Pendragon Plc v HM Revenue and Customs

This appeal is brought against a decision of the Upper Tribunal (Tax and Chancery Chamber) by which an appeal by HMRC was allowed against a decision of the First-Tier Tribunal (Tax Chamber) in favour of the present appellants. The First-Tier Tribunal (Judge Adrian Shipwright and Sheila Wong Chong FRICS) held on 31 July 2009, after a hearing lasting 8 days, that the arrangements entered into by the present appellants (whom I will call the Pendragon Group, or Pendragon for short) did not involve conduct falling within the European law principle of abuse of right. On that basis it set aside assessments and misdeclaration penalties imposed by HMRC. On 29 November 2011 the Upper Tribunal (Mr Justice Morgan and Judge Colin Bishopp) allowed an appeal by HMRC, holding that the arrangements did involve abuse of right, and reinstated the assessments. The appellants appeal with permission granted by Lewison LJ.

The Pendragon Group is said to be the largest car sales group in Europe. Its principal areas of business are the sale of new and used cars. For the purposes of this appeal the important part of its business is the sale of demonstrator cars. These are cars which potential customers are invited to use to test drive on the road, and the phrase also includes loan or courtesy cars which are lent to customers for repair services while their own vehicles are being repaired. The demonstrator cars in question were all being driven on the road for a number of months, and they counted as used cars by the time that, having ceased to be used in the business as demonstrator cars, they were offered for sale to the public. As the Pendragon Group bought these cars new and sold them as used cars, the profit margin would be either non-existent or very low.

The practice of using cars as demonstrator cars in this way, which is normal in the relevant trade, affects the economics of the car dealer’s operation not only as regards the effect on the resale price. A car dealer’s stock of cars requires funding in any event, since it has to pay the manufacturer on purchase but has to wait until a sale to a purchaser in order to realise the value of the vehicles. As regards demonstrator cars, all the more time will elapse between purchase and eventual sale by the dealer. The requirement of finance for the dealer’s operations is an ever-present factor.

In accordance with normal principles of VAT, tax would ordinarily be payable on the sale of demonstrator cars by reference to the full amount of the sale price. When the Pendragon Group bought a car new from the manufacturer, VAT would be charged on the full price. This would be the input VAT suffered by the purchaser, which would in due course be recovered by set-off against output tax for which the purchaser would be accountable on its sales and other supplies. However, special provision is made in the VAT legislation for sales of second hand goods, in some circumstances, so that they may be taxed on a different basis, namely on the profit margin. This is known as the margin scheme. Conditions laid down in legislation must be met for it to apply, but even if they are met it is a matter of choice for the relevant taxable person as to whether it shall apply.

The appeal concerns a scheme which involved the use of the margin scheme by the eventual seller of the used cars, but also involved the recovery in full of the input tax incurred on the original purchase of the new cars. I will set out the details shortly, but I will refer to it as the Scheme. It was devised by KPMG, and was put into operation by Pendragon on two occasions, involving substantial sums of money and substantial numbers of cars each time.

HMRC contend that the Scheme was wholly artificial, did not reflect economic reality and was set up with the sole or essential aim of obtaining a tax advantage of a kind which was illegitimate, that is to say one which was inconsistent with the policy of the relevant legislative provisions. That advantage was that the Pendragon Group was able to recover all of the input tax incurred on the purchase of the cars from the manufacturer, but was able to apply the margin scheme to the eventual sale of the cars to a retail purchaser. The First-Tier Tribunal did not accept this contention, holding that the essential aim of the transactions was to obtain finance, but the Upper Tribunal held that the FTT had erred in law in this respect, and that the essential aim was to obtain an illegitimate tax advantage. The principal question on the appeal to this court is whether, in doing so, the Upper Tribunal went beyond what is properly open to an appellate court or tribunal where facts have been found and evaluated by the court or tribunal from which the appeal is brought.

As in the Upper Tribunal, the case was argued by Mr Cordara Q.C. leading Ms Sloane for Pendragon and by Mr Pleming Q.C. leading Mr Thomas for HMRC, all of whom (apart from Mr Thomas) had also appeared before the First-Tier Tribunal. I am grateful to them for their assistance in a case which is complex as regards both the relevant law and the factual and procedural history.

For the reasons which I set out below, at inordinate length which I regret, I have come to the conclusion that the First-Tier Tribunal was entitled to come to the conclusion that it reached and that the Upper Tribunal was wrong to reverse the decision. I would therefore allow the appeal.

The essential basis of VAT was set out in article 2 of the First Council Directive on VAT, 67/227/EEC:

“The principle of the common system of value added tax involves the application to goods and services of a general tax on consumption exactly proportional to the price of the goods or service, whatever the number of transactions which take place in the production and distribution process before the stage at which tax is charged.

On each transaction, value added tax, calculated on the price of the goods or services at the rate applicable to such goods or services, shall be chargeable after deduction of the amount of value added tax borne directly by the various cost components.

The common system of value added tax shall be applied up to and including the retail trade stage.”

In the normal way, when the manufacturer of a new product such as a motor car sells it, it will charge VAT on the price (referred to as output tax), and will account for that to HMRC after deducting tax paid on supplies made to it (input tax), including that on supplies of goods and services made to it and referable to the process of manufacture. Assuming that the sale is not itself a retail sale, but is to another trader (such as the Pendragon Group in the case of cars), that trader will charge VAT when it sells the car on, and will deduct, against that and other VAT charged by it on the supplies which it makes, the input tax charged to it on supplies made to it, including on the purchase of the car. If and when the goods are sold to a purchaser who is not a taxable person, such as an ordinary retail purchaser, then that person suffers VAT, as it is charged on the sale to him, but he cannot recover it because he makes no taxable supplies (or no relevant such supplies) on which he charges VAT from which he can deduct the VAT charged to him on the purchase of the car. That is the point at which VAT is finally charged and paid so that the full amount of tax is received by HMRC. I need not refer to the detail of the European legislation by which this principle was promulgated, nor to that of its implementation in the UK by the Value Added Tax Act 1994 (VATA).

Where that system applied in the ordinary way to the process by which the Pendragon Group bought and sold motor cars, a company within the group would suffer VAT (input tax) on the purchase of a new car from a manufacturer (let us say, for illustration purposes, Ford), it would charge VAT (output tax) on the sale to a purchaser at a price which, assuming the car is still new and subject to the state of the market, should be higher than the price paid to Ford, and it would deduct the input tax incurred by it on the purchase against the output tax charged by it on the sale, leaving it, it would hope, with a balance in its favour on the transaction as a whole, having recovered by deduction (or by repayment from HMRC) the input VAT which it incurred.

In the case of a car which a Pendragon company uses for a time as a demonstrator car before selling it, the price on the ultimate sale to a retail buyer is likely to be less than the price paid on the initial purchase.

A car dealer may come to sell a used car to a retail purchaser in a number of different circumstances. For example, on the sale of a new car, the retail buyer may wish to sell the car he already owns, and the dealer may take it in part exchange for the new car, at an agreed price. In that situation, VAT is charged on the sale of the new car, but not (in the case of an ordinary retail buyer) on the sale of the old car back to the dealer. If the dealer then sells the old car to another retail buyer, that sale attracts VAT, but the dealer has not incurred any relevant input tax referable to that car. The margin scheme allows the dealer to charge VAT on the profit margin (if any) on the sale of the car rather than on the full sale price, but it does not allow the dealer to recover any input tax by deduction or otherwise as against the VAT charged and accounted for under the margin scheme.

By contrast, on the eventual sale by the Pendragon Group of a car which has been used as a demonstrator car, the group will have incurred input tax on the purchase but, in ordinary circumstances, will not previously have charged output tax in respect of that vehicle against which the input tax can be offset. (Of course, this is an over-simplification, for explanation purposes. Given that VAT is accounted for at least quarterly, the input tax on the purchase of vehicles (and other supplies) in one quarter will have been set off against output tax charged in the same quarter, on sales of vehicles, all or many of which will be different from those acquired in the same quarter. However, the overall economic effect for the Group of the pattern of purchases and sales can be understood in this way.)

As regards European legislation, the margin scheme was introduced by a Council Directive on VAT, 94/5/EC, which amended the Sixth Directive, 77/388/EEC, so as to include a new article 26A. This provided for special arrangements which were applicable to second-hand goods, works of art, collectors’ items and antiques. Paragraph 1 provided that, in respect of supplies of second-hand goods effected by taxable dealers, Member States were to apply special arrangements for taxing the profit margin made by the taxable dealer. Paragraph 2 listed the circumstances in which the supplies of goods were to be supplies of second-hand goods by a taxable dealer. Paragraph 3 provided that

“the taxable amount of supplies of goods referred to in paragraph 2 shall be the profit margin made by the taxable dealer, less the amount of value added tax relating to the profit margin. That profit shall be equal to the difference between the selling price charged by the taxable dealer for the goods and the purchase price.”

Paragraph 6 said that the seller of goods under the margin scheme could not deduct the tax on supplies made to it in respect of the relevant goods. Thus, while a taxable person who is able to, and chooses, to, apply the margin scheme had to charge little or no VAT on the sale of a second hand car, thereby making it easier to offer an attractive price to a purchaser, it could not recover its input tax incurred on supplies relevant to the car.

The recitals to Directive 94/5/EC refer to article 32 of the Sixth Directive, by which the Council was to adopt a Community taxation system to be applied to used goods, works of art, antiques and collectors’ items. They show, as is the case, that different Member States had operated different schemes for the taxation of second hand goods for a long time. The recitals continued as follows:

“Whereas the present situation, in the absence of Community legislation, continues to be marked by the application of very different systems which cause distortion of competition and deflection of trade both internally and between Member States; whereas these differences also include a lack of harmonization in the levying of the own resources of the Community; whereas consequently it is necessary to bring this situation to an end as soon as possible;

Whereas the Court of Justice has, in a number of judgments, noted the need to attain a degree of harmonization which allows double taxation in intra-Community trade to be avoided;

Whereas it is essential to provide, in specific areas, for transitional measures enabling legislation to be gradually adapted;

Whereas, within the internal market, the satisfactory operation of the value added tax mechanisms means that Community rules with the purpose of avoiding double taxation and distortion of competition between taxable persons must be adopted;”

The European legislation was given effect in the UK by the Value Added Tax (Cars) Order 1992, SI 1992 3122, which I will call the Cars Order. I will need to deal with some of the detail of this Order when I come to examine the Scheme. I will also need to refer to the Value Added Tax (Special Provisions) Order 1995, SI 1995/1268 (the Special Provisions Order) at that stage.

This is a principle of European law which has been applied on several occasions in the area of VAT, but is not limited to that field. I will discuss it in more detail later, but for the moment it is sufficient to quote two paragraphs from the judgment of the ECJ in Case C-255/02 Halifax v Customs & Excise Commissioners [2006] STC 919:

“74.

In view of the foregoing considerations, it would appear that, in the sphere of VAT, an abusive practice can be found to exist only if, first, the transactions concerned, notwithstanding formal application of the conditions laid down by the relevant provisions of the Sixth Directive and the national legislation transposing it, result in the accrual of a tax advantage the grant of which would be contrary to the purpose of those provisions.

75.

Second, it must also be apparent from a number of objective factors that the essential aim of the transactions concerned is to obtain a tax advantage. As the Advocate General observed in paragraph 89 of his opinion, the prohibition of abuse is not relevant where the economic activity carried out may have some explanation other than the mere attainment of tax advantages.”

The Scheme used several captive companies under the control of Pendragon plc as well as a bank resident in Jersey and independent of the Pendragon Group. The captive companies were referred to as Captive Cos 1, 2, 3, 4 and 5. The relevant parties entered into a series of contracts designed to give effect to the Scheme. The Scheme worked in five steps. It is common ground between the parties that each transaction complied with all the relevant statutory requirements under UK law. For this reason, as well as outlining each step and the documentation required for the relevant transactions in it, I also indicate briefly each transaction’s VAT consequences, absent any application of the principle of abuse of rights. This highlights the aspects of the Scheme which were advantageous in terms of VAT. I will explain the VAT treatment in more detail later.

Step 1. Pendragon plc, having bought new cars from, say, Ford, sold new cars which were destined for use as demonstrator cars, before sale to a consumer, to Captive Cos 1, 2, 3 and 4 (“the Captive Leasing Companies” or CLCs). (In fact only three companies were used, but I use the language which has been used elsewhere to describe the Scheme, in order not to generate unnecessary confusion.) Pendragon plc’s sale of the cars to a CLC was a taxable supply of goods for VAT purposes. Therefore, Pendragon plc accounted for output tax on the sale of the cars, and reclaimed input tax, including the tax incurred on the purchase from Ford.

Step 2. On the same day as Step 1, the Captive Leasing Companies leased the cars pursuant to hybrid HP/lease agreements to dealership companies in the Pendragon Group (“the Dealerships”). Each of the Captive Leasing Companies entered into a “Vehicle Demonstrator Hire Agreement” (referred to as a hybrid lease) in favour of the Dealerships. Paragraph 8(c) of Second Schedule to the hybrid leases (generally referred to as clause 8(c), as I will refer to it hereafter, so as to avoid confusion) conferred on the Dealership an option to purchase the hired vehicles. The option was exercisable 7 days after the end of the hire agreement, and not earlier.

The services provided by the Captive Leasing Companies to the Dealerships under the Vehicle Demonstrator Hire Agreement were taxable supplies at the standard rate of VAT. Input tax incurred by the Captive Leasing Companies on the purchase of the vehicles from Pendragon plc at Step 1 was therefore fully recoverable, being attributable to the making of those taxable supplies of leasing to the Dealerships. The Dealerships incurred VAT on the rental payments but recovered that VAT in full, being attributable to their taxable sale activities.

Step 3. On the day following Steps 1 and 2, the Captive Leasing Companies began assigning the hybrid lease agreements and title in the cars to SG Hambros Bank and Trust (Jersey) Ltd, known in the case as Soc Gen Jersey (SGJ), which was resident in Jersey, not in the UK. Each of the Captive Leasing Companies entered into a Deed of Assignment with SGJ. SGJ paid the Captive Leasing Companies the sum of approximately £20 million. On the same date, SGJ had entered into a facility agreement with its parent company in the UK, SG London, in relation to the facility of £20 million to finance the assignments. SGJ granted SG London an assignment of the assets to be assigned to it, as a form of security.

This step was critical to the success of the Scheme. It depended on the assignment of a lease, granted by a Captive Leasing Company to a Pendragon dealership, to a bank; according to HMRC this had to be an offshore bank, as it in fact was. No VAT was due on this transaction. The assignment by the Captive Leasing Companies to SGJ was not a supply for VAT purposes, by virtue of Article 5(4) of the Special Provisions Order, which “de-supplied” it, i.e. treated it as neither a supply of goods nor a supply of services.

Step 4. On a date envisaged as being some 30 to 45 days later, SGJ transferred as a going concern the lease agreements and title in the cars to Captive Co 5. Captive Co 5 resolved to purchase the relevant “hire business” carried on by SGJ. On the same day, SGJ contracted with Captive Co 5 to sell to it the business of the hire of cars said to have been carried on by SGJ. The consideration was in excess of £18 million and was apportioned as to £100,000 for the sale of goodwill and as to the balance (save for £2) for the sale of the motor vehicles. That agreement was completed on the same date, and Captive Co 5 paid the agreed price to SGJ.

