Cobalt Data Centre 2 LLP and Cobalt Data Centre 3 LLP v The Commissioners for HM Revenue and Customs:[2019] UKUT 0342 (TCC) [2019] UKUT 0342 (TCC)

UPPER TRIBUNAL

TAX AND CHANCERY CHAMBER

[2019] UKUT 0342 (TCC)

Case No UT/2018/0007, UT/2018/0008, UTJR/2017/003

Between

Cobalt Data Centre 2 LLP and Cobalt Data Centre 3 LLPAppellant

The Commissioners for HM Revenue and CustomsRespondent

Before

MR JUSTICE ZACAROLI

JUDGE JONATHAN RICHARDS

Ruxandu (instructed by the General Counsel and Solicitor to HM Revenue and Customs) for Respondents

Date 15 November 2019

Category: Tax

[1]The appellants (“CDC2” and “CDC3” separately and, together, the “LLPs”) are two limited liability partnerships formed under the Limited Liability Partnerships Act 2000. On 4 April 2011, they acquired, among other assets, an assignment of rights under a construction contract (the “Golden Contract”) entered into between, among others, Highbridge North Tyneside Developer One Limited (the “Developer”) and Highbridge North Tyneside Contractor One Limited (the “Contractor”). The Golden Contract related to construction works to be undertaken at the Cobalt Business Park (the “Site”) that was, from February 1996 to 18 February 2006, within an enterprise zone. As consideration the LLPs paid consideration (the “Price”) of £153,709,750 in the case of CDC2 and £109,754,500 in the case of CDC3.

[2]Before the appellants acquired their rights under the Golden Contract, that contract was amended and “change orders” issued by the Developer to the Contractor. On 1 April 2011 and 4 April 2011, the Developer made substantial advance payments to the Contractor. The appellants consider that these payments were made under the Golden Contract as advance payment for construction works to be undertaken, but HMRC dispute this. In due course, two data centres (“DC2” and “DC3” respectively and, collectively, “Data Centres”) were constructed on the Site with DC2 being completed to shell and core on 28 January 2013 and DC3 being completed to shell and core on 17 December 2012. However, efforts to find a tenant for the Data Centres have been unsuccessful. At no point since the appellants acquired their interests in the Data Centres have those Data Centres been let.

[3]The appellants contend that, by virtue of s863 of the Income Tax (Trading and Other Income) Act 2005 (“ITTOIA”), they are to be treated for income tax purposes as if they were partnerships. As a consequence, they submitted partnership tax returns pursuant to s12AA of the Taxes Management Act 1970 for the tax year 2010-11. In those tax returns, they claimed enterprise zone allowances (“EZAs”) which they contend arose on the expenditure they incurred acquiring rights under the Golden Contracts.

[4]The arrangements under which the appellants acquired their interests in the Data Centres were notified to HMRC under the disclosure of tax avoidance schemes legislation set out in Finance Act 2004, although neither party suggested that this was relevant to the question whether EZAs are available or not. HMRC opened enquiries into the partnership tax returns. Following completion of their enquiries, HMRC issued closure notices concluding that the appellants were not entitled to the allowances claimed. The appellants appealed to the First-tier Tribunal (Tax Chamber) against HMRC’s closure notices. The appellants also considered that, in denying the allowances that had been claimed, HMRC were acting contrary to their published practice which gave the appellants a legitimate expectation that EZAs would be available and they therefore also instituted judicial review proceedings.

[5]The appellants’ substantive appeals against HMRC’s closure notices have been transferred to the Upper Tribunal (Tax and Chancery Chamber) pursuant to Rule 28 of 3 the Tribunal Procedure (First-tier Tribunal)(Tax Chamber) Rules 2009 (the “FTT Rules”). The appellants’ claim for judicial review was also transferred to the Upper Tribunal by order of Ben Emmerson QC, sitting as a judge of the High Court, on 18 August 2017. The Upper Tribunal has directed that the appeal against the closure notices and the judicial review claims are to be heard together. The relevant parties to the transaction and their roles

[6]The principal parties to the transaction were as follows. The LLPs

[7]The LLPs were incorporated on 19 January 2011 and 15 March 2011 respectively. The LLPs obtained funds from a combination of investors’ subscriptions for membership interests and from loan finance (“Bank Winter Loans”) provided by Bank Winter & Co Aktiengesellschaft (“Bank Winter”). The LLPs used those funds to acquire interests in DC2 and DC3 from the Developer. The hope and expectation, both of the LLPs and investors, was that, since DC2 and DC3 were located on the Site that, until 18 February 2006, was part of an enterprise zone, the expenditure incurred on acquiring the “relevant interests” in DC2 and DC3 would qualify for EZAs under s296 of the Capital Allowances Act 2001 (“CAA 2001”). Moreover, it was hoped and expected that, because the LLPs would by virtue of s863 of ITTOIA be treated for income tax purposes like partnerships, members of the LLPs (as distinct from the LLPs themselves) would be able to benefit from the EZAs by setting them against other tax liabilities that they had. The Developer, the Contractor and Highbridge

[8]The Developer and the Contractor were both formed as special purpose companies within the Atmel group. Atmel subsequently sold those companies to the Highbridge group. Highbridge was, and is, an experienced commercial property developer which had a track record of developing properties situated in enterprise zones. Highbridge had experience of the construction and development of data centres on the Site since it had been involved in the development of the first data centre on the Site (“DC1”) which achieved practical completion in 2011. Harcourt Capital LLP, Taurus Asset Finance Limited and CDC Administration LLP

[9]Harcourt Capital LLP (“Harcourt Capital”) is a “structured finance house” which was formed in 2009. It, and its partners, have considerable experience in raising financing to fund investments and administering those investments once made. By November 2010 Taurus Asset Finance Limited (“Taurus”) had been approached by Highbridge to help to secure funding for what was to become DC2. Taurus, however, considered that it did not have the capacity to take on that task single-handed and approached Harcourt Capital to help. At the time, Harcourt Capital had no previous experience of enterprise zone investments or property investments more generally (and had not been involved in the financing of DC1). Taurus and Harcourt formed a joint venture (CDC Administration LLP) to organise and administer the Cobalt investments. 4 Bank Winter

[10]Bank Winter is an Austrian private bank. It provided finance for the transactions in the form of the Bank Winter Loans described above at [7]. It also provided more general banking facilities including bank accounts to parties to the transaction. Evidence and procedural matters

[11]HMRC did not rely on witness evidence. The appellants relied on evidence from the following witnesses:(1) Mark Fielding and Alison Brister, who are both partners at Harcourt Capital LLP;

(2) Piet Pulford, who is a director of the Developer and other companies in the Highbridge group;

(3) Douglas Smith MRICS, who is an executive Director of CBRE Ltd (“CBRE”) who gave expert evidence in relation to matters involving investment in enterprise zones as well as evidence in support of the Appellants’ claims for judicial review. (CBRE had given valuation advice to both the Developer and the LLPs in the course of the transactions that are the subject of this appeal. However, Mr Smith was not involved in the provision of that advice and HMRC did not challenge his ability to give independent expert evidence);

(4) Ian Watson FRICS, who is an Associate Partner at Ernst & Young LLP, and who gave expert evidence on valuation matters;

(5) Mark Baldwin, a solicitor and partner at Macfarlanes LLP;

(6) John Watson, a solicitor and, until his retirement, a partner in Ashurst LLP;

(7) Peter Waterman, a member of B&M Tax Accountants LLP.

[12]All the appellants’ witnesses were cross-examined with the exception of Mr Waterman who was in poor health at the time of the hearing and whose witness evidence was accepted without challenge.

[13]We found all the witnesses who gave oral evidence to be reliable and honest. We also found all the expert witnesses helpful and conscious of their duty to provide the Tribunal with dispassionate expert opinion.

[14]At the start of the hearing, Ms Nathan applied for permission to amend HMRC’s Statement of Case. We refused that application for reasons that we gave orally to the parties during the hearing and which we have not, therefore, set out in this decision. The statutory code setting out entitlement to EZAs as relevant to this appeal

[15]It was common ground that Parliament enacted the EZA regime in order to provide incentives for investment in areas perceived as disadvantaged. It did so by including within the regime that provides industrial buildings allowances (“IBAs”) for 5 “qualifying expenditure” incurred on industrial buildings, an entitlement to generous EZAs (described in the legislation as “initial allowances”) equal to 100% of that expenditure in relation to buildings located within enterprise zones. The IBA regime was, however, repealed for corporation tax purposes in relation to chargeable periods beginning on or after 1 April 2011 and for income tax purposes in relation to chargeable periods beginning on or after 6 April 2011. Since the EZA regime was an aspect of the IBA regime, that repeal meant that EZAs also ceased to be available.

[16]In a “straightforward” situation, s294 of CAA 2001 provided for IBAs to be available where “qualifying expenditure” is incurred on the construction of a building. Allowances under s294 would be available to the person incurring the construction expenditure. Section 272 excluded expenditure on the acquisition of land from the definition of “qualifying expenditure” with the result that sums spent to acquire land, as distinct from sums spent on the construction of a building, did not attract IBAs. If the relevant building was on a site in an enterprise zone, it would be an “EZ building” within the meaning of s298(2) of CAA 2001 and, provided the expenditure was incurred within the relevant time limit (discussed in more detail below) s299 would treat the expenditure as qualifying enterprise zone expenditure with the result that s305 of CAA 2001 would apply to confer entitlement to the generous 100% EZAs.

[17]Section 296 of CAA 2001 (when read together with the definition of “qualifying expenditure in s292) expanded the situations in which IBAs were available beyond the “straightforward” situation outlined at [16] above. It did so by providing that, in certain circumstances, a person who did not build a building “from scratch”, but instead purchased the building unused from a developer could obtain IBAs. Section 296 provided, so far as material, as follows: 296 Purchase of building which has been sold unused by developer(1) This section applies if— (a) expenditure is incurred by a developer on the construction of a building, and (b) the relevant interest in the building is sold by the developer in the course of the development trade before the building is first used.

(2) If— (a) the sale of the relevant interest by the developer was the only sale of that interest before the building is used, and (b) a capital sum is paid by the purchaser for the relevant interest, the capital sum is qualifying expenditure. … (4) The qualifying expenditure is to be treated as incurred by the purchaser when the capital sum referred to in subsection (2)(b) or (3)(b) became payable. 6

[18]Section 296 therefore dealt with a situation where a developer had incurred expenditure on the construction of a building and a person subsequently acquired the “relevant interest” in that building from the developer before the building had been used. In that case, if a capital sum was paid “for the relevant interest”, it attracted IBAs. The “relevant interest” for these purposes was defined in s286 as: the interest in the building to which the person who incurred the expenditure on the construction of the building was entitled when the expenditure was incurred.

[19]If the building in question was located in an enterprise zone, and provided the time limit in s298 of CAA 2001 was met, s300 of CAA 2001 would treat the “qualifying expenditure” given by s296 (i.e. the capital sum paid for the relevant interest) as qualifying enterprise zone expenditure that attracted a 100% initial allowance.

[20]In normal circumstances, a person commissioning construction work might not wish to pay their contractor the entire contract price in advance. However, in order to fall within s296, the parties considered that it was necessary for the Developer to pay the Contractor the entire contract price in advance with the LLPs then acquiring the Developer’s contractual right to have the building constructed (constituting the “relevant interest” in the building that the Developer owned). By the time of the hearing, the parties were agreed that this arrangement was capable of satisfying the requirements of s296. Indeed, they were agreed that all but one of the requirements of s296 were satisfied. The element in dispute was whether the LLPs paid capital sums “for the relevant interests” as required by s296(2)(b).

[21]As noted, s298 imposed a deadline with qualifying expenditure incurred after that deadline not attracting industrial buildings allowances. 298 The time limit for qualifying enterprise zone expenditure (1) For the purposes of sections 299 to 304, the time limit for expenditure on the construction of a building on a site in an enterprise zone is— (a) 10 years after the site was first included in the zone, or (b) if the expenditure is incurred under a contract entered into within those 10 years, 20 years after the site was first included in the zone.

[22]In the context of this appeal, s298 needs to be read together with s300 of CAA 2001 which provided as follows: 300 Application of sections 295 and 296 If –(a) expenditure is incurred on the construction of an EZ building, and

(b) all the expenditure is incurred within the time limit Any qualifying expenditure given by sections 295 and 296 in relation to that expenditure is qualifying enterprise zone expenditure 7

[23]Therefore, the combined effect of s296, s298 and s300 was that, in order for a purchaser of a relevant interest from a developer to obtain EZAs, the developer must have incurred expenditure on the construction of the building within the time limit set out in s298. In the circumstances of this appeal, the Developer incurred expenditure on the construction of the Data Centres on 1 April 2011 and 4 April 2011 when it paid the Contractor in advance to construct DC2 and DC3. Those dates fell more than 10 years after the Site was first included in an enterprise zone. It was common ground, therefore, that the LLPs who purchased the relevant interests in the Data Centres would obtain EZAs only if the Developer’s expenditure was incurred “under” a contract entered into in the period ending 10 years after the Site was first included in an enterprise zone. The relevant question in this case is accordingly whether the Developer incurred expenditure “under” the Golden Contract (which was entered into on 17 February 2006 just before the expiry of the relevant 10-year period).

[24]It is important to note that any EZAs arising under s296 of CAA 2001 could only benefit the initial purchaser of the Data Centres. If and when the LLPs came to sell the Data Centres, the person purchasing from the LLPs could not obtain the benefit of any EZAs.

[25]As we have noted, s296 of CAA 2001 provides that the LLPs are entitled to EZAs only to the extent that they paid capital sums “for the relevant interest” that the Developer sold. Section 356 of CAA 2001 deals with the situation where a capital sum paid “for” a relevant interest is “attributable” partly to some assets and partly to others with the broad effect being that EZAs are available only to the extent that, on a just and reasonable apportionment, the sum paid is attributable to assets qualifying for EZAs. We will address the detail of s356 later in this decision. The issues arising in this appeal

[26]During the course of the hearing, the parties produced an agreed list of the three principal issues arising on this appeal.

[27]Issue 1 is whether the expenditure that the Developer incurred as described at [2] above was incurred “under a contract” entered into before 19 February 2006 (being 10 years after the date from which the Site was designated as an enterprise zone) for the purposes of s298 of CAA 2001. This requires consideration of the following:(1) Were the terms of the Golden Contract sufficiently certain for the expenditure in question to be incurred “under” it or was it an “agreement to agree”?

(2) Was the Golden Contract rescinded as a result of the variations made on 4 February 2009 and 3 April 2009 and/or any one or more of the Change Orders issued on 20 November 2009, 1 April 2001 and 4 April 2001 respectively and/or construction of the Data Centres themselves?

(3) Was the construction of the buildings in question under the Golden Contract? 8

[28]Issue 2 is whether the LLPs were carrying on business with a view to profit for the purposes of s863 of ITTOIA when they acquired the “relevant interests”, in other words, on or before 4 April 2011 in the case of CDC2, and on or before 5 April 2011 in the case of CDC3.

[29]Issue 3 is whether the entirety of each capital sum was paid for the “relevant interest” within the meaning of s296 CAA 2001. This requires consideration of the following (leaving aside a sub-issue relating to the costs of fitting out the Data Centres, which HMRC abandoned in closing argument):(1) As regards the Developer’s obligation, as part of the terms on which the LLPs acquired their interests, to pay certain “Yearly Sums” to the LLPs: (a) Is any reduction in the amount of the qualifying expenditure to be made under s356 of CAA 2001 on the basis that part of the price paid is referable to rental support arrangements, in circumstances where HMRC accept that EZAs may, in principle, be available under s296 CAA 2001 in respect of parts of the purchase price paid by investors which are referable to reasonable rental support arrangements? If so, what is the reduction? (b) If the answer to Issue 3(1)(a) is yes, did the LLPs nevertheless have a legitimate expectation that HMRC would accept that the full purchase price paid by an investor (less the value of the land element) would qualify for capital allowances even if part of that price was referable to rental support arrangements (so long as, and to the extent that, such arrangements were not artificial within s10D of the Capital Allowances Act 1990 (“CAA 1990”), now s357 CAA 2001 (it being acknowledged that s357 CAA 2001 is not in issue in these proceedings))? (c) If the LLPs did have such an expectation, are HMRC permitted, on public policy grounds, to resile from the representations they made?

(2) As regards the effect of EZAs, is any reduction in the amount of the qualifying expenditure to be made on the basis that the value of the EZAs factored into the value of the relevant interest acquired by the LLPs was excessive? (For the avoidance of doubt HMRC’s position, as stated in paragraph 143(3) of their skeleton argument for the tax appeal, is that it is in principle permissible for the value of EZAs to be factored into the value of the relevant interest in the building). If so, what is the reduction?

(3) As regards developer payments and profits: (a) Is any reduction in the amount of the qualifying expenditure to be made on the basis that the Developer was obliged to pay fees to Harcourt Capital Investments Limited and CDC Administration LLP? If so, what is the reduction? 9 (b) To the extent that the Developer’s profit (if any) arises from capital sums that comprise both qualifying and non-qualifying expenditure, should there be a reduction in the amount of qualifying expenditure on this basis? If so, what is the reduction? Structure of this decision

[30]We will structure this decision as follows:(1) In Part A, we will explain, at a high level of generality, the transactions that were effected. In doing so, we will incorporate the facts set out in the parties’ Statement of Agreed Facts. Part A will not contain all our findings of fact since it will be necessary to make more detailed findings on specific aspects of the transactions when considering Issues 1, 2 and 3.

(2) In Parts B, C and D we will deal with Issues 1, 2 and 3 respectively. Part E addresses the judicial review claim. Those sections will contain detailed findings of fact relevant to the issues under consideration.

(3) Part F will contain our overall conclusion on the appeals and claims for judicial review. PART A – HIGH LEVEL DESCRIPTION OF TRANSACTIONS The entry into the Golden Contract and the subsequent variations to that contract

[31]The Site was, from February 1996 until 18 February 2006, within an Enterprise Zone pursuant to the Tyne Riverside Enterprise Zones (North Tyneside)(Designation) (No.1) Order 1996 (Statutory Instrument 1996 No.106).

[32]Prior to 2006 part of the Site had been used as a semiconductor manufacturing facility. Initially, that facility was owned and operated by Siemens but, following a downturn in the world price for computer memory chips, Siemens ceased production at the Site. Siemens eventually sold the facility to Atmel who carried on some production at the Site for a period.

[33]In 2006, Atmel realised that the enterprise zone at the Site would shortly be coming to an end and they took positive steps, including by engaging advisers and obtaining the assistance of Highbridge who were knowledgeable about EZAs, and with whom Atmel had a good relationship, to ensure that the ability to claim EZAs on future development of the Site would not cease. Atmel established the Contractor and the Developer as two special-purpose companies. On 17 February 2006, the Contractor and Developer (referred to as the “Owner” in the relevant contract), executed the Golden Contract which incorporated the conditions of the JCT Standard Form of Building Contract with Contractor's Design 1998 Edition (the “JCT Standard”) and made modifications to that JCT Standard. The Golden Contract was, therefore, entered into the day before the enterprise zone at the Site expired and formed part of arrangements that both the Developer and Contractor hoped would ensure that EZAs could still be claimed on future construction work on the Site. 10

[34]We will consider the terms of the Golden Contract in more detail when we address the parties’ arguments on Issue 1. One significant amendment to the JCT Standard embodied in the Golden Contract was that, while the JCT Standard envisages that the Contractor would be obliged to perform, and the Employer would be obliged to pay for, a single building project, the Golden Contract envisaged that the Developer would be entitled to require the Contractor to undertake any one of six building projects (defined as “Works Option 1” to “Works Option 6” in the Golden Contract).

[35]The six Works Options provided for in the Golden Contract varied significantly in size and scope. They also provided for works to be undertaken on different parts of the Site (described as Sites A, B and C). Works Option 6, for example, involved the construction of a single industrial unit on Site A for a contract price of £13,672,116. Works Option 2 involved the construction of an industrial unit to accommodate the manufacture of a 12 inch board on Site B and Site C for a Contract Sum of £183,000,000.

[36]The Golden Contract as originally drafted required the Employer to select a single Works Option by delivering a “Notice to Proceed”. In February 2009 and April 2009 the parties effected two variations to the Golden Contract (“Variation Agreement One and Variation Agreement Two”) which permitted the Employer to submit Notices to Proceed in respect of combinations of specified Works Options.

[37]Clause 12 of the Golden Contract permitted the Employer to make what were defined as “Changes in the Employer’s Requirements” that, very broadly, could operate to change the scope of what was to be built. The appellants’ case (which HMRC dispute) is that the Developer exercised its rights under Clause 12 on two occasions relevant to these appeals:(1) On 1 April 2011, the Developer’s agent, EC Harris LLP (“EC Harris”), issued “Change Order 2” which, on the appellants’ case, invoked Clause 12 and altered the scope of Works Option 1 so that, instead of involving the construction of a semiconductor manufacturing facility, Works Option 1 would involve the construction of a data centre. The Developer then served a Notice to Proceed, also on 1 April 2011 and the Contractor ultimately built DC2.

(2) On 4 April 2011, EC Harris issued “Change Order 3” which the appellants also assert invoked Clause 12, altering the scope of Works Option 1 so that it involved construction of a further data centre. The Developer then served a Notice to Proceed and the Contractor ultimately built DC3.

[38]On 1 April 2011, the Developer paid the Contractor £54,845,150 being the Contract Sum due in respect of the construction of DC2. On 4 April 2011, the Developer paid the Contractor £42,284,000 being the Contract Sum due in respect of the construction of DC3. The Developer borrowed the amounts due to the Contractor from Bank Winter by means of a short-term loan with the Developer repaying Bank Winter out of the proceeds of sale due to it under the Sale and Development Agreements as referred to below. The Developer paid the Contractor in advance for the construction of the Data Centres in order to meet the requirements of s296 of CAA 2001 as described above. 11 The LLPs’ purchase of interests in the Data Centres and the surrounding arrangements

[39]On 4 April 2011, CDC2 executed a document entitled “Sale and Development Agreement relating to the sale and development of DC2” (the “DC2 SDA”) with, among other parties, the Developer and the Contractor. This agreement was supplemented by a variety of other agreements dealing with financing and other matters (including a “Services Agreement” between, among others, the Developer, the Contractor and the relevant LLP, and various security documents). On 4 April 2011, in pursuance of the DC2 SDA, the Developer entered into a deed of assignment under which it assigned the benefit of its rights under the Golden Contract in respect of DC2 to CDC2.

[40]On 5 April 2011, CDC3 entered into a materially similar sale and development agreement relating to DC3 (the “DC3 SDA”). A materially similar suite of contractual documents to those relating to DC2 was entered into. Since these agreements were virtually identical to the counterparts to which CDC2 was party in relation to DC2 we will express most of our findings by reference to the DC2 suite of agreements and should be taken as making identical findings in relation to the DC3 suite of agreements unless the context provides otherwise. Moreover, unless it is necessary to distinguish between the DC2 and DC3 arrangements, we will refer to documents generically as, for example, an “SDA” or a “Services Agreement” without specifying whether they related to CDC2 or CDC3.

[41]Pursuant to clause 3 of the DC2 SDA, CDC2 agreed to pay the Developer consideration of £153,709,750. The consideration payable under Clause 3 of the DC3 SDA was £109,754,500.

[42]Both CDC2 and CDC3 raised 30% of the consideration payable under the SDA (i.e. £46,112,925 in the case of CDC2 and £32,926,350 in the case of CDC3) from proceeds contributed by their members by way of subscription for membership interests. Both CDC2 and CDC3 raised the remaining 70% of the purchase price (£107,596,825 in the case of CDC2 and £76,828,750 in the case of CDC3) via the Bank Winter Loans. Those loans were “cash collateralised” by means of a complicated network of security documents including a requirement on the Developer and Contractor to pay sums into a “Restricted Account” that was charged as security for the Bank Winter Loans with the result that, for so long as the Bank Winter Loans were outstanding, Bank Winter could be confident that it would hold sums, charged as security, that were sufficient to repay the entirety of those loans. The interest rate of 0.5% reflected the fact that Bank Winter was taking little, if any, credit risk in advancing those loans.

[43]The Services Agreement was concerned in large part with(i) the making of arrangements for letting the relevant Data Centre to a tenant and

(ii) the allocation of risk of a delayed letting or of a letting at rent lower than was hoped for. The Developer agreed to market the Data Centres and seek to obtain tenants for them, although the LLPs retained the ultimate right to decide whether to grant a lease to a particular tenant or not. 12

[44]Until a tenant was found, the LLPs would have no income, but would have to meet ongoing expenses (primarily interest on the Bank Winter Loans, but also ground rent on the Site and other expenses). Moreover, even once a tenant was found, the LLPs were exposed to the risk that the tenant would pay a rent lower than had been hoped. Pursuant to the Services Agreement, the Developer agreed to share in, or assume, certain of these risks by agreeing to pay “Yearly Sums” to the LLPs. The Developer had the obligation to pay Yearly Sums for a maximum of 15 years. During the period during which they were payable, Yearly Sums covered the following items (in addition to others which we will address in more detail later in this decision):(1) Until a Data Centre was let, the Yearly Sums were sufficient in amount to enable the LLP to meet its expenses (primarily interest on the Bank Winter Loan and ground rent).

(2) If a Data Centre was let to a tenant at a rent below a defined “Target Rent”, the Yearly Sums were sufficient to “top up” the rent that the LLP received to the Target Rent level.

[45]The structure of the arrangements meant that there was an in-built commercial need for the Bank Winter Loan to be refinanced once a tenant had been found. That commercial need arose because, once a tenant had agreed to take a lease of a Data Centre, the Data Centre would need to be fitted out to the tenant’s specifications. Clause 6 of each SDA imposed an obligation on the Developer to ensure that this fit-out took place. Some £36,167,000 had been earmarked to pay for this fit out. However, that sum had been paid into the Restricted Account which was subject to security interests that would not be released until the Bank Winter Loans were repaid. Therefore, if the Contractor was to receive sums needed to enable it to perform the fit-out, the Bank Winter Loan needed to be repaid with the proceeds of a new loan so that sums could be released from the Restricted Account. If the refinancing did not take place so that Bank Winter released sufficient funds to enable the Contractor to be paid, the Developer faced the unpalatable prospect of having to use its own resources to fund the fit-out.

