England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE CHANCERY DIVISION
THE HON MR JUSTICE HENDERSON

		Royal Courts of Justice Strand, London, WC2A 2LL
		20/12/2007

Sir Peter Gibson:

1. This is an appeal by the claimant, Jayam NV ("Jayam"), from the dismissal by Henderson J on 1 March 2007 of its claim against the defendant the Diamond Trading Company Ltd ("DTC"). Permission to appeal was refused by the judge. On a renewed application by Jayam to this court, Lawrence Collins LJ granted limited permission to appeal. I must return to what those limits were at a later point in this judgment.

The facts

2. The facts found by the judge have been meticulously and lucidly set out in his judgment and are not in dispute. I essay a summary only in an attempt to make this judgment comprehensible on its own.
3. DTC is the sales and marketing arm of De Beers, the colossus in the diamond world. Historically De Beers had a virtual monopoly in the production and sale of rough (ie uncut and unpolished) diamonds from its mines in South Africa. More recently other sources have emerged but in 2002 De Beers' share of diamond production was still in excess of 50% and the European Commission took the view that DTC had a dominant position in the worldwide market for the supply of rough diamonds.
4. DTC sells its rough diamonds to its customers, known as Sightholders, at periodic viewings, known as Sights, at which the diamonds allocated for sale are made available for physical inspection. There are 10 Sights each calendar year divided into two six months selling periods, known as H1 and H2 of each year. Thus the first Sight in the second selling period of 2003 is known as Sight 6, H2/2003. The number of Sightholders varies from time to time. There are about 90 worldwide.
5. Jayam is a firm of diamond manufacturers and jewellers based in Antwerp. It has been a Sightholder since 1974. Its shares are owned by Mr Mahendra Mehta and his family.
6. Demand by Sightholders for rough diamonds usually exceeds by a substantial margin the supply which DTC makes available for sale. Before 2000 Sightholders were selected at twice yearly meetings of DTC's executive committee and allocations of rough diamonds were determined by a team of account managers and a committee of senior executives.
7. In 2000 following a strategic review DTC announced the introduction of a new system of supply arrangements known as Supplier of Choice ("SOC"). This was intended to place DTC's relationships with its Sightholders on a formal basis and to introduce a new contractual framework both for the selection of Sightholders and for the allocation to them of diamonds. SOC was implemented with effect from 7 July 2003 after discussions with, and clearance (in the form of a comfort letter) from, the European Commission.
8. The SOC arrangements in the period July 2003 to June 2005 are embodied in the following documents all published by DTC at the same time and known as the SOC documentation:

(1) DTC Sightholder Policy Statement;

(2) Sightholder Criteria and Other Considerations;

(3) Notes to the Sightholder Criteria and Other Considerations;

(4) DTC Diamond Best Practice Principles;

(5) Conditions of Sale;

(6) Ombudsman Terms of Reference.

9. The Policy Statement has the following introduction:

"DTC believes it to be in the interests of all those participating in the diamond industry to encourage long-term growth in the retail market and the sustainable increase of rough diamond demand. These objectives can only be attained if the industry meets the requirements and expectations of consumers. This Policy Statement is intended to reflect these aims and the means by which they can be accomplished by industry participants while building on the best traditions of the industry."

10. Section 1 of the Policy Statement sets out the qualifications to be satisfied by Sightholders who have to satisfy the Sightholder Criteria and the Best Practice Principles. The Sightholder Criteria are 8 in number and include such matters as financial standing and ability, strong market position in relation to particular diamonds, distribution ability, marketing ability and technical manufacturing ability and general business reputation. In the same document as the Sightholder Criteria, other considerations known as the Sightholder Considerations are specified. They include availability of supplies, particularly of the kind or kinds sought by the Sightholder. Note A1 to the Sightholder Considerations includes the following:

"Availability of Supplies

…

In relation to supply decisions, DTC recognises that there will be occasions when applications for a particular type of box exceed availability. In these circumstances, the DTC will have regard to applicants' relative level of satisfaction of the Sightholder Criteria in determining to whom those boxes should be supplied."

11. Section 2 of the Policy Statement, headed "Supply", contains the following relevant provisions:

"2.2 In each calendar year, supplies to Sightholders will be made during two six month selling periods ('Selling Periods'), from January to June and July to December. Commencing with the first Selling Period in the first full calendar year after this Policy Statement comes into effect, the Sightholder will, when requested by DTC, inform DTC of the aggregate value and nature of goods it believes it may wish (but shall not be obliged) to purchase during the subsequent Selling Period. At the beginning of each Selling Period, DTC will provide the Sightholder with an intention to offer (an 'ITO') indicating the aggregate level and nature of goods it intends to make available for inspection by (but shall not be obliged to supply to) the Sightholder during the subsequent Selling Period.

