British and Irish Legal Information Institute

Freely Available British and Irish Public Legal Information

[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

United Kingdom VAT & Duties Tribunals Decisions

You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> Captial One Developments Ltd v Customs and Excise [2004] UKVAT V18642 (09 June 2004)
URL: http://www.bailii.org/uk/cases/UKVAT/2004/V18642.html
Cite as: [2004] UKVAT V18642

[New search] [Printable RTF version] [Help]

Captial One Developments Ltd v Customs and Excise [2004] UK V18642 (09 June 2004)

18642

VALUE ADDED TAX — avoidance — construction of premises for probable occupation by partially-exempt banking group — series of transactions designed wholly or partly to reduce liability for VAT — assumption that tribunal's decision in Halifax plc correct — whether transactions artificial — whether tax avoidance sole purpose

VALUE ADDED TAX — Sixth Directive arts 17, 19, 20 — VAT Regulations 1995, reg 101 — residual input tax — intention to make future taxable and exempt supplies — fixity of intention — whether necessary for application of reg 101(2)(d) — effect of change of intention — whether absence of demonstrable intention to make identifiable taxable supply renders reg 101(2)(d) inapplicable

VALUE ADDED TAX — building — meaning — VATA Sch 9 Group 1 — whether new construction attached to existing structure by link bridge with shared services and used for same purposes to be regarded as a "building" in its own right or merely an extension of the original building — whether Sch 8 Group 5 Note (16) concepts relevant — Sixth Directive art 4(3)

MANCHESTER TRIBUNAL CENTRE

CAPITAL ONE DEVELOPMENTS LIMITED

Appellant

- and -

THE COMMISSIONERS OF CUSTOMS AND EXCISE

Respondents

Tribunal: Colin Bishopp (Chairman)

Cranstoun Gill

Sitting in public in Manchester on 26 to 29 January and 17 to 20 February 2004

Paul Lasok QC and Michael Patchett-Joyce, instructed by PricewaterhouseCoopers, for the appellant

Jonathan Peacock QC, instructed by the Solicitor's Office of HM Customs and Excise, for the respondents

© CROWN COPYRIGHT 2004

DECISION

Introduction

This is an appeal against the decision of the respondent Commissioners, set out in a letter of 28 June 2001 (the "decision letter"), to reject the appellant's claim for input tax credit of just under £8 million, made in its return for the accounting period 04/00. The decision letter expounds in some detail the Commissioners' perception of a series of transactions entered into, or expected to be entered into, by the appellant and other companies within its trading group, some with each other and some with companies at arm's length. All of the transactions relate, directly or indirectly, to the acquisition of a site and the construction on it of new offices which were thereafter leased to and occupied by a company within the trading group whose activities are wholly or mainly exempt from VAT.

The decision letter in fact sets out five discrete decisions, although three are, in substance, the same decision expressed in different ways. That is, shortly put, that some of the transactions are to be disregarded when determining the incidence of VAT because their purposes are not commercial or economic, in the normal sense of those terms, but tax avoidance—essentially, the Commissioners' contention is that they constitute together a scheme whose purpose is solely the recovery of input tax which ordinarily would not be recoverable, since the end user carries on an exempt or largely exempt business. Another of the decisions sets out what the Commissioners perceive to be the correct reconstruction of the series of transactions, shorn of those which are to be disregarded. The remaining decision is of a different character, and is that the construction which is the subject of the transactions is not a building in its own right, but an extension of an existing building. The status of the construction is of critical importance to the validity of the appellant's claim.

It is accepted by the appellant that the transactions were planned in a way designed to achieve a tax saving and, to that extent, that they constitute a scheme. However, it challenges the five decisions by contending that the transactions which the Commissioners impugn were not designed solely for the purpose of avoiding tax but that each had additionally a proper commercial purpose; that the Commissioners have misconstrued the law as it is to be applied to the transactions; that a reconstruction such as the Commissioners propose is not permissible; and that it is not possible to disregard the economic effects which, even on the Commissioners' own approach to the case, the transactions have. They dispute too the Commissioners' view that the construction does not constitute a building.

We heard the appeal at a time when the law on the subject—that is, where (if it exists at all) the boundary between acceptable tax planning and unacceptable tax avoidance lies—was, as it still is, in a state of some uncertainty. Putting the same point another way, there is no clearly identifiable distinction between those transactions which do, and those which do not, constitute supplies for the purposes (or are within the "spirit") of the Sixth VAT Directive (77/388/EEC) and domestic implementing legislation.

In Halifax plc and others v Commissioners of Customs and Excise [2001] V & DR 73 (we shall refer to this and the tribunal's later decision in the same case, together, as "Halifax") the tribunal concluded that where some of the transactions within a series had been entered into, as it found, for the purpose of tax avoidance and lacked any genuine business or commercial purpose beyond tax avoidance the series was to be reconstructed (essentially by elimination of the "offensive" transactions) in order to reflect its true economic purpose, and the incidence of VAT determined by reference to the reconstructed series; see also Customs and Excise Commissioners v Reed Personnel Services Ltd [1995] STC 588. The first of the tribunal's decisions was quashed in the High Court—see Halifax plc and others v Customs and Excise Commissioners [2002] STC 402—because Neuberger J concluded that the tribunal had not asked itself all of the requisite questions. The case was remitted to the tribunal for further consideration. The tribunal thereafter made further findings, though it might perhaps be more accurate to say that it restated its original findings more fully, and decided to refer two questions to the European Court of Justice: see [2002] V & DR 117.

It is not necessary to examine the referred questions in detail for the purposes of this decision, but we do need to set them out, so far as they are of general application rather than specific to the circumstances of that individual case, in order to put the issues which arise in this appeal in their context. In Halifax, the tribunal determined that the supplies made by arm's length builders, of construction services, were to be treated as made, not to either one of two subsidiaries of Halifax plc, but to Halifax plc itself (like the appellant's associated company in this case, a company carrying on largely exempt business) on the ground that the transactions entered into by the subsidiaries had no purpose other than the obtaining of a tax advantage and were for that reason to be disregarded.

The first of the referred questions is in two parts. By the first part, the tribunal sought the Court's guidance on the narrow point, whether transactions effected "with the intention solely of obtaining a tax advantage and which have no independent business purpose qualify for VAT purposes as supplies made by or to the participators in the course of their economic activities" and, by the second part, and evidently assuming that the Court would answer the first in much the same way as the tribunal itself had done, "what factors should be considered in determining the identity of the recipients of the supplies made by the arm's length builders?". The second referred question was rather broader in scope: "Does the doctrine of abuse of rights as developed by the Court operate to disallow the appellants their claims of recovery of or relief for input tax arising from the implementation of the relevant transactions?". Further questions of a similar character, though expanding on those in Halifax, were referred by Park J in BUPA v Commissioners of Customs and Excise and have, we understand, been joined with the Halifax reference (Case C-255/02).

We were asked by the parties to determine (among other things) whether, on the assumption that the tribunal's decision in Halifax is correct (although the appellant conceded as much only for the purposes of this hearing), the scheme in which the appellant was a participant offends against the principles established in that case. Since the reference made by the tribunal to the European Court had not been determined when we heard the appeal, it was necessary for us not merely to assume that the tribunal was right in Halifax, but to take into account the possibility that the Court of Justice would approach the facts of the case in a different way and also to anticipate, in our own findings of fact, the various answers the Court might give to the referred questions. The Commissioners additionally argue that the scheme fails in its purpose for other reasons, independent of those adumbrated in Halifax, and we shall deal with those reasons in due course.

In outline the scheme we must consider is similar to that employed in Halifax: the transactions relate to the construction of new office premises to be, or expected to be, occupied wholly or partly by a company whose principal business activity is the exempt supply of financial services. However, although its purpose is similar, the appellant's scheme differs considerably in its details from that employed in Halifax. While, as we accept from the evidence we heard, there was at the outset a possibility that some or all of the new accommodation might be occupied by a company making taxable supplies (and might instead be sold or let outside the group), the scheme was quite clearly designed upon the assumption that the principal occupier would be an exempt, or partially exempt, trader (and would not have been entered into otherwise), and that turned out to be the case. The companies within the VAT group of which it is a member make, between them, some taxable supplies and the VAT group is able to recover some, but not all, of the input tax its members incur. We understood that the recovery rate varied considerably from one accounting period to another, ranging from 1% to 45%. The admitted purpose of the scheme we must consider was to ensure that the entirety of the input tax incurred in the acquisition of the site and construction of the premises would be recovered, and not clawed back.

The Commissioners maintain that the similarities between Halifax and this case are such that they should be treated in a similar fashion. The appellant, while acknowledging that there are superficial similarities, contends that Halifax is to be distinguished on the detailed facts; but even if that is not right the principles established in that case are incapable of application.

Paul Lasok QC (leading Michael Patchett-Joyce), for the appellant, argued that we should focus primarily on the third of the five decisions, to which everything else was subsidiary. This is the decision—it is reproduced in full at paragraph 131—which sets out the Commissioners' proposed reconstruction of the transactions so as to reflect what they perceive to be their true economic character. Dr Lasok's point was that, if the reconstructed series of transactions proposed in that decision would not have occurred in any event, as he contended, for valid business reasons quite apart from any tax avoidance motive, the decision could not stand and the refusal to repay the input tax claimed by the appellant, based on that decision, must fall with it; the appeal must be allowed and the Commissioners must either pay the input tax claimed, or reconsider and issue another decision.

While we agree that (if Halifax is correct) reconstruction of the transactions must follow if any are to be eliminated as offensive, it seems to us impossible to take the Commissioners' proposed reconstruction as the starting point, still less to regard it as the only issue of significance. It was not suggested in Halifax that transactions which, independently of any tax avoidance purpose, have economic substance should be revised in some way. Such a suggestion is, in any event, inconsistent with the conclusions reached in cases such as WHA Ltd v Customs and Excise Commissioners [2003] STC 648 (and upheld on this issue in the very recent Court of Appeal decision, at EWCA Civ 559) and RBS Property Developments Ltd v Customs and Excise Commissioners (2002, Decision 17789) (where a distinction was drawn between transactions which have the sole purpose of tax avoidance, and those where that purpose is no more than predominant). It is, of course, conceivable that the Court of Justice will take a contrary view and will conclude that related transactions which have a purpose other than tax avoidance are tainted and for that reason can be revised, but we do not propose to speculate in that way.

In our view it is not possible, or desirable, to deal with this single issue in isolation. It is dependent on the interpretation we place on the evidence as a whole, and we do not think it helpful to attempt to deal with the evidence relevant to Dr Lasok's argument separately. Much of the evidence related to more than one issue. Moreover, the parties required us to make as many findings of fact as possible; and even if we were persuaded to allow the appeal on this single ground the Commissioners' further reconsideration which would be the result would be dictated, to some extent, by our findings. We intend therefore to return to Dr Lasok's argument when we have considered all the evidence.

In the interests of brevity we will need to use, throughout this decision, a number of abbreviations and acronyms, which we set out for ease of reference in the following table, in which we have also noted the paragraphs of this decision in which the terms are more fully explained:

Abbreviation or acronym	Full name	See para
COB	Capital One Bank	18
COBEP	Capital One Bank (Europe) plc	18
COCI	Capital One Communications Inc	27
COCL	Capital One Contracts Limited	30
CODL	Capital One Developments Limited (the appellant)	33
COFC	Capital One Finance Corporation Inc	16
COPI	Capital One Properties Inc	47
COSI	Capital One Services Inc	20
Prologis	Prologis Kingspark Developments Limited	29
REOB	Real Estate Oversight Board	54
Woolf	Woolf Limited	37
Worldscape	Worldscape Inc	47

The Capital One group

The appellant is one of many members of a large corporate group whose structure requires some explanation. When we refer to "Capital One" we mean the group, or alternatively one or more of the companies within the group, in a context in which it is unnecessary or impossible to identify a single company or to discriminate between the individual companies. What follows in this section of our decision was, for the most part, common ground, or unchallenged.

The ultimate holding company of the entire group is Capital One Finance Corporation Inc ("COFC"). It is an American corporation, established in Delaware. As its name suggests, its principal activities lie in the financial services sector which, in the United States as in the United Kingdom, is closely regulated by bodies charged with that task; in Capital One's case, regulation of its US activities and those of its overseas activities performed by US corporations and their subsidiaries is undertaken at national level (by the Federal Reserve) and for some purposes at state level. We deal with the evidence we heard about regulation and its significance in this appeal later in our decision.