[46]The Services Agreement required the Developer to use reasonable endeavours to procure a refinancing of the Bank Winter Loan with a “Second Loan” once a tenant had been found for a Data Centre. The Services Agreement recognised that the Second Loan might not be sufficient to refinance the Bank Winter Loan completely and, to the extent that there was a shortfall, the Developer was obliged to make an interest-free subordinated loan (a “Subordinated Loan”). At an LLP’s election, that Subordinated Loan could be converted into a membership interest in the LLP. The constitution of the LLPs

[47]The LLPs’ constitutions were set out in their Members’ Agreements.

[48]Those constitutions recognised that, if the LLPs disposed of any interest in the Data Centres prior to 7 years after the LLPs’ acquisition of the Data Centres, the entirety of the EZAs would be clawed back but that after 7 years, it might be possible to assign a lesser interest in the Data Centres (for example by granting a long lease of them) 13 without triggering a clawback of allowances. To deal with this issue, the Members’ Agreements provided in clauses 13.13 and 13.14:(1) An outright sale of the LLPs’ entire interest in the Data Centres (which would, whether made before or after the relevant 7-year period trigger a clawback of allowances) required the unanimous consent of all members of the LLPs.

(2) A sale of a lesser interest in the Data Centres in the first 7 years would similarly also require the unanimous consent of all members.

(3) A sale of a lesser interest in the Data Centres after the first 7 years could be made provided a “Super Majority” of Ordinary Members (broadly 75% of those present and voting on the resolution) agreed to it.

[49]Clause 11.5 of the Members’ Agreement provided that, if the LLP had a surplus, initial investors in CDC2 (i.e. those who acquired their interests in April 2011, but not any members admitted subsequently, such as the Developer following a capitalisation of its Subordinated Loan) had a priority entitlement (the “Priority Entitlement”) to share in that surplus up to an amount equal to 20% of the Price that CDC2 paid to acquire DC2. PART B - DISCUSSION OF ISSUE 1

[50]HMRC contend that the sums the Developer paid to the Contractor were not paid “under” the Golden Contract for the purposes of s298 of CAA 2001. Their submissions fall into three parts:(1) The Golden Contract was rescinded, so that the expenditure on DC2 and DC3 was necessarily pursuant to a new contract.

(2) Even if the Golden Contract was never rescinded, it was never performed and the construction of DC2 and DC3 was “under” a contract other than the Golden Contract.

(3) The terms of the Golden Contract were not sufficiently certain for the expenditure incurred on the construction of the Data Centres to be incurred “under” it.

[51]Before addressing these submissions in turn, we first set out in greater detail the relevant provisions of the Golden Contract. The Golden Contract

[52]The core obligation of the Contractor under the Golden Contract was (per Article 1), in addition to assuming responsibility for the design of the “Works” already carried out, to “complete the design and carry out and complete the construction of the Works”. This was in consideration for the “Contract Sum or such other sum as shall become payable hereunder…” (Article 2).

[53]The “Works” were defined as: 14 the design, construction and commissioning the Employer wishes to obtain for the Works Option stated in the Notice to Proceed which for the avoidance of doubt shall either be Works Option 1 Works Option 2 Works Option 3 Works Option 4 Works Option 5 or Works Option 6 and referred to in the Employer’s Requirements and the Contractor’s Proposals for that Works Option … and any changes made to these works in accordance with this Contract

[54]Each of the “Works Options” was separately defined in broad terms by reference to a specific type of building. Works Option 1, for example, was defined as: The design, construction and commissioning work comprising an industrial unit to accommodate the manufacture of an eight inch board on Site C for which the Employer has issued to the Contractor its requirements (hereinafter referred to as the Works Option 1 Employer’s Requirements).

[55]The “Works Option 1 Employer’s Requirements” were defined as “the documents referred to in Appendix 14 as Ref: ER WO1”. (The term “Employer’s Requirements” was itself defined as the document referred to in Appendices 14 to 19, as the case may be, “setting out the requirements of the Employer in relation to each Works Option”).

[56]Documents “ER WO1” comprised a set of terms with appendices containing technical specifications and quality standards. By clause 1.2.1 of the terms, “the intention is that the Contractor shall provide a fixed lump sum price for the works described within this document which comprise the construction of an 8 inch fab semi-conductor manufacturing facility, together with all associated hardstandings and external works.” Clause 1.3.1 described the scope of works by reference to the same semi-conductor manufacturing facility.

[57]Appendix 14 also defined the Contract Sum for Works Option 1 as £102,500,000.

[58]The other five works options provided for works to be undertaken on different parts of the Site (referred to as Sites A, B and C in the Golden Contract). In summary:(1) Works Option 2 involved the construction of an industrial unit to accommodate the manufacture of a 12 inch board on Site B and Site C for a Contract Sum of £183,000,000.

(2) Works Option 3 involved the construction of an office park on Site A for a Contract Sum of £70,000,000.

(3) Works Option 4 involved the construction of a light industrial business park on Site A for a Contract Sum of £22,353,714.

(4) Works Option 5 involved the construction of a mixed use office and light industrial business park on Site A for a Contract Sum of £50,000,000.

(5) Works Option 6 involved the construction of a single industrial unit on Site A for a Contract Sum of £13,672,116.

[59]Clause 23A of the Golden Contract required the Developer to serve upon the Contractor a written “Notice to Proceed” stating(1) when the Contractor should 15 proceed with the Works and

(2) which Works Option the Contractor was to carry out and complete.

[60]Although no deadline was set for the service of a Notice to Proceed under Clause 23A, it was common ground between the parties that the Developer was obliged to give such a notice and to select a Works Option.

[61]The requirement in the Golden Contract that the Developer must select a single Works Option was subsequently varied as follows:(1) On 4 February 2009, the Developer and the Contractor entered into “Variation Agreement One”. This agreement varied the Golden Contract by adding a new Clause 23A.3 that permitted the Developer to give a Notice to Proceed in relation to both Works Option 2 and Works Option 3.

(2) On 3 April 2009, the Developer and the Contractor entered into “Variation Agreement Two” that permitted the Developer to give a Notice to Proceed in relation to both Works Option 1 and Works Option 3.

[62]On 20 November 2009, the Developer’s agent, EC Harris LLP (“EC Harris”) issued “Change Order 1”. This referenced Works Options 1 and 3 and required the Contractor to undertake the design and construction of a data centre with an area of 4,262 m2 of net technical space. Following issue of Change Order 1, DC1 was constructed.

[63]On 1 April 2011, EC Harris issued “Change Order 2” on behalf of the Developer that ultimately resulted in DC2 being constructed. Change Order 2 was expressed to relate to Works Option 1 and read, so far as material, as follows: Under Clause 12 of the contract between [the Developer and the Contractor] dated 17 February 2006 we have been authorised by the Employer to issue the following instructions. Undertake the design, construction and commissioning of 1 no. Data Centre (“DC2”) totalling 3,360 square metres net technical space together with support facilities and enclosed plant. The Works will include drainage, external works and services all in accordance with the Employer’s Requirements ref. Draft Version March 2011 and Appendices listed therein. For the sum of £54,845,150.00 exclusive of VAT. Shell & Core is to proceed from the date of a Notice to Proceed under the contract The Date of Possession shall be to be confirmed. [sic]

[64]Also on 1 April 2011, EC Harris issued what was described as a Notice to Proceed in respect of Works Option 1 which stated as follows: Works Option 1 In accordance with the contract between [the Developer and the Contractor] dated 17 February 2006, we are authorised by the employer 16 to instruct you to proceed with the works in accordance with Change Order No.2 dated 1st April 2011 reference 010/20320.

[65]On 4 April 2011, EC Harris issued “Change Order 3” that ultimately resulted in DC3 being constructed. Change Order 3 also referenced Works Option 1 and was in materially identical terms to Change Order 2 except that it required the Contractor to: Undertake the design, construction and commissioning of 1 no. Data Centre (“DC3”) totalling 2,400 meters net technical space…. For the sum of £42,284,000 exclusive of VAT

[66]Also on 4 April 2011, EC Harris provided the Contractor with a Notice to Proceed that referenced Works Option 1 and required the Contractor to proceed with the works “in accordance with Change Order No. 3 dated 4 April 2011”.

[67]Both Change Order 2 and Change Order 3 required the Contractor to proceed with Works defined in the “Employer’s Requirements ref. Draft Version” (with Change Order 2 referencing a document dated “March 2011” and Change Order 3 referencing a document dated “April 2011”).

[68]Neither the “March 2011” nor the “April 2011” document was produced in evidence. We were, however, shown two documents both dated January 2012 and entitled “Employers Requirements Cobalt DC2 – Contract Version” and “Employers Requirements Cobalt DC3 – Contract Version”. These later documents set out employer’s requirements for the purposes of a contract between the Contractor and its sub-contractors (in which the Contractor was the “Employer” and the relevant subcontractor was the “Contractor”).

[69]The LLPs contended that it was to be inferred that the requirements in the later document were materially similar to the (missing) draft Employer’s Requirements referenced in Change Order 2 and Change Order 3. The only evidence in support of that inference, however, was that of Mr Pulford. While he said that he would expect the requirements set out in the “contract versions” to be materially the same as those in the drafts, he had little day-to-day involvement in the detail of those documents and was unable to give direct evidence as to the contents of the draft employer’s requirements referred to in Change Order 2 and Change Order 3. We do not need to decide what, precisely, was in the documents, however, as we are prepared to assume that the Data Centres as built reflected the requirements in them and, in any event, HMRC’s case is that either the terms of the contract under which the Data Centres were built or the Data Centres in fact built, were so different from what was required to be built under Works Option 1 (prior to the Change Orders). It is accordingly unnecessary to consider whether there was any material difference between the Employer’s Requirements referenced in the Change Orders and the Data Centres as built. (1)(1) - Rescission

[70]HMRC contend that the Golden Contract was rescinded either by the Change Orders or by the fact of construction of DC2 and DC3, since what was agreed to be built, or actually built, was so radically different from that which was required to be 17 built under the Golden Contract as to go to the root of the Golden Contract. (HMRC had initially contended that Variation Agreement One and Variation Agreement Two also operated as a rescission of the Golden Contract, but that was abandoned in closing argument.)

[71]The LLPs’ response is twofold. First, they argue that the Change Orders were given pursuant to clause 12 of the Golden Contract. Accordingly, no question of rescission can possibly arise, as the changes to the design and specification of the building to be constructed under Works Option 1 were the result of the operation of the contractual mechanism in the Golden Contract. Second, even if the Change Orders were not given pursuant to Clause 12, then the intention of the parties at the time of the Change Orders was to effect an amendment to the Golden Contract, not to rescind it. Whether the Change Orders were made pursuant to Clause 12 of the Golden Contract

[72]Clause12.2.1 of the Golden Contract permitted the Developer (in its role as the “Employer” pursuant to that contract), to make changes to the works that the Contractor was obliged to perform. Taking into account the amendments that the parties made to the JCT Standard, that clause provided as follows: 12.2.1 The Employer may subject to the proviso hereto and to clause

12.2.2 [which is not relevant in this appeal] issue instructions effecting a Change in the Employer's Requirements. No Change effected by the Employer shall vitiate this Contract. Provided that the Employer may not effect a Change which is, or which makes necessary, an alteration or modification in the design of the Works without the consent of the Contractor which consent shall not be unreasonably delayed or withheld.

[73]Clause 12.2.1 was therefore engaged by a “Change in the Employer’s Requirements”. That term was defined in Clause 12.1 of the Golden Contract which, taking into account the amendments that the parties made to the JCT Standard, read as follows: 12.1 The term ‘Change in the Employer’s Requirements’ or ‘Change’ means:12.1.1 a change in the Employer’s Requirements which makes necessary the alteration or modification of the design, quality or quantity of the Works, otherwise than such as may be reasonably necessary for the purposes of rectification pursuant to clause 8.4, including: .1.1 the addition, omission or substitution of any work; .1.2 the alteration of the kind or standard of any of the materials or goods to be used in the Works .1.3 the removal from the site of any work executed or materials or goods brought thereon by the Contractor for the purposes of the works other than works materials or goods which are not in accordance with this contract. 18

12.1.2 the imposition by the Employer of any obligations or restrictions in regard to the matters set out in clause 12.1.2.1 to 12.1.2.4 or the addition to or alteration or omission of any such obligations or restrictions so imposed or imposed by the Employer in the Employer’s Requirements in regard to: .2.1 access to the site or use of any specific parts of the site, .2.2 limitations of working space, .2.3 limitations of working hours (save where the Employer requires the Contractor pursuant to clause 16.6 or 16.7 to remedy any defect shrinkage or other fault outside the normal working hours of the Contractor), .2.4 the execution or completion of the work in any specific order.

[74]Neither Change Order 1 nor Change Order 2 fell within Clause 12.1.2 (and the appellants did not seek to argue that they did) since those Change Orders did not vary the way in which the Contractor was to discharge its duties (of the kind set out in Clause 12.1.2.1 to Clause 12.1.2.4). The question, therefore, is whether those Change Orders fell within Clause 12.1.1. To do so, they needed to(i) set out a change in the Employer’s Requirements which

(ii) “makes necessary the alteration of modification of the design, quality or quantity of the Works”.

[75]The LLPs’ argument is relatively simple. They contend that there is no limit on the nature or scope of the changes in the Employer’s Requirements that can be made pursuant to clause 12.1. Clause 12.1.1. expressly includes, within such changes, “the addition, omission or substitution of any work”. Accordingly, for example, the Developer was entitled (subject only to the Contractor’s right reasonably to withhold consent in the case of changes which necessitated an alteration or modification in the design of the Works) to require the Employer’s Requirements to be wholly replaced with different requirements, for a completely different building to that specified within the original Employer’s Requirements.

[76]We do not accept this argument. In our judgment, there is a limit on the nature and scope of the changes that can be made pursuant to clause 12.1, namely that those changes cannot effect a change in the definition of the Works. Thus, for example, in relation to Works Option 1, the changes permitted by clause 12.1 in the Employer’s Requirements for that Works Option do not extend to changing the building to something other than “an industrial unit to accommodate the manufacture of an eight inch board on Site C.”

[77]First, clause 12.1 makes a clear distinction between the Works and the Employer’s Requirements and contemplates changes being made in the Employer’s Requirements, not the Works. It is true that it contemplates such changes might affect the Works, however, only to the extent that they necessitate alteration or modification of the “design, quality or quantity” of the Works. In other words, it does not contemplate a change in the Employer’s Requirements necessitating a change in the definition of the 19 Works themselves. The wide words relied on by the LLPs (“addition, omission or substitution of any work”) specifically apply to the changes in the Employer’s Requirements, not to the “Works”.

[78]Second, the respective definitions of the “Works” and the “Employer’s Requirements” demonstrate that the latter exist as part of, and are subordinate to, the former. Specifically, the Employer’s Requirements are defined as the requirements of the Employer “in relation to the relevant Works Option”, and in the definition of Works Option 1, they are the requirements issued by the Employer for the construction of an “industrial unit to accommodate the manufacture of an eight inch board on Site C”.

[79]Third, there is support for this conclusion in the fact that the “Works”, as a defined term, are the design, construction and commissioning the Employer wishes to obtain for the “Works Option” stated in the “Notice to Proceed”. Accordingly, until the Notice to Proceed is issued, there are no “Works” which can be the subject of any change under clause 12.1 as the Notice to Proceed is required to specify which Works Option the Contractor is required to carry out and complete. The contractual scheme therefore proceeds on the basis that the question whether there is an “alteration or modification” in the design, quality or quantity of the “Works” can only be answered once the Employer has given a firm instruction to proceed with a particular Works Option.

[80]The LLPs also contended that since the “Employer’s Requirements” are defined as (in each case) the relevant document (for example, “ER WO1” in the case of Works Option 1), and (as we have noted above) ER WO1 itself states that the works comprise “the construction of an 8 inch fab semi-conductor manufacturing facility”, a change to the document to replace those words with reference to a Data Centre is permitted under Clause 12.1. Ingenious though that argument is, we reject it. Those words exist, in our judgment, in order to identify the Works Option to which the Employer’s Requirements related but are not operative parts of the Employer’s Requirements for that Works Option. Using the language of the definition of Employer’s Requirements, the description of the Works Option which appears in the relevant document is not something which “[sets] out the requirements of the Employer in relation to [that] Works Option”. Accordingly, they are not the intended subject-matter of a change permitted under Clause 12.1.

[81]Mr Kosmin, for HMRC, made two further submissions. First, he submitted that the fact that Clause 12.1 applies where there is “a change in the Employer’s Requirements” necessarily limits the scope of permitted changes to part of, but not all of, the Employer’s Requirements. He said that, if changes to the whole of the Employer’s Requirements were permitted by Clause 12.1, then the parties would have referred to a “change of the Employer’s Requirements”.

[82]Second, he submitted that commercially unreasonable results would follow if the scope of the power in Clause 12 is as broad as the appellants argue it to be. He relied on Schuler AG v Wickman Machine Tools Sales Ltd

[1974] AC 235

where Lord Reid said: 20 … the fact that a particular construction leads to a very unreasonable result must be a relevant consideration. The more unreasonable the result, the more unlikely it is that the parties can have intended it, and if they do intend it, the more necessary it is that they make their meaning clear.

[83]Mr Kosmin noted that Clause 26.1 of the Golden Contract deals with the situation where the Contractor incurs loss or expense due to regular progress of the Works being affected by, for example, a Change in the Employer’s Requirements. In that situation, the Contractor is entitled to have a “fair and reasonable amount” added to the Contract Sum but only if it first makes a “fully reasoned analysis stating and showing that [the Contractor] has incurred or will incur direct loss and/or expense”. That fully reasoned analysis has to be made no later than 16 days after it has become, or should reasonably have become apparent, that the regular progress of the Works would be affected. Therefore, Mr Kosmin argued, if a Change in the Employer’s Requirements permitted the Developer to make fundamental changes to the very nature of the Works, the Contractor could be left in the invidious position of having just 16 days to provide the “fully reasoned analysis” required by Clause 26.1 in the context of a completely new set of Works. He suggested that was such an unreasonable result as to engage the principle in Schuler v Wickman and demonstrate that the parties could not have intended the Developer to change completely the nature of the Works under the guise of serving a notice of a Change in the Employer’s Requirements.

[84]We are not persuaded by either argument. We do not find the parties’ use of the word “in” as opposed to “of”, with reference to the Employer’s Requirements, in clause 12.1, to be particularly illuminating as to their intention on the question we have to determine. So far as Mr Kosmin’s second point is concerned, the difficulty is that even if the concept of a Change in the Employer’s Requirements does not embrace a change in the nature of the Works themselves, Clause 12.1 plainly gave the Employer the right to insist on extensive changes. If the Employer chose to make those extensive changes, the Contractor ran the risk of having to provide its analysis under Clause 26.1 within 16 days. That was simply the bargain it had struck. Understood in those terms, the requirement to provide a “fully reasoned analysis” under Clause 26.1 sheds no light on the scope of changes that the parties considered were permitted under Clause 12.1.

[85]Nevertheless, for the reasons we have identified above, we conclude that neither Change Order 1 nor Change Order 2 was made pursuant to clause 12 of the Golden Contract. It necessarily follows that the Change Orders resulted in a variation to the terms on which the parties contracted. We turn, therefore, to consider whether (as the LLPs submit) that variation constituted an amendment to the Golden Contract, or whether (as HMRC submit) it resulted in the rescission of the Golden Contract. Rescission: the law

[86]The essential question is whether the expenditure on the Data Centres was incurred “under” a contract entered into within ten years from the date the site was first included in the enterprise zone, within the meaning of s.298. In response to questions from the Tribunal, the parties indicated that they were broadly in agreement that the phrase “under a contract entered into within those 10 years” in s.298 was not to be given any 21 particular meaning, according to principles of statutory interpretation, by reason of its statutory context. Instead, the question was to be answered by reference to the common law principles which determine whether two parties to a contract had, by reason of a subsequent agreement between them, rescinded that contract or merely amended it. (We describe the parties as “broadly” in agreement because there were some areas in which the parties appeared to depart from this approach, as we explain further below).

[87]So far as the common law of contract is concerned, HMRC contend that the question whether a subsequent agreement is a variation, or a rescission, of the original contract depends on the magnitude of the changes made to the contract. The LLPs, on the other hand, contend that the question is to be determined solely by reference to the intention of the parties.

[88]Both sides relied on Morris v Baron

[1918] AC 1

. The backdrop to this case was s4 of the Sale of Goods Act 1893 which provided that contracts for the sale of goods having a value of £10 or above could not, in certain circumstances, be enforceable unless evidenced in writing. A supplier and a customer entered into a written agreement for the sale of goods worth £800 and a dispute arose in relation to that agreement. The parties agreed to compromise that dispute by way of an oral agreement.

[89]The House of Lords concluded that a contract evidenced in writing as required by the Sale of Goods Act may be impliedly rescinded by a subsequent oral agreement (itself unenforceable by reason of its non-compliance with the statute), where there was a clear intention to rescind as distinguished from an intention to amend. Of particular relevance are the passages in the speeches in that case which consider what is needed to effect a rescission, as opposed to an amendment.

[90]At page 18, Viscount Haldane, having expressed the view that there was no reason why parol evidence may not be admissible to prove rescission of a contract as much as to prove variation, said: What is, of course, essential is that there should have been made manifest the intention in any event of a complete extinction of the first and formal contract, and not merely the desire of an alteration, however sweeping, in terms which still leave it subsisting.

[91]Lord Atkinson, at page 31, said: … Moreover rescission of a contract, whether written or parol, need not be express. It may be implied, and it will be implied legitimately where the parties have entered into a new contract entirely or to an extent going to the very root of the first [and] inconsistent with it.

[92]At page 33, however, Lord Atkin concluded as follows: It is quite impossible, in my opinion, to reconcile the agreement of April 22, 1915, with that of September 24 previous. With the exception already pointed out as to price, they are in conflict in all those material and fundamental provisions which go to the root of each of them. It is, I think, impossible to arrive at any rational conclusion as to the meaning, aim, and effect of this new arrangement other than this, that it was the 22 clear intention of both the appellant and the respondents to put aside, in their future dealings, the original agreement, and to treat it thenceforth as abandoned or non-existent.

[93]Lord Dunedin, at page 25, said: The difference between variation and rescission is a real one, and is tested to my thinking by this: in the first case [i.e. variation] there are no such executory clauses in the second arrangement as would enable you to sue upon that alone if the first did not exist; in the second you could sue on the second arrangement alone and the first contract is got rid of either by express words to that effect, or because the second dealing with the same subject-matter as the first but in a different way, it is impossible that the two should be both performed.

[94]At page 28, however, in expressing his conclusion, he said: My Lords, it seems to me quite impossible to come to any conclusion on this but that the parties agreed that the old contract should be abrogated.”

[95]Finally, Lord Parmoor, at page 38, said: It is necessary… to inquire whether the conditions have been so changed in their essential character that there is a substantial inconsistency such as to lead to the inference that parties did intend to rescind the earlier contract … It is not possible to lay down any general principle, but where the alteration is such that the conditions of the earlier contract cannot be restored without placing one of the parties under a permanent and substantial disability there is a strong prima facie probability of an intention to rescind.

[96]Mr Kosmin submitted that their Lordships had laid down a principle to the effect that where the differences or inconsistencies between the later agreement and the existing agreement were so fundamental as to go to the root of the contract, then this was sufficient to constitute an implied rescission. We do not accept this submission. It seems to us that the case is authority for the proposition that the difference between rescission and variation is dependent on the intention of the parties. In the absence of other evidence, the fact that changes effected by the subsequent agreement go to the root of the existing contract, or that some changes are “inconsistent” with the contract’s original terms, might lead to an inference that the parties intended to rescind the contract, but the question remains one of intention.

[97]Mr Kosmin also relied on an obiter passage in the judgment of Toulson LJ in Samuel v Wadlow

[2007] EWCA Civ 155

. That case concerned, relevantly, whether an agreement between the parties, which it was alleged was procured by undue influence, had been varied or rescinded by a subsequent settlement agreement. If it had been only varied, then the claimant contended that he remained entitled to rescind it for undue influence. Having referred to Morris v Baron Toulson LJ pointed out the practical difficulties in determining whether parties had intended to vary or rescind an agreement: 23

[39]However, it may not be easy to determine whether the parties "intended" that the original contract should continue to exist as a matter of legal analysis but in varied form, or whether as a matter of legal analysis it was intended to be discharged and replaced, since the distinction is one of legal theory which might have little commercial meaning for the parties.

[40]In the present case it is plain what the parties intended to be the effect of the settlement agreement in terms of their ongoing financial rights and obligations; but to ascribe to them an intention to achieve that result by variation of the management agreement, as distinct from its replacement by the settlement agreement, or vice versa, is artificial. From a practical viewpoint it is a distinction without a difference.

[98]Ultimately, Toulson LJ did not need to reach a conclusion on the point, since he concluded as follows:

[45]It is a fine question. I consider it to be a sterile question. The law about undue influence is based on broader concepts and I do not believe that its application to the present case should be affected by whether technically the settlement agreement discharged the management agreement.

[46]The principle in Morris v Baron was brought into existence in order to deal with the technical problems produced by legislation analogous to the Statute of Frauds. The less that it is brought into other parts of the law to deal with problems of a different nature which do not require a formalistic approach, the better.

[99]While Toulson LJ pointed out the difficulties inherent in the test laid down in Morris v Baron, we do not consider that he intended to conclude that it was not part of the law of contract (particularly as he had already observed at [36] to [38] of his judgment that the House of Lords in Morris v Baron and British and Benningtons Ltd v NW Cachar Tea Co Ltd [ 1923] AC 48 and the Privy Council in United Dominions Corporation (Jamaica) Ltd v Shoucair

[1969] AC 340

had determined that the question had to be determined by reference to the intention of the parties).

