UK: Going Round In Circles

First published in The Tax Journal, Issue 877, 19 March 2007

It has been some four years since the first article on Missing Trader Intra- Community (MTIC) VAT fraud appeared in The Tax Journal (Issue 690, 28 April 2003). The jurisprudence on the subject has, over the past few years, spawned a growing body of case law, including no fewer than three European Court of Justice (ECJ) judgments. However, the case law still leaves much to be revealed about the applicable principles in this area. That is not surprising, since much judicial effort (in the UK) has related to deciding not to decide MTIC or carousel fraud-related cases on their merits. It would be much better for all concerned if the litigation were permitted to go ahead as rapidly and as thoroughly as possible straight away, rather than prolonging the current period of ‘phoney war’, during which the innocent are damaged by not receiving their EU law rights to immediate input tax refund, while the guilty escape exposure in the rigours of the courtroom. Justice delayed is justice denied.

In the broadest terms, MTIC or carousel fraud occurs when a VAT-registered trader in one EU Member State acquires goods at zero-rated prices from a trader in another Member State and on-sells them within his own country at standard-rated prices before disappearing without accounting for the acquisition and sales VAT to the tax authorities. The goods sometimes change hands domestically a number of times until they are finally exported at zero-rated prices by another VAT-registered trader, who reclaims the input tax he suffers. The goods could end up being re-circulated, hence the practice being referred to as carousel fraud. There is a more complex version of the fraud involving ‘contra’ trading, where parallel supply chains relating to different consignments are set up, with VAT only being accounted for on one of the chains.

These are, however, only statements of the broadest kind. It is the precise detail of the fraud or frauds in the case of each chain that is of paramount importance, both factually and legally. For example, is one dealing with a ‘missing’ trader or merely a ‘defaulting’ trader somewhere else in the chain? If defaulting, why has that trader defaulted? What precise kind of chain is it – for example, is it a closed carousel, where goods allegedly circulate round and round, or a looser, more fragmented situation? How do the steps in the chain interact, objectively speaking? In our opinion, the cases have too often taken a broad-brush approach to the actual detail. Too easily has it been somehow assumed that if there is fraud – or indeed mere non-payment of tax – somewhere in the chain, the chain itself is ‘tainted’. The use of labels such as ‘brokers’ and ‘buffers’ have not helped, since they imply that the chain is a single unit, with parties playing pre-arranged roles.

Although it has now been confirmed that tainting does not deprive any of the activities in the chain of their status as economic activities, there is a danger that mere knowledge (actual or constructive) of problems elsewhere in the chain will be treated as an equivalent. This would be wrong. If a trader buys a computer from a shop whose owner says he is going bust tomorrow, and wants to sell cheap today, should the buyer (i) not make the purchase and/or (ii) not reclaim his input tax, because he knows that HMRC might not get all the output tax in the ensuing liquidation? What of the situation where the shop is solvent but says it can offer a good price because it has bought stock cheaply from a trader on the verge of collapse (or who had collapsed) – is the shop’s customer precluded from buying or claiming his input tax because he has reason to believe that others in the chain will not fully discharge their fiscal liabilities? Does he have to ask if the collapse was an honest collapse or a dishonest one, or a combination? How precisely does that alter the trader’s rights under Article 17 of the Sixth Directive (now Articles 167, 173, 176 and 177 of Directive 2006/112/EC, which entered into force on 1 January 2007 – see Article 413 thereof)? These issues lie at the core of the coming debate concerning ‘extended verification’, though they have yet to be brought into the open.

The claims of a large number of taxpayers for the repayment of input tax have been effectively withheld by HMRC pending the completion of an ‘extended verification’ exercise in order to ascertain the extent to which (if any) the relevant claim has been ‘tainted’ by MTIC fraud.

However, this process often takes many months and occasionally in excess of a year, as the investigations typically cover the entire supply chain to which the taxpayer is connected and often include enquiries directed abroad. This places a large number of those taxpayers affected in a precarious financial position and some of them have even been forced to cease trading as a result. Such a measure would appear to be prima facie at odds with the line of ECJ cases that stresses the need for the right to deduct input tax to be exercised immediately – see Case C-62/93 BP Supergas [1995] STC 805; also Case C-439/04 Axel Kittel v Belgian State [2006] ECR I-0000 at paragraphs 47 and 48, applying BP Supergas, paragraph 18, and Joined Cases C-110/ 98 to C-147/98 Gabalfrisa and Others [2000] ECR I-1577, paragraph 43.