The sale by SGJ to Captive Co 5 of its “hire business” was the transfer of a business as a going concern (TOGC). As such the transaction was neither a supply of goods nor a supply of services; therefore no VAT was due on this transaction.

Step 5. On various dates thereafter, the cars were sold to customers by the Dealerships, acting as undisclosed agents for Captive Co 5 in which title to the vehicles was vested. VAT was charged to the purchasers on the seller’s profit margin on the sale, rather than on the total sale price, Captive Co 5 having opted to apply the margin scheme.

When Captive Co 5 sold the vehicles to the retail customer, the Cars Order applied. The tax relief provided for by Article 8 of that Order applied only where the taxable person making the sale had come into possession of the car in the circumstances set out in Article 8(2), which I will set out below. If those requirements were met, and if the option was exercised that the margin scheme should apply, then VAT was due only on the profit margin on the supply, rather than on the whole value received for the supply. This meant that Captive Co 5 accounted for VAT on the difference between the cost of the car on the purchase from SGJ, and the price at which it sold the car to the consumer. By means of the de-supplied assignment of the leases to SGJ at Step 3, and the TOGC from SGJ at Step 4, the Scheme was designed to meet the taxation requirements of the Cars Order.

Pendragon implemented the Scheme twice. The first implementation took place between 16 November and 22 December 2000 (a 37 day period). The second implementation occurred from 22 February to 5 March 2001 (an 11 day period). The Scheme might have been used again, but HMRC refused to accept that it worked, and therefore did not give credit to the Pendragon Group for the input tax incurred on the purchase of the cars from the manufacturer. That made it unattractive to Pendragon. The second implementation was over a shorter period than had originally been envisaged because the 2001 Budget day was announced as being 7 March, and it was thought advisable that the Scheme should have been completed by then in case of the announcement of legislation which might affect its validity. It was also the case that the Pendragon Group was then well advanced in negotiations for a larger-scale financing facility with another bank. It has not been suggested that different considerations applied to either of the two successive implementations of the Scheme by Pendragon.

The Captive Leasing Companies were Pendragon Company Car Finance Ltd, Pendragon Demonstrator Finance Ltd and Pendragon Demonstrator Finance November Ltd. Captive Co 5, the last owner of the cars within the group, was Pendragon Demonstrator Sales Ltd (PDS). The dealership companies were Stripestar Ltd, Pendragon Motorcycles Ltd, Pendragon Viking Ltd and Arena Auto Ltd.

HMRC’s position as regards the Scheme was set out formally in a letter dated 22 October 2001. The heading to this letter is: Arrangements for the sale of cars under the margin scheme notwithstanding recovery of VAT on their purchase. This gives an indication of the nature of HMRC’s objection to the Scheme. The letter recorded that, before November 2000, the Dealership companies were obliged to, and did, account for VAT on the full sale price of demonstrator cars when sold to a retail customer. It asserted that the margin scheme was only available to those who have not recovered VAT on the purchase of the cars they sell. The Scheme was said to have been intended to create an artificial entitlement to use the margin scheme by exploiting a loophole. The commercial reality of the Scheme was said to be that the dealership companies sold cars to retail customers, the cars having first been acquired from manufacturers and having been used in the meantime as demonstrator cars. In addition, it was noted, the Pendragon Group was provided with short term finance by a bank. HMRC contended that the proper VAT treatment of the sequence of events was that VAT was chargeable on the sale to the retail customer on the full sales price of the car. I need not go into detail as to the contentions advanced in the letter.

It was against the decision set out in this letter, and various consequential assessments and impositions, that relevant companies in the Pendragon Group appealed to the First-Tier Tribunal.

At the material time, the critical UK legislative provision was article 8 of the Cars Order, as it had been amended by the VAT (Cars) (Amendment) Order 1995, SI 1995/1269 and the VAT (Cars) Amendment) Order 1997, SI 1997/1615. Article 8(1) and (2) were then as follows:

“(1)

Subject to complying with such conditions (including the keeping of such records and accounts) as the Commissioners may direct in a notice published by them for the purposes of this Order or may otherwise direct, and subject to paragraph (3) below, where a person supplies a used motor car which he took possession of in any of the circumstances set out in paragraph (2) below, he may opt to account for the VAT chargeable on the supply on the profit margin on the supply instead of by reference to its value.

(2)

The circumstances referred to in paragraph (1) above are that the taxable person took possession of the motor car pursuant to—

(a)

a supply in respect of which no VAT was chargeable under the Act or under Part I of the Manx Act;

(b)

a supply on which VAT was chargeable on the profit margin in accordance with paragraph (1) above, or a corresponding provision made under the Manx Act or a corresponding provision of the law of another member State;

(bb) a supply received before 1 March 2000 to which the provisions of article 7(4) of the Value Added Tax (Input Tax) Order 1992 applied;

(c)

a transaction except one relating to the transfer of the assets of a business or part of a business as a going concern which was treated by virtue of any Order made or having effect as if made under section 5(3) of the Act or under the corresponding provisions of the Manx Act as being neither a supply of goods nor a supply of services,

(d)

a transaction relating to the transfer of the assets of a business or part of a business as a going concern which was treated as neither a supply of goods nor a supply of services if the transferor took possession of the goods in any of the circumstances described in this paragraph.”

Thus, the ability of the taxable person to opt to apply the margin scheme to the sale of a used car depended on the circumstances in which he took possession of the car. Under paragraphs (a) to (c) of article 8(2) attention was focussed on the transaction under which the taxable person took possession of the car. Under paragraph (d), where the taxable person took possession of the goods by a transaction relating to the transfer of the assets of a business as a going concern (TOGC), it depended on whether the transferor took possession of the goods in any of the circumstances to which paragraphs (a) to (c) apply. Thus, if the transferor could have sold the cars using the margin scheme, but sold his business as a going concern, the purchaser was in the same position.

The amendments made in 1995 were connected with the Special Provisions Order. I need to refer to article 5(1) and (4) of that order.

“(1)

Subject to paragraph (2) below, there shall be treated as neither a supply of goods nor a supply of services the following supplies by a person of assets of his business—

(a)

their supply to a person to whom he transfers his business as a going concern where—

(i)

the assets are to be used by the transferee in carrying on the same kind of business, whether or not as part of any existing business, as that carried on by the transferor, and

(ii)

in a case where the transferor is a taxable person, the transferee is already, or immediately becomes as a result of the transfer, a taxable person or a person defined as such in section 3(1) of the Manx Act;

(b)

their supply to a person to whom he transfers part of his business as a going concern where—

(i)

that part is capable of separate operation,

(ii)

the assets are to be used by the transferee in carrying on the same kind of business, whether or not as part of any existing business, as that carried on by the transferor in relation to that part, and

(iii)

in a case where the transferor is a taxable person, the transferee is already, or immediately becomes as a result of the transfer, a taxable person or a person defined as such in section 3(1) of the Manx Act.

…

(4)

There shall be treated as neither a supply of goods nor a supply of services the assignment by an owner of goods comprised in a hire-purchase or conditional sale agreement of his rights and interest thereunder, and the goods comprised therein, to a bank or other financial institution.”

Step 1 in the Scheme involved Pendragon selling new cars, just acquired from the manufacturer, to the Captive Leasing Companies. These were ordinary taxable supplies. Pendragon charged VAT on the full price, and accounted for this to HMRC, deducting its input tax including the VAT paid on the purchase from the manufacturer. The Captive Leasing Companies incurred input tax on the purchase in the ordinary way.

At Step 2, the Captive Leasing Companies made taxable supplies on entering into the hybrid leases in favour of the dealership companies. The CLCs charged VAT to the dealership companies and recovered the input tax they had suffered by deduction against that output tax. Because the hybrid leases provided that title to the goods was not to pass when the final payment was made, but only afterwards, these were treated as supplies of services, not of goods: VATA, Schedule 4, paragraph 1. The dealerships deducted the input VAT which they suffered against their own output tax from their sales business.

At Step 3, the assignment by the CLCs to SGJ was treated as neither a supply of goods nor a supply of services. This is the result of article 5(4) of the Special Provisions Order, quoted above. The CLCs were the owners of goods comprised in a hire-purchase or conditional sale agreement, and they assigned their rights and interest under the agreement, and the goods comprised in the agreement, to a bank or other financial institution. Thus no VAT was due on this transaction. The parties were not agreed as to whether at this stage it mattered that the assignee was not resident in the UK.

At Step 4, when SGJ sold its hire business to Captive Co 5 (PDS), this was a TOGC. Under article 5(1) of the Special Provisions Order this is treated as neither a supply of goods nor a supply of services. So no VAT was incurred.

At Step 5, when PDS sold a used car to a retail customer, it opted to apply the margin scheme to the sale. PDS had come into possession of the goods under a transaction with SGJ which was a TOGC, treated as neither a supply of goods nor a supply of services, and SGJ had taken possession of the goods under a transaction (not being one relating to a TOGC) which was treated by virtue of an Order under section 5(3) of VATA as being neither a supply of goods nor a supply of services. As mentioned above, that was true of what happened at Step 4, and the Special Provisions Order was indeed made under section 5(3) of the Act.

So that is how the Scheme worked, in terms of the national legislation.

As I mentioned, there was some debate as to whether it would still have worked if Step 4 had involved an assignment to SG London, rather than to the Jersey-based subsidiary. On the face of it, Step 4 would be treated in the same way regardless of the residence of the bank, since article 5(4) of the Special Provisions Order simply speaks of an “assignment … to a bank or other financial institution”. For HMRC Mr Pleming argued that the offshore status was necessary because, if the transferor was a taxable person (as the CLCs were) then the effect of article 5(1)(a)(ii) or (b)(iii), as applicable, was that the transfer of the hire business of the CLCs to SGJ (i.e. Step 3) would have been a TOGC, and for that reason it would not have fallen within the terms of article 8(2)(c). According to a warranty given in the course of the transaction, SGJ was not a taxable person, and did not become one, so the TOGC provisions did not apply. On that basis, its acquisition of the business could not be a TOGC within the terms of article 5(1) of the Special Provisions Order, and so it was within article 8(2)(c) of the Cars Order, by reference to article 5(4) of the Special Provisions Order. Mr Cordara for Pendragon did not accept this, contending that the transaction fell within article 5(4) even if it was also a TOGC. It is not necessary to resolve this debate. The point of it was that HMRC identified the use of an off-shore bank as one of the elements of artificiality in the Scheme, especially having regard to the fact that the funds were supplied by SG London, and therefore, it was submitted, the transaction could more naturally have been entered into by SG London itself.

The Halifax case, already cited, concerned a scheme under which the bank, which could only recover a very small proportion of its input tax (because financial services are exempt for VAT purposes), sought to be able to recover all of the input tax on the construction of buildings for use as call-centres. The arrangements made complied with the requirements of the national legislation, just as those adopted in the present case do. If this were not so it would not be necessary to have resort to the abuse of right principle. I do not need to go into detail as to the nature of the arrangements used by Halifax plc, which appear from the reports of the case.

We were shown some passages from the opinion of Advocate-General Poiares Maduro. After a thorough review of the circumstances in which the principle of abuse of rights had been applied as an autonomous concept of European law, he started his summary of the position in his paragraph 87, as follows:

“I am of the view therefore that the Community law notion of abuse, applicable to the VAT system, operates on the basis of a test comprising two elements. Both elements must be present in order to establish the existence of an abuse of Community law in this area. The first corresponds to the subjective element mentioned by the Court in Emsland [2000] ECR I-11569, but it is subjective only in so far as it aims at ascertaining the purpose of the activities in question. That purpose — which must not be confused with the subjective intention of the participants in those activities — is to be objectively determined on the basis of the absence of any other economic justification for the activity than that of creating a tax advantage. Accordingly, this element can be regarded as an element of autonomy. In fact, when applying it, the national authorities must determine whether the activity at issue has some autonomous basis which, if tax considerations are left aside, is capable of endowing it with some economic justification in the circumstances of the case.”

The first element which he there describes is the second of the two elements as set out by the Court in its judgment in Halifax, as quoted above. Having referred to the other necessary aspect in his paragraph 88, he then said this at paragraph 89, to which the Court referred in its paragraph 75:

“The prohibition of abuse, as a principle of interpretation, is no longer relevant where the economic activity carried out may have some explanation other than the mere attainment of tax advantages against tax authorities. In such circumstances, to interpret a legal provision as not conferring such an advantage on the basis of an unwritten general principle would grant an excessively broad discretion to tax authorities in deciding which of the purposes of a given transaction ought to be considered predominant. It would introduce a high degree of uncertainty regarding legitimate choices made by economic operators and would affect economic activities which clearly deserve protection, provided that they are, at least to some extent, accounted for by ordinary business aims.”

He continued, at paragraph 91:

“On the basis of the foregoing analysis I am therefore of the opinion that there is a Community law principle of interpretation prohibiting the abuse of Community provisions, which is also applicable to the Sixth Directive. According to that principle, the provisions of the Sixth Directive must be interpreted as not conferring the rights that might appear to be available by virtue of their literal meaning, when two objective elements are found to be present. First, that the aims and results pursued by the legal provisions formally giving rise to the tax advantage invoked would be frustrated if that right were conferred. Second, that the right invoked derives from economic activities for which there is objectively no other explanation than the creation of the right claimed.”

I have already quoted paragraphs 74 and 75 of the judgment of the Court in the same case, from which it appears that the Court adopted the same approach, applying two criteria. In the passage leading up to these paragraphs, the Court recognised that, on the one hand, according to settled case law, Community law could not be relied on for abusive or fraudulent ends, and that Community law could therefore not be applied to “transactions carried out not in the context of normal commercial operations, but solely for the purpose of wrongfully obtaining advantages provided for by Community law” (paragraphs 68 and 69). On the other hand the Court also recognised the need for legal certainty, so that those concerned, especially with rules liable to entail financial consequences, should know precisely the extent of the obligations imposed on them (paragraph 72), and it also recognised the right of the trader to choose to structure its business in such a way as to limit its tax liability (paragraph 73). At paragraph 86, and in turn in paragraph 2 of the dispositif, summarising its answer to the particular question posed by the referring court, the Court repeated the substance of paragraph 74 and the first sentence of paragraph 75. The Court went on to hold, in response to the second question referred, that a finding of abuse must not lead to a penalty, but only to an obligation to repay as a consequence of a finding that deductions of input tax had not been properly made. It held that transactions involved in an abusive practice must be re-defined so as to re-establish the situation that would have prevailed in the absence of the transactions constituting that abusive practice.

The Court also said this about the second element in the test, at paragraph 81:

“As regards the second element, whereby the transactions concerned must essentially seek to obtain a tax advantage, it must be borne in mind that it is the responsibility of the national court to determine the real substance and significance of the transactions concerned. In so doing, it may take account of the purely artificial nature of those transactions and the links of a legal, economic and/or personal nature between the operators involved in the scheme for reduction of the tax burden.”