[100]Mr Kosmin also referred us to the decision of the Special Commissioners in Shell UK v HMRC [2007] SPC 00624. That case had some similarities with this appeal. Petroleum revenue tax was introduced in 1975, but, s10 of the Oil Taxation Act 1975 provided that profits derived from a sale of oil (a defined term which included gas) to the British Gas Corporation “under a contract made before the end of June 1975” were to be disregarded for the purposes of that tax. The taxpayers and British Gas entered into a contract on 27 June 1975. That contract was due to expire in October 2002. In 1999 the parties agreed amendments to that contract that included extending its term and changing provisions about price and quantities of gas to be supplied. The question arose whether sales of gas after October 2002 were “under” the 1975 contract with the result that profits arising from them were outside the scope of petroleum revenue tax.

[101]The Special Commissioners decided the appeal on the basis that, even if the 1975 contract was merely varied and not rescinded, the taxpayer was not selling gas “under” the 1975 contract in the relevant statutory sense (see [120] of the decision). We pause 24 to note that the agreement between the parties that the issue in this case is to be determined by reference to the law of contract, and does not turn on any issue of statutory interpretation, precludes HMRC from running an argument along the lines of that which succeeded in Shell UK.

[102]The Special Commissioners went on, however, at [121] to [124], in case they were wrong in their conclusions on statutory construction, to consider whether the 1975 contract had been rescinded. At [122], the Special Commissioners noted that: … there is no doubt that the parties intended a variation and not a rescission, not least for tax reasons However, despite reaching that conclusion, at [124], the Special Commissioners concluded that the 1975 contract had been rescinded largely because the changes made to it were so crucial that they went to the root of the contract.

[103]Mr Kosmin submitted that the Special Commissioners had followed the correct approach at least in relation to what he described as a “status case” such as the one at issue in both Shell UK and this appeal (i.e. a situation where the question whether a contract has been rescinded determines the status of a payment for some other statutory purpose).

[104]We do not accept Mr Kosmin’s submission. The decision of the Special Commissioners in Shell UK is not binding on us. Insofar as they decided that, even where both parties share a common intention not to rescind a contract, that contract could nevertheless be rescinded by the parties agreeing fundamental changes to it, we consider that to be an incorrect statement of the common law given the authorities we have referred to above.

[105]In determining whether the parties intended the Change Orders to rescind the Golden Contract, we accept (in agreement with HMRC) that intention is to be ascertained objectively (see, by analogy the approach of Lord Clarke in RTS Flexible Systems v Molkerei Muller GmbH

[2010] 1 WLR 753

). Rescission – application of the test

[106]Applying a test based on the parties’ intentions, we are in no doubt that they intended the Change Orders only to vary the Golden Contract, not to rescind it. This case is far from the type of case referred to by Toulson LJ in Samuel v Wadlow, where the parties do not turn their minds to the somewhat legalistic question of whether or not a variation is intended to rescind the original agreement. Viewed objectively, the parties to the Golden Contract clearly intended that contract to preserve the entitlement for someone to claim EZAs on construction expenditure on the Site. The Change Orders were similarly intended to result in a building being constructed that was somewhat different from those provided for in the Golden Contract, but for EZAs nevertheless still to be available on the costs of construction. Mr Williamson QC was correct to submit that the parties’ common desire to ensure that EZAs would be available to the purchaser of the building ultimately constructed is strong evidence of a common intention that the Change Orders would not result in rescission. 25

[107]Mr Kosmin argued that this reasoning results in primacy being given to the parties’ subjective intentions, particularly since neither the Golden Contract nor the Change Orders refer in terms to CAA 2001 or tax benefits that were hoped to arise, but we do not agree. At all material times, the Developer and Contractor were under common control. As we have found at [33] the Golden Contract was entered into as part of arrangements specifically designed to ensure that EZAs could still be claimed on construction expenditure at the Site. The expectation that EZAs would continue was not simply an unexpressed subjective wish (or, as Mr Kosmin put it in his submissions a “contemporaneous hope”) of the parties. On the contrary, the desired tax benefits informed the willingness of the Developer and the Contractor to be party to that contract and also informed their common approach to changes to it. In those circumstances, the desire for construction expenditure to qualify for EZAs was a common goal shared by the Developer and the Contractor and is thus directly relevant to the objective ascertainment of their intentions. That is the case even though the parties did not refer expressly to EZAs in the Golden Contract or the Change Orders.

[108]Ms Shaw QC accepted in her submissions that there may be situations where the parties’ apparent common intention not to rescind is so inconsistent with the way they actually acted as to call into question whether the common intention existed at all. Mr Kosmin relied on a number of differences between the contractual scheme applicable to the Data Centres constructed following the Change Orders and the terms of the unamended Golden Contract. These included the following:(1) Works Option 1 set out in the original Golden Contract envisaged the construction of a facility for the manufacture of microchips whereas ultimately the Contractor built two Data Centres. However, while the Contractor built different buildings from those originally envisaged by Works Option 1, we accept Mr Pulford’s evidence that, from a construction perspective, the differences were not significant. To build the Data Centres, the Contractor used broadly the same subcontractors it would have used to build the microchip facility (the principal difference being that, to build the microchip facility, the Contractor would have needed to engage fluid-handling subcontractors but it did not need to do so in order to build the Data Centres). Otherwise, constructing the Data Centres involved broadly similar skills and expertise to those that would have been involved in building the microchip facility.

(2) Works Option 1 as set out in the original Golden Contract envisaged that the microchip facility would be built on “Site C”. However, while the Data Centres were built on the Cobalt Business Park and in the vicinity of Site C, they were not actually built on Site C (largely because, by the time DC2 and DC3 were constructed, DC1 was already present on Site C). We accept Mr Pulford’s evidence that ultimately DC2 and DC3 were built around 20 metres from Site C and accept his description of this change as not being a “big deal”.

(3) Notwithstanding the point we make at (1) above, there was some inconsistency between Works Option 1 as set out in the original Golden Contract and the works necessary to build DC2 and DC3. The original 26 Works Option 1 envisaged that the Contractor would have to “work around” existing buildings located on Site C. For example, the Employer’s Requirements applicable to that original Works Option 1 envisaged that an “Existing Bulk Gas Facility” would be retained to supply the newly built microchip facility. By contrast, before DC2 and DC3 could be built, existing buildings located on the site chosen for those Data Centres needed to be demolished.

(4) Because Works Option 1 envisaged the construction of buildings on Site C specifically and because, by the time of the Change Orders, DC1 was either present or in the course of construction on Site C, absent amendments to the Golden Contract it would not have been possible, at the time the Change Orders were made, for Works Option 1 to be completed as set out in the original Golden Contract.

(5) In return for building the Data Centres, the Contractor was entitled to receive a contract price of £54,845,150 for DC2 and £42,284,000 for DC3 (both sums exclusive of VAT), with payment being made in advance. If the Contractor had constructed Works Option 1 as set out in the Golden Contract, it would have received an aggregate price of £102,500,000 exclusive of VAT.

[109]In our judgment however, none of those differences, whether individually or together, was inconsistent with the parties having a common intention that the Golden Contract should be amended, as opposed to rescinded.

[110]For the above reasons we conclude that the expenditure on the Data Centres was incurred under the Golden Contract. 1(2) – A separate contract

[111]Even if the effect of the Change Orders was not to rescind the original Golden Contract (so that the Golden Contract lived on and was capable of performance), HMRC argue that those Change Orders created a new contract that stood separate from the Golden Contract. Expenditure incurred in the construction of the Data Centres was incurred “under” that new contract for the purposes of s298 of CAA 2001 with the result that, since that new contract was not entered into within 10 years of the Site first being included within an enterprise zone, the deadline in s298 was exceeded and no EZAs are available.

[112]No authority was cited in relation to this argument. Given that the parties are agreed that this question has to be resolved simply by determining whether, as a matter of common law, a new contract was created, it seems to us that it has to be resolved, in the same way as HMRC’s first argument, by reference to the parties’ intentions. Approaching the issue in that way leads inevitably to the same conclusion as we have reached on the issue of rescission. By the time of the Change Orders, all parties were clearly aware that the success of the composite transactions of which the Change Orders formed part depended on the expenditure being incurred “under” the original Golden Contract. Therefore, they intended that the works set out in the Change Orders would 27 be performed subject to the terms of the Golden Contract and not a new contract formed following the issue of the Change Orders. 1(3) Uncertainty

[113]HMRC’s pleaded case was that the Golden Contract was merely an “agreement to agree”, essentially because it simply set out a framework for six Works Options which was a “menu for future contract options” with the result that it was only after the Developer selected a Works Option (and after other conditions precedent were satisfied such as the grant of a lease of Site A, B or C to the Developer) that a contract came into existence. Since the Developer selected a Works Option on 4 February 2009 at the earliest (the date of Variation Agreement One), the sum paid to the Contractor was not “under” contract entered into within the first 10-year period of the enterprise zone. We would have rejected that argument, on the basis that the Golden Contract imposed obligations on the parties from the outset, and the fact that the Developer retained a degree of optionality within the terms of the contract did not preclude a contract from coming into existence.

[114]The argument developed in HMRC’s closing submissions was, however, different. They contended that expenditure on the construction of the Data Centres could be regarded as being “under” a contract entered into within 10 years of a site coming within an enterprise zone for the purposes of s298 of CAA 2001 only if the payment obligations under that contract were capable of being enforced within that 10 year period. HMRC argued that this requirement was not met since no enforceable payment obligation could arise until the Developer selected a Works Option from the menu available under the Golden Contract which did not happen until after the 10-year period expired.

[115]Mr Kosmin developed that submission by arguing that while HMRC accepted that the original Golden Contract obliged the Developer to select one of Works Option 1 to 6, since it did not do so within the first 10 years of the Site first being included within an enterprise zone, no party could have obtained an order for specific performance of that contract within those ten years. Similarly, if the Developer repudiated the contract it would not be possible to calculate damages for that breach until the Developer issued a Notice to Proceed. Since neither party to the Golden Contract could obtain specific performance of it, or any sensible award of damages for breach of it, in the 10-year period after the Site was first included in an enterprise zone, it followed in Mr Kosmin’s submission that, when the Data Centres were ultimately constructed, that construction cannot have taken place “under” a contract entered into within that 10 year period.

[116]While HMRC expressly disavowed any argument that the Golden Contract was an “agreement to agree” at common law of the kind considered by the House of Lords in Walford v Miles

[1992] 2 AC 128

, they maintained that for the purposes of s298 as properly construed, the Golden Contract was an “agreement to agree”.

[117]We reject this new argument. Section 298 of CAA 2001 does not require that any payment obligation under a contract should become unconditional at any point. Rather, s298 focuses on two questions: the first is when the relevant contract was “entered into” 28 and the second is whether the expenditure incurred was “under” a contract that was “entered into” within the requisite period.

[118]In our judgment, the Golden Contract contained enforceable contractual obligations from the moment it was executed. It was, therefore, “entered into” on 17 February 2006 and thus within ten years of the Site first becoming part of an enterprise zone. The fact that not all obligations under that contract were immediately enforceable does not alter that conclusion. If Parliament had, as Mr Kosmin submits, intended to focus attention on the date on which payment obligations become unconditional, it would not have framed s298 in terms of the date on which the contract (as a whole) was “entered into” and instead would have focused attention on specific payment obligations arising under it.

[119]Nor do we consider that the fact that payment obligations had not crystallised within the 10-year anniversary of the Site being included within an enterprise zone prevents expenditure incurred subsequently from being “under” that contract. Given the conclusions that we have reached on Issues 1(1) and 1(2) above, in our judgment, the parties intended the Golden Contract (as varied by the Change Orders) to regulate the terms on which DC2 and DC3 were to be both constructed and paid for. Accordingly, it follows that DC2 and DC3 were constructed “under” that contract (as varied).

[120]In light of our conclusion on the substance of the point, it is unnecessary for us to decide whether (as the LLPs contended), since it was an argument about the proper construction of s298 CAA 2001, it was contrary to the agreement between the parties that Issue 1 as a whole was concerned only with the application of common-law contractual principles and it was thus not open to HMRC to run the point at all. For completeness, however, we would have found that it was not open to HMRC to do so. PART C: Were the LLPs carrying on business with a view to profit?

[121]Pursuant to s863 of ITTOIA, if an LLP carries on a “trade, profession or business with a view to profit” then it is for income tax purposes treated as a partnership. In the jargon of tax lawyers, that would mean that the LLP is “transparent” for tax purposes and, while the LLP still needs to submit a partnership tax return pursuant to s12AA of the Taxes Management Act 1970 (“TMA 1970”), the profits or losses shown in that partnership tax return would not lead to any tax liability (or tax relief) for the LLP itself. Instead, the taxable profit or loss shown on that partnership tax return would be allocated to the individuals who are members of the partnership who would each include their share of the profit or loss on their own tax returns.

[122]The question whether the LLPs satisfy the requirements of s863 of ITTOIA is relevant in this appeal because of the withdrawal of IBAs pursuant to s84 of the Finance Act 2008. That section provided that Part 3 of CAA 2001 (which contained the IBA regime, and its subset, the EZA regime) did not apply to expenditure incurred “on or after the relevant date”. The “relevant date” was defined in s84(4) as 1 April 2011 “for corporation tax purposes” and 6 April 2011 “for income tax purposes”. 29

[123]The LLPs consider that, pursuant to s296 of CAA 2001, they are entitled to EZAs on the price they paid to acquire the relevant interest in the Data Centres. They incurred that expenditure on 4 April 2011 in relation to DC2, and on 5 April 2011 in relation to DC3. If s863 of ITTOIA applied to the LLPs, they would be treated as “transparent” partnerships for income tax purposes and so would not be companies liable to corporation tax. In that case, the “relevant date” would be 6 April 2011 and, since the LLPs incurred their expenditure before that date, EZAs could still be available.

[124]However, if s863 of ITTOIA did not apply then it was common ground that the LLPs were not transparent and were instead bodies corporate subject to corporation tax and that, accordingly, the “relevant date” was 1 April 2011 with the result that no EZAs would be available.

[125]HMRC do not accept that the LLPs carried on their business with a “view to profit”. The burden lies on the LLPs to establish that they did. The legal test

[126]There was substantial agreement between the parties as to the legal test to be applied. In particular, they were in agreement that the test was entirely subjective, notwithstanding the decision of the First Tier Tribunal to the contrary effect in Ingenious Games v HMRC

[2017] SFTD 1158

. The Upper Tribunal released its decision in Ingenious Games

[2019] UKUT 266 (TCC)

after the conclusion of the hearing in this case. We have received further submissions in writing on the impact of the Upper Tribunal’s decision from both parties.

[127]The parties were broadly agreed on the following propositions.

[128]First, the question is to be tested at the time the LLPs incurred their expenditure by acquiring the relevant interests in DC2 and DC3 (i.e. 4 April 2011 and 5 April 2011 respectively) and should not be determined with the benefit of hindsight.

[129]Second, the test is purely subjective. Accordingly, provided that the LLPs had the requisite “view to profit” it does not matter how likely, objectively, they were to make that profit. The fact that the test is subjective does not mean, however, that it is satisfied simply because it is asserted that the requisite view to profit is present. That assertion is to be tested against all the relevant evidence, including the manner in which the business was carried on. If, for example, the way in which a taxpayer conducts its business is inconsistent with the assertion that it did so with a view to profit, then it is open to the Tribunal to reject that assertion. This was recently endorsed by the Upper Tribunal in the Ingenious Games decision: see in particular the following passages from the judgment in that case:

[333]We consider the better view to be that the test is a purely subjective one. There is no need for profit to be the predominant aim. As is noted in Lindley & Banks, difficult questions can arise when any profit-making aim is subsidiary to other purposes. In those circumstances, it is necessary to consider at what point the line is crossed and there is in fact no view to profit. Some sort of “reality check” is needed. It is necessary 30 to identify whether there is a “real” intention rather than something that was not, in fact or reality, aimed for. The question as to whether a trade was carried on “with a view to profit” also cannot be answered in isolation, divorced from the context of the business in question. The context of “carries on a trade…” directs attention at least to some extent to the way in which the trade is conducted. Furthermore, an indifference to whether a profit is realised is not sufficient to meet the test. In this case, therefore, the FTT would have had to have been satisfied that the LLPs had genuinely intended to seek a profit from their activities.

[340]The test is a qualitative rather than a quantitative one and it would be wrong to prescribe a minimum percentage of probability of profit. The question is whether there is a real and serious intention to make a profit. As noted at [344] and [345] below, the likelihood of profit may be an element of relevant evidence, but no more.

[341]It is obvious that there must be evidence to support the contention that an entity genuinely had a subjective intention to carry on its business with a view to profit. Whilst the stated intention of the controlling minds of the entity is highly relevant, the tribunal is entitled to examine their witness evidence critically and decide what weight to attach to it, and in particular to decide whether that stated intention reflected the reality of the situation in the light of other available evidence and the inherent probabilities.

[344]In determining whether there is the requisite subjective intention, all the evidence must be considered. As mentioned in Gestmin v Credit Suisse at [22] which we have cited at [342] above, contemporaneous documentary evidence will always be highly relevant. Objective evidence is also relevant and, depending on the context, it may be significant. This may include evidence about whether there was, in fact, a real potential for, or likelihood of, profit. This is not because there is an objective test or override. Rather, the potential for profit is one part of the evidence that may be relevant to determine whether the requisite subjective intention exists.

[130]Third, in the circumstances of this appeal, the question whether the LLPs had a “view to profit” should be tested by reference to the subjective view of the controlling minds of the LLPs. Those “controlling minds” were Mr Fielding and Ms Brister as they represented the controlling minds behind CDC Administration LLP, the joint venture vehicle formed to organise and administer the Cobalt investments.

[131]Fourth, the test does not require the view to profit to be the sole or primary intention of the controlling minds of the LLPs. Therefore, the fact that the LLPs were clearly interested in securing the benefit of EZAs for their members (which would not of itself generate any profit for the LLPs) is not fatal to an argument that they were carrying on their businesses with a view to profit.

[132]The parties were also broadly agreed on the meaning of “profit” for the purposes of applying the test:(1) It means the excess of the LLPs’ business income over their expenditure before tax. It is not the same as the computation of “profit” for tax purposes. 31 Nor does determining “profit” require an application of the full rigours of accounting practice. However, “profit” has to be given a realistic and commercial meaning: an unworldly person in business might regard gross receipts as “profits” by ignoring the expenses that need to be incurred to generate them. Even if that view is honestly and genuinely held, a view to obtaining gross receipts is not a view to “profit”.

(2) “Profit” may be determined over a period which exceeds, or is less than, a calendar or fiscal year. Therefore, it is not necessarily fatal that the LLPs realised a significant loss in their first accounting period (represented by the difference between the actual value of the Data Centres they acquired and the price they paid for them), provided that the LLPs had a view to achieving a profit over a longer period.

[133]The only material issue on which the parties were not agreed on the formulation of the applicable legal test related to the relevance or otherwise of authorities relating to “sideways loss relief” set out in sections 380 and 384 of the Income and Corporation Taxes Act 1988 (“ICTA”) as rewritten in sections 66 and 74 of the Income Tax Act 2007 (“ITA 2007”).

[134]Those provisions set out statutory tests that must be met before an individual who incurs a loss in a trade can set that loss “sideways” against other taxable income so as to reduce his or her overall liability to income tax. To put the parties’ arguments in context, we will quote aspects of the provisions as rewritten in ITA 2007: 66 Restriction on relief unless trade is commercial(1) Trade loss relief against general income for a loss made in a trade in a tax year is not available unless the trade is commercial.

(2) The trade is commercial if it is carried on throughout the basis period for the tax year— (a) on a commercial basis, and (b) with a view to the realisation of profits of the trade.

(3) If at any time a trade is carried on so as to afford a reasonable expectation of profit, it is treated as carried on at that time with a view to the realisation of profits.

[135]Therefore, s66 of ITA 2007 and the predecessor provisions in ICTA contained two tests that a trade has to satisfy: a “commercial basis” test and a “profits test”. Ms Nathan referred us to a number of authorities on these statutory tests including Samarkand Film Partnership No 3 v HMRC

[2017] EWCA Civ 77

and Seven Individuals v HMRC

[2017] STC 874

in support of an argument that, when we are considering whether the LLPs had the requisite “view to profit” for the purposes of s863 of ITTOIA, we should necessarily consider whether they were carrying on their businesses on a commercial basis. In short, she submitted that the courts had decided that the “commercial basis” test and the “profits test” in s66 of ITA 2007 were not “hermetically sealed” and each informed the other. For example, in Samarkand Film Partnership v HMRC, Henderson LJ endorsed what the Upper Tribunal said at [96] and [97] of its decision (reported at

[2015] STC 2135

): 32

[96]Commercial’ and ‘with a view to profit’ are two different tests but that does not mean that profit is irrelevant when considering whether a trade is being carried on a commercial basis. The reference in Wannell v Rothwell to the serious trader who is seriously interested in profit is not only relevant to deciding whether a person is a serious trader or an amateur or dilettante. We consider that the FTT were right when they said, at [253], that the serious interest in a profit is at the root of commerciality. We also consider they were correct in regarding “profit” in the context of commerciality as a real, commercial profit, taking account of the value of money over time, and not simply an excess of income over receipts.

[97]The FTT were, in our view, right to conclude that a trade that involved transactions that were intended to produce a loss in net present value terms, with no compensating collateral benefits, was not conducted on a commercial basis. No-one who was seriously interested in running a business or trade on commercial lines would pay £10 for an income stream with a net present value of £7 unless there were some good reason to do so. Of course in this case the reason why the partnerships were willing to do this was because they believed that tax relief would be available to the partners.

[136]In the passage quoted above, the Upper Tribunal was considering the statutory tests applicable in that case and concluding, in essence, that a trader who lacked a serious interest in profit would fail the statutory “commercial basis” test even though profits also fall to be examined in the statutory “profits test”. The Upper Tribunal’s reasoning (endorsed by the Court of Appeal) is limited to the statutory tests it was considering and does not extend to the test set out in s863 of ITTOIA. We note that, at [332] of Ingenious Games, the Upper Tribunal reached a similar conclusion. We do not, therefore, accept Ms Nathan’s general point that authorities dealing with the statutory “commercial basis” test are of any direct relevance to the test set out in s863 of ITTOIA.

[137]We nevertheless accept a more limited proposition, namely that if a business is carried on in an uncommercial way, then that might provide evidence that it was not carried on with a view to profit within the meaning of section 863. Application of the legal test to the facts of this case

[138]This case falls squarely within the type of case referred to at paragraph 333 of the decision of the Upper Tribunal in Ingenious. We have no doubt that the principal purpose of the LLPs was to obtain the benefit of the EZAs for their members. The LLPs did not seriously contend otherwise. We accept, however, that this is not inconsistent with the LLPs nevertheless conducting business with a view to profit.

[139]Mr Fielding’s evidence was that he was aware that the LLPs needed to carry on business with a view to profit. We accept that evidence. He was experienced in investments underpinned by government incentives, and was well advised as to the requirements relating to the availability of EZAs, in particular that it was necessary – in order to enable investors in an LLP to obtain the benefit of EZAs on an acquisition by the LLP – that the LLP was carried on with a view to profit. 33

[140]Mr Fielding also understood the general principles on which real estate investments such as DC2 and DC3 should be valued. In summary, Mr Fielding understood, as was uncontroversial, that the most appropriate basis on which to value DC2 and DC3 would be the “rent and yield” approach that seeks to determine the present worth of income (in the form of rent) that would be receivable over their life or the likely holding period which would involve the following steps:(1) First, it is necessary to ascertain the “headline rent” that is payable under a lease of that building, or will be payable once the building is let. That “headline rent” would typically ignore the effect of incentives that a landlord might grant to a tenant on commencement of the lease (such as a rent-free period).

(2) Next, it is necessary to ascertain the “yield”, as a percentage, that is represented by that headline rent. A valuer would determine this yield on a comparative basis having regard to the evidence provided by other market transactions. The appropriate yield to apply in the context of a let building will take into account an assessment of the tenant’s likely ability to meet its rental and other obligations under the lease (referred to in the property industry as the tenant’s “covenant strength”). The higher the covenant strength of a tenant, the lower the yield that would be expected.

(3) Finally, it is necessary to perform some estimate of the likely costs that a purchaser would incur in purchasing the building.

[141]Mr Fielding’s evidence was that, because he realised the importance of the LLPs carrying on business with a view to profit, he modelled both the on-going position of the LLPs and the position on any expected disposal of the property interest by the LLP, in each case without taking into account the EZAs. To this end, he produced at the time of the transaction spreadsheets which modelled whether the LLPs would be expected to receive more (in the form of rental receipts and capital proceeds) than it paid out. We refer to the detailed assumptions underlying these spreadsheets in paragraph [146]. His evidence was that he genuinely believed that the assumptions were reasonable. On the basis of these numbers, he said that he felt “very confident” that each LLP would realise a profit.

[142]Ms Brister has a legal background. She was principally involved in the negotiation of the contractual documents, leaving the financial modelling to Mr Fielding. While her evidence is less relevant than that of Mr Fielding, she did confirm in cross-examination that she had satisfied herself that the LLPs were formed with a view to profit, noting that she had signed a declaration to that effect in documents relating to the incorporation of CDC3 and that, had she not so satisfied herself, she had “…actually set myself up with some pretty nasty consequences”.

[143]HMRC did not directly challenge Mr Fielding’s evidence that he believed each LLP would realise a profit. Instead, relying principally on the evidence as to various things the LLPs did (and what they failed to do) at the time of the transactions, they contend that there is ample evidence that they “did not operate the LLPs with a view to profit.” We understood this aspect of HMRC’s challenge to be that the LLPs were at best indifferent as to whether a profit would be made for two primary reasons:(i) first, 34 their true motivation, as revealed in information memoranda (the “Information Memoranda”) promoting the LLPs, was to secure EZAs for their members and

(ii) they did not conduct proper due diligence in respect of such matters as the rent that could be achieved and the capital value of the Data Centres.

[144]In considering HMRC’s arguments, the starting point is that, immediately upon acquisition of the interest in the respective Golden Contracts, each of the LLPs realised a significant loss.

