422). In the end, only one of the two closest comparables accepted by the Tribunal - Paul Y - was affected and then only to the extent of 0.1 of a Beta (See R153). 0.1 of a Beta makes a difference of approximately 1.9%, real, to the discount rate calculable by C.A.P.M. 347. Even under the most favourable circumstances, C.A.P.M. is clearly not a precision tool. It is common ground that Equity Betas are liable to large estimation errors (on this, see Brealey & Myers, op.cit, p.181 and 197). The assessment of Debt Betas is highly subjective, as demonstrated by Mr Best's zero and Mr Li's 0.3 or 0.4 for the pair of closest-comparables. Asset Betas, inevitably, will share the imperfections of the Equity and Debt Betas, from which they are, in part, compounded. 348. With Mr Best still sticking to Debt Betas of zero, the Asset Beta formula became yet another collateral issue. 349. Whether Debt Betas should be zero is only one of serveral issues the Asset Beta formula brought in its wake. 350. Brealey and Myers, as far as we are aware, the only authority known to the parties for the Asset Beta model, are the first ones to point out that the formula is an oversimplification. 351. Offsetting tax advantages from company debt are mentioned by Brealey and Myers as a source of tempering the result reached by a strict application of their Asset Beta formula. True, Brealey and Myers are writing about the United States where company tax rates are approximately twice those in Hong Kong, but that does not mean the point they are making has no validity for Hong Kong. It is only that the effect will be less. 352. Another serious doubt concerning the validity of the Asset Beta formulation is whether the degree of risk generated by debt is, in fact, in direct proportion to a company's gearing, as the formula assumes (See Glover, op.cit, page 170). The absence of empirical evidence leaves the concepts of Asset and Debt Betas as nothing better than untested hypotheses. 353. Where, for example, a company has a modest amount of debt, say, a 20% gearing, we are by no means persuaded of any likelihood that the rate of return required by investors, and, hence, the discount rate, would necessarily increase, linearly, by 20%. Much would surely depend on what the debt was funding, and investor's perceptions of the Net Present Value of the venture to which such funds were being applied. 354. As the issue of new shares has the effect of diluting existing holdings, it is well known that shareholders frequently prefer a company's expansion to be financed by borrowing, particularly in times of low interest rates. 355. We find ourselves inclined to agree with Mr Li's opinion that moderate borrowing by a company will not of itself make shareholders perceive their investment as more risky. 356. At the end of the day, we were satisfied that Mr Li had directed himself correctly on the factors to be borne in minding for assessing the Debt Betas of the closest comparables and SFI. 357. As both sides have acknowledged, the level of those Debt Betas was a matter of judgment, and, having weighed the opinions of Mr Best and Mr Li in the light of the evidence, we thought that Mr Li was the more likely to be right on this. The Debt Betas we find are 0.4 for Paul Y, and 0.3 for Kumagai Gumi. 358. To test the results of C.A.P.M., as applied on behalf of SFI, Mr Li introduced, by Exh R117, what he described as a "Reality Test". 359. At the stage when Exh R117 was first put in evidence, namely, 24th August 1989, the Asset and Debt Beta concepts had not yet surfaced in the case. In due course, to take account of Asset Betas, and also for the purpose of making some general corrections to the real discount rates shown on Exh R117, SFI produced Exh SF1250A. 360. As we understand the position, it was common ground that the second column of SF250A correctly showed the real discount rates for the indices and companies listed. The Asset Beta column was contentious, because of the dispute over Debt Betas. 361. We now set out Exh SF250A : "R1l7 - REVISED Real Discount Rate
Basically, this Reality Test of Mr Li's was meant as an appeal to common sense. It compares the discount rate of 13%, real, claimed by SFI for itself on the basis of control with, say, the 20.1% for the whole H.S.Z. or 18.3% for China Light and Power on the basis of minority interests. The question being implicitly asked by this Reality Test is whether it is sensible to suppose that a reasonable investor would feel safer owning SFI in the No-Scheme-World rather than "buying the Index", or holding shares in China Light and Power. Again, would the sensible investor consider he was getting value, owning SFI where the required rate of return is 13%, when the figure for investing in Hong Kong Bank as a minority shareholder, namely, 10.8% on Mr Best's calculation, is only 2.2% (or 0.3% on Mr Li's calculation in 42/05F, page 41) better? 362. In 42/05F, pages 41 and 42, Mr Li gives brief descriptions of Hong Kong Bank and China Light & Power, confirming the common knowledge that, by Hong Kong standards, they are the bluest of blue-chips. 363. To pretend that, in the No-Scheme-World, ownership of SFI would have been almost as good as investing in Hong Kong Bank and safer than having shares in China Light & Power is the stuff of Wonderland. 364. SFI seek to contend that this Reality Test is not an independent test of results derived by C.A.P.M. We think it is, both at the theoretical and practical level. It is only if one is willing to suspend incredulity, and assume the validity of C.A.P.M. that the Reality Test (Exh R117) has to be ignored. 365. Unless and until there is a more persuasive body of theory than that encapsulated in C.A.P.M., we find ourselves driven to accept Glover's (op.cit, pages 146 and 170) view that there is, as yet, to the best of our knowledge, no quantative, objective measure of risk. Risk, in the context of investment, can be recognized and described, but not explained. Read carefully, Brealey and Myers (op.cit.) do not appear to differ from Glover on this. At page 140 of their book, they pour cold water on the notion of ascertaining risk simply by plugging numbers into a formula. Words to the same effect can be found at their page 175. At page 189, they lament not having a more fundamental scientific understanding of which characteristics of an asset are associated with high or low betas. They then go on to consider characteristics such as cyclicality, or high operating leverage, by which risk can be recognized, but go on to conclude (at page 191), "you cannot hope to estimate the relative risk of assets with any precision". 366. At page 882, they acknowledge that :
367. If all that ultimately remains of C.A.P.M. is, "the crucial distinction between diversifiable and nondiversifiable risk", and that is, "the main idea underlying the capital asset pricing model", we fail to see how now, or, in the future Brealey and Myers contemplate, C.A.P.M., or its replacement, will be of any assistance in finding a capitalisation rate for a lone company like SFI which, divorced from any portfolio, is exposed to the full brunt of both unique/diversifiable and market/undiversifiable risk. 368. In language we ourselves would not have chosen, although we concur in the drift of its meaning, Brealey and Myers (op.cit., page 884), make the point that, "Assessing project risk is therefore still largely a seat-of-the-pants matter". The same, we feel, applies to finding a capitalisation rate for SFI. Introducing the rigmarole of Asset Betas and Debt Betas into such a situation is simply inappropriate. 369. Then, Brealey and Myers go on to note, in relation to C.A.P.M. "... but there are many puzzles left, some statistical and some theoretical". To that we would add, "and some practical, e.g. how a stock, isolated from a portfolio, can come within C.A.P.M.'s embrace". 370. Lastly, on C.A.P.M., Brealey and Myers, (op.cit 884) utter a thought which puts them in the same idealogical camp as Glover, "The statistical problems arise because the capital asset pricing model is hard to prove or disprove conclusively". That is one of Glover's central propositions :.because Modern Portfolio Theory, including C.A.P.M., can be neither proved nor disproved, it is meaningless. 371. Approached from the stand-point of scientific method, C.A.P.M. theory, being neither verifiable, nor falsifiable, falls within the category : "unrestricted general hypothesis", and, hence, is devoid of meaning. The Efficient Market Hypothesis - one of the theoretical supports underlying C.A.P.M. - is likewise meaningless. 372. If C.A.P.M.'s proponents want it to belong to the realm of science, rather than that twilit world of pseudo-science to which such bogus theories as, say, astrology, or alchemy have been consigned, it must be subjected to the touchstone of scientific method. Drawing heavily on the ideas of Professor Karl Popper, the philosopher of science, as to what is and what is not scientific, that is what Glover (op.cit., pp.108, 115-117, 122, 137 and 138) has done. 373. In chapter 6 of his book, Glover shows how threadbare the Efficient Market Hypothesis (E.M.H.), in fact, is. In particular, how E.M.H. fails to qualify as scientific is to be found in his Chapter 6, at pages 115, 116, 117, 122, 124, 125, 127 to 130, and 132 to 141. From that same chapter 6, exposure of the fallacy underlying C.A.P.M. also lays bare C.A.P.M.'s unscientific character. 374. The whole of his chapter 7 reveals the shortcomings of C.A.P.M., both in theory and practice. 375. SFI's subjective claims to be like the closest-comparables, and its claims for Asset Betas and zero Debt Betas, together with its subjective reduction by three percentage points of the 16% result it got by its application of C.A.P.M./Debt Betas/Asset/Betas, has produced an absurd result. That absurdity is the point Exh R117 drives home by a time-honoured descriptive approach, in preference to C.A.P.M.'s new-fangled contortions. 376. Too much manipulation of concepts like P/E ratios or C.A.P.M., devised, hopefully, with a view to introducing an element of objectivity into private company valuation, (in our case, a proxy for SFI's loss or damage), is bound to be self-defeating. 377. The facile way in which data was selected on behalf of SFI, so that P/E ratios were always 7 or 8, and discount rates invariably 12% or 13%, did nothing to strengthen SFI's case. 378. On P/E ratios, SFI first of all claimed a multiplier of 8 on the basis of the average of the P/E multiples of an assortment of U.S., Malaysian and Singaporean publicly-quoted companies running mini-mills, which was then halved on the basis of the Dixon prescription (op.cit., p.176) when using a quoted-company as a comparable for a private company. (See Mr Best's Report 33/02 para 261.) 379. That halving then got dropped when Mr Best subsequently decided Kah Wah Stones was the most comparable. It's P/E (prospective) was between 8.5 and 12.52 during the period January to June 1987 considered by Mr Best, and again he stuck to 7 to 8 for SFI. See 32/03, paras 391 to 422. 380. In SFI 219 page 229 and 230 (replicated in 33/08, page 32), Mr Best selected Hong Kong's seven closest-comparable quoted companies, four of which were new issues, and had been given prospective P/E ratios in their prospectuses as follows:
381. For the other three companies - Green Island Cement, Paul Y Holdings and Kier Kin Sun (Beazer) - he devised a prospective P/E ratio by dividing each one's price in January 1987 by its price at the end of the accounting period next elapsing. Elsewhere, we have commented on the inappropriateness of such a method. The resulting P/E ratios were:-
382. For the averaging exercise, the inconveniently large Paul Y Holdings was excluded, and the habitual 7 - 8 P/E ratio declared. 383. Despite recanting on Dixon's opinion that the P/E ratio of a comparable quoted company should be halved for application to a private company, Mr Best still attempted to salvage what he could from the US, Malaysian and Singaporean quoted mini-mill companies previously referred to, by saying their P/E ratios still deserved to be halved because they were historic rather than prospective. Once again, the result was a P/E ratio of 7 or 8. 384. On each occasion of a finding of a P/E ratio of 7 or 8, there was a corresponding finding of a discount rate of 12% to 13%, whether based on US or Hong Kong data. 385. In launching its first C.A.P.M. exercise in Mr Best's Report 33/02, SFI based its Market Risk Premium of 8.3% on the data collected by Messrs Ibbotson and Sinquefield for the period 1925 to 1981 on the Standard & Poor's 500-stock index. 386. A Beta coefficient of 1.24 he thought appropriate for SFI, was derived by Mr Best from the individual Betas of four publicly-quoted U.S. mini-mill companies, and a similar Singaporean and a Malaysian company. It is now conceded by Mr Best that he was wrong to include the Singaporean and Malaysian Betas since those Betas were only applicable to the Singaporean and Kuala Lumpur market indices, respectively. Nothing turns on that mistake, as luck would have it, since the average relating to the four U.S. companies is not significantly changed by exluding the Singaporean and Malaysian companies. 387. The Beta derived from that averaging exercise was 1.120. See 33/02 page 167. For the purpose of purportedly paralleling his exercise of halving the P/E ratios of those same foreign quoted-companies in accordance with Dixon's ideas, Mr Best increased the Beta average from 1.12 to 1.24 (See SF219 page 15 for the full workings of this C.A.P.M. exercise). 388. Such a small increase made no sense in terms of Dixon's halving approach for adjusting a publicly-quoted comparable to fit a private company. 389. Increasing a Beta based on Standard & Poor's 500-Stock index from 1.12 to 1.24 translated into an increase in discount rate of about 1%. 390. To mirror what he had done in halving quoted-companies' P/E ratios, a more appropriate adjustment by Mr Best to the Beta would have been a doubling. 391. By making the adjustment he did, Mr Best managed to get a 12<% discount rate, which matched his P/E ratio of from 7 to 8. 392. A later C.A.P.M. exercise, based on different U.S. material, again contrived a discount rate for SFI in the 12% to 13% range: (see Mr Best's Report 33/08, page 22 and 50). 393. This time, an updated version of the Standard & Poor's 500-stock index was used with a Market Risk Premium of 6.9% (compared with 8.3% previously). Now the Beta coefficient was 1.1 (formerly 1.124). A novel move was including 3.8% as a "Small Stock Premium". 394. Those ingredients, when blended, gave the answer 11.94% (say 12%) for the real discount rate. 395. Various mixtures of Hong Kong closest comparables and indices, subjected to C.A.P.M. constructed from Hong Kong data, invariably managed to produce a discount, rate for Mr Best much in line with the U.S.-based C.A.P.M.exercises already described. 396. Firstly, he made his own selection of publicly quoted Hong Kong "Industrials", as follows, together with their Beta coefficients, calculated against the Hong Kong Index, (which is different from H.S.I.), (See 33/03 para. 440):
397. On the strength of his belief that all those Betas were less than 1.0, he considered himself justfied in maintaining his view of the discount rate he had derived from C.A.P.M., based on U.S. material, namely, 13%. 398. Not only was it conceptually misconceived to try to use Hong Kong Betas in conjunction with C.A.P.M. based on U.S. material, but he also got the Betas wrong. 399. The true position (extracted from Exh.R142) was as follows:-
400. Far from being less than 1.0, all of those H.K. Betas, except one, were way above 1.0. Moreover, a Beta of 1.0, in terms of the Hang Seng Index, gives rise to a market Risk Premium of 19.83%, compared with 8.3% for the Standard and Poor's 500-stock index up to 1981, and 6.9% if the period be extended to 1989. 401. Yet again, in Exh. SF219, page 50, para 11, Mr Best relied on his own original selection of Hong Kong Industrials and their wrong Betas in support of a Beta of 0.8 to be used in conjunction with C.A.P.M., worked out on the basis of the H.S.I. He also used the companies allocated to the Inudstrial sector of the Hong Kong Index to come up with a Beta of 0.8 (rounded up), correctly calculated against the Hang Seng Index. 402. For ease of reference, we illustrate the Hong Kong Industrial Sector and its Beta in terms of the Hang Seng Index from the simpler table of SF 219 page 152, rather than the earlier, more complex table, at page 55. "SHUN FUN IRONWORKS LIMITED BETA COEFFICIENTS OF ALL COMPANIES INCLUDED IN THE INDUSTRIAL SECTOR OF THE HONG KONG INDEX 403. Weekly 1.3.77 or date listed to 29.5.89
404. This Beta of 0.8, (rounded up), is of the old-fashioned Equity variety. At the stage in the evidence when the Industrial Sector of the Hong Kong Index was introduced, Asset Betas had not yet been unveiled for us. 405. With a Beta of 0.8 applied to C.A.P.M. based on the H.S.I., the calculation produces a real discount rate of 16.36. (See Exh. SF 219 p.106). 406. That divergence from the familiar 12% or 13% was quickly corrected. Suddenly, we were reminded that SFI had benefits accruing to it from its relationship with New World such as a captive market and "reliable and cheap" finance. Mention was also made of C.A.P.M. being based on minority interests, whereas a discount rate appropriate for SFI should be on the basis of control (Exh. SFI 219 page 108). The result of these factors previously ignored in the C.A.P.M. exercises is that 16% becomes 13%. 407. When the argument shifts to the specific Hong Kong closest comparables - Beazer, Paul Y, Kumagai Gumi and Green Island Cement - the Beta recommended by Mr Best for use with C.A.P.M. is still 0.8, although, this time, it has been reached with the help of Asset Betas plus also the Equity Beta of the Industrial Sector of the Hong Kong Index. (See Exh. SFI 219, pages 192 to 199, and Mr Best's Report 33/08, para. 30 et seq, and page 50). 408. By way of general comment on D.H.S.'s employment of P/E ratios and C.A.P.M., we do not think it unfair to say that, figuratively, whichever way they deal the cards, the answer always turns out the same - a discount rate of 12% to 13%, or its reciprocal, a multiplier of 8. 409. How there was a prefigured result to this number-juggling is explicitly acknowledged in one instance. What we have in mind is the way D.H.S. reacted to the situation brought about by the Tribunal's pressing for specific Hong Kong comparables, instead of the Industrial Sector, as a surrogate for SFI. That prompted Mr Best to have resort to what Mr Carnwath described as a "wild-card", namely, gearing. 410. This is how Mr Best put the matter (in SF219, page 145, para. 23 - 24):-
411. Instead of working backwards by elaborately selecting data and processing it in such a way as to come up with a "discount rate which in Mr Best's judgment is reasonable", why not side-step the charade, and simply tell us what, in Mr Best's opinion, was a fair figure? II. RULE OF THUMB METHODS a. PAY-BACK PERIOD 412. On the basis of Mr Li's contention that the correct nominal rate for discounting SFI's future earnings is 37%, the implied pay-back period would be the reciprocal of that discount rate, namely, 2.7 years. 413. It is common ground it would take about 2 1/2 years to build a new steel plant of similar capacity to SFI's. On top of that, there would be a further 4 years by way of learning curve to bring such a plant up to full capacity, so that approximately 6 1/2years would be required altogether. 414. With a nominal discount rate of approximately 20%, implying a 5 year pay-back period, Mr Best contends his discount rate is the more reasonable. 415. The rate of return on any investment will be determinded on the basis of the risks of that investment compared with the risks of the innumerable other investment outlets for an investor's funds. (See Glover, op.cit., page 93). 416. If one compares Mr Li's version of SFI's rate of return, and consequential pay-back period, with that of the Hong Kong publicly-quoted closest comparables, as is done in Exh. R151, a pay-back period of 2.7 years for SFI is not out of line:-
417. Another company with a pay-back rate not dissimilar to SFI's 2.7 years is its parent, New World, which has a nominal discount rate of 34.8% (and a consequential pay-back period of 2.87 years), based on Mr Best's own calculations: (see Exh R133). 418. It might well be that a larger mill with higher technology would have risks and profitability commensurate with a longer pay-back period - If the figures were appropriate, the pay-back period could be five years or more - but, as things stand, SFI's position in the hierarchy of risk and reward is such that, we think the correct nominal discount rate to apply to it is 33%, implying a pay-back period of three years. 419. To adopt Mr Best's approach would be the equivalent of letting the tail wag the dog: it is not legitimate to ask how many years it takes to build a particular factory, and then to declare the discount rate to be the reciprocal of at least that number of years. Because it would take 8 years, say, to build an old-fashioned blast-furnace, would not mean its cash flows should be discounted at 12 1/2%. 420. A variation of the pay-back theme emerged from the evidence of Mr Stewart Leung, New World's director responsible for its subsidiaries, including SFI. 421. He explained (his second Affidavit, paragraph 8) that, from his substantial experience of buying businesses and investing in Hong Kong, he thought Mr Li was wrong in coming up with a discount rate resulting in such a short pay-back period as 2.7 years. He instanced how, for example, in developing hotels, he estimated the return would take over 10 years. (His transcript, page 180, line 8). 422. Whatever the position might be in relation to hotels, we think he has been over-sanguine in his view over SFI. In part, his opinion has no doubt been coloured by what we regard as the over-optimistic projections made for SFI by DHS. Moreover, we finished up with little faith in Mr Stewart Leung as a witness, since he left us with the impression he tailored his evidence to whatever he thought would best aid SFI's case. 423. In the light of what we know about SFI's history, and our findings on its prospects in the No-Scheme-World, we think SFI's correct position in the spectrum of available investment opportunities competing for the investor's money would be somewhat within the vicinity of a 33% nominal discount rate, implying a 3 year pay-back period. 424. In support of the proposition that for heavy industry, such as SFI's, a considerably longer pay-back period than 3 years was appropriate, the Australian case of Eastaway v. The Commonwealth [1950 - 1951] 84 CLR 328 was cited on SFI's behalf. There, the resumption of the plaintiff's land in Sydney in 1949 had resulted in the extinguishment of its business of engineers and ironfounders conducted on the land. For the purposes of that business, the plaintiffs had had a large quantity of machinery, plant and trading stock on the resumed land. 425. In upholding a lower court's capitalization rate of 6%, the High Court of Australia made the following observation, (at page 340), on which SFI now relies,
426. What might have been a reasonable discount rate for a business extinguished in Australia in 1949 will not, in our view, necessarily have any bearing on the rate for a Hong Kong business extinguished in 1986/7. 427. Certainly, the High Court's observation, relied on by SFI, is not a proposition of law, and there is reason to doubt whether it even embodies current thinking on valuation. The idea that a special valuation approach should be adopted for what have subsequently become known as "smokestack" or "sun-set" industries, such as, say, ship-building, or coal-mining, on account of the nature of their assets, has not stood the test of time. 428. Heavy investment in plant and machinery is of itself no guarantee of a low capitalization rate these days. It is, we think, a fallacy to imagine that normal considerations of risk-reward are suspended when heavy industry has to be valued (See Glover, op.cit., pages 237, 242, 260 and 264). 429. In any event, we do not set too much store by a method as crude as "Pay-back", the limitations of which quickly become exposed when, for example, as in the present case, one is working in constant dollars. Then nominal discount rates need converting into real. As we show later in the present section, the real (i.e. inflation-proofed) equivalent of 33% nominal is just under 24%, suggesting a longer "Pay-back" period. b. CATEGORIZATION (as per Schilt) 430. There is a fairly long history in the United States of guide-lines on discount rates appropriate for the acquisition of different categories of private company. On this, we draw attention to the article by James H Schilt, "A Rational Approach to Capitalization Rates For Discounting The Future Income Stream of A Closely Held Company", in Mr Li's Report 42/05D, page 19, third column. 431. Such an approach does not seek to conceal the reality that there is a fairly arbitrary element in selecting discount rates. 432. For each broad category of business, a discount rate is assigned on the basis of the risk perceived by the author for that category. 433. By this method, the discount rate for a category is added to the risk free rate to arrive at the rate to be applied in discounting the company's future earnings. For Hong Kong, the agreed risk free rate is 8.76%. 434. The article does not overlook that stock exchange prices are based on sales of minority shares, and that premiums for control are often substantial. 435. Schilt's proposed discount rates are clearly intended to apply to the acquisition of whole private companies. 436. We now set out his table (42/05D, page 20), showing his categories of company and recommended discount rates : "Risk Premiums For Discounting Projected Income Streams
437. We share Mr Li's view that the correct category for SFI would be No.4, under "Larger established businesses that are highly cyclical in nature .... future earning may be expected to deviate widely from projections." 438. A somewhat similar arbitrary approach is to be found in Glover (op.cit. p.230) who explains that, in England, private companies, in his experience, are generally sold on the basis of a capitalisation rate ranging from 20% for profitable, established, well-managed medium-sized companies, to around 30% for small, well-managed companies. Those percentages Glover proposes are based on pre-tax profits, whereas in our case all calculations have been done on the basis of after-tax profits. 439. We do not think SFI was well managed, and consider that on the Glover approach a nominal discount rate of over 30% would be appropriate for SFI, after taking into account that, for discounting purposes in the present case, SFI's profits have consistently been treated on an after-tax basis. As SFI only started paying tax in the No-Scheme-World in Financial Year 1991/2, the point is only of any significance from then onwards. 440. Needless to say, in many respects, the position in Hong Kong will not be the same as in the United States or England. Almost certainly, a higher discount rate (or lower number of year's purchase) will be appropriate for Hong Kong where risk levels, as reflected by stock-market indices, are among the highest in the world. Asian stock-markets generally reflect perceived levels of risk far higher than for the United States or England, and within Asia, Hong Kong's stock market is regarded as one of the riskiest, if not the riskiest. Once again, we refer to the Articles, The Risk and Return of Investing in the Far East Emerging Markets by Dr Ho Yan Ki (Doc 42/05A pp.66 to 70); and "Investments", by Jacob & Pettit, 42/05D, pages 17 and 18. III. THE JUDGMENTAL APPROACH 441. Answers derived by means of all or any of the approaches we have described are no more than aids in the quest for a result squaring with common sense. Only a result judged reasonable can prevail, but, all the same, one should not overlook that the reasoning process has an important role to play in ascertaining what is reasonable. 442. Of the various methods put forward as aids to assist the Tribunal in arriving at an appropriate capitalization rate, only two, in our opinion, commend themselves as in any way helpful in the circumstances of the present case. 443. Firstly, there is C.A.P.M. C.A.P.M., in our view, is, at best, marginally useful, because it is such a flawed model. Perhaps something can be salvaged from C.A.P.M. using Hong Kong data to find Equity Betas for the companies we have acknowledged to be the closest comparables - Kumagai Gumi and Paul Y. As we have previously indicated, their Equity Betas can be averaged with a view to affording collective guidance to the Tribunal. With an average Beta of approximately 1.2, the real required rate of return for those closest comparables collectively is about 24%, and nominal 33%. 444. Seen in the context of the real rate of return, based on minority interests for the HSI, being approximately 20% to cover both market/undiversifiable and unique/diversifiable risk in the light of the circumstance that H.S.I. constitutes a well diversified portfolio of thirty-three stocks and the real rate of return for New World on its own being approximately 26% for market/undiversifiable risks only, we think a fair real rate of return on the basis of control for SFI measured against such benchmarks is 25%. 445. The second method, and the one we found far and away the most helpful was the somewhat crude method described by Schilt of classifying private companies in rather broad categories. On the view we took, the correct slot for SFI was category 4, "Larger established businesses that are highly cyclical in nature", the appropriate risk premium, based on American material for such companies being a nominal 21 - 25%. If one adds the Hong Kong risk free rate of 8.76% to that, the range becomes approximately 30 to 34% nominal. A real rate of 25% would not be too far out of line with those nominal rates. 446. An even cruder form of categorization is the one referred to in Glover (op.cit p.230) of valuing private companies on the basis of capitalisation rates from about 20% for medium sized, well-managed companies to about 30% for small, well managed companies. On that basis, too, a capitalisation rate of 25% real for SFI would not be too far out of line. 447. Although those rule of thumb approaches of Schilt and Glover are based on the position in the USA and England, respectively, that, if anything, works to SFI's advantage, since the risks of doing business in Hong Kong, as reflected in rates of return generally, are higher than for America or the United Kingdom. 448. We think the fair rate at which SFI's real cash flows should be discounted is a real rate of 25%. That rate is meant to reflect the risks as they would have been perceived on the agreed valuation date of 19th January 1987 when SFI vacated the site. As at that date, SFI would not have been perceived by the market in the No-Scheme-World as one of the brighter jewels in New World Development's crown. The contract price for rebar, based on the HH/SWS surrogacy, was $1,754 per M/T in January 1987. See Exh SF216, page 145. Allowing the usual loading of 1 1/2% for special lengths results in a figure of $1,780 per M/T, which is what, in our view, SFI might reasonably have been expected to earn per ton from contracts for the sale of its rebars in the No-Scheme-World. That is a rather modest figure compared to the contract price of the same surrogate, with the same 1 1/2% loading, in the preceding three Financial Years. We set out the figures, our source being Exh SF216, p.143 and 144 :
449. Thus, it can be seen that rebar prices were on a downward trend, even in nominal dollars, at the valuation date of 19th January 1987. 450. In fact, one had to go back as far as 1982/83 to find prices even in nominal dollars, as low as in January 1987. With reflators applied, the picture looks even worse from SFI's point of view. 451. In 1986/7, SFI became marginally profitable for the first time since 1974/5, its total profits for 1986/7 being $6.506 million. SFI was still not out of the woods completely, since it lapsed back into unprofitability the following year, and did not properly hit its stride until 1988/9, from which time it became set on a course of inflation-proofed earnings $34.543 million per year before bank interest or tax. 452. Judgmentally, Mr Best has consistently stuck to a real discount rate of 12 to 13%, whereas Mr Li has always kept to 28%. 453. After weighing the evidence and opinions of that pair of experts against the background of all the evidence in the case, we think that a discount rate of the order propounded by Mr Li is the more appropriate, and that Mr Best's 12 to 13% is way too low. 454. We are not, however, prepared to accept Mr Li's proposed figure of 28% unreservedly. We were unhappy with his evidence about SFI having high operating leverage, and have considerable doubts as to whether high operating leverage should have been included as one of the factors which prompted him (in Exh. R78) to add a further 4% nominal to the 32.73% he first calculated via C.A.P.M. for SFI on the basis of five major property companies as near comparables. 455. As we do not know how much of that 4% should be apportioned to the supposed high operating leverage, and how much to other factors? We lack a rational basis for reducing it, and feel, in such circumstances, that the additional 4% should be wholly disallowed. 456. Such disallowance means re-working the formula for converting nominal to real rates :
With the necessary substitutions, as follows, (1.3273/1.071) - 1, the real rate is 23.93%. 457. To avoid any pretence to spurious accuracy, we regard 25% real as the appropriate discount rate for SFI. SECTION V : LAND VALUE LARNE VALUE AS PART OF THE VALUE OF THE WHOLE 458. In presenting their cases both parties have itemised each of the various elements of the total compensation amount under various Heads. The value of the site on which the SFI mini-mill was erected was one such Head. Both parties considered this Head as the value of the land on a formed bare-site basis, and we draw attention to our being conscious of its forming part of the value of SFI's Junk Bay steel-mill business as a whole. That is to say that the value of that business as a whole includes various elements such as land, buildings, plant and machinery and goodwill. 459. We agree with government's submission that the Ordinance (section 10(2)(a)) requires that a separate valuation of land and buildings is necessary and this should, strictly speaking, include fixed plant as well. (See our Section VI : PLANT AND MACHINERY) 460. Section 10(2)(a) states :-
