The essential issue on this appeal is whether a VAT-saving scheme is invalidated by reason of the EU principle of abuse of rights; and, if so, what the consequences are. It is an appeal from the decision of the Upper Tribunal (Rose J and Judge Sinfield), with the permission of the Upper Tribunal, reversing a decision of the First Tier Tribunal (Judge Demack) who had held that the scheme was not an abuse of rights. The decision of the FTT was itself a continuation of an earlier hearing that resulted in a decision in 2002. The reason for the long delay between the two hearings before the FTT was the need for a reference to the CJEU. The decision of the Upper Tribunal is at [2014] UKUT 438 (TCC), [2015] STC 307. The first decision of the FTT (at the time the VAT & Duties Tribunal) is decision 17854 (MAN/00/2630); and the second is at [2013] UKFTT 429 (TC), [2014] SFTD 78.

I can take the factual background from the decision of the Upper Tribunal with some additions from the findings of fact made by the FTT.

The University of Huddersfield (“the University”) makes supplies of education which are exempt from VAT. It makes a few taxable supplies and under the VAT code it is able to recover a proportion of input tax at its partial exemption recovery rate. In 1996 this rate was 14.56 per cent but it later fell to 6.04 per cent. In 1995, the University wanted to refurbish two Grade II listed derelict mills of which it had bought the leasehold. They were known as East Mill and West Mill and were both situated in Canalside, Huddersfield. The University recognised that, as its arrangements then stood, if it paid a construction company to refurbish the mills, it would have to pay VAT to that company for the refurbishment and would only be able to recover a small proportion of that input VAT. The University sought advice from its accountants KPMG as to how to save VAT or defer its liability to pay it. The University dealt with West Mill first, but this appeal is concerned only with the arrangements for East Mill.

After various iterations the eventual scheme that KPMG recommended and the University adopted was as follows.

On 17 November 1995 the University established a discretionary trust (“the Trust”). The deed was in standard form but provided that the power to appoint and remove trustees be vested in the University. The trustees appointed were three former employees of the University and the beneficiaries were the University, any student from time to time enrolled there and any charity. The FTT found that the sole purpose for the creation of the Trust was to facilitate the tax mitigation scheme suggested by KPMG. The Trust was a shell, having capital of only £10; its only other funds consisted of interest-free loans from the University. The FTT also found that the Trust was controlled by the University. There was no evidence that the trustees had ever considered exercising the discretion which the trust deed purported to confer on them. Very regrettably, on the basis of the FTT’s findings, when the University answered HMRC’s questions about the rationale for the Trust it gave bogus answers.

The essence of the scheme was as follows:

On 21 November 1996, the University opted to waive exemption from VAT in relation to East Mill, thus making any lease of it a taxable supply. This was formally permitted by domestic legislation.

On 22 November 1996, the University granted a taxable, 20 year, full repairing lease of East Mill to the Trust. The initial yearly rent was £12.50.

The Trust also elected to waive exemption from VAT in relation to East Mill. This, too, was formally permitted by domestic legislation.

Also on 22 November 1996, the Trust granted a taxable internal repairing underlease of 20 years less 3 days of East Mill to the University at an initial yearly rent of £13.

The University contracted with a company called University of Huddersfield Properties Ltd ('Properties Ltd') for the refurbishment of East Mill. Properties Ltd was a non-VAT group, wholly-owned subsidiary of the University.

Properties Ltd was registered for VAT and contracted with the University to refurbish East Mill. Properties Ltd also issued an invoice to the University in the sum of £3.5 million plus VAT of £612,500 for future construction services on East Mill.

Properties Ltd then engaged building contractors at arm's length to provide the necessary construction services for East Mill.

The University paid the £3.5 million plus £612,500 VAT to Properties Ltd. There was no intention that Properties Ltd would make a profit on the supply. Properties Ltd had no other assets of its own.

