First-tier Tribunal (Tax)

[2012] UKFTT 102 (TC)

TC01798

 

 

 

Appeal number TC/2009/10895

 

VAT – MTIC Fraud - contra trading - connection to fraud - means of knowledge

 

 

FIRST-TIER TRIBUNAL

TAX

 

FONECOMP LIMITED Appellant

 

 

- and -

 

 

THE COMMISSIONERS FOR HER MAJESTY’S

REVENUE AND CUSTOMS Respondents

 

 

 

TRIBUNAL: CHARLES HELLIER

NIGEL COLLARD

 

Sitting in public at Audit House, London  on 3-7,10-14 and 17 October 2011

 

 

Andrew Young instructed by Dass Solicitors  for the Appellant

 

Mark Cunningham QC and James Puzey, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

 

 

© CROWN COPYRIGHT 2012

DECISION

 

 

I. Introduction

1.      Fonecomp appeals against HMRC's decision to deny its input tax credit in respect of its purchase of mobile phones in two transactions (Deal 1 and Deal 2) which took place in July 2006. In each case Fonecomp immediately sold and exported, to companies incorporated in the EU, the phones it had bought. The amount of input tax in dispute is £183,951.30.

2.      In Deal 1 (Fonecomp’s documentation numbered it 1760) Fonecomp purchased:

(1)  1100 Nokia 9300i; and

(2)  1200 Nokia 8801

mobile phones from PDA Stuff Limited for £759,226.25, including VAT of £113,076.25, and sold them to Axess Denmark ApS (“Axess”), a company incorporated in Denmark, for £679,200. The purchase and sale invoices are dated 11 July 2006.

3.      In Deal 2 (Fonecomp’s documentation numbered it 1765) Fonecomp purchased:

(1)  900 Nokia N 80; and

(2)  500 Nokia N 91

mobile phones from TM Global Limited (trading as Team Mobile) for £475,875, including VAT of £70,875, and sold them to Axess. The purchase and sale invoices are dated 26 July 2006. (TM Global had sold PDA Stuff the phones it sold Fonecomp in Deal 1)

4.      HMRC rely upon the doctrine propounded in Axel Kittel v Belgium (C-439/04) (“Kittel”) to support this decision. They say that Fonecomp’s purchases were connected to the fraudulent evasion of VAT by a company called Softlink Limited, and that Fonecomp knew or should have known of the connection to VAT fraud.

5.      HMRC say that Softlink defaulted in its VAT payment in respect of transactions it undertook in August 2006. Yet the Deals took place in July 2006. HMRC say that Fonecomp's purchases are connected with Softlink’s default through Klick (UK) Limited (“Klick”). They say that Klick acted as a "Contra Trader". That is a term they use for a trader which (a) buys goods from a defaulter and exports them claiming, in what they term the "dirty chain", the input VAT (the “dirty input VAT”) on the purchase; and (b) in a "clean chain", imports goods and sells them to a third trader, and then offsets the dirty input tax against the clean output VAT on the sale to the third trader. The dirty input VAT is by this means sought to be transmuted into clean input VAT in the hands of the third trader; or at any rate the third trader is sought to be so distanced from the default that he could not know of his connection to it, or HMRC discover it

6.      In this appeal Fonecomp is alleged to be in the position of that third trader (although distanced from Klick by one or two other traders who bought and onsold the phones before Fonecomp purchased them). HMRC say that the activities of Klick provide the connection between Fonecomp’s purchase and Softlink’s default. They say that Fonecomp knew or should have known of its connection to VAT fraud.

7.      HMRC go further than this. They say that there was a wider managed scheme associated with Klick’s transactions; a scheme which involved some 37 other traders, including Fonecomp, in an organised assault on the revenue which involved reclaiming the VAT on which Softlink defaulted, amounting to some £66m. The evidence of this Klick Scheme is part of HMRC's argument for saying that Fonecomp must have known of its connection to fraud. Mr Cunningham put the argument memorably: an animate puppet cannot have his strings pulled without knowing that they are being pulled.

II.The History of the Appeal

8.      As the result of a direction made on 24 June 2010 by Judge Bishopp, this appeal was originally directed to be heard with the appeal of 12 other appellants in relation to transactions undertaken by them which HMRC had alleged were part of the Klick Scheme. In the period between that direction and the start of the hearing the appeals of the other appellants were either withdrawn or struck out for lack of pursuit.

9.      At an interlocutory hearing on 23 March 2011 Mr. Young had argued that HMRC’s statement of case was deficient because it failed specifically to argue conspiracy. The tribunal directed that this point be heard as a preliminary issue. In response to that direction the tribunal heard argument from Mr. Cunningham. Mr. Young did not appear to argue the point. We deal with this below.

10.   We were told that, as a result of the denial of its input tax claim and the refusal (in a direction made following a hearing on 13 July 2011) by this tribunal to direct that HMRC release part of it to the appellant,  the appellant had very limited funds with which to finance the costs of its appeal. Mr. Young had been instructed to appear before us for only five days over the course of a hearing which was listed for  11 days. That meant that there were days when the only party appearing before the tribunal was HMRC. The timetable for the hearing was arranged so far as possible to permit Mr. Young to be present for those parts of the evidence and argument in relation to which the appellant considered his presence would be most effective.

III. The Law

11.     The right to deduct input VAT is given domestically by sections, 1, 4 and 25 VAT Act 1994 in implementation of Article 17 of the Sixth Directive and its predecessor. In Axel Kittel v Belgium, Belgium v Recolta recycling C-439/04, C-440/04 [2006] ECR 1-661 the ECJ said [61]:

"…where it is ascertained, having regard to objective factors, that the… taxable person… knew or should have known that, by his purchase, he was participating in a transaction connected with fraudulent evasion of VAT"

 then  it was a matter for the national court to refuse to allow the right to deduct.

12.     The scope and meaning of these words has been discussed in subsequent judgements of the High Court and in Mobilx Limited v HMRC [2010] EWCA Civ 517 in the Court of Appeal. A number of issues have been raised. We have come to the following conclusions on those issues relevant to this appeal.

(1) Does a right to a VAT credit exist unless and until it is denied?

13.     At a preliminary hearing the appellant had sought a direction that HMRC make an interim payment to the appellant to permit it to fund its appeal. Mr. Young had argued that the way the ECJ had formulated its judgement meant that a right to an input VAT credit exists unless and until it is taken away by a court. The application was refused. The Decision is set out in the Appendix. For the reasons there given the ECJ was, in our view, indicating that the right does not exist if the relevant conditions are satisfied.

(2) Is the Kittel principle part of UK law?

14.     In Mobilx, Moses LJ said [47]:

“ …Accordingly, the objective criteria which form the basis of concepts used in the Sixth Directive form the basis of the concepts which limit the scope of VAT and the right to deduct under sections 1, 4 and 24 of the 1994 Act. Applying the principle in Kittel, the objective criteria are not met if a taxable person knew or should have known that by his purchase he was participating in a transaction connected with fraudulent evasion of VAT. That principle merely requires consideration of whether the objective criteria relevant to those provisions of the VAT Act 1994 are met. It does not require the introduction of any further domestic legislation.”

15.     That is a statement of law which is binding on us.

(3) The time of the requisite knowledge

16.     The taxpayer is barred from input tax credit only if he had the requisite knowledge at the time of his purchase ( Mobilx [52] and  Lewison J in Livewire  [16]: "HMRC must prove, having regard to objective factors, that the  taxable person, at the time of his transaction, knew or should have known that his transaction was connected with fraud".)

(4) The requisite knowledge

17.     The question is not whether the taxpayer had the requisite knowledge that the transaction might be connected to fraud, but whether he had that knowledge that it was connected to fraud.

(5) Meaning of "should have known"

18.     “Should have known” means the same as “having the means of knowledge”. A trader may thus lose his right to deduct if he fails to deploy the means of knowledge available to him ([50]  to [52]  Mobilx).

19.     Where a person had the means of knowing that the only reasonable explanation for his transaction is that it was connected to fraudulent evasion of VAT, then it can be said that he should have known that it was connected to such fraud. (Mobilx [59] and others)

(6) An Impenetrable Shield.

20.     If a taxpayer does not know of the connection with fraud and takes all reasonable precautions then he has an "impenetrable shield" and is entitled to the input tax credit. That in our view is equivalent to saying that if the taxpayer deploys all the means available to him and cannot thereby discover a connection to fraud (ie come to a conclusion that the only reasonable explanation of the circumstances of the transaction was connection to fraud) he is entitled to the input tax credit.

21.     If a taxpayer does not know the connection to fraud but does not take all reasonable precautions, then he is not automatically disentitled to input tax credit: it is only if, had he taken all reasonable precautions (i.e. used all the means of knowledge available to him)  he would not have discovered  the connection to fraud, that he is entitled to the input VAT.

(7) Connection With Fraud.

22.     The Appellant says in its skeleton argument that a “tenuous connection i.e. one at n- removed”, and displaced from wholly different supply chains is an insufficient one. It says that the test articulated by Lewison J [107] in Livewire is to be preferred to the approach adopted by the Chancellor in Blue Sphere at [40-46].  It says that the Court of Appeal in Mobilx does not address the tension between these views, and that Kittel concerned transactions where the fraud was committed by the seller, and not by a party at n- removes from the seller,  still less a fraud which depended upon a contra trading construct.

23.     In Mobilx Moses LJ dealt with the question of whether the fraud had to be that of the supplier to the taxpayer in order for there to be a connection. He said [62]:

"the principle of legal certainty provides no warrant for restricting the connection ... to a fraudulent evasion which immediately precedes a trader’s purchase. If the circumstances of that purchase are such that the person knows or should know that his purchase is or will be connected with fraudulent evasion, it cannot matter a jot that is that evasion precedes or follows the purchase. The trader’s knowledge brings him within the category of participant. He is a participant whatever the stage at which the evasion occurs."

24.     That as a matter of domestic law is binding on us. If we thought it a wrong or doubtful interpretation of Kittel then Mr Young urges us to make a reference to the ECJ.  He cites Rheinmuhlen-Dusseldorf v Einfuhr- und Vorratsstelle Getreide und Futtermittel Case 146-73 for the proposition that:

“The existence of a rule of national law whereby a court is bound on points of law by the rulings of a superior court cannot on this ground alone deprive the inferior courts of their power, provided for under Aricle 177, to refer questions to the Court of Justice of the European Communities for a preliminary ruling.”

25.     However for the reasons which follow we do not consider the proposition wrong. We deal further with the question of a reference to the ECJ at the end of this decision.

26.      Lewison J alluded to the question of what constitutes connection with fraud in Livewire (a contra trading case) at [107] when he said that if the connection is "an accounting connection in that the alleged contra trader offsets his input VAT in the dirty chain against output VAT in the clean chain  [then] since the whole VAT system works on the basis of constant offsetting of input tax and output tax the implication ... is that every taxable person could be connected with every other taxable person".

27.      Likewise it seems to us that if a connection arises simply by dealing in particular goods which have at some time been used in a fraud, then, once so used, the goods could never become “clean”. These examples suggest that a narrower meaning of connection was intended by the ECJ.

28.     In the same paragraph Lewison J said:

"…there is an evident factual difficulty in proving a connection with fraud in the case of contra trading where the contra trading is not part of an overall scheme to defraud the revenue. …the problem in real life is that there is no logical connection between the clean and dirty chains.",

He added; “indeed it seems to me that the whole concept of contra trading …necessarily assumes that [the clean and dirty chains are part of an overall scheme to defraud the revenue]. But that assertion is the assertion of a factual conclusion which HMRC is required to prove on the facts of individual cases”

29.     We understand the appellant’s preference to be for this test: that in alleged contra trading there is a connection only if the two chains are part of an overall scheme to defraud the revenue.

30.     In Blue Sphere Global the Chancellor stressed the importance of context[44]:

“The nature of any particular necessary connection depends on its context, for example electrical, familial, physical or logical. The relevant context in this case is the scheme for charging and recovering VAT in the member states of the EU. The process of offsetting inputs against outputs in a particular period and accounting for the difference to the relevant revenue authority can connect two or more transactions or chains of transactions in which there is one common party whether or not the commodity sold is the same. If there is a connection in that sense it matters not which transaction or chain came first."

Thus the Chancellor held that the mere contiguity of the offset of inputs and outputs was enough to constitute connection.

31.     In Kittel the ECJ said [56 and 57]:

“In the same way a taxable person who knew or should have known that, by his purchase, he was taking part in a transaction connected with fraudulent evasion of VAT must, for the purposes of the Sixth Directive be regarded as a participant in that fraud ... That is because in such a situation the taxable person aids the perpetrators of the fraud and becomes their accomplice."

32.     We note two matters. First that the ECJ does not refer to the fraud of the  seller, merely to fraudulent evasion of VAT. Although the case was in the context of evasion by the seller it seems clear to us that the ECJ did not limit the principle to evasion by the seller.

33.     Second, we must consider the context of the use of “connection”. As the Chancellor said in Blue Sphere the meaning of connection depends on its context. The context in which the ECJ uses it indicates to us that its meaning is to be found in the context of whether participation by purchase aids the fraud. If knowledge of connection to fraud makes a person a participant in the fraud because he thereby knowingly assists a fraud, the connection with his purchase must therefore be something which assists the fraud. Such a participation will necessarily require the offsetting of any inputs and outputs referred to in Blue Sphere and is not limited to fraud by the seller. 

34.     If that approach is right then the seeming anomalies referred to above are avoided. Every taxpayer does not become connected with almost every other, and the passage of goods from one trader to another connects the transactions only if it assists a fraud. Where goods pass on the same day down a chain of back to back purchases and sales each of those transactions may assist a fraud by a person at the start of the chain because they provide the funds for the sales which give rise to the fraud, but if the game of pass the parcel stops and goods languish in a warehouse for months before they are sold on, the fraud may have been assisted by the warehousing trader’s purchase but may not be assisted by his later sale. 

35.     In the context of the allegations made by HMRC in this appeal the question is: if (i) Softlink fraudulently defaulted, and (ii) if Klick (or someone managing Klick) arranged sales, the output tax on which would be set against the input tax on purchases which would be made in the future from Softlink, then is a back to back purchase of these goods from Klick, or from another person in a back to back chain after Klick, connected with fraud?

