Railtrack Plc v Guinness Ltd.

This is an appeal against a decision of the Lands Tribunal exercising its arbitration jurisdiction. The dispute arose out of a major development at Park Royal in West London, on land mainly owned by Guinness Ltd, the Respondent to the appeal. The development requires an access road direct to the A40, running over railway lines owned by Railtrack and London Underground. Negotiations for the consideration to be paid by Guinness, for the air rights needed for the construction of the road, led in due course to an agreement to refer this issue to the Lands Tribunal under a so-called “terms of reference agreement” dated 16^th May 2000 (“TOR”). The TOR set out the assumptions on which the Tribunal is to proceed (which differed in some respects from the ordinary law of compensation – see below).

The wide gap between the parties is apparent from the fact that, by the time the case reached the Tribunal, the claim was put at over £33m, while the evidence on behalf of Guinness supported a figure of just over £1.5m. After a hotly contested hearing lasting over 21 days and including evidence from 25 witnesses, the Tribunal issued a reasoned decision, running to 300 paragraphs, with appendices. I would pay tribute to its clarity and comprehensiveness. They determined the value of the rights as £5m. That decision was issued on 11^th February 2002. The last paragraph indicated that it “concludes our determination of the substantive issues in this reference”; but that it would “take effect as a decision” only when the question of costs had been decided, from which point the right of appeal would come into operation.

The costs issue was finally determined on 29^th April 2002, following an exchange of written representations. In an addendum dealing with that issue, the Tribunal made clear that it had regarded the approach of the claimants as “lacking restraint and judgment”, particularly against a background that in late 1999, on the basis of the advice of their then valuers, they had offered to settle for less than £10m. The Tribunal also regretted that agreement on so many matters had proved intractable, and that the parties found themselves in conflict at so many points. Taking account of those matters, the Tribunal made no order as to costs.

Although this was an arbitration, it is accepted that it is subject to the ordinary Lands Tribunal procedure for appeals to this Court, (and is not, for example, subject to the special rules in section 69 of the Arbitration Act 1996). Although appeals from the Lands Tribunal were formerly by case stated, they now take the form of appeals on a point of law under CPR Rule 52 (Lands Tribunal Act 1949 s 3(4), as amended by Civil Procedure (Modification of Enactment) Order 2000). Permission to appeal is required from this Court (see Girls’ Day School Trust –v- Dadak 15.3.2001 per Robert Walker LJ).

The matter first came before me on a paper application for permission to appeal on 26^th June 2002. There were four grounds. I granted permission on the first, but refused on the other three grounds. On the renewed application, I confirmed the refusal of permission on grounds 3 and 4, but adjourned the application on ground 2 to come on with the substantive hearing on ground 1. In doing so, I made clear that I was unpersuaded that there was “any error in the Tribunal’s approach, let alone one of law”, but I was concerned that, on the material available (largely computer-generated), I had been unable to satisfy myself fully on the analysis of the figures. I invited the parties to attempt to put the relevant calculations into comparable and simplified form, and to entrust the task, not to a computer, but to “a reasonably numerate human-being”.

Accordingly two grounds of appeal remain live before us, the second still requiring permission, in summary:

The claim was made jointly by Railtrack and London Underground, but London Underground are not parties to this appeal. It is not suggested by either side that the issues in the appeal are affected by the fact that only one of the claimants is a party to the appeal.

In view of the relatively narrow scope of the two outstanding grounds, it is unnecessary for the purpose of this judgment to set out the facts in any detail, nor to review most of the very complex valuation exercise conducted by the parties and the Tribunal.

The proposed development was described by the Tribunal as follows:

“Guinness, with a development partner, London and Regional Properties Ltd (“L&R”), is carrying out a major development and part-redevelopment of a substantial part of its land. This development, known as the First Central scheme, will involve the construction of 116,100 sq m of Class B1 offices, 61 residential units, a 150-bed hotel and indoor leisure facilities. Planning permission was granted on 15 July 1999. The development also includes the construction of a new road into the site from a new junction to be formed on the A40, and a new underground station on the Central Line. The road is being constructed partly over railway tracks and other land owned by the claimants and partly on land owned by Guinness. Because buildings at the brewery need to be demolished to make way for the scheme, new buildings and plant are required to replace them. The new access, together with an associated road improvement called the Concord Avenue Link, will improve the accessibility of the western part of Park Royal industrial estate and is expected to produce regenerative benefits. A Single Regeneration Budget grant of £12.252m has been approved by central government and a £4.9m interest-free loan has also been negotiated with English Partnerships.”

It was common ground that the value of the access rights should be determined by a “residual” approach: that is, assessing the fully developed value of the land with the benefit of the access rights, and deducting the construction and other costs involved in achieving that value, and then comparing it with the value of the land without those rights. The difference, subject to some incidental additions and adjustments, gave the extra value assumed to be released by the acquisition of the access rights, of which 50% was treated as the claimant’s share, or “open market value of the rights”. The Tribunal had also found admissible, and therefore had before it, evidence of the figures which had in fact been discussed by the parties, in negotiations prior to the reference.

