Organised missing trader fraud(“MTIC”) or carousel fraud

6. Perhaps out of this simple missing trader fraud, which we shall refer to as acquisition fraud, and not to be confused with it, was born a much more sophisticated fraud. It is referred to as MTIC (for 'missing trade intra-community') fraud or carousel fraud. This fraud dispenses with the genuine market: the defaulter creates an artificial market. Therefore, a genuine market does not limit the extent of the fraud: on the contrary, this fraud can be committed as often as the fraudster desires – at least until suspicions are raised. It is a pernicious fraud as it has no natural limit other than the pockets of the governments of EU member States.

7. As it relies on an artificial market, how does the fraudster realise his profit? The profit in an acquisition fraud arises by the missing trader running off with the VAT generated by a genuine sale onto a genuine market. The market in MTIC fraud is artificially generated: the fraudster organises the purchases and sales of the goods so the goods and money are likely to move in a circle of transactions beginning and ending with the fraudster or a person acting on his behalf. So merely running off with the VAT would be pointless as, logically, the money in the chain will have originated with the fraudster (even if it passes through the hands of innocent dupes caught up in the artificial chain).

8. For this fraud to be profitable, it relies on not only the VAT free acquisition by a trader of the goods within the UK but a VAT free sale of the goods out of the UK. The VAT free sale by the exporter (the 'broker') to another EU country, which entitles the broker to recover VAT paid to his supplier, is the key to this fraud.

9. Perhaps the simplest explanation of this fraud is that its object is to induce HMRC to refund to the broker VAT that was never actually paid to HMRC by the broker’s (ultimate) supplier.

10. The person making the cross-border sale is the lynchpin of the fraud, whether or not he knows it. He doesn't have to understand his role. As long as the broker, when selling the goods pays his vendor more than he receives from his buyer, the fraudster is able to extract the fraudulent profit.

11. For the fraud to work the broker has to be induced to pay more than he receives: in other words he has to be induced to put some of his own money into the chain. He may be induced to do this if there is profit in it. The broker's buying price includes VAT but his selling price does not. But if his net buying price is less then his selling price he will make a profit as long as HMRC refund the VAT.

12. And of course the fraud is lucrative for the fraudster as the fraudster causes the acquirer to default on the VAT on the importation (or 'acquisition') in the UK. So it is essential that there is still a missing trader. But the missing trader is not the lynchpin of this fraud: the object of the fraud is the broker's VAT reclaim.

13. In this artificial market, the goods are bought and sold but there is no real market for the goods. For this type of fraud it is not even necessary for the goods to actually exist. (We note in passing that there is no allegation in this case that the goods in EES's supply chains did not exist).

14. The fraud as described does not depend on the broker knowing that his role is vital to a fraud. It is possible that, so far as the broker is aware, he is simply buying and selling goods at a profit. Whether any particular alleged broker is aware of the fraud (if proved) is a question of fact.

15. As MTIC fraud and acquisition fraud both involve missing traders it is easy to confuse them although they are two very different frauds. An analysis of cases indicates that even the courts have not always appreciated the difference. For instance, Lewison J in Livewire at [96] said 'what is extracted from the public revenue is not the repayment of VAT at the end of the chain, but the VAT for which the defaulter should have accounted but did not.' which is a description of acquisition fraud and not MTIC fraud.

Why sometimes termed ‘carousel fraud’

16. The fraudster is arranging a chain of transactions in which the sale to and by the broker is essential for the fraud to work. The sale to and by the broker is the lynchpin of the fraud, its raison d'etre. So the fraudster has to arrange a sale to the broker and a sale by the broker. Therefore, ultimately a company (or companies) controlled by the fraudster must directly or through buffers sell the goods to the broker, and directly or through buffers buy the goods back from the broker.

17. As the fraud has no limit, it made sense for the fraudster to re-use the same goods and the same brokers and commit the fraud as often as possible sending the same goods round the same transaction chain. This gave the fraud its name of “carousel” fraud because the goods may go round in circle. But it is often a misnomer. Although the transaction chain (or at least the chain of money as the goods may not exist) must start and end with the fraudster or a company or person controlled by him, it is not necessarily the same person or company at the start and end of each chain. Further, the fraudster is likely to use a large number of buffers and brokers in lots of different chains in order to commit the fraud as often as possible. Therefore, although the same goods may circulate many times, they do not necessarily pass through the hands of the same broker more than once.

Variations on a theme

18. As we have said the fraud could be very lucrative and theoretically without limit. In practice though there might be a finite limit of brokers with resources to buy at one gross price and sell at a lower gross price, funding the difference from their own resources pending the VAT repayment by HMRC. The fraudster, therefore, might take a hand in this and put the broker in funds. The fraudster might arrange for loans or other funding to be made available to the broker.

Protecting the broker

19. It will be important to the fraudster (even where the broker is entirely independent of the fraudster) that the broker recovers its input tax (or at least believes that he will) because otherwise the broker will not buy the goods. The fraudster must want to protect the brokers he uses, as a fraud takes effort to organise and it must be easier if the same broker can be used in a transaction chain time and time again.

20. The first and most common method of protecting the broker’s input tax reclaim was to introduce buffers in the chain between the defaulter and the broker so that the broker was not purchasing directly from the defaulter, nor the broker selling directly back to the fraudster. Of course, the buffers themselves may not understand that their transaction was part of a series of transactions organised for the purpose of fraud.

21. Buffers offered some protection because if HMRC investigated the broker's purchase, it would not be obvious that it was connected to an earlier default, and the hallmark of MTIC trading, as described above, is a default by an acquirer or importer or goods (although, as I have said, the default is not the object of the fraud). A fraudster must realise that if HMRC did not find an earlier default, they would be very unlikely to suspect that the broker's trading was engineered for the purpose of fraud.

22. The fraudsters then invented a more sophisticated method of distancing the broker from the default. This was to remove the default from the broker's chain. As explained above, although the object of the fraud would be the broker's input VAT (or at least that in cash terms the broker would, relying on a future input tax reclaim, pay more in cash for the goods than he receives), the fraud also relies on the acquirer defaulting on the VAT due on the original acquisition (or pretend acquisition) of the goods. There are two ways the UK acquirer could avoid paying this VAT: the original method was, as already described, to evade the VAT by defaulting and going missing (thus ‘missing trader fraud’). Alternatively, it could itself act as a broker (ie a UK despatcher of goods to Continental EU) in respect of different goods, and use the input tax claim generated by that sale to offset the output tax liability generated by the acquisition in the first chain.

23. This second chain, referred to as a ‘contra chain’ or ‘dirty chain’ would involve a similar carousel of goods (existent or non-existent) with a default by the UK acquirer. The fraud was fundamentally the same fraud with the same opportunity for profit for the fraudster. But if HMRC looked at the chain of supply down to the original broker they would find that the acquirer (referred to in MTIC speak as a ‘contra trader’) had not gone missing owing substantial VAT: they would find an acquirer-cum-broker who had completed a VAT return showing output tax netted off against input tax.

24. However, if HMRC were to trace back the broker transactions which gave the acquirer-cum-broker (the contra trader) its input tax claim (which we will refer to as the 'dirty chain'), they would find that these traced back to a default.

25. In this case, the allegation is that there were both normal chains involving buffers, contra-trading chains where the contra-traders' broker deals traced back to a default and a third type of chain. That third type is alleged to be where the contra-traders' broker deals traced back to other contra-traders, whose own broker deals traced back to a default. HMRC give this trading the name of ‘second line contra-trading’. We consider later whether these allegations are made out at §XXX below.

26. At root, MTIC fraud involving contra trading (if proved) is the same as ordinary MTIC fraud. The fraudster’s object is exactly the same: to induce the broker to pay more for the goods than he receives by relying on a VAT refund from the tax authorities. And the fraud relies on no VAT actually ever being paid to HMRC, whether the default is in the same chain or a different chain. Whether the contra-trader or broker knows (or ought to know) that they are participating in a fraud are questions of fact in any individual case.

The law

27. There are 31 transactions at issue in this appeal in which HMRC have denied the appellant's claim to input tax recovery, amounting, as we have said to a denial of input tax of approximately £6.2 million. As we have said HMRC have accepted that the appellant would be entitled under EU and UK VAT law to recover this VAT except that they consider that the legal principle stated by the Court of Justice of the European Communities (“CJEU”) in Kittel v Belgium; Belgium v Recolta Recycling SPRL (Joined cases C-439/04 and C-440/04) (“Kittel”) is engaged and that such input tax is denied to the appellant because it (via its director) knew or ought to have known at the time of them that these 31 transactions were connected with fraud.

28. In Kittel the CJEU held that:

“49. The question whether the VAT payable on prior or subsequent sales of the goods concerned has or has not been paid to the Treasury is irrelevant to the right of the taxable person to deduct input tax.….

51. Traders who take every precaution which could reasonably be required of them to ensure that their transactions are not connected with fraud, be it the fraudulent evasion of VAT or other fraud, must be able to rely on the legality of those transactions without the risk of losing their right to deduct the input VAT.

….54. … preventing tax evasion, avoidance and abuse is an objective recognised and encouraged by the Sixth Directive (see Joined Cases C-487/01 and C-7/02 Gemeente Leusden and Holin Groep [2004] ECR I-5337, paragraph 76. Community Law cannot be relied on for abusive or fraudulent ends (see, inter alia, Case C-367/96 Kefalas and Others [1998] ECR I-2843, paragraph 20; 373/97 Diamantis [2000] ECR I-1705, paragraph 33; and Case C-32/03 Fini H [2005] ECR I-1599, paragraph 32).

55. Where the tax authorities find that the right to deduct has been exercised fraudulently, they are permitted to claim repayment of the deducted sums retroactively (see, inter alia, Case 268/83 Rompelman [1985] ECR 655, paragraph 24; Case C-110/94 INZO [1996] ECR I-857, paragraph 24; and Gabalfrisa, paragraph 46). It is a matter for the national court to refuse to allow the right to deduct where it is established, on the basis of objective evidence that that right is being relied on for fraudulent ends (see Fini H, paragraph 34).

56. 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 the 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.

57. That is because in such a situation the taxable person aids the perpetrators of the fraud and becomes their accomplice.

58. In addition, such an interpretation, by making it more difficult to carry out fraudulent transactions is apt to prevent them.”

29. The CJEU summarized the position as follows:

“61... Where it is ascertained, having regard to objective factors, that the supply is to a taxable person who knew or should have known that by his purchase, he was participating in a transaction connected with the fraudulent evasion of VAT, it is for the national court to refuse that taxable person entitlement to the right to deduct.”

30. In the earlier case of C-354/03 Optigen v HMRC [2006] the CJEU was asked to give a ruling on issues relating to the recoverability of input tax in circumstances where the traders were innocently caught up in a chain of transactions which were fraudulent. The CJEU concluded that:

“47 Each transaction must be regarded on its own merits and the character of a particular transaction in the chain cannot be altered by earlier or subsequent events.

…..

51 Transactions which themselves are not vitiated by VAT fraud constituted supplies of goods or services, and where an economic activity within the relevant legislation, where they fulfil the objective criteria on which the definitions of those are based, regardless of the intention of the trader other than the taxable person concerned, involved in the chain of supply, and/or the possible fraudulent nature of another transaction the chain, prior or subsequent to the transaction carried out by the taxable person of which the taxable person had no knowledge and no means of knowledge.

……

55 The right to deduct input VAT by a taxable person who carries out such a transaction can XXX be affected by the fact that in the chain of supply, of which those transactions form part, another prior or subsequent transaction is vitiated by VAT fraud, without that taxable person knowing or having means of knowing.”

31. Kittel was considered by the Court of Appeal in the conjoined appeals of Mobilx Ltd (in Administration) v HMRC; HMRC v Blue Sphere Global Ltd (“BSG”); Calltel Telecom Ltd and another v HMRC [2010] EWCA Civ 517 (“Mobilx”).

32. Moses LJ made clear that this refusal of the right to deduct does not depend on any specific UK legislation. Moses LJ stated:

“43. A person who has no intention of undertaking an economic activity but pretends to do so in order to make off with the tax he has received on making a supply, either by disappearing or hijacking a taxable person's VAT identity, does not meet the objective criteria which form the basis of those concepts which limit the scope of VAT and the right to deduct (see Halifax § 59 and Kittel § 53). A taxable person who knows or should have known that the transaction which he is undertaking is connected with fraudulent evasion of VAT is to be regarded as a participant and, equally, fails to meet the objective criteria which determine the scope of the right to deduct.”

33. On the meaning of “should have known” Moses LJ said:

“50. The traders contend that mere failure to take reasonable care should not lead to the conclusion that a trader is a participant in the fraud. In particular, counsel on behalf of Mobilx contends that Floyd J and the Tribunal misconstrue § 51 of Kittel. Whilst traders who take every precaution reasonably required of them to ensure that their transactions are not connected with fraud cannot be deprived of their right to deduct input tax, it is contended that the converse does not follow. It does not follow, they argue, that a trader who does not take every reasonable precaution must be regarded as a participant in fraud.

51. Once it is appreciated how closely Kittel follows the approach the court had taken six months before in Optigen, it is not difficult to understand what it meant when it said that a taxable person “knew or should have known” that by his purchase he was participating in a transaction connected with fraudulent evasion of VAT In Optigen the Court ruled that despite the fact that another prior or subsequent transaction was vitiated by VAT fraud in the chain of supply, of which the impugned transaction formed part, the objective criteria, which determined the scope of VAT and of the right to deduct, were met. But they limited that principle to circumstances where the taxable person had “no knowledge and no means of knowledge” (§ 55). The Court must have intended Kittel to be a development of the principle in Optigen. Kittel is the obverse of Optigen. The Court must have intended the phrase “knew or should have known” which it employs in §§ 59 and 61 in Kittel to have the same meaning as the phrase “knowing or having any means of knowing” which it used in Optigen (§ 55).

52. If a taxpayer has the means at his disposal of knowing that by his purchase he is participating in a transaction connected with fraudulent evasion of VAT he loses his right to deduct, not as a penalty for negligence, but because the objective criteria for the scope of that right are not met. It profits nothing to contend that, in domestic law, complicity in fraud denotes a more culpable state of mind than carelessness, in the light of the principle in Kittel. A trader who fails to deploy means of knowledge available to him does not satisfy the objective criteria which must be met before his right to deduct arises.”

34. He concluded:

“59. The test in Kittel is simple and should not be over-refined. It embraces not only those who know of the connection but those who “should have known”. Thus it includes those who should have known from the circumstances, which surround their transactions that they were connected to fraudulent evasion. If a trader should have known that the only reasonable explanation for the transaction in which he was involved was that it was connected with fraud and if it turns out that the transaction was connected with fraudulent evasion of VAT then he should have known of that fact. He may properly be regarded as a participant for the reasons explained in Kittel.60. The true principle to be derived from Kittel does not extend to circumstances in which a taxable person should have known that by his purchase it was more likely than not that his transaction was connected with fraudulent evasion. But a trader may be regarded as a participant where he should have known that the only reasonable explanation for the circumstances in which his purchase took place was that it was a transaction connected with such fraudulent evasion.”

35. With regard to the burden of proof Moses LJ said:

“81. It is plain that if HMRC wishes to assert that a trader’s state of knowledge was such that his purchase is outwith the scope of the right to deduct, it must prove that assertion. No sensible argument was advanced to the contrary.

Circumstantial Evidence and an overview

36. HMRC's position is that this Tribunal, when answering the questions posed by the CJEU in Kittel, should look at an overall picture. HMRC have authority on their side. Lord Justice Moses in Mobilx said

“[82] But that is far from saying that the surrounding circumstances cannot establish sufficient knowledge to treat the trader as a participant. ….Tribunals should not unduly focus on the question whether a trader has acted with due diligence. Even if a trader has asked appropriate questions, he is not entitled to ignore the circumstances in which his transactions take place if the only reasonable explanation for them is that his transactions have been or will be connected to fraud. The danger in focusing on the question of due diligence is that it may deflect the Tribunal from asking the essential question posed in Kittel, namely, whether the trader should have known that by his purchase he was taking part in a transaction connected with fraudulent evasion of VAT. The circumstances may well establish that he was.”

37. The Court went on to say:

“[84] ….circumstantial evidence …. will indicate that a trader has chosen to ignore the obvious explanation as to why he was presented with the opportunity to reap a large and predictable reward over a short space of time.

[85] A trader who chooses to ignore circumstances, which can only reasonably be explained by virtue of the connection between his transactions and fraudulent evasion of VAT participates in that fraud and, by his own choice, deprives himself of the right to deduct input tax.”

38. Lord Justice Moses also adopted the words of Clarke J in Red 12 [2009] EWHC 2563 (Ch) where he said:

“[109] Examining individual transactions on their merits does not, however, require them to be regarded in isolation without regard to their attendant circumstances and context. Nor does it require the tribunal to ignore compelling similarities between one transaction and another or preclude the drawing of inferences where appropriate....The character of an individual transaction may be discerned from material other than the bare facts of the transaction itself, including circumstantial and 'similar fact' evidence...”

39. Mr Bridge considers that Moses LJ and Clarke J were wrong. He says the Tribunal can only under EU law consider the matter on a transaction by transaction basis. He implies the tribunal could not consider anything other that the direct facts of the each transaction itself when considering the appellant's right to input tax recovery and that in particular that the Tribunal cannot rely on circumstantial evidence.

40. This Tribunal is bound by the law of precedent and even if we agreed with Mr Bridge's view of the law as stated above (and we do not) we would be bound to follow what Lord Justice Moses and Clarke J had ruled. Mr Bridge, however, says that this Tribunal is not bound by the law of precedent as CJEU decisions override Court of Appeal authority. This is not right either (see Newey J in S & I Electronics plc [2012] UKUT 87 (TCC) at $$13-19) but it does not matter as Mr Bridge cannot point to any CJEU authority that prohibits a Tribunal from considering circumstantial evidence

41. While Mr Bridge is right that the Tribunal must consider in respect of each individual transaction whether the appellant has met the conditions for input tax entitlement in respect of that particular transaction, and Moses LJ did not suggest otherwise, the Tribunal can and must consider the whole picture, including circumstantial evidence, when deciding whether in respect of each individual transaction the conditions for input tax recovery have been met. And we will do so.

