London Luton Hotel BPRA Property Fund LLP v The Commissioners for HMRC

This is our decision on appeals by London Luton Hotel BPRA Property Fund LLP (the “LLP”) and by HMRC against a decision of the Upper Tribunal (“UT”), Michael Green J and Judge Thomas Scott, reported at [2021] UKUT 147 (TCC) (the “UT decision”). The UT decision partly reversed a decision of the First-tier Tribunal (“FTT”) reported at [2019] UKFTT 212 (TC) (the “FTT decision”).

The appeals relate to a claim by the LLP for a form of capital allowance known as business premises renovation allowance (“BPRA”) in its tax return for the year ended 5 April 2011. The claim was made in respect of the conversion of a former flight training centre located near London Luton Airport and owned by the LLP (the “Property”) into a Ramada Encore hotel. The amount claimed was £12,478,201. The LLP appealed to the FTT against HMRC’s decision that £5,255,761 of the LLP’s claim should be disallowed, with the result that its claim for BPRA was reduced to £7,222,439.

Although LLPs are separate legal entities from their members they can effectively be treated as transparent for certain tax purposes. The LLP’s members included individual investors who sought to benefit from the LLP’s claim to BPRA by setting their shares of it against their total income for tax purposes for the year ended 5 April 2011 (“sideways” loss relief). We are concerned solely with the LLP’s claim for that tax year.

The LLP’s primary case was and remains that it is entitled to BPRA in the full amount claimed. It maintains that the whole of that amount reflects expenditure that the LLP incurred under an agreement (the “Development Agreement”) that the LLP entered into with a developer, OVL (Bankfield) LLP (“the Developer” or “OVL”), for the conversion of the Property.

HMRC contend that it is necessary to look beyond the terms of the Development Agreement and break down the total sum into constituent elements. Both the FTT and UT agreed with that contention, with the result that each Tribunal considered the LLP’s secondary case that each of those elements amounted to qualifying expenditure in any event. Different conclusions were reached on a number of the elements and both parties now appeal to this Court, with the permission of the UT.

HMRC maintain that the outcome of this case will be relevant to other claims for BPRA totalling in excess of £100m of expenditure.

The structure of this judgment is as follows:

The BPRA legislation is contained in Part 3A of the Capital Allowances Act 2001 (“CAA 2001”). Unless otherwise indicated, statutory references are to provisions of the CAA 2001 as in force for the year ended 5 April 2011.

The BPRA was introduced by the Finance Act 2005 as a time-limited relief intended to incentivise the bringing back into use of unused business property in certain areas that were designated as disadvantaged. In outline, a 100% initial allowance was available if a person incurred “qualifying expenditure” in respect of a “qualifying building” in which the person incurring the expenditure had an interest (ss.360A and 360G).

“Qualifying expenditure” is defined by s.360B as follows:

The terms “qualifying building” and “qualifying business premises” are defined by ss.360C and 360D respectively. In summary, a qualifying building is a building or structure (or part thereof) which is situated in an area which had been designated as a disadvantaged area, had been unused for at least one year and prior to that had last been used for the purposes of a trade, profession or vocation, or otherwise as offices. Section 360D defines “qualifying business premises” as follows:

Allowances granted were subject to adjustment if a “balancing event” occurred. This included a sale of the property or the building ceasing to be qualifying business premises. However, at the relevant time no adjustment was made if the event in question occurred more than seven years after the building came back into use or became available for use (ss.360M and 360N).

There is no dispute that the Property was and is a qualifying building, that it became qualifying business premises as a result of its conversion from being a flight training centre to a hotel, and that the LLP “incurred” the disputed expenditure prior to the expiry date. The dispute is over whether the expenditure incurred by the LLP amounted to capital expenditure on or in connection with that conversion, within s.360B(1)(a).

The facts were summarised by the UT as follows:

In what follows we will adopt the same definitions as the UT. We will also refer to the “conversion” as a shorthand for the activities referred to in s.360B(1)(a)-(c), as applied to the facts of this case.

As already indicated, the LLP’s primary case, and its first ground of appeal to this Court, is that the entire amount claimed qualifies for BPRA. The FTT and the UT were against the LLP on this point. But the LLP maintains that both Tribunals wrongly focused on how OVL was to spend the Development Sum and not on the fact that the LLP had incurred that expenditure. We will refer to this as “Issue 1”.

If that challenge fails and it is necessary to consider constituent elements of the Development Sum (“Issue 2”), then the following five elements identified by HMRC as comprised in the Development Sum remain in dispute:

HMRC appeal in respect of the UT’s decision on each of these amounts, other than the Interest Amount in respect of which the LLP appeals. We will refer to the individual issues as Issues 2(a), 2(b) etc.

The UT disagreed with the FTT in important respects. Of the amounts listed within Issue 2 above, it reversed the FTT’s disallowance of the Capital Amount, the Franchise Costs (so far as the Sanguine payment was concerned) and part of the Residual Amount. It also reversed the FTT’s allowance of the Interest Amount. It agreed with the FTT that the fees paid to the Promoters and IFAs, and the payment to Ramada, should be allowed.

In addition to the specific items referred to above HMRC also have a more general ground of appeal (ground 5 of their grounds of appeal). This covers two related points, namely what HMRC say was an incorrect rejection of an argument that the BPRA claim had been “ramped up” by reference to uncommercial and circular arrangements (particularly the Capital Account and Interest Account mechanisms discussed further below) and what is said to have been an incorrect approach in respect of valuation, where HMRC say that the UT was wrong to conclude that Tower MCashback LLP 1 and another v HMRC [2011] UKSC 19, [2011] 2 AC 457, [2011] STC 1143 (“Tower MCashback”) could be distinguished on the basis of a lack of discrepancy between the expenditure claimed and the value of the asset.

There are two further elements which were in issue before the FTT and the UT but which are not in issue on this appeal. These comprise certain legal fees, which the UT found not to be eligible and in respect of which Lewison LJ refused the LLP permission to appeal, and an amount labelled as being in respect of fixtures, fittings and equipment (“FF&E”) but which the UT commented were in reality elements of the overall construction costs and agreed with the FTT were allowable. HMRC did not seek permission to appeal in respect of the FF&E amount. It is worth noting here that this FF&E amount should not be confused with a separate arrangement under which approximately £685,000 of the funds raised by the LLP was spent on the provision of FF&E that was not the subject of any claim to BPRA. This amount is referred to as a separate loan by OVL at [25] of the UT decision (set out above), but in fact it formed part of the Developer Loan of £1,985,000 referred to at [16] of its decision.

The net effect of the UT decision was that the LLP’s claim was allowed in its entirety, with the exception of the Interest Amount and the legal fees. The LLP estimates that this amounts to approximately 96% of its original claim. As a result, it filed its application for permission to appeal to this Court only once HMRC sought permission.

The essence of the legal dispute is the breadth of the concept of capital expenditure “on or in connection with” the conversion. Simplifying somewhat, HMRC’s starting point is that it covers the cost of the physical conversion work, together with certain additional amounts regarded as sufficiently closely related to that work rather than to anything else, for example the acquisition of the land. The LLP says that HMRC’s approach disregards the requirement of the legislation to consider the LLP’s expenditure rather than OVL’s and the lack of any finding by the FTT that the Development Sum was “ramped up” to maximise the BPRA claim. The LLP also maintains that HMRC’s arguments fail to take proper account of the scope of the phrase “in connection with” and the fact that (having failed before the FTT in an attempt to amend their statement of case) HMRC are not permitted to maintain that any part of the expenditure was not incurred by the LLP.

Insofar as not summarised by the UT in the passage set out at [14] above or in the FTT’s findings in relation to individual issues (as to which see below) the following findings are worthy of note for the purposes of this appeal, references in square brackets being to paragraphs of the FTT decision.

OVL was part of the Cannock group of companies (and is generally referred to as Cannock in the FTT decision, reflecting the fact that it was renamed as Cannock Projects LLP). Michael Tracey, a chartered surveyor, was a director and shareholder. At the time the majority shareholder was Stephen Bantoft, who died in 2015 ([38]). Cannock had been involved in developments and property regeneration projects in enterprise zones since 1995, before moving into development projects that qualified for BPRA. Prior to the London Luton project Cannock had undertaken six BPRA projects with Downing ([48]).

Downing, or strictly its predecessor entity Downing Corporate Finance Ltd, was the promoter of the LLP. Its founding partner was Nicholas Lewis, who was also a partner in Blakes. There was a profit-sharing arrangement between Downing and Blakes under which Blakes was allocated most of the fees from the project, and to this end Blakes was appointed as “property adviser”. Downing and Blakes agreed with Mr Bantoft that their aggregate fees for the project would comprise 2% of the gross proceeds raised (the “Promoter fees”) and 15% of any release from the Capital Account, together with 15% of other amounts received (including interest). In the event Blakes received £260,000 out of a total of £310,000 attributable to the funds raised and was entitled to the full 15% of any other amounts received ([38] and [115]-[116]). We note that £310,000 is 2% of £15.5m. That figure corresponds to the aggregate of the debt and equity funds raised by the LLP if the £685,000 allocated to FF&E is excluded.

It was Mr Tracey who was approached by the owner of the Property, Chainridge ([48]). A conditional agreement to buy the Property was entered into on 9 December 2009 to which Chainridge, OVL and the LLP were all parties. The contract went through a number of iterations before it was finally completed on 25 March 2011. OVL paid the bulk of the deposit and also itself bought some adjoining land for use for car parking. Planning permission for the conversion was obtained during 2010 ([51]-[53]).