The respective valuations descended to an impressive degree of detail, requiring computer assistance. (For this purpose, Railtrack used a computer programme, known as the “Circle Systems” programme.) The examination of the respective residual valuations required the Tribunal to resolve a wide range of detailed issues, most of which are irrelevant for present purposes. Having resolved these issues, the Tribunal, with the agreement of the parties and the help of a neutral expert, used the Circle Systems programme to assist them in performing the calculations. This led them ultimately to assess the “open market value” of the rights at £5m, and, having also looked at the figures discussed in the actual negotiations, they concluded that the figure of £5m was “one that the parties would have agreed” (para 288).

This issue concerned the Tribunal’s application, to the somewhat special facts of the case, of the “willing seller” principle under section 5(2) of the 1961 Act. That section provides:-

“5.

Compensation in respect of any compulsory acquisition shall be assessed in accordance with the following rules:

…..

(2)

The value of land shall, subject as hereinafter provided, be taken to be the amount which the land if sold in the open market by a willing seller might be expected to realise…” (emphasis added)

I have already noted that the assumptions set out in the TOR differed in some respects from ordinary compensation principles. For present purposes, it is sufficient to refer to the Tribunal’s summary:

“41.

The rights are defined to mean Railtrack’s proprietary rights in Coronation Road, the rights to construct, maintain and use the link road across the railways, and the land required for construction of the underground station. The rights were the subject of two Deeds of Grant, of the same date as the Terms of Reference agreement.

42.

Specific assumptions as to the basis of valuation are set out in clause 8. They include the following. Valuation date is 16 May 2000. It is to be assumed that the rights had been acquired compulsorily under Part IX of the Town and Country Planning Act 1990 for the purposes of the Scheme, and that they had been acquired at market value. Rule (3) in section 5 of the Land Compensation Act 1961 is disapplied, and it is provided that the Pointe Gourde principle does not apply so as to exclude consideration of the proposed development of land owned by Guinness, its parent company Diageo PLC or any other company in the group or any other land. Account may be taken of the fact that Guinness/Diageo own land beyond the scheme land. No set-off for betterment is to be made in relation to any new station constructed as a requirement of a planning obligation, but betterment to the scheme arising in other ways can be taken into account”

The assumption that the land had been acquired compulsorily under the Town and Country Planning Act 1990 carried with it the implication that the rules in the Land Compensation Act, including rule (2), would apply, save as varied by the TOR. As to that the Tribunal said:

“43.

Under rule (2) of section 5 of the 1961 Act the value of the land is the amount which the land if sold in the open market by a willing seller might be expected to realise. There is no dispute between the parties that any purchaser who, in the real world, might be a bidder for the land can be taken into account as forming part of the market, and that Guinness, and indeed London and Regional, could thus be regarded as potential bidders. In fact, the evidence on both sides effectively assumed that Guinness would be the purchaser in this hypothetical transaction.”

More specifically, in relation to rights over the railway, clause 8((c) of the TOR required the Tribunal to determine the Dispute on the basis (among other points) that:

“… the Rights had been acquired on the basis of the open market value of the Rights required for accesses over the railway.”

The nature of the dispute which underlies the first ground of appeal, and the Tribunal’s conclusion on it appears from their decision:

“44.

The parties disagreed about the assumption that should be made about the vendor. Mr Purchas submitted that rule (2) required the assumption that the vendor is a hypothetical vendor. Thus there were to be excluded from consideration the personal characteristics of the actual vendor that might affect the negotiations, including any particular obligations or pressures to which he might be subject. Mr Ash submitted that, while it is right that the vendor is hypothetical, what is being sold - air rights over a railway line - is not. The hypothetical vendor should therefore be understood to be a hypothetical railway infrastructure company….

45.

We agree with Mr Ash’s submission…

46.

The correct assumption, in our view, is that the hypothetical vendor would be a company or authority with the function of maintaining the track and associated permanent features of a railway system. In the real world, such a company or authority would be regulated and subsidised by central government, and would be subject to pressures, in relation to a project like the present, of a sort that can be termed political. We can see no reason for disregarding these realities. At the same time the company or authority would be concerned to extract a proper value for the rights that it granted. It would not see its role as subsidising the development or any part of it. Railtrack and LUL in the real world thus seem to us to be reasonably representative of the hypothetical vendor, and we think it clearly relevant, therefore, to have regard to the negotiations that actually took place between the parties.”

As appears from the last comment, the practical significance of this point lay in the extent to which the Tribunal could gain assistance from the figures put forward by Railtrack’s representatives in the actual negotiations. Later in the decision, the Tribunal explained the use made of that information:

“91.

We have said that Railtrack and LUL in the real world seem to us to be reasonably representative of the hypothetical vendor, while the evidence on both sides assumes that Guinness would be the purchaser in the hypothetical transaction. The evidence on the negotiations between the parties on the price for the rights was ruled to be admissible, and we have had regard to it. While it has played no part in the residual valuation that we, adopting the method of the parties, have followed, it does in our view provide some confirmation of the result that we have arrived at.”