Allocation of tax loss?

42. It was accepted (and clear to the Tribunal from the evidence of the officers) that in around 2007 HMRC had a policy of allocating tax loss to brokers. In other words HMRC would seek to relate a specific denial of input tax to a specific default by a defaulter. Later HMRC dropped this policy.

43. The Tribunal considers HMRC's policy irrelevant to the questions at issue in this Tribunal. In so far as HMRC's policy was originally applied to the appellant, this was after the deals at issue in this appeal and did not in any way contribute to them. So as a matter of fact it is irrelevant. And HMRC's policy does not make the law.

44. The appellant's position is simply that HMRC's original policy was right. In the appellant's view the Tribunal should allow the appellant's appeal, even if HMRC prove knowledge of connection to fraud, unless the Tribunal is also satisfied that the disputed input tax claim can be allocated to a proved default which has been assessed and that that assessment has not been paid.

45. The legal basis on which the appellant makes this submission is that any other conclusion would make the application of Kittel penal in nature as it might result in HMRC recovering more than they had lost. The appellant refers to the general principles of EU law such as fiscal neutrality and proportionality, and the decision in Elida Gibbs (C-317/94).

46. We reject this argument. It amounts to an attack on the CJEU's decision in Kittel. The CJEU, as the ultimate interpreter of the principles of EU law, and alive to the principles of equivalence, fiscal neutrality and proportionality, decided in Kittel that a person with knowledge of connection to fraud was not entitled to recover its input tax. That decision itself elucidates a principle of EU law: it cannot be attacked as inconsistent with other principles.

47. As the CJEU did not require allocation of tax loss as a precondition of the application of the rule in Kitttel, then such allocation is not required as a matter of law. Not only was there nothing express about allocation in Kittel, allocation would be inconsistent with the reasoning on which the CJEU made the decision. That reasoning was, as set out above:

“[56] 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 the 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.

[57] That is because in such a situation the taxable person aids the perpetrators of the fraud and becomes their accomplice.

[58] In addition, such an interpretation, by making it more difficult to carry out fraudulent transactions is apt to prevent them.”

48. Thus it is clear that the reasoning of the decision was not that the member state should be compensated for tax loss but that persons who are to blame, either because they knew of the fraud or because they ought to have known of the fraud, are to be discouraged from proceeding with such transactions. In that sense it was intended to be a penal provision: a carrot and stick approach. The risk of losing the right to input tax is meant to discourage persons from entering into transactions which they know are connected to fraud.

49. As that was the CJEU's explicit reasoning, it is clear that allocation of tax loss was neither expressly nor impliedly required as a pre-condition of the application of the rule in Kittel.

50. In any event, we are bound to come to this conclusion as it is consistent with higher authority: Floyd J in Calltell [2009] EWHC 1081 (Ch) said:

“[96] In my judgment there is no principle which requires HMRC to acknowledge a claim to repayment to the extent that the claim exceeds HMRC’s tax loss. Firstly…the correct unit of fiscal analysis is not th entire chain by the individual transaction. …The question is accordingly whether the taxpayer has or does not have the right to decuct or reclaim his input tax in respect of an individual transaction. Consideration of this question does not justify recourse to the overall fiscal impact on HMRC of all the transactions in the chain.

[97] Secondly, none of the statements in Kittel suggest that the right is lost only to the extent that tax is lost elsewhere in the chain. … but once it is established that a taxpayer has, by his purchase, participated in the fraudulent evasion of VAT, it seems to me to be impossible to argue that, by withholding repayment of VAT in respect of that very purchase the taxpayer is being subjected to a disproportionate remedy. In fact, to use the VAT legislation to achieve any benefit from such a purchase seems to me to be wrong in principle.

[98] Thirdly, although fiscal neutrality is a fundamental feature of the system of VAT….the fiscal neutrality of an individual transaction will, as Kittel shows, have to give way to the objective of combating fraud.

[99] It seems to me that the objective of not recognising the right to repayment is not simply to ensure that the exchequer is not harmed by fraud: the objective includes combating fraud and discouraging taxpayers from entering into transactions of this nature. In that context, considerations of fiscal neutrality of the impugned transaction are, it seems to me, besides the point.”

51. In Moblix Moses LJ said:

[65] …It is true that there may well be no correlation between the amount of output tax of which the fraudulent trader has defrauded HMRC and the amount of input tax which another trader has been denied….

52. It is not necessary for HMRC to allocate the input tax claimed by the appellant to a specific amount of lost tax. For the first part of Kittel, they need merely to show the appellant’s transaction from which the input tax claim arose was connected to fraudulent tax loss.

53. While this is jumping ahead and refers to our findings of fact made below, and is not necessary, in any event we are satisfied that the evidence produced by HMRC shows that all the appellant's claimed input tax can be allocated to a tax loss actually suffered by HMRC. The appellant has accepted the evidence that 23 out of its 31 deal chains trace back to a fraudulent tax loss in the same deal chains. And so far as the remaining 8 'clean' chains are concerned, where connection to tax loss is not accepted, it is overwhelmingly clear that the appellant's transactions were part of an organised MTIC fraud (see §§198-200) so it is considerably more likely than not that all of them traced back to a fraudulent tax loss. In any event, we have also accepted Mr Humphries' evidence (see §§150 & 154) and Mr Humphries' evidence was that, overall, the 'Cell 5' defaults amount to virtually the same amount of money as all the 'Cell 5' broker's VAT reclaims. In other words, each input tax reclaim in 'Cell 5' including the appellant's 8 in dispute, can be matched to fraudulent tax loss. We find, were it necessary, which it is not, allocation to tax loss is proved in respect of all 31 deal chains.

Contra trading and Kittel

54. It is also, as we understand it, Mr Bridge's position that even if HMRC could show that the acquirers in the alleged contra chains were fraudulent defaulters, there is no connection to the appellant's purchases. HMRC, says Mr Bridge, having found a tax loss in one chain, are just looking to recoup their losses by pinning the loss on a company in an unrelated chain for no better reason than because someone in the appellant's chain happens to have traded at some point with someone in the acquirer's chain. And the connection would be even more remote with second line contra trade.

55. We think that this is a question of fact. If the trade by the contra trader with the chain in which the default occurred was a matter of chance, then Mr Bridge would have a point here. As a matter of law, there would not be 'connection' in the sense of facilitation of fraud which, as we discuss below at §§73-76, is what we think the CJEU meant in Kittel. However, if as a matter of fact, HMRC can show that the connection was not a matter of chance, but part of an overall organised fraud then there is nothing in what Mr Bridge says. The connection between the appellant's chain and the contra chain would have been deliberately engineered for the purpose of fraud and, however many traders and chains intervened between the broker's reclaim and the defaulter's default, they would be connected in the Kittel sense.

56. The appellant does not agree: its case, put by Mr Lasok in closing, is that as a matter of law the rule in Kittel did not extend to a connection to fraud via contra trading. For this proposition they relied on CJEU cases released after Kittel and after the Court of Appeal decision in Mobilx. In particular, Mr Lasok relied on dicta of the CJEU from Mahageben (C-80/11) that the relevant connection had to be with the supplier or another trader higher in the same supply chain.

57. Mr Lasok's first point is that none of the cases decided by the CJEU actually involved contra-trading so, he says, there is no authority to deny input tax where the connection to fraud is solely by a contra trading chain. However, we do not agree. The doctrine in Kittel applies wherever there is connection to fraud of which the claimant knew or ought to have known. That the factual position in this case is not identical to those considered by the cases raised before the CJEU does not prevent the application of the doctrine.

58. In Mahageben, the CJEU said the input tax could only be denied:

'on the basis of objective evidence, that the taxable person concerned knew, or ought to have known, that the transaction relied on as a basis for the right to deduct was connected with fraud committed by the issuer of the invoice or by another trader acting earlier in the chain of supply.’

59. In Bonik (C285/11), the CJEU said:

'a taxable person may not be refused the right to deduct VAT in relation to a supply of goods on the ground that, in view of fraud or irregularities committed upstream or downstream of that supply, the supply is considered not to have actually taken place, where it has not been established on the basis of objective evidence that the taxable person knew, or should have known, that the transaction relied on as a basis for the right of deduction was connected with VAT fraud committed upstream or downstream in the chain of supply – a matter which it is for the referring court to determine.' (our emphasis)

60. In Toth (C- 324/11) the CJEU said:

“[53]....in a situation such as that at issue in the main proceedings, the right to deduct may be refused only where...the addressee of the invoice knew or should have known that the transaction relied on as a basis for he right to deduct was connected with a fraud committed by the issuer or another operator supplying inputs in the chain of supply.”

61. Mr Lasok and Mr Bridge's interpretation of all these cases is that the doctrine in Kittel only applies if the VAT default is in respect of the same goods which the appellant later buys and sells. In respect of 8 of the transactions at issue in the appeal HMRC do not allege that the default was in the same chain of transactions. Therefore, say Mr Lasok and Mr Bridge, HMRC must repay the input tax to the appellant in respect of those 8 transactions. Is this the correct interpretation of what the CJEU said?

Interpretation of CJEU decisions

62. As repeated experiences should show, national courts must not read what the CJEU says in its decisions as if it was statute law. Great care should be taken before extrapolating what the CJEU says in one case to another case.

63. The CJEU will often summarise what it has said in earlier cases using different words: it is wrong to assume that when it does this it intended to widen or narrow the scope of what it said earlier. This is especially the case where the difference in wording concerns an issue not relevant in the appeal which the court was actually dealing with. Its judgments should not be read like statutes, particularly when the court is merely referring to an established doctrine rather than elucidating it.

64. In none of the cases mentioned by counsel was there a question of contra trading and it should not be supposed the CJEU intended to make any comment on it. Moreover, there was no decision in Mahageben or any of the other cases relied on by counsel that the doctrine in Kittel was limited to situations where the fraud was in the same supply chain: none of them concerned a situation where the fraud was not in the supply chain. On the contrary, in all the cases the fraud was committed by the appellant's immediate supplier. And so far as Bonik is concerned, this is even clearer as can be seen from the italicised part of the above quotation from the decision. In that case the CJEU made explicit that its reference to knowledge of fraud in the supply chain was limited to a case where the fraud actually was in the supply chain. The CJEU said nothing in Bonik about the requisite knowledge when the fraud did not take place in the same supply chain.

65. The appellant's case amounts to saying that in Mahageben, Bonik and Toth the CJEU chose to narrow the scope of Kittel to limit it to cases where the fraud was in the same supply chain. Yet not only is this inherently an unlikely thing for the CJEU to do as they were not being called on to consider fraud outside the supply chain, in none of these cases did the CJEU require the Advocate General to give an Opinion. Yet they ought to have done so if they thought they were making a significant decision qualifying an earlier judgment. As pointed out by Lord Reed in Aimia Coalition [2013] UKSC 15, in the context of entirely different CJEU case-law than that at issue in this case, the CJEU's constitution requires it to have the opinion of an Advocate General if a new point of law is concerned:

[34] “[the CJEU] appears to have considered that both cases alike involved the straightforward application of established principles, since it determined them without a submission from the Advocate General. In terms of article 20, paragraph 5 of its Statute, it may do so only ‘where it considers that the case raises no new point of law’.”

66. So it follows that in Mahageben, Bonik, etc as the CJEU did not ask for an opinion from an Advocate General it did not consider that these cases involved any new point of law. It follows that they did not intend anything said in those cases to qualify what they had earlier said in Optigen and Kittel.

67. While not strictly forming precedent, we note that this is what Moses LJ said in refusing the appellant's application for permission to appeal to the Court of Appeal from the Upper Tribunal decision in POWA (Jersey) Ltd (unrep 30 January 2013):

“[11]...if the [CJEU] intended to cut down the principle it had identified in Kittel and was changing the law, it would have said so. On the contrary it was not. It was merely applying it.”

68. This alone would be sufficient to dispose of the appellant's arguments that any of the later cases it relied on qualified Kittel, but there are many other reasons which make it even clearer how erroneous were the submissions of the appellant's counsel on this.

What is the rationale to the Kittel decision?

69. Notwithstanding these points of construction, Mr Lasok’s view is that the CJEU did intend to exclude contra-trading from the rule in Kittel. A line has to be drawn somewhere, says Mr Lasok, because ultimately every transaction will be connected in some way to some kind of fraud committed by someone somewhere and Mr Lasok considers it sensible to draw the line at contra trading.

70. We agree with neither proposition: there is no risk of Kittel ultimately disallowing input tax in all transactions nor that if there were such a risk, would it be sensible to draw the line to exclude contra trading.

71. On Mr Lasok’s first limb, Kittel has very natural limits. While a broad interpretation of “connection” would bring in many transactions with some remote connection to a fraud committed by someone else much earlier or later in the supply chain, or not in the supply chain at all, the requirement for the trader to have known or have the means of knowledge of the fraud would be in most cases mean there would be no grounds on which to deny input tax.

72. Nevertheless, it is possible to postulate circumstances in which a trader might know of fraud by someone else but nevertheless it would not seem rational to deny the trader input tax recovery. Counsel tried to give an example but it was rather convoluted. We would suggest an obvious example is a purchase from the liquidator of a company which was insolvent due to assessments following a fraudulent failure to account for VAT on its acquisitions. A buyer might well know of the fraud committed by its vendor prior to its liquidation, but why should that disentitle the buyer from VAT recovery? The sale does not facilitate the fraud, but instead realises assets to meet the insolvent company’s liabilities. The company’s creditors, anxious that as much value as possible be realised from the company's assets, will include HMRC.

73. Our view is that the natural limit of Kittel is not only the requirement of knowledge or means of knowledge but a restricted meaning to the word ‘connection’. We think it was meant by the CJEU to mean 'facilitate'. So the Kittel question is whether the trader seeking input tax deduction knew or ought to have known his transaction was connected to (in the sense of facilitating) fraud.

74. As already cited above, the justification given by the CJEU in Kittel for denial of input tax by someone who was not the fraudster, was as follows:

“[56]…. 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 the 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.

57. That is because in such a situation the taxable person aids the perpetrators of the fraud and becomes their accomplice.

58. In addition, such an interpretation, by making it more difficult to carry out fraudulent transactions is apt to prevent them.”

75. In [57] and [58] the CJEU refers to aiding the perpetrators, and refusing input tax recovery to knowing parties to make it more difficult to commit fraud. The underlying assumption the CJEU is making that the participation of the taxable person making the input tax reclaim actually facilitated the fraud: if it did not then the appellant would not need penalising for aiding the perpetrator of the fraud, nor would refusing the input tax recovery make it more difficult for the fraudster to carry out the fraud. Here, therefore, is the guide to interpretation of what the CJEU meant by 'connected'. It meant it in the sense of a connection which facilitated the fraud. This is therefore the natural extent of the decision in Kittel. Not every remote connection to fraud is relevant: it is only the connections which facilitate the fraud which matter.

76. On this view, the CJEU’s decision is logical and easy to apply. It makes no difference how the transaction is connected to fraud – whether by a dirty chain or a contra chain - as long as it facilitated the fraud. Indeed this analysis of Kittel shows that a connection to fraud via a contra trade chain most definitely would be connected in the Kittel sense as the broker's transaction not only facilitated the fraud but was the lynchpin of it (as explained above in §§22-26).

77. But in the example at §72 above, the trader buying from the liquidator would not be making a purchase connected to fraud, because, while he knows of the fraud, that fraud was not facilitated by his purchase.

78. Mr Lasok also postulated it made sense to draw the line to exclude contra trading from Kittel. We do not agree. Contra trading is merely a more sophisticated means than using buffer traders of distancing the broker from the default. In essence it is exactly the same fraud as 'ordinary' MTIC fraud as described above in §§6-26. So far from being sensible, it would be perverse to draw a distinction between ordinary MTIC fraud with lots of buffers and the more sophisticated version using contra trading.

79. And moreover, the CJEU is concerned with what the broker knew or ought to have known of the connection to fraud. Yet from the point of view of the broker, whatever his state of knowledge about connection to fraud, he would not necessarily perceive any difference between dirty and contra chains. Even if the broker was 'in the know' about the fraud, he might not be told by the person organising the fraud whether the chain in which he was making a purchase and sale involved a defaulter or contra trader. So far as he was concerned, his role in the chain was identical whether or not the default was in that chain or a chain linked by a contra trader.

80. Lastly, it would be a perverse place to draw the line as Mr Lasok’s proposition is that would mean (so far as EU law was concerned) that it would be effective for fraudsters even today to set up contra trading MTIC deals safe in the knowledge that even if the tax authorities discover that the broker is “in the know” about the fraud nevertheless the broker must be repaid its input tax. If Mr Lasok's interpretation of Kittel were accepted, it would be a green light to fraudsters to commit unlimited fraud and bankrupt the EU.

Is there fraud in the “clean” supply chain anyway?

81. We have roundly rejected Mr Lasok’s arguments that Kittel requires the fraud to be in the same supply chain with the sale and purchase by the broker.

82. But, not that this defunct argument needs a further nail in its coffin, just for the sake of completeness, we note that it is in any event based on a false premise. That premise is that the fraud occurs in a different supply chain. This is because the VAT default is in the ‘dirty’ chain. What Mr Lasok overlooks is that there is nevertheless still almost certainly fraud in the so-called ‘clean’ supply chain of which the broker's purchase and sale forms part.

83. This is because it would be extremely unlikely that the contra-trader could or did act as a contra-trader without knowing it was participating in and facilitating a fraud. This is because it has to offset its VAT liability by entering into broker transactions and, subject to the facts of any particular case, it is unlikely to do this by chance. Certainly in this case our finding is that the contra traders acted knowingly: see §§201-213. And if the contra-trader acted knowingly, then the contra-trader is committing fraud: it is claiming VAT refunds (by way of set-off) to which it knows it is not entitled because it knows its transactions are connected to fraud. The contra trader is committing fraud and it is in the same supply chain with the broker.