36.     It is so connected on the Blue Sphere test through the offsetting of the inputs and outputs by Klick and the other chain members. It would also be connected on the test outlined above: the purchase of goods in such a chain from Klick assists Klick’s organisation of the setting off of  the input and the output VAT by providing a market for Klick’s sales and funds for its operations; and it assists in the fraud by Softlink by providing, through Klick’s organisation, the funds and the purchases which drive the fraud. Thus on that test such a purchase is connected with Klick’s fraudulent arrangement and with Softlink’s fraudulent default. The questions for us will be: whether as a matter of fact Klick did make such arrangements, whether a back to back chain led to Fonecomp, and whether Softlink did in fact fraudulently evade VAT.

 (8) Knowledge of connection

37.     In our view, the knowledge required is not of the specifics of any particular fraud, but knowledge that by purchasing the taxpayer would be participating in some VAT fraud.

38.     We start by noting that in Kittel the ECJ does not refer to relevant knowledge of a particular fraud: it speaks of knowledge that the taxpayer "was taking part in a transaction connected with fraudulent evasion". The test propounded is that there was knowledge of connection to some fraudulent evasion and not to a particular fraud

39.     In Blue Sphere, the Chancellor said (we have used E to refer to the exporter, C to the contra trader, and D for the defaulter):

 [53]: “…The contention is that [E] ought to have known of the connection, through [C], between its transactions and the fraudulent evasion of VAT by the defaulting traders in the dirty chain. This formulation involves two separate questions, knowledge of the connection and knowledge that the connection was with the fraudulent evasion of VAT. Clearly [E] would have known that its transactions would, for the purposes of VAT, be connected with other transactions with which [C] was concerned in the sense that [C’s] output tax, paid by [E], would have to be set against input tax payable by [C] in respect of other transactions. But that is not enough. HMRC must also prove that [E] ought to have known that those other transactions involved the fraudulent evasion of VAT.

“[54] The tribunal rejected any allegation of conspiracy involving [E] or [C]. It rejected the suggestion that [E] had been manipulated. It acquitted [C] of fraud. If [C] did not know of the fraud when it happened and was not party to any arrangement that it should happen, how could [C] have known of any fraud before it happened? No amount of due diligence undertaken in respect of [C]… could have revealed it. And if [E] could not have known, how could there be circumstances in which it could properly be concluded that [E] ought to have known?

“[55] In my view it is an inescapable consequence of contra trading that for HMRC to refuse a claim by E it must be in a position to prove that C was a party to conspiracy also involving [D]. Although the fact that C is party to both the clean chain with E and the dirty chain with [D] constitutes sufficient connection it is not enough to show that E ought to have known of the fraudulent evasion of VAT involved in the subsequent dirty chain. At the time he entered into the clean chain there was no such dirty chain of which he could have known, nor was the occurrence of such dirty chain inevitable and the sense of being preplanned."

40.     In Livewire at [103] Lewison J, in relation to contra trading, said:

"…it must be established that the taxable person knew or should have known of a connection between his own transaction and at least one of those frauds [that is the failure to account of VAT by the defaulter, and the dishonest cover-up of that fraud by the contra trader]. I do not consider that it is necessary that he knew or should have known of the connection between his own transaction and both of those frauds.”

41.     In Brayfal Lewison J said at[19]:

“The essence of contra trading is that transactions in the clean chain are used to mask transactions in the dirty chain. There is no fraud in the clean chain. The dirty chain is where the fraud takes place. Accordingly in order for a trader in the clean chain to know or have the means of knowledge that his transaction is connected with fraud, he must either know or have the means of knowledge that the contra trader is a fraudster; or he must know or have the means of knowledge of the fraud in the dirty chain.”

42.     These comments could be taken to suggest that the knowledge required is of the specific  activity of the contra trader or the fraudster.

43.     But in Megtian Limited v HMRC [2010] EW HC 18 (Ch), Briggs J at [37] and [38] said:

"In my judgement there are likely to be many cases in which a participant in a sophisticated fraud is shown to have actual or blind-eye knowledge that the transaction in which he is participating is connected with that fraud, without knowing, for example, whether his chain is a clean or dirty chain, whether contra trading is necessarily involved at all, or whether the fraud has at its heart merely a dishonest intention to abscond without paying tax, or that intention plus one or more multifarious means of achieving a cover-up while the absconding takes place.

"Similarly, I consider that there are likely to be many cases in which facts about the transaction known to the broker [the exporter] are sufficient to enable it to be said that the broker ought to have known that his transaction was connected with a tax fraud, without it having to be, or even being possible for it to be, demonstrated precisely which aspects of a sophisticated multifaceted fraud he would have discovered, had he made reasonable inquiries. In my judgement, sophisticated frauds in the real world are not invariably susceptible, as a matter of law, to being carved up into self-contained boxes even though, on the facts of particular cases, including Livewire, that may be an appropriate basis for analysis."

44.     We agree with the tribunal in Ixes (UK) Ltd MAN07/0992 when it said [28] “We respectfully agree [with Briggs J], and we do not think either Lewison J or the Chancellor intended to say anything else. It may well be that the evidential problems facing HMRC in a contra trading case are greater, as Lewison J indicated, but greater evidential difficulties do not affect the underlying test, which remains that described by Moses LJM in Mobilx. All that is required is that the trader knew or should have known that his transaction was connected with fraud. There is nothing in what Moses LJ said which supports the proposition that he must know, or  have the means of knowing, any of the detail of the fraud."

 (9) Does HMRC need to plead conspiracy?

45.     Mr Young argued (at an interlocutory hearing)  that the effect of the  Chancellor’s statement at [55] in Blue Sphere, quoted above at para 39,was that HMRC could only succeed if they expressly alleged that there was a conspiracy for the evasion of VAT  to which the Appellant was a party.

46.     We do not agree.

47.     In Brayful  Lewison J said:

“As the Chancellor pertinently asked in  Blue Sphere… how can a trader who is not part of a conspiracy know of a fraud before it happens?” and then later [17]: “…since the dirty chain was created after the clean chain actual knowledge and conspiracy are likely to be interchangeable concepts.”

48.     Mr Young argued that in order to succeed HMRC had to allege conspiracy and said that they had not.

49.     Mr Cunningham says:

(1)        HMRC’s statement of case alleges knowledge of a fraudulent scheme and dishonest participation in it.

(2)        All that HMRC needed to allege were the requirements of Kittel. The ECJ did not require there to be conspiracy; when at[56] it said:

“In the same way, a taxable person who knew or should have known that, by his purchase, he was taking part in a transaction connected with fraudulent evasion of VAT must, for the purposes of the Sixth Directive, be regarded as a participant in that fraud, irrespective of whether or not he profited by the resale of the goods.”.

This was a deeming provision, not a condition for the operation of the doctrine. There was no requirement for the taxpayer to be a conspirator, no requirement that the taxpayer be dishonest; all that was required was that he knew or had the means of knowledge of the connection to fraud.

(3)        In Megtian  Briggs J made no reference to conspiracy even though that too was a contra trading case. At 41 he said:

“A person who knows that a transaction in which he participates is connected with fraudulent tax evasion is a participant in that fraud. That person has a dishonest state of mind. By contrast a person who merely ought to  have known of the relevant connection is not dishonest, but has a state of mind broadly equivalent to negligence.”

 In the should-have-known-case the taxpayer could not be a conspirator.

(4)        Mobilx contained no requirement for conspiracy. It said that the test should not be over-refined. The Court of Appeal explained that Kittel laid down  an objective test. All that was necessary was to plead those matters which were relevant to that test.

(5)        It may well have been true that on the facts of Brayful a plea of knowledge of connection to fraud was equivalent to conspiracy, but that need not always be the case, and Lewison J did not say it was, merely that it was likely.

50.     Conspiracy, as we understand it, is an agreement to pursue a course of conduct which will result in the commission of an offence or an unlawful act.  It seems to us possible that a trader, by agreeing to buy goods which he knows stem from someone who will commit a VAT fraud, is not agreeing to a course of conduct which will involve an offence in relation at least to that VAT fraud: he may, particularly where he does not buy directly from the fraudster, be simply agreeing to buy goods but to nothing else. Thus it seems to us that even in the case where a person knows of the connection to fraud, he may not be conspiring. Thus the Kittel  test is potentially broader than the conspiracy argument.

51.     We accept that a serious allegation of dishonesty or involvement in fraud must be fairly and squarely pleaded and that a person about whom it is made, if a witness, should have it put to him fairly and squarely and have the opportunity to refute it. But if what has been pleaded and what is put to the witness clearly amounts to conspiracy, then it matters not that that word has not been used. Thus even if knowledge of connection to fraud necessarily meant involvement in conspiracy it seems to us that there is no need for that word to be used so long as the allegations are clear and, where serious, squarely put.

52.     Further if what has been pleaded satisfies the Kittel test but does not amount to a conspiracy, then HMRC are entitled to succeed if the facts they rely  upon are proven.

53.     As a result we do not  consider that any part of HMRC’s case on Kittel  fails by reason of a failure to plead conspiracy.

(10) The Burden and the standard of proof.

54.     The burden of proving VAT fraud, connection to fraud, and knowledge or means or knowledge rests on HMRC. The standard of proof is on the normal balance of probabilities.

(11) The relevance of circumstances surrounding a particular transaction

55.     In Mobilx Moses LJ [83] set out with approval the part of the judgement of Christopher Clarke J in Red 12 where he said that examining individual transactions on their own merits does not however require them to be regarded in isolation without regard to their attendant circumstances and conduct.

56.     The appellant says in its skeleton that it is wrong in law to have regard to a scheme as a whole because each transaction must be regarded on its own merits, viewed objectively, individually and per se, to determine whether VAT becomes chargeable on the transaction. For this reason it says it is not appropriate (nor even permissible) for the tribunal to have regard to all the surrounding circumstances. “Adopting a transaction by transaction approach is that which is expressly required under the EC VAT legislation and jurisprudence. To the extent that domestic courts, whether at first instance or appellate, can be read as endorsing a broader approach (as the respondents appear to urge on the basis of Red 12) it is clear that such a course is impermissible”. It says that the character of the appellant’s transactions cannot be altered by earlier or later or subsequent events: it is not permissible to determine the character of a transaction by reference to the purpose or results of the whole chain.

57.     In our view the question to be determined is whether a particular transaction fails theKittel tests. In determining whether a person has knowledge or means of knowledge, factors other than the features of a single transaction must be relevant but the object must be to determine by reference to all the available evidence the nature of that single transaction. That is how we understand the judgement of the Court of Appeal and the High Court in Red 12.

(12) Double Recovery

58.     As will later appear HMRC assessed Klick on the basis that it was not entitled to input tax credit in respect of the purchases made from Softlink. The assessment was made on the basis of the Kittel principle: HMRC alleged Klick knew its transactions were connected to Softlink’s default.  The question therefore arose as to whether that assessment precluded the denial of input tax to the appellant whose input tax represented some of the very input tax which had been denied to Klick.

59.     In Bulves v Bulgaria 3991/03 the ECHR found that the denial of an input tax credit to a company because of delay in VAT compliance by its supplier was an unjustified interference with its property rights. One of the considerations the Court had in mind was that, in refusing the company's right to deduct, the state recovered VAT effectively both from the company and the supplier - two payments of VAT for the same supply. That, the court found, did not seem to be justified by the need to secure payment of taxes. The question arose as to whether the same principle applied here.

60.     We agree with HMRC that it does not. The ECJ in Kittel propounded an objective test for the question of whether a person was entitled to input VAT credit on his purchase. VAT is a transaction by transaction tax. The question is whether input tax is deductible on the particular transaction by reference to the circumstances of that transaction. If an objective test is satisfied input VAT is deductible. If it is not it is not. The actions of HMRC in relation to other persons in the chain are not relevant to that transaction.

61.     In particular Kittel does not impose a penalty. Rather it limits and circumscribes the rights to deduct input VAT. Once it is established that a taxpayer has participated in fraudulent evasion, the non existence of his right to that input VAT is not a penalty.

 

IV. The Evidence

62.   We heard oral evidence from Diane Worbey the officer of HMRC who had been responsible for liaising with Fonecomp in the period from August 2006; from Dean Jones, an officer of HMRC who had investigated transactions emanating from and connected to transactions undertaken by Klick; Russell Hall, an officer from HMRC who had investigated data obtained from computer servers belonging to First Curacao International Bank NA (“FCIB “) relating to the bank accounts of certain traders, including the Appellant; and Yasin Sharif, the sales administrator of Fonecomp.

63.   We received in evidence witness statements from a number of other HMRC officers the import of which had not been expressly disputed by the appellant.

64.   When the substantive hearing had been expected to encompass the appeals of 13 appellants, it had been listed for some 12 weeks. The documentation for that combined hearing gave rise to a large number of lever arch files. The reduction in the number of appellants meant that fewer documents were referred to. Although all the original evidence was available to the tribunal at the hearing, we were referred only to a dozen or so files.

65.   Many of HMRC's witnesses made statements of opinion in their evidence drawing their own inferences from the facts they related. Except in the case of certain aspects of Mr. Hall's evidence in relation to the FCIB evidence, we have ignored these opinions; our task was to make our own judgement from the underlying facts on these matters.

66.   We set out our findings of fact from this evidence in seven sections: (1) deals with the nature of Fonecomp’s business in relation to these transactions; (2) contains findings in relation to the two Deals including some of the FCIB evidence; (3) deals with the default of Softlink Ltd to which HMRC says the Deals are connected; (4) deals with the Klick Scheme including the FCIB evidence; (5) deals with certain aspects of the chain of supply to Fonecomp; (6) deals with the question of whether the Deals were connected to Softlink’s default; and (7) deals with evidence of statements made by the director of PDA Stuff. Under each heading we set out the facts we have found and in some cases include in a commentary the inferences we have drawn. Those inferences and findings we draw together  in the section headed “ Discussion” towards  the end of this decision.

V. Our Findings of Fact

(1) Fonecomp: its business.

67.   This section is divided into a series of topics relating to Fonecomp’s business. These topics are; (a) an introduction, (b) Fonecomp’s knowledge of MTIC fraud, (c) the due diligence it undertook on its trading partners; (d) its FCIB account.

