Field Opportunities Ltd v Revenue And Customs (VALUE ADDED TAX – denial of relief for input tax) [2019] UKFTT 0531 (TC)

FTT-Tax

Field Opportunities Ltd v Revenue And Customs (VALUE ADDED TAX – denial of relief for input tax)

[2019] UKFTT 0531 (TC) · 2019-06-13

[50]In Citibank , Sir Geoffrey Vos C said as follows: [85] The key point, in my judgment, is that, whilst HMRC can, of course, allege that a taxpayer has acted dishonestly and fraudulently in relation to the transactions to which it was a party, they do not need to do so in order to deny that taxpayer the right to reclaim input tax under the Kittel test. The exercise upon which Judge Mosedale was engaged was, therefore, inappropriate. It was simply irrelevant for the FTT to ask whether the allegations in the Statement of Case, if all proved, would necessarily lead to the conclusion that the taxpayer had been dishonest or fraudulent. It was even more inappropriate for Judge Mosedale to direct HMRC to plead dishonesty when it had expressly informed her that it did not wish to make any such allegation. It might be, of course, that if some or all of the allegations made in the Statement of Case were proved, that might (in theory, though not, of course, in practice) have allowed a tribunal to go on to make a finding that the taxpayer had been dishonest. But if HMRC does not seek such a finding, and if such a finding is not needed to support the conclusion that the taxpayer cannot recover its input tax, there is neither any need nor any utility in asking the FTT to undertake that exercise.

[86][86] The UT's concern was that, because the allegations made in the Statement of Case, might be sufficient to support a plea of dishonesty, further particulars of that plea were required. But as I have sought to show by reference to what Judge Mosedale actually found in the Citibank case, the core pleading in that case did not automatically mean that it was alleged that Citibank behaved dishonestly. As the UT said, such conduct might or might not, depending on a number of factors, be conduct that would be regarded as dishonest.

[87][87] In these circumstances, in my judgment, the UT was wrong to imply that the FTT was justified in undertaking the task of seeking to ascertain from the Statement of Case whether or not the conduct alleged automatically amounted to dishonesty or fraud. Such a process was unnecessary and inappropriate and particularly so when HMRC disavowed an allegation of that character …

[90][90] Finally, if a summary of the applicable law is required along the lines of paragraphs 86 and 87 of the UT's decision, I would simply summarise the principles as follows:-(i) The test promulgated by the CJEU in Kittel was whether the taxpayer knew or should have known that he was taking part in a transaction connected with fraudulent evasion of VAT.

(ii) Ultimately the question in every Kittel case is whether HMRC has established that the test has been met. The test is to be applied in accordance with the guidance given by the Court of Appeal in Mobilx and Fonecomp .

(iii) It is not relevant for the FTT to determine whether the conduct alleged by HMRC might amount to dishonesty or fraud by the taxpayer, unless dishonesty or fraud is expressly alleged by HMRC against the taxpayer. If it is, then that dishonesty or fraud must be pleaded, particularised and proved in the same way as it would have to be in civil proceedings in the High Court.

(iv) In all Kittel cases, HMRC must give properly informative particulars of the allegations of both actual and constructive knowledge by the taxpayer…

[97][97] It will by now be obvious that I agree with HMRC's submission as to the fundamental issue that is required to be resolved by this appeal. HMRC said that the question was whether the UT was wrong to conclude that an allegation that a taxpayer knew that its transactions were part of an orchestrated scheme to defraud HMRC required HMRC to plead and particularise, and therefore to prove, an allegation of dishonesty. I agree that that was the question raised by this appeal. I also agree that the allegation, which is a classic Kittel first limb contention, does not require HMRC to plead, particularise and prove dishonesty or fraud.

[98][98] The main point in this case was not, as the taxpayers suggested a simple pleading question. The UT failed, I think, to identify the basic error that Judge Mosedale had made in the Citibank case, where she said, in effect, that making a first limb Kittel allegation required a plea of dishonesty. It does not; even if in some cases, the findings of knowledge made by the FTT could have led the FTT to uphold a plea of dishonesty had it been made. HMRC is entitled to stop short of alleging dishonesty and content itself with pleading, particularising and proving first limb Kittel knowledge. If, however, HMRC do expressly allege dishonesty, they will be required to comply with the normal rules of pleading and disclosure applicable to such cases. In future, it might be helpful in these cases for HMRC to say expressly in their Statements of Case whether or not they set out to prove the dishonesty of the appellant taxpayer. ” 51. Hallett LJ and Sir Terence Etherton MR both agreed with the conclusions of Sir Geoffrey Vos C – see, in particular, paragraphs [109] and [120] of the decision in Citibank . 52. It follows that we consider that the submission of Ms Stephenson and Mr Carey in relation to this point is correct and that, in any case in which the Respondents plead that the taxpayer in question had actual knowledge of the relevant fraud, the Respondents are entitled to do so without alleging dishonesty or complying with the rules outlined in paragraph 47 above which relate to allegations of dishonesty. PPI LEADS 53. Each of the transactions in respect of which the repayment of VAT input tax has been denied involved the sale, by a company called Multi Level Media Limited (“MLM”) to the Appellant, of data leads in relation to claims for the mis-selling of payment protection insurance (“PPI”) In each case, the PPI leads so acquired were on-sold by the Appellant to a company located in Gibraltar called RAM Media and Advertising Limited (“RAM”) on a back-to-back basis. 54. Before turning to the facts, issues and witness evidence which are involved in this appeal, it is appropriate to set out in a little more detail the background to the market in PPI leads, and the manner in which PPI leads were traded at the time when the transactions to which this appeal relates occurred. 55. At the time when the transactions to which this appeal relates occurred, there was a significant market in the generation and sale of PPI leads as a result of the mis-selling of PPI by a number of financial institutions in earlier years. The PPI leads comprised data in relation to those who might potentially be interested in pursuing claims against the institutions involved. 56. PPI leads would be generated in a number of ways and might then be sold to entities who would use them to generate claims commissions for themselves or to profit from on-sales of the PPI leads in question. The value of a PPI lead at any time would depend on a variety of factors, such as:(1) whether the data related to a person (the “target”) who had previously expressed an interest in making a claim for PPI mis-selling or who had a less direct connection to PPI mis-selling - such as a person with multiple mortgages and/or credit cards – but was still a potentially valuable lead or who was simply on a database unrelated to PPI mis-selling, such as the electoral roll, and was therefore more of a “long-shot”;

(2) whether the target had given permission for his or her data to be used to make cold calls;

(3) whether or not the data had been properly formatted and “cleaned” – by which is meant the removal of, inter alia, obvious errors, incomplete entries and duplicates;

(4) the currency of the data – recent data’s being more valuable than old data;

(5) how many times the data had previously been used – leads that continually flowed around the market became less valuable over time; and

(6) the use which the purchaser was entitled to make of the data – data sold for a single use would be less valuable than the same data sold for multiple uses. BACKGROUND AND COMMON GROUND 57. In connection with this appeal, we have been provided with evidence from a number of witnesses, as well as several files of documents. Whilst there are various areas in which the parties disagree in relation to the salient facts, there are certain matters which are not in dispute, either because they are set out within the documents with which we have been provided or because they are common ground. 58. We think that it is helpful to start our consideration of this appeal by setting out those matters which are not in dispute. Before doing so, we would note that, as the transactions to which this appeal relates were part of an overall course of dealing between the individuals involved in those transactions starting some time before the VAT accounting periods in which those transactions occurred, it is necessary to recount in some detail the background and lead up to the formation of the Appellant and the Appellant’s participation in those transactions. 59. The following describes the background to the present dispute: (1) the Appellant was incorporated on 2 January 2013. Its sole director was, at incorporation and at the time of the transactions to which this appeal relates, Ms Philippa Sutton, the wife of Mr Shoefield. Whilst not technically a director, Mr Shoefield was the controlling mind of the Appellant and therefore, for all practical purposes, the Appellant was his company; (2) Mr Shoefield’s background is in advertising, marketing and media. Prior to the formation of the Appellant, Mr Shoefield had worked for a number of companies, one of which was a company called Smarter Communications, which subsequently changed its name to Truly Advertising Limited (“SC”). As a result of significant bad debts, SC went into administration and the media part of its operations was taken over by a new company called Smithfield and Associates Limited (“Smithfield”). As was subsequently the case with the Appellant, Mr Shoefield was not technically a director of Smithfield but, for all practical purposes, Smithfield was Mr Shoefield’s company; (3) all of the back office work for Smithfield was carried out by companies within the Havas group, a significant participant in the area of media and advertising. One of those companies in the Havas group was Media Planning Limited. Since nothing turns on exactly which member of the Havas group was involved in the activities which are the subject of this decision, we will hereafter refer to each member of the group generically as “Havas”; (4) the activities of Smithfield involved its acting as agent for various clients in relation to advertising campaigns. In that capacity, Smithfield would purchase advertising space in the press or on TV and advise its clients on how to obtain the best return from their marketing budgets. In the course of those activities, Smithfield would occasionally purchase data on behalf of the relevant client. Although the relationship of Smithfield to its clients was described as that of agent and principal, all expenses – including expenditure on data – would technically be incurred by Smithfield as principal and then on-sold to the relevant principal on a back-to-back basis. Smithfield would then earn a commission from its role; (5) in June 2011, Mr Shoefield, along with his colleague, Mr Nick Wood, were introduced to Mr Jason Butler by a mutual business associate, Mr Simon Kevan. Mr Shoefield did not know Mr Kevan particularly well. It was his employee, Mr Wood, who had the stronger connection. As a result of that introduction, Smithfield and Mr Butler’s company, Jump Money Limited (“Jump”), did some business together later in 2011; (6) in around 2008, while he was working for SC, Mr Shoefield had met Mr Neil Wright of RAM. Mr Shoefield believed that Mr Wright would be able to assist Mr Butler with the generation of data leads and accordingly introduced Mr Wright to Mr Butler in June 2011. Mr Wright also did some work with Jump although this did not amount to anything significant;

(7) another company which Mr Butler controlled was MLM. On 12 January 2012, Mr Butler sent an email to Mr Shoefield in which he outlined in brief “ the opportunity that faces MLM in relation to the above data and our objectives in setting up longer term authorised media distributors. MLM is currently producing PPI data for use by large scale call centres who are cashing in on the current PPI claims culture. They have a massive appetite for customer data and we are currently supplying data batches (500) records to international media buyers based in Gibraltar who then deal directly with various authorised centres .” Mr Butler went on:

“ MLM is restricted in the volume of data it can provide to its buyers by cash-flow. Buyers typically require 30 days credit whilst the supply chain requires a constant feed of cash to procure both data and call centre services. An opportunity exists to become an authorised media distributor via MLM. Distributors will be provided with a license to operate MLM’s bespoke software dealing platform that will enable the purchase of bulk data to be transferred to international buyers within a secure environment. MLM will sell data to the distributor for onward sale to authorised and contracted buyers. Margins across all data channels are variable but typically range from 2% to 6% depending on the product type and the cost of procurement .”

Mr Butler then explained that the sales of most data to the Appellant would be subject to VAT but that, in the case of onward sales of that data to customers abroad “ you will reclaim the duty paid via your vat return ”. He concluded by describing the terms of a specific transaction which gave rise to “ an annualised gross margin per batch of 48% ” and by attaching a draft contract. Mr Shoefield immediately forwarded Mr Butler’s email to the finance director of Havas, Mr Ben Skuse, with a request to discuss it with Mr Skuse “ when you have five mins ”. This was the first time that the proposal which would lead to the transactions to which this appeal relates – ie the sale of PPI leads – was raised and it is clear from Mr Shoefield’s reaction to the proposal that it was Havas (acting through Smithfield as its agent) which would be entering into the transactions described by Mr Butler if the proposal were to be adopted. Moreover, it is also clear from the exchanges of emails on that date and later in January 2012 that Mr Butler and Mr Wright had already agreed to enter into transactions in PPI leads before Smithfield was approached. Mr Shoefield and Mr Wright both confirmed the same in their respective oral testimonies at the hearing; (8) on 18 January 2012, Mr Shoefield wrote to Mr Butler to say that he would be reviewing the proposal with “ my CFO ” – by which he meant Mr Skuse – the next day and would then revert. Mr Butler replied later that day to say that “ [the] commercials are negotiable and subject to contract. We have orders coming out of our ears at the minute so distribution channels are a key focus ” and Mr Shoefield responded by enquiring of Mr Butler as to the total available weekly volume and payment terms. Mr Butler replied to that as follows:

“ The data is loaded in batches of 500. We have orders outstanding at the moment of 40 batches which need fulfilment before the end of Jan. Lead point Uk have just contracted to do £100k per month and I am working with 3 other possibles. It depends who makes decisions first but I suspect that we can accommodate whatever volume you decide to fulfil. In terms of credit I am happy to discuss and accommodate wherever I can but ultimately you will be required to give Rams a minimum of 14 days .”

Again, Mr Shoefield immediately forwarded the above exchange to Mr Skuse; (9) on 19 January 2012, Mr Skuse sent an email to Mr Shoefield in the following terms:

“ Looks like we can do it, but our VAT boys want to know that we are doing some level of mark up (evidence of re-work of the data) on the trade, otherwise its something that could cause suspicion. So can you confirm the value of the buy and sell side for me or a data bundle so that we know its going to work – then lets try for the first package next week :)” ; (10) on 23 January 2012, Mr Butler sent an email to Mr Shoefield to the effect that MLM needed “ to supply approximately 10 batches to Neil [ie Mr Wright] which are currently not allocated to a distributor before the month end. Please let me know ASAP if you are likely to be in a position to move forward ”.”

Mr Shoefield promptly forwarded that email to Mr Skuse, saying:

“ Please see below; Straight buy and sell with a 4% mark up. Looking to see what can be done with the data – although not keen to take out of secure environment? ”; (11) on 24 January 2012, there was an exchange of emails between Mr Shoefield and Mr Wright in which they each confirmed the details of the contracting parties to the arrangement between them and Mr Wright sent Mr Shoefield a draft contract; (12) on 25 January 2012, there was a further exchange of emails between Mr Shoefield and Mr Skuse in which Mr Shoefield forwarded an email from Mr Damian Guy of MLM describing the initial purchases of PPI leads by RAM and asked Mr Skuse to prepare invoices showing that the PPI leads in question had been bought by Havas for a price of £24.03 per PPI lead with payment terms of 7 days and sold by Havas on the same day for a price of £25 per PPI lead with payment terms of 14 days; (13) on 1 February 2012, Mr Wright sent an email to Mr Shoefield saying: “ I am trying to feed a large demand and want to know if there is a possibility of you supplying more batches in the very near future?”

. Mr Shoefield promptly forwarded that email to Mr Skuse, with the words:

“ Fyi He is 40-60 batches short a month!”