OVL explored franchise agreements for the Property, successfully concluding an agreement with Wyndham Worldwide for a Ramada Encore. There were also discussions with Sanguine in relation to anticipated trading figures and advice on the layout, design and finishes for the completed hotel, but OVL subsequently identified ThenHotels as suitable managers and operators and decided to use them rather than Sanguine to advise during the design and development period as well. The FTT recorded at [102] that the Operating Company entered into a management agreement with ThenHotels on 25 March 2011. It also recorded Mr Tracey as saying that the Sanguine payment was not a franchise cost “but a sum agreed between Sanguine and Mr Bantoft as a result of the decision to use [ThenHotels] and not Sanguine” ([54]-[55]).

As far as Ramada was concerned, the FTT found at [119] that the total of £24,862 paid to Ramada that was then in dispute comprised an initial fee of $15,000 paid for a licence to use the Ramada brand and £15,000 paid pursuant to a Technical Services Agreement to which the Operating Company was a party. The latter payment is no longer in dispute.

The FTT also found at [56] that ThenHotels prepared financial projections that were reviewed by TRI Hospitality Consulting (“TRI”), a management consultancy specialising in data analysis for the hotel sector. TRI’s final report concluded that the projections “appear reasonable, accurately reflecting the anticipated market position of the hotel and the strength of the market opportunity”. These projections were used in the Edward Symmons valuation.

OVL put the construction contract out to tender. There were initial discussions with Multibuild in April 2010. The Design and Build Contract was entered into on 24 March 2011 (so before the LLP acquired the freehold) conditional on the Co-op Loan Agreement being entered into, and with provision for the contract to fall away if that did not occur by 5 April 2011. The FTT accepted Mr Tracey’s evidence that this reflected usual practice, so as to avoid a delay in work being commenced and the risk of an increase in costs ([57]-[59]).

A FF&E Supply Agreement was also entered into between OVL and Multibuild on 24 March 2011 on a similarly conditional basis, for £685,000, although OVL also agreed under a separate deed to meet additional costs up to a cap of £817,896. The FTT described the FF&E Supply Agreement as relating to “loose” FF&E ([60]-[61]).

Bank finance for the project was difficult to obtain following the 2008 global banking crisis. OVL was responsible for obtaining it and was turned down by several banks before approaching the Co-op in December 2010. The “outline details” of the proposal (the “Debt Finance Request”) provided to David Matthews, a senior business development manager at the Co-op, included details of the various bank deposits that were proposed to be secured. This document said the following about the Capital Account:

Discussions proceeded with the Co-op, with the funds sought increasing to £7m. The FTT decision records that when a query was raised internally at the Co-op as to why £2m was to be deposited rather than the debt just being reduced by that amount, Mr Matthews said that he had explained to his immediate superior:

The valuation report by Edward Symmons was produced at the Co-op’s instruction but following an approach to the bank by that firm at the request of Mr Bantoft, Edward Symmons having already valued the hotel for investors ([75], [76], [79], [82]). The FTT recorded Mr Matthews’ evidence that he had no concerns about that, Edward Symmons being “well known and trusted by the bank” and a firm that would not risk its relationship with the Co-op by providing an erroneous valuation, and that a pragmatic approach was taken in view of the tight timeframe ([83]-[84]). Later in the decision, at [136], the FTT made clear that Mr Matthews’ evidence on this point was accepted.

The FTT found that, while the Information Memorandum was produced by Downing, Mr Bantoft of Cannock took a significant interest and played a part in its production, and was also actively involved in attracting investors and providing information to IFAs ([95]-[100]).

As already mentioned, £310,000 was paid by OVL as Promoter fees, comprising £50,000 to Downing and £260,000 to Blakes. In addition, £372,424 of the Development Sum was used to pay the fees of IFAs, comprising £209,552 of fees paid by Downing and reimbursed by OVL and £162,872 of fees paid directly by OVL ([114]). We refer to these as the “IFA fees”.

The FTT also found that between 25 March 2011 and 10 October 2012 the LLP invoiced OVL for “rent” in the total amount of £316,120 (excluding VAT), and that OVL retained the difference between that amount and the £350,000 that had been deposited ([110]). In addition, the FTT referred to a Costs Agreement between OVL and Downing under which OVL agreed to bear Downing’s costs associated with the transaction.

In a discussion of subsequent events at [123]-[128] the FTT recorded that, in the light of a meeting with HMRC on another project, the claim made for BPRA excluded an amount of £34,999 in respect of legal fees, resulting in a claim for £12,478,201 rather than the full Development Sum of £12,513,200. This was regarded as a pragmatic decision in the hope that the claim could be concluded swiftly.

The FTT found that the conversion was completed in July 2012 and trading commenced, but the results for the first two years were below what had been forecast due to unanticipated competition from another new hotel. This led to a breach of the financial covenants in the Co-op Loan Agreement, and although the loan was not formally called in the Co-op drew down the £2m from the Capital Account on 12 February 2014 “increasing the LLP’s indebtedness to [OVL]” by that amount. The FTT understood that, following a rebranding, the LLP was now generating significant profits and cash, and its performance was converging with the projections set out in the Information Memorandum. The Co-op Loan had not only been refinanced by NatWest but the £2 million added to the Developer Loan had been repaid.

At [131]-[142] of its decision the FTT considered two issues that had been the subject of focus at the hearing before it, namely the relationship between the parties and valuation. It noted that HMRC’s case “did not stop far short” of alleging collusion between OVL and the LLP to increase the BPRA claim, whereas the LLP argued that the agreements were all negotiated at arm’s length and the Development Sum was the amount required to be paid to secure the conversion of the Property to a hotel. In response to a submission that there was no record of any negotiation between Downing, the LLP and OVL the FTT commented that OVL and Downing had an established business relationship such that an absence of detailed negotiations was perhaps not surprising. The FTT also noted that the parties were not “connected” in a statutory sense and “as such, any transactions between them are to be regarded as being at ‘arm’s length’ commercial transactions”.

As to valuation, the FTT noted that the Edward Symmons valuation, while independent, related to the anticipated value of the completed conversion rather than to any particular works. HMRC submitted that that valuation could not be used to test the market value of the works undertaken or to determine the purpose for which the LLP paid the Development Sum, and maintained that its own expert evidence showed that it was in any event an overvaluation. The FTT rejected the last of these points, finding that there was no reason to doubt the integrity of the Edward Symmons valuation and preferring the evidence of the LLP’s valuation expert to that of HMRC. As regards its relevance to the value of the conversion works, the FTT agreed with the LLP’s submission that HMRC were confusing costs with value. What investors were concerned with was what they got for their money, which was the Property converted into a Ramada Encore hotel. The Information Memorandum showed that the price paid by the investors was what the market was prepared to pay.

The UT commented on the FTT’s findings about the parties being at arm’s length and about valuation. It correctly noted at [46] of its decision that there was no relevant statutory test of connection and that the question of whether parties were at arm’s length depended on the facts. It observed at [47] and [48] that the FTT did not address HMRC’s substantive submission that the transactions were deliberately structured to inflate the BPRA claim and proceeded on the basis that there were no findings of fact to support HMRC’s assertion to that effect.

It is also worth noting that, while the LLP’s expert evidence appears to have been confined to a valuation of the Property, HMRC’s expert evidence appears to have included not only a valuation of the hotel but evidence in respect of the conversion works, including the reasonableness of the costs ([38]-[40]). The FTT decision at [42]-[43] records challenges by the LLP to the weight to be accorded to at least some of HMRC’s expert evidence, but beyond what is summarised above appears to have reached no conclusions on the expert evidence.

There is no dispute between the parties that Barclays Mercantile Business Finance Ltd v Mawson (Inspector of Taxes) [2004] UKHL 51, [2005] 1 AC 684, [2005] STC 1 (“BMBF”) provides the correct starting point. That was a case in which the House of Lords held that, on the facts, the taxpayer was entitled to capital allowances on a finance leasing transaction related to an oil pipeline. Lord Nicholls gave the leading judgment, and as part of a discussion of the “new approach” heralded by the Ramsay case (WT Ramsay Ltd v Inland Revenue Commissioners [1982] AC 300), said this:

Having referred to a number of cases which had decided that elements which were inserted into a transaction without a business or commercial purpose did not prevent the composite transaction from falling within a charge to tax or bring it within an exemption from tax (at [35]), Lord Nicholls emphasised the need for courts and tribunals to make a realistic assessment of the transaction:

Tower MCashback, a case on which HMRC rely, is an example of a taxpayer not being entitled to capital allowances on the full amount claimed. The issue for the Supreme Court was whether a “loop” of borrowed money could be said in any real sense to give rise to the incurring of expenditure “on” software rights. The Supreme Court held that it could not. The money moved around in a circle merely to enable the investors to indulge in a tax avoidance scheme (at [75]), and the case was to be contrasted on its facts with BMBF where the money had been borrowed and the pipeline had been purchased on commercial terms (see [77]). But Tower MCashback does not establish any new or different principle, rather it confirms the approach in BMBF applied to the different facts of the case to lead to a different outcome. This is clear from Lord Walker’s judgment, at [80]:

A similar point was made by Lord Hope DP at [93]:

It is not in our judgment necessary to refer to other authorities following BMBF, of which there are a great number. BMBF establishes the principle, and that principle is not disputed. Tower MCashback illustrates the application of the BMBF approach to a set of transactions, tested by reference to different legislation, with the result that the transactions were held not to amount to the incurring of expenditure on something qualifying for capital allowances.