The rules governing deduction are meant to relieve the trader entirely of the burden of the VAT payable or paid in the course of all his economic activities.

Whilst it is accepted that HMRC has a right to investigate each claim, its failure to do so reasonably and proportionately could be made the subject of an application for judicial review. The question as to what is reasonable and proportionate, though, turns on the facts of each case – see R (on the application of UK Tradecorp Ltd) v Commissioners of Customs and Excise [2005] STC 138, per Lightman J at paragraphs 18-27. As the Judge in that case held, this nevertheless embraces an obligation, inter alia:

• to keep all investigations under review;

• to complete all investigations within a reasonable period of time;

• to consider whether an interim and/or part payment can and should be made;

• to inform the taxpayer as to the reasons for the payment being withheld;

• to inform the taxpayer as to the stage which the investigations have reached;

• to explain any apparent undue delay in the progress of the investigations;

• to answer sensible and proper questions relating to the investigations; and

• to raise with the taxpayer at an appropriate time matters which emerge in the investigations which call for comment or explanation by the taxpayer or on which the taxpayer may reasonably be expected to have some input and the taxpayer should be allowed to rebut an adverse inference.

But these are all procedural considerations. While they are important, they do not go to the heart of the issue, which is what HMRC should be looking for and why. This turns on a point of law that has yet to be fully addressed in the cases, namely as a matter of legal analysis what is it that actually deprives a trader of his Article 17 right to reclaim input tax on a transaction? The procedural considerations listed above also suffer from a further difficulty in terms of practical application – how does a taxpayer or a Court, neither of which is privy to the necessarily confidential enquiries of HMRC, get the material with which to judge the progress and focus of the investigation? At the end of the day, as long as HMRC can show quantities of sustained activity, that is usually enough to satisfy the Court. Yet it is the quality and focus of that activity which is critical.

In circumstances where a decision has been made by HMRC to deny a taxpayer who has dealt with a supply chain in which there is MTIC fraud his right to deduct input tax, the current judicial view appears to be that this can only be justified if it can be shown that he had actual or constructive knowledge of the fraud at the time he acquired the goods which form the subject-matter of the claim for deduction – see Kittel at paragraphs 56- 57:

‘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 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’ (emphasis added).

In similar vein, it should be noted that VATA 1994, s 77A imposes joint and several liability for the payment of VAT owed by a missing or defaulting trader on such a trader and on the taxpayer who ‘knew or had reasonable grounds to suspect that some or all of the VAT payable in respect of [the] supply [of goods to him], or on any previous or subsequent supply of those goods, would go unpaid’. However, the application of s 77A is expressly confined to the supply of telephones and computers and any equipment for use in connection therewith – see s 77A(1). In August 2003 HMRC published Notice 726: ‘Joint and several liability in the supply of specified goods’, setting out the circumstances in which such joint and several liability would be imposed. Unlike the formulations in Kittel, however, there is no reference to aiding or participating in the fraud. We believe that this is a significant omission.

The question as to the bridge between knowledge and the loss of right of reclaim is, we believe, important. Of course this question does not relate to the fraudsters themselves but to those who are not party to the fraud, yet who seek to do deals which place them in the chain where fraudsters are operating. The answer is particularly important, given the existence of a school of thought that merely to deal in mobile telephones or computer chips is somehow ‘asking for trouble’ in VAT terms.

There are differing views as to whether the test of knowledge is merely focused on what the relevant taxable person knew about the transactions to which he was a contracting party, or whether it is broad enough to encompass the other transactions further along the same supply chain and his knowledge of them. This was made the subject of lengthy analysis in Dragon Futures Ltd 19831. This is an important debate but it masks another issue: namely, whether through his knowledge the trader can be said to participate in the fraud, so as to lose his Article 17 right. We see knowledge as necessary, but not sufficient, for the deprivation of Article 17 rights: full deprivation can only occur when there is a combination of knowledge (whether actual or constructive) and participation.

Dealing first with the question of knowledge, support for a narrower approach can be found in the observations of the ECJ in Joined Cases C-354/03, C-355/03 and C-484/03 Optigen Ltd, Fulcrum Electronics Ltd and Bond House Systems Ltd v Commissioners of Customs and Excise [2006] STC 419, at paragraph 47:

‘As the Advocate-General observed in para 27 of his Opinion, each transaction must therefore be regarded on its own merits and the character of a particular transaction in the chain cannot be altered by earlier or subsequent events.’