When Mr Skuse responded by saying that Havas’s maximum credit exposure on RAM at any time was £75,000, Mr Shoefield asked:

“ is it worth trying to get his credit limit up? He is more than happy to chat about this and share any information we may want? ”

Mr Skuse responded to that email by saying that he had put in an application for more credit; (14) on 4 April 2012, there was an exchange of emails between Mr Shoefield and Ms Sunita Kaur of Havas which referred to a potential further transaction with MLM and RAM in respect of which the credit exposure of Havas to RAM would be £115,000. In the course of that exchange, Mr Shoefield made reference to the fact that Havas had insurance cover against a default by RAM of up to £100,000 and an indemnity from MLM of up to £250,000. Ms Kaur forwarded the exchange to Mr Skuse, saying that Havas’s credit limit in respect of RAM was in fact £75,000 and not £100,000 and asking Mr Skuse for general guidance on how to deal with similar requests from Mr Shoefield in the future. Mr Skuse’s response to Ms Kaur was that, as Havas had an indemnity from MLM, Havas would be sufficiently protected as long as it did not buy more than £250,000 of PPI leads at any time and RAM stuck to its terms of payment; He concluded:

“ …- at the slightest sign of delay – call adam ”; (15) the MLM indemnity to which reference was made in the above exchanges was an undertaking by MLM to compensate Havas in the event that the clients to which Havas sold PPI leads acquired from MLM defaulted in their payment obligations to Havas. The indemnity was expressed not as a payment obligation but as an obligation to issue a credit note to the value of the amount in default. The only such indemnity which was in the documents bundle for the hearing was one dated 27 July 2012 in the amount of £400,000. This differs from the £250,000 indemnity limit to which reference was made in the exchanges described in paragraph 59(14) above and it therefore seems likely that there was an earlier indemnity from MLM for the lower amount; (16) on 13 May 2012, Mr Kevin Ferguson of MLM wrote to Mr Shoefield to explain that there was a new software platform to be used in relation to the sale of the PPI leads and that Data Tech Support Limited (formerly Jump Media Limited), another company controlled by Mr Butler and the operator of the new software platform (“Data Tech”) would be calling to explain to Mr Shoefield how the system worked. In that email, Mr Ferguson informed Mr Shoefield that an outstanding amount of £330,291.36 which was due from Havas to MLM should be paid to a bank account in Cyprus of a company called International Media Distribution Limited (“IMD”) “ c/o MLM Ltd ”

. Mr Shoefield, in turn, forwarded that email to Mr David Harvey of Havas and asked him to effect the payment. A further payment to the same destination – this time in the amount of £297,000 - was made at the behest of Mr Shoefield on 9 May 2013. It is common ground that IMD was another company controlled by Mr Butler. It is also common ground that a number of payments were made by both Havas and the Appellant either to IMD or to an account in the name of MLM at a bank in Cyprus although no payments in the former category were made by the Appellant after the Respondents visited the Appellant on 9 July 2013, as described in paragraph 59(26) below; (17) on 15 June 2013, there was a further email from Mr Wright to Mr Shoefield, referring to the fact that a payment of £324,000 had been made to Havas that morning and asking Mr Shoefield to “ supply as many leads as you can as soon as possible ”. That email was forwarded by Mr Shoefield to Ms Kaur along with the details of the attendant invoicing, from which it is apparent that the price of each PPI lead sold by MLM to Havas was at that stage 26.21 and the price of each PPI lead sold by Havas to RAM was at that stage £27.00. Again, Mr Shoefield instructed Ms Kaur to pay the Cyprus account; (18) on 25 July 2012, Mr Shoefield sent an email to Mr Ritchie Marshall, who had by then replaced Mr Skuse as the finance director of Havas. In that email, Mr Shoefield referred to a meeting earlier that day with Mr Marshall and explained that he was expecting the trade in PPI leads to amount to £1m per week for approximately 7 weeks until the volume of PPI trades began to subside. He again referred to a Havas credit limit on RAM of £100,000 and an indemnity from MLM to Havas of £250,000 – which meant that Havas could take an aggregate credit exposure to RAM of £350,000 - and to a request from RAM to increase that aggregate credit exposure to £500,000 “ to facilitate this last burst of activity ”. Mr Shoefield said that he had already agreed with MLM that MLM would increase its indemnity “ in line with this increase ” – presumably referring to the fact that, by virtue of the document mentioned in paragraph 59(15) above, the MLM indemnity amount was going to be increased by £150,000 to £400,000 (although the response to this email from Mr Marshall of 1 November 2012 stated, apparently erroneously, that the MLM indemnity amount at that stage was £350,000); (19) on 30 July 2012, a further transaction in relation to PPI leads occurred. Havas acquired 15,491 PPI leads from MLM at £26.21 per PPI lead and on-sold the same number of PPI leads to RAM at £27.00 per lead. An email from Mr Harvey to Ms Tina Richmond of Havas reveals that the payment to MLM in this case was split 50/50 between MLM’s UK account at NatWest and the IMD account in Cyprus; (20) although sales of PPI leads from MLM to Havas and then from Havas to RAM continued into 2013 – indeed, during the period covered by the Appellant’s VAT accounting periods 07/13 and 08/13, the value of the sales of PPI leads made by MLM to Havas were almost exactly the same as the value of the sales of PPI leads made by MLM to the Appellant - it was clear that the demand for PPI leads from RAM and the availability of PPI leads from MLM were both greater than Havas’s appetite for those trades. Accordingly, Mr Shoefield decided to incorporate the Appellant on 2 January 2013 so that the Appellant could perform the same role as Havas in the transaction chain and thereby increase the overall volume of trades from which he, through his involvement in Smithfield and the Appellant, could benefit; (21) in the Form VAT1 which the Appellant submitted to the Respondents on 29 January 2013, the Appellant estimated that its turnover over the ensuing 12 months would be £5m and checked the box which indicated that VAT repayments were not expected. (The latter was incorrect, as the letter referred to in paragraph 59(23) below, from Throgmorton, the Appellant’s accountants, subsequently revealed); (22) on 30 January 2013, the Respondents registered the Appellant for VAT purposes with effect from 2 January 2013. Initially, the Appellant was given quarterly VAT accounting periods but, with effect from 1 July 2013, it moved onto monthly VAT accounting periods. The reason for that was to reduce the cash flow cost for the Appellant of funding the VAT input tax on its purchases from MLM and its recovery of the credit in respect of that tax from the Respondents; (23) on 18 February 2013, Throgmorton wrote to the Respondents in response to a request for further information about the Appellant made on 11 February 2013. In its letter, Throgmorton informed the Respondents that: (a) the Appellant’s activities involved the purchase of PPI leads from MLM and the on-sale of the same PPI leads to RAM; (b) those transactions were effected on a platform from which PPI leads could be downloaded; and (c) as the sales of PPI leads were being made to RAM, a company belonging in Gibraltar for VAT purposes, those sales would be outside the scope of VAT and therefore the Appellant would generally be in a VAT repayment position; (24) Throgmorton enclosed with its letter what it described as a copy of the contract between the Appellant and MLM (the “First Contract”) and a copy of the contract between the Appellant and RAM (the “Second Contract”). A number of points should be noted in relation to the contracts, as follows: (a) neither of the contracts was signed; (b) the contracts bore different dates (and, in the case of the Second Contract, the date was only on the cover sheet of the contract and not in the operative part of the document); (c) clause 6.2 of the Second Contract referred to RAM’s indemnifying “Smithfield” (instead of the Appellant, as it should have done) from any claim resulting from any use of the data which was not in accordance with law; (d) clause 4.2 of each contract provided for the provision of replacement PPI leads as the sole remedy in any case where the “ non-contact rate ” and the “ wrong product/not interested rate ” exceeded 15%. Neither contract envisaged any other remedy in such cases. In particular, no provision was made for the relevant remedy to take the form of a refund/credit note instead of replacement PPI leads; (e) although the terms of the contracts were largely the same, there were some meaningful differences; (f) first, the First Contract had a minimum initial term of 3 months and provided for additional one month renewals, whereas the Second Contract had a minimum initial term of 6 months and provided for additional one month renewals; (g) secondly, the First Contract set out a minimum purchase obligation of 5,000 PPI lead data packets per month in batches of 500 PPI leads at £33.00 per PPI lead although the price per PPI lead was subject to variation from time to time at the agreement of the parties, whereas the Second Contract had no minimum purchase amount and did not specify a price per PPI lead – it simply said that the price per PPI lead was subject to the agreement of the parties from time to time; (h) thirdly, the First Contract stated that the Appellant would be billed monthly in advance, that payment was due on delivery of the invoice and that no credit would be provided by MLM to the Appellant, whereas the Second Contract stated that RAM would be billed on a daily basis for purchases effected during any particular period and that payment was due within 14 days of the invoice date; (i) finally, there was a difference between the language used in clause 4.2 of each contract – dealing with the remedy to the relevant purchaser if the “ non-contact rate ” and the “ wrong product/not interested rate ” exceeded 15%. The relevant language in the Second Contract was ambiguous in that it could be read as requiring that, if the “ non-contact rate ” and the “ wrong product/not interested rate ” exceeded 15%, then all of the relevant PPI leads - and not simply those “ non-contact ” or “ wrong product/not interested ” PPI leads in excess of the 15% - would need to be replaced; (25) Throgmorton also enclosed with its letter the invoices in relation to the sales of PPI leads which it said had been made pursuant to the contracts referred to above; (26) On 9 July 2013, Ms Camilla Kathel and Ms Shamsun Ahmed, Officers in the section of the Respondents which dealt with Missing Trader Intra-Community (or “MTIC”) Frauds, visited the Appellant and spoke to Mr Shoefield. Ms Kathel prepared a report of that meeting but it was not shown to Mr Shoefield and he was not asked to sign it. There is a dispute between the parties as to exactly what was said at that meeting and we will return to that subject when we summarise the witness evidence below; (27) Later on 9 July 2013, Mr Shoefield sent an email to the two Officers attaching a summary of all of the transactions which the Appellant had implemented in the period from 1 March 2013 to 30 June 2013; (28) On 12 July 2013, Ms Kathel sent an email to Mr Shoefield to which she attached VAT Notice 726 and in which she said the following:

“ Thank you for your co-operation and for your swift reply. I have today finalised my report and you will be hearing from the Officer dealing with your repayment in due course. Please read and put into practice advice in this Notice 726, regarding the checks and due diligence you should be carrying out on every new and existing customer and new and existing supplier, in order to protect you and your business from being involved in MTIC Fraud as discussed at the end of the visit.”

The VAT Notice 726 which was sent to Mr Shoefield stated, in paragraphs 1.1, 1.3 and 1.4, that it applied to supplies of certain specified goods. Supplies of data, such as PPI leads, were not mentioned in those paragraphs. The notice then went on in section 6 to set out various checks which a trader should make in order to satisfy itself that its transactions were not connected with fraud. Section 6 of VAT Notice 726 is set out in full in the Appendix to this decision; (29) on the same day, Mr Michael Everett, an Officer in the Specialist Investigations Section of the Respondents, sent a letter to the Appellant in which he alerted the Appellant to the fact that MTIC fraudsters might be attempting to use alternative banking platforms (or ABPs) to facilitate or launder the proceeds of the fraud. In that letter, he explained that, by ABPs, he meant “ a kind of virtual financial institution that provides the functionality of a traditional bank without a bank’s reporting or regulatory requirements ”. The letter continued:

“ We strongly advise you that you do not ignore this warning. If you fail to carry out proper ‘know your customer’ checks (KYC) you may be putting your business at risk. It is for you as a business to demonstrate that all conditions for a taxable supply are met. You must be able to demonstrate that a payment made is for a particular supply from a particular person. If you are not able to establish the audit trail for the supply and consideration you may be denied input tax .”

The letter then outlined what it described as “ the actual characteristics of an ABP used to further fraud ”, which included the fact that the ABP would generally, but not exclusively, be held at a financial institution outside the jurisdiction of the supplier, there would be no face to face contact with the ABP personnel, fees might be abnormally high and counterparties might be required to use the same ABP. The annex to the letter then provided a summary of the characteristics of MTIC fraud. The list of factors set out in the annex were similar to a number of those set out in VAT Notice 726. However, unlike VAT Notice 726, which stated expressly in its terms that it related solely to the purchase and sale of certain specified goods – in paragraphs 1.1, 1.3 and 1.4 of the notice - the annex, having outlined the categories of goods to which MTIC frauds historically related, went on to state that “ [since] its inception MTIC fraud has mutated and moved into intangibles and services in addition to the type of commodities mentioned above ”; (30) on 7 August 2013, Ms Rachel Austin, an Officer in the Specialist Investigations Section of the Respondents, sent a letter to the Appellant in relation to the Appellant’s return for the VAT accounting period 07/13. In that letter, Ms Austin informed the Appellant that, as part of the Respondents’ efforts to combat MTIC fraud, certain returns were being targeted for verification and that the Appellant’s return for that VAT accounting period was one of them; (31) on 8 August 2013, Ms Kathel sent an email to Mr Shoefield to ask for additional information about the relevant VAT accounting period in order to expedite the verification process and to arrange a further visit to the Appellant’s offices; (32) later on the same day, Mr Shoefield responded to suggest a date for the further visit and Mr Michael Davey of Throgmorton sent the requested information to Ms Kathel; (33) on 12 August 2013, Ms Kathel and Ms Kemi Aina, another Officer of the Respondents, visited the Appellant. Once again, Ms Kathel prepared a report of that meeting but it was not shown to Mr Shoefield and he was not asked to sign it; (34) later on the same day, Mr Shoefield sent to Ms Kathel the banking information which she had requested at the meeting; (35) on 22 August 2013, Ms Kathel sent two emails to Mr Davey – the first to update him on the position in relation to the 07/13 VAT repayment (in response to Mr Davey’s chasing email of the previous day and Mr Shoefield’s chasing email of 16 August 2013) and the second to ask him to send her a copy of the Appellant’s Information Commissioner’s Office (“ICO”) registration certificate; (36) later on the same day, Mr Shoefield sent an email to Ms Kathel attaching an ICO registration form; (37) on 23 August 2013, Ms Kathel sent an email to Mr Shoefield containing further questions in relation to the Appellant’s transactions. The first asked for details about the call centre which Mr Shoefield had mentioned in the meeting of 12 August 2013, the second asked for further information about DataTech and the third asked for further information about Mr Wright; (38) later on the same day, Mr Shoefield sent an email to Ms Kathel setting out his responses to the questions raised. In response to Ms Kathel’s first question - about the call centre - he replied as follows:

“ All of my data is bought from a company called Multi Level Media (as per the invoices supplied). Although I am aware of how the data is manufactured I am not involved in this process. Once the data is sourced it is then uploaded onto the Data Tech platform which allows the data to be handled and transferred in a secure manner. The Data Tech platform allows me to request data from my suppliers and then offer it on to sell to my buyer RAM .”

In response to Ms Kathel’s second question – about Data Tech – he replied as follows:

“ Multi Level Media was set up by an existing client of mine – Jason Butler. He was a board director at a business called Jump Money. I made a television commercial for him and booked a TV campaign for him. I also produced various pieces of marketing material. I introduced him to Neil at RAM advertising. During this time we had credit insurance in place, we also had to run a compliance check for the TV campaign, his company had a ministry of justice licence. In January 2012 I was asked to handle a data buying process between a company owned by Jason – Multi Level Media & RAM Advertising. Due to the sums of money involved this process was handled by [Havas] on my behalf. Due diligence was undertaken by company lawyers with regard to the VAT status of these transactions. Contracts were in place – you have copies of my corresponding contracts. Data Tech is the platform used to transfer and adminitrate these purchases. Its role is as a software provider for the process it is not actually taking part in the transaction… My current contact for Multi Level Media & Data Tech is Kevin Ferguson – he is the current Managing Director of Multi Level Media… I speak to Kevin 2 or three times a week. My accountants carried due diligence for VAT purposes the above businesses – they are both deemed to be active VAT/companies. I regularly speak to Jason Butler…”

In response to Ms Kathel’s third question – about Mr Wright – he replied as follows:

“ I have been working with Neil Wright since 2006. He was a client and worked on behalf of Insure & Go. I have been trading with him for this period. He is subject to credit insurance and I regularly receive upto date accounts for RAM to maintain their credit limit…. I have had my lawyers review all of the contracts for the above and my accountants advised me on the structure of the business, setting up and all ongoing vat/accounting for the business .”