[145]In the case of CDC2, it paid £153,709,750 for DC2, but immediately wrote down the value in its initial audited accounts for the period ended 5 April 2011 to £77,450,000, resulting in a loss of £76,259,750. CDC3 paid £109,754,500 for DC3, but immediately wrote down its value to £54,520,000 in its accounts for the period ended 5 April 2011, resulting in a loss of £55,234,500. The write-downs are explained, in each case, by the fact that it was acknowledged by the LLPs that a substantial part of the value obtainable on purchase of their interest in the respective Golden Contracts was the value of the EZAs, but that these were intended to be solely for the benefit of the members of the LLPs, and thus did not represent an asset in the hands of the LLPs. Another way of making the same point is that, if the LLPs sold the Data Centres in the future, a purchaser would not be entitled to the benefit of EZAs and therefore that the value of EZAs should not be reflected in the balance sheet value of the Data Centres in the LLPs’ accounts.

[146]We have already referred to the spreadsheets which Mr Fielding prepared at the time of the transactions, produced in order to illustrate that the LLPs would be expected, overall, to receive more (in the form of rental receipts and capital proceeds) than they paid out. The salient features of these spreadsheets (which we will describe in relation to the one prepared for CDC2, but the conclusions apply equally to CDC3) are as follows:(1) It was assumed that DC2 would be let at a rent of £170 per square foot of the net technical area (“£170 psf”), an annual rent of £6,148,390 in the first year of letting, with rents escalating at 2% per annum.

(2) It also assumed (as was envisaged by the Services Agreement) that the Bank Winter Loan would be refinanced with a third-party bank on a letting of DC2. It was assumed that the third-party bank valued DC2 using a yield of 7%, producing a value of £87,834,143, and was prepared to lend 60% of DC2’s value, resulting in a Second Loan of £52,700,486 at an interest rate of 6%. The balance needed to repay the Bank Winter Loan (£54,896,339) would then be provided by the Developer pursuant to its obligation to provide an interest-free Subordinated Loan referred to at [46] above. It was assumed that the LLP did not exercise its right to require the Developer’s subordinated loan to be converted into equity in the LLP.

(3) The spreadsheet indicated that CDC2 would generate sufficient rental income (on the above assumptions) to service interest on the Second Loan and produce a profit each year gradually increasing from £2.9 million in the first year to £8 million after 15 years. That profit would be used to pay down 35 the Second Loan, which could be fully repaid after 13 years. Amortisation of the Subordinated Loan was assumed to commence after 16 years, with it being fully repaid after 22 years.

(4) If there was no disposal of the Data Centre, then the spreadsheet indicated that the LLP would break even (having repaid the whole of the Second Loan and the subordinated loan from the Developer) after 23 years.

(5) On the basis of the above assumptions, the spreadsheet then indicated the amount that would need to be received on a disposal of the Data Centre, and the yield which that implied. After 10 years, it indicated that £108.5 million would be needed to enable the LLP to break even, which assumed a yield of 6.77%. After 15 years, it indicated that £70.36 million would be needed to enable the LLP to break even, which assumed a yield of 11.53%.

[147]At one point in her submissions, Ms Nathan submitted that it is necessary to place a limit on the time within which the LLPs envisaged that a profit might be made and that in this case the time period of several years (which we address in more detail below) was excessive. We accept that in an extreme case, for example if an LLP asserted an expectation of making a profit after 100 years, it would be difficult to conclude that the business was being carried on with a view to profit. That is not, however, because s863 of ITTOIA requires profit to be achieved within any particular timescale. It is rather because a tribunal would be sceptical as to whether the averred view to profit was genuinely held in such circumstances. This is in contrast with provisions in the income tax acts relating to “sideways” loss relief. For example, s74 of ITA 2007 provides that individuals cannot carry back losses incurred in trades in the early years of trading and set them “sideways” against total income unless profits of the trade could “reasonably be expected to be made in the period or within a reasonable period of time afterwards”. In this case, we do not regard the length of time within which it was anticipated that the LLPs were likely to make a profit as being inherently too long to be relevant for the purposes of the statutory test. The more pertinent question – which we will consider in more detail below – is whether it was envisaged by the LLPs that they would continue to hold the Data Centres for a period long enough to enable them to make any profit.

[148]HMRC made a number of points on the contents of these spreadsheets. Some of their objections can be disposed of relatively quickly. Ms Nathan criticised the spreadsheets produced by Mr Fielding because(1) they failed to address the loss realised by the LLPs in their initial accounts as a result of the write-down in the value of the investment in the Data Centres; and

(2) in assuming that rent at £170 psf would be paid from the beginning of year 1, they failed to reflect both (a) the void in rental while the Data Centres were built and while suitable tenants were sought and (b) the fact that there were likely to be incentives provided to tenants meaning that the headline rent would not be paid for a substantial period after the date of letting.

[149]We reject these criticisms. The spreadsheets were prepared on the basis that the LLPs would break even only when they had received sufficient funds to discharge all expenses and liabilities they incurred including, most significantly, the Bank Winter Loans and loans that refinanced the Bank Winter Loans. There was accordingly no need to identify separately the amount required to reverse the loss realised in the initial 36 accounts. The remaining objections are all met by the fact the Developer was obliged, via the Yearly Sum, to cover all of the LLPs’ costs (including interest due on the Bank Winter Loans) prior to the building being let, and to top-up rent to £170 psf in the event that a tenant, or tenants, paid less than the Target Rent. While the spreadsheet did not reflect reality in assuming the Target Rent was paid from the beginning of year one, therefore, this had no impact on profitability because in the time needed to get to “year one” as illustrated in the spreadsheet, the LLP would be operating on a break-even basis. It is right to point out, however, that it is necessary (in order to reflect these likely realities) to add between two to three years to the number of years within which the spreadsheets indicated the LLPs would break even on each of the assumed bases. For example, where the spreadsheet indicated that a sale realising £108.5 million would be sufficient to enable CDC2 to break even after 10 years, the likely time period after April 2011 was more likely between 12 and 13 years.

[150]Ms Nathan also submitted that it was unrealistic to expect a new lender to advance a loan equal to 60% of the capital value of the Data Centres. We do not accept this, however, as it lacked any evidential basis, and was not put to Mr Fielding in cross-examination.

[151]Finally, on this issue, Ms Nathan submitted that, since the LLPs were seeking a Target Lease with a duration that could, in certain circumstances, be as low as 10 years, by the time of the “break even” point that Mr Fielding had identified any initial lease of the Data Centres would have at most a few years to run. That, she argued, would make the disposal proceeds identified in Mr Fielding’s spreadsheets unrealistic. This submission was not supported, however, by expert evidence and we conclude that, while there would undoubtedly have been costs associated with a further fit-out of the Data Centres if a new tenant was to be found, that did not impact on the estimates of future value which proceeded on the basis that (and were driven by) the assumption that the Data Centres had been, or could be, let on leases at the headline rent.

[152]HMRC, more substantially, contend that the spreadsheets unreasonably relied on the assumption that tenants could be found willing to pay a headline rent of £170 psf. They also contend that the spreadsheets made unreasonable, and unsupported, assumptions as to the potential capital value of the Data Centres. Headline rent of £170 psf

[153]The essence of HMRC’s objection in relation to the headline rent is that(i) this level of rent was never realistically achievable and

(ii) the LLPs failed to identify this fact by doing the sort of due diligence which would be done by someone whose purpose was to make a profit.

[154]In particular, HMRC observed that the profitability or otherwise of the LLPs’ business depended on them being able to find a tenant for a new data centre, without a track record, in the North East of England, willing to pay the Target Rent of £170 psf. They contend that before purchasing the Data Centres, the LLPs did not investigate in any great detail the prospects of such a tenant being found, or how long it would take 37 to find such a tenant. That was a material failing because, in 2011 there were significant obstacles that the Data Centres would need to overcome in order to attract a tenant:(1) As Ms Brister put it in her evidence, businesses tend to like to “hug their data centres” by locating them near their main business operations. Since there are proportionately fewer businesses located in the north of the country, and proportionately more in London and the south east, the pool of potential tenants for DC2 and DC3 was smaller than it would have been if those data centres were located near London.

(2) Users of data centres tend to attach significance to those centres’ resilience, or their ability to continue to function despite adverse events. The resilience of a new data centre will not have been tested. Since DC2 and DC3 had no track record, it was likely to be more difficult to persuade tenants to take a lease of them.

[155]The LLPs took advice as to the appropriate headline rent, and as to the likelihood of attracting tenants at that rent, from CBRE. We accept that this was the only advice on rental levels obtained. It was suggested that advice as to the capital value of the Data Centres subsequently obtained from DTZ and GVA provided corroboration of CBRE’s opinion as to the appropriate level of rent. We reject that suggestion: both DTZ and GVA were asked to assume that the headline rent was £170 psf; although Mr Ian Watson agreed that DTZ would not have accepted that assumption without question if they disagreed with it, there is no evidence that DTZ had undertaken any analysis so as to render them able to form any reasoned view on the issue.

[156]CBRE’s advice was formalised in a letter which was appended in full to the Information Memorandum for CDC2. So far as relevant it stated: We understand that the proposed quoting rent for DC2 will be £170 per square foot (psf) excluding running costs and power consumption costs and held at this level until a letting is achieved. We think that this rental level, if offered with a substantial incentive package will be competitive relative to current market levels. It is worth noting that because wholesale data centre space is generally built to order, the market rents have a natural floor linked to the cost of capital and developers return on investment requirements. Given the substantial incentive package available by virtue of the Enterprise Zone tax status we expect to be able to undercut the wider market…. We are pleased to be the letting advisor for the Cobalt Data Centres and in our capacity as one of the market leaders in this field are confident that the current tenant interest expressed in DC1 suggests that there are good prospects that tenants will be secured for DC1 and the proposed DC2.

[157]A similar view was expressed in relation to DC3 in a letter appended to the subsequent Information Memorandum for CDC3.

[158]CBRE’s view was expressed after taking account of such information as to comparable rentals as was available. They advised that they had on their books tenants interested in acquiring 140,000 square feet of data centre space. By the time DC3 was 38 proposed in March 2011 that figure had increased to 300,000 square feet. These figures represented many times the aggregate capacity offered by DC2 and DC3.

[159]CBRE did not say in their letters what they would regard as a “substantial incentive package” that would need to be given in order to secure a rent of £170 psf. Mr Fielding said in his oral evidence that CBRE indicated to him that this might mean a rent-free period of around three to six months. Mr Pulford gave similar evidence. HMRC suggested that we should be cautious about accepting this evidence given that it had not been mentioned in the witness statement of either Mr Fielding or Mr Pulford and was not backed up by any contemporaneous written evidence. Moreover, in his expert report on the value of DC2 and DC3, the appellants’ own expert, Mr Ian Watson estimated that, in order to secure a headline rent of £170 psf the LLPs would need to offer a rent-free period of 24 months to 36 months. In addition, HMRC pointed out that Mr Fielding’s contemporaneous models prepared in order to support financial illustrations set out in the Information Memoranda envisaged a rent-free period of one to two years.

[160]We accept the evidence of Mr Fielding and Mr Pulford in this respect. The fact that Mr Fielding provided illustrations based on substantially longer rent-free periods does not in our view undermine his evidence that CBRE had advised on the basis of shorter periods. It does suggest, however, that he was aware that there was a reasonable chance that greater incentives would be required, and that it was in those circumstances appropriate to prepare illustrations on a more conservative basis.

[161]HMRC criticise the LLPs for relying on CBRE for advice, because CBRE were also advising the Developer, the party on the other side of the transaction, and were thus conflicted. We reject that criticism and do not think that the LLPs’ reliance on CBRE in these circumstances demonstrates that they were indifferent to profit. We accept the evidence of Mr Fielding and Ms Brister that CBRE were recognised as having market-leading expertise in data centres.

[162]HMRC also contend that CBRE’s opinion was itself flawed, because it was based on properties in and around London, and ignored the effect on demand of siting a data centre in the north east of England. Mr Watson’s evidence was that £170 psf was an appropriate headline rent for the Data Centres as at April 2011, provided that substantial incentives were offered to prospective tenants. HMRC did not dispute his evidence that £170 psf was an achievable rent, but they contended that in calculating the “headline rent”, upon which any valuation would be based, it was necessary to take into account the impact of any rent-free period. For example, therefore, with a 24-month rent-free period over a 10-year lease, in HMRC’s submission, the “real” rent was £136 psf. Although Mr Watson’s view is that the capital value of the Data Centres (as at April 2011) would be reduced to reflect the fact that rent-free or reduced-rent periods would be offered to tenants, he maintained that such incentives did not undermine the conclusion that the appropriate headline rent remained £170 psf. In particular, he rejected Ms Nathan’s suggestion in cross-examination that, because a tenant would require a 2-year rent-free period as an inducement to sign a 10-year lease (i.e. a rent-free period of 20% of the lease term), it followed that the appropriate headline rent was just 80% of £170 psf. He reached that conclusion because, in his expert opinion, there is a material difference between a 10-year lease at a rent of £170 psf with a 2-year rent- 39 free period and a 10-year lease, with no rent-free period, at a rent of £136 psf: in the former case, the higher rent would serve as a benchmark for future upwards-only rent reviews and so produced a materially different outcome from the latter case. Therefore, Mr Watson’s clear evidence was that a rent of £170 psf was an appropriate headline rent, although the LLPs might need to grant a 2-year rent-free period over the initial 10-year lease term in order to induce a tenant to take a lease at that rent. We accept that evidence. Although Mr Watson’s evidence was not available at the time, the fact that it corroborates CBRE’s contemporaneous view as to the appropriate headline rent does to some extent undermine the contention that reliance on CBRE’s opinion alone indicated an indifference to making a profit.

[163]For the above reasons, we conclude, first, that it was reasonable to expect a tenant, assuming one could be found, to pay a headline rent of £170 psf and, second, that Mr Fielding believed this to be the case.

[164]So far as the likelihood of finding a tenant willing to pay that headline rent is concerned, while we find that the LLPs could have undertaken greater due diligence in this regard, we do not accept HMRC’s submission that this demonstrates they were indifferent to making a profit. We accept the LLPs’ contention that they were justified in going to, and relying upon assurances from, the market leader, CBRE, (which CBRE was prepared to commit to writing) that there were good prospects of securing tenants. We also accept that, having received that advice, the LLPs genuinely believed that they would be able to secure a tenant who would pay a headline rent of £170 psf (although they realised that they might need to offer a rent-free period in order to secure a tenant’s signature). Capital value of Data Centres

[165]The LLPs accept (as indicated by the immediate write-down in their accounts) that they paid substantially more for their respective interests in the Golden Contracts than they were worth in their hands. Their contention that the difference is legitimately accounted for by reference to the benefit of the EZAs does not itself answer the question whether the LLPs were carrying on business with a view to profit, because as we have already noted, the benefit of the EZAs was never an asset belonging to the LLPs. However, the capital value of the Data Centres, and the LLPs’ perceptions of their capital value, are relevant to this issue since if, for example, the LLPs realised that they were significantly overpaying for the Data Centres, that could be indicative that they lacked a view to profit.

[166]At the time the Information Memorandum for CDC2 was published (at the end of January 2011) the only written advice relating to valuation was the letter from CBRE referred to above. That did not purport, however, to provide any advice on the capital value of the Data Centres. It was solely advice as to the appropriate headline rent.

[167]CDC2 did, however, obtain some guidance on valuation of the Data Centres, at around this time, in the form of a letter from DTZ dated a few days after the issue of the Information Memorandum. This was a “broker’s opinion of value”, stated to be a desktop overview, provided on a limited basis for guidance only, without having 40 undertaken full verification or research. Based on the assumptions that the Data Centres were completed, fully fitted and available for immediate occupation, and that they were let to grade “A” tenants on a 15 year lease at a rent of £170 psf increasing at 3% per annum, the estimate for DC2 was £77,450,000. Subsequently, DTZ provided advice on a similar basis in respect of DC3, valuing it at £55,300,000. Although the advice from DTZ does not expressly identify the yield percentage on which this valuation was based, it was implicitly based on a yield of approximately 8%.

[168]Mr Fielding referred in his evidence to other valuation evidence provided by GVA Grimley Ltd (“GVA”) of between £83 million and £110 million. This is a reference to an email from GVA dated 24 March 2011 which contained a “preliminary view”, based on similar assumptions to those made by DTZ, that “the potential values” of DC2 “could be” between £83 million and £104 million. We note that in the first audited accounts for CDC2 there is reference to a range of valuation advice, with the lower value (reflecting DTZ’s advice) being adopted as net book value at 5 April 2011 on the basis that it was the “most prudent”.

[169]In fact, the higher of the GVA values assumed a lease of 20 years. The advice that had been received from CBRE (as stated in their letter annexed to the Information Memorandum) was that wholesale data centre tenants typically commit to leases of “10 years term certain”. Mr Watson’s evidence is to the same effect, namely that lease lengths for data centres were typically between 10 and 15 years, but closer to 10 years. He noted that the average lease length proposed on DC1 was 11 years. The Information Memorandum stated that the target lease length was 15 years, but with a minimum of 10 years. Accordingly, we discount the higher “potential value” indicated by GVA, as based on an assumption that was unrealistic, and must have been known at the time to be unrealistic.

[170]It is also important to recognise that all of the valuation advice received was given on the assumption that the Data Centres were completed, fully fitted, and let to ‘A’ grade tenants without any incentives having been offered. In other words, they were valuations of what the LLPs hoped to have acquired, at some point in the future, once all of the assumptions on which they were based were satisfied. As a result, as confirmed by the expert evidence of Mr Watson, the actual value of the interest in the Golden Contract acquired by the LLPs in April 2011 was significantly lower.

[171]We nevertheless accept that, as a result of the obligations of the Developer to pay the Yearly Sum under the Services Agreement, the LLPs were insulated from the risk of completion of the Data Centres and most of the risks associated with finding tenants willing to pay the Target Rent (although there was still a risk that no tenant at all would be found). Accordingly, we accept that in estimating the amounts that might be realised on a future disposal it was appropriate to do so on the basis of the assumptions built into the valuation evidence received at the time.

[172]Mr Fielding stated, in his witness statement, that he regarded (and still regards) 8% as a conservative yield. He said that, from discussions with John White (who had previously been employed by Cushman & Wakefield) and “the valuers” – as well as from documents he has seen subsequently – he thinks a more typical yield would be 41 6%. He also stated that he had good reason to believe that a lower yield, closer to 4%, could be achieved, because the Department of Works and Pensions were expressing interest in finding new data centre capacity. Given their covenant strength, they could “negotiate a lower rent relative to the data centres’ capital value” and a yield of 4% or even lower could be achievable.

[173]There are a number of difficulties with this evidence. In the first place, Mr Fielding is not a valuer, so his opinion of appropriate yield percentages carries little if any weight. Second, he provides no details of the discussions with John White or the valuers which suggested that a yield of 6% was reasonable, and expressly acknowledges that his belief is influenced by documents he has seen subsequently. Insofar as he is relying on what was later discussed with GVA then, as we have noted above, their comments in relation to yields of less than 7% were based on leases of 20 years and thus do not reflect what was anticipated in respect of the Data Centres. In contrast, the only contemporary evidence of an appropriate yield is that of 8% implied in the DTZ valuation, and this assumed the Data Centres were fully let on leases of between 10 and 15 years. Third, the DTZ valuation already assumed a tenant with the highest covenant strength, which undermines Mr Fielding’s assertion as to a potential yield of 4% if such a tenant could be found. We also note that the logical conclusion from his comment that a tenant with a stronger covenant would be able to negotiate a lower rent relative to the Data Centre’s capital value is, not that the valuation of the Data Centre should have been higher than that implied by a rent of £170 psf and a yield of 8%, but that the rent might have been negotiated down by such a tenant.

[174]In light of these points, we find that, at best, Mr Fielding may have hoped that the Data Centres would be capable of achieving higher values than that advised by DTZ, but that he had no genuine belief that they would have such a value. Nevertheless, the illustrations he produced at the time of the transaction assumed that after approximately 11 or 12 years since the property was first let, the finance debt of the LLPs could be wholly repaid from a disposal based on a yield percentage of 8%, being that implied in the DTZ valuation. Although the amount required to be raised from such a disposal was envisaged to be much higher than the DTZ valuation, that is explicable by the 2% per annum increase in rent assumed within the illustrations. It was not put to Mr Fielding that he had no genuine belief that such increase in the rent was reasonable. We note in any event that the leases of data centres which CBRE referred to at the time as comparable (albeit in the South of England) were achieving uplifts in rent of approximately 3% per annum. In addition, Mr Fielding’s spreadsheets indicated that the LLPs could break even 15 years after the Data Centres were fully let by selling the Data Centres for a price lower than that produced by the DTZ valuation.

[175]HMRC criticise the LLPs for failing to obtain a formal valuation performed to the exacting professional standards set out in the “Red Book” published by the Royal Institute of Chartered Surveyors. They contend that the LLPs’ willingness to proceed on the basis of brokers’ opinions alone is evidence that Mr Fielding did not genuinely believe that the LLPs were capable of making a profit. We accept, however, the evidence of Mr Fielding that he was advised at the time that without knowing the identity of the tenant or the terms of the lease, it would not have been possible to obtain a more precise evaluation. While Mr Watson agreed in cross-examination that it would 42 have been possible for a valuer to produce, in form, a fuller valuation, he made the point that it would have been heavily caveated due to the lack of a significant local market and the absence of an occupier or a pre-let.

[176]We accept that, given that the reason a fuller valuation report was not obtained was because of the lack of evidence on which to base such a valuation, there was an inherent uncertainty as to the reliability of the limited valuation advice that the LLPs did obtain. Nevertheless, we conclude that Mr Fielding did have a genuine belief that the valuations contained in his illustrations could be achieved. As such, we reject HMRC’s contention that the failure to obtain better valuation evidence or the material difference between the value of the interest in the Data Centres that the LLPs acquired and the Price demonstrates the absence of a “view to profit”. Lack of negotiation as to the Price

[177]HMRC rely, in support of their contention that the LLPs did not carry on business with a view to a profit, on the fact that they made no attempt to negotiate the Price. It is true, as frankly acknowledged by Mr Fielding, that he made no attempt to negotiate the Price, accepting without question the amount asked for by the Developer. When that is combined with the fact that the LLPs knew that the value – in their hands – of the interest under the Golden Contract was substantially less than the Price, it provides an indication that the LLPs were at best indifferent to making any profit.

[178]The lack of negotiation of the Price has, however, to be viewed in the context of the other rights and benefits which the LLPs negotiated as part of an overall package. We identify these in greater detail in Part D below, dealing with the question whether the Price was paid solely for the relevant interest. As we there conclude, those other rights had substantial value to the LLPs. Two examples suffice for present purposes. First, the Yearly Sum removed much of the commercial risk of being unable to locate a tenant willing to pay the Target Rent, and insulated the LLPs against the cost of delay in completing and fitting out the Data Centres. Second, the obligation of the Developer to provide cash collateral for the Bank Winter Loan, combined with its obligation to permit that cash collateral to be used to repay that part of the Bank Winter Loan that could not be refinanced, to be replaced with an interest-free subordinated loan from the Developer, had a potentially large value to the LLPs. Even on the basis of the assumptions made in Mr Fielding’s illustrations prepared at the time of the transactions, this would have involved the Developer making an interest free loan with no repayment for a period of some 16 or 17 years.

[179]We accept Ms Brister’s evidence that these additional benefits were the subject of hard negotiation by the LLPs. This goes a long way to negating the inference that would otherwise be drawn against a party that not only failed to negotiate the Price but knew that it was around double the market value of the asset it was buying. The Information Memoranda

[180]HMRC rely on the terms of the Information Memoranda as evidencing the LLPs’ indifference to making a profit. They point to a number of features which emphasised 43 the benefit of the EZAs over the importance of any commercial returns of the venture. Most strikingly, the Information Memoranda made clear, with the help of illustrations, the extent to which members would be cash positive from the beginning, as a result of the 70% gearing at partnership level and the availability of 100% capital allowances from the first year. The illustrations demonstrated that in return for an investment of £300,000 an investor could expect to obtain a more or less immediate entitlement to tax relief worth £500,000. In other words, an investment in the LLPs was expected to produce a 66.67% profit almost as soon as it was made. This is contrasted with the lack of emphasis on the ability of the LLP, as a stand-alone entity, to generate a profit. HMRC submit that in the case of such substantial investments, the Information Memoranda should have been closer to “business proposals of sufficient rigour to pass muster on a Masters of Business Administration course or to withstand scrutiny in an episode of Dragons’ Den”, in the words of Judge Wikeley in JF v HMRC

[2017] UKUT 0334 (AAC)

. They note that the importance of the Information Memoranda was emphasised by the statement that investors should make their investment decision solely on the basis of the information contained in them.

[181]We agree that the Information Memoranda are drafted in a way which suggests that the LLPs were set up primarily in order to generate EZAs. It is important, however, to view the Information Memoranda in the context that they were documents whose purpose was to encourage potential investors to become members of the LLPs. The main selling point of the LLPs to potential investors was undoubtedly the benefit of the EZAs and investors were unlikely to base their investment decision on their assessment of the likelihood of the LLPs making profits. Indeed, as we will explain in more detail below, there was a positive disadvantage to members arising from profits made by the LLPs, in the form of a ‘dry’ tax charge. It may well be, therefore, that the individual members (or even the members collectively) were indifferent to whether the LLPs made any profit. Given, however, it was common ground that the intentions of the members are irrelevant, the contents of the Information Memoranda are far from determinative. So far as Mr Fielding and Ms Brister are concerned, the most that can be said, based on the Information Memoranda, is that the principal purpose of the LLPs was to generate EZAs for the benefit of their members. As noted above, however, the statutory test does not require that an LLP carries on business with a view mainly to making a profit, provided that one of its purposes is to make a profit.

[182]HMRC also point to the fact that insofar as the Information Memoranda considered how long the Data Centres would be held, they principally focused on a period of seven years, and that even Mr Fielding’s contemporaneous illustrations of financial performance did not envisage that the LLPs would be able to dispose of the Data Centres for an amount sufficient to repay all borrowings within seven years. They rely on this to suggest that Mr Fielding and Ms Brister were at best indifferent as to whether the LLPs would carry on business for a period sufficiently long to enable a profit to be made.