They referred to the evidence of negotiations between June and December 1999, which had culminated in an apparent offer by Railtrack to sell the rights for £9.7, and to the evidence from Mr Kirby (sales director of Railtrack) of the circumstances in which those offers had been made. They commented:

92.

Mr Purchas, relying on the evidence of Mr Kirby, said that the offers, and in particular the offer of £9.7m, were the result of the political pressure that was being applied to Railtrack and LUL. While we have little doubt that the desire to achieve the regeneration of Park Royal through the Guinness development led to pressure on Railtrack and LUL to reach a settlement, we see no reason to suppose that this pressure exceeded what a hypothetical vendor would have been subjected to. Indeed we have concluded that Railtrack and LUL were representative of the hypothetical vendor.”

They rejected a suggestion by Mr Kirby that the apparent offer was not a real one. They said:

“95… In our view it represented the amount that Railtrack and LUL, on the basis of professional advice and after months of negotiation, were prepared to settle for. At the stage at which the offer was made neither party had carried out the extremely detailed and sophisticated calculations on which their evidence before us was based. But they had made assessments of the sort that, in the world of commercial negotiations, would normally be sufficiently detailed to enable agreement to be reached. We consider that the offers made in November and December 1999 suggest that the price now contended for by Railtrack/LUL is very substantially too high, and that the value that we have arrived at is one that would have been reached between a willing buyer and a willing seller.”

Finally, at the end of their decision, having arrived at the valuation for the rights of £5m, they said:

“288 Having arrived at this figure we have thought it right to take an overall view in the light of all the factors bearing upon the hypothetical negotiations, including the bargaining position of the parties, and the negotiations that actually occurred and to consider whether the amount is the one that the parties would have agreed. We are satisfied that it is….”

Before us Mr Purchas submitted that this approach was wrong in law. He said that the principle underlying rule (2) was that the market value should not be affected by the particular characteristics of the actual vendor, including any personal proclivity for or resistance to sale. The objective was to determine what the market would pay irrespective of any individual or personal factors special to the actual owner and vendor. The Tribunal had wrongly made the assumption that the vendors of the access rights were effectively Railtrack and LU, and were thus unduly influenced by their particular corporate characteristics and their particular financial position and control. The Rights should have been valued as if put on the market by a hypothetical vendor, not subject to the pressures and considerations which affected Railtrack and LUL.

For Guinness, it is submitted that it would have been wholly unreal for the Tribunal to ignore the fact that what was in issue was the value of rights over a railway line, and would therefore be owned by a railway company. The Tribunal were entitled to find as a matter of fact that the negotiations were reasonably representative. In any event, even if this ground of appeal were correct, it would have made no difference. The Tribunal’s award was based on its own residual valuation, which was not directly affected by the evidence of the negotiations.

Both parties relied on Trocette Property Co Ltd v Greater London Council (1974) 28 P&CR 408. In that case, claimants held property on lease from the predecessor of the GLC, which had only a few years to run, but which had substantial development potential if an extension were granted. One issue was whether, in assessing compensation, the Tribunal should take account of the known unwillingness of the GLC as lessor to grant an extension, or whether it should consider the likely position of a hypothetical lessor. The Court held that there was nothing in the Act to justify excluding the views of the actual lessor, so far as known to the market.

Megaw LJ contrasted this case with the position on a compulsory acquisition of a freehold interest, having regard to rule (2):

“In a case which concerns the compensation for the acquisition of a freehold interest in land where there is no leasehold interest there is, as I see it, no problem. No question can arise as to the special characteristics of the particular freehold owner. The compensation which falls to be paid does not vary according to whether the freehold owner would in fact refuse to sell, if he had the choice or whether, if he sold, he would be likely to be a hard bargainer or a soft bargainer, or whether he would insist on conditions which would reduce the value of the land in the hands of a purchaser. Questions whether he is young or old, rich or poor, miserly or spendthrift, are wholly irrelevant…

The ‘willing seller’ is a hypothetical character. There is no justification for attaching to him, so as to increase or decrease the assessment of compensation, any special characteristics. He is to be assumed to be willing to sell at the best price which he can reasonably get in the open market” (p 415-6).

He distinguished this from the case before him, where the attitude of someone other than the “willing seller” was in issue:

“The fact that this freehold owner, the G.L.C., is no longer prepared to do that which is necessary to create the “marriage value” may be described as a peculiar characteristic of this particular landlord, but I see nothing in the rules which enables it to be ignored on that account. It is a fact which would indeed have affected the value of the leasehold interest if it had been offered on the open market by the claimants on the relevant date” (p 416)

Agreeing, Lawton LJ said:

“The assessment of compensation in cases such as this is a most difficult task calling for the judicial use of fertile imagination…. It is important that this statutory world of make-believe should be kept as near as possible to reality. No assumption of any kind should be made unless provided for by statute or decided cases.” (p 420)

Of rule (2) he said:

“… When applying this rule, the Lands Tribunal has to disregard what may well be the reality of a most unwilling seller and, in the absence of evidence on the point, use its imagination to envisage how buyers in an open market would react to what was on offer…” (p 421)

Mr Purchas relied on the statement that the willing seller is “a hypothetical character” and that “no question can arise as to the special characteristics of the particular freehold owner”. Mr Ash on the other hand relies on Lawton LJ’s statement that “the statutory world of make-believe should be kept as near as possible to reality”. He refers also to the statement to similar effect by Peter Gibson LJ in Hoare (VO) v National Trust [1998] RA 391, 415, where he applied the approach of Lawton LJ in a rating context, and emphasised “the necessity to adhere to reality subject only to giving full effect to the statutory hypothesis”; he called this “ the principle of reality”.