He then offered Ms Kathel the opportunity to speak to Mr Ferguson and/or Mr Wright and confirmed to her that MLM had recently been through a VAT review; (39) on the same day, between the two emails referred to above, Mr Shoefield sent an email to Ms Kathel attaching confirmation from the ICO of his application for registration on the previous day; (40) on 28 August 2013, there was a further exchange of emails between Mr Shoefield and Ms Kathel in relation to the VAT repayment in which Mr Shoefield asked for an update and Ms Kathel said that she was hoping to conclude her review by lunchtime on the following day; (41) on 4 September 2013, Mr Shoefield sent an email to Ms Kathel attaching the Appellant’s ICO registration certificate of the same date; (42) on 11 September 2013, Ms Austin sent a “ selection for verification ” letter to the Appellant in relation to the Appellant’s VAT accounting period 08/13 in the same form as the letter described in paragraph 59(30) above; (43) on 13 September 2013, there was a further exchange of emails between Mr Shoefield and Ms Kathel in relation to the VAT repayment, in the course of which: (a) Ms Kathel requested similar information in relation to the Appellant’s VAT accounting period 08/13 as the Appellant had previously provided in relation to its VAT accounting period 07/13; (b) Mr Shoefield provided the requested information; and (c) Mr Shoefield expressed his concerns about the impact which the continuing delay in repayment was having on the Appellant’s working capital; (44) on 5 December 2013, there was a further exchange of emails between Mr Shoefield and Ms Kathel in which Ms Kathel informed Mr Shoefield that the VAT repayments for 07/13 and 08/13 “ require some further discussion ”; (45) on 6 December 2013, in response to that exchange, Mr Shoefield sent an email to Ms Kathel attaching the download certificates for the two VAT accounting periods in respect of which the VAT repayments remained outstanding; (46) on 10 March 2014, Mr Shoefield sent an email to Ms Kathel: (a) attaching a copy of the encrypted data to which the transactions that are the subject of this appeal relate; (b) advising Ms Kathel that he had requested from RAM permission from the end user of the data to give the Respondents access to the data; and (c) offering to set up and host a meeting between the Respondents’ technical team and DataTech if the Respondents wished to know more about the data. Attached to that email was an email from Mr Ferguson to Mr Shoefield attaching the encrypted data, in which Mr Ferguson informed Mr Shoefield that MLM and Data Tech had been given a clean bill of health by the Respondents following a recent visit; (47) Ms Kathel’s response to that email was not in the documents bundle for the hearing but, on 28 March 2014, Mr Shoefield sent an email to Ms Kathel attaching the de-encrypted data and saying that he had now “ sourced the data as requested ”, had signed a non-disclosure agreement and was taking Ms Kathel’s email to be confirmation that the Respondents would also accept to be bound by the same confidentiality obligation. When that email bounced back, Mr Shoefield re-sent the data as a mega bitfile later on the same day; (48) On 31 March 2014, Ms Kathel sent an email to Mr Shoefield to say that she could not open the file because of the Respondents’ IT system and asking Mr Shoefield to send the data by some other method; (49) on 2 April 2014, Mr Shoefield sent an email to Ms Kathel to confirm that “ I have burnt the data onto a disc ” and that he had dropped the disc off at the Respondents’ office the previous day. Ms Kathel acknowledged receipt of the data by an email to Mr Shoefield of the same day. On 10 April 2014, there was a further email exchange between Mr Shoefield and Ms Kathel in which Mr Shoefield confirmed, in response to Ms Kathel’s question, that the disc could be destroyed by the Respondents as it did not need to be returned; (50) on 25 September 2014, following earlier emails of 1 September 2014 and 10 September 2014 in which he had asked for an update, Mr Shoefield sent an email to Ms Kathel to express his frustration at the continuing delay in processing the relevant VAT repayments and referring to the fact that the Appellant’s business was simple in that it had only one supplier and one customer and that its supplier had itself only nine suppliers including companies like Experian; (51) following a response from Ms Kathel by email on 26 September 2014 – in which Ms Kathel requested further information about the nine suppliers to MLM, Mr Davey wrote to Ms Kathel on 29 September 2014 to ask for the Respondents’ availability over the next two weeks for a meeting; (52) on 2 October 2014, Ms Kathel responded by email to the effect that “ there is nothing that can be discussed in a meeting that will help complete the enquiries ” and informing Mr Shoefield that she was changing jobs within the Respondents and would be replaced by a colleague; (53) Mr Shoefield continued to press Ms Kathel for the name of her replacement and for an update on the enquiry by emails of 13 October 2014 and 3 November 2014. In response to the second of those emails, Ms Kathel informed Mr Shoefield that her replacement would be Mr Edmond Whelan; (54) on 3 November 2014, Mr Shoefield sent an email to Mr Whelan, with a copy to Ms Kathel, referring to a conversation earlier that day and again expressing his frustration at the continuing delay in the processing of the VAT repayments; (55) on the same day, Mr Shoefield sent a letter to Mr Whelan summarising the position in relation to the VAT repayment claims as he saw it and offering a meeting or conference call to resolve things. In the course of that letter, Mr Shoefield said:

“Our understanding is that the marketing leads are generated by a series of tv, newspaper and magazine advertisements that provide a call centre phone number who screen the enquiries and through the process of screening end up with high quality leads. These leads are uploaded onto Data Tech’s platform and then sold onto various Lead management brokers, one of which is Field Opportunities. The leads are a legitimate product, and we know the data sets we sell provide name, address, contract details and are categorised with details of the nature of the enquiries made with the call centre. As someone who has worked within media and advertising for in excess of 20 years I identified RAM who I have worked with in excess of 8 years, as a potential client for the quality leads and made the introduction to Jason Butler of Multi Level media. It was agreed that Field Opportunities Limited should arrange the sale of date between the two companies”; (56) on 8 January 2015, Mr Shoefield sent an email to Mr Whelan to say that he was aware that Mr Whelan had now been sent further information from MLM and asking whether the Respondents were now in a position to make the VAT repayments. In the absence of any response to that email, Mr Shoefield sent a further chasing email to Mr Whelan on 2 February 2015; (57) on 3 February 2015, Mr Shoefield sent an email to Mr Daniel Outram, an Officer in the Specialist Investigations Section of the Respondents, attaching a copy of his letter to Mr Whelan of 3 November 2014 and complaining that it had not even received an acknowledgement; (58) on 5 February 2015, Mr Outram sent a letter to Mr Shoefield to explain that the verification process was continuing and that the Respondents were not yet in a position to effect the VAT repayments; (59) on 17 February 2015, Mr Shoefield sent an email to Mr Whelan attaching accounts for MLM to 31 October 2013 and a chart demonstrating the Appellant’s business model and offering to set up a meeting between the Respondents and MLM. In the chart, the Appellant showed MLM as having nine suppliers, only two of which were named (Experian and DBS); (60) on 2 March 2015, Mr Jeremy Corbyn MP, Member of Parliament for Islington North, wrote to Mr Outram at the request of Mr Shoefield, expressing his concerns about the length of time which the Respondents’ investigation was taking and suggesting that the Respondents should provide Mr Shoefield with both an explanation for the delay and an anticipated end date; (61) on 16 March 2015, Mr Shoefield sent an email to Messrs Whelan and Outram, referring to an earlier email of 20 February 2015 (which was not in the documents bundle for the hearing) and to Mr Corbyn’s letter of 2 March 2015 and pointing out that neither of those communications had received a response. In that email, Mr Shoefield requested a copy of all information which the Respondents had on the Appellant and him in relation to the outstanding VAT repayments; (62) on 23 March 2015, Mr Outram responded to Mr Shoefield’s email by saying that the request for information had been handed to the team within the Respondents which dealt with information requests and saying that he was hoping that the repayment decision would be made within the next few months; (63) on 27 March 2015, Ms Kathy Ellis, an Officer in the Complaints Section of the Respondents, sent a letter to Mr Shoefield in response to his email of 20 February 2015. In that letter, Ms Ellis explained why the Respondents’ investigation was taking as long as it was and why the Respondents had refused to meet with Mr Shoefield; (64) on 1 May 2015, Mr Outram sent a letter to Mr Corbyn in response to Mr Corbyn’s letter, in which he explained why the investigation was taking so long to complete; (65) on 6 July 2015, the Respondents sent a letter to the Appellant formally notifying the Appellant that they had decided to refuse the Appellant’s claim for repayment. The figures set out in the table in that letter have subsequently been adjusted by the agreement of both parties; (66) on 11 August 2015, the Appellant’s then solicitors – RPC – formally requested an independent review of the Respondents’ decision. In the course of that letter, RPC said as follows: “ In 2013, [the Appellant] identified RAM Advertising (RAM) as a potential client for PPI leads. Mr Shoefield identified the company as he had previously provided advertising services to RAM in his role as a Director of Smithfield. Through Mr Shoefield’s business relationship with Mr Butler, Field subsequently introduced MLM to RAM. Once the parties were happy to proceed, as [the Appellant] had made the introduction, [the Appellant] arranged the sale of data between the two companies. [The Appellant] provided the businesses with an opportunity to an increased volume of leads due to the operational structure of the business. [The Appellant] as a small business was able to operate, manage and process payments more efficiently, which provided an improved range of services to RAM .”; (67) as an appendix to the above letter, RPC attached a chain of correspondence from January 2012 which included the emails described in paragraphs 59(7) to 59(9) above; and (68) in 2018, Mr Butler was tried and convicted in Leeds Crown Court of an offence of cheating the Revenue. On 26 March 2018, he was sentenced to nine years’ imprisonment. Mr Ferguson was charged along with Mr Butler but was acquitted. 60. The following general statements are also common ground: (1) a considerable number of the purchase invoices in relation to PPI leads purchased by MLM and in respect of which MLM made VAT input tax claims were false. Mr Butler was the controlling mind of the fraud; (2) the Respondents have made the VAT repayments which were claimed by the Appellant in respect of each of its VAT accounting periods ending prior to the VAT accounting period 07/13; (3) the Respondents have not sought to challenge the VAT input tax credits claimed by Havas in respect of its purchases of PPI leads from MLM; (4) when the Appellant bought and sold PPI leads, it did so through the Data Tech platform and, by virtue of the way in which that platform operated, it had no access to the de-encrypted data. Instead, the data was encrypted when it was uploaded to the platform by MLM before the sale to the Appellant, was then sold in encrypted form within the platform by MLM to the Appellant and by the Appellant to RAM and then de-encrypted by RAM before RAM sold the data on to its customers. Users of the system would be given access to specific functions – for instance, not everyone was able to upload data onto, or download data from, the system. A person might simply be able to request to make purchases and/or sales through the system; (5) despite the fact that the only remedy which was stipulated in the First Contract and the Second Contract for a “ non-contact rate ” and a “ wrong product/not interested rate ” in excess of 15% was the provision of replacement PPI leads, the remedy which was generally adopted by the parties in each case was for the relevant seller to provide the relevant purchaser with a credit note in respect of the amount paid for those PPI leads; and (6) the Appellant: (a) never applied for authorisation from the Ministry of Justice (the “MOJ”) to provide claims management services; and (b) did not register with the ICO until 4 September 2013, following the request from the Respondents on 22 August 2013 to see the Appellant’s ICO registration certificate. THE ISSUES INVOLVED 61. Before moving on to consider the witness evidence, we think that it would be helpful to summarise the questions which are at issue in this appeal. As we noted at the start of this decision, this appeal relates to the denial by the Respondents of VAT repayments which have been claimed by the Appellant in respect of the VAT input tax incurred by the Appellant on certain purchases of PPI leads from MLM. As those PPI leads were on-sold to RAM, a company belonging outside the UK, the Appellant would be entitled to a credit for the VAT input tax in question – and hence to the VAT repayments - unless: (1) there has been a loss of VAT; (2) the loss of VAT has been caused by fraud; (3) the relevant purchases were connected with that fraud; and (4) the Appellant knew or should have known that the relevant purchases were connected with that fraud. 62. It is common ground that the first three of the above conditions are met in this case. The Appellant does not dispute that its purchases from MLM were connected with a fraud which gave rise to a loss of VAT. However, the Appellant does dispute that it knew or should have known that the relevant purchases were connected with that fraud. 63. It is for the Respondents to prove, on the balance of probabilities, that the Appellant’s submissions are wrong and that it did have such actual knowledge or means of knowledge– see Mobilx at paragraphs [81] and [82]. THE WITNESS EVIDENCE 64. The Respondents called four witnesses to provide oral evidence – Ms Elaine Emery and Ms Kathel, both Officers of the Respondents, Ms Jessie Wilton, a Senior Claims Manager at the MOJ, and Mr Paul Cresswell, Head of Data Governance at Experian Marketing Services. For its part, the Appellant called two witnesses to provide oral evidence – Mr Shoefield and Mr Wright. 65. In addition, we were provided with two witness statements from Mr Whelan, a witness statement from Mr Andrew Chisman, another Officer of the Respondents, and a witness statement from Mr Michael Cooke, a Lead Case Officer at the ICO but none of those people attended the hearing to provide oral testimony. The documents bundle also contained certain other witness statements from witnesses in the criminal trial of Messrs Butler and Ferguson – for example, there were such statements from Mr Skuse and Mr Wright, along with statements from a Mr Michael Chung of Marketing Lists Limited, which sold data to MLM, and Mr David Hedges, who developed the Data Tech platform. 66. Much of the witness evidence does no more than describe the agreed facts set out above. Accordingly, we will confine this summary of the witness evidence to those points which are the subject of dispute between the parties or which add context and colour to the description of those agreed facts. 67. The key points arising from Ms Emery’s evidence were as follows: (1) Ms Emery had been working predominantly with businesses involved in MTIC fraud since 1999 and became the case officer in relation to the Appellant’s appeal on 11 May 2017. She then took over Mr Whelan’s role in relation to the appeal on 15 May 2018, when Mr Whelan retired; (2) she had reviewed Mr Whelan’s decision to deny the VAT repayments in question and agreed with it. In particular, she considered that Mr Shoefield should have realised from the circumstances of his involvement in the overall proposal that something fraudulent was going on. However, she conceded that: (a) whilst she had considerable experience in MTIC fraud and in the manner in which the commercial world in general operated, she did not have a good understanding of how the market in data leads worked specifically; (b) there was a significant market in PPI leads at the time of the transactions which were the subject of this appeal; (c) she might have made greater efforts to look into the transactions in PPI leads which had been made through Havas than she actually had done; (d) if Mr Shoefield had been given access to MLM’s PPI lead purchase ledger, which included the forged invoices, there was nothing in that ledger which would have aroused his suspicions; (e) payments made to a third party at the behest of a supplier would not always indicate fraud (but Ms Emery added that such payments could well indicate that fraud was involved and that the existence of such payments should lead to further investigation and questions); (f) the Respondents had become aware that MLM was forging invoices as early as November 2012, when they visited Phruitt Limited, one of the companies from which MLM had claimed to have made purchases, and the visit to Mr Shoefield on 9 July 2013 had been prompted by the Officers within the Respondents who were investigating MLM. However, the investigation into MLM was still ongoing at the time when the Respondents visited Mr Shoefield in July and August 2013, which explained why Mr Shoefield had not been informed at that time that there was VAT fraud in the Appellant’s supply chain; (g) as a strict technical matter, VAT Notice 726 did not apply to the sale of PPI leads, as it was expressed to be limited to specific supplies of goods (but Ms Emery added that the context in which the notice had been sent to Mr Shoefield should have put Mr Shoefield on notice that the questions set out in section 6 of the notice should be considered by him); and (h) she could not point to any particular line of enquiry which Mr Shoefield could have adopted which would have led him to discover the fraud; and (3) in Ms Emery’s view, the emails and SMS messages between the various parties involved in the sales of PPI leads - Mr Shoefield, Mr Wright, Mr Ferguson and Mr Butler – included messages which showed that they were all aware of each other’s roles in the overall fraud and confirmed their closeness because those messages extended beyond communications on purely business matters to enquiries about family and social arrangements. 68. The key points arising from Ms Kathel’s evidence were as follows: (1) Ms Kathel worked in the MTIC Fraud Section of the Respondents from the beginning of May 2006 to the end of October 2014. In that capacity, she had been allocated the Appellant from 8 August 2013; (2) she had had no direct involvement in the processing or approval of the VAT repayments which were made to the Appellant in respect of its VAT accounting periods ending prior to the 07/13 VAT accounting period. However, the officer who had conducted those tasks, a Mr Steven Harms, had suggested that she visit the Appellant because of concerns about fraud in the Appellant’s supply chain; (3) she could not recall the precise date when the reports of her meetings on each of 9 July 2013 and 12 August 2013 had been written although her practice was to compile handwritten notes during the course of a meeting and then to write up the report of the meeting shortly afterwards. She therefore considered those notes and the ensuing reports to be an accurate representation of the proceedings at those meetings even though Mr Shoefield had not been given an opportunity to comment on their veracity at the time; (4) the notes and report in respect of the meeting on 9 July 2013 showed that her visit had been prompted by the ongoing investigation into MLM. They also recorded that the subject of due diligence was discussed and Ms Kathel said in her oral testimony that reference would have been made in that context to VAT Notice 726. She added that, when she had sent the notice to Mr Shoefield following that meeting, Mr Shoefield had never written back to query the relevance of the notice to his circumstances or to ask why the notice had been sent to him; (5) although it is not recorded in either the notes or the report of the meeting on 12 August 2013, Ms Kathel said in her witness statement that she recalled telling Mr Shoefield at that meeting that it was not sufficient to rely solely on the existence of a personal relationship in the context of due diligence and that he still needed to ask for more information from Messrs Butler and Ferguson to satisfy himself that there was no fraud in the supply chain. However, Ms Kathel would not be drawn on what information Mr Shoefield could have asked for which would have alerted him to the fraud. In her view, the fact that Mr Shoefield had done no due diligence at all meant that he had deliberately shut his eyes to that possibility; (6) neither the notes nor the reports in relation to either of the two meetings suggested that Mr Shoefield was explicitly warned about the possibility of there being fraud in the Appellant’s supply chain. However, Ms Kathel said in her oral evidence that she had given Mr Shoefield that warning; (7) the contemporaneous notes of the meeting on 9 July 2013 recorded: “ The ability to fund VAT, cash etc., investments through [the Appellant] ” and that Mr Shoefield showed the two Officers how the Data Tech system worked, whilst the report of that meeting recorded that the Appellant “ does not pay on credit ”; and (8) Ms Kathel said that she could not recall ever trying to find out whether any challenge had been made to the VAT input tax credits which had been claimed by Havas in performing a similar role to the Appellant in the transaction chain running between MLM and RAM. 69. The key points arising from Ms Wilton’s evidence were as follows: (1) at the time of her witness statement, Ms Wilton had been part of the MOJ’s Claims Management Regulation Unit for four years. The unit was responsible for, inter alia, the receipt and processing of application forms in respect of businesses which were applying for authorisation under the Compensation Act 2006 (the “CA 2006”) to trade in claims management activity; (2) in Ms Wilton’s opinion: (a) the activities in which the Appellant was involved in the course of the proposal relating to PPI leads would generally have fallen within the description of services which required authorisation under the CA 2006 and Regulation 4 of The Compensation (Regulated Claims Management Services) Order 2006; and (b) relevant MOJ guidance, located on the MOJ website, would have made that clear to anyone who had looked at that guidance; and (3) having said that, Ms Wilton said that, in the period prior to 2014, non-compliance with the requirement to be authorised and with the requirement to register with the ICO was widespread. This was attributable in some cases to ignorance of the requirements but also in some cases to the deliberate flouting of the law. She added that: (a) it was not uncommon in the data industry to rely on assurances from third parties and that she had “ seen forums on Linkedin for example ‘naming and shaming’ data brokers who have either sold poor quality data or not paid for data received ”; and (b) this would be less likely to occur if checks were undertaken by businesses prior to purchasing data, (see her witness statement at paragraph [93]). 70. The key points arising from Mr Cresswell’s evidence were as follows: (1) Mr Cresswell had had thirty years of data analytics, consulting and commercial experience. Based on that experience, Mr Cresswell explained to us the basics of de-duplication (as summarised in paragraph 56(3) above), data lead generation – the process of attracting and gaining the interest of potential customers to create future sales - and data suppression - the process of identifying customers or prospects who cannot, or do not want to, respond to a specific marketing campaign (for example because they are deceased or because they have expressed the wish for privacy). The de-duplication of data, the identification of data which was incorrect or incomplete and data suppression were all important parts of the process of “cleaning” data. It might then often be followed by “enrichment” – enhancing the data by tailoring it more specifically to the intended market in question. The cleaning and enrichment processes added value to the data leads; (2) Mr Cresswell conceded that: (a) large companies would find it easier than small ones like the Appellant to comply with data protection rules. For small companies, the lack of human resources could be an issue in that regard; (b) although he did not have direct experience of dealing in PPI leads himself, he knew that the PPI claims market in 2013 was such that properly-targeted PPI leads might well have been very valuable; (c) although Experian had not in fact supplied data to MLM, the nature of Experian’s business was such that, in concept at least, that could have occurred. For instance, if a potential customer had called Experian and asked for PPI leads, Experian might well have used that as an opportunity to sell to the potential customer data which was of relevance to PPI claims – for example, data relating to people who had mortgages and/or credit cards and were therefore more likely than others to have a potential PPI claim. It was therefore reasonable for Mr Shoefield to have accepted on trust that the PPI leads which MLM was providing to the Appellant came from companies like Experian because that could have been the case. However, Mr Cresswell said that he would have expected the Appellant at least to have asked about the type of data it was buying – for example, to ascertain whether the data in question related to people who had expressed an active interest in making a PPI claim or instead related merely to people who might be interested in making such a claim; (d) although the Appellant should not have taken on trust the quality of the data which it was purchasing from MLM (because of concerns about data protection and privacy), it was not uncommon in 2013 for a company purchasing data not to have access to that data itself. At that time, it was often the case that the relevant company would just rely on contractual protection in relation to the data quality – for example, by buying and on-selling and then waiting to see if the purchaser under the on-sale complained. That was much less common nowadays because of the greater degree of regulation in the data protection field; (e) in addition, it was by no means clear that a vendor of data in 2013, if expressly asked about the provenance of its data, would have complied with such a request – it was possible that the vendor would have been concerned about being cut out of the supply chain; (f) it was also not uncommon in the data market for a person to buy and on-sell-raw data without adding significant value to that data by cleaning or enriching it. Such simple purchases and on-sales tended to be made by small and medium sized enterprises. However, Mr Cresswell added that it was unusual for a broker in that position to have only a single source and a single customer – usually, the broker in question would be aggregating data from multiple sources and then on-selling that data to more than one customer. Moreover, where the single supplier and the single customer were known to each other before the broker became part of the transaction chain, it was difficult to see what value the broker was adding to the transaction chain; and (g) it was also not uncommon for data to be sent outside the UK because that was where the call centres which would use the data were generally located. 71. The key points arising from Mr Shoefield’s evidence were as follows: (1) in the media industry, credit exposures were generally managed by way of credit insurance or by way of requiring prepayment; (2) in paragraph 27 of his first witness statement (“AS1”), Mr Shoefield referred to the fact that he was introduced to Mr Butler “ by an old associate, Simon Kavanagh ”