It would come as no surprise that the Court has had occasion to pronounce on issues of this kind on a number of occasions since then. We were shown several relevant decisions. The wording used has a good deal in common, as is normal in judgments of the Court, but there are differences.

In Part Service Srl Case C-425/06 [2008] STC 3132, a reference from Italy, the taxpayer company and a leasing company in the same group participated in leasing arrangements involving cars, under which the leasing company concluded an agreement with the customer for the use of the car with an option to purchase in consideration of lease payments, the setting up of a surety for the costs of the vehicle not covered by the lease payments, and an unlimited security. The taxpayer company entered into a separate agreement with the customer by which it insured the car and guaranteed fulfilment of the customer’s obligations to the leasing company, in return for a payment to it in advance which reduced the amount of the lease payments. The leasing company charged VAT under Italian law, but the payments by the customer to the taxpayer, and those made at the direction of the customer by the taxpayer to the leasing company, were paid without any charge to VAT, as being within the exemption for financial services supplies. The Italian tax authorities contended that there were not two separate agreements but that both were linked, and VAT should be accounted for and paid accordingly. The Italian court’s first question to the European Court was whether the concept of abuse of rights required that the transactions in question had to be carried out solely for a tax advantage. The case was assigned to Advocate General Poiares Maduro, but the Court decided to proceed to judgment without requiring an opinion from him. In paragraph 42 of the judgment, the Court quoted from paragraphs 74 and 75 of Halifax. At paragraphs 44 and 45 it said this:

“44.

Therefore, when it [i.e. the Court] stated, in paragraph 82 of that judgment, that in any event, the transactions at issue had the sole purpose of obtaining a tax advantage, it was not establishing that circumstance as a condition for the existence of an abusive practice, but simply pointing out that, in the matter before the referring court in that case, the minimum threshold for classifying a practice as abusive had been passed.

45.

The reply to the first question therefore is that the Sixth Directive must be interpreted as meaning that there can be a finding of an abusive practice when the accrual of a tax advantage constitutes the principal aim of the transaction or transactions at issue.”

The Court then went on to address the second question, referring both to the freedom of traders to choose how they structured their business, legitimately taking into account the incidence of tax on different arrangements, and also to the question whether, when a transaction involves the supply of a number of services, it is to be considered as a single transaction or rather as several individual and independent transactions. The Court referred to a number of features of the particular transactions at issue in that case. I do not need to refer to all of those, but I will quote a sentence from paragraph 52:

“In particular, a service must be regarded as ancillary to a principal service if it does not constitute for customers an aim in itself, but a means of better enjoying the principal service supplied.”

The Court referred to the first criterion laid down in Halifax, by reference to the objectives of the relevant legislation, and then turned to the second. As to this it said, at paragraph 62:

“As regards the second criterion, the national court, in the assessment which it must carry out, may take account of the purely artificial nature of the transactions and the links of a legal, economic and/or personal nature between the operators involved (Halifax and Others, paragraph 81), those aspects being such as to demonstrate that the accrual of a tax advantage constitutes the principal aim pursued, notwithstanding the possible existence, in addition, of economic objectives arising from, for example, marketing, organisation or guarantee considerations.”

On 22 December 2010 the Court gave two further judgments on which reliance was placed before us. In Weald Leasing Case C-103/09 [2011] STC 596, the arrangements again involved a group (the Churchill Group) whose main supplies (insurance) were exempt for VAT purposes. The taxpayer’s business consisted of purchasing equipment and leasing it; it was not an exempt trader. When members of the group which were exempt traders required new equipment, the taxpayer bought it, and leased it to a third party, which then leased it to the relevant exempt member of the group. The exempt trader had to bear irrecoverable input tax, but this was spread over time by being charged on the rental amounts, not on the purchase price. HMRC challenged these arrangements, contending that the use of the leasing and sub-leasing arrangements was an abuse. It also argued that the level of the rent charged was artificially low, and that the use of the third party intermediate lessor, which was not part of the Churchill group, was intended to prevent the use of statutory powers to direct that the supply be taken to be made at its open market value, that power being available only where the person making the supply and the person to whom it was made are “connected”: see VATA Schedule 6 paragraph 1(1). The third party (SUAS) was not within the statutory definition of “connected”, but it was nevertheless not entirely at arms’ length, since it was owned by a VAT consultant to the Churchill group, and his wife, and it had no trading activity other than to take leases of equipment from the taxpayer and to grant sub-leases in turn to exempt members of the group.

The VAT and Duties Tribunal held that the essential aim of the transactions was to secure a tax advantage, but that that advantage was not contrary to the provisions of the Sixth Directive, in that it was properly open to a trader to decide whether to acquire goods by purchase or by way of lease, the latter having what for some traders would be the advantage of spreading, and so deferring, the incidence of VAT. The Tribunal noted that an abuse might arise, not from the leases themselves, but from the level of the rentals under the leases and from the arrangements made to avoid being subject to a direction that the supply be treated as having been made at market value. However, the case had not been argued on this basis, as was observed at paragraph 144 of the Tribunal’s decision (see (2007) VAT Decision 20003):

“It might have been argued by Customs that the introduction of Suas into the transactions thus preventing Customs from making directions under Schedule 6, paragraph 1 resulted in the accrual of tax advantages contrary to the Sixth Directive and to the domestic legislation including Schedule 6, paragraph 1. This would have involved widening the ambit of the second answer in Halifax somewhat so as to encompass the purposes of national legislation enacted pursuant to a derogation under Article 27 of the Sixth Directive. There is no doubt that Schedule 6, paragraph 1 was enacted under such a derogation, see RBS Leasing & Services (No.1) Ltd and others v Commissioners of Customs and Excise [2000] V&DR 33. Since it was not argued by Mrs Hall that the abuse consisted of low rentals coupled with the introduction of Suas to counter Schedule 6, paragraph 1, not only were there no submissions as to this but no evidence was adduced as to what open market rentals would have been.”

Lindsay J, giving judgment on the appeal by HMRC, endorsed the view that this might have been a promising line of attack, had the case been argued on that basis: see [2008] EWHC 30 (Ch) and [2008] STC 1601, paragraph 39.

The case was referred to the European Court of Justice by the Court of Appeal, on an appeal against Lindsay J’s order. In its judgment the European Court referred to previous decisions including both Halifax and Part Service, summarising the two conditions in familiar terms, of which paragraph 30 is the more relevant for present purposes:

“Second, it must also be apparent from a number of objective factors that the essential aim of the transactions concerned is to obtain a tax advantage. The prohibition of abuse is not relevant where the economic activity carried out may have some explanation other than the mere attainment of tax advantages (see Halifax and Others, paragraph 75, and Part Service, paragraph 42).”

The Court observed that the use of leasing transactions as such was not contrary to the purposes of the Sixth Directive. However it went on to say this:

“39.

That being so, the national court will have to determine, first, whether the contractual terms of the leasing transactions at issue in the main proceedings are contrary to the Sixth Directive and of the national legislation transposing it. That would particularly be the case if the rentals were set at levels which were unusually low or did not reflect any economic reality.

40.

Secondly, the national court will also have to determine whether the involvement of an intermediate third party company, in this case Suas, in those transactions is such as to preclude the application of those provisions.

41.

In that regard, the national court will have to ascertain whether, as is apparent from certain documents in the case file and as was stated at the hearing, the involvement of Suas in those transactions precluded the Commissioners from applying Paragraph 1 in Schedule 6 to the VAT Act 1994 so far as the transactions were concerned.

42.

In that context, Weald Leasing’s argument that the principle of prohibiting abusive practices does not apply to breach of Paragraph 1 in Schedule 6 to the VAT Act 1994 because that provision is purely a question of national law cannot be accepted, because that provision was adopted on the basis of Article 27 of the Sixth Directive and forms part of the national legislation implementing that directive.

43.

Moreover, the fact that an undertaking which resorts to leasing transactions such as those at issue in the main proceedings does not engage in leasing transactions in the context of its normal commercial operations does not affect the foregoing considerations.

44.

A finding that there was an abusive practice is inferred, not from the nature of the commercial operations usually engaged in by the party which made the transactions in question, but from the object and effects of those transactions, as well as their purpose.

45.

In those circumstances, the answer to the first and second questions is that the tax advantage accruing from an undertaking’s recourse to asset leasing transactions, such as those at issue in the main proceedings, instead of the outright purchase of those assets, does not constitute a tax advantage the grant of which would be contrary to the purpose of the relevant provisions of the Sixth Directive and of the national legislation transposing it, provided that the contractual terms of those transactions, particularly those concerned with setting the level of rentals, correspond to arm’s length terms and that the involvement of an intermediate third party company in those transactions is not such as to preclude the application of those provisions, a matter which it is for the national court to determine. The fact that the undertaking does not engage in leasing transactions in the context of its normal commercial operations is irrelevant in that regard.”

Advocate General Mazák had said, at paragraph 21 of his opinion, that he did not consider that the setting up and use of a captive subsidiary to grant leases of the equipment, in order to obtain the advantage of deferral of VAT, was in itself abusive because the same advantage could be gained by entering into an arm’s length leasing agreement with a third party. That is consistent with what the Court said in its judgment just quoted.

Thus, the Court appears to have endorsed both the Tribunal’s view, accepted by the High Court on appeal, that the use of leasing arrangements to spread and defer the incidence of irrecoverable VAT on the exempt trader was not in itself abusive, even though it was adopted in order to secure a tax advantage, and also the observation that, if the case had been differently argued, the use of the interposed third party and the low level of rent charged might have provided the basis for a finding of abuse. This shows that it is necessary to take particular care in identifying the aspects of the arrangements in question in any given case that are said to amount to an abuse.

The other case decided by the ECJ on the same day was RBS Deutschland Holdings Case C-277/09 [2011] STC 345. The taxpayer, a German subsidiary of a UK bank, carried on a banking and leasing business in Germany. It did not have a place of establishment in the UK. It was registered in the UK for VAT as a non-established taxable person. It bought cars in the UK from an unconnected seller, and leased the cars back to the same group, with a put option requiring the cars to be bought back later. It paid VAT on the purchases but did not charge VAT on the leases. Under UK legislation the leases were treated as supplies made in Germany, and not subject to VAT in the UK, whereas under German law, having regard to the particular terms of the leases, they were treated as supplies made in the UK and therefore not subject to VAT in Germany. On the exercise of the put option, UK VAT was charged on the sale, but the taxpayer sought to deduct the input tax incurred on the purchase.

At paragraph 49 of its judgment, the Court reiterated without modification the propositions set out in Halifax at paragraphs 74 and 75. Of course the taxpayer in that case was, by the relevant transactions, carrying out what were for it normal commercial business operations, of banking and leasing, though Weald Leasing shows that this is not essential. It took advantage of an anomalous situation as between UK and German law, possible only in a situation where VAT legislation had not been fully harmonised. The Court concluded, at paragraph 55:

“Consequently, the answer to the second, third and fourth questions is that the principle of prohibiting abusive practices does not preclude the right to deduct VAT, recognised in Article 17(3)(a) of the directive, in circumstances such as those of the main proceedings, in which a company established in one Member State elects to have its subsidiary, established in another Member State, carry out transactions for the leasing of goods to a third company established in the first Member State, in order to avoid a situation in which VAT is payable on the sums paid as consideration for those transactions, the transactions having been categorised in the first Member State as supplies of rental services carried out in the second Member State, and in that second Member State as supplies of goods carried out in the first Member State.”

I must also mention the WHA case, in which the Court of Appeal considered the abuse of right principle in its decision in 2007 ([2007] EWCA Civ 728), having in earlier judgments upheld (partly for different reasons) my own decision allowing an appeal by the taxpayers against a ruling of the VAT and Duties Tribunal on issues of fact and construction. The Court of Appeal held that the arrangements undertaken in that case were abusive. WHA appealed to the Supreme Court of the United Kingdom. The judgment of the Supreme Court was delivered on 1 May 2013, the last day of the hearing of Pendragon’s appeal before us. The Supreme Court dismissed WHA’s appeal, but did so on the ground that I and the Court of Appeal had both been wrong to disagree with the decision of the VAT and Duties Tribunal as to the construction and effect of the arrangements involved. Accordingly it said nothing about the principle of abuse of right. It follows that the observations of the Court of Appeal on abuse of right are not binding as a matter of ratio, but they remain persuasive: see Helena Partnerships v HMRC [2012] EWCA Civ 569 at paragraph 55.

On the other hand, the Court of Appeal in WHA had the benefit only of the Halifax case, not of any of the later European Court decisions to which I have referred. It is therefore of limited assistance.

One of the critical issues is how it is to be determined whether the essential aim, or the principal aim, of the transactions concerned is to obtain an illegitimate tax advantage. Mr Cordara submitted that this is an objective question, and he pointed to the phrase “the aim of the transactions”, as distinct from the aim of the parties to those transactions. The objective approach is made clear in the opinion of the Advocate General in Halifax, in passages which I have quoted above. In addition Mr Cordara showed us an even clearer passage from the same opinion at paragraph 70:

“In this regard, what is referred to in Emsland as the subjective element of the abuse does not affect the interpretative nature of the Community law notion of abuse.  In Emsland the Court linked that subjective element to the finding that the situation giving rise to the application of a certain Community rule was purely artificial. In my view, that finding of artificiality should not be based on an assessment of the subjective intentions of those claiming the Community right. The artificial nature of certain events or transactions must certainly be determined on the basis of a set of objective circumstances verified in each individual case. This is, furthermore, in line with the Court’s reference, again in Emsland, to the ‘sole purpose’ of an activity or behaviour as a central element supporting the conclusion that there has been an abuse of Community law.  When the Court takes the view that an abuse exists whenever the activity at issue cannot possibly have any other purpose or justification than to trigger the application of Community law provisions in a manner contrary to their purpose, that is tantamount, in my view, to adopting an objective criterion for the assessment of the abuse. It is true that those objective elements will reveal that the person or persons engaged in that activity had, most likely, the intention of abusing Community law. But it is not that intention that is decisive for the assessment of the abuse. It is instead the activity itself, objectively considered. In that regard, suffice it to imagine, by way of example, a case where A confines himself without further reflection to following the advice of B and to carrying out an activity for which there is no explanation other than securing a tax advantage for A. The fact that A did not have any subjective intention of abusing Community law will certainly not be material for the assessment of the abuse. What matters is not the actual state of mind of A, but the fact that the activity, objectively speaking, has no other explanation but to secure a tax advantage.”

Mr Cordara also showed us a passage from the Opinion of Advocate General Tizzano in the Mirror Group case, Case C-409/98, quoted with approval by Jonathan Parker LJ (see paragraph 59 of his judgment, admittedly in a context other than abuse of right) in Tesco plc v. Commissioners of Customs & Excise [2003] EWCA Civ 1367, as follows:

“27.

In order to identify the key features of a contract ... we must go beyond an abstract or purely formal analysis. It is necessary to find the contract’s economic purpose, that is to say, the precise way in which performance satisfies the interests of the parties. In other words, we must identify the element which the legal traditions of various European countries term the cause of the contract and understand as the economic purpose, calculated to realise the parties’ respective interests, lying at the heart of the contract. In the case of a lease, as noted above, this consists in the transfer by one party to another of an exclusive right to enjoy immovable property for an agreed period.