In my view there is no principle of law at issue here. The need to arrive at an “open market” valuation arises under many statutes, and the general approach is not in doubt. A recent authoritative statement can be found in Walton v IRC [1996] STC 68, 83 per Peter Gibson LJ. In that case the statute, relating to capital transfer tax, required the value to be assessed as “the price which the property might reasonably be expected to fetch if sold on the open market…”. Peter Gibson LJ said:

“Although the statute says nothing about a willing seller or a willing buyer the concept of the open market automatically implies a willing seller and a willing buyer, each of whom is a hypothetical abstraction. However the willing buyer ‘reflects reality in that he embodies whatever was actually the demand for that property at the relevant time’. (see IRC v Gray [1994] STC 360, 372 per Hoffmann LJ). Whilst both the seller and the buyer are assumed to be willing neither is to be taken to be over-eager… The statue assumes a sale. That means… that the vendor, if he is offered the best price reasonably obtainable in the market, cannot be assumed to say that he will not sell because the price is too low as inadequately reflecting some feature of the property nor can the purchaser be assumed to say that he will not buy because the price is too high...” (p 85 – 86).

In substance this is no different to professional definitions of “market value”; see, for example, the RICS’ Practice Statement 4, 1/3/00:

“The estimated amount for which an asset should exchange on the date of valuation between a willing buyer and a willing seller in an arm's-length transaction after proper marketing wherein the parties had each acted knowledgeably, prudently and without compulsion.”

In the present case the Tribunal had to assume a sale of access rights over a railway between a willing seller and a willing buyer. I agree with Mr Ash that it was not required to ignore the fact that, since these were rights of access over a railway, the willing seller would by definition be a railway company. In any event, the only “political pressures” which could be relevant would be pressures to sell at a price less than that which would otherwise be obtained. Outside any special statutory or commercial context, any company (statutory or not) would normally be expected to seek the best price for its assets. It is not clear why a railway company should be any different. Not surprisingly, the Tribunal found that, whatever the pressures, the hypothetical railway company “would be concerned to extract a proper value for the rights that it granted”.

We were referred to agreed extracts from the evidence of Mr Kirby, the sales director of Railtrack Property, who was involved in the negotiations. He indicated that Railtrack were reluctant to specify a figure in the negotiations, because he thought that “Guinness were going to use political pressure to achieve their aims”, and he referred to the “political heat” which greeted the £9.7m offer. Later, in re-examination, he commented that, Guinness’s agents were seeking to use political pressure to get the rights on a non-commercial basis, but he had always made clear that he was “happy to negotiate but only a commercial basis”. He explained the £9.7m offer as:

“… not … on a commercial basis. It was an offer made really at that time to try and induce them to work on profit share.”

Insofar as that evidence shows the possibility of political pressure being used to overcome Railtrack’s reluctance to sell, it does not advance the debate, since a willing sale had to be assumed. Further, Mr Kirby made clear that, whatever Guinness’s hopes might have been, he was only prepared to negotiate on a “commercial basis”. Therefore, the only question was whether the offer of £9.7m was a genuine offer capable of being accepted or, as he implied, simply a tactical ploy. The Tribunal considered that point, and held, contrary to his evidence, that -

“… it represented the amount that Railtrack and LUL, on the basis of professional advice and after months of negotiation, were prepared to settle for”. (para 95).

Having reached that finding, there can be no doubt in my view that they were entitled to take that offer into account, as part of their overall evaluation of the evidence.

Before leaving this issue, I should mention a point made by Mr Purchas on the interpretation of rule (2). He noted that in rule (2) itself there is a reference to the “willing seller”, but not to the “willing buyer”. As Mr Purchas rightly says, rule (2) was originally enacted in the Acquisition of Land (Assessment of Compensation) Act 1919, which followed the Report of a Committee under Leslie Scott QC, relating to the Acquisition of Land for Public Purposes (Cd.9229; the background is discussed in Law Commission “Towards a Compulsory Purchase Code (1) Compensation”, CP No.165, para 2.5). Mr Purchas contrasts the wording of rule (2) as enacted, with the recommendation of the Scott Committee. They had said:

“We think it desirable that it should be definitely provided that the standard of the value to be paid to the owner is to be the market value as between a willing seller and a willing buyer…” (para 8, enphasis added)

Mr Purchas submitted that the fact that rule (2) as enacted refers only to the “willing seller” shows that the Scott recommendation was only partially accepted; the particular characteristics of the actual vendor had to be ignored, not those of the actual buyer. In the end, Mr Purchas did not rely strongly on this point in the present case. However, since it may affect other cases, I should comment briefly.