. However, on cross-examination, Mr Shoefield admitted that the person he had identified in AS1 as Mr Kavanagh was in fact Mr Kevan and that he did not know Mr Kevan very well; it was really Mr Shoefield’s colleague, Mr Wood, who had had the past connection with Mr Kevan; (3) in his letter of 3 November 2014 to Mr Whelan, which is summarised in paragraph 59(55) above, Mr Shoefield had said to Mr Whelan that he had identified RAM as a potential purchaser of PPI leads and made the introduction of Mr Wright to Mr Butler for that purpose. However, on cross-examination, Mr Shoefield admitted that that statement was not true. He accepted that Mr Wright and Mr Butler already knew each other before the trades in PPI leads commenced. He initially attributed the untruth to the stress and frustration which he was suffering as a result of the delay by the Respondents in making the VAT repayments in question to the Appellant. Subsequently, on re-examination by Mr Webster, he sought to justify the statement by saying that he was referring in the relevant passage to his introduction of Mr Butler to Mr Wright in 2011, before the transactions in PPI leads had commenced; (4) Mr Shoefield accepted that the same untruth was contained in the letter from RPC to the Respondents of 11 August 2015, which is summarised in paragraph 59(66) above. Mr Shoefield blamed RPC for that error; (5) in his response of 23 August 2013 to Ms Kathel’s questions of earlier that day, Mr Shoefield referred to the fact that he had been asked to “ handle ” the data buying process between MLM and RAM. Mr Shoefield said that, by “ handle ”, he meant managing the credit risk associated with selling to RAM and facilitating the cash flow between the two entities; (6) in addition, in paragraph [33] of AS1, Mr Shoefield had said that Mr Wright and Mr Butler had wanted his input into the PPI transactions “ to manage the process, handle the cleaning of the data and, importantly, to provide the funding that was required on exports ”; (7) in relation to the first of those functions – the management of the process – Mr Shoefield was taken to his exchanges of emails with Mr Butler in January 2012 and described in paragraphs 59(7), 59(8) and 59(10) above, which demonstrated that management of the process was very much being driven by Mr Butler; (8) in relation to the second of those functions – handling the cleaning of the data – Mr Shoefield accepted that the Appellant had never had access to the data in de-encrypted form but insisted that the Appellant had cleaned the data before the data was passed on to RAM. In response to a question from Ms Stephenson as to why the Appellant would need to clean the data before the sale to RAM when the data had come from MLM and the same cleaning process would have been implemented by MLM, Mr Shoefield said, at one point, that he didn’t know whether MLM would have had sufficient access authority to the Data Tech system to be able to clean the data before the sale to the Appellant and, at another point, that it was impossible for the data to be sold within the system without the implementation of the cleaning process; (9) Mr Shoefield was also unable to explain why there was no reduction in the number of PPI leads when the Appellant applied the cleaning process before the on-sale to RAM or why the hard copy of the data which had been provided to the Respondents as described in paragraphs 59(46) to 59(49) above contained so many duplicates and obvious errors; (10) in relation to the third of those functions – the provision of funding – Mr Shoefield had elaborated on that function in paragraph [33] of AS1 when he said:

“ Neither Mr Butler nor Mr Wright had the funds to enable them to wait for the VAT repayment. The purchases by MLM were taxable supplies from the UK and exports/despatches to RAM were zero rated. There was also a substantial amount of credit to be afforded to RAM, which caused further cash flow issues ”; (11) however, Mr Shoefield said that he had not at any time conducted an analysis which compared the margin on each transaction in PPI leads with the potential risks associated with the Appellant’s credit exposure to RAM and possible non-recovery of the VAT repayments from the Respondents; (12) Mr Shoefield was taken to his exchanges of emails with Mr Skuse in January 2012 and described in paragraphs 59(7) to 59(10) above. Ms Stephenson pointed out to him that Mr Skuse’s email of 19 January 2012 showed very clearly that Havas had reservations about the commerciality of its involvement in the transaction chain and that it thought that it ought to be justifying that involvement by reference to some sort of role in enriching the data. The words used by Mr Skuse were that suspicion could be aroused in the absence of “ evidence of re-work of the data ”

. In his response to that email on 23 January 2012, Mr Shoefield had simply replied:

“ Looking to see what can be done with the data – although not keen to take out of secure environment .”

Mr Shoefield was unable to explain why he had not pointed out to Mr Skuse at that stage that Havas (through its agent, Smithfield) was performing a valuable role in managing the transactions, cleaning the data, taking on credit exposure to RAM and providing liquidity for the transactions or why Mr Skuse had not been aware of any of those “ value-added” functions when he raised his question with Mr Shoefield; (13) Mr Shoefield was aware at the time of dealing with Jump and MLM that both companies were authorised by the MOJ, he had visited Mr Butler’s offices in Leeds and he had checked that MLM had a valid VAT registration number but: (a) he had never checked with Companies House for the details of Mr Butler’s other directorships; and (b) he had never asked to see MLM’s accounts until October or November 2013, after the end of the VAT accounting periods to which this appeal related; (14) in response to Ms Stephenson’s question as to whether, had he seen the MLM accounts at an earlier time, they would have alerted him to the fact that MLM’s net assets – at £4,385 in the financial year ending 31 October 2012 - were too low to support the volume of transactions to which MLM must have been party in order to acquire the PPI leads which had been sold by MLM to Havas and then the Appellant, Mr Shoefield said that this was not the case. He said that: (a) in the first place, it was very common in the advertising industry for companies to have net assets which were very low in comparison to their turnover. This was an inevitable result of the practice within the market for the advertising agency to act as principal in acquiring advertising space and leads and then to on-sell the relevant space and leads to its true principal; and (b) in the second place, he would have regarded MLM’s low net assets as a plausible explanation for the fact that Havas and the Appellant needed to be involved in the transactions for credit exposure and liquidity reasons – as described in paragraph 71(10) above; (15) Mr Shoefield said that he had had two reasons for incorporating the Appellant in order to implement PPI trades through the Appellant. The first was to be able to increase the overall deal flow in relation to the transactions in PPI leads given that Havas’s concerns in relation to its credit exposure to RAM were constraining the number of transactions which could be done and the second was that the transactions effected through Havas necessarily involved sharing the margin with Havas – the same transactions effected through the Appellant alone would enable Mr Shoesmith to keep all of the profit. He confirmed that Havas was aware of the existence of the Appellant but not aware that the Appellant was effecting transactions in PPI leads; (16) in relation to the terms of the written contracts, Mr Shoefield admitted that he didn’t read the contracts himself. In the case of the initial contracts of January 2012 in relation to the transactions effected through Havas, he had simply passed the draft contracts on to Havas, knowing that Havas’s lawyers would review the terms of the drafts, whilst, in the case of the contracts of January 2013 in relation to the transactions effected through the Appellant, he had relied on the review of the terms of those contracts by his lawyers, Simons Muirhead & Burton. (The documents bundle contained an invoice from the latter to the Appellant for advising on the terms of the two contracts); (17) he explained that the reason why the price per PPI lead payable by the Appellant under the First Contract was higher than the price per PPI lead which was payable by Havas under the equivalent contract in January 2012 - £33.00 as compared to £24.03 – was that MLM had informed the Appellant that the price had to be increased because PPI leads were harder to come by. He added that the prices payable for PPI leads under the First Contract and the Second Contract were not set in stone – at one point, MLM and RAM had tried to reduce the Appellant’s margin but he had refused; (18) Mr Shoefield conceded that he had not paid any attention to the value of the PPI leads that the Appellant was acquiring under the First Contract because he knew that he could on-sell each PPI lead to RAM at a margin under the Second Contract. He would have been concerned in due course if the level of PPI leads so acquired which turned out to be “ non-contact ” or “ wrong product/not interested ” regularly exceeded 15% and RAM had thus repeatedly exercised its right to demand replacement PPI leads (or asked for a credit note) but this had not occurred and so he saw no reason to doubt the value of the PPI leads which the Appellant was acquiring; (19) in relation to his knowledge as to the sources of the PPI leads, Mr Shoefield made a number of contradictory statements; (20) at first, he said that he had never enquired with MLM as to the source of the PPI leads until the Respondents had approached him in July and August 2013 – again, this was because he knew that the Appellant could on-sell each PPI lead that it was acquiring to RAM at a margin; (21) but he then said that names like Experian and DBS “ came up ” in conversations with MLM before those enquiries were made.