28.

It goes without saying that this purpose is the same for all the parties to the contract and thus determines its content. On the other hand, it has no connection with the subjective reasons which have led each of the parties to enter into the contract, and which obviously are not evident from its terms. I have drawn attention to this point because, in my view, failure to distinguish between the cause of a contract and the motivation of the parties has been the source of misunderstandings, even in the cases under consideration here, and has complicated the task of categorising the contracts.”

For his part Mr Pleming did not take issue with the need for an objective assessment. He argued that the First-Tier Tribunal had misdirected itself in other respects and that the Upper Tribunal’s assessment, on the basis of objective factors, was properly open to it. I therefore do not need to go further into the question as to the correct principle. It seems to me that, as Mr Cordara submitted, it is indeed necessary to assess the aim of the particular series of transactions objectively, and not by reference to the actual intentions of the parties.

Nor, despite some variation in the language used by the European Court, does it seem to me that there is a significant difference between the test as formulated in different judgments of that Court. In Halifax there was reference both to “sole purpose” (Judgment paragraph 82, and having “no other explanation”: Advocate General at paragraphs 70, 86 and 91) and to “essential aim”: Judgment paragraphs 75 and 86. In Part Service it was said that it could be sufficient if obtaining the tax advantage constitutes the principal aim of the transactions at issue: paragraph 45. The reference to “sole purpose” in paragraph 82 of Halifax was explained as referring to the fact that the test was already satisfied in that case. As it seems to me, Part Service did not create a substantial change from the principle set out in Halifax, not least because I find it difficult to see a significant difference between “principal” and “essential” aim: in either case there could be some other incidental or ancillary aim. This does not fit exactly with some of the reasoning of the Advocate General in Halifax. However, the fact that no Opinion was required of him before the Court proceeded to judgment in Part Service, and the fact that the Court did not express itself as intending to modify the principle set out in Halifax, seems to me to make it clear that all that was achieved by Part Service, in this respect, was to clarify a possible tension between the use of the phrase “sole purpose” in paragraph 82 and of the phrase “essential aim” in paragraphs 75 and 86 in Halifax. Nor did the Court in Weald Leasing consider that Part Service had made any significant difference to the test: see paragraph 30 of that judgment, quoted at paragraph [57] above. In particular, I do not see that there is any basis for suggesting that the formula used in Part Service involves any departure from the requirement of an objective approach laid down in Halifax.

The judgment in Part Service does expressly contemplate that there may be ancillary, collateral or subsidiary commercial benefits which do not qualify or override the essential aim as being to secure the tax advantage, but paragraph 52 of that judgment (see paragraph [52] above) gives a very clear indication of what is meant by ancillary or collateral benefits in this context. In such a case it would be possible to assess the benefits as being collateral or ancillary (or not) by an objective consideration of the nature of the benefits themselves as compared with the commercial purpose of the transaction as a whole (in that case, as seen by the consumer of the services supplied). It does not seem to me that, viewed in that way, the reference to ancillary or collateral economic objectives as being permissible, but nevertheless compatible with a finding of abuse, introduced any significant qualification to the principle as set out in Halifax.

Mr Cordara criticised the Upper Tribunal, arguing that it had gone beyond what is legitimate for an appellate body. The most familiar observations about this in respect of tax appeals are to be found in Lord Radcliffe’s speech in Edwards v Bairstow [1956] AC 14. There the issue was whether a joint venture undertaken by two individuals by way of the acquisition of assets with a view to a quick sale at a profit was such that the profit eventually made was taxable on the basis of it having been an adventure in the nature of trade. The General Commissioners held that it was not such an adventure. The Revenue’s appeal to the High Court and the Court of Appeal were dismissed on the basis that the decision was one of fact. The House of Lords disagreed and allowed the appeal. Lord Radcliffe observed that, since the law did not supply a definition of what constitutes “trade”, it was for the courts to interpret it, as a matter of law, and to apply it to the facts as agreed or found. He said at page 33:

“In effect it lays down the limits within which it would be permissible to say that a “trade” as interpreted by section 237 of the Act does or does not exist. But the field so marked out is a wide one and there are many combinations of circumstances in which it could not be said to be wrong to arrive at a conclusion one way or the other. If the facts of any particular case are fairly capable of being so described, it seems to me that it necessarily follows that the determination of the Commissioners, Special or General, to the effect that a trade does or does not exist is not “erroneous in point of law”; and, if a determination cannot be shown to be erroneous in point of law, the statute does not admit of its being upset by the court on appeal. I except the occasions when the commissioners, although dealing with a set of facts which would warrant a decision either way, show by some reason they give or statement they make in the body of the case that they have misunderstood the law in some relevant particular.”

Later, at pages 35 to 36 he said this:

“I think that the true position of the court in all these cases can be shortly stated. If a party to a hearing before commissioners expresses dissatisfaction with their determination as being erroneous in point of law, it is for them to state a case and in the body of it to set out the facts that they have found as well as their determination. I do not think that inferences drawn from other facts are incapable of being themselves findings of fact, although there is value in the distinction between primary facts and inferences drawn from them. When the case comes before the court it is its duty to examine the determination having regard to its knowledge of the relevant law. If the case contains anything ex facie which is bad law and which bears upon the determination, it is, obviously, erroneous in point of law. But, without any such misconception appearing ex facie, it may be that the facts found are such that no person acting judicially and properly instructed as to the relevant law could have come to the determination under appeal. In those circumstances, too, the court must intervene. It has no option but to assume that there has been some misconception of the law and that, this has been responsible for the determination. So there, too, there has been error in point of law. I do not think that it much matters whether this state of affairs is described as one in which there is no evidence to support the determination or as one in which the evidence is inconsistent with and contradictory of the determination, or as one in which the true and only reasonable conclusion contradicts the determination. Rightly understood, each phrase propounds the same test. For my part, I prefer the last of the three, since I think that it is rather misleading to speak of there being no evidence to support a conclusion when in cases such as these many of the facts are likely to be neutral in themselves, and only to take their colour from the combination of circumstances in which they are found to occur.”

Applying that approach he held that only one true and reasonable conclusion could be drawn from the facts as found, namely that it was an adventure in the nature of trade.

The issues arising under the abuse of rights principle are not so clear cut as those that had to be considered in that case, but it might perhaps be said that the question for the First-Tier Tribunal (corresponding to the former General or Special Commissioners) was whether it was, or was not, the essential aim of the transactions to obtain the particular tax advantage. The First-Tier Tribunal in the present case quoted the pertinent observations of the European Court in paragraph 81 of the judgment in Halifax, which I have quoted at paragraph [49] above. The task there identified is primarily that of the first instance court or tribunal, here the First-Tier Tribunal. It is for that body to determine, with the benefit of any relevant and necessary fact-finding, the real substance and significance of the transactions concerned.

HMRC’s appeal to the Upper Tribunal was not advanced on the basis of Edwards v Bairstow. That case was not cited in HMRC’s skeleton argument to the Upper Tribunal nor by the Upper Tribunal in their judgment, though it was mentioned in the course of submissions to the Upper Tribunal. Instead the contention was that the First-Tier Tribunal had erred in law in ways apparent from their decision.

There are many other familiar cases about the proper role of an appellate court in relation to an appeal of this kind. In the context of VAT they were collected and applied in the Pringles case: Procter & Gamble v HMRC [2009] EWCA Civ 407. Of course the issue in that case was rather different: were P&G’s Pringle products “‘similar to’ potato crisps and ‘made from’ the potato”? Moreover the determination of that issue was not affected by any disputed oral evidence. It was an evaluative task on the evidence which was entrusted to the VAT and Duties Tribunal, predecessor of the First-Tier Tribunal in the present case, subject to an appeal on a point of law from there to the High Court as now to the Upper Tribunal. In his judgment in that case Jacob LJ said this, at paragraphs 7 and 8:

“7.

Although Mr Vajda QC for HMRC opened the appeal by attacking the judgment of Warren J rather than concentrating upon the decision of the tribunal (which of course he contended was correct) in the end counsel were agreed that what really mattered was whether the decision of the tribunal was wrong in law. For it is the tribunal which is the primary fact finder. It is also the primary maker of a value judgment based on those primary facts. Unless it has made a legal error in that in so doing (e.g. reached a perverse finding or failed to make a relevant finding or has misconstrued the statutory test) it is not for an appeal court to interfere. This has been said in other contexts e.g. Osmani v London Borough of Camden [2004] EWCA Civ 1706 at [34], (‘… the main focus of attention on a second appeal such as this should be on the decision of the Council rather than that of the County Court Judge on appeal’ per Auld LJ and Maloba v Waltham Forest London Borough Council [2007] EWCA Civ 1281 at [19], [2008] 1 WLR 2079 per Toulson LJ). The same applies for the same reasons to appeals from this tribunal.

8.

The effect of this principle in this case is that although P&G is the respondent to this second appeal, in reality it is necessary for P&G to show that the tribunal erred in law. The judge held that it had. Of course his reasons why need to be examined, but in the end the focus is on the tribunal decision.”

Likewise Mummery LJ said this at paragraphs 73 and 74:

“73.

The tribunal’s decision in favour of Her Majesty’s Revenue and Customs (‘HMRC’) was not an absolute answer to a pure question of fact or to a pure question of law. It was a judgment of mixed fact and law on the classification of Regular Pringles for value added tax (‘VAT’) purposes. ‘Similar to’ and ‘made from’ are loose-textured concepts for the classification of the goods. They are not qualified by words such as ‘wholly’ or ‘substantially’ or ‘partly’ which have crept into the legal arguments. Those words are not in the legislation itself. The tribunal’s conclusions were on matters of fact and degree linked to comparisons with other goods and related to the composition of the goods themselves. Some aspects of the similarity of Regular Pringles to potato crisps are close to the centre, others are on the fringes. This exercise in judgment is pre-eminently for the specialist tribunal entrusted by Parliament with the task of fact finding and with using its expertise to make the first level decision, subject only to appeal on points of law.

74.

For such an appeal to succeed it must be established that the tribunal’s decision was wrong as a matter of law. In the absence of an untenable interpretation of the legislation or a plain misapplication of the law to the facts, the tribunal’s decision that Regular Pringles are ‘similar to’ potato crisps and are ‘made from’ the potato ought not to be disturbed on appeal. I cannot emphasise too strongly that the issue on an appeal from the tribunal is not whether the appellate body agrees with its conclusions. It is this: as a matter of law, was the tribunal entitled to reach its conclusions? It is a misconception of the very nature of an appeal on a point of law to treat it, as too many appellants tend to do, as just another hearing of the self-same issue that was decided by the tribunal.”

Accordingly, the first issue for us is whether the First-Tier Tribunal erred in law in reaching the conclusion that the essential aim of the transactions was not to achieve the tax advantage. Was that a conclusion to which it was entitled to come? The Upper Tribunal held that it had so erred. Of course we need to look at the basis for the Upper Tribunal’s decision but in the end our decision is as to whether the First-Tier Tribunal went wrong, not (directly) whether the Upper Tribunal went wrong. It also follows that it may not be sufficient to show (if it be the case) that the Upper Tribunal was wrong, because the First-Tier Tribunal might also have been wrong, but for other reasons. In a case where the issue is one of law on undisputed facts there is no real difference between looking to see if the lower tribunal went wrong as compared to seeing whether the intermediate tribunal erred. But if the lower tribunal carried out a fact finding task or an evaluative task, and if its conclusions in that respect are challenged, then it is important to be clear as to the basis of the challenge and to focus on whether the lower tribunal did err.

The oral submissions before us focussed principally on the second requirement under Halifax, namely whether the essential aim of the transactions was to obtain the illegitimate tax advantage. I will limit my examination of the case to that aspect as well. However, as Mr Pleming submitted, it is important to bear in mind that the second requirement arises only if the first is satisfied. If it is, then it has been shown that, although the provisions of the national legislation have been satisfied so as to secure the relevant tax advantage (subject to the abuse principle), nevertheless the grant of that advantage would be contrary to the purpose of the relevant provisions. For the relevant undertaking to have the benefit of the tax advantage in question is, therefore, contrary to the policy of the legislation. The letter of the law has been satisfied, but so as to produce a result contrary to its spirit. In considering the application of the second requirement, that conflict with the legislative policy is to be borne in mind.

It is common ground between the parties that the First-Tier Tribunal asked itself the right question, as regards the principle of abuse of right generally. It was agreed that the Scheme fell squarely within the terms of the UK legislation so that, unless the doctrine of abuse applied, the appellants were entitled to the tax treatment for which they contended. It was common ground that “other than for VAT purposes financing was obtained”, that the cars in question were properly regarded as used cars, and that there was no suggestion of any wrong-doing: see paragraph 13. The Tribunal then set out the legislation and went on to consider the evidence and to make factual findings. At paragraph 110 (Footnote: ¹) the Tribunal started to deal with the law as regards the doctrine of abuse, with relevant citations, in particular from Part Service at paragraph 42, which was then the latest European court decision on abuse in the context of VAT. The review of the law covers paragraphs 110 to 166. Paragraphs 167 to 170 dealt with the first of the Halifax questions, leading to a decision that the transactions were not contrary to the purpose of the Sixth Directive. Then at paragraphs 171 to 185 the Tribunal addressed the second Halifax question, and held that the aim of the Scheme was to obtain finance, so that it was not its essential aim to obtain a tax advantage. At paragraph 184 the conclusion is expressed as follows:

“From the evidence which we have seen and heard we find as a fact considering matters objectively and not subjectively that in the particular circumstances of this case the essential aim of the transactions was to obtain finance and not to obtain a tax advantage. The real substance and significance of the transactions was the obtaining of finance. This puts it in Mr Pleming QC’s non-abusive box and we so find.”

Mr Pleming criticised the First-Tier Tribunal for having addressed the essential aim of the Scheme early in its judgment, by which time it had done no more than refer in the briefest of terms (though accurately so far as they went) to the doctrine of abuse (at paragraph 8). It is a fair comment that the Tribunal anticipated its final conclusion at paragraphs 51 to 53 when it said this:

“51.

We have carefully considered the position here in the light of these objective factors and all the circumstances of the case from an objective perspective. We consider that the obtaining of finance in all the circumstances of the case was the predominant, principal or a central aim of the transactions and we so find as a primary fact on the basis of objective factors.

52.

This was clearly the case for the first tranche and we consider it also to be the case, though less certainly, for the second tranche. This is not to suggest that we are wavering as to the finding concerning the second tranche. We are not because the shortening was because of Budget uncertainty and not because finance was not needed. Again we find this on the basis of objective factors.

53.

We find, having considered all the evidence and circumstances, that it is not “… apparent from a number of objective factors that the essential aim of the transactions concerned is to obtain a tax advantage.” The essential aim was finance.”

I will come back to the decision of the First-Tier Tribunal in more detail shortly.