461. The user restriction under the Crown lease limited development on the site to shipbreaking and other industries operated in conjunction with this use. It is agreed that this includes the use to which the site was put as a mini-mill. 462. Therefore, provided that a mini-mill is the most valuable use to which the land might have been put - and there appeared to be agreement that it was - the value of the land must be influenced by economic factors relating to the steel-making industry in Hong Kong, at or about the valuation date (i.e. the date of resumption : 30th July 1986). 463. Assessments based on expected profits have been put forward by both parties in relation to the value of the goodwill of SFI's business at Junk Bay. However, without any actual sales of steel mills, the parties have agreed to base the value of the buildings (as well as the plant and machinery) on the depreciated replacement cost method. This method requires that an appropriate depreciation factor be applied to the estimated equivalent replacement cost new as at the date of valuation. This factor is expected to reflect the age of the buildings and plant as well as the degree of wear and tear and obsolescence. On this basis, in the course of the hearing, the parties have reached agreement on the value of the buildings at a figure of $26 million. The value of the plant and machinery, which has been assessed on the same basis, remains disputed. 464. The only other item making up the value as a whole is the value of the goodwill, which was put to us, as only existing if the value as a going concern exceeded the value of the tangible assets. This item is also disputed. Government submits that no additional amount for goodwill exists, as the value of the business was less than the value of those assets. However while government says the valuation-of-the-assets-approach may result in an over-generous measure of value, it has not sought to contend that the overall valuation should be less than the value of the assets. MATTERS AGREED 465. As the case progressed, the value as agreed on a number of significant matters which are relevant to the value of the land. Before embarking on any examination of the rest of the evidence it would be as well to mention these agreements at this stage. 466. The road access from Clearwater Bay Road along Anderson Road and Po Lam Road to the subject site at the date of resumption was described as being narrow for parts of its length with passing bays in places. Most of its surface was tarmac or similar material. It was agreed that vehicles including heavy lorries could and did traverse the whole length of the road but that its general suitability for such traffic was below average. There was also access to Kwun Tong by way of Junk Bay Road which linked with Po Lam Road which is described as above. It was agreed that the submitted aerial photograph dated 18/5/1985 probably showed the road as it existed at the relevant date. 467. There was agreement that industrial land values rose over the period from the end of 1985 to the beginning of 1987 and that this rise was greater in the second half of the period than in the first. For the period from early 1986 (February) to the date of reversion (30th July 1986) the rise in industrial land values, was about 5%. Mr Doran, SFI's expert witness on land matters, through his counsel, was not prepared to say that a 5% rise was appropriate for waterfront industrial land such as the subject site, but he was not in a position to show that it had been any greater than the 5% adopted by Mr Brown, government's land expert. Mr Brown based his assumption largely on a Jones Lang Wootton report, showing that the price of industrial properties, based mostly on the sale of flatted factories rose by about 5% over the period in question. This matter is relevant to the use of sales evidence prior to the valuation date of 30th July 1986. 468. As already mentioned, it was also agreed that the value of the buildings based on a depreciated replacement cost approach was $26 million. We were told that this amount included an allowance for interest on the depreciated cost of the buildings, and was based on an assumed 2 year construction period by allowing for interest for the mean of 1 year to allow for periodic payments over the total 2 year term. However it transpires that this agreement was based more on compromise than, calculation. This agreement is relevant to the claim for Interest on Land Value. LOCATION VALUE 469. The exercise of determining what the land element is worth as a separate entity would have been more straight forward if other sites, which were of about the same area and were subject to the same or similar restrictions on title, had changed hands around the date of resumption. If this had been so, the prices paid would be expected to take into account most of the factors mentioned and require only minor adjustment for any differences such as location or other matters which might have had a bearing on value. 470. In Hong Kong, leases of land are auctioned at a premium. These leases usually contain comprehensive conditions which determine the use to which the land may be put and, often, the extent to which the land may be developed. A point made by both of the parties was that in most countries industrial land would not be subject to such restrictive conditions as those in the instant case or, perhaps, in Hong Kong generally. In the United Kingdom, for example, a buyer of industrial land could be expected to be drawn to a location which, as well as suiting his price, would also be suitable for his particular purpose, but not strictly limited to that purpose. 471. For a purchaser wishing to erect a mini-steel-mill, the site would have to be obtained at a price which permitted a reasonable return after allowing for the costs associated with optomising the land to produce the finished product. However, the price he eventually has to pay would be influenced more by what other similar industries might be prepared to pay in that locality. 472. In respect of the subject site we believe that such a localised industrial value existed for the Junk Bay area, but the limited user restriction may well have led to the economics of steel production having a greater influence and being more relevant than in places where there is a more general level of value for suitable industrial land. RESQMPTION SETTLEMENTS 473. Similar types of industry still tend to develop in the same or similar localities in Hong Kong, but this is mostly due to the lease conditions or Town Planning restrictions imposed by the authorities. Noxious industries, and a steel-mill, (particularly one like SFI's without pollution control devices) might be expected to fall into this, are likely to be prohibited from certain areas, so they will usually tend to be found concentrated in particular localities. The subject site, being restricted to a use in conjunction with shipbreaking, adjoined similar sites being put to similar uses. Those sites were also resumed at the same time. 474. The result is that the only transactions involving similar sites at or around the valuation date were under compulsory powers relating to the same Scheme. The agreements on land value between those dispossessed owners and the Crown have been relied on to some extent by Mr Brown. He acknowledged that such settlements should be viewed with caution but they might "provide assistance in the absence of other suitable comparables". 475. Mr Brown cited this Tribunal's decision in the case of Tang Chun Ki v Director of Engineering Development MTR 2/84 in which it was stated, "The extent to which weight will be given to the resumed price, in resumptions by agreement will depend on the evidence. In some circumstances a Tribunal may give them considerable weight". It was further observed, "As with comparables based on actual market sales between individuals, it is basically a question of weight to be determined in the light of all the surrounding circumstances." and, "In the case of some resumptions contemporaneous payment of ex-gratia sums and housing and other entitlements may, of course, be complicating factors." 476. Mr Doran who valued the land for the claimant considered that Mr Brown failed to present the full picture with respect to these settlements. The Crown's offers included an ex-gratia amount in addition to the value it placed on the land. We were told, and we accept, that this ex-gratia amount related only to the land element of the total compensation. Mr Doran showed that the total figures consisting of land value plus the ex-gratia payment were between $1,475 per sq.m. and $1,507 per sq.m., when the land value element was only between $400 per sq.m. and $500 per sq.m. Although he did not seek to rely on these payments, he was of the opinion that the figures supported his valuation of $1,500 per sq.m. rather than Mr Brown's of $450 per sq.m. 477. These ex-gratia payments are outside the provisions of the Ordinance, and would therefore only be available to the dispossessed owners if they agreed to the Crown's offer. In reality, any owner will be more concerned with the total amount offered, and is unlikely to bother himself regarding how much of that total has been apportioned by the Crown to the land. Moreover, even where such a land owner considers the land value alone to be greater than the total offered by the Crown he will frequently settle so as to spare himself the trouble of a disputed claim. If he is inclined to settle, he will obviously want it to be on the basis of including the ex-gratia payment, so that government's threat to withdraw the ex-gratia element after a certain date will act as a strong inducement to settle before then. 478. As the ex-gratia proposal can play no part in any proceedings before the Tribunal, the dispossessed owner will be faced with the dilemma of having to contest, in the Tribunal, only the lesser amount which the Crown determined as the market value relating to the land element under the provisions of the ordinance. 479. With settlements such as these, it would be unrealistic to put much faith in the amount determined by the Crown as the market value of the land under the Ordinance when the real concern of those resumed would be that the total, including the ex-gratia amount, was either above or at least in the region of what their idea was of the value of the land. Accordingly, we consider that these settlements are unlikely to evidence the market value of the subject land. VALUATION PRINCIPLES ADOPTED 480. There is no other evidence which could be described as direct comparisons of sales of similar sites. Hence, the valuers had no choice other than to embark on exercises of detailed analysis of the sales of other sites which permit industrial use, in an attempt to relate them to the subject site by making appropriate allowances. There was a variety of lease conditions governing the permitted developments of the compared sites, making any conclusion as to the value of the subject site difficult in the extreme. Because of the unpromising nature of the material, the valuers could not, in practice, avoid an element of conjecture and were not always able to abide by pure valuation theory. 481. By relying principally on sales of sites permitting industrial use, the object of the exercise for both Mr Doran and Mr Brown was to show what the subject site might have been expected to realise as a vacant but formed site should it have been sold on the open market on the 30th July 1986, which is the date the land reverted to the Crown. This approach accords with section 12(d) of the Ordinance which states :
482. The evidence derived from the industrial sales on which the valuers relied, can only be described as a series of very general indicators as to what that figure might have been. It is not surprising that the opinions of the two valuers were very different. Mr Doran valued the land at $54 million based on $1,500 per square metre, while that determined by Mr Brown was $16,208,000, based on $450 per square metre. 483. Mr Doran included, under his claim for land value an amount of $10.157 million as representing interest on land value. The total claim under land value is therefore $64.157 million compared to government's figure of $16.208 million. "DIRECT" COMPARABLES AND "CHECKS" 484. The claimant has pointed out that the subject site was ideally suited to the purpose to which it was put. It had a site area of 387,700 square feet (about 36,000 square metres) and sea access along its eastern boundary which consisted of a sea wall from which Shun Fung was able to ship in raw materials and take away finished products. As Mr Brown pointed out, this water-frontage was an essential requirement for an industry associated with shipbreaking. 485. It had an adequate water supply for industrial purposes from a stream on the adjoining hillside, and was supplied with its electricity requirements by China Light and Power. An electric substation was located within the site on the road boundary. Although the adjoining road was formed, it had never been properly surfaced. Road access, however, was adequate for the claimant's requirements. The site had been formed in compliance with the Sale Conditions when it was purchased in 1962 and was a level site. 486. Mr Doran and Mr Brown sought sales of similar land with which to compare the subject site. As we have already mentioned, none of the sales on which they were obliged to rely could really be described as an ideal direct comparable. 487. Mr Doran used as his prime basis an analysis of what he described as reasonably direct comparables backed up by analysis of Crown land auctions of industrial land as a check. Mr Brown's approach was much the same. In many cases both valuers relied on the same sales evidence. Mr Brown, however, put more emphasis on Crown land auction sales and used them as direct evidence of value. 488. Mr Doran also looked at the industrial land provided by the Hong Kong Industrial Estates Corporation at its estates at Tai Po and Yuen Long which are sold to selected industries at "special" prices based on the cost of producing the land. This, according to Mr Doran, gave the minimum price at which industrial land was released, and showed what he considered should be a base price for anyone considering purchasing land for the purpose of constructing a mini-mill. 489. Originally, he sought to check his conclusions by indexing the original cost of the land in 1962 for the years which have elapsed between then and the date of resumption by reference to the change in industrial land prices in the Kwun Tong area. He later discovered a flaw in his calculations and this approach was dropped. We do not believe anything has been lost here, as this is not a method in which we have much faith and we would have been unlikely to have given it much weight, with or without the flaws in the calculations. 490. He also adopted as a check the method the government uses to determine the premium for grants to public utility companies by adjusting industrial land values to suit the lower building densities required by such companies. EUROASIA DOCKYARD SITE 491. Of Mr Doran's so-called direct comparables, he put most reliance on the sale of Euroasia Dockyard Ltd's ("Euroasia") site at Tsing Yi Island. This site was sold by Island Navigation Ltd ("Island Navigation") to the China-based China Merchants Steam Navigation Company Ltd ("China Merchants") for $170 million on 13th June 1986, which is very close to the resumption date of 30th July 1986. This land is described as Tsing Yi Town Lot 60 and has an area of 9.6 hectares or 96,000 square metres. Tsing Yi Town Lot 70, which is an adjoining sea-bed lease over which a floating dock is moored, was also included in the sale. Part of the site is steep hillside, with the result that the useable level area is only about 70,600 square metres. 492. Mr Doran considered the user condition which limited the site to ship building and ship repairing, to be "not dissimilar to that of the subject site". Like the Shun Fung site, it was a large level waterfront site with a sea wall along the whole of its water frontage. T.Y.T.L. 70 was required, under the conditions of grant, to be used in conjunction with T.Y.T.L. 60. 493. One of the complications of this sale was that the site adjoins the site of Yau Lian Shipyard Ltd, a subsidiary of the purchaser, China Merchants as, in fact, now is Euroasia. It was really the company, Euroasia, rather than T.Y.T.L.'s 60 and 70 which were sold to China Merchants. The sale was of the shares of the company, Euroasia, to get around the problem created by the conditions of grant from the Crown to Island Navigation, prohibiting any further assignment of the land. Nevertheless, Mr Doran considered it to be a "clean transaction" in the sense that Euroisia had no assets besides T.Y.T.L.'s 60 and 70. On analysis the price paid by China Merchants for Euroasia translated into a price of about $2,400 per square metre (the actual figure was $2,394 per sq.m.) for the useable area of the formed bare site. 494. Mr Brown did not use this sale, but in his analysis of it he put more value on the structures on the site than had Mr Doran. He agreed that an analysis on the same basis as Mr Doran's would result in at least a figure of $2,000 per square metre. However, he was unable to accept that this site was remotely comparable to the subject site at Junk Bay which he considered far inferior. Also he did not regard the sale as a true indication of value. He said that it was not, what he described as "an arm's length" transaction due to the adjoining owner's relationship with the purchaser. 495. Another of the many complications of this sale was the inclusion of the adjoining sea-bed lease of Tsing Yi Town Lot 70 but not the floating dock which was moored over it. No proper conclusions could be reached as to what added value this sea-bed area might have had. Government submitted that it must have some value but the claimant was of the view that it was more of a liability than an asset. The floating dock was still being used and any arrangements concerning Island Navigation's continued occupation of the sea-bed lease were not known. 496. Mr Doran believed that due to the fact that T.Y.T.L.70 was now no longer being used in conjunction with T.Y.T.L. 60, it was subject to possible government lease enforcement action, or re-entry by, the Crown. We have no knowledge of any such action being contemplated, and we are not prepared to speculate on whether such a possibility might have influenced the purchase price or not. The point seems to have been lost in the multifarious issues concerned with this sale. 497. One such issue, concerned the Conditions of Grant having included a provision for compensation on resumption to be determined by a formula which would result in the Crown only having to pay something far less than market value if the property were ever resumed. Mr Doran seemed to think that this would result in a lower price than usual being paid. On this we would only say again that we are not prepared to speculate on whether this might have influenced the price paid, but normally we would expect a purchaser to regard the possibility of resumption as being very remote. If it were otherwise, he would be unlikely to buy in the first place. In any case, Mr Doran's argument, we believe, lessens the usefulness of this sale rather than strengthens it. 498. Among the various issues raised in connection with this sale, one of the most important, we thought, is the fact that on the date of sale a waiver of the user condition by the Crown existed. This was for the use of some 9,290 square metres of T.Y.T.L. 60 for container storage at an annual fee of $577,400. The waiver was granted in February 1984, apparently to regularise the unauthorised container storage use from July 1983. It was back dated to the 1983 date. 499. It was also made apparent that from September 1985 the area actually used for container storage was over 40,000 square metres, and this was the subject of a further waiver granted shortly after the sale to China Merchants. Thus, at the date of sale more than half of the useable area was being used to store containers. By 1988 this had grown to some 58,000 square metres which, as well as container storage, included a concrete batching plant. The government continued to issue waivers to cover these uses as and when they were discovered through regular inspections of the site by the staff of The District Land office. This office had also noted that any activity related to ship-repairing had all but ceased by 1988. 500. There is no evidence that re-entry of the site was being contemplated. However the possibility of such re-entry due to breaches in the user condition, along with the expectation of waiver fees or a modification premium would, in Mr Doran's opinion, have offset any overbid China Merchants may have made to secure the site for container use. He later agreed that speculative overbids do exist and there was no way of telling in this case what it might have been if it existed. 501. Government submitted that it was clear that the purpose behind the purchase was to use the site for container storage, and it had in fact been reported in a newspaper that China Merchants planned to "take advantage of the lengthened deep water berthing 'coast line' by converting the existing facilities into a multipurpose pier to alleviate the present congestion of China Merchants loading and unloading facilities." (Exh SF 22) 502. The purpose behind the sale was never disputed. The claimant argued that it could have only been allowed on a permanent basis by government at a premium due to the Conditions of Grant limiting the use of the site to ship building and ship repairing. Such a premium could be expected to fully cover any additional value for such use. Mr Doran expressed the belief that if any overbid existed the purchaser would on payment of a premium be charged that amount again to secure the modification. This would mean, we suppose, that any overbid would be expected to be minimal or severely limited. 503. Although we accept the claimant's argument concerning the requirement for a premium to cover any permanent change of use, we cannot help but be sceptical of whether any eventual premium would really cover the true difference in value as the theory of the exercise would have us believe. There must be an element of doubt as to whether or not this sale to China Merchants would be likely to influence any premium assessment. On the facts before us, there must be a strong likelihood that a proportion of the difference in value was already included in the purchase price paid by China Merchants in anticipation of the container use being permitted, whether by way of modification, waiver or mere acquiescence. 504. Over half the area of the site was being used for container storage when the site was sold. Also the waiver fee at the time for some 9,290 sq.m. of this area was known to be $60 per sq.m. per annum. By 1988 the annual fee charged for 55,474 sq.m. of container storage was $2,971,821 (p.15 of Exh R56) or $53.57 per sq.m. The reason for the reduction is not known but the rate adopted is a discount from a base of $75 per sq.m. per annum. 505. What we can be sure of is that, at the time of sale, China Merchants was in a position to put a price on this land based on the granting of temporary waivers for the container use. On whether the waiver fee adequately covers the more valuable use, we must remain unsure, although Mr Brown thought it was relatively low. We are left with the question of what a negotiated premium might be expected to be when the beneficiary is already in occupation and is already using the site for the purpose for which the modification is sought. 506. The evidence before us shows that the property was purchased with the intention of putting it to container related uses, and it is being put to those uses without any permanent modification having been granted. We understand that a change of use was sought in September 1986 (Exh SF169) for multipurpose berthing facilities in addition to the container use. This application was still being considered when it was reported in a newspaper in March 1988 that a formal application had been made in February 1988 (Exh SF25). 507. The reason suggested by the claimant for the owner to want a permanent modification was that waivers are temporary and subject to three months notice. We do not see this as likely to give rise to any anxiety on the part of the owner. After all, the prospect of a permanent modification seems to have been left unresolved for over two years that we know of. There does not appear to be any reason why the existing arrangement should not continue until terms favourable to China Merchants are offered. 508. On the matter of premium, it would be reasonable to expect the owner to argue, as Mr Doran has done before this Tribunal, that little or no speculative element existed in the price paid at the time of purchase. Hence, government would find itself in much the same position as this Tribunal, that is, without any detailed knowledge of the container-handling business, or for that matter, the shipbuilding business, we would expect it to be difficult for anyone to determine what the true value is. 509. We strongly suspect that the price paid was based more on the container use than the shipbuilding/ shiprepairing use, but we are in no position to say with any certainty by how much more. We suspect that the government would find itself in a similar situation with insufficient information with which to counter the owner's argument. Of course the longer the negotiations drag on, the less relevant the purchase price becomes. In any case, Mr Doran agrees that people do pay prices based on a speculative use, and that in the present case we have no way of judging what that speculative element might be, if it does exist. 510. We regard it as unlikely that, what Mr Neoh, on behalf of the claimant, described as "the full incremental value" for the difference between the ship building and repairing use, and the container handling use, could in fact be assessed with any degree of accuracy by the Crown, let alone charged, given the circumstances of this case. 511. Also with respect to China Merchants' application, we could not help but note that the press statement reported in the Hong Kong Standard on 16th March 1988 (Exh SF25) quoted "Sources close to Government" as saying, "that the size of the premium would depend on whether Government regarded the proposed development as a service for Hong Kong trade or as purely profit making". While we find this difficult to accept, it certainly does not help the claimants' argument concerning the premium assessment. 512. We heard considerable argument on whether or not T.Y.T.L. 60 was similar to the subject site at Junk Bay. There is little doubt that the Euroasia Dockyard site is very suitable for a mini-mill if not for shipbreaking, but it seems to us that the location bears little resemblance to that at Junk Bay, which could aptly be described, as far as industrial locations are concerned, as something of a back water. Even if we disregard the potential for more valuable uses associated with shipping and container handling due to its proximity to the container terminal and main deep water shipping activity in Hong Kong, the Tsing Yi location is much more in the main stream of general industrial activity than that at Junk Bay. 513. There was some debate concerning traffic congestion when there was one bridge to Tsing Yi. There may well have been considerable traffic congestion on or about the date of purchase in June 1986 but it also seems to us to be evidence of the general extent of the commercial activity in the area at that time. Also the traffic situation might not be so much of an issue now that a second bridge has been completed. We noticed no unusual congestion when we inspected the site in December 1988. We find it difficult to accept that the second bridge was not in the offing and publicly known about at the time of purchase. 514. The claimant has suggested that the uses of ship repairing and shipbreaking are reasonably similar. The subject Junk Bay site was suitable for shipbreaking but we doubt that a large ship repairer would be content with that location. We would expect him to prefer to be in close proximity to the main shipping activity near Tsing Yi. Apart from the government's suggestion that the location was vastly superior by virtue of its proximity to the container port and its deep water marine access, there was really little evidence on this issue, so we could take the matter no further than to say that we would expect such docking facilities to be located close to other shipping activity, if at all possible. This seems to be borne out by virtue of the location of the floating docks in this locality. 515. On the other hand we accept the claimant's argument that a ship breaker would prefer the more sheltered Junk Bay location where beaching facilities were once available and the water is shallow. Such a situation does not mean that the Tsing Yi site should be worth about the same per square metre as the subject. It merely suggests to us that, not only are the uses different but also the requirements and also the likely scale of the operations on these sites are so different as to be beyond comparison. The price suitable for one purpose is unlikely to bear much resemblance to the price suitable for the other. 516. The subject site has an area of about 36,000 square metres while the Euroasia site is some 96,000 square metres with a usable area of 70,600 square metres. The usable area of the former is about half that of the latter. 517. Mr Doran suggested that with respect to comparisons of properties of different sizes, the usual practice of reducing the unit rate (ie the price per square foot or per square metre) as the area increases would not necessarily apply to what he described as "land intensive uses" or industries where the land area was important but only for ground floor and open areas use. 518. We appreciate what Mr Doran is trying to say and to some extent agree. Each use has its optimum size and anything less in area puts greater value on what is available, while anything greater is in the nature of a bonus which becomes less important and hence less valuable as more area becomes available. Idealy, we should be looking for comparisons which are not only for similar industrial uses but also, perhaps more importantly, for likely industrial operations which are of a similar scale and would require sites of similar dimensions to that of the subject. Here we do not believe, on the evidence, that the scale o?any suitable operations on these two sites is likely to be comparable. We will mention the question of size again in relation to other comparables. 519. Even if the relationship with the adjoining owner did not exist we would find it extremely difficult to gain much assistance from this sale. We do not share Mr Doran's opinion that it compares with the subject site, when even he has found it necessary to adjust the price paid down by over 35%, rather than the 20% in his workings, without the benefit of any proper analysis. We acknowledge that analysis was probably an impossible task, but, without it in this case, we cannot accept that the adjustments are in any way adequate. 520. The added disadvantages of it being an exchange of shares, the unresolved value which the sea-bed lease for T.K.T.L. 70 might have, the differences in size and permitted users and the intended use being for a purpose other than that permitted by the Lease Conditions, lead us to conclude that it would be most unwise to attempt to use this transaction for the purpose of determining what the value of the subject site might be. 521. On top of all this there is the adjoining owner relationship. It is generally accepted that an adjoining owner would be prepared to pay more for the opportunity to extend his holding than any other purchaser in the market might pay for the same land whether or not there were obvious benefits. The fact is simply that there is no possibility of a purchaser who is not an adjoining owner paying anything extra with respect to amalgamation, while there must always be the doubt that an adjoining owner might well be doing so. Thus, notwithstanding that no change of use on amalgamation seems to have been contemplated so far for the Yau Lian Shipyard, the mere fact that there is an adjoining owner relationship associated with the sale casts some doubt on its reliability. WAIVER SITES 522. Mr Doran produced two sales of agricultural land in the Yuen Long area of the New Territories on which the Crown had granted waivers to allow them to be used for steel rolling mills or "iron foundry purposes". One had a site area of 2,266.22 sq.m. and the other 907.1 sq.m.. 523. We do not propose to spend too much time on this evidence, for Mr Doran, after introducing these sales because they were put to uses similar to that of the subject site, then described them as difficult to analyse due to the likelihood that the declared amount of the transaction "would be the lowest the parties thought they could get away with". This is presumably to minimise stamp duty. He also agreed that the sizes of the sites were hardly similar in scale to the subject site. Neither site was suitable for a mini-mill. 524. Mr Doran in his analysis ignored any value that the buildings might have as he believed the parties were only registering the transactions as sales of agricultural land. Mr Brown analysed these transactions after allowing for the buildings and was able to come to figures which lent more support for his valuation than the $1,000 per square metre which Mr Doran concluded was the unit rate and the minimum price one would expect to pay for land suitable for any form of industrial use. 525. Although the waivers which seem to have been pegged at a standard rate of $30 per square metre per annum were granted on the basisw of allowing industrial use, since the Court of Appeal's judgment in Melhado Investment Co Ltd. v. Attorney General (1983) HKLR 327 it is really the right to construct industrial buildings on agricultural land which has been granted. It was held in that case that agricultural use as stated on the schedule of the lease was only a description of the use and not a limitation. What the leases do in most cases is prohibit buildings other than agricultural buildings on the land. As a result of the court of Appeal's judgment, it has become legitimate to use agricultural land for storage of containers or for car wrecking yards. By virtue of the waivers giving the right to place buildings on non-building land, Mr Brown's analysis which places value on the buildings seems more appropriate than Mr Doran's approach which ignored the buildings. 