(1) (a) Introduction

68.   Fonecomp was incorporated in 1998 and registered for VAT in the same year. Since about 2001 its sole director has been Ahmed Ibrahim and its company secretary Ghazala Hussain. Its shares are held equally by Mr. Ibrahim, Miss Hussain, and Iqbal Patel.

69.   The company has traded in mobile phones since 2000. It operates from 92 Leytonstone Road, Stratford London and shares these premises with a retail phone business run by Miss Hussain under the name Fone Co.

70.   Apart from a period between August 2004 and November 2005 Mr. Sharif has been employed by the company since about 2001. His duties have included buying and selling mobile phones and the associated administration.

71.   Mr. Ibrahim played an active part in the company's business, regularly attending its premises, but he did not play an active role in dealings with HMRC. Mr. Patel dealt with HMRC in many of their early visits to the company and was described in 2006 as its sales manager.

72.   The company makes monthly VAT returns and from March 2000 until July 2006 each return bar one showed a VAT reclaim. The average monthly reclaim in 2001 was £44,000; the reclaims grew fairly steadily thereafter, and in 2005 the average reclaim was £251,000; each reclaim from February 2006 to July 2006 was between £400,000 and £465,000. The pattern of these reclaims suggested that, in the period during which Mr. Sharif had not been employed by the company (in 2004 and 2005), the trade of the company had continued unabated.

73.   It was plain that a significant part of the company's business since 2001 was the purchase of mobile phones in the UK and their export.

(1)(b) Knowledge of MTIC fraud

74.   In December 2001 HMRC wrote to the company explaining that a number of cases of "significant evasion of VAT" had been identified in mobile phone dealing. The letter required the company to provide additional information to aid the "detection and prevention of serious fraud".

75.   In December 2002 HMRC wrote to the company explaining that some of its suppliers of mobile phones had failed to account for VAT, and requiring the company to check that its suppliers and customers were genuine.

76.   In May 2003 the company, with advice from its lawyers, replied to an HMRC questionnaire on VAT strategy in connection with the issue in August 2003 of notice 726 (see below). In reply to a question about the taking of reasonable commercial steps to avoid connection to fraud the company made a comment that HMRC should "focus their attention on catching the missing trader by deploying their officers under cover ...".

77.   Mr. Sharif explained that at this time they knew that there were VAT problems in the mobile phone industry with missing traders and that, although it was possible that there could be a problem with one of the company's suppliers, the problem was likely to be with persons beyond its suppliers in the supply chain.

78.   In November 2003 HMRC told the company that it would be undertaking extended verification of its VAT returns for October 2003. It appears that the repayment of VAT was delayed for this quarter but was eventually made.

79.   In August 2004 HMRC wrote to the company indicating that it disputed part of its input tax claim for December 2003. Input tax credit of £125,000 was refused. This sum related to transactions which HMRC said were part of chains of circular movements of goods which contained a defaulting trader. An appeal was made to the VAT and Duties tribunal. The appeal was stayed pending the decision of the ECJ in  Optigen Ltd v Customs and Excise Commissioners [2006] ECR 1-483. Following the decision of the ECJ in that case HMRC made payment of the input tax claim in March 2006. HMRC's review checklist in connection with this repayment contains a statement: "no strong evidence to support an assertion that Fonecomp knew or had a means of knowledge.".

80.   HMRC published Notice 726 in August 2003. Fonecomp was aware of HMRC's Notice 726 from late 2003. Notice 726 concerns the operation of section 77A VAT Act 1994 and the ability that section gives to HMRC to impose liability on the recipient of a supply for the VAT payable by a supplier if the recipient knew, or had reasonable grounds to suspect, that its supplier or any previous supplier had defaulted on VAT payment.

81.   Notice 726 said that joint and several liability had been introduced to help in HMRC's fraud strategy and explained that MTIC fraud involved an importer of goods selling the goods and defaulting on its VAT liability. It explained that the fraud generally involved wholesale export of goods from the UK. It stated that the fraud was widespread and cost some £2 billion in 2001/02. The notice implicitly accepted that there could be a legitimate wholesale trade in  mobile phones.

82.   Commentary

83.   It seems clear to us that at the time of the Deals Fonecomp knew that there was VAT fraud in the wholesale mobile phone industry and that this fraud involved an importer of phones defaulting on its VAT liability having sold the phones on to another UK trader. It also knew that the fraud could be, and was, fed by the sale of the phones along a chain of UK entities and, in many cases, by their ultimate export by a person in its position. It therefore knew that there was a possibility that its purchases could be connected to such a fraud committed by a trader which was not its immediate supplier. It knew that the fraud was widespread and involved very large amounts of money, and that HMRC were seriously concerned about it.

84.   In our estimation the information made available to it and noted above was not sufficient for Fonecomp to be able to conclude that every transaction it undertook was connected to fraud, but it was sufficient to enable it to conclude that such fraud could be present, and to consider it in the forefront of any explanations for unusual factors in its transactional trade.

85.   Mr Young notes that notice 726 suggests that part of the template for fraudulent activity is that goods are offered at a price less than market value; HMRC had not suggested that Fonecomp’s purchases were at less than market value and had never attempted to apply section 77A. No specific criticisms of its operations had been made by Mrs Worbey’s predecessor. It was entitled to conclude that its deals were untainted.

86.   We do not believe that HMRC’s failure to apply section 77A should have given Fonecomp any comfort. The operation of section 77A relies on similar types of factual conclusions to the challenges mounted in the Kittel refusal: one would expect delay before such a challenge. Although para 3.3 of notice 726 says that “In order for the fraud to be perpetrated the price has to be cut within the supply chain”, that does not mean that every purchase in a fraudulent chain will be below market value, and if the appellant knows that there was a chain of supply leading to its supply (as we later find the appellant did) , that should raise the question of whether earlier sales in that chain were at less than market value, and thereby raise a concern about fraud.

(1)(c) due diligence

87.   No due diligence was undertaken by Fonecomp on its freight forwarders , although it had used Paul’s Freight for some time. Since 2000 the company used the facilities offered by HMRC (“Redhill checks”) to verify the VAT numbers of its counterparties. In the period 2000 until 2006 it received replies from HMRC on six occasions indicating that the VAT registration number could not be confirmed, although in respect of two of those occasions the registrations were later reinstated. Fonecomp did undertake some other checks on its customers and suppliers.

88.   On 11 September 2006 there was a meeting between HMRC and Fonecomp to discuss its outstanding VAT claims for June and July 2006. Andy Monk and Diane Worbey attended for HMRC, and Mr. Ibrahim, Mr. Patel, Mr. Sharif and Mr Abrahim for Fonecomp together with Mr. Fielder from  the due diligence company Veracis.

89.   At that meeting HMRC were told that prior to trading with a counterparty Fonecomp would obtain a copy of its VAT certificate, undertake a Companies House check, a web check and a credit check with First Report which gave details of its registered office and sometimes of its directors. HMRC were told that someone from Fonecomp always visited customers and suppliers (or got Veracis to do so) at some point.

90.   Due diligence was discussed at that meeting. Mr. Young asked Mrs Worbey whether HMRC had asked to see the company’s due diligence materials. She avoided answering that question directly but repeated that due diligence was discussed. We conclude that no express request was made to see the due diligence material kept by the company.

91.   Mr. Sharif told us that due diligence was discussed but that HMRC did not expressly ask for Fonecomp's material. He said that the folder was on the shelf and could have been supplied had it been asked for.

92.   Mr. Sharif appended to his witness statement a number of documents which he said were due diligence materials obtained by Fonecomp on TM Global, Axess, and PDA Stuff (its counterparties in the Deals), and also on a company called 3G Trade SA (a company unrelated to the Deals).

93.   The materials in relation to PDA Stuff consisted of:

(1)  a note of a visit in a report by Mr. Ibrahim dated 4 May 2006

(2)  a copy of page of a UK passport of the managing director ;

(3)  a copy of a 2005 phone bill;

(4)  notepaper;

(5)  a report from First Report

(6)  a trading application;

(7)  a VAT certificate;

(8)  bank details; and

 an incorporation certificate.

94.   The materials in relation to Axess consisted of:

(1)  a European VAT number validation;

(2)  an introductory letter and bank details

(3)  the trade application form, uncompleted;

(4)  a copy of a page of a Danish passport of the chief executive officer; and

(5)  a number of documents in Danish which looked as if they related to its incorporation and VAT registration.

95.   The documents for TM Global appeared similar to those for PDA Stuff, but without a passport and phone bill, and with the addition of two due diligence reports from Veracis Ltd (“Veracis”). But included in the materials was a certificate of incorporation and a ‘standard report’ from First Report for a company called TM GB Ltd. There was no First Report or Certificate of Incorporation for TM Global; but it seems that the two companies shared the same registered address and single director.

96.   The 3G material also contained a report from Veracis dated 15 December 2005.

97.   The Veracis reports were not addressed to any particular recipient. No copy of a covering letter was provided.

98.   HMRC did not accept that the Veracis reports in relation to TM Global had been prepared for Fonecomp or that they were in Fonecomp's possession before the times of the Deals. HMRC say they were obtained later by Fonecomp to bolster its case. They say that had Fonecomp had them at the time of the September meeting (when Mr. Fielder from Veracis was present) they would have been shown to HMRC at that time. They also say that this is evident from the fact that the  two TM Global reports are missing enclosures which were listed and referred to in the body of the report; whereas the 3G report had appended to it enclosures referred to in its main text.

99.   In Fonecomp's documentation for each of the Deals was an A4 sheet which contained a checklist. The form had a number of boxes with typeset entries such as "purchase order" and "VAT verification". Mr. Sharif explained that this was a photocopied form. There was one box in the form of these Deals which appeared to have read "Report by Veracis", but the form had in each case been amended in manuscript: the box had been divided in two and, in the newly formed box, "declaration signed + stamped" was written in manuscript. These words were identical on both forms and we concluded (and Mr. Sharif agreed) that these words had been added to an original form which had then been photocopied and used for the Deals. But the other half of the original box still contained the typescript words "Report by Veracis". In each case, however, these words had been crossed out and replaced by "Credit Cheque Yes/no”. But the writing of these words and the manner of the crossing out of "Report by Veracis" was different (albeit slightly) in each of the forms. (Thus on the form for 1765 the crossing out contained seven or eight peaks which surmounted the typescript, but the form for 1760 had only one or two such peaks). We concluded that the crossing out had not been done on the form before it was photocopied for these Deals, but afterwards.

100.                  These forms so amended provided equivocal evidence in relation to the question of whether Fonecomp had the Veracis reports on TM Global in its possession at the time of the deals. That was because

(1)  the original inclusion of the reference to the reports suggests that Fonecomp was, at the stage it used the form and did the Deals, receiving such reports; but

(2)  the crossing out suggested that it had not in fact received such a report for those, although

(3)  the fact that very few of the boxes had actually been filled in suggested that the form was not in fact a record of what happened but merely an aide memoir: in fact the only entries made on the form were the invoice numbers and short details of the phones agreed to be sold.

101.                  Commentary.

102.                  We concluded from all this that Fonecomp did some checks on its customers and suppliers before dealing with them. Those checks included verification of their existence and VAT registration and some checks on whether those who claimed to act for the relevant companies were indeed the officers of those companies. We were not convinced that Fonecomp had received the Veracis reports on TM Global before the Deals, but neither were we persuaded that they had been included in Mr. Sharif's witness statement with the intention of misleading the tribunal.

103.                  However, it was clear from the report of the September 2006 meeting, from the limited checks on the creditworthiness of its counterparties, and from Mr. Sharif's evidence, that the creditworthiness of its counterparties was not a real concern of the company. So far as its customers were concerned, the company was satisfied that it had no exposure to their creditworthiness so long as the goods were not released to its customer until payment was made; and in relation to its suppliers the company did not pay until the goods were released to it; indeed Mr. Sharif said that credit was taken from suppliers.

104.                  That lack of concern, however, betrayed either a lack of interest in the question as to whether the supplier could reasonably be expected to have the resources to participate in these Deals (and, if not, how they managed to do so); or an understanding that its counterparty was also dealing on back to back terms.

105.                  Mr. Cunningham asked Mr. Sharif about the credit standing of TM Global. He referred Mr Sharif to the First Report on TM GB Ltd (which it will be recalled was a different company from TM Global) which indicated that a little while previously it had negative net assets. Mr Cunningham’s question suggested that he thought that the report was on TM Global; he asked: how could it have bought the phones which it sold to a Fonecomp? Mr. Sharif (who seemed to share the same view of the report) replied, "they hadn't bought the product because the goods weren't under their title, the goods were on hold. That is what back-to-back trading is." In other words Fonecomp knew that its supplier was buying on back-to-back terms from someone else to sell to it. It seems likely that the same was true of the purchase from PDA Stuff. In a later response Mr Sharif suggested that his reference to back-to-back chains was limited to Fonecomp’s transactions. That latter restriction did not make sense in the context of his earlier answer. We preferred that earlier answer.

106.                  The TM GB Ltd documents were exhibits to Mr Sharif’s witness statement, described there as part of Fonecomp’s due diligence on TM Global Ltd. We make two comments. First, that the confusion of the two companies does not affect the import of Mr Sharif’s statement that they knew that their suppliers were dealing back to back, and, second, that this confusion demonstrates, at best, some lack of rigour in Fonecomp’s due diligence.

(1)(d) FCIB Accounts and computer arrangements.

107.                  Fonecomp maintained an account with FCIB during 2006. This account was operated by Mr Ibrahim through a computer in Fonecomp’s office. Axess, PDA Stuff and TM Global also maintained FCIB accounts. Payment was made for the Deals through these accounts.

(2) The Deals

108.                  We described the Deals in outline in paragraphs 2 and 3. In this section we address (a) the specification of the phones in the documentation; (b) Fonecomp’s inspection of the goods; (c) terms of trade; (d) the nature of the phones dealt in; and (e) whither the phones were shipped.