COFC's policy and practice is that its group's activities, or most of them, are undertaken by wholly-owned subsidiaries. Some of those subsidiaries have been established to carry on business indefinitely while others, regarded as special purpose vehicles ("SPVs"), are intended to perform only a specific function, and then, we understood, are allowed to become dormant unless and until they are revived to undertake a similar function. Some of the companies within the group are immediate subsidiaries of COFC, while others are subsidiaries of subsidiaries, occasionally several "generations" below COFC. Most are wholly-owned subsidiaries of their parents, while the shares of others are held by more than one other company. Ultimately, however, the beneficial ownership resides entirely in COFC in every case.

One of COFC's immediate and major subsidiaries is Capital One Bank ("COB"), which is incorporated in Virginia. At the beginning of the period with which we are concerned, COB carried on business in the USA and in the UK, in the latter case by means of a designated UK branch. In September 2000, for reasons quite unconnected with the tax saving scheme, the UK branch's business was transferred to a UK company incorporated for the purpose, Capital One Bank (Europe) plc ("COBEP"). COBEP is a subsidiary of Capital One Holdings Limited which is itself a subsidiary of COB.

Despite what might be inferred from their names, neither COB nor COBEP carries on business as a bank in the conventional sense, as the term is understood in the United Kingdom. COB is what is defined by US legislation (the Bank Holding Company Act 1956, as amended by the Competitive Equality in Banking Act 1987) as a "limited-purpose credit card bank", and is authorised only to issue credit cards and to take deposits of not less than US$100,000—in other words, wholesale rather than retail deposits, the intention being that its dealings with members of the public are confined to its credit card business. Its deposit-taking must be undertaken at only one branch, though its credit card activities may be pursued at many branches. Although COBEP is a United Kingdom company, it carries on the same business as COB and, because it is a subsidiary of COB, must comply with the same regulatory requirements. That, at least, is the appellant's contention; although not altogether conceding it the Commissioners do not argue otherwise, at least in relation to COBEP.

Since COB's UK branch and COBEP carried, or carry, on financial services business in the United Kingdom, they were, and are, also subject to UK regulatory requirements. One such requirement is that each should have some employees but as few as two suffice. With limited exceptions, which we will shall identify where necessary, all other Capital One employees (although they are called "associates") have, and at all material times had, contracts of employment with another of COFC's immediate subsidiaries, Capital One Services Inc ("COSI"). This company, too, is incorporated in the United States, and carries on business there and in the UK, through a local branch. COSI is in the same VAT group as COB's UK branch, as are COBEP and COFC's own UK branch; COB and COBEP have been successively the representative members of the group. COSI is also the parent of two of the Capital One subsidiaries which feature extensively in the events which we will later describe. We will deal with those subsidiaries, CODL (the appellant) and COCL, at that stage.

The material events

In this section of our decision we relate the history of the case, about which there was no material disagreement between the parties. In later sections we will examine the evidence we heard about the motives and reasons for the various events and other matters we describe.

In January 1999 PricewaterhouseCoopers ("PwC"), the chartered accountants, delivered instructions to leading counsel at the Revenue bar, Kevin Prosser QC, to advise in consultation on the effectiveness of a scheme of transactions devised by PwC whose objective, as it was described in those instructions, was "to mitigate the VAT cost incurred by financial and educational institutions ('Exempt Institutions') when constructing buildings to be put to exempt use." The consultation took place on 4 February 1999, when Mr Prosser duly advised. We were provided with copies of the instructions to Mr Prosser, of a note prepared by PwC of the advice given at the consultation and of some supplementary exchanges between PwC and Mr Prosser. He was required to consider not only VAT, but also other forms of taxation which might be of relevance (though we do not need to consider any other possible tax here) and it is worth remembering that he advised some two years before Halifax first reached the tribunal. Neither he nor PwC could know at that stage what would be the attitude of the Commissioners to such schemes, nor what view the tribunal and the courts might take of them.

The scheme, as it was put in PwC's instructions to Mr Prosser, was summarised in these terms:

"(a) A subsidiary of an Exempt Institution constructs a new freehold building. The construction costs are subject to VAT.

(b) The subsidiary makes two part disposals of the building:

(i) an exempt 999 year lease to its parent Exempt Institution

(ii) the taxable sale of the freehold reversion to a fellow subsidiary.

(c) The first subsidiary is able to treat the VAT on the construction costs as residual since it relates to both taxable and exempt supplies. The standard partial exemption method allows full VAT recovery."

We will deal with the legal provisions on which that scheme relies for its effect, or purported effect, when we come to examine the parties' arguments. At this stage we need mention only that the scheme is designed to take advantage of regulation 101(2)(d) of the VAT Regulations 1995 (SI 1995/2518) in order to ensure that all of the input tax incurred in the acquisition of the site and the construction of the building is recovered in the period in which it is incurred even though the greater part by value of the intended future supplies will be exempt. It requires the participation of a registered trader which intends to make in the future both taxable and exempt supplies. Although an exempt supply is to be made in the future, the trader incurs (and recovers) the input tax the scheme is designed to save in a prescribed VAT period in which it makes only taxable supplies.

Mr Prosser's instructions went on to set out rather more detail of the scheme as PwC envisaged it and of the requirements it would need to meet if it were to satisfy the requirements of regulation 101. Two particular issues which were identified and which are relevant here were the need to create tax points, possibly by making payments in advance, and, although for different reasons, the incidence of the capital goods scheme. There is no need for us to describe Mr Prosser's advice save to say that he suggested some changes of detail but otherwise expressed the view that the scheme would probably succeed in its purpose, though it was not entirely without risk.

PwC refined the scheme in order to take account of Mr Prosser's comments and then began to offer it to suitable clients, including Capital One. There were discussions between PwC and a senior official of Capital One from whom we were to hear a good deal of evidence, Graham Foster. The culmination of those discussions was that Capital One "bought" the scheme from PwC.

At that time—the summer of 1999—Capital One had been trading in the UK for some time from premises in Nottingham known as Trent House, which it had recently had refurbished. Its principal activity was COB's credit card business, undertaken at that time by its UK branch, but the group was also engaged in the sale of compact discs by mail order, a business undertaken by another of COFC's immediate subsidiaries, Oakstone Ventures Inc, and in telecommunications, through Oakstone's subsidiary Capital One Communications Inc ("COCI"). Oakstone and COCI were also in COB's VAT group and, we understood, it was their activities which dictated the group's partial recovery rate. Capital One's business was expanding and it was apparent that if it continued to do so, the group would soon require more office space to accommodate its employees.

On its northern and southern sides Trent House is bordered by a canal and a main road respectively, and on neither of those sides is there room for expansion. On the east lies a building, known colloquially because of its shape as the "H-block". Some thought was given to the possibility of acquiring and refurbishing the H-block. An alternative was to acquire a vacant site to the west of Trent House, then known as Kingspark. It was the site of a building which had been demolished and was then being used (though without the benefit of planning permission and, we learnt, despite enforcement action by the planning authority) as an open-air car park. Planning permission for the commercial redevelopment of the site had been granted at some time in the past, but it had lapsed. However, the nature and location of the site was such that it was thought unlikely that there would be any difficulty in obtaining new planning permission for commercial redevelopment if a further application were made.

Three possibilities were considered: the acquisition and refurbishment of the H-block; the acquisition of the Kingspark site and the erection on it of a new office block; and both of those, as a single project. It was eventually decided to adopt only the second of those courses. (A further scheme, for the creation of another centre at Sheffield, seems to have been discarded at a fairly early stage.) It was learnt that the then freehold owners of Kingspark, Prologis Kingspark Developments Limited ("Prologis") wished to dispose of it and were seeking a price Capital One was willing to pay. At this stage, Capital One was prepared to consider the acquisition of the site in order to hold it in a land bank or to resell it, although its expectation (or, at least, the expectation of those in charge of its UK operations) was that it would itself build on the site, and occupy the resulting accommodation.

Matters came to a head when Capital One learnt that a Regional Selective Assistance grant, payable by the Department for Trade and Industry, might be available and, rather than lose the site (or risk an increase in the price demanded if Prologis heard of the possible grant) it decided to proceed despite the absence of planning permission. On 1 October 1999 Prologis entered into two agreements with Capital One Contracts Limited ("COCL"), a subsidiary of COSI, which had been incorporated in September 1998 apparently as the SPV by which the refurbishment of Trent House was undertaken. COCL has at all times been individually registered for VAT.

The first of the agreements, a put option agreement, entitled Prologis to require COCL to buy Kingspark; the option could be exercised by Prologis between 1 April and 31 December 2000. The other, a call option agreement, entitled COCL to require Prologis to sell Kingspark to it; that option could be exercised at any time from 1 November 1999 to 31 December 2000. Thus the call option could be exercised up to five months before the put option came into effect. COCL was permitted by both of the agreements to assign its interests, and to require Prologis to transfer the land to its assignee. The price of the land, including the price paid by COCL for the option agreements, was £1.75 million whichever of the options was exercised. Prologis had waived the exemption (that is, had exercised the right conferred by paragraph 2 of Schedule 10 to the Value Added Tax Act 1994) and VAT was therefore payable in addition.

The belief that a grant might be available turned out to be correct. The application process was quite lengthy, but Capital One was ultimately successful. Although we heard a good deal of evidence about the grant application, and particularly representations made by Capital One about the extent to which it might be able to recover the input tax it had incurred, and expected to incur, in the course of the project, we did not find that evidence helpful in reaching our conclusions. It seemed to us to demonstrate no more than uncertainty and misunderstanding on the part of some of Capital One's staff. In our view the grant is of little or no relevance to the matters we must decide, though we will need to make some further brief references to it.

It had originally been intended that an existing group company, Oakstone (Nottingham) Limited, then thought to be dormant, would be used as a participant in the tax-saving scheme which Capital One had acquired from PwC but it transpired that the company had been struck off (for its failure to file annual returns) and instead an "off the shelf" company with the memorable name of Shelfco (No 1709) Limited was acquired from Capital One's solicitors in July or August 1999. That company changed its name to Capital One Developments Limited ("CODL") as soon as Capital One acquired it, and applied to become registered for VAT. The application was refused since the Commissioners were not then satisfied that CODL would trade. It was agreed between Capital One and the Commissioners that the registration of CODL would be reconsidered when it had acquired, or was about to acquire, some land.

In January 2000 an application for outline planning permission was made by town planning consultants instructed by Capital One; the application was in CODL's name. We record, since it was material to the arguments advanced on behalf of the respondents by Jonathan Peacock QC, that the outline planning permission granted to CODL by Nottingham City Council on 17 March 2000, reflecting the application, identifies the site as "Land bounded by Trent Street/Station Street/Nottingham Canal and including Trent House Station Street Nottingham", in other words both the Kingspark site and the existing Trent House. At the same time, Capital One's town planning consultants procured the consent of the city council to the parking of ten cars on the site. Their letter seeking that consent (which appears to have been granted informally) states that Capital One "have identified a need for some modest car parking space on [Kingspark] associated with the proposed implementation of the development scheme …".

It was decided that COCL should exercise the call option. The information available to us (which, on this and a number of other topics, was rather surprisingly absent from the bundles of documents and from the witnesses' evidence) did not reveal how or precisely when the option was exercised, but it was apparently at the beginning of March 2000 since the option agreement provided for completion of the purchase (which took place on 15 March 2000) ten working days after the option was exercised. Even though the planning permission had not yet been granted, we imagine Capital One's officers were confident that it would be (and there was a suggestion they had been given an informal intimation that the application had been successful). By this time it was clear that Capital One had a need for new accommodation and it was almost certain that it would develop the site entirely for its own benefit. COCL did not elect to have the property transferred to an assignee, as the agreement permitted, and took the transfer itself.

On 7 March 2000—that is, in the period between its exercising the option (as we have deduced, at the beginning of March) and completion—COCL entered into an agreement for lease with CODL. It provided that COCL would grant a lease in the form of an annexed draft (no copy of the draft was produced to us) on the Completion Date, which was to be 30 April 2000 or earlier by mutual agreement. Clause 3.3 of the agreement reads: "Either the Landlord or the Tenant can terminate this Agreement at any time prior to the Completion Date by written notice to the other". That right was not exercised, and the lease was duly granted on 16 March 2000. It is a lease of the entirety of the land acquired by COCL from Prologis on the previous day, for a term of 150 years from March 2000 (the day is not specified) at a peppercorn rent in exchange for a premium of £2,150,000. COCL elected to waive the exemption and VAT was therefore payable in addition. The election had the consequence that COCL was able to recover as input tax the VAT it had paid to Prologis. We have assumed, in case it should be of any importance, that the lease actually granted is in materially the same form as the draft.