84. So the argument is based on a false premise in any event. Even in a ‘clean’ contra trading chain there is fraud, at least if the contra trader acted knowingly.

Common law

85. While it is not necessary to add even more nails to its coffin, we note in passing that the common law of this country would (irrespective of Kittel) apply to prevent a broker who knew its transactions facilitated fraud from maintaining its ‘right’ to reclaim input tax. It can not rely on its own fraud to assert a right generated by that fraud.

Reference to CJEU

86. Mr Lasok suggested that we should refer the matter to the CJEU if we were in any doubt that his propositions did not represent good law. However, we are in no doubt that they do not represent good law. We refuse to refer to the CJEU, whose answer to the question of whether a broker in a contra-trading 'clean' chain, knowing its transactions were connected to fraud, was entitled to recover its input tax, would, very predictably, be ‘no’.

Binding authority

87. We note that in any event and after the hearing in this case, the Upper Tribunal has in Fonecomp Limited [2013] UKUT 599 (TCC) also dealt with this same question. We are bound by its decision, which rejected the appellant’s arguments, as follows:

[24]…..The authoritative statement of the principle given by the Court at paras. [56] and [61] of its judgment in Kittel is not qualified in this way, and such an arbitrary and excessively narrow focus would not accord with the usual purposive approach to interpretation of EU legislation. Since it is readily possible to conceive of situations in which contra-trading occurs, whereby transactions in the so- called clean chain of supply can provide material assistance for VAT fraud in the so- called dirty chain of supply (as the FTT found had happened in this case), there is no basis at all for reading down and limiting the Kittel principle in the manner contended for by Mr Patchett-Joyce. The rationale and explanation given by the Court in Kittel for the principle it stated in that case apply with equal force in a contra-trading situation as in relation to a simpler type of case involving a single chain of supply. It should be borne in mind that the primary mechanism to control the application of the principle which disallows a trader from claiming input VAT is that the national authorities have to establish that he knew or should have known that his transaction was connected with the fraudulent evasion of VAT, and that mechanism is equally available to protect a trader in both a simple chain of supply case and in a contra- trading case.”

Knowledge of what?

88. What must the claimant be shown to know in order for HMRC to legitimately deny it input tax recovery? The appellant's position is that Lewison J established binding authority that the appellant must be shown to know of the dirty chain or be shown to know that the contra trader was fraudulent. Lewison J said at [19] in Livewire:

“[102] In my judgment in a case of alleged contra-trading, where the taxable person claiming repayment of input tax is not himself a dishonest co-conspirator, there are two potential frauds:

i) The dishonest failure to account for VAT by the defaulter or missing trader in the dirty chain; and

ii) The dishonest cover-up of that fraud by the contra-trader.

[103] Thus 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. I do not consider that it is necessary that he knew or should have known of a connection between his own transaction and both of these frauds. If he knows or should have known that the contra-trader is engaging in fraudulent conduct and deals with him, he takes the risk of participating in a fraud, the precise details of which he does not and cannot know.....”

89. However, we consider that, in so far as Lewison J was suggesting by this passage, that the claimant must be shown either to know the identity of the contra trader and know that he was fraudulently offsetting input tax against output tax or know the fact that a parallel dirty chain existed, Lewison J is out of line with other authority, which we prefer, and wrong in principle.

90. As a matter of authority, we prefer the Upper Tribunal in POWA (Jersey) Ltd [2012] UKUT 50 TCC where Mr Justice Roth said:

“[50] ...HMRC must establish that the fraudulent evasion of VAT took place, and if the form of fraud involved was contra trading then that is what they have to prove. But it is a misconception to consider that they must also establish that the party seeking to deduct input tax...should reasonably have known that its own transaction was connected to (or involved in) this particular form of missing trader fraud as opposed to another form....”

91. Other High Court authority, which we also prefer to Livewire, (such as Megtian Ltd [2010] EWHC 18 (Ch) at §§33-38 and Calltell [2009] EWHC 1081 (Ch) at §§79-82) are also clear that all HMRC have to show is that the claimant knew that his transaction was connected with fraud. There is no requirement to show that he knew the identify of the fraudster, the identity of the missing trader or even the nature of the fraud and in particular there is no requirement to show that the claimant knew that the connection to fraud was by a straight dirty chain to a default or a 'clean' chain connected to fraud by one or more contra trade chains.

92. As a matter of principle, the logic to the CJEU's decision in Kittel was that by undertaking a transaction knowing it was connected to fraud the claimant was right denied its input tax as it was aiding the perpetrators of the fraud: if the law required the claimant to know the complex details of the fraud, such as whether there was contra-trading and the identity of the defaulter, before it could be denied its input tax, then Kittel would be a hollow shell, and persons who clearly knew their transactions enabled a fraud to take place would nevertheless be able to reclaim input tax.

93. The appellant's position is that the question of knowledge is somehow different in a contra trading case than a straight connection to fraud. The appellant sees a contra trading connection as much more remote than a connection by the same goods to the default. Certainly the fraud is structured slightly differently depending on whether there is a single dirty chain, or one dirty chain combined with one or more 'clean' chains connected by contra traders. But it does not follow that, while the actual connection is different, that there is any qualitative difference. As we have explained, it is the same fraud, the broker's input tax reclaim is ultimately the object of the fraud (whether or not the broker is 'in the know'), and the only difference is the manner in which the fraudster has chosen to distance the broker from the linked default.

94. And, as we have said in §79, there may well be no qualitative difference from the point of view of the broker. A broker could be knowingly participating in the fraud, content to take, say, a large profit in return for accepting the risk that HMRC might withhold the VAT, yet be ignorant of identity of the defaulter, and ignorant of whether there are parallel dirty chains or whether the default is in his chain. He would not need to know any of this in order to actively and knowingly participate in the fraud. Whether of course EES was a knowing participator is a question of fact we address below. But in principle, a broker could be a knowing participant in the fraud without knowing whether the link to the default was through a single or multiple chains of goods and without knowing who the defaulter was.

95. We reject the appellant's case on this and chose to follow the Upper Tribunal in POWA (Jersey) Ltd.

The facts

The appellant

96. Having considered the relevant law, we move on to consider the facts.

97. EES was incorporated on 12 July 2001. Its first director was Mr Ben Landreth and Mr Anthony Wright was appointed director in July 2002. After Mr Landreth resigned in October 2003, Mr Wright continued as sole director. He is also now and at the time of the deals in issue the sole shareholder.

98. Therefore, at the critical time in this appeal, 2006, the company was the alter ego of Mr Wright. Knowledge possessed by Mr Wright is to be attributed to the company. We also find it was Mr Wright who conducted the deals at issue in this appeal.

99. Mr Wright was the director of a number of other companies, engaged, as was EES, in transactions involving mobile phones. For instance, Mr Wright became the director of Recycled 4 You Ltd (“Recycled”) in September 2004 and remains its director. Its business is, as the name suggests, refurbishing used mobile phones. In December 2003 Mr Wright became the director of Cellular Solutions (T Wells) Ltd but he sold this company (and resigned the directorship) in February 2006. There were a few other companies which he controlled.

100. Mr Wright via his various companies had been involved in mobile trading for about 13 years at the time of the transactions at issue in this appeal. His companies traded on the grey market and undertook transactions such as box breaking and arbitrage (see §238 below). After a break until 2007, EES continued to trade after the transactions at issue in this appeal and is registered for the reverse charge.

The fraud

101. This appeal relates to the input tax claimed on 31 purchases by EES in the period March-May 2006. In that period the appellant undertook 32 deals. HMRC allege all 32 are connected to fraud but have denied the input tax on only 31 of them. The appellant admits that 23 out of these 31 purchases trace back, through a chain of purchases and sale of the same goods, to an earlier sale of those goods by a person (the ‘defaulter’) who failed to account for VAT on that sale. The appellant also admits that that default by the defaulter in these 23 chains was fraudulent.

102. In other words, the appellant accepts that, in 23 out of 31 of its purchases at issue, its deals were connected to fraud. It does not accept that at the time it knew or ought to have known this.

103. With respect to the remaining 8 purchases, it does not accept that they were connected to fraud. HMRC’s case is that these 8 purchases were connected to fraud via contra-trading chains. In other words, while HMRC agree that there was no default in the chain of transactions relating to the goods purchased by the appellant, it is HMRC’s case that the output tax liability arising earlier in the appellant’s chain of transactions was offset by an input tax claim arising in another chain of transactions in which a fraudulent default did occur. In other words, the parallel chains were linked by a ‘contra trader’ who offset its output tax liability in the appellant’s chain by acting as a broker in another ‘dirty’ chain.

104. The appellant denies the connection in these 8 chains both as a matter of law and of fact. We have dealt with and rejected its submissions on the law (see §§54-87). It also denies that there was fraud in these alleged contra chains. And in any event says it did not know nor should it have known of the connection, if it existed. We make our findings of fact below.

The defaulters in the “dirty” chains

105. The Tribunal need say little about the copious evidence – which we accept - about the fraudulent default by the acquirers in the 23 “dirty” chains. This is because, as we have said, the appellant has accepted that the VAT was defaulted upon and that that default was fraudulent. We note that the defaulters were C&E Enterprises UK Ltd, CHP Distribution Ltd, Midwest Communications Ltd and Computec Solutions Ltd.

106. EES actually undertook 32 deals in this three month period. As mentioned, input tax was not denied on one of these deals: that was deal no 26 and was one of the ones that took place at the end of April. The difference with this transaction was that EES was in the position of a buffer, and the broker was Mr Wright’s brother’s company. We find HMRC has also proved to us that this transaction chain was connected to a fraudulent tax loss. It also traced back to Computec Solutions Ltd, which the evidence shows, and the appellant accepts, defaulted fraudulently on its VAT liabilities.

Connection in the alleged contra-trading chains as a matter of fact

107. Contra trading is alleged in 8 of the appellant’s 31 deal chains. These eight comprised two of the appellant’s deals in late April 2006 and all of its deals in May 2006. In three of the chains the UK acquirer (ie the company which imported the goods) was Digital Satellite 2000 Ltd trading as Powerstrip (“Powerstrip”), in four of the chains the UK acquirer was Svenson Commodities (“Svenson”) and in one was David Jacobs UK Ltd (“David Jacobs”).

108. In these eight chains, EES made the purchase direct from Svenson (four of the chains) but where the acquirer was Powerstrip or David Jacobs EES made the purchase from a 'buffer'. The alleged buffer was Svenson in one case and Team Mobile Inc in the remaining chains. Therefore, so far as EES was concerned, 5 of these 8 purchases were from Svenson and three were from Team Mobile.

109. The appellant does not accept as a matter of fact that any of these eight purchases by it were connected with fraud. In particular it denies that there was any connection between its chain of purchases and any other chains of purchases, and its position is that even if there was such a connection, it denies that there was fraud in any other connected chain.

110. It puts HMRC to proof of a connection to fraud. Its position is that HMRC is unable to prove this because, says Mr Bridge, the evidence before the tribunal was insufficient to prove it.

Evidence of 'contra' chains

111. We have already described contra trading. Simply put it relies on the contra trader avoiding accounting for output tax on its cross border acquisition by making a second UK to UK purchase the input tax on which offsets the output tax on the acquisition. In reality the position is more complicated: there is no need for the contra trader to match deals on a deal by deal basis. All it needs to do is carry out its intended acquisitions initiating the 'clean' chains that will sell the goods to the various brokers and, just before its VAT accounting period ends, carry out one or more contra transactions which are approximately equivalent in value to the total of its 'clean' chain acquisitions. It does not need to match individual transactions in order to achieve the necessary off-setting of input tax and output tax.

112. Mr Bridge's position is that therefore the Tribunal could only come to the conclusion that such a trader acted as a contra trader by looking at its position for the entire accounting period in which the 'clean' acquisition took place. And further, by tracing every 'contra' chain back to its origin to find out whether there was in fact ultimately a fraudulent default. And as this case is one where second line contra trading is alleged, the same process would need to be adopted in relation to all the acquirers in the four alleged contra traders' contra chains.

113. And it is Mr Bridge's case that HMRC's documentary evidence on the contra chains is only partial and inadequate to prove what they have to prove.

Invoices and deal sheets

114. We find that HMRC's procedures were that individual officers charged with responsibility for any particular trader whose trades were suspected of forming part of an MTIC chain would make entries on HMRC's electronic data storage system (the electronic folder) showing that trader's purchases and sales in some detail (date, amount, item, invoice number and so on). The same or another officer would use this information from a number of different officers to build up a 'deal sheet' which would show the trading route of a consignment of goods, from one trader to another. The deal packs might well trace a chain of goods back to an acquirer, and where information from customs’ authorities in Europe was available, the chains might go back beyond the UK acquisition. Similarly the chains would show the despatch by the broker to an EU customer, and where information from European customs authorities was available, the deal sheet might show subsequent sales too.

115. The deal sheets were therefore not primary evidence. They are nevertheless evidence of fact as they were simply a representation of the primary evidence. Mr Bridge's case is that the Tribunal should not consider the deal sheets to be reliable evidence unless supported by the underlying invoices. They are hearsay evidence, as different officers, not called to give evidence, would have looked at the various invoices and purchase orders on which the deal sheets are based.

116. Moreover, to the extent that the deal sheets showed the chain of goods before the UK acquisition or after the despatch by the broker, they were compiled with evidence obtained from officers in other EU countries after HMRC made mutual assistance requests. Those foreign officers have not been called to give evidence nor have HMRC in all cases been given copies of the invoices on which these entries on the deal sheets have been based.

117. We consider to what extent the evidence is reliable and should be relied on by this Tribunal.

Svenson

118. The evidence shows that Svenson had 112 purchases from a UK supplier in the relevant three month accounting period to end 06/06. Fifteen of these traced to a VAT default. This was evidenced only by deals sheets as the supporting invoices were omitted from the trial bundles by agreement between the parties. In these circumstances we accept as reliable the evidence of the deal sheets: the appellant has had disclosed the invoices and has chosen not to challenge the accuracy of the deal sheets although it could have done so by insisting that the invoices were part of the bundle.

119. Of these 15 dirty chains, three trace back to Midwest. The appellant accepted that Midwest was a fraudulent defaulter on the straight chains, and it follows, we find and HMRC's unchallenged evidence is, that Midwest defaulted on all its liabilities. In any event we accept the unchallenged evidence that Midwest was a fraudulent defaulter in respect of the 'straight' chains in which it was the acquirer.

120. The other Svenson chains trace back to Stockmart, West 1, Prompt Info and Restar. While not the allocated officer for Stockmart, Mr Williams gave unchallenged evidence in this Tribunal that Stockmart was a defaulter. We accept that evidence.

121. Mr F Lam gave unchallenged evidence about West 1 Facilities Management Ltd, which we find shows that the company was knowingly involved in massive fraudulent MTIC trading and the company has left unpaid and unchallenged an assessment for £110 million. The director has been disqualified.

122. For the other alleged defaulters there is a lack of direct evidence. We return to this point in §§134-141.

123. HMRC's case is that the remaining Svenson purchases all trace back to other alleged contra traders, whose own UK purchases trace back to a fraudulent default by an acquirer.

124. It was Mr Williams’ evidence that these purchases of Svenson's traced back to acquisitions by Export Tech and Rioni, and that Export Tech. It was also his evidence that Rioni's UK purchases which offset these acquisitions themselves traced back to two defaulting companies, Homes Sales and Lettings Ltd ('HSL') and East Midland Engineering Service Ltd ('EMES').

125. Exhibited were the deal sheets for Export Tech, but not the underlying invoices. That Export Tech's purchases originated with HSL and EMES was proved by freight forwarding documentation.

126. So far as Rioni was concerned the evidence in front of the Tribunal included the deal sheets but only the invoices for half of these deal sheets. Rioni's own contra chains were evidenced by deal sheets and invoices.

127. For Svenson, therefore, the documentary evidence of the chains of transactions and defaults in the 'clean' and 'dirty' chains is relatively extensive but not complete. It is less extensive for the other two alleged contras, David Jacobs and Powerstrip. We return to this point in §§134-141.

David Jacobs and Powerstrip

128. It was Mr Davies' evidence that in the 3 month period to 06/06 David Jacobs made 92 acquisitions and entered into 97 purchases within the UK. deals. The bundles included the deal sheets in so far as David Jacobs' deals traced back to a default, but not where they were alleged to trace to a default via a second line contra trader. The defaulters in David Jacobs' dirty chains were alleged to include Midwest and West 1 and the Tribunal had sample deal sheets with invoices tracing to both of them.

129. The allegation was that Powerstrip had in the relevant period made 77 purchases from UK business, which traced back to a defaulting acquirer and 72 which traced back to an acquirer whose own UK purchases could be traced back to a default. Where the second line contra was alleged to be Export Tech some deals sheets and invoices were in evidence.

130. Although HMRC officers gave written and oral evidence about these two alleged contra traders, in summary most of the documentary evidence of this was lacking.

Conclusions on the partial documentary evidence

131. Mr Bridge's case is that the Tribunal cannot be satisfied that there was a connection to fraud unless all the invoices for all the chains which made up each alleged contra trader and second line contra trader and defaulters trading in the entire relevant periods, because how else can it be demonstrated that overall no VAT was accounted for to HMRC? And unless there is tax loss, there is no fraudulent tax loss. And even if some tax loss is demonstrated, that is insufficient, says the appellant, as it can't be said that a partial tax loss is connected to EES's purchases rather than the purchases of some other broker.

132. In so far as it goes we agree with Mr Bridge. Where we part company with Mr Bridge is over the question whether HMRC had satisfied this Tribunal that none of the alleged contra traders, second line contra traders and defaulters had accounted for any VAT, and in particular whether there was fraudulent tax loss connected to EES's 8 purchases in which this was an issue.

133. The appellant does not challenge the veracity of the officers' evidence. We accept it was truthful. But this does not prove the matter alone. The evidence was truthful but was it accurate? Ordinarily the Tribunal would be reluctant to accept evidence based on documents without sight of the documents as this denies the other party, and the Tribunal, the opportunity to check the evidence. In so far as the evidence was disclosed but not copied to the bundles, we accept it as the appellant had the opportunity to challenge it and chose not to.