(2)(a) Specification

109.                  In Deal 1 Fonecomp provided a purchase order (numbered 1760) to PDA Stuff. The order was for 1100 Nokia 9300i and 1200 Nokia 8801 on 11 July 2006. Underneath that part of the order setting out these items and the price was the following specification in the middle of the page:

(1)  colour grey/silver

(2)  battery type  original lith-ion

(3)  operations manual C. Euro spec

(4)  menu languages: C Euro spec

(5)  software version: latest

(6)  Manufacturer's warranty: 12 month

(7)  others: no label, Sim-free, two pin plug,

(8)  delivery terms: FOB London

(9)  delivery address: Paul's Freight Services Ltd, London

(10)                   delivery time: ASAP

110.                  Mr Sharif told us that this middle part of the purchase order was part of a template for Fonecomp's purchase orders. It could be changed, but it would remain pretty much the same in every deal although sometimes the colour and the delivery address might change.

111.                  PDA Stuff signed and returned the purchase order on the same day and invoiced Fonecomp for:

"1100 Nokia 9300i  mobile phones sim free and boxed

1200 Nokia 9500 mobile phones sim free and boxed" [Note: “9500”]

at the prices set out in the purchase order from Fonecomp for the Nokia 9300i and Nokia 8801. It seems clear that "9500" was a typographical error for "8801" because another invoice on the same date from PDA Stuff makes the correction.

112.                  On the same day Axess provided two purchase orders to Fonecomp: one was for 1100 "Nokia 9300i", the other for 1200 "Nokia 8801". There was no further specification of the phones in those orders. On the same day Fonecomp provided both a pro forma invoice and an invoice to Axess for these quantities of these phones. The pro forma invoice contained the template specification panel set out above, save that the delivery address was "Pro-log, Sogaris, Rungis, France”; the invoice recorded the specification of the phones being sold less extensively:

"original Sim free

Central Euro-spec".

113.                  In Deal 2 Fonecomp provided a purchase order to TM Global on 26 July 2006 for 900 Nokia N 80 and 500 Nokia N 91. The purchase order had the same middle specification section as had the purchase order in Deal 1. TM Global signed and returned the purchase order on the same day and sent a pro forma invoice and then invoiced Fonecomp for these phones on that day; its pro forma invoice and its invoice contain the same limited specification, "boxed", and referred to Fonecomp's purchase order as being "verbal".

114.                  On the same day Axess sent purchase orders to Fonecomp for these numbers of these phones. There was no specification on these purchase orders.

115.                  On the same day Fonecomp sent Axess a pro forma invoice which contained the same template specifications as on its purchase order to TM Global, and an invoice with the less extensive specification which is noted above in relation to its invoice in Deal 1.

116.                  Commentary

117.                  Mr. Cunningham says that these documents show surprisingly little interest by the parties in the precise specifications of the phones being sold. Mr. Sharif indicated that these contractual documents would have been supplemented by oral telephone communications between the parties.

118.                  Both Deal 1 and Deal 2 were for a very large amount of money - much more than the cost of the average house. It seems clear to us that the precise specification of the assets being acquired and sold could easily affect their value. Whilst the purchase order provided by Fonecomp to its supplier provided an amount of detailed specification (in the template section in the middle of the order), the documentation for the purchase by Axess was less extensive. Axess simply ordered a number of a particular type of phone. It was only when Fonecomp sent Axess its pro forma invoice that the detailed specification could be seen. By that time Axess must have placed its order.

119.                  We find it difficult to believe that details of the specification would have been agreed orally in all cases without some form of documentation. We draw the conclusion that Fonecomp’s suppliers and Axess in both deals were not particularly interested in the detailed specification of the phones and that Fonecomp knew this.

(2)(b) Inspection

120.                  In both Deals the phones purchased and sold by Fonecomp were held by the UK freight forwarder, Paul's Freight. The freight forwarder was some two hours drive away from Fonecomp's offices and no one from Fonecomp went to check that the phones which were held there from its suppliers and for Fonecomp were of the specific type and number agreed to be purchased (and sold). Fonecomp did however obtain some comfort from Paul's Freight as related below.

121.                  The September 2006 meeting note records that Fonecomp used Paul's Freight to carry out inspections of the phones in which Fonecomp was trading until about March 2006. Those inspections related to the detail of the phones as well as their model and number. However after March 2006 Paul's Freight ceased to carry out inspections. But Fonecomp did ask it to count the number of phones in relation to both Deals and Paul’s Freight did send Fonecomp a fax setting out the number of phones of each type which had been received in the warehouse and which were sent to Paris. This document however did not provide any details of the specifications of the phones other than their model numbers.

122.                  Mr. Sharif said that Fonecomp had already asked Paul's Freight to check the contents of a few of the boxes at the same time as it did the counting exercise, and that Paul's Freight had notified Fonecomp by telephone of the results of those sample inspections. There was no note or written confirmation of this oral exchange. We declined to find that there was any form of inspection which would have given Fonecomp any real comfort that the specification of the phones in the possession of Paul's freight matched that of those it had agreed to buy or sell.

123.                  Commentary

124.                  We find Fonecomp's lack of need for reliable assurance that the goods received and sent out via Paul's Freight were of the specification it had agreed, troubling. It seems to us likely that Fonecomp knew that it would be paid whatever the precise specification was and that is the only explanation for its apparent indifference. That in turn suggests that Fonecomp knew that what mattered was that there was a sale, rather than a sale of specific goods, and that the purchaser already knew what it was getting.

(2)(c) Terms of Trade

125.                  Both Fonecomp's pro forma invoices and its invoices indicated that "goods remain the property of Fonecomp until ... paid in full". Neither TM Global's invoice and pro forma invoice to Fonecomp nor PDA Stuff’s invoice made any statement about payment or title in the phones sold (although a supplier declaration by PDA Stuff indicated that the goods were bought with "free title"); neither did Axess' purchase order.

126.                  In Deal 1 Fonecomp’s purchase order indicated that payment terms were "100% TT after inspection". We understand “TT” to mean by telegraphic transfer. PDA Stuff's invoice to Fonecomp contained no provision about time for payment. Axess' purchase order to Fonecomp contained no stipulation as to the date of payment. Fonecomp's pro forma invoice specified payment "100% after inspection TT".

127.                  In each Deal Fonecomp instructed Paul's freight to ship the phones to Pro- Log in France saying "Please do not release the goods until further instructions are received from Fonecomp limited".

128.                  In Deal 1, on 17 July 2006 payment was made by Axess and on the same day Fonecomp wrote to Paul's Freight asking them to inform Pro-Log to release the goods to Axess.

129.                  In Deal 2 it was on 29 July 2006 that Fonecomp wrote to Paul's Freight asking them to inform Pro-Log that the stock should be released to Axess.  Axess had paid Fonecomp on the preceding day (28 July) and Fonecomp paid TM Global on 31 July.

130.                  Mr. Sharif told us that where Fonecomp had agreed to buy goods from a supplier, the supplier would “allocate” them to Fonecomp. That would not enable Fonecomp to remove the goods, but it would enable Fonecomp to give instructions to its freight forwarder to ship the goods on hold to another warehouse. When Fonecomp received payment it would tell Paul's Freight to tell their agent to release the goods to the customer. But he said Pro-log would not release without instructions from Paul's Freight, and Paul's Freight would not release until it had instructions from Fonecomp's suppliers. It was possible he said that the chain went further and that Fonecomp's supplier had only an allocation of goods, and that a further release by the supplier's supplier would, therefore, also be required. Mr. Sharif said, where goods had been allocated to Fonecomp to enable Fonecomp to ship on hold, "that was sufficient to us".

131.                  Mr. Sharif could not say whether there was an agreement between a supplier and the freight forwarder, in circumstances in which the goods were held after they had been shipped on hold, under which the supplier could command the return of the goods.

132.                  Commentary

133.                  There are a number of worrying aspects of this evidence

134.                  First the lack of payment terms in the invoices from Fonecomp's suppliers makes little commercial sense. These deals were for hundreds of thousands of pounds and close attention to when payment would be made would normally be expected.

135.                  Second, even if the supplier was relying on the term in Fonecomp's purchase order that payment would be made on inspection, it was clear that payment had not been made in accordance with that term since the inspection (or piece count) had been done on the day of the deal but payment was made days later.

136.                  Third, whilst Fonecomp imposed a retention of title clause on its customer, it made no provision in its contract with its supplier as regards the transfer of title. It may well have been said that, in the absence of express terms, Fonecomp could rely upon an implied term that title would pass on delivery or the formal notice of release of the goods. But, in a transaction involving large amounts of money, one would expect some consideration to have been given to providing that title would pass when payment was made if not earlier.

137.                  Fourth, as recounted earlier at section (1)(c), Fonecomp knew that its suppliers were back to back trading, and it knew therefore that it was possible that its suppliers would obtain title only on their suppliers releasing the goods to them or upon their making payment. Thus even after Fonecomp paid it was possible that it would not obtain title or even possession of the goods in the form of a release.

138.                  Fifth there was a distinct lack of clarity in relation to the arrangements for allocation and shipping on hold. If the owner of the goods (whether Fonecomp or its supplier) wished to recall them from a freight forwarder with which it had no direct contractual relationship and which might be in another jurisdiction, it was unclear what right it had to do so, and if the goods had been shipped on hold it was not immediately clear that there were terms which ensured that release would be given only after releases had been given earlier in the chain. Fonecomp, and it seems its suppliers, were at risk that goods which had passed to another freight forwarder were beyond their control.

139.                  Sixth, Fonecomp placed a lot of trust in freight forwarders in relation to which it had done no due diligence.

140.                  The lack of terms of payment, or the ignoring of the terms which were express, the lack of clarity about the transfer of title, and the reliance upon freight forwarders of uncertain credit to abide by what can only have been informal arrangements or understandings about the release or the possession of goods have two possible explanations. The first is that Fonecomp was simply careless. The second, and to our mind more likely, is that Fonecomp knew or must have known that it was participating in a transaction whose sole purpose was the movement of a particular consignment of goods along a chain and to a non-UK counterparty in circumstances where until payment was made and flowed back along the chain the transaction could be unwound; what mattered to the parties in this transaction was not precisely what was supplied, but simply that what was expected to pass down the chain passed and when it arrived payment was made. Until payment was made the chain could be unwound.  It was akin to a game of pass the parcel where what mattered was that the parcel was passed on and arrived, but the players were not concerned with what was inside it.

(2)(d) The phones dealt in

141.                  In both Deals the phones were described (see above) as “C Euro Spec…2 pin” phones. In Deal 1 part of the transaction was in 8801s.

(i)  2 pin Euro Spec phones

142.                  Mr Sharif explained that the phones which were the subject of the two deals were “Euro spec” and that this meant that they had 2 pin chargers and had Europe wide warranties rather than UK 3 pin chargers and UK limited warranties. He said that it was fairly easy and cheap to replace 2 pin chargers with 3 pin chargers, and that with such replacement these phones could easily have been sold to UK domestic consumers. Mr Sharif’s experience of doing this conversion came from a time before he joined Fonecomp.

(ii)8801s

143.                  Mr. Jones told us that the Nokia 8801 was a telephone intended for the US market because it operated on the three mobile phone frequencies used in America. Of those three frequencies only two were used in Europe. Although it would therefore work in Europe it would not do so as widely or successfully as it would in the US or as would a phone designed for the European market. Mr Sharif said that the European market counterpart of the 8801 was the 8800.

144.                  Fonecomp’s purchase and sale documentation described the 8801s being purchased and sold as Central European Spec. Mr. Sharif told us that Fonecomp knew that these were American phones but said that it was normal for phones from other parts of the world to be traded in the UK. Mr. Sharif said they had been told by Paul’s Freight, their freight forwarder, that these phones had the round European two pin plugs rather than the American flat two pin plug and that Fonecomp had been told that by Paul's freight at the time of the deal.

145.                  Commentary

146.                  We accept that 2 pin chargers could be replaced relatively cheaply by 3 pin ones, and it seems to us that there is a possibility that someone could import 2 pin phones with the intention of changing the chargers to 3 pin UK chargers and then selling them in the UK; and that this could be an explanation for finding the occasional batch of two pin the phones for sale in the UK.

147.                  But as we have related above, Mr. Sharif explained to us that the middle part of Fonecomp's purchase order was a standard template and that part specified Central European specification and two pin plug adaptors. That indicated to us that it was the normal course for Fonecomp to deal in 2 pin phones, and that the two pin phones in these deals were not unusual but part of Fonecomp's normal trade. In our view the level of trade shows that these were not small consignments intended for conversion to the UK market but phones in the UK whose only destination must have been outside the UK.

148.                  We found the idea that these quantities of phones designed for sale in the American market should find themselves in the UK with chargers which suited European but not UK or American users very odd. We found it surprising that Mr Sharif said that he remembered this aspect of Paul’s Freight’s telephone report on the nature of the phones’ charger pins. We incline to the view that at best Mr Sharif thought that this must have been what Paul’s Freight told him given the description of the phones  in Fonecomp’s paperwork.

149.                  If, as we believe, Fonecomp did not know that the 8801 phones were designed for the American market it should in our view reasonably have asked itself (1) why they were present in the UK, (2) why there was a market for them in Europe, and (3) why they had European two pin charges (if they did). It was plain from Mr. Sharif's answers to these questions that these points had not been considered in any depth, if at all. We can think of no likely explanation other than that they were imported into the UK in order to be exported.

150.                  If Fonecomp did not know that these phones were designed for the American market then: (a) the description of the phones as Central European Spec was probably misleading because the likelihood was that the phones in fact had flat two pin chargers; (b) the question of why such phones were in the UK remains;  and (c) it suggests a degree of lack of care in relation to the subject matter of the Deals which suggests that what mattered was that something passed rather than that particular phones were sold and bought.

151.                  (2)(e) Shipping to Pro-Log in France

152.                  Axess has a Danish address. The phones in both Deals were shipped to Pro-log in France. Mr. Sharif explained why this was in several parts of his oral evidence:

(1)  he said that Pro-Log France was a neutral location;

(2)  this meant that if Axess did not pay Fonecomp, Fonecomp could easily have sold the phones to someone else;

(3)  had they been sent to Copenhagen and Axess had not paid, it would have been difficult to sell to someone in say Germany because it would have cost more to sell (or, we inferred, transport);

(4)  if the goods were in France they were at a more easily accessible freight forwarder for a later sale if Axess defaulted;

(5)  Axess had wanted the goods delivered to Copenhagen but for the above reasons Fonecomp had insisted on Pro-Log.

(6)  Axess had agreed to this. Mr Sharif said "the fact that we shipped them to a neutral location, and he agreed, was that his customer may have also wanted them there as well.".