(a) in his email to Ms Kathel of 25 September 2014, he said:

“ My supplier also has only nine suppliers – these include company’s like Experian ”; (b) in the note of a meeting between him and his adviser, Mr Ahmed, with the Respondents on 19 April 2016, Mr Shoefield is recorded as having said that he “ had been used to dealing with the likes of DBS and Experian etc. and did believe this was where the leads came from ”; (c) at paragraph [42] of AS1, he said that “ [it] was the belief of Mr Wright and I that MLM must have been obtaining the leads from such companies ”; (d) at paragraph [75] of AS1, he said that “ Experian and others would have been very credible suppliers who could source large quantities of leads. I still insist that, although I was not told specifically about Experian, it would be impossible for MLM to obtain such vast amounts of data from smaller companies ”; and (e) at paragraph [24] in his second witness statement, he said “ I have always stated that Experian and DBS were the suppliers ”; (22) similarly, he said in his oral testimony that he didn’t believe at any point that the PPI leads had been sourced in a call centre under the control of Mr Butler but, after he was directed to: (a) the report of the meeting on 12 August 2013 prepared by Ms Kathel – which referred to his saying that MLM “ has a 700-seater call centre that is sourcing values for new products ”; (b) the subsequent exchange of emails with Ms Kathel on 23 August 2013 - in which his response to the first question posed by Ms Kathel strongly suggested that the call centre mentioned in that question was the source of the PPI leads; and (c) his letter to Mr Whelan of 3 November 2014 – in which he said that his understanding was that the PPI leads derived from a call centre, he then said that his belief at the time of the meeting, that email exchange and that letter was that the PPI leads came from a call centre; (23) Mr Shoefield said that, at the meeting with Ms Kathel and Ms Ahmed of 9 July 2013, the two Officers had not taken him through the details of VAT Notice 726. He accepted that the notice had been sent to him following that meeting but said that he didn’t think that the notice applied to him given that it was expressed to be limited to certain specific sales of goods. However, he conceded that he had never replied to the email from Ms Kathel attaching the notice to ask why she had sent it to him or to tell her that he didn’t think that it applied to the Appellant; (24) in relation to the third party payments which Havas had made to IMD at the direction of MLM in 2012, Mr Shoefield said in his oral testimony that he didn’t have any concerns about paying IMD given that the payment direction came from MLM and referred to making payment to IMD “ c/o MLM ”

. He added that it seemed perfectly reasonable to him that MLM had given as the reason for asking that payment be made to IMD the fact that its bank account at NatWest had been frozen due to suspected money-laundering issues because, in his experience, dealing with banks in the UK could often be difficult. However, he was directed to a note of a meeting between him and his adviser, Mr Ahmed, with the Respondents on 19 April 2016 in which he had said that he “ was not comfortable paying a Cypriot bank account and did want to pay a UK account ”; (25) Mr Shoefield said that he had not become aware of the fact that he needed to be registered with the ICO until Ms Kathel had asked to see his ICO registration certificate on 22 August 2013 but that he had responded to that request by immediately applying for registration and then sending to Ms Kathel, in quick succession, a copy of his application form (on 22 August 2013), a copy of the confirmation from the ICO of his application (on 23 August 2013) and then the registration certificate itself (on 4 September 2013). He added that clause 6 in each of the First Contract and the Second Contract contained warranties and undertakings in relation to regulatory compliance and that these would not have been included if he had not been taking that issue seriously; and (26) finally, Mr Shoefield pointed out that he had repeatedly offered the Respondents the opportunity to attend meetings with MLM and DataTech and that he would not have done that if he had been trying to hide anything from the Respondents. 72. The key points arising from Mr Wright’s evidence were as follows: (1) Mr Wright confirmed that he was the Managing Director of RAM and that he had given evidence on behalf of the Respondents in the criminal trial of Mr Butler and Mr Ferguson; (2) he also confirmed that he had met Mr Shoefield in around 2008 when Mr Shoefield was at Truly London Limited and that Mr Shoefield had introduced him to Mr Butler in around June 2011. RAM’s first transaction for PPI leads was in November 2011 and the seller of the PPI leads in that transaction and several others which followed it up until November 2012 was a company called Leadpoint UK Limited (“Leadpoint”), which had been introduced to him by Mr Butler; (3) he had been told by Mr Butler that Mr Butler could source the PPI leads and also that Mr Butler had a buyer lined up for the PPI leads called Quantum Media LLC (which later became Delhicon LLC) (“Delhicon”). According to Mr Butler, as Delhicon had no proven track record – it was a new company – the interposition of RAM would facilitate the trades because RAM would stand in the middle and take the credit exposure on Delhicon. This made sense to Mr Wright from the commercial perspective. In addition, at the outset, he had received assurances from Mr Butler that Mr Butler would underwrite the creditworthiness of Delhicon. However, Mr Wright remained concerned about the creditworthiness of Delhicon and, in October 2013, he had asked for, and received, a business overview paper from Delhicon; (4) Mr Wright’s suppliers were Leadpoint, a company called Broadstone Limited run by a Mr Gary Smith, Havas and the Appellant. Midway through 2012, Mr Wright learned that MLM was the supplier to those suppliers. Mr Wright believed that the sources of the PPI leads coming from MLM were companies like Experian. He said that he had heard that from Mr Shoefield, who had been told the same by Mr Ferguson; (5) Mr Wright’s contacts at Delhicon were initially Ms Genevieve Pennill and subsequently Mr James Taylor. However, Mr Wright never met either of those people in person or spoke to them on the telephone. Instead, all communications with them were by email. Mr Wright’s relationship with Mr Taylor was at times very fractious as a result of Delhicon’s either not accepting or not downloading PPI leads for which it had asked or delaying in making payments for the PPI leads. On those occasions, Mr Butler would always intervene to resolve the dispute. Mr Wright would meet Mr Butler about once every quarter. In addition, Mr Wright would communicate with Mr Butler by text; (6) we were presented with an exchange of texts between the two in which Mr Wright told Mr Butler that RAM had a specified sum of money to disburse and Mr Butler told Mr Wright which of RAM”s suppliers RAM should pay. Mr Wright explained the exchange as follows. As a result of RAM’s transactions with its four suppliers, RAM owed money to each of those suppliers. Those suppliers needed cash to fund their future purchases of PPI leads and Mr Butler knew which of them was currently in the market to make more purchases. So Mr Butler would instruct RAM to pay the surplus cash to the RAM supplier which needed the cash to make the future purchase; (7) Mr Wright would negotiate on price with his suppliers and with Mr Taylor. Mr Butler was not obviously involved in those negotiations although Mr Wright now accepted that Mr Butler must have been involved behind the scenes; (8) Mr Wright said that he had three reasons for believing that no fraud was involved in the trades in PPI leads: (a) first, his initial supplier of PPI leads, Leadpoint, was a well-known and respected generator of data leads; (b) secondly, it made commercial sense to him that leads would be acquired from the UK and then sold to call centres abroad; and (c) thirdly, the fact that provision was made in the agreements for replacement leads to be provided if the “ non-contact rate ”, together with the “ wrong product/not interested rate ” exceeded a specified percentage suggested that the PPI leads were genuine; and (9) accordingly, in Mr Wright’s opinion, if Mr Butler and Mr Ferguson were found to be fraudsters, “ I would say that they are very convincing ”. THE PARTIES SUBMISSIONS The Respondents’ arguments 73. Ms Stephenson and Mr Carey, on behalf of the Respondents, submitted as follows: (1) the Respondents’ primary case was that Mr Shoefield, and hence the Appellant, knew that the transactions in question were connected with the fraudulent evasion of VAT; (2) the Respondents’ secondary case was that, even if the Appellant did not have such actual knowledge, it should have known that the transactions in question were connected with the fraudulent evasion of VAT; (3) in support of the above assertions, Ms Stephenson and Mr Carey made the following points: (a) the Appellant had made a large profit by doing little more than logging on, pressing a button, and selling hundreds of thousands of pounds worth of PPI leads at a time; (b) Mr Shoefield had given three reasons for justifying the Appellant’s margin. The first was that he managed the process, the second was that he cleaned the data and the third was that he provided working capital so that the transactions could be undertaken; (c) none of those assertions was sustainable; (d) as regards management, it was clear from the evidence that the Appellant did not upload or download the data. Instead, it simply passed the data on to RAM without de-encrypting the data. In addition, Mr Shoefield did not introduce Mr Wright and Mr Butler at the time when the opportunity to implement the transactions in PPI leads arose. Instead, Mr Wright and Mr Butler already knew each other at that time and the Appellant was simply inserted into the transaction chain between MLM and RAM. Finally, the evidence demonstrated that it was Mr Butler who was pulling the strings in relation to the transactions – it was Mr Butler who offered Mr Shoefield the opportunity to become one of the “ distributors ” in the transaction chain and it was Mr Butler who told Mr Wright which of RAM’s four suppliers should be paid out of any surplus cash in RAM; (e) as regards cleaning, the evidence suggested that whatever cleaning functions existed within the Data Tech system, those functions did not work properly because the data which had been provided to the Respondents were full of duplicates and other errors. However, leaving that aside, and assuming for the moment that those functions within the Data Tech system did work, there was no need for the Appellant to perform those functions because they had already been performed by MLM on the sale to the Appellant. Mr Shoefield said, at one point, that he didn’t know whether MLM would have had sufficient access authority to the Data Tech system to be able to clean the data before the sale to the Appellant and, at another point, that it was impossible for the data to be sold within the system without the implementation of the cleaning process. Those answers were inconsistent. In addition: (i) it would make no sense for MLM, which had developed the Data Tech system, to deprive itself of the ability to clean the data prior to a sale it was making; and (ii) if it was impossible for the data to be sold without engaging the cleaning process, then, by definition, any cleaning effected by the Appellant was not adding value to the process. Finally, Mr Shoefield appeared to have no clear idea of what cleaning function the Appellant was performing. He could not explain exactly what data was removed as a result of the Appellant’s involvement in the transaction chain and he accepted that there was never any reduction in the number of PPI leads between the sale by MLM to the Appellant and the sale by the Appellant to RAM; (f) as regards the provision of working capital, this reason appeared not to have occurred to Mr Skuse when he wrote to Mr Shoefield on 19 January 2012 to say that, following a review of the transaction, his “ VAT boys want to know that we are doing some level of mark up (evidence of re-work of the data) on the trade, otherwise its something that could cause suspicion.” By that email, Mr Skuse indicated that, so far as Havas was concerned, the provision of working capital in the transaction was insufficient to justify Havas’s role in the transaction chain and that some element of re-work of the data was necessary in order to do so. Moreover, Mr Shoefield did not respond to Mr Skuse’s email by pointing out that Mr Skuse had missed something in his analysis by failing to realise the value which the provision of working capital provided to the transaction chain. Instead, he had simply replied that he would see what he could do about re-working the data without exposing the data to any damage (see Mr Shoefield’s email of 23 January 2012). In addition, Mr Shoefield had admitted that he had not at any time conducted an analysis which compared the margin on each transaction in PPI leads with the potential risks associated with the Appellant’s credit exposure to RAM and possible non-recovery of the VAT repayments from the Respondents. That was peculiar given that an experienced businessman like Mr Shoefield might have been expected to have weighed up the potential financial risks and rewards of its provision of working capital; (g) the above showed that none of the reasons which Mr Shoefield had given to justify the Appellant’s role in the transaction chain held any substance. In short, the opportunity was too good to be true and therefore Mr Shoefield either knew or should have known that fraud was involved; (h) the Appellant’s supplier and the Appellant’s customer knew each other and the relationship between all of the parties – that is to say, Mr Butler and Mr Ferguson (MLM), Mr Shoefield (the Appellant) and Mr Wright (RAM) – was a close one; (i) the Appellant had failed to conduct meaningful due diligence on either its supplier or its customer. If the Appellant’s claim that its provision of working capital was a significant reason for its involvement in the transaction chain were to be plausible, then one would have expected the Appellant to have checked on the financial position of MLM. An obvious check would have been on the net assets of MLM, which would have led to a question mark over how a company of such limited value (less than £5,000 of net assets in the financial year ending 31 October 2012) could be making such large purchases of PPI leads (a turnover of almost £23m in the same financial year); (j) the Appellant had failed to conduct any due diligence on the PPI leads that it was buying and selling. In his testimony, Mr Cresswell, a person experienced in data governance, had expressed the view that he would have expected a company in the Appellant’s position to have ascertained the nature and value of the PPI leads that it was buying, even if the sources of the PPI leads were unnamed; (k) the transactions lacked commercial reality. In his testimony, Mr Cresswell had explained how there might be a role in a transaction chain for a company which added value to the data that was being supplied in the chain or which had a wide distribution network so that it could on-sell to customers which were unkown to its suppliers. However, where, as was the case here: (i) there was no increase in the value of the data between the time of its supply by MLM to the Appellant and the time of its onward supply by the Appellant to RAM; and (ii) there was both a single supplier to the Appellant and a single customer for the Appellant, who were known to each other and had entered into other business together before the opportunity to implement the relevant transactions arose, the interposition of the Appellant in the transaction chain made no commercial sense. Notwithstanding the evidence of Mr Shoefield and Mr Wright to the effect that there were ongoing negotiations between MLM, the Appellant and RAM in relation to the respective margins of the Appellant and RAM, the fact was that each of the transactions to which this appeal related involved the same purchase price per PPI lead and the same sale price per PPI lead and the mark up derived by the Appellant in relation to each purchase and sale was the same. In addition, if, as Mr Shoefield had alleged, the reason for the incorporation of the Appellant was to derive a share of the lucrative market in PPI leads over and above the share that Smithfield was deriving from acting as agent for Havas, then why didn’t the Appellant: (A) find sources of PPI leads other than MLM; (B) try to ascertain the identity of MLM’s sources and try to cut MLM out of the transaction chain; (C) advertise; or (D) have its own website? In the view of the Respondents, Mr Shoefield did not take all the steps which were available to him to maximise the Appellant’s profits; (l) the First Contract and the Second Contract were very similar and contained clear errors. This demonstrated that the contracts had little purpose other than as window dressing and that the work ostensibly done on the contracts by Simons Muirhead & Burton must have been minimal given the relatively low amount charged and the errors in the contracts. Mr Shoefield in his evidence had sought to distance himself from the contracts by saying that he had relied on his lawyers to review the terms of the contracts. In reality, the contracts had little bearing on the way that the parties operated, as was evidenced by the reference to “ Smithfield ” in clause 6.2 of the Second Contract. Moreover, the fact that the contracts were so similar and emanated from a single person, Mr Ferguson, should have aroused suspicion; (m) the Data Tech platform had virtually no functionality and the only function that it was supposed to have (de-deduplication) did not work; (n) Mr Shoefield had adopted a laissez-fair e attitude to the Appellant’s regulatory obligations. In her testimony, Ms Wilton had pointed out that the Appellant should have been both registered with the ICO and authorised by the MOJ at the time when the transactions which were the subject of this appeal were implemented. Whilst Ms Wilton had also said that a number of participants in the market were unaware of the need for ICO registration, Mr Shoefield in his testimony had confirmed that he knew that MLM was ICO-registered. The Respondents suggested that that should have alerted Mr Shoefield to the fact that the Appellant should also have been so registered and that the fact that Mr Shoefield had failed to do so was down to his desire not to draw attention to the Appellant and to the fact that the Appellant’s transactions were not commercially sensible; and (o) there was clearly an overall scheme to defraud, given Mr Butler’s involvement and orchestration at every stage in the transaction chain – for example, telling Mr Wright where to transfer his surplus cash - and the knowledge of Mr Shoefield and Mr Wright that that was the case. The fact that there was an overall scheme to defraud meant that there would be little incentive to insert an unknowing participant into the transaction chain because of the risk that such a participant might undo the whole fraud; (4) at the time of their meetings with the Appellant in July and August 2013, the Respondents were not in a position to warn the Appellant that there was fraud in the Appellant’s supply chain or to issue a tax loss letter to the Appellant because of their obligations of confidentiality to other taxpayers and because they were still investigating MLM at the relevant time and hadn’t yet reached any firm conclusions. In any event, any shortcomings in the Respondents’ management of the case against the Appellant and their failure to pursue similar cases against: (a) the Appellant in respect of its VAT accounting periods preceding the VAT accounting periods in question; or (b) Havas, were wholly irrelevant to the questions which were at issue in this appeal. Those might be relevant to a judicial review in the High Court of the Respondents’ conduct but not to whether the Appellant knew or should have known that its transactions were connected with fraud – see Harwich GSM Limited v The Commissioners for Her Majesty’s Revenue and Customs