526. In any case any decision on our part concerning the appropriate value of the waivers or the prices reflected in the transactions, seems to be unnecessary when Mr Doran, who sought support from them to determine a minimum industrial value, has also effectively discredited them as useful evidence of value. He said, "I don't think anyone would ever get to the bottom of these transactions", and that he believed that the true purchase price was different from that declared by the parties to the transaction. 527. Mr Brown probably summed up these sales when he said, "We don't know what was transacted here; whether it was a business; whether it was the land; whether it was the land and buildings; or indeed, the land and buildings and the equipment on the site". YAU TONG SITES 528. The only other comparable put forward as a direct comparable by Mr Doran was a sale of two adjoining lots at Yau Tong which permitted shipbuilding or saw milling. These lots had water frontage to Yau Tong Bay which is at the eastern end of Hong Kong Harbour and divided from Junk Bay by Lei Yue Mun channel. Yau Tong was once considered a fairly remote industrial area although it adjoins the very busy Kwun Tong industrial area. 529. This changed with the announcement of the construction of the Eastern Harbour Crossing which would provide within a tunnel both a road and a Mass Transit rail link between the eastern end of Kowloon and the eastern end of Hong Kong Island. The Kowloon entrance was to be in the Yau Tong area and this gave rise to dramatic increases in prices in the locality. Public knowledge of the proposed tunnel would have been fairly general from October 1984 when the press first announced it. A firm commitment to building it was made in December 1985. 530. The sale which Mr Doran produced was for Yau Tong Marine Lots 30 and 31 on 13th February 1987 at $9.5 million. The area was 1,802.3 square metres. After allowing for the value of buildings on the site Mr Doran analysed this sale as showing a price of $4,150 per sq.m. for the bare site. A later analysis led him to alter this to $3,745 per sq.m. (Exh SF27E). 531. Mr Brown produced another sale for a similar site comprising two adjoining lots nearby which also had a frontage to Yau Tong Bay. It sold on 2nd January 1986 at $7.3 million. These lots, described as Yau Tong Marine Lots 36 and 37, had an area of 3,344 sq.m. 532. Mr Doran analysed both these transactions on his SF23E, and concluded that the sale of YTML's 36 and 37 showed a rate of $1,734 per sq.m. while that of YTML's 30 and 31 showed a rate of $3,745 per sq.m. Mr Brown agreed with Mr Doran's allowances for the buildings on the sites, and agreed that $1,734 per sq.m. was a reasonable conclusion for the rate per square metre reflected by the sale he produced. 533. There was considerable discussion during Mr Brown's cross-examination as to what might have been the increase in Yau Tong values between the date of this sale and the resumption date of July 1986. This took the form of an exchange between Mr Neoh for SFI and Mr Carnwath for government, as well as the witness Mr Brown, whereby each of them attempted to interpret for the Tribunal what had been the basis of an agreement (Exh SF 151 (Matter 3)) between Mr Doran and Mr Brown for the increase in industrial land values over this period. 534. We mentioned this agreement earlier. The problem centred on what Mr Doran, who most likely drafted the agreement, understood the term "general industrial land" to mean in the agreement statement. The agreement was that "general industrial land" values rose over the period from the end of 1985 to the beginning of 1987, and that the rise in values was greater in the second half of the period than the first. The agreement was more specific for the period from February 1986 to the date of reversion where it said a rise of 5% should be adopted. 535. We were told that Mr Doran had some reservations concerning the application of the 5% to sites having water frontages but he was not in a position to say exactly what adjustment should be made. 536. Occasionally we experienced difficulty understanding some of Mr Doran's inventive terminology, which was delivered as if it were recognised technical jargon. This was one such case which, unfortunately, was debated in Mr Doran's absence. We have been led to understand that Mr Doran regarded the term, "general industrial land" as meaning land suitable for development for flatted factories, while Mr Brown believed the agreement was for industrial land generally. In any case it does not seem to us to be of any consequence, and either interpretation will serve for our purposes. 537. Mr Neoh for SFI was later prepared to accept that all industrial land came under the terms of the agreement and, in his submission, has gone further and suggested that the Tribunal adopt an approach to give a range of possible values by adding 5% to $1,734 per sq.m., which is the rate agreed for the January 1986 sale after deducting for the value of the buildings, to give the lower end of the range, and adding 5% to $2,183 per sq.m., which is the rate if there is no deduction for the buildings, to give the higher end. 538. The range which results from Mr Neoh's overly precise calculations, is from $1,821 per sq.m. to $2,292 per sq.m. He suggested that the Tribunal adopt the higher figure. This proposal or method was not advocated by any of the valuers. 539. It was agreed that Yau Tong had experienced a large degree of speculation since the announcement of the new tunnel. So the analysis and use of these sales caused considerable discussion during the cross-examination of both Mr Doran and Mr Brown. It sometimes appeared that the parties were at cross-purposes. This was more evident during the cross-examination of Mr Brown. 540. In an effort to resolve the concern over the degree of speculation inherent in these sales, the claimant produced a written submission (Exh SF170) by Mr Moffoot who is a director of the same firm as Mr Doran. Mr Doran was apparently absent from Hong Kong at this point of the hearing. 541. Mr Moffoot was not called as a witness and was therefore not cross-examined on his submission which attempted to measure the amount of the overbids due to the speculation reflected in these sales. He did this by comparing them with industrial sales mainly in the Sha Tin area and two sales in Kowloon Bay over the same period. 542. Mr Brown did not object to the use of the Sha Tin sales but he criticised the narrowness of the comparison. He was more critical of Mr Moffoot's omission of the value of the buildings in his analysis of the Yau Tong sales and doubted that his methodology would define the degree of speculative overbid even if it were correctly applied. Mr Brown, then reworked Mr Moffoot's calculations after deducting for the buildings. This resulted in considerable increases in the calculated percentages for the overbids. 543. Mr Brown's insistence on deducting for the buildings seems to be something of a change of stance for he was originally of a view that the buildings were unlikely to be of much consequence to any buyer in such a speculative market. 544. In spite of the differences of opinion, this Tribunal would prefer to see such analysis more often with respect to the application of sales in a valuation exercise rather than the far too often adoption of a series of large percentage adjustments based on nothing more than the valuer's professed experience. While Mr Moffoot's method may have been open to criticism, and his base for comparison may have been somewhat restricted, it was a commendable attempt to obtain useful conclusions through sales' analysis. The Tribunal found this analysis helpful in spite of the fact that we did not entirely agree with it. 545. We refer to Mr Moffoot's arguments concerning the value of the buildings which appear to be self-contradictory. (Exh SF 258 p.17 and 18) On the one hand, he argued against any deduction for the value of the structures, as he believed they would have no value to a speculator who purchased with a view to seeking a modification of the lease conditions to allow more intensive industrial use. He then drew attention to the fact that government has continued to refuse any lease modification applications, a factor, which any astute buyer would have to acknowledge, and accordingly would have to allow for holding the land for an unknown period under its present lease conditions. We cannot see how the buildings could have no value to a buyer who must know that he will not be permitted to redevelop for the time being for any use other than that laid down under the present lease conditions. 546. Mr Doran allowed for the value of the buildings and Mr Brown agreed with his deductions, even though they both said that the buildings would have been of little consequence in these transactions. This also seems to be something of a contradiction, but we inspected these sites and noted that the buildings were being used, and we presume that both Mr Doran and Mr Brown feel as we do that it would be wrong to disregard them. Mr Doran's allowances for the value of the buildings on these sites has been agreed by Mr Brown and we do not propose to interfere with this agreement. 547. However the conclusion reached by Mr Moffoot that the land at Yau Tong was worth about $2,000 per sq.m. at or about the resumption date does not differ substantially from, the figure of about $1,800 per sq.m. which one derives from Mr Doran's and Mr Brown's agreed rate for the sale of YTML's 36 and 37, plus the 5% suggested by Mr Neoh for the increase in value from January to July 1986. Mr Doran and Mr Brown have deducted for the buildings while Mr Moffoot has not, and Mr Moffoot, of course, has also allowed for any overbids. 548. With due respect to Mr Moffoot's efforts, the degree of the overbids, in our view, turns out to be not too significant a factor in this exercise. The sales show what was being paid for land in Yau Tong at a time when there was knowledge that the Eastern Harbour Crossing was likely to cause a change for the better in the area. Our attention was drawn to the price rise for Yau Tong from the earlier sale of YTML's 30 and 31 in February 1985 at $2.05M (this sale was introduced during the course of further analysis, p.17 of Exh R56) to the sale of the same site in February 1987 for $9.5M. The increase was of the order of 360%. Yet the graph produced by Mr Moffoot (Exh SF 170B) shows that the rate of the increase in the prices in Sha Tin compared to Yau Tong throughout 1985 and 1986 was much the same, or of only marginal difference. It seems on this, as well as the other evidence available, in particular the agreement between Mr Doran and Mr Brown concerning the price rise for industrial land, that it was only towards the end of 1986 or early 1987 that Yau Tong prices were to increase rapidly. 549. The Yau Tong market may have increased at a much greater rate than that at Shatin or elsewhere but this seems a reasonable consequence of the proposed location for the new cross-harbour tunnel. The inevitable improvement in access for transport, on its own, should have resulted in an increase in value for the Yau Tong area. 550. This leads us to reflect on whether these Yau Tong sales are a useful guide as to the value of the subject site. There is some degree of agreement between the parties concerning the values reflected in the sales of the Yau Tong sites. Although much smaller than the subject SFI site, and not suitable for mini-mill use, they are examples of the sale of industrial land with water frontages not too far removed from the subject site at Junk Bay. 551. The present permitted users under the lease conditions for shipbuilding yards, sawmills or timber yards limits development to the ground floor, with perhaps some office space above. The two Yau Tong sites since their sale, have been put to uses which Mr Doran described as "land intensive", being the storage of buses in one case and the storage of heavy machinery in the other. These may be interim uses pending modification but they are confined to the ground floor. This leads us to conclude that perhaps too much emphasis may have been put on the possibility of modification as the prime motive behind the prices paid for these sites. We are told that no modifications have been granted as yet. This is a very different situation from that concerning the Euroasia Dockyard site where the higher use was actually available by way of temporary waivers. 552. We have noted and Mr Brown pointed out that the purchaser of YTML's 30 and 31 was Chung Wah Shipbuilding and Engineering Co. Ltd. which has its head office on another of these Marine Lots at Yau Tong (Exh SF171). When we inspected YTML's 30 and 31, it was being used for the storage of buses, but Chung Wah, we understand, includes within its interests shipbuilding and ship repairing which is what the site user condition permits. 553. We are obliged to have regard to section 12(c) of the Ordinance which provides that, "no compensation shall be given in respect of any expectancy or probability of the grant ... by the Crown or by any person of any ... lease or permit whatsoever." If a comparable is affected it is obvious that it should either be rejected, or the price element associated with the prospect of any future grant should be deducted, before any comparison is made. 554. The possibility of future grants of modifications for flatted factory use causes us some concern, but we must assume that not all speculators are imprudent and that the prices should reflect the market for Yau Tong at that time. There may have been some element reflecting the expectancy of modifications in the pre-July 1986 sales, but we believe it would have been insignificant compared to the increase in value due the potentional advantages of being closer to transport and other facilities by virtue of the proposed cross harbour tunnel. 555. We do not consider that Mr Neoh's proposed application of 5% on the January 1986 sale to allow for the increase in the market up to the resumption date of July 1986, was really appropriate for Yau Tong. Neither Mr Doran nor Mr Brown suggested that it was. This 5% may have been agreed as the increase over the period for "general industrial land" on Mr Doran's interpretation or industrial land in Hong Kong generally on Mr Brown's, but it is obvious that Yau Tong was not conforming to the general pattern about that time. Of course as we previously noted the rapid increase does not seem to have developed until late 1986 or early 1987. While 5% may be an appropriate increase throughout the flatted factory market for the first half of 1986, Yau Tong is likely to have progressed little more in that time. 556. We prefer to accept that the prices paid at Yau Tong were an indication of the market there at that time. It included the influence of the proposed tunnel. The prospect of modifications being granted must have been considered by any well informed purchasers as being fairly remote throughout 1986 or at least something of a gamble. We would not wish to place as much weight on this probability as perhaps the valuers may have done. If there was any element of value associated with the prospect of modifications, it was most likely small enough to be capable of being allowed for in any comparisons within a range of values. 557. We believe that the most we can gain from these sales is that, at the date of resumption, the price of Yau Tong, with all its advantages and disadvantages, was somewhere in the region of $2,000 per square metre. Mr Moffoot came to the same conclusion but he allowed for any speculative overbid, and did not allow for the value of the buildings. 558. Mr Brown drew no conclusion as to the value of the Yau Tong sites at or about the resumption date. What government says about these sales is simply that they were prices affected by speculation. This was agreed, but speculation does not seem to have really taken hold until after the resumption date. 559. Where we have difficulty concerning the usefulness of this information as a basis for the value of the SFI site, is in relation to the operations suitable for the Yau Tong sites being of a much lesser scale than anything such as a mini-mill. The price for a use suitable in Yau Tong is hardly comparable to the larger scale use suitable for Junk Bay. 560. All we can conclude, is what Mr Brown seemed to propose at the outset when he described his sale as a useful "bench mark", which is to say it gives us a reference point as to what might be paid for waterfront industrial land in a better locality and suitable for much smaller scale operations than a mini-mill. 561. Mr Doran on the other hand would have us consider it as the value of industrial waterfront land and make no allowance for the difference in the sizes of the sites. In his opinion the larger site may well result in a higher rate per square metre than that paid for a small site. This may be so, but Mr Doran produced no evidence to show that it is. These sales on their own are insufficiently comparable to allow us to come to any firm conclusion on the value of the subject site. CLAIMANT'S "DIRECT" COMPARABLES 562. Mr Doran considered the Euroasia Dockyard site to be the best of what he described as his direct comparables which included the sale of the waiver sites and one of the two Yau Tong sites. This information gave him a range of values of between $1,000 per square metre and $2,000 per square metre from which he chose to apply a rate of $1,500 per square metre to the subject site. 563. Try as we might, we cannot gain much help or draw any conclusions from any of these sales. What emerges is the realisation that the scale of the industrial undertaking might well be more important than the actual use in a valuation exercise such as this. CROWN LAND AUCTION SITES 564. The scale of the optimum industrial undertaking as an aspect of comparability is also largely absent from Mr Doran's list of Crown land auction sales of industrial land in the New Territories. Mr Doran put these sales forward by way of a check on his valuation of $1,500 per square metre for the Shun Fung site. The list comprises eight such sales over a period from December 1985 to February 1987. Of Mr Brown's list of nine comparables, five of them were Crown land auctions in Mr Doran's list. Mr Brown used them as direct comparables. Those five also used by Mr Brown are marked with an asterisk (i.e.*)
565. We note that Sale No 7 seems to be incorrectly stated. If the analysed figures and the area are correct, the sale price would have been $41 million. Mr Moffoot also referred to this sale in his SF170. His analysed figure also suggested that the price was $41 million. No further reference was made to this sale in Mr Doran's report or evidence. 566. These were all sales of land which was to be developed as flatted factories. Sometimes, as was the case for Sale No. 4, the ground floor of such developments is taken up by vehicle parking which is required under the lease conditions. Where this occurs there is virtually nothing in the way of comparability with the subject site which can only be used for what is essentially a ground level development. 567. Mr Doran considered that these Crown land auction sales could not be used as comparables as they were for proposed developments which were too dissimilar from that of the subject site. But he said some use could be made to establish approximate ground floor accommodation values. 568. Accommodation value (often abbreviated to A.V.) is a term used to describe the unit value reflected in the price paid for a site on the basis of each square metre or square foot of building which can be put on that land. It is used to compare similar sites which permit different plot ratios (building size to site area). 569. Mr Doran pointed out that the ground floor was worth more than the upper floors and that analysis should be done to dissect the ground floor accommodation value before making any comparison with the subject site where only ground floor value was important. 570. Where the sizes of permitted or optimum developments differ, meaningful comparisons can be made by using accommodation values. However, we must agree that most flatted factory land has very little in common with land which can only be used for a mini-mill in conjunction with shipbreaking. If, we had been able to gain anything from "the direct comparables" we probably would have looked no further, or, alternatively, expected that the valuers would have concluded that this was a case where a residual valuation was necessary in order to determine what this land might be expected to realise in the market. 571. There seemed to be a reluctance to use the residual method, which suffers some neglect in Hong Kong where, we feel, it is wrongly considered to be looked upon unfavourably by the courts. Criticism has been leveled on the application of the method, but this was usually due to inadequate supporting analysis for the various elements incorporated in the calculation. This, obviously, would have minimised the usefulness of any resultant valuation. There is naturally a general preference for direct comparisons with sales of similar properties. 572. This Tribunal, in such cases as Director of Lands and Survey v. Ip Ping Chi (1978) HKLTLR 91 and Director of Lands and Survey v. Chan Tai Land Investment Ltd (1978) HKLTLR 115, where satisfactory comparables existed, has rejected valuations based on the residual method in favour of direct comparisons. On the other hand, in Hofei Estates Ltd v. Secretary for City and New Territories Administration CLR1/82 where development land was resumed, the valuers, as they have done here, limited their valuations to the comparative method. In that case, each made a large number of percentage adjustments which were largely unsupported by any analysis. The Tribunal pointed out that this was one case where it would have been better to use the residual method. 573. In the present case, there is information on the cost of setting up a mini-mill and evidence from steel experts and accountants which could have been used by the valuers to determine what someone seeking to purchase a site suitable for a mini-mill could afford to pay. We do not believe that such an exercise would have been without some problems, given the parties' differing views. We also accept that the cost of the land in such an exercise could be expected to be somewhat insignificant compared to the estimated cost of setting up the plant and this might well result in any residual value for the land being impossible to determine with any degree of accuracy. In any event because both valuers ruled out the residual method, we feel we are obliged to look more carefully at these auction sales. 574. Where we and the valuers have difficulty with these Crown land auction sales is that we do not have a true comparison of like with like. None of these auction sites is suitable for a mini-mill, but as Mr Carnwath explained they may be used as a guide as to what might have been a hypothetical negotiation between a willing seller and a willing buyer of the subject site. As we have said, in the absence of any well researched residual exercise we have little else where there is some degree of agreement between the valuers. Mr Carnwath further submitted that it was a similar approach to that, according to Mr Doran, taken by the Crown when arriving at a value for public utility land. 575. It was also Mr Carnwath who reminded us of the willing buyer, willing seller concept set out in the High Court of Australia case of Spencer v The Commonwealth of Australia (1907) 5 C.L.R., 418, to the effect that one has to imagine the negotiation between the owner and a potential purchaser at the date of resumption, putting onself in the words of Griffiths C.J., "as far as possible, in the position of the persons conversant with the subject at the relevant time, and from that point of view to ascertain what, according to then current opinion of land values, purchaser would have had to offer for the land to induce such a willing vendor to sell it, or, in other words to inquire at what point a desirous purchaser and a not unwilling vendor would come together". 576. The willing buyer willing seller concept assumes that the parties are completely conversant with all aspects of such transactions and all the factors which may influence the price at which they might reasonably be expected to agree. We see no reason to deviate from this principle but here we have the hypothetical parties in a market where there is little or no direct evidence of value. 577. If we were at first to approach this from the point of view of a willing but not anxious vendor contemplating what he might "expect to realise", (in the words of section 12(d)) for this vacant site which can only be used in conjunction with shipbreaking and steel-making, we must assume that he is well informed; but, given the fact that he may well enlist the services of a steel expert, we may be permitted to assume that he might wish to base his assessment of what his site might realise on more than just being conversant with all the facts and figures associated with the production of rebar. In the absence of sales of similar land he might well look to other sales of industrial land which might be used for, what Mr Carnwath described as a "proxy" for the value of subject site. 578. The exercise is to go from knowing what is a basic industrial value for a similar location and then try to reason as a prospective vendor might, as to what he could reasonably expect to realise for his site. Any estimate of value can only be based on the best evidence available. It does not seem unreasonable to us to assume that even with perfect knowledge on the part of the parties the figure they would be prepared to agree might well be based on such a process. Although bound to be a very strong influence on price, a purchaser does not necessarily pay what the land is worth to him, but rather what is required to secure it. He pays what the market dictates he should pay. It is the market value of the land that we are seeking to determine under section 10(2)(a) and section 12(d). 579. The user restriction ensures that the site will need to be of sufficient size to accommodate a mini-mill and have the benefit of an area which can be used to store scrap on site. As the user clause in the lease requires that the site be used for shipbreaking, it is obvious that it would have no value for this purpose if it did not have suitable water frontage. These elements would seem to be the basic requirements which distinguish it from other industrial sites. 580. Not only has Mr Brown used these auction sales as direct comparisons but also Mr Doran's colleague, Mr Terence Lo, has used one of these sales in his negotiations with the Crown, on behalf of Mr Doran's firm for adjacent lots which were the subject of the same resumption exercise. We have, his analysis (exhibit R.3) for lot 1002 in DD215 at Sai Kung (Sale No. 3 on Mr Doran's list), but Mr Doran discounted this analysis as he said Mr Lo used it merely as a "negotiating tool" to persuade the government that it should pay $790 per square metre for the value of the land. 581. Mr Lo was not called to give evidence on his analysis, but Mr Doran was adamant that the settlement sought by Mr Lo was governed by factors such as the Crown's settlement with Chip Hua Comalco which was granted a new site nearby in exchange for their resumed site. It seems a figure of about $750 per square metre was used in this arrangement. It would have been unrealistic, according to Mr Doran, for Mr Lo to seek a higher figure in the light of the government's attitude and actions at that time. 582. We are not concerned with this evidence of negotiations with other resumed parties and we place no store on it. But we do note the support exhibit R3 gives for this approach as one used by valuers at the time of the resumption. We also have noted the calculations of Mr Lo where they support or correspond with those of Mr Doran and Mr Brown. 583. Mr Carnwath pointed out that Mr Lo's adjustments were very similar to those of Mr Brown. Mr Lo arrived at an analysed land value of $889 per square metre in support of his firm's counter-offers to the Crown at that time of $790 per square metre. According to his letter of 6th August 1986 the Crown was offering $400 per square metre. 584. Mr Brown in his application from an analysed overall accommodation value of $755 per square metre derived from the same sale of lot 1002, adopted a Plot Ratio of 1 for the subject site and arrived at an overall land value figure of $490 per square metre. 585. Mr Doran on the other hand arrived at a figure of $1,298 per square metre for the land value, but he applied a ground floor accommodation value of $3,000 per square metre derived from adjustments to his analysed figure of $1,373 to only an area of 13,307 square metres which was the actual area covered by buildings on the SFI site. To the balance area of 22,711 square metres of open land he applied 1/10 value ($300 per sq.m.). 586. As Mr Carnwath observed, if Mr Doran's figures are adjusted for the now agreed 5% rise in values over the first half of 1986, his figure for the unit land value becomes $936 per square metre. (Exh R6.C). We will show these calculations in more detail when we deal with the sales analysis. 587. $936 per square metre is not too far removed from Mr Lo's $889 per square metre. Mr Lo used a 4% adjustment for the same time period. Mr Doran and Mr Lo may not have used the same adjustments or method of application of accommodation value, but the end result seems to be a figure in the region of $900 per square metre. 588. Mr Lo's application of analysed accommodation values is interesting. He, of course, was valuing adjacent lots when he assumed a plot ratio of 2.5 being used for half of his subject sites. Thus, he used an analysed upper floor accommodation value as well as a ground floor accommodation value. He then valued the other half of each site as open yard by reference to the rent that might be paid for land which was used for open storage. 589. This concept may have had considerable merit, for Mr Doran tells us that land sold by the Hong Kong Industrial Estates Corporation at Yuen Long and Tai Po and used by industries which require extensive ground floor accommodation was offered at a maximum plot ratio of 2.5. Mr Doran introduced the prices for land on these estates as one of his checks. Also lot 1002 in DD215 allowed a plot ratio of 2.5. 590. Mr Lo seemed, to accept, by making a user restriction reduction, that this plot ratio exceeded that required for Junk Bay; and he was not valuing specifically for steel-mill use. Nevertheless it is perhaps a pity that Mr Lo was not called to give evidence so that his method could have been explored in some detail. The fact that he was not called leads us reluctantly to take this approach no further. 591. What does emerge from the evidence on this method is the agreement of the valuers to adopt an approach based on the site being essentially used for ground floor development. It would probably be similar for some of those industries which have purchased on the estates of The Hong Kong Industrial Estates Corporation at Tai Po and Yuen Long where the provision for high ceilings and heavy floor loadings can be met. 592. Mr Brown's adoption of an analysed overall accommodation value for a plot ratio of 1 may have been the simplest method of comparing the sales and applying value to the subject site. Mr Doran and Mr Lo have gone a little further in analysing for what was being paid for land in relation to building the ground floor industrial accommodation which is more valuable than the upper floor accommodation. Mr Doran used a ratio of 4 to 1 between ground floor and upper floor accommodation values and Mr Brown had no argument with this. Mr Lo, we note, used a ratio of 3 to 1. 593. We favour the analysis to take out the ground floor accommodation value as it allows for the differing ratios of potential ground floor to upper floors in the comparabies. We are only really interested in what has been paid for potentional ground floor development. We are content to follow Mr Doran's 4 to 1 approach. 594. An open area is needed for scrap storage and Mr Doran's adoption of the areas actually used by SFI for buildings and open storage seems to be reasonable. We acknowledge that a hypothetical purchaser might not have exactly the same idea as SFI as to how the buildings should be designed and placed and we would not want to be too rigid when considering the area that the buildings would be expected to cover for a mini-mill. We should be content to say that it would be likely to beinn the region of what SFI had on the site. For instance the layout of the buildings would be expected to be similar to that proposed for Shunde. Also, some subsidiary requirments such as office or canteen accommodation might be met by constructing upper floors, thereby freeing more area for the scrap yard. To some degree, this had actually been the case with the SFI buildings. 595. We propose to adopt an area of 13,000 square metres as the area required for buildings and the balance area of about 23,000 square metres as the area of the open space. 596. We have from Mr Doran's list a series of analysed accommodation values for ground floors at various locations. We set out his list together with his and Mr Brown's analysed A.V.'s. Mr Doran's A.V.'s were for Ground Floor value only, and based on his 4 to 1 ratio, while those of Mr Brown were for the overall building area.