In my view, the submission reads too much into the wording of rule (2). The rule requires the assumption of an “open market” sale. As Peter Gibson pointed out in Walton (see above), the concept of an open market “automatically implies a willing seller and a willing buyer”. The reason for the specific mention of the “willing seller” is apparent from Horn v Sunderland [1941] QB 26, where Scott LJ (as he had then become) explained the background to the 1919 Act:

“The main object of the Act of 1919 was undoubtedly to mitigate the evil of excessive compensation which had grown up out of the theory, evolved by the courts, that because the sale was compulsory the seller must be treated by the assessing tribunal sympathetically as an unwilling seller selling to a willing buyer…” (emphasis added)

Similarly, later in the same passage, he referred to rule (2) as designed to “check exaggerated prices”, by reversing

“the old sympathetic hypothesis of the unwilling seller and the willing buyer which underlay judicial interpretation of the Act of 1845.”

Thus, the mention of the “willing seller” was to emphasise the departure from the previous law; the “willing buyer” was already implied. There was no hint in this judgment that Scott LJ himself thought that his recommendation had not been adopted in full.

I accept of course that the assumption of “willingness” has different implications for buyer and seller. This was succinctly explained by Lord Romer in the Privy Council (The “Indian” case [1939] AC 302, 312), where he said:

“The compensation must be determined by reference to the price which a willing vendor might reasonably expect to obtain from a willing purchaser. The disinclination of the vendor to part with his land and the urgent necessity of the purchaser to buy must alike be disregarded.”

Thus in the present case the Tribunal must disregard any “urgent necessity” for Guinness to buy the land, and any “disinclination” by Railtrack to part with it. Both are assumed to be to be willing to deal at “the best price reasonably obtainable”.

For the reasons already given, this ground of challenge fails.

The parties agreed in making allowances in their valuations for “profit” or “risk” The Tribunal explained the concept as follows (para 266):-

“… Developer’s profit represents a return for enterprise, organisation, overheads and risk. In general, the greater the risk attached to a particular development the higher the level of profit required.”

To see how “profit” or “risk” came into the Tribunal’s findings, it is necessary to set out the figures which underlay their award. Their calculation was set out in their Table 9 (para 287):

The figure of £14.8m was a computer-generated figure, produced as I have explained by the neutral expert using the Circle Systems programme. Again, it is helpful to see how it was presented by the Tribunal. Appendix 1 to the Decision was in the form of a “Report”, entitled “Circle Systems Development Appraisal – First Central”, and consisting of 9 pages of data and calculations. The main elements can be seen sufficiently from the “Appraisal summary”, which I have attached as an appendix to this judgment.

The relevant output from that table is (surprisingly, at first sight) not the final figure (“Balance”), but the figure for “Site Purchase Cost”, £14.807,388, which, as rounded to £14.8m, is the starting-point for the Tribunal’s Table 9. The unexpected order, no doubt, was due to the way the programme was set up. It was apparently designed to enable the expected “Net Realisation” (including sales and rental income) to be compared with the “Outlay” (including the costs of acquiring the site and building), in order to arrive at the expected “Balance” or “Profit”. However, in this case it has been used the other way round. A “profit” of 20% was assumed (as indicated by the “Performance Measure”); the other elements of the calculation (Net Realisation, Construction Costs etc) were determined by the Tribunal on the evidence; and the programme was used to arrive at the “Site Purchase Cost” which would be consistent with that profit.

Before returning to the body of the Decision, three points should be noted about this calculation, and its relation to the Tribunal’s reasoning:-

The Tribunal’s conclusions on this aspect were as follows:

“265.

Profit and risk – Mr Banks has deducted a profit of 20% on all costs (including land value) in his valuation of the First Central land. No further deduction has been made when valuing the access rights. Mr Whitfield has deducted 15% profit on the costs of Buildings A and C to J and 20% on Building B (assumed to be a speculative development) in arriving at his residual land value. No profit on the land value has been allowed on the grounds that Guinness would not seek a return on the value or cost of the First Central land. They would take the view that this land has been owned for 60 years and there is no cost to the company in continuing to hold it for a further period, making it available for development on a plot by plot basis. In addition, when calculating the value of the access rights, Mr Whitfield has made a deduction of 33% for risk.

“266 Developer’s profit represents a return for enterprise, organisation, overheads and risk. In general, the greater the risk attached to a particular development the higher the level of profit required. Mr Banks has used the conventional 20% profit on costs and land value. Mr Whitfield has allowed mainly 15% on costs, excluding land value due to the particular situation of Guinness.

267.

Having regard to our reservations regarding Park Royal as an office location, we prefer Mr Banks’s overall deduction of 20% on costs. We agree with Mr Whitfield that allowance for risk should be made when calculating the value of the access rights. The payment for these rights is part of the cost of development and should attract a development return or profit. In our judgment this deduction should also be 20%, not the 33% used by Mr Whitfield.”