[2012] UKFTT 279 (TC)

at paragraph [140]; (5) notwithstanding the point made in paragraph 73(4) above, the evidence of Ms Kathel was that, at the meeting of 9 July 2013, she had drawn Mr Shoefield’s attention to the requirements in VAT Notice 726. This was supported by Ms Kathel’s contemporaneous note of the meeting and the reference to that discussion in her email of 12 July 2013 attaching the relevant notice. Although Mr Shoefield claimed that that level of detail in relation to the due diligence provisions in the notice was not discussed, he had not produced any note of the meeting himself and had frequently responded to questions raised of him in cross-examination by saying that he could not remember the position because of the passage of time. Accordingly, although Ms Kathel’s note was unsigned, her testimony was to be preferred in this respect. In any event, Mr Shoefield had accepted that he had received the relevant notice under cover of Ms Kathel’s email of 12 July 2013 and had at no stage responded to Ms Kathel to query why the notice was relevant to the Appellant’s circumstances; (6) when one looked at the examples given in section 6 of the notice of circumstances which should alert a trader to the possibility of fraud in the supply chain, a number of them were pertinent in the Appellant’s case. For example: (a) the Appellant’s supplier and customer were already known to each other before the Appellant joined the transaction chain and the PPI leads which the Appellant bought from MLM were identical in quantity and specification to the PPI leads which the Appellant had sold to RAM; (b) the deals always involved the same margin irrespective of the date, quantities and specifications of the data traded on any particular date; (c) normal commercial practices were not adopted in the negotiation of prices; (d) the value of the deals was very high in comparison to the creditworthiness of MLM; and (e) the Appellant knew nothing about the derivation or value of the data; (7) the requests from MLM to make payments to an offshore bank account in Cyprus because its account at NatWest had been frozen due to money-laundering concerns should also have alerted Mr Shoefield that all was not well with MLM. Requests for payments to be made offshore were stated in the relevant notice to be an indication of fraud; and (8) finally, there had been a general lack of credibility in Mr Shoefield’s evidence, as exemplified by the number of occasions described above in summarising the witness evidence in which: (a) Mr Shoefield had given incorrect answers to the Respondents in communications over the period of the investigation; and (b) had changed his evidence in the course of cross-examination. The Appellant’s arguments 74. Mr Webster, on behalf of the Appellant, responded as follows: (1) although the evidence provided by Mr Shoefield had not been entirely satisfactory, the same criticism could reasonably be levelled at the two Officers of the Respondents who had given evidence at the hearing. In particular, the Officers who had reached the conclusion that the claim for the VAT repayment should be denied had had little commercial experience and had a minimal knowledge of the advertising industry in general and sales of data in particular. Moreover, it did not appear to have occurred to anyone within the Respondents that this case did not involve a missing trader and was not in fact a typical MTIC. Instead, it was simply a case of fraud. That failure to identify the true nature of the case had led the Officers in question to draw adverse inferences from the evidence which were not in fact justified; (2) there were errors in the manner in which Ms Kathel had dealt with the meetings of 9 July 2013 and 12 August 2013. For example, she had kept no record of the times at which the meetings took place and was unclear as to when the meeting reports were written up. In addition, she had not asked Mr Shoefield to confirm that the reports were an accurate summary of the meetings or to countersign the reports. More seriously, despite the fact that the Respondents first discovered that MLM was reclaiming VAT input tax in respect of false invoices as early as November 2012, and that Ms Kathel’s meeting with Mr Shoefield on 9 July 2013 had been prompted by the investigation into MLM, no mention was made of this to Mr Shoefield at the meeting and no tax loss letter, alerting him to the fact that there was fraud in the Appellant’s supply chain was issued. Finally, Mr Shoefield was sent a VAT notice (VAT Notice 726) which, on its express terms, did not apply to trading in data; (3) in addition, the errors and contradictions in Mr Shoefield’s evidence could reasonably be explained by reference to differing states of knowledge over a lengthy period, and the length of time which had elapsed since the events in question occurred, as opposed to an intention to mislead; (4) in relation to the allegation that Mr Shoefield had had actual knowledge that the relevant transactions were connected with fraud, the chain of transactions initially involved Havas, as opposed to the Appellant. In that case, the opportunity which was put to Mr Shoefield by Mr Butler was immediately forwarded by Mr Shoefield to Havas and Havas satisfied itself that its role in the transaction chain was appropriate; (5) the fact that Mr Shoefield had repeatedly invited Officers of the Respondents to meet the Appellant’s counterparties and had regularly provided information willingly and quickly were indications that Mr Shoefield was not aware of any fraud in the Appellant’s supply chain. Mr Shoefield was not trying to mislead the Respondents but was, on the contrary, anxious to assist the Respondents with their enquiries; (6) there were other indications that Mr Shoefield was not aware that there was fraud in the Appellant’s supply chain. For example: (a) in the Appellant’s application for VAT registration, Mr Shoefield had stated that the Appellant’s expected annual turnover was going to be £5m. Had Mr Shoefield been intending to escape the notice of the Respondents, he might have been expected to insert a much lower figure for the Appellant’s expected annual turnover; and (b) Mr Shoefield took advice from lawyers in relation to the First Contract and the Second Contract before the Appellant entered into those documents in January 2013. Similarly, Mr Shoefield had been advised by accountants before setting up the Appellant in January 2013. It was unlikely that Mr Shoefield would have engaged reputable professionals to advise the Appellant if he had known that the Appellant’s transactions were linked to fraud by the Appellant’s supplier because those professionals might well have discovered the fraud; (7) in relation to the allegation that Mr Shoefield should have known that the relevant transactions were connected with fraud, as at the start of the 07/13 VAT accounting period, Mr Shoefield was in the following position: (a) there was, demonstrably, a huge demand for PPI leads; (b) Mr Shoefield had had a satisfactory experience, through his involvement with Smithfield, of having acted as agent for Havas in buying from MLM and on-selling to RAM for 17 months. (Indeed, trading of PPI leads through Havas continued from January 2012 until well after the transactions to which this appeal related occurred); (c) even before January 2012, there had been previous business relationships between Mr. Shoefield, Mr Butler and Mr Wright and there was no evidence to suggest that there was anything questionable about those prior dealings; (d) most of the trading in PPI leads which had been done with MLM and RAM hitherto had been through Havas, a reputable multinational agency, which had its own compliance and legal departments and which had carried out its own due diligence; (e) there had been, to Mr Shoefield’s knowledge, a VAT inspection of MLM which had resulted in MLM’s being allowed to continue to trade; (f) all of the VAT input tax credits in relation to Havas’s previous trading with MLM and RAM had been allowed without any cavil. The Respondents had not challenged claims in respect of VAT input tax made by Havas, despite the fact that the value of the transactions involved in the trading chain which included Havas was much larger than the value of those involved in the trading chain which included the Appellant; (g) similarly, the Respondents had not queried the right of the Appellant to claim VAT repayments in respect of the Appellant’s previous VAT accounting periods (and indeed those repayments had been made by the Respondents); (h) no warnings had been issued about the Cypriot bank in which MLM’s account was located; (i) RAM had never defaulted on its debts to Havas or the Appellant; and (j) RAM had not raised any suggestions that the data which Havas and the Appellant had sold to it was generally not effective for its purposes. There had been some claims from RAM for credits in relation to data which were not useful but the natural inference to be drawn from those claims was that, in general, the data did have substantial and meaningful value. In other words, the credit notes did not reflect wholesale failings in the data but just failings in certain cases. There was nothing in any of the above facts to put Mr Shoefield on notice that there was anything untoward in the trading chain; (8) in addition, Mr Shoefield: (a) knew that both MLM and RAM were registered companies; (b) knew that MLM was registered for VAT; and (c) could not reasonably be expected to ask MLM for the details of the suppliers to MLM or the invoices rendered to MLM by those suppliers. Thus, subject to the arguments about the accounts of MLM to which reference is made in paragraph 74(10) below, there was nothing for which Mr Shoefield could have asked which would have alerted him to the existence of fraud in the Appellant’s supply chain; (9) the Officers of the Respondents who had given evidence had clearly reached the same view because they had been unable to identify what information Mr Shoefield would have been able to obtain from conducting due diligence which would have alerted him to the existence of the fraud. Instead, they had simply clung on obdurately to the fact that Mr Shoefield had not done any due diligence as if the absence of due diligence in itself meant that Mr Shoefield should have known that its supply chain involved fraud; (10) the Respondents had, however, made the point that, had Mr Shoefield asked for (and been provided with) the accounts of MLM at an earlier stage, he would have been able to identify grounds for suspicion because of the very low net asset value of MLM in its financial year ending 31 October 2012. There were two responses in relation to that argument. First, Mr Shoefield had testified to the effect that it was common practice in the advertising industry for an agency to contract as principal with third parties and then to make a back-to-back onward supply to its principal, with the result that the accounts of an advertising agency would generally show a level of turnover which was disproportionately high in comparison to its net assets. Thus, in the opinion of Mr Shoefield, had he checked the accounts of MLM earlier than he did, he would not have seen anything amiss in the insubstantial net asset position of MLM relative to MLM’s turnover. Secondly, Mr Shoefield had testified to the effect that, insofar as the insubstantial net asset position of MLM was a ground for concluding that MLM was a man of straw, that would only have confirmed his view that MLM lacked the working capital which was needed to implement the transactions with RAM and therefore confirmed to him that the Appellant’s involvement in the transaction chain was fulfilling a valid commercial function; (11) in response to the Respondents’ submission that the request to make payments to MLM’s account with a Cyprus bank as a result of the freezing of its account with NatWest should have aroused suspicion, Mr Shoefield had testified that he had in the past experienced difficulties himself in using the UK banking system and that he did not consider the fact that MLM’s account with NatWest had been frozen to be any cause for concern; (12) turning to the questions raised in VAT Notice 726, it was notable that: (a) Mr Shoefield already knew Mr Butler and Mr Wright even before the transactions in PPI leads commenced and that such transactions had already continued for some 17 months (largely through Havas, as opposed to the Appellant) without any difficulties before the VAT accounting periods in question. This was not a case where the supplier to the Appellant was completely unknown to the Appellant before the transactions in question; (b) there was a commercial risk in the transactions for the Appellant in that it was taking a credit risk on RAM and not entitled to defer its obligation to pay MLM until it received payment from RAM; (c) payments were made to a UK bank account of MLM as well as to MLM’s Cyprus bank account; (d) the Appellant had credit insurance in respect of its exposure to RAM; (e) there were formal contractual arrangements between the three parties which had been reviewed by the Appellant’s lawyers; (f) there was a lucrative market in PPI leads at the relevant time and no suggestion that there was any insufficiency in the PPI leads that were said to be the subject of the transactions; (g) the commercial rationale for the transactions was clear; and (h) normal commercial practices were adopted by the parties in that, as both Mr Shoefield and Mr Wright had said in their testimony, the margins were potentially subject to renegotiation at any time. In addition, the checks suggested in section 6.2 of the notice were appropriate for transactions between parties who had not traded before and not for parties such as those in this case who knew each other and had done business together before; (13) as for the failure on the part of the Appellant to register with the ICO or to obtain authorisation from the MOJ, Ms Wilton had testified to the fact that, at the time of the transactions which were the subject of this appeal, there had been a great deal of non-compliance due to ignorance of the relevant legislation. Moreover, there was no reason why, just because Mr Shoefield knew that MLM was registered with the ICO, he should have deduced that the Appellant should also be so registered. As the Appellant’s transactions in the data were limited to the purchase and sale of the encrypted data and the cleaning of the data, and therefore different from those implemented by MLM, it would have been reasonable for Mr Shoefield to conclude that the Appellant’s position under the legislation in question was different from MLM’s and that there was no reason why the Appellant needed to be registered just because MLM was so registered; (14) Mr Wright had been put forward as a witness of truth by the Respondents in the criminal prosecution of Mr Butler and Mr Ferguson. It was therefore surprising, to say the least, that, in the present proceedings, the Respondents were alleging that he was involved in the fraud; and (15) in the criminal trial, the trial judge had expressly referred to the fact that innocent third parties had been caught up in the fraud and it was difficult to see to whom the judge could have been intending to refer in that remark if it were not for Mr Shoefield and Mr Wright. DISCUSSION Introduction 75. We think that it is helpful to start this section of our decision by reiterating that the sole point which we have to decide is whether, on the balance of probabilities, the Respondents have satisfied us that the Appellant knew, or should have known, that the transactions which are the subject of this appeal were connected with the fraud perpetrated by Mr Butler. 76. We say this because, in the course of the hearing, there was considerable focus on the alleged shortcomings in the manner in which the relevant Officers of the Respondents both conducted their investigation into the Appellant and reached the conclusions which they did. We therefore agree with Ms Stephenson and Mr Carey that it is important to bear in mind from the outset that: (1) this is not an action of judicial review against the Respondents and that any shortcomings in the conduct of the Respondents’ Officers are wholly irrelevant in determining whether the Appellant, as represented by Mr Shoefield, knew or should have known that its transactions were connected with the fraud; and (2) similarly, the conclusions drawn by the Respondents’ Officers as to whether or not the Appellant, as represented by Mr Shoefield, knew or should have known that its transactions were connected with the fraud are wholly irrelevant in determining our conclusions on that subject. 77. Having said that, we feel bound to record that we do have some concerns over the way in which the Respondents have conducted themselves in the course of the proceedings leading up to this appeal. In particular, we agree with Mr Webster that: (1) this appeal has nothing to do with a missing trader and it is unhelpful that the Respondents approached the matter as if it was. The fraud in this case was simply the forging of invoices by Mr Butler and that gives rise to a slightly different factual matrix from missing trader fraud in terms of the tests outlined in Kittel. There is, after all, a difference between knowing, or having the means of knowing, that your supplier is going to disappear owing VAT to the Respondents and knowing, or having the means of knowing, that your supplier has been forging its purchase invoices; (2) it is surprising that neither of the Respondents’ Officers who gave evidence at the hearing was able to explain why the transactions in this appeal had been singled out for challenge when none of the transactions in the same form implemented by Havas and none of the transactions in the same form implemented by the Appellant in its earlier VAT accounting periods had been challenged in the same way. We would have expected the relevant Officers to have discussed those other transactions with the appropriate colleagues and been aware of the differences and similarities between them and the transactions to which this appeal relates before reaching the conclusions which they reached in relation to the latter transactions. In particular, it was incumbent on those Officers to have understood the basis for the decision not to challenge those other transactions, even if that basis was simply an absence of resources within the Respondents; (3) the Officers ought also to have been aware that, according to its express terms, VAT Notice 726 did not apply to the Appellant’s situation and, when Ms Kathel sent the notice to Mr Shoefield, she should have explained that to Mr Shoefield and gone on to say that he should nevertheless apply the tests set out in section 6 of the notice as if it did. (However, we do not accept Mr Webster’s argument that Mr Shoefield was entitled simply to ignore those tests on the basis that the notice technically did not apply to the Appellant. It was clear from the context in which the notice was sent to Mr Shoefield that the Respondents were of the view that the notice was in point and, if Mr Shoefield had read the notice – which we doubt - it would have been incumbent on him to question the relevance of the notice if that was what he believed after reading it. In our view, the argument that Mr Shoefield actually read the notice and concluded that it did not apply to the Appellant but then did nothing further about it strains credulity. Moreover, on the same day that the Respondents sent VAT Notice 726 to Mr Shoefield, they sent a letter to the Appellant in relation to ABPs (see paragraph 59(29) above). That letter, which contained similar questions to those set out in section 6 of VAT Notice 726, made it clear that it applied to a wider range of transactions than the ones mentioned in the notice and that those transactions could include the supplies which were being received and made by the Appellant. There is therefore no reasonable basis on which Mr Shoefield can credibly contend that, on and after 12 July 2013, he was unaware of the questions which needed to be considered in relation to possible fraud in the Appellant’s supply chain); and (4) the report of each of the meeting on 9 July 2013 and the meeting on 11 August 2013 should have set out the time of the relevant meeting and the date on which the relevant report was written. 