134. In so far as the evidence and in particular the deals sheets and/or the invoices underlying them were not even disclosed the position is different. We accept that HMRC considered that they had a good reason for not disclosing all the invoices and even some of the deal sheets and that reason was proportionality. We accept that it amounts to an extremely large amount of paperwork, and that copying it would have been extremely expensive and storing it difficult.

135. Nevertheless, we consider that HMRC have to live by their unilateral decision not to disclose it. A proper course of conduct would have been to ask the Tribunal for directions if HMRC considered disclosure of the invoices behind the deal sheets was not proportional. And certainly the use of electronic media may have been a solution.

136. Nevertheless, despite this, our view is that the officers' evidence should not be entirely discounted. It was supported in part with most of the deal sheets, some of the invoices, and in one case freight forwarder information. While the deal sheets (including those parts based on evidence from overseas tax authorities) are effectively like hearsay evidence, there was no suggestion that those disclosed were not accurate. But the officers' evidence should be given less weight in the absence of the complete documentary evidence, including invoices, on which it was based.

137. What do we mean by giving the evidence less weight? It does not mean that we entirely discount it: we mean that we will not accept it without corroboration.

138. So is the officers' evidence corroborated? To the extent the documents are exhibited, the evidence we find is entirely corroborated. Other than that, there is no other direct evidence of the contra chains, second line contra chains, and ultimate defaults. However, there is, as discussed (below in §§161-200) overwhelming evidence that the 'clean' chains in which EES' 8 purchases and sales featured, as well as the 23 ‘dirty’ chains, were orchestrated, and that in all 32 chains the goods and funds moved in a circle, returning to their originator. There was no missing trader in these 8 ‘clean’ chains.

139. We find that therefore we are wholly satisfied that the only purpose of this orchestration was MTIC fraud. We can think of no other explanation for what this evidence shows and certainly none credible was suggested to us. Logic dictates that if EES's 8 chains were orchestrated for the purpose of MTIC fraud and there was no missing trader in those chains, then there must have been a contra trading chain, possibly even second line contra trading chain. It shows the Tribunal that the acquirer must have been a contra trader. The FCIB and other evidence of orchestration of EES's chains does not tell the Tribunal who the defaulters were nor the identities of other parties in the contra chains, nor does it tell the Tribunal if there was second line contra trading, but it does show beyond any reasonable doubt that there would have been a fraudulent default to which EES's purchases were connected.

140. We consider this corroborates the various officers' evidence overall even if not the details. We therefore accept the evidence of the officers including the details of it despite the lack of disclosure of the complete documents on which it was based. Nothing turns on this in that HMRC need only prove a connection to fraud: it is not necessary for HMRC to be able to identify the individual defaulters. Nevertheless, we find that they have done so.

141. In conclusion, HMRC have proved the connection to fraud of EES's purchases and sales in the 8 chains in which the appellant did not accept that connection was proved.

The evidence relating to alleged 'cell 5'

142. Mr N Humphries, HMRC officer, gave evidence that he had looked at vast numbers of deal sheets on the electronic folders. He did not look at the invoices underlying these deal sheets.

143. His evidence was that these showed a large number of closed cells. In other words, certain companies would occupy a designated position within a 'clean' MTIC chain. The company would be a broker, or buffer, or contra or an EU dispatcher or EU acquirer. Goods from a particular acquirer would only be traded by the companies within this closed cell.

144. A broker or buffer company might feature in more than one closed cell, but it was Mr Humphries’ evidence that nevertheless the goods originating with the EU dispatcher would pass through companies within the closed cell and end up with the same EU company.

145. It was his evidence that all of EES's 8 contra trades were within what he chose to designate as 'cell 5'. In this cell Comica and Parasail were the EU traders who were at the top and/or bottom of every deal chain. The cell comprised about 60 companies.

Do we accept the evidence about cell 5?

146. Mr Bridge challenged Mr Humphries evidence on the grounds that Mr Humphries himself had not claimed to see the original invoices on which the deal sheets were based and in any event had not even exhibited the deal sheets to his witness statement.

147. We accept that the deal sheets alone were extremely voluminous, and that combined with the invoices would have been enough to fill a room. Nevertheless, as we said before, the proper way to have dealt with this problem would have been to approach the tribunal for directions long before the hearing.

148. While the honesty of the witness was not challenged, we do consider that we should be cautious about accepting his evidence as it was not supported with the underlying documents. Another witness, an Officer Murphy, corroborated that he had too seen the Cell 5 deal sheets. Nevertheless, the appellant and Tribunal have been deprived of the opportunity to check them. As with the contra trading evidence, we consider whether other, unconnected, evidence corroborates it.

149. We find that the FCIB evidence (§§161-168) gives overwhelming evidence that all of EES's chains at issue in this appeal were orchestrated for the purpose of fraud. This is general corroboration of Mr Humphries's evidence which also shows a highly organised fraud. Moreover, the FCIB evidence gives specific corroboration of one aspect of the Mr Humphries' evidence, and that is that the chains started and ended with the same companies, Comica and Parasail (see §163).

150. For these reasons, we accept Mr Humphries' evidence on 'Cell 5', including the details of it which are not corroborated.

151. To a large extent Mr Humphries' evidence does not really add anything to the wealth of evidence before this tribunal that all 31 of the deal chains at issue in this appeal were orchestrated by a fraudster for the purpose of MTIC fraud. But it does add some relevant details.

152. It shows that the brokers within Cell 5 included Cellular Solutions and Fonecode, as well as EES.

153. It shows that Svenson, David Jacobs and Powerstrip were part of Cell 1 as well as of Cell 5. Thus when these companies made a purchase from a Cell 1 company they sold to a Cell 1 company; when they purchased from a Cell 5 company they sold to a Cell 5 company.

154. Mr Humphries was also able to demonstrate that the contras traders' overall combined input tax and output tax very nearly netted off. We accept this evidence and have referred to it at §53.

155. Mr Shorrock gave evidence about loans to traders (not including EES) within Cell 5. The supporting evidence was not served and we were not given a satisfactory reason for this. It seems that it was originally served without exhibits to support an interim application and HMRC still relied on it in the substantive hearing but had never served any supporting exhibits. There was no corroboration of the loans from other evidence, and while not impugning the honesty of the officer, we put no weight on the evidence. Similarly large parts of Mr Murphy's statement are uncorroborated and unsupported by documentary evidence and other than to the very limited extent mentioned above at §148 , we do not place weight on it or refer to it.

What sort of fraud – the appellant's alternative explanation.

156. While the appellant has accepted 23 of the deals at issue in this appeal were connected to fraudulent tax loss, we have taken the appellant's position to be that it does not accept that that fraud was organised MTIC fraud as described in paragraphs §§6-26 above.

157. This is because the appellant’s case is that (with a reference to §59 of Mobilx) the reasonable explanation for why it entered into the 31 transactions the subject of this appeal is that, because HMRC failed to prevent fraudulent trading by catching the defaulters, the market was awash with mobile phones on which a default had earlier occurred, and the appellant could not help but buy phones connected to fraud in order to fulfil its customers' genuine trading requirements.

158. In other words, it is implicit in the appellant's case that the fraud which occurred in the 23 chains in which it has admitted that there was a default was acquisition fraud. It is implicit in its case that there was a genuine market for the phones but earlier in the supply chain an acquirer had run off without accounting for the VAT. Elsewhere it says it accepts that in the 23 'dirty' chains there are indications of organisation but we find the appellant never actually accepts that the fraud is MTIC fraud, so we consider the evidence and come to our conclusion on this below.

159. It is important to do this because the nature of the fraud is relevant. HMRC's case is that the appellant knew or ought to have known its transactions were connected with fraud even though it never purchased directly from a defaulter. They can only make out their case on this, if at all, by relying on circumstantial evidence. Therefore, the Tribunal needs to understand the circumstances of the transactions: were EES' purchases and sales on the open market or were they contrived for the purpose of fraud? If the later case is the factual position, then it raises the question whether the appellant knew this (or should have known it), whereas in the former case there would be nothing to know.

160. In the three month period at issue in this appeal EES entered into 32 transactions. Its input tax reclaim was denied in 31 of these transactions. Nevertheless, it is HMRC’s position that all 32 deals, including the 8 alleged contra deals where the appellant has not accepted connection to fraud, were an integral part of MTIC fraud. As we have said, the difference with the one exception (deal 26) is that EES occupied the position of buffer rather than broker. We look at all 32 deals and we find as follows.

FCIB evidence

161. Mr Birchfield and Mr Young gave evidence about money movements within the FCIB bank. The documentary evidence in front of the Tribunal comprised printouts from both the Paris and Dutch servers used by FCIB.

162. We find the evidence is complex consisting as it does of many pages of extracts from many traders’ bank accounts. When distilled we find it shows that in at least 30 of the 31 deals at issue the money started and ended with one of three companies. And we are satisfied that with the one remaining chain the similarities were such that it also was circular although the complete links are missing from the evidence.

163. In more detail, in the chains where Globalfone was EES' customer, the chains started and ended with either Comica or Parasail. (Parasail did not have an FCIB account but we are satisfied on the evidence that it is proper to attribute payments to a company called Negresco to Parasail. In any event, we accept the evidence that Comica, Parasail and Negresco were all owned by a certain Mr Adil Kamran).

164. When EES's customer was David Jacobs or J Corp the money started and ended with Marxman International. Both David Jacobs and J Corp were owned by Mr David Jensen. Marxman was owned by Mr Sabir Mansoor Hussain.

165. All the money for the transaction chains shown in the deal chains moved between accounts in the FCIB and did not leave the bank. Although four of the traders (out of 61) did not have FCIB accounts, payments to those traders were made to other companies which did hold FCIB accounts.

166. We find that the money originated from four BVI companies – Amira, Solutions Beyond, SM Systems and Mighty Mobiles. The first three of these companies share the same BVI address and Amira and SM Systems both have same account signatory and email address (Imran Memon). The first three are shown to be connected and we find it more likely than not the fourth company was connected with the other three.

167. We have briefly summarised this evidence but that is not to underestimate its significance. We find it conclusively proves that organised MTIC fraud was the rationale for the 31 chains of transactions at issue in this appeal. The money flows show that, for all 31 deals, one of three companies, within a space of a few days, has sold and then re-acquired the identical goods. It would be unusual if the same quantity of goods, being bought and sold through 6 or so traders would end up (in the space of only a few days) back with the original vendor: that this should happen in all 31 of EES’s deals at issue can have no other explanation than that EES’ deals were orchestrated for the purpose of MTIC fraud. There was no genuine market for these goods. So the fraudsters engineered the sales and purchases for their fraudulent purposes.

168. There is a great deal of other evidence before the Tribunal about these chains, which we move on to consider below. Our conclusion is that it, too, demonstrates that all 32 of EES's transactions in this period were orchestrated for the purpose of fraud. None of the evidence is consistent with it being acquisition fraud rather than MTIC fraud.

The chains

169. length of chains: putting aside the 8 alleged contra trading deals, all the chains of transactions taking place in the UK are long. They comprise a defaulter followed by three or four buffers, and then the broker.

170. We find that there is no rational explanation for this, other than MTIC fraud. We accept Mr Fletcher's evidence (see §231 below), consistent as it is with common sense, that markets tend to act rationally. Buyers would seek the cheapest source of identical goods and that would tend to cut out the middlemen. This is especially the case where the middlemen add no value as it appears these buffers added no value as they were simply back to back transactions in identical amounts. And further, traders advertised on trading boards, which facilitated buyers seeking out the cheapest source.

171. However, in the world of MTIC fraud, four middlemen ('buffers') makes sense as their purpose is to distance the broker from the defaulter (see §§19-21).

172. length of alleged contra-trading chains: In contrast, the length (in the UK) of the alleged contra-trading chains is much shorter, involving only the acquirer (the alleged contra trader) and the broker (EES), and a single buffer in half of the eight alleged contra deals, and no buffer in the other half (see §§107-108).

173. If none of the deals were orchestrated, the question would be why there were invariably long chains where a default occurred in the chain and short chains where a default only occurred in a connected chain? It would stretch coincidence.

174. However, in MTIC fraud it makes perfect sense: the long chains are to protect the broker by interposing buffers between defaulter and broker, but such protection was perceived to be unnecessary when the protection was given by moving the default into a parallel but separate chain. Long chains with a default and short chains involving a contra trading makes sense in MTIC 'trading' and therefore indicates to us that all the chains were MTIC trading.

175. It is of course the appellant's case that there was no connection to fraud in the 8 alleged contra chains. This would, no doubt, on the appellant's case, explain the shortness of the chains. We do not accept this. There are simply too many similarities between the 24 straight chains and 8 alleged contra chains. Of the 5 UK traders involved in the alleged contra chains, only 1 (Powerstrip) does not also feature in the “dirty” chains. Moreover the margin made by the traders who appear in both types of chain is the same, suggesting it is all part of the same organised fraud. For instance, Team Mobile makes 50p per phone in both types of chain, and David Jacobs and Svenson make £1 per phone in the majority of both types of chain. Further, we have found (see §§201-212) that David Jacobs and Svenson were knowingly involved in the fraud so it seems most improbable that the 8 contra chains were not a part of the same fraud with the 24 “dirty” chains. Lastly, and most importantly, the FCIB evidence of circularity of funds (§167) makes it quite clear that these 8 deals were all part of the same fraud.

176. patterns in the straight chains (24 chains): the chains show some patterns and repeats. There are some 8 companies which occupy the position of first buffer. They never occupy any other position in the chain. There are some 12 companies occupying the position of second buffer – sometimes they also occupy the position of third line buffer but only where there is a fourth line buffer. In other words, they are always at least one remove from EES and one remove from the defaulter or contra trader. The buffers which dealt directly with EES were more limited in number. There were only four of them: Svenson (12 deals), David Jacobs (6 deals including the EES's buffer deal), Team Mobile (2 deals), and Megantic (4 deals).

177. There were other patterns. In some cases, the line of traders was repeated with only one alteration. For instance, in deal 3 on 17 March, Anderson Cellular & Data (a line 1 buffer) sold to Xcel Solutions, who sold to David Jacobs, who sold to EES; on 24 April Anderson Cellular & Data sold to Xcel again, but Xcel sold to Megantic who sold to EES. Another example is a chain involving V2 UK Ltd, who sold to Platinum Mobiles, who sold to Xcel Solutions, who sold to Svenson, who sold to EES. This chain occurred twice, once on 17 March and once on 23 March. Not only that but the chain was repeated in another deal on 17 March, with Watts Management occupying the position of second line buffer (ie the position of Platinum Mobile) but all the other traders being the same.

These patterns make no sense in a real world where trading is commodity trading on an open market, particularly one where all traders would be known to each other because of trading boards.

178. no commercial rationale in chains: All the phones originated outside the UK and were acquired in the UK by the defaulting trader or contra trader. All the phones were then despatched by EES out of the UK to buyers who wanted the goods delivered to Belgium and Spain. Therefore, the market for the phones – if it existed – was on the Continent. So why were the goods acquired into the UK? Why didn’t the original suppliers of the phones identify the Continental market for the phones, as the Continental buyers paid considerably more for the phones than the original UK purchasers?

179. Bringing the phones into the UK only makes sense if it was organised MTIC fraud as the market is artificial and wouldn’t behave rationally. It makes no sense otherwise.

180. Mr Wright attempted to give an explanation. His explanation was that traders on the Continental market were exploiting subsidies. His experience in the UK market was that phones would be made available at a subsidy if authorised distributors ('ADs') would pair them with a SIM card. This was because the MNOs (multiple network operators) want lucrative contracts with phone users. Nevertheless, the ADs would exploit their relationship with the MNO by selling these phones to grey market traders, who would split the SIM card from the phone, resell the card in the UK (so that the network operators ('MNOs') would not know their subsidy had been wasted) and resell the phones at a profit abroad.

181. But, as Mr Fletcher explained, and is consistent with common sense, while subsidies might well be offered in European Countries, grey market traders would not sell the phones into the UK as it had heavily subsidised phones. The phones would be sold to those parts of the worlds for which the greatest price could be achieved, and that would be where there was no or only low subsidies.

182. It also makes no sense because it was crucial to the ADs, who wished to retain their preferred trading status with MNOs, that the phones did not come back onto their national market as, if SIM paired again, the MNO (keeping a record of IMEI numbers) would realise what had happened and might put an end to the AD’s preferred trading status. Mr Wright's evidence was that it was an understanding he had with the ADs within the UK from whom he had bought such subsidised phones that he would sell the phones in another part of the world. Yet, if this is what was happening with European ADs in these deals, there was a high risk of the phones returning to the country they originated from as they were being re-exported from the UK to the Continent.

183. Yet one more obvious failure with this explanation is that, even putting aside the risk of the phone returning to the country from which it originated, it was clear that the phones were both originating on and returning to Continental Europe. Someone seeking to exploit the opportunity identified by Mr Wright (if it existed), would act to cut out unnecessary journeys and buy the goods before they left the Continent. They would also cut out unnecessary middleman, so putting aside the many other objections to Mr Wright's suggestion, it fails to explain the delivery of the goods to the UK or the buffer trades.

184. Mr Wright put in evidence a few invoices which showed his companies had bought phones abroad which, he said, were box broken as the invoices showed that they had been SIM unlocked. However, only one of the invoices showed a delivery address and that was to Hong Kong, which was consistent with what Mr Fletcher said. In any event, while it might be the case that a small number of phones box broken elsewhere might make their way to the UK, we accept Mr Fletcher’s evidence that it would be counter-intuitive for large numbers to do so, and Mr Wright has not demonstrated otherwise.

185. So there is no rationale explanation for these trades other than that they were artificial trades organised for the purpose of MTIC fraud. They were not driven by any market forces.