[183]Seven years is significant because the benefit of the EZAs would be lost (and tax savings would be clawed back by HMRC) in the event of any disposal of an interest in the Data Centres within seven years from the date of the transaction. Thereafter, although an outright sale of the Data Centres would have resulted in a clawback of tax 44 benefits, the disposal of a lesser interest (such as the carve-out and sale of a sub-leasehold interest) was permitted. It is true that the Information Memoranda focused on the requirement to hold the Data Centres for seven years, precisely because of the tax disadvantage of an earlier disposal. That does not mean, however, that the Information Memoranda demonstrated any intention on the part of the LLPs that the Data Centres would be sold at the end of that period. We note, in this regard, that the LLPs’ Members’ Agreement provided that any disposal of a lesser interest in the Data Centres after seven years needed to be approved by a “Super Majority” of ordinary members (broadly 75% of those present and voting on the resolution) so that it is not surprising that the Information Memoranda contained nothing purporting to indicate when that would happen. HMRC suggested that Ms Brister had accepted in her cross-examination that there was an intention to sell a significant interest in each building in year seven, however that is a misreading of her evidence. The evidence to which HMRC refer came in a passage of cross-examination in which Ms Brister was asked why the LLPs had not sought to ensure that the Developer’s obligation to pay Yearly Sums would continue for a longer period than 15 years. It was put to Ms Brister that the LLP did not need a longer period than 15 years because “the LLP was going to sell in 15 anyway”. Ms Brister’s answer and follow-up questions then proceeded as follows: Well, it’s pretty standard and I believe it’s accepted that – by the Revenue that there will be a disposal. This is a property investment business and you realise your value by selling at some point. Q. But the “some point” was anticipated to be around 15 years… A. No, it could have been at any point after seven years. Q. And the seven years was important because you didn’t want any EZA clawback. A Correct

[184]This passage comes nowhere near establishing that there was any intention or understanding that the LLPs would sell the Data Centres at any particular point in time.

[185]We consider that on a closer analysis of Mr Fielding’s contemporaneous spreadsheets, there may have been a powerful incentive on members to agree upon a disposal of the Data Centres before the LLPs were anticipated to make a profit.

[186]As the Information Memoranda explained, in each year in which the Data Centres were let the LLPs would make a profit for tax purposes upon which the individual members would be taxed, but which would be used by the LLPs solely for the purpose of paying down the principal sums outstanding on their borrowings. In other words, for each year in which the Data Centres were let, until the Second Loan was repaid in full, the members would suffer a “dry” tax charge without receiving any cash from the LLPs.

[187]The magnitude of this tax charge depended on a number of factors including the amount of the Second Loan and the interest rate payable on it. The Information Memorandum for DC2 indicated to investors that, if DC2 was fully let from the end of year 3 (at the target rent of £170 psf) an investor investing £300,000 in the partnership could expect to be treated for tax purposes as receiving taxable income of £28,079 over 45 the first seven years of his or her investment (resulting in an aggregate tax charge for an investor of £14,040). If the Developer became a member of the LLP following capitalisation of a Subordinated Loan, it too would suffer a “dry” tax charge on its allocation of taxable income.

[188]Mr Fielding’s spreadsheet indicated that the tax charge was likely to get more pronounced each year DC2 was let since, over time, as the spreadsheet assumed that rental income would be used to repay principal on the Second Loan, the interest amount on the Second Loan reduced whereas rental income was assumed to increase at 2% per annum. That gives rise to the question (having regard to the fact that the decision whether to dispose of the Data Centres lay with the requisite majority of members of the LLPs) whether a sufficient number of investors would have been willing to continue to hold the Data Centres for a period long enough to enable the LLPs to realise a profit. HMRC did not put their case on this basis, however, no evidence was led as to the intentions of investors, and Mr Fielding was not cross-examined as to his understanding in this regard. Accordingly, this is not an element in our conclusion on Issue 2. Conclusions

[189]For the reasons we have developed above, we conclude that while the principal purpose of the LLPs was to obtain the benefit of EZAs for their members, it was their subsidiary purpose to carry on business with a view to profit. We find that Mr Fielding and Ms Brister genuinely believed, on the basis of the illustrations prepared by Mr Fielding at the time of the transactions, that making a profit was reasonably achievable for the LLPs and that it was their genuinely held intention that the LLPs were carrying on business with a view to profit. We therefore determine Issue 2 in favour of the LLPs. PART D - ISSUE 3 The relevant statutory provisions in more detail

[190]The LLPs claim that the whole of the purchase price (the “Price”) paid by them (being £153,709,750 in the case of CDC2 and £109,754,500 in the case of CDC3) is qualifying expenditure within the meaning of s.296.

[191]Section 296 applies where expenditure is incurred by a developer on the construction of a building, and the relevant interest in the building is sold by the developer in the course of the development trade before the building is first used. In that event, provided that the sale of the relevant interest by the developer was the only sale of that interest before the building was used, then if “…a capital sum is paid by the purchaser for the relevant interest”, then that capital sum is qualifying expenditure. The “relevant interest” is, by s.286, the “interest in the building to which the person who incurred the expenditure on the construction of the building was entitled when the expenditure was incurred”. In this case, the “relevant interest” consisted, in each case, of the Developer’s right under the Golden Contract to have the building constructed.

[192]It is common ground that each of the requirements of s.296 is satisfied in this case, save for the requirement that the purchase price is paid “for the relevant interest”. 46 HMRC disputes that the whole of the purchase price paid by the LLPs was ‘for’ the respective ‘relevant interests’ acquired by them.

[193]Section 296 needs to be read together with sections 356 and 357. Section 356 provides as follows: (1) If the sum paid for the sale of the relevant interest in a building is attributable— (a) partly to assets representing expenditure for which an allowance can be made under this Part, and (b) partly to assets representing other expenditure, only so much of the sum as on a just and reasonable apportionment is attributable to the assets referred to in paragraph (a) is to be taken into account for the purposes of this Part.

[194]Section 357 provides as follows: (1) If— (a) the relevant interest in a building is sold, (b) related arrangements have been entered into, at or before the time when the sale price is fixed, which had the effect at that time of enhancing the value of the relevant interest, and (c) the arrangements contain a provision which has an artificial effect on pricing (see subsection (4)), the sum paid on the sale of the relevant interest is to be treated for the purposes of arriving at qualifying expenditure as reduced to what it would have been if the arrangements had not contained the provision having that artificial effect. … (3) “Related arrangements” means arrangements between two or more persons which relate— (a) to an interest in or right over the building, or (b) to other arrangements made with respect to such an interest or right; and for this purpose it is immaterial whether the interest or right in question is granted by the person entitled to the relevant interest or another person. (4) Arrangements contain a provision having an artificial effect on pricing to the extent that they go beyond what could reasonably have been regarded as required in comparable commercial transactions by the market conditions prevailing when the arrangements were entered into. (5) “Comparable commercial transactions” means transactions— (a) involving interests in or rights over buildings of the same kind as (or of a similar kind to) the building to which the arrangements relate, and 47 (b) made by persons dealing with each other at arm’s length in the open market.

[195]The LLPs contend that the whole of the Price paid by each of them was for the relevant interest. While they accept (as we shall describe in more detail below) that they acquired various ancillary rights (as set out in the suite of documents entered into alongside the SDA), they contend that the only way in which the Price can be adjusted is if and to the extent that those ancillary rights artificially inflated the Price within the meaning of s.357. It is not, they contend, appropriate to identify each and every single right and benefit acquired, and ascribe a value to each under s.296: if that were the right approach it would render s.356 and s.357 redundant. This argument was advanced, in particular, in relation to ancillary benefits consisting of rental support arrangements, but we did not understand it to be limited to such arrangements.

[196]At an earlier stage in these proceedings, HMRC had sought to argue, in the alternative, that the ancillary rights which the LLPs obtained were indeed caught by s.357, such that the sums paid by the LLPs were to be treated as reduced by the value of those ancillary rights. That argument was, however, abandoned prior to trial.

[197]We reject the LLPs’ contention that s.357 constitutes the only route for challenging their assertion that the full purchase price was qualifying expenditure. In our view, s.296, s.356 and s.357 work together as follows:(1) Section 296 is a threshold condition. In order to be entitled to EZAs at all, the LLPs must identify a capital sum that is “paid… for the relevant interest”. Although s296 does not contain any express words of apportionment (unlike s356) some element of apportionment is necessarily contained within the concept of consideration being paid “for” a relevant interest. For example, a purchaser could pay a developer £100,000 in consideration for the developer’s agreement to transfer both a relevant interest in a building and a portfolio of shares. Section 296 would confer allowances only on the proportion of that consideration paid for the relevant interest.

(2) Section 356 recognises that a capital sum paid for a relevant interest in a building can be attributable partly to assets qualifying for allowances and partly to assets that do not. For example, a single building might consist of both offices (which could attract EZAs) and residential accommodation (which would not). A purchaser might pay a developer £100,000 for the “relevant interest” in that building (thereby satisfying s296) but s356 would restrict EZAs to the proportion of that capital sum that is attributable to the office space. The example of the person paying a developer for both shares and a relevant interest in a building above is within the scope of s296, but not s356, since s356 applies only to consideration that is paid “for” a sale of the relevant interest.

(3) Section 357 addresses a different question. Even if a capital sum is paid for a relevant interest, and even if it is attributable entirely to assets that qualify for EZAs then, if the value of the relevant interest has been enhanced by “artificial” arrangements, the sum paid on sale of the relevant interest is 48 to be treated as reduced, with a corresponding reduction in EZAs that are available.

[198]We accept that there may, in some cases, be an overlap in the operation of the three provisions and that it may be difficult to see the precise point of distinction between their operation. That is not a reason, however, to conclude that s.357 is an exclusive code for the purpose of challenging the quantum of the amount said to be paid “for” the relevant interest. It seems to us that scope for overlap is an inevitable consequence of amendments on a piecemeal basis to an already complicated statutory code. We address separately the question whether HMRC is precluded from contending that insofar as any part of the payment was made in consideration for the rental support arrangements in this case they are not for the “relevant interest” in considering the judicial review claim. Authorities

[199]We were not referred to any authority dealing directly with s296 or s356 of CAA 2001. The parties, were, however, agreed that guidance is to be found in authorities dealing with similar statutory provisions, in particular the decision of the House of Lords in Barclays Mercantile Business Finance Limited v Mawson

[2005] 1 AC 684

(“BMBF”) and the decision of the Supreme Court in Tower MCashback LLP1 v HMRC

[2011] STC 1144

(“TowerM”).

[200]BMBF involved finance leasing transactions. As part of a composite transaction, a company (“BGE”) sold an oil pipeline to BMBF for £91m. BMBF granted a lease back of the pipeline to BGE in return for lease rentals. BGE granted a sublease of that pipeline to its UK subsidiary (“BGE UK”). BGE UK agreed to assume direct responsibility to BMBF for BGE’s obligations to pay rent under the head lease. BMBF had borrowed the £91m which it had paid for the pipeline from Barclays Bank. Barclays Bank also provided a guarantee to BMBF of BGE’s obligations in respect of the lease rentals. As counter-security for its potential liability under the guarantee, Barclays Bank required BGE to provide a charge over the £91m. This was achieved via a complex series of agreements between BGE, BGE (UK), Barclays Bank and certain other entities, summarised by Lord Nicholls of Birkenhead at [17] as follows: So, as the Special Commissioners and Park J pointed out, the £91m passed from Barclays Bank to BMBF, from BMBF to BGE, from BGE to Deepstream, from Deepstream to BIoM and from BIoM back to Barclays Bank again. The effect, as Park J said, was that BGE, having sold the pipeline, was unable to get its hands on the purchase price. It had to remain on deposit with Deepstream and be paid out, year by year, partly (in the form of A payments) to discharge the liability for rent under the lease and partly (in the form of B and C payments) for the benefit of BGE. And the benefit obtained by BGE was entirely attributable to BMBF being able to pass on the benefit of its capital allowances.

[201]In BMBF, the relevant statutory question was whether BMBF had “incurred capital expenditure on the provision of machinery or plant”. The Inland Revenue argued, on the basis of the “Ramsay” principle set out in WT Ramsay Ltd v Inland 49 Revenue Commissioners

[1982] AC 300

, that in reality BMBF had incurred capital expenditure on a tax avoidance scheme to provide it with capital allowances. Before the Special Commissioners and before Park J, this argument was successful. Park J’s conclusion (recited at [23] of the speech of Lord Nicholls) was as follows: It is true that in a strictly legal sense one can say that BMBF incurred expenditure on the provision of the pipeline. That is what the two acquisition agreements said. … However, in the light of the Ramsay authorities I consider that I have to interpret and apply the statute in a wider way … I have to ask: on what did BMBF really incur its expenditure of £91m? Was it really incurred on the provision of the pipeline, or was it really incurred on something else? … My answer is that the expenditure was really incurred on the creation or provision of a complex network of agreements under which, in an almost entirely secured way, money flows would take place annually over the next 32 or so years so as to recoup to BMBF its outlay of £91m plus a profit.

[202]Park J’s decision was overturned by the Court of Appeal. In dismissing the Inland Revenue’s appeal to the House of Lords, Lord Nicholls explained (at [32]) that the essence of the Ramsay approach is “to give the statutory provision a purposive construction in order to determine the nature of the transaction to which it was intended to apply and then to decide whether the actual transaction (which might involve considering the overall effect of a number of elements intended to operate together) answered to the statutory description … the question is always whether the relevant provision of the statute, upon its true construction, applies to the facts as found.”

[203]He went on, at [39] to [41] to determine that the object of granting a capital allowance under the relevant statutory provision (s.24(1) of the Capital Allowances Act 1990) was to provide a tax equivalent to the normal accounting deduction from profits for the depreciation of machinery and plant used for the purposes of a trade. Consistently with this purpose, section 24(1) requires that a trader should have incurred capital expenditure “on the provision of machinery or plant for the purposes of his trade.” He concluded that these statutory requirements were concerned “entirely with the acts and purposes of the lessor. The Act says nothing about what the lessee should do with the purchase price, how he should find the money to pay the rent or how he should use the plant.” He rejected the Special Commissioners’ conclusion that the transaction “had no commercial reality” as depending entirely upon an examination of what happened to the purchase price after BMBF paid it to BGE, but “…these matters do not affect the reality of the expenditure by BMBF and its acquisition of the pipeline for the purposes of its finance leasing trade.” At [42] Lord Nicholls said: If the lessee chooses to make arrangements, even as a preordained part of the transaction for the sale and lease back, which result in the bulk of the purchase price being irrevocably committed to paying the rent, that is no concern of the lessor. From his point of view, the transaction is exactly the same. No one disputes that BMBF had acquired ownership of the pipeline or that it generated income for BMBF in the course of its trade in the form of rent chargeable to corporation tax. In return it paid £91m. The circularity of payments which so impressed Park J and the Special Commissioners arose because BMBF, in the ordinary course of 50 its business, borrowed the money to buy the pipeline from Barclays Bank and Barclays happened to be the bank which provided the cash collateralised guarantee to BMBF for the payment of the rent. But these were happenstances. None of these transactions, whether circular or not, were necessary elements in creating the entitlement to the capital allowances.

[204]In TowerM the relevant statutory question was whether the taxpayers had incurred capital expenditure on the provision of plant and machinery for the purposes of s11 of the Capital Allowances Act 1990.

[205]MCashback Limited (“MCashback”) had developed software that allowed manufacturers to promote products to retail customers by offering free airtime on their mobile phones. Manufacturers paid MCashback “clearing fees” for the use of that software. The taxpayer LLPs paid consideration for the grant of software licences which would entitle the LLPs to a proportion of the clearing fees that MCashback received. The LLPs funded the consideration payable out of the subscriptions paid by their members for their membership interest. 25% of the subscriptions was provided out of the individuals’ own resources. The other 75% was provided by loans made to the individual members, limited in recourse to their shares in the LLP. MCashback placed around 82% of the consideration it received on deposit which was used as security for the loans made to the individual members.

[206]The Special Commissioners found as a fact that(i) while the scheme was not a sham, it was pre-ordained and designed as a composite whole;

(ii) the market value of the software licences was “very materially below” the price that the LLPs ostensibly paid for those rights and

(iii) there was little chance that the loan to the members would be repaid in full within ten years; as much as 60% of the loans might be unpaid or waived at the end of that period. Having made those findings, the Special Commissioners concluded that the LLPs had not incurred the full amount of expenditure that they had claimed, essentially because 75% of that expenditure was “filtered back” to the LLPs.

[207]Henderson J, as he then was, reversed this conclusion. He held that, given the decision of the House of Lords in BMBF, the only conclusion open to the Special Commissioners was that the LLPs had incurred the full amount they claimed on the provision of plant for the purposes of s11. In reaching his decision, he concluded that it was “irrelevant” that the market value of the software licences was materially below the sums the LLPs claimed to have incurred in acquiring them. The Court of Appeal dismissed HMRC’s appeal concluding that the taxpayer LLPs had incurred “real expenditure”.

[208]The Supreme Court reversed the decision of the Court of Appeal. Lord Walker (at [72]) said that the Judge had been wrong to dismiss as sweepingly as he did the Special Commissioner’s scepticism about the valuation of the software rights. At [74] he considered that 75% of the capital raised, although not simply a sham, “was really being used in an attempt to quadruple the investor members’ capital allowances.” While he accepted that there was genuinely a loan, he concluded that “there was not, in any meaningful sense, an incurring of expenditure of the borrowed money in the acquisition 51 of software rights. It went into a loop in order to enable the LLPs to indulge in a tax avoidance scheme.” The relevant question was not simply whether there was “real expenditure” but whether there was real expenditure “on the acquisition of software rights”.

[209]At [77] Lord Walker distinguished BMBF on the basis that in that case the full £91m had been borrowed, and the pipeline had been acquired, on fully commercial terms. In contrast, in TowerM, “the borrowed money did not go to MCashback, even temporarily, but passed in accordance with a solicitor’s undertaking straight to R&D where it produced no economic activity … until clearing fees began to flow from MCashback to the LLPs…” The Supreme Court reduced the taxpayer’s allowances to 25% of the amount claimed.

[210]The LLPs placed heavy reliance on the reasoning of the House of Lords in BMBF, in particular that the focus should be on the acts and purposes of the entity claiming the allowances, in this case the LLPs as purchasers, and that what the Developer did with the money is irrelevant. The LLPs seek to distinguish TowerM on the basis that the two central features of the arrangements in that case (the circularity of funds and the fact that the price paid was obviously and materially excessive) are not present here. We will address the extent to which these points may be present on the facts of this case below, but for present purposes we merely note that the comparison of the facts of other cases with the facts before us is of limited assistance. What is important is the principles to be derived from the authorities. As to that, we derive the following propositions from BMBF and TowerM:(1) The first step is to construe, purposively, the relevant statutory provision.

(2) It is then necessary to analyse the contractual arrangements to determine whether, in accordance with a purposive construction of the statutory provisions, they fall within them.

(3) Matters such as the presence or absence of circularity of funding, or the valuation of the assets ostensibly acquired, are relevant considerations, but only as part of the process identified in (1) and (2) above. To the extent that funds are circulated back to the purchaser, then that may indicate that they were not paid in consideration for the relevant interest. If the value of the asset ostensibly acquired, objectively assessed, was significantly less than the purchase price, then that may indicate that some or all of the ‘overpayment’ was for a different asset. As Lord Walker said, in TowerM (at [67]) “…I cannot accept that the question of valuation was totally irrelevant in the context of a complex pre-ordained transaction where the court is concerned to test the facts, realistically viewed, against the statutory text, purposively construed.”

[211]In our judgment, the overarching purpose of s.296 and s.356 is to ensure that capital allowances are restricted to expenditure of a type which is intended under Part 3 of the CAA to attract allowances. This necessitates an inquiry, first, whether the amount claimed in any given case is, as a matter of substance and not merely form, consideration solely for the “relevant interest” within s.296, or whether some part of it 52 was consideration for a separate asset. Secondly, if it was paid for the relevant interest, it requires a further analysis as to whether the relevant interest comprises partly assets representing expenditure for which an allowance can be made, and partly other expenditure, so that an apportionment under s.356 can be carried out. The parties’ respective cases

[212]As we have noted, the LLPs’ case throughout has been that the whole of the Price, in relation to each of DC2 and DC3, was paid for the relevant interest. HMRC, for their part, have throughout denied that this was so.

[213]In their closure notices issued in July 2016, HMRC objected that “some or all of the capital sum paid by [the LLPs] claimed as qualifying expenditure was not ‘for the relevant interest’”. They indicated, by way of example only, that the remote and contingent expenditure relating to the fit-out of the building was not qualifying expenditure. Otherwise, they did not attempt to identify the amount of the Price that was to be apportioned towards payment for the relevant interest.

[214]In their Amended Statement of Case, HMRC pleaded that: The capital sum due under the Sale and Development Agreement was not paid for the relevant interest in the Building by the LLP. To the extent that the capital sum was attributable to payments made for any items which were distinct from “the relevant interest in” the Building, that amount does not attract relief. Only that part of the capital sum which represented expenditure required to complete the construction of the Building and reasonable profits and charges related to its construction attracts relief. Accordingly, the capital sum should be apportioned on a just and reasonable basis under CAA, s356(1).

[215]They went on to plead that the portion of the Price which was not paid for the relevant interest related to other assets and charges “in particular (but not limited to)” three items. The first was “A proportion of the capital sum was required to pay the Yearly Sum to the LLP, which amounts were then used to repay the [Bank Winter] Loan”. They pleaded that the only or main purpose of the Bank Winter Loan was to inflate the claim to capital allowances, noting that it was cash-collateralised. The second was that part paid for the fit-out costs and the third was that part of the Price paid to purchase the land.

[216]In their skeleton argument served shortly before the trial, HMRC maintained their case that the Price was not wholly paid for the relevant interest in each of CDC2 and CDC3. They again itemised various “other assets and charges” which they said were included in the parts of the Price that were not paid for the relevant interests. These included the first and second of the matters contained in their pleading, but added three further matters:(1) the value of the EZAs available in respect of the buildings;

(2) the proportion of the Price attributable to financial adviser and promoter fees; and

(3) any part of the Price attributable to the Developer’s profit insofar as it related to non-qualifying expenditure. 53

[217]In the list of issues prepared by the parties during the trial itself, Issue 3 was formulated as follows: “Was the entirety of each capital sum(i) paid by each LLP

(ii) for the ‘relevant interest’ within the meaning of s.296 CAA 2001”, and this was said to require consideration of five sub-issues. These, in essence, enquired whether there should be any reduction in qualifying expenditure to reflect: (1) the Yearly Sum; (2) the fit-out works; (3) the effect of the EZAs; (4) the payment of the Arranger’s Fee and (5) the Developer’s profit arising from capital sums that themselves were not qualifying expenditure.

[218]By the conclusion of the trial, HMRC’s position (as to precisely what items the capital sum had been paid for apart from the relevant interest) had evolved further. They contended that this Tribunal should adopt a “realistic and pragmatic” approach to the question of how much the LLPs paid and what they paid it for. They accepted that on this approach some part of the Price was paid for the relevant interest, but suggested that this equated to something less than 30% of the Price. They identified “specific disallowances” that should be made, as follows:(1) Funds which were required to be used as collateral for the Bank Winter Loans. In the context of CDC2, this amounted to £107 million. They contended that the Bank Winter Loans resulted in no meaningful commercial risk for the LLPs, Bank Winter or members of the LLPs, and that they generated no meaningful commercial benefits for the LLPs, the members of the LLPs or the Developer. In those circumstances, the function of the Bank Winter Loans was simply to inflate the LLPs’ claims for EZAs with the result that to the extent expenditure was financed by the Bank Winter Loans, that expenditure was not attributable to assets on which EZAs were available;

(2) Funds attributable to fees;

(3) Funds attributable to the Developer’s profit;

(4) Funds attributable to interest;

(5) Funds attributable to the post-letting Yearly Sums; and

(6) Funds attributable to the EZAs. HMRC contended that 50% of the Price was, on a realistic view of the facts, paid ‘for’ EZAs.

[219]Certain of these points are clearly made in the alternative, in particular points (1) and (5), because part of the sums paid into the Restricted Account was expressed to be security for the Yearly Sum, and points (1) and (6), because self-evidently the disallowed sum cannot be the aggregate of 70% of the Price and 50% of the Price.

[220]The LLPs mounted a strong objection to HMRC being permitted to rely on arguments relating to most of these specific disallowances. They pointed out that of the specific disallowances relied on by HMRC in closing argument, only the Yearly Sums had been pleaded. Accordingly, they submitted, they were prejudiced if HMRC were permitted to place reliance on the other matters.

[221]We are not persuaded by this objection, for the following reasons. 54

[222]First, it is important to keep in mind that the LLPs bear the burden of establishing an entitlement to the capital allowances claimed by them. Their case (and their only case) has been that the whole of the Price was paid for the relevant interest, so that the whole of the Price qualifies for EZAs. HMRC’s case has been, throughout, to deny that the whole of the Price was paid for the relevant interest, in relation to each of the LLPs. It is not for HMRC to identify and prove those matters, other than the relevant interest, which were acquired in return for the Price.

[223]Second, as we have noted above, the task of the Tribunal, as indicated by the authorities (on which both sides relied) is to analyse the contractual arrangement between the parties to determine, as a matter of substance, what was acquired in consideration for the payment of the Price. All the evidential materials necessary to undertake that analysis were available to the parties and the Tribunal. Insofar as it is relevant (as we discuss below) to consider the value of the relevant interest acquired by the LLPs, we were presented with the (essentially unchallenged) evidence of Mr Ian Watson on behalf of the LLPs.