153.                  Comment

154.                  This made very little sense. First, Mr. Sharif had agreed that the cost of shipping was on the order of £1000, not a huge amount in context. If Fonecomp had shipped to Copenhagen and then had to ship to say, Germany, the cost would have been the same as the cost Axess would incur in shipping from France to Copenhagen - where it was said that initially it wanted the phones. If Axess did not want them in France it was accepting an additional £1000 of transport costs. Axess would have asked for a  price reduction to  compensate it. If it had, would not a reasonable trader had been willing to ship to Copenhagen to avoid a certain £1000 reduction in price, even though taking a risk of a £1000 cost if Axess did not pay?

155.                  Second, if the goods had been kept in the UK rather than sent to France, would that not also have served Fonecomp's purpose? In reply Mr. Sharif says that if Fonecomp had kept the goods in the UK until payment, then Axess would not have known that Fonecomp was going to ship the goods. But Axess would have been in exactly the same position if after payment and shipment of the goods to France, Fonecomp had refused to release them from Pro-log in France.

156.                  These considerations make us doubt that Fonecomp shipped the phones to France as a "neutral jurisdiction". The answer appears to be in Mr. Sharif's comment was that "his customer may have wanted them there as well". Fonecomp must have known this and it seems to us that Mr. Sharif's description of France as a neutral location was an attempt to hide it.

157.                  We conclude that Fonecomp knew that the goods were going to be sold from Pro-log in France by Axess as part of another back-to-back deal. The alternative is that Fonecomp knew that what mattered was the export of goods and for that purpose France was as good as anywhere and that it had to play some part in a scheme.

(2)(f) FCIB evidence in relation to the Deals

158.                  Mr. Hall gave evidence of information derived from the computer records of FCIB. These included the bank accounts in the period June to August 2006 of companies identified by HMRC as participants in the Klick Scheme including Axess, Fonecomp, PDA Stuff, and TM Global. We accept that the data he produced came from FCIB's computer records. These records were contained on two servers, one in Holland and the other in Paris. The data on the Paris server replicated that on the Dutch server but contained additional information about individual transactions. He withdrew some of the inferences he had drawn from the data on the Dutch server after additional information had become available from the Paris server.

159.                  Mr. Hall's evidence included diagrams showing the money flows arising from the deals he had examined. These were based in part on inferences he had drawn from his examination of the bank accounts. Whilst it was possible that some of the conclusions Mr. Hall drew were wrong -and that what appeared to be a connection between two payments was simply an innocent coincidence, we thought that overall his conclusions as to connections were generally justified.

160.                  We deal with Mr. Hall's evidence in relation to the alleged Klick Scheme later. Here we address that part of the evidence relevant to the two Deals.

Deal 1

161.                  In relation to Deal 1 Mr. Hall produced a chart showing a circular money flow in which some £680,000 was paid on 17 June 2006:

(1)  from Mundo to I Code Corporation,

(2)  from I Code to Movi World;

(3)  from Movi World to Sano y Salvo;

(4)  from Sano y Salvo to ESM Products;

(5)  from ESM to Axess; and

(6)  from Axess to Fonecomp; followed by

(7)  a payment of about £766,000 (being the purchase price paid on Deal 1 including VAT) from Fonecomp to PDA Stuff.

162.                  Then on 18 July 2006 PDA Stuff paid about £760,000 to TM Global, and TM Global paid a similar amount to Klick which paid Mundo on the following day.

163.                  This linking of payments and receipts was in our view justified from the bank statements. It was clear that money went around in a circle.

164.                  Although the account information included a narrative for each payment (e.g. "invoice ABC"), the narratives of the payments made in the chain between Mundo and Axess enabled no conclusions to be drawn as to what the  payments were for and whether or not they represented goods moving in the reverse order.

165.                  The information extracted from the servers also gave the time at which a transfer took effect, and the time during which the IP address, from which instructions for the payment had been given, was connected to the FCIB server. We drew the following conclusions from that information:

(1)  the money movements (1) to (7) took place within 1 1/2 hours;

(2)  the same IP address was used in respect of movements (2) to (5). These movements took place within 30 minutes;

(3)  Fonecomp arranged via Mr. Ibrahim's internet connection to make its payments to PDA stuff within six and a half minutes of the receipt into its account of payment from Axess. We were told that once an instruction was given to move funds it took about three minutes before it took effect. That would mean that Fonecomp gave instructions for onward payment within three minutes of the arrival of the funds from Axess into its account.

(4)   Mr. Ibrahim's computer was logged into the system 1 1/2 minutes before the payment arrived from Axess. (Mr. Sharif told us that Axess had phoned Fonecomp to let it know that the funds were being credited to its FCIB account).

Deal 2

166.                  For Deal 2 Mr. Hall's research into the bank accounts of the participants in the chain and beyond did not reveal a circular movement of funds. The data showed that transfers of about £425,000 were instigated on 28 July 2006 by instructions from a single IP address and moved along the chain: High Port Ltd - Movi World - Sano y Salvo - Axess. The correspondence of the amounts was supported by the fact that these movements took place within 30 minutes.

167.                  On the same day Axess paid Fonecomp, and 2 1/2 days later Fonecomp payed TM Global, which in turn and within 18 minutes paid Klick. There was no identifiable related payment to ESM Products (Klick's supplier in this chain).

168.                  Commentary

169.                  In Deal 1, the speed of the monetary transactions and the circularity of the money flow provided strong pointers to a conclusion that the transactions in the chain of supply had been orchestrated.

170.                  In Deal 2 the payments down a chain to Axess may represent other sales which the bank account information suggests may have been prearranged. The somewhat more leisurely payments by Fonecomp suggests that its transactions were not necessarily part of that arrangement.

(3). The Fraudulent evasion of VAT by Softlink

171.                  From the evidence of  Farzana Malik, an officer of HMRC responsible for Softlink, we find:

(1)  Softlink's first active director and secretary were Mohammed Afzad and Margaret Styles. They were appointed soon after its formation on 29 September 2005.

(2)  Softlink was registered for VAT with effect from 1 December 2005.

(3)  On 5 April 2006 HMRC's officers were given to understand that Softlink was buying phones from UK retailers, and in July 2006 that its business was buying from UK retailers, box-breaking (i.e. freeing the phones from a tie to one network) and selling the phones to UK traders. Its sales in the period to 26 June 2006 were about £100,000.

(4)  Between 15 August 2006 and 25 August 2006 Softlink, in 194 deals, in which it acquired some £377 million of mobile phones from Mundo, and sold them to Klick. The output VAT arising on these sales was some £66 million.

(5)  On 24 August 2006 HMRC's officers called Softlink to ask about phones at the freight forwarder, Paul's Freight, which had been imported from the EU. Mr. Afzad confirmed that Softlink had purchased those phones from Mundo.

(6)  A regulation 25 notice was served on Softlink on 31 August 2006 requiring a return to be produced. No response was received by HMRC.

(7)  On 5 September 2006 a VAT assessment for £66,525,371 was made on Softlink in respect of the transactions between 15 and 25 August.

(8)  On 6 September 2006 HMRC received a letter from Softlink dated 14 August (the day before the deals with Mundo and Klick) indicating that Mr. Afzad and Miss Styles had been replaced by Nazakat Hussain and Christopher Mewies respectively.

(9)  Later Mr. Afzad explained to HMRC that he had been paid £1000 in cash for his interest in Softlink.

(10)                   In an interview with HMRC officers in January 2007 Mr. Afzad said that he had been asked to stay on at Softlink by its new owners, but said he had not been involved in the mobile phone deals with Mundo and Klick at the end of August.

(11)                   In June 2007, at a meeting with Christopher Mewies, HMRC officers were told by Mr. Mewies that he had lost his wallet, driving licence and bank cards in August. A letter showed that he had reported the loss before 4 September 2006. Mr. Mewies denied any knowledge of Softlink or of Mr. Afzad.

(12)                   In September 2007 Mr. Afzad told HMRC officers that he had met Mr. Mewies and Mr. Hussain at the office of the accountants Vaghela & Co on 14 August 2006. It seems that Mr. Assad's description of Mr.Mewies was different from that of the HMRC officer who had met him in June 2007.

(13)                   At a visit to Vaghela & Co, HMRC officers were told that there had been a meeting on 14 August attended by Mr. Afzad and Mr. Hussein in relation to the sale of Softlink.

(14)                   No payment of, or communication in relation to, the assessment of 5 September 2006 has been received by HMRC.

(15)                   A compulsory liquidation order was made on Softlink on 2 April 2008. The amount assessed was unpaid and formed part of the company’s debt to HMRC at the time of liquidation.

172.                  From the evidence of Christina Quinn, an officer of HMRC responsible for Klick, we find that she was told by Klick’s representatives that Mr. Afzad was dealing with Klick in the period between 15 and 25 August 2006.

173.                  It seems clear to us that Softlink had a liability to pay VAT on the 194 deals it undertook between 15 and 25 August, and that it did not pay, and at the time that it entered into those transactions it did not intend to pay, that VAT. The apparent replacements of Mr. Assad by Mr. Husain and of Miss Styles by Mr. Mewies appeared us to have been constructed to distance Mr. Afzad from transactions with which he was clearly involved and is further evidence of an intention to evade the tax due on those deals. We have no doubt that Softlink fraudulently evaded, and at the time of the deals intended to evade, the £66 million of VAT on its sales to Klick between 15 and 25 August 2006.

(4) Klick and The Klick Scheme

(4)(a) Klick (UK) Ltd

174.                  From Christine Quinn's evidence we find:

(1)  Following confirmation from its sole director Mr. Zumorid Hussain that Klick would not be dealing in mobile phones, Klick was VAT registered on 28 September 2005. In April 2006 Klick started wholesaling mobile phones. At an interview with Christina Quinn in May 2006 Mr. Hussain said that 12 such transactions has taken place; in fact substantially more than that had (105).

(2)  Klick's sales in the 05/06 quarter were £72 million; it exported £30 million of goods and made a VAT reclaim of £1.3 million.

(3)  Klicks premises comprised a retail shop with a rear office. It seems retail business was conducted from the shop. When Christine Quinn and her colleagues had meetings at Klick’s premises between May and September 2006 they took place in the shop; HMRC's officers were not invited into the office.

(4)  Mr. Z Hussain was the only person HMRC met at the shop who was engaged in the company's wholesale phone business. Mr. Tokeer Hussain, Mr. Z Hussain's brother, was present at some meetings but he indicated he had nothing to do with the wholesale side of the business and was present in the shop only occasionally. We were not told of any other employees .

(5)  Mr. Z Hussain, whose background was not in high value transactions, was fairly new to dealing in mobile phones and not overly rigorous in his due diligence.

(6)  In May 2006 Klick entered into wholesale mobile telephone transactions of significant value. Each transaction was part of a chain of transactions in which phones were imported into the UK and in contemporaneous transactions sold through a sequence of UK entities and then exported. Klick was variously the importer, the exporter, or a "buffer" (an intermediary trade between UK traders in a chain) in these chains, and:

(a)   when Klick was an importer it acquired from ESM Sarl and sold to TM Global;

(b)  when it was an exporter is sold to Chippy and bought inter alia from TM Global.

(7)  In 56 out of 58 of such deals it made a gross profit of 3% on each sale.

(8)  These transactions gave rise to a VAT reclaim in its 05/06 return of £1,565,107.

(9)  In June 2006 the only wholesale phone transactions undertaken by Klick were imports and the on-sale of the phones imported. In each case Klick's transaction was part of a chain in which the phones were imported by Klick and sold to a UK intermediary (or in HMRC’s terminology a “buffer”) and thence along a chain so that they were later exported by another UK company.

(10)                   In each case Klick bought the phones from Mundo. Mundo's supplies to Klick were detailed on 361 invoices with a value of some £172 million on dates between 7 and 30 June 2006.

(11)                   In July 2006 the pattern was the same: the only wholesale phone transactions undertaken by Klick were imports and the related onsales. In each case Klick’s transaction was part of a chain culminating in the export of the phones from the UK. Fonecomp's exports in this appeal were part of one of these chains. In each case Klick purchased the phones from Mundo or ESM Products. There were 150 invoices from Mundo with a value of £177 million, and 44 invoices from ESM Products with a value of £32 million.

(12)                   On 7 August 2006 Klick entered into its only wholesale import deal in August. It imported a consignment of phones from Mundo and onsold them.

(13)                   As a result of the imports described in (9), (10) and (11) above, and the subsequent on sale of the phones imported, Klick would, if it had entered into no other transactions in the 08/06 quarter, have had a VAT liability for that quarter of £66,026,805.

(14)                   But between 15 and 25 August 2006 Klick entered into 197 transactions under which it bought phones from Softlink and exported them to ESM Products. (Softlink had imported the phones from Klick’s former supplier Mundo). These transactions were valued at some £380 million, and generated an input tax claim of £66,380,151.

(15)                   This left Klick with a net input tax reclaim in its return for the period 08/06 of only some £355,000.

175.                  Commentary

176.                  It was absolutely plain to us that Klick's activities in June, July and August were part of a planned scheme which encompassed Softlink's default. The clear object of that scheme was to fuel chains of supply in June and July 2006 with input tax credits which could eventually be claimed by exporters, and to match those credits with a later fraudulent VAT default by Softlink leaving a small VAT reclaim for Klick at the end of the quarter. All Klick's sales in June and July were by this arrangement clearly connected to Softlink’s later fraudulent default. We reached this conclusion:

(1)  because of the pattern of Klick’s trading: its initial imports and onsales followed by the Softlink purchases and exports;

(2)  because of the substantial match between the accrued VAT output tax liability on the June and July transactions with the input tax credit arising from its Softlink purchases and later exports;

(3)  because of the block selling of all of its August Softlink purchases to one overseas customer

(4)  because its only UK purchases were from Softlink;

(5)  because Softlink purchased in August from the very company which had been supplying Klick in June and July;

(6)  because of the remarkably short period in which the Softlink purchases were achieved; and

(7)  because of the fraudulent default of Softlink.

177.                  These facts also make clear that Softlink’s default  was planned as part of the same operation as Klick’s activities in June, July and August 2006.