To understand the respective submissions before us, it is necessary to distinguish between the “profit” or “risk” allowances on three different elements: (a) construction and other costs (other than site value), (b) site value, (c) access rights. (For this purpose I gratefully acknowledge the assistance derived from the Tables prepared by the parties, in response to my invitation.)

Element (a) is not controversial for our purposes. Mr Banks had deducted 20% on all the costs; Mr Whitfield had deducted 15% of the costs of most of the buildings, but 20% in the case of one (Building B) “assumed to be a speculative development”. The Tribunal preferred Mr Banks’s approach, having regard to their “reservations regarding Park Royal as an office location”.

As to element (b), the Tribunal noted that Mr Banks’ profit allowance of 20% covered land value, whereas, in Mr Whitfield’s valuation, no profit on the land value had been allowed. This, in the Tribunal’s summary, was because of the special position of Guinness, who would in effect have written off the value of a site that they had owned for 60 years. The Tribunal (as appears from para 267 above) followed Mr Banks in treating land value as part of the overall cost, subject to his allowance of 20%. They did not in terms comment on Mr Whitfield’s reference to the special position of Guinness, as a reason for making no allowance on the land element.

The third element (c) reflected an aspect of the evidence of Mr Whitfield, which had been strongly attacked by Mr Banks. The Tribunal had earlier recorded the respective views. Having noted that Mr Whitfield differed from Mr Banks in making no allowance for profit on land (para 248), they said:

“249 Mr Whitfield has made a deduction of 33% for risk. This reflects the inherently volatile nature of the calculations used to arrive at site value. It acknowledges that risk is carried by Guinness, not by Railtrack…. In the market developers will speculate but, wherever possible, they do so by sharing value when realised. Where there has to be a down payment against the possibility of future profit there is substantial risk and a substantial risk allowance is required. In this case there are uncertainties regarding timing and costs and there is sensitivity inherent in the residual method of valuation…” (para 249)

Earlier they had summarised Mr Banks’ comment on this element:

“219…. A profit level of 20% on costs throughout the development would be used with no further allowance for risk. There is no risk associated with obtaining the access rights. Mr Whitfield’s adjustment for risk of 33% is without precedent and wrong in principle. Marriage value should take account of investment risks. Planning permission has been granted and it is common ground that there is a prospect of enhancement in value, even though the amount of that enhancement and the costs of development are in dispute. Development could take place without incurring technical costs. The residual method of valuation makes allowance for risk and uncertainty. The deduction of a further 33% is arbitrary and unjustified. The site could have been sold to a developer at the valuation date. This would have transferred the risk.”

The Tribunal apparently agreed with Mr Whitfield’s approach, but made a deduction of 20%, rather than 33% as proposed by him (para 267 above).

Further explanation of the Tribunal’s reasoning on this aspect is found in their response to a request from Railtrack, made after receipt of the reasoned decision, but before it became “effective”. On 2^nd April 2002, Railtrack’s solicitors wrote to the Tribunal for clarification:

“The point is whether it was the intention of the Tribunal to deduct 20% on cost, excluding the residual land value, pursuant to the first sentence in para 267 (and then a further 20% on the residual land value pursuant to the second and third sentences in that paragraph). Alternatively did the Tribunal intend to deduct 20% on cost and the residual land value pursuant to the first sentence and then the further 20% pursuant to the second and third sentences?”

The Tribunal responded on 5^th April as follows:-

“In our residual valuation to find the value of the first central land we have accepted and incorporated Mr Banks’ overall deduction of 20% for profit on total cost (including acquisition price). In our calculation of the value of the rights we have accepted Mr Whitfield’s evidence that there should be an allowance for risk and have made a deduction of 20% from the increase in value for risk. The first 20% deduction (profit) relates to the value of the land and the second deduction (risk) related to the value of the rights.”

Mr Purchas puts his complaint in various ways, but in essence his case is that the Tribunal’s final deduction of 20% was unjustifiable, or in any event insufficiently justified in the Decision. He relies on a number of points:

Mr Purchas also sought to gain assistance from the fact that the Tribunal had rejected his clients’ case on many other elements in the calculation, thereby, as he said, adopting a “cautious” approach, which made any further allowance for risk inappropriate. I do not accept that point. The rejection of those aspects of his case is equally consistent with their being examples of what the Tribunal described as “lack of restraint and judgment” (see above).

Before giving my conclusions on this ground of appeal, I should comment on two legal issues which have been raised.

First, I would accept that, if the Court is satisfied that there was “double-counting” as Mr Purchas claims, or if it is left uncertain as to that possibility, it may intervene, even on an appeal limited to issues of law, and may make an appropriate order remitting the matter to the Tribunal to deal with that point.