[224]Third, although we accept that it is necessary to guard against the LLPs suffering prejudice in circumstances where the Tribunal’s rules require the parties to set out what is in issue in statements of case and HMRC seeks to rely at trial on matters beyond those that were pleaded, in light of the second point we make above, we do not believe that there is such prejudice in this case. The LLPs were aware of the substantive case they had to meet – namely whether on a proper analysis of the contractual arrangements the whole of the Price was paid for the relevant interest. The precise characterisation of the other matters that they acquired, apart from the relevant interest, is of secondary importance. As it happens, as we explain below, our conclusion that the Price was not wholly paid for the relevant interest involves characterising the other rights and benefits acquired in return for the Price in a way that differs to some extent from the characterisation advanced by HMRC. That, however, is of no consequence. We are not limited, in determining what substantive rights and benefits were acquired by the LLPs in return for the Price, by the characterisation placed upon them by HMRC. The possibility that those other rights and benefits, as we characterise them, were acquired separately from the relevant interest was fully canvassed during the hearing, and we heard substantive arguments (which we address below) in relation to them from the LLPs.

[225]Fourth, of the “specific disallowances” relied on by HMRC which were not pleaded by them, the funds attributable to fees and funds attributable to the benefit of the EZAs were separately identified in their skeleton argument, and in the agreed list of issues, and it was not suggested until closing argument that HMRC should not be permitted to rely on these.

[226]We turn, therefore, to analyse the contractual arrangements relating to the acquisition of the relevant interest. 55 The contractual arrangements

[227]We will focus our analysis on the suite of documents relating to DC2. Unless we state otherwise, it can be assumed that the provisions of the documents relating to DC3 are materially the same (apart from the numbers).

[228]The contractual arrangements were detailed and complex. We set out here only the aspects that are directly relevant to the resolution of Issue 3. The two principal agreements were the SDA and the Services Agreement.

[229]Pursuant to clause 3 of the SDA, the “Price” (being the sum of £153,709,750) payable by CDC2 to the Developer was expressed to be in consideration of(i) the assignment of the Developer’s rights under the Golden Contract in favour of CDC2 and

(ii) the fulfilment of the Developer’s obligations under the SDA.

[230]By clause 2.1 of the Services Agreement, the obligation of the Developer to pay the “Yearly Sum” was stated to be in consideration of, among other things, the payment by CDC2 of the “Price” under the SDA. The Yearly Sum was defined in the Services Agreement as amounts which equalled: by paragraph (a) of the definition:(1) seven and a half years’ of interest payable by CDC2 under the Bank Winter Loan;

(2) the rent payable under the headlease of the property; and

(3) the annual operational costs of CDC2; by paragraph (b) of the definition, following a refinancing of the Bank Winter Loan, interest payable to the new lender, up to five years after the date of the Services Agreement; and by paragraph (c) of the definition, following a refinancing of the Bank Winter Loan, the difference between any rent payable by tenant(s) of the Data Centre and rent calculated on the basis of £170 psf (subject to increase of 2% per annum on a compounded basis). The Yearly Sum was payable during the “Services Period” which expired 15 years after completion of the SDA, or on DC2 being fully let to an occupational tenant who satisfied certain status requirements under a 15-year lease whose terms matched those of a “Target Lease” (as to which see [233] below), or on a termination of the Services Agreement or the sale of DC2.

[231]Pursuant to clause 4.2 of the SDA, the Developer and the Contractor were between them obliged to pay £107,669,075 (almost equivalent to the Bank Winter Loan) into an account referred to as the “Restricted Account”. Of that sum, £36,167,000 was in fact to be paid into the Restricted Account by the Contractor (in respect of the estimated fit-out costs of the building), however the ultimate source of those funds was the amount payable by the Developer to the Contractor under the Golden Contract. In addition, the Developer was required to pay £2,739,693 into an “Unrestricted Account”, as partial security for the amounts due under paragraph (a) of the definition of the Yearly Amount.

[232]The Developer’s obligations under the SDA also included delivering certain documents, including security in favour of CDC2 over the Restricted Account and the Unrestricted Account. The Developer was also required to enter into a guarantee (joint and severally with the Contractor) of the Bank Winter Loan, supported by a cash collateral deposit equal to the amount of the facility. 56

[233]The “Services” which the Developer agreed to perform under the Services Agreement included marketing DC2 to prospective tenants with a view to DC2 granting a “Target Lease” of the whole or part of DC2 to a “Target Tenant”. The concepts of “Target Lease” and “Target Tenant” sought to encapsulate, in general terms, the sorts of lease and tenant that were considered to be desirable. A Target Lease was expected to have a duration of at least 15 years from the date of grant (although the definition envisaged that DC2 might be prepared to accept a 10-year lease in certain circumstances) without any provision for early termination. Moreover, a Target Lease was expected to provide for an Initial Target Rent of £170 psf after expiry of applicable rent-free periods with provision for yearly rent reviews. A Target Tenant was, very broadly, to be either a public sector tenant, or a private sector tenant that satisfied certain financial and other requirements that were considered relevant to its ability to meet its rental obligations.

[234]In addition, the Developer was required to use reasonable endeavours to procure a refinancing of the Bank Winter Loan once DC2 was let. Such refinancing could not, however, provide a lender with any recourse to members of DC2 (as distinct from recourse to DC2 itself).

[235]In the event of a refinancing which was insufficient to repay the Bank Winter Loan in full, then by clause 5.4 of the Services Agreement the shortfall was to be paid from the Restricted Account, resulting in an interest-free Subordinated Loan from the Developer to CDC2. The Subordinated Loan could only be repaid out of proceeds received on a subsequent disposal of DC2, although the Developer’s right to obtain payment would also arise only once certain other obligations had been discharged (for example the Second Loan, the costs of disposal and the “Priority Entitlement” described further below). Following a refinancing, therefore, part of the sums standing to the credit of the Restricted Account could be applied in making an interest-free Subordinated Loan. The balance of the Restricted Account would then be paid into the Unrestricted Account which was in the name of the Developer and the Contractor. Despite its name, sums in that account were not freely at the disposal of the Contractor or Developer (since, as noted they were charged as partial security to pay amounts set out in part (a) of the definition of Yearly Sums). However, to the extent sums were not needed to cover that obligation, the Developer and Contractor could withdraw sums from the Unrestricted Account.

[236]By way of side letter (the “Funds Agreement”), the Developer and CDC2 agreed further terms that would apply if the Developer made a Subordinated Loan. That arrangement was broadly as follows:(1) The LLP could require the Developer to convert the Subordinated Loan into a membership interest in CDC2. The Developer’s interest in CDC2 would rank pari passu with that of other members except that the Developer would not be entitled to share in the 20% Priority Entitlement described more fully below.

(2) The Developer’s membership interest in the LLP would be treated as a different class of capital from that of other members but otherwise (and 57 except in relation to the Priority Entitlement) would carry the same rights as other members’ capital.

(3) The Developer agreed that the CDC2 Members’ Agreement could be amended so as to prevent the Developer from exercising control over CDC2 either alone or in concert with Piet Pulford or Guy Marsden and so as to ensure that Developer could not take decisions that would adversely affect the other ordinary members.

[237]Finally, by Clause 22 of the SDA, the Developer was required to pay to CDC2’s solicitors £9,555,716 (exclusive of VAT) which represented the “Arranger’s Fee” due to Harcourt Capital Investments Limited and CDC Administration LLP.

[238]It is apparent from the above provisions of the SDA and Services Agreement, that the rights and benefits acquired by CDC2 in return for the payment of the Price extended beyond the acquisition of the Developer’s interest under the Golden Contract and so extended beyond the acquisition of the “relevant interest”. A number of these additional rights and benefits acquired by CDC2 are expressly linked to the payment of the Price. For example, the payment of the Yearly Sum was expressly stated in the Services Agreement to be in consideration of (among other things) the payment of the Price under the SDA. The obligations assumed by the Developer under the SDA (which were expressly stated to be in consideration for the Price) included the payment of the Arranger’s Fee and the payment of substantial sums into the Restricted Account, where they were subjected to the provisions of the Services Agreement and the Funding Agreement. Importantly, those sums stood as cash collateral for the Bank Winter Loan and could be used to discharge the Bank Winter Loan and, in substance, converted into an interest-free subordinated loan or, at the option of CDC2, equity in CDC2.

[239]More generally, we consider that as a matter of substance all of the rights and benefits acquired by CDC2 under the suite of documents comprising the SDA, the Services Agreement and ancillary documents were in consideration of the payment of the Price of £153,709,750. No other consideration was provided by CDC2 for any of those benefits. In relation to the Services, for example, clause 3.2 of the Services Agreement provided that “for the avoidance of doubt” the Developer was not to charge a fee for them. All of the rights and benefits were negotiated as a package. It is a striking feature of the negotiations between the LLPs and the Developer that there was no negotiation at all on the Price. Mr Fielding candidly accepted that he did not concentrate on negotiating down the purchase price, but that his efforts were focused on negotiating the various benefits, such as the capital repayment support arrangements, which were capable of causing “real economic loss” to the Developer. He regarded this as “simply part of the commercial negotiations that took place between us.” This corroborates the conclusion that these other rights and benefits formed part of a package, together with the relevant interest in the building, acquired in consideration for paying the Price.

[240]In considering the substantive arguments advanced by the LLPs in relation to the package of rights and benefits which DC2 acquired we consider it helpful to characterise the rights and benefits, over and above the acquisition of the rights under the Golden Contract, as falling within the following four categories: 58(1) Support for the expenses incurred by CDC2 under the following heads (a) its operating expenses; (b) the interest payable on the Bank Winter Loan; and (c) the ground rent payable under the headlease of the premises. These were paid as part of the Yearly Sum, and we will refer to them as the “expenses support arrangements”.

(2) Rental support arrangements by way of payments to CDC2 to ensure that, once the building was let, CDC2 would receive rent equivalent to £170 psf, and the ancillary obligations on the Developer to market the building to prospective tenants. We will refer to these as the “rental support arrangements”.

(3) Arrangements designed to support CDC2 in repaying the Bank Winter Loan, comprising the funding of the Restricted Account so as to provide cash collateral for the Bank Winter Loan, the right to use the funds in the Restricted Account to cover the shortfall on any refinancing of the Bank Winter Loan, the provision of an interest-free subordinated loan and the right to convert that loan into equity. We will refer to these as the “capital repayment support arrangements.”

(4) The discharge of the Arranger’s Fee, which was funded by a payment from the Developer pursuant to clause 22 of the SDA.

[241]We are satisfied that these additional rights and benefits were of significant value to the LLPs. The Arranger’s Fee was in a fixed amount of £9,555,716 exclusive of VAT. So far as the Yearly Sum is concerned (which encompasses the arrangements we have characterised as “rent support” and parts of the “expenses support”), the LLPs adduced evidence from Mr Ian Watson as to their value. Faced with the difficulty that the amount to be paid by way of Yearly Sum depended on variables that were unknown at the time of the transaction, he approached this task by calculating the net present value of the amounts set out in the illustrations prepared by Mr Fielding at the time, which Mr Watson considered “would have been likely over the holding period.” His conclusion (which we accept) was that the present value of the right to receive those projected Yearly Sums was, as at April 2011, approximately £10.89 million in relation to DC2 and £7.86 million in relation to DC3.

[242]No evidence was led by either party, however, as to the value of the capital repayment support arrangements. The extent to which the Developer’s obligations in this respect would be called upon in practice depended upon what happened subsequently, in particular upon how quickly a tenant willing to pay the Target Rent could be found and upon the extent to which replacement financing could be found to repay the Bank Winter Loan. If, for example, a tenant paying the Target Rent could be found to occupy the building immediately upon completion (on the basis of minimal incentives, such as a three-month rent-free period) and the Bank Winter Loan could be fully repaid upon a refinancing then the vast majority of the funds in the Restricted and Unrestricted Accounts would be available, unencumbered, to the Developer. There was – to say the least – however, considerable uncertainty as to whether this could be achieved. It was highly unlikely that a bank could be persuaded to lend, on the security of the Data Centres, sufficient to repay the whole of the Bank Winter Loan, even if a Target Tenant could quickly be found paying the Target Rent for the whole of the 59 premises. In practice, banks would only be prepared to advance funds up to a proportion of the value of a Data Centre. Mr Fielding’s own financial illustrations assumed that an incoming bank would apply a loan to value ratio of 60%. On this assumption, which we regard as reasonable, the Developer would be required to make an interest-free subordinated loan to CDC2 of approximately £55 million for many years. Accordingly, while we are not in a position to determine the precise value of the contractual right to the capital repayment support arrangements, we have no doubt that they would have had significant value to CDC2. The LLPs’ substantive arguments

[243]Turning to the substance of the LLPs’ arguments, they first resist any attempt to apportion the Price between the relevant interest and the various other arrangements we have identified in paragraph [240] above. They contend that it is irrelevant to inquire what use the Developer made of the proceeds of the Price once paid to it, relying on BMBF. Accordingly, they say, the payments made by the Developer pursuant to the expenses support arrangements, the rental support arrangements, the capital repayment support arrangements and in relation to the Arranger’s Fee are irrelevant.

[244]We consider, however, that these arguments miss the point. Our conclusion that the Price was paid “for” these additional rights and benefits is not based on the fact that the proceeds of the Price were used by the Developer to fund the various arrangements we have identified. It is based on the fact that on a proper analysis of the contractual arrangements, these were rights and benefits which were provided in consideration for the Price. In other words, the critical point is that in return for payment of the Price the LLPs acquired, pursuant to the SDA and related contracts, the rights and benefits constituting those arrangements in addition to the assignment of the Developer’s interest under the Golden Contract. This was not an issue which arose in BMBF: it was not suggested that for the payment of £91 million BMBF acquired any asset other than the pipeline, or that this was anything other than a reasonable price for the pipeline.

[245]Our conclusion in this respect also disposes of the LLPs’ argument that TowerM is to be distinguished because it involved circularity of funds whereas in this case the Developer, at the point of the transaction, retained an economic interest in virtually all of the proceeds of the Price. The Developer was always going to be obliged to pay the Arranger’s Fee out of the Purchase Price and was also inevitably going to have to pay something under the expenses support arrangements (as the Data Centres would take some time to construct and the Developer had to defray certain of the LLPs’ expenses during the construction phase). However, we accept the LLP’s contention that, except in this limited respect, the Developer retained an economic interest in the proceeds of the Price, because, as we have noted above, the extent to which it might be called upon under the rental support arrangements or the capital repayment support arrangements depended on what happened subsequently in practice.

[246]While we accept, on that basis, that there is a difference between the facts of this case, and those in TowerM, we do not accept that the difference is material to our conclusion. That is because we have not reached our conclusion that part of the Price was paid for something other than the relevant interest because of a circularity of funds. 60 Rather, we have reached that conclusion because valuable rights and benefits were acquired by the LLPs, at the time of the transaction, over and above the relevant interest itself. The rights to have the Bank Winter Loan cash collateralised, to have interest on the Bank Winter Loan repaid, and the right to a guarantee for repayment of the Bank Winter Loan in return for a long-term interest-free subordinated loan, convertible into capital, all had value irrespective of the extent to which funds were subsequently actually paid pursuant to those rights (similar to the way in which the benefit of a guarantee is a valuable asset at the time it is conferred).

[247]It follows that we also reject HMRC’s argument (advanced in closing) that the whole of that part of the Price equal to the funds required to be used to collateralise the Bank Winter Loan was not paid for the relevant interest. Their argument in this respect is based on the contention that the whole of the £107 million consisted in essence of circular funds. This is answered, however, by our acceptance (in the previous paragraph) that the Developer retained an economic interest in the funds provided as cash collateral.

[248]HMRC also submitted, as a variant of this point, that even if the Developer retained an economic interest in the funds in the Restricted Account, that interest would only “crystallise” at some point in the future when its entitlement to have the funds released from the Restricted Account arose. Accordingly, HMRC submitted, those funds could not be said to have been “paid” at all until that later time. We reject that analysis. We do not think that the fact that an amount equal to the Bank Winter Loan was required to be held as cash collateral detracts from the fact that the whole of the Price due under the SDA was paid at the outset.

[249]The LLPs contend that even if the Price was referable to any of these other rights and benefits, then it was still paid ‘for’ the relevant interest. They contend that, far from being separate assets requiring apportionment under s.296, those rights and benefits are part and parcel of, or at least ancillary to, the relevant interest. They are, it is said, analogous to warranties or a tax deed which the purchaser of shares might obtain from the seller. We will examine this argument in relation to each of the different arrangements we have identified. Rental support arrangements

[250]So far as the rental support arrangements are concerned, in light of our conclusion in relation to the application for judicial review (where we accept that the LLPs had a legitimate expectation that HMRC would apply their practice to the effect that EZAs would be available to the extent that the Price was paid “for”) it is strictly unnecessary to decide whether the arrangements comprised a separate asset, or were part and parcel of the relevant interest. We accept, however, the LLPs’ argument that the rental support arrangements are properly analysed as ancillary to the relevant interest, as opposed to a separate asset, for the purposes of both s.296 and s.356.

[251]It is true that the Yearly Sum (within which are comprised the payments pursuant to the rental support arrangements) was separately provided for in the Services Agreement, and expressed to be provided in consideration for, among other things, the 61 Price. Accordingly, it is strictly something distinct from the relevant interest. However, we regard it (in agreement with the LLPs) as being analogous to a warranty of value that is typically provided in a property or share sale agreement. In essence, the LLPs were purchasing the right to have built and fitted out data centres that would be capable of attracting tenants paying a headline rent of £170 psf. The rental support arrangements, in substance, ensured that the LLPs would receive rent of that amount, even if the Data Centres in fact attracted only tenants who were willing to pay a lower rent. This is economically equivalent to a warranty by the Developer that the Data Centres would generate rental income of such an amount.

[252]Accordingly, we accept that in so far as the Price was paid in consideration for the rental support arrangements, it is appropriate to regard it as being paid ‘for’ the relevant interest under s.296, without any apportionment being required under s.356. Expenses support arrangements

[253]Our decision on the claim for judicial review also renders consideration of this issue strictly unnecessary. Nevertheless, and in contrast to the position in relation to rental support arrangements, we find it difficult to analyse the expenses support arrangements as ancillary to the relevant interest. We consider that the analogy with warranties such as those provided in a share sale agreement does not hold good in relation to the Developer’s obligation to pay any of them. The headlease was not part of the relevant interest, so it is difficult to see how the right to receive funds equal to the ground rent payable under it could be regarded as an aspect of, or ancillary to, the relevant interest itself. We do not see any basis on which the payment of the LLPs’ general expenses can be said to be ancillary to the interest purchased by the LLPs. The issue is more finely balanced in relation to the payments relating to interest on the Bank Winter Loan, however we find that the right to have these paid, for a period of time, by the Developer cannot properly be regarded as ancillary to the relevant interest. There is a difference, in our view, between a right to payments designed to guarantee the value of the asset being acquired, and a right to payments designed to assist the purchaser in servicing lending obtained by it in order to acquire the asset.

[254]Accordingly, we reject the contention that the expenses support arrangements were ancillary to the relevant interest. They constitute instead a separate asset for the purposes of s.296. Capital repayment support arrangements

[255]It follows from our conclusion in relation to the expenses support arrangements that we regard the capital repayment support arrangements as separate from, and additional to, the relevant interest. We find it impossible to categorise arrangements designed to assist the LLPs with repaying the capital borrowed from Bank Winter in order to acquire the relevant interest as either part of or ancillary to the relevant interest.

[256]Ms Shaw submitted that the right to assistance from the Developer with funding both the interest and capital repayment elements of the Bank Winter Loan was made necessary by the fact that the LLPs were purchasing a “greenfield” site in an enterprise 62 zone. The essence of her submission was that since the Contractor had to be paid “up front” (in order for the conditions of s296 of CAA 2001 to be met), it necessarily followed that finance had to be provided. Since the Data Centres were located in an enterprise zone, they were risky propositions and no bank would lend without some form of guaranteed income stream making it necessary for the Bank Winter Loans to be cash collateralised. Therefore, in Ms Shaw’s submission, the contractual rights to that assistance were essential features that enabled the transaction to take place in the first place. To hold otherwise would, she argued, undermine the EZA regime.

[257]Even if that were the case, we do not see that this would lead to the conclusion that the rights to which those obligations gave rise were to be regarded as ancillary to (and not separate from) the relevant interest. Their proper characterisation would still be as financial assistance to enable the purchase of the relevant interest to take place.

[258]In any event, we do not accept that they are essential features, without which the transaction could not take place. It is inevitable that if a substantial part of the purchase price of a greenfield site is to be borrowed, some security would be needed and, if the building has not been constructed, still less let, it would not be possible to provide security over a building or the cashflows it generates. That does not demonstrate, however, that cash collateralised lending is a necessary feature of a purchase of a “greenfield” site in an enterprise zone. Where the acquisition is funded by investors, as in this case, then there is no reason why the purchase price could not be funded solely by those investors. It is the purpose of EZAs to encourage investment in speculative ventures. It is true that such a transaction would not produce the striking result advertised in the Information Memoranda that an investment of £300,000 would produce almost immediate tax relief of £500,000, but where we part company with the LLPs is in the assumption that such a result is essential in order to encourage investment in an enterprise zone. The Arranger’s Fee

[259]In our judgment, the LLPs’ right – provided by clause 22 of the SDA – to receive from the Developer funds necessary to pay the Arranger’s Fee cannot be regarded as ancillary to the relevant interest.

[260]The Arranger’s Fee is payable in respect of the “Arranger’s Services”, defined as follows: The identification of potential members, the negotiation of banking facilities, the appointment of professionals, the co-ordination of all professionals, reviewing and analysing all due diligence and construction, reviewing all documentation on behalf of the Owner, paying the professional fees, SDLT and Land Registry Fees incurred by the Owner.

[261]These services were self-evidently provided to, and for the benefit of, the LLPs (referred to in the definition as “the Owner”). The “identification of potential members”, for example, refers to members of the LLPs and is solely for their benefit. Reviewing due diligence can only have been for the benefit of the purchasing entities, 63 the LLPs, and the review of documentation was expressly stated to be on behalf of the LLPs. Defraying the various fees can only have benefited the LLPs, since (as the definition states) these were incurred by the LLPs. Finally, the bank facilities were required by the LLPs in order to fund the Price, so arranging them was a service provided to the LLPs. It is true that the Developer also needed a bank facility, in order to fund – for just a few days – the payment of the construction costs to the Contractor, prior to receipt of the Price. It is unclear whether the arrangement of that banking facility was encompassed within the Arranger’s Services. Even if it was, however, it is de minimis in relation to the bulk of the services provided by the Arranger.

[262]Ms Shaw submitted that is not unusual for one party to a transaction to agree to pay the other’s fees: for example a borrower might agree to pay a bank’s fees incurred in connection with a facility agreement. We do not disagree, but that is irrelevant to the question we have to decide, namely whether the right to have fees incurred by the LLPs paid by the Developer is to be regarded as a right and benefit that is separate from the asset (the relevant interest) being acquired. On a proper analysis of the SDA, we conclude that the right to have the Developer fund the payment of the Arranger’s Fee was an asset that was in addition to, and separate from, the relevant interest.

[263]We turn, next, to consider the value of the relevant interest which, as was noted in TowerM, is at least a relevant consideration in determining whether the Price was paid wholly for the relevant interest. Value of the relevant interest

[264]We have already discussed, in Part C, aspects of Mr Ian Watson’s expert evidence on valuation on which the LLPs relied. HMRC produced no valuation evidence of their own and accepted the substance of Mr Watson’s conclusions. We have accepted Mr Watson’s evidence. Mr Watson concluded that – excluding the value of the EZAs – DC2 had a value of between £51,275,000 and £76,360,000 and that DC3 had a value of between £36,835,000 and £54,520,000. More specifically, he valued the Data Centres on three different bases:(1) On the assumption that the Data Centres were completed, fully fitted and let to grade ‘A’ tenants on a 15-year lease at rent of £170 psf uplifted in accordance with RPI, then DC2 was worth £76,360,000 and DC3 was worth £54,520,000;

(2) On the assumption that the Date Centres were completed, fully fitted and let to grade ‘A’ tenants on a 15-year lease at rent of £170 psf, but on the basis of a 30-month rent-free period, then DC2 was worth £63,730,000 and DC3 was worth £45,500,000;

(3) On the assumption that the Data Centres were completed, fully fitted and available with vacant possession, then DC2 was worth £51,2750,000 and DC3 was worth £36,835,000.

[265]Mr Watson noted, however, that none of the assumed bases on which he arrived at a valuation for the Data Centres reflected reality. As he explained, the LLPs were “effectively forward funding a completed data centre and the holding costs and letting 64 risks would be mitigated (albeit not completely eradicated) by the rental support arrangements.” His preferred analysis, therefore, was that the value of what was acquired by the LLPs in April 2011 is reflected in(1) the value of the Data Centres on the assumption that they were completed, fully fitted and available with vacant possession, plus

(2) the discounted value of the rental support arrangements (which, as we have noted in paragraph [241] above, he valued at £10,890,000 for DC2 and £7,860,000 for DC3). Accordingly, his conclusion as to the reasonable value of the interest acquired (including the benefit of the rental support agreements) by the LLPs was £59.22 million for DC2 and £42.58 million for DC3. These values represented the value of the Data Centres in the hands of the LLPs (it being common ground that the EZAs are irrelevant on any subsequent disposal of the Data Centres, and thus irrelevant to the value of the Data Centres, once acquired, in the hands of the LLPs). Mr Watson was not instructed to, and did not, take into account the benefit of the EZAs in reaching his conclusions.

[266]In many cases, there would be no necessary correlation between the objective value of the asset acquired and the amount paid ‘for’ the asset. The purchase price will usually be arrived at by a process of negotiation and will invariably, therefore, be either more, or less than, the objective value of the asset. A purchaser may pay more for the asset because it has a particular value to him, albeit one that would not be reflected in its market value, or because he simply made a bad bargain. We agree with the LLPs that in such cases the amount paid is paid ‘for’ the asset, even though it does not reflect its market value.

[267]In this case, however, as we have already noted, there was no negotiation in respect of the Price. The LLPs simply accepted the figure put forward by the Developer. In such circumstances, the fact that the market value is less than half the Price is strong corroborative evidence that the Price was also consideration for the various other rights and benefits acquired by the LLPs, as we have described above. As we explain below, this is not a case where the LLPs can realistically contend that they simply made a bad bargain.