…

2.4 Subject to current market conditions, DTC will use reasonable endeavours to meet applications for those boxes placed by Sightholders at each Sight taking into account (a) the requirements of DTC's other Sightholders, (b) the Sightholder's level of satisfaction of the Sightholder Criteria as compared with that achieved by other Sightholders, (c) the Sightholder Considerations, and (d) the aggregate value and nature of goods requested by the Sightholder and indicated by the DTC as those it intends to make available for inspection, pursuant to paragraph 2.2 above."

12. Paragraph 11.1 of the Policy Statement provides that Sightholders will be appointed for a period of 24 calendar months and that each of DTC and the Sightholder will be entitled to terminate the appointment with effect from the expiry of the 24 month period by giving not less than 6 months' prior written notice. If such notice is not given, the Sightholder's appointment will continue for further 24 month periods terminable in the same way.
13. Paragraph 14.4 of the Policy Statement is an entire agreement clause. Paragraph 15 provides that the Policy Statement and any arrangements entered into under it are to be governed and construed exclusively in accordance with English law. Further provisions of the paragraph relate to any dispute. Litigation in English proceedings is one method contemplated for resolution of the dispute. Another is by reference to the SOC Ombudsman if the dispute falls within the scope of the Ombudsman Terms of Reference.
14. The judge recorded that there was no discernible disagreement between the parties about the way in which the SOC machinery for allocating diamonds to Sightholders has operated since its introduction in 2003. The main features of the system were explained by DTC to Sightholders and their brokers both in writing and at presentations in the run-up to the introduction of SOC, and access to DTC's computer programmes relating to its Allocations Model was given in the course of these proceedings. The judge provided this overview of the SOC Allocations Model (in paragraph 31 of the judgment):

"At the simplest level, applications by Sightholders at the ITO stage for each supply period are scaled down by reference to various criteria, including in particular the historic purchases which the Sightholders have made from DTC in the previous ten Sights. However, the process as a whole is much more complex and sophisticated than that brief description might suggest. It depends on the detailed interaction of three basic principles which DTC likes to refer to as the 'three As', namely the application made by each Sightholder, the assessment of each Sightholder's ranking made by reference to the Sightholder Criteria, and DTC's forecast availability of goods."

15. The judge then proceeded to set out the 16 different steps involved in the application of the SOC Allocations Model. I shall do no more than indicate in outline the more important features of those steps, but I would add the comment that the various steps taken together can broadly be said to be aimed at meeting the objectives set out in the introduction to the Policy Statement.
16. In Step 1, before each Selling Period, the applicants submit ITO applications in which they indicate the band of diamonds (viz groups of boxes) which they wish to buy in the period. Once the applications have been received, the applicants for each band are ranked according to their score under DTC's assessment of the Sightholder Criteria. As the judge noted, the ranking of Jayam has at all times been relatively low in the third quartile of Sightholders.
17. Steps 2, 10 and 11 require mandatory supplies to certain diamond-producing countries.
18. Step 3 involves what is called "applications capping", but which often operates by increasing a Sightholder's application. As the judge explained (paragraph 39):

"The … applications are now capped (or, if they fall below certain levels, uplifted) by reference to the extent by which they exceed (or, as the case may be, fall short of) prescribed thresholds."

The thresholds for four categories of Sightholders are set out by reference to the average amount of purchases from DTC in the last ten Sights, the threshold being that average amount increased as specified in that step. The judge said (in paragraph 40):

"The purpose of the step, viewed in isolation, is not merely to place an upper limit on applications, but also to introduce into the system a built-in allowance for what DTC considers to be reasonable expectations of future growth by Sightholders on an across the board basis. Thus (ignoring for the moment the smallest category) Sightholders may in effect apply in each selling period at a level which exceeds their average purchases in the last ten Sights by a factor of between 100% and 50% depending on their size; and if, for whatever reason, they do not apply at or above this increased level, they are nevertheless treated at this stage in the process as if they had applied at the increased level. A further reason for the uplift in the case of the smallest category, that is to say applicants with average earlier purchases of less than $1 million per Sight, is that it is designed to allow small Sightholders to apply in sufficient volume to meet minimum band application thresholds. It is DTC's experience that applications of less than $2 million per Sight run a significant risk (after subsequent scaling down) of falling below the minimum supply threshold of $0.8 million per Sight …."