186. Back to back deals: All the deals were back to back, with exactly the same quantity of phones being sold down the line. Yet despite traders apparently all wanting the same quantity of the same phones on the same day, nevertheless the chains were long so it appeared that no one was successfully identifying the root of the demand or supply and thus able to cut out all the middlemen despite the middlemen adding no value. Indeed the goods sat in the same warehouse throughout the UK transactions. This makes no sense in a rational free market.

Failure to deal directly with existing contacts:

187. Another inconsistency in the chains with what we would expect to find in a rationale market is that some of the participants failed to take obvious steps to cut out middlemen. We find David Jacobs had a pre-existing relationship with EES. They had previously sold goods direct to EES. Yet twice in these 32 deals they sold to an intermediary who then sold to EES. Yet EES gave a higher price than the intermediary. Rationale behaviour would have been for David Jacobs to offer its goods to all its contacts, including EES and cut out the middleman. That it did not do so suggests rational market behaviour was not the driving force behind its choices.

The margins:

188. There was a great deal of uniformity in the margins on the 24 straight deals. The first line buffers made either 20p or 25p per phone. Any particular first line buffer always made the same margin eg First Associates always made 20p per phone while Wireless Warehouse always made 25p per phone. The second line buffers nearly all made 25p per phone (there were only 4 exceptions to this out of 24 deals). Third line buffers and fourth line buffers tended to make a larger margin of 50p or £1 per phone, although there were a few exceptions.

189. Similarly, there was much uniformity of margins in the 8 alleged contra deals. The margin was £1 for the acquirers in all but two of the deals. Team Mobile made 50p on the three occasions it was buffer and Svenson £1 on the other occasion that there was a buffer.

190. There is no explanation of why there should be such uniformity of margin in an open rational market. In an open market, we would expect margins to be very variable and sometimes to be negative. Yet if it was MTIC fraud this uniformity makes sense. A small fairly consistent reward is paid to those put into the chain to add distance, but the reward is small as they are not perceived as taking any risk and the fraudster does not wish to reduce his own profits by any more than necessary. They do small job and get a small reward.

191. There was a clear pattern that in all 32 deals the broker made a profit many multiples higher than any other trader in the chain. Whilst the profit per phone made by any buffer varied between 15p per phone to £1.70 per phone (and was normally 25p to £1 per phone), EES’s profit varied from £10 to £34.50 per phone. Yet there is no commercial explanation for why the broker would consistently make such a significantly higher profit than other UK traders. If the price of the phone really was higher on the Continent than in the UK, the buffers (if acting rationally) would be exporting the phones to the Continent to make the large profit EES was making, rather than selling them within the UK for a very small profit in real terms.

192. But if this was MTIC fraud, it is quite rational. The rationale for the buffers is to provide distance between the defaulter and the broker. It makes sense for them to have very low profits as they add nothing to the chain except length. They take no risk (as their output tax is very slightly higher than their input tax). But the broker is taking a risk. As explained in §11, he will make a loss unless HMRC makes the VAT refund. He is at risk. In the world of MTIC fraud, the contrasting level of profit between buffers and brokers makes sense.

Dates of deals

193. All the deals in March took place on two days (17 and 23 March). All the deals in April – including the one in which EES acted as buffer and the two alleged contra deals) took place on two days too (25 and 28 April). All six deals in May took place on two days (24 and 31 May).

194. It would be difficult to explain in a free commercial market why mobile phones were traded on only a few days each months, and those dates being late on in the month. But if it is artificial trading for the purpose of MTIC fraud, then it is quite rational. We presume it is easier to organise a number of deals to take place on the same day and we find deals later in the month make sense because brokers are on monthly returns and the later in the month the deals take place the less time the brokers have to wait before they can claim the VAT refund. The timing of the deals are driven by VAT and not market forces.

The same bank

195. Of the 61 companies which featured in the deal chains in this appeal, 57 banked with FCIB. HMRC say that this would be an extraordinary coincidence only explicable by the chains being artificially orchestrated for the purpose of fraud: the fraudster must be supposed to find it easier to orchestrate the fraud and keep control of the funds if the funds only move between accounts held in the same bank.

196. The appellant offers a different explanation. Its explanation is that UK banks, warned by HMRC of the need to file Suspicious Activity Reports for any customer trading in large quantities of mobile phones, were closing the accounts of all mobile phone traders forcing the traders to look to foreign banks. FCIB was an obvious choice, says the appellant, as it was known to mobile phone traders.

197. However, the appellant’s explanation does not explain why the traders in the chain who were not UK-based all chose to bank with FCIB. Therefore, we agree with HMRC that the fact that all traders, UK based or not UK based either had an FCIB account, or used a proxy which had an FCIB account, to carry out these trades is no coincidence. It is yet one more indication that these chains of transactions were organised for the purpose of fraud.

Conclusion

198. Even without the evidence of circular movements of money, on the above evidence about the deals themselves, we would be satisfied that EES' purchases and sales at issue in this appeal were organised by an unidentified fraudster for the purpose of committing MTIC fraud and did not take place on the open market. Combined with the FCIB evidence of circular money movements, overall the evidence for this conclusion is overwhelming.

199. This was not acquisition fraud. There was no market awash with phones on which there had been a default. There was no market at all. There was no genuine trading in mobile phones. There was no genuine demand from the appellant's customers. Therefore, there is no alternative explanation. This was organised MTIC fraud and the appellant's purchases and sales were an integral part of that fraud.

200. The question is whether the appellant knew or ought to have known that its transactions were connected with that fraud. But, before looking at the evidence on this, we consider whether any other of the traders were knowingly involved in the fraud.

Who were knowingly involved in the fraud?

201. Svenson: We accept Mr Williams' evidence on Svenson for the reasons set out above (§§131-140). His evidence shows that Svenson had a phenomenal growth in output without any corresponding increase in staff or even significant increases in pay to staff. Its small turnover of £15K in 03/03 jumped to £7milion per quarter a year later; it then jumped to over £20 million a quarter by 03/05, was up to £81m per quarter by 12/05, then £215 million next quarter and then £404million in its next and last quarter. But up to and including 09/04 its returns showed that only a very small amount of VAT was owing to HMRC of around £2000 per quarter so its astounding increase in turnover had had no impact on VAT due. From 12/04 its returns show a steady increase in VAT reclaims up to to £8 million by 30/06 although this was still very small in comparison to its turnover.

202. The owner of the company was Mr Abdul Koser. He was not a witness in this case. From the relevant officer's evidence, which we accept, we find his explanations given to HMRC were untrue or incredible. He said the company traded mostly within the UK. It did not. He said he took everything on trust and was given trust by his trading partners: he said he had no need for capital as his sellers extended credit, he said he never inspected goods, he said he was always able to find a buyer for exactly the goods he was purchasing.

203. We also find that on 10 May, Svenson sold phones to Globalfone. Yet despite this previous relationship with Globalfone, Svenson supplied EES in 4 of the contra deals later in May. EES sold the phones to Globalfone for considerably more than Svenson sold them to EES. The question arises why Svenson didn’t offer the 10 May phones direct to Globalfone, with whom it had previously traded, and who clearly wanted the phones.

204. This makes no sense in a commercial world. Moreover, it suggests that Svenson was knowingly involved in the fraud. A genuine trader, who did not know that it was participating in MTIC fraud, would seek to maximise its profit. Svenson did not, suggesting it knew it was MTIC fraud and each ‘trader’ should play the part assigned to it.

205. Svenson was denied input tax in respect of trading in 06/06. It appealed and then withdrew its appeal: it appears that therefore even Svenson accepted that it knew or ought to have known of a connection to fraud.

206. We find Svenson organised its trading so that its outputs net off against its inputs. We simply cannot see how the fraudster could manoeuvre Svenson into doing this without the full knowledge and cooperation of the trader. This factor, combined with Svenson's failure to act in a commercial fashion as outlined above, its failure to challenge its assessment, its astonishing increase in turnover from nothing, and lastly that it kept its trading within the discrete cells (see §153) leads us to conclude that Svenson must have been knowingly involved in the MTIC fraud at issue in this appeal.

207. David Jacobs: We accept Mr Davies evidence on David Jacobs (see §§131-140). We find that David Jacobs was informed that some of its 2005 transactions traced back to fraud in 2006 but it carried on trading. It clearly knew about MTIC fraud as HMRC officers often spoke to director many times. It was happy to ship its goods abroad before it was paid for them and sometimes before it had even received a purchase order. It never undertook due diligence on its trading partners. It did not hold adequate insurance.

208. Even more significantly, it did not hold the CMRs for the goods it shipped. Vehicles supposedly transporting its mobile phones – relating to deals not at issue in this appeal - were searched on three occasions and each time were found to be empty or carrying other goods (animal feed and steel coil). We find it more likely than not its failure to hold CMRs reflects the fact that the goods it traded in as broker (not directly at issue in this appeal) did not exist.

209. David Jacob's input and output tax largely netted off on its VAT return. In 09/06 its liability on £37 million sales almost exactly netted off its VAT entitlement arising on virtually same amount of purchases leaving approximately a mere £34,000 owing. In the previous period, the netting off was less exact (£392 million of sales compared to £380 purchases) leaving a £4 million reclaim.

210. David Jacobs also failed to exploit existing commercial relationships. On one day it sold phones to EES who then sold to Globalfone, while on the same day it also made a direct sale to Globalfone. Then in deal 31 David Jacobs sold to Team Mobile, who then sold to EES who then sold to Globalfone. Yet at that point in time David Jacobs had previously sold direct to both Globalfone and EES in various earlier deals.

211. David Jacobs was denied input tax in respect of trading in 06/06. It appealed and then withdrew its appeal: it appears that it accepted that it knew or ought to have known of a connection to fraud.

212. We find David Jacobs organised its trading so that its outputs net off against its inputs. We simply cannot see how the fraudster could manoeuvre David Jacobs into doing this without the full knowledge and cooperation of the trader. This factor, combined with David Jacob's failure to act in a commercial fashion as outlined above, its failure to challenge its assessment, its ability to generate a £308 million turnover in 2006 from a standing start, and its trade in non-existent goods, and lastly that it kept its trading within the discrete cells (see §153) leads us to conclude that David Jacobs must have been knowingly involved in the MTIC fraud at issue in this appeal.

213. Powerstrip: We also accept Mr Ruler’s detailed evidence about Powerstrip. Like the other contra traders, it had a phenomenal turnover (over £600 million) in 2006 which was vastly increased over its previous turnover of barely the VAT registration threshold. It made only a small margin, other than when it acted as broker. All its deals were back to back and never went wrong. Trading conditions were very easy. It never needed finance as its suppliers were content to wait until Powerstrip was paid by its customers, even though Powerstrip had a nil credit rating and offered no security. It netted off most of its input tax against output tax. For instance, netting off left it with an £2 million input tax claim from an actual £62million claim in 06/06. It also failed to challenged (or pay) its assessment. It kept its trading within the discrete cells (see §153). We find it was knowingly a party to the fraud.

214. Comica, Parasail, Negresco and Marxman: These three companies were between them at the start and end of the money flows and chain of goods in all of the deal chains at issue in this appeal. It is impossible for them to have occupied this position and not been a pawn of the fraudster or in conspiracy with the fraudster. We find they were knowingly involved in the fraud.

215. Export Tech:

216. Mr Berry's evidence, which we accept, was that Export Tech was registered for VAT in late 2005. It had no turnover until February the next year when it rockets up and by mid-2006 peaks at £200million. It is down to nothing by October 2006 at the time of HMRC's policy of extended verification. Its returns show that its output tax virtually always equalled its input tax. It had no obvious means of finance. As noted above (§4131-140) we find its contra chains traced to defaults Its profit margin was very small indeed. In summary the evidence shows it had an enormous turnover, with a tiny margin, it expended virtually no effort in achieving its profit margin, undertook no real due diligence on its trading partners, nor negotiations, it did not hold insurance, and had such favourable trading terms that it needed no finance. And as we have said, all its deals were connected to fraud and it facilitated that fraud by acting as a contra trader.

217. Could it have been a party to all this without knowing that its transactions were connected to fraud? On the balance of probabilities we find that therefore it must have known it was facilitating a fraud. It must have known its netting off did not arise by chance.

218. Rioni Lmited: The position with Rioni was much the same as with the other contra traders. It traded without the need for finance, as its suppliers were always trusting and willing to extent credit, it had a phenomenal turnover (eg £108 million in first three months of 2006). Yet it traded without carrying out due diligence or obtaining CMRs for its despatches. It netted off its inputs and outputs. It has never paid or challenged the assessments levied on it. We find it was knowingly involved in the fraud.

219. Globalfone: It traded directly with Comica and Negresco, who were at the start and end of the chains of goods and money in may of the deals and must have been directly controlled by the fraudster. Globalfone had to know that it must sell to Comica. It was clearly prepared to buy whatever stock EES offered it knowing Comica would buy it.

220. The appellant's proposition was that Globalfone was merely maneovered into the position it occupied in the chain by the freight forwarder. It was the appellant's contention that somehow the freight forwarder controlled the deals so that if EES or Globalfone looked to be about to sell the goods to a genuine buyer rather than the one intended by the fraudster, the freight forwarder would be able to step in and prevent the chain happening from the start by simply making the goods unavailable.

221. We reject this. The fraud was highly organised and it makes no sense for the organiser to leave vital parts of the fraud to chance: it would not leave EES' purchase and sale, or Globalfone’s purchase and sale to arise by chance. If it was left to chance the deal probably wouldn't happen as intended by the fraudster. Whereas the evidence is that a great many fraudulent deals involving millions in pounds took place in a closed cell of about 60 traders.

222. Further, as noted in §§203 & 210 above, Globalfone was content to buy from EES on a number of occasions when the identical phones were available at a cheaper price from another supplier with whom Globalfone had an pre-existing relationship. This un-commercial behaviour on its part is suggestive of willing participation in the fraud.

223. On the balance of probability we find it more likely than not that Globalfone was knowingly involved in the fraud.

Expert evidence on grey market trading

224. We allowed Mr Fletcher to give evidence despite the appellant's objections and the reasons for this decision are set out in an appendix (§§377-394).

225. The evidence having been admitted, it was the appellant's case that we should not rely on it. The appellant’s case amounted to saying that Mr Fletcher (if he was an expert at all which the appellant did not accept) was an expert in white and grey market mobile phone trading and not in trading which might arise where a market was flooded with phones on which the importer had failed to account for VAT. Mr Fletcher agreed he was not an expert on fraudulent trading and we agree with this too.

226. If the grey market in mobile phones was flooded with phones (probably below normal market price) on which no VAT had been paid, Mr Fletcher's ignorance of this would cause the Tribunal to doubt his expertise.

227. But the criticism is not made out. As we have said (§168), there is no evidence that any of the phones in the 31 deals were ever sold on a genuine market at all. If the appellant was right and this was merely the case of genuine traders finding the market flooded with phones on which an importer had defaulted on the VAT, it would fail to explain the circular movements of money (see §167). It would fail to explain the evidence of orchestration in the chains (see §§169-170). This is because if the appellant was right, the trading would be driven by market forces and not organised. On the contrary, the overwhelming evidence is that the chains of ‘transactions’ involved in this appeal were chains of artificial transactions organised solely for the purpose of MTIC fraud. This was not a genuine market flooded with phones: there was no genuine market at all.

228. It is not surprising Mr Fletcher was unaware of the sort of 'trading' at issue in this appeal impinging upon the genuine grey market. It was entirely separate from it. All the fraudulent transactions of which we have evidence were orchestrated for the purposes of fraud and were not driven by any market demand and did not take place on a genuine grey market. Genuine wholesale demand arising in the grey market was not satisfied by any supply of phones traded on the fraudulent market. The genuine grey market and the MTIC 'market' were entirely separate. We reject this attack on Mr Fletcher's reliability.

229. Mr Fletcher's reliability was also criticised on the basis he had revised his witness statement to reflect trading in phones to large retailers such as Tesco and John Lewis. We do not consider this in any way alters his reliability as a witness. There was never any suggestion that trading with such large retailers was the explanation of the trades in this case. He had not revised any his conclusions, but merely removed a simplification of his description of the genuine grey market.

230. We found Mr Fletcher to be a very careful and unbiased witness on the grey market. He weighed his answers carefully. For example, he accepted that a particular kind of trading which Mr Wright said he had undertaken in the past (buying subsidised business phones) was feasible in the grey market even though Mr Fletcher had no knowledge of it happening in practice.

231. His evidence was very detailed. We find it was internally consistent and described a market based on rational, self interested behaviour. We also find it was largely (if unintentionally) corroborated by Mr Wright's evidence of the sort of trading he had undertaken through his various companies outside the sorts of deals at issue in this appeal. For instance, Mr Wright's businesses had engaged in box breaking, which Mr Fletcher described in his witness statements.

232. Mr Fletcher's and Mr Wright's evidence diverged over whether the sort of trading at issue in the appeal took place on the grey market. Mr Fletcher's opinion was that it was inconsistent with rational grey market trading. Mr Wright did not agree.

233. We accept Mr Fletcher's evidence. We prefer it to Mr Wright's where they diverge. We accept that Mr Wright was very knowledgeable about mobile phone trading as (at the time of the events) he had had 13 years of experience and had clearly managed to adapt his methods of trading to suit the developing new market in mobile phones. He still traded in mobiles at the time of the hearing. Nevertheless, we did not consider that his evidence could carry the same weight as an independent expert because he lacked independence: he had a clear and substantial financial interest in the outcome of the appeal. Further, for reasons explained at §§180-185 we were unable to find his explanation for the trading at issue in this appeal rational or reliable. Lastly, the FCIB evidence and the evidence of the chains overwhelmingly show that the trading was not genuine grey market trading: this is consistent with Mr Fletcher's opinion and not with Mr Wright's. So for all these reasons we find Mr Fletcher's expert evidence to be reliable and Mr Wright's evidence on the grey market only to be reliable in so far as it was consistent with Mr Fletcher's. And we find Mr Wright's evidence tended to be consistent with Mr Fletcher's only when talking about trades that were not the type of trading undertaken in this appeal.