597. The valuers seem to have agreed that the most comparable location is at Sai Kung and the most comparable sale is that of lot 1002 in DD215 in January 1986 (Sale No. 3) which showed a ground floor accommodation value of $1373 per square metre. This is the same sale as that analysed by Mr Lo in exhibit R3. Another nearby lot 1008 in DD215, which sold in February 1987 (Sale No. 8), showed a ground floor accommodation value of $2559 per square metre. Lot 1002 has an area of 3231 square metres while lot 1008 has an area of only 1094 square metres. The plot ratios are 2.5 and 2.0 respectively. 598. It appears that it was really only these two sales which Mr Doran used. He apparently adjusted his analysed A.V. of $1,373 per square metre for Sale No. 3 which was dated 23rd January 1986, by reference to Sale No. 8 which showed an A.V. of $2,559 per square metre for 26th February 1987. The rise in value had been $1,186 per square metre over 13 months which by interpolation corresponds to about a $550 rise per square metre for the 6 months to the resumption date of July 1986.
599. Mr Doran apparently found this to be close enough for him to adopt $2,000 per square metre for July 1986 for Sai Kung. To this figure he applied an increase of 50% for sea frontage and "superior form", which is Mr Doran's term for available ceiling height, to determine a value of $3,000 per square metre as appropriate for the built over area at Junk Bay. He made no further adjustments. His calculation was as follows :
For a total site area of 36,018 sq.m., this equates to a rate of about $1,298 per square metre. (Say, $1,300 per square metre) 600. Mr Doran took this figure of $1,298 per square metre as support for the $1,500 per square metre which he had derived from his "direct comparables". 601. As we mentioned earlier, Mr Carnwath showed that if the now agreed 5% were applied for the rise in value for the first half of 1986, Mr Doran's calculation would have been :
For the total site area of 36,018 sq.m., this equates to a rate of about $936 per square metre. 602. Mr Carnwath's calculations are shown in exhibit R6C. 603. By comparison Mr Brown applied various percentage adjustments to his analysed A.V.s as follows:
From these and his other sales figures he adopted the rate of $450 per square metre for the value of the land at the date of resumption. However from Sale No. 3, (the Sai Kung comparable), he obtained an analysed rate of $490 per square metre. As Mr Brown applied his analysed A.V. of $450 to a plot ratio of 1 his final figure was simply calculated as follows:
As we have said we prefer to adopt Mr Doran's method. 604. We agree that Sale No. 3 at Sai Kung is the best comparable. It is supported by the sale of lot S.T.T.L. 261 (Sale No. 2) at Sha Tin. Sale No. 2 shows a ground floor accommodation value of $1,660 per square metre. This lot sold on the same date as Sale No. 3, but Sha Tin is a better location than Sai Kung. This lot has a plot ratio of 3.5 and a larger area of 13,350 square metres. 605. Most of the other sales are not very helpful except for two at Fanling having areas of 2,029 square metres (Sale No. 5) and 1690 square metres (Sale No. 6). Both sold on 29th October 1986 and show a ground floor accommodation value of $1,479 per square metre in each case. The plot ratios of the Fanling sites are 5.0, so these sites are subject to more intensive development, over smaller areas, but, the analysed ground floor accommodation values lend some support for both Sale No. 3 at Sai Kung and Sale No. 2 at Sha Tin. 606. However the other Sai Kung sale of lot 1008 (Sale No. 8), is not helpful because it is out of line with the others. This may be due to its being for a very small site as well as being in February 1987, by which time values were said to have risen sharply. Accordingly, we believe it should be ignored, as should Mr Doran's use of this sale. Lot 1008 was not included in Mr Brown's list of comparables. 607. The Sha Tin sale (Sale No. 2) is particularly useful in that the area is about the same as the building area which could be expected to be used for a mini-mill. The location is better than Sai Kung, but the difference between the ground floor A.V.'s of $1660 per sq.m. for Sha Tin (Sale No. 2) and $1,373 per sq.m. for Sai Kung (Sale No. 3) seems reasonable and suggests that to this extent there should be no adjustment for any difference in size. Hence, our hypothetical seller has a basic land value for a ground floor building at an A.V. of something in the region of $1373 to $1660 per sq.m. for better industrial locations at a date some six months earlier than the relevant date. 608. We have the agreement between Mr Doran and Mr Brown that land values for sites suitable for this type of development increased by about 5% over the first half of 1986. Mr Lo allowed 4% for the same period. The ground floor accommodation values could then be expected to be about $1740 per square metre for Sha Tin and about $1440 per square metre for Sai Kung by the relevant date at the end of July 1986.
609. We would not wish to put too much emphasis on it due the different lot sizes, but if we compare these accommodation values for July 1986 with those at Fanling of about $1,480 per sq.m. for October 1986, they look about right. We might also mention that by making our adjustments in sequence we are able to make comparisons such as this. The adjustments are also in line with the percentage differences in value which have either been analysed or agreed. 610. Mr Doran felt very strongly that this is the only correct method to apply such adjustments, but this Tribunal has not ruled out what has been described as "the aggregate method" where percentage adjustments are aggregated or totalled and the sum of the adjustments applied to the comparable as one overall adjustment. In most cases the results are much the same. In the case of Hofei Estates Limited v. Secretary for City and New Territories Administration CLR 1/82 the Tribunal was asked to rule on which method should be preferred. In that case the adjustments were mostly unsupported by any detailed analysis. Because the Tribunal felt that the adoption of either method depended on the way analysis for adjustments were done, it could not rule in favour of either method. Each was acceptable as long as the method adopted followed the same line as the analysis. Here we are able to state a preference for making adjustments in sequence as it complies with the differences in comparative values either analysed or agreed between the valuers. 611. It might be helpful at this stage to note that, apart for the adjustment for time which we now accept as 5%, the adjustments made to the most comparable sale of lot 1002 at Sai Kung (Sale No. 3) by each of the three valuers were as follows :
612. Mr Doran and Mr Lo adjusted in sequence while Mr Brown aggregated his percentage adjustments. 613. Having established what the ground floor accommodation values for Sai Kung and Sha Tin might have been at the relevant date, the next step would seem to be to derive, as best we can, what that value might have been for Junk Bay. 614. Mr Brown reduced his value by 20% for the difference in location between Sai Kung and Junk Bay and by 30% between Sha Tin and Junk Bay. Mr Doran made no adjustment. He looked at the location from the point of view of a mini-mill site, and said that the location and access were suitable and adequate for that purpose. He did agree that as "general industrial land", which is Mr Doran's term for land suitable for flatted factory development, Junk Bay would require some downward adjustment. 615. As we are using ground floor accommodation values for flatted factories as our proxy, we cannot, as Mr Doran would have us do, switch horses and consider location in the terms of a steel-mill. In any case, on any basis, access to Junk Bay, as well as its comparative remoteness, make it a poorer industrial location than Sha Tin or Sai Kung. We have no difficulty in accepting Mr Brown's adjustment of minus 20% to the Sai Kung comparable, particularly since Mr Lo has also used a minus 20% adjustment. 616. If minus 20% is appropriate for Sai Kung then Mr Brown's adjustment of minus 30% to the Sha Tin comparable also seems reasonable.
617. We have established that the ground floor accommodation value for Junk Bay should be somewhere in the region of $1150 to $1200 per sq.m. but there was disagreement as to whether the water frontage would add value or not. In this respect Mr Brown was of the opinion that, as a waterfront site was essential for shipbreaking, no further adjustment was necessary. He believed that any added value was cancelled out by the user restriction. 618. Mr Doran explained with respect to lot 1002 at Sai Kung that an adjustment was necessary to determine what the ground floor accommodation value would be at Sai Kung if it had waterfrontage. Mr Doran added 30%. This allowance was lumped together with an allowance for what Mr Doran termed as "form". (His adjustment for potential high ceiling development.) He said that his total adjustment of 50% included 20% for "form" and 30% for sea access. We will deal with "form" in a moment. 619. Mr Doran did supply evidence of sales of waterfront land to support his allowance of 30%, but Mr Brown accepted that 30% was reasonable for most flatted factory land. Mr Brown did argue that, if a site of the size of the subject site had waterfrontage and was used for flatted factories or go-downs, only about 25% of the site would benefit from that waterfrontage. 620. Mr Brown's argument concerning the user restriction cancelling out any added value for it having waterfrontage suffers from the same inconsistency as Mr Doran's argument concerning the location and access. We are using ground floor flatted factory accommodation value to determine what might be paid on that basis for the subject site given that the seller accepts that it would not be in a purchaser's interest to pay for any potential development in excess of the ground floor. 621. Flatted factory development over the whole of a site of this area would be unlikely, as Mr Brown said, to attract as much added value for waterfrontage as a flatted factory site of usual size. We accept that any factory or warehouse development that could adjoin the sea wall would be limited. We accept Mr Brown's argument on this point but we are only concerned with a ground floor development covering about one third of the site. The area is roughly equivalent to the area of the Sha Tin comparable where we have observed that no deduction up to that size appeared to be necessary. Mr Brown agreed that 30% was reasonable added value for waterfront flatted factory land. On that basis we accept that the accommodation value should be adjusted upwards by 30%.
622. We now come to Mr Doran's allowance for "superior form" which he made together with his allowance for waterfrontage. Mr Doran considered that an adjustment to land value was necessary for the fact that a steel-mill requires high ceilings. He added 20% for what he described as the "form" the building would take. There is no restriction on the site concerning building height or ceiling heights and we find it difficult in the absence of any supporting evidence to accept that a purchaser would pay extra for his intention to provide high ceilings. If he does produce a building with high ceilings, it would be at an extra cost. The additional building may add to the value of those buildings but there seems to be no basis for it adding to the value of the land. 623. Once a prospective purchaser's floor area requirements are met we do not believe he would pay more for his intention to provide high ceilings. 624. For example a man purchasing a home site would expect to pay no more for a piece of land whether it is a bungalow or a two storey house which he intends to erect, assuming there is no prohibition on either of these choices and that he is able to build to what he considers to be the optimum size in terms of floor area. His two storey house may cost more to build and be more valuable than a bungalow, but the land value is the same. 625. In other countries, as with the industrial estates at Tai Po and Yuen Long, the industrial developments within a locality may be many and varied but the land value is much the same. It is the norm, or what is regarded as the optimum, which usually sets the value. One industry's very large and expensive structure probably satisfies the return required in its case, while, nearby, another's small and basic structure fits its particular needs. Other issues being equal, the land value remains the same. Any additional value is in the value of the improvements. The land has a basic value which similar industries accept as the market. It is this basic value that we are seeking to establish by reference to other industrial sales. 626. We are adapting analysed A.V.'s, which are based on floor area, to the requirements of a steel mill, but we would not expect that the other similar sites adjoining SFI would not have had the same land value for similar areas even though the actual buildings may have differed. This is because the similar user restrictions dictated similar optimum development for those sites. 627. Even if we knew the A.V. for steel-mill use, which we do not, we would not have to make any adjustment for height as it should be included. We have a basic ground floor A.V. for flatted factory development and we really have no way of knowing anything further. We are concerned with what might have to be paid for an optimum ground floor development on an area basis. Our comparisons are on an area basis. There is nothing to prevent high ceilings being provided. It is a very general concept which is in danger of becoming convoluted if we try to refine it too far. 628. An example of this exists in Mr Lo's analysis of lot 1002. He made a deduction from his ground floor area by allowing 25% for parking and circulation on the ground floor. The finished development on lot 1002 is in single ownership and does have a building which only covers about 75% of the site with the balance apparently used for parking. There may be a case for no value or transferred value in the case of multi-ownership but rather than having no value as in Mr Lo's analysis, the open area would add value in this case. It could be expected to have been reflected in the the price paid. His extra refinement in deducting for parking space may look impressive, but on this occasion it was not correct. 629. While on the subject of the Sai Kung comparable, it might be as well to mention that, on inspection we also noted that the development on lot 1002 had a higher than usual ceiling for the ground floor. Thus, if there were any need for a high ceiling factor, in this instance, at least, it should be partly covered. It is, perhaps significant that it appears to have made no difference whatsoever to the analysed A.V. 630. We do not consider that there is any basis for making an adjustment for what Mr Doran termed "superior form". 631. Mr Brown made an additional adjustment for size of minus 20% which corresponded with Mr Lo's adjustment of minus 20% for user restriction. It was this minus 20% adjustment made by each for different reasons, which led Mr Carnwath to claim that Mr Lo's adjustments for lot 1002 at Sai Kung were similar to those of Mr Brown. 632. With respect to size we have already mentioned Mr Doran's view that no allowance is necessary and that for a mini-mill site an upward adjustment might be justified when comparing it with smaller industrial sites. Mr Brown drew attention to the usual and generally accepted practice of decreasing unit values as size increases. 633. If we are to look at our basic ground floor accommodation value for flatted factory use, we might expect an increase in area to conform with Mr Brown's contention that such unit area values could be expected to decrease once a certain optimum area has been satisfied. The evidence we have in this case leads us to make no such adjustment. As we previously noted, the area of 13,350 square metres for lot S.T.T.L. 261 at Sha Tin (Sale No. 2) shows a comparable ground floor accommodation value to that for lot 1002 at Sai Kung (Sale No. 3) which has an area of 3,231 square metres. The Fanling sales (Sales No.5 and No.6), for even smaller areas, also seem to conform with this observation. 634. With regard to these sales, whether or not the optimum size has been reached we cannot be sure; but it would seem that the smaller area at Sai Kung has not resulted in any premium being placed on it, or conversely any discount being evident for the larger area for Sha Tin. It is a fortunate coincidence that the area of the Sha Tin lot of 13,350 square metres is virtually the same as the 13,307 square metres calculated by Mr Doran as the area covered by buildings at Shun Fung. We are content that the accommodation values ate appropriate for the area of ground floor accommodation to which they are to be applied. 635. Mr Doran, we believe, has actually made an adjustment for size when he applied 1/10 value to the open area. It is an adjustment for a decrease in value beyond the optimum requirement and, in our view, is an appropriate way of making such an adjustment. The method adopted has therefore effectively made allowance for the size of the lot as the proxy value has only been applied to the area expected to be used for building which is of a size comparable to that in the compared sales' information. 636. Mr Lo's allowance of minus 20% for the user restriction requires some examination. We have ruled against Mr Brown's argument concerning the user restriction being accounted for by not adjusting for sea access. This exercise of using a more general industrial value to arrive at the value of this site, which has a very limited industrial use, suggests that an adjustment for this difference must be necessary. In Mr Lo's case we agree that it was necessary. Mr Lo adopted a plot ratio of 2.5 as a kind of standard. He then must have considered this to be more than that required for the Junk Bay sites by virtue of the user restrictions. He reduced his value by 20%. 637. Mr Doran and Mr Brown on the other hand, applied the flatted factory accommodation value to only ground floor development. Mr Brown used a plot ratio of 1 while Mr Doran confined his ground floor accommodation value to the actual area of buildings on the Shun Fung site. This was a more direct way of allowing for the limited use. 638. Mr Doran, by applying different values to the building area and the open land, not only allowed for the size of the site but also for the actual mini-mill use to which the site might be put. Thus Mr Doran's method, which we are adopting, adequately allows for the limited use. Therefore no other adjustment is necessary. Both the size of the site and the user restriction are covered by the method he adopted. 639. The appropriate ground floor accommodation value is therefore in the region of $1,500 per sq.m. to $1,550 per sq.m. We are prepared to adopt the higher figure. The calculation following Mr Doran's method is therefore:-
INDUSTRIAL ESTATES 640. Mr Doran drew attention to the land sold on the industrial estates of The Hong Kong Industrial Estates Corporation at Yuen Long and Tai Po. A brochure produced by the Corporation was admitted as exhibit SF29. 641. This brochure explained how land is sold to selected industries for manufacturing processes which could not be carried out in the usual multi-storey factory buildings. The object was to supply land suitable for industries requiring heavy floor loadings and high ceilings or wide areas of free space in which to install large machines. The Corporation described itself as a non-profit making organisation which sold the land at "close to cost". 642. We inspected these estates and found them to be of a high standard. The land, which was mostly reclaimed had been provided with bitumen roads, drains, sewers, electricity and water supply. 643. Mr Doran explained that the prices were pre-set and the land was only sold to screened specialised users. The prices around the date of reversion were :-
We were not told, so we do not know if we are to presume, that transactions actually took place at these prices. Until transactions actually occur there is no question of accepting these prices as being indicative of the market at that time. 644. These prices were raised on 30th October 1986 to $1,100 per square metre for Tai Po and to $900 per square metre for Yuen Long. 645. Mr Doran said he did not attempt to use this information as evidence of value, but rather as an indication of the minimum price at which industrial land was released. Mr Doran also considered these sites inferior to Junk Bay since they lacked waterfrontage. 646. We cannot agree that these sites were inferior to Junk Bay, but since the pricing is said to have no relation to market value but rather to cost, it would seem little, if anything can be gained from this information. Perhaps if there had been more data on the total cost of acquiring and producing this land compared to its price, some conclusions may have been possible. A private developer, for instance, would have had to follow a similar exercise of covering costs if he had been able to acquire the land and develop it. However, on the information we have, we do not believe we can take the matter any further. 647. Although we consider this land to be superior to that at Junk Bay, given the circumstances governing these sales and the fact that we have no evidence of actual transactions, we really have no way of telling whether these prices might have influenced the market for this type of land or not. PREMIUM FOR GOVERNMENT GRANTS TO PUBLIC UTILITIES 648. Mr Doran also checked his valuation by reference to what he believed the Crown would ask as a premimum if it were to grant the subject site to a public utility, such as an electricity power company, for a power station. 649. This approach seems to make too many assumptions to be of much use, but it does support a method based on industrial accommodation value for a development limited by the requirements of the proposed user. 650. Mr Doran's calculation proposes a plot ratio of about 2.5 for a covered area of 13,306sq.m.. This is a, slight variation from the 13,307sq.m. he used elsewhere but it is his calculation for the area of the buildings at Shun Fung. Mr Doran obtained a value for such a grant of $56 million ($1,555 per sq.m.) based on a ground floor accommodation value of $2,600 per sq.m. The A.V. of $2,600 per sq.m. was derived from his analysed ground floor A.V. of $2,000 per sq.m. plus 30% for water frontage. 651. If the Crown's valuers were to follow this method, then the ground floor accommodation value they should have used would have been about $1,500 per sq.m. based on the Sai Kung comparable with the deductions we described previously. It follows that the upper floor A.V. would be $375 per sq.m. (Upper Floor A.V. = Ground Floor A.V. divided by 4). 652. We have reworked Mr Doran's figures as set out below and arrived at a hypothetical premium of about $32.25 million or roughly $900 per sq.m.
653. Compared to our earlier exercise based on a steel-mill user, this resultis not surprising. It is the same method we adopted in the earlier exercise, but this time it is for a building having a plot ratio of 2.5 on a building area of about 13,000 square metres of the site, rather than for one with the lesser plot ratio of 1 over the same building area. The balance area of open land was valued at 1/10 the A.V. for the building area as previously. 654. It might be noted that Mr Lo's method was based on a 2.5 plot ratio plus an open area. He then deducted 20% for the user restriction. $900 per sq.m. - 20% = $720 per sq.m. Based on our adoption of this method, but for the steel-mill user, it seems his 20% was an insufficient deduction. 655. This exercise may be support for the method but it adds nothing to what has already been done to arrive at a value for the site. RESPONDENT'S OTHER COMPARABLES 656. Mr Brown produced a list of nine comparables in all. Five of them have already been mentioned with respect to the Crown land auction sites, and another was the sale at Yau Tong which we compared with that produced by Mr Doran. 657. Of the others, he produced two comparables for the purpose of showing that there was industrial land which was worth less than the minimum put forward by Mr Doran as $1,000 per square metre. Mr Doran relied on the sales of the waiver sites to support him on this point. We found that no reliance could be placed on the evidence associated with the waiver sites. 658. These two "comparables" were not completed sales. One was an unsuccessful tender for a large Crown land site of 2.9 hectares at Sheung Shui to be used for a private abattoir. The highest tender was $413.34 per square metre. The other also concerned the invitation from the Crown to tender for a small, industrial waterfront lot of 1,534 square metres at Peng Chau which is a small island on the Hong Kong side of Lantau Island. This Peng Chau lot which permitted industrial or go-down use as well as boat building and boat repairing received no applications. 659. Even if the waiver site sales had been useful, this information would not have been, as an offer to buy or sell is not evidence of value. If any figure is involved, it would be merely an estimate by the offeror of the value of the property to him. 660. This Tribunal has previously rejected such evidence in a number of cases. One such case was Chow Chi Keung v. China Light and Power Company Limited LTMR 1/82 where the evidence of an offer to purchase was rejected by reference to the High Court of Australia judgment of McDonald v. The Deputy Federal Commissioner of Land Tax for New South Wales (1915) 20 CLR 231 wherein Isaacs J. pointed out :-
661. We do not believe that anything can be gained from the evidence associated with these so-called comparables. Even as a test of demand for a certain type of property, it is suspect, as there could be innumerable reasons for tenders being refused or invitations to tender receiving no response. 662. Mr Brown also produced a sale of land by the Crown by tender in the Tai Po district. This was a small waterfront site designated Tai Po Town Lot 51. It was near Plover Cove, and adjoined the Government Fish Market on Tolo Harbour. The user condition permitted the repair, and maintenance and service of motor boats. It was presumably sold for the purpose of providing a service to the fishing boats. The site had an area of only 1,805 square metres which included 395 square metres of sea-bed to allow for the provision of a slipway. It sold for $1 million (about $550 per sq.m.) in December 1985. 663. Mr Doran considered this site to be too small for any meaningful comparison to be made. We have to agree. However, on the basis of this being a waterfront site, we cannot help but note that Mr Doran's analysis of this sale assumed the permitted built over area to be worth three times the open area. This gave him a rate of $1,300 per sq.m. for the built over area. As this site, by virtue of a height restriction, was limited to a ground floor development, this figure seems to lend some measure of support to the $1,550 per sq.m. ground floor A.V. adopted for the subject site. We would not wish to take the argument any further, and really place little or no reliance on this sale. 664. The only other comparable used by Mr Brown was a Crown land auction site at Tsing Yi. Described as Tsing Yi Town Lot 92 it was also the first (Sale No.l) of Mr Doran's list of Crown land auction sales of industrial land. It sold for $7 million on 12th December 1985 and had an area of 2,554 square metres and a permitted plot ratio of 5. Mr Doran's analysis showed it as having a ground floor A.V. of $1,370 per square metre. Mr Brown, adjusted for time, size and location, but not for sea access, and thereby converted the rate of $2,745 per square metre as shown in the sale price, or an overall A.V. of $549 per square metre to reflect an A.V. of $357 per square metre as applicable to the subject site. These adjustments were set out when we dealt with the Crown land auction sites. Little was made of this sale by either of the valuers and it was not mentioned in either counsel's final submissions. 665. We do not find it a very useful comparable having very few similarities with the subject land. However, if we were to take Mr Doran's analysed ground floor A.V. of $1,370 and adopt Mr Brown's adjustments of plus 5% for time and minus 20% for location, and ignore size, which is not relevant on this basis, but add 30% for sea access, the resultant ground floor accommodation value is $1,496 per square metre. That figure was derived on our preferred method of adjustment in sequence. On the aggregate basis, a ground floor A.V. of $1,575 per square metre results. 666. This, perhaps, might be looked at as some small measure of support for our adoption of $1,500 to $1,550 per square metre for the ground floor accommodation value for the subject lot based on the Sai Kung and Sha Tin sales. SUMMARY 667. There was not much to choose between Mr Doran and Mr Brown as witnesses, since each identified with his client too readily. Rather than opt for the opinion of either one or the other of them, we have made our decision on the basis of our own evaluation of their evidence accepting some of their propositions and rejecting others. As much as possible, we have refrained from introducing any opinion of our own as to method or level of adjustment to be applied. 668. The auction sales at Sai Kung and Sha Tin were the best evidence of value available to us in this case. The rest was of little assistance. Of the claimant's "direct comparables", only the Yau Tong Sales were evidence of value, but there existed differences which were too great for them to give a firm indication of value. A comparable value of $2,000 per square metre at or about the valuation date does not seem to be unreasonable compared to the overall value of $660 per square metre deduced for the subject site from the auction sales. There is perhaps a closer relationship with the built over value of $1,500 to $1,550 per square metre, but, as the optimum development for the Yau Tong sites is unclear, it would be unsafe to attempt any further observations on such comparative values. 669. Mr Doran's adoption of $1,500 per square metre was arguably a reasonable deduction from unreliable sources; but, on this basis, he made no adjustment for size. That is, at that stage, he made no allowance for the fact that a steel-mill site required open space as well as building area. He remedied this when he analysed the Crown land auctions for ground floor A.V.'s and applied 1/10 value to the open area. 670. Mr Brown dealt with the valuation on too broad a basis of comparison. He seemed to be overly concerned with the restrictive user condition and the basis of the Crown's earlier settlements for other resumed sites. His adoption of a plot ratio of 1 as the optimum for the subject site would seem to assume far more building than was necessary for a steel-mill on the subject site; but he did counter this by using overall accommodation values. We believe his more conventional broader analysis of the accommodation values reflected in the auction sales was not sufficient a break-down in this case to allow for the substantial differences in the comparables of the ratios of the areas for the potential ground floor development compared to that for the upper floors. 671. However, if we include an allowance for sea access which Mr Brown did not do, Mr Brown's broader concept of analysis and application based on the overall accommodation values reflected in his comparables, does give some interesting results. He adjusted for time, location and size only. In the case of Lot 1002 in DD215 (Sale No. 3 on Mr Doran's list) at Sai Kung he obtained a rate of $490 per square metre, and for S.T.T.L. 261 (Sale No. 2 on Mr Doran's list) at Sha Tin he arrived at a rate of $500 per square metre as applicable for the subject site. These were based on an aggeration of his individual adjustments. 672. If we add 30% for sea access to these adjustments, but on the sequential basis which we have adopted, we get $660 per square metre from the Sai Kung sale and $662 per square metre from the Sha Tin sale. While the comparisons are perhaps too broad and the adjustments lack supporting analysis, the end result endorses the $660 per square metre that we have obtained using Mr Doran's approach. 673. Taking all the evidence into account on the comparative basis as submitted, we determine the value of the land, as at the date of resumption at $23,750,000. This is a rate of $660 per square metre. INTEREST ON LAND VALUE 674. Whether there should be an allowance for interest on the land value remains an issue. The claimant submitted, based on Mr Doran's report and evidence, that the value of the land as "bare land" was $54 million but an amount of $10.157 million should be added for interest on this figure. This makes the total amount claimed under Land Value as $64,157,000. The sum of $10.157 million represents a rate of 9% over the 2 years estimated as the time it would take to develop the site from its bare land status to that of a completed mini-mill. On our assessment of the Land Value, 9% for 2 years represents a figure of $4,225,000. 675. Mr Brown, on behalf of the government, did not agree that interest should be allowed. However, the government had no argument with the rate of 9% being appropriate as the price of money at the time in question or the period of 2 years being a reasonable estimate of the construction duration. 676. The basis of Mr Doran's argument was developed from his proposition that, apart from the value of the buildings themselves, land when it is combined with buildings is more valuable than when it is bare land with no buildings on it. He referred to this added value as the "intrinsic value" which a building on land added to the value as a bare site. He said it was due to a "time benefit" which a purchaser saved by not having to build. Mr Doran illustrated this in his evidence by reference to Exchange Square which is a relatively recent large office complex in Central Hong Kong and is what he described as a functional building. He said that if one took the replacement cost of the building from the overall value of Exchange Square, the figure left for land would be very much higher than the figure one would expect to pay in the market if the site were vacant and awaiting development. 677. Mr Doran also set out his position in SF258, where he drew our attention to an illustration of two adjoining sites, with one having a new building and one having a building 20 years old. Mr Doran said that, "A purchaser would be expected to 'reduce his bid' on the site with the old building but, he could hardly make any adjustment for the land as both sites are identical for all practical purposes." We accept this as fairly obvious, and that the land value in both cases should be the same; but, Mr Doran was suggesting that any building, be it new or old, would add the same amount of what he describes as "intrinsic" value to the figure representing the market value of the bare site. 678. Mr Doran drew attention to the English Court of Appeal case of Oldham, Ashton and Hyde Electric Tramways Limited v. Ashton Corporation and Others (1921) 3 KB 511 which examined the depreciated replacement cost method. He said that this case supported his contention that any interest associated with land value should not be depreciated. We will deal with depreciation in due course. What we would mention at this stage is that in the lower court, (1921) 1 KB p.269, the Oldham case contained references to the interest on capital during construction being a cost associated with improvements whether or not it was actually paid. "If one spends #100 on buying something in 1919 which brings in nothing until 1920 one spends on that not only the capital laid down, but also the value of that capital lying idle for a year, and that is part of the cost." Here we have support for including interest on any capital outlay in the replacement cost estimate. Capital used to purchase the land would seem to fit the description of "capital lying idle". 679. Mr Doran believed that the RICS Guidance Notes on "The Depreciated Replacement Cost Basis of valuation" supported the allowance of interest on land value. A copy of this document was attached to his report Document 34 at appendix A. He felt that the following passage in paragraph 3 of this document supported the allowance of this interest element. "If the land was to be looked upon as a virgin site being offered for sale in the open market, the price that would be obtained would allow for the fact that it may take a purchaser many years to carry out the development." (underlining added). 680. On cross-examination it became apparent that at best this could only be regarded as a vague reference to the need to allow interest on land. In fact, with respect to these notes, Mr Doran considered paragraph 4.2, which Mr Carnwath put to him as also dealing with this aspect, as wrong when it stated;