19. Step 4 involves the capping of surplus applications by band. Where a band is more than 100% oversubscribed, applications for the band from Sightholders are eliminated in reverse order of ranking (ie the lowest ranking applicant is eliminated first) until the 100% over-subscribed level is reached.
20. In Step 5 all remaining applications are reduced on a pro rata basis until they match the quantity which DTC has determined as the sales target for that band. This ensures that demand, as modified by the earlier steps, will not exceed supply.
21. In the remaining steps, further adjustments are made and checks applied. Thus if a working allocation to a Sightholder falls below the minimum threshold for supplies for a band, that allocation is removed and reallocated to the remaining Sightholders pro rata. The purpose of this is to ensure that supplies are made at a level which DTC regards as economically efficient and to discourage Sightholders from making unfocussed applications across a wide range of bands. Similarly where a working allocation to a Sightholder is below $0.8 million per Sight (or $4 million per ITO for a supply period of 5 Sights), that allocation is removed and reallocated to the other applicants in the relevant bands.
22. The ITO allocations are communicated to Sightholders and, subject only to a category known as ex-plan sales, they form the basis of the allocations made to Sightholders at the Sights in the ensuing selling period. Ex-plan sales are sales made by DTC in addition to the ITO sales plans. They arise because the overall value of the ITO is slightly below DTC's sales target for that selling period, that target being based on forecast availabilities rather than goods actually on hand for sale. A safety margin of 6-7% is built into the ITO sales targets to allow for forecasting fluctuations in case of unexpected supply challenges at the mining level. To achieve the full sales targets ex-plan sales are made by goods being allocated for offer to Sightholders automatically by the same system that calculates the Sight allocations which fulfil Sightholders' ITO sales plans.
23. I turn now to what was found by the judge to be the factual background to what he called the critical questions whether the introduction and operation of the SOC capping system from H2/2003 onwards involved any breach of DTC's contractual obligations under paragraph 2.4 of the Policy Statement and, if so, whether such breaches caused Jayam to enter and become locked into the SOC model at too low a level.
24. Between 1986 and 1989 Jayam's purchases from DTC exceeded $200 million per annum, reaching a peak of $280 million in 1988, but thereafter declined. In 1997 and 1998, at a time when Mr Mehta's wife was ill and Mr Mehta was distracted from work, Jayam's purchases fell to $79 million and $34 million respectively. When Mrs Mehta recovered, Jayam's purchases in 1999 increased to $72 million and in 2000 to $80 million. The 2000 total compared with applications Jayam made totalling $132 million. In 2001 Jayam's purchases dropped by 37% to only $50 million compared with its applications totalling $110 million (representing a reduction in applications by Jayam of 17%). As the judge pointed out in paragraph 74 of his judgment, in a market where demand habitually exceeded supply and at a time when DTC based its allocations decisions on its own assessment of commercial considerations, Sightholders were inevitably at the mercy of DTC for what they actually received.
25. Between 2001 and 2003 DTC gradually moved from its previously discretionary system of allocations to the SOC arrangements with their more objective criteria for allocations to Sightholders. In October 2001 Sightholders were asked for the first time to submit proforma applications for the supply period H1/2002. The sales target fixed by DTC was $400 million per Sight, but applications from Sightholders amounted to $1,044 million, reflecting to a considerable extent tactical over-application. Concern over such over-application (or "gaming") remained a constant problem for DTC thereafter. One mechanism explored within DTC was a capping mechanism, in which account was taken of the Sightholder's recent purchases both from DTC and from all sources. However, the allocations actually made to Sightholders at the first Sights in H1/2002 continued to be determined by DTC as before in accordance with commercial considerations. Jayam applied for a total of $45 million and achieved purchases of $28 million, an improved success rate of 62%.
26. For H2/2002 DTC was required by the European Commission to stop its previous practice of obtaining information from Sightholders about purchases from sources other than DTC and it revised the prototype capping mechanism so that it operated by reference to previous purchases from DTC only. Because of exceptional market conditions at the time and a heavy over-application in H2/2002 DTC decided to base the capping threshold for H2/2002 on previous purchases from DTC in only the first four Sights of 2002. DTC's forecast sales target for H2/2002 was 25% lower than the average sales in the first four Sights of 2002 and DTC assumed that the average Sightholder would buy 25% less from DTC at each Sight in H2/2002 than it had done in Sights 1 to 4 of that year. The final decisions were again taken as before in accordance with commercial considerations, the proforma sales plan produced by the prototype model in H2/2002 being treated as non-binding templates for internal guidance only. Also in H2/2002 the ITO system was introduced. Jayam applied for goods totalling $55.5 million but received an ITO allocation of $22.25 million. Jayam then applied at the five Sights in H2/2002 for goods totalling $33.875 million and received allocations of $22.736 million, a success rate of only 41.4%. Jayam was upset at the low level of allocations, the more so because it had expanded a diamond polishing factory in Thailand, as it informed DTC. It told DTC in May 2002 that its production would increase at least 40% by the end of September 2002.
27. For H1/2003 the prototype model was further refined and bore a much closer resemblance to the SOC model adopted from H2/2003 onwards. Sightholders had to apply for particular bands rather than boxes, and the capping threshold was set by reference to the ten Sights of 2002. Jayam's application was left uncapped. The final allocations were again ultimately governed by commercial considerations. Jayam originally applied for goods totalling $49.75 million and received an ITO allocation of $25.75 million. It then made Sight applications of $43.75 million and received Sight allocations of $28.36 million. Its actual purchases (including ex-plan) totalled $29.104 million, a success rate of 58.5%.
28. H2/2003 was the selling period in which SOC began. Capping was based on purchases over the last 10 Sights, which for H2/2003 meant Sights 5 to 10 of 2002 and Sights 1 to 4 of 2003. Those Sights therefore included H2/2002 in which the cap had been based on purchases in the first four Sights in H1/2002. Jayam applied for $7.75 million per Sight. Its ITO was $5.23 million per Sight. It made Sight applications totalling $44.529 million and received Sight allocations of $26.649 million. Its actual purchases were $26.142 million, a success rate of 58.7%.
29. Jayam was unhappy with the allocations made to it and in September 2004 it made a complaint to the SOC Ombudsman, Mr Dermot Gleeson SC, a former Attorney General of the Republic of Ireland. Jayam alleged that DTC had followed improper procedures in respect of the diamonds which it supplied Jayam in the first six month ITO period following the implementation of SOC, viz H2/2003. The nub of Jayam's complaint was directed against the application capping mechanism in step 3 and was essentially the same as the complaint in the present proceedings. The Ombudsman dealt with the complaint on paper. In a Final Determination dated 9 March 2005 he upheld Jayam's complaint and required DTC to reconsider the allocation. He summarised his reasoning in paragraph 7.4 of his Final Determination. First, the SOC documentation makes no reference to the application capping mechanism. Second, the mechanism "operates so as to introduce additional criteria supplementary to, and potentially conflicting with, displacing or cutting across (in the sense of reducing their significance) the four criteria listed in paragraph 2.4 of the SOC Policy Statement." Third, the Policy Statement is part of the matrix of contractual provisions applying between Sightholders and the DTC; the language of paragraph 2.4 is clearly intended to specify the criteria by which the DTC will make supply decisions. Fourth, that conclusion was reinforced by the entire agreement clause. Fifth, discussions between DTC and the European Commission were immaterial. DTC took the view that it was unable to alter its allocation decisions retrospectively and informed Jayam that all it could do was to clarify the position for the future. A new paragraph 2.5 was inserted into the Policy Statement:

"For the avoidance of doubt, in taking into account the factors listed at paragraph 2.4 above, DTC will employ such mechanisms, calculations, methodology and/or workings which are reasonably required from time to time to give effect to the principles set out in this Policy Statement (and the documentation referred to herein). DTC will employ such mechanisms, calculations, methodology and/or workings in a fair, objective and consistent manner to all Sightholders."

The Proceedings

30. These proceedings were commenced on 4 October 2005. Jayam claimed damages from DTC for breach of contract in respect of the supply of diamonds made by DTC to it in the supply periods commencing H2/2003. By its Amended Claim Form it also claimed a declaration that on a proper construction of the SOC documentation DTC is not entitled to impose or use the application capping mechanism in determining allocations of diamonds to Jayam.
31. Jayam pleaded in paragraph 16 of the Reamended Particulars of Claim that in applying the application capping mechanism (the capping formula based on actual purchases over the previous 10 Sights), DTC acted in breach of contract and in breach of the terms of the SOC documentation. In the course of the hearing before the judge, on 25 January 2007 particulars were given of the alleged breach. They were, as summarised by the judge in paragraph 11 of the judgment, that (a) on its true construction, paragraph 2.4 of the Policy Statement stipulates the exclusive criteria that DTC is entitled to take into account in meeting applications placed by Jayam and other Sightholders; (b) in breach of paragraph 2.4 DTC failed to use reasonable endeavours to meet applications by Jayam and/or other Sightholders taking into account only the criteria specified in paragraph 2.4, and (c) in breach of paragraph 2.4 DTC took into account a further criterion not stated in that paragraph, namely the application capping mechanism.
32. Jayam further pleaded in paragraph 5 of its particulars of breach that the application capping mechanism is not a method to take into account the criteria in sub-paragraphs (d), (c) or (b) of paragraph 2.4 of the Policy Statement, and in paragraph 6 of its particulars of breach that it is not a method to take into account the criterion in sub-paragraph (a) of paragraph 2.4. The pleading continued:

"On the contrary, even if … it is the aim of the Application Capping Mechanism to ascertain 'genuine' (or non-gamed) requirements, for the reasons set out below it is not a fair, reasonable, non-discriminatory or non-arbitrary method to achieve that aim."

The reasons given were because (a) a cap based on prior purchases over a period of only 10 Sights is based on too short a period and is too inflexible to take account of any specific circumstances that may affect the size of a Sightholder's genuine requirements; (b) the effect of the imposition of the cap is that the Sightholder is locked in by reference to the cap imposed at the commencement of SOC; and (c) the cap imposed at the commencement of SOC was itself (i) based on subjective and discretionary decisions on allocations by DTC in H2/2002 and H1/2003, and also (ii) materially based on the use of capping by reference to a period of less than ten Sights, particularly in H2/2002.