234. In summary Mr Fletcher's evidence was that there was only four types of trading on the grey market, box breaking, arbitrage, volume shortages and dumping.

235. Box-breaking is where the UK retail price of the phone is (due to subsidies) less than the wholesale price so a business might seek to buy the phones retail and sell them wholesale. This is difficult as retails phones can normally only be bought individually in a retail shop, and as the purchase of the phones is so labour intensive box-breakers need to employ a very large staff. The phones are then normally sold abroad in countries with lower subsidies (ie a higher retail price).

236. This evidence was clearly right. It was apparent that the Mr Wright's companies (not including EES) had been heavily engaged in box breaking trading. While the appellant criticised Mr Fletcher’s evidence, Mr Wright’s evidence of the appellant’s trading apart from the type of deals at issue in this appeal was consistent with Mr Fletcher’s description of box breaking.

237. It was also clear, and not in dispute, that the 32 deals at issue in this appeal did not involve box breaking. On the contrary the phones were purchased by EES in wholesale and not retail trades, and they were purchased in bulk and not in ones and twos.

238. Arbitrage trading occurs where the phone manufacturer sets different prices in different territories. An authorised distributor in one territory might sell phones via an intermediary to an authorised distributor in a higher priced territory.

239. Mr Fletcher agreed that there were aspects of the appellant’s trading at issue in this appeal that superficially resembled arbitrage trading, in particular the back to back trading without stock holding. However there were many aspects that did not resemble arbitrage trading. In particular, no ADs were involved and there were far too many middlemen.

240. Further, there was no opportunity for arbitrage trading as most of the phones were Nokia phones and Nokia had a homogenous pricing policy in Europe. While currency fluctuations could nevertheless create the possibility of arbitrage trading, the margins achieved by EES were well in excess of the currency fluctuations at the time so this is no explanation. Further, the sale and purchase was always in sterling: for currency arbitrage EES would have needed to buy in sterling but sell in Euros or some other currency.

241. We accept his evidence that EES was not in these transactions undertaking arbitrage trading.

242. Volume shortages; Mr Fletcher also described a grey market where a distributor can source volume shortages in a market faster than the OEM . This only works where a distributor is holding stocks speculatively and the specification of the phone will be very precise and only relatively small numbers of handsets would be involved. The distributor would have a good relationship with the MNO (multiple network operator).

243. This was clearly not the sort of trading at issue in this case: no stocks were held. Very large numbers of handsets were traded. An MNO was not part of the chain. And the deal documentation was insufficiently precise in Mr Fletcher's opinion for this sort of trading.

244. Dumping The fourth type of grey market described by Mr Fletcher was dumping which occurs where authorised distributors want to off-load surplus stock. His evidence was that the authorised distributor would often dump at a loss and would normally dump into another territory. It would trade on detailed specifications of the phone.

245. EES' deal chains did not have these characteristics as an authorised distributor was not involved, some of the phone specifications were very imprecise, and the phones remained in the same territory (ie Europe). Mr Fletcher was also of the opinion that the trades at issue in this appeal were not exploiting dumping opportunities because EES did not buy speculatively but to order, its trades were not one-off but followed a pattern and were in an excessively high % of the market. It was Mr Fletcher's evidence was that in his opinion phones dumped by MNOs and ADs into the grey market would not represent a material volume of handsets. Imports of phones into the grey market was included in his estimate of a mere 12 million handsets available.

246. As we have said, while we accept this evidence, it was in the large part superfluous, as the FCIB and other evidence overwhelmingly proved that the trades at issue in this appeal were organised as part of an MTIC fraud and clearly did not arise in a genuine market. Its main relevance was to demonstrate that Fletcher's opinions were reliable as his opinion that the structure of the appellant's trade did not match any structure on the genuine market was corroborated by other, independent evidence, such as the FCIB evidence. Mr Fletcher gave other, relevant evidence, which we accept as we found for the reasons above his evidence to be reliable.

Mr Fletcher’s other relevant evidence

247. It was Mr Fletcher's opinion evidence, which we accept, that only an insignificant number of handsets would be imported on the grey market and then re-exported. This is consistent with the market behaving rationally and acting to cut out unnecessary costs such as the cost of shipping.

248. It was Mr Fletcher's evidence, which we accept, that most box breaking took place in UK as UK had in general the highest subsidies. And even if some countries did have higher subsidies on particular phones, grey market traders in the business of box breading would still make most money by selling to lowest subsidy countries, and certainly not to the UK. So in Mr Fletcher's opinion box breaking from outside UK would not be a source of phones within UK. We have referred to this above (§181).

249. It was also Mr Fletcher's evidence that 59% of EES's trades in phones was where EES's apparent market share exceeded the actual market for that particular phone and the rest of its trading (at issue in this appeal) was for an amount in excess of 5% of market, whereas Mr Fletcher would not expect anyone to achieve that % share of the grey market in any particular phone.

250. It was also his evidence, which we accept, that EES traded for one deal in Nokia 8801 handsets although there was no market for phone in Europe. We accept his evidence that the phone was designed for the US market. It was the same phone as the Nokia 8800 which was designed for the European Market. The only significant difference between the 8801 and 8800 was the frequencies on which the phones would operate. The 8801 was designed to operate on all the frequencies used by MNOs in the US. The 8800 was designed to operate on all frequencies used by MNOs in Europe. While there was some frequency overlap between the US and Europe, the 8801 would have and did have no market in Europe because it could only operate on certain of the networks making it difficult for 'roaming'. And as an identical phone (the 8800) was available and compatible with all European frequencies, in Mr Fletcher's opinion there would be no market for the 8801 in Europe.

251. The appellant sought to discredit this evidence by proving that a 8801 phone did work in Europe. But Mr Fletcher's evidence was not that it would not work, but that it did not work on all frequencies and that therefore there was no market for it. We accept Mr Fletcher's evidence on this.

252. He backed up his evidence by reference to GfK data which showed a mere 299 of the 8801 were sold in 2006 in Europe compared to half a million 8800s. The appellant sought to discredit the GfK data. Mr Bridge pointed out that it only captured 70% of retail (not business to business) sales and only in the European countries for which there was reliable data. However, we agree with Mr Fletcher that the evidence of the GfK data was sufficiently representative of the whole picture of consumer demand in Europe that we accept that the 8800 had no market in Europe in 2006 or at any other time.

253. The appellant also attacked the reliability of GfK data because certain sales it made in 2005 were not reflected in the data. But the data related to 2006. We consider that the GfK data was indicative of the levels of retail trading in mobile phones.

254. We also accept Mr Fletcher's evidence that because Nokia did not intend the 8800 for the European market, they never manufactured a 8800 with a European specification. Therefore, when the invoice on which EES bought and sold the phones described them as Nokias with a European specification, the invoices described a phone which did not exist.

255. This evidence on excessive market share and trading in Europe a phone for which there was no demand only serves to reinforce the finding we have already made that there was no genuine market for the phones traded in the deals at issue in this appeal and that therefore the deals were artificially generated for the purpose of MTIC fraud.

Conclusion on connection to fraud

256. We have found that all of the appellant's 32 deals in the period at issue, and in particular the 31 deals at issue in this appeal, were connected to fraudulent tax loss. We have found that the fraud to which they were connected was MTIC fraud and that EES' purchases and sales were organised for the purpose of facilitating that fraud.

257. The question is whether the appellant (via its alter ego Mr Wright) knew or ought to have known this at the time of the deals.

Did the appellant know of the connection to fraud?

Reliability of Mr Wright's evidence

258. Mr Wright was the only witness for the appellant. We did not find his evidence reliable. We deal with particular aspects of its unreliability below but mention a couple of matters here.

259. As discussed below (§§312-320), there was a question which Mr Wright had refused to answer when asked by HMRC when they were investigating the circumstances surrounding EES' input tax claim, and he continued to refuse to answer it in the Tribunal hearing. That question was the source and amount of a loan. Putting aside the question of what, if any, inferences ought to be drawn from his refusal, we found that there was inconsistency in the reason he proffered for his refusal to answer the question.

260. Originally his reason was that he had signed a confidentiality agreement when he took out the loan on behalf of EES. However, under cross examination he said he would have provided the information if HMRC had obtained a disclosure order. The next day at the end of cross-examination, the judge explained to Mr Wright that the Tribunal itself had the power to order him to disclose the information, and that she would make such an order if that would enable him to answer the question. His reply was that he would not answer the question in any event. Therefore, we find his answer to Mr Bryant Heron the day before that he would have answered the question if HMRC had bothered to obtain a disclosure order was unreliable.

261. A further inconsistency in his evidence was that in Mr Wright’s witness statement he said that EES monitored the market and picked trades in which he believed EES could make the most amount of profit. His oral evidence, however, was that he always offered the deal to Globalfone as they were reliable and he trusted them, and did not try to see if anyone else would offer more. When asked why, on 8 of the deals he sold to J Corp and D Jensen, when he had no previous trading connection with them, his answer amounted to little more than he had had a conversion with the director on the phone and 'warmed' to him.

262. Another inconsistency in Mr Wright's evidence was that in cross-examination his evidence was that he was offered the phones at a price by the vendor which he did not negotiate. He then offered them to Globalfone at another price, dictated, he said, by the market. In re-examination he changed tack and said that he negotiated with Globalfone and when Globalfone agreed to buy the phones he then went back to his vendor and negotiated on price. Not only had he failed to mention this negotiation in cross examination despite having the opportunity to do so, it is contradicted by the other evidence. As already mentioned, the buffers' margins were very small and very rigid (see 188). We find there would have been no room for negotiation. Further, the evidence is very improbable on other grounds: if he knew the market price as he said, why would he not negotiate at the outset if he thought the vendor's price too high and if there was a market price, why was there such a difference between what EES paid and what it was paid? It seems to us that sometimes it suited Mr Wright to say he negotiated (as in his witness statement and re-examination) and on other occasions (cross-examination) he said he preferred to sell to a trusted customer rather than negotiate the best price.

263. In conclusion we did not find Mr Wright to be a reliable witness and treated all his evidence with caution. As indicated below, in many places we were unable to accept it.

Knowledge of MTIC fraud

264. We find that HMRC sent a great deal of information about the risks of MTIC fraud to EES and to Cellular (at a time when Mr Wright was the director). He was warned the UK’s losses amounted to over £1 billion. Mr Wright as director of Cellular and then as director of EES had had many visits from HMRC about MTIC fraud and since 2003 had been submitting details of his trades to HMRC on a monthly basis. Some returns had been the subject of extended verification, Cellular received veto letters (telling it that a company whose registration it had checked was no longer registered) and on in January 2006 Mr Wright had been informed that a trade undertaken by Cellular had been traced back to fraud although Cellular’s input VAT would not be withheld.

265. Mr Wright said that when the banks started closing mobile phone traders' accounts, he believed this was because they were put under pressure by HMRC who wanted to stop MTIC fraud. After a question from his counsel Mr Wright seemed to resile from this and indicated he wasn't sure if that is what he thought at the time. However, we think his first, less guarded answer was accurate as he used past tense and referred to speaking to an HMRC officer about it at the time. He was asked again about this in cross examination and his evidence was that he thought HMRC, with its concern about fraud, was behind the banks' actions. In conclusion, we find that at the time of the deals he thought concerns over MTIC fraud was the driver for the banks closing the accounts of mobile phone traders.

266. We find that, although he tried to play down in the hearing exactly what he understood about MTIC fraud at the time of the deals in question, he was well aware that MTIC fraud occurred in mobile phone trading, was the cause of millions of pounds of losses to the UK, and was a very serious concern to HMRC.

EES' pivotal position in the fraud

267. HMRC’s case is that Mr Wright knew that EES' transactions were connected to fraud and that this is evidenced (they say) because:

· All the company’s transactions trace back to fraud;

· The fraud was highly organised and it would be fanciful to conclude that Mr Wright was an innocent dupe;

· The circularity of funds in FCIB shows that the trading was directed and that therefore Mr Wright must have known the transactions were directed;

· The circularity of funds was not coincidental and Mr Wright must have known.

· Took substantial profit.

268. In summary, HMRC’s case is that this was a highly organised fraud and the broker was the lynchpin in it and profited from it. In HMRC’s view the only credible scenario is that the fraudster who organised the fraud would have ensured that the broker knew the part it had to play in the fraud, and in particular the broker would have been told from whom to buy at what price and to whom to sell at what price. And if the broker knew this, it knew its deals were organised for the purpose of fraud and not arising on the market. It is not entitled to recover its input tax.

269. We agree with HMRC that this was a highly organised fraud. The money moved in a circle, originating and ending with the same company. Indeed, the money moved in a circle within the same bank and we find this was controlled. Where a trader did not have an FCIB bank, the money moved to a nominee which did have an FCIB account. It was a highly sophisticated fraud involving a closed cell of about 60 companies. It involved inserting buffer companies between broker and fraudster or using the even more sophisticated contra trading method of protection with parallel chains.

270. We also find that it was absolutely essential to the success of the scheme that the goods moved through their pre-ordained route else the fraudster would not receive back the money it put into the banking system. It was essential to the fraudsters that EES, having purchased the goods from the appropriate buffer, then sold the goods to the designated acquirer, Globalfone, J Corp or D Jensen.

271. It is obvious to us that the EES’ purchases and sales – as to date, price and identity of trading partners - in these 32 chains were organised and determined by the fraudster who organised the overall scheme. We have entirely rejected the appellant's suggestion that EES's sale to its customer, Globalfone, was not organised by the fraudster.

272. Mr Bridge urges the Tribunal to conclude that the mere position of a trader as broker in an MTIC chain does not mean that they necessarily have to know of the fraud. Nevertheless, we agree with HMRC that, on the facts of this case, from the point of view of the fraudster, it had nothing to gain from using an innocent dupe as broker, while doing so would not only put it to the trouble of engineering a situation where the innocent dupe thought he was entering into openly negotiated purchases and sales which were not linked, but took the risk that the innocent dupe would sell the goods to someone other than the intended acquirer. This suggests to us that EES would not have been offered the chance to be the broker in this scheme if it was an innocent dupe.

273. Mr Wright suggested that the fraudster would avoid the risk of the putative innocent broker selling to the wrong party by simply withholding the goods from the chain of transactions if that happened. But it seems to us that this was a highly organised fraud: why would the fraudster take the risk of a loose cannon, who might sue for non-completion of sale, when he could simply identify a less honest broker who was prepared to cooperate?

274. Mr Wright’s response is that the risks the broker ran (of not recovering its input tax) were so great that no one would knowingly agree to be a broker in an MTIC fraud. We don’t agree. Until Spring 2006 when HMRC carried out extended verifications before authorising repayments, it seems likely to us that that the risk of input tax being refused was considered to be low. And so far as Mr Wright himself knew, HMRC had actually refunded Cellular VAT on a transaction which HMRC had found was connected with fraud. Viewed from 2013 with hindsight, being part of an MTIC fraud chain in 2006 was considerably more risky than it would have appeared at the time. So we simply do not agree that in 2006 the risks were perceived to be so great no one would knowingly be a broker in an MTIC fraud chain. The appellant was, anyway, on his own evidence, someone prepared to take risks (he traded without insurance and checking whether his buyers were good for the money – see §293 & 296).

275. HMRC point to the profit which EES achieved on these deals. In their opinion the large profits were the trade off for knowingly taking the risk of being a broker in an MTIC chain. We agree with HMRC that, in such an organised fraud, it was more likely than not that the fraudster would not use an innocent dupe as the broker, and that the level of profit achieved by EES indicates that it was not an innocent dupe. It achieved a very high level of profit for doing virtually nothing (other than taking a risk of HMRC not refunding the VAT).

276. Mr Bridge put the case that the profit margin for EES on these deals was achieved by EES in some of its later deals after the reverse charge had come in and which HMRC do not suggest were part of a fraud. But we note that what matters is the level of profit: there were 31 of these deals and all for very large quantities. EES's profit (assuming it received the VAT refund) on these 31 deals would have been just over £2.5 million. We find that these levels of profitability arising out of a mere 6 days' trading was in absolute terms very high, as was the turnover. Its later trading was for a much smaller turnover in any comparable period of time and it follows, therefore, that in absolute terms its later trading was significantly less profitable over any comparable period of time.

277. It was put to Mr Wright that the ease of the transactions should have put him on notice that the transactions were organised for the purpose of fraud. He denied this. He said that he protected EES by always trading with the same companies he had been trading with for years.

278. We consider that EES realised a turnover of £46 million in a few months by making back to back purchases on a small number of days where it was offered the phones at a price on which it could make a very large profit by selling to an easily located willing buyer (in fact the same buyer in most of the transactions). It did this without any apparent effort and while Mr Bridge pointed out that Mr Wright's payroll company employed some 50 people we do not find that any of them were employed on these deals which were entered into by Mr Wright personally.

279. We also find that Mr Wright ought to have known that trading with the same companies was no protection as he had been notified of a loss in a chain (see XXX). And further, in any event sometimes he chose to trade with companies with whom he had had no previous dealings (D Jensen and J Corps).

280. It was put to him that he was told from whom to buy and to whom to sell. He denied this but for the reasons given above we are unable to accept the denial as reliable.

281. It is very much more likely than not that EES would have been aware that its purchases and sales were being dictated to it. And if it knew that, bearing in mind Mr Wright's background knowledge of MTIC fraud (see §§264-266), Mr Wright must have known EES was participating in a fraud.

282. This conclusion is enough to dispose of the appeal but in any event we go on to consider how Mr Wright behaved at the time: did he behave consistently with someone who knew he was participating in a fraud or as someone who thought they were trading in an open market would behave?

Choice of buyer

283. Mr Wright’s evidence was that the market was supplier-led. In other words he would be offered phones to buy and he would then try to identify a buyer for them. We know that, on the contrary, the deals were artificially contrived by a fraudster for the purpose of fraud. We do not accept that Mr Wright’s choice of customer (Globalfone, D Jenson and J Corp) arose by chance: it was dictated by the fraudster.