The sole reason for the use of the Trust in relation to East Mill was that of facilitating VAT planning, and the sole purpose of the lease by the University to the Trust was also that of facilitating VAT planning. The University contrived to achieve the rent differentials between the lease and underlease by means of the leaseback being on an internal repairing basis only. The sole purpose of the underlease by the Trust to the University of East Mill was to facilitate VAT planning. In addition it was the University's intention, in order to obtain an absolute VAT saving, to collapse the leasehold structure in respect of East Mill after 2 or 3 years, or on the 6^th or 10^th or 15^th anniversary of the commencement of the term of the lease in accordance with the terms of a break clause contained in the lease. The University had the ability to arrange the collapse of the leasehold structure at any time, because of its control of the Trust. It would have done so as soon as its advisers had said it could and thus achieve a VAT advantage in absolute terms. The perceived benefit of the scheme was that the University would be able to recover the full amount of input tax paid on the construction costs in the VAT periods in which they were incurred, while paying output tax on the rental payments as they fell due.

When the University came to complete its VAT return for the period 01/97, it had a net liability to VAT of over £90,000, disregarding the East Mill arrangement. The University completed the VAT return showing a repayment due to it of some £515,000. HMRC unconditionally paid that sum to it. Work on East Mill was completed by third party contractors on 7 September 1998 and the University occupied the building from that date for the purpose of its main economic activity of making exempt supplies. Subsequently, the rents due under the lease and the underlease were increased to £400,000 and £415,000 per annum respectively. The lease and the underlease for East Mill were both terminated on 18 August 2004. As at that date, six years' rent had accrued on both the lease and the underlease. No money had actually changed hands between the University and the Trust although the sums payable on the lease and the underlease were set out in their respective financial accounts.

All the steps in the scheme were accepted as genuine (in the sense of not being shams); and if the provisions of the VAT Act 1994 are formalistically applied, the University was entitled to deduct the input tax that it claimed in its return for the period 01/97. However, that is where the principle of abuse of rights comes into play. The critical finding of fact by the FTT in its 2002 decision at [133] was:

At [136] it added:

The first question, as it seems to me, is to identify what it is that is alleged to constitute the abuse of rights. It is, in my judgment, clear on the facts that the scheme devised by KPMG and adopted by the University consisted of all the steps set out above, including the contemplated collapse of the leasehold structure as soon as it was advised to do so. In examining whether a scheme falls foul of the principle of abuse of rights, it is the scheme as a whole that must be considered: WHA Ltd v HMRC[2013] UKSC 24, [2013] 2 All ER 907 at [26], HMRC v Pendragon plc[2015] UKSC 37, [2015] 1 WLR 2838 at [13].

The principle of abuse of rights rests on two tests, both of which must be satisfied:

The transactions concerned, although formally valid under the VAT code, result in the accrual of a tax advantage the grant of which would be contrary to the purpose of its provisions; and

The essential aim of the transactions is to establish a tax advantage.

There is no dispute about the second test. The essential aim, indeed the sole purpose, of the creation of the Trust, and of the leasehold structure was to obtain a tax advantage. That test is therefore satisfied.

Whether the first test is satisfied entails identifying (a) the tax advantage that the scheme gave the University and (b) the purposes of that part of the VAT code with which we are concerned. It is then necessary to compare the purpose and objectives of the part of the VAT code allegedly being abused with the purpose and results achieved by the activity at issue. If the tax advantage results from a choice that the VAT code intended to give the taxable person, then there is no abuse: Poiares Maduro A-G in Halifax at [88].

VAT is a tax on consumption which is levied at every stage in the chain or production or distribution, with the ultimate burden being borne by the end consumer. Thus it is an inherent part of the scheme that every taxable person who is liable to pay output tax is entitled to deduct input tax. The right to deduct is conferred by article 17 of the Sixth Directive which provides that a taxable person is entitled to deduct “in so far as goods and services are used for his taxable transactions”. The payment of output tax and the deduction of input tax are the obverse and reverse of the same coin. It is the means by which the VAT code relieves the trader of the burden of VAT. As the CJEU said in Halifax at [78]:

The principle of fiscal neutrality is a central feature of the code. However, EU law requires some economic activity to be exempt from VAT. The provision of university education is one such activity. So, too, is the provision of financial services (as in Halifax). Because a person engaged in making supplies exempt from VAT does not have to account for output tax on those supplies, it necessarily follows that he is not entitled to deduct input tax on supplies having a direct and immediate link to those exempt supplies. This, too, is part of the fundamental principle of fiscal neutrality. As Poiares Maduro A-G explained in Halifax at [93]:

He continued at [94]:

One of the three cases under consideration was the University’s own case. Indeed the Advocate-General commented on the University’s case at [95], saying:

The court itself endorsed the Advocate-General’s view at [80]:

Mr Lasok QC, for the University, emphasised that the VAT code worked by matching inputs and outputs on a transaction by transaction basis. Input tax was only deductible where there was a direct and immediate link between the input and a taxable supply. There was no ability in the VAT code to “look through” transactions to the end result. That is the point that Poiares Maduro A-G made in Halifax at [93]. That is true, as far as it goes. But as Lord Neuberger pointed out in WHA Ltd[2007] EWCA Civ 728, [2007] STC 1695 at [22] if that approach is applied to a scheme alleged to be an abuse of rights, a scheme which complied with the formal rules of VAT would never be held to be within the principle of abuse of rights. I therefore reject Mr Lasok’s argument which, moreover, is contrary to the way in which Poiares Maduro A-G went on to develop his opinion in Halifax at [94] and [95] in which he was plainly looking at the end result.

Mr Lasok relied heavily on the decision of the CJEU in (Case C-103/09) HMRC v Weald Leasing Ltd [2011] STC 596. Weald Leasing concerned companies in the Churchill Group of companies, which principally made supplies of exempt insurance services. The operating companies, CML and CARC, were entitled to deduct only 1 per cent of input tax incurred on the purchase of equipment. Another company in the group, Weald Leasing, was set up to carry on the business of buying equipment and leasing it to a company outside the group (Suas Ltd) which in turn sub-leased the equipment to CML and CARC. CML and CARC thus avoided having to purchase directly the equipment they needed and avoided having to pay in a single sum the total amount of non-deductible VAT on those purchases. The aim of the transactions was to divide and spread the payment of that amount in order to defer the Churchill Group's VAT liability. CML and CARC were not immediately liable for the non-deductible VAT on the total cost of the equipment purchased, but were so liable on the amount of rent relating to that equipment, spread over the term of the leasing agreements: see Mazák A-G at [2] to [4]. At [14] the Advocate-General said that the arrangements resulted in a cash-flow advantage to CML and CARC. At [31] the court itself said:

Thus the reference itself identified the tax advantage in quite limited terms. The critical point, however, was that neither CML nor CARC owned the equipment in question so that the choice between acquisition outright and acquisition on lease was a real choice with commercial and practical effects in the real world. In this case, by contrast, the creation of the Trust and the leasehold structure had no commercial effect, except to obtain the VAT advantage.

Mr Lasok put forward four propositions, each said to be firmly based on the decision of the court in Weald Leasing, and submitted that success on any one of the propositions was enough to win this appeal. The four propositions, as Mr Lasok formulated them (and the paragraphs from Weald Leasing on which they were based), were as follows:

Tax advantages that are intrinsic features of normal leasing arrangements are not abusive even if obtained solely for fiscal purposes: Weald Leasing at [33].

Choosing leasing arrangements which procure a fiscal advantage cannot be criticised so long as the correct VAT treatment of the leasing arrangement is followed: Weald Leasing at [34].

A comparison between a particular transaction entered into by a taxable person and a transaction that he normally enters into or needs to enter into is irrelevant in the present context: Weald Leasing at [43] and [44].

If a leasing arrangement contains an element that results in the accrual of an abusive tax advantage, redefinition involves addressing that element rather than the leasing arrangement as a whole: Weald Leasing at [49] to [52].

At [33] the court said:

Taken literally that supports Mr Lasok’s first proposition. Leasing transactions do not in themselves amount to abuse. That is obvious. But, as I have said, the case that the CJEU was considering was a case in which the taxable person wished to acquire an asset which it did not already own. Whether to lease or to buy that asset is clearly not in itself abusive. Here, however, there was no question of the University making a choice between buying or leasing an asset: it already owned the Mills that it wished to develop. I do not consider that this proposition, unexceptionable in itself, advances the University’s case.

At [34] the court said:

In this paragraph the court is discussing different types of transaction as a means of acquiring assets which the taxable person does not already have. Mr Lasok’s second proposition glosses over this point by referring to “leasing arrangements”. In my judgment Mr Lasok’s second proposition is too wide and does not reflect the fact that the court was discussing commercial choices.