[268]The LLPs seek to avoid this conclusion by contending that the Price was nevertheless solely ‘for’ the relevant interest because the value of the relevant interest was enhanced by reason of the availability of the EZAs. The value of the EZAs

[269]The LLPs rely in this respect upon HMRC’s concession that it is, in principle, possible for the value of EZAs to be factored into the value of the relevant interest. As Ms Shaw expressed the point in closing argument:

“Once you accept the proposition that a building in an enterprise zone is worth more than a building outside an enterprise zone there is nothing within the statutory regime which puts a limit or a cap on the amount of value that you can attribute to the benefit of the value of the allowances.”

On this basis the LLPs contend that the whole of the Price was paid for the relevant interest even though the objective value of the asset in the hands of the LLPs was less than half that sum. Ms Shaw further submits that HMRC’s case runs counter to its own practice as outlined in a note prepared by Mr David Cooper of the Inland Revenue in 65 1998. Mr Cooper’s note was an internal document and it is not suggested that it is capable of giving rise to any legitimate expectation in taxpayers for the purposes of a judicial review claim. Reliance is placed on it by the LLPs solely because it is said to indicate an understanding which is consistent with their argument.

[270]We reject these arguments for the reasons developed in the following paragraphs.

[271]We accept that a building (or a right to have a building constructed) in an enterprise zone is likely to have an enhanced value as a result of the availability of EZAs. That is because the availability of EZAs means that there are likely to be potential purchasers with a special interest (the ability to take advantage of EZAs), enabling them to outbid a purchaser who does not have that ability. The owner of the relevant interest (the Developer in this case) may, therefore, be in a position to extract a higher price. The extent to which it can do so, and thus the extent to which the value of the relevant interest is enhanced, however, is a valuation question. It is irrelevant that the legislation does not provide an upper limit on what the enhancement of value might be.

[272]As we have already noted, there was no negotiation in respect of the Price. Nor was there any attempt to identify how much the value of the relevant interest was in fact enhanced by the availability of the EZAs. Mr Pulford, the only person to give evidence on behalf of the Developer, did not identify in his witness statement how the Price was arrived at, other than to explain that it was based on a headline rent of £170 psf, and that the rental level was “validated” by CBRE. In his oral evidence, Mr Pulford was asked, in general terms, about arriving at a capital value for a building such as the Data Centres. He agreed that the capital value is arrived at by multiplying the headline rent by the yield percentage, that these would be guided by comparables and that “it comes down to the negotiation from the parties”. There is no evidence of any negotiation between the parties as to appropriate yield. Moreover, Mr Pulford does not suggest either that he looked at comparable yields himself, or that he was advised by anyone as to the appropriate yield percentage.

[273]The only evidence adduced by the LLPs as to the basis upon which the Price was accepted by them is that of Mr Fielding. In his witness statement he said only that it was “assumed that” a yield of 8% was reasonable, without taking into account EZAs, but that the availability of EZAs meant that investors were prepared to pay a higher price for the property:

“In this case, assuming a rental of £170 psf and an 8% yield and then adding on the benefit of the EZAs the purchase price for DC2 was £153,709,750.”

[274]In his oral evidence, Mr Fielding accepted that the purchase price was not something the LLPs arrived at, but he said that in order to test its reasonableness they looked at the valuation of the building, excluding the EZAs. He did this by deducting the value of the EZAs from the Price (on the basis that allowances available on a purchase price of £153 million were 50%), and then “sense-checked” the resulting figure (of approximately £76.85 million) against the value of the building excluding the EZAs.

[275]By the time the Information Memorandum in relation to CDC2 (in which the price of £153 million was identified) was issued in late January 2011, the LLPs had not 66 obtained any written advice as to the valuation of DC2. They had obtained written advice from CBRE that a headline rent of £170 per square foot, if offered with a “substantial incentive package”, would be “competitive relative to current market levels”. But CBRE was not asked to, and did not, provide any advice on the value of the interest being acquired by either CDC2 and CDC3.

[276]We have already referred to the valuation advice that the LLPs obtained from DTZ and from GVA. We have discounted the GVA valuation advice as being based largely upon inappropriate assumptions (most notably the assumption of a 20-year lease term which underpinned GVA’s higher estimates of the value of DC2). In any event, the Price was agreed upon, at the latest, at the time of the Information Memorandum. At that time, we do not accept that the LLPs had available to them any valuation other than the DTZ valuation.

[277]We have also explained in Part C above why we attach little significance to Mr Fielding’s hope that the Data Centres might have a value higher than the figures set out in DTZ’s advice.

[278]It follows that, at the time the LLPs agreed to the Price, they can only have had in mind DTZ’s estimate to the effect that, if DC2 was fully fitted and let to a Grade A tenant on a 15-year lease at a rent of £170 psf, it would have a value of £77,450,000 and a corresponding figure for DC3 of £55,300,000. Those valuations did not value what the LLPs were actually obtaining (the right to obtain unlet, but fully-fitted data centres) but instead valued what the LLPs hoped to obtain in the future (completed and let data centres).

[279]It is clear, in our view, that Mr Fielding was prepared to accept the Price on the basis that, with the full value of the EZAs to investors taken out, it was nevertheless roughly equal to the value that DTZ advised (albeit on a limited basis) the completed and fully let building would have to the LLPs. Beyond this, we find that he did not give any serious consideration to what the value of the relevant interest, taking into account the benefit of the EZAs, might be. We have already referred to Mr Fielding’s evidence in cross-examination that rather than negotiating the Price, he sought to negotiate instead the variety of benefits such as the capital repayment support arrangements. In a passage from his cross-examination which is particularly instructive in this regard, he said that the LLPs focused their efforts in ensuring that there was “a proper alignment of interest with the developer”, so that if their assumption as to the value of the building did not prove correct, then “…the developer guarantee, subordinated loan, the other mechanisms, would cause real economic loss to the developer that would effectively compensate for that purchase price being incorrect.” In other words, there was no need to worry about the Price, because the availability of all the other benefits meant that if it turned out to be too high, then the LLPs would not suffer the economic consequences of the overpayment, because value would flow back to them via the related arrangements.

[280]At one point in her submissions, Ms Shaw suggested that the Price had been worked out on the basis of £170 psf and a 4% yield. We reject that submission. There is no evidence that the Price was “worked out” in that way. The most that could be said 67 was that Mr Fielding may have calculated that the Price could be arrived at, mathematically, by assuming a 4% yield and a headline rent of £170 psf, and that this was equivalent to taking the value of the Data Centre on the assumption that it was completed and fully let at the Target Rent (as advised by DTZ) and adding the full amount of the benefit of the EZAs on the basis of a 50% tax rate.

[281]So far as the LLPs’ reliance on Mr Cooper’s note is concerned, Mr Cooper recognised that “people will pay more for a building in an EZ if it has the benefit of allowances than for one without such benefit” and that “we do not attempt any adjustment in respect of this”. The nature of the “contrived devices” which he identified as acceptable, however, fall far short of permitting the whole of the value of the EZAs to investors (i.e. 50% of the gross amount of the investment) to be added to the non-EZA value of the property. The matters identified by Mr Cooper as being artificial, but permitted in light of the purposes of EZAs, included up-front expenditure under Golden Contracts, so as to create an immediate relevant interest and secure allowances, pre-lets and developer’s construction leases. Ms Shaw relied in particular on two points mentioned by Mr Cooper: guarantees of construction costs and rent accounts and funding by non-recourse lending (i.e. on the basis that there is no recourse to the underlying investors). Neither of these, however, describes what occurred in the case of the LLPs in this case. So far as non-recourse lending is concerned, Mr Cooper’s note makes it clear that he is contemplating the lender having security only on the property, and that this has not been challenged “where real money is secured against a real asset without evidence of circularity of funding or the inflation of claims”. That is very far from permitting the 70% of the purchase price to be retained as cash collateral for the bank lending.

[282]For all of the above reasons, a consideration of the valuation evidence corroborates the conclusion that the LLPs did not pay the entirety of the Price “for” the relevant interest in the Data Centres.

[283]In reaching our conclusion we have also considered the evidence of Mr Smith on which the LLPs relied. Mr Smith gave evidence as to the nature of (in his experience) typical transactions undertaken in enterprise zones. In particular, he said that in his experience the benefits of enterprise zone status (mainly EZAs) were typically shared between the parties involved in the project – although how that was done would be a matter for the parties to decide. His view is that in deciding whether to invest, an investor should have regard to three important elements:(1) the non-tax value of the investment (i.e. what the building would be worth without the benefit of EZAs);

(2) the pre-tax price (being the total amount paid on which EZAs are to be calculated); and

(3) the after-tax price (being, in any case where the tax rate was 50%, a sum equal to 50% of the pre-tax price). He considers that what mattered most to an investor is that the after-tax price was less than the non-tax price. We did not find this evidence to be of assistance in respect of the essential question we have to decide. Ultimately, the factors that investors take into account in deciding what to invest in order to obtain the benefit of EZAs does not assist in determining whether the Price paid by the LLPs was paid exclusively for the relevant interest. HMRC’s characterisation of the additional rights and benefits 68

[284]As indicated in paragraph [224] above, our characterisation of the rights and benefits received by the LLPs in return for the Price differs to some extent from that advanced by HMRC. We address the specific matters relied on by HMRC as follows:(1) Funds attributable to collateral. We have rejected HMRC’s contention that the whole of the amount of the Price attributable to collateral for the Bank Winter Loan is to be disallowed (see [247] above). However, the collateralisation of the Loan forms part of what we have characterised as the capital repayment support arrangements. To that extent, we have accepted part of HMRC’s submissions in this respect.

(2) Funds attributable to the EZAs. We accept that the LLPs were prepared to pay more for the relevant interest because their members were obtaining the benefit of the EZAs. For the reasons already set out above, this enabled them to acquire the package of rights and benefits (such as the rental and capital support arrangements) over and above the relevant interest. To that extent, we agree with HMRC’s proposition. We do not accept, however, HMRC’s characterisation that this constituted the acquisition of a distinct asset, being “the EZAs”. EZAs arise only to the extent an amount is paid “for” the relevant interest. They are not separate items.

(3) So far as Fees, Interest and Yearly Sums are concerned, these are subsumed within our characterisation of arrangements for rent support and expenses support. We have addressed these when dealing with the substantive arguments above.

(4) That leaves Developer’s profit. HMRC contended that once it is accepted that part of the Price was paid for some element other than the relevant interest, then it must follow that to the extent that any profit element for the Developer was built into the Price, then such part of the profit that is referable to the purchase of that other element must also be disallowed. We do not find this analysis helpful. The relevant question is what was paid “for” the relevant interest, as opposed to “for” something else. An amount is paid “for” an asset, irrespective of the fact that part of that amount was arrived at by building in a profit element for the seller. It is not relevant, therefore, to identify such part of the price paid for an asset which reflected the seller’s profit. Conclusions

[285]For the above reasons, we reject the LLPs’ contention that the whole of the Price was paid “for” the relevant interest in the case of either DC2 or DC3.

[286]We accept, on the contrary, HMRC’s contention that only part of the Price was paid for the relevant interest and that an apportionment is necessary. Whether that apportionment is required under s.296 or s.356 makes no practical difference, but we consider the better view to be that it is required under s.296, as the rights and benefits acquired by the LLPs under the contractual arrangements were additional to, and separate from, the relevant interest. 69

[287]That conclusion is strictly sufficient to dispose of the LLPs’ appeal, since the only case advanced by them is that the whole of the Price was paid for the relevant interest. They have not advanced an alternative case that EZAs should be allowed on some lesser part of the Price. HMRC accepted in closing argument, however, that at least part of the Price was paid for the relevant interest. That seems to us, in any event, to be self-evidently correct: the rights under the Golden Contract satisfy the definition of a relevant interest and at least part of the Price was paid for the relevant interest. Notwithstanding, therefore, that the LLPs have not advanced any alternative case and notwithstanding that they have not called any evidence directly to support such a case, we consider it is necessary for us to go on and consider to what extent the Price should be apportioned in favour of the relevant interest.

[288]HMRC raised, for the first time in closing argument, the possibility that the Tribunal should limit itself to giving a decision in principle, rather than attempting to identify the exact proportion of the Price that was paid ‘for’ the relevant interest. The LLPs objected to this course, however, pointing out that the case had been some years in preparation, and that there should be no further delay in reaching a final conclusion. This was said, we note, in the context of their argument that HMRC should not be allowed to advance arguments that part of the Price was consideration for other rights and benefits where those matters had not been pleaded. They contended that if HMRC was in difficulty in quantifying the disallowed part of the Price then that was its fault for having failed to obtain directions for relevant expert evidence.

[289]Since we have rejected the LLPs’ contention in that regard, and have concluded that the Price was paid for certain rights and benefits which were indeed separate from the relevant interest (in particular the capital repayment support arrangements and the Arranger’s Fee), we consider it would be wrong to reach a conclusion on the amount of the Price to be apportioned to the relevant interest without hearing further from the parties.

[290]Accordingly, our conclusion on Issue 3 at this stage is limited to the following:(1) The whole of the Price was not paid for the relevant interest. It was also paid for expenses support arrangements, capital repayment support arrangements and the Arranger’s Fee.

(2) The Price is to be apportioned between the relevant interest, the expenses support arrangements, the capital repayment support arrangements and the Arranger’s Fee.

(3) Given our conclusion on the LLPs’ claim for judicial review, HMRC must grant the LLPs EZAs on the amount paid “for” expenses support arrangements. However, amounts that are paid for the capital repayment support arrangements and the Arranger’s Fee do not benefit from EZAs.

[291]We have separately released directions inviting further submissions from the parties on the following questions:(1) How, as a matter of law, the Price should be apportioned between the relevant interest and other rights and benefits. Specifically, the parties are 70 invited to consider whether s562 of CAA applies so as to require a “just and reasonable” apportionment between these items.

(2) Whether the parties should be given permission to adduce further evidence on the apportionment question and, if so, the scope of that evidence and an appropriate timetable for the preparation and service of reports. PART F: THE CLAIM FOR JUDICIAL REVIEW Introduction

[292]The LLPs’ claim for judicial review relates only to that part of the Price referable to what we have defined at [240] as the “expenses support arrangements” and “rental support arrangements” constituents of the Yearly Sum, and is advanced on the assumption that we do not accept their principal submission that all of the Price was paid for the relevant interest.

[293]Although we have concluded (in Part E above) that such part of the Price that is referable to rental support arrangements was nevertheless paid for the relevant interest, it is necessary to address the judicial review claim because we reached the opposite conclusion in connection with the expenses support arrangements.

[294]The judicial review claim relates, however, to both kinds of arrangement without distinction. That is because, although in the relevant correspondence with HMRC that is relied on to found a legitimate expectation for the purposes of the judicial review claim, HMRC referred to “rental support arrangements”, on a proper analysis of the correspondence that phrase encompassed both of the matters that we have defined separately as rental support arrangements and expenses support arrangements. We shall accordingly address the judicial review claim as it applies to “rental support arrangements” as that term was used by HMRC in the relevant correspondence.

[295]The LLPs contend, first, that HMRC, in making the argument that such part of the Price that was referable to rental support arrangements was not paid “for” the relevant interest, are departing from a long-standing practice to the effect that such arrangements did not represent an asset separate from the relevant interest. Second, the LLPs contend that they had a legitimate expectation that HMRC would apply the legislation in such a way as to permit the LLPs to obtain EZAs on the whole of the purchase price paid for the relevant interests including to the extent that purchase price was enhanced as a result of the these arrangements (provided that the rental levels underpinning the determination of that purchase price were not wholly unreasonable or grossly inflated). The principles to be applied

[296]There was little difference between the parties as to the applicable principles. Lord Carnwath summarised the concept of “legitimate expectation” in the decision of the Privy Council in United Policyholders Group and others v Attorney General of Trinidad and Tobago

[2016] 1 WLR 3383

, at [121], in the following terms: 71 Where a promise or representation, which is “clear, unambiguous and devoid of relevant qualification”, has been given to an identifiable defined person or group by a public authority for its own purposes, either in return for action by the person or group, or on the basis of which the person or group has acted to its detriment, the court will require it to be honoured, unless the authority is able to show good reasons, judged by the court to be proportionate, to resile from it. In judging proportionality the court will take into account any conflict with wider policy issues, particularly those of a “macro-economic” or “macro-political” kind.

[297]As a general matter, in order for a protected legitimate expectation to arise, a public body must make a statement that is “clear, unambiguous and devoid of relevant qualification” (see R v Inland Revenue Commissioners ex parte MFK Underwriting Agents Limited

[1990] 1 WLR 1545

). However, in an exceptional case, such a legitimate expectation can arise in the absence of such a clear and unequivocal representation (see, for example R v Inland Revenue Commissioners ex parte Unilever plc

[1996] STC 681

in which HMRC’s pattern of conduct over 20 years of accepting the taxpayer’s late claims for relief was held to be sufficient to establish a legitimate expectation that the pattern would continue).

[298]Difficult questions can arise where HMRC publish a statement setting out a view of the law that turns out to be incorrect. In such a case there is an apparent conflict between the principle that a taxpayer should be subject to the tax as determined by Parliament and the principle that legitimate expectations should be protected. In the MFK Underwriting case, the Divisional Court approached that conflict by concluding that (save where the law is clearly contrary to HMRC’s practice) HMRC have a managerial discretion to decide on the best way of carrying out their duty to collect tax and that it is within that discretion for HMRC to give advice and guidance to the public as to what it believes the tax position to be by which HMRC may be bound even if it results in HMRC forgoing tax that is lawfully due. In R (on the application of GSTS Pathology LLP and others) v Revenue and Customs Commissioners

[2013] STC 2017

, Leggatt J put the point in this way: … [T]here seems to me to be a significant distinction between the situation where the law is clear that the Revenue in the exercise of its managerial discretion declines to enforce and a situation where, as in this case, the true position in tax law is uncertain. There is no doubt that the managerial discretion of the Revenue may extend even to agree not to collect tax which, as a matter of law, is undoubtedly payable if it considers this to be in the overall interest of good administration and maximising the collection of Revenue. I would agree however that the discretion of the Revenue in such a case must be a very narrow one. But that, in my view, is very different from a case in which(1) the Revenue has given advice or guidance which it believed to be correct at the time that the advice or guidance was given, and

(2) what has happened since is not that there has been any material change in the law but simply that the Revenue has changed its view as to what it believes to be the correct tax position.

[299]In the light of the above authorities, the following three issues arise: 72(1) Whether HMRC’s statements made in correspondence with EZPUTA were “clear, unambiguous and devoid of relevant qualification” and given to an identifiable group for HMRC’s own purposes.

(2) Whether the LLPs acted to their detriment in reliance on HMRC’s statements.

(3) Whether HMRC have shown good, proportionate reasons why they should be permitted to resile from those statements. HMRC’s correspondence and practice in relation to rental support arrangements Events up to and including 1994

[300]The correspondence that the LLPs consider to give rise to their “legitimate expectation” took place in the 1990s between HMRC and the Enterprise Zone Property Unit Trust Association (“EZPUTA”). EZPUTA was analogous to a “trade body” and consisted of persons who were at the time active in promoting investments in enterprise zones. EZPUTA engaged in regular correspondence with HMRC on issues relevant to the tax treatment of investments in enterprise zones. Mr John Watson agreed to act on a pro bono basis for EZPUTA by assisting with correspondence and discussions with HMRC and we accept his evidence that HMRC regarded such correspondence as “an avenue through which problems which might otherwise impede the [enterprise zone] programme could sometimes be resolved”.

[301]One strand of correspondence between EZPUTA and HMRC focused on what we will term “greenfield site issues”. The LLPs’ claim for judicial review is not founded on this aspect of the correspondence with HMRC. However, since in some respects HMRC’s approach on greenfield site issues provides context for their approach to rental support arrangements we will provide a brief account of the correspondence on greenfield site issues.

[302]Very broadly, these issues arose where a new building was to be constructed on a previously undeveloped greenfield site in an enterprise zone. A property developer contemplating such a project would typically be reluctant to build such a building “on spec” and hope that it could be sold after construction to investors for two principal reasons.

[303]First, such a project would be risky, since there could be no guarantee that, once the expenses of construction were incurred, a purchaser could be found and this risk was particularly acute in the context of buildings in enterprise zones since those were, by definition, areas that were perceived to be less attractive to investors.

[304]Second, a developer might not have the capital necessary to fund a speculative construction of a building. EZPUTA’s members wanted to be able to promote structures that would enable investors to provide finance for the construction of buildings on greenfield sites in enterprise zones in such a way that they would qualify for EZAs on their investment. Therefore, EZPUTA sought confirmation from HMRC as to the circumstances in which investors would be able to claim EZAs in circumstances where those investors were not themselves party to any construction contract but instead 73 acquired rights “second hand” from a developer who had entered into a “golden contract” during the life of an enterprise zone.

[305]The other strand of correspondence between EZPUTA and HMRC that is relevant to these appeals was concerned with rental support arrangements. We have already explained the rationale behind the rental support arrangements that were present in the particular transactions involving the LLPs. The rationale behind rental support arrangements at the time of the correspondence with EZPUTA was similar.

[306]We accept Mr John Watson’s evidence that, at the time of that correspondence, investors in enterprise zones were typically members of the public. In some cases those investors would invest individually, or through syndicates. However, in other cases, a unit trust was formed with a UK resident trustee. In order for such a unit trust to be regarded as “transparent” for UK tax purposes (so that investors could obtain the benefit of EZAs themselves), it would need to meet conditions specified in the Income Tax (Definition of Unit Trust Schemes) Regulations 1988 (the “Unit Trust Regulations”). Whether investing directly, or via a transparent unit trust, investors tended to be sophisticated but lacking in expertise in the real estate market. Since they lacked experience in real estate matters, investors typically took a conservative approach to an investment in an enterprise zone and were not happy for their return to be wholly dependent on whether the building was ultimately let or the rent ultimately paid by a tenant. Therefore, in order for an enterprise zone investment to be marketed successfully to investors it was typically necessary for some form of rental support package to be in place.

[307]At the time of the EZPUTA correspondence, rental support arrangements typically took one of two broad forms:(1) In some cases the developer could achieve a “pre-let” in the form of an agreement by an institutional tenant to take a lease of the building once it had been constructed. Such a pre-let obviously insulated investors from some of the commercial risks of their investment. However, even where there was a pre-let arrangement in place, investors would typically require a guaranteed income during the construction phase and any rent-free period granted to the tenant. That guaranteed income would typically be provided by the developer, or a company connected with it, taking a licence to occupy the building in question and paying a licence fee to “fill the gap” until rent from the tenant started to be paid. Typically the obligation to make rental support payments would be fully cash collateralised with part of the money received by the developer when it sold the building to investors (or the unit trust through which they were investing).

(2) Often, however, a “pre-let” could not be achieved. In such a case, a developer, or an associated company would agree to pay a certain level of income for a set period (in much the same way as the Developer did in the transactions to which the LLP was party). Such rental support payments were often in legal form payments under a licence to occupy the building. Once the period covered by the rental support payments expired, investors’ 74 return would be dependent solely on rents actually received on a letting of the building.

[308]In December 1991, Mr Watson wrote on behalf of EZPUTA to Richard Steele at HMRC on greenfield site issues. On 16 December 1991, Mr Steele replied giving EZPUTA some degree of assurance that HMRC would be likely to accept that an investor acquiring a partly completed building from a property developer would be likely to obtain allowances by references to the price paid for that building.

[309]Until 1993 HMRC tended to accept that, where investors (established as a syndicate or as a unit trust) acquired greenfield sites situated in enterprise zones then, provided expenditure was incurred within applicable time limits and all statutory requirements met, EZAs would be available. HMRC’s practice was to apportion the purchase price paid by investors between a sum attributable to the building (which would attract EZAs) and a sum attributable to the land on which the building was located (which would not attract EZAs) in accordance with a fixed formula. Until 1993 there was no reason for EZPUTA to be concerned that the presence of rental support arrangements might cause EZAs to be available on an amount lower than the application of the formula would suggest.

[310]However, that position changed in 1993 when a promoter of EZAs (“Matrix”) began to promote a syndicate which was to acquire a property with a vacant possession value of some £8m for a price of £95m, the price being inflated by a combination of a lease at an unrealistic rent and a put option over the property. Matrix marketed the arrangement as having HMRC approval and, when HMRC denied that, issued a claim for judicial review. Ultimately, the claim for judicial review failed in the House of Lords.

[311]The experience with Matrix caused both HMRC and EZPUTA to focus on the potential for the purchase price of a building (and so the EZAs that could be claimed on that purchase) to be inflated by artificial ancillary arrangements. EZPUTA were keen to ensure that this renewed focus should not result in HMRC denying or restricting allowances simply because of the presence of what they regarded as benign rental support arrangements. These concerns within the industry resulted in two meetings taking place. The first was in early 1994 between members of EZPUTA and the Financial Secretary to the Treasury. The second took place on 4 February 1994 between representatives of EZPUTA and HMRC with Sue Crawford, a partner of Mr Watson at Ashurst, attending given Ashurst’s role in providing pro bono advice and assistance to EZPUTA.

[312]No note of either meeting was available. However, on 7 February 1994, Alastair Altham, the then Chairman of EZPUTA, sent John Gilhooly, a representative of HMRC who had been present at the meeting on 4 February, a letter summarising the outcome of that meeting (which also referred in places to the earlier meeting with the Financial Secretary to the Treasury).

[313]Mr Altham’s letter emphasised that EZPUTA regarded the discussions as important and as going to the heart of the question whether funding of enterprise zone investments remained viable, writing: 75 The uncertainty surrounding the availability of Tax Allowances for Enterprise Zone properties has knocked investor confidence considerably. If we are able to resolve the outstanding points quickly, it may be beneficial to all parties involved in this matter if we are able to achieve early tax clearances on a number of simple Enterprise Zone Trusts…Simple confirmation regarding the level of Allowances available for these Trusts would go a long way to restoring investor confidence. He also expressed the hope that: …we will be able to agree, quickly, the method by which Enterprise Zone funding might be allowed to continue for the current tax year.