See also Kittel at paragraph 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 VAT (see, to that effect, the order of the Court in Case C-395/02 Transport Service [2004] ECR I-1991, paragraph 26). According to the fundamental principle which underlies the common system of VAT, and which follows from Article 2 of the First and Sixth Directives, VAT applies to each transaction by way of production or distribution after deduction of the VAT directly borne by the various cost components (see, inter alia, Case C- 98/98 Midland Bank [2000] ECR I-4177, paragraph 29; Case C-497/01 Zita Modes [2003] ECR I-14393, paragraph 37; and Optigen, paragraph 54).’

However, it has recently been held by the Tribunal in Dragon Futures that the correct approach would be to take the entire supply chain into account. Dr David Williams (Chairman) observed as follows at paragraphs 67 and 71:

‘67. The tribunal has little hesitation in applying the test potentially across the whole of any relevant chain of transactions, and not limiting it to the contract of sale to, and the contract of sale by, the taxable person … It is oversimplistic to look only at the individual contract of sale or purchase. Of course, the European Court made it clear in Optigen that a particular claim to input tax must be based on a particular contract. Entitlement to input tax from that contract cannot be taken away simply by reference to the fact that the contract was one of a chain of transactions affected by fraud elsewhere …

‘71. The tribunal therefore does not accept that there is any specific limit "up" or "down" the chain to the knowledge to be considered by a taxable person …’

The test of knowledge is an objective one (see Kittel at paragraphs 59 and 61). The time at which knowledge is assessed is the point at which the transaction giving rise to the right of deduction of input tax was entered into; hindsight is irrelevant (see Dragon Futures at paragraphs 73 and 74(1)). The test of knowledge extends to knowledge about the entities involved in making the relevant transactions and of the transactions themselves (see Dragon Futures at paragraph 26). The concern requiring investigation is with fraud on the public revenue through the VAT system, not other forms of fraud such as fraud on a foreign trader (see Dragon Futures at paragraph 74(6)).

In circumstances where what is being alleged is constructive knowledge of fraud, it is important to note that an assessment as to what a taxable person ‘should have known’ must be applied within the principles of proportionality and certainty – see Dragon Futures at paragraphs 73, 74(2), 74(3) and 74(5). The Tribunal made, inter alia, the following observations:

• The test can take into account, but cannot be dictated by, guidance offered by HMRC or any other national tax authority.

• The formulation of the test in terms only of ignoring factors that may indicate that there may be fraud is too uncertain. It risks imposing a disproportionate burden on the taxpayer.

• It includes knowledge of fraud ‘in the market’ for the goods in question, as well as knowledge in the public domain or otherwise actually known of fraud by a specific trader. However, there can be no presumption that because there is fraud in a chain of transactions, that fraud is known, or should have been known, to all others in that chain.

• It includes information about all known counterparties in the web of transactions of which the contract forms part, and counterparties that can be identified on proportionate enquiry made within the limits imposed by market confidentiality.

• The taxpayer must take proportionate steps to use all means reasonably available to increase actual knowledge – for example, checks on the validity of VAT registration numbers, checks on customs stamps on goods going through a customs inspection, checks with and about individual suppliers and customers (including checks with national registration institutions), checks with credit agencies and inspection agencies, and the use of appropriate terms of contract.

• Where an initial enquiry gives rise to information suggesting the need for further enquiry, the test is reapplied to assess the need for that further enquiry.

However, as noted above, there is an additional element to the test of knowledge which was not canvassed in Dragon Futures – that the right to deduct input tax can only be vitiated where the material knowledge constitutes or evidences facilitation of, or participation in, the fraud complained of; the ECJ made clear in paragraph 57 of its judgment in Kittel that the denial of the right of deduction was justified on the basis that ‘in such a situation the taxable person aids the perpetrators of the fraud and becomes their accomplice’. An analogy can be drawn with the position in English law regarding the imposition of third-party accessory liability for dishonest assistance in breach of trust or fiduciary duty – the dishonest third party must have actually assisted in the breach of trust or fiduciary duty before liability can be found.