We did not hear detailed argument on the point. However, Mr Purchas relied by analogy on Aslam –v- South Beds DC [2001] EWCA Civ 514, which concerned the compensation payable for discontinuance of a slaughterhouse business. The Court had to examine evidence relating to the somewhat esoteric elements of the calculation of the profits of a slaughterhouse business. The Court emphasised that it was not seeking to lay down any new proposition of law, but felt able to intervene in relation to two matters, where the Tribunal had either proceeded on no evidence, or had failed to take account of “the whole of the evidence” on a particular point (see eg para 26).

This case is no more than illustration of the point that issues of “law” in this context are not narrowly understood. The Court can correct “all kinds of error of law, including errors which might otherwise be the subject of judicial review proceedings” (R v IRC ex p Preston [1985] 1 AC 835, 862 per Lord Templeman; see also De Smith, Woolf and Jowell, Judicial Review 5^th Ed para 15-076). Thus, for example, a material breach of the rules of natural justice will be treated as an error of law. Furthermore, judicial review (and therefore an appeal on law) may in appropriate cases be available where the decision is reached “upon an incorrect basis of fact”, due to misunderstanding or ignorance (see R (Alconbury Ltd) v Secretary of State [2001] 2 WLR 1389, 2001 UKHL 23, para 53, per Lord Slynn). A failure of reasoning may not in itself establish an error of law, but it may “indicate that the tribunal had never properly considered the matter…and that the proper thought processes have not been gone through” (Crake v Supplementary Benefits Commission [1982] 1 All ER 498. 508).

Accordingly, without wishing to lay down any definitive tests, I consider that there is in principle sufficient legal foundation for Mr Purchas’ submission on this ground of appeal, if made out on the facts.

The other point concerns the status of the letter of the Tribunal of 5^th April 2002. Although, as I have said, that was written in response to a request from Mr Purchas’s clients, he now submits that we should take no account of it, because it did not form part of the Tribunal’s formal decision. He says that the powers of the Tribunal to amend or add to an award are limited to those conferred by the Arbitration Act 1996 s 57(3) (applied by the Lands Tribunal Rules r 32(3)); that is, the power to correct a “clerical mistake or error”, or to make “an additional award” in relation to a claim “not dealt with in the award”. He also relies on the line of authorities, mainly judicial review cases in the housing field, in which the Court has considered how far it should have regard to affidavit evidence supplementing or modifying the reasoning in the letter of decision (see the useful review by Burnton J in R (Nash) v Chelsea College of Art and Design [2001] EWHC Admin 538, para 34).

With respect to Mr Purchas, this is a hopeless submission, for three reasons:

I should add that, since the hearing, I have become aware of the Tribunal’s own consideration of this issue in Shraff Tip Ltd v Highways Agency [1999] RVR 322. The President held that, under this procedure, there is no final decision until the costs are decided, and that the parties can apply to have the hearing re-opened. That seems correct in principle, although in practice no doubt the Tribunal, having given what was intended to be a fully reasoned decision, would wish to resist any general reopening of the issues at that stage. However, there may be circumstances where clarification is required, or where some material point has been overlooked.

Some guidance may be obtained by analogy from the circumstances in which the High Court may reconsider a judgment, before it has been formally drawn up. These were reviewed recently by this court (following the changes made by the CPR) in Stewart v Engel [2001] 1 WLR 2268. Sir Christopher Slade (at p 2274) referred to the judgment of Neuberger J in Re Blenheim Leisure (Restaurants) Ltd (No 3) (1999) Times, 9 November, giving “some helpful examples” of cases where the jurisdiction might justifiably be invoked before the order in question was drawn up:

“… a plain mistake on the part of the court; a failure of the parties to draw to the court’s attention a fact or point of law that was plainly relevant; or discovery of new facts subsequent to the judgment being given. Another good reason was if the applicant could argue that he was taken by surprise by a particular application from which the court ruled adversely to him and that he did not have a fair opportunity to consider.”

Sir Christopher Slade commented:

“It is to be observed that in all these instances, if the court had no power to reconsider its order before it was drawn up, the only remedy open to the party prejudiced would be by way of appeal from the order. Though on such hypothetical facts an appeal would itself have a good chance of success, common sense suggests that in such cases the judge who made the order should himself have the power to vary it before the appeal procedure has to be set in motion, with the likelihood of exposing all parties to far greater expense and delay than an application to the court of first instance.”

I see no reason why the practice of the Tribunal, before its decision has become effective, should be any more restrictive than as explained in that passage. The “common sense” of applying to the Tribunal itself to correct such faults, in order to avoid unnecessary delay and expense of an appeal, is equally strong.

As I have said, there are two aspects to Mr Purchas’ attack; the first concerns the “profit” on the land element, the second, the final “risk” allowance. As to the former, taken on its own, he has no real complaint, since the Tribunal adopted his own expert’s approach. At most, it may be said that they did not clearly explain their reasons for rejecting the alternative, but Railtrack cannot complain of the result.