284. Indeed, Mr Wright’s own evidence was that he chose to trade with Globalfone without even looking for a possible alternative customer offering a better price. His explanation, which we find unconvincing, was that this was because he had dealt with Globalfone for years and trusted them. This is no explanation of why Mr Wright did not chose to maximise profit by even looking to see if someone else trusthworthy would offer more. Further, it fails to explain the deals that were not with Globalfone. Neither EES nor any of Mr Wright's other companies had had any previous deals with J Corp and D Jensen so there was no relationship of trust built up over time. Mr Wright's own evidence suggested that he was happy to trade with these two unknown companies without a relationship, without checking their credit worthiness, but simply on the basis of phone conversations in which he 'warmed' to the companies' director.

285. We do not accept this explanation as reliable. We consider Mr Wright's behaviour consistent with the behaviour of someone who knew his deals were organised for him and not consistent with the behaviour of someone who believed he was trading on the open market and looking out for the best deal with a credit worthy counter party.

Documentation

286. EES’ invoices contained the term:

“Ownership of the goods remain the property of Cellular Solutions (T Wells) Ltd until full payment has been made.”

287. This was the only condition which EES imposed on its customers. Clearly, this was an invoice originally used by Mr Wright for the business of Cellular, although he no longer owned this company at the time of the deals in issue. Mr Wright continued to use the invoice for EES' deals: it appears that neither EES nor its customers either noticed the error or bothered to change it. This suggests and we find that Mr Wright and his customers were not really concerned with the terms on which they traded. This is consistent with knowledge that the trades were fraudulent and inconsistent with a belief that the trades were on the open market between third parties.

Contracts

288. That retention of title clause was the only condition which EES imposed on its customers in the deals at issue in this appeal although there were millions of pounds of goods at stake and the goods were being shipped abroad before payment.

289. It was Mr Wright's evidence that in all his years of trading, including the deals which HMRC do not impugn, such as the box-breaking and trading post the reverse charge, he had never signed a contract. He accepted his suppliers might impose terms and conditions on his company via the print on the back of invoices.

290. His view was that EES was not at risk in these 32 deals if the goods were lost because the ownership of the goods was, in his view, with the ultimate owner of the goods. EES did not own them, he said, because EES had not paid for them. However, he did accept that EES would be at risk if the buyer failed to pay because then EES, being committed to pay for the goods, would be obliged to find a new customer and might not be able to cover its purchase price.

291. As EES was therefore at risk, we consider a person believing themselves to be trading on the open market, would be concerned to ensure that its customers would pay for the goods when they arrived.

Due diligence

292. Yet Mr Wright agreed that he did not undertake credit checks on his two new customers, D Jensen and J Corp. EES undertook no credit checks on Globalfone either but Mr Wright's explanation for this was that they were a long established customer of his companies.

293. We find that while EES carried out some due diligence, such as checking the identity of the directors of the companies with whom it traded, it did nothing to check their credit worthiness, such as carrying out credit checks or taking up trade references.

294. We find that some of the due diligence was carried out after the first deals with the companies and in some instances after the deals at issue in this appeal. In one case, a copy of a document was completed after it had originally been provided to HMRC. Mr Wright's case was that EES continually updated its due diligence and that the latter discrepancy was just an error by a bookkeeper. We do not accept this evidence as reliable. There is no documentary evidence of earlier due diligence so we find it was not merely being updated. We also note that Mr Wright's first witness statement claimed that EES was very careful to ensure it had due diligence in place but we find that this was not a reliable statement, as EES did not.

295. Mr Wright's explanation was that he took a calculated risk in not checking credit worthiness. He accepted he was at financial risk if his buyers did not pay him as at the least he would have to find new buyers who might pay less but he offered no explanation of why he did not try to mitigate this risk by making some sort of check such as a credit check. Our opinion is that his behaviour, both in trading with J Corp and D Jensen without any checks on their creditworthiness, and in collecting due diligence after the event, is consistent with knowledge that the deals were orchestrated (as the deals were orchestrated EES did not need to protect its position) and inconsistent with belief that the trade was genuine (where a trader would believe he was genuinely at risk).

Insurance

296. EES did not hold insurance. Mr Wright said to HMRC and then to tribunal that he “self-insured” although we find that this means EES was uninsured.

297. His view, set out above, was that EES was not at risk if the goods were damaged or stolen en route as EES did not own them until they paid for them and that the freight forwarders would be insured. However, he never took advice on whether this was legally correct.

298. We also find that Mr Wright's companies never held insurance. For instance, with his box breaking trade, he took a calculated risk by holding the uninsured stock in the company's own, very secure, warehouse.

299. The goods involved in the deals at issue in this appeal were in total worth some £38 million in value and we find, rather than taking a calculated risk, Mr Wright's choice was simply to take a risk on their loss. He did not take steps to know his legal liability and therefore could not take a calculated risk. We find this behaviour consistent with knowledge that the deals were orchestrated for the purpose of fraud and that therefore there was no real risk, rather than with the behaviour of someone who thought the trades were on the open market.

Reaction to notification of fraud in chain

300. As mentioned above, Mr Wright received notification from HMRC of tax loss in a deal chain involving Cellular in January 2006. Mr Wright's reaction, according to his evidence, was to satisfy himself that his supplier did proper due diligence and, as HMRC repaid him, just carry on with the same sort of trades. He did not change suppliers. He said that he saw no problem as he was repaid his VAT and did not receive a 'joint and several liability' letter from HMRC.

301. While not conclusive by itself, we consider that this reaction was more consistent with the reaction of someone who knew that their deals were organised rather than that of someone who had thought they were trading in a genuine market.

302. We also note that in examination in chief Mr Wright said he found it “quite frightening” to look at HMRC’s evidence and realise EES was “that close” to fraud. Yet this seems a somewhat disingenuous statement to us when, in 2006, his reaction to being told that Cellular had participated in a fraudulent deal chain was to check that its supplier’s completed a due diligence checklist but otherwise to carry on as before.

303. We consider the other relevant evidence about what Mr Wright knew at the time.

Mr Wright's knowledge of mobile phone trading

304. As we have said, we accept that Mr Wright was a very experienced mobile phone trader. We have also found that he was well aware of the risk of MTIC fraud in mobile phone trading. His case was that he had satisfied himself that the deals EES undertook were not driven by fraud.

305. We consider that a trader who believed his deals were on the open market would understand the commercial rationale for the deals: he would understand what his role was in the market. And we find that Mr Wright was able to give a detailed and commercially rational explanation for the box breaking trade undertaken by his companies and explain the role that his company played in such trade.

306. When asked in respect of Mr Fletcher’s evidence whether he thought the EES trades in this appeal were arbitrage Mr Wright’s answer was that “I never thought about which particular opportunity I was taking advantage of. I was buying and selling”. This was not consistent with his ability to explain the opportunities his other companies exploited, nor with the explanation put forward by him earlier in the hearing of why (he said) he was not surprised to be able to undertake the trades at issue in this appeal. We have dealt with this explanation at §178-185. It made no commercial sense. We did not find this explanation convincing and do not accept his evidence as reliable that he thought at the time that this explained his ability to make such profitable and easy deals. We agree with HMRC that Mr Wright was an astute mobile phone trader and not the sort of person to be duped, particularly not in a market he understood so well. It was put to him by counsel that he knew his trades were organised. He denied this but we are unable to accept the denial.

Market share

307. The evidence we have already referred to at §§249-255 shows that EES traded in some cases well in excess of the retail demand for the particular phone. It was put to Mr Wright that as a knowledgeable mobile phone trader he must have known he was dealing in amounts of stock well in excess of retail demand. His answer was that he always believed his trade was legitimate. We do not accept this as reliable: Mr Wright clearly knew a great deal about the mobile phone market and he must have known was dealing in excessively high numbers of particulars models.

Choice of Bank

308. Mr Wright indicated in examination in chief that he opened EES’s FCIB bank account because in late 2005/early 2006 his company’s UK bank accounts were being closed (we have mentioned this evidence before in the context of Mr Wright’s belief at the time that this was because the UK banks were under pressure from HMRC to close the bank accounts of mobile phone traders in an attempt to stop MTIC fraud – see §265).

309. However, he opened an FCIB account for Cellular (whose trading EES took over) in June 2005. He said he could not recall if at the time any of Cellular’s bank accounts had been closed although Recycled (another of Mr Wright’s companies) had had one account closed in that same month. He said that his decision to open Cellular’s account “quite possibly” could have been because its UK banks accounts were being closed and that he would “absolutely not” have had any other reason to open an FCIB account. He denied that he opened EES’s FCIB account because he had to do that in order to participate in the MTIC scheme.

310. We take into account that all traders in the organised fraud had FCIB accounts. While the closure of UK bank accounts might explain why UK-based traders did this, no explanation (other than MTIC fraud) was offered why the non-UK based traders did this (eg Comica) or why those without FCIB accounts used proxies with FCIB accounts. The obvious explanation is that it made it easier track the funds in their circular movement from trader to trader and back to the funds originator (Comica, Parasail (Negressco) or Marxman), so the fraudsters did not lose control of their money.

311. We are unable to accept Mr Wright’s case that he only opened Cellular and EES’s FCIB accounts because of UK bank accounts being closed. There was no documentary evidence Mr Wright had been told Cellular’s accounts would be closed at the time Mr Wright opened its FCIB account, and Mr Wright was less than definite about it himself saying he needed to see the documentation before he could commit himself on this. And it was clear that the fraudsters would have had an interest in participants having FCIB accounts. We do not accept his denial that he did not open this account to participate in the MTIC scheme, and we find it he was told to open this account in the same way he was told from whom to buy and from whom to sell.

The Chanar loan

312. The facts: During extended verification HMRC discovered 5 letters from Chanar requesting EES to pay monthly service charges which were due for February to June 2006 although there was no record that they were ever paid. In answer to a query from HMRC, EES's solicitors provided HMRC with a copy of an agreement between EES and a Leichtenstein company, Chanar Porfolio Ltd ('Chanar'). This stated that Chanar was making a loan to EES of an unstated amount in return for a 50% share in the profits of the business. It is not really clear whether this was a loan agreement or a sale of 50% of the business particularly as no repayment date was specified. Mr Wright refused to answer questions to HMRC on who was behind Chanar or how much money was paid to him under this contract. He would say nothing about it in Tribunal hearing either, other than to say that he had signed a confidentiality agreement and that the loan was not connected in any way to fraud.

313. The law: The recent Supreme Court decision in Prest v Petrodel [2013] UKSC 34, dealt with the law on what inferences can be drawn from a witness’ refusal to answer questions. Lord Sumption said:

“There must be a reasonable basis for some hypothesis in the evidence or the inherent probabilities, before a court can draw useful inferences from a party's failure to rebut it.”

314. He went on to say that this is qualified in family cases because the economically dominant spouse (normally the husband) controls the evidence relating to the family finances:

“The concept of the burden of proof, which has always been one of the main factors inhibiting the drawing of adverse inferences from the absence of evidence or disclosure, cannot be applied in the same way to proceedings of this kind as it is in ordinary civil litigation. These considerations are not a licence to engage in pure speculation. But judges exercising family jurisdiction are entitled to draw on their experience and to take notice of the inherent probabilities when deciding what an uncommunicative husband is likely to be concealing.”

315. While in the normal tax case the burden of proof is on the taxpayer who normally exclusively controls the evidence, this is not so in an MTIC case. It is established by binding authority that allegations of fraud against a taxpayer must be proved by HMRC (see Mobilx at §81 – cited at §35 of this decision). Indeed, even to the extent an MTIC case does not involve an allegation of fraud against the taxpayer (such as where the allegation is of connection of fraud of which the taxpayer ought to have known) the burden, we think, is still on HMRC because it is HMRC which controls the evidence (the judge has explained this view in more detail in $10-18 in Massey t/a Hilden Park Partnership [2013] UKFTT 391 (TC)).

316. Therefore, in this case, unlike an ordinary tax case, the burden of proof is entirely on HMRC. In order to succeed they must prove the connection to fraud and that the appellant knew (or ought to have known of it). Therefore this is the sort of case, unlike an ordinary tax case, where it may be appropriate for the tribunal to draw adverse inferences from the failure of a witness for the appellant to answer questions put to him by HMRC.

317. Nevertheless, as Lord Sumption said, this is not a licence to engage in pure speculation. It is merely that the tribunal must take notice of the inherent probabilities when deciding what the witness is likely to be concealing. So we take into account what we know and the inherent probabilities.

318. Mr Wright said that the only reason for secrecy was to protect the identity of the person making the loan, although he gave no reason why the lender would wish to protect his identity. However, we cannot think of any reason why a person, with a 50% stake in the business, and with no knowledge of the fraud, would wish to remain unidentified, particularly when withholding the information risked the tribunal making adverse inferences and therefore made a successful appeal less likely. If EES's appeal fails, EES is less likely to have the funds to pay Chanar.

319. As we said at the outset (§18), MTIC fraud requires the brokers to be in funds to bridge the gap from the date of the transaction until the date of the VAT repayment. If the fraudster wants the broker to increase the deals it undertakes, it may be in the interests of the fraudster to put the broker in funds. In this case, Mr Wright will not say how much money Chanar provided. But the terms of the loan do not look to be a commercial loan: while the lender gets a 50% stake in the business the amount lent is not stated.

320. We take into account all the other evidence in tribunal, the unreliability of Mr Wright's evidence, and the inconsistency between the appellant's case that they knew nothing and the manner in which the appellant actually behaved, which indicated knowledge. The inherent probability is that this loan was made by the fraudster or someone connected to him with the intention of putting EES in funds to undertake more MTIC deals than it would otherwise have been able to do, and so we find. Mr Wright's refusal to identify the person behind Chanar was more likely than not because he knew it would indicate a connection between EES and the fraudster and we find his statement that Chanar was not connected with the fraud unreliable.

Mr Wright’s connections

321. In February 2006 Mr Wright sold his company Cellular to Mr Imran Hafeji. Mr Hafeji had been a director of, and was still an officer of, Svenson and , who, as we have found at §206 was a company knowingly involved in the MTIC fraud at issue in this appeal.

322. At the same time Mr Pooley sold Fonecode to Mr Yusaf who was a director of Powerstrip. We find Powerstrip was knowingly involved in the fraud – see §213. Some years before Mr Wright set up Fonecode for Mr Pooley who was a personal friend of his. Fonecode shared premises with Mr Wright's companies in 2004. The purpose of Fonecode was to exploit software developed by Mr Pooley for unlocking mobile phones. The company was not profitable as by 2005 it was clear that similar technology was available elsewhere for free. In February 2006 Mr Pooley sold Fonecode to Mr Yusaf, who was a director of Powerstrip.

323. EES, Fonecode and Cellular all became brokers within 'Cell 5'. HMRC's case is that at about February 2006 the fraudster behind Cell 5 was putting the organisation in place to undertake the elaborate contra trading scheme that HMRC now designate Cell 5. It is their case that it is no coincidence that Mr Wright sold Cellular and that his friend sold Fonecode to persons who were knowingly involved in this scheme at that time.

324. Mr Wright was asked about his sale of Cellular. Mr Hafeji paid £350,000 but Mr Wright was unable to identify what he paid for. The company had no assets and after the sale Mr Wright carried on Cellular’s business through EES. He was unable to explain why he thought Mr Hafeji wanted to buy the company. According to Mr Wright’s oral evidence, Mr Hafeji did not even ask him how Cellular traded. Export broking was not discussed. Further, Mr Wright did not mention the sale to HMRC at the time even though he was in regular contact with them.

325. At about the same time, EES took out the Chanar loan. It was on its face, and we find, intended to finance at least in part, EES's trading. We have found it was made by the fraudster or someone connected to him – see §320. As EES' only trade was the sort of deals at issue in this appeal, it was intended to finance the sort of deals at issue in this appeal. Those deals were orchestrated for the purpose of fraud. It is HMRC's contention that the loan, as well as the company sales, were all part of the fraudster setting up Cell 5 and that Mr Wright knew that.

326. We agree with HMRC that the purchase of Fonecode and Cellular was by persons knowingly involved in the fraud and that more likely than not the purchase was to facilitate the fraud by increasing the number of brokers available (and the logic of the fraud requires there to be a lot of brokers so that each broker's VAT reclaim would not be of a truly incredible amount of money). We agree with HMRC about the Chanar loan too: for the reasons explained above, we find it more likely than not to have been made by someone connected to the fraud.

327. Did Mr Wright know any of this? His evidence was that the sales of the two companies was unconnected. Taking into account his lack of curiosity of Mr Hafeji's reasons for paying £350,000 for a company with no assets, the fact he did not draw the sale to HMRCs intention, the obvious coincidence in all three happening at once, and all the other circumstances of this case, we unable to take what he said as reliable. We think he did know all three were connected. This too indicates that he was knowingly involved in the fraud.

328. We also note that we have found Globalfone must have been knowingly involved in the fraud (see XXX). Mr Wright was rather ambivalent about the state of Globalfone's knowledge. At one point he said that he thought they had been 'got at' but later said he had only meant that the freight forwarders would stop a deal if Globalfone did not sell the goods within the chain. He also said that at some stage in the future he planned to sit down with Globalfone's director and ask him what Globalfone's position was. His evidence was that Globalfone's director (Stephen Partoll) was a close family friend. We think his ambivalence in the Tribunal was from a reluctance to accept in the hearing what was obvious which is that Globalfone was knowingly involved in the fraud.

329. We agree with HMRC that Mr Wright was friends with, and sold his company to, and took loans from, and traded with, persons knowingly involved in the fraud. It is very unlikely in these circumstances that Mr Wright did not know what was going on: taking into account our other findings on knowledge all is clear. Mr Wright knew very well that his deals were orchestrated for the purpose of fraud.

Payment for deals 12 & 15

330. Deals 12 & 15 were some of those which took place at the end of April. We find that Globalfone (EES’ customer) in both these deals paid EES for Deal 12 (invoice 671) on 2^nd May. (The payments actually referred to invoice 693 which was obviously a mistake but totalled the exact amount outstanding at that time on deal 671.) We find Globalfone paid EES for deal 15 (invoice 675) on 8 May as that was when the exact amount outstanding on that deal was transferred.