The Upper Tribunal disagreed. The most important paragraph of its decision is paragraph 114, as follows:

“The FTT decided otherwise. We are not legally able to reverse the decision of the FTT and give effect to our own conclusion unless the FTT committed an error of law in reaching its conclusion. We consider that the decision of the FTT on this question was wrong. We also consider that the issue as to the essential aim of the transaction was not an issue as to the primary facts but one which was, at least in part, an issue of law. Accordingly, we consider the FTT’s decision was wrong in law. If it were necessary for us to do so, we would hold that the FTT’s decision was plainly wrong and was only consistent with its having committed an error of law in its approach. We consider that we can see how and why the FTT reached a conclusion which we have held was plainly wrong. The FTT do not refer to matters which we regard as of central importance, namely, the scale of the tax advantage and the fact that the finance was relatively expensive at a time when Pendragon had unused committed facilities. It is conspicuous that the FTT did not provide in its decision a comprehensive description of the arrangements, nor a detailed analysis of them, and in consequence focussed on the fact that finance was obtained, rather than on the fact that the obtaining of finance was subordinate to the essential aim of obtaining a tax advantage. Further, the FTT must have under-estimated the difficulty of leaving out of account much of what Mr Forsyth had said and of confining themselves to objective factors which would have occurred to a reasonable observer. The fact that the FTT said that it was determining the essential aim of the transaction “as a primary fact” does not prevent the Upper Tribunal reversing that finding but rather serves to emphasise that the FTT did not understand the task before it as involving an objective assessment of essential aim. The background facts may have been primary facts but an assessment of aim by reference to objective factors is not a primary fact. A finding as to Mr Forsyth’s state of mind may have been a primary fact but that was not a directly relevant matter. The FTT appears also not to have appreciated that the essential aim of the lender might have been different from the essential aim of the transaction. No doubt, the lender did intend to lend money and obtain security by entering into documents which were effective to provide such security but we do not regard that as the essential aim of the transaction. Finally, it is likely that the FTT lost sight of the fact that it was not enough for Pendragon to show that the transactions produced real commercial consequences where those were collateral matters which did not detract from the essential aim of obtaining a tax advantage; the reference in paragraph 173 of its decision to “some autonomous justification” suggests that this is so. It is not necessary for us to determine whether these explanations as to how and why FTT reached a conclusion which was plainly wrong themselves disclose further separate errors of law, although we incline to the view that at least some of them do. In our judgment, the ultimate decision of the FTT was plainly wrong and therefore the FTT committed an error of law in reaching that conclusion. Even if the decision of the FTT involved a multi-factorial valuation judgment, we are satisfied that its decision ought to be reversed.”

Mr Pleming submitted that the Upper Tribunal was correct in identifying errors of law in the First-Tier Tribunal’s decision, and that accordingly the Upper Tribunal was entitled to come to its own decision, on the basis of a correct self-direction as to the law. He identified several errors in the First-Tier Tribunal’s decision. One was its treatment of documents emanating from KPMG, who devised the Scheme. Secondly, he argued that the correct approach necessarily requires an assessment by the national court of the relative significance of the tax advantage as compared with other factors. He said that the First-Tier Tribunal’s approach treated the European principle as if it were a “sole purpose” test, not a “principal purpose” which allowed for the possibility of other aims. He identified as artificial three particular elements: the use of the off-shore bank (as opposed to its London parent), the use of the CLCs to create leases immediately before assigning the leases and the underlying assets to SGJ, and the use of the hybrid lease and in particular clause 8(c), the provision about deferring the passing of title at the end of the period of the agreement.

In his skeleton argument for the hearing in the Upper Tribunal, Mr Pleming had made points which can be summarised as follows. First, the First-Tier Tribunal had been wrong to characterise the evidence of KPMG’s advice to Pendragon as subjective and (by implication) therefore to ignore it; he argued that it was highly relevant as to the purpose of the transactions at issue, describing the reason for and therefore the purpose of the transactions. Secondly, the First-Tier Tribunal had erred in treating the question as binary: i.e. either it was a finance transaction or it was an abuse. There could be a genuine underlying commercial purpose to the transactions but they might still have an illegitimate tax advantage as their essential aim. It was therefore necessary to address the elements for which there was no commercial purpose, as those aspects provide the best objective evidence of the aim of entering into the Scheme. Collateral beneficial results were to be ignored if the Scheme was conceived and implemented purely for tax purposes. Thirdly, he identified a series of objective factors which he said showed that the essential aim was the tax advantage. I do not need to mention all of these, for some did not feature in the decision of the Upper Tribunal or in the argument before us. Those that did so feature included the fact of the Scheme having been created by KPMG, and the three points mentioned in the last paragraph: the use of the CLCs to create the hybrid leases, the use of the hybrid lease and in particular clause 8(c), and the use of the offshore bank rather than its London parent. Another factor was the marketing and pricing of the Scheme by KPMG, including the fact that KPMG was paid on a contingency basis by reference to the VAT saving, if achieved.

That last point also bears on a point relied on by the Upper Tribunal, namely that the finance obtained by Pendragon through the Scheme was relatively expensive (so it said at paragraph 114). It had set out the terms as between Pendragon and SGJ at paragraph 98. It also relied on the proposition that, at the relevant time, Pendragon had adequate other borrowing facilities so that, by implication, it did not really need to undertake this additional and expensive short term funding for the sake of arranging its credit facilities.

Furthermore the Upper Tribunal held that the First-Tier Tribunal had erred (as HMRC had argued that it did) by not taking into account the scale of the tax advantage as compared with the nature of the other commercial consequences of the Scheme.

There was argument in the Upper Tribunal about three separate aspects of the evidence: that of Mr Forsyth, finance director of the Pendragon Group at the time, who gave a witness statement and also gave oral evidence in chief and in cross-examination, that of KPMG’s part in the process, which was apparent from documents disclosed by Pendragon, and that of the expert witnesses called by Pendragon. In the end not much turned on the evidence of the expert witnesses. Much more attention was given to Mr Forsyth and to KPMG.

The issue as regards Mr Forsyth’s evidence was as to what part of it was admissible as relating to matters of objective relevance and what part should be ignored being inadmissible subjective material. It was not, and could not have been, suggested that the First-Tier Tribunal was unaware of the importance of the distinction between the two. The Upper Tribunal mentioned the issue as to Mr Forsyth’s evidence at paragraph 93, where it accepted that expert evidence could assist as to how normal or how commercial the Scheme or a particular feature of the Scheme was, and that the particular circumstances of Pendragon and its attitude to obtaining finance could be relevant when considering the essential aim of the transaction, but that evidence from Mr Forsyth which attempted to answer the question why Pendragon did enter into the transactions would be on the wrong side of the line. His report to the Board was mentioned at paragraph 97, as being relevant, at any rate in identifying the scale of the potential economic advantage (through the tax saving). His evidence more generally is covered at paragraphs 102 to 106. Some of his evidence concerned factual matters as to the background against which the transactions were entered into, which the Tribunal said was admissible. It said that much of his evidence, however, sought to explain his thinking, intentions and wishes in relation to the transactions, which was not admissible. At the end of paragraph 102 the Upper Tribunal said this:

“In our view, much of the evidence of Mr Forsyth was cast in a way which focussed on what he said were his motives and his reasons. Cast in that way, the evidence went directly to inadmissible subjective matters. On the other hand, we can see that if the evidence had been drafted in a different way, Mr Forsyth could have drawn on his own involvement in the transactions to identify the considerations which would have influenced reasonable commercial people, acting in the circumstances in which Pendragon actually found itself at the relevant time. That would have been material to the objective assessment which was required to determine the essential aim of the transaction.”

In turn, paragraph 103 is as follows:

“In the event, the FTT received all of Mr Forsyth’s evidence. It seems to us that this placed the FTT in a most difficult position. It had to guard against being directly influenced by the considerable amount of evidence given by Mr Forsyth as to subjective matters but, perhaps, it could try to distil from that evidence the different evidence which could have been given as to the type of considerations which, objectively assessed, would have been in the minds of reasonable commercial people in the actual circumstances. The decision of the FTT contains passages (notably at paragraphs 58 - 59, 172 – 174 and 180) where they state that they would leave out of account Mr Forsyth’s subjective views and look instead at what an outsider, knowing the relevant facts, would have objectively thought. However, the difficulty of the exercise, given the unhelpful way in which the evidence was presented, should not be under-estimated.”

Then the Upper Tribunal described the scope of Mr Forsyth’s evidence, with some detail at paragraphs 105 and 106 of what he said about the credit facilities available to Pendragon at the time.

It then dealt with KPMG in paragraphs 107 and 108, to which I will return, and then considered whether objectively assessed the essential aim of the transactions was to obtain a tax advantage (at paragraph 109). As to this, the Tribunal said that, at first sight, the scale of the tax advantage when set against the nature of the other commercial consequences was such that the tax advantage was overwhelmingly the real reason why Pendragon entered into the transactions. Then it sought to check that provisional conclusion against the evidence of Mr Forsyth and the expert witnesses. At paragraph 112 it mentioned Mr Forsyth’s evidence and referred to the comments earlier in the judgment on the question:

“in relation to the large part of it which was inadmissible, whether Mr Forsyth’s evidence can easily or safely be used in order to help with an assessment of the objective matters which would be considered by a reasonable observer.”

Then at paragraph 113 the Tribunal said that the evidence of the witnesses did not displace the overwhelming case for saying that the essential aim was the tax advantage. In the course of paragraph 114 which I have set out at paragraph [82] above, it made the only statement in the decision which gets near to being a holding that the First-Tier Tribunal erred in respect of its treatment of Mr Forsyth’s evidence. I repeat the single sentence here for convenience:

“Further, the FTT must have under-estimated the difficulty of leaving out of account much of what Mr Forsyth had said and of confining themselves to objective factors which would have occurred to a reasonable observer.”

That sentence is oddly formulated. It does not identify any specific misdirection on the part of the First-Tier Tribunal as regards Mr Forsyth’s evidence, or any deficient reasoning in the First-Tier Tribunal’s decision on the point. As I will show, there is a good reason for that: the Upper Tribunal could not have identified any such passage because the First-Tier Tribunal expressly directed itself correctly in these respects. Nor does even this sentence say in terms that the First-Tier Tribunal erred in law. Is the reader to understand that the First-Tier Tribunal not only under-estimated the difficulty of the task but also, because they did so, they failed to carry it out correctly? If so, in what respects did it err, and with what effect? Nor did the Upper Tribunal identify any specific part of his evidence which it considered that the First-Tier Tribunal had taken into account but was wrong to do so.

It rather looks as if the Upper Tribunal, observing that what the First-Tier Tribunal had said about Mr Forsyth was, on its face, legitimate but finding it inconsistent (so far as it went) with the Upper Tribunal’s own impression from the evidence, decided to brush aside the evidence of Mr Forsyth, and the First-Tier Tribunal’s reliance on it, on the basis that that Tribunal “must have been” wrong, although it could not say in what respect it was wrong, and did not even say in terms that it was wrong.

The First-Tier Tribunal dealt with Mr Forsyth’s evidence in a number of passages in its decision. At paragraph 29 it observed that his evidence was necessarily more concerned with the particular transactions and circumstances than that of the expert witnesses, but it mentioned the requirement of objectivity and said that it had considered his evidence in the light of that. It recorded that Mr Pleming had raised doubts, before the evidence was heard, about the relevance and admissibility of some of Mr Forsyth’s evidence, but he did not formally object to the admission of the evidence, so the Tribunal did not rule on it in advance.

At paragraphs 43 to 45 it referred to evidence from Mr Forsyth about the economic and financing issues affecting the retail motor industry. None of this could be objected to as merely subjective. Then from paragraphs 46 to 50 there is a passage headed “Mr Forsyth’s funding tenets at Pendragon”. Some of the material described here is plainly objective, such as Pendragon’s then recent expansion and its dependence on borrowing. In paragraph 46 the four tenets which he held as regards financing are set out. Insofar as these were personal to him, they were subjective. However, at paragraph 50 the Tribunal said that “although devised by Mr Forsyth we consider that they were also objective factors which we can properly take into account”. That is a fair comment in that there was nothing at all surprising or idiosyncratic about these four propositions; rather they were the sort of factors that any finance director in that industry would be likely to take into account. They were:

“1: Sufficient funding in place plus an additional safety margin to cover contingencies “headroom”;

2: Ensure that the funding is committed within a range of repayment dates;

3: Avoid dependency on a single source of financing; and

4: Always consider the overriding fact that running out of money means the liquidation or reconstruction of the company. This is a high price to be paid by the employees and other stakeholders of business.”

From this part of the decision it appears that the Pendragon Group had heavy borrowings, from a relatively small number of lenders. Mr Forsyth wanted to increase the number of lenders, as the Pendragon Group wished to expand still further. This was a subjective view but something that would have been objectively a part of sensible risk management in any event. Pendragon’s gearing and headroom were of concern to potential lenders. The Tribunal had been shown how close to various lending limits Pendragon was at various times, though it did not set these out in its decision for reasons of commercial confidentiality. The Tribunal said that it found as a primary fact that further sources of funding finance were very important objectively to the Pendragon Group (see paragraph 49).

Immediately after this passage in the decision come paragraphs 51 to 53, quoted at paragraph [80] above. Thereafter comes a section headed “Financing Proposals”, from paragraph 54 to paragraph 59. Some of this is objective material. Some of it records Mr Forsyth’s own views, including paragraph 57 which related in terms to his attitude to the Scheme. At paragraph 56 the introduction of the proposed scheme by KPMG to Mr Forsyth is mentioned. The paragraph includes this passage:

“This was at a time when Pendragon were renegotiating their borrowing facilities. The safety margin in the group’s headroom was much smaller than was desirable. Extra sources of funding were thus highly desirable. At that time two 45 day tranches of up to £20m funding with potential VAT savings were even more so from a financing perspective.”

The First-Tier Tribunal went on to say that “this”, which must mean what had been discussed in that section of the decision, was part of the setting for the transaction which has to be considered to give an objective view of the transactions and, in particular, that whether something is not commercial needs to be considered in the light of the objective commercial context in which transactions take place: paragraph 59. Then there is a passage headed “Finance and the Manufacturers Finance Arms” which is plainly objective material (paragraphs 60 to 62). Next it dealt with Mr Forsyth’s presentation of the proposal to the board of directors in October 2000. Mr Forsyth’s views of the Scheme are subjective, but the terms of the proposal are objective material. In paragraphs 64 to 82 the First-Tier Tribunal described the implementation of the Scheme and its details. In the course of that passage it said this at paragraph 72:

“There was also a £20,000,000 term facility agreement between them which was supported by the deed of assignment. This effectively gave an extra £20m facility to Pendragon for 45 days or so potentially in two tranches. This is a significant objective factor to be taken into account in considering the transactions. We consider that objectively this is an important indicator of the commerciality of the transactions and shows that they were not uncommercial operations and we so find. This is not to say that the funding was not done in a VAT efficient way but that is something that the ECJ says is permissible.”

The decision continues at paragraph 81, saying that, comparing the transactions to a normal sale and leaseback transaction or finance from a manufacturers’ finance arm, the essential elements of finance and security were the same, as was the commercial need to be able to record sales. At paragraph 82 the transactions were said not to be self-cancelling or evidently uncommercial.