681. We would comment that we found these notes to be somewhat ambiguous. We agree with Mr Doran that if the text of paragraph 4.2 were intended to mean that the estimated cost should be based on an earlier date than the valuation date, then such a concept is indeed wrong. This statement led to a later argument from Mr Carnwath on whether the depreciated replacement cost method could be applied retrospectively, and we shall deal with this later. 682. For the present we would mention that the "Guidance Notes" at appendix A, lent scant guidance on the depreciated replacement cost method. 683. Guidance was sought from an extract at page 109 of "Land Valuation and Compensation in Australia" by Rost and Collins which stated in respect of the depreciated replacement cost method:-
While this quotation gave the best support to Mr Doran's allowance for interest on land value, he also pointed out in his statement in SF258 that; "Nowhere does it say that the land carrying costs should be based on a depreciated figure. Indeed, I consider it would be illogical to do so." 684. The lack of any reference to depreciation is not surprising as only cost and not depreciation was being addressed by the authors in this paragraph. In their next paragraph they explain that, "Depreciation will be discussed later in this chapter, but it is first necessary to consider costs in some detail and to explain methods which may be used to ascertain current replacement costs of improvements." 685. What is being confirmed in the earlier paragraph from Rost and Collins is that the total replacement cost of improvements is more than the mere estimate of the contract price (i.e. "costs directly connected with construction") which might be paid for any structures on the land. It also includes all the other costs related to a development up to the time that development is completed. Such costs include interest on land. It follows that any omission of such carrying costs when deducting the cost of construction from the total overall value could easily lead one to believe, as Mr Doran does, that the land element, when there are buildings, is more valuable than a bare site, or, when no buildings exist. 686. The interest allowance on the land is an opportunity cost and does not need to be an actual payment. The allowance recognises that money outlayed can only begin to realise some gain when the development is completed, and that, therefore, interest is only lost for the time the money is tied up within the development period. In other words, when land is unproductive during a development, interest on the value of the land is forfeited over the total period of the development. It is one of the costs in providing those improvements. So as pointed out by Rost and Collins, it should be included as one of the costs in the overall replacement cost assessment. 687. Thus we have support from Rost and Collins for the inclusion of an allowance for interest on the land. What we need to examine is the belief that it would be "illogical", as Mr Doran contends, to include interest on the land with the other replacement costs when considering how much those costs should be depreciated to arrive at a market value for the property as a whole. 688. We now turn to the statement, cited to us by Mr Doran from Oldham Ashton and Hyde Electric Tramways Ltd v. Ashton Corporation and Others, (1921) 3 KB 511, that "interest on capital during construction should be subject to depreciation in so far as the capital on which it was the interest was subject to depreciation - that is to say, according to whether the capital was spent on a depreciating thing such as rails or on a non-depreciating thing such as, excavation". Land is a non-depreciating thing, so on this authority, Mr Doran says that interest on land is not subject to depreciation. 689. Rowlatt J. in his judgment at first instance, (1921) 1 KB p.269, with reference to non-depreciable items, said, "One spends the interest one has to forgo just as much as one spends the capital sum. Therefore this item is depreciable or not depreciable with the capital on which it is the interest … I conceive that much of the capital was sunk in non-depreciable things like excavation and acquisition of powers, for I suppose the expense of obtaining powers was expense provided out of the capital. On the other hand much of the capital was spent in making tram lines, which is a depreciable sum. Therefore it seems to me that the sum of interest on capital must be analysed and divided and that the award must go back to the arbitrator for that purpose if the parties cannot agree on a sum." 690. On the face of it, Mr Doran's statement that interest on the land price is a cost which cannot be depreciated seems to be correct. At first Mr Carnwath had no argument with it. We believe that the fallacy in this argument lies in the fact that interest on the land does not add to the bare site value but rather to the value of improvements which are a wasting asset. Interest is an important factor in residual and depreciated replacement cost assessments. Interest lost on money locked into a development is a cost against that development. It is misleading in our view to think of interest only in terms of a discount or deferment in valuation exercises. 691. The Oldham case centred on interpreting and determining a specific value entitled the "then value of a tramway and all lands, buildings, works, materials and plant of the promoters suitable to and used by them for the purpose of their undertaking within such district, but excluding any allowances for past or future profits of the undertaking". Thus, any alternative to the depreciated replacement cost approach was ruled out. 692. Rowlatt J. appeared to be critical of this description of the "then value", when he said, "One is met with very great difficulty in understanding it because when one is discussing the value of a commercial undertaking the element of value is its profits and there is none other. To seek to find a value without looking at the profits is like seeking to build a house without materials." 693. This is, perhaps, a pertinent remark in respect of the present case where valuation based on the itemised depreciated replacement cost approach has been favoured. 694. Rowlatt J. went on to say, "If one disregards profits the value seems to be the scrap value, but that is not what was meant. The only way in which to deal with the matter, a way which is well established, is to take an estimate of the cost of constructing the tramway at the time when the "then value" is to be ascertained, subject to depreciation in order to get at the actual value at that time". This seems to be a fair description of the depreciated replacement cost method. 695. In that case, the question of interest on land was not addressed. In fact the land element was linked under an agreed item which went to the heading of "Value of lands, buildings, office furniture and fittings, car shed track, car shed equipment etc.". The question of interest on plant did not arise, either, as the value for "Plant, tools and fittings and car shed and plant and fittings belonging to the Tramway Co Ashton Power Station" was also agreed. 696. The dispute lay with items listed as "Amount allowed for preliminary expenses", "Amount allowed for miscellaneous items", "Amount allowed for cost of raising capital" and, what concerns us because it was the only matter to address depreciating or non-depreciating items, "Amount allowed for interest on capital durincr construction". 697. Certainly, the judgment holds that, in a totalling of itemised costs approach, interest should be depreciated only if that item depreciates. This, we agree, makes sense when it relates to such costs as the example given by the court in respect of excavation cost. We can see that interest on such a cost should not be depreciated when it is not a cost associated with a wasting asset. 698. The cost of excavation like any formation costs adds to the value of the land. It is a cost which contributes to the formed bare site value which would not be expected to depreciate. In the present case we have not had to consider the costs of formation as the land has been compared with other formed sites, and we agree that direct comparison of similar properties which are improved to the same or a similar state, is less liable to error than where there is an addition of various costs particularly when they may have to be depreciated. 699. If we had had to arrive at the formed bare site value by adding the cost of formation to an unformed site value, the interest on the formation costs would not have been subject to depreciation as it would have formed part of the land value. However once that value has been determined the interest lost on the land value during the construction of the improvements is not interest associated with the non-depreciable land but a cost associated with the construction of the improvements. It is only attributable to and part of the cost one has to bear when producing a new building in the same way as the "cost of engineering" referred to in the same judgment is a cost associated with providing a new tramway. These engineering costs were subject to depreciation as the cost of a new tramway is a wasting asset just as the cost of providing a new mini-mill is a wasting asset. 700. In adopting this method, we have to establish the value of the formed bare site at the valuation date as one head and the value of the improvements as another head. The formed bare site does not depreciate but the improvements do. In this instance the time over which the interest is allowed can only be the period estimated for the construction of the improvements and forms part of their cost. It is a cost which would recur any time the improvements were replaced and therefore should be included in any assessment of their replacement cost. 701. Of course, loss of interest occurs merely by virtue of land lying idle for a number of years. It may well be one and the same thing but such loss of interest does not necessarily have to be associated with the construction of improvements over a development period. Also, while interest may be lost on the initial price, land will usually appreciate in value over a period. Therefore one gain may well cancel out the other loss. More often than not the appreciation in value will be greater than any loss based on the cost of money and hence more than the amount of interest lost over any given period. 702. Mr Carnwath, influenced by para. 4.2 of the RICS "Guidance Notes", which seemed to envisage the depreciated replacement cost method being applied retrospectively, looked at the loss of interest on land in this context and proposed alternative approaches.
He preferred the second approach. This would eliminate any need to allow for interest. Government submitted that "Mr Doran seeks the best of both worlds by taking the inflated land value at the date of resumption and adding interest to that. This is double counting". 703. Therefore Mr Carnwath's argument acknowledges that, in the course of a development one pays, in relation to the land, not only the purchase price, but also the interest which is lost during the period of the construction. He also argues, however, that, at the end of the period the land value will usually have made up for any interest which may have been lost. Thus, if the date of valuation is taken as that of the date of the purchase of the land, undepreciated interest will become part of the cost of the land. On the other hand, if, with respect to the development, the date of completion is taken as the date of valuation, the land value then may well be greater than the original price paid at the date of commencement, and it would not be correct to add interest to that figure as the land is no longer tied into the development period which is now completed. One cannot adopt the second alternative as Mr Doran has done and then add interest which applies only to the first alternative. 704. With respect to Mr Carnwath's argument, we do not accept that there can be alternatives. The depreciated replacement cost method requires the valuation to be based on one specific date, and, certainly under the provisions of the ordinance there is no option but to value the land as at the date of resumption. The value of the buildings must also be based on that date. We have said before that we agree with Mr Doran that basing the replacement cost estimate on a date earlier than the valuation date is wrong. We are not entitled to base any assessment of value or cost from any date prior to the valuation date. 705. If we were to adopt a date 2 years before resumption as the starting point, we might get the appropriate costs right, but this would only give us a total cost. It would not give us the value of either the land or for that matter the replacement cost of the improvements at the date of resumption. 706. What is required is the determination of the value of the "formed bare" site at the date of resumption and no other date. If one is valuing the land as if it were vacant, then it can only be assumed to be vacant before any construction. It is what one would expect to pay for land assuming one intends to develop the land. The starting point is the notional purchase of the land at the relevant valuation date, and the land value should relate to this. 707. The depreciated replacement cost method is based on the estimated cost of construction as determined at the valuation date and interest would still have formed part of the cost which has to be calculated at that date. It is not what the cost would have been 2 years before or, for that matter, what it might be in 2 years time. It is a replacement cost assessment at the valuation date. For that reason no inflation can be assumed but interest is a part of the cost of producing the improvements. Clearly, with respect to lost interest on cost, a building which takes a long time to build will be no more valuable to a prospective purchaser than one which is built quickly. Nor is a property more valuable because the owner happened to hold the site for some time before developing it. The interest element can only relate to what is a reasonable period for such a development. Here we are told that is two years. 708. Mr Doran's approach seeks to isolate the value which is due to the interest on land element and which he describes as "intrinsic value due to a functional building on the land" at the date of valuation. The actual building cost may be the largest part of the added value to the site, but it is only part of the total cost which affects the value of the improvements on the land. There is also the amount the owner or developer is out of pocket while the building is being erected. It is part of the total cost and becomes part of the value of the improvements. There does not appear to be any reason why any element of added value should be placed under a separate heading. 709. Mr Doran does show that his "intrinsic value" is attributable to interest on land; but the interest on the land is only one of the items which make up the carrying costs over the development period. Even if it were some previously unexplained phenomenon, it does not justify its separation from the value of the improvements which is the added value to the bare site. 710. If the exercise is to value the land as if it were vacant and ready to be developed, then the value of the improvements is the only other element making up the whole. The value of the improvements, as Rost and Collins have pointed out, should include any interest lost over the construction period, be it associated with the estimated land price or the estimated contract price of the buildings, and, therefore, there is no need to consider the point of whether old buildings or new buildings result in the same "intrinsic value", as this so-called "intrinsic value" forms part of the value of the improvements. 711. It is, we think, clear that carrying costs cannot form part of the land value and that interest lost on the estimated purchase price of the land as part of those carrying costs becomes part of the cost of the development of the improvements to be placed on that land. These improvements will depreciate and the land will not. Therefore the estimated cost of replacing these improvements including the interest lost on the land should be depreciated if the property is to be valued when those improvements are no longer new. 712. When Mr Doran sought to explain that "intrinsic value" was due to the interest on the land and existed when a site was developed, drew attention to it as the "the sometimes forgotten cost". He went on to explain his approach as a "reverse residual" exercise. 713. A "residual valuation", as its title suggests, is the process of working backwards from an overall value to determine the land element contained therein. As the reverse of a reversing process, what Mr Doran is describing is really a "summation method" whereby every element of value is included in a sum-total. Since it has been raised, let us examine the "residual method". 714. We agree that a mere deduction of the estimated contract price of a building in a residual exercise will not result in the value of the land as a vacant site. It has been demonstrated and recognised that the total value figure would be expected to embrace all incidental carrying costs including the interest over the development period and perhaps, when appropriate with respect to a new development, a margin for profit and risk which the developer could reasonably expect if he were to sell the completed development. 715. We feel it necessary to mention profit and risk because Mr Doran raised it and Mr Neoh seemed to be unsure as to its relevance when he sought to introduce the residual method into his argument on this subject. It is usually only appropriate where the object of the hypothetical optimum development is resale. We do not think that it is appropriate in the case of a hypothetical steel-mill development. 716. After deducting all the costs and the profit element, if any, one is left with what one might expect to have to pay for the land, if one were purchasing it as a vacant site suitable for development at the date of valuation. 717. If there was a problem with Mr Doran's "intrinsic value", we hope that it has now been satisfactorily identified. It is most likely the interest on the land but it could also easily include any part or all of the carrying costs as well as profit and risk. 718. Interest on land is not part of the value of the land for the same reason as any carrying costs are not part of the value of the land. There is absolutely no reason why land value as a vacant or "bare" site should be distinguished from land value when there is a building on the site. Both should be considered on the basis of a vacant site ready for development and both should be the same value as Mr Doran, actually showed in his example of the two identical sites with one having a new building and one having an old building. 719. Thus, what Mr Doran believed was a "time benefit for a piece of land which already has buildings on it" is part of the value of the improvements. It occurs with and only because of the construction of the improvements and is wholly associated with those improvements. 720. This time benefit is maximised only when the building is one which utilises the site to its highest and best use. It must also be new or near new such as Mr Doran's example of Exchange Square, for as a building depreciates so does this so-called "time benefit" until eventually the land value only is left. In fact, when the building becomes completely obsolete, the property will become worth less than its bare site value. The building has reached the stage where it has become a liability and the cost of its demolition has to be added back to determine what the property as a bare site on which to redevelop might be worth. In truth, the so-called value of the improvements can be negative. 721. This situation is not unusual, and is observable in Hong Kong in relation to pre-war buildings subject to rent control. The value based on the revenue such buildings are capable of generating from controlled rents is invariably far less than the value of the site based on its value for redevelopment. The site value is more than the value based on the income which the property can produce in its present state, because the property is not being put to its highest and best use. 722. In Shun Fung's case the site can only be used for an undertaking associated with shipbreaking, (i.e. a mini-mill). The mini-mill is the highest and best use. Thus, the method adopted, which includes all these elements of value, is appropriate, but as the mill depreciates or comes closer to being obsolete, so does any time benefit of having those improvements on the site decrease. 723. Far from it being illogical, to depreciate interest on the land as part of the carrying costs in the depreciated replacement cost method, it is, actually, illogical not to do so. Mr Doran's approach would result in the absurdity of a site with an obsolete building on it appearing to be more valuable than a bare site ready for redevelopment. 724. If we go back to Mr Carnwath's proposed alternative valuation dates, we hope we have now demonstrated there can only be one valuation date, and scope for confusion will be reduced if regard is paid to the notional land purchase as at the valuation date as the starting point. Thereafter, any interest lost on the cost of the land, or for that matter any other construction costs assessed at the valuation date for the period of construction, is part of those estimated construction costs. It is also clear, we hope, that, based on a valuation date, any depreciated replacement cost approach should include the interest on the land, as it is a cost element associated with the time it would reasonably take to complete the construction of the improvements. 725. As Mr Carnwath has observed, the only way undepreciated interest on land can form part of the cost of the land itself is for it's estimated price to be back dated for the period of construction. That is the total cost of land is equal to its original price 2 years back plus interest for 2 years. But this is not the value at the valuation date and we are only concerned with the value of the land as a vacant, but formed site at the date of resumption. The cost of the land deferred back to the commencement of a construction period plus interest is of no concern as it would be extremely unlikely to be the same as the value at the date of resumption. Cost should not be confused with value. 726. Mr Neoh drew our attention to the words of Scrutton L.J. in the Court of Appeal judgment at page 527 in the Oldham case, when he affirmed the judgment of Rowlatt J., "Some items of cost are once and for all and will never occur again, so that a purchaser does not get any depreciated value but the full value of an expenditure which will not occur again." He said this supports Mr Doran's contention that the buyer of a completed development gets the benefit of saving the carrying costs whether the buildings are new or not. Mr Neoh's argument fails, so we think, by virtue of the fact that carrying costs occur with the replacement of the improvements or each time a property is redeveloped. Therefore such costs will depreciate along with other replacement costs as a building ages. 727. Perhaps the words of Atkin L.J. at page 528 in the same judgment, expand on the point made by Scrutton L.J. and also sum up the whole matter :-
728. While providing a fitting summary this statement also endorses the same "once and for all" test for an item to be non-depreciating. Atkin L.J. further describes it as a cost which "will never require renewing" such as "the cost of excavation". As we have already pointed out, the cost created by loss of interest on land value over a development period recurs each time land is redeveloped in the same way as the cost of engineering and other carrying costs would recur. Quite simply it is not a cost which can be categorised as "once and for all". 729. As Rost and Collins have indicated, "such charges represent a substantial addition to the construction cost". Because interest is a cost, we believe that it should be possible to regard it as such, both in a "residual" exercise as well as a "summation" valuation. Perhaps if it were looked at in this light, rather than as a deferment, it would not be "the forgotten cost" as Mr Doran so aptly described it. 730. We consider that any interest element related to the land value during an assumed construction period is not part of the value of the land, but, part of the value of the improvements, which, over a period of time, will depreciate in value. 731. It was explained to us from the Bar table that the agreed figure for the buildings of $26 million did not address the amount itemised by Mr Doran as the interest on land. It was not included in their agreement. We are now left with the problem of how this interest on land element should be incorporated in an award. 732. Mr Carnwath's submission was simply that interest on land should not be applied. His argument, as we set it out earlier, described Mr Doran's approach as double counting. Mr Neoh for the claimant said that, if the Tribunal held that interest on land should be allowed, but that it should be depreciated, he would seek leave to address the Tribunal as to what that rate of depreciation should be. 733. On this point the interest element in a depreciated replacement cost assessment would be more significant in a new or near to new development. The buildings in this case may have had some remaining life, but they were mostly in the region of some 20 years old. They had been subject to fair wear and tear, and were, by comparison with the proposed layout of the mill at Shunde, undoubtedly subject to some measure of obsolescence. 734. Agreement as to the value of the buildings occurred during the course of the hearing, but before any evidence on the subject. Hence we heard no evidence on the depreciation rates to be adopted for the buildings. However we have noted that the government's experts, in their reports, gave an economic life of 20 to 25 years for the main steel buildings while the claimant's experts opted for 40 years. The ancillary concrete buildings were given longer economic lives, but we do not believe that the majority of those structures would have been retained beyond the estimated life of the main buildings which we believe were also tied to the life of the plant. 735. The evidence of the steel experts, Messrs Medley and Wilcox, as well as Mr Blickle who valued the plant for government, suggested that a different layout, similar to that proposed for Shunde, would be adopted if the plant were ever replaced. Mr Medley estimated the overall life of the plant ending in about 1999. (Doc 30/01, Section 4). On this basis, we would expect an overall depreciation rate for the main buildings, completed in about 1966 or 1967, somewhere in the region of at least 60%. 736. We note that Mr Best, who assessed the value of the goodwill on behalf of the claimant relied on a figure of $15.023 million as the value of the buildings as at 1999. On a straight-line basis, this is equivalent to a depreciation rate of only 40% for the buildings as at 1986 (the resumption date). This means he has given an overall life of 50 years for the buildings. This seems excessive particularly when, the claimant's expert on this subject, Mr John Smith of McLennans in Document 32/01 Section 6, mentioned that plant and buildings for steel mills are usually amortised over a period of 15 to 20 years. However, he qualified this when he said there were many examples of buildings erected 30 or more years ago remaining perfectly suitable for their duty. If the plant and buildings are considered to have an economic life up to 1999, the life of the buildings will have been 33 years. An overall life of 33 years seems to accord with Mr Smith's remarks. 737. The government's experts were of the view that the buildings had all but used up their economic lives. Depreciation rates of 90% were recommended for the main buildings. In such circumstances, an overall depreciation rate of 60% seems to us a minimum. On that basis the interest on the land element should have been no more than about $1.7 million ($4.225 million less 60% = $1.69 million) or, more likely, about $1.5 million. 738. With regard to the possible overall depreciation factor and its effect on any amount attributable to interest on land value, we should, perhaps, take cognisance of the arbitrariness of the agreed figure of $26 million, and note its having been determined as a compromise following what Mr Neoh described as "horse-trading" where there was "neither rhyme nor reason". 739. Accordingly, due to the fact that the parties were poles apart to start with (Some $36.5 million as against $8 million if the unauthorised buildings were included or $1.15 million if they were not), we see no reason to assume that the agreed figure of $26 million would have been any different had the parties been aware that an amount of about $1.5 million representing depreciated interest on land was included within that figure. In spite of our concern that interest should be included in such a valuation assessment, the present occasion, however, was one where it could easily have been accounted for in an arbitrary figure such as that agreed on the basis described. 740. The agreement, with which the Tribunal cannot interfere, was for the total value of the buildings, and depreciated interest on land is part of that value. 741. Mr Neoh was adamant that interest on land was excluded from the agreement. The government confirmed this. The government's contention was, all along, based on it being inappropriate to include a separate amount for interest on land. Its agreement to $26 million for the value of the buildings was on the understanding that this figure represented the total value of the buildings. So, too, was it the total value as far as the claimant was concerned, for it was the claimant's argument that interest on land was a part of the claim associated with the land value. 742. It being agreed that the value of the buildings was $26 million, then, whether, acknowledged or not, within that figure there is an element for carrying costs which includes interest on land. To allow it again as a separate item would be double counting. 743. We do not, therefore, consider it appropriate, in the circumstances, to make any separate allowance for interest on land, because it is already incorporated in the value of the buildings, which has been agreed at $26 million. SECTION VI : PLANT AND MACHINERY VALUATION METHOD : DEPRECIATED REPLACEMENT COST 744. Both the claimant and the government have adopted the depreciated replacement cost approach to value the plant and machinery. We understand that this is the approach usually relied upon as the sale of industrial plants as whole operating units is rare. AGREEMENT AND DISAGREEMENT 745. Only eight major items of plant remain in dispute, the parties having agreed on a figure of $7,290,730 for the other items, excluding the scrap yard items which were separately agreed at $3 million. 746. The disputed values of the eight major items arise mainly by virtue of disagreement on the rate of depreciation to be adopted. 747. There is also disagreement as to the value of spares, and what should be allowed for Design Engineering and Project Management as well as Contingencies and Interest. We set out the two assessments below :-