At [43] the court said that the considerations that it had discussed were not affected by the question whether the transactions were or were not normal for the trader in question. That is the foundation of Mr Lasok’s third proposition. But Mr Lasok’s third proposition is incomplete, because the court added at [44]:

I do not consider that this paragraph in the court’s judgment supports the unqualified way in which Mr Lasok put his third proposition. To the contrary it contradicts it. It positively requires an examination of the object and effects of the impugned transactions, as well as their purpose.

Mr Lasok’s fourth proposition is relevant only if the impugned transactions are found to be abusive. So I will return to that proposition in due course.

Mr Lasok argued that the Upper Tribunal were wrong to interpret paragraph [80] of the CJEU’s judgment in Halifax (quoted at [17] above) as they did. In Weald Leasing Ltd, Mazák A-G said at [31] to [33] that the court’s reference to “normal commercial operations” in Halifax was intended to deal with the second limb of the test, namely whether the operations in question were essentially designed to achieve a tax advantage; and were not directed to the first part of the test, namely whether the advantage in question was one that was contrary to the purpose of the relevant part of the VAT code. With all respect to the Advocate General I find it impossible to share that conclusion. Paragraph [80] of the judgment in Halifax is the only part of the judgment in which the court clarifies the application of the first of the two Halifax tests; and is immediately followed by paragraph [81] which begins with the words “As regards the second element” which clearly turns from the first test to the second. Moreover, the law is not made by the Advocate-General but by the CJEU. In Weald Leasing the CJEU did not specifically endorse this part of the Advocate-General’s opinion: see [44] quoted at [28] above. In addition in Pendragon at [11] Lord Sumption said that:

This was said in the context of a consideration of Weald Leasing and represents a different view from that expressed by the Advocate-General.

Mr Lasok also argued that the question whether a transaction was artificial goes only to the second Halifax test, and is irrelevant in considering the first Halifax test. I find that submission very difficult to square with (Case C-110/99) Emsland-Stärke GmbH v Hauptzollamt Hamburg-Jonas [2000] ECR I-11569 (cited in Pendragon at [6]) in which the court described the principle of abuse of rights as follows:

Clearly the artificial creation of conditions which formally comply with the requirements for obtaining a tax advantage is at the heart of the principle of abuse of rights. If Mr Lasok’s submission were correct it would simply substitute one form of formalism for another.

Mr Lasok criticised the Upper Tribunal for having described the tax advantage in what he said were ambiguous and in places inconsistent terms. At [71] the Upper Tribunal said:

But at [79] they said:

I do not consider that [71] is ambiguous in the way that Mr Lasok suggested. Read fairly and as a whole, it seems to me to identify the tax advantage as the entitlement to claim 100 per cent of the input tax as a deduction rather than the much smaller percentage which, as a supplier of mainly exempt supplies, the University would have been able to deduct under normal circumstances.

Mr Lasok then argued that the Upper Tribunal had incorrectly identified the tax advantage that the University obtained. This was not a case of an absolute VAT saving, but merely a deferral scheme with the possibility of a future absolute saving. He said this was a question of fact which the FTT had determined, and a finding with which the Upper Tribunal was not entitled to interfere. I reject this submission. First, it overlooks the FTT’s finding of fact that it was always the University’s intention to make an absolute VAT saving at the maximum amount that could be obtained. This finding of fact goes far beyond deferral with the mere “possibility” of an absolute saving at a later date. Second, the scheme had what the FTT described as a “built-in feature” that allowed the absolute saving to be made. Since it was “built-in,” that feature was part of the scheme from the outset.

Allied to this argument was the argument that the abusive tax advantage (in the shape of an absolute VAT saving) would only accrue to the University if and when the leasehold structure was collapsed. At that point HMRC would be entitled either to raise an assessment under section 73 of the VAT Act 1994 (interpreted so as to conform with the principle of abuse of rights) or to bring an action at common law for restitution of the input tax originally deducted which, in the language of the CJEU in Halifax at [93], had been “rendered undue”.