The FTT considered the meaning of the words at [147]-[154]. Its conclusion was that the words were to be given a wide construction ([154]). The FTT stated that for the expenditure to qualify, it had to be incurred on or in connection with the conversion of Blush House into hotel premises, and should not be construed so widely as to provide an entitlement to BPRA on all expenditure associated with creating a fully functioning hotel business ([155]). At [209], the FTT accepted Mr Gammie’s submission that the legislation did not refer to “sufficiently connected” but merely required a connection.

The UT considered the meaning of the words “on, or in connection with” at [145]-[148], summarising its views on the meaning of “in connection with” in four points:

The UT went on to consider the extent of the works which were the target of the relief, concluding that:

HMRC attack the conclusions of the FTT and UT. It is HMRC’s case that the words “in connection with” should be construed narrowly. There must be a close connection with the physical works that are undertaken. The words extend to expenditure which has a close connection with the conversion works but which might not be treated as incurred “on” them, such as architects’ or local authority’s fees. The purpose of the legislation is to promote the development of derelict or sub-standard business premises, with a view to those premises coming back into business use, but the relief should not be drawn any wider than is necessary to achieve that end. The UT was in error in its conclusion that the legislation was concerned with the end product, namely a functioning business. It was further in error in suggesting that HMRC’s proposed narrow construction would reduce the effectiveness of the regeneration of disadvantaged areas.

The LLP says that the UT was correct in its interpretation of the statute for the reasons it gave. The words “in connection with” are broad and intentionally so, in order to encourage precisely the kind of investment undertaken in this case. The focus is on creating qualifying business premises, the end product, namely functioning business premises. In this case, therefore, the claim should be allowed in full because it comprises expenditure incurred in connection with converting Blush House into a functioning Ramada Encore hotel.

This appeal concerns Part 3A of the CAA 2001, extracts of which are set out above. BPRA is available if a person incurs qualifying expenditure in respect of a qualifying building. Section 360B(1) defines qualifying expenditure, being expenditure incurred “on, or in connection with” various works, namely the “conversion of a qualifying building into qualifying business premises”, renovations of such buildings and incidental repairs. Section 360D specifies the meaning of “qualifying business premises” as that term is used in s.360B, in terms that require the premises to be “used, or available and suitable for letting for use” for the purposes (so far as relevant in this case) of a trade.

The legislation itself contains a number of clear exclusions or limits on the type of expenditure which can qualify. Expenditure on or in connection with (i) the acquisition of land or rights to land; (ii) extensions to qualifying buildings and the development of adjoining land; and (iii) the provision of loose plant and machinery, do not qualify by virtue of s.360B(3). In addition, expenditure on dwellings does not qualify by virtue of s.360D(1)(c). Other expenditure can qualify in principle, as long as it is incurred “on, or in connection with” the conversion, renovation or repairs in question (and assuming compliance with the other detailed requirements).

The LLP submits that the disputed expenditure was incurred “on” or at least “in connection with” the conversion of Blush House. Although the words “in connection with” have not previously, to our knowledge, been considered in the context of s.360B(1), they have been considered by courts and tribunals in different statutory contexts. A number of these cases were reviewed by the FTT at [147]-[154] and by the UT at [145]-[148].

One such case, cited by the FTT at [147] in a passage quoted by the UT at [146], is of particular assistance: Coventry and Solihull Waste Disposal Co Ltd v Russell (Valuation Officer) (“Coventry Waste”). That case concerned the rateable value of a plant originally constructed to incinerate waste which was substantially altered to create an electricity generating plant, using steam from the boilers of the incinerator plant to drive a turbine. The Electricity Generators (Rateable Values) Order 1989 permitted a rating reduction in certain circumstances, including where the “primary function” of the hereditament was “in connection with a scheme for the production for sale of both electrical power and heat” (paragraph 3(2)(a)(ii) of the 1989 Order, emphasis added). The Lands Tribunal held that the plant was originally constructed to dispose of Coventry’s waste by incineration and although the operations of the plant had been expanded so as to enable electricity to be generated, the disposal of refuse remained the principal reason for its existence. The Lands Tribunal nonetheless allowed the claim for a reduced rate on the basis that the requirements of paragraph 3(2)(a)(ii) were met.

The Court of Appeal (judgment 22 May 1998, [1998] RA 427, The Times 11 June 1998) were split, with the majority allowing the appeal from the Lands Tribunal. Robert Walker LJ was in the minority. He said:

Waller LJ, in the majority, referred to Johnson v Johnson [1952] P 47 at pp. 50-51 and In re Nanaimo Community Hotel Ltd [1944] 4 DLR 638. He said:

Hobhouse LJ, also in the majority, said:

The House of Lords in Coventry Waste ([1999] 1 WLR 2093) dismissed the taxpayer’s appeal. Lord Hope emphasised the words “primary function” in the 1989 Order. Having discussed the legislative history and the implications of each of the rival interpretations, he concluded at p. 2103 B-C (with emphasis added):

Coventry Waste therefore stands for the proposition that the words will usually take their meaning from those which surround it and the wider context, and that courts and tribunals may have to determine whether the words have a broad or a narrow meaning, understood in context. In literal terms, both meanings are possible.

In Barclays Bank plc & Trustees of the Barclays Bank Pension Fund v HMRC [2007] EWCA Civ 442, [2008] STC 476 (“Barclays”), Arden LJ observed that the words “in connection with past service”, which appeared in s.612(1) of the Income and Corporation Taxes Act 1988, could describe a “range of links” (see [18]). This fits with Coventry Waste (to which Arden LJ referred) in suggesting that different meanings are possible. Arden LJ also referred at [19] to the need to examine the function or purpose of the legislation, and at [30] to the purpose of the legislation potentially informing the court’s thinking where there is a choice of meaning.

Herons Court v Heronslea [2019] EWCA Civ 1423, [2019] 1 WLR 5849 is a decision of the Court of Appeal which was not cited to us but provides a useful illustration of the principles discussed in Coventry Waste and Barclays. It concerned s.1(1) of the Defective Premises Act 1972 which provided that “a person taking on work for or in connection with the provision of a dwelling … owes a duty to see that the work which he takes on is done in a workmanlike or, as the case may be, professional manner…”. The issue was whether an approved inspector owed such a duty. Hamblen LJ decided that the approved inspector did not owe such a duty under the statute:

Although it is not a case on the meaning of “in connection with”, some further assistance can be derived from Ben-Odeco Ltd v Powlson (Inspector of Taxes) [1978] 1 WLR 1093, [1978] STC 460. That case concerned the availability of capital allowances under section 41(1) of the Finance Act 1971, which contained the words “expenditure on the provision of machinery or plant”. The dispute related to interest and commitment fees incurred on the financing of an oil rig called the Ocean Tide and whether those payments could be said to be expenditure “on the provision of” the rig. The House of Lords held (Lord Salmon dissenting) that capital allowances were not available.

Lord Wilberforce held that they were too remote to qualify, see p. 1098 E-F:

Lord Hailsham of St Marylebone posed the question at p. 1099 D-E:

These cases show that the meaning of “on, or in connection with” is heavily dependent both on context and policy. The phrase might require what Robert Walker LJ in Coventry Waste referred to as “a strong and close nexus” or it might require “a weak and loose one”. Ben-Odeco v Powlson introduces the concept of remoteness, which is another way of considering the same question.

The issue for this Court in light of the authorities considered above is whether the disputed expenditure was incurred “in connection with” the conversion of Blush House. (While noting the LLP’s formal case that the expenditure was alternatively incurred “on” the conversion, little emphasis was placed on this point and we would not accept that, on any ordinary use of language, the disputed expenditure was incurred “on” the conversion of Blush House.) In determining the meaning of the words “in connection with” as they appear in s.360B(1) there are three points which emerge from their context. First, the words “in connection with” are followed by a short list of the types of work with which the connection must exist for the expenditure to qualify. The items on that list are linked by a common thread. They are all types of physical work on the particular building. The list comprises works of conversion (s.360B(1)(a)), renovation (s.360B(1)(b)) and incidental repairs (s.360B(1)(c)). We infer that the focus of the legislation is on the physical works undertaken. The context in which the words appear in this legislation has some similarity with the context identified in Herons Court. Physical works are at the heart of this relief.

Secondly, s.360B(1) operates when a qualifying building becomes, by a process of conversion or renovation, “qualifying business premises”. The latter term is defined in s.360D(1) in terms that extend to premises which are “used, or available and suitable for letting for use”, for business purposes (s.360D(1)(b)). The LLP makes much of the reference to “used”. But that word is followed by the words “or available and suitable for letting for use”, and the lowest common denominator is the availability of suitable converted premises, not the fact of their use. Here too, the focus is on the physical subject matter of the converted premises in a manner consistent with s.360B(1), and not on their use.

Thirdly, the purpose of the legislation is plainly to encourage the conversion and renovation of existing business premises to facilitate their return to business use. The consultation document issued by the Inland Revenue in December 2004 is not itself an aid to construction of the legislation, but it is of some general interest in understanding its broader context. Reference was made in that consultation document to the potential positive environmental impact of BPRA which would “support the recycling of existing premises that have fallen out of use” and would also lead to “increased investment in disadvantaged areas” (paragraphs C.43 and C.44). In the Regulatory Impact Assessment issued in March 2005, the policy objective was recorded as being: “To increase private investment, enterprise and employment in the UK’s most disadvantaged communities (designated as “Enterprise Areas”), by bringing longer-term vacant business properties, in those areas, back into productive use.”