33. DTC in its response to the particulars of breach averred:

"4. The Applications Capping Mechanism is an essential component of the Allocations Model under SoC and is a necessary and appropriate mechanism for the purposes of giving effect to the broad principles set out in paragraph 2.4 of the Policy Statement.

5. In particular, the Applications Capping Mechanism is necessary in order to assist in addressing the very real difficulty imposed by the fact that the demand for the DTC's rough diamonds substantially exceeds the available supplies. Sightholders have traditionally made applications for many more rough diamonds than the DTC has available to supply. In order to make decisions as to which applications can be met and which must regrettably be refused, the DTC must employ some means of determining first which of the applications made by Sightholders reflect genuine requirements and which applications are tactical or represent attempts to 'game' the Allocations Model.

6. The Applications Capping Mechanism is a fair, reasonable, non-discriminatory and non-arbitrary method for determining the requirements of Sightholders (as required by paragraph 2.4(a) of the Policy Statement)."

34. The hearing before the judge lasted 7 days. He heard evidence both from witnesses of fact and from experts. There was common ground that DTC can only take into account the four specified criteria in paragraph 2.4 of the Policy Statement in devising a system to meet applications by Sightholders. The dispute centred on whether the application capping mechanism went outside the proper scope of the four criteria and instead applied some other criterion or criteria. The judge pointed to the fact that demand for rough diamonds exceeded supply and that some machinery for matching supply and demand had to be found. He noted a concession made by Jayam in the course of the hearing that Jayam's objection was no longer to the existence or application of any cap at all, but only to the particular cap that DTC had chosen to adopt. The judge commented (in paragraph 104):

"This concession was in my judgment clearly correct, because in a situation where demand exceeds supply it must at the very lowest be reasonable for DTC to take account of specified criteria (a) and (d) by imposing an application capping mechanism. The fact that there is no express reference to such a mechanism in paragraph 2.4 is immaterial, given the latitude which it affords to DTC on how best to meet applications."

35. The judge then turned to the question whether the actual application capping mechanism employed by DTC from H2/2003 onward breached the requirements of paragraph 2.4. Leaving aside the question of the initial entry into the system, he considered that this question had to be answered in the negative, describing the application capping mechanism as "in principle a perfectly reasonable method of matching supply with demand for DTC to have adopted" (paragraph 105 of the judgment). The judge then gave a number of reasons for that conclusion. First, the use of a capping threshold based on historic allocations in the previous ten Sights, plus an uplift of between 50% and 100% (and more for the smallest applications) to allow for growth could not be stigmatised as unduly unreasonable or unfair. Special circumstances aside, a period of one year should be long enough to allow for seasonal variations in requirements and allocations while a period of more than a year would run the risk of focussing too much on past performance rather than providing a realistic indication of the potential for further growth which the system was designed to encourage. Second, at steps four and six in the process preference was given to higher-ranked applications over lower-ranked ones, the rankings themselves being determined by reference to DTC's assessment of applicants' scores against Sightholder Criteria. The judge described this feature of the system as unexceptionable because the Sightholder Criteria and the rankings were clearly intended to promote and reward the qualities that DTC considered most likely to achieve the objective of long term consumer growth. Third, there was no suggestion, still less evidence, to establish that the application capping mechanism operated in a way which was irrational or unfair or arbitrary or in breach of anti-discrimination law as between different Sightholders, provided that their initial entry into the system was at the "right" level. Further, it could not be said that Sightholders entering at the "right" level were then locked into the system without any reasonable opportunity of improving their level of purchases. The judge described the SOC system as having a good deal of flexibility built into it.
36. The judge next considered whether it could be said that Jayam entered the SOC system at too low a level in H2/2003 or whether the level at which Jayam entered the system was attributable to a breach by DTC of its obligations under paragraph 2.4. He answered that question in the negative. The judge referred to two main factors relied on by Jayam.
37. The first was the use by DTC of prototype capping mechanisms in the three selling periods preceding the introduction of SOC, which were in two cases based on purchases in the previous ten Sights and in one case (H2/2002) based on the purchases in a period of only four Sights. The judge pointed out (in paragraph 114 of his judgment) that the output of the prototypes was never determinative of allocations which were always ultimately made by senior executives of DTC in accordance with their assessment of market considerations. The output from the prototype models was a factor taken into account but was not binding on the executives. Further, in H2/2002 Jayam's applications were not capped and accordingly Jayam did not suffer in that selling period as a result of its previous purchasing history. In H2/2002 Jayam's application was capped at the ITO stage by reference to average purchases during a period of only four Sights, but none of those four Sights had been subject to capping. In H1/2003 Jayam's applications were not capped at all. The theoretical possibility that Jayam might have been prejudiced by the capping in H2/2002 by reference to a period of less than a year was negated by evidence from DTC's witness of fact, Mr Davies, that if the ITO applications of Jayam and other Sightholders for H2/2002 had been capped by reference to average purchases in the previous ten Sights, Jayam would have ended up significantly worse off.
38. The second factor relied on by Jayam was the use by DTC as a yardstick for H2/2003 of purchases made by Sightholders at a time when DTC's allocations were ultimately made at its discretion in a way which was opaque, unpredictable, unaccountable and potentially open to abuse. The judge said (in paragraph 116 of the judgment) that such a procedure could not be criticised as unreasonable, because the obvious starting point for the new system was the status quo if only to avoid disruption and to provide Sightholders with a reasonable expectation of continuity in their supplies.
39. The judge concluded that Jayam failed to make good its case on liability.
40. The judge in paragraphs 119ff considered the Ombudsman's decision and briefly explained why he disagreed with the Ombudsman's conclusion. On the Ombudsman's point that the SOC documentation did not refer to the application capping mechanism, the judge commented that the relevant question was not whether paragraph 2.4 referred to it but rather whether that was a reasonable method for DTC to have adopted in order to meet applications from Sightholders taking into account the four specified criteria. He said that the Ombudsman recognised this when he went on to consider whether the application capping mechanism could be justified by reference to any of the four criteria. The Ombudsman concluded that it could not, but the judge disagreed. The judge said in paragraphs 123 and 124:

"123. In circumstances where demand exceeds supply, there has to be some machinery for cutting down applications, and prima facie any reasonable machinery for that purpose will satisfy criterion (a). The choice between possible methods of achieving that end is for DTC alone, and as I have already pointed out paragraph 2.4 affords DTC a good deal of latitude. I agree with Mr Christopher Carr QC [counsel for DTC] that the limits on what DTC may choose to do are set by considerations of reasonableness. As Mr Carr submits, what DTC has essentially done is to use each Sightholder's purchases from DTC during the last 12 months as a proxy for its (genuine) requirements in the forthcoming selling period. This approach has the benefit of addressing the problem of gaming, or tactical over-application. It may also be said to have the possible disadvantage of curbing applications by a Sightholder whose genuine requirements in the forthcoming selling period are higher than they were in the last 12 months; but this possible prejudicial effect is mitigated by the growth allowance built into the application capping mechanism and also by the incentives at steps four and six of the model for higher-ranked Sightholders. It cannot be said, in my judgment, that this possibility of prejudice comes anywhere near making the whole system inherently unreasonable. I return to the point that, in the absence of more closely prescriptive language, DTC has a wide discretion how to proceed.

124. In short, I agree with the way Mr Carr QC puts it in paragraph 80 of his skeleton argument where he says that:

'The Ombudsman's central error was to assume (without analysis) that, because the application capping mechanism was intended to eliminate tactical application and gaming it could not have been adopted for the purpose of taking into account the requirements of other Sightholders.' "

41. In view of his conclusion on liability, the judge did not deal with the question of quantum.

The appeal

42. Lawrence Collins LJ, in considering on paper the application for permission to appeal to this court, noted that there were points on construction raised in the notice of appeal and gave permission limited to those points, which he said were arguable. However, he held that there was no realistic prospect of success on the other points raised in the notice of appeal. He referred to the judge as having decided that the capping mechanism was a reasonable method of matching supply with demand and that it was not inherently unreasonable and that it did not operate in a way which was irrational or unfair or arbitrary. The Lord Justice held that the judge's decisions on reasonableness were evaluations of the kind with which the Court of Appeal would not interfere in this case. He said that if the parties could not agree on the consequences and on what issues would be excluded from the hearing, it would be necessary for the management of the appeal to be dealt with at a separate hearing, preferably before him.
43. Jayam's bold response to Lawrence Collins LJ's order giving limited permission to appeal was not to reduce its skeleton argument and grounds of appeal on the basis of which it had sought permission but to add a new paragraph 38 to that skeleton argument. It also sought DTC's agreement that the matters raised in its skeleton argument and grounds of appeal either raised issues of construction or were matters arising out of a finding in Jayam's favour on appeal on construction. DTC did not agree. On reference of the dispute to the Lord Justice, he urged the parties to try to reach agreement, failing which he gave directions preparatory to a hearing. Leading counsel for both sides, Mr Charles Hollander QC for Jayam and Mr Laurence Rabinowitz QC for DTC, then had a discussion and agreed, in effect, to disagree. DTC agreed to respond to Jayam's skeleton as it then stood, reserving the right to argue that certain points fell outside the scope of the permission given. Jayam reserved the right to ask this court at the appeal hearing for leave to argue any such points.
44. It is in that state, which I feel bound to say I regard as less than satisfactory particularly in the light of Lawrence Collins LJ's directions for resolving any disagreement, that this appeal has come before us. However, it seems to me plain that the Lord Justice's order refused Jayam permission to argue anything other than a pure question of construction and that this precludes the taking of any point challenging the judge's conclusions on fairness and reasonableness. Moreover, Jayam itself in the new paragraph 38 of its skeleton argument recognised the severe constraints on what it could argue in this court. The paragraph reads:

"Jayam's case is that for the reasons set out above, the judge misconstrued Clause 2.4. If the Court of Appeal take the view that the judge's construction was correct, this appeal fails. The remainder of this skeleton deals with the consequences in the event the Court of Appeal accepts Jayam's criticism of the judge's approach to construction."