COST OF REPLACEMENT 748. Much of SFI's plant was not of the latest design and was no longer being produced as at the date of resumption. Thus, the valuer's task was complicated by the replacement cost of some items needing to be assessed by reference to the latest, closest equivalent equipment, which was up to date in design, and technology. In spite of this, during the course of the hearing, Mr Wood, who valued the plant and machinery for the claimant, and Mr Blickle, who carried out the valuation on behalf of governemnt, were able to agree on the replacement cost of most remaining items. These replacement cost assessments were labeled as, "Cost of Replacement New" or, "C.R.N.". 749. Mr Neoh, in his submission on SFI's behalf, appeared to cast some doubt on those agreements. Mr Wood had said he accepted the "Government's figure for these items", but then qualified this with, "I should make it clear that I do not necessarily accept their quote in its entirety since their quote does not give sufficient specifications to enable me to determine whether it is identical to the original items at Junk Bay". 750. That was a curious statement, for the quotes relied on had to be for new items of plant which were the closest in type and design to the originals, but, in many cases, were undoubtedly different in some aspects from the originals at Junk Bay. Adjustments were made for some of these differences in specifications, and Mr Wood had accepted the estimated replacement cost figures so derived. 751. Mr Neoh's argument seems to be with the circumstance that Mr Blickle considered most of his replacement cost assessments (C.R.N.'s) erred on the generous side. Mr Blickle believed there were elements of "betterment" in the quotes. Sometimes he reduced his C.R.N.'s for this. Some he was content to leave with the qualification that some betterment existed in the C.R.N. adopted. 752. We understood that Mr wood agreed with the Crown's C.R.N. figures for the listed items of plant as they stood in the Shun Fung mill at the date of resumption. His qualification does not resile from this, as it refers only to the source of determining those figures. He has accepted the Crown's C.R.N. figures even if he was unsure of, or not completely in accord with, Mr Blickle's methods of adjustment of the initial quotes. Whether or not he agreed that those figures were generous, to our mind, does not affect his valuation. 753. We fail to see how Mr Neoh's arguments can alter that agreement. If they were to, it might well make this Tribunal's task impossible. 754. We will in due course, take each item in turn, and take account of what Mr Neoh said about each of the C.R.N. figures. DEPRECIATION 755. Disagreement exists mainly on the discount for depreciation to be made to the cost of replacement new (C.R.N.) to arrive at the value of the remaining major items of plant at the date of resumption. 756. While the method adopted by each of the valuers was essentially the same, they tended to work from opposite directions when it came to considering depreciation. 757. Mr Wood, who actually saw the plant in operation, at the outset took a view on the remaining life of each item of plant without knowing its actual installation date, since he was not given access to the plant register at the time he made his inspection. 758. Mr Blickle allowed for depreciation by working from an estimated economic life, and the actual age of each item as disclosed by the plant register or other reliable sources. He used the date of installation to calculate depreciation due to age on a straight line basis i.e. in the ratio age bore to life. Mr Blickle, where appropriate, made an adjustment for obsolesence by reducing the physical life, or by prolonging it where there had been upgrading or refurbishment since installation. 759. Mr Blickle, who never saw the plant in operation, had, nonetheless, inspected the stored items of dismantled plant at Liuzhou where they awaited installation, having been purchased at auction from the Crown following resumption. Mr Blickle was also assisted by a report and photographs compiled by Mr Ian Phillis, a Chief Building Services Engineer with government. Mr Phillis like Mr Wood, had visited the Junk Bay site with the plant and machinery still in situ prior to the date of resumption. 760. Each valuer adopted the standard practice, as described in National Telephone Company Limited v His Majesty's Postmaster General [1913] T.L.R. 190 of taking the cost of replacement new, and depreciating it by taking the value as reduced in the ratio age bore to life of the item in question. For example, the age against economic life depreciation adjustment for an item of plant with an economic life of 20 years and 5 years remaining life, would be 75% (20 - 5 as a percentage of 20 = 75%). As value is not a matter of how bad something is, but rather how much good remains in it, it is perhaps better put as 25% of its replacement cost new would be its remaining value. Further adjustment might sometimes be necessary if, for one reason or another, this straight line method of depreciation was not appropriate for a particular item. LEGAL SUBMISSIONS 761. The claimant was concerned at the fact that it had ceased improvements to its plant from the time it was informed of the proposed resumption in November 1981. Mr Neoh was prompted to ask Mr Blickle if the value would be different if 1982 technology, rather than 1986 technology, was in place at the resumption date. 762. While SFI had more or less completed a comprehensive upgrading programme by the end of 1981, there seems to be little doubt that it would, under normal circumstances, have continued to upgrade its plant. 763. Mr Neoh was concerned that it might not be right in law to value SFI's plant on the basis of 1986 technology. That argument was fortunately not pursued in that form, but, on a somewhat similar tack, Mr Neoh submitted that, "The Tribunal has to value the loss to Shun Fung's business resulting from the loss of its plant and not a plant with the latest technological advances". 764. If Mr Neoh is intending to say that we are to determine the value of the plant as it stood at the resumption date without the benefit of the latest technology which was available at that time, he and government will have both adopted what we regard as the correct approach. But if, as we suspect, he is suggesting that we should attach no significance to what is state-of-the-art technology when it comes to valuing SFI's plant and machinery, we cannot agree with him. 765. A valuation should not disregard deficiencies or inadequencies. Up-to-dateness in technology was a factor which had to affect the value of SFI's plant. If it fell short of the technology available at the valuation date, that most surely had to be reflected in the valuation. 766. As Mr Wood himself showed in his Report (Doc't No 37) in his Section 4, Valuation, starting at page 9, one has to take into account the nature of the plant and machinery and the state of the market (Mr Wood's para.4.2.2). One also has to take into account obsolescence. (Mr Wood's para.4.3). The price the willing seller can expect to receive will be affected by the buyer's knowledge that the plant and machinery has not been upgraded to 1986 technology, and that such upgrading will entail cost. 767. Various possibilities have been suggested as the valuation date for the plant and machinery. Mr Wood's valuation purports to be as at September 1986, whereas Mr Blickle opted for the actual date of resumption, 30th July 1986. An argument could also be made for 19th January 1987, being the date SFI physically vacated the Junk Bay site. No one has suggested there is any difference in value amongst any of those dates. We consider the most convenient date to choose is 30th July 1986, and interest can run from then. 768. Government correctly pointed out that, technically, the plant and machinery in the present case were fixtures as at the date of resumption since SFI did not dismantle them, so that, strictly, speaking, they fell to be valued the same as the land under section 10(2)(a) and section 12(d) of the Ordinance. 769. However, quite properly in our view, government indicated it had no objection to the plant and machinery being valued as if it fell under section 10(2)(d) of the Ordinance. 770. Section 10(2)(d) which, as we have previously indicated, is concerned with damage to a business requires the valuation to be made on the basis of value to the owner which means, in effect, the owner is compensated for consequential losses arising from the resumption. 771. In practice, however, in the present case, it makes no difference whether the plant and machinery is valued under s.10(2)(a) or (d) since, in fact, SFI failed to identify, any particular consequential loss to it arising from the loss of its plant and machinery. 772. Unless SFI can put its finger on some special pecuniary loss in relation to any item of plant and machinery being valued, SFI does not begin to gain any advantage from the value to the owner concept. See The Bodlewell [1907] P.286, 290. 773. Unless SFI can point to some such special loss in relation to the item of plant and machinery being valued, the, measure of SFI's loss, by default, will be market value. What it will be worth to Shun Fung as a possible buyer will be no more nor less that what any discerning prospective buyer would expect to pay. 774. We hasten to add that we have not overlooked that SFI is entitled to the benefit of the plant and machinery being already installed. Each item of plant and machinery is to be valued on the basis of continued use in the business as part of a going-concern. Profit or loss is not attributable to individual items of plant or machinery, but to the business as a whole which produces a stream of earnings. We have valued that stream of earnings in the context of goodwill. 775. Spencer v. The Commonwealth of Australia (1907) 5 C.L.R. 418, which we cited previously with respect to the value of the land, dealt with the determination of compensation under a statute which provided for "value to the owner". It was held that this value was market value, or in the words of Griffiths C.J. what "a purchaser would have had to offer for the land to induce such a willing vendor to sell it." Mutatis, mutandis, the same principle applies to plant and machinery. 776. In Re The Hobart Bridge Co Ltd v. The Government of Tasmania (1946) The Valuer 62, Morris C.J. agreed with the Spencer Case, when he simply stated, "The problem then is to find the value to the owner of the physical assets which value is the amount a 'willing but not anxious purchaser' would give and which 'a willing but not anxious vendor' would accept." (underlining added) 777. Lord Moulton's following statement in Pastoral Finance Association v. The Minister (1914) A.C. 1083 dealing with the meaning of value to the owner, explains it in even clearer and more detailed terms which apply mutatis mutandis to plant and machinery: "Probably the most practical form in which the matter can be put is that they were entitled to that which a prudent man in their position would have been willing to give for the land sooner than fail to obtain it. Now it is evident that no man would pay for land in addition to its market value the capitalised value of the savings and additional profits which he would hope to make by the rise of it. He would no doubt reckon out these savings and additional profits as indicating the elements of value of the land to him, and they would guide him in arriving at the price which he would be willing to pay for the land, but certainly if he were a business man that price would not be calculated by adding the capitalised savings and additional profits to the market value." 778. While the value to the owner concept assumes the dispossessed owner as the possible purchaser it does not envisage him as being willing to pay, in terms of value, any more than any other willing and astute prospective purchaser in the market. However the owner, if he occupies the property, is assumed to be prepared to pay more than the market value only where he would be faced with additional costs or losses arising out of his dispossession. For example, there could be the costs associated with having to remove to other premises or there could be additional loss in value to any remaining land when only part of a property is resumed. "What he would be willing to give .... sooner than fall to obtain it." 779. Therefore, the concept allows for any consequential losses due to the resumption to be claimed in addition to market value. If no such additional losses exist, then the market value only is the measure of the value to the owner, as in the Spencer Case. 780. Apart from valuers, economists and accountants also use the concept of value to the owner, and in the same way as lawyers. In Brealey and Myers' Principles of Corporate Finance, 3rd Edition, pages 11, 238, 241, 247, and 248 the authors, in effect, ask the reader why some particular piece of properly such as, say, an aeroplane, or a block of flats, should be worth more to him than to anyone else. If it can earn more for you than others, there will be value to you as owner over and above the market price. If not, the value of that item in your hands is simply the market price. 781. On the evidence in the case before us, there was nothing to show that the value of the plant and machinery at Junk Bay was worth more in SFI's hands than to anyone else. 782. Nor was it shown that SFI suffered any special loss by being deprived of it. 783. Hence, the only basis on which SFI's plant and machinery could be valued was market value. 784. The same concept is a recurring theme in Glover's Valuation of Unquoted Securities at, for example, pages 21-25, 47, 246. 785. Although the words "value to the owner" were bandied about endlessly on SFI's behalf in relation to plant and machinery, none was, in fact, ever identified to us. VALUATION PRINCIPLES 786. A valuation process usually commences with a selection of transactions, and by dissection and anaylsis each is reduced to its elements which, either singly, or in combination, lead to a determination of market value. 787. In the present exercise, for the items of plant still under dispute, the valuers had no transactions on which to base their valuations. Their starting point was the cost of replacement new, and their valuations depended on subjective opinion as to the remaining worth based on the remaining economic life. 788. Where evidence of transactions is available it must be the preferred basis. The vehicles used in the scrap yard, for example, were correctly compared with market transactions by the government's valuer, but Mr Wood, endeavouring to represent what he described as "value to the claimant", stuck to his age/life formula. 789. That approach of his was simply wrong since he ignored the most direct evidence of value in favour of theory. 790. In the end, the parties did manage to agree a value for all the items in the scrap yard, including vehicles, so the Tribunal was spared that exercise. We mention it to demonstrate the problem which arises, when an exercise as subjective as one undertaken by the valuers here purports fairly to represent what could reasonable be expected as the price each item of plant would realise as part of the sale of the whole mill at the relevant date. It is an exercise so based on theory that it poses a danger of losing sight of the object of the exercise which is to find Open Market Existing Use Value (also known as "In Situ Value") and, intead, allowing theory to prevail over reason. 791. With any improvement on land, (and a plant is such an "improvement"), cost will never equate with value unless the "improvement" is new, and exactly suited to the full economic development of the site on which it is located. Without the benefit of knowledge of actual sales, one can only make a reasonable guess as to what that value might be when the improvement is no longer new, and wear and tear, as well as more advanced technology, has detracted from its original worth. 792. The more information, knowledge and experience a valuer has in dealing with particular articles or property, the more likely he is to make a reasonable assessment of its value. This will be even more important in a depreciated replacement cost assessment than when evidence of sales is readily available. 793. In this respect, although Mr Wood had the advantage of seeing the plant in operation, and had the assistance of an experienced steel expert in Mr Medley of McLellan and Partners, Mr Blickle's greater experience of valuing steel mills in various parts of the world was a factor weighing in the latter's favour. Moreover, Mr Blickle impressed us as taking a more independent and rational approach than Mr Wood who struck us as too anxious to please his firm's client, and more or less oblivious to the possible effects of or even the existence of obsolescence. Mr Wood also often experienced difficulty reconciling his estimated remaining lives with his adopted economic lives as he had not worked from commissioning dates. OBSOLESCENCE 794. Mr Neoh expressed concern that, in some instances, Mr Blickle may have deducted for obsolescence twice: first, when he adjusted the replacement cost for what was termed as "betterment", and, again, when he allowed for obsolescence in addition to his deduction for age. 795. As we have said, if technological advancement at the valuation date (i.e. 30th July 1986) affects the value of the machinery, it is a proper consideration to be taken into account. It may form part of the so-called betterment allowance which may have to be made concerning any refinements which are included in the quoted-new price, but do not exist on the item to be valued. It may also have to be considered with respect to the anticipated economic life of an item of plant. For example, the attachment of any additions, such as, say, water-cooled panels to the E.A.F.s, might be accounted for in the C.R.N. of the furnace, while it would also have the effect of lessening obsolescence or extending economic life. Conversely, the lack of such additions shortens what would otherwise have been the economic life. 796. Mr Blickle agreed that the fitting of water-cooled panels to the E.A.F.s would prolong their useful life, but, as they were not, in fact, so equipped, the remaining life should reflect this. Thus, if the quote used for the replacement cost assessment for an E.A.F. included attachments which were more expensive than those on the subject furnace or the subject furnace did not have them, it follows as a matter of reason that the quote should be adjusted for the extra cost relating to these additional items. 797. That is what Mr Blickle correctly referred to as "functional obsolescence". Functional obsolescence may be simply measured by the cost of bringing an item of plant up to date, but this cost must be economically feasible. 798. There is also the prospect of newer, and more up-to-date items being made from cheaper materials than the originals. The use of more expensive materials does not necessarily make the originals any more valuable. This was explained by Mr Blickle as a form of "functional obsolescence", his illustration of it being older mills of brick construction while newer ones have steel-frame structures with relatively cheaper cladding. The brick buildings may be more substantial and cost more to reproduce, but are no more valuable than the more efficient, yet cheaper, modern structures. In fact the brick buildings are likely to be less valuable. Accordingly, the cost of replacement should be based on the latest materials and methods. Any difference in cost is a form of obsolescence, (but we don't agree with Mr Blickle that this is necessarily "functional obsolescence". We would put it in the category of "economic obsolescence"). 799. We understand economic obsolescence to be the loss in value, which cannot be restored by the addition of refinements, such as the attachment of water-cooled panels to an E.A.F., and would usually be due to outmoded design, or technological advancements which cannot be readily adapted to an older unit. It may be difficult to measure precisely, but it would inevitably be an important consideration in the eventual agreed price of any item of plant. 800. As Mr Blickle explained, even when a unit, such as an E.A.F., is completely refurbished, it is still not worth the same as a new unit. It is, we think, unrealistic to treat a unit as new, after refurbishment, in the way Mr Medley has at page 21 of his report, Document 30/01. 801. We have found no double counting in Mr Blickle's assessments. 802. Physical deterioration is another factor which may, or may not, be curable. We accept that when it came to maintenance, SFI was reasonably diligent, and this was not a significant issue with respect to most of the outstanding items. THE CLAIM 803. The claim for plant and machinery totals $89,285,395, while the government assesses it as worth $46,528,051.50. (Ex 54E). 804. We are somewhat surprised at the pretence to precision in these assessments when the adopted valuation method is nothing better than an approximation of value based on an estimated cost of replacement against a surmised useful remaining economic life. 805. We do not suppose a willing buyer and seller concluding a deal in this range of tens of millions would include the last 50 cents or $5. The parties own agreements for the buildings at $26 million or the scrap yard at $3 million, are evidence of that. We would have preferred to have been told that the claimant believed the plant to be worth, say, $89 million or even $90 million, and for the government to have set its figure at say $46.5 million or better still at $45 or $50 million. 806. Valuation has never been presumed to be an exact science. It is the process of making an informed estimate on the best information available. In the case of Singer & Friedlander Ltd v. John D. Wood & Co. (1977) 293 E.G. 212, 295, which dealt with a claim against a valuation firm for negligence, it was observed by Watkins, J. that :
807. With this in mind, not only Is the purported accuracy in respect of the present case surprising, but, equally remarkable, is the difference in the values so determined, Mr Wood's valuation working out at almost double Mr Blickle's. The difference between the valuations was in the region of $42.75 million. This difference is not inconsiderable, and, in this instance, surprising, given the degree of agreement which already existed for the value of the majority of the plant and machinery items, and the replacement costs of the others. 808. Of course, some $22 million of this difference arises through the claim for additional amounts under the heading of "Extras", but the difference for the eight disputed items of plant amounts to over $20 million which is very nearly 40% of the claimant's figure. THE ELECTRIC ARC FURNACES 809. We now turn to the assessments for the two Electric Arc Furnaces which had similar 22 to 25 metric tons nominal capacity and were both manufactured by the Italian firm of Taligaferri. E.A.F.1 with its 8250 KVA transformer was commissioned in 1967 and E.A.F.3 with a 12000 KVA transformer in 1975. The cost of replacement new was agreed at $13,356,000 for E.A.F.1 and at $14,276,000 for E.A.F.3. 810. The quotes relied on for new but similar E.A.F.'s, according to Mr Blickle would include betterment for such features as water-cooled panels and roofs, and other refinements such as computerised controls. Mr Blickle made no adjustment for those, but in his opinion they could represent between 11% to 18% of the overall cost. 811. Mr Wood agreed that the quote which he used and which was obtained by Shun Fung, included betterment in excess of the 4% he originally allowed. We presume from what Mr Neoh said that Mr Wood's agreement with Mr Blickle's C.R.N. is on the basis that the furnace has no water-cooled panels, although Mr Neoh says that with or without water-cooled panels the respondent's quotes are for furnaces with a capacity of 20 tons. That is a lower capacity than Shun Fung's furnaces which as operated by SFI were capable of 22 tons of liquid steel per heat. We were told by Mr Medley that with water-cooled panels Shun Fung's furnaces would have been able to melt capacities of up to 30 tons. Shun Fung's quote was for a furnace with a nominal capacity of 25 tons. 812. Mr Blickle explained to Mr Neoh during cross-examination that both the claimant's and the government's quotes were much the same once adjustment was made for the different transformer sizes. We do not doubt this. The specifications for the claimant's quote for a 25 ton nominal capacity furnace included water-cooled panels and roof. So this should make it about the same as Shun Fung's existing E.A.F. of 22 tons liquid steel capacity without the addition of water-cooled panels which improve capacity. 813. The claimant's quote also included computer controls. We have no evidence on how much such controls might be worth, but we understand that it should not make much difference to the C.R.N. Mr Blickle gets approximately the same C.R.N. whether he uses his own quote or adjusts that of the claimant for the difference in the transformer sizes. 814. The point about this C.R.N. is that Mr Blickle has not deducted for any betterment in the water-cooled panels or computer controls. Because he chose to ignore this betterment factor, it has no bearing on what his depreciation factor should be, but it could allow him to further decrease his C.R.N. for each E.A.E. by at least $0.75 million. We know from SF219 page 221 that this is the cost of adding the water-cooled panels to the E.A.F.s. Mr Blickle has merely said that his C.R.N.s are generous. We agree, and find that his quote and the agreed C.R.N.s do include water-cooled panels. This is confirmed by the comparison with the quote used by the claimant which includes water-cooled panels and roof, and computer controls. 815. For E.A.F. No.1 Mr Wood has allowed for 30 years economic life. Mr Blickle allowed 20 years. Mr Medley in his report put the normal life of an E.A.F. after installation or major refurbishment at 25 years. He said that there is ample evidence throughout the world of furnaces lives in excess of 25 years by "uprating". That is by adding new parts as technological advances are made. 816. On the evidence, we regard it as reasonable to expect that Shun Fung would have added water-cooled panels by Financial Year 1985/86. Mr Medley's firm, McLellans, had described water-cooled panels as "rapidly becoming standard practice" in an article in 1979 (Doc30/03, p.30). Shun Fung already had water-cooled rings in the roofs of the E.A.F.'s. 817. Leaving for the moment the lives estimated by Mr Wood and Mr Blickle, we will now look at Mr Medley's basis of about 25 years plus with continuous uprating. 818. Among the first steps likely to be taken by a buyer of Shun Fung's plant would be increasing the transformer size, and adding these water-cooled panels, as they improve capacity and lessen costs. The E.A.F.'s would therefore require at least about $0.75 million spent on each of them by 1985/6, the year we have found as a fact SFI would probably have added these. 819. E.A.F. 1 was improved to the same specifications as E.A.F. 3 in 1976. This included improved hydraulics, but, as we accepted from Mr Blickle, that should not be construed as a major refurbishment. 820. The E.A.F. 1. C.R.N. of $13,356,000, depreciated by Mr Wood's factor of 60%, which is using Mr Wood's 30 years economic life against an age of 19 years, gives a value of $5,340,000. But, to achieve this life, it would seem that upgrading with water-cooled panels, and, perhaps, other advancements such as better controls, would be necessary. Hydrogen burners we are told, would also be a likely addition. The cost would be at least $0.75 million per E.A.F. for the water-cooled panels, so an astute buyer would be expected to allow for having to spend something of the order of $0.75 million to, say, $1 million on each E.A.F. The result, as we see it, is that, at best,. E.A.F. 1 should be worth something in the region of $4.5 million using Mr Wood's estimated economic life. 821. Mr Blickle took a more robust approach towards E.A.F.1. He merely gave it a value of 20% of its cost new, because it could be made to last a bit longer through refurbishing. Otherwise, in his view, it was virtually obsolete by 1986. He arrived at a figure of $2,671,600. 822. Taking Mr Medley's 25 years "usual life" gives a remaining life for E.A.F.1 of 6 years or 24% (say 25%). This gives a value of about $3.34 million. To get this 25 years, we would expect to have to spend at least $0.75 million on the water-cooled panels. The price a purchaser would pay on this basis would therefore seem to be about $2.6 million. 823. These figures are what we consider to be the alternatives based on the evidence and submissions we have heard. We also place more credence on the Medley alternative than on Mr Wood's 30 years life which we regard as excessive. 824. In the light of the foregoing, we consider that Mr Blickle's figure of about $2.6 million looks reasonable. However, by adding a few years to take account of Mr Medley's observations concerning the upgrading of mills in Taiwan and New Zealand, we are prepared to put a figure of $3 million on E.A.F. No.l. 825. E.A.F.3 was installed in 1975 and was overhauled in 1980 when a modification of the piping system was carried out. As we have previously noted, E.A.F.3 had a 12,000 K.V.A. transformer, while E.A.F.1's was 8,250 K.V.A. The estimated C.R.N. was agreed at $14,276,000. The difference between this figure and that of $13,356,000 for E.A.F. 1 lies mainly in the K.V.A. rating for the transformers. Mr Wood based his assessment on his economic life of 30 years. 826. Against this, we have Mr Medley's statement that new or completely refurbished furnaces of this type usually last 25 years, while Mr Blickle opted for 20 years once again. We believe that Mr wood's 30 years is excessive. Mr Wood's estimate is out of line with Mr Medley's 25 years for a more modern unit. 827. The agreed C.R.N. of $14,276,000, we find, includes water-cooled panels. As such, Mr Medley's estimated life of 25 years would be appropriate only if the furnace were so equipped. Otherwise, it would be less. 828. There was an overhaul and modification of the piping system in 1980. Mr Blickle believed that this may have been a form of maintenance, and as it was carried out after 5 years of installation, it was likely that it was due for a similar overhaul by 1986. He also suggested that another reason might have been that it was substandard to start with. 829. We think it is reasonable to accept that this furnace has already spent 11 years of its estimated 25 years life, and the remaining life was 14 years at the date of valuation. However, we are prepared to extend this by 2 years to take account of any upgrading Shun Fung may have done on this furnace. This gives a remaining life of 16 years from an overall economic life of 27 years. The residual value is therefore about 60% of the C.R.N. which equals $8.5 million. To achieve this water-cooled panels would have to be installed at a cost of $0.75 million which we deduct. This gives us a figure of $7.75 million as its remaining worth. 830. Mr Blickle valued this furnace at $6.424 million, while Mr Wood saw it as being worth close to $9 million. We are prepared to round up to a figure of $8 million which allows about 45% depreciation on the agreed C.R.N. 831. While we do not agree with Mr Neoh that Mr Blickle was "double counting" when he used an economic life of 20 years, our approach of adopting, as a basis, the 25 years given by Mr Medley as the usual life should cover any suggestion that there might have been double counting. 832. We do not agree with the Medley thesis one can start afresh the life of these furnaces from the date of any overhaul, since such an approach completely disregards economic obsolescence. The age/life adjustment must take the date of installation as the point of departure. Functional obsolescence has been accounted for to some degree, by deducting for the cost of installing water-cooled panels. 833. We determine the total value of the two E.A.F.'s at $11 million. THE CONCAST 834. The original single strand concast was installed in 1967, but was never successfully put into operation. It was replaced by the present two strand one which was commissioned in 1978. 835. The Concast was agreed to have a C.R.N. of $15,992,000. Mr Wood determined its value at $10,234,900 while Mr Blickle valued it at $8,479,800. 836. The original estimated economic lives for the concast were the subject of some changes of opinion by both Mr Blickle and Mr Medley. Mr Blickle went from 20 years to 15 years, while Mr Medley after seeing the concast at Liuzhou was prepared to do the opposite, changing from 15 years to 20 years. At first, Mr Medley only said that he was tempted to put the life of this concast at 20 years rather than his original 15 years but he bowed to pressure by his client and yielded to the temptation. Mr Blickle thought a concast would be "in tatters" after 20 years. 837. Concasters can still be regarded as something of an innovation, having first appeared in the 1960's. There have been swift improvements in concast technology since then, rendering the earlier versions obsolete within a short space of time. We accept what Mr Blickle said on how they have virtually replaced ingot casting in "First World" countries. As Mr Blickle said, ingot casting is now "a technologically and economically obsolete system of casting and rolling steel". 838. There was little in the way of evidence or expert knowledge to support Mr Wood's remaining economic life of 16 years for SFI's concast. Mr Blickle's extensive experience and knowledge on this subject satisfied us that his fifteen years was the correct usual economic life of a concast. We do feel, however, that he did not sufficiently acknowledge the substantial alterations which were made to this concast in 1982, including the replacement of the mould oscillating system. However, we certainly do not agree with Mr Medley that the life of the concast should begin again from this 1982 date. 839. If we were to take a life of 20 years, but from the 1978 installation date, the calculated remaining life at the resumption date is 12 years. Mr Blickle's 15 years life gives a remaining life of only 7 years which we feel on the whole of the evidence might be too short a period. It is more likely to be somewhere between those two figures, twelve and seven years, so we propose adopting 10 years as the remaining life. In doing that, we accept that the 15 years first spoken to by Mr Medley and confirmed by Mr Blickle is the usual economic life. The extra period is allowed on the evidence of the upgrading and Mr Medley's observations at Liuzhou. In taking this approach, we are setting the overall life for this particular machine at a total of 18 years and the remaining life at 10 years. This gives a depreciation rate of 45% which results in a figure of about $8.8 million, which we are prepared to round up to $9.0 million. 