The purpose of the measure is to be taken from the legislative scheme, read as a whole. The legislation did not require the premises actually to be used. Allowances could be obtained and retained provided the work was done and the premises became available for letting for at least seven years (that is, the premises satisfied the definition of qualifying business premises for that period), even if they were never actually used. The purpose was therefore to encourage the conversion or renovation of disused properties to make them available for business use. Further, the purpose was not to provide tax efficient investment opportunities for high net worth individuals. That may be the consequence of such individuals investing in works which qualify for BPRA – accepting that BPRA aimed to attract investment and the tax relief it offered would be most attractive to those paying higher rates of tax – but the cart must not be put before the horse, and the legislative aim should not be overstated or mischaracterised.

In our judgment, these contextual features point towards the words “in connection with” being construed relatively narrowly. The connection must, by inference, be with the particular works of conversion (or renovation or repair) which lead to the building being, at least, “available and suitable for letting”. It is not necessary that it be used in fact; availability and suitability are sufficient. The scope of expenditure capable of qualifying for allowances cannot differ according to whether it results in premises that are actually in use, as opposed to being available and suitable for use.

We test that conclusion by asking ourselves what the outcome would be if the opposite view were taken and a broad meaning were adopted. Two problems come into view. First, that would open the door to an obvious risk of abuse and avoidance of tax because BPRA could be claimed on expenditure of all sorts, even where the connection with the conversion (or renovation or repair) works was tenuous. Secondly, that could lead to unfairness between taxpayers, because the position could differ fundamentally between cases where the property comes into use for the purposes of the taxpayer’s own trade (or, as in this case, that of a related party) and where it is let or available for letting to a third party tenant; it could also differ between cases where a structured arrangement is put in place as it was in this case and other cases where works are funded more conventionally. The desirability of construing tax legislation in a way that leads to fairness as between taxpayers was emphasised by Lord Wilberforce in Ben-Odeco v Powlson. He referred at p. 1098 B-C to “the principle of the laws of taxation … that, in the absence of clear contrary direction, taxpayers in, objectively, similar situations should receive similar tax treatment”. Lord Hailsham made a similar point at p. 1100 G. We do not consider that Parliament can be taken to have intended this legislation to be construed in a way which leads to the sort of perverse outcomes we have identified.

It follows that the words “in connection with” in s.360B(1) are to be construed relatively narrowly, as requiring a strong and close nexus with the physical works of conversion, renovation or repair that enable the building to become available and suitable for business use. Whether particular expenditure meets that requirement is to be assessed realistically, applying the principle set out in BMBF.

We summarise our approach as follows:

We find the UT and the FTT to have erred in their construction of the legislation as part of the first task summarised above. Both tribunals thought the words “on, or in connection with” carry a broad meaning (most clearly articulated by the FTT at [154] and the UT at [147]). This is not so, in our judgment. The words must be construed in their statutory context and in the light of the purpose of the legislation. The error of both tribunals was to proceed on the basis that the words carry a broad meaning and thereby to assume the answer to the very question posed by Robert Walker LJ in Coventry Waste.

Further, the UT was wrong to conclude that the target of the measure was a functioning building which is open for business (see UT at [151]); it is not. The measure has as its focus the works of conversion, renovation or repair which lead to business premises being either used or available and suitable for letting.

The UT was materially in error in its construction of s.360B(1). In consequence, and to the extent that it concluded that the LLP was entitled to BPRA on disputed items of expenditure and those items remain in dispute on this appeal, the UT’s decision must be set aside.

We now turn to the individual issues in dispute.

Having considered case law to which it was referred, including Tower MCashback and BMBF, the FTT concluded at [165]-[169] that it was necessary to adopt a realistic view of the facts, taking account of the economic realities. The LLP not only knew but (as demonstrated by the Intercreditor Deed) intended how the money transferred to OVL would be used, and for that reason it was necessary to consider the “constituent elements” of the Development Sum.

The UT determined that the FTT had wrongly approached the issue on the basis that it was required to scrutinise the individual payments made by OVL, whereas the correct question was whether the LLP incurred the Development Sum on or in connection with the conversion. In answering that question, all relevant matters and circumstances should be taken into account, including matters such as the presence or absence of circularity of funding and valuation (UT decision at [51]-[56]; [73]-[94]).

The UT went on at [95]-[132] to remake the FTT’s decision and dismiss the appeal on Issue 1. It adopted the reframing of the question by Mr Gammie and Mr Bremner (for the LLP) as being “what did the LLP get for its money?” and answered that question by reference to a construction of the contractual documents, specifically the Development Agreement and the Intercreditor Deed, rather than by reference to issues of valuation and commerciality. It concluded that the LLP did not simply obtain an obligation to carry out the construction works, but also obtained a series of specific obligations pursuant to the Intercreditor Deed, being obligations which supplemented the liabilities owed to the LLP under the Development Agreement. In reaching its conclusion the UT relied on the FTT’s finding that the LLP both knew and intended how the money paid to OVL would be used, and also placed some reliance on the fact that the LLP had excluded £34,999 of legal fees from its claim.

Mr Gammie, for the LLP, repeated the submission advanced to the UT that the question was what the LLP got for its money. For the £15.5m it paid it got (a) the freehold of Blush House and (b) the development obligation, for which it paid a market price. The UT essentially fell into the same error as the FTT in focusing on what OVL would do with the money it received, whereas BMBF showed that that was immaterial. Further, the Intercreditor Deed did not set out obligations owed by OVL to the LLP.

HMRC’s position was that, viewing the facts realistically, the UT’s approach was plainly correct.

The UT was correct to identify the question as being whether the LLP incurred the Development Sum on or in connection with the conversion. In answering that question the facts must be viewed realistically. It is true that the Development Agreement provided for the LLP to pay the Development Sum in exchange for carrying out the conversion work. However, the Development Agreement was only one of a number of documents entered into on 25 March 2011. Their overall effect was to confer a number of important additional benefits on the LLP. The LLP’s case on Issue 1 effectively requires these to be ignored, and the enquiry restricted to the terms of the Development Agreement. That would require a blinkered rather than a realistic approach.

The starting point is clause 16 of the Intercreditor Deed. Although other parts of that deed contained unexceptional provisions regulating the competing claims of the Co-op and the LLP against OVL and providing for those of the LLP to be subordinated, clause 16 was a bespoke provision governing the operation of various accounts set up by OVL with the Co-op. Clause 16.5 made specific provision as to the destinations of the funds raised by the LLP (then sitting in a “Subscribers Account” in the LLP’s name), as follows:

The list of fees and expenses in Schedule 1 relevantly included “Sponsors and IFA fees” and “Franchise fees”.

The effect was that there was never any straightforward payment of the Development Sum to OVL. Rather, sums were transferred from the LLP’s account as specifically directed by clause 16.5.

Clause 16.5 did not simply benefit the Co-op but also the LLP. The LLP had an obvious and material interest in the destination of the funds. As discussed further below, the Capital Account provided a significant element of support for the Co-op Loan, including as a source of funds to enable the LLP to avoid a default. Withdrawals from the Capital Account were regulated by the Capital Account Deed, to which the LLP was a party. The Interest Account secured funds required to meet the LLP’s obligations to pay interest on the Co-op Loan during the development phase. Withdrawals from that account were regulated by clause 16.9 of the Intercreditor Deed and by the Licence Deposit Deed between OVL and the LLP, under which OVL also granted a second charge over the Interest Account in favour of the LLP. Withdrawals from the Construction Account, the Construction Cost Overruns Account and the FF&E Account were regulated by clause 16.6-16.8 of the Intercreditor Deed, again for the benefit of OVL as well as the Co-op. In addition, the LLP took second charges over those accounts to secure OVL’s liabilities to it more generally. Among the benefits to the LLP was a specific provision in clause 16.6 of the Intercreditor Deed entitling the LLP to access the funds in the Construction Account to pay VAT due on its purchase of the Property, pending recovery of the VAT from HMRC.

In summary, this was not a situation where the use of the funds by the payee was of “no concern” to the taxpayer (BMBF at [42]). Rather, the LLP not only knew and intended how the funds would be used by OVL (as the FTT found), but its material interest in their use was reflected in the significant contractual protections that it obtained.

Further, a realistic appraisal of the facts requires regard to be had to the entire factual matrix, extending beyond the confines of the contractual documents. This may include questions of valuation as well as identifying the “real purpose” or object of the expenditure (Tower MCashback at [67] and [80]). As to the latter, it is clear that elements of the Development Sum were intended to be “earmarked” for specific purposes.

As to the former, the LLP’s case is not assisted by its assertion that it paid a “market price” for the Property and its conversion. It is true that individuals were prepared to invest the amounts that they did in the LLP on the basis of arrangements that would result both in the acquisition of the Property and its conversion into a hotel, but it does not follow that the LLP incurred all its expenditure on (or in connection with) (a) the Property and (b) the conversion. The £15.5m that investors in the LLP were prepared to pay (see [26] above) reflected a premium of around 15% over the highest valuation produced by Edward Symmons. That value was £13.4m and was itself a valuation of the business in a “stabilised” third year of trading. In other words, it was based on established trading. The valuation of the converted hotel, ready to trade but before trade commenced, was £11.9m. The tax advantaged nature of the product was therefore clearly a material feature for investors. To the extent that valuation is relevant, it is also worth bearing in mind that Multibuild agreed to complete the physical conversion works for a little under £6m.