The decision then dealt with KPMG and other aspects of the case. So far as Mr Forsyth is concerned, it next dealt with him at paragraph 171, returning to the question of the essential aim of the Scheme. Paragraphs 172 to 174 are as follows:

“172.

We have been careful in considering Mr Forsyth’s evidence to look to the commercial realities objectively as to the position of Pendragon and to consider the terms of the transactions.

173.

We consider (even ignoring Mr Forsyth’s evidence) that the obtaining of finance provided some autonomous basis which if tax considerations are left endows some economic justification in the circumstances of this case and we so find.

174.

This is so not because Mr Forsyth said that the company needed finance but from the position of the officious bystander it was clear that the company in this business would need considerable finance available to it. A company in Pendragon’s position as regards headroom and gearing in particular would clearly need finance and on the finest terms available.”

Again at paragraph 180 the Tribunal reiterated that it had not based its decision on Mr Forsyth’s personal views as regards the subjective reasons for the transactions.

I will come back to some of the things that the First-Tier Tribunal said about the evidence and the conclusions drawn from it, but from this review of what the First-Tier Tribunal said about Mr Forsyth, it seems to me that there was no basis on which the Upper Tribunal could hold that the First-Tier Tribunal had erred in law in its treatment of his evidence. That Tribunal accepted that his evidence was inadmissible insofar as it went to his subjective attitude to the transactions or to the situation in which the transactions were entered into, though it was admissible as regards objective factors relevant to the transactions. At paragraph 102, in the passage quoted above, the Upper Tribunal commented on the way in which his evidence was “cast” and suggested that it might have been “drafted in a different way”. I do not understand that comment. What matters, surely, is the relevant content of the evidence. Of course the way in which it is presented in a witness statement may be more, or may be less, helpful in terms of clarity, organisation or the like. But once the witness is giving evidence, or when he has given evidence, the court or tribunal has to get to grips with the subject and the evidence, decide what of it is helpful, with regard to issues of relevance or weight or both, and make whatever findings can properly be made on the basis of that evidence and all other relevant evidence before the court or tribunal. Moreover, given that Mr Pleming had not pressed the First-Tier Tribunal to rule on the issue of admissibility before Mr Forsyth gave evidence, it is somewhat unfair to criticise the Tribunal for allowing the evidence to be given, subject to later submissions as to its effect, as it is also to criticise it on the basis that it under-estimated the difficulty of distinguishing between what was admissible and what was not among his evidence.

For these reasons I do not regard the First-Tier Tribunal’s treatment of the evidence of Mr Forsyth as erroneous in law, as the Upper Tribunal seems to have implied, though without saying so in terms. This did not afford any reason for the Upper Tribunal to interfere with the First-Tier Tribunal’s conclusion.

The evidence before the First-Tier Tribunal included material passing between KPMG and the Pendragon Group. There was no witness from KPMG, but Mr Forsyth gave evidence in relation to some of this material. The First-Tier Tribunal and the Upper Tribunal took different views as to the admissibility and effect of this material. The Upper Tribunal did not identify an error on the part of the First-Tier Tribunal as regards the KPMG material in paragraph 114 of their judgment. It is reasonably clear from earlier paragraphs that they did regard the First-Tier Tribunal’s position in relation to this material as having been wrong, but not at all clear that they found that to be significant in any respect.

The First-Tier Tribunal mentioned KPMG’s involvement at paragraph 56 in the section (mentioned above) headed “Financing Proposals”. Then, starting at paragraph 83, it dealt with KPMG in detail in a section headed “KPMG’s involvement”. KPMG had devised the Scheme which it described as a VAT planning arrangement, using the heading (for the engagement letter) “VAT Advisory Services: Margin Scheme”. It offered the Scheme to Mr Forsyth, on behalf of the Pendragon Group, in March 1999. The Scheme was described in a briefing note which included passage about risk evaluation. That described the Scheme as “a very aggressive arrangement”, and as almost certain to be challenged. The fees payable to KPMG included an element contingent on the Scheme working and calculated by reference to the VAT saving achieved.

The First-Tier Tribunal commented at paragraph 94 that this material, while very interesting, was just as subjective as some of Mr Forsyth’s evidence. It “sets out the subjective view of what KPMG thought it was selling”. At paragraph 96 it said:

“KPMG were, we find, engaged to advise on a VAT arrangement. We do not consider that because Pendragon took VAT advice it necessarily follows that the essential aim of the arrangements was abusive.”

In summarising the submissions on behalf of HMRC at paragraph 104 the First-Tier Tribunal recorded several points which were based on the KPMG material, including that the scheme was designed as a VAT saving scheme or a tax avoidance scheme. In the part of the judgment devoted to the essential aim issue, from paragraph 171 to paragraph 184, the First-Tier Tribunal referred to the Pendragon Group having taken advice as to the VAT position; see paragraph 177:

“We find that Pendragon was fully aware of the VAT position. It would be surprising if they were not. They had a significant in-house tax team and had taken advice from leading accountants and practitioners on the matter. The fact that they took advice does not make the transactions abusive.”

KPMG is not otherwise referred to in this section, and that is no doubt explained by the view that had already been expressed that what KPMG said was subjective and therefore inadmissible in determining, on the necessary objective basis, what the essential aim of the transactions was.

As I have already stated, this treatment of KPMG’s part in the process was one of the grounds relied on by Mr Pleming in support of his appeal to the Upper Tribunal. This was reflected in the Upper Tribunal’s decision at paragraphs 107 and 108. At paragraph 107 the Upper Tribunal referred to some of the correspondence from KPMG, from which it was clear that KPMG had developed the scheme with the aim of securing a VAT advantage, and which showed KPMG promoting the scheme by reference to the tax advantage to be obtained, though in another letter there was reference to the funding benefit of the transaction as well as to the indirect tax advantage. The Upper Tribunal also noted that the First-Tier Tribunal had put all or most of this material to one side on the ground that these communications showed KPMG’s subjective view of what KPMG thought it was selling, and therefore not relevant to the objective assessment of the essential aim. The Upper Tribunal then set out its own view as to the KPMG material in paragraph 108 which I quote in full:

“In our judgment, some of the contents of the communications between Pendragon and KPMG are admissible and relevant to the question which needs to be decided as to essential aim. We consider that it is an objective characteristic of the scheme which was implemented that it had been devised by tax advisers to obtain a tax advantage, without regard to any other commercial benefits which might result. However, that characteristic is not of itself conclusive. It is legally possible for a scheme of such a kind to be implemented in a range of different circumstances and so that, in some circumstances, the essential aim of the particular transaction would not be to obtain the tax advantage. One would expect a tax adviser to concentrate on the tax treatment of the arrangements and to leave it to its client to assess the commercial circumstances in which the arrangements might be implemented. Furthermore, the mere fact that a taxpayer takes tax advice as to how to structure an intended transaction does not of itself establish that the essential aim of the transaction is to obtain a tax advantage. The decided cases make it clear that a taxpayer can engage in tax planning and select a permissible method of proceeding which is more tax efficient than alternative methods. Further, we do not consider that the opinions expressed by KPMG as to the likely success of the scheme and whether it was an “aggressive” scheme and whether HMRC would be likely to challenge it are admissible when objectively assessing the essential aim of the transaction. Further, even if such opinions were admissible, we would not regard them as of any real help in assessing the essential aim of the transaction.”

The Upper Tribunal made no further reference to KPMG or to this material in their decision. What I draw from this is that, although it disagreed with the First-Tier Tribunal as to the status of some of the KPMG material, it did not find its own different view of that material to be significant in resolving the issue as to essential aim. What it said in paragraph 108, after the first two sentences, seems to me to be sound and realistic. Parties to transactions which come to be examined in relation to the abuse of right principle will inevitably have entered into them with a view to economic advantage, whether by increasing their receipts or by reducing their outlay or both. Such parties will have done so with the benefit of advice, internal or external or both, as regards factors relevant to the economics. In all cases where it is said that the abuse of right principle comes into play in relation to taxation, a fiscal advantage will have been obtained (subject to the abuse of right principle) by the use, in accordance with their terms, of the relevant legislative provisions. That is always likely to have been done with the benefit of professional advice as regards tax, among other things. The relevant advisers, if external, are always likely to have commended their advice – both in general and in particular – as beneficial to their clients, so as to justify instructing them to advise in the first place, and paying them the fees charged in due course.

In that context, I agree with the Upper Tribunal that the fact that a tax adviser draws attention to tax advantages and points out the tax consequences of the course which his advice recommends is of no assistance at all in determining the objective aim of the transactions if they are later entered into. I also agree with the Upper Tribunal that the views expressed by the adviser are of no assistance either. I have difficulty in seeing what is left, once these matters are disregarded, of the communications between the tax adviser and the client that could be relevant and admissible, such as to justify the Upper Tribunal’s apparent disagreement with the First-Tier Tribunal, expressed in the first two sentences of paragraph 108. At all events, I agree with the view which seems to me to emerge from the Upper Tribunal’s paragraph 108 as a whole that, even if the First-Tier Tribunal was wrong in some respect in the characterisation of the KPMG material, it had no effect on the outcome of the case. Accordingly I do not accept Mr Pleming’s submission that this was a point on which the First-Tier Tribunal fell materially into error, even if it erred at all, as to which I am very doubtful.

One point on which the Upper Tribunal clearly did hold that the First-Tier Tribunal had gone wrong was in failing to have regard to the scale of the tax advantage. In paragraph 109 it said: “In the present case, if one contrasts the scale of the tax advantage and the nature of the other commercial consequences, it seems to us prima facie that the obtaining of the tax advantage was overwhelmingly the real reason why Pendragon entered into the transactions.” In paragraph 114, quoted above at paragraph [82], having said that it considered that it could see how and why the First-Tier Tribunal reached a conclusion which it had held was plainly wrong, the first thing mentioned was the First-Tier Tribunal’s failure to refer to matters “which we regard as of central importance”, of which the first is the scale of the tax advantage.

It is true that the First-Tier Tribunal did not refer to the amount of the potential tax advantage in their reasoning as to whether the essential aim was to secure the tax advantage. It cannot have been unaware of it, but it is not mentioned as a relevant factor. Nor was it evaluated in any way which would make it possible to carry out any kind of balancing exercise, comparing the tax advantage with the significance of the commercial benefit of the transactions to Pendragon in other respects. I am not aware of any other case in this area in which such an exercise has been carried out, or suggested as being relevant. It seems to me that it can be taken as read that there is likely to be a significant tax advantage in any case where the fiscal authorities invoke the abuse of right principle. To point out that this is so in any given case does not really carry the analysis any further as regards whether the essential aim of the transaction was to secure the tax advantage, so as to bring the abuse of right principle into application.

Mr Cordara submitted that, given that the transaction will necessarily have been undertaken in such a way as to obtain a tax advantage, the real question is not about that fact, nor about the scale of that advantage, but about whether there is some explanation for the transactions other than the mere attainment of tax advantages: see Halifax at paragraph 75, quoted above at paragraph [19]. As mentioned above, the Court’s decision in Part Service is expressed in slightly different terms, but its later decision in Weald Leasing (see paragraph [57] above) shows that the test is still that laid down in Halifax.

Moreover, as I have said at paragraph [69] above, the judgment in Part Service itself gives a clear indication that the existence and significance of ancillary or collateral commercial aims can be determined in an objective fashion. This provides no support for Mr Pleming’s proposition (and that of the Upper Tribunal) that a balancing exercise of the tax advantage against the commercial benefits is necessary or appropriate in this or any other case.

It seems to me that Mr Cordara is right to argue that the tax advantage should, as it were, be taken as read, and that the focus of attention should be on whether there is another explanation for the transactions, or at any rate for the elements in them which are said to be artificial. The court or tribunal needs to be realistic in assessing any explanation which is put forward, so that one which is not substantial may not suffice, any more than did the suggested justifications put forward in Part Service for adopting a structure of the transactions which was plainly of an artificial nature.

If it were necessary or appropriate to carry out an assessment of the balance of significance or value between, on the one hand, the tax advantage obtained and, on the other hand, the other commercial or economic justifications for the transaction, then it seems to me that the requirement of legal and economic certainty, which the European Court has repeatedly emphasised in the cases about abuse of right, from Halifax onwards, would be seriously prejudiced. That is supported by what Advocate General Poiares Maduro said at paragraph 89 of his opinion in Halifax, quoted at paragraph [46] above, to the effect that it would be very unsatisfactory, as regards the need for legal certainty, to allow an approach under which it was relevant to consider what purposes of a given transaction ought to be considered predominant.

For those reasons I cannot accept this particular criticism by the Upper Tribunal of the First-Tier Tribunal’s approach, nor Mr Pleming’s submission that the First-Tier Tribunal erred in this respect.

The transactions with which we are concerned did afford to the Pendragon Group short term credit facilities amounting to some £20 million. The Upper Tribunal recognised this and that, accordingly, the transactions did have real commercial consequences. But later in paragraph 109 it said that it provided short term finance on relatively expensive terms and at a time when Pendragon had undrawn committed facilities for sums comfortably above the amount advanced. In effect, it said that Pendragon did not need to enter into the transactions in order to get the funding which they provided. Then, in paragraph 114 (immediately after referring to the point about the scale of the tax advantage) it identified as an error by the First-Tier Tribunal the failure to refer to “the fact that the finance was relatively expensive at a time when Pendragon had unused committed facilities”.

It seems to me (and Mr Cordara did not argue to the contrary) that this is a point which was relevant to the objective assessment of the essential aim. If Pendragon really did not need an extra £20 million of short term credit, and if it was obtained on particularly expensive terms, then it might be legitimate to draw the inference that the object of the transaction was not the short-term funding but some other advantage, and the tax saving would be the obvious explanation.

However, the First-Tier Tribunal had taken a different view of the facts. I have referred to the review of the relevant evidence given by Mr Forsyth in its decision, at paragraphs [96] and [97] above, and I have quoted paragraph 56 at paragraph [98] above. The decision therefore proceeded on the basis of findings that, at the relevant time, Pendragon was renegotiating its borrowing facilities and that the safety margin in the group’s headroom was then much smaller than was desirable.

Paragraph 56 of the First-Tier Tribunal’s decision is not quoted in the Upper Tribunal’s decision (unlike paragraphs 51 to 53 and 57 to 59), nor is it referred to as such. The Upper Tribunal did make observations about the terms of the finance, at paragraph 98, and about the facilities available to Pendragon, at paragraph 105 and 106. The matters to which reference is made in those paragraphs are no doubt relevant, but I fail to see on what basis the Upper Tribunal felt able to disregard the findings of the First-Tier Tribunal set out at its paragraph 56. These were not entirely findings of primary fact, though in part they were, but insofar as they were not primary findings, they were findings of fact based on an evaluation of the evidence of Mr Forsyth (which in this respect was not merely subjective) and of the other material which was in evidence.

Accordingly, it does not seem to me that the Upper Tribunal was able to identify any error in the First-Tier Tribunal’s assessment of Pendragon’s position as regards borrowing needs, as set out in its decision at paragraph 56, nor has any such error been shown to exist in Mr Pleming’s submissions to us. For that reason, I would reject this part of the Upper Tribunal’s reasoning in paragraph 114 for holding that the First-Tier Tribunal had erred in law.