840. We consider, on the evidence adduced, that the concast would realise about $9.0 million within the value of the plant as a whole. THE ROLLING MILLS 841. A wealth of detail was traversed in the submissions on the value of the rolling mills. 842. Mr Blickle worked from a quote for a substantially superior type of mill, while Mr Wood's "Mercon" brand rolling mill quote was closer to that used by SFI. We are content to work from Mr Wood's quote which is slightly lower than Mr Blickle's in any case. The figure for this Mercon is $31,885,400. 843. For present purposes, it is sufficient to note that depreciation rates had to take account of the circumstance SFI's mills being of the "cross country" type which is not so common these days. "Cross Country" mills have loopers which turn the bar through 180 degrees, and are more likely to produced "cobbles" i.e. bars which have to be discarded for re-cycling as scrap, since they have not survived their passage through the full length of the rolling mill and will not conform to the British Standard. We note that the proposed new mill at Shunde would be the more usual, and in all respects preferable, single or "straight through" mill rather than the complicated combination of three mills used by SFI. The estimated cost of installing the "straight through" single mill at Shunde was $45 million based on December 1987 prices. 844. Mr Blickle drew attention to what he considered were largely antiquated stands, some of which had survived from the time SFI was located at Ma Tau Kok in the 1950's. It is also to be noted as relevant that Mr Roy Leung, himself has said he intended to replace the finishing stands in each of the mills. Other matters which diminished the value of this mill in Mr Blickle's opinion included the home-made design (rather than from a reputable mill manufacturer) and construction of some of the stands and run-out tables, the old fashioned electrics and controls used, and the set up and design of the cooling beds as well as the excess in manning required for this type of mill. 845. The claimant considered that rolling mills can be used for many years with proper maintenance, and a degree of updating when necessary. This may well be so, but it does not prevent them from becoming obsolete, a condition which we regard as having affected SFI rolling mills to a large extent. 846. Having said that, there is not really all that great a difference in the depreciation rates adopted, the claimant applying 60% and the government 80%. 847. Rather the same as with E.A.F.1, we agree with Mr Blickle on SFI's rolling mills that, despite obsolescence, they are, nonetheless, still in use and can still be kept going, so that they still deserve a remaining value of at least 20% of the C.R.N. Presumably the same general criteria apply to SFI's rolling mills as to E.A.F.1, despite MR Blickle's choice of a somewhat complicated age/life calculation for each of the mills. 848. The mills are of a type not usually used these days, and most of the stands are somewhat ancient. From the vast deluge of material and arguments presented to us on the rolling mills, we believe it impossible to do better than robustly accept that a purchaser would almost certainly replace these rolling mills with something more modern to save wastage from overweight bars and to save labour costs through reducing manning. This is one of the more difficult of the assessments from the point of view of trying to be precise and we have inevitably been driven to adopting a broad-brush approach. 849. We intend to adopt a value of $8 million. That is a depreciation rate of about 75% on a replacement cost in the region of $32 million. We do not believe that, with the best will in the world, anyone can, in all honesty, be any more accurate than that, given the nature and degree of obsolescence inherent in these mills. THE REHEATING FURNACES 850. Shun Fung had three reheat furnaces. Two of them "B" and "C", had been rebuilt in 1981, thereby increasing their capacity to between 14 to 16 tons per hour. Reheat furnace "All was installed in 1967 and, on any view, was clearly redundant to Shun Fung's needs. Mr Wood regarded it as a "standby", but Mr Blickle ascribed only value for salvaging the parts at $157,000. Mr Wood depreciated his C.R.N. of $4,777,000 by 90% to give a value of $477,700 for furnace A. 851. Furnaces B & C were each valued by Mr Wood on a straight line depreciation basis on a C.R.N. of $4,777,000 for a remaining life of 15 years of a total economic life of 20 years. This gave Mr Wood a value of $3,582,800 for each of these two furnaces. 852. Mr Blickle agreed with the economic life of 20 years, although we note that Mr Medley considered 15 years to be the economic life of a refurbished reheat furnace. (Doc 30/01 p.21). Mr Blickle valued furnaces B and C at $2,100,000 each. He depreciated his higher C.R.N. by 60% on the grounds of obsolescence. 853. The reheating furnaces are, in effect, an extension of the rolling mills which we have held to be outdated and largely obsolete. These reheat furnaces are suitable, and, as Mr Medley put it, "well matched" to those rolling mills. They are small by usual reheating furnace standards, but suit the relatively short Shun Fung billet lengths which are cut to suit the size of the bars produced. B and C Furnaces can more than cope with a capacity of 110,000 metric tons of good product for the works. 854. Mr Wood saw no measure of obsolescence in these two furnaces stating that they were not obsolete, "in terms of Shun Fung Ironworks. They were perfectly adequate for the product they were marketing". 855. There is no doubt regarding these RHFs, as to their capability of producing the product to the specifications required by SFI or its customers. There is also no doubt as to their being able to remain in use for some 15 years after refurbishment, provided the rolling mills remained as "cross country" or "repeater" type mills. 856. What we do note from the evidence is the unlikelihood of this type of layout being employed in any new mill where a single reheat furnace with an "in line" or "straight through" rolling mill is almost certainly to be preferred. The fact is that a set-up of two or more of these "cross country" type mills, with these sort of furnaces, is, by modern standards, not only obsolete, but, also, with respect to the furnaces, more expensive to construct. 857. This is a prime example of what Mr Blickle's terminology described as "functional obsolescence", but we regard as "economic obsolescence". Mr Blickle's example was that of the value of an older and more expensive to reproduce brick-built factory being no greater and maybe even less than that based on the cost of a modern and cheaper steel-framed one. 858. In terms of value, these furnaces should be considered in the light of what it would cost to have a single reheat furnace of double the capacity, rather than the two furnaces, (not to mention the additional "standby" furnace), which are more expensive and require more fuel and men to run them. In saying, this we have taken account of Mr Neoh's submission concerning the supposed advantages of having two furnaces over one; but such arguments had little if any support from the experts, and struck as in the nature of trying to make a virtue out of necessity. 859. We were guided also by Mr Medley's statements in paragraph 8.2.2 on page 48 of section 8 of his Document 30/01. There, he explained his proposals for the new rolling mill at Shunde, which included a single 30 ton capacity reheating furnace, and, that in our opinion, would have been the bench mark for any astute prospective purchaser when considering the value of the SFI rolling mills and reheating furnaces at Junk Bay. 860. We consider, on the evidence, that the value of the Shun Fung rolling mills and the furnaces was affected by a strong possibility of their needing replacement within a few years, in spite of the fact that reheating furnaces B and C had been refurbished as recently as 1981. SFI's arrangement of multiple RHFs and rolling mills was obsolete, and if SFI wanted to become competitive it needed a single straight through mill fed by a single RHF. 861. These three furnaces, or at least the two that were in use and working, cannot be worth more than the costs associated with replacing them with a single 30 ton furnace. Even if Shun Fung or any prospective buyer might have chosen to continue to use the two furnaces for a further 15 years, their value and hence their cost of replacement new (C.R.N.) must be based on the cheaper alternative of one furnace. 862. Mr Medley's estimated cost for such a furnace was $4,550,000 (Doc 30/01 section 8 p.64). This figure does not include the 19% he estimated for shipping and erection for the plant at Shunde. No other allowances such as interest or contingencies are appropriate if they are to be considered elsewhere in the overall assessment. It follows that after allowing for shipping and installation, which we acknowledge in the case of reheat furnaces may not be line with the 19% usually allowed for other items of plant, the total C.R.N. for the furnaces should be no more than say $5.5 million. On the view we take, we regard Mr Wood's estimate of $14.331 million or Mr Blickle's $16.104 million as excessive. 863. If we are correct in assuming that the rolling mills are likely to be replaced eventually by a single straight-through mill, we should depreciate these furnaces at the same rate as for the rolling mills. This means the remaining worth of the furnaces should be somewhere in the region of 25% of $5.5 million which is $1,375,000. On this basis, the total value of the rolling mills and furnaces should be of the order of somewhere between $9 million and $10 million. 864. Having said this, we are conscious of the likelihood of there being no alternative to "cross country" mills due to the restrictions of the Shun Fung site and the layout of the mill. On this basis the furnaces could be retained and only the rolling mill replaced with a new "cross country" type mill. We do not believe that an astute prospective purchaser would relish such a prospect, but would probably settle for it, so in fairness to the claimants the valuations should be done on that basis. 865. The furnaces, on the basis of their being retained for their agreed normal life of 20 years, (as opposed to Mr Medley's 15 years (Doc 30/01 p.21)) have a remaining life of 15 years. This is a residual value of 75%, based on the Cost of Replacement for a single 30 ton furnace. 75% of $5.5 million is $4.125 million. 866. Mr Blickle by a somewhat more arbitary approach valued the three furnaces at $4.357 million. Given that he ascribed $157,000 to the redundant "standby" Furnace A, which, on our approach attracts no value, we see that our figure, arrived at by a different route, agrees closely with Mr Blickle's. Our figure of $4.125 million compares with Mr Blickle's $4.2 million if $157,000 is deducted for Furnace A. However, Mr Blickle was willing to give some value to "standby" Furnace A, and we have no wish to go below his overall figure of $4.357 million. 867. We believe the RHFs, and the rolling mills should be valued as a single entity. As a matter of simple arithmetic, the combined assessment for the rolling mills and RHFS works out at $12.357m. 868. We recoil from that sort of spurious accuracy, and propose allowing a figure of $5 million for the reheating furnaces on the basis that the total value of the rolling mills together with the furnaces should be in the region of $13 million. This may, at first sight, appear to err in favour of the claimants, as arguably, a total figure of $4.5m for the furnaces would more than adequately include any value Furnace A may have had. 869. If, however, the furnaces are treated as part of the rolling mills in the way we think they should be, the $13million figure represents only a variation of about 1.5% in the depreciation rate for the mills to make up the additional $500,000. 870. $8 million for the rolling mills was a very general figure. $13 million for the mills including the furnaces must also be regarded in the same light. THE CASTING (LADLE) CRANES 871. These are two "Demag", electric, overhead, 40 tons, double girder, cab-operated cranes. There was little difference between the valuers on their estimated C.R.N.'s. Mr Wood, using a quote from the firm of Jebsens for a similar "Demag" crane adopted a C.R.N. of $2,852,000 for each crane, while Mr Blickle working from a handbook for the American "Whiting" cranes used a C.R.N. of $2,839,000 for each crane. To our minds, these two C.R.N.'s are virtually the same, and each estimate supports a figure of about $2,850,000. 872. The dispute is really on the estimated economic life for such cranes. Mr Blickle's economic life of 25 years was supported by Mr Medley in his Document 30/01 page 21. Mr Wood used 30 years. We are satisfied on the correctness of Mr Blickle's 25 years, bearing in mind his greater experience together with the corroboration from Mr Medley's estimate. 873. Mr Wood valued the older (1966 vintage) crane at $1,045,700, and the newer (1974) at $2,091,500. The total for the two was, therefore, $3,137,200. Mr Blickle's corresponding values were $567,800 and $1,420,000, for a total of $1,987,800. 874. There is nothing in the evidence to incline us to depart from a simple age/life depreciation. The 1966 crane was 20 years old at the date of resumption, with an estimated remaining life of 5 years. The depreciation rate is therefore 80%. The other was commissioned in 1974. It is therefore 12 years old with a remaining life of 13 years. This crane's hoist was replaced in 1978, but although, this could be expected to extend its useful life, it was done so long ago as to be of no significance by the time of resumption. Mr Blickle applied a depreciation rate of 50% with which we agree. 875. The older crane depreciated by 80% on a C.R.N. of $2,850,000 results in a remaining value of $570,000. The other from the same C.R.N. depreciated by 50% gives a value of $1,425,000. 876. The total calculated value of these two cranes is, therefore, $1.995 million. We consider that those two cranes should add a value of $2 million to the overall value of the plant, and we adopt this figure in respect of these cranes. THE CHARGING CRANES 877. These two cranes were home-made, having a 20 ton capacity. They were electric, overhead, double girder and cab-operated. Both were commissioned in 1966, but just the one had a winch crab replaced in 1981 as well as the installation of air-conditioning to its control cabin. Mr Blickle increased his original valuation for those improvements. 878. Mr Wood valued each crane at $784,000, while Mr Blickle valued one at $333,200, and the other, which had been improved in 1981, at $500,000. Mr Wood's total for the two is therefore $1,568,000 compared to Mr Blickle's $833,200. 879. Here we have disagreement on the cost of replacement (C.R.N.). Mr Wood used a C.R.N. of $1,960,000 which he obtained through Jebsens for a "Demag" crane. Mr Blickle used a C.R.N. of $1,666,000 which was calculated from the handbook for "Whiting" cranes based on the weight of the material used in the construction of the crane. 880. Mr Wood had originally used a C.R.N. of $2,364,000, which was a Jebsen quote for a 25 ton scrap yard crane. He reduced that when Mr Kat on behalf of government drew his attention to a charging crane being listed as some 80,000 Deutch Marks cheaper than a scrap yard crane in a letter from Jebsen's (Doc 27/02 pages 105 to 107). Mr Wood's estimation method was never described to us. Mr Blickle, on the other hand, described to Mr Neoh in cross-examination (CPB-T 725) that he worked from a weight of 54,000 pounds at US$4.00 a pound. US$3.00 was for material and US$1 was for freight. 881. Using a conversion rate of HK$7.8 to US$1, a calculation of 54,000 x 4 x 7.8 results in a figure of $1,684,800 (say $1,685,000). The C.R.N. Mr Blickle presented to the Tribunal was, actually $1,666,000. 882. Mr Blickle explained his method for calculating the C.R.N. for these charging cranes was exactly the same as for the 40 ton Casting Cranes, and, on that occasion, he and Mr Wood were practically on the same figure. As we were left mystified by Mr Wood's method of estimation, but found Mr Blickle's approach sensible, we, in principle, accept Mr Blickle's C.R.N. but round it up to $1,700,000. 883. In accepting this C.R.N., we also have taken account of the fact that these charging cranes were manufactured by SFI in-house. We also accept Mr Blickle's opinion that although the Whiting cranes are based on a short ton weight, they, nonetheless, are every bit as robust as the "Demag" with the same number of tons, albeit of the long variety. For these reasons, we believe that a figure of $1,700,000 is likely to be somewhere in the region of what these cranes would cost to replace new. 884. Mr Wood adopted an economic life of 30 years compared to Mr Blickle's 25 years. Again, Mr Medley's estimate agreed with Mr Blickle (Doc 30/01 p.21). On a straight age/life depreciation Mr Blickle at first depreciated each of these cranes by 80%. Mr Wood used 60%. Mr Blickle revised his depreciation factor to 70% for the crane which had been improved in 1981. 885. We accept Mr Blickle's depreciation factors. 886. $1,700,000 depreciated 70% gives a figure of $510,000, while 80% results in a figure of $340,000. The total for the two cranes is, therefore, $850,000, which we adopt. POWER AND OTHER DISTRIBUTION 887. This heading covers the electrical power distribution cables as well as the compressed air, gas, oil and water distribution pipework within the mill. 888. The cost of replacement new was agreed at $4,832,500 based on the inventory compiled by Mr Wood. 889. Once again the difference in the depreciation rate to be adopted resulted in disagreement as to value. Mr Wood's valuation was $3,220,000 compared to Mr Blickle's figure of $1,691,400. 890. We admit to some surprise that this item is included as we would have thought that such ducting and cables formed part of the installation costs associated with the various items of plant. If that were the case, then there would be double counting. However both parties have included it in their respective valuations, so we must assume that it has not been accounted for elsewhere. 891. Mr Wood used an economic life of 30 years and a depreciation rate of 33%, compared to Mr Blickle's economic life of 25 years and a depreciation rate of 65%, which he arrived at by weighting 1/3 of his C.R.N. from an installation date of 1974 and the remaining 2/3 from 1966. Mr Blickle based this on the commissioning dates for the major items of plant. Mr Wood's installation date was calculated as a simple average. Mr Blickle, correctly in our view, criticised Mr Wood's failure to take account of the commissioning dates of major users of power such as the E.A.F.s and the rolling mills. 892. Mr Neoh produced SF207 and SF207A to show, firstly, the average date of commissioning, and, secondly, the amount of money associated with the installation of items of plant between 1967 and 1982. Mr Neoh contended that SF207A showed that any weighting should be greater for the period after 1975 than before it. Mr Blickle pointed out these were historical costs, and allowance for the value of money was not included in Mr Neoh's exercise. Unarguably, there is a vast difference between what $1 was worth in 1967 compared to 1982. 893. Mr Blickle said these figures would require indexing. We agree, but, even then, the exercise would still not take into account the major users of power or the existence of any redundancy in the installation. Mr Blickle considered that some redundancy was inevitable when plant was added to over a number of years, and we agree with him on that. 894. Having considered such evidence as there is on this matter, we are driven to the conclusion that accuracy is an impossibility. Mr Blickle's arguments have considerable merit, but we do not believe that his weighting can be regarded as anything better than guess. We find ourselves perplexed on how anyone can hope to have a rational view on an economic life for an assortment of cables and ducts of unknown ages and of unknown condition. For these reasons, we believe that a robust approach is the only one available. We adopt a figure of $2 million which is, roughly, a discount rate of 60%. TRANSFORMERS AND SWITCHGEAR 895. This item covers two "Tamini" brand 2,500 KVA oil-cooled transformers and one "Vertano" 3,500 KVA oil-cooled transformer, each with associated switch gear for electric power distribution throughout the mill. 896. The cost of replacement or C.R.N. was agreed at $2,656,000. Mr Wood's valuation was $1,770,000 while Mr Blickle's was $930,000. 897. The two 2,500 KVA transformers were installed in 1966 (Doc 21/04) and the 3,500 KVA transformer in about 1974 or 1975. Mr Blickle based the date for the largest transformer on the electricity demand at that time, as well as an entry in Doc 21/04 for a "Switch Board" in November 1975. Thus, at the resumption date, the two smaller transformers were 20 years old, and the larger about 11 years old. 898. Mr Wood used an economic life of 30 years. By calculation, this would mean that the older transformers should each have a reamining economic life of 10 years, and the newer one, 19 years. Yet Mr Wood gave all three transformers a remaining life of 20 years, based on the fact that he believed they were in very good condition. He took no tests but based his assessment of remaining life on what he observed during his inspection of the Shun Fung Mill. 899. Based on his own assessment of age and economic life, Mr Wood should have concluded from his installation date of 1966 for the two small transformers, and his remaining life of 20 years that he had really given this equipment a useful life of 40 years, and half of it had been used up. His depreciation factor should have been 50%. On the other hand, if he had really used an economic life of 30 years, based on his installation date of 1966 his depreciation rate should have been 66 2/3%. His depreciation rate of 33% is therefore insufficient on the basis of the figures he adopted. 900. Mr Blickle said that switch gear tends to be modernised from time to time as it becomes obsolete. The type of switch gear used in the 1960's and 70's is not the same as that built in 1986. There was no evidence of any replacement of the switch gear. Mr Blickle has once again depreciated on a weighted basis with 2/3 value depreciated from 1966 and 1/3 from 1975 based on an economic life of 25 years. His depreciation rate was 65%, although by calculation this could have been increased by 2% or 3%. 901. On the evidence, there seems to be little reason to deviate from a simple age/life depreciation, but we are prepared to adopt Mr wood's 30 years economic life in this case as it corresponds closely with Mr Medley's estimate for the life of the plant as a whole of 13 to 18 years or "at least to 1999 with some refurbishment". (Doc 30/01 p.22). The refurbishment in this case would simply be regular maintenance and replacement of components, as and when it became necessary. 902. We intend to adopt Mr Blickle's depreciation weighting of 2/3 from 1966 and 1/3 from 1975 which we consider reasonable, and, if anything, favours the claimant. This results in an overall depreciation rate of between 56% and 57%, and a value of about $1,150,000, at which figure we determine the value of these transformers and switchgear. SPARES 903. We now turn to the claim under "Extras". The first item was Spares. On the claim for Spares, there was scant evidence, and even that made little sense. 904. Mr Wood based his claim for spares on 2 1/2% of his total C.R.N., on the assumption that the normal level of spares would have been in the region of 5% of the total capital cost, and that this normal level had been reduced by about half at the resumption date. 905. Mr Wood was guided by Mr Medley on this aspect of his assessment. Mr Medley assumed a figure of 5% of the capital cost of the plant, before shipping and installation, as the usual cost for the spares required for a new plant. Hence, Mr Wood's estimate should have been reduced for the additional 19% which Mr Medley has allowed for shipping and installation on the cost of the plant. 906. On our figures, the total C.R.N. is in the region of $135 million, so the figure on which to base a percentage estimate for spares should have been about $113 million, or roughly somewhere between $100 million and $115 million. On his basis of 2 1/2% of the capital cost of the equipment, Mr Wood's claim should, therefore, be adjusted to about $2,825,000, rather than the $3,610,072 set out in the claim. 907. Much confusion characterised the evidence of Mr Ho Chi Ming on this subject. Mr Ho had been Shun Fung's mechanical engineer in charge of maintenance uptil 1984 when he left after 30 years with the company to take up the post of Chief Instructor, in the machine shop of the Vocational Training Council at Kowloon Bay. Mr Ho had held a number of posts during his time with SFI, including a lengthy spell in SFI's Planning Department where he was much involved with the design and layout of the Junk Bay works. 908. From Mr Ho we heard that manufacturers recommended maintaining a level of spares which would work out at about 10% to 15% of the capital cost of the equipment. This conflicts with Mr Medley's estimate of 5% of the initial cost before shipping and installation. Mr Blickle also confirmed the 5% figure as being the usual rule of thumb for the value of spares for a new plant. 909. Mr Ho went on to say that SFI generally accepted the manufacturers' recommendation, but had reduced the level of spares to about 4% to 5% of the capital cost at the time he left in 1984. He later altered this to about half of the 10% to 15% originally kept, saying this was what he had meant to convey. 910. He produced a list of spares (SF124), compiled from his memory, which purported to be what SFI usually kept at the time he was there. No estimate of the cost or value of the items, in Exh.SF124, was produced. We were told that most of those spares were acquired before 1982. Mr Ho also mentioned it was Shun Fung's policy to manufacture, "in house", many of the spares it required, particularly those needed for the rolling mills. These spares were usually left unfinished. He said the reason for them requiring further machining was a shortage of manpower. 911. A further fact mentioned by Mr Ho was that some of the major spare parts for the E.A.F.'s, overhead cranes and concaster had actually been used, and then repaired in order to be used again. In fact we gathered the parts for a third strand on the concaster had previously been used at some stage, but had become available for reuse. 912. By the end of Mr Ho's evidence, the picture was even less clear regarding what spares SFI had at the time of resumption. On cost or value, his estimate of 14% to 15% of the capital cost was totally at variance with the usual level of spares for a new plant spoken to by Mr Wood, Mr Medley and Mr Blickle. We feel impelled to assume Mr Ho got it wrong. 913. The part of Mr Ho's evidence which revealed that many of the spares kept by Shun Fung were unfinished or used, pointed to the value of spares kept by Shun Fung being less than that recommended by manufacturers for a new plant. On the not unreasonable assumption that the usual level of spares recommended by manufacturers had, in SFI's case, been reduced to half by 1984, then, the adjusted figure of $2.825 million for Mr Wood's estimate for 1986 would have been excessive. 914. A point raised by Mr Ho in favour of SFI having a higher level of spares than might at first appear that spares were located throughout the plant for immediate use in the event of a breakdown. Such spares, according to Mr Ho, were regarded as having been allocated to the relevant department, and were no longer part of any reserve stock. This might go some way towards explaining why SFI's records, referred to by government in SF125 and page 90 of R22, show so little by way of spares still available. As with so much of Mr Ho's evidence, his point about spares being located throughout the plant was so general and vague that it offered little help in quantifying what level of spares SFI in fact had at the date of resumption. 915. Government drew attention to there being no inventory or comprehensive record of capital spares held by Shun Fung in 1986. The records submitted by Shun Fung (SF125) for 1984, and the similar list for 1986 accompanying the Jones Lang Wotton letter of 10th September 1986 to the Director of Buildings and Lands (page 90 of R22), show capital spares as well as consumables worth a little over $3 million in the first instance, and a little over $700,000 in the second. It is agreed that only capital spares should be included in the claim. 916. Mr Wood's firm was never instructed to include any valuation of spares in the assessment of the value of the plant and in fact never did so. Mr Wood had no inventory of spares and was only able to support his claim by reference to a number of crates in the go-down at Junk Bay. They were presumed by Mr Wood to contain spares. 917. Mr Blickle said that it was impossible to value spares without an inventory, and government drew attention to Mr Wood's agreeing with Mr Kat that an inventory would be, "a very good start", and Mr Wood going on to say, "But really you need an inventory plus invoices". 918. Government has simply submitted that, without any inventory or evidence of what the crates at Junk Bay and later observed at Liuzhou contained, no allowance can be made for spares in the assessment. Our attention was also drawn to the claimant having taken no account of the fact that some of the spares had already been used, were homemade, or had depreciated in value since purchase. 919. As we understand the position it would have been necessary for SFI to carry at least some spares to cover the possibility of breakdowns; but the level was likely to have been as low as possible, with the level continuously waning from the time SFI learnt of the threat of resumption in November 1981. 920. The total value of the disputed items of plant together with the amounts agreed for the other items, works out at about $49.3 million. This is by no means an exact figure. If we round it up to $50 million, we believe, from the sketchy evidence adduced, such a figure should cover the value of any spares SFI may have had at the date of resumption. DESIGN, ENGINEERING AND PROJECT MANAGEMENT 921. Mr Wood included in his claim the sum of $6,495,463 to cover the value of Design, Engineering and Project Management. This was also an "Extra" which Mr Wood admitted was made on the advice of Mr Medley. It is certainly a cost associated with new plants such as that proposed at Shunde where Mr Medley estimated the cost of the fee to be paid for this service as 10% on top of the cost of the plant. 922. Mr Wood applied that 10% factor to his valuation of $64,954,630 to arrive at his figure. 923. In his estimate of the cost of the Shunde plant Mr Medley applied his 10% to his total cost which included spares. 924. Mr Blickle agreed that it was a proper cost associated with the cost of providing a "green field" plant. An estimate of 10% of the total cost of the plant was what he described as a starting point for such a fee. He said it depended on the extent of the work involved and could range between 7% and 12%. 'However, he made the point once again that cost is not synonomus with value, and that, as he put it, "A willing buyer is not too willing to pay for costs that could have been incurred that are not in his best interests in further operating the plant". 925. Mr Blickle was willing to allow 5% of his "in situ" value because in his opinion about half of the plant was "thoroughly well designed". He believed that rather than go piece by piece throughout the plant, deciding which items should attract 10% or 0%, it was simplier to allow 5% overall. 926. As we understood him, Mr Blickle was contrasting the melt shop, which he considered was well designed, with the scrap yard and rolling mills which he considered were not. The scrap yard and rolling mills, so he contended, were outmoded to such a degree that any value associated with design and layout, and, for that matter, any associated project management, had completely dissipated. An astute purchaser would be expected seriously to consider their replacement with something more up to date according to Mr Blickle. As Mr Blickle explained, "Why should I pay at a premium for things that are going to cost me extra to do?" 927. We think that what Mr Blickle said makes good sense. 928. At first, Mr Blickle was not prepared to make any allowance under this head, but, on further consideration, conceded that 5% should be allowed. 929. Although Mr Blickle thought it cumbersome to go through the plant piecemeal, we would have preferred this allowance to have been applied to the C.R.N.'s, in the same way as allowance would be made for installation and shipping. In making such an allowance by way of an end adjustment, as has been done here, the arithmetic may be correct, but there is always the risk of double counting in the depreciation factor. Moreover, by making this an end adjustment, our task becomes more difficult concerning the agreed figures for the other items of plant and the scrap yard. 930. We say this, because by making the claim in this way, the claimant now seeks to alter an agreement for the total value of the scrap yard. The claimant seeks an additional 10% while government proceeded on the assumption that the total worth was $3 million. This also holds for the agreement of $7,290,730 for the other items of plant. These agreements should not be altered as the design factor is part of their value. 