As discussed further below in the context of the Residual Amount, the FTT found that the £15.5m was paid for a “package” put together by OVL. This went beyond the development obligation and included provision of the Property and other elements which did not qualify for BPRA. Even if the UT’s construction of the effect of the contractual documents had been incorrect that would not preclude regard being had to the substance of what OVL provided to the LLP.

The LLP also relied on Ben-Odeco v Powlson, where it was recognised by Lord Hailsham and Lord Russell that the position would have been different if the supplier had borrowed the necessary working capital and reflected the cost of doing that in its price for the rig. However, in that event the taxpayer would have incurred all of its expenditure on the provision of the rig, whereas in this case the LLP got significantly more for its money than the obligation under the Development Agreement to do the conversion work.

We would therefore dismiss the LLP’s appeal on Issue 1.

As already indicated, the purpose of the Capital Account Deed was to regulate withdrawals from the Capital Account. The key provision governing withdrawals is clause 3.5, headed “Payments from the Capital Account”. Clauses 3.5.1 and 3.5.2 permitted OVL to withdraw sums after the third anniversary of the date of the deed (25 March 2011) to the extent that the principal amount of the Co-op Loan had been repaid (subject, broadly, to the Co-op being satisfied with the LLP’s financial position). It was not disputed that any sums so withdrawn would be for OVL’s benefit. Clauses 3.5.3 and 3.5.4 provide as follows (the “Bank” and “Borrower” being the Co-op and the LLP respectively):

The Developer Loan Agreement is a relatively informal document in the form of a countersigned letter from OVL to the LLP. The opening paragraph reads:

OVL provided the Co-op with a guarantee of the Co-op Loan. The guarantee was secured by a charge over the Capital Account, by charges over other accounts held by OVL with the Co-op (namely the Construction Account, the Interest Account, the Construction Cost Overruns Account and the FF&E Account) and by various assignments including an assignment of OVL’s rights in the Design and Build Contract. The guarantee was given on a limited recourse basis, limited to the realisable value of the secured assets. The charge over the Capital Account provided that effect would be given to the security by a release of the deposit to the bank. However, although the Capital Account Deed cross-refers to the guarantee there is no explicit provision in the Capital Account Deed for withdrawals to be made pursuant to the guarantee or by way of enforcement of the security. Rather, clause 3.6.2 provides that “No withdrawals from the Capital Account may be made except as permitted by this Deed”. Unless that is read as being subject in some way to the guarantee and charging arrangement, the Co-op would therefore need to seek to rely on clause 3.5.3 if it wished to call under the guarantee.

The FTT discussed the Capital Amount at [182]-[196]. It concluded that “in reality the nature of the Capital Account was circular and self-cancelling cash flow” starting and ending with the Co-op, and was not incurred on or in connection with the conversion or renovation of the Property. In reaching that conclusion:

The UT held that the FTT’s reasoning was flawed, and allowed the LLP’s appeal on this issue ([163]-[225]). There had been too much focus on clause 3.5.3 rather than other scenarios, but in any event the UT preferred the LLP’s construction that sums withdrawn under clause 3.5.3, as well as sums withdrawn under clause 3.5.4, would increase the Developer Loan. In reaching that conclusion it took into account that the Development Sum constituted “a market price within a reasonable range” for the converted Ramada hotel ([208]) and the terms of the Developer Loan Agreement and related charge, the Information Memorandum and the Debt Finance Request. It rejected HMRC’s suggested commercial rationale for distinguishing between clause 3.5.3 and clause 3.5.4 and found that a withdrawal by the Co-op pursuant to clause 3.5.3 effectively amounted to a call on OVL’s guarantee of the LLP’s obligations to the Co-op.

The UT also concluded that, even if that was wrong, the funds in the Capital Account remained OVL’s funds unless and until withdrawn by the Co-op, saying this at [221]:

Mr Brinsmead-Stockham, who made submissions for HMRC on this issue, based his primary case on what HMRC consider to be the UT’s incorrect conclusions about the effect of sums being withdrawn under clause 3.5.3 and (relatedly) an erroneous assumption that the funds in the Capital Account were OVL’s money. While the funds remained in the Capital Account OVL could not derive any benefit from them and they were at risk of being withdrawn pursuant to clause 3.5.3, in which event (contrary to what the UT had concluded) OVL would never benefit from them. He also submitted that there was no commercial reason for the arrangement and the FTT was entitled to conclude that it was circular and self-cancelling. On the basis of the approach set out in Tower MCashback allowances were not available.

Mr Gammie submitted that the UT reached the correct conclusion. As the Developer Loan Agreement made clear, any withdrawal under clause 3.5.3 would be added to that loan. This was what happened in fact and had also been anticipated in the Information Memorandum and the Debt Finance Request. OVL had simply applied part of its profit as security for the project. It made no commercial sense for it to be deprived of that profit with no recourse to the LLP simply because there was a withdrawal under clause 3.5.3, and even if it could be that was just one of several possible outcomes.

We agree with the UT that the FTT’s reasoning was flawed, but disagree with the UT that the FTT reached the wrong conclusion.

Much of the argument before us was about whether, as the UT held, a withdrawal under clause 3.5.3 gave rise to an increase in the Developer Loan, whether it should be treated as amounting to a call on OVL’s guarantee and whether it should be regarded as giving rise to some form of restitutionary claim against the LLP. The issues are not straightforward, but in our view the appeal on this issue does not turn on them and it is unnecessary to decide them.

There are two key points to bear in mind. First, as already discussed, the facts must be viewed realistically. Secondly, there is no dispute that the legislation must be applied as at the date on which the expenditure is incurred (“day 1”), rather than by reference to what may or may not happen later. In particular, the fact that OVL might benefit from the Capital Amount in the future, whether in the form of withdrawals under clauses 3.5.1 and 3.5.2, an increase in the Developer Loan under clause 3.5.4 or (possibly) some form of right of recovery further to a withdrawal under clause 3.5.3, does not of itself answer the question whether the expenditure that it is accepted was incurred by the LLP on day 1 was incurred on or in connection with the conversion.

On a realistic appraisal of the facts the Capital Amount was not incurred on or in connection with the conversion of the Property. The contractual documents required it to be used, and the parties knew and intended that it would be used, to support the LLP’s borrowing from the Co-op. It is not necessary to state positively on what the expenditure was incurred, but if it were we would characterise the expenditure as being on a “support package” in respect of the Co-op Loan, enabling that borrowing to be maximised.

The support was provided a number of ways. The Capital Account mechanism not only provided security to the Co-op but, under the terms of the Co-op Loan Agreement, the balance in the Capital Account was netted against the loan for the purposes of the LLP’s covenant as to the maximum loan to property value ratio. The Co-op also required the LLP to enter into interest rate hedging arrangements for £5m rather than the full £7m. Further and importantly, the Capital Account mechanism mitigated the risk of the Co-op enforcing its charge over the Property, an enforcement which would have led to a clawback of capital allowances as well as a loss of the investment. Clause 3.5.4 allowed the LLP to avoid a payment default by accessing funds in the Capital Account whenever it considered that to be necessary to meet its obligations to the Co-op, subject only to a limited veto by the bank. Essentially, as far as OVL was concerned the LLP could help itself to the funds when it required them. This feature, together with the netting for covenant purposes, were regarded as sufficiently important to be highlighted in the Information Memorandum, clearly with a view to providing comfort to potential investors as to the robustness of the structure.

In contrast, and although the account was in its name, OVL had no access to the Capital Account for at least three years, and then only to the extent that the Co-op Loan was repaid. If funds were withdrawn under clause 3.5.3 or 3.5.4 the most that OVL would have would be an increase in the Developer Loan (or possibly some other right in restitution), but any such right would obviously rank behind the Co-op. Indeed, under the terms of the Developer Loan any such increase was explicitly interest free.

It makes no difference that interest accrued on the balance in the account. It was clear that any such interest would be added to the account and subject to the same arrangements as the principal amount. In any event the treatment of any interest that might accrue does not provide material assistance in determining whether the statutory test is met.

While we accept that the correct tax treatment must depend on the arrangements actually entered into rather than alternative arrangements that might have been entered into, it is relevant in undertaking a realistic appraisal to recognise that the substantive effect of the mechanism was really no different to the LLP retaining the Capital Amount, using it to support its borrowing by placing it on deposit in an account with the Co-op and agreeing that in certain circumstances it would pay deferred consideration to OVL. The LLP’s case amounts to saying that it makes all the difference that the deposit was in the name of OVL rather than the LLP, and that that is so despite the fact that, at least for the first three years, it was the LLP (or its lender) and not OVL that could benefit from the funds.

Characterising the arrangement as expenditure on a support package for the Co-op Loan might be said to raise the question whether it is nonetheless sufficiently closely related to the conversion to be “in connection with” it. We do not accept that it is. The reality is that the £2m was not needed for the conversion work and cannot properly be viewed as sufficiently related to it to satisfy the “in connection with” test.