[314]Mr Altham also recorded HMRC’s views as follows: 2 Availability of Allowances Your colleagues indicated that as the regulations governing the availability of Capital Allowances for Enterprise Zone buildings could be open to various interpretations, you were inclined to agree to a continuation of the previous procedures for Enterprise Zone investments (effectively allowing Tax Allowances on the whole purchase price, less an amount applicable to the land), at least until 6th April 1994, whereupon the whole area would be reviewed again. We agreed that the Inland Revenue would provide a Statement of Practice regarding this matter but, notwithstanding this, it was generally agreed that until 6th April 1994, Enterprise Zone property investment would continue to attract 100% Initial Tax Allowances in respect of the purchase price for the building, less an amount applicable to the land (see 6 below). In order to avoid investors claiming excessive levels of Tax Allowances, Inland Revenue were keen to ensure that rental levels underpinning the transactions could be proved to be reasonable and that they were not grossly over-inflated. In the case of an Enterprise Zone Trust, we agreed that a member of the RICS should be required to give the Trustee, not only his report on value, but also an independent confirmation that the rental level for the development is reasonable in the context of the market. We agreed that it did not seem sensible for a valuer to have to provide a report on rental levels addressed to Inland Revenue but that if the Inspector of Taxes (in determining the quantum of allowances) wished to refer the matter to the District Valuer he would be entitled to do so. With regard to pre-let transactions, we agreed that a similar rental level report should be made available to the Trustee. In this case, it was stressed that the “headline” rent payable must be proved to be reasonable in the light of the circumstances prevailing at the time of the letting. The object here was to avoid payment of substantial reverse premiums to Occupational Tenants in order to induce them to sign leases at unrealistic rents. We agreed that as long as the headline level of rent was reasonable the level of reverse premiums that the Occupational Tenant had been 76 able to extract from the developer should remain a commercial matter between themselves. 3 Licence Fees Inland Revenue were concerned that the investment value for a speculative development might be artificially inflated where the Developer agrees to pay a licence fee equivalent to the passing rent for the building during the construction process. It was pointed out however that the Purchaser was contracting for the purchase of a building (including the rental income stream attached to that building) and, having paid over his money (and made an irrevocable commitment to proceed), we would normally expect to receive a return equivalent to the initial passing rent, during the construction process. Outside the tax market, this procedure is not uncommon where buildings are being forward funded. It was generally agreed that as this practice was commercially commonplace, this issue was unlikely to cause further concern.

[315]Mr Altham’s letter also included, in section 4, a record of a discussion that had taken place on greenfield site issues and specifically the issue whether investors acquiring an interest in a building that had not yet been constructed were entitled to allowances under s1 of the Capital Allowances Act 1990 (now s294 of CAA 2001) or s10 of the Capital Allowances Act 1990 (now s296 of CAA 2001).

[316]Mr Gilhooly of HMRC sent Mr Altham a response in a letter dated 23 February 1994. He prefaced his response to the detailed points that Mr Altham had made with the following remarks:

[4]Turning now to the other issues we discussed, I hope that you will find the following proposals helpful. But I should perhaps make a couple of points first.

[5]The approach which I set out below represents an attempt at reaching a common understanding for the period up to and including 5 April 1994.

[6]You are of course right that the Revenue looks at the entirety of a transaction in determining the availability of capital allowances and, where applicable, compliance with the [Unit Trust Regulations].

[7]However, we must of course reserve the right to challenge any schemes which do not meet any of the qualifying requirements for enterprise zone allowances (not covered below); or which include any artificial arrangements aimed at securing allowances which not otherwise be available, or any scheme which cannot be justified under the arrangements set out in this letter. This would, for example, include any schemes where the principles established in Ensign Tankers are in point. In particular there is likely to be difficulty with any proposal which involves the provision of finance to investors otherwise than on full recourse terms. 77

[317]Mr Gilhooly then provided a response to points Mr Altham had made. In response to the section of Mr Altham’s letter headed “Availability of Allowances”, Mr Gilhooly wrote: Availability of allowances

[9]I confirm that, for building purchases completed on or before 5 April 1994, and subject to the points made elsewhere in this letter, we will accept claims for capital allowances on the expenditure incurred by the trustees (on behalf of the investors) less any disallowance for land as provided for in the formula attached; and subject also to the proviso that rentals underpinning transactions are at a reasonable level. To the extent that rentals underpinning the transactions are perceived to be unreasonable, the Revenue may restrict the expenditure qualifying for capital allowances accordingly. The question of whether such rentals are at a reasonable level will be determined by our having sight of a rental level report from a member of the RICS, addressed to the Trustee. We reserve the right to refer the matter to the Valuation Office; and in practice will generally do so.

[10]All this does, of course, assume the exercise of common-sense and self-restraint on the part of promoters. Otherwise there is likely to be difficulty and, perhaps, delay in considering proposals.

[318]Mr Gilhooly then quoted Mr Altham’s proposal in relation to pre-let transactions with particular reference to risk of significant reverse premiums being paid to tenants and said:

[12]We are prepared to go along with the proposals you put forward; but this is a difficult and sensitive area and a package which involves material reverse premiums may cause delay while we establish whether the rent is reasonable.

[319]Responding to Mr Altham’s proposal regarding licence fees, Mr Gilhooly wrote: Licence Fees We note what you say about licence fees. Our conclusion was that this is a matter which will be covered as need be by the Valuation Office’s opinion on the reasonableness of rents payable, as set out in our previous paragraphs.

[320]In parallel, there was ongoing correspondence between Ashursts and HMRC relating to greenfield site issues. It is not necessary to refer to all of that correspondence. It is sufficient to note that on 19 October 1994, Mr Watson sent Mr Gilhooly an analysis of a four-step transaction involving the construction of a new building in an enterprise zone. Mr Watson wrote: Let us assume for the moment that under the Golden Contract between Developer and Contractor the total consideration for works is a single lump sum payable before building commences. Contracts are then entered into under which the following things occur in the following

[321]Mr Watson commented that in his view the purchaser at Step 3 should be entitled to EZAs computed by reference to the “net price paid” for the relevant interest (the formulation set out in s10A of CAA 1990 which was the statutory predecessor to s296 of CAA 2001 in force at the time). Mr Watson went on to canvass some issues that might arise as a consequence of variation of the Golden Contract under his hypothetical transaction in the following terms: If we are agreed [that allowances were available to the purchaser], the remaining difficulty with this structure is that its success is clearly dependent on the amount paid under the Golden Contract being payable up front and upon the benefit of the Golden Contract being transferable to the Trust. In some cases where professional advice was taken when it was entered into, a Golden Contract will cover these points. In many cases, however, the contract will be in a standard industry form which simply provides for interim payments. Accordingly, it will often be necessary to vary the Golden Contract as a pre-condition to the transaction. It is a moot point as to when variations to a contract cease to be variations and result in it being replaced by a new agreement (which, being entered into after the zone would not protect allowances). Because the issue is one of degree there will often be some uncertainty on this score and that uncertainty will make it impracticable to use this route unless Revenue confirmation can be obtained in specific cases. If, however, arrangements could be made for confirmation to be given in appropriate cases that variations to the Golden Contract did not result in a new contract for the purposes of Section 10A, I think the above structure might well provide a way forward which would avoid any need to distort the law, on the one hand, and any need to distort the commercial structure on the other.

[322]On 11 November 1994, Alan O’Brien of HMRC replied to that letter. He agreed that, in principle, allowances would be available to the purchaser in Mr Watson’s four-step transaction. As regards the question of variation, he wrote:

[3]As you then go on to say, it is a moot point whether the changes made to any ‘golden contract’ will merely result in a variation of that contract or the creation of an entirely new contract. Each will turn on its own facts. Our view is that revised arrangements for payment of a purchase price need not result in a new contract but building contracts are complex documents and, for example, it is possible that changes in the terms of 79 payment would require consequential changes which might lead to the conclusion that a new contract has been created.

[4]We are prepared to look at particular cases in advance (that is before the trust is launched) until 5 April 1995…

[323]There was some further correspondence between Mr Watson and HMRC on greenfield site issues in 1994 but that was clarificatory only. In a letter dated 18 November 1994, Mr Watson said that he was “glad to see we are ad idem on the technicalities” and in his response to that letter dated 21 November 1994, Mr O’Brien did not suggest otherwise. The enactment of s10D of CAA 1990 (the predecessor to s357 of CAA 2001)

[324]The correspondence between Ashursts, EZPUTA and HMRC contained a number of suggestions that HMRC would publish a Statement of Practice (that would be published generally) setting out their approach to enterprise zone transactions. However, that never happened, primarily because targeted anti-avoidance legislation was enacted in s10D of CAA 1990 (later re-written as s357 of CAA 2001). The introduction of s10D was announced in a Budget press release issued on 29 November 1994. That press release gave summary details of the new provision in the following terms: Industrial buildings allowances The Chancellor proposes in his Budget to clarify the amount upon which capital allowances for buildings may be claimed including buildings in enterprise zones. The Chancellor’s intention is to confirm the current interpretation of the law. Details

[1]Purchasers of buildings can pay more than just the cost of the building. The purchase price often includes the provision of associated benefits such as rental guarantees.

[2]The intention is to define the amount qualifying for allowances as the price paid for the relevant interest minus(a) the value of the land element and

(b) any value attributable to elements over and above those which would feature in a normal commercial lease negotiated in the open market.

[3]The proposed measure is in line with existing practice. HMRC’s practice after 1994

[325]The correspondence with EZPUTA was expressed to set out a practice that was time-limited: Mr Gilhooly’s letter of 23 February 1994 was expressed to apply only in relation to the period ending 5 April 1994. Mr O’Brien had only confirmed that HMRC would be prepared to issue confirmations on certain greenfield site issues up to 5 April 1995. We have concluded that this was in part because the parties to the correspondence thought that a formal Statement of Practice would be issued and published generally. 80

[326]However, the enactment of s10D of CAA 1990 obviated, from HMRC’s perspective, the need for any Statement of Practice since it provided them with a statutory mechanism to deny allowances where the purchase price of relevant interests was artificially inflated. HMRC have not sought to deny that from 1994 to 2011, as a matter of practice, they followed the approach set out in the correspondence with EZPUTA and, where they had concerns as to whether purchase prices had been artificially inflated, they considered the application of s10D of CAA 1990 (which was subsequently rewritten as s357 of CAA 2001).

[327]Since we did not understand HMRC’s continued application of the practice set out in the correspondence with EZPUTA to be controversial, we can set out our findings in this regard relatively briefly:(1) Even though the EZPUTA correspondence was expressed to set out a practice applying only up to 5 April 1994, or 5 April 1995 in relation to the greenfield site issues that were discussed, they continued to apply the principles set out in that correspondence up to, and including, 2011 when the LLPs acquired their interests in the Data Centres.

(2) If HMRC were satisfied that the assumed rental levels underpinning a greenfield site transaction were reasonable, there were no artificial arrangements that might justify an application of s357 of CAA 2001 and no other technical impediments to the availability of allowances, they would proceed on the basis that a purchaser of an interest in an uncompleted building in an enterprise zone that followed Mr Watson’s “four-step” process set out at [320] would be entitled to EZAs on the purchase price paid less an amount attributable to the land on which the building was to be located.

(3) Virtually all transactions that involved investors purchasing an interest in an unconstructed building on a greenfield site in an enterprise zone between 1994 and 2011 would have involved some element of rental support arrangement. If HMRC were satisfied that the assumed rental levels in a greenfield site transaction were reasonable, they would not in practice seek to assert that EZAs should be restricted on the basis that part of the purchase price was paid “for” those rental support arrangements rather than “for” the relevant interest.

(4) In practice, between 1994 and 2011 HMRC made relatively few challenges to the availability of EZAs on “greenfield site” transactions that followed Mr Watson’s four-step approach. We have inferred that this was because Section 10D of CAA 1990 (now s357 of CAA 2001) had a deterrent effect which meant that promoters would not promote transactions that could have fallen foul of those provisions. Moreover, in accordance with the correspondence with EZPUTA, HMRC were typically given sight of advice from a member of the RICS on assumed rental levels and, occasional queries aside, HMRC tended to be satisfied with that advice. 81 The LLPs’ reliance on the EZPUTA correspondence

[328]Again, we did not understand HMRC ultimately to challenge the LLPs’ assertion that, when they were structuring their acquisitions of the Data Centres, they relied on their understanding of HMRC’s practice as set out in the correspondence with EZPUTA. While neither Mr Fielding nor Ms Brister had themselves read that correspondence, we accept Mr Baldwin’s evidence that in the course of advising the LLPs he had that correspondence firmly in mind and that his understanding of HMRC’s practice as set out in that correspondence informed his advice to the LLPs. The respective positions of the parties in the judicial review

[329]The LLPs consider that the requirements that Lord Carnwath set out in United Policyholders Group are all met:(1) HMRC’s practice set out in the EZPUTA correspondence was “clear, unambiguous and devoid of relevant qualification”. HMRC gave that representation to an industry body for HMRC’s own purposes knowing that it would be relied upon.

(2) The LLPs were entitled to, and did, act to their detriment in reliance on HMRC’s guidance. They were within the terms of HMRC’s practice since the assumed rental level of £170 psf that underpinned their transactions was reasonable. The LLPs therefore had a legitimate expectation that HMRC would act consistently with that practice.

(3) By asserting that part of the LLPs’ purchase price was paid for rental support arrangements (and so did not qualify for EZAs), even though the underlying assumed rental levels were commercially reasonable, HMRC are acting inconsistently with their practice. They are not acting inconsistently because their practice has been shown to be wrong in law. They are doing so simply because they have changed their mind. In all the circumstances, there is no good reason why they should be permitted to resile from their practice.

[330]HMRC argue that the judicial review claim should be dismissed for the following broad reasons:(1) The correspondence with EZPUTA did not set out a representation that was clear, unambiguous and devoid of qualification primarily because Mr Gilhooly’s letter of 23 February was heavily caveated.

(2) In any event, the assumed rental level of £170 psf that underpinned the LLPs’ transactions was unreasonable, so the LLPs were outside the terms of the practice.

(3) Even if the EZPUTA correspondence did set out a clear representation to the effect that the LLPs’ entitlement to EZAs would not be restricted by reference to an amount paid “for” rental support arrangements, the LLPs cannot rely on the EZPUTA correspondence to substantiate a claim for allowances to which they are not in law entitled. 82

(4) HMRC should be permitted to resile from the statements in the EZPUTA correspondence on public policy grounds.

(5) Section 31(2A) of the Senior Courts Act 1981 provides that the Upper Tribunal must refuse the LLPs the remedy of judicial review since it is highly likely that their claims for EZAs would have been refused for the reasons HMRC gave in connection with the LLPs’ substantive appeals. Discussion of the claim for judicial review Whether the assumed rental level was unreasonable

[331]We reject HMRC’s argument set out at [330(2)]. The effect of our conclusion at [163] is that the assumed rental level of £170 psf that underpinned this transaction was reasonable. Moreover, we have accepted Mr Ian Watson’s expert evidence that, even though a tenant would require the LLPs to offer a rent-free period of around 24 months (in a 10-year lease) in order to take a lease of either Data Centre at a headline rent of £170 psf, that does not call into question the reasonableness of the headline rent (see [162] above). Whether HMRC made a representation that was clear, unambiguous and devoid of relevant qualification

[332]In their skeleton argument served prior to the hearing, HMRC relied strongly on the caveats in Mr Gilhooly’s letter of 23 February 1994. Given those caveats, HMRC submitted, in paragraph 35 of that skeleton argument: At most the correspondence [with EZPUTA] could only amount to an acceptance that qualifying expenditure might, in principle, include amounts referable to rental support income arrangements.

[333]However, in paragraph 112 of their written closing submissions, HMRC retreated from that position saying: 112 HMRC accept that their practice (set out in the relevant correspondence with EZPUTA) has since the 1990s been to accept that amounts which might be said to be attributable to rental support arrangements are paid “for” the relevant interests in question, provided that the rental support arrangements are reasonable…. 113 In this case the rental support arrangements were not “reasonable”. This is because they were based on a rent of £170 per square foot which was itself not reasonable…

[334]Ms Nathan’s oral closing submissions on this issue similarly focused on what she submitted was the “unreasonable” nature of the rental support arrangements because the figure of £170 psf was excessive. In the course of those submissions, we canvassed with Ms Nathan the example of a transaction which contains both “reasonable” rental support arrangements (based on a reasonable estimate of rent payable) but also other aspects which had an artificial effect on value which HMRC considered to be objectionable. We asked Ms Nathan whether HMRC’s position is that the other 83 arrangements would “infect” the reasonable rental support arrangements such that it would be open to HMRC, notwithstanding the practice set out in the EZPUTA correspondence, to argue that part of the purchase price was paid “for” reasonable rental support arrangements and so did not qualify for allowances. Ms Nathan’s response was that HMRC’s practice permitted the other arrangements to be challenged (under s357 or otherwise) but that these arrangements did not “infect” the rental support arrangements which HMRC’s practice treated as being not subject to challenge.

[335]Given that articulation of HMRC’s position, we have concluded that HMRC did make a representation that was clear, unambiguous and devoid of relevant qualification that, provided the assumed rental level underpinning the transaction was reasonable, HMRC would not seek to argue that part of the purchase price was paid “for” rental support arrangements so as not to qualify for EZAs. That leaves the question of what “rental support arrangements” were included within the scope of HMRC’s representation. That must be determined by reference to the kind of “rental support arrangements” that HMRC had in mind at the time of the EZPUTA correspondence. We have summarised, at [307] conclusions that we have drawn from Mr John Watson’s evidence as to the “rental support” arrangements with which both EZPUTA and HMRC were concerned. It can be seen that those arrangements embraced both the payment of sums sufficient to defray expenses while a building was being constructed, or before it was let and the payment of sums to “top up” rent received after a building is let to an agreed level. As such, we have concluded that both “rental support arrangements” and “expenses support arrangements” as defined by us at [240] were covered by HMRC’s practice.

[336]We do not in any event understand the LLPs to be arguing that the representation extended beyond these. Since HMRC’s practice was limited to what we have termed “expenses support arrangements” and “rental support arrangements”, even if the assumed rental levels underpinning the transaction were entirely reasonable, HMRC’s practice did not preclude them from arguing that the LLPs nevertheless paid an amount “for” the capital repayment support arrangements and the Developer’s agreement to discharge fees referred to at [240] and in our conclusion on Issue 3 set out at [290].

[337]HMRC’s representation was given to an identifiable group. In the first instance it was given to EZPUTA. However, HMRC were aware that EZPUTA was an industry body that represented persons active in promoting investments in enterprise zones. Therefore, HMRC were aware that the representations they made would be relied upon by persons who were making, or proposing to make, investments in such zones. As such, HMRC’s representation was made to investors such as the LLPs.

[338]Moreover, HMRC made their representations “for their own purposes”. An important aim of the EZA regime was to stimulate investment in areas perceived as disadvantaged. As a government department, therefore, HMRC wished to ensure that the necessary stimulus was delivered while, at the same time, avoiding undue risk to the exchequer should the reliefs granted be abused. 84

[339]As we have noted above, the LLPs were within the terms of that representation since the assumed rental levels underpinning their transactions were commercially reasonable.

[340]Accordingly, we conclude that the requirement set out at [299(1)] above was met. The requirement set out at [299(2)] was also met given the conclusions we have reached at [328] above. Whether there is a good reason why HMRC should not be bound by their practice

[341]HMRC advance two arguments why they should not be bound by their practice.

[342]First, as a matter of law, the part of the purchase price that the LLPs paid “for” the rental support arrangements did not attract EZAs and therefore, they should not be held to a statement in their correspondence with EZPUTA to the effect that EZAs should be available on that part of the purchase price. In their skeleton argument, HMRC went as far as submitting that HMRC had no power to confirm that EZAs would be available when the legislation provided that they were not.

[343]We reject this argument. In 1994, there was no clear answer to the question whether a person in the position of the LLPs could be regarded as paying part of the purchase price “for” those rental support arrangements (with the result that allowances were not available on that part). That remained an open question at the time of the hearing before us. In the event, we have answered it, in part, in favour of the LLPs. Moreover, we were able to provide an answer only following a detailed examination of the arrangements to which the LLPs were party.

[344]This is not a case, therefore, where HMRC were giving a confirmation that EZAs would be available in a situation where Parliament had clearly legislated to say that they would not. Instead, the situation is analogous to that examined in the GSTS Pathology LLP case referred to above. Accordingly, we conclude that HMRC were acting well within their “managerial discretion” as to how the EZA regime should be administered when giving that confirmation to EZPUTA.

[345]HMRC’s second argument is that, if the guidance that HMRC gave was wrong as a matter of law, that would represent a good proportionate reason for HMRC to resile from that guidance. We also reject that argument. EZPUTA had approached HMRC for confirmation as to how the law would be applied in an area where its application was open to debate. Moreover, EZPUTA and HMRC had a shared interest in a publication of HMRC’s view of the law. From HMRC’s perspective, publication of that view might be expected to enable the EZAs to provide a stimulus to investment with appropriate safeguards to the exchequer and, from EZPUTA’s perspective, investors would be provided with comfort as to the circumstances in which allowances would be available. In those circumstances, it would be unjust and disproportionate for HMRC to be permitted to disavow their guidance even if, with the benefit of hindsight, it did not set out an accurate statement of the law. 85

[346]We therefore conclude that there is no good, proportionate reason why HMRC should be permitted to resile from their guidance. Section 31(2A) of the Senior Courts Act 1981

[347]Section 15(4) of The Tribunals Courts and Enforcement Act 2007 provides that, in deciding whether to grant the remedy of judicial review, the Upper Tribunal must apply the principles the High Court would apply in judicial review proceedings.

[348]Section 31(2A) of the Senior Courts Act 1981 provides: (2A) The High Court – (a) must refuse to grant relief on an application for judicial review… if it appears to the court to be highly likely that the outcome for the applicant would not have been substantially different if the conduct complained of had not occurred.

[349]In their skeleton argument, HMRC referred to other reasons, in addition to their assertion that part of the purchase price was paid “for” rental support arrangements, why, in their submission, the LLPs are not entitled to EZAs. Those other arguments were:(1) HMRC’s argument that, because the LLPs were not carrying on business with a view to profit, they fell to be treated as “opaque” bodies corporate and that their expenditure on the Data Centres was incurred after the cut-off date of 1 April 2011 applicable to bodies corporate (Issue 2);

(2) HMRC’s argument that expenditure was not incurred “under” the Golden Contract (Issue 1); and

(3) HMRC’s argument that part of the amount on which the LLPs were claiming allowances represented the price payable for land, on which EZAs are not available, an argument that HMRC withdrew by the time of the hearing.

[350]We reject HMRC’s arguments for the following reasons:(1) To the extent that HMRC pursued the above arguments at the hearing, we rejected them. Therefore, we are not satisfied that these other arguments make it “highly likely” that the LLPs would have been denied allowances altogether.

(2) In any event, HMRC’s reliance on s31(2A) is misconceived. The LLPs correctly accept that, if HMRC’s arguments on either Issue 1 or Issue 2 succeed, that would be a bar to any claim for EZAs. Therefore, the judicial review application only needs to be considered (i) if HMRC’s arguments on Issue 1 and Issue 2 fail and (ii) if we are wrong in our conclusion that the LLPs did not, as a matter of law, pay any of their purchase price “for” rental support arrangements as opposed to the “relevant interests” in the Data 86 Centres. In that scenario, the “conduct complained of” (being HMRC’s refusal to apply their guidance) would make all the difference since, if HMRC applied that guidance, the LLPs would be entitled to EZAs without any reduction for an amount said to have been paid “for” rental support arrangements whereas, if HMRC do not apply that guidance, the LLPs’ entitlement to allowances would be reduced. Conclusion on the claim for judicial review

[351]For the above reasons, we conclude that HMRC is precluded from denying the LLPs’ claim for EZAs by reference to an amount of expenditure said to have been paid “for” the matters we have identified as “rental support arrangements” and “expenses support arrangements”, notwithstanding that in relation to the latter (but not the former) we have concluded that as a matter of law it was not paid “for” the relevant interests in the Data Centres. PART F – DETERMINATION OF THE APPEAL AND THE CLAIM FOR JUDICIAL REVIEW

[352]Our overall conclusion is that the LLPs’ appeals against HMRC’s closure notices are allowed in part, and the claim for judicial review is allowed, as follows:(1) While neither of the LLPs was, in the tax year ended 5 April 2011, entitled to EZAs on the basis of the whole of the Price paid by them for their respective Data Centre, each of them was entitled to EZAs on some part of the Price paid.

(2) As a matter of principle, no reduction to the LLPs’ claim for allowances is to be made on the basis that part of the Price was paid for rental support arrangements (as we have defined that concept at [240] above).

(3) In any event, HMRC is precluded from denying the LLPs’ claim for EZAs by reference to an amount of expenditure said to have been paid “for” the matters we have defined (at [240] above) as “rental support arrangements” and “expenses support arrangements”.

(4) As a matter of principle, part of the Price that the LLPs paid was paid “for” capital repayment support arrangements and “for” the Developer’s agreement to discharge the Arranger’s Fee referred to at [240] above. To this extent, the Price paid by the LLPs does not qualify for EZAs.

(5) We invite further submissions from the parties as to how the apportionment of the Price between that which does, and that which does not, qualify for EZAs is to be calculated.

[353]We will extend the time limit for applying for permission to appeal against this decision until resolution of the further issues identified above, or further order in the meantime. 87 MR JUSTICE ZACAROLI JUDGE JONATHAN RICHARDS RELEASE DATE: 15 November 2019

order

78 Step 1. It is agreed to begin work so that the lump sum becomes immediately payable under the terms of the Golden Contract. Step 2. The Developer pays that lump sum to the Contractor. The Developer is now entitled to have the building constructed for no further consideration. Step 3. The Developer transfers to the Trust its interest in land with the benefit of its rights under the Golden Contract to have the building constructed. Step 4. Work begins.

Cobalt Data Centre 2 LLP and Cobalt Data Centre 3 LLP v The Commissioners for HM Revenue and Customs:[2019] UKUT 0342 (TCC) [2019] UKUT 0342 (TCC)

order

Cited in 6 later judgments