931. As with interest, this is a cost which need not be actually spent for it to influence value. Good design adds value in the same way as poor design detracts from it. 932. It makes sense for any premium for good design to be allowed for by way of an additional 10%, which is the usual level of such fees. With time, this added value will depreciate along with the value of the plant. If we accept Mr Blickle's statement that the melt shop was well designed and would attract additional value for the cost of providing the design and layout, then 10% might well be a proper additional factor to be applied. For the depreciated value of the rolling mills or the scrap yard it would not. If an item has been depreciated to include an element for poor design, it would be ludicrous to suggest that 10% for good design should be added to that figure. 933. This is a cost incurred together with any other installation costs. It is an initial cost and therefore is applicable to all items of plant as part of the cost of replacement new (C.R.N.). It may have been applicable to the rolling mill and scrap yard when they were installed, but the value assoicated with their design has simply depreciated at a much greater rate than the other factors which have contributed to their depreciated value. If it could be shown that this was a case of poor design in the first place, this cost might well have depreciated to nothing or detracted from the value from the start. 934. In our view it would have been better to include the 10% factor in the C.R.N.'s and then apply the appropriate depreciation factor for each item or group of items to take care of what must be part of an obsolescence factor. 935. In this way, any risk of double counting by making a separate end adjustment is eliminated. If we had started with C.R.N.'s 10% higher, the final figure for the scrap yard and rolling mills should have been no greater because, by 1986, any value associated with design engineering and project management was most probably negative. On the other hand, as far as the melt shop items were concerned, they could have been expected to have had more value ascribed to them. 936. If we delete the value of $13 million for the rolling mills and $3 million for the scrap yard from our total of $50 million we are left with $34 million. However, the scrap yard items would have had to come out in any case as their total value has already been agreed at $3 million. For the same reason we should delete the already agreed figure of $7,290,730 for the other items of plant. The remainder is $26,709,270, 10% of which is $2.67 million. 937. Alternatively, if we follow Mr Blickle's recommendation and apply 5% to the total value of $50 million to allow for half the plant being of good design we obtain a figure of $2.5 million. 938. On the evidence it is impossible to obtain a more precise figure than either of those, but we believe that the remaining value for this factor probably lies somewhere between $2.5 million and $3 million. We propose to allow $3 million to cover this aspect of value. CONTINGENCIES 939. This is another of Mr Wood's "Extras" which, as he stated, was the result of "drawing on the expertise of McLellan and Partners Ltd", and, "a reasonable allowance for unforeseeable contingencies was in the order of 8% of the plant and machinery replacement new value". (Mr Wood's Report, Doc 37/01 page 11.) Mr Wood applied 8% to his in situ value of $64,954,630 and arrived at a figure of $5,196,370. 940. Mr Medley in his estimate for the Shunde plant applied 8% to his total cost after allowing for installation, spares and fees. On Mr Medley's basis, 8% of our total figure of $53 million is $4.24 million, while on Mr Wood's method, 8% of our $49.3 million for the plant value before spares and design, engineering and project management, is $3.94 million. We feel the claim under this heading should, therefore, be adjusted to about $4 million. 941. In justification of this allowance, Mr Wood cited the costs which may occur due to delays in meeting delivery targets by suppliers, strikes, or delays in installation caused by adverse weather conditions. Also mentioned was the possibility of installation contractors going out of business. 942. As Mr Blickle explained, no one is likely to pay for another's misfortune. If an original project were to cost extra due to any of these reasons, it is most unlikely that someone else would be prepared to pay anything extra. It cannot be worth any more to the purchaser by virtue of any of these events having occurred in the past. 943. Mr Neoh drew attention to the Hobart Bridge Case (1946) The Valuer 62, where, he said, an allowance for contingencies was included. This case dealt with the acquisition by the Tasmanian Government of a bridge only one year after its completion, its construction having taken place under the abnormal circumstances which existed during World War II. The equivalent replacement cost was estimated on a unit basis and actual historical costs were available as a check, but were found to be unreliable by reason of the bridge having been constructed during the war when costs were esculating. 944. No definite reasons were put forward for the contingency allowance, but in the circumstances of that case and on the basis of the estimate, the allowance may have been justified. Morris C.J. made this comment concerning historical cost compared to the original contract price: "In a normal case historical cost of a structure completed within one year of the date of acquisition would be a matter of very great importance indeed. Its importance in this case, however, is somewhat diminished by the circumstance that the period of construction was an abnormal one owing to war conditions affecting man-power, labour and materials - a fact which caused a considerable increase in cost during the period of construction." 945. He went on to say, "I attach more importance to the contract price of 1938 (with extras #331,000) since I think one might with better hope of a satisfactory result estimate the necessary increase to that cost if reconstruction were to be effected at the time of acquisition." (underlining added) 946. This, ultimately, was the method adopted by Morris C.J. Herein lies the clue that, the use of the 1938 contract price when contingenices would have undoubtly have been used up, may have given rise to the contingency factor being applicable in the final assessment. But we can only surmise. Morris C.J. did find that the unusual conditions of the time causing additional expense had no bearing on the valuation. The cost of unforseen misfortune could not be expected to be passed on to a prospective purchaser. It should also be noted that the assessment in this case provided for "value to the owner". 947. Contingencies are a part of a proposed project cost estimate to cover any unforseen expenses. Mr Wood acknowledged this. Where they are part of a contract quote, they are not necessarily expected to be spent. This money is drawn as and when unforseen additions have to be covered. To our mind, a depreciated replacement cost valuation is not based on a cost estimate for something yet to be built like the Shunde plant, but rather it is based on an estimate of the cost of replacing something which is already in place. Unforseen items of expense in a valuation exercise should not exist. Everything to be assessed is in place at the date of valuation and should form part of the replacement cost estimate. 948. In any case, when the replacement cost of an improvement is depreciated, some initial cost allowances, however reasonable in the first instance, tend to become insignificant against a depreciation factor of some magnitude. That is why we believe that any element of cost should not be included as an end adjustment when it forms part of an initial cost estimate. The arithmatic may be correct, but the exercise seems to us to be merely an excuse to add something on to a figure after it has already been determined as a reasoned estimate of remaining worth. 949. The depreciation factor is nothing more than the best estimate one can make, and any cost additions such as the "Extras" claimed here might well have been absorbed within the depreciation factors if they had been included in the C.R.N.'s in the first place. 950. If there was a necessity to allow for any underestimation of cost, (for that is what contingencies are), we believe it should have been, and in this case was, covered in the cost of replacement new figures. It is not appropriate that any further allowance should be made after an in situ valuation has been arrived at. 951. With the exception of spares which ideally should have been valued against an inventory, we believe that it is not appropriate to make an end adjustment for the other "Extras" in this claim. In normal circumstances, we would have expected them to have been included in the replacement cost estimates together with any other associated costs, before a depreciation factor was applied. 952. While there may be some argument for the inclusion of design and project management costs as well as interest, there is none, in our opinion, for a cost to cover those things (i.e. contingencies) that cannot be readily identified in a proposed project, but should have been obvious and readily identifiable in a completed project. 953. Normally there should be no need for a contingency factor in any valuation based on depreciated equivalent replacement cost, and we hold that, in this case, there is certainly no justification for any allowance under this head of contingenies. INTEREST 954. The subject of interest was dealt with in some detail in relation to the land valuation. In the same way as interest was claimed in Mr Doran's assessment for land and buildings, the claimant now puts it forward with respect to the value of the plant, as a cost item to be allowed as an end adjustment. Again we say, as an initial cost it should have formed part of the replacement cost estimate, and should be depreciated together with all the other associated costs of replacement. 955. A value is determined by the application of a reasonable depreciation rate to an estimated equivalent replacement cost to give what should represent the remaining worth. As that appears to have been done in Mr Wood's assessment, we are reluctant to allow any further tinkering with that figure even if it purports to allow for cost which may have been overlooked or not allowed for previously. We also believe that interest associated with any item which has used up most of its economic life may be so negligible as to have been reasonably accounted for in the applied depreciation rate. Many of the items included in the Shun Fung plant were well into or nearing the end of their economic lives. 956. Neither Mr Wood nor Mr Blickle specified the inclusion of interest in either of their assessments and only Mr Blickle gave evidence on the subject of whether it should be included or not. 957. Mr Blickle said that it was the first time he had heard of interest being applied to valuations of plants of this age. He also said, "I wouldn't feel that interest during construction would have any bearing on the value of the plant. Presuming that this interest occurred some in 1966, some in 1974 and presumably some in 1981." (Mr Blickle's Transcript p282). 958. Mr Blickle does not deny the exitence of the interest element; he just believes that it is of no significance in a plant of this age built over a period of time. Mr Blickle agreed that interest may be of some significance in the value of a new plant or one of about 5 to 6 years old. This seems to be another way of saying that interest for an old plant is taken care of in the depreciation factor. 959. According to Mr Blickle, if it had to be applied, then it certainly should not be applied to contigencies; but he did not consider it inappropriate for it to be applied to all the other items where inventories existed, or to the cost associated with design and engineering. On the basis of periodic payments being made over a total estimated installation period of 24 months or 25 months, he would only allow interest over half the period, i.e. 12 1/2 months at the most. 960. The claim is for a period of 15 months at 9% as set out in the claimant's instructing solicitor's letter of 7th July 1989 (Ex SF197). These time estimates were taken from Mr Medley's estimated installation programme for the proposed Shunde plant on FIG 8.4 at page 74 of his Report, Doc 30/01. The total estimated period from when orders for plant are made to "start up" is 25 months, but the claimant bases the interest allowance on 3 months as the approximate mean period for plant deliveries and the additional 12 months from completion of deliveries to commencement of production. 961. Mr Kat has submitted on behalf of the government that the Tribunal must consider whether interest is a factor, which, in the eyes of a notional willing buyer, would add value to a plant, particularly where that plant has been built and rebuilt in phases over a considerable period of time or where there has been no evidence of such an interest cost having actually been incurred. 962. Firstly, Mr Kat's reference as to how and when this particular plant was installed is irrelevant. Interest forms part of the estimated equivalent replacement cost (C.R.N.) which has no bearing on the age of the plant, or whether it was constructed in stages. 963. We have acknowledged that interest, as Mr Doran has described it, is a "sometimes forgotten cost" in many valuation exercises. It should not be so, for, as an opportunity cost, it is unavoidable and exists even where no actual payment is made. Unlike contingencies, it will occur on each and every occasion a project is constructed over any period of time. Thus, interest is an element of cost which, in our view, should be included in any depreciated replacement cost valuation as a means of allowing for the time value of money. 964. The answer to Mr Kat's specific question as to whether interest would add value in the eyes of a notional purchaser is, "Yes". It is, as Mr Doran described it, a measure of the time benefit for being able to purchase a plant already installed rather than having to forgo interest on capital during construction and installation and before there is any beneficial return from production. 965. As we agreed that the design factor adds value and should be covered in a valuation, so, too, do we agree that the time factor should be similarly covered. 966. In our opinion, the inclusion of interest in the valuation should have been provided for through an addition to each C.R.N. before depreciation, as part of the cost of installation. This is no more cumbersome than the inclusion of an amount to cover installation and shipping. It belongs as part of the equivalent replacement cost, (C.R.N.). 967. However, as we see the position, there is no scope for adding interest to the miscellany of items for which a value of $7,290,730 was agreed between the parties, nor to the agreed $3 million for the scrap yard. 968. Interest is part of value, and, unless the claimant had expressly excluded interest from both of those agreed sums, we see no basis on which it can now be claimed. 969. If we were now to apply interest to the agreed figures as an end adjustment we are of the view this would probably amount to double counting. 970. We therefore disallow interest on both those agreed figures. Apart from that we propose allowing interest for the rest of the plant and machinery on the basis of the mean period of a reasonable total installation period being one year, and the appropriate interest rate being 9%. We do not agree with the 15 months period proposed by the claimant, as periodic payments could be expected to be made over the whole project period of about 2 years. (FIG 8.4, p.74 of Doc 30/01) 971. Interest at 9% for one year on our estimated total of $42.7 million gives a figure of $3.84 million which we round up to $4 million. ASSESSMENT 972. We cannot emphasise too strongly that this is no more than an estimate of the value of the plant and machinery. It is not a figure based on hard evidence such as that of actual sales. It is the best that can be achieved on the information available, and the total of the figures obtained cannot be regarded as precise or accurately representing the total value of the plant. 973. As we have already indicated, in almost any valuation exercise a margin of error of about 10% is an acknowledged fact. In a depreciated replacement cost estimate it is bound to be greater particularly where depreciation factors discount cost by 50% or more. 974. The result of our analysis of the evidence as well as the agreements reached by the parties, leads us to the conclusion that the in situ value of plant and machinery at the date of resumption was of the order of $60 million at which figure we determine its value. The details of this assessment are set out below.
SECTION VII : LOSS ON STOCKS 975. On 31st July 1986, which is to say, the day following resumption, SFI's stock of billets and ingots awaiting rolling was 2,060 metric tons (34,060 piculs). See Doc 02/97, being Schedule 2 of SFI's Financial Statement for July 1986, which can also be found in CCS, Sect. 12, opposite page 9 of Part I. 976. As there is only one day's difference in time, we treat that 2,060 metric tons as the volume of SFI's stock of billets and ingots on the date of resumption. 977. From the date of resumption, SFI did no more melting. It did, however, continue to roll, and by 22nd August 1986 had only 1,166 metric tons of billets and ingots left to roll. 978. On 22nd August 1986, SFI conformed with an order from government to cease rolling that day, despite permission having been granted to SFI by government on 12th August to continue rolling for four weeks from 12th August. 979. As at 22nd August 1986, SFI would have needed to roll another 3,793 metric tons to fulfil all outstanding contracts. 980. If governemnt had not gone back on its word allowing SFI to continue rolling for four more weeks from 12th August 1986, SFI would easily have rolled the remaining 1,166 metric tons of billets and ingots within that four weeks period, and would have had no problem selling the resultant rebar. Then, there would have been no need for SFI to make the present claim (Head 8 on both the relocation and the extinguishment claim) for $1,797,086 which is the difference between the price for which SFI sold the remaining 1,166 metric tons of billets and ingots, and the price SFI would have received had it been allowed to roll them into rebar. The total claim under this Head is $3,584,122, but government does not dispute $1,787,036 of it. 981. Government's behaviour in not letting SFI finish rolling its billets and ingots strikes us as capricious, and unreasonable in the circumstances. 982. There was no hurry to stop SFI's rolling. SFI's staff were allowed to remain on the Junk Bay site up until January 19th 1987, and there has been no explanation, reasonable or otherwise, from government as to why SFI were suddenly forbidden to use the rolling mills. 983. The only issue relating to the $1,797,086 in dispute is the factual one of whether SFI built up its stocks of billets and ingots to an inflated level. See Exh SF51, para.1(ii). 984. A stock level of 2,060 metric tons of billets and ingots for the date of resumption strikes us as moderate. If government had behaved less unreasonably, that would have been whittled down to nothing by early September. 985. Unfortunately, those representing government appear to have become somewhat fixated by the following passage from the workpapers of SFI's auditors, Price Waterhouse, in respect of SFI's accounts for Financial Year 1985/6. 986. We now set out the text of that workpaper :
987. That extract from the workpapers bristles with all sorts of collateral issues such as whether in fact SFI's stock balance had increased "15 times" - in fact it was 5 times - and whether charging $180/pcl when the market price was $120/pcl was to SFI's disadvantage by reducing its disturbance claim against government, or to its advantage by making out the rebar price to be higher than it in fact was. 988. Based on that workpaper the Crown took the Tribunal on a wild goose chase. 989. We found Mr Roy Leung's explanations of why he still had some billets and ingots at the date of resumption satisfactory, and we deplore the sort of "fishing expedition" conducted by the Crown in cross-examining him over this. 990. As we neither think the 1,166 metric tons of billets and ingots that SFI finished up with, indicated any unreasonable conduct by SFI, nor do we think that the Crown managed to show a scintilla of failure to mitigate by SFI, we hold that SFI is entitled to an award of the full $1,797,086 in dispute under Head 8.
Appendix 1 FUTURE LOSS OF PROFITS CONSEQUENTIAL TO RELOGATION
Note: This schedule is the same as page 5 of the Deloitte revised schedules, except that the discount rates used are the ones indicated previously by the Tribunal (Tribunal Document C4 24th October 1991).
Notes: (i) Net cash flow from January 20, 1987, in the year 1986-87. (ii) No adjustment made for annual decrease in the depreciation allowances used in calculating the taxation charge & liability in this and subsequent years. (iii) Per John E. Medley of McLellan and Partners Limited. -------------------------------------------------------------- REVISED APPENDIX XXXIV DISCOUNTED CASH FLOW OF FUTURE EARNINGS AND VALUE OF NET ASSETS IN 1999
Notes - (i) A nominal rate of 33% per annum has been used in 1986-87 & 1987-88, as inflation has been taken into account in the estimation of profits for these years. (ii) Cash flows are discounted to January 19, 1987, the date the land was vacated. (iii) It has been assumed that the cash flow accrues evenly during a year. Cash flows have therefore been discounted from the mid point of each period. REVISED APPENDIX XXXIV.1 ESTIMATE OF VALUE OF GOODWILL IN 1999
* Working capital at July 1, 1986 of 11,000, as per SF226, plus increases in working capital during the period from July 1, 1986 to January 19, 1987, being 8,075 x 203/365 = 4,491. CONCLUSIONS 991. Before formally pronouncing our conclusions in the light of the findings embodied in the judgment we have just delivered, we deem it desirable for the sake of the parties and the public to put on record why it took so long to hear the present case, and to deliver judgment. 992. Hearings took place on two hundred and sixty three days, and judgment has been reserved for twenty-two months. 993. In a variety of ways, the case has suffered from giantism. 994. A huge amount of money has been claimed by SFI a sum, which, with ongoing items, exceeds one thousand million Hong Kong dollars. 995. As government's positon has been that SFI was entitled to less than one hundred million Hong Kong dollars, the gulf between the parties has been enormous. 996. A period stretching back to 1951, when SFI started in the steel business, and forward to 1999, (and even into the third millennium, for some purposes), by when SFI's Junk Bay plant, machinery and buildings would have needed replacement, came under the Tribunal's close scrutiny. 997. Far from being left to conduct a wide-ranging, generalised survey of SFI's past, present and future during that period of almost fifty years, the Tribunal had to engage in intensive study of endless pernickety details involving, say, particulars down to the diameters and lengths of the rebars of every contract SFI made with its customers starting 1978/9 through to 1985/6 (See Exh SF216 pages 72 to 79 and 125 to 136), and every rebar contract between SWS and HH from 1982/3 to 1988/9 (See SFI 216, pages 117 to 123 and 142 to 149). 998. Similar exercises of tiny detail were done for scrap and sundry other topics. 999. Besides giantism, then, the case also involved prolonged study of vast quantities of minute detail. 1000. In addition to (and probably because of) giantism and excessive detail, the case was unusually complex, not only in relation to the law, but, also, factually. The facts spanning almost fifty years of the life of a company engaged in an activity like steel-making, which has seen great technological change over the past few decades, will inevitably appear complex to anyone not actually working in that field. 1001. That complexity was compounded by the considerable divergence in the opinions of the experts called by the two sides. SFI's land valuation expert, for example, valued the Junk Bay site at $64 million while the government's expert said $16 million; SFI's plant and machinery valuer contended the Junk Bay plant and machinery was worth $90 million, while the government valuer said $46.5 million; SFI's experts (Mr Gillett and Mr Best) came up with trend figures in constant 1987/8 dollars of $2,412 for rebar and $514 for scrap in the No-Scheme-World for 1988/9 onwards, whereas SFI's expert (Mr Li) said $2,211 for rebar and $650 for scrap; on the rate for discounting SFI's future cash flows (real) in the No-Scheme-World, Mr Best's opinion was 12 1/2%, while Mr Li said 28%; one steel expert, Mr Medley, for SFI, was of the opinion that, with its 1982 installation, SFI's rebar-making capacity was at least 110,000 metric tons, and that it could build up to that level in 1983/4, whilst Mr Willcox for government stated capacity to be 100,000 metric tons, with the build-up to that level even, taking until 1985/6. 1002. If the unmodified opinions of SFI's experts were followed, SFI's claim on an extinguishment basis was $474 million, and government's experts indicated a result of $94 million. On the relocation basis, if you followed SFI's experts the result was $844 million, plus ongoing items which by now will have taken the claim to over $1,000 million, whilst the government's position has been that SFI is not entitled to anything on the relocation basis, with SFI's claim being limited to the extinguishment basis. 1003. The Tribunal was used as a tilting ground by some of the experts to test their theories. 1004. There was a hard-swearing match between the steel experts, each declaring how conservative and reasonable he was, and between them deluging the Tribunal with reams of what was, basically, arithmetic. Some of the presentations such as Exh SF81 showing tap-to-tap-times for three days' operations of one EAF in early 1982, and Exh SF175, a spread-sheet showing utilisation/availability factors, as well as each side's build-up, required many hours of patient study for full understanding. 1005. Another area where our minds were greatly exercised involved the relationship of the Average Wholesale Price Index (A.W.P.I.) for high tensile rebar in the 10mm to 40mm range, published monthly by the Census and Statistics Department, and SFI's and/or SWS's contract and/or delivery prices. There were times when we hardly knew whether we were coming or going when we were asked by SFI to find its rebar income from its contracts by, firstly, applying a four month time-lag to allow for the average delay between contract and income from consequent deliveries, secondly to deduct 1.28% from the contract price to reflect the relationship between SFI's contract prices and A.W.P.I., and then, thirdly, add back 1.66% to the price in recognition of a premium for SFI's special lengths. 1006. Most mind-boggling of all was the Capital Asset Pricing Model (C.A.P.M.) with its strange jargon of unique risk, also known as "unsystematic", "residual", "specific" or "diversifiable", and market risk, sometimes also called "systematic" or "undiversifiable". Brought to our attention were Beta Coefficients of various flavours such as "Equity", "Asset" or "Debt". Strange concepts like geometric deduction, the Efficient Market Hypothesis, and histograms hovered on the periphery. Regression Analysis made a reappearance. After wrestling for weeks with C.A.P.M. in case it might be the key to the vital issue of whether SFI's future cash flows were to be discounted at 12 1/2% or 28%, or some figure in between, it gradually dawned on us that C.A.P.M. is in the nature of a charade, belonging more to scientism than science, raising more problems than it solves, and, generally, hopeless as a source of enlightenment for the task in hand of deciding a reasonable capitalisation rate for SFI's income-stream in the No-Scheme-World. 1007. In the end, we decided that SFI's claim on the relocation basis was in the nature of a house of cards, and its extinguishment claim inflated. 1008. That was not a conclusion we could reach in a hurry. It involved endless sifting and weighing of the material with which we were deluged. 1009. The transcript of the proceedings exceeded seventeen thousand pages. There were nine experts' reports, some of them running into hundreds of pages (e.g. the 33/01-9, and 42/01 - 5A-F series), over four hundred documentary exhibits, some like SF216 and 219 amounting to two hundred or more pages. SF211, 216, and 229 were not exhibits as conventionally understood, but mixtures of law, fact, opinion, assertion, argument etc, and a great affliction to anyone trying to understand them. Over one hundred authorities, filling five volumes of photostats, were cited to us. 1010. The final submissions were in writing, amounting, in all, to thirty-eight volumes. They were amplified by final oral submissions on seventy-four days on behalf of SFI and on 21 days for government. 1011. When the case was set down, the parties estimated one month for the whole hearing. As events turned out, SFI's opening submission alone took twelve days, the Crown's four. 1012. One witness, Mr Meocre Li, the government's expert on accounting matters, was thirty-four days in the witness box, twenty-three of them being cross-examined; another, Mr Willcox, the government's steel expert, eighteen days, twelve of which were cross-examination. Mr Best, SFI's accounting expert, was twenty-five day's in the box, nine being under cross-examination, while Mr Medley, their steel expert was eleven days, on four of which he was cross-examined. 1013. If parties want to litigate on this epic scale, and have sufficiently deep pockets to do so, there is little the Tribunal can do about it, beyond making disapproving noises, cracking down on any obvious time wasting, and making orders for costs which reflect the court's views about steps unnecessarily taken or witnesses unnecessarily called. 1014. A virtue frequently extolled for legal systems in free societies is that they allow the litigant his day in court. When that gets stretched to two hundred and sixty-three days in court, one starts to wonder whether some curbs need introducing on grounds of practicality. 1015. Hong Kong's legal system is not alone in experiencing a tendency for cases of recent years generally taking longer to try, plus the odd block-buster which goes on for months or even years. 1016. Until the second half of this century, the longest case in Britain, and also, we believe, in the territories it controlled, was that of the Tichborne Claimant which, in the, middle of the last century, went on for about six months, a period which, nowadays, would not be regarded as anything too extraordinary. 1017. In the Tichborne Claimant's day, the court's record would have been kept by hand (perhaps even with a feather quill), documents would have been copied sparingly because the only way to do it was by hand, and there obviously were no computer-based information retrieval systems which allow lawyers to trawl around the world for authorities, and likewise allow experts to research the literature relevant to their subject. Faxes and telephones make information instantaneously available in Hong Kong from the other side of the globe. 1018. So long as a document has the slightest relevance, no matter how marginal, it will these days be copied for adding to the pile of the Tribunal and the lawyers. 1019. An insufficiently-noticed culprit for lengthening trials is the running transcript. Instead of being an unmixed blessing which serves only to speed up proceedings, it can also extend them by providing ammunition for cross-examination upon cross-examination, ad infinitum. When advocates had to rely on their memories or the notes they or their instructing solicitors had taken, cross-examination tended to be far briefer. 1020. The wonders of the electronic information age, which, at first, offered hope for speeding up the preparation and hearing of trials have turned out to be a false dawn. 1021. In a way not too dissimilar from how a vehicle capable of speeds of over one hundred miles per hour is likely nowadays to take longer getting through Central at rush hour than a person on a push-bike or in a rickshaw seventy-five years ago, a judge snowed under nowadays with a blizzard of electronically generated material is likely to be handicapped rather than helped in comparison with his low-tech predecessors. 1022. In our case, we found that the sheer mass of material with which we were blitzed meant that progress could only be incremental. 1023. Like our judicial predecessors from bygone eras, our quest was still the jewel in the mud. Why, compared to them, it took us so much longer to find it was because there was so much more mud. 1024. For arm-chair critics who cannot understand how judges, applying themselves conscientiously, can take so long to hear a case and write a judgment, as we have on this occasion, we suggest three steps. Firstly, proceed to Court No.15 on the seventh floor of the Supreme Court to see the sheer mass of material the Tribunal has had to marshal. Secondly, browse through some of that material, starting perhaps with Exh R98, or Exh SF175, and Report 42/05F or 33/08 and continue until it is fully understood. Thirdly, actually read this judgment, for otherwise, any comment, favourable or otherwise, is ignorant. 1025. For reasons stated earlier, we reject SFI's claim on the relocation basis, and make an award to SFI, on the extinguishment basis, for $131,030,728.
Representation: Mr David Widdicombe, Q.C. and Mr Anthony Neoh, Q.C. (inst'd by M/s McKenna & Co) for the claimant Mr Robert Carnwath, Q.C. and Mr Nigel Kat (inst'd by M/s Lovell White Durrant) for the respondent | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||