As I have mentioned, Mr Pleming focussed part of his challenge to the transactions on three elements which he said were artificial: the use of the CLCs to create leases, the terms of the hybrid leases, in particular as to the stage at which ownership would pass to the lessee, and the use of the offshore SGJ rather than its onshore parent. He was right to do so, because it is clearly relevant to identify whether any aspect or element of a series of transactions is artificial, and if so what, before deciding whether the essential aim of the transaction, or of that part of the transaction, was to obtain the tax advantage. Thus, in Halifax what was artificial was the interposition of a subsidiary company between Halifax itself and the builder of the call centres, in circumstances in which that subsidiary did not carry out any separate function or activity, and did not undertake any commercial risk. It was not there to do anything except to pay the builders (and with that the VAT charged) and to recover the VAT as input tax from HMRC. In WHA what the Court of Appeal considered to be was artificial was the interposition of an added offshore entity in a chain of retrocession of claims handling liabilities, that entity having no independent economic function to perform. In Part Service the artificial aspect was the division between two different companies in the same group (but with different VAT status) of the obligations under what was essentially a single contract with a third party. In Weald Leasing, the decision to acquire the equipment by lease instead of by an outright purchase was not artificial, because it was a legitimate commercial choice between proper alternatives. What might have been artificial was the interposition of a nominally independent third party and the setting of the rental payable by the acquiring company at an uneconomically low level.

The Court’s recognition, in Weald Leasing, of the trader’s freedom to choose between different types of transaction, and to take into account the different VAT consequences of the options available to him, and in particular that he does not have to choose the one which involves paying the highest amount of VAT, was a clear application of the approach articulated by the Court in paragraph 73 of its judgment in Halifax. That approach, together with the emphasis on the need for legal certainty and for the foreseeability of the application of Community legislation, and its especial importance in the case of “rules liable to entail financial consequences” (see Halifax paragraph 72), must be respected in the application of the principle of abuse of right. It seems to me that it is of the highest importance to bear these factors in mind when considering a contention by HMRC that elements in a series of transactions are artificial and such as to attract the principle of abuse of right. RBS Deutschland shows, for example, that it is not contrary to the principle of abuse of right to choose to carry out a leasing transaction through one company in a group, established and trading outside the UK, where the application of its local VAT legislation is such as to give a more beneficial tax treatment than would have been the case if a UK member of the group had been used. The choice of the foreign subsidiary rather than a UK subsidiary might be regarded as artificial, but it was not so treated by the European Court.

The use of the Captive Leasing Companies. Mr Pleming argued that there was no independent reason for Pendragon to have sold the relevant cars to the CLCs and then have arranged that the CLCs would grant hybrid leases to the Dealerships immediately before assigning the leases and the cars to SGJ. This is not the same sort of point as arose in WHA, where the interposed Gibraltar entity did nothing other than undertake a liability which it immediately passed on, nor in Halifax as regards the interposed subsidiary. However, it could be said that the use of the CLCs was not strictly necessary. Pendragon plc could have retained the cars, or it could itself have entered into leasing arrangements with the Dealerships, and then assigned them to the bank, or if it assigned them to the CLCs, the latter could have assigned them, in turn, to SGJ without first entering into the hybrid leases. Other financing arrangements were available from manufacturers but the evidence was that the Pendragon Group wanted to explore other possibilities so as not to be over-dependent on the manufacturers.

The terms of the hybrid leases. In this respect the point is said to be that it was artificial to set up the agreements on terms that the ownership of the car would not pass until after all the money had been paid. This is the effect of paragraph 8(c) of the Second Schedule to the hybrid lease, generally referred to as clause 8(c):

“When the Hirer has made all the payments under the Hire Agreement to the Owner the Hirer will have the option, seven days after the Hire Agreement ends, of purchasing the Goods from the Owner for the Option to Purchase Price specified in the Agreement Schedule. The Hirer will not have this option if all payments have not been made or if the hiring of the Goods has been terminated. The option will remain open for seven working days only. Until this option has been exercised the Goods remain the property of the Owner and, for the avoidance of doubt at no point during or after the Period of Hire will the Hirer acquire any Ownership in the Goods whether legal, equitable, beneficial, economic or otherwise. An option to purchase fee of £10 will apply per vehicle or such other amount from time to time notified in writing by the Owner and is payable upon exercise of the option to purchase. Title to the Goods will pass from the Owner to the Hirer fourteen days after exercise of the option to purchase.”

Mr Pleming’s point is that it is unusual, uncommercial and, he said, therefore artificial to defer the moment at which the Hirer acquires title to the goods, having made all payments due under the hire agreement, until (a) the exercise, no earlier than seven days later, of an option to purchase, and (b) a further 14 days have expired thereafter. He submitted that this unusual feature was required solely in order to fit into the tax legislation and that this is therefore an artificial element such as to bring into play the abuse of right principle.

The use of the offshore bank. The third point relied on is the use of SGJ rather than its UK parent. In this respect too, Mr Pleming argued, the structure of the transaction was driven by the need to bring the case within the VAT legislation. The parties were not agreed as to whether this feature was necessary in order to achieve the tax saving. I have mentioned this debate at paragraph [43] above.

I turn to what the First-Tier Tribunal said about these points. At paragraph 175 there is this:

“It is permitted to arrange affairs to take advantage of the relevant tax provisions provided it is not abusive. Here we find that the financing was necessary but was done in a tax-efficient but non-abusive way. The ECJ has not prevented this. It specifically says that one may choose the more tax-efficient way of carrying out a transaction. We consider that this was what Pendragon did and we find this as a primary fact. The obtaining of finance provided a sufficient autonomous basis and economic justification.”

At paragraph 178, having referred to what had been said by the Court of Appeal in WHA, the First-Tier Tribunal pointed out that any lender would require security and, as regards chattels, would want ownership of the goods as well as the right to the income stream. It said: “Given the need for finance from a third party who would require such security we find this to be part of the normal commercial operations in these particular circumstances”.

Later, at paragraph 182, having referred to WHA and to the artificial insertion of the additional Gibraltar company, it differentiated the present case, saying this about the Scheme as a whole:

“Given that outside financiers would require security over both the vehicles and the income flow it is not an artificial step to carry out a transfer of the business as a going concern. Given that there was to be a hire of the vehicles drafting an HP contract in such a way as to give the taxpayer the choice of carrying out the normal commercial operation in a tax-efficient way cannot be said to be abusive in the current circumstances. It was not something inserted as a purely artificial step which could be disregarded. The cars were to be sold and leased back such that title to the goods, i.e. the cars and the income (i.e. the rentals), would be in a third-party bank. The vehicles had to be got back and the short-term finance repaid. In the circumstances of this case there was no artificial insertion of steps or the creation of a wholly artificial set of transactions rather the necessary financing was carried out in a tax-efficient way. This seems to be contemplated by the ECJ when discussing how a taxpayer can structure his or her business.”

Earlier, at paragraph 68, they had said this, about the hybrid leases:

“It was what was described as a ‘hybrid agreement’ as it was designed to obtain a particular VAT treatment. It provided that title would not pass to the dealership company until a period after payment in full. This is not that unusual in an HP type of agreement nor uncommercial.”

After setting out clause 8(c) the Tribunal said that in its experience this was not an unusual or uncommercial term in an HP agreement, and it so found.

It referred to the use of an offshore bank, for example at paragraph 104(r) near the end of the summary of Mr Pleming’s arguments, but did not deal separately with the point made by HMRC that it was artificial. At paragraph 56 it referred to the introduction to Société Generale, rather than to SGJ as such; later it referred to the dealings with SGJ and those between that company and its parent. Thus, the First-Tier Tribunal’s decision as regards that aspect of the case is to be found in the more general observations, at paragraphs 175 and 182 already quoted.

The Upper Tribunal addressed these three points at paragraph 111. At this stage in the judgment it had expressed the prima facie view that, because of the scale of the tax advantage and because of the apparent lack of need on Pendragon’s part to have recourse to this arrangement for its short-term funding requirements (see paragraph 109), there was an overwhelming case for saying that, objectively assessed, the essential aim of the transaction was to gain the tax advantage. It then went on to check that preliminary view against the evidence of the witnesses and against Pendragon’s submission that the structure of the transactions was normal and commercial. It was on the latter point that it looked at whether elements in the transaction were artificial or not. It said that the use of the CLCs “was wholly artificial and served no commercial purpose”. The provisions of clause 8(c) of the hybrid lease appeared uncommercial but it said they were not necessary to the efficacy of the scheme, so it did not make a finding that this was artificial. As for the offshore bank, it said that it served no independent commercial purpose and, at the end of the paragraph, that it was artificial and driven by the perceived need to achieve the desired VAT result and no other reason.

Thus, the First-Tier Tribunal made specific findings about the use of the CLCs and the hybrid lease, it did not deal specifically with the offshore bank issue, and it found that the arrangements as a whole were not abusive or artificial. The Upper Tribunal found that the use of the CLCs was wholly artificial and served no commercial purpose, and that the use of the offshore bank was also artificial. Although it said that clause 8(c) of the hybrid lease appeared uncommercial, it did not say in terms that it was artificial. On that point, therefore, the two Tribunals were not at odds, whereas as regards the other two points the Upper Tribunal took a quite different view from that of the First-Tier Tribunal.

That difference of view was not, in terms, part of the Upper Tribunal’s reasoning in paragraph 114 where it explained why it considered that the First-Tier Tribunal had erred in law. Nevertheless, to the extent of the Upper Tribunal’s specific disagreements with the First-Tier Tribunal (on the two points), it seems to me right to proceed on the basis that this did form part of the reasoning. I will therefore bring this aspect into the discussion of the Upper Tribunal’s disapproval of the First-Tier Tribunal’s conclusion generally, to which I now turn.

As I have already said, I do not consider that the First-Tier Tribunal went wrong in its treatment of the evidence of Mr Forsyth, or of the KPMG material, or in failing to weigh the size of the tax advantage in the scale against the other commercial aspects of the transactions.

In its paragraph 114 the Upper Tribunal criticised the First-Tier Tribunal for expressing as a finding of primary fact something that was not a matter of primary fact. There is force in this comment. At paragraph 49 the First-Tier Tribunal held as a primary fact that further sources of funding were very important to the Pendragon Group (see paragraph [97] above). I see nothing wrong with that, and the Upper Tribunal did not criticise it. However at paragraph 51 (see paragraph [80] above) the First-Tier Tribunal expressed itself as finding as a primary fact that the obtaining of finance was the predominant, principal or a central aim of the transactions. I agree with the Upper Tribunal that this conclusion was not a matter of finding primary facts. Instead it was an evaluative process. Even as such, however, it did involve an assessment of the facts. That the First-Tier Tribunal should have misdescribed what they were doing, by calling it the finding of primary facts when it was not, is unfortunate but this should not obscure the fact that it did have to consider and assess the relevant facts, as it had found them to be, and to evaluate them on an objective basis. The same can be said of the First-Tier Tribunal’s statement at paragraph 175 that it found as a primary fact that the financing was necessary but was done in a tax efficient but non-abusive way. That financing was necessary is, or at least may be, a matter of primary fact, but whether it was obtained in a non-abusive way is a matter of evaluation and assessment, not of primary fact-finding. In turn at paragraph 184, quoted at paragraph [79] above, the First-Tier Tribunal expressed itself to find as a fact (though not in terms a primary fact) that on an objective consideration of the relevant circumstances the essential aim of the transactions was to obtain finance, not to obtain a tax advantage. That is much the same as it had said at paragraph 51, where it is expressed to be a finding of primary fact.

As to these passages, I agree with the Upper Tribunal in its paragraph 114 where it said that the issue as to the essential aim of the transaction was not an issue as to the primary facts but one which was, at least in part, an issue of law. I also agree with the statement later in that paragraph that the background facts may have been primary facts but an assessment of aim by reference to objective factors is not a primary fact. Immediately before making that comment, the Upper Tribunal said that the First-Tier Tribunal’s reference to finding as a primary fact something that was not a matter of primary fact “serves to emphasise that the FTT did not understand the task before it as involving an objective assessment of essential aim”. If it were apparent from the First-Tier Tribunal’s reasoning that it had failed to understand the nature of the task which it had to undertake in this respect, then the Upper Tribunal would be entitled to say that there was an express, or at least implicit, misdirection of law in the First-Tier Tribunal’s decision. However, nothing else in the First-Tier Tribunal’s decision discloses any such misunderstanding of its task or their function, nor did the Upper Tribunal identify any other indication of such an error in this respect. In my judgment it is not a fair reading of the First-Tier Tribunal’s decision to say that, because it described as a finding of primary fact something which was either not at all, or not entirely, such a finding but rather an exercise in the objective evaluation and assessment of the facts as found, therefore it must have misunderstood, and misdirected itself as to, the nature of the task it was required to undertake, and thus erred in law.

The Upper Tribunal’s other criticisms of the First-Tier Tribunal in paragraph 114 include points that I have already considered and rejected, in particular the Upper Tribunal’s reliance on the First-Tier Tribunal having failed to refer to the scale of the tax advantage, and that the finance was relatively expensive at a time when Pendragon had unused committed facilities, which amounted to an implicit rejection of the First-Tier Tribunal’s finding that Pendragon needed the finance that the Scheme afforded. It commented on the fact that the First-Tier Tribunal did not provide in its decision a comprehensive description of the arrangements, nor a detailed analysis of them. So far as that is concerned, on the one hand it is to some extent a fair comment on the presentation and the content of the First-Tier Tribunal’s decision. The elements of the Scheme are all explained in the section headed “The transactions in question”, starting at paragraph 63 and continuing to paragraph 82. They could have been set out more clearly, as the Upper Tribunal did at paragraphs 20 to 22, and they could have been accompanied by an analysis of the VAT consequences of the scheme, as the Upper Tribunal did at paragraphs 23 to 36.

But it is one thing to say that a court or tribunal could have expressed its judgment more clearly; it is quite another to say that this shows an error of law or misdirection. The Upper Tribunal relied on this feature to say that the First-Tier Tribunal “in consequence focussed on the fact that finance was obtained, rather than on the fact that the obtaining of finance was subordinate to the essential aim of obtaining a tax advantage”. With respect, it seems to me that here the Upper Tribunal fell into an error which it had already identified on the part of the First-Tier Tribunal. It was not a “fact” that to obtain finance was subordinate to the essential aim of obtaining a tax advantage. What was the essential aim is also not a question of “fact”, but of an evaluative determination on the basis of the objective facts. It contains an element of law, in that a given set of facts may be such as do, or do not, permit in law a conclusion that the essential aim is to obtain a tax advantage. As I have indicated above, given that there always will be a tax advantage in these cases, otherwise the question of abuse of right will not arise, really the question is whether “the economic activity carried out may have some explanation other than the mere attainment of tax advantages” (see Halifax judgment, paragraph 75) or in other words “whether the activity at issue has some autonomous basis which, if tax considerations are left aside, is capable of endowing it with some economic justification in the circumstances of the case” (Halifax Advocate General paragraph 87), or whether “the economic activities are, at least to some extent, accounted for by ordinary business aims” (Halifax, Advocate General, paragraph 89), though this is subject to the qualification that it may not be sufficient for this purpose to show “the possible existence, in addition [to the tax advantage], of economic objectives arising from, for example, marketing, organisation or guarantee considerations” (Part Service judgment paragraph 62).