The starting point for the funds raised was the figure of £15.5m that it was (correctly) determined that investors in the LLP would be prepared to pay for the tax-advantaged “package” of the converted hotel. As already mentioned, this reflected a premium of around 15% over the highest valuation produced by Edward Symmons (see [94] above). In order to raise the full £15.5m, borrowings needed to be maximised, irrespective of whether the funds were actually required for business purposes. Doing so allowed the maximum potential profit to be secured for those who put the structure together. The Co-op would undoubtedly have been content simply to lend £5m. The conversion could still have been done but the potential profit (and the claim for allowances) would have been lower. In reality the additional £2m raised by the LLP was attributable to the ability of those putting the structure together to secure a greater level of funding than would otherwise have been achievable, due to the perceived tax advantages of the structure. There was no need for that £2m to be raised from investors to do the conversion work. Furthermore, and crucially, the £2m was not simply taken as profit. It had to be expended to support the additional funding that was raised as a result of the perceived tax advantages. As such it cannot properly be regarded as incurred on or in connection with the conversion.

The FTT accepted HMRC’s characterisation of the arrangement as circular and self-cancelling. It is understandable that it did, but some care is needed in applying those labels. There is no doubt that there was a real borrowing of £7m from the Co-op, on which the LLP bore interest in full. There was a real transfer of funds from the LLP to an account in OVL’s name, and it was OVL that charged the funds in the Co-op’s favour. As things turned out, OVL ultimately benefitted from both the Capital Amount and the interest accrued on the Capital Account. However, those features do not determine whether the expenditure was incurred on or in connection with the conversion. Rather than describe the arrangement as circular or self-cancelling, we would observe that it would not have been possible to determine on day 1 whether OVL rather than the LLP (or its lender) would actually end up benefitting from the deposit. If OVL did then it might at that stage be characterised as additional profit derived from its commitment to the LLP to undertake the conversion, but that does not determine that the LLP’s expenditure on day 1 was on or in connection with the conversion.

By paragraph 1.1 of Schedule 1 to the Development Agreement, the LLP granted a licence to OVL (and specified other parties, including Multibuild) to enter upon and to occupy the Property in order to carry out the development works. By paragraph 2 of Schedule 2 to the Development Agreement the LLP was obliged to pay the Interest Amount “into a deposit account to be drawn down in accordance with the terms of a Licence Deposit Deed entered into between [OVL] and the [LLP]”. Paragraph 1.2 of Schedule 1 then required OVL to use the Interest Amount to pay a quarterly licence fee equal to each quarter’s interest charged to the LLP on the Co-op Loan. That obligation endured until the hotel opened, with any shortfall being the liability of OVL, to be paid by it either directly or from the Cost Overruns Account (as to which see [88] and [91] above).

By clause 16.5.2(d) of the Intercreditor Deed, the Interest Amount was required to be transferred into the Interest Account. By clause 16.9 of the Intercreditor Deed, OVL could not withdraw any amount from the Interest Account save to allow the LLP to comply with its payment obligations to the Co-op. The Interest Account was also charged to the Co-op as part of the security for OVL’s limited recourse guarantee (see [100] above).

The Licence Deposit Deed defined OVL as the “Chargor” and the LLP as the “Chargee”. Under that deed the LLP was granted a second charge over the Interest Account (defined in that Deed as the Deposit Account), after the Co-op’s first charge. The “Licence Fee” was defined as the sum of £350,000 “to cover interest payable on the funding advanced by the [Co-op] and payable in accordance with the terms of the Development Agreement”. By clause 4, interest accruing on the account was to be “released” to OVL (although we note that clause 16.9 of the Intercreditor Deed appears to have precluded this). Clause 5 provided for withdrawals and was in terms that matched the Development Agreement, permitting withdrawals to pay the quarterly licence fee in sums equal to the interest charged to the LLP on the Co-op Loan up to the date when the hotel opened for business.

Clause 6 of the Licence Deposit Deed provided that:

The LLP relied on evidence from one of its experts, Mr Douglas Smith, a chartered surveyor, that it is quite usual in tax-driven property investment to have a licence fee which provides a commercial return after a sale has been paid for up front (a “forward sale”). In his report, Mr Smith had explained that by tax-driven he meant projects where the investor would look to the availability of capital allowances as part of the pricing or return for the investor.

Mr Lewis gave evidence that a similar arrangement using a licence fee to fund interest payments had been used in Downing’s Enterprise Zone funds from 1997 to 2005 and it seemed logical that the arrangement should be continued for BPRA funds. He suggested that this was a commercial arrangement between unconnected parties, designed to meet the interest on the Co-op Loan for the period until the conversion was complete and the hotel was open for business. Similar points were made by Mr Tracey.

HMRC pointed to oral evidence given by the LLP’s own valuation expert (Mr David Harper) that he had not seen a licence fee arrangement such as this before. HMRC’s own expert chartered surveyor, Mr Anthony Williams, considered it normal practice to grant the developer a licence over the land to be developed for a nominal fee, but his view was that the licence fee arrangements in this case, designed to meet the LLP’s interest costs, were most unusual.

The FTT agreed with the LLP’s submission that the arrangement under which the LLP granted a licence to OVL and received periodic payments equal to the interest payable to the Co-op during the development phase was “a commercial arrangement, the effect of which was not to increase the tax deduction available to the LLP as the income it received from the licence fee was taxable” (FTT decision at [179]). It also rejected at [177] HMRC’s submission that, if the Interest Amount was properly characterised as a payment in return for a licence, it was incurred on or in connection with an interest in land, being the licence. This was because it was OVL and not the LLP which acquired the licence. The FTT therefore allowed the Interest Amount.

The UT allowed HMRC’s appeal on this aspect (UT decision at [226]-[271]). The UT noted that the Intercreditor Deed only permitted sums to be withdrawn to make payments to the Co-op. Whilst it was the case that under the Development Agreement OVL bore the risk that the Interest Amount was insufficient to bear all the interest costs incurred before the hotel became operational, the UT rejected the LLP’s argument that the effect of the Licence Deposit Deed was that any excess over the amount required to meet interest costs was required to be passed back to OVL (as occurred in fact: see [38] above), rather than paid to the LLP as its express terms provided.

The UT concluded that while the arrangements overall were commercial and “broadly at market price” it could not discern a credible commercial reason for a sum to be paid only to be returned. The FTT had not referred to the expert evidence produced by HMRC. It was clear that the level of the fee was not determined by reference to the value of the licence but rather a completely unrelated factor, being the interest payable to the Co-op. The risk of overruns could easily have been provided for by another mechanism. The UT concluded that the Interest Amount was a circular and self-cancelling arrangement that had no real discernible commercial purpose, and the FTT’s finding that there was a commercial arrangement was an Edwards v Bairstow error of law.

The UT also noted that if the interest had been paid by the LLP it would have been an income expense which could have been deducted against the LLP’s future profits. However, it was in investors’ interests for the BPRA claim to be as high as possible to maximise sideways loss relief in the first year of the scheme.

If the UT’s primary conclusion on this issue was wrong, it concluded in the alternative that the FTT was also wrong to decide that the land-related exception in s.360B(3)(a) was not engaged, on the basis that the Interest Amount was paid at least in part to enable OVL to fund the acquisition of a licence over the land.

The LLP challenged the UT’s decision on the Interest Amount. It denied the arrangements were circular, but even if they were, that was not a reason to disallow BPRA, citing BMBF at [42] and Tower MCashback at [77]. Further, the arrangements could not be described as self-cancelling in circumstances where it was accepted that the Development Sum had been incurred by the LLP. It was in any event not open to the UT to reverse the FTT which had found that the arrangements were commercial, which they plainly were; this was a means of laying off the risk of overrun onto OVL, who would remain liable to pay the licence fee quarterly until the hotel opened. It was irrelevant that the effect of the arrangements was to convert income expenditure into a capital expense; in any event, the tax consequences remained neutral because either way the LLP would be entitled to deduct the cost of servicing the Co-op Loan as a business expense. It was irrelevant that the same outcome could have been achieved by the LLP simply holding onto the money because that is not what the LLP in fact did. The sole question was whether the interest amount was incurred “on, or in connection with” the conversion; the FTT had been correct to find that the Interest Amount did come within those words. The UT’s alternative conclusion was wrong because this expenditure was not attributable to the purchase of rights in land, because those rights were purchased by a third party and not by the LLP so that the exclusion in s 360B(3)(a) did not in any event apply.

HMRC sought to uphold the UT’s decision on the Interest Amount. The arrangements were plainly circular because the LLP might just as well have retained the money in its own bank account and paid the interest itself. It was open to the UT to decide that the arrangements were also self-cancelling, which was a different question from whether expenditure had been incurred. These arrangements gave a tax advantage to the LLP’s investors who achieved sideways relief on the whole cost of servicing the Co-op Loan in the first year of the scheme, by capitalising the interest cost into the Interest Amount. There was no sufficient connection between the Interest Amount and the conversion works. The real connection was with the tax mitigation aspect of the scheme, which was designed to claim BPRA on as much expenditure as possible in order to maximise the benefit to the investors. Alternatively, any connection that did exist was with the purchase of rights in land and excluded from BPRA.

It is common ground that funds in the Interest Account were used to service the interest on the Co-op Loan. The Interest Amount was provided by the LLP to OVL on terms that OVL put it in the Interest Account, which account was charged to the bank first and the LLP second. Under clause 6 of the Licence Deposit Deed any residue reverted to the LLP (and, like the UT, looking at its plain terms, we reject the suggestion that OVL had a contractual right to any residue; the second reference to “Chargee” is not a mistake). In no real sense did the money become available to OVL. Although OVL was at risk of having to continue making payments even after the Interest Amount was exhausted, that was a future risk which would only materialise if the works in fact overran; before then, OVL carried no risk because it was pre-funded to make the quarterly payments. Further, there was no need for the Interest Account structure to allocate that potential risk to OVL. The obligation to meet any excess of interest costs over the £350,000 that comprised the Interest Amount was simply a contractual obligation on the part of OVL to pay it either from the Cost Overruns Account or from OVL’s other resources.