178.                  The proceeds of this fraudulent operation were the sums payable to Softlink by Klick which represented the VAT on which Softlink defaulted. The evidence thus far recounted does not indicate who benefited from these sums. The FCIB evidence below sheds some light on this (see (4)(c) below).

179.                  What this evidence does not (at this stage) show is that those who purchased from Klick in June and July were parties to the arrangement. Certainly it was the VAT element in the purchase price which they paid to Klick on their purchases (which became their input VAT) which would have funded Klick's ability to pay the VAT element in its purchases from Softlink,  and represented the liability on which Softlink defaulted; but on the evidence related so far it is possible that could have been the case without their transactions having been arranged by the same minds which organised Klick and Softlink.

180.                  However, the number of transactions undertaken by Klick in June and July does indicate to us that these transactions cannot have been negotiated and executed only by Mr. Z Hussain in the back office of his shop. Mr. Hussain's background appears to have been in small-scale transactions: we do not believe that within a few months he would acquire understanding and capacity to deal in transactions worth hundreds of millions of pounds per month. There was someone else. The uniformity of Klick’s mark-up on its export deals indicates as well that someone had planned these trades.

181.                  Finally we should note that on 18 May 2009 Christina Quinn wrote to Klick denying its claim for input tax of £66,320,738 for the 08/06 period on Kittel grounds.

(4)(b) The alleged Klick Scheme

182.                  We have referred to HMRC’s contentions that there was a wider "Klick Scheme" of which Fonecomp’s transactions formed part. HMRC assert that there was a managed scheme under which a group of traders exploited VAT stagger periods so that their VAT returns claiming input VAT could be made in periods before Softlink defaulted on the VAT which gave rise indirectly to the input VAT claims. Because the returns were made and would have been checked before Softlink defaulted, HMRC would have difficulty denying the payment of the input VAT.

183.                  In outline, HMRC's case is that there was a scheme under which :

(1)  Trader A, whose VAT period ended on the 30 June, would buy phones from trader B whose VAT period ended on 31 July. The purchase would take place on 29 June and A would export the phones on 30 June. A would reclaim its input VAT on its purchase from B in its June return;

(2)  Then trader B would on 29 July buy phones from Klick which had imported them and whose VAT period ended on 31 August. Trader B would immediately export the phones. In trader B's VAT return for July the input tax on its purchase from Klick would be offset against the output tax on its sale to A and no VAT would be payable;

(3)  Then in August Klick bought phones from Softlink and exported them. In its August VAT return Klick would set the VAT due on its sale to B against the input tax on its purchase form Softlink: the return would show only a small net amount of VAT.

(4)  Then Softlink, owing the output VAT on its sale to Klick, which represented Klick’s, and B’s and C’s input tax credit and indirectly C’s reclaim, defaulted on that tax.

184.                  HMRC do not say that in every case there were traders in the position of B between the exporter and Klick. In some cases they say that there was only one such contra trade between Klick and trader A, and in some cases that there were more than two. They also say that there were in some cases "buffer" traders between these participants who bought and sold within the UK in the same VAT period: thus on 30 June B might have sold to buffer X who sold to trader A.

185.                  In the case of Fonecomp HMRC say that there was no contra trader between Fonecomp and Klick although there were one or two buffers, but that the chain in which Fonecomp purchased exploited the same VAT period advantage because Fonecomp’s VAT return was made for the month to 31 July 2006 and Klick’s (and also Softlink’s) corresponding returns in which its sales and purchases were accounted for was for the three-month period to 31 August 2006.

186.                  HMRC summarise their contention thus in their statement of case:

"Accordingly, by orchestrating a series of supplies between such traders subject to different accounting periods, it is possible by exploiting the different staggers of these periods to create an individual (and overall) net repayment by the Commissioners in each month over a number of successive months. This is achieved by ensuring that each trader involved in the Scheme, by the end of its accounting period, has built up an entitlement (or purported entitlement) to input tax credits on transactions in which it acts as a broker [that is to say a trader who purchases from the UK and exports], which exceeds its liability for output tax on transactions on which it acts as an acquirer [i.e. importer] or a buffer.

“The continued operation of such a scheme necessarily entails other traders involved in making (or purporting to make) an ever-increasing value of acquisitions of goods which can be supplied to the brokers to export. This generates an ever-increasing output tax liability on the part of those traders approximately equivalent to the sums of input tax claimed from the Commissioners.

“Those acquirers then, in turn, enter into contra-trades of their own to pass on to another trader and thereby defer crystallisation of the liability of the traders within the Scheme to make any payment of output tax due to the Commissioners.

“The Scheme comes to fruition when one trader, namely Softlink ... defaults on this liability to the Commissioners ...".

187.                  We approach the contention that there was a Klick Scheme in the following way. We start with the definition of a set of transactions. These are the transactions which HMRC, through Mr. Jones, say were part of a scheme. We then ask whether the nature and circumstances of those transactions indicate that they were indeed part of the scheme, that is to say orchestrated or organised by one or more directing minds.

188.                  In what follows in this section we refer to chains of supply. By this we mean a sequence of purchases and sales of a consignment of mobile phones. When we speak of a transaction "tracing" to another we mean that the two transactions can be shown to be part of the same chain.

189.                  The scheme is defined as comprising all those transactions in any chains of supply which:

(1)  started with Klick's imports and onsales in 08/06;

(2)  included a supply made by a member of the chain in (1) in the VAT period in which that member participated in that chain, where that supply did not trace back to a default other than that of Softlink, or

(3)  included a supply made by a member of a chain in (2) in the VAT period in which it participated in that chain and which did not trace back to other defaults, and so on.

190.                  Under this definition supplies made before the beginning of Klick’s 08/06 period could be defined transactions if the relevant trader’s VAT period commenced before 1 June 2006.

191.                  Mr. Jones described the identification process in (2) and (3) differently. He categorised certain traders as engaging in contra trades within the scheme and took as part of the scheme all those contra trades and their chains and any chains connected with them by other contra trades. However the result of this approach was that after May all the transactions of those contra traders (barring those tracing back to other defaulters, with a few exceptions which we will return) were treated as scheme transactions. Thus his approach gives rise to the identification of the same trades as in the definition above.

192.                  We use the definition above because it does not require us to investigate whether the trades Mr. Jones describes as contra trades had features which enabled their classification as such. Since the transactions within the definition are the same that question does not have to be answered by us

193.                  Mr Jones was not able to trace forward defined transactions which were sales of two members of the relevant group of traders as we have defined it. In respect of these sales he was therefore unable to show whether the goods had eventually been exported. The input tax on the purchases which fuelled these sales was only 0.3% of the total VAT input tax reclaims.

194.                  We accept that Mr Jones correctly identified the transactions which fell within the defined transactions. We find from Mr Jones’ evidence that in these transactions: .

(1)  there were 1622 chains made up of 5800 transactions;

(2)  3,623,698 phones were traded with the selling price of some £1.5 billion;

(3)  all the fully traced chains (i.e. all the chains  apart from chains of 0.3% in value) ended with an export of the phones which had been imported;

(4)  36 companies were involved of which 12 were importers of phones in the defined transactions;

(5)  of the 12 importers, 10 made their first ever transactions in May 2006 as part of the defined transactions, and six of those transactions were made on 3 May;

(6)  some 36% of the phones sold were Nokia 8801. These phones are designed primarily for the US market. A further 10% were 8801LT, also designed for that market.

(7)  Mr. Jones was supplied with data on the freight forwarders used in 1009 of the 1622 deal chains. In each case the goods (2,146,129 mobile phones in all) were held at Paul's Freight when in the UK;

(8)  in 899 chains the phones were delivered to Pro-log in Paris;

(9)  in 201 chains these phones came to the UK from Pro-log and at the end of the chain were dispatched to Pro-log in Paris;

(10)                   the net VAT input reclaims made by the exporting traders in relation to the defined transactions was £66,556,102. Soft link’s default was in relation to an amount of £66,525,371.

195.                  HMRC say that these factors show that the defined transactions were all part of a managed scheme to defraud.

(4)(c) The FCIB evidence

196.                  An introduction to Mr Hall’s evidence is at (2)(f) above.

197.                  Mr. Hall told us that a good deal of time was spent analysing the FCIB Datastore to determine whether there were any links between the various companies in the Klick Scheme transactions. He presented the links which had been found but said that no link was found between Fonecomp and the other participants.

198.                  Mr. Hall was given details of the Klick Scheme transactions by colleagues at HMRC. He selected a sample of the chains of transactions and researched the cash payments between the members of the transaction chains. He then sought to determine whether it was possible to identify payments to, or by, the persons in a chain which could be said to represent the payments respectively by and to those participants and so on. This process involved treating receipts and payments which were of similar amounts as being connected; in particular where they were made consecutively or where a payment exhausted the amount standing to the credit of an account after a receipt had been credited to it.

199.                  Mr. Hall produced money transfer diagrams from some of the transactions in which Klick had purchased from Softlink and onsold to ESM. These showed that there were movements of funds between the accounts of some seven companies, some of which took place within about 30 minutes, which were instigated by instructions from as few as one IP address, and which resulted in a broadly circular movement of funds.

200.                  Whilst on occasion the timing of the funds’ movements did not wholly support Mr. Hall's conclusion of sequential circularity, there was consistency in the closeness of timings and in the correspondence of related amounts sufficient for us to accept Mr. Hall's general conclusion of frequent circularity.

201.                  In all the Softlink deals Klick made payments to Softlink of a VAT inclusive amount which Softlink then paid on to Mundo without retaining the VAT element of its receipt to fund its output VAT liability. From that we conclude that Mundo was the initial recipient of the profits of Softlink’s fraudulent evasion of VAT.

202.                  These are the aspects of the FCIB evidence which further fortify us in our conclusion that Softlink fraudulently evaded VAT as part of a fraudulent scheme to which Klick was a party.

203.                  Mr. Hall also investigated funds’ movements in the case of a sample of other transactions within the defined transactions. He found circularity of funds in a substantial majority of the transactions investigated. We accept that such was the case: even if in one or two cases Mr. Hall might have made unjustified inferences of connection, our overall impression was that his conclusion was substantially justified.

204.                  Commentary

205.                  If Klick had sold its goods to traders which had in turn had sold to others who had exported them, and if the exporters had not engaged in other transactions, then Klick’s VAT output tax liability, and the corresponding output tax liability of Softlink would, as a matter of the logical application of the tax, have been approximately equal to the aggregate VAT input tax reclaims by those exporters (assuming that sales and purchases down each chain were at approximately the same prices). In those circumstances the coincidence of the Softlink VAT default amount and the amount of the aggregate VAT reclaims would not have been surprising or particularly indicative of any management or contrivance.

206.                  What distinguishes the pattern for which HMRC contend is the contra trading activity of the intermediate traders and in particular the cases where the only transactions of those contra traders were balanced transactions in which they acquired goods from Klick, or from someone who had acquired from Klick, and then exported them, balancing input VAT with an output tax liability on a sale of other imported goods to a person who then exported them.

207.                  The facts recorded in (4)(b) above, together with those recorded in this section, make plain that within the defined set of transactions most were part of an organised scheme. It seems unlikely that all the transactions were organised by one person: to set up 5600 transactions in three months would be beyond most people. But given the frequency of the circularity of funds, the use of the same freight forwarders, the often similar movements of goods, and above all that all the chains involved both the import and export of phones, it seems to us that at the very least the activities of those participants which Mr. Jones describes as contra traders must have been part of an arrangement in which they knew they had a part. It is in our view not inappropriate to call the defined set of transactions a scheme.

208.                  The transactions could be likened to a tree with Softlink as its roots, Klick as the trunk, the contra traders and buffer traders at its branches, and the exporters as its leaves. The factors listed above must mean that the branches knew they were part of the tree and undertook their actions to further the scheme.

209.                  The simple exporters’ (those who were not also contra traders) transactions took place at the edge of the scheme. It did not seem to us that it was a necessary consequence of the planned architecture of the tree that they knew that they were part of it.

(5) Aspects of the chain of supply to Fonecomp

210.                  We noted the following aspects of the chain of supply from Mundo or ESM to Fonecomp and on to Axess in each of the two Deals:

(1)  Deal 1, Klick, TM Global, and PDA Stuff made gross profits of 25p or 50p per phone; Fonecomp made a profit of £11.50 (N 9300i) or £17 (N 8801) per phone;

(2)  in Deal 2, Klick and TM Global made profits of 25p and 50p per phone; Fonecomp made a profit of £17 (N 91) or £14 (N 80) per phone;

(3)  all the transactions were invoiced on the same day in back to back transactions; and

(4)  the phones were all held at Paul's Freight during their stay in the UK.

211.                  We also noted that the profit made by TM Global on Deal 2 was £650, that made by PDA Stuff on Deal 1 was £1150. For these small profits, they seemed to take a large risk on Fonecomp’s credit.

212.                  The first of these features points to some sort of contrivance of the chain or at least to some recognition by the companies before Fonecomp that they knew their place in some grander scheme: they took a fixed profit for passing the phones along. The second suggests to us that these companies were told what to do and it. Otherwise the coincidence of such risk taking seems too good to be true.

(6)Connection to Softlink’s default

(6)(a) connection to Klick

From Mrs Worbey’s evidence we find as follows.

213.                  In relation to Deal 1, as recorded by invoices all dated 11 July 2006:

(1)  Mundo sold 1200 (675+525) Nokia 9200 and 1100 (600+500) Nokia 8801 to Klick;

(2)  Klick sold 1200 Nokia 9300i and 1100 Nokia 8801 to TM Global. Although one invoice showed the sale being of 1100 Nokia 9500 rather than Nokia 8801, there was a second invoice which showed the sale being Nokia 8801. We conclude that Klick sold the 1200+1100 phones it acquired from Mundo to TM Global;

(3)  TM Global sold 1200 Nokia 9300i and 1100 Nokia 8801 to PDA Stuff; and

(4)  PDA Stuff sold  1200 Nokia 9300i and 1100 Nokia 8801 to Fonecomp..

214.                  We find that Fonecomp's acquisition of these phones was connected with Klick’s acquisition and disposal of them in the senses: (a)  that they were the same phones and sold and acquired in transactions close in time, (b) that the input and output VAT  were offset, and (c) that the acquisition aided Klick’s sales. 