78. We now turn to address the substantive questions which are in issue in this case. Actual knowledge 79. It is our view that the Respondents have not established, on the balance of probabilities, that the Appellant, as represented by Mr Shoefield, had actual knowledge that its transactions with MLM were connected with fraud. Indeed, we believe that it is highly unlikely that Mr Shoefield had such actual knowledge. We say this for the reasons which follow in paragraphs 80 to 98 below. An absence of evidence 80. First, we have been unable to detect anything in the voluminous written evidence with which we have been presented to suggest that Mr Shoefield was aware that Mr Butler was committing fraud. We have been presented with a number of email and text communications between the two men and there is no indication in any of those communications that Mr Shoefield was colluding with Mr Butler in the latter’s perpetration of the fraud or was aware that the fraud existed. On the contrary, those communications can best be described as anodyne - they either involve communications about the opportunity for Havas, and subsequently the Appellant, to act as a “ distributor ” (as Mr Butler put it in one of his emails) of the data to RAM, details about specific transactions in PPI leads or personal exchanges unrelated to business matters. 81. It is of course theoretically possible that, outside of any of the documentation with which we have been provided, Mr Shoefield and Mr Butler were entering into oral communications which would have indicated that Mr Shoefield was aware of Mr Butler’s fraud but, if that had been the case, one would have expected to see some vestige of those communications in the documentation with which we have been provided – for example, a stray suspicious cross-reference. We have not detected any such vestige. In any event, we have been presented with no evidence that Mr Shoefield and Mr Butler entered into oral communications by reference to which Mr Shoefield’s actual knowledge of Mr Butler’s fraud can be established and we think that, in view of the tone of the written communications between the two which we have seen, it is somewhat fanciful to suppose that oral communications of such a nature existed. 82. In this regard, we should say expressly that we reject the argument which was made by the Respondents’ Officers at the hearing that the fact that the communications between Mr Shoefield and Mr Butler were not strictly confined to business matters - and also extended to social communications and arrangements such as barbeques – meant that the relationship between the two men was such that Mr Shoefield must have had actual knowledge of Mr Butler’s fraud. We think that, in this, as in their approach to certain other matters which we will shortly address, the Respondents have revealed an inexperience of the way in which business in the commercial world is generally conducted. The fact is that business people largely operate on the basis of personal relationships and mutual trust which is built up in the course of dealings over many years. It is perfectly understandable that Mr Shoefield, having done business with Mr Butler for some years, would develop a personal relationship with Mr Butler which extended beyond purely business matters. 83. For that reason, the suggestion made by the Respondents that the failure by Mr Shoefield to investigate how Mr Shoefield might be able to go straight to MLM’s sources of data and thereby cut MLM out of the transaction chain somehow demonstrates that Mr Shoefield must have been complicit in Mr Butler’s fraud is, in our view, wide of the mark. A business will not survive for very long if it develops a reputation for underhand practices. We therefore consider that the failure on the part of Mr Shoefield to explore ways of cutting MLM out of the transaction chain does not reveal complicity in the fraud in any way. 84. A similar point may be made in relation to the Respondents’ observation that the similarities between the First Contract and the Second Contract and the various errors which were observed in those contracts should be regarded as evidence that the contracts were effectively “ window dressing ” and designed to mislead any observer into thinking that the transactions involved in the transaction chain were normal commercial arrangements. 85. In our experience, when individuals have been doing business together for a number of years, they often pay very little attention to the precise terms of any written contract which may exist between their respective companies. They know that their relationship depends on mutual trust and on behaving in accordance with the terms of what has actually been agreed orally, even if those terms are not properly recorded in the written contract. The differences between clause 4.2 in each contract in this case exemplify this point. The possibility that those differences might have opened the way to one party’s seeking to take advantage over another as a result of an ambiguity in the drafting is unrealistic; the relevant party would know that any such behaviour would lead to the end of the relevant relationship with the other party and all possibilities of future business with that party. Worse still, that behaviour might well damage the first-mentioned party’s business as a whole because word would spread within what is a relatively small community that that party was not to be trusted. 86. Another example of this may be found in the fact that the sole remedy for exceeding the specified percentage of “ non-contact rate ” and “ wrong product/not interested rate ” in both written agreements was the provision of replacement PPI leads whereas, in practice, the seller in each case provided a credit note instead of replacement PPI leads. The commercial outcome to which that remedy gave rise was exactly the same as if replacement PPI leads had been provided and the fact that no provision for that alternative remedy was set out in the written contracts is, in our view, irrelevant. 87. We agree that the terms of the two contracts are not a model of clarity but we do not ascribe that lack of clarity to any adverse motives on the part of Mr Shoefield. Instead, we think that a more likely explanation is that people who draft contracts sometimes make mistakes. An example of this is the reference to “ Smithfield ” in clause 6.2 of the Second Contract. Given the involvement of Smithfield (as agent for Havas) in the earlier contract chain, we believe that this is much more likely to have been an erroneous hangover from the contract which was in place between Smithfield (as agent for Havas) and RAM than an indication of any mala fides on the part of the Appellant or RAM. We therefore do not consider that the reference to “ Smithfield ” in that clause is evidence of any misfeasance or wrongdoing on the part of the Appellant. Instead, it is simply an understandable human error. Co-operation with the Respondents 88. The second reason for our conclusion on this point is to be found in the manner in which Mr Shoefield conducted himself both before the transactions in question and during the course of the investigation. We tend to agree with Mr Webster’s submission that, if Mr Shoefield had been aware that the Appellant was entering into transactions connected with fraud, he would not have put the figure of £5m as the expected annual turnover of the Appellant in the Appellant’s VAT registration form (but would instead have inserted a much lower number, so as not to excite suspicion) and he would not have engaged a firm of accountants and a firm of lawyers to represent the Appellant because the fraud might easily have been uncovered by one of those firms. 89. Similarly, although we have noted in the course of this decision that Mr Shoefield’s responses to the questions raised by Ms Stephenson and Mr Carey were not always truthful - and we will discuss below what we have made of that fact - he was always quick to respond to the Respondents’ requests for information and his relationship with Ms Kathel appears to have been cordial and co-operative. We were particularly struck by Mr Shoefield’s repeated offers to the Respondents to arrange meetings between the Respondents, Data Tech, MLM and RAM. It seems unlikely to us that a person who knew that the Appellant was a party to fraudulent transactions would invite the Respondents to attend a meeting at which the fraud could easily have been uncovered. 90. Finally in this context, we have noted that, when the investigation by the Respondents continued into 2015 without any sign of a resolution, Mr Shoefield wrote to his MP, Mr Jeremy Corbyn, to ask Mr Corbyn to intercede with the Respondents on his behalf. Again, it seems unlikely to us that a person who was aware that the transactions in question were connected with fraud would have asked Mr Corbyn to intervene in this way. We infer from the fact that the request was made to Mr Corbyn that Mr Shoefield was completely unaware that the transactions in question were connected with fraud. Mr Wright 91. The third reason for our conclusion on this point relates to the evidence of Mr Wright. 92. We think that it is highly unlikely that Mr Shoefield could have had actual knowledge that the Appellant’s transactions with MLM were connected with fraud without Mr Wright’s being encumbered with similar knowledge. Both parties to this appeal implicitly recognised that fact when the Appellant called Mr Wright as a witness and the Respondents sought to challenge Mr Wright’s ignorance of the fraud in cross-examination. 93. In our view, it was somewhat opportunistic on the part of the Respondents that they sought to rely on the evidence of Mr Wright as a witness of truth in the criminal prosecution of Mr Butler and Mr Ferguson and then sought to impugn Mr Wright’s reputation by challenging his evidence at the hearing of this appeal. Either a man is to be believed or he is not. To rely on his evidence at one hearing and then challenge his evidence in another hearing which pertains to the same subject matter is curious, to say the least. 94. Be that as it may, we considered that Mr Wright was a reliable and credible witness. One may well question whether he was naive in failing to identify that RAM was part of a transaction chain involving fraud - particularly when one of the parties with which RAM was dealing in India was named “ Delhicon ” - but we believed him when he said that Mr Butler was very convincing and that he had had no knowledge of the existence of the fraud at the relevant time. 95. And, if Mr Wright did not have actual knowledge of the fraud, then, as we have noted above, it seems highly unlikely that Mr Shoefield had such actual knowledge. Orchestrated fraud 96. Finally, we have considered the allegation made by the Respondents that, because the transactions in question were part of a highly orchestrated chain of transactions, with Mr Butler as the prime mover, that should be regarded as an indication that Mr Shoefield (and, for that matter, Mr Wright) must have known that the Appellant (and RAM) were participating in a fraud because it was improbable that Mr Butler would have involved an unknowing party in the transactions for fear of having the fraud uncovered. 97. Whilst we have taken into account the fact that Mr Butler was clearly the prime mover in an orchestrated chain - for example, his instructions to Mr Wright as to which of RAM’s suppliers ought to have its receivable from RAM discharged first - this is merely something which is part of the circumstances in which the transactions took place. As such, whilst it should be regarded as informing the answer in relation to whether or not the Appellant had actual knowledge of the connection to fraud, it is not determinative of that question - see the extract from Pacific Computers cited at paragraph 44 above. We do not think that the mere fact that there was an orchestrated fraud by Mr Butler necessarily means that Mr Shoefield cannot have been an innocent party in the transactions. Mr Shoefield’s complicity is merely a possible inference from the fact of orchestration - and we do not think that that possible inference should be regarded as outweighing all of the factors described above which point in the opposite direction. Conclusion on actual knowledge 98. For the reasons set out above, we consider that the Respondents have not satisfied us that the Appellant, as represented by Mr Shoefield, had actual knowledge that the transactions in question were connected with fraud in the Appellant’s supply chain. Means of knowledge 99. Although we have a high degree of certainty that the Appellant, as represented by Mr Shoefield, did not have actual knowledge that the transactions to which this appeal relates were connected with fraud in its supply chain, we consider that the answer to the question of whether the Appellant, as so represented, should have had that knowledge is much more finely-balanced. 100. In that respect, Mr Webster made a compelling case to the effect that Mr Shoefield was simply an innocent victim of Mr Butler’s fraud, who could not reasonably have known that Mr Butler was up to no good. 101. The key plank in Mr Webster’s submissions was that, although Mr Shoefield did not do any meaningful due diligence in relation to MLM and Mr Butler: (1) he had no reason to do so, for all of the reasons cited in paragraph 74(7) above; and (2) even if he had conducted the requisite degree of due diligence, he would not have been able to discover the fraud because there was no question which he could reasonably have asked which would have uncovered the true nature of Mr Butler’s activities. 102. We think that Mr Webster had a reasonable point in asking whether there was any due diligence which Mr Shoefield could have done which would have revealed the connection with the fraud. When put to the challenge to identify what due diligence would have put Mr Shoefield on notice, the Respondents were able to identify only the accounts of MLM in respect of its financial year ending 31 October 2012. 103. Ms Stephenson made a fair point when she observed that the negligible net asset position of MLM at the end of that financial year, in comparison to the volume of turnover which was going through MLM in that financial year, would have put a reasonable observer on notice that something was amiss. Mr Shoefield countered that suggestion with the two arguments noted in paragraph 74(10) above – namely, by saying that: (1) it was common practice in the advertising industry for the accounts of an advertising agency to show a level of turnover which was disproportionately high in comparison to its net assets; and (2) MLM’s lack of net assets would only have confirmed his view that MLM lacked the working capital which was needed to implement the transactions with RAM and therefore confirmed to him that the Appellant’s involvement in the transaction chain was fulfilling a valid commercial function. 104. Whilst we accept that both of those arguments have some merit, we ultimately prefer the views of Ms Stephenson on this point. We believe that the paucity of MLM’s net assets at the end of the relevant financial year was so extreme and so low in comparison to the enormous turnover of MLM in that financial year that any consideration of those accounts would have caused a reasonable person to conclude that there was fraud in the supply chain. 105. However, be that as it may, we think that the debate in relation to what information would have been disclosed by any due diligence is ultimately irrelevant because the case law described in paragraphs 34 to 42 above demonstrates that it is wrong to focus on due diligence as an end in itself. The relevant question in this context is ultimately not whether Mr Shoefield conducted the appropriate due diligence but rather whether Mr Shoefield should have known that the relevant transactions were connected with fraud and, in relation to the latter question, the case law demonstrates that Mr Shoefield was not entitled to disregard the obvious inferences to be drawn from the circumstances in which the relevant transactions took place. In the words of Moses LJ in Mobilx , if the relevant trader “ chooses to ignore obvious inferences from the facts and circumstances in which he has been trading, he will not be entitled to deduct ” and “ [if] it is established that a trader should have known that by his purchase there was no reasonable explanation for the circumstances in which the transaction was undertaken other than that it was connected with fraud then such a trader was directly and knowingly involved in fraudulent evasion of VAT .” 106. In applying that test in the present case, we have concluded that, notwithstanding the acute forensic skill of Mr Webster, there were enough indications in the terms of the transaction chain and the manner in which the transactions were conducted to indicate to Mr Shoefield that the transactions were simply too good to be true and must therefore have been associated with fraud. 107. In this regard, we start with the observation that, by his own admission, Mr Shoefield did not take any steps whatsoever to check on the sources of the PPI leads which were the subject of the relevant transactions or to ascertain whether there was any correlation between the value of each such PPI lead and the price that the Appellant was paying for it. This is because, as he candidly admitted, he knew that, no matter what the value of each such PPI lead, he could immediately on-sell the relevant PPI lead to RAM for a substantial margin. In the words of Mr Butler in his original email to Mr Shoefield of 12 January 2012, that margin amounted to “ an annualised gross margin per batch of 48% ”. Accordingly, Mr Shoefield knew that, unless and until he received wholesale complaints from Mr Wright about the quality of the PPI leads which were on-sold to RAM, he had no reason to concern himself with the value of the PPI leads. 108. It is therefore clear that, so far as Mr Shoefield was concerned, this was just an opportunity to make a substantial profit with the assistance of Mr Butler and Mr Wright. The subject matter of the transactions, and the value of that subject matter, were neither here nor there unless it became clear at some point that there was something seriously wrong with the subject matter or that value because of adverse feedback from Mr Wright. And, of course, the circularity of the transaction chain meant that that never occurred. 109. Given that the Appellant was being given the opportunity to make this substantial margin, we consider that Mr Shoefield should have asked himself why that was the case or, to put it another way, what the Appellant was contributing to the transaction chain in terms of value to justify the largesse of MLM and RAM in allowing the Appellant to stand between them and take some of the profit to which the transaction chain was giving rise. 110. In that respect, as Mr Cresswell pointed out in giving his evidence, the Appellant was not adding any value to the transaction chain by enhancing the quality of the data as it passed through it. Moreover, although Mr Creswell added that it was not uncommon in the data market for a broker to buy and on-sell raw data, he made it clear that it was unusual for a broker in that position to have only a single source and a single customer, In this case, the Appellant was not adding value to the transaction chain either by collecting data from multiple sources or by identifying future potential customers for the data. So Mr Shoefield could not have been relying on any of those factors to explain the margin which the Appellant was being offered. 