We agree with HMRC that the evidence was clear that a licence fee of this magnitude, put to this purpose, was not a regular feature of property development projects. It conferred a tax advantage on the individual investors by converting the interest cost into capital expenditure on which BPRAs could be claimed in the first year of the scheme.

That feature alone would not prevent the arrangements from achieving their aim. However, the Interest Account mechanism was devoid of any real commercial purpose. Although (as with the Capital Amount) there was a real transfer of funds into an account in OVL’s name and a charging of those funds by OVL, the arrangement was entirely circular and in substance self-cancelling. The fact that the LLP could have achieved the same economic result by holding onto the money itself and depositing it with the Co-op as security is relevant in undertaking a realistic appraisal of the facts because it highlights the circularity and self-cancelling nature of the arrangements and the lack of a strong and close nexus with the conversion works. These points serve to underline the lack of any real benefit to OVL from the Interest Amount.

It follows that we agree with the UT’s finding that the arrangements lacked commerciality. Further, it was, in our judgment, open to the UT to reverse the FTT on this point for the reasons given by the UT: the FTT had made an “illogical and unjustifiable leap” from its conclusion that the parties were not connected in a statutory sense to finding that the arrangements were commercial (UT decision at [250]); the FTT had failed to explain why such a large amount was charged for the licence, and in that regard had failed to refer to the evidence that this was a very unusual arrangement (UT decision at [253]); and the FTT had failed to acknowledge the artificiality in the LLP meeting its interest obligations by means of money which went from the LLP to OVL and back again (UT decision at [254]). Overall, the UT characterised the FTT’s failures as containing insufficient analysis and ignoring undisputed facts; the result was an Edwards v Bairstow error of law by the FTT (UT decision at [257], and see [1956] AC 14 per Lord Radcliffe at p. 36).

The central issue is whether the Interest Amount was expenditure incurred in connection with the conversion of Blush House. The UT held it was not, applying the term “in connection with” in the broader sense in which the UT had construed it. We have found the legislation to be rather narrower in its ambit, and it follows that on our construction of the legislation the conclusion is even more compelling that the Interest Amount was not incurred “in connection with” the conversion works.

It is not necessary to deal with HMRC’s alternative argument that the Interest Amount was connected with the acquisition of rights in land.

We would therefore dismiss the LLP’s appeal on Issue 2(b).

The FTT found that both the IFA fees and the Promoter fees qualified for BPRA (FTT decision at [197]-[203]). It commented that if the IFA fees had been paid direct by the LLP then that would either be deductible revenue expenditure or expenditure incurred “in connection” with the conversion on the wider interpretation of that phrase adopted by the FTT. It rejected HMRC’s argument that any part of the fees was incurred in connection with the acquisition of land. The FTT’s reasons for allowing the Promoter fees were similar.

The UT agreed with the FTT’s conclusions ([272]-[290]). Taking account of the purpose of the legislation, the phrase “on, or in connection with” extended to costs of raising equity finance. Neither HMRC’s more restricted interpretation nor its renewed attempt to overcome the lack of any finding by the FTT that the BPRA claim was “ramped up” justified interfering with the FTT’s decision on this point. The UT also rejected HMRC’s alternative argument that the fees should be apportioned so as to exclude an amount attributable to the finance raised to purchase the land, on the basis that it was not clear how any apportionment could be carried out on a principled basis. The FTT was entitled to reach the conclusion it had on the facts, namely that the fees were incurred for the general purpose of raising equity for the project rather than partly in connection with the acquisition of land.

HMRC argue that the IFA fees and Promoter fees are not expenditure incurred on or in connection with the conversion of Blush House. The degree of connection is too remote to qualify; rather, these costs were incurred in connection with setting up the LLP as the funding structure for the development and should be characterised as fees for marketing that structure. The LLP’s own evidence from Mr Smith confirmed that IFA and promoter fees were not normally associated with non-tax driven activity and were not incorporated as a cost by the Argus software package which was used to provide a market valuation. The UT was in error in adopting a broad construction of the statute and in equating these fees, which related to the element of equity funding for the development, with bank arrangement fees more typically associated with such property development projects. Further, the UT was in error in suggesting that HMRC’s analysis would restrict the effectiveness of the relief, given the evidence from Ms Katherine Nash of HMRC to the effect that the vast majority of BPRA claims were small in size and came from individuals, raising no issue about IFA or promoter fees. Alternatively, if there was a sufficient connection between these fees and the conversion works, they must be apportioned to reflect the fact that a mixed fund comprising equity and debt was partially used to purchase property rights. If no such apportionment is possible, then these fees should be disallowed in their entirety as falling within the exception in s.360B(3)(a).

The LLP maintains that the UT (and before it, the FTT) were right to allow these fees. They are connected with the conversion of Blush House. They should be treated in the same way as a bank arrangement fee which is a standard feature of this sort of project and permitted to benefit from BPRA as part of the cost of funding. They were not marketing fees or anything like marketing fees.

We have already determined that the UT’s construction of the statute was too broad. That was an error of law which infects the UT’s conclusion about these fees. We therefore approach the question of whether these fees were incurred in connection with the conversion afresh.

The obligation to pay the Promoter fees derived from the agreement with Mr Bantoft referred to at [26] above. OVL’s obligation to pay the IFA fees derived from the Costs Agreement between Downing and OVL, clause 3 of which required OVL to pay Downing’s transaction costs, including the costs of IFAs who advised the individual investors. The Intercreditor Deed contemplated that part of the Development Sum would be used by OVL to meet the fees: see [89] above.

So far as the Promoter fees are concerned, Downing was responsible for finding investors for the scheme. To that end, it produced an Information Memorandum which advised prospective investors in its Executive Summary that BPRA should be available which would “provid[e] higher rate taxpayers with significant relief on the cost of their investment”. A further passage in the summary stated that approximately 81% of the total purchase price to be paid for the project would qualify for BPRA, and that after taking account of the tax relief, the net equity cost was approximately £6,087 for each £100,000 investment. Thus, the key selling point for investment in the scheme was undoubtedly the tax benefits of doing so.

As to the IFA fees, we were taken to a number of examples of communications by IFAs to their clients, explaining the tax advantages to their clients of participation in the scheme. We also saw a document called “Tax Shelter Guide” prepared by one IFA, Ward Consultancy plc. That document considered a number of different forms of tax shelter, and contained a section describing how BPRA LLPs worked and what the tax advantages of investing in them might be. These documents demonstrated the role that IFAs played in advising their clients on the suitability of investment in the LLP for that client, and the likely benefits (and risks) of doing so.

Both types of cost were associated with encouraging individuals to invest in the LLP. They were akin to fees for marketing the scheme. The LLP was obviously the precursor to the conversion works and was the vehicle through which funds were raised, but the legislation does not pose a test of causation; rather, the words “on, or in connection with” require a strong and close nexus to be demonstrated. Costs associated with marketing the LLP to investors were at one remove from the works themselves. What is more, a significant element of the work associated with this expenditure was obviously incurred in order to investigate, explain and illustrate the tax advantages of participation in the scheme. The focus was on the position of prospective investors in the LLP. In reality the fees represented costs of OVL which it had agreed to pay in order to enable the scheme that it had put together to proceed.

We are not persuaded by the LLP’s argument that the IFA fees and the Promoter fees were analogous with a bank fee for arranging debt finance. No issue has arisen before us as to the eligibility of bank fees so the argument by analogy does not rest on solid foundations. But there is an obvious difference of fact, in any event. Any such loan finance would have been provided directly to the LLP to enable it to undertake the conversion works, and the fee would have been charged to the LLP by the bank for that service. The IFA fees and Promoter fees were paid for services of another kind that were provided to third parties, being individual investors and OVL. They are therefore very different in character from bank fees incurred by the taxpayer.

We conclude that, on its proper construction, the statute does not extend to expenditure of these types. The policy behind the legislation was to stimulate the redevelopment of derelict business property by offering generous tax incentives in the form of 100% allowances on capital expenditure. But it stretches the language of the statute too far to extend relief to costs of third parties which are associated with identifying prospective investors in the LLP and persuading them to invest. Those actions were a clear step away from the conversion works. They were focused on marketing the tax shelter structure through which the conversion works would be undertaken but were not focused on the conversion works themselves. On a realistic appraisal of the facts these fees do not have a strong and close nexus with the physical works. They are too remote to qualify.

It is not necessary for us to address HMRC’s alternative arguments for apportionment or exclusion which are predicated on the existence of a sufficient connection with the conversion works.

We therefore conclude that HMRC’s appeal on Issue 2(c) should be allowed.

The FTT distinguished at [206]-[209] between the payments to Ramada and Sanguine. The payments to Ramada (of £15,000 and $15,000) were made “to ensure the Property complied with the requirements and branding to enable its operation as a Ramada Encore hotel” and qualified for BPRA. In contrast, the payment to Sanguine (the “Sanguine payment”) was made to remove it from any involvement in the project and could not qualify.