On the same day—7 March—CODL granted a car parking licence to Woolf Limited ("Woolf"), which is a United Kingdom construction management company, at arm's length to the Capital One group. We will return to the licence shortly. Woolf had undertaken the refurbishment of Trent House, and it appears that it had already been decided that, assuming the development did proceed, Woolf was to be engaged by Capital One to manage the construction of the new building on Kingspark, to be called Loxley House. Woolf was in due course engaged; we did not see a copy of the agreement for the construction management (another striking omission from the documentary evidence) and we do not know when it was made, nor for certain which of the Capital One companies entered into it although, for reasons which will emerge, the probability is that the contracting company was COCL. CODL's interest in the land on 7 March was limited to its rights under the agreement for lease (and could therefore be terminated at any time). The car parking licence in favour of Woolf was expressed to run from 16 March 2000 (the day on which CODL in fact took the lease) until terminated by either party.

In the meantime, on 3 March, CODL renewed its application for VAT registration, supporting its request with copies of the agreement for lease with COCL and the car parking licence with Woolf (at that time still in draft). On this occasion the application was granted. Registration was effected from 7 March 2000, the date which the completed agreements bear. CODL requested that its VAT year end date should be 31 March. For reasons which were not made known to us (but which are probably immaterial) CODL was initially required to make its first VAT return to cover the period from 7 March to 31 May 2000, but that date was later amended to 30 April 2000, and its VAT year end was set at 30 April. Like COCL, CODL has at all times been individually registered for VAT.

On 16 March 2000, in addition to taking the lease from COCL, CODL entered into a further car parking licence in favour of COSI. It was in all material respects identical to the licence in favour of Woolf, with one exception. That exception related to the purposes for which Woolf and COSI could respectively use the parking facilities. We shall return to the significance of this factor when we deal with the evidence about the licences. Each agreement was for an indefinite term, terminable by either party on a month's notice, at an annual rent of £1000 plus VAT, payable half-yearly in advance. Each licence related to five of the ten spaces which the city council had authorised, which were to be "allocated to the Licensee by the Landowner from time to time." On 16 March 2000 CODL produced VAT invoices, one for each of Woolf and COSI, for £293.15 plus VAT, the apportioned charge for the period from 16 March to 30 June 2000.

On the following day, 17 March, COCL and CODL entered into two further transactions. First, COCL transferred the freehold reversion in the site to CODL for the sum of £25,000. The transfer included a declaration by CODL that the leasehold interest created on the previous day should not merge in the freehold. This was done partly in order to enable CODL later to assign the leasehold interest to one group company, and to transfer the freehold to another, and partly in order to avoid the application of the capital goods scheme, as Mr Prosser had advised. The decision letter reserves the Commissioners' position regarding the applicability of the capital goods scheme. For the purposes of this appeal we were not asked to consider whether it might apply, either for its own sake or in order to evaluate its relevance as a factor in determining whether the overall scheme comes within the Halifax principles. Mr Peacock merely invited us to leave it out of account altogether, and (with Dr Lasok's acquiescence) we do so.

Secondly, CODL contracted with COCL that the latter would "procure the design and the construction and fit out the Property [ie the Kingspark site] to provide call centre and office facilities …". The agreement (the "building agreement") described the development by reference to a specification and some plans which had already been prepared and stated that the estimated (or "prime") cost of the work was £42,975,840 plus VAT. COCL's fee for procuring the construction was to be 1% of the prime cost. CODL was required to, and did, pay the entire amount, including VAT, immediately. There was provision for adjustment of the prime cost (and with it the fee and the VAT) in the event of variations, and each party had the right to terminate the agreement on one month's notice. That option was not exercised, and the agreement remained in effect until completion of the construction. We did not have any detailed information about variations but it emerged that the actual cost of the development was rather less than had been predicted.

CODL, as a newly formed company which had not previously traded, had no assets of its own and borrowed the money which it needed to pay to COCL—that is, the consideration for the freehold, the premium due under the lease, the cost of the construction and COCL's fee with the appropriate VAT—from its parent, COSI. COCL had no immediate need for the money paid for the cost of construction—work had not then started and there was no need to make any payment to contractors—and COCL therefore paid the money to its parent which, as we have indicated, was also COSI. We deduce that it was later drawn upon in order to pay for the development as it proceeded, but in March 2000 it effectively went round in a circle.

On 7 April 2000 COSI entered into an agreement with CODL by which CODL agreed to supply security services to COSI. It was recorded in the agreement that COSI was the occupier of the Kingspark site for which it required the security services which CODL was willing to provide, though it had the right to do so by the use of sub-contractors. The agreement includes the warranty by CODL "that it has experience and expertise in performing services of the type required under this Agreement". On 26 April 2000 CODL entered into a further agreement with Wilson James Limited (a security company at arm's length to the Capital One group) by which Wilson James would provide security services to CODL in respect of premises which CODL occupied—again defined as the Kingspark site. There are some necessary consequential differences, such as in the names of the parties, between these two agreements, and the significant difference that CODL was to charge COSI rather more for the services than it was to pay Wilson James to provide them. Otherwise the agreements are very similar and we suspect Capital One had "borrowed" the wording of Wilson James' standard form agreement. In the event, these agreements were abandoned and either COSI or COCL—we did not learn which—entered into a direct contract with Wilson James in their place but the fact that the agreements were made at all is a matter to which we will need to return in due course.

CODL submitted its VAT return for period 04/00—its first return—on 3 May. It disclosed an output tax liability of £102.60, which was the VAT payable on the consideration for the two car parking licences for the period from 16 March to 30 June. It made no other supplies in the period. The return included a claim for repayment of input tax of £7,933,066.09, incurred in the acquisition of the freehold, on the premium paid for the lease, and on the payment to COCL made pursuant to the building agreement. The net sum reclaimed was therefore £7,932,963.49.

The magnitude of that claim led, not surprisingly, to enquiries by the Commissioners into its validity. Those enquiries in turn resulted (after lengthy correspondence and some meetings between the parties) in the decision letter of 28 June 2001 setting out the Commissioners' five decisions which are the subject-matter of this appeal. At the time the decision letter was written, some of the events had already occurred while others were still in the future—and might not occur at all—and the author of the letter, a Mr Colford (from whom we had no evidence) necessarily made some assumptions, not all of which turned out to be entirely correct. This was a point on which Dr Lasok was to focus in his submissions, and we shall return to it.

Construction work began in about June 2000 and continued until March 2002, when the practical completion certificate was issued. During that period, it became apparent to Capital One that COCI's telecommunications business was not expanding as quickly as it had hoped, while COBEP, which had by now taken over the business and assets of COB's UK branch, was continuing to expand and was in need of all the space within Loxley House. Accordingly the decision was taken that COBEP's staff (although they were, as before, formally employed by COSI) would occupy the entirety of Loxley House. On 13 March 2002, to coincide with the completion of construction, CODL disposed of its interests in the site and the building.

The freehold was sold to Worldscape Inc ("Worldscape"), a US corporation in which the shares were held as to 95% by COB and as to the remaining 5% by COSI, for the sum of £30,000 (that is, £5,000 more than CODL had paid to COCL for the freehold interest). CODL treated this as a taxable supply and added VAT to the consideration. CODL assigned its leasehold interest to Capital One Properties Inc ("COPI"), also a US corporation and owned in exactly the same way as Worldscape, for £44 million. That was treated as an exempt transaction since CODL did not waive the exemption.

We spent some time at the hearing exploring the manner in which the consideration for the two interests had been determined since it seemed at first that CODL had made a substantial loss when disposing of its freehold and leasehold interests, after taking into account its costs of acquisition and development and offsetting the sums it received by way of grant. It transpired however that the construction costs were significantly lower than had been projected and, although we did not hear complete evidence on the matter, we are willing to accept that overall CODL did not make a loss.

On the same day, 13 March 2002, COPI entered into an agreement for underlease with COBEP. It had originally been envisaged (at least, from the time when it became clear that the whole of Loxley House would be required for the credit card business) that there would be an underlease to COB, but that intention changed when COB's UK branch business was transferred to COBEP. The copy of the underlease produced to us is undated but we assume it was granted in March or April 2002 (there was a necessary delay after the agreement for underlease while COBEP and COPI obtained the order of a court, in accordance with section 38(4) of the Landlord and Tenant Act 1954). The term of the underlease is 20 years at an annual rent of £2.75 million, to be reviewed at five-yearly intervals. COPI did not then and, as far as we understood the matter, has not since waived the exemption in respect of this underlease. Thus COBEP has received an exempt supply of the new accommodation, and has incurred no irrecoverable VAT.

The evidence

We heard oral evidence from the following witnesses:

Graham Foster, COBEP's director of international tax;

Michael Brierley, COBEP's UK director of finance;

Michael Says, chief financial officer for Capital One's European business, and a director of CODL;

Carl Longworth, COBEP's director of European corporate real estate;

William Yontz, now retired but formerly Capital One's worldwide head of corporate real estate and director of some of the relevant companies;

Anders Navarette, associate general counsel (US-qualified) for regulatory affairs and legal risk management of the Capital One group;

Ronald Davis, a director of Jonathan Edwards Consulting Limited, property management consultants engaged by Capital One for the Loxley House project;

Victoria Mitchell, associate general counsel (UK-qualified) for, and company secretary of, COBEP;

Paul Jepson, a senior Customs officer; and

Gary McMahon, formerly employed by both COSI and COCL as project manager for the construction of Loxley House.

Each of the witnesses had made one or more statements which were disclosed before the hearing and which stood as their respective evidence in chief. We had in addition an expert's report prepared by Bradley Sabel, a US attorney who has particular knowledge of and expertise in United States financial regulation legislation and practice. Mr Sabel's evidence was not challenged, and the respondents put in no expert evidence of their own.

We do not propose to embark on the arid exercise of setting out each witness's evidence seriatim, thereafter listing our findings of fact, but instead to relate the evidence—identifying the individual witnesses only as necessary—to the events as we have described them, in order to ascertain the purpose of and motive behind each step in the sequence. Since there was no dispute about the events we have concentrated on the essential issue identified by Halifax, namely whether each step had a legitimate commercial justification, or was undertaken for no purpose other than the avoidance of tax. It is necessary to begin, however, with some more general observations, and then to set out the relevant law, before examining the transactions. We will deal with some points of evidence when we come to the parties' submissions, since that evidence needs to be put in its context if it is to be understood.

It became abundantly clear from the evidence that the Capital One group is a large organisation with a closely defined management structure. Each member of its staff had a particular role and, although he or she would (and probably must) be aware of what others in the organisation were doing when they were engaged together on a project, none strayed outside their respective roles, but instead relied on colleagues to perform the functions allocated to them.

A particular example is Mr Yontz, at the time a very senior member of Capital One's management team (although, despite his seniority, employed by COSI). His task was to ensure that Capital One had suitable and sufficient accommodation for the carrying on of its various activities. In that capacity he was the chairman of Capital One's Real Estate Oversight Board ("REOB"), a committee which brought together a number of disciplines although, as its name implies, its primary role was to assume responsibility for the group's property portfolio. It was the REOB which approved the construction of Loxley House, which Mr Yontz told us was one of Capital One's major, and high-profile, projects at that time. It was clear from his evidence that Mr Yontz devoted a great deal of his time and energy to scrutinising the written submission which was made to the REOB in support of the project (with which we will deal in more detail shortly); he needed to satisfy himself and his fellow REOB members—and thereafter COFC's main board, which would authorise the allocation of funds—that the project was necessary and suitable for Capital One's needs, that the design was to an appropriate standard and that the anticipated cost was acceptable. He considered the submission in draft at several stages as well as in its final form, and made some contribution to defining that final form.

However, although he was aware from the submission (in which it is explicitly mentioned) that a tax planning scheme was a feature of the project, he candidly conceded that he did not really understand it. He is, of course, an American with expertise in property and is not a British tax lawyer, but his lack of understanding extended beyond the mechanics of the scheme; he realised no more than that there was the potential for mitigating the cost of the project. We say that not by way of criticism of Mr Yontz, but in order to illustrate the fact that, in accordance with what we perceive to be Capital One's usual policy, he left matters of that kind to those of his colleagues within whose expertise they fell.

Similarly Mr Brierley and Mr Says, although they were directors or authorised signatories of one or more of the corporate vehicles participating in the scheme, and although they signed some of the relevant documents, had little understanding of the tax saving scheme. Mr Longworth too knew of its existence but it was not in his province; he directed his attention to the nature and quantity of the space required, and left the tax saving scheme to others, particularly Mr Foster. Despite his being the head of UK real estate at the time, Mr Longworth made no contribution to the decision that the transactions should be structured as they were. Ms Mitchell (employed by COBEP rather than COSI), though she concerned herself with those formal matters which are the responsibility of a company secretary, and in that capacity witnessed the execution of some of the later documents, had very little knowledge of the mechanics of the scheme (although we should perhaps add in fairness to her that she was on maternity leave for a large part of the relevant time).