111. During the course of the investigation, Mr Shoefield twice attempted to convey to the Respondents the impression that he had introduced Mr Butler to Mr Wright for the purposes of the transactions in the PPI leads – once in his letter to Mr Whelan of 3 November 2014 and then again in the letter from RPC on his behalf of 11 August 2015. It is easy to see why Mr Shoefield would have been anxious to convey the impression that he had linked the two men for that purpose. It was an obvious bit of “ added value ” which might have justified at least some of the margin. However, at the hearing, Mr Shoefield was compelled to admit that, contrary to the impression which he had sought to convey to the Respondents during the course of the investigation, he had not introduced Mr Butler to Mr Wright for the purpose of the transactions in PPI leads but had in fact done that well before those transactions were in contemplation, with the result that, rather than playing an important role in bringing MLM and RAM together for the purposes of the transactions in PPI leads, the Appellant - and before that, Havas - had in fact been inserted into a pre-existing relationship between MLM and RAM after those companies were already in a position to trade together. So, in his own mind, Mr Shoefield could not have been attributing the margin which the Appellant was being offered to his introduction of MLM to RAM. 112. As Mr Shoefield was not adding any value to the transaction chain by enhancing the quality of the data, by collecting data from multiple sources, by using its connections to sell data to multiple customers or even by effecting the introduction of Mr Butler to Mr Wright, what then was the Appellant doing to justify its substantial margin? 113. In paragraph [33] of AS1, Mr Shoefield gave three reasons to explain the Appellant’s role in the transaction chain. Those three reasons are set out in paragraph 73(3)(b) above and are, in short, managing the process, cleaning the data and providing working capital. 114. The first two of those three reasons simply fell apart under Ms Stephenson’s cross-examination of Mr Shoefield at the hearing, for the reasons summarised in paragraphs 73(3)(d) and 73(3)(e) above, with all of which we agree. It is, in our view, untenable that either of those activities would have led Mr Butler and Mr Wright - ie MLM and RAM - willingly to have surrendered such a substantial margin – or indeed any margin whatsoever - to Mr Shoefield - ie the Appellant. In the case of management, it is clear that all of the management was being provided by Mr Butler, who was in control of the entire chain. In the case of cleaning, it became apparent at the hearing that, if any cleaning was in fact effected by the Appellant – and there was evidence to suggest that no cleaning ever took place – then that cleaning would simply have duplicated the cleaning effected by MLM prior to MLM’s sale of the data to the Appellant, with the result that there was effectively no role for the Appellant to play in that regard. 115. The third of those reasons did potentially appear to have more to it in that the Appellant’s role in the transaction chain clearly did involve its expending working capital on making payments to MLM and then having to extend credit to RAM and to await the VAT repayments from the Respondents. However, for the reasons set out by Ms Stephenson in paragraph 73(3)(f) above, we think that the evidence suggests that, at the time when the opportunity to carry out the trades in PPI leads arose and then during the course of the transactions by Havas (and then the Appellant) which occurred pursuant to that opportunity, Mr Shoefield himself did not consider that role to be a sound reason for a margin of the size that was being offered to the Appellant. 116. It is true that, in his initial email to Mr Shoefield of 12 January 2012 in which he first introduced the opportunity, Mr Butler said that MLM was “ restricted in the volume of data it can provide to its buyers by cash-flow ” and explained that MLM needed “ authorised media distributors ” to fill that gap and that, in his email of 18 January 2012, Mr Butler said that MLM was being overwhelmed by orders and that “ distribution channels are a key focus ”. In addition, in that same email, Mr Butler expressly referred to the issue of credit and told Mr Shoefield that, whilst Mr Butler would be “ happy to discuss and accommodate wherever I can…you will be required to give Rams a minimum of 14 days ”. 117. Similarly, the contemporaneous notes of the meeting between Mr Shoefield, Ms Kathel and Ms Ahmed of 9 July 2013 recorded that Mr Shoefield referred to: “ The ability to fund VAT, cash etc., investments through [the Appellant] ”, whilst the report of that meeting recorded that the Appellant “ does not pay on credit ”; 118. So there is evidence to suggest that both: (1) Mr Shoefield was being told at the time when the opportunity to become part of the transaction chain first arose, in January 2012, that Havas’s (and, subsequently, the Appellant’s) role in the transaction chain was to provide working capital to bridge the gap; and (2) Mr Shoefield was aware in July 2013 that the Appellant was providing working capital in the transaction chain. 119. And yet the evidence as a whole suggests that Mr Shoefield did not regard the provision of working capital into the transaction chain by the Appellant as being an adequate explanation for the substantial margin which the Appellant was earning from its involvement in the transaction chain. 120. In the first place, it was not a reason which occurred to Mr Skuse or his “ VAT boys ” in January 2012, when Havas was first presented with the opportunity to play the same role in the transaction chain as the Appellant later did. In his email of 19 January 2012, Mr Skuse was clearly concerned that the absence of any improvement to the quality of the data when it passed through Havas on its way to RAM might arouse suspicion. He did not appear to have realised that the provision of working capital into the transaction chain might be a good reason why no such suspicion should arise. It also did not appear to be at the forefront of Mr Shoefield’s mind when he responded to Mr Skuse’s question about the need to provide “ evidence of re-work of the data ” in order to justify the margin by saying that he would see what could be done about reworking the data in order to do just that. Mr Shoefield did not immediately reply to Mr Skuse’s question by pointing out that, far from worrying about reworking the data to justify the margin, Havas (and subsequently the Appellant) was providing valuable working capital into the transaction chain and could rely on that fact to explain its role and avoid suspicion. 121. In addition, we were struck by the fact that, when we asked Mr Shoefield about the analysis which he had conducted in comparing the margin on each transaction in PPI leads with the financial risks and costs associated with its exposure to RAM and the Respondents in relation to receiving payments, he responded by saying that he had conducted no such analysis. We found it a little odd that, as an experienced businessman, Mr Shoefield did not consider the risks and rewards to which the transaction gave rise and, in particular, the size of the margin which the Appellant was being offered, relative to the cash flow costs and risks involved in providing credit to RAM. Had he done so, we believe that he would rapidly have reached the conclusion that the Appellant’s role as the provider of working capital did not justify a margin of the size on offer. 122. In the circumstances, we consider that Mr Skuse and his “ VAT boys ” were quite right to identify that there was something very suspicious about the opportunity to derive such easy money for doing so little and that, even if that fact had not dawned on Mr Shoefield prior to the time of his exchange with Mr Skuse, he should have been alerted to the point by virtue of Mr Skuse’s expressed concern. At that point, he should have realised that there was no credible explanation for the size of the margin which Havas, and later the Appellant, was being offered in the transaction chain apart from the fact that fraud was somehow involved. 123. Moreover, we consider that the untruths which Mr Shoefield has told during the period of the investigation and in the course of giving his evidence in the proceedings in relation to the appeal reveal his own belated recognition of that fact. In an effort to obtain the VAT repayments in question, he has sought to find reasons for the Appellant’s significant margin other than the obvious one – ie the provision of working capital - and, in so doing, has selected factors which demonstrably did not justify the margin– such as managing the process and cleaning. If Mr Shoefield had genuinely believed that the Appellant’s role in providing working capital to the transaction chain justified a margin of the size of the one which the Appellant received, why would he not have said that from the start of the investigation and dispensed with explanations which demonstrably held no water? 124. In addition, during the course of the investigation, Mr Shoefield sought to obscure the fact that he did not know the true sources of the data which the Appellant was buying from MLM or the values of that data. For instance, at the meeting with Ms Kathel on 12 August 2013 and then again in his letter to Mr Whelan of 3 November 2014, Mr Shoefield stated that the PPI leads emanated from a call centre, whilst in his email to Ms Kathel of 25 September 2014 and his email to Mr Whelan of 17 February 2015 and then again in his witness statements, he referred to the fact that the suppliers to MLM were companies like Experian. The truth was that Mr Shoefield did not know the true sources of the PPI leads despite the assertion in his email of 23 August 2013 to Ms Kathel that he was aware of how the data was manufactured. He did not care about those sources because, as far as he was concerned, as long as he could on-sell all the PPI leads which he purchased to RAM and did not suffer a wholesale rejection of those PPI leads, then the sources and values of the PPI leads were irrelevant. He said exactly that at the hearing. 125. Similarly, in his email of 23 August 2013 to Ms Kathel, he referred to the fact that the Data Tech platform allowed the Appellant to request data from the Appellant’s “suppliers” (plural), when in fact he knew that the Appellant had only one supplier, MLM. 126. These untruths suggest to us that Mr Shoefield was aware of the fact that there was something very wrong with the terms of the transactions and that there was information about the sources and values of the PPI leads which he should have obtained but had shown no interest in obtaining. 127. We consider that the overall circumstances in which the transactions occurred were such that, no matter how plausible Mr Butler may have been – and we have noted Mr Wright’s testimony to the effect that Mr Butler was very convincing – Mr Shoefield should have been aware that the transactions were connected with fraud. In the words of Moses LJ in Mobilx , if Mr Shoefield “ should have known that the only reasonable explanation for the transaction in which he was involved was that it was connected with fraudulent evasion of VAT then he should have known of that fact...” 128. We have therefore concluded that although the Appellant, as represented by Mr Shoefield, did not have actual knowledge that the transactions which are the subject of this appeal were connected with fraud, it should have known that that was the case and would have done so if Mr Shoefield had not deliberately closed his eyes to the obvious. 129. For completeness, we should note that this conclusion does not mean that we are seeking to transfer to the Appellant the investigative tasks which are properly the domain of the Respondents, as mentioned in paragraphs [60] to [65] of the CJEU decision in Mahagében and discussed in paragraph 31 above. Instead, we consider that this is a situation “ when there are indications pointing to an infringement or fraud ”, as mentioned in that case, and therefore where the Appellant should itself have been aware of the connection with fraud. 130. Given the above conclusion, we do not think that a detailed consideration of the answers to the various questions which are set out in section 6 of VAT Notice 726 in the context of the Appellant’s circumstances is worthwhile, particularly as each party appeared to consider that those answers supported its own case. It suffices to say that, for the reasons set out above, we think that Mr Shoefield should have known that the margin which had been offered to the Appellant in this case was too good to be true and therefore that the transactions in question were connected with fraud. CONCLUSION 131. For the reasons set out above, we consider that the Appellant, as represented by Mr Shoefield, should have known that the transactions which are the subject of this appeal were connected with fraud higher up its supply chain and we therefore dismiss the Appellant’s appeal. RIGHT TO APPLY FOR PERMISSION TO APPEAL 132. 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 Rules. 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. TONY BEARE TRIBUNAL JUDGE RELEASE DATE: 5 August 2019 APPENDIX SECTION 6 OF THE VAT NOTICE 726 “ 6. Dealing with other businesses – How to ensure the integrity of your supply chain 6.1 What checks can I undertake to help ensure the integrity of my supply chain The following are examples of indicators that could alert you to the risk that VAT would go unpaid: 1) Legitimacy of customers or suppliers.For example: - what is your customer’s or supplier’s history in the trade? - has a buyer and seller contacted you within a short space of time with offers to buy/sell goods of same specifications and quantity? - has your supplier referred you to a customer who is willing to buy goods of the same quantity and specifications being offered by the supplier? - does your supplier offer deals that carry no commercial risk for you – eg, no requirement to pay for goods until payment received from customer? - do deals with your customer/supplier involve consistent or predetermined profit margins, irrespective of the date, quantities or specifications of the specified goods traded? - does your supplier (or another business in the transaction chain) require you to make 3rd party payments or payments to an offshore bank account? - are the goods adequately insured? - are they high value deals offered with no formal contractual arrangements? - are they high value deals offered by a newly established supplier with minimal trading history, low credit rating? - can a brand new business obtain specified goods cheaper than a long established one? - has HMRC specifically notified you that previous deals involving your supplier had been traced to a VAT loss and/or had involved carousel movements of goods? - has HMRC specifically notified you that HMRC date stamps have been present on goods offered for sale by your supplier, or that there is evidence of HMRC date stamps being removed from packaging. This would strongly suggest that the goods had been subject to carousel movement, which should alert you to a significant risk that the transactions entered into with that supplier may be connected with the non-payment of VAT; - has HMRC specifically notified you that other MTIC VAT fraud characteristics (such as third party payments) have occurred in transaction chains involving your supplier? 2) Commercial viability of the transaction. For example: - Is there a market for this type of goods, such as superseded or outdated mobile phone models or non-UK specific models? - What research have you done to test whether these goods are available as described and in the quantities being offered? - Is it commercially viable for the price of the goods to increase within the short duration of the supply chain? - Have normal commercial practices been adopted in negotiating prices? - Is there a commercial reason for any third party payments? - Are normal commercial arrangements in place for the financing of the goods? 3) Viability of the goods as described by your supplier. For example: - Do the goods exist? - Have they been previously supplied to you? - Are they in good condition and not damaged? - Do the quantities of the goods concerned appear credible? - Do the goods have UK specifications yet are to be exported? - Is your supplier unwilling to provide IMEI or other serial numbers? - What recourse is there if the goods are not as described? HMRC recommends that sufficient checks be carried out in each of the above categories to ensure that you are not caught in a fraudulent supply chain. 6.2 Checks carried out by existing businesses The following are examples of specific checks carried out by businesses that took part in the consultation exercise in 2003 when these rules were introduced. These may also help you to decide what checks you should carry out, but this list is not exhaustive and you should decide what checks you need to carry out before dealing with a supplier or customer: - obtain copies of Certificates of Incorporation and VAT registration certificates; - verify VAT registration details with HMRC; - obtain signed letters of introduction on headed paper; - obtain some form of written and signed trade references; - obtain credit checks or other background checks from an independent third party; - insist on personal contact with a senior officer of the prospective supplier, making an initial visit to their premises whenever possible; - obtain the prospective supplier’s bank details, to check whether: (a) payments would be made to a third party; and (b) that in the case of an import, the supplier and their bank shared the same country of residence. - check details provided against other sources, for example website, letterheads, BT landline records. Paperwork in addition to invoices may be received in relation to the supplies you buy and sell. This documentation should be kept to support your view of a transaction’s legitimacy. The following are examples of additional paperwork that some businesses retain: - purchase orders; - pro-forma invoices; - delivery notes; - CMRs (Convention Merchandises Routiers) or airway bills; - allocation notification; - inspection reports Again this is not an exhaustive list, but does show some of the more common subsidiary documentation. 6.3 What will HMRC look out for when considering the extent of my checks? In each case, HMRC will be seeking to identify what actions or precautions you took in response to any indicators of risk. This will focus on the due diligence checks you undertook and, most importantly, the actions taken by you in response to the results of those checks. In each case, HMRC will consider: - What due diligence checks were performed? This includes any checks designed to address the specific risks of a specific case; - To what extent were your checks appropriate, adequate and timely in relation to addressing the risks identified? - What the results of the checks indicated? - Did you take appropriate action in response to the results of the checks? If you have genuinely done everything you can to check the integrity of the supply chain, can demonstrate you have done so, have taken heed of any indications that VAT may go unpaid and have no other reason to suspect VAT would go unpaid, the joint and several liability rules will not be applied.”;