215.                  In Deal 2, invoices, all dated 26 July 2006, recorded:

(1)  ESM Products sold 900 (500+400) Nokia N80 and 500 Nokia N91 to Klick;

(2)  Klick sold 900 Nokia N80 and 500 Nokia N91 to TM Global; and

(3)   TM Global sold those phones to Fonecomp.

216.                  We find that Fonecomp's acquisition of these phones was in the same sense connected with Klick’s acquisition and disposal of them.

(6) (b) connection to Softlink’s fraudulent default

217.                  We have found that Fonecomp's acquisition of the phones was in that sense connected to Klick’s purchase and sale of them in July 2006. We have also found that all Klick's sales and purchases in July 2006 were executed with the object of fuelling chains of supply with the input tax credits and of matching those credits in aggregate with the planned VAT  default by Softlink. This was an arrangement connecting Klick’s supplies to its purchases from Softlink and Softlink’s default. We have found Softlink fraudulently defaulted on its VAT liability.

(6)(c) Conclusion

218.                  For each Deal we find that Fonecomp's purchase was connected with Klick’s fraudulent arrangements in relation to, and with, the fraudulent evasion of VAT by Softlink. That is for the following reasons.

219.                  On the Blue Sphere test there is a connection of inputs and outputs along the chain from Klick to Fonecomp, and an offsetting connection within Klick of all its outputs against the inputs from Softlink.

220.                  If the test is whether Fonecomp’s purchase assists in Softlink’s fraud, then it does. It was a transaction which helped fuel Klick’s arrangements and Klick’s arrangements aided, encompassed  or even caused Softlink’s fraud.

221.                  We have already found that the Klick Scheme transactions were part of an organised scheme. If the test is whether Fonecomp’s purchase was part of an overall scheme to defraud the revenue, then that test too is satisfied.

 (7) The statement of PDA Stuff’s director

222.                  On 6 August 2007 HMRC's officers Siddle and Mysuik visited PDA Stuff. Mr. Khan, its director, was there. Mr. Khan, on behalf of PDA, has dealt with Fonecomp in the Deals. There was some discussion of MTIC fraud. Mr. Siddle recorded in his notebook the following:

“Officer Mysuik asked Mr. Khan if he had heard of the recent Calltell case. Mr. Khan explained that he had and continued to explain his understanding of the fraud that exists within the trade sector. Mr. Khan believes that traders within this sector are more than aware of the fraud that exists and he believes they know exactly what they are doing. Mr. Khan agrees that all deals undertaken are completed with the sole purpose of stealing from the VAT system but believes traders, as he did, become embroiled and overwhelmed by the profits that can be achieved ... Officer Siddle and Mysuik explained that they believed that hardened  criminals backed the fraud. Mr. Khan agreed with this."

223.                  The appellant did not wish to cross examine Mr. Siddle and we heard neither from Mr. Siddle nor from Mr. Khan. Mr. Khan's reported comments were made in 6 August 2007, a year after the Deals took place. They are not precise about when the traders had the knowledge he is recorded as imputing to them: he says that he believes that the traders are aware of the fraud which exists and knew what they were doing, but it is unclear whether by that he means that the traders knew that every trade they undertook traced back to fraud rather than that some trades might do. Although he is recorded as saying that he thought all deals were  undertaken with the sole purpose of “stealing” VAT, he does not say that the traders had the same view a year ago or that they then knew that their deals were all  linked to VAT fraud.

224.                  We find this evidence confirms our conclusion that Fonecomp knew that VAT fraud was prevalent and that it could be an explanation for its deals, but we do not find it of sufficient weight to help as to any other conclusion.

VI The Appellant’s arguments

225.  Mr Young said:

(1)        Mr. Cunningham set a high threshold in his questioning of Mr. Sharif. Mr. Cunningham likened participation in mobile phone fraud to theft and equated Fonecomp's unwillingness to give up its business to a thief’s reluctance to give up stealing. Mr. Young said that participation did not imply guilt: it was possible to be an innocent participant.

We accept that participation does not imply knowledge. The question for us is whether there was knowledge or means of knowledge.

(2)        There was inequality of arms in the appeal. The appellant had funds for only five days’ representation by counsel; the respondents’ fielded leading and junior counsel for over two weeks. The appellant did not have funds to enable its advisers to look, or to look in any detail, at all the available documents and statements. The tribunal should bear in mind that a witness statement may not have been challenged, not because it was accepted, but because it had not been fully considered or considered at all. The presence of unnecessary statements of opinion in HMRC's officers' witness statements increased their length and the time needed to consider them: it compounded the unfairness in the trial of an allegation of participation in fraud - a grave and serious allegation.

In making our decision we have, so far as we can, taken into account the inequality of arms which Mr. Young mentions. In particular, where Mr. Cunningham made points about the drafting of Mr. Sharif's witness statement we have allowed for the problems which arise when representation is limited.

(3)        Although the drafting of the appellant’s terms of business was a little shaky the appellant’s staff were not lawyers. Its terms of business remained the same today.

(4)        When considering knowledge the tribunal should bear in mind that, in July 2004, HMRC's own officers said that there was no strong evidence of knowledge or means of knowledge of fraud in connection with earlier transactions by the appellant.

It seems to us that the opinion of the HMRC's officers of July 2004 is of very little relevance. Our concern is with the facts known to the appellant as we find them, not with what an officer from HMRC thought.

(5)        Even if Fonecomp’s due diligence had been perfect it could not have found out about the fraud by Softlink. If it had been able to chase the chain back it would have found VAT properly accounted for at all stages. There was no missing trader in the clean chain.

We accept that had purchases been merely chased down the chain it would have been seen that there was at each stage the proper accounting for VAT in the sense of recognising inputs and outputs relating to the phones. But the problem arises when one gets to Klick. Klick can only be said to have accounted properly for VAT if it was entitled to the input tax credit it claimed to offset its output. It was not so entitled because it knew of, and in our view was probably involved in, Softlink’s fraud. Further, even if one can regard that chain as wholly clean the question, as we see it, is not simply whether the appellant would have discovered that the chain was clean, but whether looking at all the circumstances of his transactions he should have known of the connection to fraud.

(6)        Even if the appellant’s sales were part of a fraudulent scheme, that did not mean that the appellant knew it was part of that scheme: it could well have been taken in by Klick.

We agree.

.

VII. Discussion

226.                  The evidence in relation to the Klick Scheme convinced us that there was an orchestrated, contrived or managed series of transactions - an arrangement - which encompassed the Softlink/Klick deals, and the deals between Klick and the contra traders, and between contra traders. The FCIB evidence of circularity of cash movements and the central control of some cash movements supported the conclusion that many of the transactions  were preplanned or arranged.

227.                  However, it seemed to us that it is possible there may have been export sales by traders at the fringes of the scheme made by persons who were not party to the management of the scheme but whose willingness to buy and then export meant that they fitted well into the scheme and aided it.

228.                  If a true scheme participant wished to balance his VAT books by importing phones and selling them to a third company, it may well have made its usual contacts aware that it had phones for sale; and the scheme organiser may well have arranged for a non-UK buyer to pass out a message that it was looking to purchase those very phones. Sooner or later an exporter would find the two offers before him and come to the deal the organisers had expected. And for this to happen there would be no need for the exporter to know anything about the nature or planning of the scheme. At least in relation to an isolated deal this could be a case where an animated puppet had his strings pulled without knowing it.

229.                  Thus it seems to us that the existence of the scheme does not necessarily mean that the exporters at its edges like Fonecomp were knowing participants in the scheme itself. They could well have been the squirrels which buried the oak’s acorns and which enabled them to germinate.

230.                  Therefore we do not find that Fonecomp's participation in the scheme as an exporter meant that it necessarily knew that it was part of the scheme.

231.                  On the other hand the Deals had characteristics which, taken together, did not make ordinary commercial sense. Put cursorily, the appellant’s story is that: Fonecomp heard its customer was seeking phones, miraculously it found a supplier who had them, entered into a series of documents for their purchase and sale which it should have known did not make commercial sense, bought and sold on the same day, and shipped the phones to a "neutral location" where none of the parties had any presence or need for the phones; payment followed several days later. Overall it sounds fanciful, and Fonecomp should have known that.

232.                  But that is a broad impressionistic overview. We now set out the detail of our analysis.

233.                  The following findings indicated that the parties to the deals were not concerned with precisely what phones they were buying and selling:

(1)  the lack of specification, particularly on purchase orders from Axess

(2)  the lack of proper inspection of the phones despite their value

(3)  the purchase and sales of phones designed for the US market and sold with "Central European spec"

(4)  the lack of defined clear contractual terms (time of payment, title etc).

234.                  The only explanation of this is that what Fonecomp's customer (and possibly Fonecomp and its supplier) wanted was that there would be a sale and export, and was indifferent to precisely what was said to be exported because the recipient already knew precisely what it was getting because the chain had been prearranged.

235.                  The fact that the deals were  European specification 2 pin phones or American 2 pin phones is, for the reasons in section V (2)(d) above, explicable only if the phones had been imported into the UK in order that they could be exported again, that is to say imported for the sake of creating, at the time of import, a VAT liability on the importer’s later sale which could be either defaulted upon or passed through contra traders to cover another default.

236.                  The only reasonable explanation of these features was that these Deals were part of a scheme for the import, onsale, and later export of the phones, and the only explanation for such a scheme was that the transactions in the phones were part of, and aided, a VAT fraud.

237.                  Fonecomp knew the mechanics of MTIC fraud. Fonecomp knew that its suppliers were back-to-back trading: it knew there could be VAT fraud affecting the chains of supply to it. In those circumstances it should have known that the factors listed above lead ineluctably to the conclusion that these Deals assisted or were connected to such a fraud

 

238.                  We do not find that that Fonecomp knew or should have known that Softlink would default or that there was a defaulter whose fraud Klick would arrange to cover up. We find instead that Fonecomp must have come to the conclusion that somehow its purchase and sale were assisting a VAT fraud by an importer which had happened or was planned.

239.                  Fonecomp knew all the facts from which this conclusion is drawn when it made its purchases. It therefore should have known that its purchases were connected with fraud at the time it made them.

240.                  The due diligence that Fonecomp undertook could not have affected the drawing of this conclusion. It had no impenetrable shield because the information available to it led to the conclusion that its purchase was connected with a fraud.

VIII. Conclusion

241.                  We conclude that Fonecomp had the means to know that the only reasonable explanation of the Deals was that they were connected to VAT fraud. We have found them to be so connected.

242.                  On this basis we would have  dismissed the appeal.

IX.  A reference to the ECJ?  

243.  After the hearing had finished Judge Bishopp gave directions that certain MTIC appeals to the Upper Tribunal be held over pending the promulgation of decisions of the ECJ in a number of references which had been made to it which might affect the way in which the principle in Kittel is to be applied. Following that decision the appellant's solicitors wrote to the tribunal indicating that they thought that Judge Bishopp's decision had a significant effect upon this appeal. They suggested that the parties should put in written submissions setting out any argument arising from that decision. They indicated that the appellant wished to develop the argument not only that its appeal should be stood over, but also that the tribunal itself should make a reference to the ECJ.

244.  During the course of the hearing the appellant had suggested that the approach taken in Mobilx was wrong. It referred to the apparent conflict between interpretation of connection adopted by the French legislature, which appeared to regard the connection required by the test as existing only if the seller was fraudulent, and the broader test adopted by the English Courts. It was suggested that we should consider making a reference to the ECJ if this or any other issue was not acte claire. Mr Young referred us to the reference to the ECJ in Bonik.

245.  We were not convinced that the decision of Judge Bishopp affected the conclusions we had reached on the law in this case. In his directions he expresses no view on the principles and limits himself to saying that the cases referred to the ECJ raise obliquely questions in relation to the application of Kittel. We are therefore not persuaded that it would be fair or just to prolong this appeal by permitting such representations.

246.  We are also not presently convinced that it would be right to make a reference to the ECJ. We are satisfied that the principles enunciated in Mobilx and those we have applied reflect the proper interpretation of Kittel. It seems to us that the better course is for us to make our decision on the basis of those principles and, if a party considers we are wrong and seeks permission to appeal to the Upper Tribunal, to give that permission readily. In that way this and other cases would be conveniently handled together. However this is a matter on which it would be fair to hear the parties’ views. We therefore direct that if a party wishes to make written representations on this issue it must do so within 28 days of the release of this document. If no representations are made in that period this appeal shall then be treated as dismissed; if such representations are received the tribunal may give further directions.

X. Rights of Appeal

247.  If no such representations are received then this document contains full findings of fact and reasons for the decision. In those circumstances ny party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009.  The time limit for making that application in these circumstances is extended. The application must be received by this Tribunal not later than 84 (=56 +28) days after this decision is sent to that party.  The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.

 

CHARLES HELLIER

 

TRIBUNAL JUDGE

 

RELEASE DATE:  3 February 2012

 

 

 

Appendix:  Prior Decision in relation to interim credit application

 

DECISION IN RELATION TO AN APPLICATION FOR DIRECTIONS

 

1.      This is a decision on an application made by Fonecomp Ltd in an “MTIC” appeal. The essential features of the appeal for present purposes are these:

(1)        the Appellant has paid for goods supplied to it under a VATable supply;

(2)        the Appellant has claimed relief for the VAT borne on that supply;

(3)        HMRC have refused to give that credit (and to make repayment of the input VAT) because they say (a) that the supply to the Appellant was connected with fraudulent evasion of VAT, (b) that the Appellant knew or should have known of that fraud, and (c) under the doctrine in Axel Kittel v Belgian State  C-439/04 the input VAT is therefore not creditable.

2.        The Appellant sought a direction:-

“a) that an amount of input tax in the sum of £100,000 be credited to the Appellant and or in the alternative;

b) that the Respondents be debarred from adducing evidence of fraud;

c) such other directions as the tribunal sees fit”.

3.        In this decision I consider each part of that application in turn.

a) An amount of input tax be credited to the Appellant

4.        Mr Young told me that the Appellant sought this direction because the refusal by HMRC to repay its input tax has meant that it does not have the funds to pay its solicitors and counsel to pursue its appeal properly. In broad terms it says that the only way it can get a fair trial of its appeal is if some such payment is made. Mr Cunningham says that this tribunal does not have the power to make the direction sought.