Instead, all of those witnesses took it for granted that Mr Foster, whose role was to concern himself with Capital One's tax affairs within the United Kingdom, would deal entirely with matters of that kind. It was clear to us, whatever their underlying understanding might have been, that the other members of Capital One's management team who were involved in the project all accepted that Mr Foster was acting in the group's best interests, and that the documents he asked them to sign were appropriate, and ones they could properly sign, although we recognise that they did consider them before signing, even if only in a general sense. It was conspicuous that Mr Says, who was then Mr Foster's immediate superior, and who had attended some of the meetings at which the scheme was discussed, had only a very general understanding of it, and was not involved in the making of decisions about its structure.

Again, we do not criticise those witnesses; they were being guided by Mr Foster, whom they knew to be responsible for Capital One's tax affairs, and who, as they also knew, was receiving advice not only from PwC but also from Capital One's solicitors, who were responsible for drafting the various legal documents they were asked to approve and sign. What was abundantly clear to us was that, alone of Capital One's senior employees, Mr Foster understood the tax saving scheme and it was ultimately he who took the decisions about how the scheme (which, as we have already indicated, was modified from the generic scheme devised by PwC and approved by Mr Prosser) was implemented by Capital One.

One particular point on which we should comment is that there was no clear evidence available to us about the manner in which the consideration for any of the intra-group transactions had been determined. It was apparent that no formal (nor, probably, informal) advice had been obtained from external consultants. Mr Yontz told us that Capital One took a relaxed attitude about intra-group transactions, requiring much less detail of the purposes behind them than it would of a transaction involving a third party and, we deduce, that relaxed attitude extended to the calculation of the consideration since the money would remain within the group. There was certainly no evidence that the consideration for any of the intra-group transactions had been scrutinised by Mr Yontz, Mr Brierley, Mr Says or Mr Longworth. It was conspicuous too that Mr Foster was unable to explain how any of the figures had been decided upon, although he was responsible for ensuring that the necessary transactions were completed. He mentioned that he had been concerned about transfer pricing (oddly, in relation to the transactions between UK-registered companies, although at the time the transfer pricing rules related only to international transactions). The conclusion we have reached is that it was Mr Foster, principally though perhaps not alone, who dictated the value of the consideration in each case. We are sure he did so in a responsible fashion, but it was plain that determining the value of intra-group transactions was not an objective in itself.

Mr Foster could not, of course, treat the scheme in isolation. The most important requirement of all was that Capital One should obtain additional accommodation suitable for its needs; we accept from Mr Yontz's evidence that the acquisition of the site and the construction of Loxley House would have proceeded even without the tax saving scheme, and that the potential saving it offered, though by no means insignificant, was only one of several relevant factors. All the material considerations were brought together in the written submission to the REOB which we have mentioned. It was prepared for a formal presentation on 11 April 2000. By that time, of course, COCL had exercised the call option, having been permitted to enter into the agreement and thereafter to exercise the option by Mr Yontz (it was within his personal authority limits to do so). If the plans for the construction of Loxley House had not been approved by the REOB, Capital One would have been left with a marketable development site with no commitment to develop the site itself.

The submission, about which we heard a good deal of evidence, contains contributions from several Capital One employees, and some from outsiders such as Mr Davis and his colleagues. It went through several drafts and evidently took quite some time to prepare. It contains a detailed analysis of Capital One's recent trading experience in the United Kingdom, and identifies the need for additional accommodation to cater for its projected continuing growth. The document discusses the suitability of constructing additional accommodation on a site adjacent to Trent House, against the alternatives of leasing or buying an existing building, as well as the possibility of locating new offices elsewhere. Capital One was by then established in Nottingham, but the practicalities of its being located on several remote sites were considered. The availability of sufficient staff of appropriate calibre in the Nottingham area was a matter of some importance. There are extensive descriptions of the design and proposed construction methods, and a discussion of planning considerations. Another matter of significance included in the paper was the need to provide an "exit strategy", that is, a means by which Loxley House could be disposed of (or even partially sublet) if it proved to be wholly or partly surplus to Capital One's requirements.

As one might expect, the estimated cost of the project receives extensive and comprehensive treatment in the submission. Several subsidiary issues are identified: the most economical means of financing the acquisition and construction; the means by which Loxley House should be owned, and how its ownership might be financed, in the longer term; the availability of the development grant; and the tax saving scheme. All of these issues impacted to some extent on others. For example, it was desirable for accounting reasons that in the longer term, Loxley House should be an "off balance sheet" asset within the Capital One group. That requirement fitted without difficulty into the tax saving strategy, but created some problems in the context of the grant application; we do not need to dwell on this issue, but merely make the point that the paper addressed a multiplicity of issues. The description it contains of the tax saving scheme is written in very general terms, and does not include any detail about the transactions necessary for its implementation, whether already undertaken or in contemplation. That fact too supports our conclusion that Mr Foster alone was responsible for the detail of the scheme; no-one else, at least at senior management level within Capital One, seems to have been privy to that detail.

It was clear to us that there were many considerations which Capital One needed to, and did, take into account when deciding how to approach the acquisition of Kingspark and the construction of Loxley House, and that many of those considerations had nothing to do with any possible tax saving plan. The question remains, however, whether those considerations could have been dealt with adequately if those of the transactions which the Commissioners maintain to be artificial, or contrived, had not been entered into—in other words, did those transactions have any purpose beyond the intended saving of tax. In order to understand the significance of the transactions, and put them in their context, it is necessary at this stage to examine the relevant law and the means by which the scheme is intended to operate before returning to the evidence.

The law underlying the scheme

The success or otherwise of the appellant's scheme depends entirely upon the application to the various transactions of regulation 101 of the VAT Regulations 1995, which is entitled "Attribution of input tax to taxable supplies", and is intended to implement articles 17(5), 19 and 20 of the Sixth Directive. Dr Lasok did not suggest that it failed to do so in any way. In its entirety the regulation reads (and read at the material time) as follows:

"(1) Subject to regulation 102, the amount of input tax which a taxable person shall be entitled to deduct provisionally shall be that amount which is attributable to taxable supplies in accordance with this regulation.

(2) In respect of each prescribed accounting period—

(a) goods imported or acquired by and goods or services supplied to, the taxable person in the period shall be identified,

(b) there shall be attributed to taxable supplies the whole of the input tax on such of those goods or services as are used or to be used by him exclusively in making taxable supplies,

(c) no part of the input tax on such of those goods or services as are used or to be used by him exclusively in making exempt supplies, or in carrying on any activity other than the making of taxable supplies, shall be attributed to taxable supplies, and

(d) there shall be attributed to taxable supplies such proportion of the input tax on such of those goods or services as are used or to be used by him in making both taxable and exempt supplies as bears the same ratio to the total of such input tax as the value of taxable supplies made by him bears to the value of all supplies made by him in the period.

(3) In calculating the proportion under paragraph (2)(d) above, there shall be excluded—

(a) any sum receivable by the taxable person in respect of any supply of capital goods used by him for the purposes of his business,

(b) any sum receivable by the taxable person in respect of any of the following descriptions of supplies made by him, where such supplies are incidental to one or more of his business activities—

(i) any supply which falls within item 1 of Group 5, or item 1 of Group 6, of Schedule 8 to the Act,

(ii) any grant which falls within item 1 of Group 1 of Schedule 9 to the Act,

(iii) any grant which falls within paragraph (a) of item 1 of Group 1 of Schedule 9 to the Act,

(iv) any grant which would fall within item 1 of Group 1 of Schedule 9 to the Act but for an election having effect under paragraph 2 of Schedule 10 to the Act, and

(v) any supply which falls within Group 5 of Schedule 9 to the Act,

(c) that part of the value of any supply of goods on which output tax is not chargeable by virtue of any order made by the Treasury under section 25(7) of the Act unless the taxable person has imported, acquired or been supplied with the goods for the purpose of selling them, and

(d) the value of any supply which, under or by virtue of any provision of the Act, the taxable person makes to himself.

(4) The ratio calculated for the purpose of paragraph (2)(d) above shall be expressed as a percentage and, if that percentage is not a whole number, it shall be rounded up to the next whole number."

The appellant's case is that in its accounting period 04/00 it received various supplies. These were the supply to it by COCL of the leasehold, for which it paid a taxable premium; the further supply to it by COCL of the freehold reversion, the consideration for which was also taxable; and the taxable supply of the building contract for which it paid the entire consideration, albeit subject to later adjustment, in that period. It intended to use those supplies for the purpose of making both taxable and exempt supplies: in particular, the exempt supply of a leasehold interest, and the taxable supply of the freehold reversion.

In those circumstances, it claims, the identification required by regulation 101(2)(a) regulation must be made: that poses no difficulty. Neither paragraph (b) nor (c) can apply, since there was no intention to make exclusively taxable or exclusively exempt supplies, nor could the respective values of such taxable and exempt supplies it might make be determined at that time; thus paragraph (d) applies and governs the extent to which the input tax it incurred on acquiring those supplies is recoverable. The proportion to be recovered is dictated by the proportion which CODL's own taxable supplies bore to all of the supplies it made in the accounting period in which the relevant input tax was incurred. The only supplies CODL made in period 04/00 were of the car parking licences, which were taxable. Hence its taxable supplies amounted to 100% of all its supplies in the period, and it was correspondingly entitled to recover the whole of the input tax.

None of paragraph (3) of regulation 101 is relevant in this case, although it will be observed that sub-paragraph 3(a) contains the reference to the capital goods scheme which Mr Prosser advised should be avoided. However, paragraph (1) is of some importance, since it provides that the attribution made under regulation 101 is no more than provisional (regulation 102, we should mention, applies to special schemes for partially-exempt traders and is not relevant here). The attribution has later to be adjusted, if it turns out to be incorrect, in accordance with regulation 107. We do not need to set that regulation out, but merely observe that the adjustment is made by reference to a "longer period". The longer period is defined by regulation 99, and particularly sub-regulations (1)(d) and (5). The combined effect of those provisions is that CODL's longer period ended on 30 April 2000—that is, on the same day as the end of the period by reference to which it was to make its first return. Regulation 107 requires that the adjustment, if any is required, must be made by reference to the exempt and taxable supplies the trader made in the longer period. Thus if the appellant's case is correct no adjustment was required, nor indeed possible. We add for completeness that regulation 107A, which might have affected the success of the scheme, did not come into force until 17 April 2002 (after the last of the relevant transactions) and does not have retrospective effect.

The respondents' contention is that (assuming Halifax to be correct) the scheme fails in its purpose because the car parking licences were an artificial device of no true economic substance and are to be disregarded; the pre-payment by CODL to COCL of the estimated construction costs was likewise a device with no purpose other than tax avoidance; and the transactions within the Capital One group by which the leasehold interest was separated from the freehold were also entirely artificial, driven by no motive other than the saving of tax. In the decision letter the argument advanced is that the transactions between CODL and COCL were not (in the case of those which had already occurred) and would not be (in the case of those yet to happen) economic activities (the "first decision"), or supplies for VAT purposes (the "second decision"). Adopting an analysis of the substance of the transactions similar to that used by the tribunal in Halifax, it was contended that the true supplies were by Prologis, of the site, and by the arm's length builders, of the construction, to COB, for use in COB's exempt business activities (the "third decision", also referred to in the decision letter as the "correct supply position"). Alternatively, it was said, CODL's "purported exercise of its right to deduct is an abuse of that right" (the "fifth decision"). As we mentioned at the beginning of this decision, the letter also asserted (as the "fourth decision") that Loxley House is not a building. We propose to deal with this issue, and its significance, separately.

Mr Peacock also argued that the scheme failed, irrespective of Halifax principles, because the appellant could not bring itself within regulation 101(2)(d) at all. For that argument he relied on the emphasis laid by several witnesses on Capital One's requirement for flexibility (so that, they said, a final decision about the use of Loxley House would not be made until it was complete, or nearly so), on the indications in the REOB submission—which proceeded upon the assumption that COB or COBEP would occupy Loxley House—that the exempt supply of a lease was to be made, while there was no mention of the taxable disposal of the freehold, and on the differing statements made by Capital One to the DTI in connection with the grant—sometimes that the VAT would be recovered, sometimes that it would not. More particularly, he relied on a letter written to the Commissioners by PwC in August 2000, during the course of the debate between the parties which led ultimately to the decision letter. In that letter PwC referred repeatedly to the fact, as they put it, that Capital One had not yet determined what it was to do with the property—"… the company cannot give a definitive explanation about the proposed use of the property …"; "… it is very difficult for the group to make even tentative plans about property use even as little as one year in advance …"; "The only use that is likely is that the company will sell the freehold once the building is completed or soon after … However there is as yet no firm intention about this"; and "… the company is unable to establish whether the goods or services will be used to make taxable or exempt supplies …".