The question to be asked is whether, with the actual or constructive knowledge, the trader then participated in the fraud. Much will turn on what kind of fraud it is. In a closed carousel, all participants in the chain may well facilitate the process, since the intent is to get the goods back to the start. But in a more open-ended chain, or in a situation of contra-trading, facilitation by simply buying and selling at one point along the chain is far less obvious. Thus, in cases where the fraud is ‘upstream’ and not directly proximate in the supply chain to the relevant taxpayer, with no reason to think that it is dependent on subsequent transactions, it is difficult to see how later deals constitute participation. If the fraud has yet to occur and is ‘downstream’ in the supply chain, and somehow there is information to be had about it, it may be clearer that by taking part in the supply chain the taxpayer would or might be facilitating the fraud, even by just passing goods through on a fully commercial basis, taking a proper turn. Is constructive knowledge that one might be facilitating a fraud enough for a trader to lose his Article 17 right? We are doubtful.

Wherever the line is ultimately drawn, we take the view that very precise evidence may be needed of the nature and extent of the fraud before relevant knowledge and participation can be shown.

Applying the test set out in Kittel as discussed above, the three elements necessary to deny a taxpayer’s right to deduct input tax on account of MTIC fraud are:

• there was a fraudulent evasion of VAT with which the transaction giving rise to the entitlement to deduct was causally connected;

• the taxpayer actually or constructively knew about the fraudulent evasion of VAT complained of and the resultant causal connection; and

• by entering into the transaction said to give rise to the input tax claim, the taxpayer thereby participated in or facilitated the fraud.

It is clear that the onus will be on HMRC to demonstrate that the first element – that is, the existence of a fraudulent evasion of VAT – is present. One would have imagined that it logically follows that the burden in proving the existence of the second element – knowledge – would also fall to be discharged by HMRC. However, the Tribunal in Dragon Futures came to the opposite conclusion – see paragraph 85 of the decision. We would respectfully disagree. Whilst it is correct that the burden is generally on the taxpayer to establish his right to deduct input tax, this should not extend beyond showing that an entitlement to do so has arisen under Article 17 of the Sixth Directive. The burden should then shift to HMRC to demonstrate that there is an exception to that entitlement on the basis of the taxpayer’s knowledge of MTIC fraud. To hold otherwise would not only be contrary to the directly-effective right conferred upon the taxpayer by Article 17 but would also leave the taxpayer in the invidious position of being presumed guilty of fraud unless he was able to demonstrate otherwise. Furthermore, it is unacceptable for a taxpayer to have to prove a negative in order to avoid a finding of fraud, especially if it is a fraud by others.

In view of the serious nature of the allegation involved – see, for example, Kittel at paragraphs 56 and 57, where the ECJ describes the taxpayer who possesses the requisite knowledge of the fraudulent evasion of VAT complained of as ‘a participant in that fraud’ and as one who ‘aids the perpetrators of the fraud and becomes their accomplice’ – the standard to which HMRC has to prove its case is not, we suggest, the balance of probabilities but a heightened civil standard – see R (McCann) v Crown Court at Manchester [2003] 1 AC 787, where it was held at paragraph 82 (per Lord Hope of Craighead) as follows:

‘[I]t is not an invariable rule that the lower standard of proof must be applied in civil proceedings. I think that there are good reasons, in the interests of fairness, for applying the higher standard when allegations are made of criminal or quasicriminal conduct which, if proved, would have serious consequences for the person against whom they are made.’

Notwithstanding that the judgment in McCann arose in the context of an application for judicial review in relation to anti-social behaviour orders made under the Crime and Disorder Act 1998, it was found by the Commercial Court in Sphere Drake Insurance Ltd v Euro International Underwriting Ltd [2003] 1 Lloyd’s Insurance and Reinsurance Reports 525 to be applicable in a case where allegations of conspiracy, dishonesty and/or fraud have been made – see paragraphs 106 and 107 of the judgment, per Thomas J (as he then was).

The heightened civil standard was found by the House of Lords in McCann to be virtually indistinguishable from the criminal standard – see paragraph 37 of the judgment (per Lord Steyn).

Where it is established that HMRC has acted wrongfully in denying a taxpayer’s claim for the deduction of input tax, the relief to which the aggrieved taxpayer is entitled extends beyond the sum which has been withheld and includes a repayment supplement pursuant to VATA 1994, s 79, an order for the payment of interest pursuant to VATA 1994, ss 78 and/ or 84(8), an order for costs and, possibly, an award for damages in respect of the consequential loss caused by HMRC’s unlawful conduct.

It is clear that although there has been considerable judicial exposition on the principles and rules that are to be applied in dealing with a MTIC fraud case, the exact contours of liability have yet to be worked out precisely and will no doubt be the subject of vigorous debate, both in and outside the courtroom.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.