The sort of point which Jayam would wish to argue if it succeeds on construction is exemplified by the question it raised in the heading to paragraphs 39ff of its skeleton argument, viz "Can the cap imposed by DTC be justified as a fair and reasonable means of ascertaining the actual, non-gamed requirements of Sightholders." Among the matters on which Jayam would then seek to rely are some answers given by Mr Davies for DTC in cross-examination said to show that a Sightholder's requirements were not regarded by DTC as material to the cap, the taking into account by the judge (in paragraph 116) of a Sightholder's expectation on the introduction of SOC of continuity in the supplies to it and the lock-in effect of a cap system based on previous purchases. None of these points, in my judgment, goes to questions of construction.

45. In the first part of Jayam's skeleton argument ending with paragraph 37, Jayam took a number of points on construction designed to support the submission that the judge's approach was that it was sufficient merely that demand exceed supply and that in consequence of the reference to supply and demand in paragraph 2.4(d) it was open to DTC to impose any reasonable cap for the purpose of matching supply with demand. The obvious starting point was said to be that the SOC arrangements did not refer to the existence of a cap still less one depending on the value of recent purchases and the second of the points taken by the Ombudsman in paragraph 7.4 of his Final Determination (see paragraph 29 above) was repeated. The judge was said to have erred in conflating the need to reduce demand to match supply with the criteria by which that need was to be satisfied. Jayam acknowledged that sub-paragraph (a) of paragraph 2.4 could justify a cap and that DTC had relied on this sub-paragraph with sub-paragraph (d) in order contractually to justify the cap, but said that Jayam's objection was not to a cap but to this cap which was not a genuine or reasonable anti-gaming mechanism. Further, it introduced an additional criterion for allocations, viz that the cap was based upon the values of the Sightholder's purchases in the previous year.
46. Before us Mr Hollander launched a wide-ranging attack on the judge's judgment. He submitted that the judge went wrong by asking himself the wrong question in paragraphs 105ff. Instead of looking to see whether the application capping mechanism was a way of giving effect to one or more of the criteria at paragraph 2.4 the judge had detached the question of reasonableness and simply asked himself the question: is the application capping mechanism a reasonable method of matching supply with demand for DTC to have adopted? Mr Hollander pointed to sentences from the judgment where the judge had pronounced on reasonableness without making express reference to any of the criteria in paragraph 2.4. Thus in paragraph 105 the judge said:

"In my judgment the application capping mechanism … is in principle a perfectly reasonable method of matching supply with demand for DTC to have adopted."

Again in paragraph 106 the judge said:

"… the use of a capping threshold based on historic allocations in the previous ten Sights, plus an uplift of between 50% and 100% (and more for the smallest applicants) to allow for growth, cannot in my view be stigmatised as inherently unreasonable or unfair."

47. Mr Hollander acknowledged that the judge had addressed the right question in that part of his judgment where he discusses the Ombudsman's decision. However, he pointed out that this was after the judge had expressed his conclusion on liability.
48. Mr Hollander made other criticisms of the judge. They included that the judge in paragraphs 53, 108 and 112ff had considered a meaningless question whether Jayam's initial entry into the allocations system was at the "right" level, a question which he submitted was an issue related to damages, not liability.
49. Mr Rabinowitz protested that Mr Hollander in his oral submissions to this court had not been addressing the questions of construction to which Lawrence Collins LJ's order confined him but had concentrated on why the application capping mechanism was not a reasonable way in which to take the criteria in paragraph 2.4 into account, and that, he said, was why Mr Hollander had submitted that the judge had asked himself the wrong question on reasonableness.
50. I see force in Mr Rabinowitz's protest, but it is sufficient to deal with Mr Hollander's argument head on. Reading the judge's judgment as a whole and not isolating particular sentences or paragraphs, I find it quite impossible to accept that the judge lost sight of the question before him, viz whether it was permissible for DTC to take account of the criteria in paragraph 2.4 by adopting the application capping mechanism. The judge had noted in paragraphs 10ff the way Jayam had pleaded the alleged breach of contract and in paragraphs 15ff DTC's response. He therefore knew that the central pleaded issue was whether the application capping mechanism was a fair and reasonable method of determining the requirements of Sightholders as required by sub-paragraph (a) of paragraph 2.4. In paragraphs 92 and 93 the judge had noted DTC's acceptance of the proposition that it could only take into account the four specified criteria in devising a system to meet applications by Sightholders. In paragraphs 99 and 100 the judge again had noted DTC's acceptance that on the true construction of paragraph 2.4 DTC had to take criterion (a) into account in a way that was fair and reasonable and avoided discrimination and was not arbitrary. The judge was well aware that he also had to decide whether the application capping mechanism was in fact unfair, unreasonable, discriminatory or arbitrary in its operation. It was not necessary for the judge to distinguish between questions of construction and questions of reasonableness. That is only important for this court in the light of Lawrence Collins LJ's order.
51. It seems to me that the judge explained his decision impeccably. In paragraph 102 he noted the wording and content of paragraph 2.4, pointing out that the obligation of DTC to use "reasonable endeavours" to meet applications was a relatively low hurdle and that provided it made reasonable efforts to meet applications in a way that took account of and did not go outside the scope of the four specified criteria, it would not be in breach of its obligation. In paragraph 103 he noted the purpose which the Policy Statement was intended to promote and the need to construe its provisions to give effect to that purpose. In paragraph 104 he noted that demand exceeds supply, that some machinery for matching supply and demand had to be found and that one way of doing this was by placing a cap on applications. Having referred with approval to Mr Hollander's concession that Jayam no longer objected to the existence or application of any cap at all, the judge explained that approval as being:

"because in a situation where demand exceeds supply it must at its very lowest be reasonable for DTC to take account of criteria (a) and (d) by imposing an application capping mechanism."

52. Paragraph 105 commences with the question: "Can it then be said that the application capping mechanism actually employed by DTC from H2/2003 onward breaches the requirements of paragraph 2.4?" The judge's negative answer to that question and his comment that the mechanism was a reasonable method of matching supply with demand must be read in the light of that question with its express reference to the criteria in paragraph 2.4. In paragraph 106 the judge shows that he has the requirements of Sightholders, which form the subject of criterion (a), very much in mind. He says that a year would be long enough to allow for seasonal variations in requirements. He refers to the possibility that a Sightholder's "genuine requirements" may be significantly higher in the next selling period than they were in the preceding year. In paragraph 109 he considers what at first sight he accepted was a valid criticism that no allowance is made for exceptional circumstances reducing "a Sightholder's requirements in the preceding year to a level unrepresentative of its underlying business." In paragraph 115 the judge refers to a period of less than a year as running "the risk of being unrepresentative because of seasonal variations in a Sightholder's purchasing requirements." In paragraph 116 the judge says that clear and specific contractual language would have been needed to make clear that DTC was obliged to start SOC with "a fresh assessment of the requirements of Sightholders uninfluenced by their previous purchasing history."
53. I conclude from the paragraphs leading to the judge's conclusion that Jayam's case on liability failed that the judge never lost sight of criterion (a) throughout his discussion and that he cannot fairly be accused of asking himself the wrong question. However, in any event when one takes into account paragraphs 122-124, where the judge deals with the Ombudsman's views, as in my judgment one must, there cannot be any doubt on the matter. As Mr Hollander accepted, there the judge was expressly asking himself the right question.
54. I can now revert more briefly to the questions on construction on which Jayam was permitted to appeal. To my mind the judge was plainly right to hold that the absence of any express reference in paragraph 2.4 to the application capping mechanism does not preclude DTC from adopting it, provided that thereby account is taken of one or more of the very broad criteria in paragraph 2.4. That paragraph is wholly silent as to how account is to be taken of those criteria. That is left to DTC. Given the chronic problem of demand exceeding supply and the justified concern felt by DTC over gaming, there can in principle be no objection to a capping mechanism to take account of criteria (a) and (d). That cannot properly be said to be the introduction of a new and inconsistent criterion into paragraph 2.4. Jayam itself had accepted that DTC could incorporate some kind of application cap and was entitled to translate the criteria to create a more complex system. In my judgment, the judge did not simply proceed on the basis that demand exceeded supply. He rightly concluded that the application capping mechanism was part of the means by which DTC took into account criteria (a) and (d) and that it was a reasonable and fair way of taking account of those criteria.
55. As far as the other criticisms by Mr Hollander of the judge's judgment, I do not agree that the judge, in considering whether Jayam's entry into the SOC system was at the "right" level, was asking himself a meaningless question on an issue which went only to the question of damages. The judge was addressing Jayam's complaint that as a result of the operation of the application capping mechanism which relied on DTC's previous allocations Jayam entered the SOC system at an inappropriately depressed level, what Mr Carr for DTC in his closing submissions described as "the wrong rung of the ladder." If DTC could not reasonably use previous purchases as indicative of Jayam's requirements at the start of the SOC system, that went to liability as well as to damages. But in any event that seems to me not so much a question of construction as one of reasonableness. Nor do any of Mr Hollander's other points show that as a matter of construction of the SOC arrangements the judge erred in his conclusions.

Conclusion

56. For these reasons, which owe much to Mr Rabinowitz's admirable arguments, I would dismiss this appeal.

Wilson LJ:

57. I agree.

Keene LJ:

58. I also agree.