Those passages, he said, even though written a few months after the end of period 04/00, showed clearly that Capital One in general and CODL in particular did not have the intention in that period of making both taxable and exempt supplies. Without that intention, the condition on which the application of regulation 101(2)(d) depended was not satisfied, and CODL's claim that it came within that provision must fail and, with it, CODL's contention that, because it made only the taxable supplies of the car parking licences—assuming for this purpose that they were of economic substance—in that period, it was entitled to reclaim all of the input tax it had incurred. There are, we think, three things to be said about this argument.

First, we think Mr Peacock is asking us to read too much into PwC's letter. It was written for the purpose of challenging the Commissioners' view, at that time expressed only informally, that CODL was engaged in a planned tax avoidance scheme and it was designed, in part, to suggest that the arrangements were not planned in the manner, or with the degree of pre-determination, they suspected. In any event, albeit perhaps with a degree of exaggeration, it reflects the reality (as we accept) that Capital One was not at that stage—when the construction had barely started—entirely certain that COB, or COBEP, would occupy Loxley House when it was completed. That, in essence was what the witnesses said. As will emerge, we have no doubt that the intention was that COB or COBEP would occupy most if not all of Loxley House, and that it was at all times the most likely outcome, from which it follows that, even with some reservations in case its businesses did not expand as it hoped, Capital One (and correspondingly CODL) did intend, all being well, to make the taxable and exempt supplies which the scheme required.

Secondly, it seems to us that Mr Peacock was seeking to advance two incompatible arguments: either this was (as the decision letter maintains) a carefully contrived and managed scheme designed to achieve an unwarranted tax advantage; or it was a much less organised series which fails to achieve a tax advantage because of that lack of organisation and, particularly, the absence of an identifiable intention in CODL, in period 04/00, to make in the future both taxable and exempt supplies. That is not an argument even touched upon in the decision letter, which proceeds wholly on the former basis, and it seems to us (this being an appeal against the refusal to allow input tax credit for the reasons advanced in the decision letter) the Commissioners' arguments should be confined to those set out in that letter. In any event, as we have indicated (and we shall expand further on the point), we are satisfied that there was an intention to make both taxable and exempt supplies, and Mr Peacock's argument fails on the evidence.

Thirdly, even if Mr Peacock's argument were factually sound, it does not in our view succeed in law. It is necessary to set regulation 101(2)(d) in its context: the relevant supplies are first identified—sub-paragraph (a)—then those attributable to exclusively taxable—sub-paragraph (b)—and exclusively exempt—sub-paragraph (c)—supplies are set to one side. What remains necessarily falls into sub-paragraph (d), for want of any other home. That is not surprising; the purposes of regulation 101, as paragraph (1) makes clear, is to make a provisional attribution. Later regulations—and not merely regulation 107—provide for adjustment of that provisional attribution, and allow both for cases in which it is difficult or impossible to make an accurate attribution at the outset and for changes of intention.

The suggestion, necessary to Mr Peacock's argument, that the trader must have a fixed, or at least identifiable, intention at the time he incurs the input tax for which he claims credit under regulation 101(2)(d) is, in our view, wrong. He supported his argument by referring us to the decision of the Court of Justice in Finanzamt Goslar v Breitsohl (Case C-400/98) [2001] STC 355, but it does not seem to us that the case is in point. It certainly supports the proposition that an intention to make supplies in the course of business is essential, but there has never, we think, been any dispute that CODL intended to make some supply by way of business. It is the uncertainty whether those supplies will be taxable or exempt (and, as in this case, what their respective values will be)—issues on which Breitsohl does not touch—which excludes the application of sub-paragraphs (b) and (c) and brings sub-paragraph (d) into play.

We did not hear argument about the Commissioners' view that CODL's claim is an abuse of right, since it is impossible to deal with the issue without the Court of Justice's answer to the second of the referred questions in Halifax. We proceed, therefore, to review the evidence we heard about the purpose behind the various transactions we have identified. Although the appellant acknowledges that tax avoidance was a material factor, it says that all the transactions were undertaken for other reasons in addition and that, accordingly, none offends the Halifax test.

The purpose of the transactions

We have already made the comment that Mr Foster alone of Capital One's management team understood the detail of the tax saving scheme, and of the witnesses it was he who dealt almost entirely with the purpose of the transactions. As we have mentioned we heard evidence from others, particularly Mr Says and Mr Brierley, about Capital One's need to retain flexibility in order that it could accommodate to changes in circumstances—most importantly the possibility that Capital One's expansion would not continue as rapidly as was predicted—and to the effect that account had to be taken of the needs of both the credit card and other business interests then pursued by Capital One in the UK. We have taken that evidence, as well as the evidence we heard about preferred accounting methods and the need to satisfy certain ownership requirements if the grant application were to succeed, into consideration in reaching our findings. We will deal separately with the evidence we heard about the regulatory requirements which affected Capital One's operations.

Some of the witnesses (but particularly Mr Foster) mentioned that the provision of parking facilities for staff employed at Trent House was poor. The application for planning permission for the construction of Loxley House, as we have recorded, included a reference to Trent House, in part in order that the two might be considered together in the hope that the planning authority would allow the creation of more parking spaces than it might have done if faced with an application relating to Loxley House alone. There was some evidence that staff working at Trent House were parking in the street, and we imagine some would have parked their cars on Kingspark before COCL bought it and closed the car park which had been on it. We accept that there was a genuine need for additional, long term parking space.

Woolf had been the managers of the refurbishment of Trent House and, it appears, some of their employees were still on site in early 2000, using parking facilities within the H-block. There was an indication, though no firm evidence, that they may have been doing so without the owner's permission, although it also seems they had been doing so without objection and for some time. The car parking licence between CODL and Woolf entitled the latter to use the five allocated spaces in the temporary car park on Kingspark "only [in] connection with their responsibilities as Project Managers of the construction works", the "construction works" being the erection—at that time still only a proposal—of Loxley House. Construction began in June 2000, although the precise starting date would not have been known in March. The appellant's case is that in the period before construction work started, Woolf would need to undertake some site investigations and required access to the site, and facilities for its employees to park their cars there, for that purpose.

Mr McMahon was Capital One's own project manager, employed simultaneously (and unusually) by both COCL and COSI. For the purposes of the construction project itself, he was the point of contact between Woolf and Capital One. He knew of the tax planning scheme and, although it was evident he did not understand all of its detail or the legal provisions which underlay it (we think it unlikely they were explained to him), he was well aware of its importance. Among the documents we saw was a schedule prepared by Mr Foster setting out the sequence of events which must be respected if the tax saving scheme were to succeed. It mentioned, for example, the need to ensure that CODL became registered for VAT in good time before it began to enter into the transactions. One copy of the schedule bore manuscript notes added by Mr McMahon; it was clear to us that Mr Foster had gone to some trouble to ensure that all those involved understood the need to adhere strictly to the timetable set out in the schedule.

Mr McMahon's evidence, which we accept, was that he was in frequent contact with Mr Foster, and was well aware from his discussions with him and from the schedule of CODL's need to create tax points during March 2000. He realised that (whatever other purpose it might have served) the car parking agreement with Woolf was intended to create a tax point. He told us that Mr Foster asked him to arrange for Woolf to enter into the agreement, which he did; Woolf had not asked for parking facilities (nor, so far as Mr McMahon was aware, did it need any) but was quite willing to enter into the agreement. Part of Mr McMahon's instruction from Mr Foster was to explain to Woolf that although the agreement provided that Woolf would pay for the parking facility, it would in fact cost it nothing. Mr Peacock made the point that the agreement in favour of Woolf was limited to use in connection with the construction, and not with pre-construction site preparation; and that once construction started it would be impossible, for safety reasons alone, to park private cars on the site. That, we suspect, is a subtlety which escaped the notice of the draftsman of the agreement rather than the fundamental incompatibility which Mr Peacock suggested, though it does suggest that the agreement may have been prepared hastily, and it also supports the conclusion that the agreement was not designed to satisfy an identified need.

Mr McMahon was not instrumental in arranging the similar parking agreement between CODL and COSI, which allowed COSI's employees to use the site "only in connection with their employment" but without other restriction, and we did not hear clear evidence about how that agreement came into existence. The copy we had showed that it was signed for CODL by Mr Yontz (who had no real recollection of the document) and for COSI by another senior member of Capital One's management who did not give evidence.

Some time was spent at the hearing examining whether, and if so how, the consideration for the first period of the parking licences, for which CODL issued VAT invoices, had been paid. The evidence was not altogether clear, though it did appear that COSI had paid the sum due from it. By a somewhat elaborate process Mr Patchett-Joyce was able to show that Woolf had made a payment (although it had been credited by mistake to COCL's bank account). Nevertheless, we are satisfied, from Mr McMahon's evidence and from the rather limited documentation we saw—and it was not, in the end, challenged by the appellant—that the parking facility was in truth free of cost to Woolf, because, having paid, it recovered the money as an expense of the construction contract which was reimbursed to it.

Although, as we have said, there was a need for parking space in a general sense, the question arises whether either of the agreements satisfied that need, even in part. Mr McMahon's evidence was that he did not see any cars parked on Kingspark between its acquisition in March and the start of construction work in June, after which the use of the site for car parking would be impossible. He told us that Woolf had no real need of parking facilities since its staff were using the H-block, and continued to do so, and any site investigations it undertook in the period before construction began would involve the use of machinery, implying little or no need for the parking of private cars. Woolf had signed the agreement not because of any perceived need for parking space for what would in any event be only a short period, but because the company was cooperative and willing to do what it was asked to do in the context of its securing a large construction project.

Mr Foster, who obviously was taking an interest in the site and who, like Mr McMahon, had his office in Trent House, could not tell us that he had seen cars on the site in the period from March to June 2000, though he was reluctant to concede that he had seen none. Neither Mr Foster nor any other witness was able to tell us to which of COSI's employees the spaces it took under its parking agreement were allotted, nor why COSI should have thought it worthwhile entering into a parking agreement which would almost certainly be of short duration. It was also not explained why COSI, which was CODL's parent, needed to enter into a formal agreement with CODL for this purpose.

Mr Peacock shrank from contending that the parking licences were shams, but he came very close to doing so. He reminded us that the cost to CODL of the land exceeded £2 million whereas the aggregate annual income generated by the licences, assuming they had remained in effect, was only £2,000. Although there is no formal requirement that there should be a correlation in value between inputs and outputs, the enormous difference in this case suggested, he said, that there could be no real economic purpose to the licences. The licence in favour of Woolf was granted on 7 March 2000, when CODL's interest in the land was limited to that granted by the agreement for lease. There was no need to enter into the licence on that day, save that it assisted CODL in its application for VAT registration. He added that payment in advance, as the agreements demanded, for parking facilities which it was expected would be terminated in the near future, was rather odd. Mr Foster's explanation that a refund would be made if it transpired that too much had been paid struck us as rather lame. It also struck us that the limited income which might be generated by the agreements in the interval between their being granted and the expected start of construction would be comfortably exceeded by the cost, in solicitors' fees and staff time, of putting them in place.

Dr Lasok argued that Mr McMahon's and Mr Foster's evidence showed no more than that they had not seen cars on the site between March and June 2000, and did not go so far as to establish that the parking facilities had not been used. In a literal sense that is right, but we have no real doubt that if cars had been parked on the site, Mr McMahon, whose work related to the site, would have noticed them and Mr Foster would have made it his business to do so. It was clear that the allocation of spaces for which the agreements provided had not taken place. There was no evidence that arrangements had been made by which authorised users could gain access to the site while others were excluded. Had any such arrangements been made, we are sure Mr Foster or Mr McMahon would have been aware of them, but neither could tell us what, if anything, had been done. We deduce that no access arrangements were ever made. Dr Lasok also argued that, even if the facilities were not in fact used, the agreements nevertheless conferred on Woolf and COSI the benefit that the facilities could be used, and he pointed out that the fact of payment was itself a good indication of economic activity; it is, as he put it, "classically economic".

We are satisfied on the evidence we heard that neither of the parking agreements satisfied any real need, and that no cars were parked on the site as a result of the grant of the licences. The benefit identified by Dr Lasok was, we have concluded, entirely hypothetical: neither Woolf nor COSI had any true need of the space, nor any intention of using it, and Capital One had done nothing to make it available in a practical sense. Moreover, if Woolf was engaged to undertake work on the site, it would take control of it and would itself be able to dictate who could park there; it is quite unrealistic to suggest that, if its employees attended to undertake site investigation or preparation work or to begin the construction and parked their cars on the site, safety permitting, Capital One would have objected, or raised a genuine arm's length charge. COSI, as CODL's parent, could easily have arranged for its employees to park on the site without the need for a formal agreement or for payment. An agreement of this kind between parent and subsidiary seems to us self-evidently artificial. The fact of payment may be a good indicator of economic activity when it is a true payment; but where, as here, it is refunded (in Woolf's case) or money which goes round in a circle (as in COSI's case), it seems to us to be of no evidential value.