Mr Whitfield’s approach does in any event appear surprising at first sight. It seems odd that, in an assumed arms-length negotiation, Guinness should be treated as generously allowing the current value of their land to be ignored, whatever the historic cost. Further light was thrown on this aspect by what we were shown of the evidence and submissions before the Tribunal. Although Mr Whitfield had indeed justified his approach by reference to the special position of Guinness, he indicated in evidence that this was “a concession on the part of Guinness and a deviation from a normal accepted market approach” (Transcript, Day 21, p 121). It was necessary, on his figures, to avoid arriving at a negative figure for the value of the rights. In Guinness’ closing submissions, it was submitted that, if the Tribunal’s findings were to lead to a higher valuation, the concession should be treated as withdrawn. It seems therefore to have been common ground that, on a “normal market approach”, Mr Banks was correct.

The critical issue, therefore, relates to the final “risk” allowance. The Tribunal’s stated reasons are brief:

“We agree with Mr Whitfield that allowance for risk should be made when calculating the value of the access rights. The payment for these rights is part of the cost of development and should attract a development return or profit. In our judgment this deduction should also be 20%, not the 33% used by Mr Whitfield.”

The actual figure raises no material issue. The percentage allowance was a matter of judgment, which was probably not susceptible to more detailed explanation, and cannot be said to be unreasonable in law (cf Scottish Exhibition Centre v Strathclyde Regional Assessor [1994] RA 209, 214 per Lord Clyde).

As to the principle, it is a fair comment that the Tribunal’s reasoning on this point is somewhat compressed. Further, the statement that the extra deduction should be made, because the payment for the access rights was “part of the cost of development”, might be thought to imply that no deduction had yet been made for the rights as part of that cost. This would appear to overlook the fact that the access rights had been included, as part of “Site Purchase Cost”, in the Circle Systems appraisal, and therefore had been subject to the same deduction.

However, I think the key lies in the previous sentence, where the Tribunal made clear that they were adopting Mr Whitfield’s reasoning. As has been seen, he explained the extra deduction in this way:

“It acknowledges that risk is carried by Guinness, not by Railtrack…. In the market developers will speculate but, wherever possible, they do so by sharing value when realised. Where there has to be a down payment against the possibility of future profit there is substantial risk and a substantial risk allowance is required…” (para 249, see above))

As I understand it, the point is that the “profit” taken into account in the Circle Systems appraisal is the allowance which would be made by any prospective developer, needing to acquire the land (with access rights) from a third party. It takes no account of the different positions of Railtrack and Guinness, as vendor and purchaser respectively of those rights. Railtrack is receiving cash; Guinness is making “a down payment against the possibility of future profit”, which carries a separate element of risk. The additional 20% reflects that additional risk.

This interpretation is, in my view, reflected in the Tribunal’s letter of 5^th April, where they confirm that this deduction was based on Mr Whitfield’s evidence, and said:

"The first 20% deduction (profit) relates to the value of the land and the second deduction (risk) related to the value of the rights.”

More compendiously, the first is developer’s “profit”, built into the Circle Systems appraisal, and based on the total outlay, including the value of the land; the second is the additional “risk” (as explained by Mr Whitfield) borne exclusively by Guinness as purchaser of the access rights.

In conclusion, the second ground of appeal raises a reasonably arguable issue, which justifies the grant of permission to appeal. However, for the reasons given, I think that the Tribunal’s conclusion was justified on the material before them, and adequately reasoned. Furthermore, there was no breach of natural justice. It is clear from the summaries made by the Tribunal of the respective positions of the experts (see above), that the critical point was fully discussed at the hearing.

Accordingly, I would dismiss the appeal on this ground also.

There may be some lessons to be learnt from the progress of this reference and appeal.

If, as here, the parties are intending to rely on a complex valuation exercise, based on a computer model, it is of the utmost importance that they should seek to agree a common model. If necessary, the Tribunal should give directions in order to achieve this. At the same time, all concerned should be aware of the inherent imprecision of such an exercise, which cannot be more than a guide to the exercise of professional judgment. In their pre-hearing discussions, the experts should seek as far as possible to narrow the number of variables, and the range of variation. A degree of give-and-take on the less critical elements is essential, so that the evidence and cross-examination can concentrate on the key issues and variables. The pre-planning of the case also needs to take account of the possibility that the Tribunal may need its own computer assistance (as happened here) in evaluating the case.

As far as concerns the form of the decision, there is generally no objection to the Tribunal appending a computer-generated table to the decision, provided its purpose and effect are clear from the body of the decision. They are entitled to regard the decision (particularly in an arbitration) as directed principally to the parties, who will be familiar with the content and form of presentation. However, when it comes to an application for permission to appeal, the position changes. Such an application will be considered in the first instance on paper, by a Lord Justice who will have no previous knowledge of the case, and may have no experience of the valuation techniques involved. It is incumbent on those preparing the application, on behalf of the prospective appellant, to ensure that the court is able fully to understand the material relevant to the grounds of appeal. If detailed understanding of the figures is needed, as was the case here, some further explanation, either in a witness statement or in the skeleton argument, may be required. If so, it should be kept as simple and uncontroversial as possible, and the figures used should be clearly traceable to those used in the decision.

In conclusion, I would grant permission to appeal on ground 2, but dismiss the appeal on both grounds.

I agree.

I also agree