Mr Young’s argument

5.        Mr Young does not seek a direction that HMRC pay £100,000 to the Appellant: he agrees that such a direction would be outside the tribunal’s jurisdiction. But he says that the tribunal may direct that input tax be credited to the Appellant because it is given jurisdiction to do so by section 83(1)(c) VAT Act 1994:

(1)        “…an appeal shall lie to the tribunal with respect to any of the following matters-

…(c) the amount of input tax which may be credited to any person…”

Such a direction he admits would in practice have the same effect as a direction to make payment.

6.        Mr Young says that the Appellant has a right to an input tax credit which is conferred by the Directive. He relies on Garage Molenheide v Belgian State C-286/94 Belgian State v Ghent CoalTerminal BV -37/95 1998, Kittel and Alicja Sosnawska C-25/07. He says, in reliance on Bulves AD v Bulgaria no 3991/03, that that such right is an “asset” or at least a legitimate expectation of a claim which belongs to the Appellant. That right cannot be taken away by HMRC but may only be denied after a hearing by this tribunal. Kittel, he says, permits the removal by the national court of a thitherto extant right, but does not limit or reduce the nature of that right.

7.        In the hearing of an appeal to determine whether or not that right should be removed, Mr Young says that Article 6 of the Human Rights Convention entitles the Appellant to a fair trial. He says that Ferrazzini v Italy [2001] STC 1314, in which the ECHR decided that proceedings regarding taxation did not relate to civil rights and obligations for the purposes of Art 6, was distinguishable: that case related to taxing the subject, not to a decision on whether his right should be removed. A fair trial, he says, requires the ability to have legal representation in a case where HMRC had solicitors, leading and junior counsel.

8.        Further, since the input VAT was “property” for the purposes of Art 1 of Protocol 1 to the Convention it could not be taken away by the state in a disproportionate manner. He cites the ECJ cases noted above as authority for the proposition that the same requirement for proportionality arises under EU law. The refusal by HMRC to make any payment was disproportionate: it went further than was necessary to achieve a legitimate aim.

9.        He says that Rule 2 of the tribunal’s rules must be read in light of the right to a fair trial and the requirements for  proportionality. The overriding objective of dealing with a case fairly and justly included:

“(a) dealing with the case in ways which are proportionate to the importance of the case, the complexity of the issues, the anticipated costs and the resources of the parties;”

10.     Read in the light of the Convention rights and the requirements for proportionality, the tribunal had a duty to make directions to ensure a fair trial and redress disproportionate action. The direction sought did this.

 

Mr Cunningham’s reply

11.     Mr Cunningham says that the right to input tax is circumscribed by the Kittel decision. If the Appellant knew or should have known of fraud his right is impaired. He relies on Moses LJ at [36] and [47] in Mobilx v HMRC  [2010] EWCA Civ 517. He says that the Appellant does not have a property right if it knew or should have known of fraud. There is no asset to be protected by Art 1 or by proportionality.

12.     He says that, unfortunately, there are many cases where a party does not have the firepower of his opponent, but that does not mean he does not have a fair trial.

13.     He refers to the High Court’s power under CPR25 to make orders for interim payments on account. He notes that Dyson LJ in R(on the application of Teleos) v CCE [2005] EWCA Civ 200 says at [12] that this is a departure from the strict principle that a defendant has a right not to be held liable until the liability has been established by final judgement. The absence from the tribunal’s rules of a circumscribed power such as that in CPR 25 points to the intentional absence of any such power.

14.     He says that section 83 does not confer on the tribunal a power to make the direction sought. Section 83 deals with appeals: an “appeal” means a determination on the merits, not what, given the Appellant’s reported straightened circumstances, would be in effect a final direction to pay without any determination.

Discussion

(a) a right to input tax

15.     If the Appellant has a present right to input tax credit, that has two effects. First it might make it appropriate to regard section 83 as permitting the declaration of that present right as part of the appeal. Second it may bring into effect those requirements of proportionality which relate to the actions of the state in depriving an individual of a right.

16.     In Garage Molenheide, the ECJ considered certain Belgian provisions which permitted the state to withhold payments of input tax if the authorities had serious grounds for considering that VAT returns had been inaccurate or if there was a VAT debt claimed by the authority and disputed by the taxpayer. The retention was a precautionary measure to block payment until a determination of amounts at dispute.

17.     The ECJ held that the Directive did not preclude such measures, but (at [46]) that the state must employ means to achieve its aims which were least detrimental to the objectives of the legislation: it must not systematically undermine the right to deduct input VAT which was a fundamental principle ([47]). If the measure went further than was necessary it would undermine that principle.

18.     The Court held inter alia: (a) that an irrebuttable presumption of necessity for the retention, and (b) that provisions which would prevent a judge from lifting the retention even if there was evidence before him to justify so doing, were not necessary - that is to say were not proportionate. It was for the national court to determine whether a particular measure was proportionate but national provisions which barred effective judicial review should be disapplied or construed less strictly.

19.     I accept that Garage Molenheide makes clear that the right to deduct is a fundamental feature of the VAT system. I return to other aspects of that case later.

20.     In Ghent  the ECJ said that the right of deduction of input tax:

(1)        “must be exercised immediately in respect of all taxes charged on transactions relating to inputs…derogations are permitted only in the circumstances expressly provided in the Directive…”

21.     But in Kittel the ECJ, having recognised that “the right to deduct input tax is an integral part of the VAT system and in principle may not be limited” ([47]), observed that preventing evasion was an objective of the Directive and held that that right could be affected in cases of fraud or where there was relevant knowledge of connection to fraud. This conclusion was phrased by reference to action by the national court:

(1)        “[55] It is a matter for the national court to refuse to allow the right to deduct where [it is shown that it is used for fraudulent purposes];

(2)        [59]…it is for the referring court to refuse entitlement to the right to deduct where it is ascertained [that there was relevant knowledge of fraud]…”

22.     Mr Young relies on this formulation to say that a right to a credit exists unless and until it is taken away by the court. I do not read these passages in the same way. The court was giving a preliminary ruling on a question referred by the national court. It was thus giving guidance to that court on how it should interpret the law. Hence the judgement instructs the “referring court to refuse entitlement…”. This language does not indicate that the right exists until taken away, instead it indicates a conclusion that the law to be applied is that the right does not exist if the relevant conditions are satisfied.

23.     That conclusion is consistent with the judgement of Moses LJ in Mobilx . At [25] he says that the principle of legal certainty requires that a person should know his position when he enters into a transaction, and that that required the identification of objective criteria for the deduction of input VAT: ”Once a transaction meets those criteria the right to deduct for which Art 17 provides must be recognised”. At [28]: “The objective criteria determine both the scope of the tax and the circumstances in which the right to deduct arises. “. At [41] he explained that Kittel set out the principle that the objective criteria are not met where the taxpayer knew or should have known of connection to fraud. He makes the same points at the end of [41], and at [43], [44], and [47]. His conclusion is that a right to deduct arises only if the objective criteria are met: it is not a right which arises just because a taxpayer incurs input VAT and which may be taken away by a later court decision. Such a scheme would not provide legal certainty.

24.     Nor in my view does Bulves help Mr Young. In that case a taxpayer’s right to input VAT arose under the Bulgarian provisions only if his supplier had accounted for the corresponding output VAT. The ECHR held that in so far as the taxpayer had fully complied with its own obligations and had no knowledge of its supplier’s lack of compliance and no means of enforcing its compliance, the taxpayer had a legitimate expectation of obtaining enjoyment of the input VAT, and that amounted to a “possession” within Art 1 Protocol 1 (see [57]). In its conclusion (at[71]) the Court extended its caveats thus:

(1)        “considering the timely and full discharge by the company of its obligations, its inability to secure compliance by its supplier, and the fact that there was no fraud in relation to the VAT system of which the applicant company had knowledge or means to obtain such knowledge, the Court finds that [it] should not have been required to bear the full consequences of its supplier’s failure…”[my italics]

25.     The facts in this appeal are different. In Bulves the taxpayer had satisfied all the objective criteria which bore on it; in this appeal HMRC allege that the taxpayer did not satisfy all the objective criteria which bore on it because they say it knew or should have known of fraud. In other words the substance of HMRC’s allegation is that the taxpayer had no legitimate expectation. The words italicised above serve in my mind to emphasise this distinction.  For me to proceed on the basis that the taxpayer had a legitimate expectation of payment would be to prejudge the appeal.

26.     Thus I cannot proceed with this application on the basis that the Appellant has a right. Instead that is the very question to be determined in this appeal.

27.     As a result, however broadly the section is construed, I cannot conclude that there is any jurisdiction under section 83 to declare the existence of a right without a hearing on the merits. That however may not make proportionality an irrelevance for, as Garage Molenheide shows, the actions of the state must not systematically undermine the right to deduct.  There therefore remains the argument that the direction sought should be made, not because the right exists, but because proportionality or Art 6 so require.

Proportionality

28.     Garage Molenheide makes clear that fruition of the right to deduct may be delayed but only if such action is proportionate – that is to say necessary for the aim to be achieved. In that case the national provisions permitted retention in relation to the VAT of one period by reference to the VAT position in other periods.

29.     In the same vein Mr Young relies on Sosnowska.  In that case national legislation delayed the repayment of input tax to new traders. It did so because of concerns over evasion. The ECJ found that such a restriction on repayment was not proportionate to the aim of preventing evasion since it did not make provision for the trader to demonstrate the absence of risk of evasion, and so obtain his VAT repayment in a reasonable time.

30.     The national provisions in these cases were of general application. In this appeal the state contests the very tax credit which is claimed. Whether the withholding of its claim is proportionate must be judged by reference to these circumstances.

31.     I accept that the apparatus of the state, its Acts and the acts of HMRC, are subject to a requirement that they be proportionate. The issue is whether section 83 or the tribunal’s rules should be read so as to enable the tribunal to make the direction sought if proportionality so required.

Section 83 VATA

32.     It seems to me that the plain words of section 83(1)(c) do not confer on this tribunal the power to make an interim determination of the type sought by Mr Young. Section 83 confers on the tribunal the power to determine what input tax should be credited. The “matter” is the amount of that input tax credit. I do not regard “may be credited” as indicating that the tribunal may at one stage say that £40 should be credited , and then at a later stage say that nothing or £100 should be credited. The object is to determine, once and for all, an amount which is to be credited in accordance with the Act: “may” to my mind reflects that a credit may be permitted under the Act.

33.     Should section 83 be given a wider interpretation in accordance with the dictates of proportionality of community law and the requirement of the Convention for proportionality and possibly a fair trial?  In this context I note that where the matter to be decided involves a consideration of whether the state’s action is proportionate the tribunal has accepted jurisdiction to apply that principle. But that is different from construing section 83 itself more widely.

34.     So far as proportionality is concerned it seems to me that if there were not another forum in which that issue could be judicially considered, it would be necessary to consider whether section 83 should be read as conferring on the tribunal a duty and power to consider whether the apparatus of the state produced a proportionate result in relation to a claimed tax credit. But the High Court provides such a forum, and I can see no need to give section 83 a strained construction in order to provide another. The judicial framework provides for the determination of the amount by the tribunal and for the issue of proportionality of HMRC’s action in relation to the tax credit to be dealt with by the High Court.

35.     To the extent that Art 6 of the Convention is applicable to this case, it seems to me that it cannot impinge on the meaning to be given to section 83. Section 83 specifies the types of appeals the tribunal can hear, not how matters relating to the appeal should be treated: an extended meaning would not go with the grain of the section.

The Tribunal Rules.

36.     Rule 5 gives the tribunal wide powers “to regulate its own procedure.” But this to my mind relates to the way in which an appeal is conducted. On its face these words do not extend to matters outside the conduct of an appeal. The making of an interim direction that an amount be credited to an appellant seems to me to a matter outside the conduct of the appeal. I agree with Mr Cunningham that in the light of CPR 25, one would expect express provision for such a power.

37.     Rule 2 sets out the objective that the tribunal deal with cases justly and fairly. That reflects to my mind the requirement for a fair trial in Art 6 of the Convention in relation to the conduct of the case before it. (As a result in relation to the conduct of an appeal it may provide greater protection to an individual than Ferrenzi says is provided by Art 6.) Does Rule 2, read as reflecting Art 6, require a construction of Rule 5 which would enable the tribunal under its Rules to make the type of direction Mr Young seeks?

38.     If it were necessary to ensure a fair trial that such a reading be made, then, because such a reading would go with the grain of the legislation, it seems to me that it could be permissible to read into the rules, and their enabling legislation such a power.  However it does not seem necessary to do so.

39.     Given that it was accepted in R (on the aplication of Teleos) that HMRC have some power to make a payment of disputed input VAT, and that the exercise of that power is subject to judicial review by the High Court, it seems to me that there is a forum in which the substantive effect of what Mr Young seeks can be made subject to effective judicial review, and there is no need for that effect to be read into the tribunal’s rules.

Conclusion

40.     I conclude that there is no power stemming from section 83 VATA or from the tribunal’s rules to make the direction for input tax credit which is sought.

41.     Even if the tribunal had such a power, I would not have exercised it in this case. This was not a case where there were national provisions of general application which could be disproportionate, but one dependant on particular facts. There was no evidence before me of the impecuniosity of the Appellant, and I think that there would have to be some consideration of the merits of the Appellant’s case before such a direction should be made.

b) that the Respondents be debarred from adducing evidence of fraud.

42.     It seems clear that this direction is within the power of the tribunal. Indeed the rules themselves provide in certain case for the Respondents to be debarred from taking any further part in the proceedings (see Regulation 8(7)).

43.     But in a Kittel appeal the onus is on the Respondents to prove the essential elements – fraud and knowledge (see Mobilx). Such a direction would therefore automatically cause the appeal to be allowed. That in my view would be neither fair nor just in the present circumstances.

44.     I decline to make this direction.

c) such other directions as the tribunal thinks fit.  

45.     The principle that the onus is on the Respondents to prove their case in this appeal places a different complexion on the Appellant’s difficulties from that found in a more usual tribunal case where the onus is on the Appellant. An Appellant in such circumstances may need less help in the marshalling of his evidence.

46.     I shall make in a separate document direction for the preparation and hearing of this appeal. Those directions are made in the light of an expectation that the Appellant is short of funds with which to prosecute its appeal, and in the light of the principle discussed in the preceding paragraph.