We are left in no doubt that the two parking agreements had no real purpose other than the creation, or in reality purported creation, of taxable supplies and with them tax points in March 2000 and, as Mr Peacock argued, the support of CODL's registration application. Neither COSI nor Woolf had any genuine motive of its own for entering into the agreements, but did so merely in order to assist CODL in its tax saving scheme. We discern no true economic or business purpose in either agreement.

Although the agreement for the provision by CODL of security services to COSI was abandoned, the fact that it was contemplated at all seems to us to support the conclusion that there was a large measure of artificiality about the transactions. It was conceded by Mr Says, who signed the agreement in his capacity of director of CODL, that, despite the warranty to the contrary in the agreement itself, CODL had no expertise in the provision of security services. Even if it had any such expertise, it would repose in the employees of COSI, since CODL had none of its own and relied on COSI for its personnel. It defies belief that there could be any true business purpose to an agreement by which CODL, using COSI's employees to undertake their procurement from a third party, would supply to COSI services which COSI could as easily obtain for itself by using those same employees. Any profit CODL made would be for the benefit of its parent, COSI, from which the same profit would have been earned. The fact that the agreement was abandoned (we were not told why that decision was taken and the fact of the abandonment emerged almost by accident at the hearing) and COSI obtained security services directly from the arm's length supplier, rather than through CODL, seems to us to demonstrate a belated recognition by Capital One that this was a transparently artificial device. The irresistible conclusion, in our view, is that the only true purpose of the agreement was to create a tax point.

The agreement which gave rise to the greater part of CODL's input tax claim was the building agreement with COCL for the construction of Loxley House. This agreement was also signed by Mr Yontz on behalf of CODL, but again he had no recollection of it, nor of its purpose. It was acknowledged for the appellant that COCL, like CODL, had no employees of its own but was dependent for any staff it might require on COSI. The effect of the agreement was therefore that one subsidiary of COSI, CODL, arranged with another subsidiary of COSI, COCL, that COCL would procure the construction of Loxley House in circumstances in which COCL would use the employees of COSI to undertake the necessary work, when CODL could have used exactly the same employees to do so without the need for any involvement on the part of COCL. The only justification we heard of the arrangement came from Mr Foster, who told us that COCL, which had arranged and entered into the contracts for the refurbishment of Trent House, had thereby gained a track record which gave it credibility with suppliers, whereas CODL, a newly formed company, had none.

We consider that explanation quite incredible. It is, furthermore, not consistent with the evidence. Although, for reasons which did not emerge, we did not see the building contracts, many of the documents we did see, such as design drawings, bore the name of the "client", or, more accurately, assumed client. Sometimes that name was COSI; commonly it was not, and the client was described as "Capital One", without elaboration. On one occasion at least the name used was of a non-existent company. We did not detect any document on which the name used was COCL, though do not rule out the possibility that there were some. It was quite clear that the professionals (architects and structural engineers) who had produced the documents did not discriminate between companies in the Capital One group, and that they had no evident wish to contract with COCL to the exclusion of any other Capital One company; on the contrary, as one would expect, all the indications are that those at arm's length to the group viewed "Capital One" as their true client and were relying on it to ensure payment. It is quite clear to us that none was concerned which of its subsidiaries might be used as the contractual vehicle.

The suggestion that CODL might have encountered some difficulty in persuading arm's length companies to contract with it (and any implied suggestion that the terms of any business it could secure would be less attractive than those available to COCL) do not stand up to scrutiny. No other reason for it was advanced, nor is it easy to imagine any plausible commercial explanation of the building agreement between CODL and COCL. Such profit as COCL made was as illusory as that CODL would have earned from the security agreement—it was funded by COSI, and eventually found its way back to COSI, as COCL's parent. The agreement is mentioned in the schedule of events prepared by Mr Foster, and from the outset was planned to be entered into one day after the grant of the lease by COCL to CODL, and on the same day as the transfer of the freehold reversion. Had the money been paid to Woolf (in exchange for a discount on the construction cost, for example) it would, we think, be difficult to regard the payment as one not made in the ordinary course of business, even though it might be a little unusual. Mr Foster accepted that payment to Woolf was not in contemplation since Capital One would not release such a large sum of money out of the group; and he conceded that the agreement between COCL and CODL had the dual advantages of generating an input tax claim while keeping the money within the group. We are entirely satisfied that there was no purpose behind the building agreement beyond the desire to create a large input tax claim in CODL's hands in March 2000.

We have mentioned that Capital One's need for flexibility was emphasised by several witnesses, and we accept that it was a genuine concern. In particular, we are satisfied that in the early stages of the project there was no certainty that COB or, later, COBEP would need all of the new space and that it might have been necessary to share it with the telecommunications business or even dispose of some or all of Loxley House. Although the submission to the REOB with which we have already dealt painted an optimistic picture of Capital One's, and particularly COB's, future United Kingdom operations, prudence alone demanded that some plan—the "exit strategy"—should be in place if those expectations were not realised. The need for flexibility was advanced as the reason for the various transactions in the freehold and leasehold interests in Loxley House and its site.

Again, we were almost entirely dependent on Mr Foster's evidence about those transactions. Mr Yontz and Mr Says, who were directors of one or more of the companies, Mr Brierley, as finance director of the UK part of the Capital One group, and Mr Longworth, responsible for Capital One's European property portfolio, were all conscious of the need for flexibility and for an exit strategy, but none could give detailed evidence about the reasoning behind the individual transactions. As we have indicated before, they relied on Mr Foster's expertise and accepted his advice that the documents they were asked to sign or to approve were appropriate, without any real understanding of the underlying purpose of any individual transaction. It is conspicuous, as Mr Peacock was to establish, that there were no minutes of the boards of the various companies recording the decisions that the transactions should be entered into nor, inevitably, the reasons for those decisions. It was quite clear to us that, since the relevant transactions were between group companies, they received little scrutiny.

Some of the transactions, if taken by themselves rather than as part of a scheme, do not seem to us to merit criticism. If one accepts that, as its name implies, CODL was intended to be a development company (and particularly so if it is regarded as a SPV designed to undertake only one development) it is, we think, unremarkable that it transferred its freehold interest to another group company, Worldscape, whose role was the long-term ownership of property, once the development was complete. The point was made, and we think it has some merit, that Worldscape's owning the freehold carried with it the benefit of retaining some control of the site within the Capital One group. If one assumes for present purposes that there was good reason to segregate the freehold and leasehold interests, we see no objection to the separate transfer of CODL's leasehold interest to COPI. Even disregarding the regulatory reasons to which we will come shortly, the grant of an underlease of Loxley House by COPI to COBEP (instead of the assignment of the existing lease or the transfer of the freehold) does not, of itself, appear to be objectionable—COPI, with a long leasehold at a peppercorn rent granted a much shorter lease at a significant rent, and COPI was thereby left with a saleable leasehold reversion. These transactions, despite their being between related companies, all had, or could reasonably be regarded as having had, some business purpose. We bear in mind too that the decision letter does not attack these parts of the scheme.

However, the same cannot, we think, be said of some other transactions. It is understandable that COCL rather than CODL entered into the put and call options with Prologis since at that time—October 1999—CODL had only just been acquired and was not yet registered for VAT. It is certainly possible (and, we think, probable) that it could have secured registration at that time in order to make a successful claim for input tax credit but we can accept that there was a justified reluctance to take an unnecessary risk. However, by March 2000, when the call option was exercised and the freehold was acquired, that risk no longer existed since CODL had procured VAT registration. The option agreement entitled COCL to have Prologis transfer the land to any other group company. If the decision had been taken—as the events we have described show is obviously the case—that CODL was the proper vehicle to hold the title to the land while the development proceeded, COCL could without any difficulty have required Prologis to transfer the freehold to it.

Alternatively, COCL could itself have taken and retained the land; as we have just explored, it was to procure the construction of Loxley House. CODL did not need to be involved for that purpose; nor was it necessary for both COCL and CODL to participate in the onward sale of the freehold or the creation of a lease to the eventual occupant of the premises. Either of the companies, alone, could have entered into all of those transactions.

Mr Foster told us that it had not occurred to him that COCL could have required Prologis to transfer the land to CODL, and it was for that reason COCL had taken the transfer before passing on the land to CODL. We do not accept that explanation. If Mr Foster had genuinely overlooked the possibility of arranging the transaction in that way, we are sure it would have occurred to Capital One's solicitors, who prepared all of the relevant documentation. But the explanation he offered is quite inconsistent with his own schedule of necessary events, which explicitly states that separate freehold and leasehold interests are to be created and are not to merge (which necessarily implies that the same company would own both titles since merger would otherwise not be possible). We are satisfied that the real reason why the possibility did not occur to Mr Foster is that it did not fit in with the tax saving scheme. The scheme required for its success that CODL should have both freehold and leasehold titles; that result could not be achieved if the land were transferred by Prologis directly to CODL.

That the splitting of the title was necessary for the success of the scheme does not, of course, necessarily mean that it had no other purpose. It was suggested for the appellant that the need for flexibility, the conditions which had to be satisfied in order to obtain the grant, and accounting requirements all contributed to the decision to separate the titles. We found all of these explanations unconvincing, but even if any of them had satisfactorily explained the creation of a leasehold interest, none came close to explaining why (independently of the tax saving scheme) there was any economic or business reason why CODL should hold both titles. In so far as the existence of two interests was relevant to it, flexibility could have been achieved if COCL had retained the freehold after granting the lease to CODL. Far from simplifying or facilitating the grant application, the manner in which the title was held created difficulties which eventually led to a novation of the offer. The accounting requirements of which mention (itself somewhat nebulous) was made might be of relevance to the long-term ownership of the site and the building; but it was evident that those requirements could have been met by the creation of a leasehold interest (in COBEP or, if necessary, in COPI with a sub-lease to COBEP) when CODL disposed of its interests at the conclusion of construction. We can detect no purpose to the creation in CODL of two separate titles to the land other than that it was necessary for the tax saving scheme. We are satisfied that it too was a device with no true economic purpose.

We might add in this context that we could also see no economic purpose to the agreement for lease between COCL and CODL which, it will be remembered, provided that either party could rescind the agreement at will and without penalty. Mr Peacock argued, in our view correctly, that the only discernible purpose of the agreement was to support CODL's renewed application for VAT registration—by this time urgently required if the scheme were to be implemented—by demonstrating an intention to commence trading; a copy of the agreement was submitted with the renewed application. It is difficult, if not impossible, to understand why, but for that requirement, two companies in common ownership and with common management might consider it necessary to enter into such an agreement which in reality bound neither to anything.

Mr Patchett-Joyce pointed out that the agreement created an interest capable of registration in accordance with the Land Registration Acts. That may be right, though no application for registration was made and it is quite obvious it had not occurred to Mr Foster (or, so far as one can tell, Capital One's solicitors) to make one. Even if registration had been effected, it could have been cancelled as readily as the agreement could have been rescinded. The resort to arguments of this kind—advanced at the very end of the hearing and with the obvious intention of demonstrating some substance, however slight, to the agreement—in our view fails in that objective and instead underlines its lack of true business purpose. Its existence also somewhat undermines Mr Foster's assertion that he had not thought of the possibility that CODL might take the land from Prologis, but is consistent with the other evidence that the transactions between COCL and CODL which immediately followed the purchase were carefully planned in advance.

The regulatory framework

We have hitherto touched only briefly on the regulatory requirements to which the Capital One companies were subject. We had oral evidence on the topic from Mr Navarette and Ms Mitchell, with Mr Sabel's report. Mr Navarette's evidence was supplemented by a letter from Derek Bush, another US attorney with experience in the field of US regulation of financial institutions.

Their evidence was that the restriction of COB and, later, COBEP to the activities of a limited purpose credit card bank as that expression is defined by US legislation was quite strictly applied. The advantage to COB of restricting itself to those activities was that it was subject, overall, to a less stringent regulatory regime. In particular, as we understood the evidence, provided there was adequate separation within the group of regulated activities, from each other and from non-regulated activities, detailed regulatory supervision by the Federal Reserve was not imposed on COFC and its subsidiaries. COFC is, we were told, subject to the same US regulation in respect of its overseas activities and its overseas subsidiaries as applies to its US activities and subsidiaries.