331. EES, however, we find used the funds from Globalfone on deal 12 on 2 May to pay its supplier (Svenson) on deal 15. We find this because the funds paid to Svenson was exactly the amount due on deal 15 and in any event this was not in dispute.

332. In other words, EES used funds from deal 12 to pay for deal 15, before it had received the funds from D Jensen on deal 15. In both cases its supplier was Svenson, so this makes little sense: deal 12 took place earlier in time than deal 15 so where the supplier was the same one would expect the first in time to be paid first. Mr Wright’s explanation was that he paid the most “pressing” invoice first but this makes no sense. It could make no difference to Svenson to which invoice EES allocated the money: Svenson would be free to allocate the money it received however it wished and in any even Deal 15 (paid first) was for a slightly lower amount than Deal 12, so its more likely to have wished Deal 12 to be paid first.

333. We find from the FCIB evidence that the funds Globalfone used to pay EES on deal 12 originated and ended with Negresco. The funds which D Jensen used to pay EES on deal 15 originated and ended with Marxman.

334. It is not clear to us why the organisers of the fraud routed the funds in this fashion. It is clear that EES’ decision to pay for Deal 15 instead of 12 (and vice versa a few days later) meant that the funds completed their loops, so that funds which originated with Negresco ended with Negresco, and funds which originated with Marxman ended with Marxman. If EES had not made that decision, we find funds originating with Negresco would have ended up with Marxman and vice versa.

335. It was put to Mr Wright that he knew he was taking part in directed trading. He denied this but we are unable to accept the denial. There is simply no credible explanation of why EES paid Deal 15 before Deal 12 unless Mr Wright was instructed to do so by the fraudsters. And the fraudsters had a good reason to do this as (we find) they wanted to complete the financial loops.

336. Conclusions

337. Therefore, we find that on many occasions (§§ 285, 287, 295, 299, 301, 306, 307, 311, 320 & 335) EES acted in a way consistent with knowledge of fraud and inconsistent with being an innocent dupe. We had found in any event that it was highly unlikely that it was an innocent dupe (see §281). Therefore, for all the reasons given above, we find that EES by its director Mr Wright knew that all 31 of its transactions at issue were connected to fraud.

338. It was put to Mr W that his large mark up (virtually 50% of the VAT) was not real trade but a reward for broking an MTIC transaction. It was put to him that he was the one who had to ‘carry off’ the fraud as he had to ‘front up’ the fraud. It was put to him that he knew his deals were part of a fraud. His denied knowledge but for the reasons given we do not accept this.

339. We dismiss the appeal in its entirety.

Means of knowledge

340. The allegation that, if the appellant did not know, it had the means of knowledge of fraud connected to its transactions, was very much HMRC’s secondary case. We have found that the appellant knew that its transactions were connected to fraud so it is unnecessary for us to consider means of knowledge.

341. Nevertheless, for the sake of completeness, we find that the appellant ought to have known that its transactions were connected to fraud. It knew that it was offered phones for sales on really easy terms (such as not having to pay until it had been paid) and that it would be able to sell them immediately very profitably. It knew that the identify of its buyer was dictated to it. For these and all the other reasons mentioned at XXX, it ought to have made the inference that it deals were connected to fraud

342. Mr Bridge says that the Tribunal is unable to make a finding of constructive knowledge against EES because it was not put to Mr Wright that he ought to have known. Mr Bridge is mistaken. Means of knowledge is an objective test and not an allegation of fraud: it does not have to be specifically put to the appellant.

Footnote - Pleadings

343. Mr Bridge relied on the requirement of natural justice stated in many cases (such as Three Rivers DC v Bank of England No 3 [2003] AC 1 per Lord Millett at §183) that fraud must be distinctly alleged and sufficiently particularised:

[183] [fraud must be] distinctly alleged and as distinctly proved; that it must be sufficiently particularised; and that it is not sufficiently particularised if the facts pleaded are consistent with innocence.

344. We understand it is Mr Bridge's submission that the appellant did not understand, from HMRC's statement of case and the evidence on which they relied, on what factors HMRC relied to put its case that EES' transactions were connected to fraud where that connection was said to be by contra trading, nor how in any of the 31 deals at issue, it was said that Mr Wright had knowledge or means of knowledge.

345. Firstly, the pleading requirements for fraud would not apply to the pleading of connection as connection to fraud can exist without knowledge of the fraud. However, if we are wrong on that, we find in any case that it was sufficiently pleaded because HMRC’s statement of case had annexed the deal chains to show alleged connection in both straight and contra trading transactions.

346. However, for the reasons given at §159 above we do think that HMRC needed to plead the nature of the alleged fraud, as that impacts on the question of knowledge. We find that HMRC did plead that the alleged (and now proved) fraud was MTIC.

347. So far as the pleading of knowledge is concerned, HMRC pleaded:

· Mr Wright’s general knowledge of MTIC fraud in mobile phone trading;

· Failures in due diligence;

· The failure to provide details about the Chanar loan;

· All traders including EES holding an FCIB account;

· Back to back “easy” trading;

· Appellant’s lack of enquiry into rationale for why European phones came into UK only to be sold back to Europe;

· Lack of written contracts;

· Lack of commercial behaviour by Svenson, Globalfone;

· Sale to Mr Hafeji of Cellular;

348. Cell 5 and the relevance of the FCIB evidence was pleaded in a supplementary Statement of case.

349. While the Statement of Case was not particularly detailed, we find that bearing in mind the detail of the allegations contained in the various witness statements served, the appellant can have been in no doubt of the allegations made against it. In any event Mr Bridge’s complaint mostly lacks any specifics.

350. What he does say is that the Tribunal can’t make adverse inferences about refusal to answer questions about Chanar loan there was no suggestion in the Statement of Case that the loan was of any signficance. Apart from the fact it seems difficult to see how HMRC could anticipate that a witness would refuse to answer a question put to him in a tribunal hearing, HMRC did in fact plead that Mr Wright had refused so far to answer questions about it. It was clear significance was attributed to this.

351. We agree that HMRC did not plead that the appellant knew of the identity of the contra traders or that it knew the identity of the defaulters. And we agree that HMRC has not shown that the appellant knew the identity of the contra traders or defaulters nor has it shown that the appellant knew that the connection to fraud (in those 8 contra trading chains) was via a contra trade rather than in the same chain of transactions. We consider it quite possible that EES did not know the identity of the defaulters or whether it was a contra trade transaction.

352. But as we have said, HMRC only have to prove knowledge of connection to fraud. They do not have to prove the appellant knows precisely how its transaction is connected to a default by a particular defaulter. Therefore, the failure to plead the appellant knew these precise details is of no importance. They don't have to be proved so they don't have to be pleaded.

353. We find HMRC did plead the matters which we have relied upon to found our finding of knowledge. The appellant did therefore know the case that it had to answer.

354. Footnote – decision on procedure

355. The appellant asked us to record in writing the reasons for various interlocutory decisions made during or just before the hearing of this appeal, although the reasons were given orally at the time.

356. The first concerned an application for disclosure. The appellant applied to the Judge sitting alone shortly before the commencement of the substantive hearing for the disclosure of HMRC’s internal policy documents especially those in use in 2006 at the time of input tax denials. HMRC’s policy in 2006 is a matter of fact.

Mr Stone's latest witness statement:

357. Mr Stone was a witness in this case and in many other cases in which HMRC allege MTIC fraud. The judge was told his witness statement has evolved over time and in particular has evolved since it was served in this case. HMRC agreed to disclose the latest full version of Mr Stone’s report.

MTIC Policy Guidance:

358. The appellant applied for disclosure of HMRC's policy guidance on MTIC on the grounds it was relevant because, said the appellant, it was unlawful.

359. The appellant wanted to put the case that HMRC acted unlawfully by discriminating against taxpayers after the event: they chose to ‘go after’ brokers in MTIC fraud chains rather than the buffers or defaulters.

360. The judge refused the disclosure application on this basis. Such an allegation could only be a matter of judicial review. This is not a judicial review case as the Tribunal has no jurisdiction to consider the lawfulness of HMRC's behaviour (other than where it founded the appellant's liability and that is irrelevant here where the claim is merely that, where a number of taxpayers have tax liability arising out of the same matter, it is unfair to pursue some taxpayers rather than others).

361. The Tribunal is concerned only with whether HMRC were right to withhold the appellant's input tax, not whether it was fair to make repayments (or at least fail to assess) other taxpayers involved in the same fraud.

362. It also appeared to be the appellant's case that it was complaining that irrespective of HMRC's published policy, HMRC had a private policy to refuse all input tax reclaims over a certain amount and carry out extended verification.

363. The judge was not shown anything to substantiate existence of such a policy so this looked like a fishing expedition in any event.

364. So far as relevance was concerned, again in was a claim that, despite any question of the appellant's liability, the strict legal position should not be enforced on the grounds of public law. Irrespective of any chance of such a claim succeeding (which appeared to the judge to be very remote), such a claim could not be made in this Tribunal for the reasons given above.

365. Mr Bridge suggested that HMRC's policy might be the explanation for how EES behaved at the time and that therefore it was relevant. But this is only true if EES knew about the policy at the time, and so in so far as EES' application was for disclosure of internal documents of which they were and are still unaware, the application could not be allowed on this basis. And they could in any event give evidence about what it believed HMRC’s policy at the time to be if that was the explanation of how it acted as it did.

366. Mr Bridge then claimed that the market was flooded with goods connected to fraud because of HMRC’s alleged (unlawful) policy. Again the Judge was unable to see the relevance of HMRC's policy to such a case: while it would be relevant if as a matter of fact a genuine open market was flooded with goods on which they had been an earlier default, the reason for that would be irrelevant. In any event Mr Bridge did not make any credible case of how HMRC's policy could have been responsible for such a scenario.

367. Mr Bridge also suggested that HMRC's policy documents were relevant as their policy was (he said) inconsistent with the CJEU's decision in Mahageben as HMRC required traders to undertake due diligence. But even if the appellant was refused input tax deduction because it failed to undertake due diligence rather than on the grounds set out in Kittel (of which there is no evidence), the Tribunal was not concerned with why HMRC had denied input tax, it was concerned only with whether the appellant was entitled to recover its input tax.

368. So the internal policy documents had no relevance and the Judge refused to order their disclosure.

MOK submissions and response

369. Mr Bridge also applied for what is known as the “MOK submission” by EES' HMRC officer to policy setting out the grounds on which that officer thought EES should be denied its input tax. He also wanted HMRC's policy unit's response to this submission.

370. Mr Bridge's position was that this was relevant because he believed the officer was instructed to deny input tax rather than forming her own view on the matter. There was nothing to which Mr Bridge drew the judge's attention which indicated that this might have been the factual position and therefore this looked like a fishing expedition and was refused.

371. In any event, more fundamentally, it was refused on the grounds it would lack relevance even if the allegation could be substantiated. As stated above, the Tribunal was not conducting a judicial review of the decision reached by HMRC to deny input tax and the reasons why input tax was denied are therefore irrelevant. The question for the Tribunal is simply whether the appellant was entitled to recover its input tax. HMRC's decision making process is irrelevant to that question.

Policy on double recovery

372. Mr Bridge also asked for disclosure of HMRC's policy on double recovery (ie HMRC's policy on whether they must be able to demonstrate that the input tax denied to the appellant is no more than the loss they have suffered from a default which has not been recovered from any other taxpayer). Mr Bridge drew the judge's attention to another case in which the judge in that case ordered disclosure of this policy.

373. However, in that other case the other judge did not apply the right test (the test of relevance) in making the order for disclosure, and, having ordered the material to be disclosed, then found it to be irrelevant in reaching her decision.

374. In EES, the judge's conclusion was that the policy was self-evidently irrelevant as even on the appellant's case it post-dated the deals in issue and could have no bearing on what the appellant knew at the time of the deals in question.

375. Mr Bridge's real point is that he considered HMRC's original policy on double recovery to be right as a matter of law, and that was a matter for submissions.

376. The applications for disclosure were refused (save to the extent for the agreed disclosure of Mr Stone's latest witness statement). It was right to refuse at this late stage (and indeed at any stage) broad applications for disclosure of material that would not be relevant and the consideration of which would unnecessarily extend the length of the hearing. The test the Tribunal would apply to the appellant's appeal would be that of Kittel, and the material sought was therefore not relevant to the issues in the appeal.

Admissibility of Mr Fletcher's expert evidence

377. It was agreed that the appellant would make its challenge (if it wished) to Mr Fletcher's relevance, credibility, independence and expertise at the hearing.

378. At the interlocutory hearing before the substantive hearing commenced, the appellant applied for the evidence to be excluded on the grounds that CPR 35 had not been followed. In particular the appellant asked the Judge to follow her decision in Narain [2012] UKFT 188 (TC) at §14 and exclude Mr Fletcher's evidence unless he provided a further witness statement in which he swore he was complying with CPR 35.

379. HMRC objected to this course of action on the grounds of expense of obtaining a further witness statement.

380. So far as CPR 35 is concerned, the Judge considered that §14 of Narain as correct. These proceedings are akin to high court proceedings, with allegations akin to those of fraud made against the appellant, and there is every reason why the Tribunal would only wish to hear from a properly qualified and independent expert.

381. The judge took the view, however, that in view of the shortness of time before the substantive hearing and the expense (neither at which were mentioned as issues in Narain), the appropriate solution would be for HMRC's counsel to ask Mr Fletcher at the start of his examination in chief whether his witness statements complied with CPR 35. If they did not then his evidence would be excluded.

382. At the substantive hearing, Mr Fletcher did confirm as part of his evidence that his witness statements were compliant with CPR 35. And as Mr Bridge had forewarned HMRC, he challenged the admissibility of Mr Fletcher's evidence on the grounds:

· he was not in the appellant's opinion an expert as he was not experienced as a mobile phone trader nor an academic;

· Mr Fletcher's evidence was not, in the opinion of the appellant, relevant to issues in appeal;

· and he was not, in the appellant's opinion, independent because his employer, KMPG, acts for Nokia and has acted for the anti-grey market alliance. And Mr Fletcher, in his report, had relied on material from Nokia which had not been disclosed.

Relevance – was the evidence reasonably required?

383. Mr Bridge's case was that market was flooded with phones on which VAT had not been accounted for on acquisition into the UK as the VAT system effectively invited acquisition fraud. It was a market subsidised by fraud, said Mr Bridge. Mr Stone's evidence was that the 'true' grey market unaffected by fraud was tiny, but the appellant's deals (ignoring the contra trades) were connected to a fraudulent default so it was clear, said Mr Bridge, that Mr Fletcher could have nothing relevant to say about a market about which he knew nothing as his evidence concerned only the tiny 'true' grey market.

384. Mr Fletcher agreed that he was not an expert on fraud and gave no evidence about how a fraud might have created market opportunities.

385. The Tribunal ruled that Mr Fletcher's evidence was relevant. It was HMRC's case that the appellant's trades had not taken place on an open market at all, and that there was no acquisition fraud nor an open market albeit one 'subsidised' by fraud, but rather that the appellant's transactions were all entirely artificial and generated for the purpose of MTIC fraud. Mr Fletcher's evidence was relevant because it was adduced by HMRC to support their case that the transactions involving the appellant did not involve the sort of trading that would occur on an open market.

386. That was relevant to the question of whether there was fraud: while fraud was admitted on 23 of the deals, it was not admitted on 8 of the deals and was therefore a live issue in the hearing. Knowledge of fraud was denied on all 31 days and whether the deals were artificially generated by a fraudster was clearly potentially relevant to the question of knowledge.

Expertise

387. Mr Fletcher worked for OEMs (original equipment manufacturers) and large accountancy firms where his clients were OEMs. He had never been employed by a grey market mobile phone trader. Nevertheless his work while at Orange included advice to his employer on how to avoid box breaking and dumping of Orange's products.

388. We were very satisfied by seeing his CV and hearing Mr Fletcher speak that he was an expert by experience on the grey market even though never been employed by a grey market trader.

389. He admits he consulted with colleagues on aspects of the report but we were satisfied report was his and he was qualified to give the expert analysis.

Independence

390. KPMG still acts for Nokia, and co-authored a report with the anti grey market alliance about how manufacturers could avoid box breaking and dumping. The anti grey market alliance does not include mobile phone manufacturers in its membership (other than Samsung a division of which is a member but not its mobile phone division).

391. Mr Bridge also drew to the Tribunal's attention a statement by Mr Fletcher in the case of Purser [2011] UKFTT 860 (TC) that he perceived a conflict of interest if he were to give evidence on behalf of the appellant.

392. Mr Fletcher explained to us that in that case he had not understood the legal term 'conflict of interest'. He had meant merely that as KPMG had a contract with HMRC he would, as a matter of courtesy have checked with HMRC before agreeing to act as an expert for an appellant in an MTIC case. He had since discussed the matter with HMRC who had said that they would not object to him giving evidence on behalf of an appellant. Mr Fletcher also said that in any event his evidence would be the same irrespective of who was his client.

393. We were satisfied that neither KPMG nor Mr Fletcher had any interest in the outcome of this appeal. There was no suggestion that Nokia could have an interest in the outcome of the appeal.

394. We considered that Mr Fletcher was independent and could properly give expert evidence in this case. We permitted him to give expert evidence. Less weight would be placed on any part of the evidence the source of which was kept confidential.

395. This document contains full findings of fact and reasons for the decision. Any 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 application must be received by this Tribunal not later than 56 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.

BARBARA MOSEDALE

	ELECTRICAL ENVIRONMENTAL SERVICES LIMITED	Appellant

	- and -

	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	First-tier Tribunal (Tax)
You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> Electrical Environmental Services Ltd v Revenue & Customs [2014] UKFTT 129 (TC) (29 January 2014) URL: http://www.bailii.org/uk/cases/UKFTT/TC/2014/TC03269.html Cite as: [2014] UKFTT 129 (TC)

TRIBUNAL:	JUDGE BARBARA MOSEDALE
	MICHAEL SHARP

First-tier Tribunal (Tax)

Outline of case

MTIC fraud