Reversing the FTT’s decision, the UT concluded at [317] that the Sanguine payment was “in connection with” the conversion, albeit “at the outer limits of that phrase”. Establishing a hotel business would involve design and development stages which required management, it was not uncommon for longer term projects to involve parties being replaced and the payment was in substance in recognition of the replacement of Sanguine.

As regards Ramada, HMRC pursued its appeal only in respect of the $15,000 paid for a licence to use the Ramada brand (the “Ramada payment”). The UT dismissed that appeal, pointing at [324] to the policy behind the relief as being to encourage unused buildings to be converted into buildings that were open for business, such that the legislation was concerned with the “end product”. A hotel operated under a particular franchise would not be open for business unless it was entitled to use the franchise.

So far as the Sanguine payment is concerned, HMRC point to errors in the UT’s construction of the legislation, and assert that the UT was not at liberty simply to substitute its own view for that of the FTT without first identifying some error which infected the FTT’s approach, which the UT failed to do. The FTT had been right to conclude that the Sanguine payment was too remote.

The LLP says that the UT was right for the reasons it gave. A payment to remove a contractual counterparty has the same character as work done towards the conversion of the building. The Sanguine payment was connected to the conversion.

So far as the Ramada payment is concerned, HMRC renew their argument that this was expenditure incurred by the LLP for the purpose of allowing a different entity, the Operating Company, to operate a business under a particular franchise following conversion of the property. It was not expenditure incurred on or in connection with the conversion, applying that phrase, properly construed. As evidenced by the License Agreement entered into between Ramada and the Operating Company the Ramada payment was the initial fee for the licence to use the Ramada brand, and there was provision for recurring fees thereafter. This payment was for intellectual property, the Ramada Encore branding, and HMRC say it is too remote from the conversion to qualify.

The LLP argues that the Ramada payment was for the right to use the brand name and that it is eligible for BPRA because it was necessary in order for the hotel to function as a hotel under the Ramada brand. It is artificial to divorce this payment from the conversion works, when the reality was that Blush House was converted into a Ramada Encore branded hotel and the branding, as well as layout and other aspects of the conversion work, were inextricably linked with that end point. That was the deal actually done, and was the focus of the legislation. The FTT and the UT were correct in their analysis.

We deal first with the Sanguine payment. The FTT said that this payment was made to remove Sanguine from involvement in the project. The UT elaborated on the reasons for the payment at [317] but in essence did not disagree with the FTT’s description. Further, this was not merely the removal of a firm working on the conversion project but the removal of a potential manager and operator of the hotel, to be substituted by ThenHotels which was appointed as the manager and operator of the hotel (see [28] above). The issue is whether a payment of that character was incurred “on, or in connection with” the conversion. The FTT and the UT both thought that those words should be construed widely, a conclusion with which we respectfully disagree (although even on the wide construction, the FTT thought the Sanguine payment did not qualify). The UT thought that the target of the relief extended to a functioning building, open for business and in use for a trade, a conclusion with which we respectfully disagree. Once a narrower meaning is given to the words in the statute, recognising that the target of the measure is the conversion works and not the business which may lie beyond, it is clear that the Sanguine payment is ineligible for BPRA. It lacks a strong and close nexus with the conversion works. Its connection is with the Developer’s wider business interests, or possibly with the specific business of running the hotel from Blush House. The LLP got no real benefit from this payment. Accordingly, the FTT was right (albeit for reasons with which we do not entirely agree) to conclude that the Sanguine payment did not qualify for BPRA.

We turn to the Ramada payment. We accept HMRC’s submissions and conclude that the FTT and the UT were wrong to treat that payment as eligible for BPRA. The UT’s reasons for reaching that conclusion are set out at [324]. Again, we respectfully disagree with the UT’s expansive reading of “in connection with” and we do not accept that it was relevant that the hotel could not open for business if it was not entitled to use the relevant franchise. The legislation does not pose a test of causation; rather it is concerned with establishing a strong and close connection with the physical works, not with the end product in the sense of a hotel or other premises actually “open for business”. The Ramada payment was paid, in essence, for intellectual property rights to be conferred on the Operating Company (not on OVL or the LLP) to enable that company to operate its business as a Ramada Encore hotel. The payment had nothing to do with the conversion works, at least not in any direct way or by means of any strong or close nexus. It was too remote, by some margin, to qualify for BPRA.

We would therefore allow HMRC’s appeal on both aspects of Issue 2(d).

HMRC’s position before the FTT was that the “profit” represented by the Residual Amount should be apportioned between allowable and non-allowable items. This should include the land because OVL put the whole “package” together for the LLP on a “cradle to grave” basis (including securing the freehold), and its profit was attributable to all elements of the package and was calculated by reference to a valuation predicated on freehold ownership of the hotel. The FTT agreed with HMRC for the reasons given by them (see [228]-[231]).

The UT held this approach was wrong ([350]-[358]). The Residual Amount did not represent OVL’s actual profit and as a result HMRC’s submissions were “built on sand”. More importantly, HMRC’s approach was misconceived. It was based on the fallacy that what mattered was what OVL did with the money received from the LLP. The Residual Amount did not reflect a right or asset acquired by the LLP at all.

HMRC submit that the FTT was right. The Residual Amount was not the mere “mathematical difference” that the UT characterised it as. It was OVL’s “fee” and a fundamental element of the deal between the LLP and OVL. It was undisputed that OVL’s return related to the entire package and that the total expenditure on its provision was £15.5m. For example, if OVL had explicitly charged a finder’s fee for the Property that would clearly be excluded. It should make no difference that it was “bundled” with other elements.

The LLP submit that HMRC’s apportionment exercise is wrong in principle. The object of scrutiny of the legislation is the LLP’s expenditure. The Residual Amount is simply an arithmetical calculation which does not reflect any obligation owed to the LLP. In any event, if an apportionment is required the approach suggested by HMRC is inappropriate.

In our view the UT was wrong to reverse the FTT’s decision on this issue.

It is true that the Residual Amount does not represent OVL’s actual profit from the project. That would depend on its overall costs, in particular its internal costs and overheads. It would also depend on whether and to what extent OVL ultimately received a benefit from the Capital Account mechanism. However, HMRC’s case is not that OVL’s profit must be apportioned. It is that the Residual Amount represented OVL’s fee for the “cradle to grave” package.

As the FTT found, that package included securing the freehold. It also covered every other aspect of the arrangements that ultimately resulted in a converted hotel that was ready for letting to the Operating Company, which arrangements were structured with a view to securing the maximum available capital allowances. The LLP incurred expenditure totalling £15.5m on the package. Part of that amount was expenditure that did not qualify for allowances.

The UT’s error was to determine that, because the Residual Amount could not be attributed to a specific right acquired by the LLP, the entire amount should be treated as expenditure on or in connection with the conversion. But, with respect, that does not take sufficient account of the FTT’s finding that the LLP paid for a package assembled by OVL. There is no principled basis to treat the entire difference between the expenditure incurred by the LLP and the identified costs as expenditure qualifying for BPRA, in circumstances where elements of that expenditure have been found not to qualify.

As far as the land was concerned, it is clear from the FTT’s findings that it was OVL that identified the Property and secured it. OVL also obtained planning permission. That earlier work put in place a structure that allowed the purchase to be completed by the LLP on the date that it signed the transaction documents and incurred the expenditure, 25 March 2011. There was no need to impose any specific obligation on OVL to procure the Property because that had already been done, and the relevant documents could focus instead on ensuring that the price was paid and the related costs (SDLT and legal fees) were met. But it would be wrong to conclude that the lack of a specific obligation meant that it was impermissible to conclude that part of what was paid to OVL could be attributed to its services in procuring the Property. A realistic view of the facts is required. There was no dispute that the expression “on, or connection with” has a similar scope in s.360B(1) and (3). A fee attributable to procuring the Property would fall within the scope of the express exclusion in s.360B(3) for expenditure incurred on or in connection with the acquisition of rights in land.

It is no answer to this to point out that the Property was acquired from an unconnected third party. That is so, but it does not mean that the expenditure incurred by the LLP includes no element attributable to what OVL did to enable that purchase to proceed. The price paid to Multibuild was no doubt also an arm’s length amount, but that does not prevent what is in substance OVL’s fee for that element of the transaction from being capable of qualifying for BPRA. The latter is properly expenditure on or in connection with the conversion. The former is expenditure on or in connection with the acquisition of the freehold.

It was submitted for the LLP that, if some apportionment was required, then it should exclude items from which OVL derived no “profit”, such as the IFA fees and Promoter fees, and should also exclude the Capital Amount because that was an application of profit by OVL. We have concluded that there is no warrant for distinguishing between different categories of non-qualifying expenditure. The nature of the individual elements and the reasons why those elements are determined not to be incurred on or in connection with the conversion, and whether by reason of the explicit exclusion in s.360B(3) or for some other reason, do not provide a principled basis for a distinction to be drawn. The “package” included all those elements.

The precise apportionment was a matter for the FTT, as it recognised when leaving the matter to the parties to attempt to agree it with provision for either party to apply to the Tribunal if agreement was not reached (FTT decision at [231]). We therefore do not need to determine it. However, we would observe that we did not detect any real objection to HMRC’s proposal that it can be done straightforwardly by identifying the percentage of the total expenditure (excluding the Residual Amount) that has been found to be non-qualifying and applying that percentage to the Residual Amount.

We therefore conclude that HMRC’s appeal on Issue 2(e) should be allowed.

We have addressed the substance of this ground of appeal already and do not need to comment on it further.

In conclusion:

I agree.