The essential requirement of the regulatory regime is that the core activity should be carried on in accordance with the terms of the restriction, and that the restriction should not be avoided or diluted by the carrying on of additional activities. There was a de minimis allowance of non-regulated activity, as a rule of thumb one seventh of turnover, which was disregarded, but that level was no more than a concession by the Federal Reserve, with no statutory foundation, which might not necessarily be allowed indefinitely or if it was perceived that the purpose of the restrictions was being circumvented. It would not therefore be appropriate, and consistent with the regulatory restrictions, for COB or COBEP to share its premises (save possibly to a very limited extent) with another company, whether or not within its group, which carried on activities which were not those of a limited purpose credit card bank. That would include the letting of part of Loxley House to the telecommunications company, COCI. Since it was not clear, when the project of constructing Loxley House was planned, to what extent it would be occupied by COB or COBEP, and to what extent by COCI—and the proportions might in any event vary as the different businesses developed—it was decided that COB could not be a vehicle for any of the transactions, save that it should be the final lessee of so much of Loxley House as it occupied for its own purposes. That approach was in any event consistent with what we understood to be Capital One's normal practice.

Mr Peacock sought to argue that the extent of the precautions the appellant needed to take in order to avoid offending the US federal and state regulatory requirements which affected the operations of Capital One group members was overstated. We can certainly accept that the approach Capital One adopted was cautious, and it may be that it went beyond what was strictly necessary. Nevertheless, we are satisfied from the evidence that it was Capital One's policy, quite independently of the tax saving scheme, to be cautious, and to avoid any risk of regulatory difficulty. We accept Mr Sabel's and Mr Navarette's evidence that there were doubtful areas which it was preferable to avoid, and we do not think it can be said that, even if exceptionally cautious, Capital One's approach was unreasonable or was not businesslike.

However, what was also apparent from the evidence was that, while there are clear restrictions on what can be done by a regulated body if it is not to risk jeopardising its own permission to carry on its chosen activities, what may be done by a subsidiary or associated company is much less clearly defined. None of the witnesses told us, in terms, that what a regulated body was not permitted to do itself it could likewise not do by means of a subsidiary. We can, nevertheless, accept that carrying on restricted or prohibited activities through a subsidiary might be seen as an avoidance of the restrictions applied to the parent, and we understand Capital One's reluctance to have any of the transactions undertaken by a subsidiary of COB or of COBEP.

By contrast, there was no evident obstacle in the way of an associated company which was not a subsidiary of the regulated body undertaking activities which, had they been carried on by the regulated body itself, would have prejudiced its authorisation. Neither COCL nor CODL is a subsidiary of COB or of COBEP; they are both subsidiaries of COSI which, like COB, is an immediate subsidiary of COFC. They each have, therefore, the same relationship with COB and COBEP. Thus if COCL's activities do not put COB's or COBEP's authorisation at risk, nor do CODL's, and vice versa.

We are satisfied from this evidence that there is substance in the appellant's contention that it would have been unwise to use COB or COBEP, or a subsidiary of either, as the vehicle for any of the transactions, save for the underlease once it had been decided that COBEP would occupy Loxley House. We do not, however, accept that there was any other regulatory impediment which was of relevance to the acquisition of the site and the construction and letting of Loxley House. In particular, there was no reason for the use of both CODL and COCL. We are satisfied that any regulatory objection would have been met, however cautious the approach to be adopted, if one only of those companies had been involved in the scheme. In other words, the use of both CODL and COCL cannot be justified on regulatory grounds, and we regard US federal and state regulation of the Capital One group and its individual members as a factor irrelevant to the decision that both should be involved.

Summary

We are satisfied that the structure of the series of transactions we have described was dictated by the desire to recover the input tax incurred on the acquisition of the site and the construction of Loxley House which, without the scheme, would have been irrecoverable. We have reached that conclusion after taking into account the artificial character of many of the transactions, with which we have dealt above, the purchase by Capital One from PwC (at substantial cost) of the generic scheme and its implementation, after some modification, by the clearly choreographed sequence of transactions; that they were choreographed is apparent not merely from what in fact happened but also from the schedule of events to which we referred when dealing with Mr McMahon's evidence. We accept, as we have indicated, that Capital One might have been required, and for good reasons, to change its plans, but we have been left in no doubt by the evidence that the intention throughout, absent unforeseen circumstances, was to follow a sequence carefully planned and dictated by Mr Foster whose objective, and only objective, was the saving of tax. It was similarly clear from the evidence that the directors of Capital One's participant companies did not direct their minds to the purpose of the various transactions, beyond satisfying themselves that they were proper, and that the controlling mind of all those participants was Mr Foster; thus his objective of saving tax was the companies' motivation. We add for completeness that we are satisfied that, absent unforeseen circumstances, the eventual supply of exempt leasehold and taxable freehold interests was at all times in Mr Foster's—and therefore CODL's—contemplation.

We will return to the individual transactions when we come to consider the possible reconstruction of the series. Before we do so we need to examine the Commissioners' "fourth decision", that Loxley House cannot be regarded as a building.

Is Loxley House a building?

It was a necessary part of the tax avoidance scheme that the disposal of the freehold of Loxley House by CODL to Worldscape should be a taxable supply. Without that, the intention to make both taxable and exempt supplies in the future—an essential requirement if regulation 101(2)(d) were to come into effect—would not be established. It was not, or at least might not be, enough that the car parking licences (if effective) were taxable supplies since they were of the land but not of Loxley House; thus they could not be onward supplies of the construction services, on which had been incurred by far the greater part of the input tax the scheme was designed to recover. (We do not, however, deal in this decision with the question whether there has to be identification of supplies received with supplies made.)

Ordinarily, dealings in land are exempt from VAT by the combined effect of section 31 of, and Group 1 of Schedule 9 to, the 1994 Act, unless the right to waive the exemption, conferred by paragraph 2 of Schedule 10 to the Act, has been exercised. CODL did not adopt that course (nor, if its view of the matter is correct, could it have done so since the conditions on which an election might be made were not present: there was no exemption to waive) but instead relies on the exception from exemption contained in Item 1 of Group 1 of Schedule 9 which, so far as material, provides that the exemption shall apply to:

"The grant of any interest in or right over land … other than—

(a) the grant of the fee simple in—

(i) …

(ii) a new building which is neither designed as a dwelling or a number of dwellings nor intended for use solely for a relevant residential purpose or a relevant charitable purpose after the grant …".

By Note (4) to the Group, "[a] building … is new if it was completed less than three years before the grant". It is accepted by the Commissioners that Loxley House was completed less than three years before the disposal of the fee simple to Worldscape and that it is neither designed nor intended for residential or charitable use. Thus if Loxley House is a "building" its sale to Worldscape cannot have been exempt but, by reason of its exclusion from Item 1 and correspondingly from the exemption, must have been a taxable supply. The Commissioners maintain, however, that it is not a building: this is the "fourth decision". It is, they say, an extension of Trent House, and that is not a building for the purposes of this provision.

Since the word "building" is not defined by Group 1 of Schedule 9, or the Notes to the Group, resort must be had, Mr Peacock argued, to the ordinary meaning of the word. He began with what was said by McCullough J in Customs and Excise Commissioners v London Diocesan Fund [1993] STC 369 at 380:

"Building is not easily defined. Ask the question: 'Is there an existing building?' before any work has begun and the definitional uncertainty is unlikely to matter. But at the second stage those who think of a building as something which is (or was, before it became ruinous) substantially complete will be more likely to say that no building remained, whereas those who think the word wide enough to embrace virtually any erection of substance are more likely to say that one did survive.

Where, as will ordinarily be so, it is beyond argument that a building was in existence before the work began, all that para (a) of note (9) requires is to consider the building as it was, to consider the end result and to ask whether the work done amounts to the conversion, reconstruction, alteration or enlargement of the original building in the sense in which those words are commonly used, or whether the end result is a new building. If a number of buildings existed before the work began the question will be whether the work amounted to the conversion, reconstruction, alteration or enlargement of one or more of them. The matter is one of fact and degree."

The circumstances of that case were rather different from those with which we are concerned, and the statutory provision which the judge was required to consider was the forerunner of Group 5 of Schedule 8 to the Act, which deals extensively with buildings, distinguishing between those which are, and those which are not, zero-rated. It, and particularly the Notes to the Group, differentiate between buildings properly so-called, and structures which, although built, are not to be regarded as buildings, but as enlargements of, or extensions or annexes to, existing buildings. Mr Peacock relied in particular on Note (16) to that Group, which reads:

"For the purposes of this Group, the construction of a building does not include—

(a) the conversion, reconstruction or alteration of an existing building; or

(b) any enlargement of, or extension to, an existing building except to the extent the enlargement or extension creates an additional dwelling or dwellings; or

(c) … the construction of an annexe to an existing building."

Mr Peacock illustrated his argument by the example of a new structure added to the end of a terrace of houses. If it constituted a new, self-contained house it would, both colloquially and technically, be regarded as a building; but if it were an extension or enlargement of the original end house, it would not be a building, but merely an extension or enlargement: that was what paragraph (b) of the Note made clear. He did not suggest that Loxley House was an enlargement of Trent House, but it was, he said, an extension.

The proper approach when applying Note (16), he said, is to identify any building which existed before construction started (Trent House, in this case) and then to ask oneself, once the works are complete, whether, as a matter of fact, the new structure is a building in its own right, or merely an extension of the existing building: see Customs and Excise Commissioners v Marchday Holdings [1997] STC 272. The question is to be asked at the time of supply, by comparing the pre-existing building and the structure in the course of construction, and the comparison must include an examination of their physical characteristics (taking into account their similarities and differences in appearance) and of their functional capabilities: Cantrell and another (trading as Foxearth Lodge Nursing Home) v Customs and Excise Commissioners [2000] STC 100, and Cantrell and another (trading as Foxearth Lodge Nursing Home) v Customs and Excise Commissioners (No 2) [2003] STC 486. Mr Peacock relied particularly on what was said by Lightman J in the first Cantrell decision ([2000] STC 100 at 103, as subsequently corrected):

"First the question is to be asked as at the date of the supply. It is necessary to examine the pre-existing building or buildings and the building or buildings in course of construction when the supply is made. What is in the course of construction at the date of supply is in any ordinary case (save for example in case of a dramatic change in the plans) the building subsequently constructed. Secondly the answer must be given after an objective examination of the physical characters of the building or buildings at the two points in time, having regard (inter alia) to similarities and differences in appearance, the layout, the uses for which they are physically capable of being put and the functions which they are physically capable of performing. The terms of planning permissions, the motives behind undertaking the works and the intended or subsequent actual use are irrelevant, save possibly to illuminate the potential for use inherent in the building or buildings."

Trent House is a large building, designed to be used as offices, and built, we were told, in about 1930. Loxley House is of similar size. It was, of course, built recently, using modern methods and materials. Between their respective sites runs a private roadway over which there is reserved a right of way in favour of the owners of the H-block. It appeared that it also provided a means of vehicular access to the northern side of Trent House. It was necessary to respect the right of way—so far as we are aware no attempt was made to buy it out—but it was also desired that Trent House and Loxley House should be linked. A simple covered bridge, passing over the right of way and connecting the two at one or two levels and providing only for pedestrian access, could have been constructed. Instead, and (as he told us) contrary to Mr Yontz's views, a much more substantial structure was decided upon.

It is tied into the fabric of Loxley House, though without cantilevering, and is borne on its own foundations. At ground floor level there is a void, allowing for vehicles to pass along the right of way and beneath the structure; above, there are four floors of meeting and waiting rooms as well as the connection providing for internal pedestrian communication between the two main structures. Part of the façade of Trent House has been removed in order to create the necessary openings into the link, but there is no structural integration of the link with Trent House, beyond what is necessary to ensure a weather-tight junction. Some, as we understood quite modest, structural work was necessary within Trent House to provide access to the new staircases constructed within the link, leading to the passageways which cross it; the floor levels of the link and of Loxley House are the same, but those of Trent House differ, and access from Trent House to the link is available at only two of the four levels.

Among the documents produced in connection with the REOB presentation were some drawings illustrating a "campus" approach to the development (one of the schemes assumed that both the H-block and Loxley House would be part of the end result) of which the features included not only continuity of external appearance but also integration of the interiors of the buildings in order to create the impression that they formed a single whole. We had only a set of photographs illustrating, externally, the finished construction of Loxley House and the link (the photographs also depicted Trent House) but our impression from the photographs, from the plans and from the evidence of, in particular, Mr Davis and Mr McMahon was that a casual observer, looking at the southern façade and knowing nothing of the history, would conclude fairly quickly that Trent House and Loxley House had been built at different times. It would be apparent that Trent House was significantly the older of the two. It would also be clear that the link bridge had been built with Loxley House. Nevertheless, the observer would, we think, accept that a serious attempt had been made to give the impression that the three structures were a unified whole. Although we had no photographs of the interior of any of the structures, the evidence suggested that (and despite the differences in floor levels) a conscious effort had been made to create the illusion that there was a single structure. The submission to the REOB, indeed, described this as an objective of the design; and we bear in mind that the planning application referred to Trent House with the intended new structures as if they were a single whole.