In this respect I find it revealing that, later in paragraph 114, the Upper Tribunal referred disparagingly to the First-Tier Tribunal’s use of the phrase “some autonomous justification” (in its paragraph 173 – and see also in paragraph 175 “a sufficient autonomous basis and economic justification”) as suggesting that the First-Tier Tribunal had “lost sight of the fact that it was not enough for Pendragon to show that the transactions produced real commercial consequences where those were collateral matters which did not detract from the essential aim of obtaining a tax advantage”. Given that the phrases “autonomous basis” and “economic justification” are taken straight from the opinion of the Advocate General in Halifax, and that immediately after the use of these phrases in paragraph 175 the First-Tier Tribunal referred in terms to Part Service and distinguished it, I am unable to regard its use of these phrases as indicative of any error of law.

In the course of its paragraph 114 the Upper Tribunal made it plain that it disagreed with the conclusion of the First-Tier Tribunal, but it knew that it could not reverse the decision unless it was based on one or more errors of law. Because the determination of whether the essential aim of the transactions involved issues of law, it held that the First-Tier Tribunal’s error was one of law. It said that, if it had to go so far, it would hold that the First-Tier Tribunal’s conclusion was one that was not open to a tribunal properly directing itself: that is, in effect, an Edwards v Bairstow basis for finding a misdirection. It is the conclusion stated at the end of paragraph 114. It also sought to identify errors of law. For reasons which I have given, I would reject those that are identified in that paragraph. I do not consider that any of these points showed that the First-Tier Tribunal had made an error of law.

The Upper Tribunal did not refer in paragraph 114 to the three artificial elements discussed above. However, these elements (or at least two of them) had formed part of the reasoning process leading up to that paragraph, and they formed part of Mr Pleming’s criticisms before us of the First-Tier Tribunal’s decision. The Upper Tribunal was clear in categorising the use of the CLCs as artificial, and likewise the use of the offshore bank rather than its UK based parent. As to the latter, Mr Cordara pointed to the decision in RBS Deutschland that, where an offshore member of a banking group was used for particular transactions, which had particular advantageous tax consequences, it was not sufficient for HMRC to point out that an onshore member of the group could also have been used, which would have had different fiscal consequences. Why should it make a difference, he asked rhetorically, in the present case according to whether a Jersey bank or a UK bank, each in the same group, was used?

So far as the CLCs are concerned, the First-Tier Tribunal commented on the structure adopted in general terms, without dealing with the particular issue of the CLCs. If finance was to be obtained from a bank, rather than, for example, from a manufacturer’s financing subsidiary, security would have to be provided for the loans over the cars. That security would need to be given in a way which enabled the Pendragon Group to continue to use the cars for demonstrator purposes and, in due course, to acquire the cars back from the bank when the security was either discharged entirely or released on an individual basis. Therefore, although it would have been possible, no doubt, for Pendragon plc to assign the cars directly to the bank, it was necessary, or at least desirable, that before such an assignment took place, a separate transaction be set up under which the cars would continue to be used by the dealerships despite the assignment to the bank.

Mr Pleming submitted that Pendragon could have assigned the cars to the bank which could then have leased them to the relevant dealership company. Even if that approach had been adopted it would have been necessary for Pendragon and the bank to agree on the terms of the lease in advance, and for the lease to have been created immediately after the assignment to the bank, since otherwise the Pendragon Group could not continue to use the cars freely. The choice could therefore be between one arrangement under which the lease was put in place by the Pendragon Group internally immediately before the assignment, on terms which suited the group (and which were no doubt also cleared in advance as being satisfactory to the bank), and another under which the form of the lease was also agreed in advance, but the lease was effected immediately after the assignment to the bank. As between those alternatives it does not seem to me likely that one could be regarded as normal and the other as artificial. Considered in that way, it does not seem to me that there was anything evidently abnormal or artificial in the creation of leases within the Pendragon Group before the assignment to the bank.

This point does not seem to have been explored in evidence with the expert witnesses, but it is supported, as it seems to me, by what one of them (Mr Ingram) said in his witness statement, in particular at paragraph 19. The experts gave useful evidence about the practice as regards financing in the car industry and as regards the patterns of transactions used for financing either in that sector or generally. This was, of course, somewhat general, as is often the nature of expert evidence, but it seems to me that it supported the contentions put forward by Pendragon and the conclusions reached by the First-Tier Tribunal.

The alternative point relied on by HMRC relies on the fact that the leases were created by the CLCs rather than by Pendragon plc itself. That issue was addressed by the expert witnesses called by Pendragon who said that the use of such subsidiary companies was rational and normal. That evidence seems to me to have provided a sound basis for the First-Tier Tribunal’s conclusion that the transactions by which the financing was obtained were tax efficient but not abusive (paragraph 175) and that there was no artificial insertion of steps (paragraph 182). It clearly regarded the case as quite different from the insertion of the otherwise unnecessary company in the Halifax case, or the additional Gibraltar re-insurer in WHA, or the artificial sub-division of what was really one transaction in Part Service.

It seems to me that the Upper Tribunal’s categorisation of the use of the CLCs as wholly artificial and serving no commercial purpose was an assertion unsupported by any reasoning or analysis. The tribunal made no attempt to address the reasons why a group might choose to set up lease or HP agreements for the cars before assigning the cars (with the benefit of the agreements) to a bank by way of security, nor why, if it wished to do that, it might choose to use a subsidiary to grant the leases other than the company which had bought the cars from the manufacturer.

Nor did the Upper Tribunal address the question whether the use of an independent offshore bank was in itself artificial rather than a matter of reasonable and normal commercial practice. In this instance it had at least referred to the expert evidence on the point (at paragraph 101), but it did not consider the relevance of RBS Deutschland, despite having discussed that case at paragraph 52. That case supports the proposition that, if it would be normal for a bank to undertake a particular transaction, it should not, in the ordinary way, matter whether that bank is incorporated or carries on business in one place rather than in another.

For those reasons it does not seem to me that the Upper Tribunal’s criticisms of the First-Tier Tribunal’s decision in these respects justified the Upper Tribunal in holding that the First-Tier Tribunal was not entitled to come to the conclusion that it did, namely that no artificial factors had been inserted into the transactions (see paragraph 182).

The determination of whether a case falls within the second limb of the abuse of right principle, as set out in paragraph 75 of the European Court’s judgment in Halifax, is a task which may well require a tribunal to make findings of primary fact, as regards the facts relevant to the objective assessment of the transactions. In the present case some primary facts had to be found, including as to the need on the part of the Pendragon Group for short-term finance at the relevant time. Otherwise, the task involves an assessment and evaluation of the transaction, by reference to the relevant facts as found or undisputed, and on an objective basis, guided by the observations of the European Court as to what the correct principle is. It may not be an easy task. There may be cases in which it is clear that some element in the arrangements, which is needed to obtain the tax advantage, has no independent commercial rationale or explanation. Halifax was such a case, there being no point at all (other than the fiscal motive) for the interposition of the subsidiary company which entered into the construction contract for the building of the call-centres, but which took no economic or commercial risk, all of which was borne by Halifax itself. Equally, there was no non-fiscal explanation for the insertion of the additional Gibraltar re-insurer in the chain considered in the WHA case.

In other cases, it is a much more difficult task to see whether the requirements of the abuse of right principle are satisfied. From the start the European Court has made it clear that it is not abusive for an undertaking to choose to enter into commercial arrangements in a way which involves a better fiscal treatment (of whatever kind) than alternative possible arrangements. Advocate General Poiares Maduro in his Opinion in Halifax was emphatic about the strictness of the test for abuse, saying that if there were any commercial explanation for the transactions “other than the mere attainment of tax advantages” then it was not within the abuse principle: see his paragraphs 87 and 89, the latter passage endorsed in terms by the Court in its judgment at paragraph 75. The same approach appears from the judgment of the Court in Halifax at paragraphs 72 and 73, and from the use of the phrase “an abusive practice can be found to exist only if” in paragraph 74 (I have added the emphasis). That phrase applies directly to the first part of the test, as to the tax advantage being contrary to the purpose of the relevant provisions, but it must, as a matter of both language and sense, apply also to the second part of the test, set out in the following paragraph. Only if both tests are satisfied can an abusive practice be found to have been undertaken.

The European Court in paragraph 81 of Halifax refers to taking account of “the purely artificial nature” of the transactions. In many, perhaps all, cases, there is likely to be a real commercial exercise which underlies the transactions, as the building of the call centres was in Halifax and the reinsurance of the underwriting liabilities in WHA. For that reason it is natural to focus on the elements which are said to be artificial in the particular scheme, even though those may be only a part of the transactions as a whole. Thus the scheme as a whole may not be artificial and may have a genuine commercial purpose, but one or more discrete elements in the arrangements may be artificial, without any independent economic or commercial justification. But if the focus of attention is on one element in a series of arrangements, then while there may be cases in which one can isolate a particular step and characterise it as completely unnecessary from an ordinary commercial point of view (as in Halifax itself, or in WHA), in other cases the element in question may be one of several possible ways of carrying out something which would itself be a normal part of the arrangement. It may be unusual, but it may also be difficult to say that there is no economic justification for it at all. For something of that kind to be undertaken would be a normal course. What would have been a normal thing to be done may have been done in a relatively unusual way. Given that parties are allowed to choose in what way they organise their affairs, and can legitimately take account of the natural wish of a commercial undertaking to reduce its fiscal exposure in so doing, it may be more difficult to treat such a step as abusive and artificial if it can be regarded as no more than one of a number of possible ways of carrying out a stage in the arrangements which is, in itself, normal and responsive to ordinary commercial concerns.

Considerations of this kind seem to me to show that, in all but the clearest of cases (such as Halifax and Part Service may very well have been), a significant exercise of assessment, evaluation and judgment is necessary, once the relevant facts have been identified and, in case of any dispute, found as facts, in order that the national court (here the First-Tier Tribunal) can determine what the real substance and significance of the transactions is, whether the essential aim, or the principal aim, of the transactions is to obtain a tax advantage or whether the transactions have some other adequate economic or commercial explanation. This may not be by any means an easy task, requiring a judgment to be made in order to resolve the inherent tension in the European Court’s enunciation of the principle as between the freedom of economic undertakings to organise their business affairs to their own best advantage, and in so doing to be able to rely on the principle of legal certainty, on the one hand, and the proposition that they may not obtain tax advantages which are contrary to the purpose of the relevant provisions by entering into purely artificial arrangements, on the other hand.

The task is made no easier by the fact that we are concerned not with a specific legislative provision of an anti-avoidance nature, but with the enunciation by the European Court of a principle for which differing language is used in different contexts and at different times. It would be wrong to treat any given formulation by the Court as if it were a legislative text. The difficulty to which this contributes in applying the abuse of right principle seems to me to call for the adoption of a cautious approach as regards finding that the principle does apply, in any case which is not really clear on its facts.

For my part, looking at the facts of the present case, I can imagine the possibility that, if the members of the Upper Tribunal had been sitting as the First-Tier Tribunal hearing Pendragon’s appeal at first instance, they may well have decided the case differently from the actual decision of the First-Tier Tribunal, and may have been able to do so in a way which involved no misdirection of law. It does not follow from this that the actual decision of the actual First-Tier Tribunal was one which it was not entitled to reach. It is in the nature of an evaluative exercise that, on given facts, two different tribunals, properly directed as to the law, may each be able to come, entirely properly, to different conclusions. That is not a situation in which the appellate tribunal is entitled to say that the first instance tribunal has erred in law, and has reached a conclusion which it was not entitled to come to.

That, therefore, brings me back to the principles discussed earlier in this judgment about the circumstances in which an appellate tribunal or court can disturb the finding of a first instance tribunal on an appeal limited to errors points of law. As Jacob LJ said in the Pringles case, the first instance tribunal is the primary maker of a value judgment based on the primary facts, and unless it has made a legal error in so doing it is not for an appeal court to interfere.

In the present case the Upper Tribunal considered that the First-Tier Tribunal had made legal errors. Part of its reasoning in paragraph 114 was that the First-Tier Tribunal must have misdirected itself, which amounts to saying that its conclusion was legally impossible. I disagree as to that.

In addition the Upper Tribunal identified some points which it said were specific examples of legal error. I have reviewed these already. On these points too I disagree with the Upper Tribunal. I agree that the issue as to the essential aim of the transaction was not a question of primary fact, and that there are elements of law in that question. I do not regard it as an error on the First-Tier Tribunal’s part not to refer to the scale of the tax advantage, or not to balance the other commercial factors against that tax advantage. The Upper Tribunal’s reference to the short term finance being obtained “at a time when Pendragon has unused committed facilities”, and thus by implication holding that Pendragon did not have a need for this finance, is inconsistent with the First-Tier Tribunal’s finding which was one of primary fact in this respect, and it was not open to the Upper Tribunal to override that finding. The First-Tier Tribunal’s failure to set out so clear a description of the arrangements as the Upper Tribunal had done, or a detailed analysis of them, does not seem to me to reveal any error of law. The reference to the First-Tier Tribunal focussing on the fact that finance was obtained (clearly a fact, and not in dispute) rather than the “fact” that the obtaining of finance was subordinate to the essential aim of obtaining a tax advantage demonstrates to me an error on the part of the Upper Tribunal rather than of the First-Tier Tribunal. That proposition was not a matter of fact but of evaluation.

No error of law is shown on the part of the First-Tier Tribunal in relation to Mr Forsyth’s evidence. The First-Tier Tribunal’s reference to determining the essential aim of the transaction as a primary fact was incorrect but does not show, taken in context, that it failed to understand that an objective assessment of the aim was what was required. Equally, I reject the Upper Tribunal’s proposition in paragraph 108 (which may or may not have influenced its final decision) that the KPMG material was relevant and admissible. I also disagree with the observation that both the use of the CLCs and the use of an offshore bank were artificial. It seems to me that the First-Tier Tribunal was entitled, on a comprehensive objective evaluation of the arrangements, to come to the conclusion that no element of the arrangement was inserted artificially, and that the arrangements were not abusive or artificial.

Accordingly, it seems to me that no error of law has been shown on the part of the First-Tier Tribunal, whether by the Upper Tribunal’s reasoning or by Mr Pleming’s submissions to us on this appeal. The First-Tier Tribunal was entitled to come to the conclusion that it expressed. A differently constituted tribunal may have come to a different result on the same material. But that is not enough. The conclusion reached was open to the First-Tier Tribunal on a proper understanding of the law, and was not reached as a result of any misdirection.

It follows that the Upper Tribunal should not have allowed HMRC’s appeal against the First-Tier Tribunal’s decision, and Pendragon’s appeal to this court should be allowed, restoring the First-Tier Tribunal’s decision of the several appeals that were before it.

I agree.

I also agree.