On the facts found by the FTT I agree with the Upper Tribunal that the tax advantage was the ability to deduct the whole of the input tax, which the University did in its return for the period 01/97. This is quite unlike Weald Leasing, where in the view of the CJEU all that happened in fiscal terms was a spreading of the payment of irrecoverable VAT over a period. The advantage in that case was no more than a cash flow advantage. Mr Lasok argued that if the leases in the present case had run their course, then there would have been no more than a cash flow advantage accruing to the University. But that is an unrealistic appreciation of the scheme because on the FTT’s findings of fact it was always the intention to collapse the leasehold structure by using a built-in feature as soon as possible. There was never any real possibility that the leases would run their course. The scheme was abusive from the outset, and there was no need to wait for the inevitable collapse of the leasehold structure.

Nor do I accept that the characterisation of the tax advantage is a pure question of fact. The Upper Tribunal was entitled to find that, on the basis of the primary facts found by the FTT (which included its finding that it was always intended to collapse the leasehold structure by means of a built-in feature as soon as possible), the FTT had mischaracterised the nature of the tax advantage. In addition since the Upper Tribunal had held that the FTT was wrong in law not to have considered the scheme as a whole (a conclusion which is not challenged on this appeal), it was entitled to set aside the FTT’s decision and remake the decision itself. In remaking the decision the Upper Tribunal was not bound by the FTT’s findings of fact. It was entitled to come to its own conclusions: Pendragon at [50].

I do not agree that the Upper Tribunal’s statement at [79] is inconsistent with what it said at [71]. The Upper Tribunal’s statement at [79] is, on proper analysis, directed to a different way in which HMRC put their case, namely that exercising the option to tax was a separate and different abuse. It does not in my judgment conflict with what the Upper Tribunal said at [71].

Having identified the tax advantage the Upper Tribunal then considered whether that tax advantage (namely the deduction by the University of the full amount of input tax) was incompatible with the purposes of the VAT code. They said at [72] that HMRC put their case in two ways:

Having identified that purpose the Upper Tribunal then considered whether that could be said to be the relevant purpose. In answering that question affirmatively they relied on paragraph [93] of the opinion of Poiares Maduro A-G in Halifax, and paragraph [80] of the court’s judgment in the same case. Although the Upper Tribunal did not refer in terms to paragraph [94] of the Advocate-General’s opinion in this part of their decision, it is clear that he thought that the circumvention of the general principle stated at [93] was abusive; and that in itself shed light on the relevant principle.

Mr Lasok argued that if that was the principle, it had never been articulated before in any of the many cases that had come before the CJEU about the deduction of input tax by traders who made exempt or partially exempt supplies. He may be right about that, but in my judgment that does not detract from the principle articulated by the Advocate-General and the court in Halifax. There is, after all, a first time for everything; and the decision of the Grand Chamber in Halifax was intended to deal for the first time with the question of abuse of rights in the context of VAT. Paragraph [80] of the court’s judgment makes the position clear.

The Upper Tribunal said at [74]:

In my judgment the Upper Tribunal were correct at paragraph [74]. As required by Weald Leasing at [44], in considering whether the first Halifax test was met it examined the object of the transactions, their effect and their purpose. The object of the scheme in the present case was, as is conceded, to obtain a VAT advantage by the interposition of actors and transactions which had no other purpose than to secure that VAT advantage; and those transactions had no commercial effect other than to secure that advantage. I can see no flaw in the Upper Tribunal’s reasoning in this respect.

Having found that the tests for abuse of rights were both met, the Upper Tribunal then had to redefine the transactions. This is where Mr Lasok’s fourth principle comes into play. They decided that the way to do this was to disregard the artificial steps (i.e. the creation of the Trust and the creation of the leasehold structure). Mr Lasok took issue with this, arguing that the University could have achieved the same tax advantage by entering into a similar leasehold arrangement as part of an arms’ length financing package. However, in my judgment the question of redefinition does not enable past history to be completely rewritten. The fact is that the University did not enter into any financing package. On the facts found it paid Properties Ltd out of its own funds. Properties Ltd was merely the conduit through which monies passed from the University to the contractor. The fact that other, non-abusive, structures could have been adopted does not undo the abusive nature of what the University in fact did. The task for the Upper Tribunal was to remove the abusive elements of the scheme; not to replace them with a different and wholly fictional scheme. That is precisely what they did. I cannot see any legal error in their approach.

HMRC raised an additional point by way of Respondent’s Notice, which the Upper Tribunal found unnecessary to decide. Deciding that point is not necessary for this appeal either; and I would leave it for a case in which it matters.

I would dismiss the appeal.

I agree.