One can only speculate how the design might have been approached had there been no right of way which it was necessary to respect running between Trent House and the site of what was to become Loxley House. As it is, the link between the two is a substantial structure in its own right and is much more than a simple connecting bridge. We accept that the connection between the link and the fabric of Trent House has been kept to the minimum consistent with breaking through the exterior wall of Trent House in order to allow for pedestrian access and with the need to make the joint between the two weather tight, and that those parts of the façade of Trent House which were removed have been preserved in case a restoration should ultimately be undertaken; the appearance, however, is of rather greater integration, and we are satisfied that is what was intended.

The evidence also revealed a good deal of dependence by Loxley House on the services provided by Trent House, particularly the electrical and communications systems. It is not surprising, in our view, that there was integration of the telephone and computer circuits; that is, nowadays, commonplace in many organisations, including those located on several geographically separate sites. We can understand, too, why the decision was taken to integrate, at least in part, the electrical supplies and some other services, for reasons of resilience (although most of the common supplies were taken from the existing supplies in Trent House, a new electricity supply routed into Loxley House was thereafter shared with Trent House). The shared supplies have been routed through the link. The degree of integration of which we heard evidence supports the conclusion that those taking the decision to utilise the link in this way did not have foremost in their minds the possibility that the two structures might later not be in common occupation.

It appears to us that, although segregation of the services would be possible, it would be expensive and difficult. It was even more obvious (as the architects indicated in the correspondence and Mr Davis confirmed in his evidence to us) that removal of the link itself, though possible, would be a major operation, in terms of disruption and expense, and one not to be undertaken lightly. A more sensible course would be, we think, to block the access between the link and one of the buildings, and effectively incorporate it in the other.

At first Mr Peacock emphasised the use of the link to carry cables for the electricity supply and for the telephone and computer circuits, but he eventually accepted that these were not critical (he acknowledged that if there were no link and the cables had run beneath the roadway, Loxley House could not be considered an extension of Trent House) and that it was the presence of the link structure itself, together with their common purpose, which made one an extension of the other.

Although, as we accept, there was for some time, including part at least of the period while Loxley House was in the course of construction, doubt about the use to which it would ultimately be put, it was nevertheless clear that the most likely possibility, throughout, was that all of it would be occupied by COB (or, later, COBEP). That was the entire thrust of the REOB submission. The second most likely possibility was occupation in part by COB or COBEP and in part by Capital One's telecommunications business, but that and the other possibilities were, we find, considered primarily for reasons of financial safety. We are also satisfied that Capital One would not, for example, have begun the construction speculatively with the intention of letting Loxley House to an unconnected tenant. That possibility was considered only as a fall-back in case Capital One's projected expansion did not materialise. In the event, as we have recorded, the entirety of Loxley House is used for the purposes of COBEP's business, a purpose for which it is, and was always intended to be, entirely suitable, and in operational terms the entire complex of Trent House, the link and Loxley House form a single office.

This is, perhaps, a case close to the borderline but we have concluded that Loxley House should be considered as an extension to Trent House if the test to be applied is that set out in Note (16) to Group 5 of Schedule 8 to the 1994 Act. The intentional similarities of appearance, though not complete, the degree of structural integration and the intended and actual common use all point in that direction while the only factor which might be regarded as pointing to the opposite conclusion is the size of Loxley House, which is a good deal bigger than what might normally be considered an extension, though we do not think there is, in Note (16) or elsewhere, any express or implied limit on the size of an extension by reference to the size of the building of which it is an extension.

However, Dr Lasok argued that the Note (16) test is not appropriate. We agree, and have set out our comments in relation to Schedule 8 only in case it should be found elsewhere that we are wrong in that conclusion. Note (3) to Group 1 of Schedule 9 provides that Notes (2) to (10) and (12) to Group 5 of Schedule 8 shall apply to Group 1 of Schedule 9; by necessary implication the others, including Note (16), do not. Moreover, Note (16) is designed to exclude from zero-rating (among other things) extensions to existing buildings; even if it were incorporated into Schedule 9, it could be taken as no more than statutory guidance about what is excluded from zero-rating. It does not, and, we think, is not intended to, contain any definition of the word "building" nor is it, in our view, helpful in determining what is and what is not a building. We discard Note (16) as a guide and look elsewhere.

Dr Lasok referred us to article 4(3)(a) of the Sixth Directive, which reads as follows:

"(3) Member States may also treat as a taxable person anyone who carries out, on an occasional basis, a transaction relating to the activities referred to in paragraph 2 and in particular one of the following:

(a) the supply before first occupation of buildings or parts of buildings and the land on which they stand; Member States may determine the conditions of application of this criterion to transformations of buildings and the land on which they stand.

Member States may apply criteria other than that of first occupation, such as the period elapsing between the date of completion of the building and the date of first supply or the period elapsing between the date of first occupation and the date of subsequent supply, provided that these periods do not exceed five years and two years respectively.

'A building' shall be taken to mean any structure fixed to or in the ground."

The definition set out at the end of that passage is extremely wide—it is apt to include structures such as street furniture, though we doubt whether that was intended—and, if it is to be regarded as one of general application, must include a structure as substantial as Loxley House and, for that matter, the link. However, Mr Peacock objected to the adoption of the definition because it is used in the directive, not in defining buildings as such, but in determining who is to be regarded as a taxable person.

That is true, but even so the reasoning behind Mr Peacock's objection to the adoption of the article 4(3)(a) definition of "building" was not altogether clear. There is no hint in article 4, or for that matter in Item 1 of Group 1 of Schedule 9, that in the context of the provisions they contain any distinction is to be drawn between a building which stands entirely alone, and one which is an extension to another. Note (16) states that "[f]or the purpose of this Group" extensions and similar structures are excluded from treatment as "the construction of a building"; in our view the fact that it was felt necessary to include the Note suggests that, without it, extensions and the like might be regarded as buildings. It may be—as we have concluded—that Loxley House is an extension of Trent House and, were Schedule 8 relevant, it would be excluded from zero-rating, but this seems to us an immaterial consideration. Its being an extension for the purposes of Schedule 8 does not, in our view, preclude Loxley House from being a building for the purposes of Schedule 9. An educated layman, looking at Loxley House, and taking what was said by Lightman J in the passage from Cantrell which we have quoted into account, would describe it as a building, whether or not he also thought it was an extension of Trent House. In our view he would be right to do so, and we have concluded that it is a building for the purposes of the legislation. Accordingly the supply of the freehold by CODL to Worldscape is, assuming it is not discarded as offensive to Halifax principles, a taxable transaction.

Can the transactions be reconstructed?

The Commissioners' case on reconstruction—that is, the correct analysis of those transactions which are to be regarded as supplies for the purposes of VAT—is set out as the "third decision" which at this point we need to set out in full:

"Further, I have decided that the correct supply position in respect of the supply of the Kingspark site by Prologis and the supplies of construction work by the Arm's Length Contractors is that these are not properly treated as supplies to COCL, but rather as supplies to COB."

As we mentioned earlier in this decision, Dr Lasok contended that the Commissioners' approach to the case could not be justified because a reconstruction of the transactions, such as the tribunal undertook in Halifax (in order to reflect what it perceived to be the true economic supply), was not possible. Thus even if we were to conclude—as we have—that some of the transactions lacked economic substance, they are nevertheless still to be treated as supplies for VAT purposes with the result that CODL's claim must be met. He supplemented his more general argument with the specific comment that the decision letter proceeded on the false premise that there was ultimately to be a supply of an under-leasehold interest to COB or, alternatively, as the argument advanced, the COB VAT group—the former was incorrect as the supply was to COBEP, and the latter was impossible since a VAT group could not be a lessee.

We deal first with the more particular argument which, in our view, is without merit. The decision letter was written on the strength of information provided to the Commissioners by Mr Foster at a time when some of the transactions had yet to take place and when Capital One itself had not finally determined precisely what form they would take. We do not need to deal with the evidence on this point in detail, save to repeat the observation that there was uncertainty for some time whether the entirety of Loxley House would be used for financial services activities, and to comment that when the letter was written in June 2001 its author does not appear to have been made aware that COBEP had already taken over COB's UK branch activity, though whether Capital One or the Commissioners themselves are to blame for that lack of awareness is not apparent. The errors and omissions in the letter are of form rather than of substance and there cannot have been any real doubt in Mr Foster's mind what was intended. Even if there were, it seems to us that the appropriate test is not whether the arrangements can be reconstructed in precisely the manner advanced then, but whether they are capable of any reconstruction. It would be absurd if the Commissioners were held to a suggested reconstruction put forward, as here, part-way through the process, at a time when the taxpayer had the opportunity of making an otherwise insignificant change which would, if Dr Lasok's argument is right, defeat their objection to the claim for input tax deduction.

The sale by Prologis and the construction management agreement entered into by Woolf had obvious economic substance, from their perspective and from the perspective of the Capital One group. The transactions we have found to be lacking in economic substance and which, assuming the Halifax approach is right, are to be disregarded, are the grant of the lease by COCL to CODL, the subsequent transfer of the freehold reversion to CODL, the two car parking licences and the building agreement between CODL and COCL. Even though one of the car parking licences involved a company outside the Capital One group, we are satisfied that, from that company's perspective as well as CODL's, it had no economic purpose.

There may, however, be several ways of determining the true (or VAT-significant) supplies. The most obvious, taking into account those of the regulatory and other requirements which we have concluded to be material, is that COCL acquired the land, arranged for Woolf to construct Loxley House, then leased the building to COBEP, assigning its reversion thereafter to Worldscape. It may be that there was a sound economic reason for creating a leasehold interest, to be vested in COPI (as original lessee) in order that COPI could grant an underlease to COBEP—the evidence did not enable us to make a finding on the point though we have already suggested the possible sale by COPI of its intermediate interest. One could instead substitute CODL for COCL—in other words, disregard the transfer by Prologis to COCL and the lease and transfer of the freehold to CODL, and treat the true transaction as a transfer of the freehold from Prologis to CODL.

What is abundantly clear, however one approaches the matter, is that the involvement of both COCL and CODL had no economic purpose but was a pure tax avoidance device, and any reconstruction should eliminate one or the other of those companies from the chain. Thus although we differ from the Commissioners in detail, we agree with the thrust of the decision letter.

Nevertheless, we are not persuaded that the Commissioners' preferred reconstruction—that the supplies of the freehold by Prologis and of the construction services were to COB (one would now have to substitute COBEP)—can be supported. Were it not for the regulatory requirements that might be an appropriate conclusion, but we accept that Capital One would not have structured the transactions in that fashion, regardless of any tax saving scheme, because of the danger of infringing US regulatory requirements, which we regard as a legitimate business reason for its adopting a different structure. We are satisfied that COB and COBEP would not have taken more than an occupation lease.

Conclusions

We are satisfied that some of the transactions which were entered into in the course of acquiring the site and construction Loxley House on it had no purpose other than the avoidance of tax, and were entered into for no other reason. Thus, applying the principles determined by the tribunal in Halifax, and assuming that they are correct, those transactions are to be disregarded. It is possible to reconstruct the arrangements so as to reflect their true economic substance, but there is no single correct way of doing so, and we have been persuaded that the respondents' preferred reconstruction, that is of direct supplies to an exempt financial institution, would not have occurred, even in the absence of the tax saving plan. We conclude that Loxley House is a "building" for the purposes of Item 1 of Group 1 of Schedule 9.

We recognise that our findings may not be sufficient to determine the appeal finally, since the outcome of the Halifax reference remains unknown, and there may be a dispute about the method of reconstruction to be adopted, assuming the decision of the Court of Justice indicates that reconstruction is appropriate. Accordingly we give the parties permission to apply for the hearing to be continued for the resolution of any outstanding matters.

Both parties sought a direction for costs if successful. We have not, however, determined the appeal entirely in favour of either party, and there may also be matters yet to be resolved. We have concluded that we should not make any direction at this stage, but instead we reserve the costs of the appeal which (unless the parties agree otherwise between themselves) are to be the subject of a further direction to be made either at the conclusion of any further hearing or on application for the purpose by either party.

COLIN BISHOPP

CHAIRMAN

Release date:09/06/2004

MAN/01/0624

About BAILII - FAQ - Copyright Policy - Disclaimers - Privacy Policy amended on 25/11/2010